JTION NOIlfUllSNI NVINOSH1IWS S3ldVdan LIBRARIES SMITHSONIAN INSTITUTION NOIIDIIJ
— co

dal! LIBRARIES SMITHSONIAN INSTITUTION NOIlfUllSNI NVINOSHlllNS S31dVdan LIBRAR

rn gey ™ '$0" rn !!fj rn ™ rn 'sv°<y

05 £ (/) \ £ 05

UTION NOIlfUllSNI NVIN0SH1IINS S3ldVdan LIBRARIES SMITHSONIAN INSTITUTION NOlinil.

O) Z CO Z .... CO z

>

ra a l"|Z LI B RAR I ES^SMITHSONIAN INSTITUTION NOIlfUllSNI NVINOSHIIWS^SB I d Vd 8 IT LIBRAR

— CO ~ rn — 'O

(X.

<

__ o;

O pcjx7 ^ q ^ O

UTION ~ NOIlfUllSNI “nVINOSHIMS^S 3 IdVdaiT^LIBRARI ES^ SMITHSONIAN^ INSTITUTION NOIlfUL
z i- z ~ ~ -

dail LIBRARIES SMITHSONIAN INSTITUTION NOIlfUllSNI NVIN0SH1I1AIS S3ldVdan LIBRAR

Z 05 Z V CO Z CO _ Z

05 A' Z ^ Z W5 2 CO

JTION NOIlfUllSNI NVIN0SH1IWS S3ldVdan LIBRARIES SMITHSONIAN INSTITUTION NOIlfUI.

— - co

9 2 2 W 2 3 %J| 3 2

o -.,W _ o ~ o

z -J Z _J z

dail LIBRARIES SMITHSONIAN INSTITUTION NOIlfUllSNI NVIN0SH1MS S3ldVdan LIBRAR

r r *7

UTION NOIlfUllSNI NVIN0SH1IIAIS S3ldVdan LI B RAR I E S ^SMITHSONIAN INSTITUTION ^ NOIlfUI.

05 Z _ 05 z .... 05 z vv. 05

s^! '■#'!#= I #1,1 1.^

.-4# ?% iO)i% im

'd a llZ LI B RAR I ES SMITHSONIAN INSTITUTION NOIlfUllSNI NVIN0SH1IINS°°S3 I d Vd a IT LIBRAR
Z to 5 ^ ^ co — „ <o ^ _ z

-j? yj 2 ^ 2 (/)

ioshiiwsv;S3 i a va a n libraries Smithsonian institution NoiiniiisNi_NviNOSHiiws S3ia\

fHSONIAN INSTITUTION NOlinillSNI NVINOSH1MS S3iavaai3 LIBRARIES SMITHSONIAN INSTIT

josHiiiNS S3iavaan libraries Smithsonian institution NoiinmsNi nvinoshhins S3ia\

to 2 to 2 v- to Z _ ^

^ ^ 2 'V >

rHSONIAN INSTITUTION NOlinillSNI NVINOSHlIWS^'s 3 I a V a B 13 “ LI B R AR I E S SMITHSONIAN _ INSTI1

to — go . 2 <s> z

W X^^^x “ UJ XvWtT'x 7?,

161 z! x < (oft 36 z! UF < c Ufc

gosHims S3 lava a n libraries Smithsonian institution Nouniiism nvinoshiiins S3ia

— X^titx^

K '

2 _

THSONIAN INSTITUTION^NOlinillSNI- NVINOSHlI^s” S3 I a Va 3 11 LI B RAR I ES SMITHSONIAN INSTI

2 CO 2 <0 2

2 ,/dfarr. 2 ±

* W | | 'tj^ - 1 vW' | W' |

NosHiiws^sa i ava a nZLi b rar i es^smithsonian institution NoiiniusNi_NviNosHiiws sb i a

ITHSONIAN INSTITUTION NOlinillSNI NVINOSH1IIAIS S3iavaai3 LIBRARIES SMITHSONIAN INSTI

m \ z

NOSHIIINS” S3 I aVaa n~L! BRAR I ES^SMITHSONIAN'iNSTITUTIOn’ NOlinillSNI NVINOSH1IIAIS S3 I a

in z c/> z ^ cn z . ^

ITHSONIAN INSTITUTION NOlinillSNI NVINOSH1IWS S3IBVaail LIBRARIES SMITHSONIAN _ I NSTI

Palaeontology

VOLUME 23 ■ PART 3 AUGUST 1980

Published by

The Palaeontological Association London

Price £1 5

THE PALAEONTOLOGICAL ASSOCIATION

The Association publishes Palaeontology and Special Papers in Palaeontology. Details of membership and subscription rates
may be found inside the back cover.

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 on all-aspects of
palaeontology and stratigraphical palaeontology are invited. They should conform in style to those already published in this
journal, and should be sent to Dr. R. A. Fortey, Palaeontological Association, Department of Palaeontology, British Museum
(Natural History), Cromwell Road, London SW7 5BD, England, who will supply detailed instructions for authors on request
(these were published in Palaeontology 1977, 20, pp. 921-929).

Special Papers in Palaeontology is a series of substantial separate works ; the following are available (post free).

1. (for 1967): Miospores in the Coal Seams of the Carboniferous of Great Britain, by A. H. v. smith and m. a.
butterworth. 324 pp., 72 text-figs., 27 plates. Price £8 (U.S. $19.50).

2. (for 1968): Evolution of the Shell Structure of Articulate Brachiopods, by a. williams. 55 pp., 27 text-figs., 24 plates.
Price £5 (U.S. $12).

3. (for 1968): Upper Maestrichtian Radiolaria of California, by Helen p. foreman. 82 pp., 8 plates. Price £3 (U.S. $7.50).

4. (for 1969): Lower Turanian Ammonites from Israel, by R. freund and M. raab. 83 pp., 15 text-figs., 10 plates. Price £3
(U.S. $7.50).

5. (for 1969): Chitinozoa from the OrdoVician Viola and Fernvale Limestones of the Arbuckle Mountains, Oklahoma,
by w. A. M. jenkins. 44 pp., 10 text-figs., 9 plates. Price £2 (U.S. $5).

6. (for 1969): Ammonoidea from the Mata Series (Santonian-Maastrichtian) of New Zealand, by r. a. henderson. 82 pp.,
13 text-figs., 15 plates. Price £3 (U.S. $7.50).

7. (for 1970): Shell Structure of the Craniacea and other Calcareous Inarticulate Brachiopoda, by a. williams and
a. d. wright. 51 pp., 17 text-figs., 15 plates. Price £1-50 (U.S. $4).

8. (for 1970): Cenomanian Ammonites from Southern England, by w. j. Kennedy. 272 pp., 5 tables, 64 plates. Price £8
(U.S. $19.50).

9. (for 1971): Fish from the Freshwater Lower Cretaceous of Victoria, Australia, with Comments on the Palaeo-
environment, by M. waldman. 130 pp., 37 text-figs., 18 plates. Price £5 (U.S. $12).

10. (for 1971): Upper Cretaceous Ostracoda from the Carnarvon Basin, Western Australia, by r. h. bate. 148 pp., 43 text-
figs., 27 plates. Price £5 (U.S. $12).

11. (for 1972): Stromatolites and the Biostratigraphy of the Australian Precambrian and Cambrian, by m. r. Walter.
268 pp., 55 text-figs., 34 plates. Price £10 (U.S. $24).

12. (for 1973): Organisms and Continents through Time. A Symposium of 23 papers edited by n. f. hughes. 340 pp.,
132 text-figs. Price £10 (U.S. $24) (published with the Systematics Association).

13. (for 1974): Graptolite studies in honour of O. M. B. Bulman. Edited by r. b. ricKards, d. e. jackson, and c. p. hughes.
261 pp., 26 plates. Price £10 (U.S. $24).

14. (for 1974): Palaeogene Foraminiferida and Palaeoecology, Hampshire and Paris Basins and the English Channel, by
j. w. Murray and c. A. wright. 171 pp., 45 text-figs., 20 plates. Price £8 (U.S. $19.-50).

15. (for 1975): Lower and Middle Devonian Conodonts from the Broken River Embayment, North Queensland, Australia,
by p. G. telford. 100 pp., 9 text-figs., 16 plates. Price £5-50 (U.S. $13.50).

16. (for 1975): The Ostracod Fauna from the Santonian Chalk (Upper Cretaceous) of Gingin, Western Australia, by
j. w. neale. 131 pp., 40 text-figs., 22 plates. Price £6-50 (U.S. $16).

17. (for 1976): Aspects of Ammonite Biology, Biogeography, and Biostratigraphy, by w. j. Kennedy and w. a. cobban.
94 pp., 24 text-figs., 11 plates. Price £6 (U.S. $14.50).

18. (for 1976): Ostracoderm Faunas of the Delorme and Associated Siluro-Devonian Formations, North West Territories,
Canada, by d. l. dineley and e. j. loeffler. 218 pp., 78 text-figs., 33 plates. Price £20 (U.S. $48).

19. (for 1977): The Palynology of Early Tertiary Sediments, Ninetyeast Ridge, Indian Ocean, by E. m. kemp and w. k. Harris.
74 pp., 2 text-figs., 8 plates. Price £7 (U.S. $17).

20. (for 1977): Fossil Priapulid Worms, by s. c. morris. 159 pp., 99 text-figs., 30 plates. Price £16 (U.S. $38.50).

21. (for 1978): Devonian Ammonoids from the Appalachians and their bearing on International Zonation and Correla-
tion, by M. R. house. 70 pp., 12 text-figs., 10 plates. Price £12 (U.S. $29).

22. (for 1978, published 1979): Curation of palaeontological collections. Ajoint colloquium of The Palaeontological Associa-
tion and Geological Curators’ Group. Edited by m. g. bassett. 280 pp., 53 text-figs. Price £25 (U.S. $60).

23. (for 1979): The Devonian System. A Palaeontological Association International Symposium. Edited by M. R. house, c. t.
scrutton, and m. g. bassett. 353 pp., 102 text-figs., 1 plate. Price £30 (U.S. $72).

24. (for 1980): Dinoflagellate Cysts and Acritarchs from the Eocene of southern England, by j. b. bujak, C. downie, g. l.
eaton, and G. l. williams. 104 pp., 24 text-figs., 22 plates. Price £15 (U.S. $36).

© The Palaeontological Association, 1980

Cover: Edriophus levis (Bather, 1914) from the Middle Ordovician Trenton Group of Kirkfield, Ontario. x2-5.
Specimen in the Smithsonian Institution; photograph by H. B. Whittington.

DINOFLAGELLATE CYSTS FROM THE
UPPER EOCENE-LOWER OLIGOCENE OF THE
ISLE OF WIGHT

by M. LIENGJARERN, L. COSTA, and C. DOWNIE

Abstract. The Upper Eocene and Oligocene succession of the Isle of Wight, southern England (Headon Beds
to Hamstead Beds) has been studied palynologically. Seventy-one forms of dinoflagellate cysts are recorded,
including two new genera, Gerdiocysta and Vectidinium, and ten new species, Distatodinium scariosum,
Eocladopyxis tessellata, G. conopeum, Glaphyrocysta paupercula , Phelodinium pachyceras, P. pumilum, Phthano-
peridinium amiculum, P.flebile, Thalassiphora fenestrata, and V. stover i. The dinoflagellates (with the exception
of Vectidinium) are marine and indicate six marine incursions or partial incursions in the sequence; the mid-
Headon Beds, the Oyster Bed of the Bembridge Marls, the Nematura Band, and three episodes of the Upper
Hamstead Beds. Correlation with the Paris Basin indicates that the base of the Stampian lies near the
Nematura Band.

The importance of dinoflagellate cysts in the stratigraphy of the Palaeogene has been emphasized
in several recent papers. Many long-standing problems in the Upper Palaeocene and Lower Eocene
have been resolved by their application, but problems of correlation at the Eocene/Oligocene
boundary remain. This account describes the dinoflagellate cysts from the classical section on the Isle
of Wight in southern England. The initial work was done by M. Liengjarern (1973) and has been
revised recently by L. Costa.

STRATIGRAPHY

The sequences in the Isle of Wight span the Eocene/Oligocene boundary, and the placings of this
boundary have varied according to the interpretation of different authors (see Curry et al. 1978) from
the base of the Headon Beds to the base of the Hamstead Beds. The difficulties in correlation and
interpretation are largely the consequence of the paralic nature of the deposits, which varied from
open-sea to freshwater lacustrine in a complex coastal geography.

Two main localities are reported here. In the east of the island, the lower part of the succession,
from the base of the Lower Headon Beds to the Bembridge Marls, is exposed continuously in the
sea cliffs at WhitecliffBay. In the west, the upper part of the succession (Bembridge Marls-Hamstead
Beds) is exposed in Bouldnor and Hamstead cliffs as a continuous sequence (text-figs. 1 and 2).

PALYNOLOGY

All the samples were prepared by standard palynological methods. Only a few samples of fluvial sands were
barren, the remainder yielded rich assemblages of palynomorphs, including pollen and spores, plant tissue,
freshwater algae, dinoflagellate cysts, and acritarchs. Only the dinoflagellate cysts are dealt with in detail in
this paper, but in each sample the proportions of pollen and spores, Pediastrum, dinoflagellates, and acritarchs
based on counts of 200 individuals were noted. These results are shown in Tables 1 and 2. It should be noted
that these counts were made after sieving through a 20 ^m sieve and that consequently pollen is under-
represented.

A complete list of the dinoflagellate taxa recorded and their distribution and relative abundances are shown
in Table 1 . Only new taxa or combinations, or taxa necessitating further comment are described here. The genera
discussed are arranged in alphabetical order; suprageneric dinoflagellate cyst-taxa are not employed here.
(Palaeontology, Vol. 23, Part 3, 1980, pp. 475-499, pis. 53-54.1

476

PALAEONTOLOGY, VOLUME 23

text-fig. 1. Stratigraphic location of samples collected at Hamstead Cliff (prefix H) and at Bouldnor Cliff

(prefix B).

The terms employed in the descriptions are those of Williams et al. (1973) and Evitt et al. (1977). In some
species, the arithmetical mean of the measurements is indicated as a figure in parenthesis. The reference for
holotypes and illustrated specimens is given with reference to their location in the ‘England Finder’ grid system.

Division pyrrhophyta
Class DINOPHYCEAE Fritsch 1935
Order peridiniales Haeckel 1894
Genus distatodinium Eaton 1976

Type species. Distatodinium craterum Eaton 1976

LIENGJARERN ET AL .: EOCENE/OLIGOCENE DINOFL AGELL ATES

477

Distatodinium scariosum sp. nov.
Plate 54, fig. 3

Name derivation. Latin, scariosus, thin, papery.

Diagnosis. Distatodinium with broad, hollow, intratabular processes (usually one per paraplate),
oblate to subtriangular in cross-section, distally expanded, and bearing a variable number of thick
secae on their distal margin. Cingular area devoid of processes.

Description. The central body ambitus is oval, antero-posteriorly elongate. Apex and antapex are rounded;
the antapex may be prolonged into a corona formed by the expanded bases of the antapical processes.

The insertion of the processes on the central body is subcircular, oblate, or triangular. The processes occur
one per paraplate, except on the antapical paraplate (l'"'), where there may be two or more processes. The
degree of compression of the processes varies on a single specimen; some processes are taeniate, but more
commonly they are oblate to subtriangular and are open distally. The distal margin of the processes extends into
a variable number of robust secae, sometimes prolonged into fine strands which might connect with those from
near-by processes.

Two of the apical processes are considerably smaller than the other two. Cingular and sulcal zones are free
of processes. When more than one antapical process occurs, their proximal sections coalesce, forming a corona
which is apparently hollow.

Holotype. Slide ML 1456, R37/0, sample B 1 1 , Upper Hamstead Beds, Lower Oligocene, Bouldnor Cliff, Isle
of Wight.

Upper

Headon

Beds

Middle

Headon

Beds

Lower

Headon

Beds

-WC24

-WC21

WC16
WC15

Brockenhurst Bed

Bembridge
Marls

Bembridge

Limestone

text-fig. 2. Stratigraphic location of samples collected at Whitecliff Bay (prefix WC).

table 1. Distribution of dinoflagellate species

478

PALAEONTOLOGY, VOLUME 23

o §

S S

3 T*

sia

wa

na

88

Z.8

98

WH

iZH

61H

9H

PH

SSDAV

SfDAV

KDM

SZDA\

PZD/A

£ZDAY

IZDAV

OZDM

6IDAV

8IDAV

Z.IDM

9IDAV

SIDAV

HDAV

UDAV

l/~> — i

m

XXX

r— < v (N

^vo o
r-

— < vo ©

?2

si

^ ^ ^ ^ ^ ^

s a £ ^

' §!-
-5

If'

^ I

a ^

s S a

5 5 JS

^ Cl,-0 ~

S' Sf s ?

^ -a ■* 3

Q Q

& 73 "a v5 .5

-s.-a . : s s s

qqqqqkjiiifens)

s * Ps .

LIENGJARERN ET AL.\ EOCENE/OLIGOCENE DINOFLAGELLATES

<n v r- Tt

x

i/~> m r-

S g
% -S

i |-§

Its I
il i
1 H
1 1I

"a -S .2
bo 5 as

s s

£ o
R 60

3 ‘5 ,®
•3 33 ~
.a <5 5
^ .3
- -2 h-K
3 3 0’S

tu ®

ft,^ §

O a, a.

S3 1

as a

v a -a

ffS is C

®

3 S

1 1 £ Si. g I w 2 I S I 1

®>§tj

j .p o' a a ® <5 a 'C S 's be
J ’as ^ g 2 a .S -5 -® «*> a s s? js

SaSSob-'Sb-S^SS'

5, "5 § § ft, £ .g

Q *g s o fe £ § *S5 Sd 2 -ft ^

3 -3 I -§ ^gsl-S-S^oS
" I 81 5 Is S.g-1 ^ ^ bo®
- -: ^ -^co Co S hi ^ S ^ ^

I I
pi ps

tfcs*

indicates that presence is below 1%.

-E 1. Distribution of dinoflagellate <

: —indicates that presence is below 1%.

PLATE 53

liengjarern et a!., Eocene/Oligocene dinoflagellates

LIENGJARERN ET AL EOCENE/OLIGOCENE DINOFL AGELL ATES

Measurements. Holotype, central body length 51 ju.m including operculum (43 not including operculum),
breadth 31 /urn; process length 5-15-5 /nm.

Range. Central body length 38-49 p.m (not including operculum), breadth 26-31 /urn; process length 5-16 y.m.
Specimens measured— 8.

Comparisons. The broad, usually hollow and distally open processes, commonly unconnected
distally, distinguish D. scariosum from other species in the genus.

Distribution. Samples Bll, B15.

Genus emslandia Gerlach 1961
Type species. Emslandia emslandensis Gerlach 1961

Emslandia sp.

Plate 54, fig. 5

Remarks. This species of Emslandia has a bulging ventral hypocyst surface. The ambitus is sub-
circular to ovoid. The epicyst is distally rounded and is prolonged into a very short apical horn,
subrectangular in outline, with distal ending truncate, bifid or sometimes produced into a variable
number of short solid processes. The hypocyst may be rounded or somewhat pointed medially
(?compression) and sometimes bears a very short, solid antapical projection.

The autophragm is robust but does not exceed 2 ^m in thickness, it is apparently spongy, perforate,
and its outer surface is scabrate. Linear thickenings of the wall appear scattered randomly on the
autocyst; sometimes these coalesce on portions of the cyst producing irregular reticulate structures.
Two parallel thickenings of the autophragm mark the cingular margins.

The archeopyle is large, type P. The operculum may remain attached along its cingular suture.
Emslandia sp. differs from E. emslandensis by its thinner autophragm and randomly scattered
ornament of linear thickenings, in part reticulate. It is clearly a distinct species, but the material is too
badly preserved to provide satisfactory types.

Distribution. Samples WC 19-21, 23; Middle Headon Beds, Whitecliff Bay, Isle of Wight.

Genus eocladopyxis Morgenroth 19666
Type species. Eocladopyxis peniculata Morgenroth 19666

Eocladopyxis tessellata sp. nov.

Plate 53, fig. 6

Name derivation. Latin, tessellatum, tessellated.

Diagnosis. Eocladopyxis distinguished by abundant, long, solid, intratabular processes which end
distally in fine spines repeatedly furcated and reflexed. The central body is moderately compressed
dorso-ventrally and its ambitus is circular. Archeopyle type A + 3A + 6P. Additional sutures may
occur randomly between any pair of paraplates.

EXPLANATION OF PLATE 53

Fig. 1. Gerdiocysta conopeum gen. et sp. nov., SEM showing the membrane connecting the distal ends of the
processes, x785.

Fig. 2. Gerdiocysta conopeum gen. et sp. nov., holotype, dorsal view showing apical archaeopyle, x 500.

Fig. 3. Glaphyrocysta paupercula sp. nov., holotype, x 1000.

Fig. 4. Phthanoperidinium amiculum sp. nov., holotype, x 1000.

Fig. 5. Glaphyrocysta paupercula sp. nov., specimen with reduced processes, x 1000.

Fig. 6. Eocladopyxis tessellata sp. nov., holotype, x 1000.

482

PALAEONTOLOGY, VOLUME 23

Description. The autocyst is moderately to strongly compressed dorso-ventrally with a circular ambitus. The
autophragm is scabrate and is produced into solid intratabular processes, two to four, sometimes more, per
paraplate. The processes are only slightly flexible, simple, somewhat expanded proximally, circular in cross-
section; distally they flare into a number of fine spines which fork repeatedly; more rarely some of the processes
may end in simple bifurcations. They are more or less strongly reflexed.

The archeopyle appears to be of the type A + 3A + 6P although it is possible that all apical plates separate
in the formation of the archeopyle. Additional sutures commonly develop, apparently at random, between any
other pair of paraplates, both on the epicyst and on the hypocyst.

The paratabulation formula may sometimes be determined on the basis of plate separation, and is 4', 6", 6c,
?5", 1 p.v., 1 " ", ?Xs. Two of the apical paraplates appear to be larger than the other two. The precingular
paraplates are of roughly the same size, antero-posteriorly elongate, and pentagonal in outline. The cingular
paraplates are narrow and subrectangular, and frequently bear only two processes each. The hypocyst appears
to be formed by five large postcingular paraplates, a prominent posterior-ventral paraplate and an antapical
paraplate, but these are only rarely evident since secondary sutures are uncommon on the hypocyst; a number
of smaller sulcal paraplates also appear to be present.

Holotype. Slide ML 1451, T51/2, sample WC25, Middle Headon Beds, Upper Eocene, Whitecliff Bay, Isle of
Wight.

Measurements. Holotype, central body diameter, 37/xm; process length 8-15 /xm.

Range. Central body diameters 31-39 x 35-43 /xm; process length 4-5-10 /xm. Specimens measured— 11.

Comparisons. The solid processes, paratabulation, and archeopyle type leave no doubt as to the
generic allocation of E. tessellata-, however, the archeopyle is not always observable, in which case
the specimens closely resemble some species of the genus Impletosphaeridium Morgenroth 19666.

E. tessellata differs from E. peniculata Morgenroth, the only other species so far allocated to the
genus, in its larger size and longer processes. The process terminations in E. tessellata are more
complex than in E. peniculata.

Genus gerdiocysta gen. nov.

Name derivation. Latin, gerdius, weaver.

Type species. Gerdiocysta conopeum sp. nov.

Diagnosis. Cyst ambitus subcircular, posteriorly bilobed or rounded; dorso-ventral compression
moderate to strong. Pericyst bearing solid penitabular to intratabular processes arranged into
annular, soleate, or linear complexes. The process complexes support a reticulate or membraneous
ectophragm, which on the dorsal face and laterally simulate the outline of the paraplates. On the
ventral face, a median area of variable size is free of ornament and ectophragm. The processes on
either side tend to be linearly oriented more or less parallel to the ambitus; the ectophragm on the
ventral face may link processes from different paraplates.

Inferred tabulation formula: 4', 6", 6c, 5'", 1 p.v., 1" ", Os.

Archeopyle type A, with zig-zag margins including a slightly offset sulcal notch. Operculum tetra-
tabular, commonly free.

Comments. Gerdiocysta is similar to Areoligera Lejeune-Carpentier but differs strongly in the posses-
sion of an ectophragm, which, on parts or all of the dorsal surface of the cyst, simulates the shape
of paraplates. In Areoligera the processes may be joined distally or laterally by trabeculae, but
these are sparse and are loosely interconnected and do not constitute an outer reticulum or
membrane.

The genus Riculacysta Stover 1977, resembles Gerdiocysta in shape and in possessing a
membranous perforate to reticulate ectophragm. However, in Riculacysta the processes are not in
complexes, and are restricted to the ventro-lateral and lateral zones of the cyst. The ectophragm on
the dorsal surface of Riculacysta lies very close to or touches the autophragm and extends across
the paraplate sutures in that region. In contrast there are the simulate dorsal complexes in
Gerdiocysta.

LIENGJARERN ET AL EOCENE/OLIGOCENE DINOFL AGELL ATES

483

Gerdiocysta conopeum sp. nov.

Plate 53, figs. 1, 2

Derivation of name. Latin, conopeum, mosquito net.

Diagnosis. Gerdiocysta characterized by a finely reticulate to membranous perforate simulate ecto-
phragm developed over paraplates V-4', l"-5", 2" '-4'", 1 p.v., and \""\ an arcuate to soleate
complex of very reduced processes, distally free, may be developed on paraplate 6". The process
bases are connected by microgranular thickenings of the cyst wall which form low ridges within
the complexes; these thickenings are often further developed into an intratabular irregular, coarse
reticulum. Individual processes are solid, slightly fibrous, and distally furcated. The median ventral
area is large.

Description. The antapical bilobation of the central body may be moderately or only weakly marked. The dorsal
convexity and ventral depression are moderate. The endophragm is finely granulose, apparently perforate. The
periphragm, as seen on the process walls is slightly fibrous.

The process complexes are determined proximally by basal granulose thickenings on the cyst wall, which
form a more or less continuous basal ridge. Distally, the simulate ectophragm is well developed over paraplates
l'-4', l"-5", 2" '-4" ', 1 p.v., and 1" ". The cingular paraplates 2c-4c may bear linear complexes of processes
which may or may not be distally united. A narrow ectophragm may also be developed on the ventral sur-
face, forming an arcuate wing bordering the central area free of ornament. The ectophragm is closely perforate
and finely reticulate or membranous; both types may combine in the same species.

On some individuals, the processes are greatly reduced, no ectophragm is developed, but a coarse granulate
basal reticulum extends over the dorsal plate surfaces; intermediate forms between these and normal specimens
with well-developed processes and ectophragm are common.

Holotype. Slide ML 1456, E 29/2, sample Bll, Upper Hamstead Beds, Lower Oligocene, Bouldnor Cliff, Isle
of Wight.

Measurements. Holotype, central body length (operculum not included) 64 jim, breadth 73 /urn, processes height
up to 20 /un.

Range. Central body length (operculum not included) 47(54-7)64 fim, breadth 63(68)79 pm, process length
6-23 pm. Specimens measured— 15.

Comparisons. No granulate proximal wall thickenings have been mentioned in the description of
the only other species in the genus G. cassicula (Drugg) comb, nov., which also appears to differ
from G. conopeum in having considerably longer processes and a more prominently bilobed
antapex.

Benedek (1972, pi. 1, figs. 1 1 a-c) illustrated examples as Cyclonephelium pastielsii which appear to
be conspecific with G. conopeum.

Distribution. Samples B6, 7, 8, 11, and 15. Also in Lower Lintforter Beds and Ratinger Beds (early
Rupelian), Germany and Calcaire de Sannois (early Stampian), France (Chateauneuf, pers. comm.).

Other species allocated to the genus: G. cassicula (Drugg) comb. nov. = Areoligera cassicula Drugg 1 970, p. 8 1 1 ,
figs. 2b, 3a-b.

Genus glaphyrocysta Stover and Evitt 1978
Type species. Glaphyrocysta retiintexta (Cookson 1975)

Glaphyrocysta pauper cula sp. nov.

Plate 53, figs. 3, 5

Name derivation. Latin, pauperculum, diminutive of pauperculus, poor.

Diagnosis. Central body compressed, ambitus subcircular to quadrangular, with or without antapical
indentation. Autophragm microgranular, finely reticulate. Processes developed along a peripheral

484

PALAEONTOLOGY, VOLUME 23

band of varying width, leaving relatively prominent mid-dorsal and mid-ventral areas free. Processes
solid, fibrous, simple or bifurcate. The processes may be isolated or arranged into linear, arcuate,
soleate, or annular complexes. When in complexes the processes are joined by their expanded
proximal parts; a few lateral (rarely distal) trabeculae may occur. The complexes have a ragged
appearance distally. Processes from different complexes may be joined by basal ridge and/or
medially by sparse trabeculae. Processes may be considerably reduced in number and in size.

Processes may occur on some or all of the paraplates T- 4', l"-5" (rarely on 6"), 1" '-5" ', 1 p.v.,
and l"".

The archeopyle is apical tetratabular, type A; the operculum may be free or remain attached. The
archeopyle suture has a sulcal notch a little offset from the mid-body line.

Description. The central body is moderately to strongly compressed; the ambitus varies from subcircular to
quadrangular, the antapex is rounded, somewhat indented or produced into one or two unequal lobes. The
autophragm appears microgranular in optical section and is finely reticulate in surface view.

The processes are variable in number, size, and shape, and are developed along an ambital line of variable
width. The mid-dorsal and especially the mid-ventral areas are free of ornament and relatively prominent.
Individual processes, when well developed, are solid, slightly fibrous (most noticeable at and near the base),
slender, simple or bifurcate.

The processes may be isolated, although some alignment may often be evident, or arranged into complexes
on parts of the cyst. When in complexes, the processes are joined proximally by low ridges formed by their
expanded bases; sparse ribbon-like trabeculae with smooth margins may also occur laterally, and only rarely
distally. Processes from different complexes may also be united proximally by ridges and laterally by sparse
trabeculae. Process complexes are normally present and better defined on the apical, dorsal precingular, and
antapical zones of the cyst.

All apical paraplates bear processes, normally arranged into four or three annular or soleate complexes;
when four, two are smaller and tend to coalesce into a single elliptical complex. Linear to arcuate complexes
may occur on the precingular paraplates l” -5" (occasionally, processes occur on paraplate 6"). Towards the
periphery of the dorsal face (2” and 4") the complexes may be soleate. On the ventral face, linear or
somewhat arcuate complexes may be clear but sometimes the peripheral processes may coalesce with those
from postcingular paraplates and become part of a more or less continuous complex parallel to the ambitus.
On the postcingular paraplates process complexes tend to lose definition and to form a number of lines
running antero-posteriorly near the periphery of both dorsal and ventral faces. The posterior ventral processes
may join in these lines or be separate as an arcuate complex. A soleate complex is frequently observable
on paraplate 1"

These forms with more or less well-defined complexes of well-developed processes constitute one end of the
range of variation observed in this species. The other end includes forms with some isolated processes reduced
to simple spines scattered along the peripheral and dorsal precingular zones, tending to form two to four
loosely defined lines parallel to the cyst ambitus. The variability between both extreme types is continuous in
the same assemblage and cannot be applied to further taxonomic division.

The archeopyle is apical, tetratabular; the opercula may be free or may remain in place. A rather shallow
sulcal notch, relatively little offset from the mid-cyst line is observable on the archeopyle margin.

Holotype. Slide ML 1455 P44/1, sample B8, Upper Hamstead Beds, Lower Oligocene, Bouldnor Cliff, Isle of
Wight.

EXPLANATION OF PLATE 54

Fig. 1. Thalassiphora fenestrata sp. nov., holotype, dorsal view, showing archaeopyle and fenestrations, x 250.
Fig. 2. Phelodinium pumilum sp. nov., holotype, dorsal view showing archaeopyle and small cavities at the horns,
x 1000.

Fig. 3. Distatodinium scariosum sp. nov., holotype, x 1000.

Fig. 4. Phelodinium pachyceras sp. nov., holotype, x 1000.

Fig. 5. Emslandia sp. Middle Headon Beds, sample WC20, showing precingular archaeopyle and cingulum,
x 500.

Fig. 6. Phthanoperidinium flebile sp. nov., holotype, x 1000.

Fig. 7. Vectidinium stoveri gen. et sp. nov., holotype, x 1000.

^ V

PLATE 54

liengjarern et al., Eocene/Oligocene dinoflagellates

486

PALAEONTOLOGY, VOLUME 23

Dimensions. Holotype, central body length 50 ^ m , breadth 59 ^m, maximum length of processes 10 ^m.

Range. Central body length 41(47-6)52 ^ m , breadth 48(57-4)64 ^m, processes length (maximum) 6-20 ^m.
Specimens measured— 20.

Comparison. In the ragged distal appearance of the ornament, this species resembles Glaphyrocysta
divaricata (Williams and Downie 1966), but no process complexes are defined in the latter where the
processes are united distally by trabeculae bearing free aculei and/or by perforated membranes in a
more complex fashion than in G. paupercula.

G. paupercula also resembles G. intricata (Eaton 1976), G. texta (Bujak 1977), and G. micro-
fenestrata (Bujak 1977), where individual process complexes may also be distinguished. However,
the distal connections between processes in those species are always more complex than in
G. paupercula, while the processes are rarely, if at all, united distally. G. paupercula may be a
degenerate offshoot of this lineage.

Genus impletosphaeridium Morgenroth 19666
Type species. Impletosphaeridium transfodum Morgenroth 19666

Impletosphaeridium severinii (Cookson and Cranwell 1967) comb. nov.

1967 Baltisphaeridium severinii Cookson and Cranwell, p. 208, pi. 3, figs. 1, 2.

Comments. This species is transferred to Impletosphaeridium in view of its solid processes. Some
specimens appear to show archeopyle sutures; if these eventually prove to be consistent, then
I. severinii may have to be transferred once more possibly to Eocladopyxis.

Genus phelodinium Stover and Evitt 1978
Type species. Phelodinium pentagonale (Corradini 1973) Stover and Evitt 1978

Phelodinium pachyceras sp. nov.

Plate 54, fig. 4

Name derivation. Greek, pachys, large, keros, horn.

Diagnosis. Phelodinium characterized by apical and antapical horns, triangular in outline, proximally
broad, and distally rounded. Thin-walled cysts moderately compressed dorso-ventrally. Endocyst
sub-circular, with low apical and antapical lobes. Apical and antapical pericoels well developed;
a narrow ambital pericoel may occur between the horns.

Pericyst ornament atabular of reduced spinules. Pericingulum margins indicated by folds on the
periphragm. Perisulcus broad and shallow.

Description. The cyst is thin-walled and usually compressed dorso-ventrally. The ambitus has convex sides and
is projected into three prominent horns; these are triangular, with a broad base and a blunt distal ending,
and are subequal in size. The epipericyst is more or less conical and somewhat larger than the hypopericyst;
the posterior margin of the hypopericyst is straight or slightly concave.

The endocyst is rounded, only weakly bilobed posteriorly; a rounded, low projection into the base of the
apical horn may occur. The pericoels are well developed beneath the horns, a narrow pericoel is commonly
present between the antapical horns. The ornament is reduced to small spinules or granules, apparently
atabular in distribution. Cingulum relatively wide, not indented; its margins are marked by two parallel folds
on the periphragm. The sulcus is very broad posteriorly but narrows markedly towards the cingular zone.

The archeopyle is difficult to observe due to the opercula remaining nearly always in place, but the wide
posterior archeopyle suture (H4), lying very close to the cingular margin, is evident on most specimens observed.

Holotype. Slide ML 1454, H19/0, sample B6, Upper Hamstead Beds, Lower Oligocene, Bouldnor Cliff, Isle of
Wight.

LIENGJ ARERN ET AL.: EOCENE/OLIGOCENE DINOFLAGELLATES

487

Dimensions. Holotype, pericyst length 75 /am, breadth 53 /xm, endocyst length 46 ^m, breadth 53 /on, apical
horn 12 /un, left antapical horn 15 /xm, right antapical horn 13 /%m.

Range. Pericyst length 57(65)77 /xm, breadth 45(51-6)56 /xin, apical horn 6(9)12 /xm, left antapical horn
9(12)14 /xm, right antapical horn 8(10)13 /xm. Specimens measured— 12.

Distribution. Upper Hamstead Beds (B6, B8), ?Middle Headon Beds WC19.

Comparisons. The prominent broad horns and reduced ornament, as well as a strong dorso-ventral
compression, distinguish P. pachyceras from the other species allocated to this genus.

Phelodinium pumilum sp. nov.

Plate 54, fig. 2

Name derivation. Latin, pumilus, dwarf.

Diagnosis. Phelodinium of small size, ambitus bilaterally asymmetrical with reduced antapical horns,
right antapical broadly rounded, may be absent. Apical horn small, cylindrical with prominent distal
pore. Pericingulum relatively wide, marked by folds. Sulcus distinct.

Description. The ambitus varies from subcircular to distinctly peridinioid: the bilateral asymmetry of the cyst
is nearly always evident. The dorso-ventral compression is strong. The pericoels, if observable, are restricted to
the cavities beneath the horns. The cylindrical apical horn is distinctive, its truncated distal tip bears a
prominent pore bordered by a thickening of the periphragm. The left antapical horn is always developed and
is sharply pointed distally. The right antapical horn is often absent but commonly it is represented by a broad
lobe.

The periphragm is very thin and transparent and is often folded. The cingulum is only very slightly
helicoid, wide in relation to the over-all size of the cyst; anterior and posterior cingular sutures are indicated
by low smooth ridges formed by folding of the periphragm. The perisulcus is distinct.

The archeopyle is of a type and shape seen in species of Phelodinium. Peri- and endoperculum are indistin-
guishable. The operculum may remain attached along its posterior suture.

Holotype. Slide ML 1450, Q45/4, sample WC 23, Middle Headon Beds, Upper Eocene, Whitecliff Bay, Isle of
Wight.

Dimensions. Holotype, pericyst length 64 /xm, breadth 54 /am, apical horn 6 /xm, left antapical horn 5 /xm,
right antapical horn 8 /xm.

Range. Pericyst length 50(55)62 fim, breadth 41(46-5)54 /xm, apical horn 3-5(4-5)6-4 (xm, left antapical horn
2-7(4-5)6-5 /xm, right antapical horn 0(1)3 /xm. Specimens measured— 11.

Comparisons. The small size, rounded ambitus, bilateral asymmetry, and distinctive apical horn
distinguish this species from all known Phelodinium species. Allocation to Phelodinium is based on
the archeopyle shape and relative size, the absence of well-defined pericoels and the very strong
dorso-ventral compression.

Distribution. Samples WC18, 20, 21, 23, and 25.

Genus phthanoperidinium Drugg and Loeblich 1967
Type species. Phthanoperidinium amoenum Drugg and Loeblich 1967.

Phthanoperidinium amiculum sp. nov.

Plate 53, fig. 4

Name derivation. Latin, amiculum , cloak.

Diagnosis. Phthanoperidinium with ambitus rounded-pentagonal to suboval. Epicyst with convex
sides, terminating in a short apical horn, hypocyst also rounded, produced into one, very occasion-
ally two, antapical horns. Peri- and endophragm very closely appressed except beneath the horns.

PALAEONTOLOGY, VOLUME 23

where restricted pericoels develop. Periphragm ornamented with intratabular spinules and peni-
tabular to hyaline sutural ridges with smooth to slightly denticulate free edges. Laevigate to striate
pandasutural lines may be distinct. Pericingulum and perisulcus laevigate, bordered by membranes.

Description. The pericyst is fusiform in lateral view; the ambitus is rounded-peridinoid to subcircular or sub-
oval. The apical horn is short, trangular and distally blunt. The left antapical horn is usually well developed.
On some specimens, a right antapical horn, very much reduced, may occur; on most specimens, a projection
of the sutural ridges takes the place of the right antapical horn.

The intratabular spines are small and solid, distally short or somewhat capitate; those closer to the paraplate
periphery may be arranged in a penitabular ring. The ridges are hyaline and imperforate, their free margins
are entire or very slightly serrate to denticulate; the height of the ridges normally does not exceed 3 /j.m, except
along the cingular sutures where they may be up to 5 ^m in height. The ridges may be parasutural or peni-
tabular in position. Narrow laevigate pandasutural zones are normally observable on parts of the pericyst
and, on some specimens, very faint striations, perpendicular to the margin of the paraplate, may be observable.

The paratabulation formula and shape of the paraplates are normal for the genus. The pericingulum is
helicoid, its ends being offset about one pericingular width; its surface is laevigate. The perisulcus is relatively
narrow, moderately excavated, extending anteriorly to nearly a half of the epicyst height. The archeopyle is
formed by the detachment of paraplate 2a, but it is only rarely observable. Occasionally, additional sutures
occur along the margins of all three intercalary plates.

Holotype. Slide ML 1451, K23/4, sample WC25, Middle Headon Beds, Upper Eocene, Whiteclilf Bay, Isle of
Wight.

Dimensions. Holotype, pericyst length 63 /im, breadth 48 p.m, apical horn 7 /im, left antapical horn 5-5 /un.

Range. Pericyst length 47(55-5)63 ^m, breadth 40(43)48 /urn, apical horn 3(5-5)7 /un, left antapical horn
3(5-5)7 p.m. Specimens measured — 10.

Comparisons. P. eocenicum (Cookson and Eisenack 1965) appears to have sutural ridges and intra-
tabular granules, and thus resembles P. amiculum in the style of ornament; but the ambitus in
P. eocenicum is fusiform to subpolygonal, less rounded than P. amiculum and the left antapical
horn lies closer to the median axis; in addition both intratabular granules and sutural ridges are
much more reduced than on the present species.

P. alectrolophum Eaton 1976 resembles P. amiculum in possessing sutural-penitabular ridges, but
these bear well-developed spines on their free margins and the intratabular paraplate surfaces are
smooth.

Distribution. Only in sample WC25.

Phthanoperidinium flebile sp. nov.

Plate 54, fig. 6

1978 Geiselodinium cf. geiseltalense Krutzsch, Chateauneuf 1978.

Name derivation. Latin, flebilis, pathetic.

Diagnosis. Phthanoperidinium with ?partial (not continuous) endophragm occasionally developed
beneath the horns. Ornament intratabular of small echinae or setae, laevigate sutural bands may be
observable. Cingulum indicated by a relatively broad equatorial band free of ornament.

Description. The autocyst ambitus is subcircular to oval, but is frequently folded and the ambitus may appear
somewhat fusiform; the ambital outline is little affected by the horns. The apical horn is very short, sub-
triangular to rectangular in outline; its apical margin may be smooth or may bear a tuft of short spines, to
which sometimes the entire horn is reduced. The hypocyst is posteriorly rounded, and may bear a very short,
sharp, antapical horn slightly to the left of the median line.

The autophragm is thin and bears a variable number of small setae or echinae, sometimes reduced to
granules, atabular to intratabular in distribution; on some specimens the number of spines is reduced, and
these may adopt a penitabular arrangement. Sutural bands, when observable, are smooth and of variable width.

LIENGJARERN ET AL.: EOCENE/OLIGOCENE DINOFL AGELL ATES

489

The cingulum, observable on some specimens, appears as a relatively wide band free of ornament; it is not
indented. The sulcus has only been seen on one specimen, appearing as a very broad, slightly depressed area with
ornament more sparse than on the rest of the ventral autocyst face.

The archeopyle, rarely observable, is intercalary and formed by the loss of paraplate 2a; additional splitting
may sometimes develop along the lateral sutures of paraplate 3", but only very rarely, along the sutures of the
remaining paraplates in the intercalary series.

Holotype. Slide ML 1453, X27/3, sample H24, Lower Hamstead Beds, Lower Oligocene, Hamstead, Isle of
Wight.

Dimensions. Holotype, autocyst length 39 /xm, breadth 28 /xm, apical horn 5 ^m, antapical horn 1 /xm.

Range. Autocyst length 31(35)42 /xm, breadth 22(27)31 /am, apical horn 1(3)5 /xm, antapical horn 0(1)2 /xm.
Specimens measured— 20.

Distribution. Sample H24; Lower Hamstead Beds.

Discussion. P. echinatum most closely resembles P. flebile in its ornament of spines, but in
P. echinatum these are sutural to penitabular (distribution as a single simulate ring), whereas they
are intratabular to atabular in P. flebile.

Occurrence. Sample H24, and at base of Sannoisian in Paris Basin (Argile Verte de Romainville).

Genus thalassiphora Eisenack and Gocht 1960
Type species. Thalassiphora pelagica (Eisenack 1938) Eisenack and Gocht 1960

Thalassiphora fenestrata sp. nov.

Plate 54, fig. 1

Name derivation. Latin, fenestratus, windowed.

Diagnosis. Thalassiphora with partial fenestration of the periphragm. The fenestration is restricted
to the lateral and ventral areas of the periphragm. The extent of the fenestrated area is variable,
but it never extends over the whole dorsal region. The perforations are large, more or less circular,
and may be closely packed forming an irregular reticulum. The ventral flange of the pericyst is narrow
and is fenestrated throughout.

Description. This species is similar to T. pelagica in shape and in wall structure but the extension of the
periphragm on the ventral side appears to be more reduced than is common in T. pelagica, that is, the
ventral lacuna is larger. Perforations develop in the periphragm in ventral and lateral areas and disappear
towards the mid-dorsal area. Between these perforations, the fibres are more loosely packed. A large number
of smaller perforations occur between the larger fenestrations, the latter are of variable diameter tending to be
larger closer to the ambitus. Ventrally, the pericyst occurs as a relatively narrow flange which is strongly
fenestrate throughout. The antapical keel may often be reduced or, sometimes, absent.

Holotype. Slide ML 1449 U16/2, sample WC14, Middle Headon Beds, Whitecliff Bay, Isle of Wight.
Measurements. Holotype, endocyst 81 x 67 /xm, pericyst diameter 150 /xm.

Range. Endocyst 73(77)89 x 59(67)77 /xm, pericyst diameter 126(154)182 /xm. Specimens measured— 10.

Comments. This species, which is apparently restricted in distribution to the latest Eocene and ?early
Oligocene, seems to be an intermediate form between T. reticulata Morgenroth 1966a, which is
characteristic of younger Oligocene deposits and whose pericoel is fenestrate virtually all over, and
T. pelagica.

Distribution. Samples WC 13-23.

490

PALAEONTOLOGY, VOLUME 23

Genus vectidinium gen. nov.

Name derivation. Latin, Vectis, Roman name for the Isle of Wight.

Type species. Vectidinium stoveri sp. nov.

Diagnosis. Single-walled proximate peridinioid cysts, moderately compressed dorso-ventrally,
ambitus subpentagonal or subcircular to oval or somewhat fusiform. Epicyst and hypocyst of
approximately equal size. Epicyst may or may not extend into a short apical horn; apical pore always
present. Hypocyst semicircular or bilobed; left antapical horn present or absent, right antapical horn
commonly present.

Autophragm with atabular or intratabular to penitabular ornament of small granules, spinules or
baculae, which may be reduced in size and/or number. Narrow laevigate pandasutural zones may be
observable. Paratabulation formula, when determinable, 4', 3a, 1", Oc, 5", 2'"', Os. When
observable paraplate 1" is rhombic, antero-posteriorly elongate, and relatively large.

Cingulum and sulcus distinct. The cingulum is wide relative to over-all autocyst size, not indented,
non- or moderately helicoid. Sulcus shallow and broad on the hypocyst. Archeopyle combination
type 31 3P 3"-5", accessory sutures may occur along cingular margin of the remaining precingular
paraplates. Opercula free.

Comparisons. Vectidinium differs from Palaeoperidinium Deflandre 1934, and from Saeptodinium
Harris 1975, in that the apical paraplate 3' is not included in the archeopyle. From Saeptodinium it
also differs in being single walled and usually having intratabular or penitabular ornament. From
Palaeoperidinium it differs in the presence of ornament and its much smaller size.

Ginginodinium Cookson and Eisenack 1960, Laciniadinium McIntyre 1975, and Lunatodinium
Brideaux and McIntyre 1973, all have a 31 3P 3"-5" archeopyle, and they also resemble Vectidinium
in the type of ornament. Ginginodinium is double walled, and in the formation of the archeopyle the
three dorsal precingular paraplates (3"-5") always remain attached along their cingular margins
(Lentin and Williams 1975, p. 95). Laciniadinium has a single opercular piece 31 3P 3"-5" which
always remains attached to the cyst along its posterior margin, like a flap. In Vectidinium whenever
the archeopyle is present, the operculum is detached and some doubt remains as to whether this is
simple or compound. Lunatodinium (a Lower Cretaceous genus) was described as having an archeo-
pyle formed by the loss of the three dorsal precingular paraplates. However, Lentin and Williams
(1975, pp. 96 and 116) included this genus in the pericysts, possessing a 31 3P archeopyle. This
appears to be so from the original illustration of Lunatodinium (Brideaux and McIntyre 1973,
figs. 1-13). The genus is stated to have a circular or subcircular outline.

Cysts of the Recent freshwater dinoflagellate Peridinium resemble Vectidinium in the type and
distribution of the ornament, but they are normally cavate and the archeopyle is formed by the
detachment of plates along a transapical suture, type A3I3P.

Vectidinium stoveri sp. nov.

Plate 54, fig. 7

Name derivation. This species has been named after Lew Stover.

Diagnosis. As for the genus.

Description. The dorso-ventral compression of these cysts is normally slight, and some specimens may be
oriented in apical or antapical view; in lateral view the cysts are somewhat fusiform or oval. The epicyst has
strongly convex sides which may merge imperceptibly in a very short, blunt apical horn with a solid tip on
which sits a pore; the apical horn may be absent, and the epicyst apex is then invaginate. The hypocyst is com-
monly broadly rounded posteriorly, but some specimens may show a weak bilobation on the antapex. The
short, eccentrically located left antapical horn may be present or absent.

The ornament varies in density and shape. When the ornament is baculate or of short processes their distal
endings are often T-shaped and may be linked to those from near-by processes, giving the appearance of a

LIENGJ ARERN ET AL EOCENE/OLIGOCENE DINOFL AGELLATES

491

tectum supported by columellae in optical section; sometimes the ornament is very reduced in size and mostly
consisting of granules. The ornament may be densely or sparsely arranged on the paraplate surface, the most
peripheral elements tending to be arranged along simulate rings. Laevigate pandasutural zones, usually narrow,
are present but are not always clearly visible.

Cingulum and sulcus are distinct, both being marked by low ridges or folds on the autophragm. The
cingulum is relatively wide, slightly helicoid or circular, not indented; intratabular ornament and smooth
pandasutural zones may be observable on the cingular surface, but the number of cingular paraplates has not
been determined with certainty. The sulcus is also broad and shallow, and extends approximately half-way to
the apex. The shape and relative size of individual paraplates are difficult to determine because of very small
size and transparent autophragm of these cysts.

When present, the archeopyle is formed by complete detachment of plates la-3a, 3"-5". On some specimens,
accessory archeopyle sutures develop along most of the anterior margin of the cingulum, but both portions of
the cyst usually remain attached along a narrow band, presumably corresponding to the sulcus. The operculum
is always free, but it has not been possible to determine whether this is formed by a single piece or is
compound, since isolated opercula have not been observed— a fact suggesting that the operculum may be
compound, disintegrating into the very small individual paraplates which would easily be lost in sieving of the
organic residue during preparation.

Holotype. Slide ML 1452, U43/3, sample WC34, Upper Headon Beds, Upper Eocene, Whitecliff Bay, Isle of
Wight.

Measurements. Holotype, autocyst length 37 ^m, breadth 42 ^m, apical horn 1 /un, left antapical horn 1 ^m,
width of cingulum 4 ^m.

Range. Autocyst length 30(35-5)41 ^m, breadth 24-5(31)42 ^m, apical horn 0(2)4-2 /urn, left antapical horn
0(l)4-5 /xm, width of cingulum 2-7(3-6)4 ^m. Specimens measured — 24.

Distribution. The distribution of Vectidinium stoveri in the section studied deserves some special
attention since it constitutes monospecific assemblages at some horizons, and has not been found in
association with any other dinoflagellate cysts. These horizons yield ostracod assemblages of type III
(Keen 1972, 1977); these have been stated by Keen to indicate brackish-water conditions (salinity
3-9%). V. stoveri is thought to be a non-marine dinoflagellate cyst, and possibly a good indicator
of oligohaline conditions; it is recorded from samples WC34, 35, and HI 9.

PALYNOLOGICAL ASSEMBLAGES AND DEPOSITIONAL ENVIRONMENTS

The Upper Eocene-Lower Oligocene of the Isle of Wight was deposited under widely variable
environmental conditions. The area of deposition has been likened to an embayment, limited to the
north and south by the Portsdown and the Sandown-Brixton anticlines respectively, and opening
towards the sea to the east and south-east. At times this sea penetrated into the basin. At other
times an eastward flowing river system occupied the area (Keen 1977). The conditions ranged from
shallow, near-shore open sea, to brackish-water lagoons— with or without connection to the sea—
to freshwater lacustrine or fluviatile environments. These changes are reflected in the palyno-
assemblages, and are especially noticeable in the relative proportions of different classes of palyno-
morphs as well as in the composition of the microplankton assemblages where these occur.

Palaeoecological studies of palynomorph assemblages and particularly of dinoflagellate cysts are
currently in their preliminary stages, and no work on the palaeoenvironmental interpretation of
Tertiary palyno-assemblages from paralic areas has yet been published. However, the assemblages
recovered here may be correlated to particular environmental conditions by using, as a control, the
existing information on the distribution of dinocysts in Tertiary to Recent sediments, as well as the
sedimentological and faunal evidence available from the sections studied. The foraminifera
(Murray and Wright 1974), molluscs (Daley 1973), and ostracods (Keen 1972, 1977; Haskins 1969)
from the Upper Eocene-Lower Oligocene sections of the Isle of Wight have yielded a considerable
volume of data that can be used in assessing the meaning of the palynological assemblages recovered.

492

PALAEONTOLOGY, VOLUME 23

The major components of the palynological assemblages are indicated in Table 2. They clearly
fall into two groups, one with marine dinoflagellates present; the other non-marine samples contain
only terrigenous freshwater or lagoonal elements.

The non-marine group shows considerable variation, particularly in the proportions of Pediastrum
Meyen, which may contribute from 0 to over 90% of the assemblage. In some samples there is also a
considerable contribution from non-marine dinoflagellates. These non-marine samples are asso-
ciated with various lithologies ranging from limestone through to sands and no particular pattern
has so far been determined. It is evident, particularly from the work of Keen, that the salinities vary
from fresh to oligohaline water. The environments of deposition include evidently freshwater
lacustrine, fluvial, flood-plain, and bay-head situations. The control over the relative abundance of
Pediastrum Meyen is not understood. It is notably more common in the Bembridge Marls in the
west of the island. In marine sediments it is present only in very small numbers and is probably
allochthonous. It is most abundant in situations that could be interpreted as oligohaline water.

table 2. General character of palynological assemblages. P & S— pollen and spores; Ped —Pediastrum spp.;
MD- marine dinoflagellates; fd— freshwater dinoflagellates; ‘r’ indicates that dinoflagellates are all reworked

from older strata.

Sample % P & S % Ped % MD % fd
Whitecliff Bay

Bembridge Marls

WC67

99

—

l(r)

—

WC66

99

1

x (r)

—

WC65

94

2

4(r)

—

WC64

99

1

x(r)

—

WC63

91

8

l(r)

—

WC62

74

26

—

—

WC61

100

—

—

—

WC60

100

—

—

WC59

97

3

—

—

WC58

100

_

—

WC56

99

1

—

WC55

13

80

7

—

WC55A

16

84

X

—

WC54

100

—

—

—

WC53

100

—

—

—

Bembridge Limestone
WC51 100

WC49

100

—

—

—

Osborne Beds
WC47

100

WC46

58

42

—

—

WC45

36

64

—

—

WC44

100

—

—

—

WC43

100

—

—

—

WC42

10

90

—

—

WC41

97

3

—

—

Upper Headon Beds
WC40 22

78

WC39

86

14

—

WC38

97

3

—

WC37

93

6

—

1

Sample

% P & s

% Ped

% MD

% fd

WC36

99

1

x

WC35

62

1

—

37

WC34

73

14

—

13

WC33

94

6

—

_

WC31

62

38

—

—

WC30

100

—

—

—

WC29

61

39

—

—

WC28

90

10

—

—

WC26 100

Middle Headon Beds

—

—

—

WC25

74

—

26

—

WC24

84

16

—

WC23

54

46

—

WC21

58

2

40

—

WC20

40

—

60

—

WC19

31

—

69

—

WC18

28

—

72

—

WC17

38

—

62

—

WC16

29

—

71

—

WC15

25

75

—

WC14

20

80

—

WC13 37

Lower Headon Beds

X

37

WC12

72

28

—

—

WC9

80

20

—

—

WC8

45

55

—

—

WC7

25

75

—

—

WC6

100

—

—

—

WC5

7

93

—

—

WC4

80

20

—

—

WC3

96

4

—

—

WC2

100

—

—

—

WC1

98

2

—

—

LIENGJ ARERN ET AL .: EOCENE/OLIGOCENE DINOFL AGELLATES

493

Sample % P & S % Ped % MD % fd

Hamstead Cliff
Lower Hamstead Beds

H36

54

46

—

—

H35

55

45

—

—

H34

88

12

_

—

H33

82

18

—

—

H32

75

25

—

—

H31

63

37

—

—

H30

81

19

—

—

H29

41

59

—

—

H28

64

36

—

—

H27

97

3

—

—

H26

96

4

—

—

H25

96

3

—

—

H24

94

1

6

—

H23

42

13

45

—

H22

84

16

-

—

H21

87

13

x

—

H20

78

22

—

—

H19

63

29

—

8

H18

76

24

—

—

H17 73

Bembridge Marls

27

—

—

H16

76

24

—

—

H15

34

66

—

—

H14

67

33

—

—

H13

64

36

—

—

H12

36

64

—

—

H10

34

66

—

X

Sample

% P & S

% Ped

% md

% fd

H9

57

43

_

x

H8

75

25

_

x

H7

73

27

_

_

H6

35

6

59

_

H4

49

X

51

_

H3 44

Bembridge Limestone

56

H2

16

84

_

—

HI 100

Bouldnor Cliff
Upper Hamstead Beds

B15

92

—

8

_

B14

94

—

6

_

B13

96

4

_

_

B12

100

—

_

_

B 1 1

24

2

74

_

BIO

75

25

—

_

B9

42

58

_

_

B8

8

3

89

—

B7

28

2

70

_

B6

52

29

19

_

B5 63

Lower Hamstead Beds

37

x

—

B4

100

_

_

_

B3

100

_

_

_

B2

100

—

—

—

B1

52

48

—

_

Non-marine dinoflagellates are represented by a single species, Vectidinium stoveri which is present
only in three samples, WC34, 35, and H19. It is associated with ostracod assemblage III of Keen,
indicating brackish-water conditions.

Marine samples are characterized by the presence of marine dinoflagellate cysts and acritarchs.
They can be classified into a number of types according to their diversity and the dominant species.
Since these types occur in stratigraphic order and are associated with a series of marine incursions it is
convenient to discuss them in stratigraphic sequence.

The Middle Headon Beds transgression
Four assemblage types are present:

Assemblage 1. The Brockenhurst Bed and Psammobia Beds (samples WC13-21) are characterized by
assemblages with forty or more species of dinoflagellate cysts dominated by Homotryblium plectilum
which makes up 30-70% of the microplankton; other abundant species are Spiniferites ramosus,
Adnatosphaeridium reticulense, and Phthanoperidinium cometum. These assemblages are associated
with ostracod assemblage type VI and indicate open-sea conditions, the major transgressive episode
in the sequence studied.

Assemblages 2-4. The succeeding Venus Bed contains three different assemblage types showing a
marked reduction in the number of species present and in their relative abundance.

Type 2, occurring in sample WC23, has less than thirty species and is dominated by H. pallidum
and P. cometum, the latter a species evidently tolerant of reduced salinities in estuarine or lagoonal
environments.

494

PALAEONTOLOGY, VOLUME 23

Type 3, occurring in sample WC24, has only seventeen species and is dominated by broken species
of H. plectilum associated in assemblage 1 with open-sea conditions. Here these are thought to be
allochthonous. H. pallidum is the next most common species.

Type 4, occurring in sample WC25, is dominated by Eocladopyxis tessellata and P. cometum.

These three assemblages appear to indicate a period of regression with restriction of marine access
to the area. Keen refers the ostracod assemblages in these beds to his type V, indicating salinities in
the range of 16-5-33%.

The Lower Bembridge Marl transgression

Assemblage types 5-7 are associated with the Oyster Bed.

In the east, sample WC55 yielded assemblage type 5, where dinoflagellates made up only 7% of the
palynomorphs. No clearly dominant species was present, the commonest being Chiropteridium
aspinatum, Glaphyrocysta microfenestrata, Homotryblium pallidum, and Paralecaniella indentata.

In the west, assemblage type 6 is monospecific; Phthanoperidinium levimurale makes up 51% of the
palynomorphs in sample H4. Assemblage type 7 is also monospecific, G. microfenestrata making up
59% of the palynomorphs in sample H6.

The significance of these three diverse assemblages from the Oyster Bed is made clearer by
consideration of the fauna. Molluscs, foraminifera, and ostracods all indicate brackish estuarine
conditions. Assemblage type 5 is associated with Keen’s type V indicating near-marine conditions;
the assemblages from the west, however, are associated with his type IV, indicating lower salinities
(9-16%). This seems to mean that the monospecific assemblages with P. levimurale and G. micro-
fenestrata are composed of more or less stenohaline species, since both also occur in open marine
conditions. They appear to have flourished in this estuarine situation since they are particularly
abundant, more so than any of the species in the east, where the assemblage, although poorer in
relative numbers, has a greater variety of marine species and, although still estuarine, appears to have
better connection with the open sea.

The Lower Hamstead Bed transgression

Assemblage types 8 and 9 are associated with a marine incursion at the horizon of the Nematura
Band.

Assemblage type 8, sample H23, contains only four species and is dominated by Adnatosphaeridium
reticulense. Only 13% of the palynomorphs are dinoflagellates. Assemblage type 9, an even poorer
assemblage from H24 immediately above, is on the other hand dominated by P. flebile. Ostracods
from the Nematura Band show the presence of assemblage type IV characteristic of mesohaline
conditions.

The Upper Hamstead Bed transgressions

Six different dinoflagellate assemblages (types 10-15) have been found in the Upper Hamstead Beds
and the palynology appears to show the presence of three different invasions of saline water.

The first incursion corresponds to the Cerithium Bed and contains assemblage types 10-12.
Assemblage type 10, sample B6, contains 19% dinoflagellates with only a few species represented and
is dominated by G. microfenestratum and P. cometum, both of which, although known from other
marine sediments, have previously been noted in assemblage types 7 and 4 and 2, with reduced
salinities associated with Keen’s types IV and V. Keen (1972) finds that the Cerithium Bed also yields
assemblages of types IV and V. Assemblage type 1 1 in sample B7 is also impoverished in species, but
is dominated by small acritarchs of the Micrhystridium group, which accounts for about 60% of the
palynomorphs. Assemblage type 12, sample B8, is more varied and richer in numbers, but G. pauper-
cula accounts for most of these.

Taken together these three samples indicate a marine influence, which, however, did not achieve
fully marine conditions in this locality, the area remaining meso- to polyhaline.

The second incursion is represented only by assemblage type 13, sample Bll. That it is a
separate episode is indicated by the intervention of samples B9 and 10 which contain only terrigenous

LIENGJARERN ET AL.. EOCENE/OLIGOCENE DINOFL AGELL ATES

495

pollen and spores and the ?freshwater alga Pediastrum. Assemblage type 13 appears to represent
more fully marine conditions with many new species appearing. The dominant species is H. pallidum,
which also dominates in assemblage type 2 ( Venus Bed) and is abundant in type 5 (Oyster Bed,
WhiteclifF Bay). Here it is associated with Gerdiocysta conopeum. The conditions indicated are still
not yet fully marine, but must closely approach that condition.

The third incursion is represented by assemblage types 14 (sample B14) and 15 (sample B15). That
this is a separate episode is indicated by the intervention of the purely terrigenous palynological
assemblages in samples B12 and 13. The second and third incursions together form the Corbula Bed.
Assemblage type 14 is a poor monospecific one comprising only Phthanoperidinium cometum. It
probably indicates low salinities. Type 15, however, is somewhat richer and is particularly so in the
variety and lack of any clearly dominant species. Micrhystridium , Lejeunia tenella, Hystrichokolpoma
salacium, and P. amoenum are prominent, the last three being known only from open marine
sediments. It is believed that these two samples B14 and B15 represent the beginning of a major
transgression, the culmination of which is not represented due to erosion of the succeeding beds.

DINOFLAGELLATE CYST STRATIGRAPHY

The distribution of dinoflagellates is shown in Table 1.

The first dinoflagellate assemblages appear in the Brockenhurst Bed associated with the Middle
Headon transgression. Detailed comparison between the dinoflagellate assemblages from the Solent
Formation and the marine sediments of the underlying Barton Formation is not possible at present,
since little information on the dinoflagellate content of the Barton Beds has so far been published
(Bujak 1976). However, from unpublished evidence (Bujak 1973), it appears that, notwithstanding
the intervening regression represented by the Becton and Lower Headon Beds, only minor changes
take place in the composition of the assemblages between the uppermost marine beds of the Barton
Formation and the lower part of the Solent Formation (Middle Headon Beds). The number of species
that first appear in the Middle Headon Beds is very small, but they include Rhombodinium perforatum
and Thalassiphora fenestrata, and the possibility remains that some of these may also occur in the
Barton Beds; the number of apparent extinctions is also limited, and their stratigraphic significance,
which may be only local, cannot be assessed at this stage.

As the assemblages become impoverished towards the upper part of the Middle Headon Beds,
among the dinoflagellate species disappearing from the assemblages are Areosphaeridium diktyo-
plokus, Cordosphaeridium funiculatum, Distatodinium ellipticum, Palaeocystodinium golzowense,
R. draco, R. perforatum, and T. velata.

Other taxa, Emslandia sp., Eocladopyxis tessellata, and Phelodinium pumilum, make their first
appearance in the section here. These species first appearing within the upper part of the Middle
Headon Beds are all new and so their stratigraphic value, if any, cannot be stated.

The Bembridge transgression, represented by the Oyster Bed, yields poorly diversified assem-
blages. These, in terms of their species content, show a somewhat closer relationship to the Middle
Headon Beds than to the Upper Hamstead Beds. The Bembridge Oyster Bed at Whitecliff Bay
registers the last known occurrence in England of Chiropteridium aspinatum, Impletosphaeridium
severinii, Homotryblium oceanicum, and Leptodinium incompositum.

The Lower Hamsted Bed transgression, represented by a thin sequence including the Nematura
Bed, also provides a poor assemblage consisting mainly of long-ranging species. One species,
Phthanoperidinium flebile is, however, apparently confined to this horizon.

A very pronounced break in the dinocyst succession is evident in the final transgressions of the
Upper Hamstead Beds. Out of a total of sixty-eight dinoflagellate species recorded, only nineteen
are common to the Solent and Hamstead Formations; thirty-four species disappear below the base
of the Hamstead Beds, and fifteen species are first recorded within the latter. The marked renewal
of the assemblages registered between the two main marine episodes in the sequence is to some extent
environmentally controlled, since some of the species missing in the Headon Bed are known to persist
elsewhere into the Oligocene, such as C. aspinatum, Cordosphaeridium cantharellum, D. ellipticum.

496

PALAEONTOLOGY, VOLUME 23

Hystrichokolpoma rigaudiae, Kisselovia coleothrypta, R. draco, T. velata, and T. pelagica. Two
species, however, which fail to reappear are R. perforatum and A. diktyoplokus, whose absence seems
to be stratigraphically important.

A number of species make their first appearance here and some of them are thought to be
stratigraphically important. These are Gerdiocysta conopeum, Heteraulacacysta cf. companula,
Phthanoperidium amoenum, Wetzeliella gochtii, and W. symmetrica incisa. Other appearances of
possible significance are Phelodinium pachyceras and D. scariosum.

CORRELATION WITH OTHER EUROPEAN AREAS

Paris Basin

Curry et al. (1978) correlate the Middle Headon Beds with part of the Marnes a Pholadomya
ludensis, i.e. with the deposits of the Ludian transgression of the Paris Basin. Both formations yield
rich dinoflagellate assemblages. A description of those from France has been given by Chateauneuf
(1978). Most of the species recorded by him are present in the Middle Headon Beds but there is
none of sufficiently restricted range to allow confident correlation on the basis of the dinoflagellates,
except that R. perforatum (which appears for the first time in the mid-Headon Beds in England)
also appears for the first time in small numbers in the top Marinesian and more commonly in the
Ludian. R. perforatum, previously mentioned from the Barton Beds (Costa and Downie 1976) is in
fact a separate species (Bujak, in press). A marked distinction between the Ludian assemblages and
those from the Headon Beds is the remarkable abundance of H. plectilum in the Isle of Wight and
its apparent absence from the Ludian.

The impoverished assemblages from the Bembridge Oyster Bed yield little of correlative value, but
the abundance of C. aspinatum does correspond with the prominence of this species in assemblages
from the Ludian Marnes a Lucines (Chateauneuf 1978).

The equally poor assemblages from the vicinity of the Nematura Band do, however, show some
marked similarities to those of the Argile Verte de Romainville at the base of the Stampian. The lower
of the English samples (H23) is dominated by Adnatosphaeridium reticulense, which is also a
dominant form in the Argile Verte. The upper English sample (H24) is dominated by Phthanoperi-
dinium fiebile, which is restricted to this horizon in England and has also been found to be
abundant in the Argile Verte by Chateauneuf (1978) and recorded by him under the name of
Geiselodinium cf. geiseltalense. This strongly suggests a correlation between the Nematura Band and a
horizon within the Argile Verte de Romainville.

The Upper Hamstead Beds can be correlated with the Calcaire de Sannois and the lower part of the
Marnes a Huitres. This correlation is supported by the appearance of Gerdiocysta conopeum
( = Cyclonephelium reticulosum Gerlach, Chateauneuf 1978), W. gochtii (Chateauneuf, pers. comm.),
P. amoenum, and the increased abundances of W. symmetrica and Pentadinium taenigerum
(Chateauneuf 1978) in both areas.

The overlying Sables de Fontainebleau have a rich and varied dinoflagellate assemblage with
species such as Chiropteridium lobospinosum and C. partispinatum (Chateauneuf 1978). In England
there is no representative of this assemblage, which has marked similarities to those from the
Rupelian of Germany (Benedek 1972).

Belgium

Weyns (1970) described two assemblages from the Sables de Grimmertingen (Lower Tongrian). He
listed forty-seven forms of dinoflagellate cysts. Of these thirty-six are apparently present in the
Middle Headon Beds, and the assemblages have a general similarity, particularly in the prominence
of Homotryblium and Spiniferites.

In comparison with the Hamstead Beds assemblages, there are major differences. The many
species appearing for the first time in the Hamstead Beds are not listed in Weyns’s assemblages. Only
a few of the species listed by Weyns appear to have stratigraphic significance. Glaphyrocysta micro-
fenestrata (= C. semicirculatum in Weyns) does not appear until late in the Chama Beds of the

LIENGJARERN ET AL.\ EOCENE/OLIGOCENE DINOFL AGELL ATES

497

Bartonian (Bujak 1976). G. exuberans ellipsoidalis and Areosphaeridium diktyoplokus are absent
above the Middle Headon Beds. The correlation that best fits these circumstances is between the
Sables de Grimmertingen and the Middle Headon Beds. This is in agreement with recent work on the
nanoplankton correlation (Cavelier 1975). A notable difference between the Belgian and English
assemblages is the presence of Leptodinium and Nematosphaeropsis in the former. These are forms
found to be more prominent in open-sea situations.

Two samples, one from 20 m and the other from 30 m above the base of the Rupel Clay in the
type section, yielded rich dinoflagellate assemblages. These showed marked similarities to those from
the Upper Hamstead Beds, in particular containing W. gochtii. However, they also contain C. lobo-
spinosum, C. partispinatum and other species which are not present in the Isle of Wight, but are
characteristic of the Sables de Fontainebleau in the Paris Basin, and the Rupelton in Germany.
These samples are clearly younger than any from the Isle of Wight.

THE EOCENE/OLIGOCENE BOUNDARY IN THE ISLE OF WIGHT

Establishment of a standard for this stratigraphic boundary is the subject of continuing debate. In
France, it has commonly been placed at the base of the Stampian Stage, i.e. at the base of the
Argile Verte de Romainville (Chateauneuf 1978). Accepting this, the correlations between the Isle of
Wight succession and the Paris Basin based on dinoflagellates indicate that the boundary lies closely
below the Nematura Band. The boundary clearly lies between the Nematura Band and the Middle
Headon Beds. The Oyster Bed, although it has a poor assemblage, has greater similarity to the
Headon Beds than to the succeeding assemblage.

Therefore, if the French view is accepted the boundary lies between the base of the Nematura
Band and the top of the Oyster Bed. Since the Argile Verte de Romainville marks the first
important marine incursion after the episode of the Marnes a Lucines it seems very likely that the
Nematura Band represents the same transgression. The Bembridge Marls then correlate with the
Supra- and Upper Gypsiferous Groups (1st and 2nd mass) and the Osborne Beds with the 3rd mass
of gypsum. The base of the Oligocene could conveniently be taken at the base of the Hamstead Beds,
some 9 km below the Nematura Band.

An alternative, widely held, view is that the base of the Oligocene originally selected in Germany
should be adopted. This is marked by the transgression associated with the Latdorf (Lattorf) Sands
(NP21), which correlate readily with the Sables de Grimmertingen in Belgium.

Dinoflagellates have not been described from the Latdorf Sands, but from the Sable de Grimmer-
tingen assemblages very like those from the Middle Headon Beds have been described by Weyns
(1970). If this correlation is accepted the Middle Headon Beds would be Oligocene. However, the
Brockenhurst Bed has given evidence of an NP20 age, which indicates that the base should be
higher. There is, however, no apparent break in the Middle Headon Beds sequence, only a progressive
increase in terrigenous influence in the Venus Beds (samples WC22-25). No suitable location for a
boundary is evident.

The next marine incursion in the Isle of Wight succession, the Bembridge Oyster Bed, did not
yield any dinoflagellates of much value in correlation. Those that are present are not inconsistent with
a correlation with the Sables de Grimmertingen and consequently with the placing of the base of the
Oligocene immediately above the Bembridge Limestone, as is done by Curry et al. (1978).

Acknowledgements. We particularly thank Dr. J. Bujak for information regarding the Barton Beds and for
assisting Dr. Liengjarern in the field; Dr. J. J. Chateauneuf for much unpublished data on the Paris Basin; and
Professor D. Curry for helpful comments. Dr. Liengjarern acknowledges the support of a Columbo Plan
Scholarship enabling her to do this research. The collections are housed in the Department of Geology,
University of Sheffield.

PALAEONTOLOGY, VOLUME 23

REFERENCES

benedek, p. N. 1972. Phytoplanktonten aus dem Mittel-und Oberoligozan von Tonisberg (Niederrheingebiet).
Palaeontographica, Abt. B., 137, 1-71.

brideaux, w. w. and mcintyre, d. j. 1973. Lunatadinium dissolution gen. et sp. nov., a dinoflagellate cyst from
Lower Cretaceous rocks, Yukon Territory and northern District of Mackenzie. Bull. Can. Petrol. Geol. 21,
395-402.

bujak, j. p. 1973. Microplankton from the Barton Beds of the Hampshire Basin, England. Ph.D. thesis
(unpubl.), 455 pp., University of Sheffield.

— 1976. An evolutionary series of late Eocene dinoflagellate cysts from southern England. Mar. Micro-
paleontology, 1, 101-117.

Cavelier, c. 1975. Le diacronisme de la zone a Ericsonia subdisticha (Nannoplankton) et la position
de la limite Eocene-Oligocene en Europe et en Amerique du Nord. Bull. B.R.G.M. (2), 4, 3, 201 -
225.

chateauneuf, j-j. 1978. Upper Eocene and Oligocene Dinophyceae of the Paris Basin (France). Proc. 4th
Internat. Palynological Conf., Lucknow (in press).

costa, l. i. and downie, c. 1976. The distribution of the dinoflagellate Wetzeliella in the Palaeogene of north-
western Europe. Palaeontology, 19, 591-614.

CURRY, D., ADAMS, C. G., BOULDER, M. C., DILLEY, F. C., EAMES, F. E., FUNNELL, B. M. and WELLS, M. K.

1978. A correlation of Tertiary rocks in the British Isles. Geol. Soc. Lond. Special Report, No. 12,
72 pp.

daley, B. 1973. The palaeoenvironment of the Bembridge Marls (Oligocene) of the Isle of Wight, Hampshire.
Proc. Geol. Ass. 84, 83-93.

drugg, w. s. 1970. Some new genera, species, and combinations of phytoplankton from the Lower
Tertiary of the Gulf Coast USA. Proc. North Amer. Paleont. Convention, Chicago, 1969, G, 809-
843.

eisenack, a. 1964. Katalog der fossilen Dinoflagellaten, Hystrichospharen und verwandten Mikrofossilien.

Band 1. Dinoflagellaten, E. 888 pp., Schweizerbart’sche Verlagsbuchhandlung, Stuttgart.

EVITT, w. r., LENTIN, j. k., millioud, M. E., stover, L. E. and williams, G. l. 1977. Dinoflagellate cyst
terminology. Geol. Surv. Pap. Can. 76-24, 1-11.

gerlach, E. 1961. Mikrofossilien aus dem Oligozan und Miozan Nordwestdeutschlands, unter besonderer
Beriicksichtigung der Hystrichosphaeren und Dinoflagellaten. Neues Jb. Geol. Pal'aontol. Abh. 5, 112,
143-228.

Harris, w. k. 1973. Tertiary non-marine dinoflagellate cyst assemblages from Australia. Spec. Pubis, geol.
Soc. Aust. 4, 159-166.

haskins, c. w. 1969. Tertiary Ostracoda from the Isle of Wight and Barton, Hampshire, England. Part IV.
Revue Micropaleont. 12, 149-170.

keen, M. c. 1972. The Sannoisian and some other upper Palaeogene Ostracoda from north-west Europe.
Palaeontology, 15, 267-325.

— 1977. Ostracod assemblages and the depositional environments of the Headon, Osborne and Bembridge
Beds (Upper Eocene) of the Hampshire Basin. Ibid. 20, 405-445.
lentin, J. K. and williams, G. L. 1975. A monograph of fossil peridinioid dinoflagellate cysts. Bedford Institute
Oceanography, Report Bl-R-75-16, 1-237.

liengjarern, m. 1973. Dinoflagellate cysts and acritarchs from the Oligocene Beds of the Isle of Wight. Ph.D.
thesis (unpubl.), 220 pp., University of Sheffield.

mcintyre, d. J. 1975. Morphologic changes in Deflandrea from a Campanian section, District of Mackenzie,
N.W.T., Canada. Geosci. Man, 11, 61-76.

morgenroth, p. 1966a. Mikrofossilien und Konkretionen des nordwesteuropaischen Untereozans. Palaeonto-
graphica, Abt. B., 119, 1-53.

— 1966 b. Neue in organischer Substanz erhaetene Mikrofossilien des Oligozans. Neues Jb. Geol. Palaont.
Abh. 127, 1-12.

Murray, j. w. and wright, c. a. 1974. Palaeogene Foraminiferida and palaeoecology, Hampshire and Paris
Basins and the English Channel. Spec. Pap. Palaeontology, 14, 1-171.
stover, l. E. 1977. Oligocene and early Miocene dinoflagellates from Atlantic Corehole 5/5B, Blake Plateau.
Am. Assoc. Stratigr. Palynol., Contrib. Ser. 5A, 66-89.

— and evitt, w. R. 1978. Analyses of Pre-Pleistocene organic walled Dinoflagellates. Stanf. Univ. Pubis,
Geol. Sciences, 15, 1-300.

LIENGJ ARERN ET AL.. EOCENE/OLIGOCENE DINOFLAGELLATES

499

weyns, w. 1970. Dinophycees et acritarches des ‘Sables de Grimmertingen’ dans leur localite-type, et les
problemes stratigraphiques du Tongrien. Bull. Soc. beige Geol. Paleont. Hydrol. 79, 247-268.
williams, G. l., sarjeant, w. a. s. and kidson, e. j. 1973. A glossary of the terminology applied to dino-
flagellate amphiesmae and cysts and acritarchs. Am. Assoc. Stratigr. Palynol., Contrib. Ser. 2, 1-222.

M. LIENGJ ARERN
L. COSTA
C. DOWNIE

Department of Geology

Manuscript received 21 December 1978 University of Sheffield

Revised manuscript received 18 July 1979 Sheffield SI 3JD

DICTYODORA FROM THE SILURIAN OF
PEEBLESSHIRE, SCOTLAND

by m. j. benton and n. h. trewin

Abstract. The meandering trace fossil Dictyodora Weiss, 1 884 occurs in deep water greywacke/shale sequences
in the Gala Group (lower Silurian) of Thornylee and Grieston Quarries, Galashiels. Two species are recognized;
D. scotica (M‘Coy, 1851) and D. tenuis (M‘Coy, 1851); the former is distinguished by a more regular meandering
form. These traces were originally named Crossopodia scotica and Myrianites tenuis. It is suggested that
C. scotica be rejected as the type species of Crossopodia.

Thornylee Quarry (Grid ref. NT 4200 3635) (formerly spelt Thornyly, Thorney Lee,
Thornielee, Thornilee) is situated on the north bank of the River Tweed, 8 km east of Galashiels and
8 km west of Innerleithen. The quarry is located on a steep slope above a layby on the A72
(Peebles-Galashiels) road. Between the quarry and the road is a dismantled railway with cuttings
which provide a 300 m long section through Upper Llandovery greywackes and shales (Gala Group
of Lapworth 1870). The first geological description of Thornylee was given by Nicol (1850) who
noted some graptolites and abundant ‘annelid impressions’.

Grieston Quarry (NT 3130 3618) was also described by Nicol (1850), who noted the abundant
graptolite fauna and the trace fossils. More recently the fauna and sediments of this quarry have been
described by Toghill and Strachan (1970) and Trewin (1979). The thin greywackes and shales of
Grieston also lie within the top of the Gala Group of Lapworth (1870), but are not exactly the same
age as those at Thornylee on the basis of the graptolite fauna.

This study stemmed from work on H. A. Nicholson’s trace fossil collection in Aberdeen (Benton
and Trewin 1978). The following descriptions are based on large collections made at Thornylee and
Grieston in April and June, 1977. Comparisons have been made with the type material of M‘Coy and
Nicholson. Repository abbreviations used are: AUGD, Aberdeen University, Department of
Geology and Mineralogy Palaeontology Collection; BMNH, British Museum (Natural History);
GSM, Geological Survey Museum, I.G.S., London; HM, Hunterian Museum, Glasgow; SM,
Sedgwick Museum, Cambridge.

DEPOSITIONAL ENVIRONMENT AND ASSOCIATED FAUNA

At both localities deep water, interbedded greywacke/shale sequences are exposed in which the
coarser lithologies are of turbidite origin. The trace fossils at Thornylee are more abundant in the
shale-rich parts of the sequence rather than in association with greywacke beds. There seems to be a
greater frequency of meandering traces in the purple rather than the green shales. At Grieston the
greywackes are fine-grained and contain abundant ripple-lamination, possibly the results of
reworking; other beds are characterized by numerous transported graptolites which produced
delicate tool marks on bed bases (Trewin 1979). The greywackes at Thornylee are usually medium
grained, graded, and sometimes show tool marks and load casts on the sharp bed bases. Internally,
Bouma sequences of structures are frequently seen. The general aspect of the lithofacies is of a low-
energy turbidite environment with thin greywacke turbidites and abundant shale.

At both localities graptolites are present but they are much more abundant in the finer grained
rocks of Grieston Quarry, where the majority have been transported and deposited in thin turbidites.

[Palaeontology, Vol. 23, Part 3, 1980, pp. 501-513.1

502

PALAEONTOLOGY, VOLUME 23

Tail spines of Ceratiocaris occur at Grieston, but no other fauna was noted. The ichnofauna
dominated by meandering feeding burrows is typical of deep water muds and belongs to Seilacher’s
Nereites facies.

THE ICHNOFAUNA

Introduction. The ichnofauna is dominated by the meandering burrows of two species of Dictyodora,
which are described below. The small burrow Caridolites Etheridge, Woodward and Jones, 1890 is
common at both localities. Rare examples of Nereites were found at Thornylee and stuffed burrows,
cf. Planolites, are also present. The meandering traces are described below with more emphasis placed
on Dictyodora scotica in view of its taxonomic importance. A redescription is given of Caridolites and
the association with Nereites briefly discussed.

Genus dictyodora Weiss, 1884
Taxonomic discussion of Dictyodora

Geinitz (1867) founded the species Dictyodora liebeanum for a ‘plant’ from the Culm (Lower
Carboniferous) of Gera, East Germany, and Weiss (1884a, b) proposed the genus Dictyodora for this
species. He was unable to decide if it was of plant or animal origin.

Zimmermann (1889, 1891) discussed the taxonomic problems associated with German Carbon-
iferous Dictyodora, noticing that as with the British examples, different horizontal (bedding parallel)
sections had been given distinct names at different times. Zimmermann (1892) gave a detailed account
of the type species D. liebeana, and considered that the vertical wall contained no infill, but noted
longitudinal and oblique streaks. Zimmermann noted that the wall tends to slope inwards towards
the top, giving tighter loops than those of the basal burrow, but was puzzled by walls intersecting
without disturbance. D. liebeana has vertical walls up to 1 80 mm high and a well-defined over-all cone
shape distinguishing it from D. scotica and D. tenuis. Zimmermann (1892) briefly described a species,
D. hercynica, which has a looser structure and walls 1 -3 cm high, found in the Upper Devonian of the
Harz mountains. It has apparently not been figured.

D. simplex Seilacher, 1955 from the Lower Cambrian of the Salt Range of Pakistan is a simple,
loose structure about 6 mm deep. However, this is a structure built from successive sloping layers and
Seilacher proposed that the trace was produced by a worm-like animal travelling through the
sediment in an oblique position. There is no basal burrow in Seilacher’s reconstruction and the
‘vertical wall’ is of equal width from top to bottom. We consider that these differences are sufficient to
exclude D. simplex from the genus Dictyodora. No alternative generic assignment is suggested
without examination of the original material.

Seilacher (1967, p. 77) figured a Dictyodora evolutionary sequence from relatively loosely
structured forms in the Lower Palaeozoic to tightly spiralling patterns in the Carboniferous. In grade
of organization, D. tenuis appears similar to Seilacher’s most primitive type (a) and D. scotica is
slightly more advanced.

Pfeiffer (1959) reviewed previous work on D. liebeana and gave good three-dimensional
reconstructions of Carboniferous examples. Muller (1962) also described the morphology of German
Lower Carboniferous Dictyodora in detail with many figures, and Ruchholz (1967) gave further
examples from the Harz mountains. Pfeiffer (1968) gave a synonomy list for D. liebeana (Geinitz,
1867). Muller (1971) discussed the formation of Dictyodora meanders, emphasizing that the trace was
a feeding structure formed relatively rapidly, since the basal burrow does not change in diameter in
any single specimen and since it maintains a constant depth and does not rise gradually to keep up
with sedimentation.

There is thus an extensive, mainly German, literature on Dictyodora which establishes the
characteristic features of the genus as the meandering basal burrow and the dorsal striated wall. The
species D. scotica, described below, has previously been given the name Myrianites tenuis for sections
for the vertical wall and Crossopodia scotica for the basal burrow.

BENTON AND TREWIN: DICTYODORA

503

The genus Myrianites MacLeay, 1839 was established for a meandering track with small leaf-like
extensions at the sides. The type species, M. macleayii Murchison, 1839 (type specimen: GSM Geol.
Soc. Coll. 6824) appears to be a small Nereites. Species from Spain described by Delgado (1910) as
Myrianites are certainly Dictyodora but are not described or figured well enough to establish
synonomy with the material described here.

M‘Coy (1851a, b ) founded the species M. tenuis based on specimens of small meandering traces
from Grieston Quarry. Nicholson (1978, pp. 42, 43) identified wall sections of D. scotica from
Thornylee as M. tenuis, but the specific name tenuis is retained here for M‘Coy’s original material
redescribed below as D. tenuis.

M‘Coy (1851a, b) also founded the genus Crossopodia for two Silurian trace fossils. C. lata from
Llandeilo, Wales, is a 2 cm wide trail with clear transverse striations and a ‘fringe’ which better
resembles the Crossopodia of modern usage. C. scotica, however, is the form redescribed here as
D. scotica and M‘Coy’s type (SM A45575a-c) clearly shows the diagnostic features (text-fig. 2). The
figure of the type of C. scotica in M‘Coy 18516, pi. ID, fig. 15, appears to be a composite of the three
specimens SM A45575a-c. Fortunately all are of the same species and A45575a is more suitable as the
lectotype showing well all the major features. M‘Coy’s figure has been reversed in the engraving
process. Unfortunately, Hantzschel (1962, p. W189) designated C. scotica as the type species of
Crossopodia and repeated this with a mislabelled figure of ‘C. scotia' (sic) in Hantzschel (1975, fig. 34,
2b). This figure is derived from Schimper and Schenk (1879, p. 52, fig. 40) and is clearly not the
C. scotica of M‘Coy (1851a, b) and Nicholson (1978).

In order to preserve the normally accepted usages of Crossopodia and Dictyodora we propose that
C. scotica be rejected as the type species of Crossopodia. C. lata M‘Coy (1851) (type specimen
SM A37733) would then become the type species of Crossopodia. An application to this effect will be
made to the I.C.Z.N. or other appropriate body, when agreement has been achieved on the rules of
trace fossil nomenclature. Further revision of the genus Crossopodia is required, but is outside the
scope of this paper.

Dictyodora scotica (M‘Coy, 1851)

Text-figs. 1, 2, 3

v* 1851a Crossopodia scotica M‘Coy, p. 395.

v* 18516 Crossopodia scotica M‘Coy; M‘Coy, p. 130, pi. ID, fig. 15.

71855 Crossopodia scotica M‘Coy; Harkness, p. 475.
non 1879 Crossopodia scotica (M‘Coy); Schimper and Schenk, p. 52, fig. 40.
non 1962 Crossopodia scotia (M‘Coy) (sic); Hantzschel, p. W189, fig. 118, 2.
non 1975 Crossopodia scotia (M‘Coy) (sic); Hantzschel, p. W54, fig. 34, 2b.
vl978 Crossopodia scotica M‘Coy; Nicholson, p. 36, pi. 3, fig. 1, pi. 6.

vl978 Myrianites tenuis M‘Coy; Nicholson, p. 42, text-fig. 7, non pi. 4, fig. 1. [The same specimen as in
Benton and Trewin 1978, pi. 2, fig. 2.]
vl978 Crossopodia scotica M‘Coy; Benton and Trewin, p. 8, pi. 2, fig. 1.

Lectotype. Here designated, SM A45575a, the original of M‘Coy (18516, pi. ID, fig. 15). Gala Group, Upper
Llandovery, lower Silurian, Thornylee Quarry, nr. Innerleithen, Peeblesshire, Scotland. Refigured here,
text-fig. 2.

Other material. More than two hundred examples from the type locality, a representative selection of which are
catalogued as AUGD 10693 to 10710. Also: AUGD 8819, 8820, 10606, 10723, Mus. Coll. 956, 957; BMNH
39451, 58169 (1, 2); GSM 104247, 104249, 104250, RU 2970; HM X871/1-2, X1003/1-7.

Description. The burrow system illustrated in text-fig. 1 consists of a basal burrow, generally preserved with a
lenticular cross section, and having a vertical or inclined longitudinal wall arising from the dorsal mid-line of the
basal burrow. The basal burrow varies from 1-5-6 mm wide and up to 3 mm high in slate lithologies, but when
developed in fine sand may have a nearly circular cross section due to the small degree of compaction. The wall is
up to 13 mm high and tapers upwards from a width of 1 -2 mm at the base. The taper is most rapid in small
examples. The typical burrow system (text-fig. 3c, d, e) consists of 5-10 parallel meanders each 10-80 mm long

504

PALAEONTOLOGY, VOLUME 23

text-fig. 1. Scale bars 10 mm at front faces of figures. Arrows indicate direction of travel of Dictyodora
organism, a, general morphology of Dictyodora meanders showing basal burrow and wall; wall curves inwards
at meander bends. B, section of burrow to show features of burrow and wall fill, horizontal striations and curved
vertical/oblique striations of wall surface, c, block diagram illustrating different preservational aspects of the
burrow in plan and section; a, narrow sections at top of wall; b, wider sections near base of wall; c, convex top of
basal burrow with base of wall fill preserved on top; d, concave impression of underside of burrow with fill
removed, a weak median ridge may be present; e, smaller example showing effect of sectioning the inclined wall
at meander turn; /, juvenile burrow in section. The style of ripples and fine parallel lamination present is also
illustrated on the front face of c.

BENTON AND TREWIN: DICTYODORA

505

(usually 30-50 mm) and internally measured at basal burrow level as 0-20 mm apart (usually 5-15 mm). Where
successive meanders touch, a tight turning circle is present at the meander turn. The meanders may also be
irregular and broad as in text-fig. 3a, b. The relevant features of the type specimen are illustrated in text-fig. 2.

The burrow shows various preservational aspects (text-fig. lc) dependent on the level at which it is sectioned.
Sections of the wall appear as meandering lines up to 2 mm wide, occasional sharp turns are seen in sections close
to the top of the wall (text-fig. 3e) but nearer the basal burrow the wall displays smooth curves. The wall has a
finite thickness and the burrow may break either side of the wall as shown in text-fig. 3b. Sections at the top of the
basal burrow show the entire infill with a median ridge marking the base of the wall (text-fig. lc). Specimens
showing the lower surface of the basal burrow display a smooth groove which is sometimes double, with a weak
median ridge (text-fig. lc). The burrow may also split within the burrow fill giving very little relief to the
preserved trace. Internally, a distinct pattern is frequently seen in polished or etched cross-sections of the burrow
fill resulting from reorientation of platey minerals (text-fig. 1b).

text-fig. 2. Sketch of lectotype of Dictyodora scotica, SM A45575a showing the lower surface of the specimen.
Trace A shows the typical meander pattern. Most of the specimen displays the lower surface of the burrow but at
a the burrow fill is broken out to show a mould of the upper surface of the basal burrow. The wall of A is 5 mm
high and is not seen on the top of the slab. Trace B is larger than A and later since it clearly crosses A. At b the
transition from basal burrow to wall can be seen. The wall passes through the full 8 mm thickness of the slab and
is seen on the top of the specimen (not illustrated).

The burrows are indistinct in places due to the presence of several crossing burrows, and fracture irregularities
on the surface of the slab which have been omitted for clarity.

The wall is normally vertical above straight stretches of burrow, but curves inwards at meander bends (text-
figs. 1a, c, 3b, e). Fine bedding parallel striations are present on the surface of the wall closely spaced at 4 per mm.
A similar bedding parallel banding due to platey mineral orientation occurs within the wall fill, and is not related
to sedimentary laminae. Curved vertical/oblique striations are also present on the wall surface normally spaced
at 3-5 per mm. Internally the wall may show fine curved structures marked by reoriented platey minerals and
resembling backfill within the wall (text-figs. 1 b, 3b). Detailed observation of features is difficult in the wall fill but
it is likely that the possible backfill structures seen normal to bedding occur between the bedding parallel bands.

The smallest forms recognized have a basal burrow 1.5 mm wide and a wall only 1 mm high, and a full
gradation exists up to the larger forms with a progressive increase in wall height relative to burrow width (text-
fig. 4). Detailed measurement of the morphology and meander patterns of over 170 specimens using principal
components analyses failed to differentiate any groups with significantly different characters, and we consider
that all the meandering burrows of this type are growth stages of a single species.

text-fig. 3. Dictyodora scoticcr, examples of burrow morphology, a, irregular meanders (section of burrow wall)
with example of avoidance of previously formed burrow at a, AUGD 10693. b, irregular burrow which crosses
previously formed burrow; plan view shows wall above basal burrow to be partly broken away, and inward slope
of wall at meander curves; thickness of slab 10 mm; AUGD 10697. c, D, typical regular meander forms, hooked
ends to meanders seen in c; both on AUGD 10694. e, plan view of basal burrow (stipple) and position of top of
wall (solid line); sharp bends present at top of wall become smooth curves at lower levels close to the basal
burrow; AUGD 10698. All examples from Thornylee Quarry.

BENTON AND TREWIN: DICTYODORA

507

Occurrence. Dictyodora scotica is common at Thornylee Quarry and scarce at Grieston Quarry. It is probably
common in the Llandovery strata of the Southern Uplands since Peach and Horne ( 1 899) mention ‘ Crossopodia ’
and ‘ Myrianites ’ from at least twenty localities in the Galashiels-Hawick region. It also occurs in the Llandovery
of Penwhapple Glen, Girvan (Nicholson and Etheridge 1880, pp. 304-318). P. Doughty (pers. comm.) also
records Dictyodora from the Silurian of Co. Down, Northern Ireland.

H 1 1 1 1 r

1 2 3 4 5 6 W mm

text-fig. 4. Dictyodora scotica. Relationship of width of basal burrow W with burrow
height H to show range of variation and the relative increase in wall height in the larger
examples.

Dictyodora tenuis (M‘Coy, 1851)
Text-fig. 5

v* 1851a Myrianites tenuis M‘Coy, p. 394.
v*18516 Myrianites tenuis M‘Coy; M‘Coy, p. 130, pi. ID, fig. 13.
vl978 Myrianites tenuis M‘Coy; Nicholson, pi. 4, fig. 1, non text-fig. 7.
vl978 Myrianites murchisoni Emmons; Nicholson, p. 43, pi. 5, fig. 1 .

Lectotype. Here designated, SM A45579a, the original of M‘Coy (18516, pi. ID, fig. 13). Gala Group, Upper
Llandovery, Lower Silurian, Grieston Quarry, nr. Innerleithen, Peeblesshire, Scotland (text-fig. 5a).

Other material. AUGD 9224, 10329, 10607, 10612, and 10711 to 10720 from Grieston Quarry and AUGD 10710
from Thornylee Quarry.

508

PALAEONTOLOGY, VOLUME 23

Description. Dictyodora with broad irregular meanders, as in text-fig. 5, which frequently have a secondary
sinuosity with a wavelength of 3-15 mm which may develop into meanders with length roughly equal to breadth
in larger examples. The basal burrow is from T5 to 3 mm wide and the wall has not been observed to exceed
10 mm in height. The wall is 0-2-0-7 mm wide and striated in the same manner as in D. scotica. Traces range from
tiny ‘scribbles’ (text-fig. 5e) up to large examples as in text-fig. 5b, d.

Trace endings are observed as in text-fig. 5b where lengths of trace as short as 10 mm occur between inclined
circular burrows 3 mm in diameter; other traces can be followed for over 200 mm without interruption.

Discussion. The distinction of D. tenuis from D. scotica can be made on maximum size and on the
meandering pattern, which is more regular and smooth in D. scotica compared with the irregular
meanders with secondary sinuosity displayed by D. tenuis.

In the past specimens displaying sections of the wall have been identified as Myrianites and
specimens showing the basal burrow as Crossopodia or Nemertites. The specimens from Grieston
called M. murchisoniby Nicholson (1978, p. 43, pi. 15, fig. 1) are not synonymous with the American
form described by Emmons (1844) and are ascribed here to D. tenuis.

Occurrence. Common in the Upper Llandovery ( griestonensis Zone) of Grieston Quarry, nr. Innerleithen,
Peeblesshire, and also present in association with much commoner D. scotica at Thornylee Quarry. The form
illustrated by Raup and Seilacher (1969, fig. la) from the Ordovician of Barrancos, Portugal, appears to be
D. tenuis.

THE DICTYODORA ANIMAL AND ITS BEHAVIOUR

The meandering burrow of Dictyodora resembles the meandering burrows and trails produced by
worms and molluscs efficiently utilizing an area as a food source. The tightly packed meanders of
Dictyodora were probably formed during feeding, and the looser irregular meanders may have been
the result of searching for areas rich in food. We assume that the body of the animal occupied the
basal burrow, and probably progressed by peristaltic movement. Since individual burrows cannot be
traced from small to large size, and considering that the burrows are sometimes seen to end by rising
through the sediment it is likely that the animal moved from place to place on or above the sediment
surface. Thus the burrows are considered to be produced by short periods of food search and
utilization at a constant level within the sediment.

The animal appears to have maintained contact with the surface by means of an organ which was
responsible for the production of the striated wall on the dorsal burrow surface; this we term the wall-
organ to avoid assumptions implicit in the use of known zoological terms such as ‘siphon’. The
curved vertical striations on the wall and the fill of the wall indicate that the wall-organ moved
regularly through the sediments, maintaining a constant convex-forward edge and followed the
movement of the animal in the burrow; thus wall-organ traces occasionally touch or cross each other
while the corresponding burrows do not.

The behaviour of animals that form meandering traces has been discussed by several authors.
Seilacher (1967) suggested that the Dictyodora animal measured its meander length by the length of
its body. It maintained contact with a previously formed burrow (thigmotaxis) until its body was
straight and then the animal was ‘programmed’ to make a sharp U-turn (homostrophy) as its tail
straightened, and to follow beside the last-formed portion of the burrow. However, this explanation
does not satisfactorily explain individual burrows where meander length varies, or the Carboniferous
Dictyodora where the meanders spiral out from a central point, each meander being longer than its
predecessor.

Seilacher based his interpretation on the classic work of Richter (1924, 1928), who studied the
Cretaceous/Tertiary Helminthoida labyrinthica Heer, 1865 which forms similar meandering feeding
traces. Richter’s interpretation differs from Seilacher’s in one important way: he defined the
homostrophic turning stimulus as caused by loss of contact with a former trace and not by tail
straightening. The animal followed a former trace and could at times curve in front of previous
meander ends (e.g. text-fig. 3c) before turning back when it lost contact with disturbed mud. In text-
fig. 3 meander length varies from 30 to 80 mm and was clearly not measured by the body length of the

BENTON AND TREWIN: DICTYODORA

509

text-fig. 5. Dictyodora tenuis. Examples of burrow morphology shown by sections of the wall of the burrow.

a, small meandering trace with irregular meanders showing secondary sinuosity; part of lectotype SM 45579a.

b, parts of typical irregular meanders, together with short lengths of burrow terminated by inclined sections of
basal burrow; AUGD 10719. c, d, e, irregular meanders of various sizes to show variation in meander

morphology; c, E AUGD 10716; d AUGD 10718. All from Grieston Quarry.

510

PALAEONTOLOGY, VOLUME 23

animal. The reactions of the animal while feeding in meanders as listed by Seilacher (1967) and Raup
and Seilacher (1969) may be modified to:

(1) Move horizontally keeping within a single stratum of sediment (? controlled by wall-organ length);

(2) Always keep in touch with previously formed burrow while feeding (thigmotaxis);

(3) Never come closer to a previously formed burrow than a particular distance ‘d’ (phobotaxis);

(4) If contact is lost with a former burrow, make a 180° turn (homostrophy/strophotaxis).

These ‘rules’ appear to apply reasonably well, and obvious cases of burrow avoidance can be found
(text-fig. 3a). Traces made by individuals at different levels in the sediment frequently cross each
other, but the basal burrows in such cases are normally at different levels. In the Thornylee examples
population density was probably low and thus there was no need for attempting to utilize an area
more than once.

If the meandering burrows are formed during feeding then the question arises of how feeding was
accomplished. The wall-organ could have been a food collector at the surface, with the animal
protected in its burrow, or the animal could have fed by sediment ingestion at burrow level leaving the
wall-organ to perform a respiratory function. The second of these suggestions seems most favourable
since the basal burrow has a definite burrow fill which corresponds to the sediment type at basal
burrow rather than surface level. The apparently passive motion of the wall-organ does not accord
with a function as a feeding organ, and it is more likely to have had a respiratory function and to have
controlled burrow depth.

In laminated sediment the fill of the wall roughly matches the characteristics of the immediately
adjacent sediment, with only slight downward movement of sediment during filling occasionally seen
in thin section. Thus the wall-organ does not seem to have had a significant sediment transport
function. No annulation of the burrow fill is seen and the constant fine spacing of the striations
formed by the wall-organ would seem to indicate a slow regular movement through the sediment. The
wall-organ may have been ciliated to facilitate its progress through the sediment. The striations and
structured fill of the wall indicate that the organ was not merely dragged through the sediment but
that the thin wall of sediment was packed in both horizontal and vertical increments by the
wall-organ.

The Dictyodora animal was probably a worm or shell-less mollusc which fed by sediment ingestion
and maintained contact with the over-lying water by means of the wall-organ which controlled
burrow depth and possibly aided respiration.

OTHER TRACES PRESENT
Caridolites wilsoni Etheridge, Woodward and Jones, 1890
Text-figs. 6, 7

The name Caridolites wilsoni was first mentioned in Nicholson (1873) and a brief description
appeared in Etheridge, Woodward and Jones (1890), which must rank as the type description.
Nicholson’s original (1872) manuscript with a description and figure of C. vw/som'has been published
recently together with a discussion (Benton and Trewin 1978, p. 10, pi. 3) in which Nicholson’s
interpretation that the trace was made by the tail spines of shoals of swimming Ceratiocaris is
rejected.

The traces are generally about 1 mm wide and may consist of a slight central ridge bounded by
hollows or a single ridge, or the counterpart of either. The traces are generally nearly straight for from
10-50 mm before disappearing or turning fairly sharply on a new course. Typical examples are shown
in text-fig. 6 a-j and typical profiles in text-fig. 61. In cross section the traces are seen to be burrows
with a vertical depth of up to 5 mm and consist of a basal tunnel with a narrower vertical extension
(text-fig. 6k). These traces thus resemble minute Dictyodora without the meanders. Caridolites
frequently covers bedding surfaces with a confusion of burrows as in text-fig. 7.

BENTON AND TREWIN: DICTYODORA

511

text-fig. 6. Caridolites wilsoni. a-j, typical burrow traces as
seen on bedding surfaces; a-c, AUGD 10675; d-f, AUGD
10748; i, j, AUGD 7055, Grieston Quarry; g, h, AUGD
10723, Thornylee Quarry, k, typical vertical cross sections of
burrows. /, profiles of surface expressions of the burrows.

text-fig. 7. Caridolites wilsoni. Bedding
surface covered with typical examples, x 1 ,
AUGD 10674, Grieston Quarry.

Caridolites is abundant at both Grieston and Thornylee and is frequently associated with both
D. scotica and D. tenuis. It seems possible that Caridolites represents the activities of juvenile
Dictyodora animals which had not developed sufficiently to meander. Certainly the observed size
ranges of the traces fit this possibility.

Genus nereites MacLeay, 1839

Nereites is rare in the Thornylee-Grieston assemblage, with only two clear examples of this surface
trace seen. Sediment surface texture was probably not suited to preservation of surface trails and
most were probably removed by turbidity currents. The slaty muds and silts generally do not split at
the top surfaces of beds. The common association of Nereites surface traces in sequences with
Dictyodora burrows of similar width raises the speculation that Nereites could be a surface trace of
the Dictyodora animal moving from one feeding spot to another.

512

PALAEONTOLOGY, VOLUME 23

CONCLUSIONS

The deep water ichnofauna of the greywacke/shale turbidite facies of the Llandovery in southern
Scotland is dominated by two species of Dictyodora. The small burrow Caridolites is probably the
juvenile burrow of the ‘ Dictyodora ’ animal. Nereites is also present but rare, probably owing to
original sediment texture and preservation.

Crossopodia scotica is shown to be a Dictyodora, and it is suggested that it should be rejected as the
type species of Crossopodia, being replaced by C. lata.

Acknowledgements. We thank the following for the loan of specimens and study facilities: Dr. R. B. Rickards,
Sedgwick Museum, Cambridge; Dr. A. W. A. Rushton, Geological Survey Museum, I.G.S., London; Dr.
W. D. I. Rolfe, Hunterian Museum, Glasgow; Dr. R. Wilson and Mr. P. J. Brand, I.G.S., Edinburgh; and Mr.
D. N. Lewis, British Museum (Natural History).

REFERENCES

benton, m. j. and trewin, n. h. 1978. Discussion and comments on Nicholson’s 1872 manuscript ‘Contributions
to the study of errant annelides of the older Palaeozoic rocks’. Pubis. Dep. Geol. Miner. Univ. Aberdeen, 1,
1-16.

delgado, j. f. n. 1910. Terrains paleozoiques du Portugal. Etude sur les fossiles de schistes a nereites de San
Domingos et des schistes a nereite et a graptolites de Barrancos. Commis. Serv. geol. Portugal, 56, 1 68,
pis. 1-51.

emmons, E. 1844. The laconic system. 68 pp., 6 pis., Albany, New York.

Etheridge, R., woodward, H. and jones, T. R. 1890. Seventh report of the committee on the fossil Phyllopoda of
the Palaeozoic rocks. Rep. Br. Ass. Advmt. Sci. for 1889, Trans. Sections, 1890, 63-68.
geinitz, h. b. 1867. Uber Dictyophytonl Liebeanum Gein. aus dem Culmschiefer vom Heersberge zwischen Gera
und Weyda. Neues Jb. Miner. Geol. Palaont. 1867, 286-288.
hantzschel, w. 1962. Trace fossils and problematica. Pp. W177-W245. In moore, r. g. (ed.). Treatise on
Invertebrate Paleontology, Part W, Miscellanea. Geol. Soc. Am. and Univ. Kansas Press.

— 1975. Trace fossils and problematica. In teichert, c. (ed.). Treatise on Invertebrate Paleontology, Part W,
Miscellanea, Supplement 1, W1-W269, Geol. Soc. Am. and Univ. Kansas Press.

harkness, R. 1855. On the anthracitic shales and the fucoidal remains occurring in the Lower Silurian rocks of
the south of Scotland. Q. Jl Geol. Soc. Lond. 11, 468-476.
lapworth, c. 1870. On the Lower Silurian rocks of Galashiels. Geol. Mag. 7, 204-209, 279-284.
m‘coy, f. 1851a. On some new Protozoic Annulata. Ann. Mag. nat. Hist., ser. 2, 7, 394-396.

— 18516. A systematic description of the British Palaeozoic fossils in the Geological Museum of the
University of Cambridge. In Sedgwick, a. A synopsis of the classification of the British Palaeozoic rocks.
Parker, London [publ. 1851-1855, pp. 1-184 in 1851],

macleay, w. s. 1839. Note on the Annelida. Pp. 699-701. In Murchison, r. i.. The Silurian System, pt. II,
Murray, London.

muller, A. H. 1962. Zur Ichnologie, Taxiologie und Okologie fossiler Tiere. Freiberger ForschHft. C. 151, 5-49.

— 1971. fiber Dictyodora liebeana (Ichnia invertebratorum), ein Beitrag zur Taxiologie und Okologie
sedimentfressender Endobionten. Deutsch. Akad. Wiss. Berlin, Monatsber. 13, 136-151.

nicholson, H. A. 1873. Contributions to the study of the errant annelides of the older Palaeozoic rocks. Proc. R.
Soc. 21, 288-290.

— 1978. Contributions to the study of the errant annelides of the older Palaeozoic rocks. Pubis. Dep. Geol.
Miner. Univ. Aberdeen, 1, 17-47.

— and etheridge, r., jun. 1878-1881. A monograph of the Silurian fossils of the Girvan district in Ayrshire.
Vol. 1, Blackwood, Edinburgh and London.

nicol, J. 1850. Observations on the Silurian strata of the southeast of Scotland. Q. Jl geol. Soc. Lond. 6, 53-65.
peach, B. N. and horne, J. 1899. The Silurian rocks of Britain. 1, Scotland, xviii + 749 pp. Mem. geol. Surv. U.K.
Glasgow.

pfeiffer, H. 1959. fiber Dictyodora liebeana (Weiss). Geologie, 8, 425-439.

— 1968. Die Spurenfossilien de Kulms (Dinants) und Devons der Frankenwalder Querzone (Thuringen).
Jb. Geol. 2, 651-717.

raup, d. m. and seilacher, A. 1969. Fossil foraging behaviour: computer simulation. Science, N.Y. 166,
994-995.

BENTON AND TREWIN: DICTYODORA

513

richter, r. 1924. Flachseebeobachtungen zur Palaontologie und Geologie. IX. Zur Deutung rezenter und
fossiler Maander-Figuren. Senckenbergiana, 6, 141-157.

1928. Psychische Reaktionen fossiler Tiere. Palaeobiologica, 1, 225-244.

ruchholz, k. 1967. Zur Ichnologie und Fazies des Devons und Unterkarbons im Harz. Geologie, 16, 503-527.
schimper, w. p. and schenk, a. 1879-90. Palaeophytologie. In zittel, k. a. von (ed.). Handbuch der
Palaontologie, II, 1-152 (1879).

seilacher, a. 1955. Spuren und Fazies im Unterkambrium. Pp. 86-143. In schindewolf, o. h. and seilacher, a.
Beitrage zur Kenntnis des Kambriums in der Salt Range (Pakistan). Akad. fViss. Lit. Mainz, math.-nat. Kl.,
Abh. 10.

1967. Fossil behaviour. Scien. Am. 217 (2), 72-80.

toghill, p. and strachan, i. 1970. The graptolite fauna of Grieston Quarry, near Innerleithen, Peeblesshire.
Palaeontology, 13, 511-521.

trewin, N. H. 1979. Transported graptolites and associated tool marks from Grieston Quarry, Innerleithen,
Peeblesshire. Scott. J. Geol. 15, 287-292.

weiss, E. 1884a. Vorlegung des Dictyophytum Liebeanum Gein. aus der Gegend Von Gera. Sitz.-Ber. Gen. naturf.
Freunde, Berlin, 1884, 17.

— 18846. Beitrag zur Culm-Flora von Thiiringen. Jb. Preuss. Geol. Landesanst. 1883, 81-100.
zimmermann, E. 1889. Uber die Gattung Dictyodora. Z. dt. geol. Ges. 41, 165-167.

1891. Neue Beobachtungen an Dictyodora. Ibid. 43, 551-555.

1892. Dictyodora liebeana (Weiss) und ihre Beziehungen zu Vexillum (Rouault), Palaeochorda marina

(Gein.) und Crossopodia henrici (Gein.), Jb. Ges. Freunde Naturwiss. Gera, 32-35, 28-63.

M. J. BENTON

Department of Geology
University of Newcastle
Newcastle-upon-Tyne, NE1 7RU

N. H. TREWIN

Department of Geology and Mineralogy
Marischal College
University of Aberdeen
Aberdeen, AB9 IAS

Manuscript received 15 June 1979

Revised manuscript received 3 September 1979

LOWER CRETACEOUS TEREB R ATULI DAE
FROM SOUTH-WESTERN MOROCCO
AND THEIR BIOGEOGRAPHY

by FRANK A. MIDDLEMISS

Abstract. The terebratulid brachiopods contained in the Gentil and Whitaker Collections from the Lower
Cretaceous of south-west Morocco have been revised. Although the majority of the species are confined to
south-west Morocco, the affinities of the fauna are with the faunas of the shallow marine regions bordering
Tethys, such as the Jura region, eastern Spain, the Crimea, and the northern Caucasus; the Tethyan pygopid
brachiopods characteristic of the Rif in northern Morocco are almost absent. The fauna thus constitutes a Jura-
type assemblage situated on the southern side of Tethys. In the systematic section a new genus Paraboubeithyris
is erected; also seven new species: Loriolithyris melaitensis, L. marocensis, Boubeithyris tibourrensis , B. pleta,
Paraboubeithyris plicae, Kutchithyris kennedyi, and Juralina ecruensis. The genera Kutchithyris and Juralina,
previously described from the Jurassic, are shown to have survived into the Lower Cretaceous. Terebratula
subsella Leymerie is referred to Kutchithyris.

This paper consists mainly of a revision of the terebratulids contained in two important collections,
the Gentil Collection in the Collection de Paleontologie of the Universite Pierre et Marie Curie, Paris,
and the Whitaker Collection in the British Museum (Natural History), London. All the specimens
came from the Lower Cretaceous (Berriasian to Aptian inclusive) of an area at the seaward end of the
High Atlas in south-western Morocco, extending some 40 kilometres inland between Agadir in the
south, Essaouira (Mogador) in the centre, and Safi in the north.

Louis Gentil, who was born at Algiers in 1868 and died in Paris in 1925, was a pioneer in the study
of the geology of Morocco. His first major contribution was the exploration of the Tafna basin. Later
he became a member of the Segonzac exploratory mission to the Atlas Mountains and eventually
head of the mission. He was the author of numerous publications, particularly on the geology of the
Atlas, almost up to the time of his death including, most notably, the first geological map of Morocco,
which appeared in 1923. J. J. S. Whitaker was not a geologist but a Christian missionary who worked
in Morocco during the early years of this century. His collection was made at one locality only (see
p. 519 below) and very probably on one occasion. Figured specimens are in the British Museum
(Natural History) (BM) or the Collection de Paleontologie, Universite Pierre et Marie Curie, Paris
(Gentil Coll.).

THE LOWER CRETACEOUS OF SOUTH-WESTERN MOROCCO

The Lower Cretaceous geology of the area was described by Roch (1930) and that of the southern
part by Ambroggi (1963); Gigout (1951) included the extreme northern part, around Safi, in his
survey; Ager (1974) gave a brief summary in English. All agree that south-western Morocco was, in
Lower Cretaceous times, a marine depositional basin opening westwards towards the ocean, cut off
from the marine deposits of the same age, but quite different lithofacies and fauna, in the Rif arc to
the north by the interposition of the positive block of the Moroccan Meseta and from the marine area
of the Algerian high plateaux by the emergent central massif of the High Atlas. At each stage of the
Lower Cretaceous the most fully marine conditions, presumably indicating the deepest water, are
found in the extreme west, around Cap Ghir and northwards to the neighbourhood of Cap Tafelney.
Passing north-eastwards, eastwards, and south-eastwards from this region one finds increasingly

(Palaeontology, Vol. 23, Part 3, 1980, pp. 515-556, pis. 55-61.]

516

PALAEONTOLOGY, VOLUME 23

shallow-water lithofacies and biofacies and, usually within 40 or 50 kilometres, non-marine
deposits.

The deep-water facies around Cap Ghir consists of green marls and marly limestones with
ammonites. These pass eastwards into the more sandy and calcareous beds, with brachiopod and
mollusc faunas, of what Roch significantly calls a ‘jurassian facies’. These pass eventually into sub-
continental red beds. The lithological succession differs markedly from the monotonous lithofacies
of the ‘bathyal’ Lower Cretaceous, seen in the Rif and the Betic region, and has a general resemblance
to the successions seen in the Pre-Betic zone of Spain, north-east Spain (Sitges), east-central Sardinia,
Provence, and Portugal. It exhibits a very striking difference from these, however, in the absence of
the massive urgonian limestones, which are characteristically developed in the Barremian and Aptian
of those regions, and of the rudists. In these respects, the south-west Moroccan succession is most
comparable to the Lower Cretaceous of central Texas and parts of Coahuila (Mexico). The Aptian,

MIDDLEMISS: CRETACEOUS TEREBRATULIDAE

517

as in northern Spain and England, is transgressive, the Gargasian overlapping the earlier divisions on
to the flanks of the High Atlas. To the south lies the coastal Cretaceous basin of Tarfaya, at first sight
similar in situation to the Agadir-Essaouira basin, but here the earlier part of the Cretaceous is non-
marine, marine sedimentation starting only with the Apto-Albian (Choubert et al. 1967).

PALAEOBIOGEOGRAPHICAL RELATIONSHIPS OF THE
TEREBRATULID FAUNA

Endemicity. The fauna contains a high proportion of endemic species: of the eleven species described
eight are new and seven of these are so far known only from south-west Morocco. This is not unusual.
The terebratulids tend to produce local, allopatric species. For example, of the sixteen terebratulid
species in the English Aptian thirteen are known only in south and south-central England, of which
three occur at one locality only (Middlemiss 1959). I have recently (Middlemiss 1979) pointed to the
contrast between such local species and widespread species such as (in the Moroccan fauna)
Loriolithyris valdensis and suggested that these differences were probably due to differing lengths of
the free-swimming larval stage. Evidence for the palaeobiogeographical relationships of the fauna
comes mainly from the occurrence elsewhere of the widespread species but also from the taxonomic
relationships of the local species.

Loriolithyris. L. valdensis is the most widespread species of this genus, occurring in the Lower Cretaceous of
eastern Spain (and the Balearic Islands), Sardinia, southern France, the Jura, south-east Paris Basin, north-east
Bulgaria, the Crimea, northern Caucasus, Kopet Daga, and perhaps Algeria. L. russillensis shares the western
part of this distribution— eastern Spain, the Balearic Islands, southern France, the Jura, and south-east Paris
Basin. L. melaitensis and L. marocensis are local offshoots from the stock, not at present known outside the
south-west Moroccan basin.

text-fig. 2. Palaeobiogeographical relationships of the Lower Cretaceous terebratulids of south-west Morocco.
Distribution of south-west Moroccan Lower Cretaceous species which occur elsewhere: ■ Loriolithyris
russillensis, □ Loriolithyris valdensis, ♦ Cyrtothyris middlemissi, O Kutchithyris kennedyi. Distribution of other
Lower Cretaceous species of Cyrtothyris : + . Distribution of Aptian-Cenomanian species of Boubeithyris: ☆ .
Generalized occurrence of Kutchithyris subsella in the Upper Jurassic and Lower Cretaceous: * . Generalized
occurrence of Jurassic species of Juralina: + . Generalized boundary of the Tethyan pygopid fauna shown by

diagonal shading.

518

PALAEONTOLOGY, VOLUME 23

Boubeithyris and Paraboubeithyris. The three species here ascribed to these genera are all local to south-west
Morocco but the genus Boubeithyris, of which Paraboubeithyris is perhaps a specialized development, is
represented by a species in the Aptian of the Jura, by two species in the Albian of England and by one species in
the Cenomanian of Belgium and western France.

Cyrtothyris. C. middlemissi, the south-west Moroccan species, is known also in eastern Spain and southern
France. The genus is more widespread, being represented by species in the early Cretaceous of north Germany,
north-east England, and east Greenland and the Aptian of the Jura and southern France. Imlay’s species
Terebratula sillimani and T. tamaulipana (Imlay 1937), from the Valanginian-Hauterivian of northern Mexico,
probably belong to this genus.

Kutchithyris. K. brivesi is a highly distinctive form confined to south-west Morocco but K. kennedyi is known also
in the Lower Cretaceous of eastern Spain, the Balearic Islands, and southern France, the southern part of the
same distribution area as L. russillensis. Other species of the genus are found in the Middle Jurassic of India and,
according to Buckman (1918), Europe. I here refer Terebratula subsella Leymerie to this genus. This species has a
widespread occurrence in the Upper Jurassic of Europe and is known (but undescribed) in the Lower Cretaceous
of eastern Spain.

Juralina. This genus, as interpreted by recent authors (especially Boullier 1976), occurs in the Upper Jurassic of a
wide area of Europe north of the Alps from England to Russia and also of Crete ( J . immanis—see Bonneau,
Beauvais, and Middlemiss 1975) and Sicily (Boullier 1976). J. ecruensis is the first species of the genus to be
described from the Cretaceous.

Discussion. The Lower Cretaceous terebratulids of Europe can be divided into three geographical
faunas: the boreal fauna in the north, the Tethyan fauna with its distinctive Pygopinae, and between
them the Jura fauna. The last is so named after the area in which the fauna is richest and best known,
but the character of the Jura fauna is essentially that of a neritic assemblage occupying an optimum
situation on the border of the deeper-water Tethyan region and extending approximately parallel to
the border of Tethys from the Iberian Peninsula eastwards to Turkmenistan. In this sense, the Lower
Cretaceous fauna of south-west Morocco falls into place as an extension of the Jura fauna to the
south of the Tethyan fauna which is so strongly developed in the Rif.

The affinities of our terebratulids are essentially with the Jura brachiopod fauna. This is generally
true of the cephalopods listed and figured by Roch, Ambroggi, and Gigout. Characteristic Tethyan
genera such as Lytoceras (Valanginian-Hauterivian), Phylloceras (Hauterivian), Desmoceras
(Barremian), Pulchellia (Barremian), Duvalia (Valanginian), Hibolites (Valanginian) occur but are
almost confined to the deep-water region of the extreme west. Further east the cephalopods are noted
by Roch as being of ‘Jura type’ and include such genera as Acanthodiscus and Leopoldia. There is
scarcely a trace in the pre-Aptian Cretaceous of the Tethyan pygopines which characterize the Rif
and the Betic region (Geyssant 1966). The ‘jurassian’ affinities of the faunal facies were clearly
recognized by Roch and Gignoux (1955). Ager (1974) has recorded the discovery of Nucleata cf.
jacobi in the Aptian or Albian near Tamzargout. This seems to be the only recorded occurrence of
pygopine brachiopods in the Lower Cretaceous of south-west Morocco— a feeble sign of southward
Tethyan spread’ simultaneous with those transgressions which were causing northward movement
of southern species into north Spain, England, and north Germany (Middlemiss 1979). The specimen
from Safi figured by Gigout (1951, pi. 9, figs. 35-38) as T. euthymi is a terebratellidine related to
‘ Terebratula ’ moreana d’Orbigny.

Kutchithyris, in the Lower Cretaceous, does not occur north of southernmost France and is one of
those sub-Tethyan forms (Middlemiss 1979) which are sensitive indicators of the advance and retreat
of the Tethyan fauna. K. subsella shows this well. In the Oxfordian, a period of major expansion of
the Tethyan fauna (Arkell 1956), it is found throughout a large part of central Europe— England,
northern France, northern and south-western Germany, southern Poland, the Russian Platform. By
Kimmeridgian times it extended no further north than the Boulonnais. The Volgian saw a further
southward retreat to the Pays de Bray, its place in England and the Boulonnais being taken by boreal
forms. In the Lower Cretaceous it has so far been found only in the Pre-Betic region of Spain, on the

MIDDLEMISS: CRETACEOUS TEREBRATULIDAE

519

margin of Tethys. Juralina may also be a sub-Tethyan genus whose history is possibly similar to that
of K. subsella.

Reconstruction of plate positions as they were in Lower Cretaceous times shows the area of the
Jura faunas as much more linear than it is now. Provence, eastern Spain, Sardinia, the Balearic
Islands, and south-west Morocco form a linear belt which, extended westwards, would include the
western Gulf region of the U.S.A. and the northern parts of Mexico. The neritic Lower Cretaceous of
these latter regions is in this sense an extension of the area of the Jura fauna. Unfortunately
brachiopods are rare but Imlay (1940) remarked of the Neocomian faunal assemblage of northern
Mexico that it was remarkably similar to that of France, England, and Switzerland and belonged
decidedly to the ‘Mediterranean’ province. His species T. coahuilensis is certainly close to and
probably synonymous with Sellithyris carteroniana d’Orbigny, one of the most characteristic Jura
species. It seems a reasonable forecast that neritic Lower Cretaceous brachiopod assemblages of
‘Jura fauna’ affinities will some day be found in the south-eastern or Gulf continental shelf deposits of
the U.S.A. or the north-western continental shelf deposits of Africa. Unfortunately those of the
offshore part of the Tarfaya basin have yielded no brachiopods.

STRATIGRAPHIC AGES OF SPECIMENS IN THE WHITAKER AND
GENTIL COLLECTIONS

Whitaker left no record of the age of the strata from which he made his collection and it has not so far
proved possible to trace the exact locality. All the specimens were obtained from one locality,
recorded as: ‘Ecru, Mogador, Morocco. 500 ft. on plateau edge of 1000 ft. elevation’. The age can
only be assessed on the internal evidence of the fauna and appears to be either Hauterivian or
Barremian. The species represented all occur elsewhere in south-west Morocco in both the
Hauterivian and the Barremian, whereas not all occur in the Yalanginian or Aptian.

Four species are represented in the Whitaker Collection, in the following numbers: Loriolithyris
russillensis, 57; L. valdensis, 39; Juralina ecruensis, 46; Kutchithyris kennedyi, 1 . The predominance of
L. russillensis would suggest, on analogy with the occurrence of the species in Switzerland and
France, a Barremian age. The distribution of these four species in the Gentil Collection is as follows:

L. russillensis

Hauterivian

7

J. ecruensis

Valanginian

25

Barremian

39

Hauterivian

2

Aptian

9

Barremian

15

L. valdensis

Valanginian

25

K. kennedyi

Hauterivian

2

Hauterivian

63

Barremian

1

Barremian

104

Aptian

16

In general these statistics again support a Barremian age for the Whitaker Collection but they may
reflect nothing more than the accidents of collection.

I have followed stratigraphic ages given on the labels of the Gentil Collection because it was not
possible to check each locality in the field, but there are some arguments supporting the general
validity of these labels, even though there must be a number which are wrong. Analysis of all the
localities given on the labels shows that all the specimens from any one locality are assigned
consistently either to a single stage or to two, or rarely three, adjacent stages. Thus a logical series of
localities can be set out, ranging from those credited with yielding only Berriasian and Valanginian
fossils to those credited with yielding fossils only of Clansayesian age.

520

PALAEONTOLOGY, VOLUME 23

SYSTEMATIC PALAEONTOLOGY

Order terebratulida Waagen, 1883
Suborder terebratulidina Waagen, 1883
Superfamily terebratulacea Gray, 1 840
Family terebratulidae Gray, 1840
Subfamily sellithyridinae Muir-Wood, 1965

Remarks. Loriolithyris and Boubeithyris are closely related sellithyridine genera. The corniced hinge
plates which are the most distinguishing feature of Boubeithyris are essentially the same in detailed
structure as the piped hinge plates of Loriolithyris. Both genera essentially have small crural bases
(attached to the inner edges of the hinge plates) which become encased in successive layers of
secondary skeletal tissue (PI. 60, fig. 2; PI. 61, figs. 2, 3). The function of this is presumably to
strengthen the junction of hinge plates and crural bases. These structures are not inner hinge plates,
which some authors claim to be present in Terebratula, although Muir-Wood (1965, p. H775) denies
their presence in that genus, because they show no sign of having taken part in any way in the
attachment of the dorsal pedicle muscles. Boubeithyris and Loriolithyris differ mainly in the shape of
the hinge plates— concave and corniced in Boubeithyris , concave to sigmoid and piped in
Loriolithyris. Externally Boubeithyris is distinguished especially by the close spacing of the plicae of
the anterior commissure. Both differ from Sellithyris in having accessory structures (cornicing or
piping) on the hinge plates and in their much less pentagonal external form.

Paraboubeithyris has an internal structure which is closely related to that of Boubeithyris.
Externally P. plicae looks different at first sight from Boubeithyris spp. but similarities include the
convex cardinal slopes, small size of the median sinus, and the late development of folding. The
external differences, however, seem too great to allow the species to be included in Boubeithyris.
P. plicae is here regarded as a specialized local offshoot from the Boubeithyris stock.

Genus loriolithyris Middlemiss, 1968
Type species. Terebratula russillensis de Loriol, 1866.

Species included. T. russillensis de Loriol, T. valdensis de Loriol, L. melaitensis nov., L. marocensis nov. Range:
Berriasian to Aptian.

explanation of plate 55

Figs. 1-4. Loriolithyris russillensis (de Loriol). Whitaker Coll. \a-d, typical form, plaster cast of specimen
sectioned (see text-fig. 5), BM BB 76544. 2 a-c, wide latifrons- like form, plaster cast of specimen sectioned (see
text-fig. 7), BM BB 76552. 3 a-d, small sharply folded form, plaster cast of specimen sectioned (see text-fig. 6)
BM BB 76543. 4 a-d, thick latifrons- like form, BM B 17293.

Figs. 5-9. Loriolithyris valdensis (de Loriol). 5 a-c, typical form, plaster cast of specimen sectioned (see text-
fig. 11), BM BB 76545. Whitaker Coll. 6 a-d, juvenile rectimarginate form, BM BB 76546, Whitaker Coll.
la-d , juvenile incipiently biplicate form BM BB 76549, Whitaker Coll. 8 a-d, elongate adult form, BM BB
76554, Whitaker Coll. 9 a-d, wide adult form, S. 546/1/12, Gentil Coll., Upper Hauterivian, loc. unknown.

Fig. 10 a-c. Loriolithyris melaitensis sp. nov. Plaster cast of specimen sectioned (see text-fig. 12), S.556/1,
Gentil Coll., Hauterivian, Tizi Ouarioum.

Figs. 11 a-d. Loriolithyris melaitensis sp. nov. Holotype, S. 556/2, Gentil Coll., Barremian, Ait Ben Melait,
Ida ou Guelluill.

All natural size.

PLATE 55

middlemiss, Cretaceous Terebratulidae

522

PALAEONTOLOGY, VOLUME 23

Loriolithyris russillensis (de Loriol)

Plate 55, figs. 1-4; text-figs. 3-7

* 1866 Terebratula russillensis de Loriol, p. 88, pi. E, figs. 12-15.

1867 Terebratula russillensis de Loriol, p. 393, pi. C, figs. 28-31.

1869 Terebratula russillensis de Loriol, p. 28, pi. 4, fig. 1 .
vl872 Terebratula russillensis de Loriol; Pictet, p. 68, pi. 202, figs. 1-8.
vl872 Terebratula latifrons Pictet, p. 67, pi. 201, figs. 16-17.

71964 Sellithyris (7)russillensis (de Loriol); Ager, p. 340.
non 1966 Sellithyris russillensis (de Loriol); Bogdanova and Lobacheva, p. 53, pi. 5, figs. 5-6.
vl968 Loriolithyris russillensis (de Loriol); Middlemiss, p. 176, pi. A, figs. 1-4.

Lectotype. Museum d’Histoire Naturelle, Geneva (Pictet Collection), no. CB 1520. Designated Middlemiss
1968. Fig. Pictet and de Loriol 1872, pi. 202, fig. 4; from the urgonian of La Russille, Yaud, Switzerland.

Material. Fifty-seven specimens from the Whitaker Collection. About fifty-five specimens in the Gentil
Collection.

Remarks. Specimens from Morocco tend to be wider and thinner, in relation to length, than the
typical members of the species from La Russille and Orgon and many have the characters of the form
described by Pictet (1872) as Terebratula latifrons. I have previously (Middlemiss 1968a) believed the
latter form to be a variety of Loriolithyris russillensis and experience of the Moroccan fauna has
reinforced this belief. Forms from the Jura region which Pictet recognized as T. latifrons (Geneva
Museum) are distinct because of their decidedly small umbones and foramina, not because of their
wide depressed shape. They usually display well-developed russillensis-like folding of the shell and as
regards shape there seems to be a complete gradation between the two species. In both south-west
France and south-west Morocco forms apparently referable to L. russillensis show continuous
variation, in the same assemblages, into other forms with the same characters except for the
proportions of shell shape, which are those of T. latifrons. The forms with decidedly small umbones
and foramina do not occur in these regions. The internal skeletal arrangements revealed by serial
sectioning are the same in all these forms: the concave piped hinge plates, situated close to the floor of
the brachial valve, and the sigmoid passage from inner socket ridge to hinge plate, are unmistakeable.
Pictet records his typical T. latifrons forms only from the Upper Valanginian of Villers-le-Lac and
Vesency. L. russillensis was apparently a species-group very variable in proportions of length, width,
and thickness, some members of which, in part of the Jura region and for a short time in the Upper
Valanginian, became locally sufficiently differentiated to deserve recognition as a subspecies
Latifrons' .

30

20

text-fig. 3. Scatter diagrams of relationships of width to length and thickness to length in Loriolithyris
russillensis from the Whitaker Collection.

MIDDLEMISS: CRETACEOUS TEREBRATULIDAE

523

n

10

9

u 7
O

ffl 6
c

< 5

4

3

2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Posterior

5

15 20 25 30

Length

text-fig. 4. Scatter diagrams of the posterior/anterior ratio in Loriolithyris russillensis from the Whitaker Coll.

The main differences between this species and L. valdensis are that L. valdensis is longer is relation
to both width and thickness and has a higher P/A ratio than L. russillensis. These points are
graphically illustrated, as far as the Moroccan specimens are concerned, in text-figs. 3, 4, 8, 9, and 10.
Internally, a point of distinction is that in L. russillensis the hinge plates are close to, or even in
part in contact with, the floor of the brachial valve, whereas in L. valdensis they are raised clearly
above the floor of the valve for their whole width. It can be added that, internally, L. russillensis
has a very short loop, little more than 1 mm from the crural processes to the transverse band in
adult shells. Unfortunately it is characteristic of species of Loriolithyris that the transverse band
is delicate and seldom preserved and I have never yet seen this structure in L. valdensis.

text-fig. 5. Transverse sections through a small, strongly folded specimen of Loriolithyris russillensis. Sections
1.8 and 2.0 are enlarged in order to show the shape of the juvenile hinge plates enclosed within the cardinal
process (punctate tissue is stippled in section 1 .8). Section 4.2 is enlarged in order to show the structure of the
piped hinge plates. BM BB 76544, Whitaker Coll. A — scale for sections, 1.8, 2.0 and 4.2. B -scale for the

remaining sections.

524

PALAEONTOLOGY, VOLUME 23

Distribution. Ager (1964) claims this species in the Berriasian of the southern Jura and Pictet (1872)
notes it in the Valanginian of Sainte-Croix (Vaud). It certainly occurs in the Hauterivian of Vaud,
Doubs, Haute-Marne, and Yonne and of Les Corbieres (Aude). It is at its most abundant, however,
in the Barremian of Vaud, Jura, the south-east Paris Basin, Bouches-du-Rhone, Gard, Aude, eastern
Spain, and Ibiza. It occurs very rarely in the Aptian of Aude. In south-west Morocco it ranges from
the Hauterivian to Aptian inclusive.

text-fig. 6. Transverse sections through a small, strongly folded specimen of Loriolithyris russillensis to show
the short loop. Section 1.8 is enlarged in order to show the shape of the juvenile hinge plates enclosed in the
cardinal process. Sections 2.2 and 2.6 are enlarged in order to show the primary hinge plates (stippled). The
maximum height of the crural processes is seen in section 3.4. BM BB 76543, Whitaker Coll. A— scale for
sections 1.8, 2.2, and 2.6. B— scale for the remaining sections.

Loriolithyris valdensis (de Loriol)

Plate 55, figs. 5-9; text-figs. 8-11

v*1868
vl872
non 1939
1960
pars 1966
vl968
1972
v!975

Terebratula valdensis de Loriol, p. 52, pi. 4, figs. 9-12.

Terebratula valdensis de Loriol; Pictet, p. 66, pi. 201, figs. 11-15.

Terebratula valdensis var. kentugajensis Moisseev, p. 200, pi. 2, fig. 6.

‘ Terebratula ’ valdensis de Loriol; Smirnova, p. 374, pi. 1, fig. 1.

Sellithyris valdensis (de Loriol); Bogdanova and Lobacheva, p. 55, pi. 5, fig. 7 ( non pi. 7, fig. 11).
Loriolithyris valdensis (de Loriol); Middlemiss, p. 182, pi. A, fig. 5.

Sellithyris valdensis (de Loriol); Smirnova, p. 81, pi. 7, fig. 5.

Loriolithyris valdensis (de Loriol); Dieni and Middlemiss, p. 182, pi. 36, figs. 9-10.

Lectotype. Museum d’Histoire Naturelle, Geneva (Arzier Collection), no. CB 1505. Designated Middlemiss
1968. Fig. de Loriol 1868, pi. 4, figs. 9 a-d, from Bed B, Valanginian, Arzier Quarry, Vaud, Switzerland.

Material. Thirty-nine specimens in the Whitaker Collection. About 200 specimens in the Gentil Collection.
Eight specimens from Barremian or Aptian, Tizi ou Elma, Agadir (D.V. Ager Collection).

MIDDLEMISS: CRETACEOUS TEREBRATULIDAE

525

text-fig. 7. Transverse sections through a broad, latifrons- like specimen of Loriolithyris russillensis. Sections 2.8
and 3.2 are enlarged in order to show the shape of the juvenile hinge plates. The structure of the piped inner
margin of the hinge plate is enlarged at section 4.6 (see plate 60, fig. 5). The transverse band was not preserved in
this specimen. BM BB 76552, Whitaker Coll. A— scale for sections 2.8, 3.2, and 4.6 (inset). B— scale for the

remaining sections.

Description. Text-figs. 8 and 9 compare the thirty-nine specimens in the Whitaker Collection with a collection of
227 specimens made at the type locality of Arzier by Monsieur Roessinger and preserved at the Geneva Natural
History Museum. The isometric development of length and width is well shown in text-fig. 9. Thickness in
relation to length develops allometrically, although with a very small differential growth ratio (text-fig. 8). Text-
fig. 10 shows that the P/A ratio develops allometrically with a very wide range of variation (about double the
width of that shown by Sellithyris sella from the Isle of Wight Aptian (Middlemiss 1968ft, fig. 9)). The smallest
shells, less than 5 mm in length, are subcircular in ventral profile but posterior length increases allometrically
with growth, at the expense of anterior length. There is a marked tendency for Moroccan specimens to have a
lower P/A ratio, i.e. to have a relatively greater anterior length than those from the type area; in this respect the
lectotype has an anomalous position.

The anterior commissure remains rectimarginate until the shell is about 12 mm in length. It then passes
through a well-marked uniplicate stage until the shell reaches a length of about 16 mm, after which plicae and
sinuses are rapidly developed, shells from 17 mm upwards being normally sulciplicate. The episulcate stage is
occasionally seen at Arzier but is very rare in Morocco.

Remarks. Differences between this species and L. russillensis were discussed above. Roch remarks on
the abundance of this species in the Valanginian and Barremian of south-west Morocco, especially in
the Barremian of Jebel Graa and Aghbalou.

526

PALAEONTOLOGY, VOLUME 23

L.valdensis

Lectotype of L.valdensis
K.kennedyi

=i:Sr

. *

\ %.>•

5 10 15 20 25 30

Length

text-fig. 8. Scatter diagrams of the relationship of thickness to length in Loriolithyris valdensis (Arzier and
Whitaker Colls.) and Kutchithyris kennedyi (all available specimens).

text-fig. 9. Scatter diagram of the relationship of width to
length in Loriolithyris valdensis (Arzier and Whitaker Colls.).

MIDDLEMISS: CRETACEOUS TEREBRATULIDAE

527

Posterior Length

text-fig. 10. Scatter diagrams of the posterior/anterior ratio in Loriolithyris valdensis from Arzier.

35

Distribution. Berriasian and Valanginian of Vaud and Haute-Savoie; Valanginian and Hauterivian of
the south-east Paris Basin; Valanginian of Georgia and Hauterivian of the northern Caucasus
(Smirnova 1972); Neocomian of the Kopet Daga (Bogdanova and Lobacheva 1966); Hauterivian of
north-east Bulgaria. Valanginian and Hauterivian of eastern Spain; Barremian of Basses- Alpes and
Alpes-Maritimes. Aptian of La Presta (Neuchatel). In south-west Morocco the range is Valanginian
to Aptian inclusive.

text-fig. 1 1 . Transverse sections through Loriolithyris valdensis. Sections 2.8-4.4 are enlarged in order to show
the shape of the juvenile hinge plates enclosed within the cardinal process and the structure of the crural bases
within the piped inner margins of the hinge plates. Maximum development of the crural processes is seen in
section 7.2. The transverse band was not preserved in this specimen. BM BB 76545, Whitaker Coll. A— scale for
sections 2.8-4.4. B— scale for the remaining sections.

528

PALAEONTOLOGY, VOLUME 23
Loriolithyris melaitensis sp. nov.

Plate 55, figs. 10, 11; text-fig. 12

vl951 Terebratula salevensis de Loriol; Gigout, p. 360, pi. 9, figs. 15-18.

Types. Holotype, Gentil Collection specimen no. S. 556/2, from the Barremian of Ait Ben Melait. Dimensions:
L 31, W 28-5, T 18-5. Paratype, Gentil Collection specimen no. S. 556/1 (locality as holotype).

Material. Ten specimens in the Gentil Collection; nine from the Hauterivian of Tizi Ouarioum, one from the
Barremian of Ait Ben Melait, Ida ou Guelluill.

Diagnosis. Loriolithyris of elongate oval ventral profile, becoming thick in adult stage (thickness nearly equal to
width); P/A ratio slightly more than 1 . Valves equally convex. Umbo suberect. Foramen mesothyrid, attrite,
slightly labiate. Beak ridges rounded. Symphytium very short, but visible. Lateral commissure strongly arched;
anterior commissure sulciplicate. Shell not folded except at extreme anterior. Small pedicle collar present. Hinge
plates concave, piped. Crural bases well developed. Crural processes slightly incurved. Transverse band high-
arched, rounded.

Remarks. The thick, well-filled appearance of the shell, the arched lateral commissure, and the
relative lack of folding give this species a superficial resemblance to Tropeothyris salevensis
(de Loriol) and it is likely that Ambroggi’s (1963) record of T. salevensis in both Lower and Upper
Barremian of south-west Morocco refers to this species.

text-fig. 12. Transverse sections through Loriolithyris melaitensis. Section 4.8 is enlarged in order to show the
shape of the juvenile hinge plates enclosed within the cardinal process and the boundary between punctate tissue
(stippled) and impunctate laminated tissue. Section 5.2 is enlarged in order to show the primary hinge plates
(stippled). The crural bases, unusually large for Loriolithyris, are well shown in sections 6.4-7. 6. Section 9.6
shows the maximum development of the crural processes. S.556/1, Gentil Coll., Hauterivian, Tizi Ouarioum.

A— scale for sections 4.8 and 5.2. B— scale for the remaining sections.

MIDDLEMISS: CRETACEOUS TEREBRATULIDAE

529

It is distinguished from other species of Loriolithyris especially by the unusually large size of the
crural bases attached to the inner edges of the hinge plates (text-fig. 12), but also by its external
appearance.

Distribution. Hauterivian and Barremian of south-west Morocco.

Loriolithyris marocensis sp. nov.

Plate 56, figs. 1, 2; text-fig. 13

Types. Holotype, Gentil Collection specimen no. S. 547/2; age given as Upper Hauterivian (locality unknown).
Dimensions: L 49-75, W 32, T 26-25. Paratype, Gentil Collection specimen no. S. 547/1.

Material. Sixteen specimens in the Gentil Collection: four from the Hauterivian (including Oued Tidzi), two
from the Barremian, Chaine d’Azour, ten from the Barremian of Oued Aghbalou.

Diagnosis. Elongate Loriolithyris , attaining large size; P/A ratio slightly more than 1. Valves equally convex.
Umbo erect. Foramen mesothyrid, labiate. Beak ridges rounded. Symphytium hidden in adult stage. Lateral
commissure very strongly arched. Anterior commissure sulciplicate with shallow median sinus, rarely episulcate.
Shell folded only at extreme anterior, marked by strong concentric growth ridges. Small pedicle collar present.

text-fig. 13. Transverse sections through Loriolithyris marocensis. Sections 5.6 and 6.0 are enlarged in order to
show the detailed structure of the cardinal process, with juvenile primary hinge plates (fine stipple) surrounded
by laminated thickening and the body of the cardinal process infilled with punctate skeletal tissue (coarse
stipple). Section 6.4 is enlarged to show the primary hinge plates (stippled). Maximum development of the crural
processes is seen in section 1 1.2. Note the height of the transverse band above the floor of the valve in section
14.4. S. 547/1, Gentil Coll., Hauterivian, locality unknown. A— scale for sections 5.6, 6.0, and 6.4. B— scale for

the remaining sections.

530

PALAEONTOLOGY, VOLUME 23

Hinge plates initially concave, becoming rounded L-shaped, piped. Cardinal process extends along the hinge
plates, leaving small dorsal umbonal cavity. Transverse band high-arched, with somewhat pointed crest, high
above floor of valve.

Remarks. As all the specimens available are fully adult or gerontic little can be said about the
ontogeny, except that biplication of the anterior commissure and folding of the shell appear to
develop very late. L. marocensis differs from most species of the genus in the large size attained when
adult and the massive, little-folded form of the shell; in those respects it is nearest to L. melaitensis but
differs markedly from that species in its internal structures: L. melaitensis is distinguished by the large
size of its crural bases whereas in L. marocensis the crural bases are small and enclosed within the
piped edge of the hinge plate as usual in Loriolithyris. L. marocensis is also distinct from other species
of the genus in the L-shape developed by the hinge plates as seen in transverse section (text-fig. 13).
Another Moroccan Lower Cretaceous species which closely resembles L. marocensis is Cyrtothyris
middlemissi ; the latter is broader in relation to length, and has a less erect umbo, and lacks the
loriolithyrid boldly arched lateral commissure of L. marocensis, besides the internal differences.

Distribution. Hauterivian and Barremian of south-west Morocco.

Genus boubeithyris Cox and Middlemiss, 1978
Type species. Terebratula boubei d’Archiac, 1847.

Species included. T. boubei d’Arch. Boubeithyris buzzardensis Cox and Middlemiss, B. tibourrensis nov.,
B. pleta nov. Range: Hauterivian?, Barremian to Cenomanian.

Boubeithyris tibourrensis sp. nov.

Plate 56, figs. 3, 4; text-fig. 14

Types. Holotype, Gentil Collection specimen no. S. 548/2/1, from Butte de Tibourr’m; labelled Aptian (more
likely Barremian). Dimensions: L 20-5, W 16-25, T 12-5. Paratype, Gentil Collection specimen no. S. 552/3/1,
Barremian, Tibourr’m.

Material. Two specimens in the Gentil Collection from Butte de Tibourr’m, one labelled Aptian, the other
Barremian.

Diagnosis. Boubeithyris regularly oval as seen in ventral profile, apart from short straight anterior (between the
lateral plicae). Valves equally convex. P/A ratio slightly greater than 1. Umbo suberect; beak ridges moderately
well defined. Foramen mesothyrid, marginate, slightly telate. Lateral commissure arched. Anterior commissure
sulciplicate; lateral plicae close together; median sinus narrow. Plication reflected by small folds and sulci in
extreme anterior part of brachial valve only. Hinge plates thin, concave, piped to strongly corniced. Inner socket

EXPLANATION OF PLATE 56

Figs. 1, 2. Loriolithyris marocensis sp. nov. 1 a-d, holotype, S. 547/2 Gentil Coll., Upper Hauterivian, loc.
unknown. 2 a-c, plaster cast of specimen sectioned (see text-fig. 13), S. 547/1, Gentil Coll., Upper
Hauterivian, loc. unknown.

Figs. 3, 4. Boubeithyris tibourrensis sp. nov. 3 a-d, holotype, S.548/2/1, Gentil Coll., Barremian or Aptian,
Butte de Tibourr’m. 4 a-c, plaster cast of specimen sectioned (see text-fig. 14), S.522/2/1, Gentil Coll.,
Barremian, Tibourr’m.

Figs. 5, 6. Boubeithyris pleta sp. nov. 5 a-d, holotype, S. 553/3, Gentil Coll., Barremian, Sidi Bou Rjaa. 6 a-c,
plaster cast of specimen sectioned (see text-fig. 15), S.553/1, Gentil Coll., Barremian, Sidi Bou Rjaa.

Fig. 7 a-d. Boubeithyris pleta sp. nov. Large typical specimen, S. 557/6, Gentil Coll., Barremian, Igueni Ouram.

Fig. 8. Paraboubeithyris plicae gen. et sp. nov. 8 a-d, holotype, S. 548/1/3, Gentil Coll., Barremian, Vallee Asif
Ait Ameur.

All natural size.

PLATE 56

middlemiss, Cretaceous Terebratulidae

532

PALAEONTOLOGY, VOLUME 23

text-fig. 14. Transverse sections through Boubeithyris tibourrensis. Sections 3.0 and 3.3 are enlarged to show the
initial shape of the juvenile hinge plates within the cardinal process. Cornicing of the hinge plates is best seen in
sections 4.2-5.4. Section 7.8 shows the maximum development of the crural processes. The transverse band was
not preserved in this specimen. S.552/2/1, Gentil Coll., Barremian, Tibourr’m. A— scale for sections 3.0 and 3.3
B— scale for the remaining sections.

ridges narrow. Accessory articulation slightly developed. Euseptoidum short, confined to posterior part of hinge
plates, flanked by lateral ridges.

Remarks. This species closely resembles the type species in general shape, the close-set lateral plicae
being particularly characteristic of both species. B. tibourrensis differs from B. boubei in being more
oval, less pentagonal, in ventral profile and somewhat more convex in lateral profile. Like B. boubei, it
differs from B. buzzardensis in being narrower and thicker, having a higher P/A ratio and folding
almost confined to the brachial valve. Internally the hinge plates are more deeply concave and the
cornice-structure better developed than in either B. boubei or B. buzzardensis. A species of
Boubeithyris which occurs in the Aptian of the Jura region, so far undescribed, differs from B.
tibourrensis in being still more convex and in having a lateral commissure still more strongly arched,
lateral plicae even closer together, and a longer symphytium. Although only two specimens are
available, this species is important because it extends back to the Barremian the time-range of the
typical oval form of Boubeithyris, which can thence be traced through the undescribed Aptian species
from the Jura to B. boubei itself in the Albian and Cenomanian.

Distribution. Barremian of south-west Morocco.

Boubeithyris pleta sp. nov.

Plate 56, figs. 5-7; text-fig. 15

Types. Holotype, Gentil Collection specimen no. S.553/3, from the Barremian of Sidi Bou Rjaa, Oued Tidzi.
Dimensions: L 25-5, W 23-75, T 15. Paratypes, Gentil Collection specimens S. 553/1 (age and locality as
holotype) and S. 557/6, Barremian, Igueni Ouram.

MIDDLEMISS: CRETACEOUS TEREB R ATU LI D AE

533

Material. Twenty-one specimens in the Gentil Collection.

Name. Latin pleta, ‘filled’, from the well-filled appearance of the shell.

Diagnosis. Boubeithyris almost as broad as long, with thickness less than two-thirds of width. Subcircular in
ventral profile. Valves equally convex. P/A ratio about 1 . Umbo short, suberect. Beak ridges rounded. Foramen
mesothyrid, attrite. Lateral commissure arched. Anterior commissure sulciplicate; median sinus low. Shell little
folded. Hinge plates concave, piped to strongly corniced. Euseptoidum short and weak. Transverse band
moderately high.

Remarks. In external appearance this species could be taken for a sulciplicate species of Sellithyris but
the extremely gentle folding imparts to the shell a tumid or ‘well-filled’ appearance which is
distinctive; also the ventral profile is less pentagonal than in most species of Sellithyris , even
S. deningeri which is a particularly rounded species of that genus. It differs from other species of
Boubeithyris mainly in being relatively wide and flat in comparison with its length and in the wider
spacing of the plicae of the anterior commissure.

Distribution. Hauterivian(?) and Barremian of south-west Morocco.

text-fig. 15. Transverse sections through Boubeithyris pleta. Sections 2.8-4.8 are enlarged to show details of the
structure of the hinge plates and of the cornicing. Maximum height of the crural processes is seen in section 6.4.
S.553/1, Gentil Coll., Barremian, Sidi Bou Rjaa. A— scale for sections 2. 8-4.8. B— scale for the remaining

sections.

Genus paraboubeithyris gen. nov.

Type species. Paraboubeithyris plicae sp. nov.

Diagnosis. Ventral profile rounded pentagonal, as wide as, or wider than, long. Depressed. P/A ratio slightly
more than 1. Umbo suberect to erect. Beak ridges rounded. Foramen mesothyrid, marginate, becoming labiate.
Lateral commissure strongly arched. Anterior commissure deeply uniplicate, or sulciplicate with very small
median sinus. Brachial valve has a strong median fold extending from the umbonal region to the anterior;
corresponding to a deep, wide sulcus in the anterior half of the pedicle valve. Hinge plates concave, thin, sharply
differentiated from the inner socket ridges; piped to strongly corniced. Transverse band high-arched.
Euseptoidum weak, flanked by two low lateral ridges.

534

PALAEONTOLOGY, VOLUME 23
Paraboubeithyris plicae sp. nov.

Plate 56, fig. 8; Plate 57, figs. 1-3; text-fig. 16

Types. Holotype, Gentil Collection specimen no. S. 548/1/3, from the Barremian of the Vallee Asif Ait Ameur.
Dimensions: L22, W 22-5, T 10. Paratypes, Gentil Collection specimens S.546/1/1, S. 546/1/2, and S.546/1/3; age
given as Upper Hauterivian (locality unknown).

Name. Genitive of Latin plica , ‘a fold’.

Material. Thirty-three specimens in the Gentil Collection, of which ten are from the Barremian of Vallee Asif Ait
Ameur and twelve from the Barremian of Ida ou Tanan, the remainder being unlocated.

Description. This species has a deep and dramatic uniplication, especially in the more gerontic specimens. Some
of the smaller specimens have a very small median sinus, so that the anterior commissure is strictly sulciplicate,
but the sinus is always extremely small and usually asymmetrically placed. We lack juvenile representatives of the

text-fig. 1 6. Transverse sections through Paraboubeithyris plicae. Section 1 .2 shows the pedicle collar (stippled).
Section 3.2 shows a dorsal umbonal cavity. The corniced hinge plates are well seen in sections 4.8 and
5.2. S.546/1/1, Gentil Coll., Hauterivian, locality unknown.

EXPLANATION OF PLATE 57

Figs. 1-3. Paraboubeithyris plicae gen. et sp. nov. 1 a-c, plaster cast of specimen sectioned (see text-fig. 16),
S.546/1/1, Gentil Coll., Upper Hauterivian, loc. unknown. 2 a-d, adult but uniplicate form, S. 546/1/2, Gentil
Coll., Upper Hauterivian, loc. unknown. 3 a-d, elongate form showing incipient biplication, S.546/1/3,
Gentil Coll., Upper Hauterivian, loc. unknown.

Fig. 4 a-c. Cyrtothyris middlemissi (Calzada), plaster cast of specimen sectioned (see text-fig. 1 8), BM BB 76564,
D.V. Ager Coll., Aptian, Ait Abaid, Agadir.

Figs. 5, 6. Cyrtothyris middlemissi (Calzada). 5 a-c, plaster cast of specimen sectioned (see text-fig. 1 7), BM BB
76565, Calzada Coll., Aptian, La Roqueta, Spain. 6 a-c, BM BB 76566, Calzada Coll., Albian, Peracals,
Spain.

All natural size.

PLATE 57

middlemiss, Cretaceous Terebratulidae

536

PALAEONTOLOGY, VOLUME 23

species but specimens in the Gentil Collection indicate that the sinus appears late, following juvenile
rectimarginate and uniplicate stages, when the shell has attained a length of about 15 mm, and is then lost again
in the gerontic stage. Some individuals show no sign of biplication, however.

Remarks. This species is almost certainly the form that both Roch and Ambroggi identified as
Terebratula collinaria d’Orbigny, which it resembles in general shape. The principal differences
between these two species are (a) T. collinaria is always uniplicate, never biplicate; ( b ) the cardinal
slopes of T. collinaria tend to be concave in dorsal profile, with a sharply produced umbo, those of
P. plicae are convex, with an umbo which does not protrude beyond the curve of the cardinal slopes;
(c) T. collinaria has relatively flat hinge plates with no trace of the corniced structure characteristic of
Paraboubeithyris.

Distribution. Barremian of south-west Morocco.

Subfamily rectithyridinae Muir-Wood, 1965
Genus cyrtothyris Middlemiss, 1959

Type species. Terebratula depressa var. cyrta Walker, 1868.

Species included. T. depressa var. cyrta Walker, T. depressa var. uniplicata Walker, T. depressa var.
cantabridgiensis Walker, T. seeleyi Walker, T. dallasi Walker, Cyrtothyris middlemissi Calzada, C. cyrta arminiae
Middlemiss, ‘ Cyrtothyris ’ maynci Owen. Range: Valanginian to Albian.

Cyrtothyris middlemissi Calzada
Plate 57, figs. 4-6; text-figs. 17, 18
* 1972 Cyrtothyris middlemissi Calzada, p. 66, fig. 1.

Holotype. Geological Museum of the Seminario de Barcelona, specimen no. 23.346, from the Aptian of La
Roqueta, Garraf, Barcelona.

text-fig. 17. Transverse sections through Cyrtothyris middlemissi. Sections 6.0 and 6.4 are enlarged to show the
initial horizontal cuneate shape of the hinge plates. BM BB 76565, Coll. S. Calzada, Aptian, La Roqueta, Spain.
A— scale for sections 6.0 and 6.4. B— ^ scale for the remaining sections.

MIDDLEMISS: CRETACEOUS TEREBRATULIDAE

537

text-fig. 18. Transverse sections through Cyrtothyris middlemissi. Maximum height of the crural processes is
seen at 16.4. BM BB 76564, Coll. D. V. Ager, Aptian, Ait Abaid, Agadir, Morocco.

Material. Nineteen specimens in the Gentil Collection (seventeen from the Clansayesian of Sidi Bou Rjaa, one
from the Clansayesian of Imi ou Tanant, one from the Aptian of Ait Moujjout). Three specimens from probable
Aptian, Ait Abaid, north-east of Agadir (Ager Collection). Also nineteen other specimens: three from the
Aptian of La Roqueta (Calzada Collection); four from the Upper Aptian, Plan de Coloubret, Taura, Aude
(Charriere Collection); six from the Aptian of Combe Longue, Taura, Aude; two from the Albian of Peracals,
Lerida, Spain (Calzada Collection); four from the Albian of Pic du Seigneur, Tuchan, Aude (Debuyser
Collection).

Original diagnosis (after Calzada 1972). Large forms (maximum L 53, W 36, T 24; L/W ratio 1-1-1 -6; L/T ratio
1-7-21) of subpentagonal to oval ventral profile. Maximum width and thickness in middle of length. Valves
convex, pedicle valve much more so than brachial valve. Valves may show folding (but this character is very
variable). Lateral commissure inclined ventralwards at about 20° and arched. Anterior commissure uniplicate to
slightly sulciplicate. Umbo wide, massive, suberect to erect. Foramen wide, labiate, circular, mesothyrid.
Interareas somewhat concave; beak ridges moderately rounded. Deltidial plates small but visible, fused into a
symphytium. Growth lines visible. Hinge plates concave, somewhat clubbed, becoming anteriorly persistently
virgate or even V-shaped. Angle between the crural bases and the crural rami 70°-100°. Loop strongly recurved
in a posterior direction so that no one serial section includes the whole of the arch of the transverse band.

Remarks. Specimens from Morocco and from the Albian of north-east Spain exceed Calzada’s stated
maximum width (up to 43 mm); nevertheless all specimens available fall into the range of L/W ratios
given in his diagnosis. On the other hand specimens from both areas, and including the type locality,
fall outside the range of L/T ratios given (extremes are specimen MDA 2/1 , from Morocco, 1 -57 and
CaP2, from the Albian of Peracals, 2-12). Calzada understates the plication of the anterior
commissure, which is normally gently sulciplicate in the adult stage. The foramen should be described

538

PALAEONTOLOGY, VOLUME 23

as strongly marginate, labiate in the adult stage. The wide triangular shape of the loop and the strong
recurvature of the transverse band are generic features in Cyrtothyris (Middlemiss 1976).

Distribution. Aptian of Aude and north-eastern Spain; Aptian (including Clansayesian) of south-
western Morocco; Albian of Aude and north-eastern Spain.

Subfamily uncertain
Genus kutchithyris Buckman, 1918

Type species. Terebratula acutiplicata Kitchin, 1900.

Original definition (Buckman 1918). ‘Permesothyrid (beak stout, broad, quite short, thickened with callus,
obliquely truncate, foramen large, circular, attrite, close to umbo, symphytium very short); morphogeny,
biconvex to strongly sulciplicate; muscle-tracks obliterated posteriorly, not reaching far down valves, rather
sharply divergent, starting not from the umbo but from about midway of the posterior half of the shell, showing
little more than scars; dorsal septum feeble— ovarian areas large, mammillate on cast. The muscle scars
posteriorly obliterated and diverging from a point well removed from the umbo, the short beak with little
exposure of symphytium: these characters at once distinguish the genus.’

Diagnosis. Umbo suberect to incurved. Foramen mesothyrid to epithyrid; may be slightly labiate. Development
of anterior commissure uniplicate to sulciplicate, more rarely to episulcate. Hinge plates wide, concave,
flattening anteriorly, very little differentiated from the laterally deflected inner socket ridges. Crural bases low
where attached to hinge plates, rapidly elongating anteriorly and passing into high, thin, slightly flanged crural

text-fig. 19. Transverse sections through Kutchithyris acutiplicata (type species of the genus). Sections 4.8-6.0
are enlarged in order to show details of the structure of the cardinal process. The crural bases first appear at 6.0.
The transverse band at 13.6 is broken and partially displaced. BM 52420, Putchum Group (Upper Jurassic),
Jumara, Kutch, India. A— scale for sections 4.8-6.0. B— scale for the remaining sections.

MIDDLEMISS: CRETACEOUS TEREBRATULIDAE

539

5 MM

text-fig. 20. Transverse sections through Kutchithyris subsella. Sections 4.7-5.9 are
enlarged to show the initial shape of the hinge plates at 4.7 and 5.1, the primary hinge
plates (stippled) at 5.5, and the first appearance of the crural bases at 5.9. Maximum
height of the crural processes is seen at 7.9. The transverse band was not preserved in
this specimen. BM BB 76555, Kimeridgian, Le Havre, France. A — scale for sections
4.7-5.9. B — scale for the remaining sections.

processes. Hinge plates and crural processes usually clubbed. Descending lamellae thin. Transverse band high-
arched, ogival. Euseptoidum present but usually weak; may be bounded by two low euseptoidum-like ridges
bounding the adductor impressions.

Remarks. The species here ascribed to this genus differ one from another considerably in external
proportions, from the highly convex globular form of Kutchithyris brivesi, through the pentagonal
ventral profile of K. acutiplicata and K. subsella to the elongate form of K. kennedyi. They are linked,
however, by close similarity in the internal characters, especially those of the hinge plates, inner
socket ridges, and crural bases. Buckman erected the genus Kutchithyris mainly to accommodate six
species from the Bathonian and Callovian of India previously established by Kitchin but he also
included two European species of Deslongchamps and two newly established species of his own from
the English Great Oolite (Bathonian) of Bradford-on-Avon, K. fulva and K. egregia.

I here refer to Kutchithyris the species T. subsella Leymerie, a familiar Upper Jurassic species in
Europe, which has been previously referred to Sellithyris by Barczyk (1969). I exclude it from
Sellithyris mainly because of the lack of differentiation between hinge plates and inner socket ridges,
the detailed form of the hinge plates (as seen in transverse section they are like hockey sticks), and the
form of the crural processes; these are features which it shares with other species of Kutchithyris.
K. subsella survived into the Lower Cretaceous and occurs in the Upper Valanginian of La Querola

540

PALAEONTOLOGY, VOLUME 23

. 70

text-fig. 21 . Transverse sections through Kutchithyris subsella. Sections 4.2 and 4.6 are enlarged to show detail
of the primary hinge plates. The crural bases are first seen at 5.0. The crural processes are at their maximum
height at 9.0. The transverse band was not preserved in this specimen. BM BB 76558, Coll. M. Durand Delga,
Niveau 14A, Valanginian, La Querola, Spain. A — scale for sections 4.2 and 4.6. B — scale for the remaining

sections.

EXPLANATION OF PLATE 58

Figs. 1 -6. Kutchithyris kennedyi sp. nov. 1 a-d, holotype, BM BB 76556, Y. Champetier Coll., Hauterivian or
Barremian, Oliva, Valencia, Spain. 2 a-c, plaster cast of specimen sectioned (see text-fig. 23), BM BB 76557,
Y. Champetier Coll., Hauterivian or Barremian, Oliva, Valencia, Spain. 3 a-d, BM BB 76559, Durand Delga
Coll., Valanginian, La Querola, Alicante, Spain. 4 a-c, typical specimen, BM BB 76562, W. J. Kennedy Coll.,
Lower Barremian, Les Moulins, Mont Chauve, Nice, France. 5 a-c, large adult specimen, plaster cast of
specimen sectioned (see text-fig. 24), BM BB 76561, Y. Rangheard Coll., ?Hauterivian, Punta Torreta, Ibiza.
6 a-c, plaster cast of specimen sectioned (see text-fig. 22), S. 552/1/1, Gentil Coll., Hauterivian, Ifrech-Oued-
Igouzoulen.

Figs. 7-9. Kutchithyris brivesi (Roch). la-c, plaster cast of specimen sectioned (see text-fig. 26), S. 549/2, Gentil
Coll., Hauterivian, Ifrech-Oued-Igouzoulen. 8 a-d, uniplicate specimen, S.549/3, Gentil Coll., Hauterivian,
Ifrech-Oued-Igouzoulen. 9 a-d, gerontic episulcate specimen, S. 549/4, Gentil Coll., Hauterivian, Ifrech-
Oued-Igouzoulen.

All natural size.

PLATE 58

middlemiss, Cretaceous Terebratulidae

542

PALAEONTOLOGY, VOLUME 23

north of Alcoy, Alicante, Spain (Durand Delga Collection). The other Cretaceous species of the
genus, which are described here, are new.

Species included. Bathonian: T. hypsogonia Kitchin, T. acutiplicata Kitchin, T. propinqua Kitchin, T.
circumdata Deslongchamps, IK. fulva Buckman, IK. egregia Buckman. Callovian: T. aurata
Kitchin, T. jooraensis Kitchin, IT. longicarinata Kitchin, T. subcanaliculata Deslongchamps.
Oxfordian to Valanginian: T. subsella Leymerie. Valanginian to Barremian: K. kennedyi nov.,
K. brivesi (Roch).

Range of the genus. Bathonian to Barremian.

Kutchithyris kennedyi sp. nov.

Plate 58, figs. 1-6; text-figs. 22-24

Types. Holotype, BM BB 76556, from Oliva, Valencia, Spain (Champetier Collection). The horizon is dubious
but is probably Hauterivian or Barremian. Dimensions: L 30, W 20, T 18-5. Paratypes. BM BB 76557, Oliva,
Valencia, Spain; BM BB 76559, Upper Valanginian, La Querola, Alicante, Spain; BM BB 76561, PuntaTorreta,
Ibiza; BM BB 76562 and 76563, Lower Barremian, Mont Chauve, Alpes Maritimes, France; Gentil Collection
S. 552/1/1, Hauterivian, Ifrech Oued Igouzoulen, Morocco.

Material. Three specimens from Oliva, Valencia, Spain (Champetier Collection, horizon uncertain). Five
specimens from niveau 14A at La Querola, north of Alcoy, Alicante, Spain (Busnardo and Durand Delga 1960)

b

text-fig. 22. Transverse sections through Kutchithyris kennedyi. Section 3.6 is enlarged to show the juvenile
primary hinge plates within the cardinal process. The crural bases are first seen at 4.4 and maximum development
of the crural processes at 7.2. Sections 3. 6-6.0 S.552/1/1; sections 6.8-10.0 S. 552/1/2. Both specimens Gentil
Coll., Hauterivian, Ifrech-Oued-Igouzoulen. A— scale for section 3.6. B— scale for the remaining sections.

MIDDLEMISS: CRETACEOUS TEREBRATULIDAE

543

(Durand Delga Collection, probably Valanginian). Two specimens from the Lower Barremian of a stream
section 800 m north of Les Moulins, east of Mont Chauve, north of Nice, Alpes Maritimes (Kennedy
Collection). One specimen from Ecru, Morocco (Whitaker Collection). One specimen from Punta Torreta, Ibiza
(Rangheard Collection, probably Hauterivian). Four specimens in the Gentil Collection (three from the
Hauterivian of Ifrech Oued Igouzoulen, one from the Barremian of Asif Ait Ameur).

Name. Named after Dr. W. J. Kennedy, who supplied some of the specimens.

Diagnosis. Kutchithyris of elongate oval ventral profile (width about 0-7 length); thickness more than half length.
P/A ratio 1 -3—1-6. Umbo suberect to erect in adults. Symphytium very short or invisible. Foramen mesothyrid,
labiate. Beak ridges rounded. Anterior commissure sulciplicate to episulcate. Folding of the shell, corresponding
to the plicae and sinuses of the commissure, weak and confined to the anterior third of the shell except in gerontic
stage.

Description. Because of the few specimens available little can be said about the ontogeny of this species except
that the width/length ratio appears to be isometric and to remain constant during growth at a little less than 0-7,
whereas the thickness/length ratio is allometric.

text-fig. 23. Transverse sections through Kutchithyris kennedyi. Sections 3. 2-4.0 are
enlarged in order to show the juvenile hinge plates within the cardinal process (at 3.2) and
the crural bases (at 3.6 and 4.0). Maximum height of the crural processes is seen at 8.0. The
transverse band was not preserved in this specimen. BM BB 76557, Coll. Y. Champetier,
Oliva, Spain. A — scale for sections 3.2-4.0. B— scale for the remaining sections.

544

PALAEONTOLOGY, VOLUME 23

Remarks. This species is easily distinguished from other members of Kutchithyris by its elongate
form. The species with which it is most likely to be confused is Loriolithyris valdensis. K. kennedyi is
thicker in relation to its length than L. valdensis, because the differential growth ratio of this character
is slightly bigger, giving the allometric distribution a slightly steeper slope (fig. 8). In addition, the
brachial valve of K. kennedyi is slightly concave in anterior third, that of L. valdensis uniformly
convex in lateral view. Internally the characters of the hinge plates, inner socket ridges, and crural
bases are all quite different in the two species.

Distribution. ?Valanginian of south-east Spain; Hauterivian and Barremian of south-west Morocco;
?Hauterivian of Ibiza; Lower Barremian of south-east France.

text-fig. 24. Transverse sections through a large, adult specimen of Kutchithyris kennedyi. Sections 5.2 and 5.6
are enlarged to show the juvenile hinge plates (at 5.2) and the primary hinge plates (stippled at 5.6). The crural
bases are already visible at 5.6. BM BB 76561, Coll. Y. Rangheard, Punta Torreta, Ibiza. A — scale for sections
5.2 and 5.6. B — scale for the remaining sections.

Kutchithyris brivesi (Roch)

Plate 59, figs. 1, 2; text-figs. 25, 26

v* 1930 Terebratula brivesi Roch, p. 259, pi. 22, figs. 12-13.
vl951 Terebratula brivesi Roch; Gigout, p. 361, pi. 9, figs. 27-34.

Lectotype. Roch figured two specimens but there is confusion in the numbering of the figures; figs. 12a and 13 b j[

represent one specimen, figs. 1 2b and 1 3 a the other. The specimen represented by figs. 1 2a and 1 3 b is here chosen
as lectotype. It is in the collection of the Service de la Carte Geologique du Maroc at Rabat, bearing the number
Ci 55, and is from the Valanginian of Zauouia Embarek des Ida ou Troumma. The label describes it as ‘Coll. E. ;
Roch’ but Roch in his caption gives it as ‘Brives Coll.’.

Paratypes. The specimen figured by Roch as figs. 12 b and 13a (at Rabat, bearing the same number as the lecto-
type and from the same horizon and locality). A specimen in the Roch Collection at Rabat bearing number P 62
and coming from the Berriasian of Dar Caid Tigzirin. Six specimens in the Roch Collection at Rabat bearing the

MIDDLEMISS: CRETACEOUS TEREBRATULIDAE

545

number P 50 and coming from the Valanginian of Oued Igoulouzen. The following specimens in the Gentil Col-
lection: S. 549/1, S. 549/2, S. 549/3, S. 549/4, S. 549/5, S. 559/1, all labelled Hauterivian, Ifrech-Oued-Igoulouzen.
The two specimens figured by Gigout (both numbered 720 in the Gigout Collection, Universite Mohamed V,
Rabat).

Material. Nine specimens from the Roch Collection (detailed above). Forty-eight specimens from the Gentil
Collection (forty-five labelled Hauterivian of Ifrech-Oued-Igoulouzen; three labelled Barremian, Chaine
d’Azour).

Diagnosis. Kutchithyris highly obese in lateral profile, oval in ventral profile. P/A ratio slightly more than 1 .
Brachial valve more convex than pedicle valve. Umbo erect to incurved. Symphytium very short to invisible.
Foramen mesothyrid, labiate in older individuals. Beak ridges rounded. Lateral commissure arched. Anterior
commissure rectimarginate to sulciplicate or episulcate. Shell tumid and little folded, or not folded.
Euseptoidum well developed in the region of the hinge plates and flanked by two lateral ridges.

25r

text-fig. 25. Scatter diagrams of the relationships of thickness to length and thickness to width in Kutchithyris

brivesi (Gentil Coll.).

Description. The growth of this species is accompanied by rapid increase in the thickness/length ratio. In the most
adult individuals thickness can exceed width. The smallest specimens available (L 1 8-5) are either rectimarginate
or gently uniplicate but the later development of the commissure is the most variable character of the species.
Some specimens of 29 mm in length are clearly and deeply uniplicate, while other specimens of similar size are
sulciplicate or, rarely, episulcate. In other specimens again a clearly episulcate commissure is developed at a shell
length of as little as 19-5 mm.

Remarks. This species is distinguishable at once from other species of Kutchithyris and from all the
other species considered here by its globular form and the tumid appearance of both valves.
Internally it differs from other species of Kutchithyris in having a well-developed, although short,
euseptoidum. Both Roch and Gigout underestimate the plication which the anterior commissure
may show in this species. Roch states: ‘La commissure frontale est pratiquement droite, sauf deux
petits plis a peine marques.’ According to Gigout: ‘Commissure frontale droite ou tres legerement
convexe vers la petite valve.’ The larger specimens (L 25-5) in Roch’s own collection, however, are
strongly uniplicate. The form of the anterior commissure of the larger specimens in the Gentil
Collection is very variable, suggesting that Roch and Gigout may have seen only small, relatively
juvenile specimens such as the lectotype. Roch, Gigout, and Ambroggi all give the main occurrence of
this species as of Valanginian age, Roch and Ambroggi recording some also from the Berriasian,
whereas the great majority of the Gentil Collection specimens are labelled Hauterivian, with a few
labelled Barremian. It is possible that strong sulciplication or episulcation was developed in this
species only after the Valanginian. The unity of the species is demonstrated by the remaining

PLATE 59

middlemiss, Cretaceous Terebratulidae

MIDDLEMISS: CRETACEOUS TEREBRATULIDAE

547

text-fig. 26. Transverse sections through Kutchithyris brivesi. Section 4.4 is enlarged to show
detail of the structure of the cardinal process. The crural bases are first seen at 4.8. The crural
processes are at their maximum height at 7.2. S.549/2, Gentil Coll., Hauterivian, Ifrech-Oued-
Igouzoulen. A— scale for section 4.4. B— scale for the remaining sections.

EXPLANATION OF PLATE 59

Figs. 1, 2. Kutchithyris brivesi (Roch). 1 a-d, juvenile but incipiently biplicate specimen, S. 549/5, Gentil Coll.,
Hauterivian, Ifrech-Oued-Igouzoulen. 2 a-d, adult but uniplicate specimen, S.559/1, Gentil Coll.,
Hauterivian, Ifrech-Oued-Igouzoulen.

Figs. 3-7. Juralina ecruensis sp. nov. 3 a-d, holotype, BM BB 76547, Whitaker Coll. 4 a-d, typical uniplicate
form, BM BB 76548. 5 a-d, plaster cast of specimen sectioned (see text-fig. 28), BM BB 76550. 6a-d, juvenile
specimen, BM BB 76551. la-d, elongate adult form, BM BB 76553.

All natural size.

548

PALAEONTOLOGY, VOLUME 23

characters both external and internal. A specimen from Roch’s collection (from the Valanginian of
Oued Igouzoulen) was serially sectioned and differed slightly from the Gentil specimen shown in text-
fig. 26 in having hinge plates less concave in their earlier stages, a less developed euseptoidum, and in
lacking any clubbed thickening of the hinge plates and crural processes. These are signs of
immaturity, confirming that the specimens described by Roch were comparatively juvenile.

Distribution. Berriasian to Barremian of south-west Morocco.

Genus juralina Kyansep, 1961
Type species. Juralina procerus Kyansep.

Original diagnosis (from Kyansep 1961). ‘Shell plano-convex to biconvex. Anterior commissure rectimarginate
to uniplicate. Umbo massive, straight to erect. Deltidium high. Socket ridges high. Cardinal process well
developed and separated from the floor of the dorsal valve. Hinge plates divided, very narrow, in close proximity
to the socket ridges. Crural bases given off ventrally from the hinge plates. Crura narrow, with well-developed,
sharp-pointed crural processes. Loop about one-third of the length of the dorsal valve, triangular, with arched
transverse band. Pedicle collar shaped like a ring valve. Hinge teeth massive, without denticulae. Adductor
muscle impressions oval triangular, narrowing to fine lines posteriorly. Euseptoidum small. Shell smooth,
punctate.’

Emended diagnosis. Shell plano-convex to biconvex, depressed (thickness/length ratio low), subcircular in
ventral profile. Umbo straight to erect. Foramen mesothyrid, slightly labiate. Lateral commissure oblique to
arched; anterior commissure rectimarginate to squarely uniplicate or slightly sulciplicate. Cardinal process well
developed. Hinge plates rectangularly virgate (that is, L-shaped in cross-section with an inner lamina at right
angles to the outer lamina); clubbed. Crural bases given off from the anterior ventral extremities of the hinge
plates. Crural processes high, sharp-pointed, incurved at their extremities. Loop broad; transverse band high-
arched, arcuate to trapezoidal.

Remarks. Kyansep considered that his new genus strongly resembled Lobothyris Buckman but
Juralina differed in having very narrow hinge plates, high socket ridges, and well-developed crural
processes, in lacking a septum to its pedicle collar, and in the elliptical shape of its ventral umbonal
cavity. Boullier (1976) has, however, pointed out several additional differences. Kyansep also
correctly pointed to a marked external resemblance, but equally marked internal differences, between
Juralina and Rectithyris Sahni. In addition to his new species, Kyansep included in Juralina several
species from the Jurassic of Europe: Terebratula rauraca Rollier, T. repelliniana D’Orbigny,
T. censoriensis Rollier, T. bullingdonensis Rollier, T. cotteaui Douville, and T. moravica Glocker. Of
these, T. moravica was referred to a new genus Weberithyris by Smirnova (1969). In her discussion of
the genus Boullier (1976) rejects affinities with Lobothyris , Weberithyris , Tropeothyris Smirnova, and
Postepithyris Makridin but finds considerable resemblance to Cyrtothyris Middlemiss. Boullier
added three more previously established species — T. bauhini, T. valfinensis, and T. subformosa.

Barczyk (1969) added the following species from Upper Jurassic rocks of the Holy Cross
Mountains of Poland to Juralina: T. insignis insignis Schiibler, 1 830, T. insignis maltonensis Oppel,
1858, T. immanis immanis Zejszner, 1856, T. immanis speciosa Schlosser, 1882. Of these, Boullier
(1976) has since referred T. insignis var. maltonensis Oppel to the genus Galliennithyris as
G. maltonensis.

I introduced the terms inner and outer lamina in 1959 and defined them as follows: ‘A virgate hinge
plate is divisible into two parts, the outer lamina from the socket ridge to the virgation and the inner
lamina on the inner (median) side of the virgation.’ The accompanying figure (Middlemiss 1959, text-
fig. 1 j), however, showed cuneate hinge plates with large crural bases. Because of this confusion I later
withdrew the terms inner lamina and outer lamina (Dieni et al. 1975; Middlemiss 1976). Now that
more is known about the detailed structure of terebratulid hinge plates (Cox and Middlemiss 1978)
the terms are seen to be useful in their original sense and I use them here.

Species included. J. procerus Kyansep, ITerebratula rauraca Rollier, IT. repelliniana d’Orbigny,
J. graciosa Kyansep, IT. censoriensis Rollier, T. bullingdonensis Rollier, J. naklivkini Kyansep,

MIDDLEMISS: CRETACEOUS TEREBR ATULIDAE

549

T. cotteaui Douville, J. babugani Kyansep, J. earns Kyansep, T. bauhini Etallon, T. valfinensis de
Loriol, T. subformosa Rollier, J. ecruensis nov.

Range of the genus. Middle Oxfordian to Barremian.

Juralina ecruensis sp. nov.

Plate 59, figs. 3-7; text-figs. 27, 28

Types. Holotype, BM BB 76547, Whitaker Collection. Dimensions: L 34-5, W 27, T 18. Paratypes: Whitaker
Collection specimens BM B 17273, B 17277, BB 76548, BB 76550, BB 76551, BB 76553.

Material. Forty-six specimens in the Whitaker Collection. Forty-two specimens in the Gentil Collection (twenty-
five from the Berriasian or Yalanginian of Tinirt Ait Ameur, two from the Hauterivian of an unnamed locality,
three from the Barremian of Igueni Ouram, twelve from the probable Barremian of Oued Aghbalou).

Diagnosis. Juralina of subcircular to oval ventral profile; maximum width about the mid-line; valves equally
convex. Umbo erect. Foramen mesothyrid, marginate, becoming labiate. Beak ridges rounded. Symphytium
short, hidden in adult stage. Shell smooth, with faint growth lines. Lateral commissure oblique to arched.
Anterior commissure rectimarginate to squarely uniplicate or slightly sulciplicate. Euseptoidum absent or
negligible. Transverse band high-arched, rounded.

Description. Juvenile specimens resemble the adults except in being rectimarginate. At a length of about 22 mm
the characteristic adult uniplicate commissure begins to develop. In adults over about 30 mm in length the

30 -
25

10

“5 10 15 20 25 30 35 40 45

Length

i 10~ 15"

20 25 30

Length

35 40 45

25

20

5 10 15 20 25 30 35

Width

text-fig. 27. Scatter diagrams of the relationships of simple dimensions in Juralina ecruensis (Whitaker Coll.).

550

PALAEONTOLOGY, VOLUME 23

text-fig. 28. Transverse sections through Juralina ecruensis. The first two sections (upper left) are enlarged in
order to show detail of the structure of the cardinal process. Sections 3. 6-5. 2 are enlarged to show the form of the
hinge plates and of the crural bases. Maximum height of the crural processes is seen at 8.0. BM BB 76550 except
that 4. Ox is from BM B 1 7273 and 3.2 from BM B 1 7277 as these showed better the details of the cardinal process
(Whitaker Coll.). A — scale for sections 3.2 and 4.0x. B — scale for sections 3. 6-5. 2. C — scale for the remaining

sections.

uniplica may be angular, the commissure horizontal in the centre; or it may develop a gentle sinus in the centre,
giving a slightly sulciplicate stage. The other main gerontic development is that the foramen becomes labiate in
specimens over about 30 mm in length. Text-fig. 27 shows that there are a few long, narrow variants and others
that are exceptionally thick.

Remarks. This species is referred to Juralina because of ( a) its external appearance, the distinctive
elements of which are the biconvex but moderately depressed form and the erect umbo; ( b ) the
internal characters, especially the L-shaped form of the hinge plates in transverse section, with the
crural bases developed in the extreme ventral tips of the inner laminae in the anterior parts of the
hinge plates only. All these characters appear closely comparable to those described and figured by
Kyansep (1961), Barczyk (1969), and Boullier (1976).

Distribution. Valanginian to Barremian of south-west Morocco.

EXPLANATION OF PLATE 60

Fig. 1. Loriolithyris melaitensis sp. nov. Section 4.8 of text-fig. 12 photographed to show the shape of the
juvenile hinge plates and the distinction between punctate and inpunctate skeletal tissue within the cardinal
process.

Fig. 2. Loriolithyris marocensis sp. nov. Part of section 6.0 of text-fig. 13 photographed to show the primary
piped hinge plate with its secondary clubbed thickening and the structure of the cardinal process.

Fig. 3. Boubeithyris pleta sp. nov. Part of section 4.0 of text-fig. 1 5 enlarged to show the detailed structure of
the junction between hinge plate and inner socket ridge.

Fig. 4. Kutchithyris acutiplicata (Kitchin). Part of section 6.0 of text-fig. 19 enlarged to show the primary hinge
plate with its clubbed thickening and the incipient crural base, all enclosed within the cardinal process.

Linear scale = 2 mm.

PLATE 60

middlemiss, Cretaceous Terebratulidae

552

PALAEONTOLOGY, VOLUME 23

TEREBRATULID SPECIES OF MORE DOUBTFUL OCCURRENCE IN THE
LOWER CRETACEOUS OF SOUTH-WEST MOROCCO

Terebratula sueuri Pictet is recorded by Gigout from the Valanginian and Hauterivian at Safi and by
both Roch and Ambroggi from the Barremian. T. sueuri is a Jura species which is also found rarely in
the Hauterivian of the Lower Saxon Basin. Three specimens in the Gentil Collection, S. 544/1 (from
Safi), S. 547/2/1, and S. 547/2/2 (both from the Barremian of Ait el Faci) have a close external
resemblance to this species and probably represent the form to which the name was applied by
previous authors. Serial sectioning proved these to be an undescribed species of terebratellidine,
which also occurs in the Jura region (Collections of the Institut de Geologie, Neuchatel). Gigout’s
figured specimen (Gigout 1951, pi. 9, figs. 19-22) has a well-developed dorsal median septum and is
almost certainly the same terebratellidine species. The occurrence of these two externally similar but
quite unrelated species together in the Jura region is a good example of homochronous
homoeomorphy.

Terebratula collinaria d’Orbigny is recorded by both Roch and Ambroggi from the Hauterivian
and Barremian and by Roch from the Valanginian also. The records probably refer to Para-
boubeithyris plicae, although the Gentil Collection contains specimens of this species only from the
Barremian.

Tropeothyris salevensis (de Loriol). This is recorded by Gigout from the Valanginian of the
environs of Safi and by Ambroggi from the Barremian of his area. On first viewing the collections I
referred to T. salevensis the specimens which I have here named Loriolithyris melaitensis; Gigout’s
figured specimen (Gigout 1951, pi. 9, figs. 15-18) is apparently similar to these externally except that
it is a gerontic specimen. The records probably refer to L. melaitensis.

Moutonithyris moutoniana (d’Orbigny) is recorded by Roch from the Barremian and by Gigout
from the ‘Neocomian’ and Aptian of Safi and Sidi Bou Zid. Although Gigout gives in synonymy
Pictet’s (1872) figure of the species, not d’Orbigny’s original, his own figured specimen looks
reasonably convincing (Gigout 1951, pi. 9, figs. 23-26). In the Gentil Collection are four specimens
from the Hauterivian of Oued Tidzi, one from the Hauterivian of Ifrech Oued Igoulouzen, four from
the Barremian of Ait el Faci, and seven from the Barremian of Asif Ait Ameur which are probably
this species. M. moutoniana is a sub-Tethyan species of very widespread occurrence throughout the
Lower Cretaceous (see Middlemiss 1976, 1979) and it would indeed be surprising if some specimens
were not to be found in south-west Morocco.

EXPLANATION OF PLATE 61

Fig. 1 . Loriolithyris melaitensis sp. nov. Section 6.8 of specimen S. 556/1 (not included in text-fig. 12) enlarged
to show the development of the crural base with secondary clubbing. The primary hinge plate has a cuneate
relationship to the crural base.

Fig. 2. Loriolithyris melaitensis sp. nov. Section 4.8 of text-fig. 1 2 photographed to show the internal structure
of the cardinal process, especially the distribution of punctate and impunctate skeletal tissue. The juvenile
primary hinge plates have a secondary clubbed thickening which was deposited prior to the incorporation of
the hinge plates into the cardinal process.

Fig. 3. Loriolithyris melaitensis sp. nov. Section 5.2 of text-fig. 12 enlarged to show the primary hinge plate
surrounded by secondary tissue and the first sign of development of the crural base within the piped inner
margin of the hinge plate.

Fig. 4. Loriolithyris russillensis (de Loriol). Section 4.6 of text-fig. 7 enlarged to show the structure of the piped
inner margin of the hinge plate.

Fig. 5. Paraboubeithyris plicae gen. et sp. nov. Part of section 4.8 of text-fig. 16 enlarged to show the structure
of the corniced inner margin of the hinge plate.

Linear scale = 2 mm.

PLATE 61

middlemiss, Cretaceous Terebratulidae

554

PALAEONTOLOGY, VOLUME 23

Sellithyris carteroniana (d’Orbigny) is recorded by Roch from the Berriasian and the Barremian,
by Gigout from the Valanginian (of Safi) and by Ambroggi from the Hauterivian. In the Gentil
Collection there is one specimen from Tinirt Ait Ameur (probably Hauterivian) which has some
resemblance to S. carteroniana in being obese, equidimensional, and strongly episulcate but the
resemblance is closer, in fact, to the Algerian variety or subspecies of S. sella (see below). The same
can be said of Gigout’s figured specimen (Gigout 1951, pi. 9, figs. 11-14). S. carteroniana is an
interesting species from the palaeobiogeographical point of view as (a) it is a characteristic member of
the Jura fauna which is also found in north Germany during the time of the Valanginian-Hauterivian
transgression (Middlemiss 1976, 1979) and ( b ) Terebratula coahuilensis of the Neocomian of
northern Mexico is probably synonymous with it. In view of my thesis of the Jura affinities of the
south-west Moroccan fauna the occurrence of this species would be significant. Unfortunately there
is no evidence that all the records do not refer to S. sella, although some may refer to Boubeithyris
pleta.

Sellithyris sella (J. de C. Sow) is recorded by both Roch and Ambroggi from the Barremian and
Gargasian and by Roch from the Bedoulian also. This almost ubiquitous Lower Cretaceous species
would be expected to occur in south-west Morocco, especially as an undescribed form of it is
certainly known from the Lower Cretaceous of the High Plateaux region of Algeria. In the Gentil
Collection are twenty-three specimens from Tinirt Ait Ameur (labelled Berrisian-Valanginian but
more likely Hauterivian) which appear to be this obese Algerian variety of the species. There is also
one specimen from the Hauterivian of Oued Tidzi, one from the Barremian of Ida ou Troumma, and
two from the Barremian of Tibourr’m; these resemble the more normal somewhat depressed
Neocomian form of the species.

Moutonithyris dutempleana (d’Orbigny). This almost ubiquitous Albian species is recorded by both
Roch and Ambroggi from both the Clansayesian and the Albian. Its occurrence in the Albian would
not be surprising. Doubts are raised, however, by two circumstances: (a) M. dutempleana is very rare
in the Clansayesian and known certainly from that stage only in Sardinia (Dieni et al. 1975). On the
other hand if, as is likely, the species spread from south to north, it could well occur in the
Clansayesian of Morocco. ( b ) Cyrtothyris middlemissi certainly occurs in both Clansayesian and
Albian and is easily mistaken for M. dutempleana (Calzada 1972, p. 66). The specimen figured by
Gigout (1951, pi. 13, figs. 5-8) as T. biplicata is a Concinnithyris cf. obesa.

To summarize: T. sueuri, T. collinaria, T. salevensis, T. carteroniana, and M. dutempleana have
probably been misidentified by previous authors. M. moutoniana and S. sella probably do occur
rarely in south-west Morocco.

Acknowledgements. I particularly thank Mr. E. F. Owen (British Museum, Natural History), Monsieur
D. Pajaud and his staff at the Universite Pierre et Marie Curie, and Monsieur J-P. Thieuloy (Grenoble). I thank
Mademoiselle S. Willefert (Service de la Carte Geologique du Maroc, Rabat) for her help in locating and sending
to me specimens from the Roch Collection; also the Head of the Laboratoire de Geologie-Paleontologie,
Universite Mohamed V, Rabat, and Monsieur G. Cogne for lending me specimens figured by Gigout.
Additional specimens were lent by: Professor D. V. Ager (Swansea), Senor S. Calzada Badia (Barcelona),
Monsieur Y. Champetier (Nancy), Monsieur A. Charriere (Paris), Monsieur M. Debuyser (Paris), Professor
M. Durand Delga (Toulouse), Dr. W. J. Kennedy (Oxford), Monsieur E. Lanterno (Geneva), Monsieur
Y. Rangheard (Besan9on), Dr. J. Remane (Neuchatel), and Monsieur Weidmann (Lausanne).

REFERENCES

ager, d. v. 1974. The western High Atlas of Morocco and their significance in the history of the North Atlantic.
Proc. geol. Ass., Lond. 85, 23-41, London.

— and evamy, b. D. 1964. The geology of the southern French Jura. Ibid. 74 (for 1963), 325-355, pi. 9,
London.

ambroggi, r. 1963. Etude geologique du versant meridionel du Haut Atlas occidental et de la Plaine du Souss.
Notes et Mem. Serv. geol. Maroc, 157, 321 pp., 181 figs., Rabat.

MIDDLEMISS: CRETACEOUS TEREBRATULIDAE

555

arkell, w. J. 1956. Jurassic Geology of the World , Edinburgh.

barczyk, w. 1969. Upper Jurassic terebratulids from the Mesozoic border of the Holy Cross Mountains in
Poland. Pr. Muz. Ziemi, 14, 1-82, pis. 1-18, Warsaw.

bogdanova, t. n. and lobacheva, s. v. 1966. Neocomian Fauna of the Kopet-Daga. Min. Geol. U.S.S.R., Inst.

Econ. Sci. Leningrad, n.s. 130, 1-140, pis. 1-13, Leningrad. [In Russian.]
bonneau, m., beauvais, l. and middlemiss, f. a. 1975. L’unite de Miamou (Crete-Grece) et sa macrofaune d’age
Jurassique superieur (Brachiopodes, Madreporaires). Ann. Soc. geol. Nord, 94 (for 1974), 71-85, pis. 10-11,
Lille.

boullier, A. 1976. Les terebratulides de VOxfordien du Jura et de la bordure sud du Bassin de Paris. Thesis,
Besangon.

buckman, s. s. 1918. The Brachiopods of the Namyau Beds, northern Shan States, Burma. Pal. Indica, n.s. 3
(for 1917), Mem. no. 2, 1-254, pis. 1-21, Calcutta.

busnardo, R. and durand-delga, m. 1960. Donnees nouvelles sur le Jurassique et le Cretace inferieur
dans Test des Cordilleres Betiques (Regions d’Alcoy et d’ Alicante). Bull. Soc. geol. Fr. (7), 2, 278-287,
Paris.

calzada, s. 1972. Cyrtothyris middlemissi, n. sp. del Aptiense de Garraf (Barcelona). Acta geol. Hisp. 7, 66-68,
2 figs., Barcelona.

choubert, g., faure-muret, a. and hottinger, l. 1967. Apergu geologique du bassin cotier de Tarfaya. Notes
Mem. Serv. geol. Maroc, 175 (1), 7-106, Rabat.

cox, Margaret m. and middlemiss, f. a. 1978. Terebratulacea from the Cretaceous Shenley Limestone.

Palaeontology, 21, 411-441, pis. 40-42, figs. 1-13, London.
d’archiac, a. 1847. Rapport sur les fossiles du Tourtia. Mem. Soc. geol. Fr., (2) 2, 291-351, pis. 13-25, Paris.
dieni, I., middlemiss, f. a. and owen, e. f. 1975. The Lower Cretaceous Brachiopods of east-central Sardinia.

Bol. Soc. Pal. Ital. 12 (for 1973), 166-216, pis. 32-38, Modena.
geyssant, jeannine. 1966. Glossothyris et Pygope (Terebratulidae)— essai de repartition de ces especes dans la
domaine mediterraneen. Notes Serv. geol. Maroc, 26, 75-98, pis. 1-3, figs. 1-7, 8 tables, Rabat.
gignoux, m. 1955. Stratigraphic Geology (English edn.), San Francisco.

gigout, m. 1951. Etudes geologiques sur la Meseta marocaine occidentale (arriere-pays de Casablanca,
Mazagan et Safi). Notes Mem. Serv. geol. Maroc, 86, 1-507, pis. 1-18, Rabat.
imlay, R. w. 1937. Lower Neocomian fossils from the Miquihuana region, Mexico. J. Palaeont. 11, 552-574,
pis. 70-83, Menasha.

— 1940. Neocomian faunas of northern Mexico. Bull. G. S. Amer. 51, 117-190, pis. 1-21, New York.
kyansep, n. p. 1961 . Terebratulids of the Lusitanian Beds of the Lower Kimmeridgian of the south-west Crimea.

Akad. Nauk. U.S.S.R. 8, 1-101, 8 pis., Moscow. [In Russian.]
loriol, p. de. 1866. Description de fossiles de V oolite corallienne, de Vetage valangien et de Vetage urgonien du
Mont Saleve, Geneva.

1867. In favre, A. Recherches geologiques dans la Savoie, Paris and Geneva.

— 1868. Monographie des couches de l’etage valangien des Carrieres d’Arzier (Vaud). Mater, pour Paleont.
suisse, ser. 4, Geneva.

— and gillieron, v. 1869. Monographie paleontologique et stratigraphique de l’etage urgonien inferieur du
Landeron (Neuchatel). Mem. Soc. helv. Sc. nat. 23 (5), 1-123, pis. 1-18, Zurich.

middlemiss, F. A. 1959. English Aptian Terebratulidae. Palaeontology, 2, 94-142, pis. 15-18, London.

1968a. Brachiopodes du Cretace inferieur des Corbieres orientales (Aude). Ann. Paleont. (Invert.), 54,

173-197, pis. A-C, Paris.

1968 b. Observations on the ontogeny of the brachiopod Sellithyris sella. Bull. Ind. geol. Ass. 1, 1-17, pi. 1,

Chandigarh.

1973. The geographical distribution of Lower Cretaceous Terebratulacea in western Europe. In casey, r.

and rawson, p. f. (eds.). The boreal Lower Cretaceous: Geol. J. Spec. Issue no. 5, 1 10-129, Liverpool.

— 1976. Lower Cretaceous Terebratulidina of northern England and Germany and their geological back-
ground. Geol. Jb. A30, 21-104, pis. 1-11, Hanover.

— 1979. Boreal and Tethyan brachiopods in the European early and middle Cretaceous. Kreide Europas
IUGS ser. A.

muir-wood, h. M. 1965. In moore, R. c. (ed.). Treatise on Invertebrate Paleontology, pt. H, New York.

Pictet, f-j. and loriol, p. de. 1 872. Description des fossiles du terrain cretace des environs de Sainte-Croix,
pt. 5. Mater. pour la Paleont. suisse (6), Geneva.

ROCH, E. 1930. Etudes geologiques dans la region meridionale du Maroc occidentale. Notes et Mem. Serv. Mines
Maroc, 9, 1-542, pis. 1-26, Macon.

556

PALAEONTOLOGY, VOLUME 23

Smirnova, t. n. 1960. Brachiopoda. In drushchitz, v. v. and kudriavcheva, m. p. Atlas of the Lower Cretaceous
Fauna of the northern Caucasus and Crimea. Trudy, VNII Gaz., 370-387, 6 pis., Moscow. [In Russian.]

— 1969. A new terebratulid genus from the Tithonian-Valanginian. Pal. Zh., 3, 144-146, Moscow. [In
Russian.]

— 1972. Brachiopods from the Crimea and northern Caucasus. Akad. Nauk. U.S.S.R. 140 pp., 13 pis.,
Moscow. [In Russian.]

Typescript received 5 March 1979

Revised typescript received 15 November 1979

F. A. MIDDLEMISS

Department of Geology
Queen Mary College
Mile End Road
London

COLLIGNONICERATID AMMONITES FROM
THE MID-TURONIAN OF ENGLAND AND
NORTHERN FRANCE

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

Abstract. Collignoniceras Breistroffer, 1947 is represented by five species in the mid-Turonian of England and
Touraine (the type area of the Turonian stage) in northern France. The cosmopolitan and highly variable type
species C. woollgari (Mantell) is shown to be a senior synonym of C. schlueterianum (Laube and Bruder) and
C. mexicanum (Bose) amongst others, and shows features indicating that Selwynoceras Warren and Stelck, 1940
(the type species of which S. boreale (Warren), is also redescribed) is a synonym of Collignoniceras sensu stricto.
Other species referred to the genus are C. carolinum (d’Orbigny), C. papale (d’Orbigny), C. canthus (Sornay) and
C. turoniense (Sornay). Ammonites fleuriausianus d’Orbigny, 1841 is a senior synonym of A. vielbancii d’Orbigny,
1850 and is made the type species of Lecointriceras gen. nov., to which two further species, L. carinatum sp. nov.
and L. costatum sp. nov. are also referred.

Collignoniceras woollgari (Mantell) is one of the most widely cited mid-Cretaceous ammonite species,
giving its name to the middle zone of the Turonian standard sequence (Wright in Arkell et al. 1957;
Rawson et al. 1978). As with other classic species, the type material has never been adequately figured
and is of uncertain horizon, although it has at least survived the vicissitudes of a century and a half
since its original description (Mantell 1822, p. 197; pi. 21, fig. 16; pi. 22, fig. 7). In England, where it
was first described, the species is rare and the lectotype remains the only good adult specimen known.
Elsewhere, however, it is recorded abundantly, especially in the U.S. Western Interior region, where
it formed the basis of one of the early accounts of intraspecific variability in Cretaceous ammonites
(Haas 1946), although as Haas and Meek before him (1876, p. 455) noted, authors have questioned
whether the great majority of specimens referred to this cosmopolitan species are indeed conspecific
with Mantell’s types.

We have studied hundreds of European, American and Japanese Collignoniceras in connection
with this project, and encountered an initially bewildering range of variation, both in adult ornament
and the size at which ontogenetic changes occur. We have relatively few juveniles from Europe but
many from the U.S. A.; conversely, large complete adults are common in European collections, but
those from the U.S. are usually fragmentary. Whilst it would be possible to select individuals with
differences that could be framed into diagnostic features for specific or subspecific separation, this
would be misleading and conceal the over-all common features of the species recognized below. In
C. woollgari in particular we have no doubt that a series of local races of the species existed over its
wide spread, but to separate formally the successive or local populations, differing in the extent of
morphological variation but overlapping, would serve no useful purpose. The broad, variable species
described below not only represent reality but are adequate for detailed correlation and discussion of
the evolution of the genus.

SYSTEMATIC DESCRIPTIONS

Location of specimens. The following abbreviations are used to indicate the repositories of specimens studied:
AM Museum de Paleontologie d’Angers.

BMNH British Museum (Natural History), London.

CS Chateau de Saumur

EMP Ecole des Mines, Paris (now housed at the Universite Claude Bernard, Lyon).

IPalaeontology, Vol. 23, Part 3, 1980, pp. 557-603, pis. 62-77.|

558

PALAEONTOLOGY, VOLUME 23

FSM Faculte des Sciences, Le Mans; chiefly collections formerly housed in the Musee de Tesse, Le Mans.
FSR Institut de Geologie, Universite de Rennes.

GK Department of Geology, Kyushu University, Fukuoka.

MNHP Museum National d’Histoire Naturelle, Paris.

OUM University Museum, Oxford; unless stated otherwise, these are collections made by Hancock and
Kennedy.

SP Collections of the Sorbonne, now Universite de Paris VI.

WW C. W. and E. V. Wright collection.

Dimensions. All dimensions are given in millimetres; figures in parentheses are the dimensions as a percentage of
the total diameter. D = diameter; Wb = whorl breadth; Wh = whorl height; U = umbilicus; Ic = intercostal;
c = costal; R = number of ribs per whorl.

Suture terminology. The suture terminology of Wedekind (1916; see Kullman and Wiedmann 1970 for a recent
review) is followed here: I = Internal lobe, U = Umbilical lobe, L = Lateral lobe, E = External lobe.

Suborder ammonitina Hyatt, 1889
Superfamily acanthocerataceae de Grossouvre, 1 894
[nom transl. et correct. Hyatt 1900, ex Acanthoceratides de Grossouvre, 1894]

Family collignoniceratidae Wright and Wright, 1951
Subfamily collignoniceratinae Wright and Wright, 1951
Genus collignoniceras Breistroffer, 1947
{non Van Hoepen, 1955)

Type species. Ammonites woollgari Mantell, 1822 by the original designation of Meek (1876) as type species of
Prionotropis Meek, 1876 {non Fieber, 1853), for which Breistroffer (1947) proposed Collignoniceras as nomen
novum.

Diagnosis. Medium to large, moderately involute to evolute ammonites. Early whorls compressed,
parallel sided, ornamented by crowded or sparse, prorsiradiate, straight or flexuous ribs, mostly long,
with weak to strong umbilical bullae. All ribs bear in the early stages outer ventrolateral tubercles in
addition to siphonal clavi.

This style of ornament is, in some species, retained to maturity. In most, however, the ribs coarsen,
become widely spaced, with strong to weak umbilical tubercles (which migrate progressively
outwards from the umbilical margin), prominent inner and outer ventrolateral tubercles which may
fuse into a massive horn or flared rib, from which commonly arise pairs of low ribs, joining siphonal
clavi more numerous than the ventrolateral and linked into a more or less continuous keel. Rarely the
ornament is greatly reduced on the body whorl.

The sutures are little incised, with massive saddles.

Discussion. The diagnosis given above summarizes the rather wide variation seen in species referred
to this genus, which include C. boreale (Warren), C. papale (d’Orbigny), C. canthus (Sornay),
C. turoniense (Sornay) and C. carolinum (d’Orbigny). The nomenclatorial history of the genus is
somewhat complex. Meek introduced a subgenus Prionotropis in 1876, with Ammonites woollgari
Mantell as type species. Breistroffer (1947) pointed out the prior usage of Prionotropis by Fieber
(1853) and proposed Collignoniceras as nomen novum. Meanwhile Warren and Stelck (1940) had
proposed the genus Selwynoceras with P. borealis Warren, 1930 as type species, distinguishing it from
Meek’s Prionotropis by the presence of a row of nodes instead of a keel on the inner whorls and the
marked alternation in length and strength of the ribs. Wright (in Arkell et al. 1957, p. L426) regarded
Selwynoceras as a subgenus of Collignoniceras , whilst Powell (1963, p. 1223) considered the two
synonymous. Following an application by Matsumoto and Wright in 1966, the International
Commission on Zoological Nomenclature ruled in 1968 (Opinion 861) that Collignoniceras
Breistroffer, 1947, should be given priority over Selwynoceras Warren and Stelck, 1940, by those
who regard the two as synonyms.

KENNEDY ET AL.. COLLIGNONICER ATI D AMMONITES

559

From a comparison of the types and other specimens of C. woollgari and S. bore ale, we would agree
with Powell that the two do not bear even subgeneric separation: boreale is simply a small species of
Collignoniceras in which the flared ribs appear at a relatively early stage. The ventral tuberculation
visible on the outer whorl of the lectotype (here designated), which is refigured here as PI. 70, figs. 1 -2,
is on exactly the same plan as in English woollgari, whilst, as Haas (1946), Powell (1963) and
Matsumoto (1965) have shown, the style of ribbing of juvenile Collignoniceras is very variable.

Collignoniceras differs from Prionocyclus Meek, 1876 (type species P. wyomingensis Meek) in that
the latter has very fine dense irregular ribs through most or all of its ontogeny and a broader venter
with an entire or serrated keel. C. woollgari and P. hyatti (Stanton) overlap in time in the southern
U.S. and some late C. woollgari there and also in Europe show a low siphonal keel at maturity,
emphasizing the intimate relationship between the two. Ribbing is usually dominant over
tuberculation in Prionocyclus, although some species bear finger-like ventrolateral spines, fore-
shadowing the development seen in the later Prionocycloceras (Young 1963, pi. 23, figs. 1-6; pi. 27,
figs. 2-4). Matsumoto (1965, p. 19) discusses other differences between these two genera.

Subprionocyclus Shimizu, 1932 was originally separated from Collignoniceras [Prionotropis] on the
basis of minor differences between the internal sutures. As Matsumoto (1959, p. 109) notes, however,
distinguishing features also include the paired or alternately long and short sigmoidal ribs of
Subprionocyclus which may flatten on the outer whorl, greater persistence of outer ventrolateral
tubercles and absence of massive horns. Like Prionocyclus, Subprionocyclus has a continuous
persistent keel which varies with the density of the ribbing from finely to coarsely serrate.

Germariceras Breistroffer, 1947 is perhaps only doubtfully separable from Prionocyclus', known
only from juveniles, it may be separated from Collignoniceras by the possession of fine dense narrow
ribs with small sharp umbilical, inner and outer ventrolateral tubercles and a finely serrated
continuous keel with more serrations than the number of ventrolateral tubercles.

Reesidites Wright and Matsumoto, 1964, which should perhaps be placed in Barroisiceratinae, is
compressed and involute, high whorled, with a fastigiate venter; sinuous ribs branch in groups of two
or three from small umbilical bullae, with single ventrolateral and siphonal clavi only. The largest
individuals barely exceed 100 mm diameter (Matsumoto 1965).

Subprionotropis Basse, 1950, known only from specimens a few centimetres in diameter, differs
from Collignoniceras in being involute with compressed whorls, with ribs arising in pairs from
umbilical bullae (with additional intercalated ribs) bearing only ventrolateral and siphonal clavi and
forming strong chevrons on the fastigiate venter. At the end of the body chamber, ribs and tubercles
weaken and the venter becomes rounded.

Lymaniceras Matsumoto, 1965 and Niceforoceras Basse, 1948 are both compressed and involute,
with weak, dense flexuous ribs or striae, a single ventrolateral tubercle and a finely serrated keel.

Collignoniceras is the earliest genus of Collignoniceratidae to appear in the Turonian, and, as
Matsumoto (1965) has noted, some individuals in variable United States Western Interior
populations show early whorls which foreshadow Prionocyclus, Subprionocyclus and thence the
remaining late members of the group.

With respect to the evolutionary origins of the genus, Wright (in Arkell et al. 1957, p. L426) and
Matsumoto (1965, p. 10) have suggested that the diminutive late Cenomanian acanthoceratinid
Protacanthoceras Spath, 1923 might be the ultimate ancestor, with Neocar diocer as Spath, 1926 as an
intermediate. Recent collecting from the latest Cenomanian/early Turonian faunas of the condensed
Neocar diocer as Pebble Bed of Devon (see Hancock, Kennedy and Wright 1977, fig. 2 for details) has
now produced a range of specimens that provisionally we refer to Thomelites Wright and Kennedy,
1973, among which are individuals with siphonal clavi tending to form a continuous serrated keel. In
addition, a few poorly preserved fragments seem already to have reached the stage of Collignoniceras
in some features of decoration. There remains, however, a gap in the European successions,
corresponding to most of the Mammites nodosoides assemblage Zone, in which the genus is absent
apart from a single possible example in the collection of Colonel O. H. Bayliss, from Shapwick,
Devon; W. A. Cobban (in litt., 1978) tells us that Collignoniceras appears at the top of the North
American correlatives of this zone.

560

PALAEONTOLOGY, VOLUME 23

Occurrence. Collignoniceras is widespread in the middle of the Turonian stage, the classic woollgari Zone. There
are records from England, France, Germany, Czechoslovakia, Poland, Rumania, Turkestan, Japan, California,
Texas, the U.S. and Canadian Interiors, Greenland, north Africa, Colombia, and northern Australia.

Collignoniceras woollgari (Mantell)

Plates 62-67 ; Plate 69, figs. 3-4; Plate 71, figs. 1-3; text-figs. 1a, 2-4

1822
non 1841
1850
1855
1860
1867
1872
1872
1872

1887

1887

1902

1902

1902

1907

1925

1928

1931

1931

1931

1946

1963

1971

1972

1975

1977

1977

Ammonites Woollgari Mantell, p. 197, pi. 21, fig. 16; pi. 22, fig. 7.

Ammonites Woollgari Mantell; d’Orbigny, p. 352, pi. 108, figs. 1-3.

Ammonites Woolgarii d’Orbigny, p. 189 (pars).

Ammonites Woollgari Mantell; Sharpe, p. 27, pi. 11, figs. 1, 2.

Ammonites carolinus (d’Orbigny); Courtiller, p. 251, pi. 3, fig. 2.

Ammonites Woolgarii Mantell; Courtiller, p. 7, pi. 8, figs. 1-4.

Ammonites Woolgari Mantell; Schliiter, p. 25, pi. 9, figs. 1-5; non pi. 12, figs. 5, 6.
Ammonites Woollgari Mantell; Geinitz, p. 184, pi. 33, figs. 1, 2 (?), non 4-5.

Ammonites Woolgari Fritsch, p. 30 (pars), pi. 3, figs. 1-3; pi. 4, figs. 1-2; pi. 14, fig. 6; non pi. 2,
tigs. 1-2; pi. 15, fig. 6.

Acanthoceras Woollgari (Mantell); Laube and Bruder, p. 235, text-fig.

Acanthoceras Schliiterianum Laube and Bruder, p. 236, pi. 29, figs. 2-3.

Acanthoceras Woollgari (Mantell); Petrascheck, p. 149, text-figs. 7-8.

Acanthoceras cfr. Woollgari (Mantell); Petrascheck, p. 148, pi. 12, figs. 2-3.

Acanthoceras Schliiterianum Laube and Bruder; Petrascheck, p. 150, pi. 10, fig. 3; pi. 11, fig. 3;
pi. 12, fig. 1.

Prionotropis Schliiterianum Laube and Bruder; Pervinquiere, p. 275.

Prionotropis Schliiteriana Laube and Bruder; Diener, p. 156.

Prionotropis woollgari Mantell var. mexicana Bose, p. 262, pi. 1 1, figs. 11, 12.
Pseudaspidoceras(l) chispaense Adkins, p. 51, pi. 3, figs. 1-2.

Pseudaspidocerasl sp. Adkins, p. 53, pi. 2, fig. 2.

Pseudaspidoceras(l) n.sp. A; Adkins, p. 53, pi. 3, figs. 3-4.

Prionotropis woollgari Meek (? non Mantell); Haas, p. 150, pis. 11, 12; pi. 13, figs. 1-3, 5-18;
pi. 14, figs. 1-10, 12-16; pi. 15, figs. 1-6, 9, 10; pis. 16, 17; pi. 18, figs. 1-2, 7-9; text-figs. 1-91.
Selwynoceras mexicanum (Bose); Powell, p. 1225, pi. 166, figs. 2-7; pi. 167, figs. 1, 3-8; pi. 168,
fig. 4; text-figs. 2-4.

Collignoniceras woollgari (Mantell); Matsumoto, p. 130, pi. 21, fig. 4, text-fig. 1.
Collignoniceras woollgari (Mantell); Cobban and Scott, p. 94, pi. 14, fig. 5; pi. 30, fig. 1; pi. 37,
figs. 9-10 (with additional synonymy).

Collignoniceras woollgari (Mantell); Hattin, pi. 10, figs. N, P, Q, R.

Collignoniceras (Selwynoceras) schlueterianum (Laube and Bruder); Hancock, Kennedy and
Wright, p. 156.

Collignoniceras (Collignoniceras) cf. C. woollgari sensu Matsumoto, 1965, group E; Hancock,
Kennedy and Wright, p. 156.

Types. The lectotype, designated by Wright and Wright (1951, p. 35), is BMNH 5682, from the Middle Chalk of
Lewes, Sussex, refigured here as Plate 62, figs. 1-2; Plate 63, fig. 9. Two additional specimens from Mantell’s
collection, BMNH C5742 a-b (Plate 69, figs. 3, 4), are presumed to be paralectotypes.

Other specimens studied. These include: BMNH 4863 a-b, from the Middle Chalk ‘near Lewes, Sussex’; BMNH
43963 ‘Lower Chalk, near Lewes’ (J. de C. Sowerby Collection); BMNH C30394 ‘Turonian Mount Caburn Pit,
near Glynde, Sussex’ (labelled aff. woollgari by L. F. Spath); BMNH C40152 from the Middle Chalk,
Terebratulina lata Zone, Mickleham Bypass, Surrey (C. W. and E. V. Wright Collection); WW 16682, 14792-4,
from the Middle Chalk, top of the T. lata Zone Middle Chalk, Mickleham Bypass, Surrey; WW 22925-7, Middle

EXPLANATION OF PLATE 62

Figs. 1 -2. Collignoniceras woollgari (Mantell). The lectotype, BMNH 5682, from the Middle Chalk of Lewes,
Sussex.

PLATE 62

KENNEDY, wright and Hancock, Collignoniceratid ammonites

562

PALAEONTOLOGY, VOLUME 23

Chalk, Lewknor Crossroads, Lewknor, Oxon. (ex R. E. H. Reid Collection); OUM K 10273, K 10275-76 from
no more than 5 m below the top of the Chalk Rock at Fognam Barn, Berkshire, 3 km WNW of Lambourn.

BMNH 88988 b, 88989 a-c from the Turonian of the White Mountain, near Prague, Czechoslovakia.

French specimens include the following: OUM KZ 741, 743-4, 746, 748-9, 753, from the St. Cyr-en-Bourg
Fossil Bed, Champignonniere Les Rochains, 7 km south of Saumur and north-east of Montreuil-Bellay, Maine-
et-Loire, and numerous specimens in the Museum de Paleontologie, Angers, from this bed and adjacent levels in
the Tuffeau Blanc (Couffon Collection etc.) variously labelled Saumoussay, St. Cyr-en-Bourg, Saumur, and
elsewhere, including AM 57, AM 59, AM 116.

There are numerous specimens from Ponce, Sarthe, and others from Bourre in the Cher Valley, Loir-et-Cher,
including BMNH C74803.

Dimensions

D

Wb

Wh

Wb: Wh

U

Ribs

Lectotype

130(100)

40 (31)

40 (31)

1

50 (38)

13

FSR, C273

67-3 (100)

20-4 (30)

25-0 (37)

0-81

23-9 (36)

24

MNHP W7

58-5 (100)

21-0 (36)

-(-)

—

22-8 (39)

19

MNHPW18

61-0(100)

24-0 (39)

23-7 (39)

101

22-0 (36)

22

MNHP X’

86-0 (100)

29-0 (34)

34-9 (41)

0-83

29-5 (34)

~20

MNHPW15

81-0(100)

32-0 (40)

32-0 (40)

1-0

27-8 (34)

19

MNHPIc

141-0(100)

52-0 (37)

49-5 (35)

1-05

55 (39)

—

Ic

MNHP 6778
MNHPW20

133-0(100)

41-5(29)
45-0 (34)

49- 5 (35)

50- 0 (38)

0-83

0-9

44-5 (33)

18

Ic

162-0(100)

55-0 (34)

52-0 (32)

1-05

67 (41)

—

MNHP W4 c

137-0(100)

60-0 (43)

51-0(37)

1-18

- (-)

15/16

Ic

39-5 (29)

39-5 (29)

1-0

MNHP W10

109.0(100)

47-5 (44)

42-0 (39)

M3

37-0 (34)

—

33-5(31)

38-8 (36)

0-86

MNHPW19

175-0(100)

74-8 (43)

65-0 (37)

1-15

65-0 (37)

15

Description. The inner whorls of our smallest specimens show coiling to have been moderately evolute, with
compressed whorls and a shallow umbilicus. At about 10-15 mm diameter, there are 27-32 ribs per whorl; the
density decreases with increasing size. The ribs are even, bar-like, prorsiradiate, straight and clearly demarcated
from the flat interspaces. As size increases, ribs become much more widely spaced; at 40-50 mm diameter there
are only 17-24 ribs per whorl. They are of variable strength, arise from weak to strong umbilical bullae and are
narrow, high and separated by wide, flat interspaces; they are markedly prorsiradiate and straight to concave on
the flanks, always single, with no intercalated ribs. At the ventrolateral shoulder they bear conical to feebly
clavate inner ventrolateral tubercles. From these the ribs are either weakly or strongly projected forwards to
elongate outer ventrolateral clavi. A broadened swelling connects these in turn to a sharp, continuous siphonal
keel, strengthened into sharp high clavi at the peak of the variably angled ventral chevron formed by the
termination of the ribs.

This type of ornament may extend to diameters of 100 mm, but typically, as size increases, a series of changes
in ornament occur, more or less independently of each other. The umbilical bullae move outwards and come to
occupy a lower flank position, whilst the ribs are differentiated into long bullate ones and (in most specimens)
from one to four shorter ribs, restricted to the outer flank and venter and sometimes lacking ventrolateral
tubercles. The inner ventrolateral tubercles may at this stage develop into a distinctive conical horn which
supports, on the outer flank of its base, the outer ventrolateral clavus; some specimens present a ventral aspect in

EXPLANATION OF PLATE 63

Figs. 1-12. Collignoniceras woo//gan'(Mantell). 1-4, OUM KZ 746; 11-12, OUM KZ 748, from the St. Cyr-en-
Bourg Fossil Bed, Champignonniere les Rochains, south of Saumur and north-east of Montreuil-Bellay,
Maine-et-Loire. 5-6, MNHP 6778 (d’Orbigny Collection), Ponce, Sarthe; 7-8, OUM KT 1160, from the
Ojinaga Formation at Cannonball Hill, northern Chihuahua, Mexico. 9, Apertural view of the lectotype,
BMNH 5682; see explanation of Plate 62 for details. 10, MNHP Wl, ‘Le Mans, Sarthe’ (from Ponce?).
Figures 1 -2 are x 2; the remainder are x 1 .

PLATE 63

Kennedy, wright and Hancock, Collignoniceratid ammonites

f-ff-

564

PALAEONTOLOGY, VOLUME 23

text-fig. 1. Sutures of Collignoniceras species. A, C. woollgari (Mantell), from BMNH C74803; B, C. carolinum
(d’Orbigny), from the Sorbonne specimen (de Grossouvre Collection); c, C. papale (d’Orbigny), from a
Sorbonne specimen (de Grossouvre Collection). Bar scale is 2 cm.

which siphonal tubercles greatly outnumber ventrolateral, whilst others show a more or less equal number; no
two specimens agree in details of ornament.

Mature specimens show two broad types of decoration, but again no two specimens agree in detail. In the first
group the umbilical bullae move outwards and fuse with the inner ventrolateral tubercles to form a strong to
massive horn (if broad) or flange (if narrow). This supports a long, low, narrow outer ventrolateral clavus, and
the front and rear of the horn strengthens into a pair of ribs which loop to the pair of siphonal clavi
corresponding to each horn. Some specimens may develop a low siphonal horn at this stage and at the adult
aperture up to three ventral ribs may appear between the primary ribs, although in other specimens these may be
absent, the spaces between the major ribs being smooth. The second type is a more evolute form, retaining long,
straight, distant flank ribs with bullae of variable strength, connected by weak or almost effaced ribs to strong
conical ventrolateral horns which bear the outer ventrolateral clavus. A low siphonal ridge is present and there
are pairs of clavi corresponding to the horns as well as additional clavi in the interspaces. This form differs most
obviously from the first in the retention of bullae and in being somewhat larger.

The suture line is simple, with a massive, slightly incised, asymmetrically bifid E/L, narrow L and narrow, bifid
L/U2.

EXPLANATION OF PLATE 64

Figs. 1-3. Collignoniceras woollgari (Mantell). The lectotype of Acanthoceras schlueterianum (Laube and
Bruder), from the Turonian of the White Mountain near Prague, Czechoslovakia. Pictures supplied through
the courtesy of Dr. V. Housa (Prague).

PLATE 64

Kennedy, wright and Hancock, Collignoniceratid ammonites

566

PALAEONTOLOGY, VOLUME 23

Discussion. The above description is based upon the available English material, the large suite of
specimens from Touraine and a few Czechoslovakian specimens before us. It must be stressed that no
two specimens are alike and that description is inevitably generalized. Mantell’s original figures of
Ammonites woollgari give a clear and accurate representation of the juvenile form, but only suggest
the very different adult form in general terms, better shown in Sowerby’s (1828, p. 165; pi. 587, fig. 1)
beautiful watercolour and Sharpe’s (1855, p. 27; pi. 11, figs, la-b) slightly inaccurate reconstruction.

The lectotype is, in fact, a moderately distorted composite internal mould only 130 mm in
diameter, as can be seen from our photographs (PI. 62, figs. 1-2; PI. 63, fig. 9), showing no trace of
sutures or any indications of how much is body chamber. In terms of the description given above, it
falls into the first group of specimens. It is distinctive in the small size at which the massive horns are
developed and the brevity of the stage with intercalated ribs.

text-fig. 2. Collignoniceras woollgari (Mantell) BMNH 88989a, a crushed specimen
from the Turonian of the White Mountain, near Prague, Czechoslovakia.

text-fig. 3. Collignoniceras woollgari (Mantell) a, b, MNHP W14, 6778 (d’Orbigny Collection), from Ponce,
Sarthe. a. tuffeau specimen agreeing closely with the type. Reduced x 0-5 approx, c. d. MNHP 1946-19, from St.
Maure de Touraine. A hypernodose adult of the first type. Reduced x 0-4 approx.

568

PALAEONTOLOGY, VOLUME 23

At the beginning of the outer whorl the ribs bear strong umbilical bullae, strong conical inner
ventrolateral and long, low, clavate outer ventrolateral tubercles and a strong elongate siphonal
clavus. Between these long primary ribs are one or two shorter intercalated ribs which extend across
the venter and bear small siphonal clavi. By 90 mm diameter these are lost and the ornament consists
of an umbilical bulla which moves out progressively to occupy a mid-lateral position, linked by a
broad rib to a massive inner ventrolateral horn which bears, at its base, the outer ventrolateral clavus.
From this clavus two poorly defined, low, rounded ribs link to two ventral clavi.

The best-preserved horn on the lectotype is at 120 mm diameter, and here the bulla on the flank and
the inner ventrolateral horn have merged into a massive horn bearing a much weakened outer
ventrolateral clavus and subdued weakened ribs.

D’Orbigny (1841, p. 352, pi. 108, figs. 1-3) figured under the name A. woollgari a distinctive form
which he subsequently (1850) named A. vielbancii; it is redescribed below as a junior subjective
synonym of Lecointriceras fleuriausianum (d’Orbigny). D’Orbigny also described in Paleontologie
Franqaise a related form, A. carolinus (1841, p. 310, pi. 91, figs. 5-6), which he subsequently (1850)
regarded as a synonym of A. woollgari , a view followed by most later authors. Sharpe (1855, p. 27,
pi. 11, figs, la-b, 2a-b) clearly recognized the differences between young woollgari and carolinum
(\ • ■ the French shell has twice as many ribs, is less compressed, and has the keel more completely
separated from the ribs by two regular channels, than in our species’), and, as we describe below, the
two are indeed specifically distinct.

Fritsch (1872) provided a very clear discussion of Mantell’s species, and recognized three variants;
his descriptions are loosely translated as follows:

(a) Typical form, which agrees exactly with the illustrations of Mantell and Sharpe. It has very
strong tubercles on the siphonal side (pi. 4, figs. 1, 2).

(b) Form with slender ribs and weaker tubercles (pi. 3, fig. 2).

( c ) More involute form with an inverse egg-shaped mouth opening. There are tubercles close to the
umbilical seam, which remain there for a long period, and are stronger and more widely separated
than in the typical form; there are only six, even on the inner whorl (pi. 3, fig. 1).

He also described a variety lupulina from Mecholup [Michelob] near Saatz, close to Prague (1872,
p. 31, pi. 2, figs. 1, 2; pi. 15, fig. 6), which was said to be very similar to woollgari when young, but
when old, has a different venter, large sparse tubercles and an almost square cross-section. It is,
in fact, a Mammites nodosoides (Schliiter).

Schliiter ( 1 872) figured a similar range of variants; his pi. 9, figs. 1 -3 correspond to Fritsch’s form c
and his pi. 9, figs. 4-5 to the typical form. His variety (pi. 12, figs. 5-6) is, as he suggested, close to the
papale group in many respects and it could well be referred to as Collignoniceras aflf. canthus (Sornay).

Laube and Bruder (1887) reviewed a similar range of central European specimens but referred
Fritsch’s typical form (var. a) to a new species, Acanthoceras schlueterianum; they regarded the
involute form (var. c) as typical C. woollgari and var. lupulina as a Mammites, which they renamed
Mammites michelobensis. Petrascheck (1902) followed Laube and Bruder and described forms he
called woollgari, schlueterianum, and aff. woollgari.

From our study of the type material and the Touraine populations, it is quite clear that no two
adult Collignoniceras of these types are the same. The lectotype of C. woollgari, showing as it does an
early loss of umbilical bullae, which move out to mid flank, fuse into ventrolateral horns, with much
elongated outer ventrolateral clavi and subdued ribs looping to low siphonal clavi is clearly of the
same general morphology as Fritsch’s typical form (e.g. 1872, pi. 4, figs. 1-2) and the lectotype (here
designated) of Acanthoceras schlueterianum (Laube and Bruder 1887, pi. 29, figs. 2a-b) (PI. 64). It
differs, however, in showing a decline in ventral ribs and clavi at only 90 mm diameter, whereas the
Czechoslovakian examples retain umbilical bullae and intercalated ribs (particularly on the venter) to
a much greater size and in consequence have a longer middle growth stage with umbilical bullae,
conical inner ventrolateral and outer ventrolateral and siphonal clavi, like the specimen illustrated
here (text-fig. 4 c-d), Fritsch’s pi. 14, fig. 6 and Laube and Bruder’s smaller paralectotype (1887,
pi. 29, fig. 3). This stage is virtually suppressed in the lectotype of C. woollgari, which in these respects

text-fig. 4. Collignoniceras woollgari (Mantell) a, b. MNHP W22, 6778 (d’Orbigny Collection), from Ponce,
Sarthe. An adult of the second type, retaining long ribs and moderately evolute coiling. Reduced x 0-4 approx,
c, D. BMNH 88988b, from the Turonian of the White Mountain, near Prague. Reduced x0-5 approx.

570

PALAEONTOLOGY, VOLUME 23

is atypical. Other specimens show that the intercalated ventral ribs are accompanied by weak flank
ribs in middle growth but that there is great variation at this stage. The Touraine populations, which
yield specimens that both match the lectotype of C. woollgari (text-fig. 3 a-b) and show every
gradation to the other forms (PI. 66, figs. 1-3; text-fig. 3 c-d) with strong intercalated ribs and
tubercles, show that C. woollgari and C. schlueterianum should be treated as synonyms. Indeed, a
specimen from Fritsch’s own collection, now in the British Museum (Natural History) (no. 88989a)
and labelled in his own hand ‘Weisser Berg’, the type locality of C. schlueterianum (text-fig. 2),
exhibits the fusion of umbilical bullae with strong horns seen in the lectotype of woollgari but with
more persistent intercalated ribs on the venter of the last whorl. The specimen is, furthermore, adult
at only 150 mm, showing a rapid decline in ornament and loss of horns on the outer whorl.

In Germany (?), Czechoslovakia and Touraine (but not England where only one adult is known)
this hypernodose, horned form, enormously variable in its adult ornament, is accompanied by the
evolute, square-whorled forms which correspond to Fritsch’s form C, to Laube and Bruder’s
‘typical form’ and Petrascheck’s A. woollgari + A. cfr. woollgari. Inner whorls of this type are
inseparable from typical juvenile English C. woollgari, but again the variable adult whorls are quite
distinctive, as Fritsch described, and as outlined above in our description; we conclude that these are
probably sexual dimorphs.

C. woollgari var. mexicana (Bose) (1928, p. 262, pi. 11, figs. 11, 12) was originally described on the
basis of a single, crushed specimen from the Turonian Ojinaga Formation equivalent, near Jimenez,
Coahuila, Mexico, reillustrated here as Plate 65, figs. 1-3. Powell (1963) has redescribed this form (as
Selwynoceras mexicanum ) and discussed the intraspecific variation on the basis of large collections of
fragmentary material. From large additional collections from the same area (OUM KT 1 160-1183,
1200-1222, 1264-1313) and Chispa Summit, Jeff Davis County, Texas and specimens in the Adkins
Collection (preserved in the Texas Memorial Museum) we conclude that it too is a synonym of
C. woollgari. Juveniles, as Powell himself noted (op. cit., p. 1225), include individuals which cannot
be separated from the English C. woollgari (PI. 63, figs. 7-8), in addition to those which are more
compressed, finely and densely ribbed.

Powell (1963, pi. 168, fig. 4) has figured a specimen in middle growth, showing the irregularly
ribbed stage with development of inner ventrolateral horns as seen in Bohemian and Touraine
specimens and we have other slender fragments which match Petrascheck’s (1902, pi. 10, figs. 3a-b)
juvenile A. schliiterianum. Larger fragments show a wide range of variation, from robust fragments
having essentially equal numbers of inner and outer ventrolateral and siphonal tubercles to those
with multiple ventral tuberculation. Adult body chambers show clear dimorphism, as in European
material, the one form with flanges or flared horns produced by amalgamation of umbilical and
ventrolateral tubercles, the other more quadrate, retaining to maturity umbilical bullae and distant
ribs of variable strength. As can be seen from our and Powell’s figures, distinction on the basis of the
nature of the less complex suture, the finer ribbed juveniles and the coarse ornament of adults, by
which Powell separated it from C. schlueterianum, cannot be upheld in the light of the variation seen
in European specimens (not known to Powell); there is a clear overlap. We note the relatively
frequent occurrence of individuals with flares and a compressed whorl, rarely seen in Europe,
suggesting the Texas/Mexico material belongs to a local population more variable than their old
world contemporaries.

EXPLANATION OF PLATE 65

Figs. 1-8. Collignoniceras woollgari (Mantell). 1-3, the holotype of Prionotropis woollgari (Mantell) var.
mexicana Bose, from near Jimenez, Coahuila, Mexico. University of California, Berkeley, Collections. 4-6,
BMNH 4863a, from the Middle Chalk near Lewes, Sussex. 7-8, a juvenile U.S. Western Interior specimen in
the U.S. Geological Survey Collections, Denver, from USGS Mesozoic locality 21792, the mid-Turonian
Carlile Shale of the Black Hills.

PLATE 65

Kennedy, wright and Hancock, Collignoniceratid ammonites

572

PALAEONTOLOGY, VOLUME 23

The relationship of European specimens to the widely documented U.S. Western Interior material
referred to C. woollgari has been complicated by the relatively few illustrations of English juveniles.
Adults such as Meek’s specimen (1876, pi. 7, fig. lg) from the Black Hills, Dakota, would certainly
fall within the concept of C. woollgari outlined here, although differing from the lectotype most
obviously in the retention of umbilical bullae to a greater diameter. Dr. W. A. Cobban (Denver) has
also shown us medium-sized specimens in which all ribs are long and the ventrolateral and siphonal
clavi are equal in number, a feature uncommon in European material. American juveniles, described
by Haas (1946) and Matsumoto (1965) amongst others, show a much wider range of variation than
European material. This may be merely a consequence of the small number of juveniles known from
Europe: indeed, the latter fall closest to Matsumoto’s group E, one of the commonest forms in the
Western Interior. Nevertheless, there is a clear overlap with European C. woollgari. The presence of
similar individuals would also seem to preclude subspecific separation and we regard them as
conspecific, but with a different population structure. Specific differentiation of the American fauna
from their European contemporaries occurred later, with the evolution of the early members of the
Prionocyclus hyatti group.

W. A. Cobban (in lift.) has suggested to us that forms with more siphonal than ventral nodes pre-
date those in which the numbers are equal in the U.S. Western Interior, but, as we do not know the
precise horizon of the holotype of woollgari in relation to these, we prefer to unite them here, leaving
revision of these faunas to Dr. Cobban.

According to Matsumoto (1959, p. 107; 1965, p. 16, pi. 3, figs. 3-4) C. woollgari bakeri Anderson is
a subgroup of C. woollgari that characterizes the north Pacific region. All described specimens are
small, compressed, evolute Subprionocylus- like densely ribbed shells, close to subgroup D of
C. woollgari of Matsumoto (1965) from the U.S. western Interior, but more evolute and with less
prorsiradiate ribs. These differences probably do not merit separation, but without more and adult
specimens further comment is inadvisable.

C. woollgari is easily separated from the remaining species of the genus. C. carolinum (d’Orbigny)
(p. 574) is usually more densely ribbed and even in sparsely ribbed juveniles (PI. 68, fig. 1 1) the ribs are
low and subdued rather than bar-like. Adults are quite distinct; C. carolinum reaches maturity at only
100-120 mm diameter, never develops the coarse umbilical bullae, ribs, and horns of woollgari, nor
the complex looped ventral ornament. Instead, it remains compressed and flat sided, with weak ribs
and tubercles and a persistent, crenulated siphonal ridge. C. canthus (Sornay) (p. 582) has coarsely
and sparsely ribbed and tuberculate inner whorls but a virtually smooth body chamber with only
faint ribs and many tiny siphonal tubercles. C. turoniense (Sornay) (p. 584) has similarly coarsely
ornamented early whorls, is adult at a much smaller size with more massive whorls, coarse sparse
bullae, weak ribs and ventrolateral horns and the inner ventrolateral tubercles disappear at an early
stage.

There is a closer resemblance to C. papale (d’Orbigny) (p. 578) but here juveniles have fewer,
coarser ribs with strong bullae displaced well out from the umbilical shoulder, with much more
prominent inner ventrolateral tubercles. In middle growth C. papale lacks the prominent
ventrolateral horns of many C. woollgari and the inner and outer ventrolateral tubercles merge into a
pinched clavus, retained to much greater diameters in C. woollgari. Other differences are noted on
p. 582.

C. boreale (p. 586) is a genuinely small form, showing adult features at only 100 mm diameter in the
holotype. It has narrow, distant ribs and retains umbilical bullae to the end of the phragmocone,
showing early development of flared ventrolateral flanges and traces of looped ventral ribs.

EXPLANATION OF PLATE 66

Figs. 1-3. Collignoniceras woollgari (Mantell). Adult phragmocone showing intercalation of flank and ventral
ribbing, multiple ventral tuberculation and early stages of horn development. MNHP W10, from either
Ponce (Sarthe) or Bourre (Loir-et-Cher).

PLATE 66

Kennedy, wright and Hancock, Collignoniceratid ammonites

574

PALAEONTOLOGY, VOLUME 23

Occurrence. Few C. woollgari from England are well dated. Through the courtesy of the Director of the Institute
of Geological Sciences and Mr. C. J. Wood we have been able to examine the precisely positioned material from
the Leatherhead (Fetcham Mill), Surrey, Borehole (Gray 1965). Here C. cf. woollgari occurs at a depth of
570' 6” (GSM.WN 1979-80, 1982-3), 73' 1" (22-28 m) above the base of the Melbourn Rock and 17' 6" (5-33 m)
above a specimen of IMytiloides hercynicus ; at 535' 10" (GSM.WN 1942), 12' (3-66 m) above the level of large
Inoceramus of inequivalvis type, and at 518' 9" (GSM.WN 1900, 1901), 26' 9"(8-15 m) below specimens of
Mytiloides sp. and /. cf. apicalis (inoceramids determined by Mr. P. Woodroof). This range, through 51' 9"
(15-8 m) of section, includes the top of the Inoceramus labiatus/Orbirhynchia cuvieri and the lower part of the
Terebratulina lata Zones. Other English specimens have been recorded from both labiatus and lata Zones.
Specimens from Sussex, the type area, come mostly from the Lewes region. One specimen (BMNH C30394) is
said to be from Mount Caburn; unfortunately the classic pit here extends from the Melbourn Rock to basal
Upper Chalk {labiatus -planus Zones).

Specimens from the upper part of the lata Zone of Surrey (e.g. WW 14792-4, 16682), and OUM K 10273,
K 10275-6 from no more than 5 m below the top of the Chalk Rock at Fognam, Berkshire, indicate the upper
limit of its relatively long range. This is confirmed by occurrences in Sarthe and Touraine through the middle and
upper part of the Tuffeau Blanc, in the St. Cyr-en-Bourg Fossil Bed, Bourre and Ponce faunas. In the United
States the species occurs rarely in the top of Cobban and Scott’s (1972) Mammites nodosoides Zone (Cobban in
lilt.) and overlaps with the succeeding Prionocyclus hyatti (Powell, 1963).

Elsewhere the species is known to occur widely in Europe, the U.S.S.R. west to Transcaspia, Japan, California
and Oregon, Texas, Mexico, the U.S. Western Interior and northern Australia.

Collignoniceras carolinum (d’Orbigny)

Plate 68, figs. 1-11; Plate 76, figs. 1-2; text-figs. 1b, 5

1841 Ammonites Carolinus d’Orbigny, p. 310, pi. 91, figs. 5-6.

1850 Ammonites Woolgarii Mantell; d’Orbigny, p. 189 (pars).

1860 Ammonites Carolinus d’Orbigny; Pictet and Campiche, p. 316.

1872 Ammonites carolinus d’Orbigny; Schliiter, p. 27, pi. 9, fig. 6.

1881 Ammonites Carolinus d’Orbigny; Windmoller, p. 33.

71887 Acanthoceras Carolinum d’Orbigny; Laube and Bruder, p. 232, pi. 27, fig. 1.

1902 Prionotropis carolinus d’Orbigny; Petrascheck, p. 152.

71912 Prionotropis woolgari var. Carolinus (d’Orbigny); Arkhanguelsky, p. 72, pi. 3, figs. 20-22 ( fide
Arkhanguelsky, 1916).

1925 Prionotropis Carolina (d’Orbigny); Diener, p. 156 (pars).

1977 Collignoniceras (Collignoniceras) carolinum (d’Orbigny); Hancock, Kennedy and Wright,
p. 156.

Types. D’Orbigny’s original account of this species is as follows: ‘Je l’ai recueillie en place aux Martrous, pres
de Rochefort (Charente-Inferieure), dans la craie que je rapporte aux gres verts superieurs ou aux craies
chloritees. Elle y est rare a l’etat de moule. M. d’Archiac l’a aussi rencontree a Sainte-Maure (Indre-et-Loire),
dans le meme couche.’ By 1850 d’Orbigny had concluded that carolinus was a synonym of woollgari ( Prodrome ,
p. 189), and in consequence no specimens are represented in his collections under the name carolinus. Under
Ammonites woollgari, however, there is a specimen from Martrous with the label 6778a which is clearly the basis
of the original figure (PI. 68, figs. 4-8), and this is here designated lectotype of the species.

Other specimens studied. OUM KZ 747, from the St. Cyr-en-Bourg Fossil Bed, Champignonniere les Rochains,
7 km south of Saumur and north-east of Montreuil-Bellay, Maine-et-Loire. An unregistered specimen in de
Grossouvre’s collection (Sorbonne, Paris) from either Ponce (Sarthe) or Bourre (Loir et Cher). MNHP W8,
from an unknown locality in the Tuffeau. WW 14791 from the Terebratulina lata Zone, Mickleham Bypass,
Surrey.

EXPLANATION OF PLATE 67

Figs. 1-3. Collignoniceras woollgari (Mantell). Adult phragmocone of sparsely and robustly ribbed variant
with equal numbers of umbilical, ventrolateral and siphonal tubercles. MNHP W2. 1904-32. ‘Le Mans,
Sarthe’.

PLATE 67

Kennedy, wright and Hancock, Collignoniceratid ammonites

576

PALAEONTOLOGY, VOLUME 23

Dimensions

D

Wb

Wh

Wb: Wh U

Lectotype
MNHP 6778a
Sorbonne spec.

46-0(100) 14-0(30) 15-0(33) 0-93 16-7 (36)

108-5 (100) 28-2(26) 37-5(35) 0-75 34-3 (32)

Description. The lectotype from Martrous (Charente-Maritime) is a fragment with juvenile body chamber
preserved in calcarenite typical of the Calcaires a Cephalopodes of the Rochefort area. Coiling is relatively
evolute, with a broad, shallow umbilicus (36% of the diameter). The umbilical wall is low and rounded. The
whorl section is compressed (whorl breadth to height ratio is approximately 0-93), with flattened, convergent
sides, the maximum breadth close to the umbilical shoulder and the venter fastigiate. Ornament consists of
strong, dense, narrow ribs (nineteen on last half-whorl), arising at the umbilical shoulder without clear bullae
after the first two visible ribs. They are straight or slightly flexed and prorsiradiate on the inner flank, curving
strongly forwards across the ventrolateral shoulders and venter. Single, shorter intercalated ribs occur
commonly on the early part of the specimen but there are only two in the last half- whorl. The ribs are
strengthened into distinct if small inner ventrolateral tubercles at the beginning of the body chamber, but these
are lost beyond a diameter of about 34 mm. There are well-marked outer ventrolateral clavi, connected by
forwards-directed weak ribs to elongate siphonal clavi borne on a low, rounded keel. Other specimens show both
denser and sparser ribbing of the same style, as in other Collignoniceras juveniles (PI. 68, figs. 10, 1 1).

Body chambers show the species to have been adult at small diameters (100-120 mm). The adult whorls are
compressed (whorl breadth to height ratio as little as 0-75) with gently inflated inner, and flattened outer flanks,
with a fastigiate venter. Ornament consists of numerous (about thirty) rather low, rounded, prorsiradiate ribs,
arising at the umbilical shoulder without bullae and flexed strongly forwards, concave on the outer flank and
ventrolateral shoulder, where they bear blunt, clavate tubercles. The ribs are narrow as they sweep forwards
from these to long siphonal clavi. Rarely ribs branch from the umbilical seam or are intercalated, so that there
are more siphonal clavi than long ribs.

The sutures are indifferently exposed (text-fig. 1 b), but are typically collignoniceratid, with broad, simple, bifid
elements.

Discussion. D’Orbigny’s figure is partly idealized: in addition the figure lacks the abrupt start of the
ribs at the umbilical shoulder, shows too many short ribs and makes the species appear too inflated
(text-fig. 5). Pictet and Campiche (1860, p. 316) and de Grossouvre (1894, p. 75) regarded this species
as a juvenile C. woollgari, but Sharpe (1855, p. 27) had already noted that ‘the French shell has twice

Figs. 1-11. Collignoniceras carolinum (d’Orbigny). 1-3, SP, de Grossouvre Collection, probably from Bourre
(Loir-et-Cher). 4-8, the lectotype, MNHP 6778a, from the Calcaire a Cephalopodes of Martrous, near
Rochefort (Charente-Maritime). 9-10, OUM KZ 747, from the St. Cyr-en-Bourg Fossil Bed, Champignon-
niere les Rochains, south of Saumur and east of Montreuil-Bellay (Maine-et-Loire). 1 1, MNHP, from an
unknown locality in the Tuffeau Blanc de Touraine.

text-fig. 5. Collignoniceras carolinum (d’Orbigny).
Copies of d’Orbigny’s original figures (1841, pi. 91,
figs. 5-6).

EXPLANATION OF PLATE 68

PLATE 68

Kennedy, wright and Hancock, Collignoniceratid ammonites

578

PALAEONTOLOGY, VOLUME 23

as many ribs, is less compressed, and has the keel more completely separated from the ribs by two
regular channels’. Schliiter (1872, p. 27) maintained the species, as did Laube and Bruder (1887,
p. 232), although their specimen is only doubtfully referable to it. Meek (1876, p. 457) regarded
d’Orbigny’s Ammonites bravaisianus as the juvenile of carolinum, which he in turn treated as a
synonym of woollgari.

In the last 50 years the name has dropped out of currency. The most recent reference was by
Matsumoto (1971, p. 131) who upheld the view that it was possibly an immature example of C.
woollgari in which the appearance of strong distant ribs was delayed, in this respect being
intermediate between C. woollgari woollgari and C. woollgari bakeri.

C. carolinum is in fact quite distinct from C. woollgari. As early authors noted, the type of the
species is consistently more finely and densely ribbed than European C. woollgari and at comparable
diameters the ribbing is much more subdued and the ventral tuberculation finer. Other examples
before us show much sparser ribbing (PI. 68, fig. 11), but even here the ribbing is more subdued. When
adult the species are very distinct; C. carolinum reaches maturity at only 100-120 mm and never
develops the coarse umbilical bullae and ribs, the massive ventrolateral horns or the complex looped
ventral ribbing and tubercles of C. woollgari.

The delicately ribbed inner whorls immediately distinguish the species from the grossly tuberculate
young of C. canthus, C. turoniense and C. papale. Adult C. canthus are broader whorled and retain
massive bullae and ribs, whilst C. papale has strong ribs with conspicuous looping as well as being
more inflated. The feebly ornamented body chamber of C. turoniense is superficially similar, but is
much broader, virtually lacks ribs but has a row of small siphonal tubercles.

C. boreale, although adult at a similarly small diameter, has much coarser ribbing when young, and
develops distant coarse flared ribs when adult.

The confusion of C. carolinum with C. woollgari stems from the similarity of the former to finely
ribbed forms of the latter known from Japan and the United States. These have been described by
Haas (1946) as Prionotropis woollgari vars. regularis and tenuicostata, and by Matsumoto (1965) as
his Group B of C. woollgari. These finely ribbed forms are distinguished from the type of C. carolinum
in always developing relatively coarse ribs at a diameter of 20 mm or less and by ribs that are sharp
rather than subdued, straight rather than flexuous.

Occurrence. This is a rare species. Apart from the Touraine records above, it is known in France from the
environs of La Rochelle in Charente; in England from the Terebratulina lata Zone of Surrey; in north Germany,
Bohemia and Turkestan.

Collignoniceras papale (d’Orbigny)

Plate 69, figs. 1, 2; Plate 70, figs. 3-5; text-figs, lc, 6-7

1841 Ammonites Papalis d’Orbigny, p. 354, pi. 109, figs. 1-3.

1850 Ammonites papalis d’Orbigny, p. 189.

1887 Acanthoceras papaliforme Laube and Bruder, p. 237, pi. 27, figs. 3-4.

1925 Prionotropis papalis d’Orbigny; Diener, p. 156.

1925 Prionotropis papaliformis Laube and Bruder; Diener, p. 156.

1977 Collignoniceras ( Selwynoceras ) aff. papale (d’Orbigny); Hancock, Kennedy and Wright,
p. 156.

1977 Collignoniceras ( Selwynoceras ) gr. papale (d’Orbigny); Hancock, Kennedy and Wright, p. 156.

Holotype. By monotypy the specimen in the Requien Collection (Musee d’ Avignon), presumed to come from the
'craie tuffeau ou chloritee du departement de Vaucluse’ (d’Orbigny 1 841 , p. 356). We have not seen the holotype,
but d’Orbigny’s figure (text-fig. 6) is little more than two-thirds natural size.

Specimens studied. There is a series of specimens in the Museum d’Histoire Naturelle, Paris; five recorded in the
d’Orbigny Collection as coming from Montrichard (Loir-et-Cher), reg. no. 6780; MNHP W.9, unlabelled but
probably from Bourre; MNHP ‘3’, from Montrichard; MNHP ‘A’ B’ D’-‘E’ from Bourre. MNHP ‘F’
unlocalized but from the Tuffeau de Touraine.

KENNEDY ET AL.\ COLLIGNONICERATID AMMONITES

579

There are several unregistered specimens in the de Grossouvre Collection, housed in the Sorbonne, from either
Bourre or Ponce; a specimen labelled Bourre showing the inner whorls; and a small body chamber, also
unregistered, is labelled Bourre.

OUM KZ 738 and 745 are from the St. Cyr-en-Bourg Fossil Bed, Champignonniere les Rochains, 7 km south

of Saumur and north-east of Montreuil-Bellay, Maine-et-Loire.

Dimensions

D

Wb

Wh

Wb: Wh

U

MNHP W ‘9’

112-3(100)

36-4 (32)

41-8(37)

0-87

- (-)

MNHP'B’

111-7(100)

- (-)

41-0(37)

—

39 (35)

SP, de Grossouvre

160-0(100)

51-0(32)

60-0 (37)

0-85

53-0(33)

Collection

at 135-0(100)

54-5 (40)

58-0 (43)

0-94

44-5(33)

SP, Bourre

120-0(100)

40 (33)

46-0 (38)

0-87

38-8(32)

Description. The inner whorls of this species are

best displayed

by the specimen from Bourre in the

Collections illustrated as Plate 70, figs. 3-5. Up to a diameter of 55 mm the coiling is relatively evolute, with a
medium-sized umbilicus (30% of diameter), quite shallow, showing on the mould a rounded and undercut wall.

text-fig. 6. Collignoniceras papale (d’Orbigny). Copies of d’Orbigny ’s original figures ( 1 84 1 , pi. 1 09, figs. 1 -3) of
the holotype from the ‘Craie Chloritee ou Craie Tuffeau du departement de Vaucluse’. The specimen is said to be

120 mm in diameter.

580

PALAEONTOLOGY, VOLUME 23

text-fig. 7. Collignoniceras papale (d’Orbigny). Adult specimen in the Sorbonne Collections (de Grossouvre
Collection), from either Ponce or Bourre. Reduced x 0-6.

The intercostal whorl section is slightly compressed (Wb:Wh is 0-9), with convergent flanks, broadly rounded
ventrolateral shoulders and a flattened venter. The costal section is polygonal, with the greatest breadth at the
umbilical bulla. There are thirteen umbilical bullae per whorl. At the smallest diameter visible, they are very
elongate and lie close to the shoulder. With growth the maximum development migrates outwards leaving a
weak development only at the umbilicus, with the main bulla low on the flank. Broad, strong, straight, slightly
prorsiradiate ribs arise from the bullae, cross the flanks and connect to strong, conical inner ventrolateral
tubercles, from which a broad, strong rib sweeps forwards to strong outer ventrolateral clavi. These are in turn
connected to elongate siphonal clavi by a broad, low, forwardly directed rib. Between long ribs there are some
four intercalatories, usually with outer ventrolateral and siphonal clavi only.

From 50 mm onwards the ribs connecting the inner and outer ventrolateral tubercles strengthen and at 55 mm
they have fused into blunt, oblique clavi.

During middle growth, ornament consists of distant, weak to strong umbilical bullae, displaced progressively
outwards to a low or even mid flank position (not shown on d’Orbigny’s figure), which give rise to one or rarely a

EXPLANATION OF PLATE 69

Figs. 1-2. Collignoniceras papale ( d’Orbigny). SP, from Bourre (Loir-et-Cher) (Saemann Collection).

Figs. 3-4. Collignoniceras woollgari (Mantell); BMNH 5742a-b, paralectotypes from the Middle Chalk near
Lewes, Sussex.

PLATE 69

Kennedy, wright and Hancock, Collignoniceratid ammonites

582

PALAEONTOLOGY, VOLUME 23

pair of narrow, straight, prorsiradiate ribs, whilst single intercalated ribs arise at varying levels on the flank. All
ribs bear a pinched ventrolateral bulla (if weak) or horn (if strong). These are commonly limited before and
behind by narrow ribs, which loop across the venter, although the extent of this looping varies widely from
specimens in which it predominates (PI. 70, fig. 4) to those where it is simple (PI. 69, fig. 1).

Over the last half whorl of adult body chamber the tubercles decline markedly, leaving rather weak, relatively
crowded ribs without umbilical bullae, a weak, oblique to radially elongate ventrolateral tubercle (which may
disappear several ribs before the aperture) and a small blunt siphonal tubercle (text-fig. 7).

The suture is rather simple, with a broad E which tapers apically; broad, rather simply incised and
asymmetrically bifid E/L, narrow L and smaller, bifid L/U2. U2 is small (text-fig. lc).

Discussion. The material before us shows considerable variation in the relative strength of umbilical
bullae and ribs, as well as being adult (and showing typical decline in ornament) over a range of
120-180 mm diameter. Nevertheless, it forms a compact species group.

Collignoniceras canthus is immediately distinguishable on the basis of its massively tuberculate
inner whorls and feebly ribbed, almost smooth body chamber with many fine ventral clavi, as
discussed on p. 584. There are closer similarities to C. turoniense, but here the massive bullae of the
inner whorls and general dominance of tuberculation over ribbing is diagnostic, as discussed on
p. 586.

There are also similarities between juveniles of C. papale and C. woollgari , but papale have fewer,
coarser ribs (compare PI. 69, figs. 3-4 and PI. 70, fig. 3), with strong bullae displaced well out from the
umbilical shoulder and much more prominent inner ventrolateral tubercles. C. papale in middle
growth is more sharply and distantly ribbed and does not have the prominent ventrolateral horn of
many woollgari. The inner and outer ventrolateral tubercles merge into pinched, radially elongated
clavi during middle growth in papale ; in woollgari they are distinct to a much greater size. The venter
of C. papale may bear strong narrow looped ribs at a much earlier stage than woollgari and is mature
at a much smaller diameter, never developing the spectacular distantly ribbed, hypernodose body
chamber of the latter.

C. carolinum has some common features, particularly its rather small adult size. It differs in having
densely and evenly ribbed inner whorls without strong bullae, and a compressed flat-sided body
chamber without the umbilical bullae, strong ventral tubercles and broad venter with looped ribbing
of papale.

C. papaliforme (Laube and Bruder) (1887, p. 237; pi. 27, figs. 3-4), from the Turonian Greensand
of the White Mountain, near Prague, is no more than a deformed C. papale.

Occurrence. This is a relatively long-ranging species in the Tuffeau Blanc of Touraine, first appearing in the St.
Cyr-en-Bourg Fossil Bed of the Saumur region, and also occurring at Montrichard, Bourre, and Tourtenay
(Deux Sevres). Elsewhere in France there are records from Uchaux (Vaucluse). The species also occurs in the
Turonian of Czechoslovakia.

Collignoniceras canthus (Sornay)

Plate 73, figs. 1-4

1951 Ammonites canthus d’Orbigny in lift.; Sornay, p. 629, text-figs, le, 2.

1955 Ammonites ( Selwynoceras ) canthus d’Orbigny ms; Sornay, fiche 8, figs. 1-2.

1977 Collignoniceras ( Selwynoceras ) canthus (Sornay ex d’Orbigny ms); Hancock, Kennedy and
Wright, p. 156.

EXPLANATION OF PLATE 70

Figs. 1-2. Collignoniceras boreale (Warren). Cast of the holotype, Alberta Museum Collections no. CT 468,
from the basal beds of the Smoky River Shale, Grimshaw, near Peace River, Alberta.

Figs. 3-5. Collignoniceras papale (d’Orbigny), nucleus, showing coarse juvenile ornament; SP, from Bourre
(Loir-et-Cher).

PLATE 70

Kennedy, wright and Hancock, Collignoniceratid ammonites

584

PALAEONTOLOGY, VOLUME 23

Holotype. By monotypy the original of Sornay’s (1951), text-figs, le, 2, from the Tuffeau Blanc de Touraine of
Bourre (Loir-et-Cher), Museum d’Histoire Naturelle, Paris, no. 6793.

Dimensions D wb wh wh: wh v

Holotype

MNHP6793 126(100) 40-8 (32) 49-5(39) 0-82 48-6(39)

Description. The holotype and only known specimen consists of the internal mould of a body chamber 126 mm in
diameter and an external mould of the umbilicus of the inner whorls. The umbilical mould shows that the species
bore seven massive conical umbilical bullae at the smallest diameter visible (PI. 73, fig. 3) and a similar number on
the following whorl, supplemented by three ribs lacking bullae but extending to the umbilicus. From the bullae
arose rather strong ribs, usually in pairs, with occasional shorter intercalated ribs. The external mould of the
dorsum of the last part of the phragmocone shows each of these ribs to have borne a conical ventral tubercle
whence arose a pair of feeble ribs, connecting to feeble siphonal tubercles in the same looped style seen in
Collignoniceras papale (d’Orbigny).

The body chamber shows coiling to have been moderately evolute, with a small umbilicus comprising 39% of
the diameter. The umbilical wall is low and rounded, the flanks flattened and convergent, with a low fastigiate
venter which tends to become rounded towards the aperture. The maximum whorl breadth is low on the flanks,
close to the umbilical shoulder.

On the early part of the body chamber there are weak umbilical bullae, which give rise to pairs of low, broad,
radial ribs, almost insensible save to touch, as are occasional shorter, intercalated ribs. The ribs become pro-
gressively finer, denser and more subdued towards the mature aperture, and are gently flexed.

All ribs bear faint, low, rounded ventrolateral clavi which give rise to pairs of low ribs which loop forwards
and across the venter to low siphonal clavi linked into a semi-continuous serrated ridge.

The rather poorly preserved sutures of the holotype are approximated, confirming it as an adult.

Discussion. The strongly ornamented inner whorls of C. canthus place it in the same group as
C. papale and C. turoniense. It differs from both of these in the marked decline and virtual
disappearance of ornament on the outer whorl. We have seen no intermediate forms. C. carolinum
(d’Orbigny) has delicately and densely ribbed, rather than coarsely bullate inner whorls. The body
chambers of the two are more similar, especially in the marked decline in ornament, but carolinum is
much more compressed and flat-sided, the ribs are stronger, with quite thick ventral development,
and stronger siphonal clavi.

Occurrence. C. canthus is known only from the Tuffeau Blanc de Touraine of Bourre.

Collignoniceras turoniense (Sornay)

Plate 71, figs. 4-5; Plate 72, figs. 1-3
1951 Prionotropis turoniense Sornay, p. 630; pi. 21, figs. 1-3.

1977 Collignoniceras ( Selwynoceras ) turoniense (Sornay); Hancock, Kennedy and Wright, p. 156.
Holotype. MNHP unregistered, Peron Collection, from Bourre (Loir-et-Cher), by monotypy.

Other specimens studied. MNHP ‘A’, from Bourre, and two unregistered specimens in the de Grossouvre
Collection (Sorbonne, Paris), probably from Bourre.

Dimensions

D

Wb

Wh

Wb: Wh

U

Holotype

120 (100)

48 (40)

48-3 (40)

1-0

~ (-)

MNHP ‘A’

107 (100)

52 (49)

44 (41)

118

34-5(32)

Sorbonne, 1

125 (100)

43-5 (35)

48-5 (39)

0-9

34-5(28)

at 107-5 (100)

52-5 (49)

43-5 (43)

1-2

23-0(21)

EXPLANATION OF PLATE 71

Figs. 1-3. Collignoniceras woollgari (Mantell) FSR C273, from Ponce, Sarthe.

Figs. 4-5. Collignoniceras turoniense (Sornay), the holotype, MNHP, Peron Collection, from Bourre (Loir-et-
Cher).

PLATE 71

Kennedy, wright and Hancock, Collignoniceratid ammonites

586

PALAEONTOLOGY, VOLUME 23

Description. All known specimens are adults, with two-thirds of the last whorl being body chamber, and none
show the early whorls. Coiling is involute on the phragmocone, becoming relatively evolute at maturity, with a
deep umbilicus. On the phragmocone the whorl section is depressed, with convergent flanks and a fastigiate
venter intercostally. The costal section is even more depressed, the greatest breadth being at the umbilical bullae,
and subcarinate. There are five massive blunt conical umbilical nodes per whorl. These give rise to groups of two
or three broad, low ribs, with additional ribs intercalated low on the flank between the groups. At the smallest
diameters visible- these bear blunt conical inner ventrolateral tubercles and small clavate outer ventrolaterals,
with a broad low rib connecting them to stronger siphonal clavi borne on a blunt keel. On the last part of the
body chamber the intercalated ribs decline, the inner and outer ventrolateral tubercles combine into a blunt
transversely elongate tubercle, which gives rise to pairs of ribs which loop to strong siphonal clavi, which become
first rounded, then transversely elongate. Some short ventral ribs with a siphonal tubercle are intercalated, to
give a serrated blunt keel; there are three to five siphonal nodes to each pair of umbilicals.

On the body chamber the umbilical nodes decline in strength and disappear towards the aperture; intercalated
ribs are lost and the primary ribs weaken and become irregular and closely spaced. There are irregularly spaced,
clavate ventrolateral nodes, which also decline towards the aperture, with many more ventral ribs and siphonal
tubercles than ventrolateral.

The body chamber uncoils markedly and the shell becomes much more evolute as a result. Whorl
height: breadth ratio decreases, so that the aperture appears relatively constricted.

None of the specimens shows the suture well but they appear to have comprised broad, plump, rather simple
bifid lobes and saddles.

Discussion. The inner whorls of Collignoniceras turoniense are easily distinguished from those of
C. woollgari and C. carolinum, which are densely and evenly ribbed by comparison, lacking massive
bullae. In middle growth, C. turoniense has a much more massive whorl, broad and low rather than
narrow ribs and stronger ventrolateral than umbilical nodes. The adults are quite distinct (compare
PI. 62, figs. 1-2 and PI. 71, figs. 4-5).

C. canthus has similar inner whorls, but becomes virtually smooth in middle and later growth,
lacking massive umbilical bullae and strong ventrolateral tubercles.

C.papale juveniles (PI. 70, figs. 3-5) have many more (typically 9-1 1) and smaller umbilical bullae,
narrow and widely spaced ribs and more markedly clavate ventrolateral and siphonal tubercles. In
middle and later growth the differences between the two lie in the predominance of tuberculation in
C. turoniense and of ribbing in C. papale, the latter having the bullae displaced outwards to a lower
flank position and strong, narrow, well-differentiated ventral ribs looping between the ventrolateral
and siphonal tubercles with intercalatories.

C. carolinum is compressed, parallel-sided and feebly ribbed without strong bullae in middle and
later growth.

Occurrence. C. turoniense is known only from the Tuffeau Blanc de Touraine of Bourre.

Collignoniceras boreale (Warren)

Plate 70, figs. 1-2

1930 Prionotropis borealis Warren, p. 25, pi. 3, figs. 1-4; pi. 4, fig. 1.

1940 Selwynoceras borealis Warren; Warren and Stelck, p. 151.

Types. The holotype is the original of Warren 1930, pi. 3, fig. 1, University of Alberta Museum Collections
no. CT 468. Paratypes are CT 469-76, all from the basal beds of the Smoky River Shale, Grimshaw, near Peace
River, Alberta.

EXPLANATION OF PLATE 72

Figs. 1-3. Collignoniceras turoniense (Sornay) SP, de Grossouvre Collection, probably from Bourre (Loir-
et-Cher).

PLATE 72

KENNEDY, wright and Hancock, Collignoniceratid ammonites

588

PALAEONTOLOGY, VOLUME 23

Description. The holotype, a cast of which is before us, is a slightly distorted mould retaining traces of shell and
consists of half a whorl of body chamber and one quarter of a whorl of phragmocone with the following
dimensions:

D Wb Wh Wb.Wh U

costal 92-5(100) 40 (43) 33-5 (36) 1 19 35-2(38)

intercostal 90-2(100) 29-5(33) 31 (34) 0-98 35-2(39)

Coiling is moderately evolute, the umbilicus comprising 38% of the diameter, broad and rather shallow. The
umbilical wall slopes gently outwards and the whorl section is a compressed oval (whorl breadth to height ratio is
0-98) with flattened flanks. The phragmocone bears three long, straight, prorsiradiate distant ribs. These arise
from small umbilical bullae and also bear conical inner and clavate outer ventrolateral tubercles; there is a
siphonal row of distant clavi corresponding in position to the ventrolateral tubercles. Two shorter, intercalated
ribs are also present, bearing outer ventrolateral and siphonal clavi only. This same style of ventral ornament is
shown on the penultimate whorl, preserved in the dorsum of the body chamber, and in two of the paratypes
(Warren 1930, pi. 3, figs. 2-3).

On the body chamber the umbilical bullae decline and the ribs become high, distant, and flared into a
ventrolateral horn which supports the outer ventrolateral clavus. There is a poorly defined siphonal ridge,
accentuated into siphonal clavi, and the upper ventrolateral and siphonal clavi are linked by broad transverse
ribs which show incipient doubling with a riblet developing at both front and rear.

The suture is simple and little incised, with broad bifid saddles.

Discussion. Small size and even ventral tuberculation are the features by which Warren’s species is
most easily distinguished from C. woollgari ; other differences are noted on p. 572. There are no other
species with which it is likely to be confused. Of interest, however, is the striking resemblance of the
holotype to specimens of C. woollgari from the Black Hills area of the U.S. Western Interior, which
also show a very even and equal number of upper ventrolateral and siphonal clavi, never, apparently
developing the intercalated ribs and tubercles of what we take as typical woollgari. These specimens
(so far as we have seen) are much larger when adult and have horns with a triangular outline in ventral
view rather than flares. These Interior examples are obviously close relatives of the Canadian form,
although their precise relative ages are not known.

Occurrence. As for types.

Genus lecointriceras gen. nov.

Type species. Ammonites fleuriausianus d’Orbigny, 1841, p. 350.

Diagnosis. Medium-sized, involute during early and middle growth, becoming evolute at maturity. Whorls
trapezoidal when young, with sparse conical umbilical tubercles giving rise to pairs of low broad ribs, with
occasional intercalatories. All ribs bear outer ventrolateral and siphonal clavi on a fastigiate venter, but the
appearance and persistence of inner ventrolateral tubercles is variable. In middle growth the venter often
broadens and flattens, the ventrolateral tubercles fuse into a blunt horn and there is a low continuous undulant
siphonal ridge, strengthened between horns. The last part of the adult body chamber is contracted, tubular and
unornamented except for growth lines, and the aperture is simple.

The suture is simple with broad, asymmetrically bifid E/L, narrower L and smaller bifid L/U2.

Discussion. The whorl section, massive umbilical tubercles and sparse low ribs of early middle
growth, the blunt horns and the tubular body chamber distinguish Lecointriceras from all other
collignoniceratids and the persistence of short ribs on the sides from contemporaneous Collignoni-
ceras. Some C. woollgari develop a short, smooth terminal portion to the body chamber but their

EXPLANATION OF PLATE 73

Figs. 1-4. Collignoniceras canthus (Sornay). The holotype, SP 6793, from Bourre (Loir-et-Cher). 3 is the
external mould of the nucleus; 4 shows the decline in ornament over the last part of the body chamber.

PLATE 73

Kennedy, wright and Hancock, Collignoniceratid ammonites

590

PALAEONTOLOGY, VOLUME 23

compressed, finely ribbed inner and middle growth stages, much narrower flank ribs, retention of
multiple siphonal ribs and clavi is distinctive. This ventral ribbing and retention of clavi also
distinguish C. canthus and C. papale\ C. turoniense has a smooth body chamber, but lacks the
massive umbilical tubercles and ventral horns in middle growth and on the first part of the body
chamber. The phragmocone of some Lecointriceras and the adult shell of C. boreale are superficially
similar, but Warren’s species has compressed finely ribbed inner whorls and on the outer whorl,
which is slender and rounded intercostally, the ribs lack a massive bulla, are narrower and produced
into a narrow flared bituberculate horn rather than the single broad protuberance seen in
Lecointriceras.

As is discussed below. Ammonites vielbancii d’Orbigny, 1850 is a synonym of A. fleuriausianus.
Schliiter (1871, pp. 21-22) believed the former might be a synonym of Mammites nodosoides
(Schliiter), and Collignon (1939) and Wiedmann (1960, 1964) referred it to Mammites. As
Pervinquiere (1907, p. 31 1) noted, the siphonal tubercles are quite distinctive.

In Europe Lecointriceras first appears in the mid-Turonian St. Cyr-en-Bourg Fossil Bed,
accompanying typical Collignoniceras. Its origins may lie in one of the undescribed Thomelites-Mke
forms occurring in the earliest English Turonian.

Occurrence. Widespread in the French Turonian (Touraine and Aquitaine); also occurring in northern Spain,
Czechoslovakia, north Germany and southern England.

Lecointriceras fleuriausianum (d’Orbigny)
Plate 74, figs. 1-10; Plate 75, figs. 1-5; text-figs. 8, 9

1841 Ammonites Fleuriausianus d’Orbigny, p. 350, pi. 107, figs. 1-3.

1841 Ammonites Woollgari d’Orbigny, p. 352 (pars), pi. 108, figs. 1-3.

1850 Ammonites Vielbancii d’Orbigny, p. 189.

1860 Ammonites Fleuriausianus (d’Orbigny); Courtiller, p. 250, pi. 3, fig. 1.

1867 Ammonites Fleuriausianus d’Orbigny; Courtiller, p. 7, pi. 7, figs. 1-4.
non 1869 Ammonites Fleuriauanus d’Orbigny; Schloenbach, p. 291.

1871 Ammonites Vielbancii d’Orbigny; Schliiter, p. 19 et seq.

?1872 Ammonites Fleuriausianus d’Orbigny; Schliiter, p. 28, pi. 10, figs. 1-3.

1887 Acanthoceras Fleuriausianum d’Orbigny; Laube and Bruder, p. 234.
non 1902 Acanthoceras Fleuriausianum d’Orbigny; Petrascheck, p. 147, pi. 11, figs, la-b, 2.

1907 Ammonites Vielbancii d’Orbigny; Pervinquiere, p. 311.

1939 Mammites Vielbancii d’Orbigny; Collignon, p. 81, pi. 11, figs. 1, 2.

1946 Ammonites vielbancii d’Orbigny; Sornay, p. 213.

1946 Ammonites fleuriausianus d’Orbigny; Sornay, p. 214.

1960 Mammites vielbanci (d’Orbigny); Wiedmann, p. 721 .

1977 Collignoniceras ( Selwynoceras ) fleuriausianum (d’Orbigny); Hancock, Kennedy and Wright,
p. 156.

Type series. Ammonites fleuriausianus has been a poorly understood species, although the type figure (if taken to
be natural size) is an accurate representation of the middle growth stages and the type series survives. In his
original description d’Orbigny recorded the species ‘en place dans la craie chloritee ou craie tufau des Martrous,

EXPLANATION OF PLATE 74

Figs. 1-10. Lecointriceras fleuriausianum (d’Orbigny). 1 -2, the lectotype of ‘ Mammites ’ vielbancii (d’Orbigny),
MNHP 6779, (d’Orbigny Collection) from Saumur (Maine-et-Loire). 3-5, CS 629b, from the environs of
Saumur (Maine-et-Loire), a juvenile of moderate inflation. 6-7, the lectotype, MNHP 6777b (d’Orbigny
Collection) from the Calcaire a Cephalopodes of Rochefort (Charente-Maritime). 8-10, FSM 125, from
Ponce, Sarthe, a hypernodose juvenile.

PLATE 74

Kennedy, wright and Hancock, Collignoniceratid ammonites

*2$

592

PALAEONTOLOGY, VOLUME 23

pres de Rochefort (Charente-Inferieur); M. Dufrenoy l’a aussi du meme lieu; M. d'Archiac l’a observee a
Gourdon (Lot); MM. Dufrenoy et Graves font trouvee, aux environs de Saumur’ (d’Orbigny 1841, p. 352). In
the posthumous catalogue of his collection (dating from 1858-60) the following are recorded:

6777 Saumur, Maine-et-Loire, 3 specimens (4 are present).

6777a Martrous, 1 specimen (missing).

6777b Rochefort, Charente-Inferieur, 2 specimens (3 are present).

6777c Chatellerault, Vienne, 2 specimens (1 missing).

The Saumur specimens belong to at least two species. The first, 34-5 mm in diameter, is a crushed tuffeau
specimen, and is labelled [La] Fleche. It has rather flattened flanks, with umbilical bullae giving rise to 2-3
flexuous ribs with some intercalatories, giving a total of sixteen ribs per whorl. There are distinct conical inner
ventrolateral tubercles and subequal outer ventrolateral and siphonal clavi, which show it to be afleuriausianum,
as is a second individual with an estimated diameter of 55 mm, but having little indication of inner ventrolateral
tubercles and weak siphonal clavi.

A third specimen, 71 mm in diameter, and labelled Saumur, is a worn, wholly septate Jeanrogericeras
reveliereanus. The final specimen has ‘Rochefort’ written on it in pencil and is also a J. reveliereanus, with a
diameter of 104 mm. Superficially it could be the basis of d’Orbigny’s side view but it lacks all signs of a
siphonal clavus.

Two specimens from Rochefort are associated with a plaque labelled 6777b. Both are well preserved on one
side, the larger 55 mm in diameter, the smaller 35 mm, and appear to be part of d’Orbigny’s original suite. The
larger of these, the most typical in the series, is here designated lectotype.

The single specimen to survive of those originally labelled 6777c is a very battered, crushed, distorted specimen
in yellow tuffeau. Umbilical bullae give rise to pairs of ribs, terminating in rounded ventral clavi, with no sign of
siphonal nodes, suggesting it to be a mammitid or Jeanrogericeras. Chatellerault was not mentioned as a locality
by d’Orbigny in his original description and thus this specimen is not a syntype.

The types of A. vielbancii, herein regarded as a synonym, also present a confused situation. It is a Prodrome
species introduced (d’Orbigny 1850, p. 189, no. 11) as follows: ‘ Vielbancii , d’Orb., Paleont., 1, p. 352, pi. 108,
figs. 1-3. Sous le faux nom de Woolgarii, Mantell. Martrous, Saumur, Tourtenay.’

In Paleontologie Franqaise (1841, p. 354) he cites the species as occurring more widely, but we take these
references (which include England) to be to the true Collignoniceras woollgari.

The d’Orbigny catalogue lists the following:

6779 Saumur, Maine-et-Loire, 3 (4 specimens).

6779a Bords de la Vienne, 2 (1 missing).

6770b Rochefort, (illegible) (missing).

whilst d’Orbigny notes that his lateral view (pi. 108, fig. 1) is of a specimen in his collection and the apertural
view is of a specimen in the Ecole des Mines.

Inspection shows that the d’Orbigny specimens have become mixed. The Rochefort specimen is present, but
labelled 6779. It is poorly preserved, but may be the basis of d’Orbigny’s side view. The specimen from the Bords
de la Vienne is not a syntype; it is a large Mammites nodosoides. As Sornay has discussed (1946, p. 214), the
specimen figured in side view by d’Orbigny does not look like any of the poor specimens which survive in his
collections under the name vielbancii, and certainly there is little resemblance between d’Orbigny’s figures and
the specimen re-figured by Collignon as ‘type’— which we take to be a valid lectotype designation. Even the
specimen in the School of Mines upon which d’Orbigny (1841, p. 354) said his apertural view is based (no. A35.3,
locality unknown: ‘Bassin de la Loire, achete de Stur’ reads the label) does not correspond to the figure (compare
text-figs. 8 A-c and 9 a-b). We would suggest, in fact, that the illustrations are composite, the side view being
based on the poor Rochefort specimen of appropriate size, combined with the ornament of the huge Mammites
no. 6779a from the ‘Bords de la Vienne’, the apertural view being based on the School of Mines specimen plus the
Mammites.

Description. The smallest individuals we have seen are approximately 30 mm in diameter. At this size the coiling
is fairly involute (umbilicus = 25% or less of diameter) and the umbilicus quite deep, with a rounded wall. The

EXPLANATION OF PLATE 75

Figs. 1-5. Lecointriceras fleuriausianum (d’Orbigny). 1-3, FSM 120, 4-5, FSM 121, compressed and inflated
middle-aged individuals from the Turonian of Sarthe.

PLATE 75

Kennedy, wright and HANCOCK, Collignoniceratid ammonites

text-fig. 8. A-c, copies of d’Orbigny’s original figures of ‘ Ammonites Woollgari Man tell’ (1841, pi. 108, figs. 1-3)
= Ammonites vielbancii d’Orbigny, 1850. The illustration is said in the text to be reduced by a third and on the
plate by a half, d-f, copies of d’Orbigny’s original figures of Ammonites fleuriausianus (1841, pi. 107, figs. 1-3).
The illustration is said to be reduced by a third.

text-fig. 9. Lecointriceras fleuriausianum (d’Orbigny) a, B. EMP A35.3, ‘Bassin de la Loire, achete de Stur’ — the
original of d’Orbigny’s (1841) pi. 108, fig. 2. Reduced x 0-66. c, d. FSM 1 19, an adult from Ponce, Sarthef?)
showing the smooth, tubular termination to the body chamber. Reduced x 0-6 approx.

596

PALAEONTOLOGY, VOLUME 23

intercostal whorl section is typically compressed, with the greatest breadth low on the convergent flank and with
rounded shoulders and venter. In the costal section the greatest breadth is at the umbilical bulla and whorl
breadth to height ratios vary greatly up to 1:2, with concave inner flanks and a fastigiate venter.

Ornament consists of weak to strong conical umbilical bullae, 7-9 per whorl, giving rise to pairs of low, broad
straight ribs, with occasional intercalated ribs arising low on the flank. The ribs decline somewhat in strength on
the mid-flank but then strengthen into rounded inner ventrolateral tubercles. These are connected by a
strengthened rib to strong clavate outer ventral tubercles, from which a broad subdued rib sweeps forwards to
a subequal clavate siphonal tubercle.

This general style of ornament varies from individual to individual, with slender, feebly bullate forms with
weak ribs (PI. 74, figs. 6-7) and strongly bullate inflated forms with strong ribs (PI. 74, figs. 8-10). In many
individuals, including the lectotype, there are no inner ventrolateral tubercles below diameters of 35-42 mm;
occasionally they do not appear until 55 mm.

From 50 mm onwards there is usually a change in ornament; the bullate umbilical tubercles become more
distant, the associated ribs lower and broader, effaced at mid-flank in some specimens. There are usually 7-9
bullae and 16-22 ribs per whorl. The outer ventral tubercles weaken rapidly and disappear; at the same stage the
inner ventrolateral tubercles strengthen without joining the weakening ventral tubercles (PI. 77, fig. 4). The
former inner ventrolateral tubercles gradually develop into strong to massive horns on the shoulder, triangular
when viewed ventrally and relatively narrow when viewed laterally, developed both upwards and outwards.
At this, the ‘ vielbancii' stage, the venter becomes relatively broad, with a continuous low undulant siphonal
ridge, strengthened between horns at what corresponds to the site of the now coalesced siphonal clavi. The
shell now closely resembles a Mammites in all but the siphonal ridge.

This style of ornament extends onto the first half of the adult body chamber, by which stage the siphonal ridge
may become very reduced (text-fig. 9 c-d). On the last half of the body-chamber, extending for just over a
quarter whorl, all ribs and tubercles are lost and there is a relatively smooth, compressed and constricted
terminal portion with convergent sides, broadly rounded shoulders and a flattened venter, ornamented only by
low, prorsiradiate growth striae. The aperture is simple and entire.

The suture line is relatively simple, with a broad medial element to E; broad, asymmetrically bifid E/L;
narrow, symmetrically bifid L; smaller asymmetrically bifid L/U2; and small and narrow U2.

Discussion. D’Orbigny ’s original figure is idealized and bears little relationship to the surviving
syntypes in his collection; in his explanation of the plate he says the figure is reduced by a third, so
that the specimen is far larger than the proposed lectotype, being, presumably, the Martrous
specimen which is now lost. The lectotype agrees well with the dimensions given by d’Orbigny for his
smaller specimen (1841, p. 350). Juveniles of this species vary in the strength of the umbilical
tubercles; the lectotype is worn but was probably a slender, weakly tuberculate variant. This
variation continues into middle growth, where both slender and robust individuals are known (PI. 74,
figs. 3-10).

The striking contracted tubular termination of the body-chamber of adults of this species occurs at
disparate sizes. Most specimens we have seen appear to be juveniles of individuals that would have
been adult at approaching 1 50 mm diameter, but a specimen in the collections at Le Mans is complete
at only 100 mm, with half a whorl of the body chamber so modified. Unfortunately our sample of
adults is too small to show if the species shows a size dimorphism.

Some of the early references to this species are doubtful. Schloenbach’s (1869) material probably
belongs to Barroisiceras, whilst Schluter’s specimen (1872, p. 28; pi. 10, figs. 1-3), if indeed a true
L. fleuriausianum, has suffered great post-mortem crushing to give a very compressed whorl section.

Lecointriceras carinatum sp. nov., described below, differs from L. fleuriausianum in its smaller
adult size, early loss of umbilical tubercles and ribs, together with retention of a fastigiate venter on
the adult body chamber, which bears an undulose siphonal and flanking, semi-continuous lateral

EXPLANATION OF PLATE 76

Figs. 1-2. Collignoniceras carolinum (d’Orbigny), MNE1P W8, an adult body-chamber from an unknown
locality in the Tuffeau Blanc de Touraine.

Figs. 3-5. Lecointriceras carinatum sp. nov. The holotype, EMP, Ponce(?), Sarthe.

PLATE 76

Kennedy, wright and Hancock, Collignoniceratid ammonites

598

PALAEONTOLOGY, VOLUME 23

keels formed by coalescence of ventral and siphonal clavi. Differences from L. costatum sp. nov. are
discussed below.

The combined features of L.fleuriausianum as here described are so distinctive that confusion with
any other collignoniceratid is unlikely. Juveniles have a passing similarity to some Barroisiceratinae;
species of Barroisiceras have less prominent umbilical tubercles and many strong, narrow ribs at a
comparable size; whilst Forresteria and similar genera have an additional, lateral row of tubercles. In
middle growth there is a superficial resemblance to Mammites, but that genus never develops a
siphonal tubercle.

Occurrence. This species is common at the level of the mid-Turonian St. Cyr-en-Bourg Fossil Bed in the Saumur
area in Touraine, occurs in northern Aquitaine, Vaucluse, Provence, northern Spain, north Germany(?), and
Devon, England.

Lecointriceras carinatum sp. nov.

Plate 76, figs. 3-5

Holotype. A body-chamber in the Collections of the School of Mines, Paris, labelled Ponce(?) and in pencil
Choffaticeras ’ typique; ’’Thomasites"' . It is clearly from either Ponce or Bourre.

Description. The holotype and only known specimen is a half whorl, largely body-chamber and in typical rather
coarse tuffeau preservation. Coiling is very involute with a tiny umbilicus (10% of diameter). The dorsum of the
specimen (PI. 76, figs. 3-5) shows the whorl section of the inner whorls to have been slightly depressed, with the
greatest breadth at the umbilical shoulder, concave, convergent flanks and a fastigiate venter. There were sparse
umbilical bullae giving rise to low, broad ribs which terminate at elongate ventrolateral clavi, with a sharp
siphonal keel, accentuated into clavi which correspond to the ventrolaterals.

On the first part of the body chamber ornament is similar. There are low broad flank ribs which terminate in
long clavi linked into undulant keels, flanking a similarly undulant keel in which clavi merge towards the
aperture.

The poorly preserved suture shows a typical broad bifid E/L, narrow L, and broad L/U2, all with only minor
incisions.

Discussion. The single known individual is so distinctive that erection of a new species is justified. The
inner whorls are typical of a Lecointriceras, differing from L.fleuriausianum in the sparse, low, broad
ribs and presence of keels. Absence of a quadrate-whorled vielbancii stage makes the body chamber
equally distinctive. There is a striking similarity to Masiaposites Collignon, 1965, a late Turonian
form best known from Madagascar and currently regarded as a vascoceratid; however its siphonal
keel is entire and its sutures are much more deeply incised, rather like that of Neoptychites, and the
siphonal keel continuous throughout ontogeny.

Occurrence. The species is known only from the type occurrence at Ponce(?), Sarthe (mid-Turonian).

Lecointriceras costatum sp. nov.

Plate 77, figs. 1-3

1902 Acanthoceras Fleuriausianum d’Orbigny; Petrascheck, p. 147, pi. 11, figs. 1-2.

Holotype. AM 55 from the Tuffeau Blanc of Saumoussay, Maine-et-Loire, France.

Other specimens studied. AM 22 from Montsoreau, Maine-et-Loire; AM 53, 54, 60, 101, and 102 from
Saumoussay, Maine-et-Loire, France.

EXPLANATION OF PLATE 77

Figs. 1-3. Lecointriceras costatum sp. nov. 1-2, the holotype, AM 55, from Saumoussay, Maine-et-Loire;
3, AM 53 from Saumoussay, Maine-et-Loire.

Fig. 4. Lecointriceras fleuriausianum (d’Orbigny). AM 36 from Saumoussay, Maine-et-Loire; oblique view to
show the concurrent weakening of the outer and the strengthening of the inner ventrolateral clavi.

PLATE 77

Kennedy, wright and Hancock, Collignoniceratid ammonites

600

PALAEONTOLOGY, VOLUME 23

D

Wb

Wh

Wb: Wh

U

R

AM 55 (Holotype)

125-5

(100)

~ (— )

53 +

33

c. 21

AM 53

95

(100)

c. 36 ( )

c. 44-5

0-81

AM 60

183

(100)

71

54

0-76

AM 101

165

(100)

14

at 129

(100)

50-5

56

0-90

35

17

AM 102

109

(100)

40

45

0.89

30

Description. This is a moderately evolute and relatively compressed Lecointriceras, with the greatest whorl-
breadth still at the umbilical tubercles in costal section. Of the fourteen to twenty-one ribs slightly less than half
are long; the shorter ribs start about halfway up the sides. Each long rib bears an umbilical bulla, a clavus high on
the sides rather than in the normal position of an inner ventrolateral, an outer ventrolateral clavus and a siphonal
clavus. The siphonal clavi are elevated above the shoulder clavi and up to a diameter of 125 mm may form a
nodose keel. During the earlier ontogeny the high lateral clavi are weaker than those on the shoulders, but at
diameters which may be anything from 70-110 mm the upper lateral clavi strengthen and the shoulder clavi
weaken; the upper lateral clavi eventually become ventrolateral horns on the body-chamber. Similarly the
umbilical bullae become weak and are not present on all long ribs beyond diameters of 100 mm. The adult body-
chamber begins at about 125 mm diameter, but none of the specimens seen has well-preserved sutures.

Discussion. L. costatum differs from L. fleuriausianum in having a more compressed whorl section
with flatter sides, weaker umbilical tubercles (which are, however, still stronger than in typical
Collignoniceras spp.), siphonal clavi elevated above the outer ventrolateral clavi and persistent outer
ventrolateral and upper lateral clavi through much of ontogeny, certainly from a diameter of 40 mm
to about 125 mm.

Occurrence. All known French specimens are from the mid-Turonian Tuffeau Blanc of the Saumur region. In
that formation ammonites are most common in the St. Cyr-en-Bourg Fossil Bed, but we have not found any
specimens of L. costatum ourselves; as Amedro and Badillet (1978) have pointed out, ammonites do occur at
other levels in the Tuffeau Blanc. The specimens figured by Petrascheck were from Labiatus-Planer at Leubnitz
and Briessnitz near Dresden in the German Democratic Republic.

EVOLUTIONARY AND STRATIGRAPHIC CONCLUSIONS

The origins of Collignoniceras and the Collignoniceratidae seem to lie in late Thomelites of
Acanthoceratidae, the transition involving a raising of the mid-venter and forwards displacement of
siphonal clavi and ribs to give a ventral chevron ornament. This is indicated by a few scraps we have
seen from the Cenomanian-Turonian boundary beds in Devon. Lecointriceras may also arise in this
way, or be a slightly later offshoot from already distinct Collignoniceras : the low Turonian record is
too poor to be certain. In the United States C. woollgari overlaps late Mammites nodosoides (W. A.
Cobban, in litt.); in Europe C. woollgari and L. fleuriausianum co-occur in the earliest of the French
Tuffeau faunas. C. woollgari is a long-ranging species which occurs throughout the mid-Turonian
zone of which it is the index species. In Europe we have detected no evolutionary changes in the
successive Collignoniceras faunas studied. In contrast, W. A. Cobban’s work on western interior
sequences allows recognition of an early form, in which both long and short ribs persist in middle and
later growth, and a late form in which long ribs dominate. That this is not seen in Europe suggests that
typical individuals had reached the U.S. Western Interior by the beginning of woollgari Zone time,
and underwent subsequent local differentiation which did not occur in European populations. The
other collignoniceratids described here are mostly long ranging: C. carolinum, C. papale and
L. fleuriausianum range through most of the woollgari Zone. L. costatum is restricted to the lower
part, L. carinatum, C. turoniense and C. canthus to middle and low upper levels in the Zone.

These disappointingly meagre stratigraphic conclusions mean that any subdivision of the broad
woollgari Zone must be based on other groups. We have already suggested that a local sequence of
Romaniceras can be used in Touraine: R. (R.) kallesi (oldest) -> R. ( Yubariceras ) ornatissimum -*•
R. (R.) deverianum (youngest) (Hancock, Kennedy and Wright 1977; Kennedy, Wright and

KENNEDY ET AL .: COLLIGNONICERATID AMMONITES

60 :

Hancock, this volume). The lower two of these are clearly correlated with the xvoollgari Zone, but we
are not entirely certain whether R. deverianum marks a level at the very top of the woollgari Zone or at
the base of the succeeding Subprionocyclus neptuni Zone. Ammonites are too scarce at this level in
both England and northern France for us to be sure either way; Romaniceras appears to be absent
from the rich neptuni Zone fauna of the Chalk Rock (Wright 1979) but occurs in the Uchaux
(Vaucluse) faunas.

Acknowledgements. We are grateful to the following colleagues for allowing us to examine specimens in their
care, and/or for much useful discussion: Dr. J. Sornay, Dr. D. Pajaud, the late General M. Collignon, Dr. R.
Busnardo, Dr. J. P. Lefranc, Dr. J. Lovail, Mr. M. Gruet, Professor K. Young, Dr. C. Duerdon, Dr. M. R.
Cooper, Dr. W. A. Cobban, Dr. W. A. Popenoe, Dr. E. G. Kauffman, Dr. V. Housa, Dr. R. Zazvorka,
Dr. M. K. Howarth, Mr. D. Phillips, Professor T. Matsumoto and Dr. I. Hayami. The financial support of the
Royal Society, British Association for the Advancement of Science and N.E.R.C. is gratefully acknowledged by
Kennedy and Hancock. We thank the staff of the Geological Collections, University Museum, Oxford, and
of the Department of Geology, King’s College, London for their help and assistance.

REFERENCES

adkins, w. s. 1931. Some Upper Cretaceous ammonites in western Texas. Univ. Tex. Bull. 3101, 35-72, pis. 2-5.
amedro, f. and badillet, G. 1978. Repartition des ammonites dans quelques coupes du Turonien des environs de
Saumur (Maine-et-Loire). C.r. hebd. Seanc. Acad. Sci., Paris, (D) 286, 323-325.
arkhanguelsky, A. D. 1912. The fossil fauna from the shores of the Aral Sea. 1: Upper Cretaceous deposits,
part 1. Izv. Turkest. Otd. imp. russ. geogr. Obshch. 8 (2). [In Russian.]

— 1916. The molluscs of the Upper Cretaceous of Turkestan. Trudy geol. Komit. S.-Peterburg, 152, 1-57 + vi,
pis. 1-8. [In Russian.]

arkell, w. J. et al. 1957. Cephalopoda Ammonoidea. Treatise on Invertebrate Paleontology , L, Mollusca 4,
490 + xxii pp.

Basse, e. 1948. Quelques ammonites nouvelles du Cretace de Colombie (Am. Sud). Bull. Soc. geol. Fr. (5) 18,
691-698, pi. 25.

— 1950. Quelques mollusques du Cretace de Colombie. Bull. Soc. geol. Fr. (5) 20, 245-255, pi. 11.

bose, e. 1928 (mis-dated 1927). Cretaceous ammonites from Texas and northern Mexico. Univ. Tex. Bull. 2748,
143-357, pis. 1-18.

breistroffer, m. 1947. Notes de nomenclature paleozoologiques. Proc. verb. mens. Soc. Sci. Dauphine , 26 (195),
5 pp.

cobban, w. A. and scott, G. R. 1972. Stratigraphy and ammonite fauna of the Graneros Shale and Greenhorn
Limestone near Pueblo, Colorado. Prof. Pap. U.S. geol. Surv. 645, 108 + v pp., 41 pis.
collignon, M. 1939. Fossiles cenomaniens et turoniens du Menabe (Madagascar). Annls. geol. Serv. Mines
Madagascar, 10, 59-105, pis. 1-11.

— 1965. Atlas des fossiles caracteristiques de Madagascar (Ammonites), 12 (Turonien), 82+iv pp.,
pis. 376-413. Tananarive, Service Geologique.

courtiller 1860. Description de trois nouvelles especes d’ammonites du terrain cretace. Memoires de la
Societe imperiale d' Agriculture Sciences et Arts d' Angers. 3rd series, 3, 246-252, pis. 1-3.

— 1867. Les Ammonites du tuffeau. Annales de la Societe Linneenne du Departement de Maine-et-Loire, 9th
year, pp. 1-8, pis. 1-8.

diener, c. 1925. Ammonoidea neocretacea. Fossilium Catalogus 1: Animalia, 29, 244 pp.

fieber, f. x. 1853. Synopsis der europaischen Orthopteren mit besonderes Riicksicht der Bohmischen Arten.

Lotos, 3, 90-104, 115-129, 138-154, 168-176, 184-188, 201-207, 232-238, 252-261.
fritsch, a. 1872. Cephalopoden der bohmischen Kreideformation. 52 pp., 16 pis. Prag, Fr. Rivnac.
geinitz, H. b. 1871-1875. Das Elbthalgebirge in Sachsen. Part 1, Der untere Quader. Palaeontographica, 20,
319 pp., 67 pis.

gray, d. A. 1965. The strati graphical significance of electrical resistivity marker bands in the Cretaceous strata of
the Leatherhead (Fetcham Mill) borehole, Surrey. Bull. geol. Surv. Gr. Br. 23, 65-1 15, pis. 8-9.
grossouvre, a. de. 1894 (mis-dated 1893). Recherches sur la craie superieur. 2: Paleontologie— Les ammonites
de la craie superieur. Mem. Serv. Carte geol. det. Fr. 264 + (ii) pp., 39 pis.
haas, o. 1946. Intraspecific variation in, and ontogeny of, Prionotropis woollgari and Prionocyclus wyomingensis.
Bull. Am. Mus. not. Hist. 86, 141-224, pis. 11-24.

602

PALAEONTOLOGY, VOLUME 23

Hancock, j. m., Kennedy, w. j. and wright, c. w. 1977. Towards a correlation of the Turonian sequences of
Japan with those of north-west Europe. Spec. Pap. palaeont. Soc. Japan , 21, 151-168.
hattin, D. e. 1975. Stratigraphy and depositional environment of Greenhorn Limestone (Upper Cretaceous) of
Kansas. Bull. Kans. Univ. geol. Surv. 209, 128 + (iii) pp., 10 pis.
hoepen, E. c. n. van. 1955. Turonian-Coniacian ammonites from Zululand. S. Afr. J. Sci. 51, 361-377.
kullmann, j. and wiedmann, J. 1970. Significance of sutures in phylogeny of Ammonoidea. Paleont. Contr.
Univ. Kans. 47, 32 pp.

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

logan, w. N. 1898. The invertebrates of the Benton, Niobrara and Fort Pierre Groups. Univ. geol. Surv. Kans. 4
(Paleont.) (Upper Cretaceous 8), 431-518, pis. 86-120.

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

matsumoto, T. 1959. Upper Cretaceous ammonites of California, part II. Mem. Fac. Sci. Kyushu Univ.
(D. Geol.), Special Vol. 1, 1-172, pis. 1-41.

— 1965. A monograph of the Collignoniceratidae from Hokkaido, Part 1. Mem. Fac. Sci. Kyushu Univ.
(D. Geol.), 16, 1-80, pis. 1-18.

— 1971. A monograph of the Collignoniceratidae from Hokkaido, Part 5. Ibid. (D. Geol.), 21, 129-162,
pis. 21-24.

— and wright, c. w. 1966. Collignoniceras Breistroffer, 1947 (Mollusca, Ammonoidea): application to place
on the official list of generic names in zoology with priority from 1876. Bull. zool. Nom. 23, 57-59.

meek, F. B. 1876. A report on the invertebrate Cretaceous and Tertiary fossils of the upper Missouri country. In
hayden, F. v. Report of the United States Geological Survey of the Territories, 9, 629 + lxiv pp., 45 pis.
orbigny, a. d\ 1840-1842. Paleontologie Franqaise. Description des Mollusques rayonnes fossiles. Terrains
Cretaces, 1 (Cephalopodes), 662 pp., 151 pis. Paris, Victor Masson.

— 1850. Prodrome de Paleontologie stratigraphique universelle des animaux mollusques et rayonnes . . ., 2,
428 pp. Paris, Victor Masson.

pervinquiere, L. 1907. Etudes de paleontologie tunisienne I. Cephalopodes des terrains secondaires. Carte geol.
Tunisie, 438 + v pp., 27 pis.

petrascheck, w. 1902. Die Ammoniten der sachsischen Kreideformation. Beitr. Palaont. Geol. Ost.-Ung. 14,
131-162, pis. 7-12.

pictet, f. J. and campiche, G. 1858-1860. Description des fossiles du terrain cretace des environs de Sainte-
Croix, part 2 (1), Description des fossiles. Materiaux pour la paleontologie suisse (2) 2, 29-380, pis. 1-43.
Geneva, J. Kessmann and H. Georg for F. J. Pictet.

powell, j. d. 1963. Turonian (Cretaceous) ammonites from northeastern Chihuahua, Mexico. J. Paleont. 37,
1217-1232, pis. 166-171.

RAWSON, P. F., CURRY, D., DILLEY, F. C., HANCOCK, J. M., KENNEDY, W. J., NEALE, J. W., WOOD, C. J. and

worssam, B. c. 1978. A correlation of Cretaceous rocks in the British Isles. Spec. Rep. geol. Soc. Lond. 9, 70 pp.
schloenbach, u. 1869. Bemerkungen fiber einige Cephalopoden der Gosaubildungen. Jb. K.-K. geol.
Reichsanst., Wien, 19,291-294.

schluter, c. 1871-1876. Cephalopoden der oberen deutschen Kreide. Palaeontographica, 21, 1-120, pis. 1-35;
24,3-143 -t-xii, pis. 36-55.

sharpe, d. 1853-1857. Description of the fossil remains of Mollusca found in the Chalk of England. Cephalo-
poda. Palaeontogr. Soc. ( Monogr .), 68 pp., 27 pis.

shimizu, s. 1 932. On a new type of Senonian ammonite, Pseudobarroisiceras nagaoi Shimizu gen. et sp. nov. from
Teshio Province, Hokkaido. Jap. J. Geol. Geogr. 10, 1-4, pi. 1.
sornay, j. 1946. Remarques sur deux especes de d’Orbigny: Ammonites vielbancii et A. fleuriausianus. Bull.
Mus. Hist. nat. Paris. (2) 18, 213-216.

— 1951. Sur deux especes d’ammonites inedites de d’Orbigny et sur une espece nouvelle du tuffeau de
Touraine. Bull. geol. Soc. Fr. (6) 1, 627-631, pi. 21.

— 1955. Ammonites ( Selwynoceras ) canthus d’Orbigny. Palaeont. univers. (n.s.) 8.
sowerby, j. de c. 1828. The mineral conchology of Great Britain . . ., 6, pis. 581-597.

spath, l. f. 1923. On the ammonite horizons of the Gault and contiguous deposits. Summ. Progr. geol. Surv.
Lond. for 1922, 139-149.

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

warren, p. s. 1930. Three new ammonites from the Cretaceous of Alberta. Trans. R. Soc. Can. (3) 24 (4), 21-26,
pis. 1-4.

KENNEDY ET AL COLLIGNONICERATID AMMONITES

603

warren, p. s. and stelck, c. r. 1940. Cenomanian and Turonian faunas in the Pouce Coupe District, Alberta
and British Columbia. Trans. R. Soc. Can. (3) 34 (4), 143-152, pis. 1-4.

Wedekind, r. 1916. Uber Lobus, Suturallobus und Inzision. Zentbl. Miner. Geol. Palaont. (B) 1916 (8),
185-195.

wiedmann, j. 1960 (mis-dated 1959). Le Cretace superieur de l’Espagne et du Portugal et ses cephalopodes.

C.r. Congr. Soc. sav.-Dijon 1959: Colloque sur le Cretace superieur franqais, 709-764.
windmoller, t. 1882. Die entwickelung des Plaeners im nordwestlichen theile des Teutoburger Waldes bei
Lengerich. Jb. preuss. geol. Landesanst, BergAkad. [2] (for 1881), Abh. 3-54, pi. 19.
wright, c. w. 1979. The ammonites of the English Chalk Rock. Bull. Br. Mus. nat. Hist. (Geol.), 31, 281-332,
7 pis.

— and Kennedy, w. J. In juignet, p., Kennedy, w. J. and wright, c. w. 1973. La limite Cenomanien-Turonien
dans la region du Mans (Sarthe): stratigraphie et paleontologie. Annls. Paleont. (Invert.), 59, 207-242.

— and matsumoto, T. 1954. Some doubtful Cretaceous ammonite genera from Japan and Saghalien. Mem.
Fac. Sci. Kyushu Univ. (D. Geol.), 4, 107-134.

— and wright, e. v. 1951. A survey of the fossil Cephalopoda of the Chalk of Great Britain. Palaeontogr. Soc.
(Monogr.), 40 pp.

young, K. 1963. Upper Cretaceous ammonites from the Gulf Coast of the United States. Univ. Tex. Pubis. 6304,
373 + viii pp. (82 pis.).

W. J. KENNEDY
C. W. WRIGHT
University Museum
Parks Road, Oxford OX1 3PW
and Wolfson College, Oxford OX2 6UP

Typescript received 28 March 1979
Revised typescript received 8 November 1979

J. M. HANCOCK

Department of Geology
King’s College, Strand
London WC2R 2LS

THE TRILOBITE ECCOPTOCHILE FROM THE
ORDOVICIAN OF NORTHERN PORTUGAL

Abstract. The eccoptochilinid trilobite fauna from the Ordovician of the Valongo area, north Portugal, is
revised. The holotype of Eccoptochile (1 Eccoptochile) mariana (Verneuil and Barrande, 1855) is redescribed and
figured and the species is restricted to the type specimen and two specimens from Valongo. Specimens pre-
viously described as E. (IE.) mariana from Spain, north Portugal, and southern England, together with other
and new material from Portugal are here included within the new species E. ( Eccoptochile ) almadenensis.
E. ( Eccoptochile ) cf. clavigera (Beyrich, 1845) is recorded from the Valongo area.

This revision of the genus Eccoptochile from the Ordovician of north Portugal forms part of a
larger project concerned with systematic description and distribution studies of the Ordovician
trilobite faunas of that region. The faunas from the Valongo area about 10 km east of Porto (text-
fig. 1) have been well known since Delgado published extensive faunal lists from the beds (1908,
pp. 106-109); only ‘ Uralichas Ribeirof (Delgado, 1892, 1897) was described. Delgado listed
‘ Cheirurus claviger Beyrich’, ‘ Cheirurus Guillieri Tromelin (aff. C. clanger Beyrich)’, and ''Cheirurus
sp. n. (aff. C. Sedgwicki McCoy)’ from his uppermost division, the ‘Schistes a Uralichas Ribeirof ,
of the ''Ordovician moyen' from the Valongo area. Prior to this Sharpe (1849) had recorded
‘ Chirurus ’ from the Porto region but this specimen was later recognized by Salter (1853) as
‘ Placoparia Zippei, Boeck’. The most recent systematic work on this group was by Curtis (1961)
who apparently regarded all three of the species listed by Delgado as conspecific and referred them
to Eccoptochile mariana (Verneuile and Barrande).

The ‘Schistes a Uralichas Ribeirof have generally been regarded as Llandeilo in age (Costa 1931;

by M. ROMANO

text-fig. 1. Simplified geological map of the area
south of Valongo (after Delgado 1908), showing
localities (asterisks) where the species of Eccoptochile
described in the text have been recorded.

IPalaeontology, Vol. 23, Part 3, 1980, pp. 605-616, pis. 78-79.)

606

PALAEONTOLOGY, VOLUME 23

Teixeira 1955; Thadeu 1956) and more recent work on certain elements of the fauna, notably
harpids (Romano 1975), placopariids (Romano 1976), and dionidids (Henry and Romano 1978),
suggests a possible Lower Llandeilo age, equivalent to the Placoparia ( Coplacoparia ) tournemini
biozone of Spain and Brittany (Hammann 1971a; Henry and Clarkson 1975). The ‘Schistes a
Uralichas Ribeirof are included within the upper part of the Valongo Formation (Romano and
Diggens 1973-1974) which is a thick sequence of argillaceous sediments ranging in age from Upper
Llanvirn ( Didymograptus murchisoni Zone) to ?Upper Llandeilo. The formation crops out about
10 km east of Porto and it is from this area that the bulk of the collections were made by Delgado,
Wattison, and the present author with J. N. Diggens. The problem of accurately locating the material
collected by Wattison was outlined earlier (Romano 1976) and similar difficulties arise with some of
the specimens from the Delgado collection.

The collections used in this paper are housed in the British Museum (Natural History), London
(Wattison Collection — (BM In)); Ecole Nationale Superieure des Mines, Paris (T); Servigos
Geologicos, Lisbon (Delgado Collection— SG); Institute of Geological Sciences (GSM), and in the
Geology Department, University of Sheffield (SU).

SYSTEMATIC PALAEONTOLOGY

General remarks. E. ( Eccoptochile ) clavigera (Beyrich, 1 845), E. (I Eccoptochile) mariana (Verneuil
and Barrande, 1855) and E. (? Eccoptochile) guillieri (Tromelin in Guillier, 1873) form a relatively
homogeneous group within which the north Portuguese specimens clearly belong. The first two are
generally regarded as valid species but E. (IE.) guillieri has fairly recently been placed into synonymy
with E. (IE.) mariana by Hammann (1974, p. 105). E. (IE.) guillieri was compared with
E. (E.) clavigera by Tromelin and Lebesconte (1876, p. 637) who noted that the glabella of the
former differed from that in E. (E.) clavigera in being smooth, more convex, with the posterior end
of the axial furrows curved inwards more strongly. The outline diagrams and locality information
of the specimens of E. (IE.) guillieri shown in text-fig. 2 (g, h, i) were kindly sent to me by Dr. J.-L.
Henry; they are of the holotype (2g) and a topotype (2h, i from the Kerforne collection). The latter
is an incomplete but undeformed specimen, preserved in a nodule, from the type locality ‘la Butte
du Creux’, near Saint-Denis-d’Orques (Sarthe); Dr. Henry informed me that it is Llanvirn or
Llandeilo in age. This topotype shows a very narrow (sag.) frontal area and a strongly and evenly
curved glabella in lateral view. From these two specimens E. (IE.) guillieri warrants retention as a
separate species and is treated as such in this paper.

The most commonly reported species of Eccoptochile in Iberia and the Armorican Massif is
E. mariana (Curtis, op. cit.; Hammann 1971, 1974; Lindstrom, Racheboeuf and Henry 1974), but a
recent study of the holotype of this species by the author suggests that the species has been interpreted
too widely in the past. The holotype of mariana is redescribed and figured here.

Prantl and Pribyl (1948) erected the subgenus Eccoptochile ( Eccoptochiloides ) on the basis of the
thorax containing only ten segments and the four pairs of pleural spines on the pygidium. As the
thorax and pygidium of E. (IE.) mariana are unknown the subgeneric status of mariana is still in
doubt.

The morphological terms used are essentially those listed by Harrington et al. (in Moore, 1959).
Lateral glabellar lobes and furrows are labelled ‘L’ and ‘S’ respectively and are numbered from the
posterior forwards. The classification employed is that of Henningsmoen (in Moore, 1959) and Lane
(1971).

Family cheiruridae Hawle and Corda, 1847
Subfamily eccoptochilinae Lane, 1971
Genus eccoptochile Hawle and Corda, 1847

Type species. Cheirurus claviger Beyrich, 1845

ROMANO: ORDOVICIAN TRILOBITE ECCOPTOCHILE

607

text-fig. 2. Outline sketches of the cephala or cranidia of the holotype and other material of
selected species of Eccoptochile : a-c, Eccoptochile (Eccoptochile) almadenensis sp. nov. a, b,
Holotype (selected), from Hammann, 1974, pi. 12, fig. 192c and 192b (reversed for comparison);
c, from Curtis, 1961, pi. 2, fig. 1. d-f , Eccoptochile ( Eccoptochile ) clavigera (Beyrich); d, Holo-
type, from Beyrich, 1845, pi. (unnumbered), fig. 2; e, /, from Barrande, 1852, pi. 40, figs. 1, 2.
g-i, Eccoptochile (? Eccoptochile) guillieri Tromelin in Guillier); g, Holotype, h, i, Topotype.
Both drawings taken from photographs and drawings supplied by Dr. J.-L. Henry. y-/, Eccopto-
chile (? Eccoptochile) mariana (Verneuil and Barrande); y, k , Holotype, from Verneuil and
Barrande, 1855, pi. 23, fig. 4 and present paper, pi. 1, figs. 1-4; /, Paratype, from Curtis,
1961, pi. 1, fig. 1 and refigured here, pi. 1, figs. 5, 6. Sketches drawn to approximately the same

size.

Eccoptochile (? Eccoptochile) mariana (Verneuil and Barrande, 1855)

Plate 78, figs. 1-7; text-fig. 2 j-1

*1855 Cheirurus marianus Verneuil and Barrande, p. 970, pi. 23, fig. 4 (not p. 972, pi. 28 as stated by
Hammann, 1974, p. 105).

1961 Eccoptochile mariana (Verneuil and Barrande); Curtis, p. 6, pi. 1, fig. 1 (not pi. 1, fig. 2, pi. 2;
figs. 1, 2, ?pl. 3, fig. 1).

1974 Eccoptochile cf. mariana (Verneuil and Barrande); Hammann, p. 105 (referring to Curtis, 1961,
pi. 1, fig. 1).

Diagnosis. (Modified from Verneuil and Barrande, 1855, p. 970.) A species of Eccoptochile with the following
characteristics: strongly arched glabella with evenly curved longitudinal profile and, with occipital ring vertical,
highest part level with the anterior part of L2. Wide frontal area over 12% of the glabellar length (excluding
occipital ring) and consists of a more or less flat preglabellar field and a gently rounded anterior border.
Palpebral lobe level with the posterior part of L2 to the posterior part of L3. Eye ridges are faintly visible
running from the anterior of the palpebral lobe towards S3. Hypostoma, thorax, and pygidium unknown.

Type and figured material. Holotype: T 150 (Plate 78, figs. 1-4). Internal mould of incomplete cranidium
(Verneuil and Barrande, 1855, pi. 23, fig. 4). Other figured material. BM In49177 (Plate 78, figs. 5, 6) (Curtis,
1961, pi. 1, fig. 1); BM In49182 (Plate 78, fig. 7).

608

PALAEONTOLOGY, VOLUME 23

Horizon and locality. Holotype from ‘Puente de las Ovejas’ near Ciudad Real, Spain; Upper Llandeilo
(Hamman, 1974, p. 105). BM In49177 and In49182 from Covelo, near Valongo, north Portugal; upper part
of Valongo Formation, probably Lower Llandeilo.

Description of holotype. Measurements with occipital ring vertical: length (sag.) of glabella (excluding occipital
ring) and frontal area, 15-75 mm; length of glabella, 14 00 mm. Glabella longer than wide with even, out-
wardly curved lateral margins, slightly indented at S3, and a broadly rounded anterior margin; widest part
of the glabella just anterior to the S2 furrows. LI lobes subtriangular in outline, about one-quarter glabellar
length and delimited by deep, well-marked SI furrows which have an S-shaped trace and die out just under
one-third glabellar width from axial furrows. L2 lobes rectangular in outline, shorter (trans.) than LI and
about the same length (exsag.). S2 furrows shorter and less well-marked than SI, evenly curved, parallel to the
abaxial part of SI, starting just posterior to the midlength of the glabella. L3 similar in shape and orientation
to L2, but appear to be very slightly longer. S3 furrows parallel to S2 but do not reach as far towards the
midline. S3 start at nearly two-thirds the glabellar length from the posterior margin.

Glabella strongly arched transversely with a subtriangular cross section. Longitudinally (occipital ring
vertical) the glabella is evenly curved dorsally, highest part lying above the anterior part of L2. Median
glabellar lobe, L2 and L3 without independent convexity but LI lobes are slightly bulbous. Frontal area wide
(sag., exsag.), of more or less constant width around the frontal glabellar lobe but increasing at anterolateral
corners where anterior margin of fixed cheek turns back rather sharply to give a more angular, although still
rounded outline. Frontal area consists of an inner preglabellar field which is more or less flat or very slightly
upwardly concave which grades into the frontal lobe of the glabella without a marked furrow. Preglabellar
field also grades into anterior border which is gently rounded and lies horizontally. Anterolaterally the border
is slightly wider. At anterolateral comers border appears to be directed more upwards but this may be an
effect of deformation. Axial furrows well-marked from the occipital furrow to S3 where there is a deep pit
just abaxial to axial furrow. Anterior to this pit axial furrow rapidly dies out. Occipital furrow curved forwards
behind the median glabellar lobe and where it runs into the axial furrows, deep posterior to the LI lobes and
wide and shallow in the median part. Occipital ring not complete: posterior to the LI lobes ring curves
forwards. Incomplete free cheeks are narrow (trans.) opposite the palpebral lobes and fairly flat. Posterior
border furrow deep, starting from the axial furrow just posterior to the occipital furrow. Posterior border
narrow (exsag.) and convex. Convex (tr.) palpebral lobe slightly curved, lying oblique to sagittal line and
separated from fixed cheek by a well-marked palpebral furrow which dies out anteriorly along length of lobe.
Faint eye ridge extends from palpebral lobe to axial furrow at S3. Palpebral lobe level with posterior part of L2
to the posterior part of L3. Faint granular ornament on glabella but the distribution is not clear. On the
fixed cheeks there is an irregular distribution of pits.

The figured material from Valongo assigned to this species is virtually identical to the holotype, differing
mainly in convexity. The Portuguese specimens are flattened dorso-ventrally and slightly distorted obliquely.
The transverse and longitudinal profiles of the cranidia do not show the high convex glabella of the holotype
but the relative proportions of the cranidia are the same. This species is discussed further below.

EXPLANATION OF PLATE 78

Figs. 1-7. Eccoptochile (?. Eccoptochile) mariana (Verneuil and Barrande). 1-4, holotype, internal mould;
T 150. ‘Puente de las Ovejas’ near Ciudad Real, Spain; Upper Llandeilo. 1-3, dorsal, frontal, lateral views
respectively, x 3. 4, detail of cheek ornament, x9. 5, 6, internal mould; In49177. Covelo, north Portugal.
Upper part of Valongo Formation; Lower Llandeilo. 5, dorsal view. 6, frontal view. Approximately x 2.
7, internal mould; In49182. Covelo, north Portugal. Upper part of Valongo Formation; Lower Llandeilo.
Dorsal view, x 1.

Figs. 8, 9. Eccoptochile {Eccoptochile) almadenensis sp. nov. Internal moulds. 8, GSM CR 1526. Gorran
Quartzites, Perhaver Beach, Cornwall; Llandeilo. Dorsal view, x2. 9, SG 3A2. 1400 m S 32° E of Covelo
church, north Portugal. Upper part of Valongo Formation; Lower Llandeilo. Dorsal view, x 1.

PLATE 78

romano, Ordovician trilobite Eccoptochile

610

PALAEONTOLOGY, VOLUME 23

Eccoptochile ( Eccopotchile ) almadenensis sp. nov.

Plate 78, figs. 8, 9; Plate 79, figs. 1-7; text-fig. 2 a-c

1896 Cheirurus ( Eccoptocheile ) marianus (De Verneuil); Reed, p. 164.

1907 Cheirurus sedgwicki M'Coy; Lake in Reid, p. 39.

1908 Cheirurus claviger Beyrich; Delgado, ? p. 57 (refigured by Thadeu, 1947, pi. 3, fig. 2), ? p. 80,

p. 106.

1908 Cheirurus guillieri Trom. (aff. C. claviger Beyr.); Delgado, p. 106.

1908 Cheirurus sp. n. (aff. C. sedgwicki McCoy); Delgado, p. 106.

1916 Eccoptochile mariana (Verneuil and Barrande); Barton, p. 106.

* 1918 Cheirurus claviger var. marianus Verneuil and Barrande emend. Born; Born, p. 351, pi. 27, fig. 1.
1947 Cheirurus claviger Beyrich; Thadeu, p. 228, pi. 3, fig. 3.

1958 Eccoptochile clavigera (Beyrich); Whittard, p. 115 (specimen from Perhaven Beach, Cornwall).
1961 Eccoptochile mariana (Verneuil and Barrande); Curtis, p. 6, pi. 1, fig. 2 ( non fig. 1), pi. 2, figs. 1, 2,
pi. 3,? fig. 1.

1969 Eccoptochile ( Eccoptochile ) sp. indet; Racheboeuf, p. 74, pi. 2, figs. 3 a, b.

19716 Eccoptochile marianus (Verneuil and Barrande); Hammann, pp. 267, 270.

1974 Eccoptochile clavigera (Beyrich)?; Sadler, p. 73.

1974 Eccoptochile (. Eccoptochile ) mariana (Verneuil and Barrande); Lindstrom, Racheboeuf, and
Henry, ? pp. 20, 21.

1974 Eccoptochile mariana (Verneuil and Barrande); Hammann, p. 105, text-fig. 39, pi. 11, figs. 188-
191, pi. 12, figs. 192-198.

1978 Eccoptochile mariana (Verneuil and Barrande); Henry and Romano, p. 335.

Diagnosis. (Modified from Hammann, 1974, p. 106.) Species of Eccoptochile with glabella strongly convex,
anterior lobe descending almost vertically to preglabellar field. Frontal area relatively narrow (sag.); anterior
border steeply upturned forming an angle with lateral borders of free cheeks (viewed dorsally). Eyes start
approximately level with S2 and reach back to SI. Fixed cheeks narrow (sag.). Anterior thoracic segments
pointed, becoming gradually more rounded posteriorly. Internal surface of exoskeleton smooth except for pits
on cheeks.

Type and figured material. Holotype: (SMG X 337a) Internal mould of cephalon with seven thoracic segments
(figured Born, 1918, p. 351, pi. 27, fig. 1; Hammann, 1974, p. 105, pi. 12, figs. 192 a-c). Paratypes: (BM
In49 178-80) Curtis, 1961, p. 6, pi. 1, fig. 2, pi. 2, figs. 1 and 2 respectively; (SMF 24779-82, 24783^3, 24784,
24785a, 24787) Hammann, 1974, p. 105, pi. 11, figs. 188, ?189, 190-191, pi. 12, figs. 193-198. Other material:
GSM GR 1526; SG 1704, SG 171 1, and three unnumbered specimens in drawer labelled 3A2 in SG (figured here
PI. 78, fig. 9, PI. 79, figs. 6, 7).

Horizons and locality. Holotype from Valdemosillo, approximately 16 km ENE of Almaden, Spain; Upper
Llandeilo. Paratypes. BM In49178-80 from Covelo, near Valongo, north Portugal; upper part of Valongo
Formation, probably Lower Llandeilo. SMF 24779, 24785a from Corral de Calatrava (near Ciudad Real,
Spain); Co Illf, Upper Llandeilo: SMF 24780-82, from Corral de Calatrava; Co Hie, Upper

EXPLANATION OF PLATE 79

Figs. 1 -7. Eccoptochile ( Eccoptochile ) almadenensis sp. nov. 1 -6, internal moulds, 7, external impression. Upper
part of Valongo Formation; Lower Llandeilo. 1-4, SG 1704. 1650 m S 20° W of the summit of Santa Justa,
Valongo, north Portugal. 5, SG 1711, 6, 7 (both in drawer labelled 3A2), 1400 m S 32° E of Covelo
church, north Portugal. 1 -3, dorsal, lateral, frontal views respectively, x 2; 4, detail of thoracic segment, x 4.
5-7, dorsal views, x 1, x0-75, x 1 respectively.

Fig. 8. Eccoptochile (? Eccoptochile) cf. mariana (Verneuil and Barrande). Internal mould; (no number, same
box as SG 1704). 1650 m S 20° W of the summit of Santa Justa, Valongo, north Portugal. Upper part of
Valongo Formation; Lower Llandeilo. Dorsal view, x 1.

Fig. 9. Eccoptochile ( Eccoptochile ) cf. clavigera (Beyrich). External impression. SG (no number, in drawer
labelled 3A2). 1400 m S 32° E of Covelo church, north Portugal. Upper part of Valongo Formation; Lower
Llandeilo. Dorsal view, x 1-5.

PLATE 79

romano, Ordovician trilobite Eccoptochile

612

PALAEONTOLOGY, VOLUME 23

Llandeilo: SMF 24784 from Navatrasierra (Montes de Toledo, Spain); Na la. Lower Llandeilo: SMF 24784
from Navatrasierra (Montes de Toledo, Spain); Na la. Lower Llandeilo: SMF 24787 from Navatrasierra
(Montes de Toledo); Na la, basal Llandeilo.

Description. The types from Spain and Portugal have been well described and figured by Hammann (1974)
and Curtis (1961). No further comments are necessary.

Discussion. Verneuil and Barrande erected Eccoptochile (l Eccoptochile) mariana (1855, p. 970, pi. 23,
fig. 4) on the basis of it having a more dorsally convex glabella and a wider, flat anterior border than
Eccoptochile ( Eccoptochile ) clavigera (Beyrich, 1845). They stated that the eye occupied the same
relative position in both species. Curtis (1961, p. 8) listed four differences between the two species,
including that in E. (IE.) mariana ( sensu Curtis and Hammann) the frontal lobe is relatively shorter,
the eye ridge starts level with the anterior glabella furrow and the eye is situated farther back. The
specimen figured by Curtis (1961, pi. 1, fig. 1) as E. mariana , and later referred to E. cf. mariana
by Hammann (1974, p. 105) possesses a wide frontal area which distinguishes it from other specimens
of E. (IE.) mariana as understood by Curtis and Hammann. A reinvestigation of the holotype of
E. (IE.) mariana also revealed the presence of a wide frontal area and it is thus clearly distinct
from the majority of specimens previously assigned to that species. The evenly curved longitudinal
profile of the glabella of the holotype (text-fig. 2k and PI. 78, fig. 3) is also unlike that in
E. (IE.) mariana sensu Hammann where maximum curvature occurs in the anterior part of the
glabella (Hammann 1974, pi. 12, fig. 1926). Thus E. (IE.) mariana is restricted in this paper to
include, with the holotype, only the two specimens from the Valongo area; that figured by Curtis
(1961, pi. 1, fig. 1 and refigured here, PI. 78, figs. 5, 6) and a previously unfigured specimen
(PI. 78, fig. 7).

The relative lengths (sag.) of the frontal glabellar lobe and frontal area appear to show significant
differences in the species almadenensis, clavigera , and mariana. In an attempt to quantify these
differences the three parameters B, F, and b5 (text-fig. 3a) (symbols from Shaw, 1957 and Temple,

text-fig. 3. a. Outline of cranidium of Eccoptochile
(Eccoptochile) almadenensis sp. nov. (after Ham-
mann, 1974, text-fig. 39) showing parameters used in
(b), (c) and text-fig. 4; b, c. Scatter diagrams of F
against B and F against b5 respectively with calculated
regression lines for the species almadensis, clavigera
and mariana.

ROMANO: ORDOVICIAN TRILOBITE ECCOPTOCHILE

613

1975) were selected since it is assumed the ratio of these measurements taken along a constant
orientation will be virtually unaffected by deformation. When the three parameters are plotted on
size frequency and scatter diagrams the species plot out in isolated and relatively restricted fields.
Size/frequency histograms of the B:F and B : b5 ratios (not illustrated) serve to distinguish
E. (IE.) mariana from E. (E.) almadenensis and E. ( E .) clavigera quite markedly. The regression
lines of B against F and b5 against F (text-fig. 3b and 3c) show that for mariana at least the lines
appear to be clearly distinguishable and although few specimens were available to construct the
graphs (almadenensis— 14; clavigera— 7; mariana— 3) the contrast in gradient suggests the difference
in growth rate is a useful criterion for separating this species. When the three parameters are plotted
as ratios on a triangular graph (text-fig. 4) the three species plot out in discrete fields and the

B

text-fig. 4. Triangular plot for the species almadenensis, clavigera, and mariana
using the three parameters B, F, b5 (see text-fig. 3). For material and references
used to construct the graph see text. Additional sources include Dr. J.-L. Henry
(pers. comm.) and author’s collection, University of Sheffield.

selected holotype for E. (E.) almadenensis occurs near the centre of scatter for that species. Since
the number of specimens is small the fields have not been numerically defined. The species
clavigera is clearly distinguishable by the presence of a long (sag.) frontal glabellar lobe (see text-fig. 2)
and the flat profile of the glabella in lateral view. This difference in the relative length of the frontal
lobe is shown in the groupings in text-fig. 4.

Using the methods outlined above, E. (IE.) guillieri cannot be distinguished from E. (E.)
almadenensis since measurements taken from the photographs supplied by Dr. J.-L. Henry plot out
near the middle of the E. (E.) almadenensis field (text-fig. 4). However, the strong glabella convexity

614

PALAEONTOLOGY, VOLUME 23

and subrounded outline of the glabella in dorsal view of E. (IE.) guillieri are characteristic
enough to suggest it is a valid species. The specimen listed by Delgado (1908, p. 106), as
‘ Cheirurus sp. n. (aff. Ch. Sedgwicki McCoy)’, from 1400 m S 32° E of Covelo church (SG 1711)
appears to show no important differences from E. (E.) almadenensis. The size of the free cheek,
position, and structure of the eye in Placoparina sedgwicki (Whittard, 1958, pp. 112, 115) are
distinctive, and although the Portuguese specimen listed by Delgado is imperfectly preserved
(PI. 79, fig. 5) it is assigned to E. (E.) almadenensis. Delgado (1908, p. 106) also recorded
‘ Cheirurus Guillieri Trom. (aff. Ch. claviger Beyr.)’ from the Valongo area, 1650 m S 20° W from
the hill of Santa Justa (SG 1704), but the forwardly expanding and relatively longer glabella (PI. 79,
fig. 1) is unlike that of the holotype of E. (IE.) guillieri and this specimen is also identified as
E. (E.) almadenensis. Another specimen (PI. 79, fig. 8) identified by Delgado (op. cit.) as ‘ Cheirurus
Guillieri ’ is here referred to E. (IE.) cf. mariana because, although it closely resembles the holotype,
the deformed specimen precludes a definite identification. The eccoptochilinid from Perhaver Beach,
Cornwall, tentatively identified as E. (E.) clavigera by Whittard (1958, p. 115) and Sadler (1974,
p. 73) is an incomplete cranidium (PI. 78, fig. 8) which can now be confidently assigned to E. (E.)
almadenensis.

Eccoptochile ( Eccoptochile ) cf. clavigera (Beyrich, 1845)

Plate 79, fig. 9

Figured material. One external impression of an incomplete flattened pygidium; specimen housed in Serv^os
Geologicos, Lisbon; drawer 3A2.

Horizon and locality. 1400 m S 32° E of Covelo church, Valongo; probably from upper part of Valongo
Formation, probably Lower Llandeilo.

Description. Pygidium nearly twice as wide as long. Anterior margin gently rounded with nearly straight
median portion and more strongly rounded posterior margin. Axis subtriangular in outline (articulating half
ring not preserved) with outwardly curved axial furrows. Axis probably slightly wider than long, reaching
back to about one-half length of pygidium; three axial rings and a small triangular terminal piece; rings decrease
in length posteriorly, ring furrows shallow medially (except third axial ring furrow). Axial furrows shallow and
weakly defined and not present posterior to the second axial ring furrow. Three pairs of broad, bluntly
rounded, spinose pleural ribs. First and second ribs start opposite first two axial rings and curve gently
outwards and backwards; third pair directed posteriorly. 7-8 shallow pits on first pleural ribs situated at about
midlength (exsag.) of rib and extend for about one-quarter along the rib. Only 1-2 pits are present on the
second rib and none on the third. Surface of pygidium covered with fine, closely spaced tubercles except in the
shallow rib pits.

Discussion. The poor preservation of this specimen makes it difficult to compare length to width
ratios with the type material of E. (E.) clavigera (Beyrich, 1845, plate (unnumbered), fig. 3), which
appears to be relatively wider. In all other respects it closely resembles the holotype. The present
material is very similar to the specimen referred to E. (E.) clavigera by Pribyl and Vanek (1969,
p. 3, fig. 8) except that in the latter the rows of pits on the pleural ribs extend further along the
rib, although this is not so apparent in other specimens figured by those authors (op. cit. pi. 3,
figs. 6, 7).

RANGE AND DISTRIBUTION OF E. ( E .) ALMADENENSIS, E. (E.) CLAVIGERA,
AND E. (IE.) MARIANA

E. (E.) almadenensis is the most widespread species in Iberia and the Armorican Massif and
probably also occurs in southern Cornwall. It first appears in the basal Llandeilo of Navatrasierra
in central Spain (Hamman 1974, p. 15) and occurs in the Lower Llandeilo of north Portugal, the
Armorican Massif, and probably southern England. There is evidence that the species possibly also
persists into the Caradoc in the region south of Rennes, Brittany (Lindstrom et al., 1974, p. 20). There

ROMANO: ORDOVICIAN TRILOBITE ECCOPTOCHILE

615

is no record of it continuing into the Ashgill. E. (IE.) mariana (as understood in this paper) is a
relatively restricted species, recorded only from the Ciudad Real region in south central Spain where it
is of Upper Llandeilo age and from the area around Covelo, near Valongo in north Portugal
(Lower Llandeilo). E. ( E .) clavigera is poorly represented in Spain and north Portugal; E. (E.)
cf. clavigera (a deformed pygidium) occurs in probably Lower Llandeilo beds in the Valongo area
and E. ( E .) aff. clavigera (a hypostoma and pygidium) is recorded from Caradoc beds north of
Almaden, central Spain (Hammann 1974, p. 1 1 1). A deformed eccoptochilinid cranidium from the
?Caradoc of central Portugal, 50 km SSE of Coimbra (A. H. Cooper collection), is probably referable
to E. ( E .) clavigera and the Delgado collection housed in the Servigos Geologicos, Lisbon, contains
large specimens of E. (E.) clavigera from the Magao region 80 km SSE of Coimbra. The age of the
Magao specimens is not known but the associated fauna contain Actinopeltis and Eoharpes and could
indicate an Upper Llandeilo to Caradoc age. E. (E.) clavigera is common in Bohemia where it
ranges from the Liben Formation to the Bohdalec Formation (Havlicek and Vanek 1966) and is
associated with Actinopeltis. Havlicek and Marek (1973) have revised the chronostratigraphic
terminology for the Bohemian sequence and they recognize a Beroun Series of middle Llandeilo to
upper Caradoc age which includes the range of E. (E.) clavigera. In Bohemia Eoharpes dies out in the
Dobrotiva Formation which is considered by these authors to be equivalent in age to the lower part
of the Llandeilo.

Any conclusions regarding faunal migrations and phylogeny within the group must await further
work in particular on the existing collections in Lisbon.

Acknowledgements. I thank Dr. R. A. Fortey (British Museum), Dr. A. W. A. Rushton (Institute of Geological
Sciences), and the Director of the Ecole National Superieure des Mines for loaning material in their care.
I also thank Dr. Jean-Louis Henry for supplying me with photographs and outline drawings of eccoptochilinid
trilobites from the Armorican Massif. He and Professor H. B. Whittington kindly read and criticized the
manuscript. Mr. M. Cooper redrew the diagrams and Miss P. Mellor typed the manuscript. The work was made
possible by a N.E.R.C. grant.

REFERENCES

barrande, j. 1852. Systeme silurien du centre de la Boheme. lere partie. Recherches paleontologiques, vol. 1.

Crustaces, Trilobites. xxx + 935 pp., 51 pis. Prague and Paris.
barton, d. c. 1916. A revision of the Cheirurinae with notes on their evolution. Wash. Univ. Stud, scient. Ser.
3(1), 101-152.

beyrich, E. 1845. Ueber einige bohmische Trilobiten. 47 pp., 1 pi. Berlin.

born, a. 1916. Die Calymene Tristani- Stufe (mittleres Untersilur) bei Almaden, ihre Fauna, Gliederung und
Verbreitung. Abhandl. Senck. Naturf. gesell. 36, 309-358, 4 pis.
costa, j. s. Carrington da, 1931. O Paleozdico Portugues ( Sintese e Critica), 1-141, Porto.

Curtis, m. l. K. 1961. Ordovician trilobites from the Valongo area, Portugal. Cheiruridae, Pliomeridae, and
Dionididae. Bol. Soc. Geol. Port, xiv, 1-16, 4 pis.

delgado, J. F. n. 1892. Fauna Silurica de Portugal. Descripgao de uma forma nova de trilobite — Lichas
(Uralichas) Ribeiroi. (With French translation.) Comm. Trab. Geol. Port. Lisbon, 1-32, 6 pis.

— 1897. Fauna Silurica de Portugal. Novas observagoes acerca de Lichas ( Uralichas ) Ribeiroi. (With French
translation.) Dir. Trab. Geol. Port. Lisbon, 1-34, 4 pis.

— 1908. Systeme silurique du Portugal; etude de stratigraphie paleontologie. Mem. Commn geol. Port.
Lisbon, 1-245, 8 pis.

guillier, a. 1873. Faune seconde silurienne entre Saint-Denis-d’Orques et Chemire-en-Charnie. Bull. Soc.
Agric. Sci. Arts Sarthe, 21, 633-636.

hammann, w. 1971a. Die Placopariinae (Trilobita, Cheirurina; Ordovizium). Senck. Lethaea, 52, 53-75, 3 pis.
— 1971 b. Stratigraphische Einteilung des spanischen Ordoviziums nach Dalmanitacea und Cheirurina
(Trilobita). Colloque Ordovicien-Silurien Brest 1971. Mem. Bur. Rech. geol. minier. 73, 265-272, 1 pi.

— 1974. Phacopina und Cheirurina (Trilobita) aus dem Ordovizium von Spanien. Senck. Lethaea. 55, 1-150,
12 pis.

Harrington, H. j. 1959. In moore, R. c. (Editor). Treatise on Invertebrate Paleontology. Part O, Arthropoda 1.
i-xix, 560 pp. Geol. Soc. Amer. and Univ. Kansas Press.

616

PALAEONTOLOGY, VOLUME 23

havlicek, v. and marek, L. 1973. Bohemian Ordovician and its international correlation. Cas. Miner, geol. 18,
225-232.

— and vanek, J. 1966. The biostratigraphy of the Ordovician of Bohemia. Shorn, geol. ved., pal. 8,
7-68, 16 pis. (Czech summary, p. 69.)

henningsmoen, G. 1959. In moore, R. c. (Editor). Treatise on Invertebrate Paleontology. Part O, Arthropoda 1.

i-xix, 560 pp. Geol. Soc. Amer. and Univ. Kansas Press.
henry, j.-l. and clarkson, e. n. k. 1975. Enrollment and coaptations in some species of the Ordovician
trilobite genus Placoparia. Fossils Strata, 4, 87-95, 3 pis.

and romano, m. 1978. Le genre Dionide Barrande, 1847 (Trilobite) dans l’Ordovicien du Massif Armoricain

et du Portugal. Geobios, 11, 327-343, 2 pis.

lane, p. d. 1971. British Cheiruridae (Trilobita). Palaeontogr. Soc. ( Monogr .), 95 pp., 16 pis.
lindstrom, m., rachebouef, p. r. and henry, j.-l. 1974. Ordovician conodonts from the Postolonnec Forma-
tion (Crozon peninsula, Massif Armoricain) and their stratigraphic significance. Geol. et Palaeont. 8,
15-23, 2 pis.

prantl, f. and pribyl, a. 1948. Rostrideni nekterych ceskych Cheiruridu. (Trilobitae). (Classification of some
Bohemian Cheiruridae.) Sb. nar. Mus. Praze, ( B ) Geol. ( Paleont .), 1, 1-44, 6 pis.
pribyl, A. and vanek, j. 1969. Uber einige Trilobiten des mittelbohmischen Ordoviziums. Vestnik. Ustr. ust.
geol. 44, 365-374, 6 pis.

rachebouef, p. r. 1969. Generalites sur quelques trilobites des schistes Ordoviciens de la Mayenne. Bull.
Bayenne-Sci., 66-86, 6 pis.

reed, f. r. c. 1896. Notes on the evolution of the genus Cheirurus. Geol. Mag. (4) 3, 117-123, 161-167.
reid, c. 1907. Explanation of Sheet 353. The geology of the country around Mevagissey. Mem. geol. Surv. Eng.
Wales, vi + 73 pp. 7 pis.

romano, M. 1975. Harpid trilobites from the Ordovician of North Portugal. Comm. Serv. Geol. Port. 59,
27-36, 1 pi.

— 1976. The trilobite genus Placoparia from the Ordovician of the Valongo area, north Portugal. Geol.
Mag. 113(1), 11-28, 1 pi.

and diggens, J. N. 1973-1974. The stratigraphy and structure of Ordovician and associated rocks around

Valongo, north Portugal. Com. Serv. Geol. Port. (edit. 1976), 57, 23-50, 2 pis.
sadler, p. m. 1974. Trilobites from the Gorran Quartzites, Ordovician of south Cornwall. Palaeontology, 17,
71-93, pis. 9, 10.

Salter, J. w. 1853. Notes on the trilobites. (Appendix C to ‘On the Carboniferous and Silurian formations of
the neighbourhood of Bussaco in Portugal’ by C. Ribeiro.) Quart. J. Geol. Soc. 9, 158-60, 3 pis.
sharpe, d. 1849. On the geology of the neighbourhood of Oporto, including the Silurian coal and slates of
Vallongo. Ibid. 5, 142-153.

shaw, A. b. 1957. Quantitative trilobite studies. II. Measurement of the dorsal shell of non-agnostidean
trilobites. J. Paleont. 31, 193-207.

teixeira, c. 1955. Not as sobre geologia de Portugal. O Sistema Silurico. Lisboa.

temple, j. t. 1975. Standardization of trilobite orientation and measurement. Fossils Strata, 4, 461-467.
thadeu, D. 1947. Trilobites do silurico de Loredo (Bu5aco). Bol. Soc. Geol. Port. 6, 217-236, 3 pis.

1956. Note sur le Silurien Beiro-Durien. Ibid. 12, 1-38, 9 pis.

tromelin, G. and lebesconte, p. 1876. Essai d’un catalogue raisonne des fossiles silurien des departements de
Maine-et-Loire, de la Loire-Inferieure et du Morbihan, avec des observations sur les terrains paleozoiques
de l’ouest de la France. C.R. 4e Congr. Assoc, franc. Avancem. Sci. 601-661.
verneuil, p. e. p. and barrande, J. 1855. Descriptions des fossiles trouves dans les terrains silurien et devonien
d’Almaden, d’une partie de la Sierra Morena et des montagnes de Tolede. Bull. Soc. Geol. France, 12,

whittard, w. f. 1958. The Ordovician trilobites of the Shelve Inlier, West Shropshire. Palaeontogr. Soc.
{Monogr.), 71-116, pis. x-xv.

904-1025.

Typescript received 13 June 1979

Revised typescript received 22 November 1979

M. ROMANO

Department of Geology
Beaumont Building
University of Sheffield
Sheffield S3 7HF

THE MIOCENE HORSE HIPPARION FROM
NORTH AMERICA AND FROM THE TYPE
LOCALITY IN SOUTHERN FRANCE

by BRUCE J. MACFADDEN

Abstract. The three-toed horse Hipparion is diagnosed by the presence of a preorbital facial fossa that
anteriorly is poorly defined and posteriorly is moderately pocketed with a well-developed and continuous rim.
The concept of the genus Hipparion sensu s trie to ( s.s .) is presently restricted in the Old World to H. prostylum
from the genotypic locality at Mt. Leberon, France, and the species H. tehonense and H.forcei from New World
localities with a similar configuration of the preorbital facial fossa. It has previously been stated that, although
Hipparion was common in the Old World Neogene, this genus was very rare in equivalent-aged sediments in the
New World. Based on the concept of the genus presented here, Hipparion s.s. is found at numerous New World
localities. There apparently was a generic-level continuity of Hipparion s.s. that existed throughout Holarctica
during part of the Neogene. Hipparion horses ( sensu lato) appear to represent a polyphyletic assemblage of
several genera that arose independently from more than one merychippine ancestor during the Miocene. The
presence of hipparion horses in the New and Old Worlds probably resulted from more than one dispersal
event across Beringia.

For more than a century, the genus Hipparion has been used as a horizontal taxon, or ‘form genus’,
to include Holarctic Mio-Pliocene horses with isolated protocones in the upper molars, and tridactyl
limbs. The great geographic and geological abundance of this horse has made it biostratigraphically
very useful for Neogene intercontinental correlations. More than one hundred species of ‘Hipparion'
{sensu lato) have been named primarily on dental and postcranial characters. This large complex
of species is so unwieldy that, rather than comparing a new sample to all the existing species,
palaeontologists often propose new species out of despair and therefore perpetuate this taxonomic
problem.

In recent years, several studies have been presented that attempt to sort out some of the different
hipparion groups based principally on cranial morphology. Skinner and MacFadden ( 1 977) analysed
relatively large quarry samples from the North American mid-continent and showed that the
development of the preorbital facial fossa appears to be a taxonomically valid character complex at
the generic rank. In their study they proposed the genus Cormohipparion for hipparions with a
diagnostic preorbital (also termed nasomaxillary) facial fossa that is pocketed posteriorly and has
well-developed and continuous anterior and posterior rims. Skinner and MacFadden (1977) con-
centrated mostly on North American forms but also tentatively referred some Eurasian hipparions
to this genus. MacFadden and Bakr (1979) studied the Siwalik hipparions from the Indo-Pakistan
subcontinent and refer the large species theobaldi to the genus Cormohipparion. Woodburne and
Bernor (1980) studied numerous museum collections of Eurasian hipparions and proposed several
distinct groups, which probably represent separate lineages, based principally on their analysis of
cranial characters. There is general agreement among students of equid systematics that one or more
members of this polyphyletic hipparion assemblage arose in North America during the medial
Miocene. Subsequently, it appears that more than one hipparion group (i.e. a few genera) dispersed
into the Old World during the later Miocene. Many workers have suggested that the presence of
hipparions in the Old World resulted from the dispersal of one monophyletic group or ‘species’ of
‘ Hipparion ’ (e.g. Forsten 1968; Hussain 1971). Skinner and MacFadden (1977) suggested, based on
different cranial morphologies, that the dispersal of hipparions from the New to the Old World was
not monophyletic and probably involved several forms (or genera).

[Palaeontology, Vol. 23, Part 3, 1980, pp. 617-635.|

618

PALAEONTOLOGY, VOLUME 23

The concept of the genus Hipparion sensu stricto ( s.s .) is based on the species H. prostylum
described from the Turolian Mt. Leberon locality in southern France (de Christol 1832). One of the
important problems in the study of hipparion systematics has been recognition of the genus
Hipparion s.s. in North America. Gidley (1903) proposed the genus Neohipparion for most of the New
World species that had been previously included in the genus Hipparion, and Hipparion s.s. was
almost exclusively used for Old World forms. Osborn (1918) did not strictly follow Gidley’s
dichotomy between Neohipparion and Hipparion for New versus Old World forms, respectively. Since
the early studies, many workers believed that Hipparion s.s. was abundant in the Old World Miocene
and rare in the New World. Stirton (1940) stated that in North America Hipparion s.s. was
represented by only a few species distributed in California, Oregon, Washington, and Florida.

The purpose of this report is to describe Hipparion sensu stricto from several localities in North
America and to compare these samples with the material from the genotypic locality in southern
France. This study shows that Hipparion s.s. was more widely distributed in North America than has
been previously thought. Only the North American localities with well-preserved cranial material are
discussed here. Hipparion s.s. is undoubtedly present at numerous other localities in North America,
however, without relevant cranial material, it is difficult to distinguish these occurrences. It is not the
purpose of this paper to revise the taxonomy of all species of Hipparion and related forms, as that task
would certainly require a monograph. Therefore, the specific diagnoses and assignments essentially
rely on previous studies. The phylogenetic and palaeogeographic implications presented at the end of
the present study will focus on the recognition of a generic-level continuity of Hipparion s.s. through-
out Holarctica during the late Miocene.

The following institutional abbreviations are used in the text: AMNH, Department of Vertebrate
Paleontology, American Museum of Natural History, New York; BMNH, Department of Palaeonto-
logy, British Museum (Natural History), London; CIT, California Institute of Technology Collec-
tion, now housed at the Los Angeles County Museum of Natural History, Los Angeles; F:AM, Frick
American Mammals, The American Museum of Natural History, New York; MNHNP, Museum
National d’Histoire Naturelle, Institut de Paleontologie, 8 rue de Buffon, Paris 5, France; UCMP,
University of California Museum of Paleontology, Berkeley; UF, Florida State Museum, University
of Florida, Gainesville. The dental nomenclature follows Stirton (1940, 1941), Skinner and Taylor
(1967), and Skinner and MacFadden (1977).

SYSTEMATIC PALAEONTOLOGY

Class mammalia Linnaeus, 1758
Order perissodactyla Owen, 1848
Family equidae Gray, 1821
Genus hipparion de Christol, 1832

Text-figs. 1-14

Type status. When de Christol (1832) first proposed the genus Hipparion based on material from Mt. Leberon
in southern France (also called Mt. Luberon, Cucuron), no holotype was indicated. Later, Gervais (1849)
designated a syntypic series of Hipparion from Mt. Leberon, including H. prostylum, H. mesostylum, and
H. diplostylum. Osborn (1918) considered H. prostylum to be the type species for the genus Hipparion. Sondaar
(1974) stated that the holotype of H. prostylum, which consists of a fragmentary palate with P4-M2 (see Gervais
1849, pi. 19, fig. 2), is probably contained in the collections in the Musee Requien, Avignon.

Revised generic diagnosis. Medium-sized, mesocephalic, and moderately hypsodont tridactyl horses. Nasal
notch moderately developed and extends posteriorly to a position anterior to, or lying over, P2. Infra-
orbital foramen lies over P3. Preorbital facial fossa lies dorsal to P3-M* on the nasal and maxillary bones
well forward of the anterior rim of the orbit. The posterior portion of the fossa is usually developed on the
nasal and maxillary bones, anterior to the lacrimal. Anteriorly the fossa is poorly defined and is confluent with
the facial region. Posteriorly the fossa is moderately pocketed and has a well-developed and continuous rim.
There is no ventral fossa associated with the malar crest as is the case in some other horses. In the upper
cheek teeth the protocones vary from rounded to oval to lenticulate. There is a tendency for the protocone to

M ACFADDEN: MIOCENE HORSE HIPPARION

619

be connected to the protoloph in earlier wear stages than some other hipparions, e.g. Neohipparion. The hypo-
conal groove is moderately developed and is distinct to the base of the tooth. In the lower cheek teeth there is a
progressive deepening of the ectoflexids posteriorly. The metaconids and metastylids are widely separated. The
parastylid (also termed ectoparastylid or protostylid) is often developed and is either connected to the proto-
conid or is isolated. In both the upper and lower cheek teeth the enamel plications vary from simple to
moderately developed.

Distribution. Late Miocene (Clarendonian-?early Hemphillian) of North America, late Miocene-?Pliocene
(Vallesian-?Villafranchian) of Eurasia, and possibly Miocene-Pliocene of Africa. Note. The questionable ranges
listed here are taken from previous studies in which relevant cranial material is lacking. Therefore, it is difficult
to allocate certain Old World species to the genus Hipparion s.s.

Included species. At this point it is impossible to list all the species that should be included in Hipparion s.s.
(particularly in the Old World) because of the problems in recognition of this genus without cranial material. In
the present report H. tehonense and H. forcei are described from North America and these are compared to
H. prostylum from Europe.

Hipparion prostylum Gervais, 1849
Text-figs. 1-5, 13, 14

Selected synonymy

1849 Hipparion prostylum Gervais, pp. 284-285.

1873 Hipparion gracile Gaudry, pp. 32-42, pi. 5, figs. 7-10; pi. 6, figs. 1-11; pi. 7, fig. 1.

1956 Hipparion mediterraneum (in part), Pirlot, p. 28.

1968 Hipparion mediterraneus (in part), Forsten, pp. 40-53, 83-129, tables 12-15.

1974 Hipparion prostylum Sondaar, pp. 289-290, 296-299, 301-306, tables 2-4, pi. 46, figs. 1, 2; pi. 48,
figs. 2, 3, 8-10; pi. 49, figs. 3, 4, 8, 9, 10.

Type specimen. See generic discussion.

Specific diagnosis. Same as for the genus with the limitation that H. prostylum has rounded (and infrequently
oval) protocones in the upper molars. Sondaar (1974, p. 297, adapted from Gromova 1952) diagnoses
H. prostylum as follows: ‘Average size, length of the upper molar series P2-M2 123-145 mm. Enamel with
little foldings, slender footbones with relatively long metapodials.’ See discussion below.

Referred material. This description is based on the collections of H. prostylum housed in Paris (NMNHP) and
London (BMNH). These collections consist of four skulls, numerous dentitions, isolated teeth, and postcranials.

Distribution. Hipparion prostylum is recognized at the type locality, Mt. Leberon, which is of Turolian (late
Miocene) age. This species is also part of the ‘hipparionine Group 3’ complex of Woodburne and Bernor (1980).
Therefore, H. prostylum is probably represented at several other Old World localities of Turolian age listed in
that publication. Pending a revision of hipparions from other Old World localities, H. prostylum is presently
only known to occur for certain at the type locality, Mt. Leberon.

Description. Although H. prostylum has been described elsewhere (e.g. Gaudry 1873; Gromova 1952; Sondaar
1974) it is redescribed in this report in order to compare it to the North American representatives of this
genus.

The description of skull morphology is based on four specimens; NMNHP Luberon 156, NMNHP ‘un-
numbered’ (illustrated by Gaudry 1873, pi. 6, fig. 1, and Skinner and MacFadden 1977, text-fig. 3a), BMNH
M33603, and BMNH M26617 (three of these are illustrated in text-fig. 1).

The skull is mesocephalic and of moderate size. The premaxillary and nasal regions are preserved in one
specimen, BMNH M26617 (text-fig. lc). It is unfortunate that in BMNH M26617 the nasal region is covered
with matrix and therefore it is difficult to determine the posterior extent of the premaxillary bone and nasal
notch. However, the reconstructed nasal region in this specimen suggests a well-retracted nasal notch. In the four
skulls studied the buccinator fossa is either not preserved or it is covered with reconstructive material and
therefore nothing can be said about the development of this region.

620

PALAEONTOLOGY, VOLUME 23

text-fig. 1. Skulls of Hipparion prostylum from the late Turolian of Mt. Leberon, France, a, NMNHP
‘unnumbered’; b, NMNHP Lub. 156; c, BMNH M26617. Shading represents reconstruction or matrix.

M ACFADDEN: MIOCENE HORSE HIPPARION

621

0

2

3

4

5cm

text-fig. 2. Deciduous upper cheek teeth (right dP2-dP4) of Hipparion prostylum, NMNHP Lub. 94, from the
late Turolian of Mt. Leberon, France.

The preorbital facial fossa lies on the dorsal half of the cheek region. Anteriorly the fossa is poorly defined
and it is confluent with the adjoining facial region. Posteriorly this fossa is usually moderately pocketed and has
a well-developed continuous rim. The fossa lies in front of the lacrimal bone (as preserved in BMNH M26617,
text-fig. 1) and well forward of the orbit. Postero-ventral to the nasomaxillary fossa is a moderately developed
malar crest. There is no fossa associated with the malar crest as is the case in some other horses (e.g.
Pliohippus). The teeth are moderately hypsodont, slightly curved, and covered with cement.

The upper incisors have cement-filled infundibula (cups). The precanine diastema is smaller than the post-
canine diastema. DP2~dP4 are more rectangular in cross-section than the corresponding P2-P4 (text-fig. 2).
The deciduous premolars are similar in dental pattern to the corresponding permanent premolars. In particular,
the fossettes are moderately plicated, the protocones are usually rounded, and there is a tendency for the
protocone of the dP2 and P2 to become connected to the protoloph during relatively early wear stages.

In the permanent upper dentition the protocone is isolated from the protoloph until late wear stages (except
in the P2 as noted above) when these two structures frequently connect. The protocone is characteristically
rounded but infrequently varies to oval or lenticulate in shape with anterior and posterior spurs (text-fig. 3).
The hypoconal groove is relatively well developed until late wear stages. The enamel plications are simple to
moderately well developed. The posterior border of the anterior lake (prefossette) and the anterior border of the
posterior lake (postfossette) show the most complexity of plications within a given tooth or tooth row. As in
North American hipparions, the anterior border of the prefossette and posterior border of the postfossette
lack complex foldings. The plicaballin consists of either a single or double loop.

text-fig. 3. Permanent right upper cheek teeth (P2-M3) of Hipparion prostylum, BMNH 27590, from the late
Turolian of Mt. Leberon, France.

0 12 3 4 5cm

622

PALAEONTOLOGY, VOLUME 23

text-fig. 4. Deciduous lower cheek teeth of Hipparion prostylum from the late Turolian of Mt. Leberon, France.
a, NMNHP Lub. 14, right dP2-dP4; b, NMNHP Lub. 26, left dP2-dP4.

The lower incisors have cement-filled infundibula. The precanine diastema is very small and the canine is
nearly appressed to the I3. The postcanine diastema is moderate in length, with the mental foramen situated
approximately midway between the C and P2. The premolars are larger in cross-section than the molars. As
exemplified by NMNHP Luberon 14 and 26 (text-fig. 4), the lower deciduous teeth are similar to the
permanent premolars in dental pattern. In the anterior region of dP2 and P2 there is a moderately developed
anterior projection of the paralophid-parastylid complex characteristic of hyposodont horses. The P2 through
M3 are generally similar in dental pattern except as noted below (text-fig. 5). There is a well-developed
parastylid on the antero-external portion of the cheek tooth. This structure is similar to that seen in some
other hipparions, e.g. Cormohipparion. The metaconids and metastylids are well separated and vary from equal
to subequal in size. The entoconid is significantly larger than the hypoconulid. On the M3 the posterior
portion of the tooth is expanded to form a projection of the hypoconulid or ‘heel’. The protoconids and hypo-
conids are crescentic. In contrast to e.g. Neohipparion eurystyle and Pleistocene hipparions from Africa, the
ectoflexid is moderately developed in the premolars. In the molars the deep ectoflexid almost separates the
metaconid and metastylid. The plicaballinid and other enamel plications are usually absent or infrequently they
are poorly developed.

0 12 3 4 5

cm

text-fig. 5. Permanent left lower cheek teeth (P2-M2) of Hipparion prostylum, NMNHP Lub. 40, from the late
Turolian of Mt. Leberon, France.

M ACFADDEN: MIOCENE HORSE HIPPARION

623

The metapodials of H. prostylum from Mt. Leberon are of moderate size relative to other Eurasian
hipparions. Sondaar (1974) studied the metapodials of H. prostylum and concluded that this species was
smaller than the slender form from Pikermi, H. gracile. As is the case in Eurasian hipparions of Turolian
age, H. prostylum usually has a well-developed ectocuneiform facet on the MT III (Sondaar 1974, Sondaar, pers.
comm. 1979).

Discussion. Woodburne and Bernor (1980) and Woodburne (pers. comm. 1980) suggest that two
forms of hipparions are represented at Mt. Leberon. This assertion is based on the fact that, besides
the facial morphotype described as Hipparion s.s., Pirlot (1956) described one skull from the BMNH
collection that had a well-developed preorbital facial fossa. From his description, one might be
concerned that this skull possibly represented Cormohipparion. If that were true, then the validity and
proper assignment of the species prostylum to Hipparion would be questionable. Pirlot (1956)
unfortunately did not refer to the skull in question by its catalog number. I have studied the BMNH
collection, and unless this skull has been lost, it seems almost certain that based on Pirlot’s
description, he was referring to BMNH M26617 (text-fig. lc). It is not necessary to refer this skull
to another taxon besides H. prostylum because BMNH M26617 appears to be the same facial
morphotype as the other cranial specimens from Mt. Leberon.

Hipparion tehonense (Merriam 1916), new combination
Text-figs. 6-8, 13, 14

Selected synonymy

1907 ? Hipparion lenticularis (in part), Gidley, pp. 915-917. Synonymy restricted to Clarendonian
sample from Texas Panhandle.

1918 Hipparion lenticulare (in part), Osborn, pp. 184-185, text-figs. 147, 148; pi. 32, fig. 2; pi. 33,
figs. 5-7. Synonymy restricted to Clarendonian sample from Texas Panhandle.

1916 Neohipparion gratum tehonense, Merriam, pp. 118-120, text-figs. 1-7.

1918 Hipparion lenticulare Osborn, pp. 184-185, text-figs. 147, 148; pi. 32, fig. 2; pi. 33, figs. 5-7.
1939 Nannippus tehonensis Stirton, pp. 347-352, text-figs. 13, 24.

1942 Nannippus tehonensis Drescher, pp. 11-15, text-fig. 3.

1969 Nannippus tehonensis Webb, pp. 130-135.

Type specimen and locality. UCMP 21780, right upper M1?, described by Merriam (1916, p. 1 19, fig. 1), Chanac
(‘Santa Margarita’) Formation, south Tejon Hills, California, early Clarendonian.

Diagnosis. Characters same as for other species of the genus Hipparion s.s. In particular, the preorbital facial
fossa is well developed posteriorly, but anteriorly becomes poorly defined (text-figs. 6, 7). The nasal notch is
retracted to a position that lies above P2. In addition, H. tehonense is characterized by very simple enamel
plications and the anterior region of the P2 is not as well developed as some other Hipparion s.s.

Referred material. H. tehonense from the California localities is represented by numerous specimens in the
UCMP and CIT collections (see, e.g., Merriam 1916 and Drescher 1942). The Texas occurrence of this species
is represented by F:AM 74400-74585 and also numerous uncatalogued F:AM specimens from MacAdams
Quarry (locality 17), collected by the Frick Laboratory between 1934-1960, Donley County, Texas Panhandle
and also specimens from other localities in Donley County, e.g. AMNH 10854 (see Osborn 1918, pi. 32, fig. 2).

Distribution. Besides the type locality, H. tehonense is also known from the Orinda Formation, early Claren-
donian, San Francisco Bay Area, California, and the ‘Clarendon Beds’, Ogallala Group, early Clarendonian,
Donley County, Texas.

Description. In most characters, H. tehonense is similar to H. prostylum. Only those characters that show
certain important similarities and differences between H. tehonense and H. prostylum or characters not repre-
sented in the hypodigm of H. prostylum will be discussed here.

The description of skull morphology of H. tehonensis is based on a large sample from MacAdams Quarry,
as exemplified by F:AM 74478 (text-fig. 6a), F:AM 74537 (text-fig. 7a), and AMNH 10854 (‘neotype’ of
H. ‘ lenticulare ’, see Osborn 1918, pi. 32, fig. 2) from the ‘Clarendon Beds’ of the Texas Panhandle. The skull
of H. tehonensis is small relative to other species of Hipparion s.s.

624

PALAEONTOLOGY, VOLUME 23

The premaxilla extends postero-dorsally to above the P2-P3. The nasal notch, which lies above P2, is well
retracted in contrast to other hipparions such as Neohipparion whitneyi (see Osborn 1918, pi. 32, fig. 1) but
certainly less retracted than e.g. proboscideum (see Sondaar 1971, pi. III).

The infraorbital canal lies above P3. As seen in H. prostylum, the preorbital facial fossa lies on the dorsal
half of the facial region. Anteriorly, the fossa is poorly defined and it is confluent with the adjoining facial
region. Posteriorly, this fossa is usually moderately pocketed and has a well-developed continuous rim. The fossa
lies well forward of the lacrimal bone and orbit. As evidenced by the MacAdams Quarry sample, there is no
significant morphological change in the preorbital facial fossa during ontogeny (compare text-figs. 6a and 7a).

The dentition of H. tehonense is similar in pattern to other species of this genus. The enamel plications are
very simple relative to other hipparions. The protocones are rounded to oval and these structures tend to
become connected to the protoloph during later wear stages, particularly in the P2. There are well-developed
parastylids, and the ectoflexids are deep with few, if any, plicaballinids (text-figs. 6b, 7b, and 8).

Discussion. The large sample from MacAdams Quarry is assigned to H. tehonense as defined by the
topotypic material from the Tejon Hills based on the following distinctive characters; (1) small size
relative to other Hipparion s.s., (2) extreme simplicity of the enamel plications, (3) a poorly developed
anterior extension of the parastyle on P2, and (4) similar degree of hypsodonty.

Because of its distinctively small size, the species H. tehonense from California has in the past been
assigned to two different taxa of small hipparions. Merriam (1916) originally named the topotypic
material from Tejon Hills a subspecies of the tiny Pseudhipparion gratum. Subsequent workers have
assigned tehonensis to Nannippus, a genus of dubious monophyletic significance. Skinner and
Hibbard (1972, p. 117) stated that: ‘The practice of assigning all small forms of Hipparion- like

text-fig. 6. Adult specimen of Hipparion tehonense , F:AM 74478, from the Frick MacAdams Quarry, early
Clarendonian of the Texas Panhandle, a, left lateral view of skull; B, occlusal view of left upper dentition.
Shading represents reconstruction or matrix.

MACFADDEN: MIOCENE HORSE HIPPARION

625

B

text-fig. 7. Immature specimen of Hipparion tehonense, F:AM 74537, from the Frick MacAdams Quarry, early
Clarendonian of the Texas Panhandle, a, right lateral view of skull; b, occlusal view of dP'-dP4. Shading
represents reconstruction or matrix.

equids to Nannippus without careful consideration of other characters clouds the relationship of
many of the dwarf forms and prevents the recognition of true Nannippus. For example, Griphippus
[= Pseudhipparion ] gratus, which has quite different skull, dental, and postcranial characters, has
often been assigned to Nannippus.'

Although there are no skulls preserved for the Californian sample of H. tehonense , the MacAdams
Quarry specimens clearly demonstrate a similarity in facial morphology with Hipparion s.s.
MacFadden and Waldrop (1980) described the facial morphology of N. phlegon from Mt. Blanco in
the Texas Panhandle, which is the genotypic locality and therefore central to the concept of that
genus. N. phlegon has a smooth preorbital cheek region with no facial fossa. Therefore, there is no
doubt that the small hipparion species tehonense is best referred to the genus Hipparion s.s.

0 1

4

5 cm

text-fig. 8. Right lower cheek teeth (P2-M3) of Hipparion tehonense, F:AM 105440, from the Frick
MacAdams Quarry, early Clarendonian of the Texas Panhandle.

626

PALAEONTOLOGY, VOLUME 23

lenticularis, as it is used for Clarendonian hipparions from Donley County, Texas, is
synonymized here with H. tehonense. The species lH.' lenticularis has been inconsistently used in the
literature and it is appropriate to comment on its nomenclature here. In 1893 Cope assigned the
species lenticularis to Protohippus based on material of late Hemphillian age from Mulberry Canyon,
near Goodnight, in the Texas Panhandle (see Schultz 1977). Gidley (1907) referred material from
the Clarendon beds of Donley County in the Texas Panhandle to H. lenticularis. Osborn (1918)
designated a well-preserved skull (also described previously by Gidley 1907), AMNH 10584, as the
neotype of H. lenticularis. This judgement was apparently made by Osborn because the early workers
thought that the Clarendon and Goodnight beds were correlative and the topotypic material from
Mulberry Canyon was not abundant. Despite these previous taxonomic decisions, it remains to be
demonstrated that ‘Hf lenticularis from Donley County is conspecific with the material from
Mulberry Canyon and numerous other late Hemphillian localities, e.g. Coffee Ranch (Matthew and
Stirton (1930). It is unfortunate that no skulls are known of late Hemphillian lenticularis. The
Clarendonian H. tehonense and Hemphillian lenticularis are remarkably similar in dental
pattern, however, the younger species is noticeably more hypsodont. In this report the species
lenticularis is restricted to the late Hemphillian forms. Based on dental and temporal similarities,
the Clarendonian lenticularis as used by workers such as Gidley and Osborn is synonymized with
H. tehonense.

Hipparion forcei Richey 1948
Text-figs. 9, 13, 14

Selected synonymy

1919 Hipparion mohavense Merriam, pp. 549-553, text-figs. 163-170.

1948 Hipparion forcei Richey, pp. 9-25, text-figs. 4-12, pi. 2, figs, a-c; pi. 3, figs. a-d.

1969 Nannippus forcei Webb, pp. 130-135.

Type specimen and locality. UCMP 33051, P3, from Green Valley Formation, Black Hawk Ranch Quarry,
Mount Diablo area, California, late Clarendonian.

Diagnosis. Characters same as for other species of the genus Hipparion s.s., in particular, configuration of the
preorbital facial fossa and nasal region listed above for H. prostylum and H. tehonense. Specific characters for
H. forcei include an apparently higher frequency of connection of the protocone to the protoloph in P2
(Richey 1948). Also the protocone-protoloph connection is very well developed with less of a constriction
between these parts than is seen in many other hipparions. The protocone is smaller in H. forcei than in
H. tehonense relative to the occlusal area of the tooth. H. forcei has higher crowned cheek teeth with larger
occlusal cross-sectional areas than in H. tehonense (Webb 1969).

Referred material. Numerous UCMP specimens from the Black Hawk Ranch Local Fauna, Green Valley
Formation, San Francisco Bay region, California, and the Dove Springs Fauna, Ricardo Formation, Mohave
Desert, California (see Richey 1948).

Distribution. Besides the type locality, H. forcei is probably represented in the Ricardo Formation (Dove
Springs Fauna), Mohave Desert, California. These localities are late Clarendonian in age (Tedford et al., in
press).

Description. In most characters the material of H. forcei is similar to other species of this genus, including
H. prostylum and H. tehonense. The cranial morphology of H. forcei is known from one crushed but relatively
complete skull from Black Hawk Ranch (text-fig. 9a), UCMP 34511, originally described in detail by Richey
(1948). The important characters that are similar among these species include a relatively well-developed nasal
notch that is retracted to a position that lies over P2. The infraorbital canal lies above P3. As is diagnostic of
Hipparion s.s., the preorbital facial fossa is poorly defined anteriorly but posteriorly it is characterized by a well-
developed continuous rim that is pocketed. The fossa lies well forward of the lacrimal bone and orbit.

The most complete dentition of H. forcei is known from the skull, UCMP 34511 (text-fig. 9b). However, the
dental pattern in this specimen is not characteristic because it represents an old individual in late wear stage.
There are numerous isolated teeth known from the type locality and Richey (1948) described them in detail.

MACFADDEN: MIOCENE HORSE HIPPARION

627

The following characters are diagnostic of H. forcer, relatively simple enamel plications, small protocone, high
frequency of protocone-protoloph connection in the P2, and lowers with deep ectoflexids but without
plicaballinids. Richey (1948, p. 15) stated that: ‘Another character that distinguishes H.forcei from many other
species is the frequency of connection of the protocone with the protoconule [protoloph], Many hipparions
have a connected protocone in the P2. This is particularly true of H. forcei. In fact, in none of the specimens thus
far studied is the protocone separate.’

Richey (1948) studied the limbs of H. forcei and concluded that they were of moderate size in contrast to
smaller forms such as Nannippus and larger, more robust, forms such as ‘ H .’ (= Cormohipparion) theobaldi
from the Siwaliks and H. gracile from Pikermi.

text-fig. 9. Hipparion forcei, UF 22656 (cast of UCMP 3451 1) from the late Clarendonian Black Hawk Ranch
Local Fauna, California. A, left lateral view of skull; b, occlusal view of right upper cheek teeth.

Discussion. H.forcei and H. tehonense are very similar in many characters. Webb (1969) has suggested
an ancestral-descendent relationship between these two species. The samples from Tejon Hills-
Chanac Formation-Black Hawk Ranch appear to approximate a morphocline in characters such as
hypsodonty. However, in other characters such as the high frequency of protocone-protoloph con-
nection, H.forcei seems more primitive than H. tehonense. The relative primitiveness of certain dental
characters in H.forcei would, as Richey (1948) suggested, imply independent evolution in parallel of
H. forcei and H. tehonense from a common ancestor rather than a single ancestral-descendent
sequence as suggested by Webb (1969). It is not within the scope of this paper to resolve the phylo-
genetic relationships of the species H. tehonense and H.forcei. This short note is included as an intro-
duction to the next section below, i.e. the provisional assignment of forms from the mid-continent
of North America to H. cf. tehonense or forcei.

628

PALAEONTOLOGY, VOLUME 23

Hipparion cf. tehonense or forcei
Text-figs. 10-14

Referred material. Numerous specimens in the F:AM collection including well-preserved skulls, e.g. F:AM
107664, from Trail Side Kat Quarry Channel, Cherry County, Nebraska, late Clarendonian; F:AM 107663,
Rosebud Agency Quarry, Todd County, South Dakota, late Clarendonian; F:AM 71887, Olcott Quarry,
Hipparion Channel, Olcott Hill, Sioux County, Nebraska, late Clarendonian.

Distribution. Snake Creek and Ash Hollow Formations, Ogallala Group, north-central Nebraska and adjacent
South Dakota, and north-western Nebraska, late Clarendonian (see Skinner et al. 1977; Tedford et al. in press).

Description. Well-preserved cranial material from the northern Great Plains localities are referred to Hipparion
s.s. based on the configuration of certain skull characters, particularly the preorbital facial fossa.

There appears to be a significant size difference among the individuals of H. cf. tehonense or forcei. In
F:AM 107664 (text-fig. 10) and F:AM 107663 (text-fig. 11) the premaxillary extends posteriorly to a position
that lies over the P2. There is some variation in the posterior extent of the nasal notch. In F:AM 107664
and F:AM 71887 (text-fig. 12) the nasal notch extends to a position that lies over the buccinator fossa, which
is slightly less retracted than in other skulls of Hipparion s.s. described here. Although the nasal bones are not
preserved in F:AM 107664, the nasal notch appears retracted to a position over P2 similar to that seen in other
skulls of Hipparion s.s. In the skulls illustrated in text-figs. 10-12 the infraorbital foramen lies above the P3
just ventral to the antero-ventral margin of the preorbital facial fossa. This fossa is poorly defined anteriorly but
posteriorly it consists of a well-defined continuous rim. Posteriorly there also is a moderately well-developed
pocket. This fossa lies well forward of the lacrimal bone and orbit. There is a moderately developed malar crest.

The dentitions are similar to other species of Hipparion s.s. In particular, the enamel plications are relatively
simple. The protocone is oval and relatively large. In the P2 of F:AM 107663 and F:AM 107664 the protocone
is strongly connected to the protoloph.

text-fig. 10. Hipparion cf. tehonense or forcei, F:AM 107664, from the late Clarendonian Trail Side Kat Quarry
Channel, Nebraska, a, left lateral view of skull; B, occlusal view of left upper cheek teeth.

M ACFADDEN: MIOCENE HORSE HIPPARION

629

B

text-fig. 11. Hipparion cf. tehonense or forcei, F:AM 107663, from the late Clarendonian Rosebud Agency
Quarry, South Dakota, a, left lateral view of skull; b, occlusal view of left upper cheek teeth.

text-fig. 12. Hipparion cf. tehonense or forcei, F:AM 71887, from the late Clarendonian Olcott Quarry,
Hipparion Channel, Olcott Hill, Nebraska, a, left lateral view of skull; b, occlusal view of left upper cheek teeth.

630

PALAEONTOLOGY, VOLUME 23

Discussion. The configuration of the skull, particularly in the development of the preorbital facial
fossa, justifies the allocation of the material from these mid-continental sites to Hipparion s.s. How-
ever, the specific allocation is, at this point, somewhat uncertain. It is not implied that the sample
from these three localities represents one discrete species. For example, the smaller size of F:AM
107664 and F:AM 107663 possibly indicates an affinity with H. tehonense, whereas the larger size of
F:AM 71887 possibly indicates an affinity with H.forcei (following Webb 1969). On the other hand,
the very strong connection of the protocone and protoloph in both F:AM 107663 and F:AM
107664 indicates an affinity with H. forcei (following Richey 1948). The resolution of this species-
level problem would require further study beyond the scope of the present paper. The important
point is that this mid-continental sample is referred to Hipparion s.s. Therefore, this genus was
relatively widespread in North America during the Clarendonian.

BIOSTRATIGRAPHY AND PALAEOBIOGEOGR APH Y

The temporal and geographic distribution of Hipparion s.s. from North America and Mt. Leberon is
summarized in text-figs. 13 and 14. In these text-figs, several other Eurasian localities of Hipparion
s.s. have been added, based on recent studies of cranial morphology (MacFadden and Bakr 1979,
Woodburne and Bernor 1980). Undoubtedly Hipparion s.s. (as recognized by cranial morphology)
occurs at other Holarctic localities and possibly also in Africa. For this discussion only the selected
localities shown in text- figs. 13 and 14 will be presented.

The radiometric time scale and European Stages in text-fig. 14 were taken from several works,
including Berggren and Van Couvering (1974), Aguirre (1975), Fahlbusch (1976), and Van
Couvering and Berggren (1977). The age of the ‘Hipparion Datum Plane’ is shown to range from
about 12 0 mybp to about 10-8 mybp. This range is a result of alternative interpretations of radio-
metric dating of critical Old World sites, particularly Howenegg (e.g. Berggren and Van Couvering
1974; Van Couvering and Berggren 1977; Becker-Platen et at. 1977; Barndt et al. 1978, also see
discussion in MacFadden and Bakr 1979). The boundary between the Astarcian ( sensu Fahlbusch
1976) and Vallesian, which is usually taken as the first appearance of ‘Hipparion’, is dashed in
text-fig. 14 in order to indicate the uncertainty involved in the calibration of this event. The calibra-
tion and nomenclature of the North American localities is taken from Tedford et al. (in press).

Sondaar (1974) and Sen et al. (1978) consider the Mt. Leberon locality, and, therefore, the type
material of Hipparion s.s. to be of late Turolian age. Woodburne and Bernor (1980) studied facial
morphotypes from selected Old World localities. Their ‘Group 3’ consists of a morphologically
distinct group including H. prostylum from Mt. Leberon and forms from several localities in Greece
and Iran. This facial morphotype agrees with the concept of the genus presented in this report for
Hipparion s.s. The Hipparion s.s. from Saloniki, Greece (Group 3 of Woodburne and Bernor 1980,
here referred to as H. ‘ prostylum ’), and Pikermi, Greece (also Group 3, here referred to as H. ‘ gracile ’),
are considered to be of medial Turolian age and slightly older than Mt. Leberon (Berggren and Van
Couvering, 1974; Sen et al. 1978).

For a long time it was thought that at Samos, Quarries 1-4 were older than Quarry 5, and that this
succession spanned medial to late Turolian time (e.g. Berggren and Van Couvering 1974; Aguirre
1975; Sen et al. 1978). Recent field work at Samos (Solunias, pers. comm. 1977), suggests that all the
quarries are approximately contemporaneous. Therefore, depending upon the stratigraphic inter-
pretation, the Hipparion s.s. (Group 3, here referred to as H. ‘ dietrichC ) from Samos either spans
medial to late Turolian time or is restricted to the late Turolian. Woodburne and Bernor (1980) state
that Hipparion Group 3 (here referred to H. sp.) is found in the middle and upper parts of the
Maragheh, Iran, sequence. Based on this range, the Maragheh Hipparion s.s. spans medial to late
Turolian time.

The Siwalik hipparions of Pakistan and adjacent India have been the subject of numerous publica-
tions because of their association with a very rich Neogene sequence including the oldest-known
hominoid fossils (Pilbeam et al. 1977). Hussain (1971) presented the most recent revision of Siwalik
hipparions. MacFadden and Bakr (1979) recognize two, or perhaps three, supraspecific taxa of

MACFADDEN: MIOCENE HORSE HIPPARION

631

MacFadden and Bakr (1979) and Woodburne and Bernor (1980). North American localities (7-13) are taken from this report. The locality
numbers in this text-fig. correspond to the numbers of the columns in text-fig. 14.

v

SELECTED HOLARCTIC LOCALITIES OF HIPPARION S- S.

EURASIA

NORTH AMERICA

*S»

ass

tea

a

cSS±

*

F

.ffiSn

13

/ats

i

c

E ^
>

.

rn

-12

A«

t

\

1

JL.

-

—

2

JL

i

i

s

1

\

632

PALAEONTOLOGY, VOLUME 23

z <->?*<

— — i— i — <Z u-^<aiuZQOZ-<Z <°^

CD

SELECTED HOLARCTIC LOCALITIES OF HIPPARION S. S.

NORTH AMERICA

13

Southern
S. Dakota

H.

cf.

tehonense

or

forcei

12

North-

Central

Nebraska

H.

cf.

tehonense

or

forcei

11

Sioux

County,

Nebraska

H.

cf.

tehonense

or

forcei

10

Clarendon

Beds,

Texas

H.

tehonense

9

S.Tejon

Hills,

California

H.

tehonense

8

Black Hawk
Ranch,
California

H.

forcei

7

Orinda,

California

H.

tehonense

EURASIA

6

Siwaliks,

Pakistan

4

i

H.

cf.

antilo-

pinum

1

1

?

i

l

t

5

Maragheh,
1 ran

H.

sp-

NE

4

Samos,

Greece

H.

"dietrichi"

7

\

SIAN

UUM PLA

3

Pikermi,

Greece

H.

"gracile"

EURA
’ARION D/

2

Saloniki,

Greece

'E a-

x

1

Mt.

Leberon,

France

H.

prostylum

T

M

E

M.

Y.

B.

P.

O IV OO O' 2 — —

till II

. </>i— <Ouj</>

lU^OC

ttIUlOU •- Z>£*0— ' — <Z

uja-OUI

M

O

c

E

N

E

text-fig. 14. Temporal and geographic distribution of the Holarctic Hipparion s.s. localities shown in text-fig. 13. Epoch, Stage, North
American Land Mammal ‘Ages’ (NALMA), and time (mybp) calibrations are taken from numerous references cited in the text. Dashed zone
between Astarcian and Vallesian European Stages indicates the uncertainty involved in the calibration of the Hipparion Datum Plane. The
arrows in columns 4 and 6 indicate questionable ranges depending upon stratigraphic interpretations (for Samos) and lack of well-preserved

cranial material (for the Siwaliks). See discussion in text.

M ACFADDEN: MIOCENE HORSE HIPPARION

633

Siwalik hipparions. Some specimens of their ‘small hipparion complex’ (here termed H. cf. anti-
lopinum) are tentatively referred to Hipparion s.s. Based on teeth, hipparions are known to range in
the Siwaliks from the early Vallesian (roughly 10-5 mybp following Barndt et al. 1978) to the early
Villafranchian (roughly 3 0 mybp following Keller et al. 1977). It is impossible at present to deter-
mine the exact range of H. cf. antilopinum because the relevant cranial material either has poor strati-
graphic data (particularly from the early collections) or is limited to the upper Dhok Pathan
Formation, which is probably late Turolian in age.

The stratigraphic range in North America of Hipparion s.s. as recognized by cranial morphology
is from early to late Clarendonian. The individual ages of each locality are represented in text-
figs. 13 and 14.

There are several important conclusions based on the present study of Hipparion s.s. First, in
contrast to the hypotheses of earlier workers, Hipparion s.s. was widespread in North America as well
as Eurasia during the Miocene. The stratigraphic distribution (text-fig. 14) of the several species of
Hipparion s.s. demonstrates a generic-level continuity throughout Holarctica during the medial to
late Turolian and early to late Clarendonian. The slightly older occurrences of H. tehonense in North
America may be significant in a phylogenetic context depending upon the accuracy of the inter-
continental correlations. If H. tehonense is older than the other Hipparion s.s., at present there is no
implication of primitiveness or ancestry for this species. The interspecific relationships of the species
assigned to Hipparion s.s. require a detailed analysis beyond the scope of this paper.

Another interesting conclusion, based on the limited number of localities and relevant cranial
material discussed here, is that Eurasian Hipparion s.s. appears to be common in Turolian age
localities but is not recognized from the Vallesian. Therefore, it appears that Hipparion s.s. was not
involved in the Eurasian ‘Hipparion Datum Plane’ that defines the base of the Vallesian. MacFadden
and Skinner (1977) and Skinner and MacFadden (1977) suggested that the presence of hipparions in
the Old World could have been a result of more than one dispersal event rather than only one event
as has been suggested by some other workers (e.g. Forsten 1968; Hussain 1971).

As was stated in the Introduction, the species-level taxonomy of Holarctic Hipparion s.s. needs to
be revised in light of cranial morphology. Based on numerous cranial characters it appears that
Hipparion s.s. is composed of a monophyletic group of several species. It is important to determine
the ancestral stock from which the Hipparion s.s. species were descended. It is clear that the closest
relative of Hipparion s.s. is within the merychippine horses (e.g. Matthew 1924; Colbert 1935; Stirton
1940; Forsten 1968; Skinner and MacFadden 1977). A study is needed that demonstrates the
relatedness of several hipparion groups that apparently arose independently from the horizontal
merychippine complex. A striking consequence of recent studies of hipparion cranial morphology is
that these three-toed horses are certainly polyphyletic and arose from more than one merychippine
lineage. In short, several distinct supraspecific groups of hipparions originated independently and
evolved in parallel during the Neogene in the Old and New Worlds.

Acknowledgements. Special thanks are extended to Drs. Morris F. Skinner and Richard H. Tedford of the
American Museum of Natural History for their expertise and guidance during my studies of equid palaeontology.
I also thank Dr. Tedford for allowing some illustrations drawn at the American Museum of Natural History
to be used in this study. Anthony J. Sutcliffe, Jeremy Hooker, and Andrew Currant greatly aided in my studies
of the hipparions at the British Museum (Natural History). Many other colleagues have contributed to the
present study, including Vera Eisenmann, Paul Y. Sondaar, S. David Webb, Ronald G. Wolff, and Michael
O. Woodburne. Nancy Halliday and Lauren Keswick skilfully prepared many of the illustrations. The
manuscript was typed by Mrs. S. Sidaway. This report is University of Florida Contribution to Vertebrate
Paleontology number 179.

634

PALAEONTOLOGY, VOLUME 23

REFERENCES

aguirre, e. 1975. Division estratigrafica del Neogeno continental. Estudios Geol. 31, 587-595.
barndt, j. et al. 1978. The magnetic polarity stratigraphy and age of the Siwalik Group near Dhok Pathan
village, Potwar Plateau, Pakistan. Earth Planet. Sci. Letters, 41, 355-364.
becker-platen, ]. d., benda, l. and steffans, p. 1977. Litho- und biostratigraphische Deutung radiometrischer
Altersbestimmungen aus dem Jungtertiar de Tiirkei (Kanozoikum und Braunkohlen der Tiirkei, 18). Geol.
Jb. B25, 139-167.

berggren, w. a. and van couvering, J. a. 1974. The late Neogene: Biostratigraphy, geochronology, and
paleoclimatology of the last 15 million years in marine and continental sequences. Palaeogeogr.,
Palaeoclimat., Palaeoecol. 16, 1-216.

de christol, j. 1832. [No title: Description of Hipparion .] Sci. lnd. Ann. Midi, France, 1, 180-181.

Colbert, e. h. 1935. Siwalik mammals in the American Museum of Natural History. Trans. Amer. Phil. Soc.
N.s. 26, 1-401.

cope, E. D. 1893. A preliminary report on the vertebrate paleontology of the Llano Estacadao. Texas Geol.
Survey, 4th Ann. Report. Pp. 1-137.

drescher, a. b. 1942. Later Tertiary Equidae from the Tejon Hills, California. Contrib. Paleont., Carnegie
Inst., Washington, 530, 1-23.

fahlbusch, v. 1976. Report on the International Symposium on mammalian stratigraphy of the European
Tertiary. Newsl. Strat. 5, 160-167.

forsten, a.-m. 1968. Revision of Palearctic Hipparion. Acta Zool. Fennica, 119, 1-134.
gaudry, a. 1873. Animaux fossiles du Mont Leberon ( Vaucluse) : Etude sur les vertebres. F. Savy (ed.), Libr. Soc.
Geol. France. Pp. 1-112.

gervais, p. 1949. Note sur la multiplicity des especes d’hipparion (genre de chevaux a trois droits) qui sont
enfouis a Cucuron {Vaucluse). Compt. Rend. Acad. Sci., Paris, 29, 284-286.
gidley, j. w. 1903. A new three-toed horse. Bull. Amer. Mus. Nat. Hist. 19, 465-476.

— 1907. Revision of the Miocene and Pliocene Equidae of North America. Ibid. 23, 865-934.
gray, j. e. 1821. On the natural arrangement of vertebrose animals. London Med. Repository Rev. 15, 296-310.
Gromova, v. 1955. Le genre Hipparion. Inst. Paleont. Acad. Sci. URSS 36. [Original in Russian.] French transla-
tion by St. Aubin, Bur. Rech. Min. Geol. Ann. C.E.D.P. 12, 1-473.
hussain, s. t. 1971. Revision of Hipparion (Equidae, Mammalia) from the Siwalik Hills of Pakistan and India.
Verlag. Bayer. Akad. Wiss. N.s. 147, 1-68.

keller, H. m. et al. 1977. Magnetic polarity stratigraphy of the Upper Siwalik deposits, Pabbi Hills, Pakistan.
Earth Planet. Sci. Letters , 36, 187-201.

linnaeus, c. 1758. Sy sterna Naturae per Regna Tria Naturae, Secundum Classes, Or dines. Genera, Species cum
Characteribus Differentiis Synonymis Locis. Editia decima, reformata. Laurentii, Stockholm, Salvi, 1, 1-824.
macfadden, b. J. and bakr, A. 1979. The horse Cormohipparion theobaldi from the Neogene of Pakistan, with
comments on Siwalik hipparions. Palaeontology, 22, 439-447.

— and skinner, m. f. 1977. Earliest known Hipparion from Holarctica. Nature, London, 265, 532-533.

— and waldrop, j. s. 1980. Nannippus phlegon (Mammalia, Equidae) from the Pliocene (Blancan) of Florida.
Bull Fla. State Mus., Biol. Sci. 25, 1-37.

matthew, w. d. 1924. Third contribution to the Snake Creek fauna. Bull. Amer. Mus. Nat. Hist. 50, 59-210.

and stirton, r. a. 1930. Equidae from the Pliocene of Texas. Univ. California Pub. Bull. Dept. Geol. Sci.

19, 349-396.

merriam, J. c. 1916. Mammalian remains from the Chanac Formation of the Tejon Hills, California. Ibid. 10,
111-127.

— 1919. Tertiary mammalian faunas of the Mohave Desert. Ibid. 11, 437-585.
osborn, h. f. 1918. Equidae of the Oligocene, Miocene, and Pliocene of North America. Iconographic type
revision. Mem. Amer. Mus. Nat. Hist. N.s. 2, 1-326.

owen, R. 1948. Description of teeth and portions of jaws of two extinct anthracotheroid quadrupeds
{Hyopotamus vectianus and H. bovinus) discovered by the Marchioness of Hastings in the Eocene deposits on
the N.W. coast of the Isle of Wight, with an attempt to develop Cuvier’s idea of the classification of
pachyderms by the number of toes. Quart. Jour. Geol. Soc. 5, 380-383.
pilbeam, d. r. et al. 1977. Geology and palaeontology of Neogene strata of Pakistan. Nature, London, 270,
684-689.

pirlot, p. r. 1956. Les formes Europeennes du genre Hipparion. Mem. y Comun. Inst. Geol. Barcelona, 14,
1-122.

MACFADDEN: MIOCENE HORSE HIPPARION

635

richey, k. A. 1948. Lower Pliocene horses from Black Hawk Ranch, Mount Diablo, California. Univ. California
Pub. Bull. Dept. Geol. Sci. 28, 1-44.

schultz, G. r. (ed.) 1977. Guidebook: Field conference on late Cenozoic biostratigraphy of the Texas
Panhandle and adjacent Oklahoma, 4-6 August 1977. Kilgore Res. Center, Dept. Geol. and Anthropol., West
Texas State Univ., Spec. Pub. 1, 1-160.

sen, s., sondaar, p. Y. and staesche, u. 1978. The biostratigraphical application of the genus Hipparion with
special references to the Turkish representatives. Proc. Kon. Ned. Akad. Wet. B81, 370-385.

skinner, m. f. and hibbard, c. w. 1972. Early Pleistocene pre-glacial and glacial rocks and faunas of north-
central Nebraska. Bull. Amer. Mus. Nat. Hist. 148, 1-148.

— and macfadden, B. J. 1977. Cormohipparion N. Gen. (Mammalia, Equidae) from the North American
Miocene (Barstovian-Clarendonian). J. Paleont. 51, 912-926.

skinner, s. M. and gooris, R. j. 1977. Stratigraphy and biostratigraphy of late Cenozoic deposits in

central Sioux County, western Nebraska. Bull. Amer. Mus. Nat. Hist. 158, 263-370.

— and taylor, B. E. 1967. A revision of the geology and paleontology of the Bijou Hills, South Dakota.
Amer. Mus. Nat. Hist. Novitates, 2300, 1-53.

stirton, r. a. 1939. Cenozoic mammal remains from the San Francisco Bay region. Univ. California Pub. Bull.
Dept. Geol. Sci. 24, 339-410.

1940. Phylogeny of North American Equidae. Ibid. 25, 165-198.

— 1941. Development of characters in horse teeth and the dental nomenclature. Jour. Mammal, 22, 434-446.

sondaar, p. y. 1971. The Samos Hipparion. Proc. Kon. Ned. Akad. Wet. B74, 417-441.

1974. The Hipparion of the Rhone Valley. Geobios, 7, 289-306.

tedford, r. h. et al. (in press). Faunal succession and biochronology of the Arikareean through Hemphillian
interval (late Oligocene through late Miocene Epochs), North America. Univ. California Press, Berkeley.

van couvering, j. a. and berggren, w. a. 1977. Biostratigraphical basis of the Neogene time scale. In
kauffman, E. G. and hazel, j. e. (eds.), Concepts and Methods of Biostratigraphy. Dowden, Hutchinson, and
Ross, Inc., Stroudsburg. Pp. 283-306.

webb, s. D. 1969. The Burge and Minnechaduza Clarendonian mammalian faunas of north-central Nebraska.
Univ. California Pub. Dept. Geol. Sci. 78, 1-191.

woodburne, m. o. and bernor, r. l. 1980. On superspecific groups of some Old World hipparionine horses.
J. Paleont. in press.

BRUCE J. MACFADDEN

Florida State Museum
University of Florida
Gainesville, Florida 32611
U.S.A.

Typescript received 25 July 1979

THE TOARCI AN AGE OF THE UPPER PART OF
THE MARLSTONE ROCK BED OF ENGLAND

by M. K. HOWARTH

Abstract. The ‘Transition Bed’ of Oxfordshire, Northamptonshire, and Leicestershire is the weathered or
altered top of the Marlstone Rock Bed. In the top 0-05 m-0-3 m of the bed, the green ferrous minerals were
oxidized to limonite, partly before deposition of overlying beds, partly recently in some areas. In another type of
alteration, best seen at Harston, Leicestershire, much granular iron-pyrites was deposited in a highly irregular
zone up to 008 m thick at the top of the bed. In these Midland counties the whole of the Tenuicostatum Zone, the
basal zone of the Toarcian, is represented in the top 1 -3 m of the Marlstone Rock Bed, the lower 3-6 m of which
belongs to the Spinatum Zone. Regardless of the depth of weathering or alteration, Tiltoniceras antiquum and
Dactylioceras semicelatum of the Semicelatum Subzone occur widely in the top 0T m of the bed, D. tenuicostatum
occurs more locally at a slightly lower horizon, and lower still one D. crosbeyi is evidence for the Clevelandicum
Subzone. Ammonites from the Semicelatum, Tenuicostatum, and Paltum Subzones occur in the Dorset coast
Marlstone Rock Bed. North of Lincoln the top of the Bed is at about the top of the Spinatum Zone, while the
Tenuicostatum Zone is divided between an overlying hard mudstone and higher grey shales. The change from
the underlying ironstone/limestone facies to the overlying clays/shales-with-nodules facies took place at the top
of the Spinatum Zone in Yorkshire, but at the top of the Tenuicostatum Zone in the Midland counties.

T he Marlstone Rock Bed is one of the most distinctive lithological horizons in the English Lias. It is
typically an oolitic limestone, though the sand content becomes significant in a few places, and the
ferrous iron content is high enough in two areas for it to be used as an iron-ore. The outcrop (text-
fig. 1), known from numerous building-stone and iron-ore quarries in the past, extends from the
Dorset coast generally north-eastwards to north of the Humber. Although it is represented
immediately north of the Mendips and also at Dundry, south of Bristol, it is generally absent around
and to the south of Bath. There is another gap in north Northamptonshire and south Leicestershire,
where it is absent altogether or only about 0-3 m thick, owing to a combination of thin deposition and
subsequent erosion. For several miles north and south of Lincoln it disappears and is represented by a
layer of phosphatic pebbles and possibly some thin overlying shales. North of the Humber it thins out
against the Market Weighton block, and it does not reappear farther north in Yorkshire, where
equivalent beds are developed in a different facies.

Traditionally the top of the Marlstone Rock Bed was the junction between the Middle and Upper
Lias, or more specifically the junction between the Spinatum and Tenuicostatum Zones (Arkell 1933,
pp. 153-159; Whitehead et al. 1952, pp. 97, 105, 144-150). In fact the Marlstone Rock Bed was
referred to a single zone, the Spinatum Zone, because in some areas it has a rich fauna of species of
Pleuroceras (Howarth 1958, pp. ix-xi). The beds above the Marlstone Rock Bed are clays and shales
in most areas, but in some parts of the Midlands the ‘Transition Bed’, a bed of oolitic limestone up to
015 m thick, is the immediately overlying bed. It was first described from the Banbury area, north
Oxfordshire, and later at Tilton, Leicestershire, and contains a rich fauna of the Upper Lias
ammonites Dactylioceras and Tiltoniceras ‘ acutum ’ (Blake). It was called the Acutum Zone or
hemera by Buckman (19106, p. 86) and the Acutum Subzone by Arkell (1933, p. 179), and placed at
the base of the Upper Lias, below the Tenuicostatum Subzone, as the lower of the two subzones of the
Tenuicostatum Zone. This subzonal position of the Transition Bed has not been challenged until
recently, though Spath ( 1 942, p. 265; 1 956, p. 1 43) did not accept the validity of this subdivision of the
Tenuicostatum Zone. The identification of the horizon to which the Transition Bed belonged was
made more difficult by the naming of an atypical representative of its Dactylioceras fauna as

IPalaeontology, Vol. 23, Part 3, 1980, pp. 637-656, pis. 80-82.|

638

PALAEONTOLOGY, VOLUME 23

Orthodactylites directus Buckman (1926a, pi. 654), and also by the inability of anyone to find either
this ammonite or Tiltoniceras in the Yorkshire coast Upper Lias succession. In a more recent
investigation of the succession at Tilton, Hallam (1955, p. 21) discovered that D. directum occurred in
the top 0-9 m of the Marlstone Rock Bed as well as in the Transition Bed, and suggested that it would
be necessary to place the base of the Toarcian (i.e. the Upper Lias) at least 0-9 m below the top of the
Marlstone Rock Bed. A more conservative view was taken by Howarth (1958, p. xi), and followed

text-fig. 1 . Sketch map of the outcrop of the Marlstone Rock Bed, showing the principal localities described in

the text.

later by Hallam ( 1 967, p. 397), that it was best to retain the Middle/Upper Lias boundary at the top of
the bed, because the relationships between the last Pleuroceras and the first Dactylioceras in Britain
were not known at that time.

No further advance could be made until the succession of ammonites in the Tenuicostatum Zone of
the Yorkshire coast was worked out. When this was done (Howarth 1 973, enlarging on the collecting
of the late Professor P. C. Sylvester-Bradley, whose preliminary results were published in Dean,

HOWARTH: AGE OF MARLSTONE ROCK BED

639

Donovan, and Howarth 1961, p. 476) the following sequence of ammonites, and of subzones
derived from them, was established:

Zone

Subzone

Ammonite faunas

Dactylioceras

semicelatum

Tiltoniceras antiquum and
Dactylioceras semicelatum

D. semicelatum

Dactylioceras

D. tenuicostatum

D. tenuicostatum

tenuicostatum

D. clevelandicum

D. clevelandicum

D. crosbeyi

Protogrammoceras
pal turn

Protogrammoceras paltum

Abundant faunas of Tiltoniceras were found at the top of the Tenuicostatum Zone in Yorkshire, not
at the base of the zone where the genus had always been expected before (e.g. Hallam 1967, p. 415).
This alone was sufficient to suggest that the Transition Bed of the Midlands belonged to the top
subzone of the Tenuicostatum Zone, and confirmation of this correlation was obtained when it was
found that most of the Dactylioceras in that bed, to which the name D. directum had always been
given before, were typical examples of D. semicelatum. In fact the populations of the latter species in
the Transition Bed and in the Semicelatum Subzone of the Yorkshire coast are very similar, having
almost identical ranges of variation. One end of the variation consists of evolute specimens with fine
rectiradiate ribs, and Buckman gave the name D. directum to the most extreme example of this type
from the Transition Bed.

Shortly after the Yorkshire coast Tenuicostatum Zone succession had been described, an
abundant ammonite fauna was found in the top of the Marlstone Rock Bed in a quarry at Harston,
north Leicestershire. The top 0 08 m of the bed contained many D. semicelatum and a few
Tiltoniceras , and was clearly equivalent to the Transition Bed, though it was not developed as a
distinct bed at Harston. The main discovery, however, was the presence of D. tenuicostatum in
abundance in the next 0-05 m below, an ammonite that had hardly ever been found in the Marlstone
Rock Bed before. This proved the presence of the Tenuicostatum Subzone in the bed, and further
minor discoveries showed that the Clevelandicum Subzone occurred lower still in the bed.

The presence of the whole of the Tenuicostatum Zone in the Marlstone Rock Bed at Harston, and
the discovery in existing museum collections of specimens of Tiltoniceras from Tilton preserved in
green oolitic limestone typical of the Marlstone Rock Bed, led to further investigation of the Tilton
Railway Cutting. It was found that, just as had been originally described by Wilson and Crick ( 1 889),
the Transition Bed is not a lithologically distinct bed, it is merely the weathered top of the Marlstone
Rock Bed. There is no lithological break or disconformity that marks off a distinct bed at the top,
only a highly irregular zone of oxidation of the green ferrous-iron content of the oolite to brown
limonite. The Semicelatum Subzone ammonite fauna occurs in the top 0-9 m of this complete
Marlstone Rock Bed (i.e. including the ‘Transition Bed’) at Tilton. Unfortunately there is no
evidence for lower subzones of the Tenuicostatum Zone at Tilton, though there is plenty of room for
them above the highest recorded Pleuroceras at about 3 m below the top of the Marlstone Rock Bed.

The ‘Transition Bed’ and its distinct ammonite fauna is also well developed in the Banbury-Byfield
area of north Oxfordshire and west Northamptonshire. Although more constant in thickness, it
appears possible to interpret it similarly in that area as the weathered top of the Marlstone Rock Bed,
weathering that probably occurred before deposition of any overlying beds. Between north

640

PALAEONTOLOGY, VOLUME 23

Oxfordshire and south Somerset Tenuicostatum Zone ammonites are rarely found and the presence
of the zone within the Marlstone Rock Bed has yet to be demonstrated. On the Dorset coast,
however, where the Marlstone Rock Bed is very thin (0-0-6 m), ammonites are frequent and prove the
presence of the Paltum, Tenuicostatum, and Semicelatum Subzones. The new stratigraphical work
and ammonite collections, and reinterpretation of older collections are described in detail below.

STRATIGRAPHICAL DESCRIPTIONS

1 . Dorset coast. The Marlstone Rock Bed forms the lowest part of the Middle and Upper Liassic Junction Bed in
the cliffs between Seatown and Eype, and has been described in detail by Buckman (19226), Jackson (1922,
1926), and Howarth (1957). The bed is never more than 0-6 m thick and consists of three layers, the lithological
differences and ammonite faunas of which were discussed by Howarth (1957, pp. 192-193). The lowest layer R is
a coarse conglomeratic and oolitic limestone that contains many Pleuroceras indicative of the Apyrenum
Subzone of the Spinatum Zone. The middle layer Px is a hard grey and pink limestone with scattered ooliths that
contains only a few P. cf. spinatum and probably belongs to the Hawskerense Subzone. The top layer P is a
brown finely oolitic limestone that contains a rich ammonite fauna. Previously (Howarth 1957, p. 193) it was said
to be of Hawskerense Subzone age only, but now that the sequence within the Tenuicostatum Zone is known in
Yorkshire, it is clear that layer P is a highly condensed bed that contains most horizons from the Hawskerense up
to the Semicelatum Subzones. The following is a list of the ammonites that have been collected from layer P:

Dactylioceras semicelatum (BM C.17548, C.74719; IGS GSM 22475, 22514; SM J.44225-44226; NMW 26.135
G123), D. tenuicostatum (NMW 26.135 G5-8 (9 specimens), G124), Protogrammoceras paltum (BM 67939,
C.2200, C.30769, C.68536; IGS GSM 47160-47161, 49291; SM J.44789), Pleuroceras spinatum (Bruguiere),
P. spinatum var. buckmani (Moxon), P. yeovilense Howarth, P. hawskerense (Y oung and Bird), P. apyrenum
(Buckman).

These ammonites are characteristic of the Semicelatum, Tenuicostatum, Paltum, and Hawskerense Subzones,
and the only horizon for which there is no evidence is the Clevelandicum Subzone. The examples of D. semi-
celatum (PI. 8 1 , figs. 3, 4, and Howarth 1 957, pi. 1 7, figs. 5, 6) are typical of the species and match Y orkshire coast
examples closely. The ten specimens of D. tenuicostatum (PI. 82, figs. 5-8) are small and very similar to examples
from near the top of the Marlstone Rock bed at Harston, Leicestershire; they all came from a layer of fine brown
oolite which also contained one specimen of D. semicelatum, many gastropods, and a unique Terebratulid that
was later described as ‘ Terebratula ’ reversa Ager (1956a, p. 4, pi. 1, fig. 6) (possibly a Lobothyris). This
association of fossils found in only a single block was the basis for the proposal of the layer At by Jackson (1926,
p. 497) (At was derived from Buckman’s hemera ‘ athleticum , a term used for the Transition Bed of the Midlands
that was said to contain a similar Terebratulid). However, the lithology is not different from layer P, the
brachiopod has no special age significance, and the ammonites are intermediate in age between the Semicelatum
and Paltum Subzones ammonites that are found in many other blocks of layer P. Therefore, there is no
justification for the recognition of a separate At layer. Well-preserved specimens of Protogrammoceras paltum in
layer P include the holotype and paratype (Buckman 1922a, pi. 362A; 1923a, pi. 362B), an example figured by
Wright (1884, pi. 81 , figs. 4-6), and the specific synonym Platyharpites platypleurus Buckman (1927a, pi. 698). So
layer P, though never more than 0-3 m thick, contains highly condensed representatives from the Hawskerense to
Semicelatum Subzones. The Marlstone Rock Bed of the Dorset coast, i.e. layers R, Px, and P, belongs to the
whole of the Spinatum and Tenuicostatum Zones, so it is approximately equally divided between the Middle and
Upper Lias. The next higher blocks of the Junction Bed are the layers N, O, and D, which are lateral equivalents
of each other, and contain specimens of Harpoceras exaratum, from about the middle of the Exaratum Subzone.
There are no records of Eleganticeras that would indicate the presence of the lower part of the Exaratum
Subzone.

2. North Dorset, Somerset, Avon, and Gloucestershire. Northwards from the Dorset coast the Marlstone Rock
Bed thickens quickly, and the term Junction Bed is now applied to the overlying sequence of clays and limestones
of the Upper Lias. Both beds are very rich in ammonites in the Ilminster area of south Somerset. In the well-
known Barrington succession described by Hamlet ( 1 922), Spath ( 1 922), and Pringle and Templeman ( 1 922), the
Marlstone Rock Bed contains many Pleuroceras, and is overlain by bed 1 (of Hamlet), a 0-175 m bed of ‘sandy
marl’ which contains D. cf. tenuicostatum in addition to more examples of Pleuroceras. Bed 2, a 0- 1 m bed of grey
oolitic limestone, contains D. semicelatum, of which an example is figured here (PI. 8 1 , figs. 1 , 2). The overlying
bed 3 is clay containing argillaceous limestone nodules, and is of Exaratum Subzone age. So the Tenuicostatum

HOWARTH: AGE OF MARLSTONE ROCK BED

641

Zone is confined to bed 2 and part of bed 1 , and these may be a local lithological variation of the Marlstone
Rock Bed.

The Marlstone Rock Bed is well developed around Batcombe and Evercreech, near Shepton Mallet on the
south side of the Mendips (Richardson 1906, 1909), but evidence for the presence of the Tenuicostatum Zone has
not been obtained. After a gap north of the Mendips, the bed reappears north of Bath, thickens quickly, and was
formerly extensively quarried along the western escarpment of the Cotswolds in Gloucestershire. There is little
ammonite evidence for the age of the top of the bed. Species of Pleuroceras from both subzones of the Spinatum
Zone are common at some localities (e.g. Alderton Hill), but data about their stratigraphical position in the
Marlstone Rock Bed are lacking. There are no Tenuicostatum Zone ammonites in existing museum collections
from this area. One record is intriguing, however: in an exposure of the Marlstone Rock Bed near Stow-on-the-
Wold, about 25 km east of the Cotswolds escarpment, Hull (1857, pp. 19, 20) saw a ‘band of deep reddish purple
ironstone’ 015 m thick at the top of the bed ‘filled with good specimens of Ammonites annulatus ’. It is likely that
these were examples of D. tenuicostatum or D. semicelatum and they would show that most of the Tenuicostatum
Zone was in the Rock Bed. The exposure was not extant in 1929 when Richardson (1929, p. 31) quoted the
record, and the ammonites are not preserved in the Institute of Geological Sciences, so the occurrence cannot be
investigated further. In the Stowell Park bore-hole, 1 8 km south-west of Stow-on-the-Wold, the Marlstone Rock
Bed did not yield any ammonites, but 1 m of overlying shales contained Tiltoniceras and Dactylioceras of the
Semicelatum Subzone. This shows that at least some of the Tenuicostatum Zone is above the Marlstone Rock
Bed, though it need not be more than the upper half of the Semicelatum Subzone.

3. Oxfordshire and Northamptonshire. The Marlstone Rock Bed used to be extensively quarried for iron-ore and
building stone over a large area between Banbury and Northampton, and details of the many former quarries
can be found in Whitehead et al. (1952). It was in this district that the term ‘Transition Bed’ was first proposed by
Walford (1878, p. 2) for a pale-brown oolitic and ferruginous ‘marl’ 0 050-0-075 m thick that forms the top of
the Marlstone Rock Bed. The type area is around Banbury, and the best-known localities were quarries at
Adderbury, King’s Sutton, and Middleton Cheney, south and east of Banbury. Large numbers of T. antiquum
(Wright) and D. semicelatum (including the holotype of D. ‘ directum ’ Buckman) and many small gastropods
were obtained from the Transition Bed in these quarries, and they show that the bed belongs to the Semicelatum
Subzone. Arkell (1947, p. 21) proposed that the term ‘Acutum Bed’ (after Tiltoniceras ''acutum' Blake, the
holotype of which came from Adderbury) should supercede Transition Bed in the north Oxfordshire area, but
this change of name has not been adopted by other authors (‘Acutus Subzone’ had been used earlier by Walford
(1899, p. 33)). This lithology and 0-050-0-075 m thickness is fairly constant over the whole of north Oxfordshire
and west Northamptonshire as far north as Daventry. At Iron Cross Farm, Byfield, the last locality at which it
was well exposed (Howarth 1978, p. 240), it forms the upwards continuation of the Marlstone Rock Bed, with
which it has a sharp and irregular junction. The Transition Bed appears to be the altered top of the Marlstone
Rock Bed, alteration that is mainly oxidation and leaching of the green ferrous iron, and which probably took
place before deposition of the„overlying Abnormal Fish Bed. The latter is separated by a parting from the top of
the Transition Bed, and is of mid and upper Exaratum Subzone age (Howarth 1978, p. 241).

In areas further east, and especially around Milton and Bugbrooke west of Northampton, a series of beds up
to 0-35 m thick has been referred to the Transition Bed (Thompson 1889, 1892). This is due to the inclusion of an
overlying sandy or shaly clay that does not contain the characteristic Transition Bed ammonites or gastropods.
The age of the clay is not accurately known, but it may bridge the small disconformity that occurs everywhere
else between the Transition Bed and the Abnormal Fish Bed, and it should not be included in the Transition Bed.
At Bugbrooke Thompson (1892, p. 337) said that the Transition Bed was not present as a distinct bed, but was
nevertheless clearly shown by the altered character of the top of the Marlstone Rock Bed which contained the
Transition Bed fossils. Thus it appears that throughout the area the Transition Bed is the altered top of the
Marlstone Rock Bed, alteration which probably took place before deposition of the overlying beds.

Evidence for the subzonal age of the remainder of the Marlstone Rock Bed is meagre in this area. Only one
specimen of D. tenuicostatum has been found (PI. 82, figs. 3, 4), at Rothersthorpe, 5 km south-west of
Northampton, from an unrecorded horizon, but judging from the grey-green finely oolitic matrix, probably
from immediately below the altered top of the Marlstone Rock Bed. This ammonite is evidence for the
Tenuicostatum Subzone, but there are no ammonites to prove the presence of lower subzones. The majority of
the Marlstone Rock Bed belongs to the Spinatum Zone and the characteristic brachiopods Tetrarhynchia
tetrahedra and Lobothyris punctata are abundant except at the extreme top. Pleuroceras is rare in
Northamptonshire: a few P. spinatum have been found and at least one P. apyrenum is known, but their horizons
are not recorded. The Middle/Upper Lias junction occurs near the top of the Marlstone Rock Bed, probably
within the top 0-25 m.

642

PALAEONTOLOGY, VOLUME 23

4. Tilton, Leicestershire. The well-known section at Tilton Railway Cutting (SK 762055) was first described by
Wilson and Crick (1889) and there is a good photograph of it in its original state in Fox-Strangways (1903, p. 30,
pi. 2). Wilson and Crick saw the Marlstone Rock Bed soon after it was uncovered below a thickness of up to 9 m
of Upper Lias shales and it had been little affected by recent subaerial weathering. The ‘Transition Bed’ was
described as a flaggy limestone 0T 5-0-23 m thick containing a distinctive fauna, especially the ammonite
Tiltoniceras, even though they said that ‘it possesses the mineral characters of and is welded to the top of the
Marlstone Rock Bed’ (Wilson and Crick 1889, p. 297). Woodward (1893, p. 236) repeated this interpretation of
the Transition Bed, but Whitehead et al. (1952, p. 135) made no mention of a Transition Bed at Tilton, nor
anywhere in the surrounding area. The railway section was again described by Hallam (1955; 1968, p. 208) who
recognized the 01 5-0-23 m Transition Bed, and observed that it lapped ‘over minor irregularities at the top of
the ironstone’ and ‘rested non-sequentially on the ironstone’. The lithology of the Transition Bed has been
described as a pale-brown or cream finely oolitic limestone, sometimes flaggy, and sometimes passing up into
sandy marl. Tiltoniceras preserved in such brown oolitic limestone is very common, but many others also occur
preserved in the deep-green oolitic ironstone that is typical of the Marlstone Rock Bed at Tilton. Hallam (1955,
p. 21) explained the latter by saying that the genus occurred rarely in the ironstone immediately below the
Transition Bed.

Examination of the Tilton Railway Cutting exposures in recent years shows that the Transition Bed does not
exist as a separate bed. It is the weathered top of the Marlstone Rock Bed, in which the siderite and chamosite of
the deep green oolitic limestone are oxidized to limonite; partial decalcification gives it a friable, granular
texture, which has been described as sandy, though the bed is not arenaceous. The depth of weathering varies
greatly between 0-01 m and 0-25 m below the top surface, and the lowest extent is marked by an undulating thin
sheet of brown limonite. Many specimens of T. antiquum (Wright), D. semicelatum (PI. 81, figs. 5, 6) and
Gibbirhynchia tiltonensis Ager, and many small gastropods (Wilson and Crick 1889, pp. 298-305, pi. 9) occur in
the top 0-2 m of the Marlstone Rock Bed. Whether they occur preserved in deep-green ironstone or pale-brown
oolitic limestone depends entirely on how deeply the weathering has penetrated at any particular point. Several
fine examples of Tiltoniceras preserved in green ironstone were obtained from only 0-025 m below the top, but
most of the green specimens occur lower down. In some specimens that are orientated approximately vertically
in the bed, the upper half of the ammonite is pale brown and the lower half deep green. Weathering also
penetrates deeply down some of the vertical joints and can convert fossils much lower down into pale-brown
friable limestone. In a few places horizontal bedding planes lead to greater penetration of weathering, and rarely
the whole of the top 0-2 m is affected giving the appearance of a distinct lithological bed at the top of the iron-
stone. Such beds fade out rapidly laterally, and the usual state is dark-green Marlstone Rock Bed weathered
brown to a highly variable depth.

D. semicelatum is commonest in the top 0-2 m, but unlike Tiltoniceras it also occurs lower down to depths of
0-9 m below the top of the ironstone. This is the amount of the Marlstone Rock Bed that must be referred to the
Semicelatum Subzone. No Upper Lias ammonites belonging to lower subzones occur at Tilton. The only
Pleuroceras found in situ is a specimen of P. cf. hawskerense (Young and Bird) 3-0 m below the top of the
Marlstone Rock Bed, and it indicates the Hawskerense Subzone of the Spinatum Zone. The brachiopods
Tetrarhynchia tetrahedra (J. Sowerby) and L. punctata (J. Sowerby) range higher in the ironstone, the last ones
being about 1-2 m below the top (Hallam 1955, p. 20). These two are usually held to be good indicators of the
Spinatum Zone in England, but there are rare records from the Upper Lias, the genus Tetrarhynchia ranges up
into the Bajocian (Ager 19566, p. 3), and T. tetrahedra occurs in the Upper Lias, Bifrons Zone, in Spain (Hallam
1972, p. 408). So in the absence of ammonites, it does not seem safe to take the highest occurrence of these
brachiopods as unequivocal evidence of the Spinatum Zone. The evidence available at present suggests that the
Spinatum/Tenuicostatum Zone boundary occurs between 1 m and 3 m below the top of the Marlstone Rock Bed
at Tilton. There is room in this thickness for condensed representatives of the three lower subzones of the
Tenuicostatum Zone, and a disconformity need not be postulated to explain their absence. There is also no
lithological evidence for such a disconformity.

The beds above the Marlstone Rock Bed are clays and shales with a few thin beds of limestone or limestone
nodules. The basal 2-8 m belongs to the Exaratum Subzone, and uncrushed examples of Harpoceras elegans
(J. Sowerby) and H. serpentinum (Schlotheim) occur at about the 2 m level. These indicate the top part of the
Exaratum Subzone, and the absence of H. exaratum suggests that the non-sequence between the top of the
Marlstone Rock Bed and the shales represents at least the lower half of the Exaratum Subzone. All the higher
shales up to the top of the cutting belong to the Falciferum Subzone and contain the index ammonite commonly
throughout. The following is a summary of the section exposed in the Tilton Railway Cutting (SK 762055),
Leicestershire:

HOWARTH: AGE OF MARLSTONE ROCK BED 643

Zone and subzone of Harpoceras falciferum

Grey shale, with two rows of small limestone nodules about 0-5 m and 0-6 m below the top. H. falci-
ferum 5-50 m

Grey clay containing large calcite ooliths. H. falciferum, Phylloceras heterophyllum (J. Sowerby) 0-70 m

Subzone of Harpoceras exaratum

Grey clay, oolitic. Large specimens of H. serpentinum (Schlotheim) 0-80 m

Grey oolitic limestone. H. elegans (J. Sowerby) (BM C.8048 1-80483) and H. serpentinum common

(BM C.80484-80485), and many Dactylioceras sp. indet 0-20 m

Grey shales, paper shales, and clays. H. serpentinum in top 0-5 m l-80m

Zones of Dactylioceras tenuicostatum and Pleuroceras spinatum

Marlstone Rock Bed:

a. Ironstone. Dark-green finely oolitic limestone, containing chamosite and siderite, weathered
brown to an irregular depth, and sometimes more deeply along joints and bedding planes.
Tiltoniceras antiquum (Wright) (BM C. 10265-10267, C.41733, C.48753-48757, C.80242-80276,

C. 80470-80480) and D. semicelatum (BM C.36186-36188, C.49766, C.80277-80282, C.80466-
80469) are abundant in the top 0-2 m and 0-9 m respectively and indicate the Semicelatum
Subzone; P. cf. hawskerense (Young and Bird) (BM C.73686) occurs at the bottom and indicates
the Hawskerense Subzone 3-0 m

b. Green oolitic limestone, containing numerous specimens of Tetrarhynchia tetrahedra and

Lobothyris punctata, and many bivalves (band B of Hallam 1955, p. 18) 0-45 m

c. Sandrock. Green massive calcareous sandstone . . . 1 -4 m

d. Calcareous sandstone as bed c, but with many nests of the brachiopods T. tetrahedra and

L. punctata {band AofHallam 1955, p. 18) 0-3 m

e. Sandrock, as bed c 0-75 m

Other exposures of the Marlstone Rock Bed in the Tilton area were in iron-ore quarries, where the bed had
been less deeply buried than in the railway cutting. Consequently the top of the bed had been more strongly
weathered. Of those described by Whitehead et al. (1952) and Hallam (1955, 1968), few now remain. One that
can still be seen is the old quarry 1-3 km east of Tilton (SK 756056), where the top 0-2 m of the Marlstone Rock
Bed is highly weathered into a pale-brown oolitic limestone that contains Tiltoniceras antiquum and
Dactylioceras semicelatum. Other quarries, now obscured, were similar, and it is thought that the ‘Transition
Bed’ is, in all cases, the weathered top of the Marlstone Rock Bed.

5. Grantham area, north Leicestershire and south Lincolnshire. Two quarries working the Marlstone Rock Bed as
iron-ore existed, until closed down and filled in in 1975, at Harston, 12 km south-west of Grantham. Here the top
of the Marlstone Rock Bed contains more Dactylioceratidae than any other exposure of the bed in England, and
it is the most important section for dating the Upper Lias part of the bed. A section for the Upper Lias shales
above the bed was given in Hallam (1968, p. 210), but a more detailed description is now given, so that the
position of the disconformities can be established. Section at Harston Quarry (SK 843305), 1 -5 km south-south-

east of Harston:

Zone and subzone of Harpoceras falciferum

Clay. Impressions of//, falciferum and Dactylioceras sp. indet. . 2-00 m

Brown rubbly limestone, oolitic or pisolitic in places. H. falciferum ...... 0-20 m

Subzone of Harpoceras exaratum

Grey shale 1 -20 m

Scattered flat nodules of blue limestone, weathered red-brown and white. H. elegans (J. Sowerby)
abundant, Dactylioceras anguiforme Buckman abundant, Nodicoeloceras crassoides (Simpson),
Phylloceras heterophyllum (J. Sowerby), Coelodiscus minutus (Scbub\er) . . . . . OTOm

Grey shale c. 10-00 m

Grey calcareous clay, forming a hard massive bed. A few limestone nodules occur in a row at the top.

Large specimens of H. elegans, H. serpentinum (Schlotheim) and Hildaites murleyi (Moxon) . l-30m

Shale, with a row of 0-025 m thick flat limestone nodules at the top. H. serpentinum, H. elegans . . 0-05 m

Scattered lenticles of coarse sandstone, cross-bedded, with many granules of iron pyrites and some
small pebbles. Much shell debris broken into small fragments. Fragments of Harpoceras (?//. cf.
exaratum) 0-0-05 m

644 PALAEONTOLOGY, VOLUME 23

Zone of Dactylioceras tenuicostatum

Marlstone Rock Bed:

a. Pale-brown limestone, consisting of numerous calcite ooliths and minute shell fragments in a

calcareous matrix; the top 0-05 m contains patches of crystalline calcite and occasional pebbles of
brown limestone; the lower half becomes green-coloured, more coarsely oolitic, with chamosite
and siderite, and much recrystallized calcite; the top 0-025-0-080 m is full of fine granules of iron-
pyrites and is grey-green in colour; its very uneven lower boundary is marked by a solid line of
iron-pyrites, and the bed below is pale brown with only a few granules of iron-pyrites . . . 1 -20 m

Subzone of Dactylioceras semicelatum

The top 0-08 m contains Tiltoniceras antiquum (Wright) (BM C.80237-80241), D. semicelatum
common (BM C. 80169, C.80171-80235), Acrocoelites vulgaris (Young and Bird), Gibbirhynchia
sp. and gastropods.

Subzone of Dactylioceras tenuicostatum

Between 0 08 m and 013 m below the top D. tenuicostatum is abundant (BM C.80099-80168) and
Gibbirhynchia sp. occurs.

Subzone of Dactylioceras clevelandicum

0-23 m below the top one specimen of D. crosbeyi (Simpson) (BM C.80170) was found.

Zone of Pleuroceras spinatum

b. Deep-green oolitic limestone, with much chamosite and siderite, and many bands of recrystallized
calcite. Abundant Tetrarhynchia tetrahedra and Lobothyris punctata in nests. One Pleuroceras cf.
spinatum (Bruguiere) 0-25 m below the top, and several other specimens not in situ

300 m

This quarry contained one of the best sections of the Marlstone Rock Bed for demonstrating that the top part
that contains Tiltoniceras is a typical part of the bed that has been diagenetically altered. The alteration is due to
pyritization from the top surface downwards. It consisted of the deposition of a large amount of fine granular
iron-pyrites, which penetrated to a depth varying between 0-025 m and 0 080 m and the very uneven lower
boundary is marked by a thin sheet of solid iron-pyrites. Tiltoniceras and D. semicelatum (PI. 81, figs. 10, 11;
PI. 82, figs. 11,12) occur in the top 0-080 m, so some of the Semicelatum Subzone is in the pyritized part and some
in the unaltered part below. Most ammonites lie parallel to the bedding plane, but a few are at a high angle and
occasionally the lower boundary of pyritization has reached half down an ammonite. There is no lithological
break or change other than the pyritization, except for a few pebbles just below the top surface.

D. tenuicostatum (PI. 82, figs. 1, 2, 9, 10) occurs in abundance between 0-08 m and 0-13 m below the top of the
Bed, and this is the extent of the Tenuicostatum Subzone, which is below the pyritized zone. A single D. crosbeyi
0-23 m below the top is evidence for the presence of the Clevelandicum Subzone. There are no ammonites to
prove the presence of the Paltum Subzone, but there is plenty of room for it between 0-25 m and 1-2 m below the
top of the Bed. Marlstone Rock Bed division b of the above section is a natural downward continuation of the
upper part where it becomes richer in iron, and several specimens of Pleuroceras occur of the Spinatum Zone.

The whole of the Tenuicostatum Zone is in the top 1 -2 m of the Marlstone Rock Bed at Harston, and there are
no major lithological discontinuities within that part of the bed. The main disconformity is at the top of the bed
where the lithology changes to shale facies, and the bottom one-third of the Exaratum Subzone is missing
(because of the absence of Eleganticeras). The middle third of that subzone is represented only by the lenticles of
sandstone that contain Harpoceras, and continuous deposition starts only in the upper third of the subzone
where more than 1 1 m of shales contain H. elegans and H. serpentinum.

At Denton Park Quarry (SK 856316), 1-5 km north-east of Harston Quarry, a similar succession was seen at
the top of the Marlstone Rock Bed, though ammonites were much less common. The top 0-03-008 m of the bed
contains much granular iron-pyrites as at Harston, but it is more of a shell bed containing a great number of
broken bivalve shells and large numbers of the belemnite Acrocoelites vulgaris (Young and Bird). A few
fragments of large specimens of Tiltoniceras were also seen. The lower non-pyritized part of the bed is similar to
that at Harston, but Tenuicostatum Subzone ammonites were not found.

6. North Lincolnshire and south Humberside. North of Grantham the Marlstone Rock Bed thins steadily and
disappears altogether before Lincoln. At Lincoln the Spinatum Zone is absent or is represented by a bed of

HOWARTH: AGE OF MARLSTONE ROCK BED

645

phosphatic pebbles (Trueman 1918; Howarth 1958, p. xii, bed 11), but there is no ammonite evidence for the
presence of the zone nor for the lowest three subzones of the Tenuicostatum Zone. The presence of the
Semicelatum Subzone is shown, however, by specimens of T. antiquum (Trueman Coll., Nottingham University,
and BM C.48429-48432) in the top 0-15 m of the 0-60 m of overlying shales (Howarth 1958, p. xi, bed 12).
D. semicelatum also occurs in these shales, and probably in the shales of bed 14 above.

The Marlstone Rock Bed reappears north of Lincoln and it was well exposed in recent years in quarries at
Kirton in Lindsey (Howarth and Rawson 1965) and Roxby (Penny and Rawson 1969, pp. 194-197). The
distribution of ammonites in the Upper Lias shales was poorly known in these quarries, and better records have
been obtained from boreholes in the same area. Most information came from boreholes near Worlaby, 8 km east
of Roxby, where many specimens of T. antiquum occurred in shales between 4 m and 5-7 m above the Marlstone
Rock Bed (Richardson 1979). The following succession for part of the Middle and Upper Lias in this area
incorporates details of the Kirton in Lindsey Quarry (Howarth and Rawson 1965, pp. 262-263), the north end of
the Roxby Quarry (Penny and Rawson 1969, p. 196), and some records from the Worlaby boreholes.
Thicknesses and lithology show little variation, though the rows of doggers are more obvious in the quarries,
especially at Kirton in Lindsey. The ammonite distribution is the same, and records from all three places are
included.

Subzone of Harpoceras exaratum

Shale, with two rows of doggers and a band of limestone. Beds 25-29 at Kirton in Lindsey; bed 29,
a row of doggers 0T3 m from the top, contains many H. elegans (J. Sowerby); bed 27, a bed of
limestone 1-2 m from the top, contains H. cf. exaratum (Young and Bird); bed 25 is a row of

doggers at the base 4-60 m

Subzone of Dactylioceras semicelatum

Shale, close-bedded, but sandy in basal 0-3 m. Beds 23 and 24 at Kirton in Lindsey. Many crushed

Tiltoniceras antiquum (Wright), sometimes in shell beds, through most of the thickness . 3-50 m

Shale, with scattered limestone nodules, especially near the base. Crushed D. semicelatum in the
shales in the borehole, and well-preserved solid specimens in the basal nodules at Roxby 1 • 70 m

Subzone of Dactylioceras tenuicostatum

Shale. A few D. cf. tenuicostatum 1 -30 m

Subzones of Dactylioceras tenuicostatum (part), D. clevelandicum, and Protogrammoceras paltum
Hard, pale-grey, calcareous mudstone, silty and micaceous, with some phosphatic and calcareous
nodules. Bed 21 at Kirton in Lindsey. Many well-preserved ammonites and belemnites:

D. tenuicostatum and D. clevelandicum common; one large P. paltum (Buckman) known from

Roxby 0-40-1 TOm

Zone of Pleuroceras spinatum

Marlstone Rock Bed. Green oolitic limestone. Rare Pleuroceras cf. hawskerense (Young and Bird)

(level unknown). Many brachiopods 3 00 m

The Dactylioceratidae that were recorded previously (Howarth and Rawson 1965, p. 262) in bed 21 at Kirton
in Lindsey have been reassessed in the light of the succession of species now known in the Yorkshire coast Grey
Shales (Howarth 1973). Dactylioceras tenuicostatum and D. clevelandicum are both present, and with the single
large Protogrammoceras paltum at Roxby, they show that the Paltum, Clevelandicum, and part of the
Tenuicostatum Subzones are present in that bed. D. tenuicostatum also occurs in the shales above, and then
D. semicelatum and Tiltoniceras antiquum occur higher up. So it can be shown that the Tenuicostatum Zone lies
wholly above the Marlstone Rock Bed in the area north of Lincoln. Approximately the lower half of the zone is
condensed in a bed up to I T m thick, but unlike the ‘Transition Bed’ of Oxfordshire to Leicestershire, it is a
calcified silty mudstone significantly different in lithology from the Marlstone Rock Bed. The upper half of the
zone occurs in shales 6-5 m thick that resemble the Grey Shales of the Yorkshire coast in thickness and lithology,
except for the absence of pyritized doggers.

646

PALAEONTOLOGY, VOLUME 23

SYSTEMATIC DESCRIPTIONS
Family dactylioceratidae Hyatt, 1867
Genus dactylioceras Hyatt, 1867
Subgenus orthodactylites Buckman, 1926
Dactylioceras ( Orthodactylites ) semicelatum (Simpson)

Plates 80, 81; Plate 82, figs. 11, 12; text-figs. 2, 3

1819 Ammonites annulatus J. Sowerby, p. 41, pi. 222, figs. 1, 2 ( non figs. 3-5) (non Ammonites
annulatus Schlotheim, 1813).

1843 Ammonites semicelatus Simpson, p. 20.

1855 Ammonites semicelatus Simpson, p. 50.

1884 Ammonites semicelatus Simpson, p. 81.

1911a Dactylioceras semicelatum (Simpson); Buckman, pi. 31.

1926a Orthodactylites directus Buckman, pi. 654.

1927a Kryptodactylites semicelatus (Simpson); Buckman, pi. 31A.

1927a Orthodactylites mitis Buckman, pi. 738.

1957 Dactylioceras directum (Buckman); Howarth, p. 197, pi. 17, figs. 5, 6.

1957 Dactylioceras semicelatum (Simpson) and spp.; Maubeuge, figs. 1-3, 718-21, 41, 42, 44, 746, 47,
48,749, 758, 759(1), 59 (2).

1957 Dactylioceras pseudocrassoides Maubeuge, p. 201, pi. 13, fig. 28.

1957 Dactylioceras densicostatum Maubeuge, p. 202, pi. 13, fig. 29.

1960 Dactylioceras sp. indet.; Hoffmann and Martin, p. 114, pi. 9, fig. 5; pi. 10, figs. 2a, 2b.

1968 Dactylioceras cf. toxophorum (Buckman); Hoffmann, p. 4, pi. 2, figs. 3, 4; pi. 3, fig. 1.

1968 Dactylioceras ( Orthodactylites ) semicelatum (Simpson); Hoffmann, p. 6, pi. 2, figs. 1, 2.

1968 Dactylioceras ( Orthodactylites ) eikenbergi Hoffmann, p. 8, pi. 1, fig. 2.

1968 Dactylioceras ( Orthodactylites ) wunnenbergi Hoffmann, p. 7, pi. 1, fig. 1.

1968 Dactylioceras ernsti Lehmann, p. 46, pi. 17, figs. 5, 6; pi. 19, figs. 2, 4.

1971 Dactylioceras ( Orthodactylites ) anguinum (Reinecke); Pinna and Levi-Setti, p. 90, pi. 2,
figs. 1,2,5.

1971 Dactylioceras ( Orthodactylites ) semicelatum (Simpson); Pinna and Levi-Setti, p. 90, pi. 2,
figs. 3,4, 15.

1973 Dactylioceras ( Orthodactylites ) semicelatum (Simpson); Howarth, p. 262, pi. 6, fig. 1; pi. 7,
figs. 1, 2; pi. 8, figs. 1-4; pi. 9, figs. 1-3.

Occurrence. Dorset coast: Marlstone Rock Bed layer P, fairly frequent; Somerset: bed 2 (Hamlet 1922) at
Barrington, Ilminster, about four specimens known; north Oxfordshire and Northamptonshire: abundant in the
top of the Marlstone Rock Bed at many localities from south and east of Banbury to Byfield, Daventry, and
Northampton; Leicestershire: abundant in the top of the Marlstone Rock Bed in the Tilton area and common at
Harston; Lincoln: shales of beds 12 and 14 (Howarth 1958, p. xi); north Lincolnshire and south Humberside:
shales 2-3-4 0 m above the Marlstone Rock Bed at Kirton in Lindsey and Roxby.

Discussion. The occurrence of Dactylioceras ( Orthodactylites ) semicelatum in the Grey Shales
Formation of the Yorkshire coast has already been described in detail by Howarth (1973, p. 262), and

EXPLANATION OF PLATE 80

Figs. 1-12. Dactylioceras (Orthodactylites) semicelatum ( Simpson). All from top 0-1 m of the Marlstone Rock
Bed (‘Transition Bed’), Semicelatum Subzone, Tenuicostatum Zone, of the Banbury area, Northamptonshire.
1 , 2, 5, 6, King’s Sutton, 6 km SE of Banbury, BM C. 67697, C.67376. 3, 4, Middleton Cheney, 4 km ENE of
Banbury, originally figured Buckman (1926a, pi. 654) as holotype of Orthodactylites directus, IGS GSM
47847. 7, 8, Adderbury, 6 km SSE of Banbury, IGS GSM 22566. 9, 10, Chipping Warden, 10 km NE of
Banbury, BM C. 67388. 11, 12, Copredy, 6 km north of Banbury, originally figured J. Sowerby (1819, p. 41,
pi. 222, fig. 1), paralectotype of Ammonites annulatus , BM C.40125. All figures x 1.

PLATE 80

howarth, ammonite Dactylioceras

648

PALAEONTOLOGY, VOLUME 23

reference should be made to that paper for an account of the type specimen, the diagnosis and the
general description of the species. Outside Yorkshire, the commonest occurrence is in the topmost
part of the Marlstone Rock Bed (the ‘Transition Bed’) in Northamptonshire and Leicestershire. The
name Orthodactylites directus Buckman (1926<z, pi. 654) has always been used for these examples
previously (including Howarth 1973, pp. 266-267). However, analysis of the west Northamptonshire
fauna shows that it agrees closely with the Yorkshire fauna of D. ( O .) semicelatum in all characters.
Whorl dimensions and rib-density are expressed graphically in text-figs. 2 and 3, where it can be seen
that there are no significant differences from the Yorkshire fauna, and that the Yorkshire holotype
occupies an approximately central position within the variation of the Northamptonshire fauna. An
average specimen from the Marlstone Rock Bed of Northamptonshire is figured in Plate 80, figs. 1,2,
an example with higher whorls and more rectiradiate ribs in Plate 80, figs. 5, 6, and a more involute
example with higher whorls in Plate 80, figs. 9, 10. Text-figs 2 and 3 also show that the holotype of D.
directum (PI. 80, figs. 3, 4) is an extreme form being more evolute, more compressed, and more finely
ribbed than most Northamptonshire specimens. Nevertheless, it does fall within the range of
variation of the population, and it matches some Yorkshire specimens closely (e.g. Howarth 1973,
pi. 8, fig. 1), so the specific name directum should be placed in synonymy with semicelatum.

The only west Northamptonshire specimen that is more finely ribbed is one of the paralectotypes of
Ammonites annulatus J. Sowerby (1819, pi. 222, fig. 1). Previously (Howarth 1973, p. 262) it was
determined as D. ( O .) tenuicostatum, but, although it is evolute and finely ribbed (PI. 80, figs. 11, 12),
it has the characteristic compressed oval (not near-circular) whorls and prorsiradiate ribs of D. ( O .)

text-fig. 2. Scatter diagrams of whorl dimensions
(whorl height, whorl breadth, and umbilical width,
plotted against diameter) for fifty-nine specimens of
Dactylioceras ( Orthodactylites ) semicelatum (Simp-
son) from the top of the Marlstone Rock Bed in north
Oxfordshire, Northamptonshire, and Leicestershire.
The dashed lines are the envelopes of these points,
while the solid lines are the envelopes of the scatter
diagrams of the Yorkshire coast population of the
same species (from Howarth 1973, p. 259, fig. 5).

HOWARTH: AGE OF MARLSTONE ROCK BED

649

semicelatum and is matched closely by several specimens from bed 30 in Yorkshire (e.g. Howarth
1973, pi. 8, figs. 1, 2, and BM C. 77304). Another west Northamptonshire specimen was made the
holotype of O. mitis Buckman (1927a, pi. 738): it also is not typical of the Northamptonshire fauna,
being more evolute than most specimens and it has flat whorl sides and widely spaced ribs near the
aperture (text-figs. 2, 3; PI. 8 1 , figs. 7-9). It is an incomplete immature specimen 44 mm diameter, and
it is matched very closely by two specimens from bed 28 in Yorkshire and by some from Harston,
Leicestershire (e.g. PI. 82, figs. 11, 12). These are only another form in the variation of the species,
with a different combination of characters, being evolute with fewer ribs, and O. mitis should also be
placed in synonymy with D. ( O .) semicelatum. A larger west Northamptonshire example with similar
widely spaced ribs is figured in Plate 80, figs. 7, 8. It is one of the few complete adults that are known
from the Marlstone Rock Bed, and has a mouth border at 55 mm diameter. A specimen from Tilton,
Leicestershire (C. 80278), has a mouth border at 54 mm diameter, and two other Northamptonshire
and Harston specimens are 97 and 99 mm diameter at their adult mouth borders respectively. This
54-99 mm range compares with an adult diameter range of 75- 1 20 mm for the Y orkshire coast fauna.
A typical example from the top 0T m of the Marlstone Rock Bed at Tilton is figured in Plate 81,
figs. 5, 6. Two small and indifferently preserved Dorset coast specimens were figured previously
(Howarth 1957, p. 197, pi. 17, figs. 5, 6); a large, better preserved example is figured in Plate 81,
figs. 3, 4, which is a typical involute specimen with the high, oval whorls of the species. At Barrington,
Somerset, specimens occur in a bed about 0-2 m above the Marlstone Rock Bed, and the best one is
figured in Plate 81, figs. 1, 2.

The only occurrence of D. ( O .) semicelatum outside Britain that was not dealt with in the Y orkshire
coast paper (Howarth 1973) consists of those specimens in north-west Germany described as D. ernsti
by Lehmann (1968, p. 46, pi. 17, figs. 5, 6; pi. 19, figs. 2, 4; also figured by Hoffmann 1968) and smaller
specimens figured by Hoffmann and Martin (1960). These show all the usual characters of D. (O.)
semicelatum, and the holotype of D. ernsti has whorl proportions and rib-density that are close to the
average of the Yorkshire and Northamptonshire populations. All the north-west German specimens
come from the Semicelatum Subzone, and D. ernsti is considered to be a synonym of D. ( O .)
semicelatum.

text-fig. 3. Scatter diagram of number of ribs per
whorl for seventy-one specimens of Dactylioceras
0 Orthodactylites ) semicelatum (Simpson) from the top
of the Marlstone Rock Bed in north Oxfordshire,
Northamptonshire, and Leicestershire. The dashed
line is the envelope of these points; the solid line is the
envelope of the scatter diagram of the Yorkshire coast
population of the same species (from Howarth 1973,
p. 261, fig. 6).

650

PALAEONTOLOGY, VOLUME 23

Dactylioceras ( Orthodactylites ) tenuicostatum (Young and Bird)

Plate 82, figs. 1-10, 13, 14

1822 Ammonites tenuicostatus Young and Bird, p. 247, pi. 12, fig. 8.

1828 Ammonites annulatus Sowerby; Young and Bird, p. 253, pi. 12, fig. 11.

1884 Stephanoceras annulatum (J. Sowerby); Wright, p. 475, pi. 84, figs. 7, 8.

1920a Dactylioceras tenuicostatum (Young and Bird); Buckman, pi. 157.

1927a Tenuidactylites tenuicostatus (Young and Bird); Buckman, pi. 157A.

1933 Dactylioceras tenuicostatum (Young and Bird); Arkell, pi. 32, fig. 6.

1956 Dactylioceras tenuicostatum (Young and Bird); Arkell, pi. 33, fig. 6.

1957 Dactylioceras tenuicostatum (Young and Bird); Maubeuge, p. 208, figs. ?41, 42, 43.

1961 Dactylioceras tenuicostatum (Young and Bird); Dean, Donovan, and Howarth, pi. 72, fig. 1.

1973 Dactylioceras ( Orthodactylites ) tenuicostatum (Young and Bird); Howarth, p. 258, pi. 5,
figs. 1,2; pi. 6, figs. 2,3.

Occurrence. Dorset coast: Marlstone Rock Bed layer P, uncommon; Somerset: bed 1 (Hamlet 1922) at
Barrington, Ilminster, poorly preserved crushed specimens; Northamptonshire: Rothersthorpe, one specimen;
Leicestershire: 008-0T3 m below the top of the Marlstone Rock Bed at Harston, abundant; north Lincolnshire
and south Humberside: hard mudstone and 1 m of shales above the Marlstone Rock Bed at Kirton in Lindsey
and Roxby, common.

Discussion. A full account of the type specimen, diagnosis, and the Yorkshire coast fauna is found in
Howarth (1973, pp. 258-262). Outside Yorkshire, D. tenuicostatum is much less widely distributed
than D. semicelatum, and the only substantial collection from the Marlstone Rock Bed was that
obtained from Harston, Leicestershire. About seventy specimens were collected, all of them
immature and less than 60 mm maximum diameter. Most have part of their body chambers preserved
but they are incomplete, and no adult specimens, indicated by constricted mouth borders or
approximated final suture-lines, were found. All have the typical rounded whorl section and fine ribs
of D. tenuicostatum. An immature of average size is figured in Plate 82, figs. 9, 10, and the largest of 58
mm diameter in Plate 82, figs. 1 , 2. The top part of the Marlstone Rock Bed at Harston is highly
condensed, and although the main occurrence of D. semicelatum is higher up, a few specimens of the
latter species are found at the same level as the highest D. tenuicostatum. Specimens of D. semicelatum
are always separable by their higher whorls, oval whorl section, more widely spaced ribs, and by the
considerably thicker whorls in some individuals.

A single well-preserved specimen has already been referred to (p. 641) from the top of the
Marlstone Rock Bed at Rothersthorpe, 5 km south-west of Northampton (PI. 82, figs. 3, 4). It is
immature, 46 mm diameter, and has a body chamber one whorl long. About ten examples of D.
tenuicostatum are known from layer P of the Marlstone Rock Bed on the Dorset coast. Again they are
all immature or inner whorls of less than about 60 mm diameter, and two of the best specimens are
figured in Plate 82, figs. 5-8. In north Lincolnshire the species occurs in the hard mudstone that
overlies the Marlstone Rock Bed, and one of the more complete, though small, specimens is figured in
Plate 82, figs. 13, 14.

EXPLANATION OF PLATE 81

Figs. 1-11. Dactylioceras ( Orthodactylites ) semicelatum (Simpson). 1, 2, bed 2 (Hamlet 1922), 0-2 m above
Marlstone Rock Bed, Barrington Quarry, near Ilminster, Somerset, IGS GSM 31612. 3, 4, Marlstone Rock
Bed layer P, Seatown, Dorset, BM C. 17548. 5, 6, Marlstone Rock Bed, top 01 m, Tilton Railway Cutting,
Leicestershire, BM C. 80277. 7-9, top of Marlstone Rock Bed (‘Transition Bed’), Byfield, Northamptonshire,
originally figured Buckman (1927a, pi. 738) as holotype of Orthodactylites mitis, IGS GSM 38384. 10, 11,
Marlstone Rock Bed, 0 08 m below top, Harston Quarry, north Leicestershire, BM C.80169. All figures x 1.

PLATE 81

howarth, ammonite Dactylioceras

652

PALAEONTOLOGY, VOLUME 23

Dactylioceras ( Orthodactylites ) clevelandicum Howarth
Plate 82, figs. 15, 16

1973 Dactylioceras ( Orthodactylites ) clevelandicum Howarth, pp. 257-258, pi. 3, figs. 1-3; pi. 4,
figs. 1,2; pi. 5, fig. 3.

Occurrence. About eight specimens known in bed 21 at Kirton in Lindsey, north Lincolnshire, and an equivalent
horizon in the near-by Worlaby borehole.

Discussion. The most difficult problem in describing the Tenuicostatum Zone Dactylioceratidae that
occur in the Marlstone Rock Bed area in England is the identification of the well-preserved specimens
in the hard mudstone and the calcareous nodules (bed 21) that overlie the Rock Bed at Kirton in
Lindsey, north of Lincoln. Specimens, though well preserved, are not very numerous, and considered
on their own they could be a condensed mixture of D. semicelatum, D. tenuicostatum , and D. cleve-
landicum. Some limit to the age range can be obtained, however, from the ammonites in the overlying
beds, for crushed Dactylioceras that appear to be D. tenuicostatum occur in the overlying 1-3 m of
shale, then D. semicelatum appears in the next higher 1 -7 m of shale, and unmistakable specimens of
Tiltoniceras occur in the next 3-5 m. These ammonites are in the correct stratigraphical sequence for
the Tenuicostatum and Semicelatum Subzones. So it is probable that only the lower part of the
Tenuicostatum Subzone, together with lower horizons, occurs in the hard mudstone. This mudstone,
and especially the calcareous nodules within it, contains a number of small or fragmentary specimens
of D. tenuicostatum (PI. 82, figs. 13, 14), and also a small collection of larger and better-preserved
individuals that are identified with D. clevelandicum. One of the best specimens from the mudstone at
Kirton in Lindsey is figured in Plate 82, figs. 15, 16. Although it is like D. semicelatum in some
respects, it has rectiradiate ribs, not the prorsiradiate ribs of compressed specimens of D.
semicelatum. Nor does it have the typical oval whorl section of the latter species. Other examples in
this bed have a range of variation from rounded whorls with fine ribs, to depressed whorls with coarse
ribs and tubercles. The few that are measurable all fall within the ranges for whorl dimensions and
rib-density measured for the Yorkshire coast population of D. clevelandicum (Howarth 1973,
pp. 259, 261). No examples of this species have been found anywhere else in the Marlstone Rock Bed
area.

EXPLANATION OF PLATE 82

Figs. 1-10, 13, 14. Dactylioceras ( Orthodactylites ) tenuicostatum (Young and Bird). 1, 2, 9, 10, Marlstone
Rock Bed, 0T m below top, Harston Quarry, north Leicestershire, BM C.80 1 22, 80 100. 3, 4, Marlstone Rock
Bed, immediately below ‘Transition Bed’, Rothersthorpe, 5 km SW of Northampton, BM C.82051. 5-8,
Marlstone Rock Bed layer P, Doghouse Cliff, Seatown, Dorset, NMW 26.135 G124and G5.2. 13, 14, bed 21
(Howarth and Rawson 1965), 0-3 m above Marlstone Rock Bed, quarry 2 km north of Kirton in Lindsey,
north Lincolnshire, BM C. 73560. All figures x 1.

Figs. 11, 12. Dactylioceras ( Orthodactylites ) semicelatum (Simpson). Marlstone Rock Bed, 0-08 m below top,
Harston quarry, north Leicestershire, BM C.80 173, x 1.

Figs. 15, 16. Dactylioceras ( Orthodactylites ) clevelandicum Howarth. Bed 21 (Howarth and Rawson 1965),
0-3 m above Marlstone Rock Bed, quarry 2 km north of Kirton in Lindsey, north Lincolnshire, BM
C. 73561, x 1.

Figs. 17, 18. Dactylioceras ( Orthodactylites ) crosbeyi (Simpson). Marlstone Rock Bed, 0-23 m below top,
Harston quarry, north Leicestershire, BM C.80 170, x 1.

PLATE 82

'/////%

howarth, ammonite Dactylioceras

654

PALAEONTOLOGY, VOLUME 23

Dactylioceras ( Orthodactylites ) crosbeyi (Simpson)

Plate 82, figs. 17, 18

1843 Ammonites crosbeyi Simpson, p. 22.

1855 Ammonites crosbeyi Simpson, p. 58.

1884 Ammonites crosbeyi Simpson, p. 90.

1912a Coeloceras crosbeyi (Simpson); Buckman, pi. 60.

71957 Dactylioceras pseudosemicelatum Maubeuge, p. 193, pi. 3, fig. 6.

71957 Dactylioceras podagrosum Maubeuge, p. 193, pi. 4, fig. 7.

1973 Dactylioceras ( Orthodactylites ) crosbeyi (Simpson); Howarth, p. 255, pi. 1, figs. 2-4; pi. 2,
figs. 1-4.

Occurrence. North Leicestershire: 0-23 m below the top of the Marlstone Rock Bed, Harston quarry, one
specimen.

Discussion. This broken half ammonite is about 74 mm diameter, and the final one-third of a whorl is
probably body-chamber. It has relatively high and broad whorls that are about one-quarter involute,
and the whorl section has an evenly rounded venter. The preservation is mainly as an internal cast, so
the ribbing is of very low relief, and consists of prorsiradiate primary ribs, about half of which
bifurcate at the ventro-lateral edge. The ribs on the venter swing slightly more forwards, and only the
slightest traces of ventro-lateral tubercles are present. At 74 mm diameter the whorl height is 21 -5 mm
and the breadth is 2 TO mm, and these whorl dimensions agree well with those of Yorkshire coast
specimens of D. ( O .) crosbeyi. It compares well with the more compressed, more finely ribbed
examples of the species such as were figured by Howarth (1973, pi. 1, fig. 2; pi. 2, fig. 2). The whorl
height and the amount of overlap of the whorls are both too large for D. ( O .) clevelandicum. The
specimen occurs 0- 1 m below a rich population of D. ( O .) tenuicostatum in the Marlstone Rock Bed at
Harston, a stratigraphical position that agrees with its occurrence in Yorkshire. No trace was found
at Harston of the intervening species D. (O.) clevelandicum. No other examples of D. (O.) crosbeyi
have been found outside Yorkshire.

CONCLUSIONS

The ‘Transition Bed’ is the weathered or altered top of the Marlstone Rock Bed. The main change is
oxidation of the green ferrous minerals to limonite, and associated partial decalcification leaves the
bed crumbly or ‘sandy’ in some places. The weathering occurred partly before deposition of the over-
lying beds in some areas, e.g. Banbury and west Northamptonshire, though at Tilton most of the
weathering is more recent. Another type of alteration that took place before deposition of overlying
beds, was the pyritization of the bed in the Harston area, Leicestershire. There is no evidence that the
bed is otherwise mineralogically different from the Marlstone Rock Bed, and there is no sedimentary
discontinuity at its base. The term Marlstone Rock Bed should be applied to the whole of the bed.

In south Dorset and from north Oxfordshire to south Lincolnshire the Marlstone Rock Bed was
deposited during all the period represented by the Spinatum and Tenuicostatum Zones, and there is
no lithological division between the two zones. The ammonite faunas at the boundary are poor, but
generally the top 1-3 m belongs to the Tenuicostatum Zone and the bottom 3-6 m to the Spinatum
Zone. From north Somerset to south Oxfordshire there is no ammonite evidence for the age of the top
of the bed.

The following ammonite faunas have been found in the Marlstone Rock Bed:

(a) Dactylioceras semicelatum (D. directum is a synonym) and Tiltoniceras antiquum of the
Semicelatum Subzone. This is abundant at many localities and is the fauna of the ‘Transition Bed’.

( b ) D. tenuicostatum of the Tenuicostatum Subzone. Abundant only at Harston, Leicestershire,
present on the Dorset coast, and rare elsewhere.

(c) D. crosbeyi of the Clevelandicum Subzone. One specimen at Harston.

(d) Protogrammoceras paltum of the Paltum Subzone. Only on the Dorset coast.

HOWARTH: AGE OF MARLSTONE ROCK BED

655

(e) Pleuroceras spp. of the Spinatum Zone. Abundant in Dorset, Somerset, and Gloucestershire.
Much rarer from Oxfordshire to north Humberside, but sufficient are known to show that both the
Apyrenum and Hawskerense Subzones are present.

From north of Lincoln to north Humberside deposition of the Marlstone Rock Bed stopped at the
end of the Spinatum Zone, and ammonites of all four subzones of the Tenuicostatum Zone occur in
an overlying, lithologically distinct, hard mudstone and in shales above. The latter are similar to the
Grey Shales Formation of the Yorkshire coast.

The change from the lower, regressive ironstone/limestone facies to the upper transgressive
clays/shales-with-nodules facies (Hallam 1967, pp. 431-440) did not occur simultaneously in Britain.
In Yorkshire it occurred at the top of the Spinatum Zone at the Upper Pliensbachian/Toarcian (i.e.
Middle/Upper Lias) boundary, but from Dorset to south of Lincoln it occurred at the top of the
Tenuicostatum Zone. In a small transitionary area between north of Lincoln and Market Weighton
the change took place in the middle of the Tenuicostatum Subzone. The extent of the time disparity
for the facies change may be judged from the fact that 14 m of Grey Shales Formation on the York-
shire coast were being deposited while 1 -3 m of Marlstone Rock Bed was being deposited in England
south of Lincoln.

Acknowledgements. The author wishes to thank the British Steel Corporation who originally allowed access to
the Harston, Denton Park, and Roxby Quarries. The paper has benefited from discussions with Professor
A. Hallam, and with Mr. M. D. Jones of Leicester City Museum regarding the Tilton and Harston exposures. As
well as the collections at the British Museum (Natural History) (specimens with prefix BM), specimens were also
examined at the Institute of Geological Sciences, London (prefix IGS), through the kindness of Dr. H. C. Ivimey-
Cook, and borrowed from the National Museum of Wales, Cardiff (prefix NMW) through Dr. D. A. Bassett.

REFERENCES

ager, D. v. 1956a. Some new Liassic Terebratuloids. Proc. Geol. Ass. 67, 1-14, pi. 1.

— 19566. A monograph of the British Liassic Rhynchonellidae. Part 1, i-xxvi, 1-50, pis. 1-4. Palaeontogr.
Soc. [Monogr.].

arkell, w. j. 1933. The Jurassic System in Great Britain. Oxford, xii + 681 pp., 41 pis.

— 1947. The Geology of Oxford. Oxford. 267 pp.

— 1956. Jurassic Geology of the World. Edinburgh and London, xv + 806 pp., 46 pis.

buckman, s. s. 1909a- 1930a. Yorkshire Type Ammonites , 1, 2, and Type Ammonites , 3-7. London. 790 pis.

— 19106. Certain Jurassic (Lias-Oolite) strata of south Dorset; and their correlation. Q. Jl geol. Soc. Lond. 66,
52-89.

— 19226. Jurassic Chronology; II. Preliminary studies. Certain Jurassic strata near Eypesmouth (Dorset);
the Junction Bed of Watton Cliff and associated rocks. Ibid. 78, 378-475.

dean, w. T., donovan, d. t. and howarth, M. K. 1961. The Liassic ammonite zones and subzones of the north-
west European Province. Bull. Br. Mus. nat. Hist., Geol. 4, 435-505, pis. 63-75.
fox-strangways, c. 1903. The Geology of the country near Leicester. Mem. geol. Surv. U.K. 122 pp., 2 pis.
hallam, a. 1955. The palaeontology and stratigraphy of the Marlstone Rock-bed in Leicestershire. Trans.
Leicester lit. phil. Soc. 49, 17-35.

— 1967. An environmental study of the Upper Domerian and Lower Toarcian in Great Britain. Phil. Trans.
R. Soc. (B) 252, 393-445, pi. 20.

— 1968. The Lias. In sylvester-bradley, p. c. and ford, t. d. (eds.). The Geology of the East Midlands.
Leicester. 201-211.

— 1972. Diversity and density characteristics of Pliensbachian-Toarcian molluscan and brachiopod faunas
of the north Atlantic margins. Lethaia, 5, 389-412.

hamlet, j. 1922. On sections in the Lias exposed in two quarries at Barrington. Proc. Somerset, archaeol. nat.
Hist. Soc. 67, 72-75.

hoffmann, k. 1968. Neue Ammonitenfunde aus dem tieferen Unter-Toarcium (Lias e) des nordlichen
Harzvorlandes und ihre feinstratigraphische Bedeutung. Geol. Jb. 85, 1-32, pis. 1-5.

— and martin, p. R. 1960. Die Zone des Dactylioceras tenuicostatum (Toarcien, Lias) in NW- und
SW-Deutschland. Palaeont. Z. 34, 103-149, pis. 8-12.

howarth, m. k. 1957. The Middle Lias of the Dorset Coast. Q. J! geol. Soc. Lond. 113, 185-204, pi. 17.

656

PALAEONTOLOGY, VOLUME 23

howarth, m. k. 1958. A monograph of the ammonites of the Liassic family Amaltheidae in Britain. Part 1, i-xvi,
1-26, pis. 1-4. Palaeontogr. Soc. [ Monogr .].

— 1973. The stratigraphy and ammonite fauna of the Upper Liassic Grey Shales of the Yorkshire coast.
Bull. Br. Mus. nat. Hist., Geol. 24, 235-277, pis. 1-9.

— 1978. The stratigraphy and ammonite fauna of the Upper Lias of Northamptonshire. Ibid. 29, 235-288,
pis. 1-9.

— and rawson, p. F. 1965. The Liassic succession in a clay pit at Kirton in Lindsey, north Lincolnshire. Geol.
Mag. 102, 261-266.

hull, E. 1857. The geology of the country around Cheltenham. Mem. geol. Surv. U.K. 104 pp., 2 pis.
jackson, J. F. 1922. Sections of the Junction Bed and contiguous deposits. Q. Jl geol. Soc. Lond. 78, 436-448.

— 1926. The Junction-Bed of the Middle and Upper Lias on the Dorset coast. Ibid. 82, 490-525, pis. 33, 34.
lehmann, u. 1968. Strati graphie und Ammonitenfiihrung der Ahrensburger Glazial-Geschiebe aus dem Lias

epsilon (= Unt. Toarcium). Mitt. geol. Stlnst. Hamb. 37, 41-68, pis. 17-20.
maubeuge, p. L. 1957. Les ammonites de la zone a Dactylioceras semicelatum-tenuicostatum dans l’Est de la
France et plus specialement dans le Grande-Duche de Luxembourg. Archs Inst, gr.-duc. Luxemb. (n.s.), 24,
189-226, pis. 1-30.

penny, l. f. and rawson, p. f. 1969. Field meeting in east Yorkshire and north Lincolnshire. Proc. Geol. Ass. 80,
193-218.

pinna, g. and levi-setti, f. 1971. I Dactylioceratidae della Provincia Mediterranea (Cephalopoda,
Ammonoidea). Memorie Soc. ital. Sci. nat. 19, 47-136, pis. 1-12.
pringle, J. and templeman, a. 1922. Two new sections in the Middle and Upper Lias at Barrington, near
Ilminster, Somerset. Q. Jl geol. Soc. Lond. 78, 450-451.

Richardson, G. 1979. The Mesozoic stratigraphy of two boreholes near Worlaby, South Humberside. Bull,
geol. Surv. Gt Br. 58.

Richardson, l. 1906. On a section of Middle and Upper Lias rocks near Evercreech, Somerset. Geol. Mag. (5)
3, 368-369.

— 1909. On some Middle and Upper Lias sections near Batcombe, Somerset. Ibid. (5) 6, 540-542.

— 1929. The country around Moreton in Marsh. Mem. Geol. Surv. U.K. 162 pp.
simpson, M. 1843. A Monograph of the Ammonites of the Yorkshire Lias. London. 60 pp.

— 1855. The Fossils of the Yorkshire Lias; Described from Nature. London and Whitby. 149 pp.

1884. Ibid. London and Whitby. 2nd edn., xxiii + 256 pp.

sowerby, j. 1819. The Mineral Conchology of Great Britain. London. Vol. 3, pis. 222-253.

spath, l. f. 1922. Upper Lias succession near Ilminster, Somerset. Q. Jl geol. Soc. Lond. 78, 449-450.

— 1942. The ammonite zones of the Lias. Geol. Mag. 79, 264-268.

— 1956. The Liassic ammonite faunas of the Stowell Park Borehole. Bull. geol. Surv. Gt Br. 11, 140-164,
pis. 9, 10.

Thompson, b. 1889. The Middle Lias of Northamptonshire. London. 150 pp. (Reprinted from Midi. Nat. 8
(1885)— 12 (1889)).

— 1892. Report of the Committee ... to work on the very fossiliferous Transition Bed between the Middle
and Upper Lias in Northamptonshire. Rep. Br. zlsi. Advmt Sci. for 1890 (Cardiff 1891), 334-351 (Reprinted
J. Northampt. nat. Hist. Soc. 7, 35-57 (1892)).

trueman, a. E. 1918. The Lias of south Lincolnshire. Geol. Mag. (5) 5, 64-73, 101-111.
walford, a. e. 1878. On some Middle and Upper Lias beds, in the neighbourhood of Banbury. Proc. Warwick.
Nat. Archaeol. Fid Club, suppl. for 1878, 1-23.

— 1899. The Lias Ironstone of North Oxfordshire ( around Banbury). London and Banbury. 36 pp.
whitehead, t. h. et al. 1952. The Liassic Ironstones. Mem. geol. Surv. U.K. 21 1 pp., 8 pis.

wilson, e. and crick, w. d. 1889. The Liassic Marlstone of Tilton, Leicestershire. Geol. Mag. (3) 6, 296-305,
337-342, pis. 9, 10.

woodward, H. B. 1893. The Lias of England and Wales (Y orkshire excepted). The Jurassic Rocks of Britain, 3.
Mem. geol. Surv. U.K. 399 pp.

wright, T. 1884. A monograph on the Lias ammonites of the British Islands. Part 7, 441-480, pis. 78-87.
Palaeontogr. Soc. [Monogr.].

young, G. M. and bird, j. 1822. A Geological Survey of the Yorkshire Coast: Describing the Strata and Fossils
occurring between the Humber and the Tees, from the German Ocean to the Plain of York. Whitby. 336 pp.,
17 pis.

1828. Ibid. 2nd edn., enlarged. Whitby. 368 pp., 17 pis.

M. k. howarth

Department of Palaeontology

Typescript received 20 March 1979 British Museum (Natural History)

Revised typescript received 10 October 1979 Cromwell Road, London, SW7 5BD

JURASSIC ARAUCARI AN CONE FROM
SOUTHERN ENGLAND

by RUTH A. STOCKEY

Abstract. A well-preserved araucarian cone measuring 4-5 x 5-0 cm is described from Jurassic age limestone
from near Osmington Mills, Dorset. Four pieces of cone material representing a single specimen are somewhat
flattened and lignitic, with intact seed and cone-scale tissues. The cone axis and bract apophyses are replaced
with a calcitic matrix. Helically arranged cone-scale complexes with a prominent ligular sulcus surround a wide
pith. One recurved wingless ovule 0-8 cm long is deeply sunken into the cone-scale tissue. Seed integuments are
relatively mature and contain three distinct layers the most prominent of which is the sclerotesta constructed of
interlocking zig-zag sclereids. The nucellus, in some cases still cellular, is free from the integuments except at the
chalaza and has the characteristic wavy apex common to extant araucarians at a comparable developmental
stage. A well-developed vascular system like that in Araucaria bidwillii Hooker is present near the seed chalaza.
Cellular megagametophyte and embryos are present within some seeds. The specimen is described as a new
species, A. brownii sp. nov. in which the cone structure most closely resembles that of the section Bunya of the
genus Araucaria. This discovery extends the range of this section to the Northern Hemisphere during the
Mesozoic.

The Araucariaceae, an extant conifer family with a very restricted distribution, has often been con-
sidered primitive among conifer families. Two genera, Agathis and Araucaria, grow as natives in
South America, Australia, New Caledonia, New Guinea, and a few South Pacific islands. Although
the group has only a few relict species today, it was at one time widespread and included
numerous species in the Northern Hemisphere during the Mesozoic Era. Araucarian cones display
what have been suggested as primitive characters that readily distinguish them from those of other
conifer groups (Wieland 1935; Thomson 1905a, 1907, 1913; Eames 1913; Wilde and Eames 1948,
1952; Burlingame 1913, 1914, 1915; Chamberlain 1935; Hirmer 1936; Seward and Ford 1906). The
genus Araucaria, usually believed to be the more primitive of the two genera, has ovulate cones with
large bracts and partially fused ovuliferous scales. Agathis exhibits cone-scales composed of com-
pletely fused bracts and scales, believed to be a derived condition (Eames 1913). Well-preserved fossil
conifer cones of any type are rare; however, a few well-preserved araucarian fossils have been found,
and these have revealed important information about the geologic history of this family, its
distribution, and reproductive biology (Kendall 1949; Wieland 1935; Darrow 1936; Calder 1953;
Stockey 1975, 1977, 1978; Vishnu-Mittre 1954). The uniquely preserved fossil conifer cone reported
here is closely compared with other fossil and living araucarians. The information obtained has
proved useful in elucidating phylogenetic trends within the family, and in particular, within the genus
Araucaria.

MATERIALS AND METHODS

The cone was found in 1973 by Mr. P. A. Brown of Dorset in a block of limestone lying on the
beach under Black Head, west of Osmington Mills. Specimens are lignitic, in a matrix best described
as a compacted bio-pel-micrite with pelecypod shells, fecal pellets, tests of foraminifera, and corals
which have all undergone a considerable amount of diagenesis. The recrystallized calcite composing
these fragments is held together with a CaC03 cement that has infiltrated many of the preserved cone
parts including the seeds. Specimens were prepared for study by a modified coal ball peel technique
(Joy, Willis, and Lacey 1956) using 76 ^m cellulose acetate sheets and by thin sections after epoxy

I Palaeontology, Vol. 23, Part 3, 1980, pp. 657-666, pis. 83-86.)

658

PALAEONTOLOGY, VOLUME 23

infiltration of the cut face. Some cone parts were examined after gold sputter coating using an
AMR 1000 scanning electron microscope at 20 kV. A few whole seeds and cone fragments were
demineralized in 2% HC1 overnight and washed in distilled water. These parts were then embedded in
glycol methacrylate and sectioned with a rotary microtome after a technique by Robison and Miller
(1975).

Since the cone was not found in place its exact age needs some discussion. The area west of
Osmington Mills, Dorset, to Black Head where the cone was found consists of cliffs composed of
Corallian and Kimmeridge Clay sediments. These correspond to the Oxfordian and Kimmeridgian
Stages respectively of the Upper Jurassic (Arkell 1947). The cone most likely comes from the
Osmington Oolite Series which is exposed along the shore west to Shortlake. The beds within this
series contain several clay layers with nodules, in addition to the oolites and marlstones (Arkell 1947).
The matrix surrounding the specimen closely resembles these beds. Its age is therefore certainly
Upper Jurassic, and probably Upper Oxfordian.

SYSTEMATIC DESCRIPTION

Order coniferales
Family araucariaceae
Genus araucaria de Jussieu, 1789
Section bunya Wilde and Eames, 1952
Araucaria brownii sp. nov.

Plates 83-86

Diagnosis. Ovulate cone, 4-5 x 5-0 cm diameter, pith 1-4 cm diameter near point of attachment to peduncle,
expanding to 3-2 cm wide near centre of cone. Cortical resin canals present. Winged cone-scales 1-7 cm long x
1 • 1 cm wide. Bract and ovuliferous scale free for two-thirds of length, both with a system of resin canals.
Ovules 0-8 cm long x 0-3 cm wide, wingless, embedded in ovuliferous scale tissue with micropyle oriented
towards cone axis; one seed per cone-scale complex. Seed integuments with prominent branched sclereids of
sclerotesta arranged in a zig-zag pattern. Complex system of vasculature at ovule chalaza. Nucellus with wavy
apex free from integuments except at base, 0-2 mm thick. Megaspore membrane thin (7 ^m) and discontinuous.
Megagametophyte composed of polygonal cells 30-50 /xm in diameter.

Holotype. British Museum (Natural History) London, V59205, and one fragment housed at Corfe Castle
Museum, Dorset.

Etymology. This cone is named after Mr. P. Anthony Brown of Corfe Castle, Dorset who discovered the
specimen and made it available for study.

EXPLANATION OF PLATE 83

Figs. 1-8. Araucaria brownii sp. nov. Holotype, BMNH V59205 from Osmington Mills, Dorset, b, bract;
end, endotesta; i, integument; Is, ligular sulcus; m, megagametophyte; n, nucellus; os, ovuliferous scale;
scl, sclerotesta. 1 , cone axis region showing position of ovules, and arrangement of cone-scale complexes,
x 1. 2, reverse side of cone in fig. 1, showing cone axis region, rhomboidal cone-scale complexes, and
flattened nature of cone, x 1. 3, tangential portion showing seed transverse sections, xl. 4, cone portion
with part of axis and many closely spaced cone-scale complexes, x 1 . 5, reverse side of portion in fig. 4, show-
ing limestone nodule matrix and numerous cone-scale complexes with ovules, x 1. 6, V59205 B 25. Cone
tangential section with ovule transverse sections. Note large calcite crystals replacing most bract tissue, x 7.
7, V59205 B 21, longitudinal section of cone-scale showing the separation of bract and scale resulting in a
wide ligular sulcus, x23. 8, peel of V59205 B 6, ovule micropylar end showing well-developed seed

integuments, megagametophyte tissue and wavy nucellar apex, x 15.

PLATE 83

stockey, Araucarian cone

660

PALAEONTOLOGY, VOLUME 23

Description. Cone represented by four pieces which are slightly flattened. Plate 83, figs. 1 and 2 show both sides
of central portion of cone revealing position of cone axis and numerous helically arranged, seed-bearing cone-
scales. Plate 83, figs. 4 and 5 show an external portion of the cone. This piece was attached to that shown in
PI. 83, fig. 1 with an epoxy before sectioning. Plate 83, fig. 3 shows a third cone fragment, a tangential cone
piece belonging to the same cone. The remaining part of this specimen was retained by Mr. Brown and is
housed at Corfe Castle Museum in Dorset.

Cone measures 4-5 x 5-0 cm in diameter and was probably spherical in shape before burial. Pith of cone axis
reaches a width of 3-2 cm and approaches 1-4 cm in the most basal portions. Little organic material pre-
served in central part of cone, and no peduncle is present in available material. Organic remains are found in
some cases in a region corresponding to the cortex of the cone axis (PI. 86, fig. 6). Small parenchymatous cells are
often found in groups surrounding resin canals, some of which contain a dark opaque substance (PI. 86,
fig. 6, arrows). A few isolated tracheids have been identified in the cone axis region; these exhibit scalariform
secondary wall thickenings. Unfortunately these tracheids are isolated and their exact position within the axis
stele is questionable.

Cone-scales measure 1-7 cm long and IT cm wide, and are distinctly winged. On first examination the cone
appears similar in size, shape, and appearance to those of the genus Agathis Salisbury. The bract and scale,
however, are free for most of their length, a character typical of species of Araucaria (PI. 83, fig. 7). Vascular
system of the ovuliferous scales consists of at least four bundles (PI. 84, fig. 5). Bract apophyses are replaced by
coarsely crystalline calcite making vascular bundles generally difficult to observe. There is a system of resin
canals in the scale as well as the bract, although their number and placement is difficult to determine because
the cone-scales are flattened.

Ovules of A. brownii measure 0-8 cm in length by 0-3 cm in diameter (PI. 84, figs. 1, 4). One wingless seed
per cone-scale complex deeply embedded in ovuliferous scale tissue with its micropyle oriented towards the cone
axis. Although both Agathis and Araucaria have one seed per scale, only seeds of Araucaria are wingless. Ovules
show an advanced state of integumentary development. The sarcotesta, or outer layer, is represented by a thin
layer of crushed cells 20 pm thick (PI. 85, fig. 1). The middle sclerotesta or stony layer is quite thick (up to
0-2 cm) and is composed of thick-walled branched sclereids, each about 30 pm in diameter (PI. 83, fig. 8;
PI. 84, fig. 2; PI. 85, fig. 5). These cells are hexagonal in transverse section (PI. 85, fig. 3) with very small lumens
and thick walls. In many cases the walls are no longer distinguishable, but the entire cell, or layer, has been
replaced by calcite (PI. 85, figs. 1, 3). The endotesta, or inner integumentary layer, is thin, up to three cells
in thickness and often crushed (PI. 85, fig. 1). In places where it is present, the cells are short and often barrel-
shaped with relatively thin walls (PI. 85, fig. 2). Integumentary differentiation within the ovules indicates a nearly
mature developmental stage.

The non-adnate nature of the nucellus and integument is shown in all of the ovules examined (PI. 83, fig. 8;
PI. 84, fig. 7). Where well preserved the nucellus has a thick cuticle (PI. 83, figs. 6, 8; PI. 84, figs. 3, 6;
PI. 86, figs. 1, 2). It is basally attached to the inner integumentary layer and appears somewhat shrunken
(PI. 86, fig. 2). The nucellar apex appears convoluted as in living araucarians (PI. 83, fig. 8). In extant plants
it protrudes out of the micropyle at the time of pollination and later may retract by drying or by subsequent
integumentary growth (Eames 1913). Some authors (Darrow 1936; Eames 1913) have suggested that the con-
voluted apex was the result of pollen tube damage; however, on examining some living araucarian ovules, a
disruption of the apex by pollen tube damage seems unlikely. Plate 85, fig. 4 shows the wavy nucellar apex of the
extant A. montana Brongn. et Gris, and is typical of most known araucarians at this stage of development. The
configuration of the apex appears to represent a drying phenomenon rather than the result of extensive pollen

EXPLANATION OF PLATE 84

Figs. 1-7. Araucaria brownii sp. nov. Holotype, BMNH V59205. b, bract; i, integument; m, megagameto-
phyte; n, nucellus; os, ovuliferous scale; s, seed. 1 , isolated seed, x 1 0. 2, paradermal section of sclerotesta

cells showing zig-zag cell arrangement. Arrows indicate branched sclereids, x 380. 3, isolated nucellus with

apex removed, x 17. 4, V59205 B 6, cone longitudinal section showing sunken nature of seeds within the
cone-scale complex, x 7. 5, V59205 B 25 transverse section of ovule showing lateral vascular bundle of

ovuliferous scale outside of the seed integuments, x 37. 6, V59205 A 40, ovule transverse section showing

thick wavy nucellus and cellular megagametophyte tissue, x 90. 7, V59205 B 26, ovule transverse section

showing the relationship of integument, nucellus, and cellular megagametophyte, x 85.

PLATE 84

stockey, Araucarian cone

662

PALAEONTOLOGY, VOLUME 23

tube damage. The ease with which the nucellus is removed from isolated ovules is due to its narrow attachment
as well as its thick cuticle (PI. 84, fig. 3). External surface shows little cellular detail while internally outlines of
elongate, rectangular cells are visible (PI. 85, fig. 7).

Ovule attachment and vascularization are difficult to determine with peels. However, thin sections of the
chalaza of some ovules where preservation of the bract and ovuliferous scale is partial show a well-developed
system of conducting cells (PI. 86, fig. 5). Using scanning electron microscopy, the chalazal end of each ovule
exhibits a series of small holes penetrating the sclerotesta (PI. 84, fig. 6, arrows), as in the living A. bidwillii
Hooker. There appears to be what Wilde and Eames (1948, p. 326) term a vascular ‘plexus’, a complex system of
vascularization near the ovule base. Holes in the mature seed integuments correspond to points of entry of
numerous vascular bundles. The same type of attachment also occurs in A. mirabilis from the Jurassic Cerro
Cuadrado Petrified Forest (Stockey 1975).

Most ovules show some tissue remains inside the nucellar cavity. The megaspore membrane of some ovules
is thin (7 fin l) (PI. 84, fig. 7), and similar to that in living araucarian cones at a comparable stage of
development (Eames 1913; Burlingame 1915; Thomson 19056). Many of the ovules reveal preservation of tissues
within this membrane. In most, megagametophyte tissue is either poorly preserved or represented by a free
nuclear stage of development at the time of preservation (PI. 84, fig. 8; PI. 86, figs. 2, 7). Other ovules show
cellular preservation of the megagametophyte (PI. 84, fig. 6; PI. 86, fig. 4). These polygonal cells (30-50 fim in
diameter) occur in the outer portions of the megagametophyte proper. No ovules have been found with solid
megagametophyte tissue preserved within the seed cavity. However, some ovules do show two distinct regions
of poorly preserved tissue (PI. 86, fig. 2). The boundary between the two regions appears to be a discontinuous
layer. In other ovules (PI. 86, fig. 4) the cellular megagametophyte and a centrally located region probably repre-
senting the embryo are replaced by calcite. Other specimens contain a four-parted cellular structure that may
represent an embryo with four cotyledons (PI. 86, fig. 3, arrows). An alternate possibility is that this may have
been a partially formed megagametophyte at the time of preservation. The seed itself is not crushed, even
though parts of the integuments are very crumbly.

DISCUSSION

The spherical shape of the Osmington Mills cone with helically arranged cone-scales, large pith in the
cone axis region, cortical resin canals, and presence of one seed per ovuliferous scale are general
characteristics of cones from the family Araucariaceae. The presence of a ligular sulcus (space
between the ovuliferous scale tip and the bract) and wingless seeds suggest affinities with the genus
Araucaria. The cone of A. brownii was apparently in a relatively mature state of development at the
time of fossilization. The appearance of the three-layered integument with a thin sarcotesta, thick
sclerotesta composed of elongate branched sclereids, and a thin layer of endotesta composed of thin-
walled cells is characteristic of araucarian cones near maturity (Eames 1913; Burlingame 1915; Wilde
and Eames 1948; Stockey 1978). The seed contents including the configuration of the nucellus and
thin megaspore membrane support this view. A cellular megagametophyte is present with a hollow
central cavity (PI. 86, figs. 2, 4) that probably represents the remains of an embryo rather than free
nuclear megagametophyte. This embryo may have aborted or more likely deteriorated prior to
preservation since it is likely that the cone remained floating in water for some time prior to its

EXPLANATION OF PLATE 85

Figs. 1 -7. Araucaria brownii sp. nov. Holotype, BMNH V59205, scanning electron micrographs, end, endo-
testa; n, nucellus; sar, sarcotesta; scl, sclerotesta. 1 , ovule transverse section showing three integumentary
layers at a late developmental stage, x 90. 2, cells of the endotesta, x 425. 3, transverse section of integument
showing hexagonal sclerotesta cells completely replaced by calcite, x 400. 4, A. montana Brongn. et Gris.,
nucellar apex from a mature seed, x 100. 5, surface view of sclerotesta with sarcotesta removed showing

elongate interlocking sclereids, x 425. 6, seed chalaza, surface of sclerotesta. Arrows indicate holes of the

ovular vascular bundles in the plexus, x 90. 7, elongate nucellar cells, x 450.

PLATE 85

stockey, Araucarian cone

664

PALAEONTOLOGY, VOLUME 23

burial. Plate 86, fig. 3 may show the cellular remains of such an embryo. Extant araucarian cones,
at the time when free nuclear megagametophyte is present (about the time of pollination), show
ovules with the integumentary layers of approximately equal thickness with little if any expansion
and thickening of the sclerotesta (Wilde and Eames 1948; Stockey 1978).

According to Wilde and Eames (1952) the living genus Araucaria may be divided into four sections:
Columbea , Bunya, Eutacta, and Intermedia based on seedling morphology, foliage type, and cone
morphology. The differences in cone structure between the section Columbea and the other three
sections is distinct, while the differences between the other three are more subtle. The Columbea
species found only in South America have wingless cone-scales, the bract and scale are nearly com-
pletely fused and never separate after the scales are shed from the cone axis. Sections Eutacta,
Intermedia, and Bunya have winged cone-scales, those of Intermedia being widest, up to 10 cm in
A. klinkii Lauterb. (White 1947). Seeds in the closely related Eutacta and Intermedia sections are
never removed from the tightly fused bract and scale. The winged cone-scale complex is the unit of
dispersal for these species. Seeds from A. bidwillii, the only living member of the section Bunya, are
easily removed from the cone-scales which are more fleshy than those of the Eutacta and Intermedia
sections. These seeds are often dispersed by birds and small animals in Queensland but are easily
removed from the cone-scale complex even though the cone disaggregates as in the other species.

A. mirabilis from the Jurassic Cerro Cuadrado Petrified Forest is also included within the Section
Bunya (Calder 1953; Stockey 1975, 1978). These ovulate cones are only one-third as large as those of
A. bidwillii at maturity and have winged cone-scales, a zig-zag pattern of sclereids in the sclerotesta
of the seed integument, a complex system of vasculature at the seed chalaza, and a dicotyledonous
embryo characteristic of A. bidwillii (Wilde and Eames 1948; Stockey 1975, 1978). There is some
evidence to suggest that these cones shed their seeds and not their scales at maturity (Stockey
1978).

Another ovulate cone which can be considered to be closely related to these three Bunya species is
Araucarites bindrabunensis (Vishnu-Mittre 1954) from the Jurassic of India which shows a slightly
larger size than most Araucaria mirabilis cones, and is also larger than the cone from Osmington
Mills. The origin of the cone-scale complex vascular supply, nature of the winged cone-scales,
presence of a ligular sulcus, and vascularization of the ovuliferous scale tip (ligule) place it in the
section Bunya.

The cone described here from Osmington Mills in Dorset should also be considered under the
section Bunya of the genus Araucaria. The winged cone-scales, zig-zag sclereid pattern and vascular
plexus, and deep ligular sulcus are similar to A. mirabilis and A. bidwillii. These comparisons extend
the range of the section Bunya into the Northern Hemisphere where it was probably widespread
during the Jurassic and Cretaceous.

Acknowledgements. I thank Dr. C. R. Hill, British Museum (Natural History), and Mr. P. A. Brown for making
the material available for study; and Dr. T. N. Taylor for review of the manuscript and use of laboratory
facilities. This publication is dedicated to the memory of the late James M. Schopf. I acknowledge NSERCC
Grant A-6908.

EXPLANATION OF PLATE 86

Figs. 1-7. Araucaria brownii sp. nov. Holotype, BMNH V59205. e, embryo; i, integument; m, megagameto-
phyte; n, nucellus; p, plexus. 1, V59205 B 24, ovule transverse section showing lateral extensions of the seed
integuments and well-preserved endotesta cells, x9. 2, V59205 A 40, seed transverse section with well-
preserved nucellus, megagametophyte, and a possible embryo, x 25. 3, V59205 B 23, transverse section

of seed showing possible embryo with four cotyledons, x 35. 4, Y59205 B 25, transverse section of seed with
cellular megagametophyte and crystalline central area, possibly representing an embryo, x 50. 5, V59205

B 26, transverse section near seed chalaza showing vascular plexus leading into ovule, x 15. 6, V59205

C 8, longitudinal section of resin canal (arrows) in cortex of cone axis, x 30. 7, transverse section of ovule
showing disorganized megagametophyte tissue, x 925.

PLATE 86

stockey, Araucarian cone

666

PALAEONTOLOGY, VOLUME 23

REFERENCES

arkell, w. J. 1947. Geology of the Country around Weymouth, Swanage, Corfe and Lulworth. Mem. Geol. Survey
of Great Britain, 1-386.

Burlingame, L. L. 1 9 1 3. The morphology of Araucaria brasiliensis. I. The staminate cone and male gametophyte.
Bot. Gaz. 55, 97-114.

- — 1914. The morphology of Araucaria brasiliensis. II. The ovulate cone and female gametophyte. Ibid. 57,
490-508.

— 1915. The morphology of Araucaria brasiliensis. Fertilization, the embryo and the seed. Ibid. 59, 1-38.
calder, M. G. 1953. A coniferous petrified forest in Patagonia. Bull. Brit. Mus. (Nat. Hist.) Geol. 2, 99-138.
chamberlain, c. j. 1935. Gymnosperms: Structure and Evolution. Univ. Chicago Press.
darrow, B. s. 1936. A fossil araucarian embryo from the Cerro Cuadrado of Patagonia. Bot. Gaz. 98, 328-337.
eames, A. J. 1913. The morphology of Agathis australis. Ann. Bot. 27, 191-204.

hirmer, M. 1936. Die Bliiten der Coniferen. I. Entwicklungsgeschichte und Vergleichende Morphologie des
weiblichen Bliitenzapfens der Coniferen. Biblio. Bot. 23, 1-100.
joy, K. w., willis, A. j. and lacey, w. s. 1956. A rapid cellulose peel technique in palaeobotany. Ann. Bot.
(n.s.), 20, 635-637.

kendall, m. w. 1949. A Jurassic member of the Araucariaceae. Ibid. 13, 151-161.

robison, c. r. and miller, c. n. jr. 1975. Glycol methacrylate as an embedding medium for lignitic plant
fossils. J. Paleont. 49, 559-561.

seward, A. c. and ford, s. o. 1906. The Araucarieae, recent and extinct. Phil. Trans., London ( B ), 198,
305-411.

stockey, R. A. 1975. Seeds and embryos of Araucaria mirabilis. Am. J. Bot. 62, 856-868.

— 1977. Reproductive biology of the Cerro Cuadrado (Jurassic) fossil conifers: Par araucaria patagonica. Ibid.
64, 733-744.

— 1978. Reproductive biology of Cerro Cuadrado fossil conifers: ontogeny and reproductive strategies in
Araucaria mirabilis (Spegazzini) Windhausen. Palaeontographica, B 166, 1-15.

Thomson, R. B. 1905a. Preliminary note on the Araucarineae. Science (N.S.), 22, 88.

— 19056. The megaspore-membrane of the gymnosperms. Univ. Toronto Studies Bio. Ser. 4, 85-146.

— 1907. The Araucarieae— A ‘Proto-Siphonogamic’ method of fertilization. Science (n.s.), 25, 271-272.
— 1913. On the comparative anatomy and affinities of the Araucarineae. Phil. Trans. R. Soc. Lond. (B),
204, 1-50.

white, c. t. 1947. Notes on two species of Araucaria in New Guinea, and a proposed new section of the
genus. J. Arnold Arboretum, 28, 259-260.

wieland, G. R. 1935. The Cerro Cuadrado Petrified Forest. Publ. Carnegie Inst., Washington. 449 pp.
wilde, m. h. and eames, a. j. 1948. The ovule and ‘seed’ of Araucaria bidwilli with discussion of the taxonomy
of the genus. I. Morphology. Ann. Bot. (N.S.), 12, 311-326.

— 1952. The ovule and ‘seed’ of Araucaria bidwilli with discussion of the taxonomy of the genus.
II. Taxonomy. Ibid. 16, 27-47.

DR. R. A. STOCKEY
Department of Botany
The University of Alberta
Edmonton, Alberta

Typescript received 24 June 1979

Revised typescript received 20 November 1979

NOMENCLATURE AND HOMOLOGY IN
PERIDINIALEAN DINOFLAGELLATE
PLATE PATTERNS

by GEOFFREY L. EATON

Abstract. The apical and antapical series of peridinialean dinoflagellate thecal plates are redefined relative to
the cingulum. They are then compatible with the Kofoidian pre- and postcingular series, and the need to
recognize anterior and posterior intercalary series is removed. The concept of apical closing and antapical
closing series is introduced. Homologous and corresponding plates are recognized in fifteen selected modern and
fossil dinoflagellates by comparing interseries relationships with respect to a model plate pattern. The differences
between the selected patterns are due to three variable effects. First, the reduction in plate number through
simplification, where one plate in one pattern corresponds to two or more plates in another pattern. This
critically affects interseries relationships. Secondly, the primary development of fewer plates without affecting
interseries relationships. Thirdly, the variation in the relative size of certain plates. The interaction of these three
effects resulted in the comparatively independent evolution of epithecae and hypothecae. Reduction in over-all
plate number, particularly through the primary development of fewer plates, may well represent a fundamental
trend in the evolution of peridinialean plate patterns.

The dinoflagellates of the Order Peridiniales Haeckel 1894 are often informally described as
‘armoured’. They are so called because their cell covering includes a layer of rigid, polygonal,
suturally united, cellulosic plates, termed the theca. Text-fig. 1 shows thecal morphology and
nomenclature in two typical peridinialean dinoflagellates, Protoperidinium depressum (Bailey) Balech
1 974 and Gonyaulax spinifera (Claparede and Lachmann) Diesing 1866. The theca is divided into two
parts, epitheca (anterior) and hypotheca (posterior), which are separated by an equatorial groove
termed the cingulum. The two ends of the cingulum are separated on the ventral surface by a more or
less longitudinal groove termed the sulcus. Motility is achieved by the beating of two flagella (not
shown in text-fig. 1) which originate from the sulcus. The transverse flagellum lies within the
cingulum, while the longitudinal flagellum lies within, and extends posteriorly beyond, the sulcus. The
thecal plates are arranged in roughly parallel transverse series. Differences in tabulation; that is the
number, shape, and arrangement (plate pattern) of the thecal plates, have long been used as the main
criterion for taxonomic separation within the Peridiniales. The accepted system of thecal plate
nomenclature was defined by Kofoid (1907, 1909, 1911).

The fossil peridinialean dinoflagellate record ranges back at least 200 million years into the Late
Triassic period. However, in terms of representing the absolute geological history of the Peridiniales,
this record has only limited effectiveness. This is because all fossilized dinoflagellates attributed to the
Peridiniales are non-motile cysts rather than motile thecae, and modern studies show that only a very
small proportion of living peridinialeans produce potentially fossilizable cysts. Comparisons
between modern thecae and fossil cysts show that not all modern plate patterns have been recognized
in the fossil record, and some fossil plate patterns are unknown in modern dinoflagellates. Accepting
the limitations of the fossil record, and the fact that cysts only rarely show full details of their parent
thecal tabulation, it is still possible that the relative distribution of the different plate patterns through
geological time may provide some evidence of trends in plate pattern evolution. The recognition of
such trends is dependent on the critical assessment of the similarities and differences between different
plate patterns. Such an assessment will involve the recognition of homologous plates in different
patterns. In my own studies on fossil dinoflagellates I have found that a strict application of Kofoid’s

[Palaeontology, Vol. 23, Part 3, 1980, pp. 667-688.|

668

PALAEONTOLOGY, VOLUME 23

plate nomenclature often results in apparently homologous plates in different patterns, being
assigned to different transverse plate series. I believe that compatibility between nomenclature and
homology is essential for the recognition of evolutionary trends, and that it can only be achieved by
modifying certain aspects of Kofoid’s system. Discussion of the need for this modification and a way
of effecting it, forms the basis of this paper.

According to Evitt et al. (1976) fossil cyst plate patterns should be discussed in terms of their
paratabulatory nomenclature (paraplates, parasutures, etc.). However, in this paper on modern

VENTRAL SURFACE DORSAL SURFACE

text-fig. 1 . Thecal morphology of two modern peridinialean dinoflagellates. Upper, Protoperidinium depression
(Bailey) Balech 1974. Lower, Gonyaulax spinifera (Claparede and Lachmann) Diesing 1866. Interpretation of
the transverse plate series is conventional Kofoidian. The distribution and number of cingular and sulcal plates
in P. depressum is assumed to be typical of the genus. In G. spinifera only the anterior (a.s.) and posterior (p.s.)

sulcal plates are annotated.

EATON: DINOFLAGELLATE PLATE PATTERNS

669

thecae and fossil cysts I wish to avoid the use of a dual ‘tabulation/paratabulation’ nomenclature.
Therefore I assume that the fossil cyst paraplate patterns are a fair representation of their parent
thecal plate patterns, and treat them all, modern and fossil, simply as plate patterns.

ORIGIN AND DEVELOPMENT OF THE KOFOID SYSTEM OF
THECAL PLATE NOMENCLATURE

Although the system of peridinialean thecal plate nomenclature which has been generally used for the
past seventy years is attributed to Kofoid, it should be remembered that he was clearly influenced by
several nomenclatural systems proposed by earlier workers, e.g. Stein, Biitschli, Schiitt, Paulsen,
Faure-Fremiet (see Kofoid 1909, p. 44). All these earlier workers recognized that thecal plates are
arranged in transverse rows, and that there are four major plate series, two anterior to the equator
and two posterior to the equator. Various names had been applied to these series (see Kofoid 1909, p.
44), but those used by Biitschli (1885) were closest to Kofoid’s subsequent terminology. Biitschli
described the most anteriorly positioned series as apical, and the most posteriorly positioned as
antapical. The two intervening series were termed pre-equatorial (anterior) and post-equatorial
(posterior).

Kofoid recognized seven transverse plate series, comprising the four major series plus the cingular
series and two incomplete intercalary series. Each series was designated by superscript acute accent
marks, figures or letters, or simply by letters. The series were named from apex to antapex as: apical
('), anterior intercalary (a), precingular ("), cingular (c), postcingular ("'), posterior intercalary (p),
antapical (""). The plates in each series were numbered in sequence, anticlockwise (in apical view)
from the ventral surface. Additional plates at the extreme apex or within the sulcus were individually
designated, e.g. apical closing plate (cl. pi.). During subsequent use, Kofoid’s system has remained
unchanged except for the designation of the sulcal plates (s) and the use of various abbreviations to
designate additional individual plates. The typical application of Kofoid’s nomenclature to P. depres-
sion and G. spinifera is shown in text-fig. 1 . These two forms together illustrate all seven of Kofoid’s
transverse plate series. Their respective tabulation formulae are: 4', 3a, 7", 3c, 5"', Op, 2"", 6-7s, and
1 ap. cl., 4', Oa, 6”, 6c, 6"', lp, 1 5s.

The most extensive discussion of thecal plate nomenclature is given in Kofoid (1909, pp. 40-45),
but definitions of the various plate series are also found in Kofoid (1907, 1911).

Kofoid (1907, p. 179) defined the four major plate series with reference to modern Ceratium
Schrank 1793. Kofoid stated: ‘I shall use the term apical for the anterior series of plates only, and
shall designate the series anterior to and contiguous to the girdle [cingulum] as precingular (prec.),
and that posterior to and contiguous to it as postcingular (postc.) and the posterior ones as antapicals
(antap.).’

Kofoid (1909, pp. 26-28) next applied his nomenclature to modern peridiniacean dinoflagellates,
using P. steini (Jorgensen) as an example. His nomenclatural interpretation of P. steini is equally
applicable to P. depressum (text-fig. 1). Kofoid interpreted the apical plates as ‘those whose apical
ends border the apical pore’ (ap. po. in text-fig. 1). The combination of this interpretation of the
apicals and Kofoid’s earlier interpretation of the precingulars leaves three plates unaccounted for on
the dorsal surface. These plates ‘intercalate’ between the apicals and precingulars and were referred to
the anterior intercalary series (la-3a), a term Kofoid had previously used in his original description
of Heterodinium Kofoid 1906.

In his studies on modern Gonyaulax Diesing 1866, Kofoid (1911, p. 194) interpreted the apical
plates as ‘those in contact with the apex’. He recognized that in this genus the apex does not have an
open pore, but is occupied by a small apical closing plate ( 1 ap. cl. in text-fig. 1 ). Kofoid designated as
anterior intercalary those plates anterior to the precingular series but not in contact with the apex.
This series was not recognized in all species of Gonyaulax. Kofoid also introduced the concept of
a posterior intercalary series with reference to Gonyaulax. The single plate (lp) assigned to this
series lies posterior to postcingulars Y" and 2"', and anterior to antapical V" which occupies the
antapex.

670

PALAEONTOLOGY, VOLUME 23

Kofoid realized that thecal plates are arranged in rows roughly parallel to the cingulum. This led
him to use the cingulum rather than the geometric equator as a basis for defining transverse plate
series. This approach recognized the fundamental importance of the structure which divides the theca
into epitheca and hypotheca. Kofoid (1909, p. 43) believed that his recognition of transverse series
throughout the theca clarified the confused situation that had previously existed over the
nomenclature of plates anterior to the precingulars. His beliefs would seem to have been justified
by the subsequent application of his nomenclature to modern dinoflagellates and to fossil forms
ranging back to the Triassic period.

C. HIRUND1NELLA P. DEPRESSUM G. SPINIFERA

text-fig. 2. Conventional Kofoidian interpretation of tabulation in polar views of Ceratium hirundinella
(Muller) Schrank 1793, Protoperidinium depressum (Bailey) Balech 1974, Gonyaulax spinifera (Claparede and
Lachmann) Diesing 1866. Upper, epithecae. Lower, hypothecae.

A problem in designating apparently homologous plates

Although Kofoid did not recognize any intercalary plates in Ceratium, a case can sometimes be made
for a single anterior intercalary in Ceratium hirundinella (Muller) Schrank 1793 (text-fig. 2). In a
particular form of this species. Wall and Evitt (1975, p. 21) designated as apical 4' a plate which does
not reach the tip of the apical horn. They admitted that strictly speaking this plate should be
designated anterior intercalary, but to do so would only lead to confusion. They argued that since the
homology of this plate and the fourth apical of other species of Ceratium is so obvious, it is better to
consider this plate as a shortened apical. In this particular case Wall and Evitt considered the
recognition of ‘obvious’ homology to be more important than the strict application of a definition or
rule.

This approach can also be applied to other dinoflagellates, for instance, the fossil taxon
Hystrichogonyaulax cladophora (Deflandre) Stover and Evitt 1978 and certain species of fossil

EATON: DINOFLAGELLATE PLATE PATTERNS

671

Ctenidodinium Deflandre 1938. In some well-preserved specimens of H. cladophora (text-fig. 3) a plate
can be recognized anterior and adjacent to 3" and 4". This plate does not touch the apical closing
plate (1 ap. cl.) and is therefore designated anterior intercalary la. Two such plates are recognizable
in Ctenidodinium pachydermum (Deflandre) Gocht (1970, pi. 29, fig. 5) and Ctenidodinium sp. (text-
fig. 3), and 2a in this pattern appears to be homologous with la in H. cladophora. Also, 2a in
Ctenidodinium sp. and la in H. cladophora appear to be homologous with apical 3' in Gonyaulax
polyedra Stein 1883 (text-fig. 3). If this interpretation of homology is correct, then this particular
anterior intercalary plate in the two fossil taxa could be interpreted as a shortened apical.

text-fig. 3. Conventional Kofoidian interpretation of epithecal tabulation in Hystrichogonyaulax cladophora
(Deflandre) Stover and Evitt 1978, Ctenidodinium Deflandre 1938 sp., Gonyaulax polyedra Stein 1883.

Kofoid’s (1911, pp. 194-195) own comments on the anterior intercalaries in gonyaulacacean
dinoflagellates are significant here. Kofoid designated as anterior intercalary, plates in the apical
region which are ‘crowded away from contact with the apex ... as well as other plates lying between
the apical and precingular series’. He also considered the two anterior intercalaries lying laterally and
ventrally to the right of the greatly reduced apical 4' in G. polyedra (text-fig. 3) to be plates which had
been ‘crowded away’ from the apex. Kofoid remarked further (Kofoid 191 1, p. 239) that the area of
la had probably ‘split off’ from the edge of apical 4', and he also illustrated one specimen of G.
polyedra (Kofoid 191 1 , pi. 14, fig. 29) in which intercalary 2a actually touches the apical closing plate.
There is no doubt that Kofoid considered intercalaries la and 2a in G. polyedra to be territorially
apical, but their spatial relationship with the extreme apex required their designation as anterior
intercalary.

The anterior intercalary plates in H. cladophora and Ctenidodinium sp. seem to be apicals which
have been shortened and crowded away from the apex, and according to Kofoid’s comments on G.
polyedra their designation as intercalary is entirely justified. Also, it can be argued that Wall and Evitt
should have adopted this approach with C. hirundinella. Apical 4' could be interpreted as being
crowded away from the apex to occupy an anterior intercalary position, and this plate could then be
designated la. This would not lead to the confusion Wall and Evitt suggested. It would simply reflect
the strict application of a universally recognized rule, and any discussion of homologous
relationships with the apical plates of other taxa would be of secondary importance. However,
against this it can be argued that the recognition of homologous plates in different dinoflagellates is in
fact of primary importance, and is critical to the understanding of the evolution of thecal plate
patterns. Therefore, since the Kofoid rules require that apparently homologous plates in different
taxa are assigned to different plate series, Kofoid’s method of defining these series should be
re-evaluated.

672

PALAEONTOLOGY, VOLUME 23

An inconsistency in plate series definition

The foregoing comments are specifically concerned with the anterior intercalary and apical series in
gonyaulacacean epithecae. More important is the concept of these series in peridiniacean dino-
flagellates. The partially developed anterior intercalary series is a characteristic feature of the
peridiniacean plate pattern, and there can be little doubt that Kofoid considered these intercalaries to
be additional plates between the precingulars and apicals. However, a polar view of the peridiniacean
epitheca does not support this interpretation. In Protoperidinium depressum (text-fig. 2) for instance,
the three anterior intercalaries la-3a and apicals T, 2', and 4' form a perfect ring of plates, effectively
concentric with the precingular series. The interpretation of these six plates as the apical series would
leave only Kofoidian apical 3' unaccounted for. Thus Kofoid’s concept of the anterior intercalary
series seems to be an artificial one which resulted directly from his interpretation of the apical series in
peridiniacean dinoflagellates.

I can only speculate on the reasons why Kofoid defined the apical series in the way he did. He may
simply have believed that the apical plates should occupy or at least touch the morphological apex.
He may have been influenced by the fact that his concept of the apical series resulted in the
recognition of four apical plates in Ceratium, peridiniacean taxa, and certain species of Gonyaulax,
and this consistency might be significant. Whatever reason is suggested, one major criticism is
inescapable: Kofoid’s concept of the apical series in peridiniacean dinoflagellates is incompatible
with his basic statement on plate series definition. That is, since the division of the theca into epitheca
and hypotheca is of such fundamental importance, the intervening cingulum should be used as the
basis for defining the transverse plate series (Kofoid 1909, pp. 41, 43).

For consistency, after the precingular series had been defined as the plates anterior to and
contiguous to the cingulum, the next series should have been defined as the plates anterior to and
contiguous to the precingulars. This consistent definition of the apical series would not have affected
Kofoid’s interpretation of Ceratium, but it would have greatly affected his interpretation of the
peridiniacean plate pattern. In P. depressum there would be six apicals rather than four, a residue of
one plate at the apex (Kofoid’s apical 3'), and no anterior intercalaries. In species of Gonyaulax such
as G. polyedra there would be five apicals (Kofoid’s l'-3', la, 2a) rather than four, a residue of two
plates at the apex (1 ap. cl. and Kofoid’s apical 4'), and again no anterior intercalaries. In H. clado-
phora there would be five apicals, in Ctenidodinium sp. there would be six, and the conventional
intercalary plate in both patterns previously suggested to be homologous with apical 3' in G. polyedra
would now be designated apical. Also, apical 4' in C. hirundinella would be designated apical,
independent of its relationship with the morphological apex.

A similar argument can be made against Kofoid’s interpretation of the antapical series in
Gonyaulax and his resulting concept of a posterior intercalary series. After the postcingular series had
been defined as the plates posterior to and contiguous to the cingulum, the next series should have
been defined as the plates posterior to and contiguous to the postcingulars. This consistent definition
of the antapical series would not have affected Kofoid’s interpretation of Ceratium or the basic
peridiniacean plate pattern, but it would have affected his interpretation of Gonyaulax. In the latter
genus, the plate conventionally designated posterior intercalary 1 p would become first antapical 1
conventional \"" would become 2"", and there would be no posterior intercalaries.

Thus initially on the grounds of consistency in plate series definition and some limited evidence of
plate homology, redefinition of the apical and antapical series is justified.

MODIFICATION OF KOFOID’S SYSTEM OF PLATE SERIES NOMENCLATURE

Definition of all the transverse plate series relative to the cingulum generally results in the recognition of two
major plate series on both the epitheca and hypotheca. Any remaining plates occur at or near the poles of the
theca and can be accommodated in a third epithecal or hypothecal series. Although this approach differs from
Kofoid’s concept of transverse plate series, only the apical and antapical series need to be redefined. Also, almost
all of Kofoid’s terms are still applicable and there is only one completely new plate series.

I would emphasize here that the following definitions are only intended to be broad guides to the recognition

EATON: DINOFL AGELLATE PLATE PATTERNS

673

of the various plate series and the designation of individual plates. I do not believe that such definitions should be
rigidly applied. Subjective interpretation is unavoidable, and interplate relationships must be considered for
each plate pattern before individual plates can be assigned to the various plate series.

Definition of the transverse series

Three transverse series are recognized on the epitheca: precingular, apical, apical closing; and three are also
recognized on the hypotheca: postcingular, antapical, antapical closing.

The precingular series (") was satisfactorily defined by Kofoid (1907, p. 179) as the row of plates anterior to
and contiguous to the cingulum. This definition is retained here.

The apical series (') is redefined as the row of plates anterior to and contiguous to the precingular series. Also
included is the plate (or plates) anterior to and contiguous to the sulcal area, as suggested by Kofoid. The apical
series may be interrupted by a posterior extension of the apical closing series and in certain circumstances apical
plates may touch the cingulum (e.g. Helgolandinium subglobosum, text-fig. 7).

The apical closing series (ap. cl.) is defined as the plates anterior to and contiguous to the apical series. The
concept of apical closing plates was discussed by Kofoid (191 1, p. 194) with respect to the small plate occupying
the extreme apex of Gonyaulax. My idea of the apical closing series includes this and any other plates anterior
and contiguous to the apical series, with the term ‘closing’ being used in a geometric rather than a biologically
functional sense. This series may be represented by a distinct row of plates, and in certain circumstances apical
closing plates may interrupt the apical series and touch the precingular series (e.g. Shublikodinium arcticum, text-
fig. 8).

The postcingular series ('") was satisfactorily defined by Kofoid (1907, p. 179) as the row of plates posterior to
and contiguous to the cingulum. This definition is retained here.

The antapical series ("") is redefined as the plates posterior to and contiguous to the postcingular series.

The antapical closing series (an. cl.) is proposed as a new series, and is defined as the plates posterior to and
contiguous to the antapical series. Again, ‘closing’ is used in a purely geometric sense. So far this series has been
recognized only in S. arcticum and Rhaetogonyaulax rhaetica (both text-fig. 8).

Application to selected modern and fossil dinoflagellate plate patterns

The plate patterns of five modern and ten fossil dinoflagellates are illustrated in text-figs. 4-8 as diagrammatic
polar (epithecal and hypothecal) views, in which I have tried to retain true interplate relationships with minimum
distortion of observed plate geometry. This type of illustration is used rather than conventional ventral and
dorsal views (text-fig. 1) because it allows a better appreciation of the geometric relationship between individual
plates or groups of plates. Hypothecae are illustrated with the sulcus to the south rather than the conventional
north, to emphasize certain similarities with epithecae. The plates are numbered in terms of the modified plate
series nomenclature, and the direction of numbering is conventional. Individual cingular and sulcal plates are
not indicated.

The listed data for each pattern include the specific name with its authorship, geological age (where relevant),
the source of the plate pattern, and the modified tabulation formula. The formula is expressed in terms of the
epithecal (E) and hypothecal (H) transverse series only. Changes in plate designation are indicated, with the
reinterpreted designation first, followed by the conventional Kofoidian designation in parentheses. Where
necessary, changes in dinoflagellate cyst archaeopyle nomenclature are indicated for the fossil taxa, in terms of
the notation previously discussed by Evitt (1967) and Stover and Evitt (1978).

Gonyaulax spinifera (Claparede and Lachmann) Diesing 1 866, Recent, text-fig. 4.

From: Kofoid (1911, text-figs, a-d) and Wall and Dale (1970, text-figs. 19-22).

Modified tabulation formula, E: 1 ap. cl., 4', 6"; H: 6"', 2'"'.

Changes in plate designation: 1"” (lp), 2"" (1"")-

Gonyaulax polyedra Stein 1883, Recent, text-fig. 4.

From: Kofoid (1911, pi. 12, figs. 16-20).

Modified tabulation formula, E: 2 ap. cl., 5', 6"; H: 6'", 2"" .

Changes in plate designation: 2 ap. cl. (4'), 4', 5' (la, 2a), V" (lp), 2"" (1"").

Hystrichogonyaulax cladophora (Deflandre) Stover and Evitt 1978, Late Jurassic, text-fig. 4.

From: Deflandre (1938, text-figs. 5, 6) and my own observations.

Modified tabulation formula, E: 1 ap. cl., 5', 6"; H: 6'", 2"" .

Changes in plate designation: 3' (la), 4', 5', (3', 4'), 1"" (lp), 2"" (1'"')-

674

PALAEONTOLOGY, VOLUME 23

Ctenidodinium Deflandre 1938 sp., Middle Jurassic, text-fig. 5.

From: my own observations.

Modified tabulation formula, E: 2 ap. cl., 6', 6"; H: 6'", 2"" .

Changes in plate designation: 3', 4' (la, 2a), 5', 6' (3\ 4'), 1"" (lp), 2."" (1"").

Paragonyaulacysta Johnson and Hills 1973 s.l. , Middle Jurassic, text-fig. 5.

From: Johnson and Hills (1973, text-fig. 9) and my own observations.

Modified tabulation formula, E: 1 ap. cl., 5', 6"; H: 6"', 2"" .

Changes in plate designation: 3'-5' (la-3a), 1"" (lp), 2"” (1"").

Cyst archaeopyle type: dorsal apical, type A, 2A or 3A (conventionally intercalary, type I, 21 or 31).
Luehndea spinosa Morgenroth 1970, Early Jurassic, text-fig. 5.

From: Morgenroth (1970, pi. 9, figs. 1-4), Evitt (unpublished data).

Modified tabulation formula, E: 1 ap. cl., 6', 6"; H: 6"', 2"" .

Changes in plate designation: 1 ap. cl. (3'), 3'-5' (la-3a), 6' (4'), 1"" (lp), 2"" ( 1

Canninginopsis denticulata Cookson and Eisenack 1962, Mid Cretaceous, text-fig. 6.

From: Cookson and Eisenack (1962, text-fig. 2), Wall and Evitt (1975, text-fig. 11).

Modified tabulation formula, E: 1 ap. cl., 4', 6"; H: 6'", 2”''.

Changes in plate designation: 1"" (lp), 2"" (1"").

text-fig. 4. Modified interpretation of tabulation in polar views of Gonyaulax spinifera (Claparede and
Lachmann) Diesing 1866, Gonyaulax polyedra Stein 1883, Hystrichogonyaulax cladophora (Deflandre) Stover
and Evitt 1978. Upper, epithecae. Lower, hypothecae.

EATON: DINOFLAGELL ATE PLATE PATTERNS

675

Ceratium hirundinella (Muller) Schrank 1793, Recent, text-fig. 6.

From: Wall and Evitt (1975, text-figs. 5, 6).

Modified tabulation formula, E: 4', 6"; H: 6"', 2"".

Changes in plate designation: 1"" (lp), 2"" (1""), relative to Wall and Evitt (1975).

Thalassiphora delicata Williams and Downie 1966, Eocene, text-fig. 6.

From: Eaton (1976, text-figs. 18, 20).

Modified tabulation formula, E: 1 ap. cl., 4', 6"; H: 6"', 2'"'.

Changes in plate designation: 1 ap. cl. (4'), 4' (la), 2'"' (lp). Also, the sixth pre- and postcingulars are now
recognized.

Protoperidinium depressum (Bailey) Balech 1974, Recent, text-fig. 7.

From: Gocht and Netzel (1974, text-fig. 1).

Modified tabulation formula, E: 1 ap. cl., 6', 7"; H: 5'", 2"".

Changes in plate designation: 1 ap. cl. (3'), 3'-5' (la-3a), 6' (4').

Phthanoperidinium tritonium Eaton 1976, Eocene, text-fig. 7.

From: Eaton (1976, text-fig. 24).

Modified tabulation formula, E: 1 ap. cl., 6', 7"; FI: 5'”, 2"".

Changes in plate designation: 1 ap. cl. (3'), 3'-5' (la-3a), 6' (4').

Cyst archaeopyle type: dorsal apical, type A (conventionally intercalary, type I).

text-fig. 5. Modified interpretation of tabulation in polar views of Ctenidodinium Deflandre 1938 sp.,
Paragonyaulacysta Johnson and Hills 1973 s.l., Luehndea spinosa Morgenroth 1970. Upper, epithecae.

Lower, hypothecae.

676

PALAEONTOLOGY, VOLUME 23

Helgolandinium subglobosum von Stosch 1969, Recent, text-fig. 7.

From: von Stosch (1969, text-fig. 3).

Modified tabulation formula, E: 5', 7"; H: 1"', 3"”.

Changes in plate designation: T (1"), 2'-5' (l'-4'), T'-7" (2"-8").

The very small plate designated 9" by von Stosch (1969, text-fig. 3/) which lies anterior to the anterior sulcal
plate and touches the first cingular plate is here referred to the sulcus as a second anterior sulcal plate, 2 a.s. A
similarly positioned plate can sometimes be recognized in Paragonyaulacysta s.l.

Dapcodinium priscum Evitt 1961, Early Jurassic, text-fig. 8.

From: Evitt (1961, text-figs. 1-20).

Modified tabulation formula, E: 1 ap. cl., 1', 7"; H: 7'", 3"".

Changes in plate designation: 1 ap. cl. (3'), 3'-6' (la-4a), 7' (4'), 2"'-7"' (T"-6,,,)J 1"", 2"" (lp, 2p), 3"" (1"").
On the hypo theca, Evitt originally recognized only six postcingulars, but several of his drawings (Evitt 1961,
text-figs. 5- 10, 1 5- 1 7, 1 9) show an undesignated plate comparable in position to the reduced first postcingular of
Gonyaulax. This plate is designated T" here, and the number of postcingulars is increased from six to seven.

Shublikodinium arcticum Wiggins 1973, Late Triassic, text-fig. 8.

From: Wiggins (1973, text-fig. 3).

Modified tabulation formula, E: 6 ap. cl., 6', 7"; H: 7"', 3"", 1 an. cl.

Changes in plate designation: 1 ap. cl. (va. cl.), 2-6 ap. cl. ( 2'-6 '), 2'-6' (la-5a), 1 an. cl. (ppl).

Cyst archaeopyle type: combination apical closing-apical, type tACL tA (conventionally combination apical-
intercalary, type tA tl).

text-fig. 6. Modified interpretation of tabulation in polar views of Canninginopsis denticulata Cookson and
Eisenack 1962, Ceratium hirundinella ( Muller) Schrank 1793, Thalassiphora delicata Williams and Downie 1966.
Upper, epithecae. Lower, hypothecae.

P. DEPRESSUM

text-fig. 7. Modified interpretation of tabulation in polar views of Protoperidinium depressum (Bailey) Balech
1974, Phthanoperidinium tritonium Eaton 1976, Helgolandinium subglobosum von Stosch 1969. Upper, epithecae.

Lower, hypothecae.

text-fig. 8. Modified interpretation of tabulation in polar views of Dapcodinium priscum Evitt 1961,
Shublikodinium arcticum Wiggins 1973, Rhaetogonyaulax rhaetica (Sarjeant) Loeblich and Loeblich 1968.
Upper, epithecae. Lower, hypothecae.

678

PALAEONTOLOGY, VOLUME 23

Rhaetogonyaulax rhaetica (Sarjeant) Loeblich and Loeblich 1968, Late Triassic, text-fig. 8.

From: Harland, Morbey and Sarjeant (1975, text-fig. 2).

Modified tabulation formula, E: 6 ap. cl., 7', 7"; H: 7'", 3"", 1 an. cl.

Changes in plate designation: 1-6 ap. cl. (l'-6'), 1' (a.v.), 2'-7' (la-6a), 1"", (lp), 1 an. cl. (1"").

Cyst archaeopyle type: combination apical closing-apical, type tACL tA (conventionally combination apical-
intercalary, type tA tl).

HOMOLOGOUS AND CORRESPONDING PLATES IN SELECTED
DINOFLAGELLATE PLATE PATTERNS

Method

The method used here for recognizing homologous and corresponding plates in different plate
patterns involves critical comparisons of their interplate relationships. These comparisons are made
with respect to a model plate pattern (text-fig. 9) whose epitheca and hypotheca both show a very high
degree of radial symmetry and plate regularity. The model plate series are interpreted in terms of the
modified nomenclature, but to emphasize the model nature of the pattern, the Kofoidian style
notation is not applied. Instead, the series are referred to as ‘ap.’ (apical), ‘prec.’ (precingular),
‘postc.’ (postcingular) and ‘antap.’ (antapical). Also, the plates in each series are simply numbered
consecutively 1, 2, 3 etc., and referred to as ap. 1, postc. 3, etc. The idea of this model plate pattern
is based on the following observations.

In the fifteen epithecal patterns illustrated in text-figs. 4-8, the maximum number of plates in any of
the transverse series is seven (7', e.g. D. priscum\ 7", e.g. P. depressum). Also, the most regular
interseries relationship involves groups of three plates (3-plate relationship) with one plate in one
series touching two plates in an adjacent series. This relationship is fully developed between the
precingular and apical series in L. spinosa and D. priscum for instance, and between six of the apicals
(2'-T) and five of the apical closing plates (2-6 ap. cl.) in R. rhaetica.

The model epitheca shows a full development of the 3-plate interseries relationship in a pattern
with seven plates in all three transverse series. The direction of numbering the epithecal plates is
conventional, i.e. anticlockwise relative to the apical pole. When an observed epithecal pattern does
not show counterparts of all the model plates, it is the highest numbered model plates which are
considered to be unrepresented.

A similar interpretation of such observations on the fifteen hypothecal patterns illustrated in text-
figs. 4-8 would lead to a model pattern closely comparable to S. arcticum and R. rhaetica. The

text-fig. 9. The model plate pattern. Left, epitheca, E: 7 ap. cl., 7 ap., 7 prec. Right,
hypotheca, H: 7 postc., 7 antap., 1 an. cl.

EATON: DINOFLAGELLATE PLATE PATTERNS

679

maximum number of plates in any of the transverse series is again seven, but this maximum only
occurs in the postcingular series (e.g. H. subglobosum). The maximum number of antapicals is only
three (e.g. D. priscum), with a single antapical closing plate developed only in S. arcticum and
R. rhaetica. All four patterns with seven postcingulars and three antapicals ( H . subglobosum,
D. priscum, S. arcticum, R. rhaetica) show a constant relationship between these two series. This
involves groups of four plates (4-plate relationship) with one antapical touching three postcingulars.
However, this arrangement can be interpreted in terms of the fundamental 3-plate relationship (fully
developed in the model epitheca), if each of the three antapicals is treated as two plates, and the
intervening mid-ventral area is also considered to be an antapical plate.

Thus the model hypotheca shows a full development of the 3-plate relationship between the
postcingulars and antapicals, with seven plates in both series. The model hypothecal plates are
numbered anticlockwise relative to the antapical pole. This is the reverse of convention, but is more
convenient for the discussion of observed hypothecal patterns which do not show counterparts of all
the model plates. As with the epithecae, it is the highest numbered model plates which are considered
to be unrepresented.

In text-fig. 9, four plates on both the epitheca and hypotheca are ornamented. I consider these to be
key reference plates in the discussion of homologous plate relationships and the comparison of
different plate patterns.

For ease of comparison with the model pattern, the epithecae and hypothecae of the selected
modern and fossil dinoflagellates are discussed and illustrated separately (text-figs. 10-12). The
modified system of plate series nomenclature is applied throughout with a Kofoidian style notation.

text-fig. 10. Epithecal plate patterns of six selected dinoflagellate taxa. The key reference areas are ornamented.

680

PALAEONTOLOGY, VOLUME 23

The direction of numbering the plates is conventional. In the text-figures the interseries boundaries
are thickened for emphasis, and only those plates critical to the discussion are numbered. A
maximum of four such plates on both the epitheca and hypotheca are also ornamented, either in full
or in part. Each ornamented area corresponds to one key plate in the model pattern.

Epithecal plate patterns (text-figs. 9, 10, 11)

The model epithecal pattern (text-fig. 9) has twenty-one plates arranged in three series (7 ap. cl., 7 ap.,
7 prec.). This discussion is primarily concerned with the apical and precingular series in which the key
reference plates are ap. 1, ap. 4 (stippled), and prec. 4, prec. 6 (shaded).

Rhaetogonyaulax rhaetica and Dapcodinium priscum (text-fig. 10) both have a 7', 7" pattern, and
also show a full development of the 3-plate relationship. Because of this, the seven apicals and seven
precingulars in both patterns are considered to be respectively homologous with ap. 1-7 and
prec. 1-7.

In Shublikodinium arcticum (text-fig. 10) which has a 6' 1" pattern, the 3-plate relationship is lost in
the vicinity of precingular 2" which is touched by only one apical, 2'. Compared with the model
pattern, plate 2' occupies the position of two plates, ap. 2 and ap. 3. Thus apical T in S. arcticum
corresponds to two plates in the model pattern, and as a result, apical 3' in S. arcticum is homologous
with key ap. 4. The interruption of the apical series in S. arcticum by a posterior extension of the
apical closing series between T and 2', does not affect the over-all interpretation of the 3-plate
relationship. Phthanoperidinium tritonium (text-fig. 10) also has a 6', 7" pattern, but the 3-plate
relationship is lost in the vicinity of 4”. Thus apical 4' corresponds to two plates in the model pattern,
and only its stippled area corresponds to key ap. 4. In the 6', 1" pattern of Protoperidinium depressum
(text-fig. 10), the 3-plate relationship is lost in the vicinity of three consecutive precingulars, 3" -5”.
Even so, I still consider that only apical 4' corresponds to two plates in the model pattern, and this
accounts for the loss of the 3-plate relationship between the apical series and 4”. The further loss in
the vicinity of 3" and 5" is due to a relative lateral displacement of the two intra-apical sutures which
border 4'. Thus only the stippled area of 4' in P. depressum corresponds to key ap. 4. In
Helgolandinium subglobosum (text-fig. 10) which has a 5', 7" pattern, the 3-plate relationship is lost in
the vicinity of two separated precingulars, 2" and 5". Thus apicals 2' and 4' each correspond to two
plates in the model pattern, and 3' is homologous with key ap. 4.

In Luehndea spinosa (text-fig. 11) there are only six plates in both the apical and precingular series
(6', 6" pattern), but the 3-plate relationship is maintained throughout. Because of this the twelve
plates comprising the apical and precingular series are considered to be respectively homologous with
ap. 1-6 and prec. 1-6. Thus the highest numbered model plates, ap. 7 and prec. 7, have no
counterparts in L. spinosa. This last comment also applies to the eight remaining patterns discussed
here.

Ctenidodinium sp. (text-fig. 1 1) has a 6', 6" pattern, but the 3-plate relationship is lost in the vicinity
of two consecutive precingulars, 2" and 3". Plate 2" has one apical touching it, 2', while 3" has three
apicals touching it, 2,-4/. The loss of the 3-plate relationship is due to the critical shortening of 3',
accompanied by the relative enlargement of 2'. Plate 4' is also shortened, but not critically. As in
L. spinosa , the twelve plates comprising the apical and precingular series in Ctenidodinium sp. are
respectively homologous with ap. 1-6 and prec. 1-6.

In Gonyaulax polyedra and Hystrichogonyaulax cladophora (text-fig. 11) which both have a 5', 6”
pattern, the 3-plate relationship is lost in the vicinity of precingular 2". Applying the principle used in
the interpretation of S. arcticum, P. tritonium etc., apical 2' corresponds to two plates in the model
pattern. Thus 3' in G. polyedra and H. cladophora is homologous with key ap. 4. The shortening of 3'
in H. cladophora is not critical, and is comparable to 4' in Ctenidodinium sp. In Par agony aulacysta s.l.
(text-fig. 11) which also has a 5', 6" pattern, the 3-plate relationship is lost in the vicinity of
precingular 6". Thus only the stippled area of 1 ' corresponds to key ap. 1 , and the unornamented area
of T corresponds to ap. 6. The shortening of apicals 3' -5' in Par agony aulacysta s.l. is not critical.

Gonyaulax spinifera, Canninginopsis denticulata, and Ceratium hirundinella (text-fig. 1 1) all have a
4\ 6” pattern and identical homologous and corresponding plate relationships. Loss of the 3-plate

EATON: DINOFLAGELLATE PLATE PATTERNS

L. SPINOSA CTENIDODINIUM S P. G. POLYEDRA

H. CLADOPHORA PA R A G O N Y A U L A C Y S T A S.L. G. SPINIFERA

C. DENTICULATA

C. HIRUNDINELLA

T. DELICATA

text-fig. 1 1 . Epithecal plate patterns of a further nine selected dinoflagellate taxa. The key reference areas are

ornamented.

682

PALAEONTOLOGY, VOLUME 23

relationship occurs in the vicinity of precingulars 2" and 4". Thus 2' and 3' each correspond to two
plates in the model pattern, and therefore the stippled area of 3' corresponds to key ap. 4. The further
loss of the 3-plate relationship in G. spinifera is due to the critical shortening of 6". The shortening of
4' in C. hirundinella is not critical.

In Thalassiphora delicata (text-fig. 1 1) which also has a 4', 6" pattern, the 3-plate relationship is lost
in the vicinity of three precingulars, 2", 5", and 6". Thus 2' and 4' each correspond to two plates in the
model pattern, and 3' is homologous with key ap. 4. Plate 6” is critically shortened and is not touched
by any of the apical series.

Only limited comments can be made on possible homologous relationships in the apical closing
series, because of the great variation in its development. It is reasonable to suggest that 1 ap. cl. in
R. rhaetica and S. arcticum corresponds to the first and seventh apical closing plates in the model
pattern, and that the five remaining plates in all three patterns are respectively homologous. Also that
the large single apical closing plate in D. priscum, L. spinosa, P. tritonium, and P. depressum
corresponds to all seven apical closing plates in the model pattern. However, the relationship between
these two extremes of development, and the apical closing plates in G. spinifera and Ctenidodinium sp.
for instance, is undetermined.

Hypothecal plate patterns (text-figs. 9, 12)

The model hypothecal pattern (text-fig. 9) has fifteen plates arranged in three series (7 postc., 7 antap.,
1 an. cl.). This discussion is only concerned with the postcingular and antapical series in which the key
reference plates are postc. 4, postc. 6 (shaded) and antap. 3, antap. 6 (stippled). The fifteen selected
hypothecae do not show the same range of variation as their epithecae, and can be discussed in terms
of only eight patterns.

The ‘rhaetogonyaulacacean’ type ( Rhaetogonyaulax rhaetica, Shublikodinium arcticum), Helgo-
landinium subglobosum, and Dapcodinium priscum (text-fig. 12) all have a 1"' , 3"" pattern. The seven
postcingulars in these patterns and the model pattern are respectively homologous. The antapical-
postcingular interseries relationship is constant, with each antapical touching three postcingulars. As
stated earlier, each antapical corresponds to two plates in the model pattern. Since occupy

the position of antap. 2-1, corresponds to antap. 6-7, 2"" to antap. 4-5, and 3"" to antap. 2-3.
Thus, only the stippled area of 3"" corresponds to key antap. 3, and only the stippled area of Y'"
corresponds to key antap. 6.

Although no counterpart of antap. 1 is shown in the 1'", 3"” patterns, it may be represented by the
posterior area of the sulcus. However, in fossil dinoflagellates the individual sulcal plates are often
very poorly defined, and it is by no means certain that all areas designated posterior sulcal (p.s.) are
respectively homologous. Consequently this relationship is only questionably applied to the fifteen
selected patterns (Table 1).

A significant feature of D. priscum is the anticlockwise rotation of its plate pattern relative to the
model pattern, and the associated reduction of 2"' and Y". These reduced plates correspond to the
highest numbered model postcingulars, postc. 6, postc. 7, and this influenced my earlier statement
that when a hypothecal pattern does not show counterparts of all the model plates, it is the highest
numbered model plates which are unrepresented. Consequently it is convenient to discuss the plate
relationships of the five remaining patterns in an anticlockwise direction, i.e. in terms of 6' "-Y"
and 2""-Y"'.

In Luehndea spinosa there are only six postcingulars and two antapicals ( 6 2"" pattern). Plates
6”'-Y" are respectively homologous with postc. 1 -6, and there is no counterpart of postc. 7. Thus 3"'
is homologous with key postc. 4, and 1 is homologous with key postc. 6. Antapical 2'"' touches four
postcingulars and therefore corresponds with three plates in the model pattern, antap. 2-4. Thus only
the stippled area of 2'"' corresponds to key antap. 3. Plate l"" touches three postcingulars and
therefore corresponds to two plates in the model pattern, antap. 5, 6. Thus only the stippled area of
Y'" corresponds to key antap. 6, and there is no counterpart of antap. 7.

In the ‘peridiniacean’ type pattern ( Protoperidinium depressum and Phthanoperidinium tritonium )
there are only five postcingulars and two antapicals (5'", 2"" pattern). Plates 5"'-r" are respectively

EATON: DINOFL AGELL ATE PLATE PATTERNS

683

TY PE H. SUBGLOBOSUM

D. PRISCUM

PERIDINIACEAN
L. SPINOSA TYPE

TYPE

C. HIRUNDIN ELLA

DELICATA

text-fig. 12. Eight hypothecal plate patterns representative of the fifteen selected dinoflagellate taxa. The key
reference areas are ornamented.

684

PALAEONTOLOGY, VOLUME 23

table 1. Homologous and corresponding plates in the fifteen selected
plate patterns and the model pattern. The modified interpretation of
transverse series is applied throughout. Plates with an asterisk would be
designated intercalary using the conventional Kofoidian system.

EATON: DINOFLAGELLATE PLATE PATTERNS

685

homologous with postc. 1 -5 and there are no counterparts of postc. 6, 7. Thus 2"' is homologous with
key postc. 4. Antapicals 2"" and V"' each touch three postcingulars and therefore each corresponds
to two plates in the model pattern. Thus only the stippled area of 2"" corresponds to key antap. 3, and
there are no counterparts of antap. 6, 7.

The ‘gonyaulacacean’ type pattern ( Gonyaulax spinifera, G. polyedra, Hystrichogonyaulax
cladophora, Ctenidodinium sp., Canninginopsis denticulata, Paragonyaulacysta s.l.) with six post-
cingulars and two antapicals (6"', 2"" pattern) differs from L. spinosa only in showing considerable
reduction of 1 2"', and 1 The interseries relationships of the gonyaulacacean pattern are directly
comparable with L. spinosa. Thus 3'" is homologous with key postc. 4, and Y" is homologous with
key postc. 6. Also, only the stippled area of 2"" corresponds to key antap. 3 and only the stippled area
of Y"' corresponds to key antap. 6.

Ceratium hirundinella also has a 6'", 2"" pattern, but is unusual in that postcingular 6"' touches
V" as well as 2"" . A possible explanation of this relationship is that 6'" in C. hirundinella actually
represents two plates, 6'" s.s. and the conventional posterior sulcal plate (p.s. in text-fig. 1). The latter
plate has not been precisely identified in C. hirundinella , although it must be admitted that the sulcal
area of Ceratium in general is poorly known (Wall and Evitt 1975, p. 19). Plates 5 ”'-Y" are
respectively homologous with postc. 2-6, and there is no counterpart of postc. 7. Thus 3"' is
homologous with key postc. 4, and 1 is homologous with key postc. 6. Antapical 2''" touches five
postcingulars and therefore corresponds to four plates in the model pattern, antap. 2-5. Thus only
the stippled area of 2"" corresponds to key antap. 3. Plate \"" touches V" and 2"' and is therefore
homologous with key antap. 6. There is no counterpart of antap. 7.

Thalassiphora delicata also has a 6"', 2"" pattern, but its interseries relationships effectively
represent a lateral reversal of the gonyaulacacean arrangement. This affects the antapicals, where 2”"
only touches three postcingulars and therefore corresponds to two plates in the model pattern, and
V" touches four postcingulars and therefore corresponds to three plates in the model pattern. Thus
only the stippled area of 2"" corresponds to key antap. 3, and only the stippled area of 1""
corresponds to key antap. 6.

The interpreted homologous and corresponding plate relationships of the fifteen selected plate
patterns and the model pattern (excluding the apical closing, cingular and antapical closing series) are
summarized in Table 1 . This emphasizes the fact that plates conventionally designated intercalary (*)
in one pattern, are homologous with or partially correspond to conventional apical or antapical
plates in another pattern. Examples include apical 4' (conventional 2a) in Protoperidinium depressum
homologous with apical 3' in Gonyaulax spinifera , and antapical 1 "" (conventional 1 p) in G. spinifera
which partially corresponds to part of antapical V” in Helgolandinium subglobosum. This table also
shows that when a series is represented by the same number of plates in different patterns, the plates
need not all be respectively homologous. Examples include the four apicals in G. spinifera and
T. delicata, and the two antapicals in G. spinifera, T. delicata, and P. depressum.

DISCUSSION

The differences between the fifteen selected plate patterns (text-figs. 10-12) reflect an over-all trend of
reduction in the total number of plates. This reduction is effected in two particular ways. There may
be simplification through the development of a single plate in one pattern which spatially corresponds
with two or more plates in another pattern. This critically affects interseries relationships.
Alternatively there may be a primary development of fewer plates in particular series, without
affecting interseries relationships. These two styles of reduction may occur independently or together.
They may also be accompanied by variation in the relative size of certain plates which may affect
interseries relationships through critical shortening or lateral reduction.

Reduction through simplification affects the apical closing, apical and antapical series. The
development of a single large apical closing plate in D. priscum for instance, represents simplification
of the six-plate arrangement in R. rhaetica and S. arcticum. In the apical series, simplification occurs

686

PALAEONTOLOGY, VOLUME 23

in specific areas, e.g. mid-ventral (L) in Par agony aulacysta s.l.; left lateral (2') in S. arcticum; mid-
dorsal (4') in P. tritonium; left and right lateral, (2', 3') in G. spinifera, (T, 4') in T. delicata; left ventral
and right lateral (2', 4 ') in H. subglobosum. In the antapical series, simplification is best defined with
reference to the model antapicals, antap. 2-7. For instance, the three antapicals in H. subglobosum
reflect simplification of antap. 2-7 in the form 2-3, 4-5, 6-7. In the patterns with only two antapicals
this simplification takes several forms, e.g. 2-3, 4-5 (peridiniacean type), 2-4, 5-6 (gonyaulacacean
type), 2-5, 6 (C. hirundinella), 2-3, 4-6 (T. delicata ).

Reduction through the primary development of fewer plates is best defined with reference to the
position of plates which are homologous with or in part correspond to model key reference plates.

In the six epithecae in text-fig. 10 ( R . rhaetica etc.) the position of these plates is virtually constant.
In particular, the equivalent of key prec. 4. is invariably mid-dorsal. A similar constancy is shown by
eight of the epithecae in text-fig. 1 1 ( L . spinosa etc., but not T. delicata). However, in these patterns
the equivalent of key prec. 4 is invariably right dorso-lateral in position. Compared with R. rhaetica
etc. (text-fig. 10), this represents a rotation of the epithecal pattern, anticlockwise relative to the
apical pole. This accommodates the primary development of one less apical and one less precingular
plate (i.e. no counterparts of ap. 7, prec. 7).

In the eight hypothecae in text-fig. 12 there is considerable variation in the position of the key
reference areas. In particular, the equivalent of key postc. 4 rotates from mid-dorsal in the
rhaetogonyaulacacean type and H. subglobosum, through left dorso-lateral in D. priscum and
L. spinosa, to left lateral in the gonyaulacacean and peridiniacean types. This rotation, which affects
the over-all hypothecal pattern, is anticlockwise relative to the antapical pole. This accommodates
the primary development of fewer postcingulars and antapicals (i.e. no counterparts of two or more
of postc. 6, 7, antap. 6, 7).

The effect of variation in the relative size of certain plates is well shown by the epithecae of
Ctenidodinium sp., H. cladophora. Par agony aulacysta s.l. and G. spinifera (text-fig. 1 1), in which there
is enlargement of the lateral and left ventral precingulars (compared with L. spinosa ). This is at the
expense of the apical series which becomes longitudinally aligned, and 6” which is reduced. In the
somewhat bizarre pattern of T. delicata (text-figs. 11, 12) the considerable enlargement of 1 " and 2"',
3"' is accommodated by the displacement of the sulcus, 1 ', 4', 1 6"' and \"" , 2"" to a right lateral
position. There is also reduction of 5" and 6"', and 6 " is critically shortened. Other examples of
critical shortening include 3' in Ctenidodinium sp. and 6" in G. spinifera, while 4' in P. depressum is
critically reduced laterally.

Reduction in the total number of thecal plates may well represent a fundamental trend in the
evolution of peridinialean plate patterns. If this is so, then available evidence from the fossil record
suggests that primary development of fewer plates was the most important means of achieving this
reduction. This evidence is provided for epithecae by the appearance in the Late Triassic of patterns
with counterparts of ap. 7, prec. 7 ( R . rhaetica, S. arcticum), followed in the Jurassic by the
appearance of patterns without counterparts of these two plates (e.g. L. spinosa, Ctenidodinium sp.).
The hypothecal evidence is provided by the successive appearance of the rhaetogonyaulacacean type
(Late Triassic), D. priscum and L. spinosa (Early Jurassic), gonyaulacacean type (Middle Jurassic)
and the peridiniacean type (Late Jurassic). The great range of variation shown by Late Triassic and
younger plate patterns resulted from the effects of reduction through simplification, and variation in
relative plate size, being superimposed on the effect of primary reduction. The interaction of these
three variables resulted in epithecae and hypothecae evolving comparatively independently and this
is emphasized by the way in which the epitheca and hypotheca accommodated the effects of primary
reduction. In both, rotation is anticlockwise relative to their respective pole, and therefore the
hypotheca rotates in the opposite direction to the epitheca relative to the polar axis of the theca. The
model plate pattern appears to represent an evolutionary base to which all Late Triassic and younger
peridinialeans of the selected type are related. This type is characterized by having up to seven plates
in each of its epithecal and hypothecal transverse series. In view of this relationship, the complex
model pattern could be representative of a pre-Late Triassic ancestral peridinialean.

EATON: DINOFL AGELL ATE PLATE PATTERNS

687

Acknowledgements. I wish to acknowledge the considerable advice given by Barrie Dale (Oslo) and his critical
reading of the manuscript. The ideas expressed in this paper originally formed the basis of a contribution to the
GSA Penrose Conference on ‘Modern and Fossil Dinoflagellates’, held in Colorado Springs, Colorado, U.S.A.,
April 1978. 1 wish to thank the many workers at that conference who provided me with helpful comments, and
particularly Professor W. R. Evitt (Stanford) who also critically read the manuscript. I am grateful to the
chairman and directors of British Petroleum Company Limited for permission to publish, to Miss Dorothy
Watson for typing the manuscript, and to my wife for her encouragement during the preparation of this paper.

REFERENCES

balech, e. 1974. El genero Protoperidinium Bergh, 1881 ( Peridinium Ehrenberg, 1831, partim). Rev. Mus. Cienc.
natur., Hidrobiol. 4, 1-79.

butschli, o. 1885. Unterabtheilung (Ordnung) Dinoflagellata. Pp. 906-1029. In bronn, h. g. Klassen und
Ordnung des Thier-Reichs, wissenschaftlichen dargestellt in Wort und Bild, Band 1 Protozoa, Abt. 2 Mastigo-
phora. Leipzig and Heidelberg.

cookson, i. c. and eisenack, a. 1962. Additional microplankton from Australian Cretaceous sediments. Micro-
paleontology, 8, 485-507, pis. 1-7.

deflandre, G. 1938. Microplancton des mers Jurassiques conserve dans les marnes de Villers-sur-Mer
(Calvados). Etude luminaire et considerations generates. Trav. Stnzool. Wimereux, 13, 147-200, pis. 5-11.
diesing, K. M. 1866. Revision der Prothelminthen, Abtheilung: Mastigophoren. Sber. Akad. Wiss. Wien, 52,
287-401.

eaton, g. l. 1976. Dinoflagellate cysts from the Bracklesham Beds (Eocene) of the Isle of Wight, southern
England. Bull. Br. Mus. nat. Hist. (Geol.), 26, 225-332, pis. 1-21.
evitt, w. R. 1961. Dapcodinium priscum n.gen., n.sp., a dinoflagellate from the Lower Lias of Denmark. J.
Paleont. 35, 996-1002, pi. 119.

— 1967. Dinoflagellate studies II. The archeopyle. Stanf. Univ. Pubis (Geol. Sci.), 10 (3), 1-88, pis. 1-11.

— lentin, J. K., millioud, M. E., stover, l. e., and williams, G. L. 1976. Dinoflagellate cyst terminology. Geol.
Surv. Pap. Can. 76-24, 1-9.

GOCHT, h. 1970. Dinoflagellaten-Zysten aus dem Bathonium des Erdolfeldes Aldorf (NW-Deutschland).
Palaeontographica, B 129, 125-165, pis. 26-35.

— and netzel, h. 1974. Rasterelektronenmikroskopische Untersuchungen am Panzer von Peridinium (Dino-
flagellata). Arch. Protistenk. 116, 381-410, pis. 43-52.

harland, r., morbey, s. J., and sarjeant, w. a. s. 1975. A revision of the Triassic to lowest Jurassic dinoflagellate
Rhaetogonyaulax. Palaeontology, 18, 847-864, pis. 100-104.

JOHNSON, c. D. and hills, l. v. 1973. Microplankton zones of the Savik Formation (Jurassic), Axel Heiberg and
Ellesmere Islands, District of Franklin. Bull. Can. petrol. Geol. 21, 178-218, pis. 1-3.
kofoid, c. A. 1906. Dinoflagellata of the San Diego region. I. On Heterodinium, a new genus of the Peridinidae.
Univ. Calif. Pubis Zool. 2, 341-368, pis. 17-19.

— 1907. The plates of Ceratium with a note on the unity of the genus. Zool. Anz. 32, 177-183.

— 1909. On Peridinium steini Jorgensen, with a note on the nomenclature of the skeleton of the Peridinidae.
Arch. Protistenk. 16, 25-47, pi. 2.

— 1911. Dinoflagellata of the San Diego region, IV. The genus Gonyaulax , with notes on its skeletal
morphology and a discussion of its generic and specific characters. Univ. Calif. Pubis Zool. 8, 187-286,
pis. 9-17.

loeblich, a. r., Jr. and loeblich, a. r. hi. 1968. Index to the genera, subgenera, and sections of the
Pyrrhophyta, II. J. Paleont. 42, 210-213.

morgenroth, p. 1970. Dinoflagellate cysts from the Lias Delta of Luhnde Germany. Neues. Jb. Geol. Palaont.
Abh. 136, 345-359, pis. 9-13.

schrank, F. von p. 1793. Mikroskopische Wahrnehmungen. Der Naturforscher, 27, 26-37, pi. 3.
stein, f. r. von. 1883. Der Organismus der Infusionsthiere nach eigenen Forschungen in systematischer
Reihenfolge bearbeitet. Abt. 3, Hf. 2. Die Naturgeschichte der arthrodelen Flagellaten. Leipzig. 30 pp., 25 pis.
stosch, h. a. von. 1969. Dinoflagellaten aus der Nordsee II. Helgolandinium subglobosum. gen. et spec. nov.
Helgolander wiss. Meeresunters, 19, 569-577.

stover, L. E. and evitt, w. R. 1978. Analyses of pre-Pleistocene organic-walled dinoflagellates. Stanf. Univ.
Pubis (Geol. Sci.), 15, 1-300.

688

PALAEONTOLOGY, VOLUME 23

wall, d. and dale, B. 1970. Living hystrichosphaerid dinoflagellate spores from Bermuda and Puerto Rico.
Micropaleontology , 16, 47-58, pi. 1.

— and evitt, w. r. 1975. A comparison of the modern genus Ceratium Schrank, 1793, with certain Cretaceous
marine dinoflagellates. Ibid. 21, 14-44, pis. 1-3.
wiggins, V. d. 1973. Upper Triassic dinoflagellates from arctic Alaska. Ibid. 19, 1-17, pis. 1-5.
williams, G. L. and downie, c. 1966. Further dinoflagellate cysts from the London Clay. Pp. 215-235. In
davey, r. j., downie, c., sarjeant, w. A. s., and williams, G. l. Studies on Mesozoic and Cainozoic dino-
flagellate cysts. Bull. Br. Mus. nat. Hist. (Geol.), Suppl. 3, 1-248, pis. 1-25.

Typescript received 12 July 1979
Revised typescript received 25 October 1979

GEOFFREY L. EATON
BP Petroleum Development Limited
Farburn Industrial Estate
Dyce, Aberdeen AB2 OPB

MODE OF LIFE OF A GIANT CAPULID
GASTROPOD FROM THE UPPER CRETACEOUS
OF SAGH ALIEN AND JAPAN

by itaru hayami and YASUMITSU kanie

Abstract. The life habits of a huge Campanian patelliform gastropod, hitherto called ‘ Helcion giganteus', from
Saghalien and Japan are discussed on the basis of several specimens adhering to enormous shells of Inoceramus
( Sphenoceramus ) schmidti. This gastropod is here transferred to the Capulidae of Mesogastropoda, and a new
generic name, Gigantocapulus, is proposed for it. Its ecological relation with 7. (S.) schmidti is regarded as
parasitic by analogy to some living species of Capulus that attach to the valves of pectinids. This interpretation is
supported by stratigraphic and geographic distribution patterns and by its functional morphology.

‘ Helcion giganteus', originally described by Schmidt (1873) from the Upper Cretaceous at Cape Dui
near Alexandrovsk, north Saghalien, is probably the largest patelliform gastropod known. Its shell
sometimes exceeds 400 mm in maximum length, and shows a wide range of morphological variation.
This species, though restricted to the lower to middle Campanian (Zone of Inoceramus schmidti),
occurs at various localities in Saghalien, Japan (mainly Hokkaido), Koryak Highland of eastern
Siberia (Dundo and Efremova 1974), Southern Alaska (Jones, pers. comm.), and British Columbia
(Whiteaves 1903). The association of this species with I. ( Sphenoceramus ) schmidti Michael, 1899, is
important. Almost all the specimens of ‘7. digitatus' described by Schmidt (1873) together with
‘77. giganteus ' from Cape Dui seem to be referable to 7. (S'.) schmidti, as revised in Michael (1899) and
Nagao and Matumoto (1940). Their coexistence in the same fossil bed (commonly fine-grained
sandstone) was also recorded at many other localities: Naibuchi (=Naibuti) (Matumoto 1942,
p. 167) in south Saghalien, Abeshinai (Matumoto 1942, p. 205), Hetonai (Matumoto 1942, p. 251),
Urakawa (Matumoto 1942, p. 268; Kanie 1966, p. 322; 1977, p. 54) and some other places in
Hokkaido, and Dogo-Himezuka, Matsuyama City (Kashima 1972; Matsumoto 1973) in Shikoku.

Summarizing the classification and evolutionary history of Cretaceous patelliform gastropods in
the northern Pacific region, Kanie (1975) concluded that ‘77. giganteus ' belongs to the Meso-
gastropoda and that they possibly attached to some other shelled organism. Since ‘77. giganteus' is
seldom accompanied by molluscs assumed to have lived on near-shore rocky substrates, it was
assumed that it may have been attached to large bivalves such as 7. ( S .) schmidti, but at that time there
was no direct evidence. Subsequently Hayami found a specimen of ‘77. giganteus', in growth position
attached to the shell surface of Inoceramus, in the collection of the University Museum, University of
Tokyo. We have now examined the relation between the two molluscs on the basis of many specimens
stored at various institutions. In the present article we describe some of these specimens, discuss the
interpreted life habit of this gastropod, and compare it with some living species of similar habit. The
taxonomic position of ‘77. giganteus' is also reconsidered.

SYSTEMATIC PALAEONTOLOGY

Order caenogastropoda Cox, 1959
Suborder mesogastropoda Thiele, 1925
Superfamily calyptraeacea Lamarck, 1809
Family capulidae Fleming, 1822
Genus Gigantocapulus Hayami and Kanie, gen. nov.

IPalaeontology, Vol. 23, Part 3, 1980, pp. 689-698, pi. 87.1

690

PALAEONTOLOGY, VOLUME 23

Type species. Helcion giganteus Schmidt, 1873, northern Pacific region, Campanian.

Diagnosis. Shell very large, cap-shaped or conical, bilaterally symmetrical but more or less irregular in outline;
apex located anteriorly from the centre, sometimes marginal; surface commonly ornamented with irregularly
disposed radial costae in addition to concentric rings on the apical region; anterior elevated sector and internal
septum absent; outermost layer prismatic, while other and inner layers are crossed-lamellar; some species living
upon the shells of Inoceramus.

Remarks. The taxonomic position of ‘ H . giganteus ’ and its allied species from the Cretaceous of
northern Pacific has been debatable; Capulus, Patella, Scurria, Acmaea, and Brunonia also have been
used as their generic names. Living patelliform gastropods occur in various unrelated taxonomic
groups, e.g. the Patellacea of Archaeogastropoda, the Neritacea and Calyptraeacea of Meso-
gastropoda and the Siphonariacea of Pulmonata. Because their shell forms sometimes show
remarkable convergence, such essential characters as muscle impression, presence or absence of
internal septum and shell structure as well as inferable life habit may be important for determination
of the taxonomic position of fossil species.

Kanie (1975) assigned these Cretaceous species in question to the genus Anisomyon Meek and
Hayden, 1860, which had been included in the Basommatophora (an order of Pulmonata), and
proposed a new family Anisomyonidae in the Mesogastropoda. This treatment was primarily based
on the resemblance of muscle impressions and shell form of some species to the Capulidae and the
difference of shell structure from the Siphonariidae. As noted elsewhere (Hayami and Kase 1977,
p. 55), however, one of us (I. H.) doubted if the type species of Anisomyon [H. patelliformis Meek and
Hayden, 1856] should be transferred from the Basommatophora to the Mesogastropoda, and
presumed that ‘ H . giganteus ’ may represent an unnamed genus of the Capulidae. This is proposed
here, which modifies the previous classification (Kanie 1975) of Cretaceous patelliform gastropods
from the northern Pacific region.

Kanie (1975) distinguished two ‘morphotypes’ in lH. giganteus': type A is characterized by the
relatively small size, small apical angle, and irregularly noded ornament, while type B has relatively
large size, large apical angle, and almost persistent and not noded radial ribs. Of the originally figured
specimens of H. giganteus, most individuals including the lectotype (Schmidt 1873, pi. 2, fig. 17,
designated by Kanie (1975) as ‘holotype’) belong to type B, and only two small specimens (Schmidt
1873, pi. 3, figs. 8, 9) may belong to type A. Numerous individuals of type A are preserved in various
Japanese institutions, but none of them actually shows any intimate relation to the shell of
Inoceramus. All the observed specimens attached to the surface of I. (S.) schmidti belong to type B.
Moreover, significant morphological differences are newly recognized between the two ‘morpho-
types’. First, a trace of an internal septum is often seen in type A (see Kanie 1975, p. 9, fig. 2), but has
never been observed in type B. Secondly, the apex is always located subcentrally or even posteriorly in
type A, while it is commonly located very anteriorly or even near the anterior margin in type B.
Thirdly, a tongue-like projection, as described later, occurs only in type B. Host-determined non-
genetic variation is actually known in a living capulid species (Thorson 1965), and dwarf males are
also seen in such semi-parasitic gastropods. Yet, such great differences of essential characters are
hardly explicable by individual variation. At present, we consider that the two ‘morphotypes’ belong
to different species, and that the use of the specific name Gigantocapulus giganteus should be restricted
to the type B of Kanie (1975). The specimens of type A seem to be close to lA. transformis' Dundo and
Efremova (1974) from the Koryak Highland. The presence of an internal septum may suggest that
they belong to the Calyptraeidae.

EXPLANATION OF PLATE 87

Figs. 1, 2. Gigantocapulus giganteus (Schmidt, 1873). UMUT MM5535 attached to the surface of Inoceramus
(, Sphenoceramus ) schmidti Michael, 1 899. Loc. N469, north-west of Miho (gorge of Ryugase), Naibuchi area,
south Saghalien. Collected by T. Matsumoto. 1, upper view, x 0-42; 2, left lateral view, x 0-42. (See also
text-fig. 1.)

PLATE 87

hayami and kanie, Cretaceous patelliform gastropod

692

PALAEONTOLOGY, VOLUME 23

In the Western Interior of the United States some specimens of Anisomyon have also been found
adhering to the shells of Inoceramus (Sohl 1967a). The association may be comparable with the
present case. According to Sohl’s (19676) redescription of A. patelliformis (Meek and Hayden, 1856),
however, one of the paratypes reveals clearly asymmetric muscle impression, which resembles that of
Siphonaria, although the posterior carination of Siphonaria- type is undeveloped in that species. No
specimen of G. giganteus shows clear muscle impression, but Capulus- like horseshoe-shaped muscle
scars are recognized in C. cassidarius Yokoyama, 1890, which is considered to be ancestral to
G. giganteus (Kanie 1975, p. 9, fig. 2). The genus Anisomyon is represented by much smaller species
without radial costae, and we are now inclined to consider that it is not directly related to
Gigantocapulus.

The genus Brunonia Muller, 1898, may be another Late Cretaceous patelliform gastropod
comparable with our new genus from morphological and paleoecological standpoints. The genus was
generally referred to the Siphonariidae, but in the Treatise (Knight et al. 1960) it was doubtfully
included in the suborder Patellina. The concentrically ornamented shell of its type species [B. grandis
Muller, 1898, from the Santonian of Germany] resembles the apical part of G. giganteus.
Unfortunately, Muller’s original specimen of B. grandis is said to have been lost, and further
comparative study is now difficult. At present we think that Gigantocapulus is at least generically
separable from Brunonia by the developed radial costae on the surface. Judging from the original
figures of B. grandis , the apex is more constantly located near the centre of shell, and no projection is
developed on its anterior periphery.

DESCRIPTION OF SELECTED SPECIMENS

Among a large number of specimens of G. giganteus in the collection of the University Museum, University of
Tokyo (UMUT), and the Institute of Geology and Palaeontology, Tohoku University, Sendai (IGPS), several
show an intimate association with the shells of I. ( S .) schmidti. UMUT MM5535 (PI. 87; text-fig. 1) has well-
preserved shells of the two species, and shows the position and orientation of attachment. It was found in
T. Matsumoto’s collection from a greenish fine-grained sandstone of the Ray 1 Member of the Ryugase Group
at loc. N469 (gorge of Ryugase), about 4-5 km north-west of Miho, Naibuchi area, south Saghalien (see the
locality map in Matsumoto 1942). The following description is entirely based on this specimen.

The shell of G. giganteus, though a considerable part of the marginal area is broken off, exceeds 290 mm in
maximum length and 250 mm in breadth, showing a suboval, nearly bilaterally symmetrical, cap-shaped outline
with a somewhat irregularly undulating marginal area. The apex is located at about one-fifth of maximum length

LV MRV

text-fig. 1 . Sketch of Specimen I (UMUT MM5535) of Gigantocapulus giganteus (Schmidt) from the left side.
LV: Fracture of the left valve (prismatic outer layer) of Inoceramus ( Sphenoceramus ) schmidti Michael; MRV:
Internal surface of the right valve of the same individual, on which characteristic divergent ribs are impressed.

HAYAMI AND KANIE: GIANT GASTROPOD

693

from the anterior end, but the growth-lines indicate that it was situated near the centre of shell in the early growth
stage. The pre-apical area steeply descends towards the anterior margin, while the post-apical area is widely
expanded and broadly convex. The maximum inflation of shell lies far behind the apex. The surface is
ornamented with several subconcentric ribs and about fifty radial costae. Subconcentric ribs are distinct and
widely spaced, and their distribution is confined to the apical area (within 50 mm from the apex). On the
contrary, radial costae are at first indistinct but become prominent after the effacement of subconcentric ribs.
They are commonly irregularly dichotomous but sometimes convergent. The thickness of test does not exceed
10 mm. The outermost layer is thin and prismatic, and the outer and inner layers are crossed-lamellar and solid.

The associated shell of Inoceramus is evidently a part of an articulated individual. The shell of G. giganteus
adheres closely to the surface of its left valve. The shell margin of G. giganteus , though its right side is incomplete,
fits perfectly the undulating surface of Inoceramus without any perceptible gap (text-fig. 1). The opposite valve of
Inoceramus has been almost entirely exfoliated and lost, but the divergent ribs impressed on the internal surface
are unmistakably characteristic of I. ( S .) schmidti. The radial ribs of G. giganteus are evidently denser than the
divergent ribs of 7. ( S .) schmidti; the former does not necessarily correspond with the latter. Judging from the
orientation of the divergent ribs as well as the nearly closed valves of I. ( S .) schmidti below the anterior margin of
G. giganteus, this gastropod appears to have sat on the antero-ventral area of the living shell of 7. (S.) schmidti
with the apex located on the antero-ventral side. The axis of symmetry of G. giganteus forms an angle of about
30° with the line of maximum length of 7. ( S .) schmidti. The prismatic layer of Inoceramus, which represents the
outer layer of the shell, is about 7-0 mm and 3-0 mm thick below the anterior and posterior margins of
G. giganteus, respectively. The original size of this inoceramid shell would exceed 500 mm, provided that the
thickness of this layer increases isometrically to the attained shell length. If the allometric growth indices
calculated by Tanabe (1973, p. 177) on some specimens of 7. ( S .) schmidti from Hokkaido are applied, the
restored shell of this individual may exceed 700 mm in maximum length.

UMUT MM57 1 1 ( = Cr. 1217) (text-fig. 2) is interesting because its right-anterior margin is nearly complete. It
belongs to an old collection from the Cape Khoi Beds at Cape Jonquiere near Alexandrovsk, north Saghalien.

text-fig. 2. Gigantocapulus giganteus (Schmidt). Specimen II (UMUT MM5711) attached to a
crushed shell of Inoceramus ( Sphenoceramus ) schmidti Michael. Loc. Cape Jonquiere near
Alexandrovsk, north Saghalien. a, upper view, x 0-36; b, bird’s-eye anterior view of the anterior
part of the specimen showing a tongue-like projection and nearly complete right-anterior margin
of shell, x 0-55; c, anterior view of the same specimen, xO-55.

694

PALAEONTOLOGY, VOLUME 23

This individual is also closely associated with an enormous articulated shell of I. (S.) schmidti, which, however, is
so strongly crushed that the original state of attachment is difficult to restore.

This specimen is about 250 mm long and 190 mm wide, and the shell of post-apical part has been considerably
exfoliated and lost. The matrix was successfully removed from the pre-apical part of shell, and both the external
and internal characters are well exhibited. The marginal area of the pre-apical part is remarkably depressed and
gently folded like a brim (text-fig. 2c). Furthermore, there is a curious tongue-like projection at the anterior
extremity, which is unusually thickened with a rounded edge. The internal surface is nearly smooth, and neither a
septum nor a muscle scar is observed below the apical area. Radial costae are not impressed on the internal
surface even near the margin.

The following specimens of G. giganteus are also intimately associated with some crushed shells of /. (S'.)
schmidti: UMUT MM5710 (=Cr.l418) and UMUT MM5709 (=Cr.998): old collection from the Zone of
I. schmidti in Naibuchi area, south Saghalien (exact locality unknown). UMUT MM5713 (=Cr.l228): old
collection from the same locality as UMUT MM5711. UMUT MM5712 (= Cr.1218): old collection from the
Zone of I. schmidti in Alexandrovsk area, north Saghalien (exact locality unknown). These specimens show a
wide range of morphological variation. One of the illustrated paralectotypes of Helcion giganteus from the type
locality (Schmidt 1873, pi. 3, fig. 2) may be another example of an attached specimen, because a fragmentary
prismatic shell was indicated below it.

VARIABILITY OF SOME MORPHOLOGICAL CHARACTERS

When H. giganteus was originally described by Schmidt (1873), four varieties were distinguished by
the different position of the apex. All the original specimens are included either in var. a depressa, var.
j8 nasuta, var. y retracta or var. 8 centralis. Kanie (1975, p. 23) designated a specimen of var. depressa
as the ‘holotype’ of H. giganteus, but (Hayami and Kase 1977, p. 56) this procedure can be regarded
as constituting valid lectotype designation. Dundo and Efremova (1974) regarded some of these
varieties as distinct species, and referred centralis and nasutus to Patella and Helcion, respectively.
However (Kanie 1975; Hayami and Kase 1977), none of these varieties (except for two small
specimens of var. depressa (Schmidt 1873, pi. 3, figs. 8, 9) seems to constitute a distinct taxon, because
the difference of apical position as well as other characteristics appears to be gradational within a
single fossil population. The growth-lines of the present specimens show that the variability of apical
position is partly due to ontogenetic transformation: the apex evidently shifts from the central part to
the anterior portion of shell with growth. There is also a change of the direction of apex in the young
stage. As shown in UMUT MM 5709 and some other small specimens, the very apex, if preserved,
seems to point in the direction opposite to the expansion of shell. Although the apex is generally
located posteriorly in many living species of Capulus and related genera, this ontogenetic change is
one of the main reasons why we suspect here, unlike a previous interpretation (Kanie 1975), that the
shorter end is actually anterior.

The shell form and surface ornamentation are also quite variable. Among the forty specimens we
have observed at various institutions in Japan, the angle of ultimate apex in lateral view varies from
120° to 145°. The number of radial costae ranges from thirty-five to sixty-five. The thickened tongue-
like projection at the anterior end of shell is also observable in some other specimens, e.g. one of the
paralectotypes (Schmidt 1873, pi. 3, fig. 10) and IGPS no. 50910 (Kanie 1975, pi. 15, fig. 1 a, b\ Kanie
1977, pi. 2, fig. 4). It may be a widespread character in this species, but, as shown by the growth-lines
on UMUT MM5535, 5711, its development is seen only in the later ontogenetic stage.

The range of morphological variation of G. giganteus is thus unusually wide. Such a great
variability is unknown in any living species of the Patellacea, but comparable with that of some
species of the Capulidae. The variable shell form and ornamentation of this species were probably
influenced by the nature of the surface of the host.

INTERPRETATION OF MODE OF LIFE

From our observation on in situ specimens, it is likely that at least some individuals of G. giganteus
grew on living shells of Inoceramus ( Sphenoceramus ) schmidti. This is supported by the fact that the
associated inoceramid shells are, even if crushed, commonly articulated. Moreover, the stratigraphic

HAYAMI AND KANIE: GIANT GASTROPOD

695

and geographic distribution of the two species is identical, which suggests not only their intimate
ecological relation but also that the evolutionary history of the former depended on the latter.

A large number of malacologists and marine ecologists have paid attention to the parasitic or semi-
parasitic life of Capulus species and their hosts (Orton 1912, 1949; Yonge 1938; Otuka 1939;
Teramachi 1942; Kuroda 1951; Sharman 1956; Burch and Burch 1961; Orr 1962; Thorson 1965;
Kosuge and Hayashi 1967; Habe 1967). The hosts are commonly epifaunal bivalves, especially large
species of the Pectinacea, though in a few cases the epibionts rest also on the surface of certain
gastropods, brachiopods, and annelids. Sometimes an almost exclusive relation exists between the
epibiont and host species (e.g. C.tosaensis on Propeamussium sibogae in Japan), but in other cases an
epibiont species can grow on various hosts (e.g. C. dilatatus on Amusium japonicum, Pecten albicans,
Decatopecten striatus, Chlamys nobilis, etc. in Japan; C. ungaricus on P. maximus, Aequipecten
opercularis. Modiolus modiolus, Monia patelliformis, Turritella communis, etc. in Great Britain and
North Sea). According to Yonge (1938) and others, C. ungaricus is a ciliary feeder. It intercepts the
food, which has been collected on the gills of a bivalve, by inserting its long proboscis inside the
bivalve shells. The ecological relation was regarded as semi-parasitic by Sharman (1956) and as
commensalistic by Thorson (1965). Although the epibiont does not seem to cause the bivalves any
mortal harm, this state is most certainly disadvantageous to the host. We consider that this is a case of
external parasitism, but the term ‘semi-parasitic’ may be more appropriate for this species, because it
also attaches to dead shells and rocks.

The life habits of such parasitic individuals of Capulus can be classified into two types. One is
represented by C. dilatatus, in which (Kosuge and Hayashi 1967) the epibiont bores a small hole
through the pectinid shell (commonly up-facing valve) in order to insert its proboscis. The boring
position is concentrated on the anterior half of the disc (corresponding to the position of gills) and
sometimes on the anterior wing. The orientation of attachment seems to be almost random. The other
type is exemplified by C. ungaricus, which rests preferentially on the anterior and ventral marginal
part of down-facing valves of living pectinids. Sharman (1956) examined the attaching position and
orientation of many individuals of this species on the shells of A. opercularis from off the coast of the
Isle of Man, noting: ‘in its characteristic position the gastropod sits at the edge of the valve with the
front margin of the shell projecting a little over it and the apex pointing inwards.’ C. ungaricus never
makes a borehole, but the edge of the valve margin is said to be frequently chipped so that this
gastropod can easily insert its proboscis into the pectinid valves. Somewhat similar feeding habits are
known in C. tosaensis from the Japanese deep waters, although this species is said to attach
preferentially to the left (? up-facing) valve of Propeamussium sibogae.

On the shell surface of the many specimens of I. ( S .) schmidti neither a borehole nor a scar of
attachment has been recognized, and it may be difficult to know whether the valve margin was
actually chipped or not by other organisms. However, the attaching position and orientation in the
specimens on text-fig. 1 seem to indicate that the life habit of G. giganteus was analogous to the second
type, especially to the case of C. ungaricus as illustrated by Sharman ( 1 956, figs. 1-3). We interpret the
function of the curious tongue-like anterior projection as protecting the head of the gastropod which
presumably protruded a little beyond the edge of the valve margin, because otherwise the remarkable,
declined margin of this projection could not adhere closely to the surface of inoceramid shell. Such a
hanging front margin of the shell is also commonly seen in C. ungaricus. Text-fig. 3 shows a putative
living position of G. giganteus on the left valve of I. ( S .) schmidti, although it is still unknown whether
the valve is actually up-facing or down-facing.

SUMMARY

The observation of in situ specimens and the functional interpretation of the shell shows that
G. giganteus was a parasitic gastropod to I. ( S .) schmidti. Considering the much smaller size of other
associated molluscs, only this inoceramid seems to have offered the solid ground of attachment for
such a large patelliform gastropod. Although complete specimens of I. ( S .) schmidti can seldom be

696

PALAEONTOLOGY, VOLUME 23

text-fig. 3. Reconstruction of the living position of Gigantocapulus
giganteus (Schmidt) on Inoceramus ( Sphenoceramus ) schmidti Michael.

Their periostracum is not drawn, because nothing is known about its
development. This is not a sketch but chiefly based on Specimens I and II.

obtained, we have actually observed several extraordinarily large specimens of this species (exceeding
700 mm in maximum length) in the collections from Saghalien and Hokkaido. The gigantism of this
gastropod is evidently related to the unusually large size of the host. If such a parasitic relation was
developed, it can be readily imagined that an ecologically specialized epibiont was compelled to
become extinct by the decline of the host species. G. giganteus seems to have shared its evolutionary
lot with I. ( S .) schmidti, because their stratigraphic and geographic distribution is identical.

The history of this external parasitism possibly goes back to earlier times. As interpreted previously
(Kanie 1975), G. giganteus may have been derived from C. cassidarius Yokoyama through some
intermediate form. C. cassidarius is common in the Turonian to Santonian strata of the same region
and is frequently accompanied by I. ( S .) naumanni Yokoyama, which seems to be ancestral to I. (S'.)

HAYAMI AND KANIE: GIANT GASTROPOD

697

schmidti, and some other small-sized species of Inoceramus. Therefore, it is possible that the
parasitism was already established between the ancestors, although in situ preservation has not been
found. The inferred mode of life also explains shell orientation, the wide range of morphological
variation, and the curious tongue-like anterior projection in G. giganteus. Its taxonomic reference to
the Capulidae of Mesogastropoda is also consistent with the parasitic mode of life. There is still a
shortage of in situ material showing life orientations, which will provide more evidence of the
paleoecological relation between this peculiar limpet and other organisms.

Acknowledgements. We are grateful to Dr. Norman F. Sohl (U.S. Geological Survey) for his helpful suggestions,
reading the manuscript, and loan of Anisomyon; to Professor Emeritus Tatsuro Matsumoto (Kyushu
University) for the Saghalien specimens. Professor Tamio Kotaka and Dr. Kenshiro Ogasawara (Tohoku
University) made available material at their institute. We also thank Professor Tetsuro Hanai, Professor
Masuoki Horikoshi, and Mr. Paul Frydl (University of Tokyo), and Dr. Masayuki Tashiro (Kochi University)
for drawing text-figure 3.

REFERENCES

burch, J. Q. and burch, r. l. 1961. A new Capulus from Gulf of California. Nautilus , 75, 19-20.
dundo, o. P. and efremova, v. i. 1974. [Field Atlas of the Senonian Index Fauna in the Northeastern Part of
Koryak Highland .] Sci. Res. Inst. Arct. Geol. (NIIGA), Leningrad. 28 pp., 12 pis. [In Russian.]
habe, T. 1967. A new capulid snail, Capulus spondylicola, from Japan. Venus, 26, (2), 37-38, pi. 4.
hayami, i. and kase, t. 1977. A systematic survey of the Paleozoic and Mesozoic Gastropoda and Paleozoic
Bivalvia from Japan. Univ. Mus. Univ. Tokyo, Bull. 13, 1-155.
kanie, y. 1966. The Cretaceous deposits in the Urakawa district, Hokkaido. J. Geol. Soc. Japan, 72 (7), 315-328.
[In Japanese with English abstract.]

— 1975. Some Cretaceous patelliform gastropods from the northern Pacific region. Sci. Rept. Yokosuka City
Mus. 21, 1-44, pis. 1-20.

— 1977. Succession of the Cretaceous patelliform gastropods in the northern Pacific region. In matsumoto, t.
[org.]: Mid-Cretaceous Events— Hokkaido Symposium, 1976. Palaeontol. Soc. Japan, Spec. Papers, 21,
53-62, pi. 2.

kashima, n. 1972. [Fossils from the Dogo-Himezuka, Matsuyama City, No. 2.] Ehime no Shizen, 14 (8), 12-15.
[In Japanese.]

knight, J. B. et al. 1960. Treatise on Invertebrate Paleontology. Part I. Mollusca 1. Geol. Soc. America and Univ.
Kansas Press. 351 pp.

kosuge, s. and hayashi, s. 1967. Notes on the feeding habits of Capulus dilatatus A. Adams, 1 860 (Gastropoda).

Sci. Rept. Yokosuka City Mus. 13, 45-54, pis. 1, 2. [In Japanese with English abstract.]
kuroda, T. 1951. [ Capulus tosaensis living on Propeamussium sibogae .] Yumehamaguri, 6, 19. [In Japanese.]
matsumoto, T. 1973. Further notes on the Dogo-Himezuka fauna. J. Geol. Soc. Japan, 79 (7), 496. [In Japanese.]
matumoto [matsumoto], t. 1942. Fundamentals in the Cretaceous stratigraphy of Japan. Mem. Fac. Sci.
Kyushu Univ. [D] 1 (3), 129-280, pis. 5-20.

meek, f. b. and hayden, f. v. 1857. Description of a new species of Gastropoda from the Cretaceous formations
of Nebraska Territory. Proc. Philad. Acad. Nat. Sci., for 1856, 8, 63-69.

— 1860. Systematic catalogue with synonymy of Jurassic and Cretaceous, and Tertiary fossils collected
in Nebraska. Ibid. 12, 412-432.

Michael, r. 1899. Ueber Kreidefossilien von der Insel Sakhalin. Jahrb. k. preuss. geol. Landesanst. 18, 153-164,
pis. 5, 6.

Muller, G. 1898. Die molluskenfauna des Untersenon von Braunschweig und Ilsede. I. Lamellibranchiaten und
Glossphoren. Abh. preuss. geol. Landes, n.f., 25, 1-142, pis. 1-18.
nago, t. and matumoto [matsumoto], t. 1940. A monograph of the Cretaceous Inoceramus of Japan. Part 2.

J. Fac. Sci. Hokkaido Imp. Univ. [4] 6, 1-64, pis. 1-22.
orr, v. 1962. The drilling habit of Capulus danielli. Veliger, 5, 63-67.

orton, j. h. 1912. The mode of feeding of Crepidula, with an account of the current-producing mechanism in the
mantle cavity, and some remarks on the mode of feeding in gastropods and lamellibranchs. J. Mar. Biol. Ass.
U.K. 9, 444-478.

— 1949. Notes on the feeding habit of Capulus ungaricus. Rept. Mar. Biol. Sta. Pt. Erin, 61, 29-30.

698

PALAEONTOLOGY, VOLUME 23

otuka, Y. 1939. Non-sculptured species of the genus Capulus. Venus , 9, 89-98, pi. 4.

schmidt, M. F. 1873. Ueber die Petrefakten der Kreideformation von der Insel Sachalin. Mem. Acad. Imp. Sci.
St.-Petersb., ser. 7, 19 (3), 1-37, pis. 1-8.

sharman, M. 1956. Note on Capulus ungaricus (L.). J. Mar. Biol. Ass. U.K. 35, 445-450.
sohl, n. f. 1967 a. Upper Cretaceous gastropod assemblages of the Western Interior of the United States, in
Paleoenvironments of the Cretaceous Seaway — a symposium, Colorado School of Mines, Golden, 1-37.

— 1967 b. Upper Cretaceous gastropods from the Pierre Shale at Red Bird, Wyoming. Prof. Pap. U.S. geol.
Surv. 393-B, 1-46, pis. 1-11.

tanabe, K. 1973. Evolution and mode of life of Inoceramus ( Sphenoceramus ) naumanni Yokoyama emend., an
Upper Cretaceous bivalve. Trans. Proc. Palaeontol. Soc. Japan, [n.s.] 92, 163-184, pis. 27, 28.
teramachi, a. 1942. [Ecology of Capulus tosaensis .] Venus, 12 (1-2), 106. [In Japanese.]
thorson, G. 1965. A neotenous dwarf-form of Capulus ungaricus (L.) (Gastropoda, Prosobranchia) com-
mensalistic on Turritella communis Risso. Ophelia, 2, 175-210.
whiteaves, j. f. 1903. On some additional fossils from the Vancouver Cretaceous, with a revised list of the
species therefrom. Geol. Surv. Canada, Mesozoic Fossils, 1 (5), 309-416, pis. 40-51.
yokoyama, M. 1890. Versteinerungen aus der japanischen Kreide. Palaeontographica, 36, 159-202, pis. 18-25.
yonge, c. M. 1938. Evolution of ciliary feeding in the Prosobranchia, with an account of feeding in Capulus
ungaricus. J. Mar. Biol. Ass. U.K. 22, 453-468.

ITARU HAYAMI

University Museum
University of Tokyo
Hongo 7-3-1, Bunkyo-ku
Tokyo 113, Japan

YASUMITSU KANIE

Typescript received 19 July 1979

Revised typescript received 20 November 1979

Yokosuka City Museum
Fukadadai 95
Yokosuka 238, Japan

TWO NEW JURASSIC BRYOZOA FROM
SOUTHERN ENGLAND

by PAUL D. TAYLOR

Abstract. Reptomultisparsa tumida sp. nov. and Reptoclausa porcata sp. nov. are described respectively from
the Bathonian Bradford Clay of Bradford-on- Avon and the Aalenian/Bajocian Inferior Oolite of the Cotswold
Hills. The genus Reptoclausa was previously known only from the Cretaceous. Reptoclausa colonies have
autozooecia located on longitudinal ridges which are separated by furrows composed of kenozooecia. This
unusual arrangement of zooecia can be explained by the action of physiological growth gradients during colony
life.

During a revision of some Jurassic Bryozoa from England and Normandy (Taylor 1977) two new
species were recognized and are described for the first time in this paper. Both species are encrusting
forms belonging to the Order Cyclostomata, the most abundant group of bryozoans during a period
of relatively low species diversity. A tentative estimate of known Jurassic bryozoan diversity based
on available data suggests the existence of about ninety species belonging to approximately thirty
genera, although these figures are undoubtedly an underestimate of total worldwide diversity for
three main reasons. First, almost all known Jurassic bryozoans have been described from either
England, France, or Germany; they are extremely poorly known from other parts of the world (see
Taylor 1977, pp. 336-338). Secondly, high levels of phenotypic variation within species of simple
morphology has hindered taxonomic discrimination and quite probably has led to undersplitting.
Genetic studies (e.g. Thorpe et al. 1978) of living bryozoans are beginning to reveal the presence of
‘cryptic’ species that are difficult to distinguish morphologically. Thirdly, non-calcified ctenostome
bryozoans with a low fossilization potential were perhaps much more common during the Jurassic
than is immediately obvious (Voigt 1977; Pohowsky 1978; Taylor 1978). Despite their low apparent
diversity Jurassic bryozoans are ubiquitous in marine sediments, which accumulated in aerobic
environments containing firm substrates (e.g. brachiopod shells) suitable for colony attachment.
Some species developed erect growth from an attached base but the commonest species were totally
encrusting, as are the two new species described here.

Type and figured specimens are housed in the British Museum (Natural History) (BMNH).

FAMILIAL CLASSIFICATION OF JURASSIC TUBULOPORINID BRYOZOA

The order Cyclostomata was represented in the Jurassic by two suborders, the Tubuloporina and the
Cerioporina. Six major Jurassic families of the Tubuloporina may be distinguished: Stomatoporidae,
Oncousoeciidae, Macroeciidae ( = Multisparidae), Plagioeciidae, Theonoidae, and Frondiporidae.
The stomatoporids are typically encrusting uniserial or narrow multiserial (‘ribbon-shaped’) genera,
apparently lacking the larval brooding polymorphs known as gonozooids which characterize the
other tubuloporinid families. Oncousoecids have branching adnate colonies and gonozooids with
minute ooeciopores. Macroecid and plagioecid genera developed a variety of convergent colony
forms, both erect (e.g. ‘ Entalophora' , ‘ PustuloporcC) and encrusting (e.g. ‘ Berenicea'), but the two
families may be distinguished from one another by the structure of their gonozooids. Macroecid
gonozooids are longitudinally elongate and possess comparatively large ooeciopores (the orifice
through which the larvae were released). Plagioecid gonozooids are broad, bulbous and possess

[Palaeontology, Vol. 23, Part 3, 1980, pp. 699-706, pi. 88.|

700

PALAEONTOLOGY, VOLUME 23

ooeciopores which are considerably smaller than the apertures of autozooids in the same colony. In
both the theonoids and the frondiporids, groups of contiguous autozooidal apertures form fascicles
elevated above the general level of the colony surface. Frondiporid zooids are typically longer than
those of theonoids and, whereas zooidal budding occurred within the lengthening frondiporid
fascicles, within-fascicle zooidal budding is not known in the theonoids.

SYSTEMATIC PALAEONTOLOGY

Phylum bryozoa Ehrenberg, 1831
Class stenolaemata Borg, 1926
Order cyclostomata Busk, 1852
Suborder tubuloporina Milne-Edwards, 1838
Family macroeciidae Canu, 1918
Genus reptomultisparsa d’Orbigny, 1853

(see Walter 1969, p. 75, for a revised generic diagnosis)

Reptomultisparsa tumida sp. nov.

Plate 88, fig. 1; text- fig. 1

Derivation of name. The trivial name tumida refers to the broad, swollen appearance of the gonozooecia.
Holotype. BMNH D 13346 Bathonian, Bradford Clay ( discus Zone), Bradford-on-Avon, Wiltshire.

Paratypes. BMNH D52651a-c, Bathonian, Bradford Clay, locality unknown. Other specimens in the authors
collection are from the Bathonian White Limestone Formation of Foss Cross, Gloucestershire.

Diagnosis. Reptomultisparsa with delicate unilamellar zoaria; autozooecia with maximum width
midway along their frontal walls and possessing small apertures; gonozooecia broad and inflated.

Description. Zoaria (PI. 88, fig. 1) are unilamellar, fan-shaped, or discoidal (bereniciform). Zooecia arise at
divisions of existing interzooecial walls on a basal lamina.

Autozooecia have moderately long frontal walls characteristically attaining maximum width midway along
their length. Interzooecial walls form conspicuous traces on the relatively fiat zoarial surface. The small, circular
autozooecial apertures are widely spaced and have raised rims but lack distinct peristomes.

Kenozooecia may occur around gonozooecial borders. Their proximal portions are identical to those of
autozooecia but the kenozooecia are truncated distally by gonozooecial dilation and consequently lack an
aperture.

Gonozooecia (text-fig. 1) have narrow proximal portions and well-defined inflated distal portions with a
circular to oval shape. The sub-terminal ooeciopores lack ooeciostomes and are transversely elongate and
slightly smaller than autozooecial apertures.

EXPLANATION OF PLATE 88

Fig. 1. Reptomultisparsa tumida sp. nov. BMNH D13346, Holotype colony, Bathonian, Bradford Clay,
Bradford-on-Avon, x 7.

Figs. 2-6. Reptoclausa porcata sp. nov. 2, BMNH B4855, immature colony prior to ridge development,
Bajocian, Lower Ragstone, Cold Comfort, x 7. 3-5, BMNH D8724, Holotype, Aalenian, Pea Grit, Birdlip.
3, intracolony overgrowth by a new layer of zooecia, x 1 1. 4, furrow occupied by kenozooecia (centre right),
x 13. 5, autozooecia with rounded distal terminations, x30. 6, BMNH D10091, zoarium showing

conspicuous furrows and ridges with a ridge dichotomy, Aalenian, Pea Grit, Crickley Hill, x 8.

Figs. 1 and 2 are of ammonium chloride coated specimens.

PLATE 88

taylor, Jurassic Bryozoa

702

PALAEONTOLOGY, VOLUME 23

text-fig. 1. Reptomultisparsa tumida sp. nov.
BMNH D13346 (holotype), Upper Bathonian,
Bradford Clay, Bradford-on-Avon, Wiltshire.
A group of autozooecia and a bulbous gono-
zooecium. x 38.

Dimensions. See Table 1.

Remarks. The relatively broad and inflated gonozooecia of Reptomultisparsa tumida distinguish it
from other species in the genus, and the subterminal position of the ooeciostome contrasts with the
terminal ooeciostomes developed in plagioecids with similar bereniciform colonies.

Stratigraphical range. Upper Bathonian.

table 1. Dimensions (in mm) of Reptomultisparsa tumida zooecia.
Abbreviations of morphological characters: law, longitudinal auto-
zooecial aperture width; taw, transverse autozooecial aperture width; fwl,
autozooecial frontal wall length; fww, autozooecial frontal wall width
(maximum); tgl, total gonozooecial frontal wall length; igl, length of
inflated portion of the gonozooecial frontal wall; gw, gonozooecial
frontal wall width (maximum); low, longitudinal ooeciopore width; tow,
transverse ooeciopore width. Abbreviations of statistical functions: Nc,
number of colonies from which measurements were taken; Nz, number of
zooecia measured; x, mean value; Rc, range of colony means; Rz, total
range of values.

Nc

Nz

X

Rc

Rz

law

3

55

008

008

0-06-0-10

taw

3

55

008

008

0-06-0-09

fwl

3

55

0-67

0-63-0-69

0-46-0-88

fww

3

55

018

017-019

0 14-0 21

tgl

3

3

1-44

1-20-1-82

1-20-1-82

igl

3

6

Ml

0-92-1-34

0-86-1-34

gw

3

6

0-62

0-53-0-67

0-51-0-80

low

3

5

006

0-06

0-05-007

tow

3

5

0-07

0-06-0-08

006-0- 10

TAYLOR: JURASSIC BRYOZOA

703

Genus reptoclausa d’Orbigny, 1853
Reptoclausa porcata sp. nov.

Plate 88, figs. 2-6; text-fig. 2

71894 Berenicea allaudi (Sauvage); Gregory, p. 60.

1896a Berenicea Allaudi (Sauvage); Gregory, p. 44 (partim .).

1896ft Berenicea allaudi (Sauvage); Gregory, p. 77 (partim.), pi. 3, fig. 6.

1969 Idmonea triquetra Lamouroux; Walter, p. 52 (partim.), pi. 3, figs. 11-13 only.

Derivation of name. The trivial name porcata refers to the ridged and furrowed form of the zoarium.

Holotype. BMNH D8724, Aalenian, Pea Grit (murchisonae Zone), Birdlip, Gloucestershire.

Paratypes. BMNH D7526a-b, Aalenian, Pea Grit (murchisonae Zone), near Stroud, Gloucestershire. BMNH
D31586, Aalenian, Lower Limestone (murchisonae Zone), Crickley Hill, Gloucestershire. BMNH B2290a-c,
Inferior Oolite, Crickley Hill, Gloucestershire. BMNH B4855, Lower Ragstone (discites Zone), Cold Comfort,
Gloucestershire. BMNH D1795, Inferior Oolite, ?locality. BMNH D 10091, Pea Grit (murchisonae Zone),
Crickley Hill, Gloucestershire. BMNH D30002a-c, Lower Limestone (murchisonae Zone), Kimsbury,
Painswick, Gloucestershire.

Diagnosis. Reptoclausa with continuous autozooecial ridges separated by furrows of kenozooecia;
zoaria commonly unilamellar, occasionally multilamellar.

Description. Zoaria are adnate, fan-shaped (PI. 88, fig. 2) to discoidal, commonly unilamellar but occasionally
multilamellar (PI. 88, fig. 3). Zooecia arise where existing interzooecial walls divide on a basal lamina at the
colony growth margin. Rounded ridges of low profile cross the zoarial surface parallel to the direction of growth
and form lobate projections where they meet the colony growth margin (PI. 88, fig. 6). Ridge crests are about
2 mm apart and new ridges appear at dichotomies of established ridges. Ridges are occupied by autozooecia
orientated with their long axes slightly divergent from the ridge crest. Zooecium size, particularly width,
decreases progressively away from ridges towards intervening furrows occupied by kenozooecia (text-fig. 2).
Multilamellar growth was achieved either by spiral overgrowth around irregularly distributed pivot points
(Taylor 1976), or by a process, comparable with the frontal budding known in cheilostomes (Banta 1972), in
which an overgrowing zooecium arose from an autozooecial aperture to initiate a fan-shaped expansion on the
zoarial surface. The first zooecium of each new frontally budded layer has a short frontal wall and a
longitudinally elongate aperture.

text-fig. 2. Semidiagrammatic representation of part of a
Reptoclausa porcata colony showing ridges composed of
autozooecia and furrows of kenozooecia lacking aper-
tures. Open autozooecial apertures are shown in black
and the occluded apertures of autozooecia bordering
kenozooecial furrows are stippled. The large zooecium on
the left-hand ridge is a gonozooecium. The lobate distal
growth margin is evenly stippled. Approx, x 18.

704

PALAEONTOLOGY, VOLUME 23

Frontal walls of autozooecia are thick, have rounded distal terminations (PI. 88, fig. 5), and are clearly defined
by traces of vertical interzooecial walls on the zoarial surface, Autozooecial apertures are slightly transversely
elongate. Thin-walled peristomes are preserved only when immured by intracolony overgrowths. Terminal
diaphragms, level with the frontal walls, frequently occlude zooecia, particularly those situated at boundaries
between ridges and furrows (text-fig. 2). Ontogenetic zonation (Silen and Harmelin 1974) of autozooecia is not
apparent.

Kenozooecia, occurring regularly in furrows between autozooecial ridges (PI. 88, fig. 4), have narrow frontal
walls defined by the faint traces on the zoarial surface of their vertical interzooecial walls. Less elongate
kenozooecia may occur at growth margin anastomoses and in the vicinity of zoarial lateral walls.

Gonozooecia are developed in about 50% of the zoaria examined. They are elongate, slightly dilated in width
and inflated, and are situated on zoarial ridges. The transversely elongate ooeciopores are about the same size as
autozooecial apertures.

Dimensions. See Table 2.

table 2. Dimensions (in mm) of Reptoclausa porcata zooecia. Abbrevia-
tions as in Table 1.

Nc

Nz

X

Rc

Rz

law

5

125

010

009-0- 10

0-07-0-11

taw

5

125

010

0-09-0-11

0-08-013

fwl

5

125

0-61

0-52-0-66

0-40-0-80

fww

5

125

0-22

0-22-0-23

0-18-0-29

tgl

4

34

1-70

1-62-1-77

1-17-2-25

gw

4

36

0-43

0-37-0-46

0-35-0-59

low

3

19

009

0-08-0-09

0-07-0-13

tow

3

19

0-12

0-11-0-14

0-10-0-15

Remarks. Among the specimens included by Gregory (18696) in Berenicea allaudi (Sauvage) are two
(BMNH D1794, D1795) belonging to this new species. Rosacilla allaudi of Sauvage (1888) is a simple,
multiserial tubuloporinidean lacking ridged zoaria and quite distinct from the species figured as
Berenicea allaudi by Gregory (18966, pi. 3, fig. 6). Walter (1969, p. 52) includes specimens of
Reptoclausa porcata within Idmonea triquetra Lamouroux 1821. Reptoclausa porcata, however,
differs from Idmonea triquetra in the following features:

1 , R. porcata zoaria have a fan-shaped to discoidal form whereas zoaria of /. triquetra consist of dichotomising
narrow multiserial branches. 2, The branches of I. triquetra have a well-defined triangular cross-section distinct
from the rounded ridges of R. porcata. 3, Ooeciopores of I. triquetra are about half the diameter of R. porcata
ooeciopores. 4, I. triquetra zooecia are usually arranged in distinct rows. Those of R. porcata are not usually
arranged in rows and have larger frontal wall dimensions.

Furthermore, R. porcata is known only from the Upper Aalenian and Lower Bajocian, whereas the
probable range of I. triquetra is Upper Bajocian to Lower Callovian (Walter 1969).

Hillmer (1971, p. 42) noted the similarity between Lower Cretaceous Reptoclausa and two of
Walter’s (1969) figured Idmonea triquetra specimens (BMNH D 10091, D31586) which are here
included in R. porcata. R. porcata differs from the Lower Cretaceous type-species of Reptoclausa,
R. neocomiensis d’Orbigny (redescribed by Hillmer 1971), which has autozooecial ridges discon-
tinuous in the direction of colony growth and kenozooecia occupying a larger proportion of the
zoarial surface. The known range of the genus Reptoclausa is extended back from the Lower
Cretaceous into the Middle Jurassic by the description of R. porcata.

R. porcata is abundant in the Lower Inferior Oolite of the Cotswolds where, along with
Reptomultisparsa cricopora and R. ventricosa, it is found encrusting a variety of substrates including
the large terebratulid brachiopod Pseudoglossothyris simplex and limestone intraclasts. Some of the
brachiopod-encrusting colonies may represent associations with a living brachiopod because

TAYLOR: JURASSIC BRYOZOA

705

bryozoan growth is frequently found to terminate at a growth line on the brachiopod shell suggesting
that growth of both bryozoan and brachiopod were checked simultaneously but, whereas
brachiopod growth later recommenced, bryozoan growth did not (Ager 1961).

Stratigraphical range. Upper Aalenian-Lower Bajocian.

Discussion. The morphology of Reptoclausa porcata contrasts with that of most other multiserial
encrusting tubuloporinideans (e.g. Reptomultisparsa tumida ) and deserves further consideration. In
colonies of Reptoclausa porcata zooecium size, particularly frontal wall width, decreases progres-
sively passing from the crests of the ridges to the bottoms of the furrows (text-fig. 2). Ridge crests
bear broad autozooecia, furrows are composed of narrow kenozooecia, and the intervening regions
between ridge crests and furrows possess comparatively narrow autozooecia typically occluded by
terminal diaphragms. This type of morphological gradient perpendicular to the growth direction of
the colony suggests the presence during life of a physiological gradient (Bronstein 1939; Anstey et at.
1976), perhaps hormonal, which determined zooid structure according to position of budding.
Comparatively large zooids were budded at regularly spaced loci along the growing edge of the
colony. Here, the zoarium was differentially thickened to give a ridge which formed a lobate
projection where it intersected the colony growing edge. These large zooids displayed a dominance
over zooids budded between loci causing them to be crowded and reduced in size. The smallest zooids
budded between loci were too small to support a functional polypide (gut and tentacles) and thus
became kenozooids. It seems that zooids of intermediate size could support only a short-lived
polypide whose degeneration was followed by early occlusion of the zooecial aperture by a terminal
diaphragm. The functional significance of the unusual arrangement of autozooecia and kenozooecia
in Reptoclausa is unclear but may relate to the maintenance of efficient colony feeding, with
autozooid exhalent currents departing from the colony surface above regions of non-feeding
kenozooids (see Taylor 1979).

Acknowledgements. I thank Dr. G. P. Larwood (University of Durham) for constructive criticism of the
manuscript and for his supervision during the course of a N.E.R.C. funded research project. Continued support
from N.E.R.C. in the form of a research fellowship is also gratefully acknowledged. Miss J. Darrell kindly
arranged the loan of specimens from the British Museum (Natural History). Valuable photographic assistance
was provided by Dr. J. C. W. Cope (University College of Swansea).

REFERENCES

ager, d. v. 1961. The epifauna of a Devonian Spiriferid. Q. Jl geol. Soc. Lond. 117, 1 10.
anstey, R. L., pachut, J. f. and prezbindowski, d. r. 1976. Morphogenetic gradients in Paleozoic bryozoan
colonies. Paleobiology, 2, 131-146.

banta, w. c. 1972. The body wall of cheilostome Bryozoa. V. Frontal budding in Schizoporella unicornis
floridana. Mar. Biol. 14, 63-71.

borg, F. 1926. Studies on Recent cyclostomatous Bryozoa. Zool. Bidr. Upps. 10, 181-507.
bronstein, G. 1939. Sur les gradients physiologiques dans une colonie de Bryozoaires. C.r. hebd. Seanc. Acad.
Sci., Paris, 209, 602-603.

busk, G. 1852. Catalogue of marine Polyzoa in the collection of the British Museum. 1. Cheilostomata. 54 pp.,
British Museum (Natural History), London.

canu, f. 1918. Les ovicelles des Bryozoaires Cyclostomes. Etudes sur quelques families nouvelles et anciennes.
Bull. Soc. geol. Fr. 4th series, 16, 324-335.

ehrenberg, c. g. 1831. Symbolae physicae, seu leones et Descriptiones Mammalium, Avium, Insectarum et
Animalium Evertebratorum. Pars Zoologica, Dec. 1 . 4. Berlin.

Gregory, J. w. 1894. Catalogue of the Jurassic Bryozoa in the York Museum. Rep. Yorks, phil. Soc. for 1893,
58-61.

— 1896a. A revision of the British Jurassic Bryozoa. Part III. The Genus Berenicea. Ann. Mag. nat. Hist.
6th series, 17, 41-49.

— 1896 b. Catalogue of the fossil Bryozoa in the Department of Geology, British Museum ( Natural History).
The Jurassic Bryozoa. 239 pp., 1 1 pis., British Museum (Natural History), London.

706

PALAEONTOLOGY, VOLUME 23

hillmer, G. 1971. Bryozoen (Cyclostomata) aus dem Unter-Hauterive von Nordwestdeutschland. Mitt, geol-
palaont. Inst. Univ. Hamburg , 40, 5-106.

milne-edwards, H. 1838. Memoire sur les Crisies, les Horneres et plusieurs autres Polypes vivans ou fossiles
dont l’organisation est analogue a celle des Tubulipores. Annls Sci. nat. Ser. 2, 9, 193-238, pis. 6-16.
orbigny, a. d\ 1853. Paleontologie franqaise , terrains cretaces, 5 Bryozoaires. 1, 1191 pp., 2 (atlas), pis. 600-800.
Paris.

pohowsky, R. A. 1978. The boring ctenostomate Bryozoa: taxonomy and paleobiology based on cavities in
calcareous substrata. Bull. Am. Paleont. 73, No. 301, 1-192.
sauvage, h. e. 1888. Note sur les Bryozoaires Jurassiques de Boulogne. Bull. Soc. geol. Fr. 3rd series, 27, 38-53.
silen, L. and harmelin, J.-G. 1974. Observations on living Diastoporidae (Bryozoa Cyclostomata) with special
regard to polymorphism. Acta zool., Stockh. 55, 81-96.
taylor, p. d. 1976. Multilamellar growth in two Jurassic cyclostomatous Bryozoa. Palaeontology, 19, 293-306.

— 1977. The palaeobiology and systematics of some Jurassic Bryozoa. Ph.D. thesis (unpubl.), University of
Durham .518+ lviii pp .

— 1978. A Jurassic ctenostome bryozoan from Yorkshire. Proc. Yorks, geol. Soc. 42, 211-216.

— 1979. The inference of extrazooidal feeding currents in fossil bryozoan colonies. Lethaia, 12, 47-56.
thorpe, j. p., ryland, j. s. and beardmore, j. a. 1978. Genetic variation and biochemical systematics in the

marine bryozoan Alcyonidium mytili. Mar. Biol. 49, 343-350.
voigt, E. 1977. Arachnidium jurassicum n. sp. (Bryoz. Ctenostomata) aus dem mittleren Dogger von Goslar am
Harz. NeuesJb. Geol. Palaont. Abh. 153, 170-179.

Walter, B. 1969. Les Bryozoaires Jurassiques en France. Etude systematique. Rapports avec la stratigraphie et
la paleoecologie. Docums Lab. Geol. Fac. Sci. Lyon, 35, 1-328, pis. 1-20.

PAUL D. TAYLOR

Department of Palaeontology
British Museum (Natural History)

Manuscript received 2 October 1978 Cromwell Road

Revised manuscript received 20 June 1979 London, SW7 5BD

ANOMALOUS OCCURRENCES OF THE LOWER
PALAEOZOIC BRACHIOPOD SCHIZOCRANIA

by m. g. lockley and D. D. j. antia

Abstract. There are rare occurrences of Ordovician and Silurian species of the inarticulate brachiopod
Schizocrania attached to orthoconic cephalopod shells. These were probably transported considerable distances
prior to their deposition in onshore sediments, in which Schizocrania is not normally found. Relationships
between host and encruster are discussed with a view to elucidating both encrustation sequences and inferred
ecological associations.

During the course of studies of Upper Llanvirn, Ordovician (MGL) and Whitcliffian, Silurian
(DDJA) successions in the Anglo-Welsh region, we noted rare occurrences of orthocones with
Schizocrania (Trematidae) attached to either the inner or outer walls of their body chambers; in both
cases the associated clastic sediments are of a coarse arenaceous type associated with demonstrably
shallow-water facies (Williams 1953; Antia 1979). Havlicek (1972, p. 230) reported that the Upper
Ordovician trematid Ptychopeltis incola Perner from Bohemia \ . . lived attached only to the
cylindrical shells of orthocone nautiloids’; he also noted that its ancestor P. hornyi Havlicek
sometimes encrusted orthocones. We therefore consider that these examples of apparent host-specific
relationships may be paralleled elsewhere amongst the Trematidae (e.g. Schizocrania) by similar
associations between host and encruster.

MATERIAL

The Upper Llanvirn specimen is an incomplete, poorly preserved internal mould of a body chamber of an
orthoconic nautiloid of unknown taxonomic affinity. It was recovered from a shell-bed in the upper part of the
Flags and Grits Formation of the Ffairfach Group exposed at Coed Duon, 3 km south of Llangadog, Dyfed
(Grid Ref. SN 709256), where it lay parallel to bedding. The orthocone has three specimens of Schizocrania cf.
salopiensis Williams attached to the inner surface of the body chamber; the brachial valves all face inwards (text-
fig. 1a) but show no obvious preference for any particular attachment site although two of the specimens are
aligned subparallel to each other near the anterior end.

The Whitcliffian specimens are represented by poorly preserved fragmentary moulds of Orthoceras sp.
(diameters c. 20 mm and > 30 mm respectively) from the Lower Whitcliffe Beds of Mortimer Forest, south of
Ludlow (Grid Ref. SO 497725) and the Upper Whitcliffe Beds near Broadstone Farm (SO 544900). The older
specimen, an internal mould of a large portion of the conch (text-figs. 1b, 2b) has three specimens of S. striata
(Sowerby) attached to the anterior part of its external surface. The specimens all occur close to each other on the
exposed section of the orthocone mould which faces downwards from the undersurface of a bedded unit; relative
to the final entombment position of the orthocone the Schizocrania specimens occur on its ‘underside’ and
following the dissolution of the cephalopod shell have become impressed on to the preserved mould. The
younger (upper Whitcliffian) specimen consists of the internal and external moulds of a curved fragment of a
large body chamber; it has five poorly preserved specimens of S. striata attached to its inner (concave) surface
which faces downwards. The specimens are aligned transversely, parallel to the peristome (text-figs. 1, 2c).

The lectotype (Geol. Surv. Mus. No. 6631) of A. striata (Sowerby) from the Leintwardinian-Whitcliffian beds
of Delbury, Salop (Grid Ref. SO 501854) is the only other known British Schizocrania which we have discovered
attached to an orthoconic nautiloid fragment; the specimen is attached to the convex surface of the free part of a
septum, probably the last one; it differs from the other examples in its larger size (length 9 mm) and posterior
attachment site (text-fig. 2a).

[Palaeontology, Vol. 23, Part 3, 1980, pp. 707-713.1

708

PALAEONTOLOGY, VOLUME 23

■ection of for

'ward growth

of

Schizocrani

>o specimens

length

width(n

BB 92492

55

60

BB 92493

42

4-2

BB 92494

30

30

BB 94078

3-7

3-7

BB 94079

(3 6)

3-6

BB 94080

2-3

2-5

BB 94081

(40)

(4-0)

BB 94082

3-5

3 7

BB 94083

(4-0)

(4-0)

BB 94084

(2-5)

(2-5)

BB 94085

(2-5)

(2-5)

text-fig. 1. Scale drawings of Schizocrania encrusted orthoconic nautiloids
from Upper Llanvirn strata exposed near Llangadog, Mid Wales (a) and from
lower (b) and upper (c) Whitcliffe strata exposed near Ludlow, Salop. All
Schizocrania specimens have British Museum numbers, specimens BB94078-
94080 are attached to the outer surface of the shell mould (b) whilst the
remaining specimens are attached to the inner surfaces of the shell moulds.

Approximate length, width measurements, listed bottom right.

OBSERVATIONS AND INTERPRETATION

All twelve of these Schizocrania specimens exhibit only their convex brachial valves facing away from
the cephalopod shell surface. Schizocrania is known to attach to substrates by its flat pedicle valve
(Hall and Whitfield 1875; Rowell in Williams et al. 1965, p. H283). However, pedicle valves are
exceptionally rare, being either altogether absent from assemblages or hidden from view beneath the
brachial valve. The three orthocone specimens shown in text-fig. 1 indicate that the anterior edge of
the phragmacone was the preferred encrustation site for all but two of the Schizocrania specimens.
The orientation of these Schizocrania inside the phragmacone and on the shell exterior is apparently
not random since all adjacent shells are aligned with their umbones pointing in approximately the
same direction (i.e. transverse or oblique to the orthocones’ long axis).

LOCKLEY AND ANTIA: DISPERSAL OF SCHIZOCRANIA

709

The orthocones may have been encrusted while they were alive and mobile, or when they were dead
and floating, or dead and semi-buoyant, being washed around on the sea floor, or dead and settled on
the sea floor, or, finally, when being reworked.

In addition to the numerous examples of fossil cephalopod (ammonoid) encrustation recorded
from Mesozoic assemblages (e.g. Seilacher 1960; Meischner 1968) and the few broadly analogous
Lower Palaeozoic examples involving orthoconic nautiloids (Holland 1971; Havlicek 1972), we have
noted Ordovician and Silurian collections containing several varied and undescribed examples of
orthocone encrustation (e.g. National Museum of Wales specimen NMW 79. 5G. Map loc. 771;
Hunterian Museum specimens S. 25129/1-3 and S. 251 14 a/b). Schizocrania is ornamented by
numerous radial capillae (Williams 1974, p. 44). According to Williams and Wright 1963, p. 19 and
Williams and Rowell (in Williams etal. 1965, p. H81) such radial ornament probably supported setal
follicles at the commissure, and it is reasonable to assume that Schizocrania was particularly
setiferous. Sudarson (1969, p. 65) noted that Discinisca larvae have well-developed principle setae
and that ‘there may be a prolonged larval stage . . . with chaetae increasing in number to facilitate
floatation’. Both the Schizocrania species discussed here exhibit high capillae densities at the same

text-fig. 2. a. Schizocrania striata lectotype showing attachment to mould of orthocone septum from upper
Ludlow beds, Delbury, Shrops., x 3. B, Detail of S. striata specimens BB94078 {left) and BB94079 (right) from
Lower Whitcliffe Beds, Mortimer Forest, Ludlow, x 12-5; see also text-fig. 1b. c, S. striata specimens BB94081
(top) to BB94085 showing attachment to orthocone body chamber fragment, the edge of which is arrowed, from
Upper Whitcliffe Beds, Broadstone farm, Ludlow, x 6. Text-fig. lc is a scale-drawing of the counterpart of this

specimen.

710

PALAEONTOLOGY, VOLUME 23

growth stage (i.e. 10-12 per mm, 5 mm antero-medianly of the dorsal umbones) and probably
therefore had a juvenile epiplanktic stage.

Holland (1971, p. 18) considered that strophomenid (aegeromenid) and rhynchonellid ( Micro -
sphaeridorhynchus nucula ) brachiopods might have attached to living orthocone hosts but concluded
that due to the size of the brachiopods this was ‘unlikely’. Havlicek (1967, p. 21) demonstrated the
attachment of epiplanktic strophomenids to the ‘stems of algae’ (Havlicek 1967, p. 21). He
subsequently suggested (Havlicek 1972, p. 230) that aegeromenids attached to live orthocones and
considered that inarticulates such as Ptychopeltis incola ‘were attached to the shells of living
nautiloids’ (Havlicek 1972, p. 230) whilst related trematids attached both to orthocones and other
specific ‘freely moving organisms’ (Havlicek 1972, p. 229). An orthocone encrusted with Conchiolites
(Ordovician) was described by Prantl (1948, p. 6). Seilacher (1954, 1968) concluded (1968, p. 284) that
the preferentially orientated epizoans on this specimen were adjusted to the ‘head-on motion of their
host’. Both Havlicek (1972, p. 230) and Seilacher (1968) suggested that preferred orientation of
encrusters is of prime importance in testifying to pre-mortem attachment. This suggests that the
majority of known Schizocrania specimens were attached at various stages in the orthocone’s post-
mortem history. Although Havlicek (1972, p. 229) presumed that aegeromenid brachiopods such as
those depicted by Holland (1971, fig. 1 b) attached to live orthocones, direct evidence for this is
insubstantial. Although these authors, and Bergstrom (1968) have shown such brachiopods attached
in rows along orthocones and ‘algal stems’ such arrangements do not constitute the type of preferred
orientation referred to above.

Since modern spirorbids are known to be host specific and capable of seeking a preferred
attachment site and orientation (Knight-Jones 1951), it is almost certain that the occurrence of fossil
spirorbids aligned along the growth margins of orthocones (Holland 1971) indicates a comparable
relationship. This may mean that the similar alignment of Schizocrania specimens noted here (text-
fig. 1) could also be indicative of a host-specific relationship. Such a contention tends to be supported
by our observation that the Anglo-Welsh Schizocrania have not been found attached to any other
host organisms and would also offer a possible explanation for the virtual absence of pedicle valves,
which could have either remained attached to a host when the brachial valve disarticulated, or
become obscured during fossilization by the substrate to which they were attached.

The Schizocrania on the internal surface of the body chambers of the Llanvirn and upper
Whitcliffian specimens indicate encrustation beginning no earlier than the post-mortem drifting
phase (following decay of mantle lining the body chamber) but prior to the infilling of the body
chamber. Interpretation of the lower Whitcliffian orthocones’ pre-entombment history is prob-
lematical; it could have been encrusted at any one of a number of stages in its history as a live or dead
mobile organism. However, since the Schizocrania are attached to its ‘underside’ they must have
settled and had time to grow prior to its final entombment in this position. The S. striata lectotype
must have become attached to the posterior side of its septal substrate after the separation of the
orthocone’s body chamber from the remaining posterior part of the shell (i.e. at a relatively late stage
in the orthocones’ post-mortem history).

On the lower Whitcliffian orthocone the internal mould (text-fig. 1 b) is covered by numerous
irregular markings consisting mainly of small elongated raised protruberances averaging about
0T mm in height and width and between 0-3 and 0-7 mm in length. These apparently represent the
internal moulds of bryozoan borings on the inner surface of the orthocone shell although it is not
altogether clear whether some of the flatter or even slightly indented markings may not result from
the fossilization of external borings. In any event where the Schizocrania shells are slightly broken,
and around their edges, it is evident that the borings affect the orthocone shell beneath.
Unfortunately the absence of a counterpart of this specimen renders this evidence inconclusive.

Distribution of Schizocrania

The Llanvirn orthocone and Schizocrania discussed here are virtually the only representatives of
these taxa known from the predominantly arenaceous and rudaceous Ffairfach Group of the
Llandeilo area. Since S. salopiensis is common in penecontemporaneous, argillaceous successions

LOCKLEY AND ANTIA: DISPERSAL OF SCHIZOCRANIA

711

elsewhere in South Wales and the Welsh Borderlands (Williams 1974; Bassett et al. 1974, p. 9;
Lockley and Williams, in press) where there are different benthic and pelagic faunas (i.e. trilobites,
graptolites, and cephalopods), it is reasonable to assume that the exotic Ffairfach occurrence may
have been related to the drifting or migration of a stray cephalopod beyond the normal limits of its
indigenous environment. Such post-mortem drifting of modern cephalopods is well known (House
1973; Kennedy and Cobban 1976; Hewitt and Pedlay 1978) and may result in individual specimens
being transported for hundreds or even thousands of kilometres.

Similarly S. striata is rare in the Whitcliffe Beds of the Ludlow region where it constitutes only
about 0-01 to 0 005% of the total fauna with specimens generally occurring in a fragmentary
condition and random orientations. It is more common in unbioturbated, parallel-laminated,
alternating light and dark siltstones (rhythmites) of deeper-water facies (e.g. Upper and Lower
Leintwardinian Beds, Holland et al. 1963, p. 154; Lawson 1973, p. 274) and is recorded only rarely in
shallow-water bioturbated siltstones (Facies B sensu Antia 1979). Again, the Whitcliffian cephalo-
pods drifted into inshore deposits from an offshore source, although limited evidence also points to
later phases of encrustation (e.g. lectotype). Williams (1969, p. 143) discussed the potential range of
larval dispersal and its bearing on brachiopod migration during the Ordovician. Clearly his suggested
figure (up to 250 km) is only a fraction of the range potential for brachiopods capable of encrusting
live or drifting orthocones.

Trematid hosts

Encrusting Trematidae such as Schizocrania, Drabodiscina, and Ptychopeltis appear to have been
host specific. S. salopiensis, S. striata, and P. incola have hitherto only been observed attached to
orthoconic nautiloids generally presumed to have been alive or floating at the time of their
encrustation. Other members of the family, e.g. P. hornyi Havlicek and D. grandis Barrande, are
commonly attached to conularids which are considered by Havlicek (1972) to have been mobile
during life, and the American species S. filosa Hall frequently attached to the brachiopod
Rafinesquina (e.g. Cooper 1956 and Rowell in Williams et al. 1965). With respect to trematid -
nautiloid associations, it is intriguing to note that Titus and Cameron (1976) record S. filosa only in
their deep-water Geisonoceras (Orthocerida) community. Dr. R. A. Hewitt and Mrs. D. Evans (pers.
comm. 1979) inform us that they know of no Silurian or Ordovician examples of cephalopod
encrustation by brachiopods other than those reported here, which is suggestive of host-specific
relationships.

CONCLUSIONS

Faunal associations with abundant Schizocrania in the Ordovician and Silurian of the Anglo-Welsh
region are almost invariably confined to argillaceous deep-water facies where species of the genus are
represented almost exclusively by assemblages of brachial valves. Such exceptionally dispropor-
tionate valve ratios are considered to result from their encrustating habits which might account for
the obscuring or removal of pedicle valves. Known associations between trematid encrusters and
hosts such as those reported here and elsewhere (e.g. Havlicek 1972; Rowell in Williams et al. 1965)
point to some form of host-specific relationship between representatives of the family and other
larger invertebrate hosts. Whether such relationships could be termed symbiotic, commensal, or
parasitic is unclear because we lack evidence which demonstrates that hosts were encrusted during
life. However, we can establish that encrustation of orthocones, which may in some cases have begun
during their life, often began no earlier than the post-mortem drifting phase, and may have continued
or begun at a time when the orthocones were resting or rolling on the sea floor. Since encrustation of
many of these orthocones could not have taken place when they were in the final ‘resting’ position it
must have occurred during the middle phases of their pre-entombment history.

The following suggestions on the time of encrustation can be made: (1) The encrusting
Schizocrania noted here are not currently known to attach to non-orthocone skeletal components
within the deposits from which they were recovered and are therefore likely to have settled

712

PALAEONTOLOGY, VOLUME 23

preferentially on orthocone shells prior to their final deposition. (2) The apparent high-density,
orderly clustering of Schizocrania towards the anterior of the conch suggests that possibly the
orthocone was colonized as a specific host whilst it was floating. (3) Since both Schizocrania and its
nautiloid hosts are normally indigenous to sparsely fossiliferous, low-density offshore facies, it is
probable that encrustation occurred in an offshore region before the orthocones finally became
entombed in more diverse, fossiliferous, onshore facies where Schizocrania is invariably rare. This
inference is supported by the observation that the setiferous Schizocrania may well have been adapted
to a prolonged larval stage which would have enhanced its chances of encountering a suitable
encrustation site. If Schizocrania even occasionally encrusted orthocones in a manner analogous to
the attachment of epiplanktic aegeromenids to buoyant organisms noted by Bergstrom (1968), then
the combined effect of nautiloid mobility during life and drifting after death would offer an
explanation for exceptionally widespread occurrences of certain kinds of brachiopods.

Acknowledgements. We thank Dr. G. E. Farrow and Dr. R. A. Hewitt for critically reading the manuscript; Dr.
M. G. Bassett, Dr. A. Williams, Dr. J. D. Lawson, Dr. D. Atkins, Mrs. D. Evans, Mr. I. Jarvis, and Dr.
P. Sheldon are also thanked. Both authors acknowledge the receipt of N.E.R.C. grants.

REFERENCES

antia, d. d. j. 1979. Bone-Beds: A review of their classification, occurrence, genesis, geochemistry, ecology,
diagenesis, weathering and micro-biotas. Mercian Geol. 7, 93-174, 6 pis.
bassett, D. a., ingham, j. K. and wright, a. d. (eds.). 1974. Field Excursion Guide to Type and Classical Sections
in Britain. (Ordovician System Symposium, Birmingham 1974). The Palaeontological Assoc. London. 66 pp.
bergstrom, J. 1968. Some Ordovician and Silurian Brachiopod Assemblages. Lethaia, 1, 23-237.
cooper, G. a. 1956. Chazyan and related brachiopods. Smithson. Misc. Coll. 127, 1-1245, pis. 1-169.
hall, J. and whitfield, r. p. 1875. Descriptions of invertebrate fossils mainly from the Silurian System. Rep.
Geol. Surv. Ohio , 2, 65-179.

havlicek, v. 1967. Brachiopoda of the Suborder Strophomenida in Czechoslovakia. Rozpr. ustr. Ust. geol. 33,
1-235.

— 1972. Life habit of some Ordovician inarticulate brachiopods. Vestnik ustred ust. geol. 47, 229-233.
hewitt, R. a. and pedley, H. M. 1978. The preservation of the shells of Sepia in the Middle Miocene of Malta.

Proc. Geol. Ass. 89 (3), 227-237.

Holland, c. H. 1971. Some conspicuous participants in Palaeozoic symbiosis. Sci. Proc. R. Soc. Dublin, Ser. A,
4, 15-26.

— lawson, j. D. and walmsley, v. g. 1963. The Silurian rocks of the Ludlow district, Shropshire. Bull. Brit.
Mus. Nat. Hist., Geol. 8, 95-171.

house, m. r. 1973. An analysis of Devonian Goniatite distributions. Spec. Pap. Palaeont. 12, 305-317.
Kennedy, w. J. and cobban, w. D. 1976. Aspects of Ammonite Biology and Biostratigraphy. Ibid. 17, 1-94.
knight-jones, e, w. 1951. Gregariousness and some other aspects of the settling behaviour of Spirorbis.
J. Marine Biol. Ass. U.K. 30, 202-222.

lawson, J. d. 1973. Facies and faunal changes in the Ludlovian rocks of Aymestry, Herefordshire. Geol. J. 8,
247-278.

lockley, m. G. and williams, A. Lower Ordovician Brachiopoda from Mid and South Wales. Bull. Br. Mus. Nat.
Hist. Geol. (In press.)

meischner, d. 1968. Perniciose Epokie von Placunopsis auf Ceratites. Lethaia, 1, 156-174.
prantl, F. 1948. The genus Conchicolites Nicholson (Serpulimorpha) in the Ordovician of Bohemia. Vestn.
Krai. Ceske Spolecn. Nauk. (tr.) 9, 1 -7.

seilacher, a. 1954. Okologie der triassichen Muschel Lima lineata (Schloth) und ihrer epoken. Neues Jahrb.
Geol. Palaontol. Monatsh. 4, 163-183.

— 1960. Epizoans as a key to Ammonoid Ecology. J. Paleont. 34, 189-193.

— 1968. Swimming habits of Belemnites— recorded by Boring barnacles. Palaeogeog. Palaeoclimatol.
Palaeoecol. 4, 279-285.

sudarson, A. 1969. Brachiopod larvae from the west coast of India. Proc. Ind. Acad. Sci. 68B, 59-68.
titus, r. and cameron, b. 1976. Fossil Communities of the Lower Trenton Group (Middle Ordovician) of
Central and North Western New York State. J. Paleont. 50, 1209-1225.

LOCKLEY AND ANTIA: DISPERSAL OF SCHIZOCRANIA

713

williams, A. 1953. The geology of the Llandeilo district, Carmarthenshire. Q. Jl geol. Soc. Lond. 108, 177-208.

— 1969. Ordovician faunal provinces with reference to brachiopod distribution. In wood, a. (ed.). The Pre-
Cambrian and Lower Palaeozoic Rocks of Wales. Univ. of Wales Press, Cardiff.

— 1974. Ordovician Brachiopods from the Shelve district, Shropshire. Bull. Brit. Mus. Nat. Hist. Geol. Suppl.
11, 1-163, pis. 1-28.

— and wright, A. D. 1963. The Classification of the ‘ Orthis testudinaria Dalman’ Group of Brachiopods.
J. Paleont. 37, 1-32.

— et al. In MOORE, R. c. (ed.). 1965. Treatise on Invertebrate Paleontology. Part H. Brachiopoda. Univ. Kansas
Press.

M. G. LOCKLEY
Department of Geology
University of Glasgow
Glasgow, G12 8QQ
Scotland

Manuscript received 12 September 1979
Revised manuscript received 12 December 1979

D. D. J. ANTIA

B.P. Development Co. Ltd.
Fairburn Estate
Dyce
Aberdeen

THE PALAEONTOLOGICAL ASSOCIATION

The Association was founded in 1957 to further the study of palaeontology. It holds meetings and demonstrations as well
as publishing Palaeontology and Special Papers in Palaeontology. Membership is open to individuals and to institutions on
payment of the appropriate annual subscription. Rates for 1980 are:

Institutional membership ..... £25 (U.S. $60)

Ordinary membership £12 (U.S. $29)

Student membership ..... £7-50 (U.S. $18)

There is no admission fee. Correspondence concerned with Institutional Membership should be addressed to Dr. C. H. C.
Brunton, Department of Palaeontology, British Museum (Natural History), Cromwell Road, London SW7 5BD, England.
Student members are persons receiving full-time instruction at educational institutions recognized by the Council. On first
applying for membership, an application form should be obtained from the Membership Treasurer. Subscriptions cover one
calendar year and are due each January; they should be sent to the Membership Treasurer (for name and address see
Council list below).

PALAEONTOLOGY

All members who join for 1 980 will receive Volume 23, Parts 1 -4. All back numbers are still in print and may be ordered from
B. H. Blackwell, Broad Street, Oxford 0X1 3BQ, England, at £15 (U.S. $36) per part (post free).

SPECIAL PAPERS IN PALAEONTOLOGY

Members may subscribe to this by writing to the Membership Treasurer (for name and address see Council list below);
the subscription rate for 1980 is £20 (U.S. $48) for Institutional Members, and £10 (U.S. $24) 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 prices shown in the list inside the front cover. Non-members may obtain copies, at the
listed prices, from B. H. Blackwell, Broad Street, Oxford OX1 3BQ, England.

COUNCIL 1980-1981

President: Dr. W. H. C. Ramsbottom, Institute of Geological Sciences, Leeds LSI 5 8TQ
Vice-Presidents: Dr. C. T. Scrutton, Department of Geology, The University, Newcastle upon Tyne NE1 7RU
Professor C. B. Cox, Department of Zoology, King’s College, Strand, London WC2R 2LS
Treasurer: Mr. R. P. Tripp, High Wood, West Kingsdown, Sevenoaks, Kent TNI 5 6BN
Membership Treasurer : Dr. J. C. W. Cope, Department of Geology, University College, Swansea SA2 8PP
Secretary: Dr. R. Riding, Department of Geology, University College, Cardiff CF1 1XL

Editors

Dr. M. G. Bassett, Department of Geology, National Museum of Wales, Cardiff CF1 3NP
Dr. K. C. Allen, Department of Botany, Bristol University, Bristol BS8 1UG
Dr. R. A. Fortey, Department of Palaeontology, British Museum (Natural History),
Cromwell Road, London SW7 5BD

Dr. A. L. Panchen, Department of Zoology, The University, Newcastle upon Tyne NE1 7RU

Other Members of Council

Dr. R. J. Aldridge, Nottingham

Dr. M. D. Brasier, Hull

Dr. D. E. G. Briggs, London

Dr. C. H. C. Brunton, London

Dr. S. Conway Morris, Open University

Dr. R. Harland, Leeds

Dr. M. B. Hart, Plymouth

Dr. P. M. Kier, Washington
Dr. J. Miller, Edinburgh
Dr. T. J. Palmer, Oxford
Dr. M. Romano, Sheffield
Dr. D. J. Siveter, Leicester
Dr. N. H. Trewin, Aberdeen
Dr. J. Watson, Manchester

Overseas Representatives

Australia : Professor B. D. Webby, Department of Geology, Sydney University, Sydney, N.S.W., 2006
Canada: Dr. B. S. Norford, Institute of Sedimentary and Petroleum Geology, 3303-33rd Street NW., Calgary, Alberta
New Zealand: Dr. G. R. Stevens, New Zealand Geological Survey, P.O. Box 30368, Lower Hutt
West Indies and Central America : Mr. John B. Saunders, Geological Laboratory, Texaco Trinidad, Inc., Pointe-a-Pierre,
Trinidad, West Indies

Western U.S. A. : Professor J. Wyatt Durham, Department of Paleontology, University of California, Berkeley 4, California

Eastern U.S.A. : Professor J. W. Wells, Department of Geology, Cornell University, Ithaca, New York

South America: Dr. O. A. Reig, Departmento de Ecologia, Universidad Simon Bolivar, Caracas 108, Venezuela

Palaeontology

VOLUME 23 ■ PART 3

CONTENTS

Dinoflagellate cysts from the Upper Eocene-Lower Oligocene
of the Isle of Wight

M. LIENGJARERN, L. COSTA, and C. DOWNIE

475

Dictyodora from the Silurian of Peeblesshire, Scotland
M. J. BENTON and N. H. TRF.WIN

501

Lower Cretaceous Terebratulidae from south-western
Morocco and their biogeography
F. A. MIDDLEMISS

51.5

Collignoniceratid ammonites from the Mid-Turonian of
England and northern France

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

557

The trilobite Eccoptochile from the Ordovician of
northern Portugal

M. ROMANO

605

The Miocene horse Hipparion from North America and
from the type locality in southern France
B. J. MACFADDEN

617

The Toarcian age of the upper part of the Marlstone Rock
Bed of England

M. K. HOWARTH

637

Jurassic araucarian cone from southern England
R. A. STOCKEY

657

Nomenclature and homology in peridinialean dinoflagellate
plate patterns
G. L. EATON

667

Mode of life of a giant capulid gastropod from the Upper
Cretaceous of Saghalien and Japan

I. HAYAMI and Y. KANIE

689

Two new Jurassic bryozoa from southern England

P. D. TAYLOR

699

Anomalous occurrences of the lower Palaeozoic brachiopod
Schizocrania

M. G. LOCKLEY and D. D. J. ANTIA

707

Printed in Great Britain at the University Press. Oxford
by Eric Buckley , Printer to the University

ISSN 0031-0239

Palaeontology

VOLUME 23 • PART 4 DECEMBER 1980

Published by

The Palaeontological Association London

Price £1 5

THE PALAEONTOLOGICAL ASSOCIATION

The Association publishes Palaeontology and Special Papers in Palaeontology. Details of membership and subscription rates
may be found inside the back cover.

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 on all. aspects of
palaeontology and stratigraphical palaeontology are invited. They should conform in style to those already published in this
journal, and should be sent to Dr. R. A. Fortey, Palaeontological Association, Department of Palaeontology, British Museum
(Natural History), Cromwell Road, London SW7 5BD, England, who will supply detailed instructions for authors on request
(these were published in Palaeontology 1977, 20, pp. 921-929).

Special Papers in Palaeontology is a series of substantial separate works; the following are available (post free).

1. (for 1967); Miospores in the Coal Seams of the Carboniferous of Great Britain, by A. H. v. smith and m. a.
butterworth. 324 pp., 72 text-figs., 27 plates. Price £8 (U.S. $19.50).

2. (for 1968): Evolution of the Shell Structure of Articulate Brachiopods, by a. williams. 55 pp., 27 text-figs., 24 plates.
Price £5 (U.S. $12).

3. (for 1968): Upper Maestrichtian Radiolaria of California, by Helen p. foreman. 82 pp., 8 plates. Price £3 (U.S. $7.50).

4. (for 1969): Lower Turonian Ammonites from Israel, by R. freund and m. raab. 83 pp., 15 text-figs., 10 plates. Price £3
(U.S. $7.50).

5. (for 1969): Chitinozoa from the Ordovician Viola and Fernvale Limestones of the Arbuckle Mountains, Oklahoma,
by w. a. m. jenkins. 44 pp., 10 text-figs., 9 plates. Price £2 (U.S. $5).

6. (for 1969): Ammonoidea from the Mata Series (Santonian-Maastrichtian) of New Zealand, by r. a. Henderson. 82 pp.,
13 text-figs., 15 plates. Price £3 (U.S. $7.50).

7. (for 1970): Shell Structure of the Craniacea and other Calcareous Inarticulate Brachiopoda, by A. williams and
a. D. wright. 51 pp., 17 text-figs., 15 plates. Price £1 - 50 (U.S. $4).

8. (for 1970): Cenomanian Ammonites from Southern England, by w. j. Kennedy. 272 pp., 5 tables , 64 plates. Price £8
(U.S. $19.50).

9. (for 1971): Fish from the Freshwater Lower Cretaceous of Victoria, Australia, with Comments on the Palaeo-
environment, by m. waldman. 130 pp., 37 text-figs., 18 plates. Price £5 (U.S. $12).

10. (for 1971): Upper Cretaceous Ostracoda from the Carnarvon Basin, Western Australia, by r. h. bate. 148 pp., 43 text-
figs. , 27 plates. Price £5 (U.S. $12).

11. (for 1972): Stromatolites and the Biostratigraphy of the Australian Precambrian and Cambrian, by M. R. Walter.
268 pp., 55 text-figs., 34 plates. Price £10 (U.S. $24).

12. (for 1973): Organisms and Continents through Time. A Symposium of 23 papers edited by N. F. hughes. 340 pp.,
132 text-figs. Price £10 (U.S. $24) (published with the Systematics Association).

13. (for 1974): Graptolite studies in honour of O. M. B. Bulman. Edited by R. b. rickards, d. e. jackson, and c. p. hughes.
261 pp., 26 plates. Price £10 (U.S. $24).

14. (for 1974): Palaeogene Foraminiferida and Palaeoecology, Hampshire and Paris Basins and the English Channel, by
j. w. Murray and c. a. wright. 171 pp., 45 text-figs., 20 plates. Price £8 (U.S. $19.'50).

15. (for 1975): Lower and Middle Devonian Conodonts from the Broken River Embayment, North Queensland, Australia,
by p. G. telford. 100 pp., 9 text-figs., 16 plates. Price £5-50 (U.S. $13.50).

16. (for 1975): The Osfracod Fauna from the Santonian Chalk (Upper Cretaceous) of Gingin, Western Australia, by
j. w. neale. 131 pp., 40 text-figs., 22 plates. Price £6-50 (U.S. $16).

17. (for 1976): Aspects of Ammonite Biology, Biogeography, and Biostratigraphy, by w. J. Kennedy and w. a. cobban.
94 pp., 24 text-figs., 11 plates. Price £6 (U.S. $14.50).

18. (for 1976): Ostracoderm Faunas of the Delorme and Associated Siluro-Devonian Formations, North West Territories,
Canada, by D. L. dineley and e. j. loeffler. 218 pp., 78 text-figs., 33 plates. Price £20 (U.S. $48).

19. (for 1977): The Palynology of Early Tertiary Sediments, Ninetyeast Ridge, Indian Ocean, by E. m. kemp and w. K. Harris.
74 pp., 2 text-figs., 8 plates. Price £7 (U.S. $17).

20. (for 1977): Fossil Priapulid Worms, by S. c. morris. 159 pp., 99 text-figs., 30 plates. Price £16 (U.S. $38.50).

21. (for 1978): Devonian Ammonoids from the Appalachians and their bearing on International Zonation and Correla-
tion, by m. r. house. 70 pp., 12 text-figs., 10 plates. Price £12 (U.S. $29).

22. (for 1978, published 1979) : Curation of palaeontological collections. A joint colloquium of The Palaeontological Associa-
tion and Geological Curators’ Group. Edited by m. g. bassett. 280 pp., 53 text-figs. Price £25 (U.S. $60).

23. (for 1979): The Devonian System. A Palaeontological Association International Symposium. Edited by M. R. house, c. t.
scrutton, and m. g. bassett. 353 pp., 102 text-figs., 1 plate. Price £30 (U.S. $72).

24. (for 1980): Dinoflagellate Cysts and Acritarchs from the Eocene of southern England, by J. B. bujak, C. downie, g. l.
eaton, and G. l. williams. 104 pp., 24 text-figs., 22 plates. Price £15 (U.S. $36).

© The Palaeontological Association, 1980

Cover: Edriophus levis (Bather, 1914) from the Middle Ordovician Trenton Group of Kirkfield, Ontario, x 2-5.
Specimen in the Smithsonian Institution; photograph by H. B. Whittington.

THE TRILOBITE TRETASPIS FROM THE UPPER
ORDOVICIAN OF THE OSLO REGION, NORWAY

by ALAN W. OWEN

Abstract. All known Norwegian species of Tretaspis are described. Six are established taxa: T. ceriodes
(Angelin) angelini Stormer, T. seticornis (Hisinger), T. anderssoni Stormer, T. hadelandica hadelandica Stormer,
T. sortita (Reed) broeggeri Stormer, and T. kiaeri Stormer. Three are new: T. hisingeri, T. askerensis , and T.
latilimbus (Linnarsson) norvegicus. Most of these taxa have a broad range of variation encompassing two or
more morphs. The relative proportions of these morphs are used to distinguish T. latilimbus norvegicus and T.
sortita broeggeri from their nominate subspecies. The British form T. convergens Dean and its subspecies are
reinterpreted as subspecies of T. hadelandica. Ingham’s concept of species groups within Tretaspis is revised with
the North American species and, provisionally, T. kiaeri and T. calcaria Dean being recognized as a distinct
group centred on T. sagenosus Whittington. Neoteny is considered to have played a part in the evolution of
Tretaspis.

Species of the trinucleid Tretaspis have played an important part in the correlation of late Caradoc
and Ashgill successions in Britain over the past two decades (Dean 1961, 1963; Ingham 1970; Price
1973, 1977; McNamara 1979). The classical studies by Stormer (1930, 1945) on the Scandinavian
trinucleids include a number of species of Tretaspis , most of which are closely related to British forms.
The present study is part of a broader project aimed at revising the late Caradoc and Ashgill
stratigraphy and trilobite faunas of the Oslo Region. This area was divided into eleven districts by
Stormer (1953, text-fig. 1) and Tretaspis is known from four of them (text-fig. 1). Of these, the upper
Ordovician stratigraphy of two, Hadeland and Ringerike, has been revised by Owen (1978, 1979) and
summaries of the successions in the other two, Oslo-Asker and Skien-Langesund, are given by Strand
and Henningsmoen (1960, pi. 7). In the case of Oslo-Asker, the youngest Ordovician units were
redescribed by Brenchley and Newall (1975). The present study includes an examination of all
available museum material and samples of Tretaspis collected by the writer from 1 1 7 localities in
Oslo-Asker, Hadeland, and Ringerike, now housed in the Paleontologisk Museum, Oslo (PMO) and
the Hunterian Museum, Glasgow (HM).

TREATMENT OF DATA

The distribution of pits on the bilamellar fringe of trinucleids is one of the major taxonomic features
of the group (Hughes, Ingham, and Addison, 1975, pp. 550-552) and the terminology applied herein
is that advocated by Hughes et al. (1975, pp. 543-545, text-figs. 3, 4). Hughes (1970) demonstrated
that although individual specimens of Trinucleus fimbriatus Murchison may show slight asymmetry
in the development of fringe pits, there is no significant statistical difference between the left and right
sides of the fringe when populations are considered. It has thus become standard practice to present
data in terms of half-fringe pit counts and this is followed herein. Moreover, Hughes also
demonstrated that the distribution and number of pits is independent of holaspid specimen size and
this also is assumed for other trinucleids.

The number of arcs of pits and the number of pits in each arc has been used for defining species
and subspecies in various trinucleids, not least Tretaspis. These features can be determined even in
heavily distorted material and lend themselves to simple univariate techniques of display and
analysis. Such an approach is adopted here and enables direct comparisons to be made with data
presented in other studies. Moreover, few horizons in the Oslo Region have yielded more than a

[Palaeontology, Vol. 23, Part 4, 1980, pp. 715-747, pis. 89-93.|

716

PALAEONTOLOGY, VOLUME 23

text-fig. 1. Stratigraphical ranges and suggested phylogeny of Norwegian and closely related British species of
Tretaspis in terms of the standard British succession. The geographical distribution of Norwegian forms is given
also: O-A = Oslo-Asker (Oslo is in the eastern part of this district), H = Hadeland, R = Ringerike, S-
L = Skien-Langesund. The British species are revised to some extent herein.

dozen or so specimens and although many thousands of specimens have been examined, these
comprise relatively few complete half-fringes, let alone entire fringes and thus most specimens have
provided information on only a small proportion of the possible parameters. Univariate analysis
therefore is preferred.

PROBLEMS OF POLYMORPHISM

Hughes et al. (1975, p. 590) noted that in many trinucleid stocks there is a progressive increase in the
number of I arcs between 1 1 and In. This is broadly the case in Tretaspis and in general terms the fringe
criteria used in defining species and subspecies are, in decreasing order of importance: (1) the number
of arcs present, (2) whether these arcs are complete anteriorly and/or posteriorly, (3) the range of
variation in pit number per arc and along the posterior margin of the fringe. These features are closely
related in most forms in that there is a threshold value (4-7 pits in Norwegian forms) for the number
of pits present in the I arc adjacent to In before that arc can be complete anteriorly and a greater
threshold (7-13 pits in Norwegian forms) before another incomplete arc can be developed between it
and In. Other taxonomic fringe features are more dependent on preservation and include the extent of
pits in adjacent arcs sharing sulci, the size of individual pits and the development of lists between arcs.

Many of the species and subspecies of Tretaspis described from Britain appear to have a fairly
narrow range of variation with a purely typological concept based on characters 1 and 2 listed above
being sufficiently diagnostic for both the taxon and all the individuals within it. In some cases this
may be simply an artefact of small sample sizes. In contrast, Price (1977, pp. 764-772) found that
some populations of Tretaspis from Wales have a range of variation in fringe characters which
encompasses that seen in two named taxa which he considered to be end-member subspecies.
Similarly, Lesperance and Bertrand (1976) distinguished a number of different morphotypes within

OWEN: TRILOBITE TRETASPIS

717

Cryptolithus although this needs reassessment in relation to the development of F pits on the
posterior part of the fringe (Owen 1980).

Most of the Norwegian populations of Tretaspis have a broad range of variation, in some instances
comprising morphotypes which correspond to the type specimens of described taxa. This presents
great problems which, if taken to extremes could produce a taxonomy which is either very divisive
and unweildy with each population sample comprising a number of formally named taxa or one
which is grossly simplified to the extent of masking potentially useful affinities.

A fairly conservative approach is therefore adopted with different phenotypes within populations
being recognized by the neutral term ‘morphs’ (Mayr 1969, p. 46). In the case of T. ceriodes angelini
there is evidence for the progressive establishment of distinct phenotypes (see Hayami and Ozawa
1975 for a discussion of this process). In all the other Norwegian taxa only slight non-directional
temporal and geographical changes in the relative proportions of constituent morphs are seen. To
some extent the morphs are simply the product of variation exceeding the threshold values noted
above. Thus, for example, it is not surprising that T. ceriodes angelini morph D (see below), which has
arc I4 developed, commonly has more pits in I3 than does morph B. Moreover, I3 is always
continuous anteriorly in morph D but, by definition, is incomplete frontally in morph B.
Nevertheless, the recognition of morphs is found to be very useful in describing variation and in
making comparisons with named taxa from elsewhere which typologically resemble particular
portions of the Norwegian range of variation.

The Norwegian species and subspecies therefore are defined in terms of recurrent associations of
morphs (Table 1) and to some extent the relative percentages of these morphs. The boundaries
between formally named taxa in some instances are ones of convenience, allowing for maximum
stability of present usage within the new framework. Thus as Table 1 shows, the Norwegian taxa
T. latilimbus norvegicus sp. nov. and T. sortita broeggeri are distinguished by the relative abundance
of two morphs and the presence in the latter taxon of a third morph. This allows for formal expression
of the greater similarity of T. latilimbus norvegicus to the Swedish T. latilimbus latilimbus (in which
morph B is virtually absent) and of T. sortita broeggeri to the coeval T. sortita sortita from Scotland
which is composed almost entirely of morph C.

13

14

15

Taxon

Morph

Radii

E2

*2

% P.

%c.

%c.

% p.

%c.

%c.

%P

%c.

%c.

°/

Pr.-.i

Ant.

Post.

Ant

Post.

Ant.

A

34

1

c

96

8

48

T. ceriodes angelini

B

13

1

C

c

90

0

0

C

26

|

C

c

100

18

100

D

27

1

C

c

100

66

100

100

6

14

T. hisingeri sp. nov

2

Inc

•vine

T. seficornis

2

vine

c

T. oskerensis sp. nov

2

Inc

c

29

0

0

T. anderssoni

2

Inc

C

100

0

0

T. hadelandica

hade/andico

A

37

2

'■Jnc

c

100

0

61

B

56

2

;nc-C

c

100

100

23

C

7

2

Inc

c

ioo

67

100

100

o'

0

T. latilimbus norvegicus

A

59

2

c

100

100

100

100

10

29

subsp. nov

B

41

2

Inc

c

100

100

100

100

0

17

T. sortita broeggeri

A

5

2

c

100

100

100

100

0

0

B

58

2

Inc

c

100

100

100

100

0

75

C

37

2

Inc

c

100

100

100

100

0

93

100

ncT

T. kiaeri

A

65

2

c

c

100

100

100

100

3

100

B

35

2

c

c

100

100

100

100

9

100

100

0

22

table 1 . The basic fringe development in the Norwegian species and
subspecies of Tretaspis. Where more than one morph is recognized, the
pit distribution in each morph is given. Note that the range in pit
number in each arc also serves to differentiate between the morphs and
this is detailed in text-figs. 2-9. The shading marks the complete
absence of an arc, Inc = incomplete, C = complete, P = present.
Post. = posteriorly, Ant. = anteriorly

718

PALAEONTOLOGY, VOLUME 23

The revision of the Norwegian Tretaspis material has entailed a reassessment of some of the well-
documented British forms. In addition, Dr. J. K. Ingham of Glasgow University has given me access
to his data on some of the Swedish forms. Most of the numerous citations in the literature of Swedish
and other European material are based on very limited collections, as is the North American
T. clarkei. The polymorphic Norwegian material indicates that individual specimens with a
particular fringe morphology could belong to one of a number of taxa. Large samples are required to
determine the range of variation and presence of morphs before taxonomic assignment can be carried
out with any confidence. Thus whilst the known material of Tretaspis from outside Norway and
Britain is discussed, it would be premature to make more than general comments on its affinity.

SYSTEMATIC PALAEONTOLOGY

Family trinucleidae Hawle and Corda, 1847
Subfamily trinucleinae Hawle and Corda, 1847
Genus tretaspis McCoy, 1849

Type species. Asaphus seticornis Hisinger, 1 840, p. 3, pi. 37, fig. 2; from the Fjacka Shale (early Ashgill), Dalarna,
Sweden; by subsequent designation of Bassler (1915, p. 1285).

Discussion. Ingham (1970, pp. 41-45) divided Tretaspis into three species groups centred on
T. moeldenensis Cave, T. seticornis (Hisinger), and ‘7V granulata (Wahlenberg). Hughes et al. (1975,
pp. 503-505) reassigned the species constituting the last-mentioned group to Nankinolithus Lu and
slightly revised the other two groups.

The T. seticornis group was originally stated to be characterized by an incomplete or absent E2, the
1 1 — E j_2 radii ‘out of phase’ with those containing the remaining I arcs, the number of pits in Ex
ranging from 1 6 to 23, rarely up to 27 (half-fringe), the thoracic rachial rings relatively broad (tr.) and
bearing a median tubercle and the pygidium never having more than six pairs of apodemes.
Populations described below as T. hadelandica hadelandica include specimens with E2 complete and
up to ten pairs of pygidial apodemes. Similarly, populations of T. anderssoni have seven pairs of
apodemes. In all other respects these forms correspond to the T. seticornis group. T. persulcatus from
the Upper Drummuck Group at Girvan, south-west Scotland, has a complete E2 but otherwise
corresponds to the T. seticornis group and was almost certainly derived from an unnamed form which
has E2 incomplete (see discussion of T. hadelandica below). Thus the extent of E2 and the number of
pygidial apodemes are not, per se, indicative of the T. seticornis group.

Ingham (1970, pp. 44-45) had difficulty in assigning T. kiaeri Stormer to his groups but Hughes et
al. (1975, p. 563) assigned it to the T. moeldenensis group. T. kiaeri is redescribed here and has E2
complete frontally, two sets of radii, up to 27^pits in Ex and up to ten pairs of pygidial apodemes. It is
therefore intermediate between the T. seticornis and the T. moeldenensis groups. T. kiaeri and its
probable derivative T. calcaria Dean resemble a number of middle Ordovician species from North
America: T. canadensis Staiible, T. reticulata Ruedemann, and T. sagenosus Whittington and broadly
coeval allied species from Scotland and Ireland. (Hughes et al. 1975, pp. 564-565). These middle
Ordovician forms are older than all other known species of Tretaspis and have a single set of radii, a
large number of pit arcs and in most cases a high pit count in most arcs. They were assigned to the T.
moeldenensis group by Hughes et al. (1975, pp. 563-564). Specimens from the low Carodoc of
Belgium assigned to Tretaspis by Hughes et al. (1975, p. 564) belong to Nankinolithus (= N. sp. of
Hughes et al. 1975, p. 559).

The T. seticornis group as presently defined seems to be a natural grouping derived in the earliest
Ashgill from T. ceriodes (Angelin), a member of the T. moeldenensis group. The removal of T. kiaeri,
T. calcaria, and the middle Ordovician species listed above would leave the T. moeldenensis group as a
close grouping within which phylogenetic relationships are fairly clear. The American province forms
are poorly known but probably closely related and are here termed the T. sagenosus group. They
almost certainly gave rise to T. ceriodes, the earliest known member of the revised T. moeldenensis
group possibly by neoteny (giving a much simplified fringe morphology) and at a time of major

OWEN: TRILOBITE TRETASPIS

719

immigration into the Scandinavian area (Bruton and Owen 1979). T. kiaeri and T. calcaria have the
typical large number of arcs and high pit counts of the T. sagenosus group and whilst having two sets
of radii developed the possibility exists that they are more closely related to that group than to the
other two groups and thus are provisionally included in it.

Tretaspis moeldenensis group
Tretaspis ceriodes ( Angelin, 1854) angelini Stormer, 1930
Plate 89, figs. 1-12; text-fig. 2

1887 Trinucleus; Brogger, p. 23.

1930 Tretaspis cerioides [sic] (Angelin); Stormer, pp. 44-48, pi. 9, figs. 1-4; text-fig. 21 b.

1930 Tretaspis cerioides var. angelini Stormer, pp. 48-50, pi. 9, figs. 5-10.

1934 Tretaspis cerioides; Stormer, p. 331.

1945 Tretaspis ceriodes (Angelin); Stormer (pars), p. 402, pi. 1 , fig. 6; non pp. 387, 404-405, pi. 1 , fig. 7; pi.

4, fig. 16 (= T. hadelandica hadelandica).

1945 Tretaspis ceriodes var. angelini Stormer; Stormer, p. 402, pi. 1, fig. 5.

1945 Tretaspis ceriodes var. donsi Stormer, pp. 388, 402, 405, pi. 1, fig. 8.

1953 Tretaspis ceriodes; Stormer, pp. 68, 87, 94.

1953 T. c. angelini; Stormer, p. 68.

1973 Tretaspis cerioides; Lauritzen, p. 29.

1978 Tretaspis ceriodes (sensu lato) (Angelin); Owen, pp. 9, 14, 15.

1979 Tretaspis ceriodes; Owen, pp. 250, 251.

1979 Tretaspis ceriodes (Angelin) ( sensu lato); Bruton and Owen, text-figs. 3-6.

Holotype. A cranidium (PMO H226) from 2 m below the top of the Upper Chasmops Limestone on
Terneholmen, Asker.

Material, localities, and horizons. The subspecies has a short stratigraphical range and, although no complete
specimens are known, a large number of disarticulated skeletal elements are known from the uppermost parts of
the Upper Chasmops Limestone in Baerum and Asker in the western part of Oslo- Asker (see Bruton and Owen
1979 for detailed information), from 0-85-1 02 m above the base of the Lower Tretaspis Shale on Nakholmen,
Oslo, from the uppermost parts of the Solvang Formation throughout Hadeland and at Norderhov in
Ringerike, and from the lowest part of the Gagnum Shale Member of the Lunner Formation in the northern part
of Hadeland.

Description. Sagittal length of glabella equal to 50-60% of width between posterior fossulae. Occipital ring
arched gently upwards and rearwards and defined anteriorly by a shallow furrow which bears deep slot-like pits
laterally. Occiput short (sag., exsag.), very weakly swollen. Ip furrows deep, transversely oval. 2p furrows large,
deep, situated a very short distance in front of lp furrows and diverging forwards at approximately 90°.
Composite lateral glabellar lobes very narrow (tr.) adjacent to 2p furrows, anteriorly and posteriorly to which
they are very weakly developed. 3p furrows developed as very shallow depressions on the pseudofrontal lobe
directly in front of the mid-length of the glabella. Pseudofrontal lobe very strongly swollen, almost circular in
dorsal view, occupying approximately 70% of the sagittal glabellar length. Median node situated on the highest
part of the glabella at 60% of the sagittal glabellar length. Dorsal furrows broad (tr.) and shallow posteriorly,
narrowing and deepening a little frontally, diverging forwards at approximately 30° to a level a short distance in
front of the 2p furrows, anteriorly to which they are gently convex abaxially and bear deep fossulae frontally.
Genal lobes quadrant-shaped, gently inclined from the dorsal furrows, more steeply declined towards the
fringe. Lateral eye tubercles situated opposite or slightly in front of 2p furrows. Low but distinct eye ridges
converge adaxially forwards at about 145° from the eyes to the outer parts of the dorsal furrows. Posterior
border furrows deeply incised, transversely directed, bearing deep fossulae distally. Posterior borders ridge-like,
transversely directed to behind posterior fossulae abaxially to which they are deflected steeply downwards and
rearwards at approximately 60°. External surface of glabella and genal lobes bears a variable but usually strong
reticulation which is coarsest around the glabellar node and lateral eye tubercles. On internal moulds the glabella
is commonly smooth and the genal lobes bear a very subdued reticulation. Fringe flat-lying over the inner one or
two I arcs anteriorly and anterolaterally, otherwise almost vertical.

All specimens have arcs E1? I x , I2, and In complete but there is considerable variation in the development of
arcs E2, 13, and I4. On the basis of these arcs, four morphs are recognized (Table 1). Morph A lacks E2 and I4 and

720

PALAEONTOLOGY, VOLUME 23

has I3 continuous in front of the glabella in 48% of 3 1 specimens where this could be determined and extending to
the posterior margin in 8% of 24 specimens. I3 is absent in 4% of 24 specimens. Morph B has a complete E2, 13
developed in 90% of 20 specimens but never continuous anteriorly or posteriorly and I4 is absent. Morph C has a
complete E2, no I4, and an I3 arc which is always continuous anteriorly and extends to the posterior margin in
18% of 36 specimens. Morph D has E2 complete, I3 invariably complete anteriorly and complete posteriorly in
66% of 25 specimens and a short I4 developed. The range of variation in arcs Ex, I3, In, and the number of pits
along the posterior margin of the fringe for T. ceriodes angelini as a whole and in the constituent morphs, is given
on text-fig. 2 along with data on the development of I4 in morph D. With the exception of the number of pits in I3,
these ranges are very similar for all the morphs although the mean values for morph A are lower than those of the
other morphs. Arcs I1; Ex, and E2 (when present) commonly share sulci on the anterior and lateral parts of the
fringe in most specimens. Although the extent of this feature was recorded wherever possible, there is often some
difficulty in assessing the precise extent of the sulcation which may also be partially dependent on preservation
and consequently this is not presented in histogram form. In a few specimens the sulcation does not extend
laterally beyond the dorsal furrows and in a few it extends almost to the posterior margin. The mean extent is to
about bR9 (fifty specimens, standard deviation 4) and there is no apparent difference between the morphs. Only
one set of radii is developed. On external surfaces, lists are developed between all the I arcs. Genal spines parallel,
length unknown.

Hypostoma and thorax not known.

Pygidium sub-semicircular in outline with sagittal length equal to approximately 35% of the anterior width.
Rachis occupies 25% of the anterior width of the pygidium, tapers rearwards at about 30°, and is composed of an
anterior articulating half-ring and five or six rings. Ring furrows progressively less well-defined rearwards along
the rachis, bearing deep apodemal pits a short distance in from the weakly incised dorsal furrows. Pleural lobes
flat-lying, bearing four pairs of very broad furrows which define three or four ribs which die out some distance
from the weakly developed marginal rim. Pygidial border very steeply declined, broad, maintaining constant
width. Antero-lateral corners of pygidium bear steeply declined facets which diverge abaxially backwards at
about 120°.

Discussion. The absence of arc E2 from morph A clearly distinguishes it from the other morphs where
this arc is not only present, but complete. Morphs B, C, and D could be viewed as representing a
single morphological type with a broad range of variation. However, three morphs are recognized
because two, B and C, are similar to, or correspond to, the holotypes of named taxa, and there is also
some evidence for a progressive development of levels of phenotype organization from morph A
through B and C to D. In Hadeland, a sample of thirty-three specimens from an exposure of the
Lieker Member of the Solvang Formation illustrated by Owen (1978, text-fig. 6) from a level near the
first appearance of the species has the following morph composition: A88%, B6%, and C6%. Higher
levels in the formation in the nearby stratotype section have yielded morph D, and ten specimens
from broadly equivalent levels in the Gagnum Shale (including the holotype of T. ceriodes donsi )
comprise B10%, C80%, and D10%. Similar results have been obtained from Oslo-Asker with early
populations having morph A dominant over B and C; morph D being restricted to the later
populations where A is rare or absent.

It can be argued, therefore, that morph A represents the primitive condition, the development
of a complete E2 arc in some members of the population giving morphs B and C and individuals of
morph D type developed from morph C parents. It must be stressed, however, that the morphs
are regarded as representing fairly broad portions of the range of variation in interbreeding
populations.

Angelin’s original material of T. ceriodes (1854, p. 65, pi. 34, fig. 2-2 b) from the Upper Mossen
Formation (late Caradoc) at Kinnekulle, Vastergotland, Sweden, was reported by Stormer (1930, p.
45) to be lost and a neotype from the Solvang Formation in Ringerike was chosen. This neotype could
not have any standing as it was not from the type locality and recently Angelin’s probable syntypes
have come to light in the collections of the Riksmuseum, Stockholm. A full examination of the E pit
development can be made in only one of these and E2 is not developed. Two specimens show the
development of I3 which in both cases is short (3-4 pits) and not present anteriorly. I4 is absent. Thus
these probable syntypes resemble T. ceriodes angelini morph A. Two other specimens in the
Riksmuseum collections from the Upper Mossen Formation (locality not known) show an extensive
I3 development and while one lacks E2, the other has it developed mesially but not beyond R4. This

OWEN: TRILOBITE TRETASPIS

D n = l8
C n = 21
B n = l3

A I3 complete anteriorly n=8
A I3 incomplete anteriorly n=9
A all specimens n=20

A 13 incomp. ant. 1 — M ^ n = 10
A all specimens 1 — mm M n=25

text-fig. 2. Histograms showing the range of variation in fringe characters of all
available specimens of Tretaspis ceriodes angelini with a comparison of the
range, mean, and one sample standard deviation on each side of the mean of the
four morphs (A, B, C, and D) present in the subspecies. Morph A is also
subdivided to compare these parameters in specimens with arc I3 incomplete
anteriorly (i.e. like morph B) with those in which this arc is complete anteriorly
(i.e. like morphs C and D). It may prove useful to define separate morphs on this
basis once more material is available. In all instances n = number of specimens
in the sample.

722

PALAEONTOLOGY, VOLUME 23

condition is not known from any Norwegian specimen. Detailed comparisons of the Swedish and
Norwegian forms must await the documentation of more material from Kinnekulle.

T. ceriodes alyta Ingham, 1970, from the upper part of the Onnian Stage in northern England has
arcs E2 complete, I4 absent, and I3 extensive or complete posteriorly but incomplete anteriorly. It
thus resembles T. ceriodes angelini morph B, differing only in having a more extensive I3 arc and the
I1-E1_2 sulci commonly extending almost to the genal angles. Examination of specimens from the
Onnian Stage in the Gross Fell Inlier in northern England figured by Dean (1961, 1962) shows that
Ingham was correct in suggesting that they belong to T. ceriodes alyta (1970, p. 5). Some of the
specimens of supposed Onnian age in Dean’s collections in the Cross Fell Inlier (localities A12 and
A1 5 of Dean, 1959, text-fig. 1) have a very large number of pits in In (25^-28^) and up to 9\ pits in I4
and most closely resemble T. moeldenensis Cave, 1960 (see Price 1977, pp. 764-772 for a discussion of
this species).

T. ceriodes favus Dean, 1963, is a poorly known form based on specimens from the upper part of
the Actonian Stage and the lowest beds of the Onnian Stage in the Onny River section and supposed
Actonian strata near Cardington, Salop, England. The subspecies was diagnosed as having arc E2
developed only laterally and I3 complete anteriorly but not posteriorly. Whilst the latter is true for the
holotype and other specimens from the Onny River, the material is too poorly preserved for the E pit
development to be discerned fully although E2 is certainly present. The I3 development is closest to
that seen in T. ceriodes angelini morph C. All of the sixteen specimens from near Cardington in the
British Museum (Natural History) (including Dean collection) and the Hunterian Museum (Owen
and Ingham collection) in which the E arc development is clear, undoubtedly have E2 complete. I3 is
incomplete anteriorly in this material (eleven specimens) and has 2-14 pits. Arcs \x-YLl_2 are sulcate
over almost the whole fringe. The Cardington material therefore is similar to both T. ceriodes angelini
morph B and T. ceriodes alyta and, as noted by Bruton and Owen (1979, p. 220), its association with
Onnia gracilis may indicate an Onnian age for the strata here.

T. ceryx Lamont, 1941, from the Raheen Shales (late Caradoc-early Ashgill) of Co. Waterford,
Eire, differs from T. ceriodes angelini morph C only in having very long, slot-like I1-E1_2 sulci
anteriorly and anterolaterally. The Irish form is probably best viewed as a geographical subspecies of
T. ceriodes.

T. colliquia Ingham, 1970, from the Pusgillian Stage in the Cautley district of northern England is
probably a derivative of T. ceriodes alyta and some specimens, like T. ceriodes angelini morph D have
a short I4 developed. The English species is distinguished by its very large, deep, extensive I1-E1_2
sulci and in having a very high E pit count (twenty-eight in the two specimens showing this feature).

Dr. J. K. Ingham of Glasgow University has informed me of an undescribed form of T. ceriodes
similar to T. ceriodes angelini morph A from the Upper Whitehouse Group (late Caradoc-early
Ashgill) at Girvan, south-west Scotland (Ingham 1978, pp. 170, 171).

EXPLANATION OF PLATE 89

Figs. 1-12. Tretaspis ceriodes (Angelin) angelini Stormer. 1, 3, 5, morph D, PM0100826, dorsal, anterior, and
lateral views of internal mould of cranidium, 5-3-54 m below top of Solvang Formation, Norderhov,
Ringerike, x 4. 2, 4, morph D, PMO101552, dorsal and anterolateral views of external surface of cephalon,
approximately 1-7 m below top of Upper Chasmops Limestone, East Raudskjer, Asker, x 6. 6, holotype,
morph C, PMO H226, oblique anterolateral view of partially exfoliated cranidium, 2 m below top of Upper
Chasmops Limestone, Terneholmen, Asker, x 6^; also figured by Stormer (1930, pi. 9, fig. 5). 7, 10, morph A,
PMO H593, posterolateral and frontal views of partially exfoliated cephalon, same horizon and locality as 6,
x 5; also figured by Stormer (1930, pi. 9, fig. 10). 8, morph B, PMO H250, anterolateral view of partially
exfoliated cephalon, same horizon and locality as 6, x 10. 9,PMO103952,dorsalviewofpygidium,upperpart
of Solvang Formation, Lunner, Hadeland, x 4J. 11, morph C, PM081 100, anterolateral view of partially
exfoliated small cranidium, same horizon and locality as 2, x 20. 1 2, morph A, PMO H495, anterolateral view
of partially exfoliated cephalon, 0-85-1-02 m above base of Lower Tretaspis Shale, Nakholmen, Oslo, x 4.

PLATE 89

B I

v m

OWEN, trilobite Tretaspis

724

PALAEONTOLOGY, VOLUME 23

Tretaspis seticomis group
Tretaspis seticornis (Hisinger, 1840)

Plate 90, figs. 1 -4

1840 Asaphus seticomis Hisinger, p. 3, pi. 37, fig. 2.

1840 Asaphus cyllarus Hisinger, p. 3, pi. 37, fig. 3.

71845 Trinucleus seticomis (Hisinger); Loven, p. 107, pi. 2, fig. 2.

71854 Trinucleus seticornis (Hisinger); Angelin, p. 84, pi. 40, fig. 19.

71869 Trinucleus seticornis (Hisinger); Linnarsson, p. 79.

1883 Trinucleus seticornis (Hisinger); Tornquist, p. 43.

71884 Trinucleus seticornis (Hisinger); Tornquist, pp. 84-87.

71887 Trinucleus seticornis (Hisinger); Brogger, p. 24.

1930 Tretaspis seticornis (Hisinger); Stormer (pars), pp. 55-67, ?pl. 7; ?pl. 8; ?pl. 1 1, fig. 4; text-figs. 27, 28
(pars), 729, 33a, 346 (pars), 34c, 736, ?37a, b, 742.

1934 Tretaspis seticornis', Stormer (pars), p. 330.

1936 Tretaspis seticornis (Hisinger); Asklund (pars), p. 4, pi. 1, figs. 1-3, 75, 76, non 4.

71959 Tretaspis seticornis (Hisinger); Whittington in Moore, text-fig. 323.2.

1979 Tretaspis seticornis seticornis (Hisinger); Owen, pp. 250, 251, 252, text-fig. 6.

1979 Tretaspis seticomis seticornis (Hisinger); Bruton and Owen, text-fig. 6.

This synonomy only includes references to material which actually, or very probably, belongs to T. seticornis. A
more complete list, comprising forty-seven entries, was given by the writer (1977, pp. 243-245) in an unpublished
thesis and includes reidentifications wherever possible.

Material, localities, and horizons. Hisinger’s syntypes of Asaphus seticornis from the Fjacka Shale in well diggings
at Furudal in Dalarna, Sweden, have not been identified unequivocally in the collections of the Riksmuseum,
Stockholm, and as noted by Tornquist (1883, p. 43) may not have been collected in situ. The species, as here
defined, is known from the lower part of the Fjacka Shale (J. K. Ingham, pers. comm. 1 976), the lower part of the
Lower Tretaspis Shale at Ole Deviks Vei (lowest 5-86 m), Astaddammen (lowest 4-65 m at least), S. Grakommen
and between Fossung and Hogstad in Oslo-Asker, and from the Hogberg Member of the Solvang Formation on
Frognoya, Ringerike.

Description. Most of the available material is crushed to some extent. Glabella and genal lobes similar to those of
T. ceriodes angelini except that the pseudofrontal lobe is more elongate. External surface of glabella and genal
lobes smooth or bearing a faint reticulation. Internal moulds smooth. Steeply declined fringe bears complete arcs
Els Ix, I2, and In, and an incomplete E2 arc. Arcs Ix -Ex_2 are out of phase with radii comprising the other two I
arcs. Pits in Ix and Ex share sulci anteriorly and anterolaterally. There is insufficient material to assess the range
of variation in pit distribution. Only one Norwegian specimen is sufficiently well preserved for the number in Ex
to be determined (18), and whilst one specimen clearly lacks E2, others show minimum values of 4, 7, 9 (3
specimens), and 1 0 pits. In is seen completely in 4 specimens where it comprises 1 5, 17), 18, and 1 8) pits and there
are 6 (3 specimens) or 7 (3 specimens) pits along the posterior margin of the fringe. Lists are not developed. One
specimen (pi. 90, fig. 4) does not conform to the typical T. seticornis development in having a stronger
reticulation and in having pits developed in I3 on the lateral parts of the fringe at aR6, 7, 9, 11-17. Such a
development is most unusual for any species of Tretaspis and may reflect hybridization with T. hadelandica
hadelandica which includes morphs with this arc complete posteriorly.

Hypostoma unknown.

Thorax barrel-shaped, comprising six segments of which the third and fourth are slightly broader (tr.) than the
rest. Rachis occupies 30% of the width of each segment and is bounded laterally by very weakly incised dorsal
furrows. Rachial rings strongly convex in transverse view and each bears a small median tubercle on its anterior
edge and is separated from its articulating half-ring by a transversely directed furrow which bears deep apodemal
pits laterally. Pleurae parallel-sided proximally, tapering slightly over the distal 25% where they are deflected
gently downwards and rearwards. Pleural furrows shallow, each directed transversely and broadening (exsag.)
from near the anteromesial corner of the pleura such that the posterior band tapers abaxially and the anterior
band expands a little.

Pygidium broadly similar to that of T. ceriodes angelini. Rachis composed of six, possibly seven rings and the
pleural lobes bear up to three poorly defined ribs.

OWEN: TRILOBITE TRETASPIS

725

Discussion. Hisinger (1840) described two species of Tretaspis, ‘ Asaphus ’ seticornis and ‘A’. cyllarus,
from the Fjacka Shale. His illustrations of both show the development of four complete arcs of pits
and there is a well-developed list between the inner and the outer pairs of arcs on his drawings of T.
seticornis. Dr. J. K. Ingham informs me (pers. comm. 1976) that in the probable syntypes of both
species and all other available specimens from the Fjacka Shale at Furudal which have the fringe
preserved, arcs Ex, Ix, I2, and In are complete and a short E2 is developed posteriorly. Thus it seems
reasonable to assume that this is indeed the case with the syntypes and, in order to stabilize the
species, it is advocated that this be assumed to be the case. Dr. Ingham has examined material from
Dalarna described by Angelin (1854) as T. seticornis and considers that this identification probably is
correct. Angelin’s originals of T. a ffinis have a complete I3 arc developed and thus are excluded from
T. seticornis.

Stormer (1930) assigned a large number of specimens to T. seticornis from the Fjacka Shale and
various horizons in Norway. Many of these are reassigned herein to T. anderssoni Stormer and T.
hisingeri sp. nov. It is clear that at least three forms are present in the Fjacka Shale and so references to
T. seticornis in this unit by Linnarsson (1869) and Tornquist (1884) are only tentatively included in
the above synonymy. Further discussion of material previously assigned to T. seticornis is given
below in the discussions of T. anderssoni and T. hadelandica.

Tretaspis anderssoni Stormer, 1945
Plate 90, figs. 5-10; text-fig. 3

?«oh1894 Trinucleus seticornis ( Hisinger); Andersson, p. 532, figs. 1-5.

1930 Tretaspis seticornis (Hisinger); Stormer (pars), pi. 1 1, figs. 2, 5; pi. 12, figs. 1-5; pi. 13, figs. 1,2,
5-7; ?pl. 14, figs. 4, 5; text-figs. 33 b, c (pars), d, 11c.

71936 Tretaspis seticornis (Hisinger); Asklund (pars), p. 4, pi. 1, fig. 4.

1945 Tretaspis seticornis (Hisinger) var. anderssoni Stormer, p. 401, pi. 1, fig. 2.

1959 Tretaspis seticornis (Hisinger); Harrington in Moore, text-figs. 52, 67.

«o«1965 T. seticornis anderssoni Stormer; Cave, p. 296 [? = T. hadelandica brachystichus Ingham].

1975 Tretaspis seticornis anderssoni Stormer; Hughes et al., p. 563, pi. 4, figs. 52, 53.

1976 Tretaspis seticornis (Hisinger); Miller, text-fig. 2 h.

1979 Tretaspis seticornis anderssoni Stormer; Owen p. 253 text-fig. 8.

71979 [specimens resembling] T. hadelandica Stormer; Owen, p. 253.

Holotype. A cranidium (PM065196) from the Frognoya Shale, on Frognoya, Ringerike.

Material, localities, and horizons. Specimens from low in the Frognoya Shale tentatively compared with T.
hadelandica by Owen (1979) probably belong in T. anderssoni in which case cephala, cranidia, lower lamellae,
and pygidia are known from throughout the type unit on Frognoya and from the overlying Sorbakken
Limestone (except the lowest 9 m and the uppermost 17 m) on Frognoya and at Norderhov, Ringerike. Two
poorly preserved cranidia from the Venstop Shale in Skien-Langesund may belong here also.

Description. Cephalic proportions similar to those of T. ceriodes angelini. The fine structure of the median
glabellar tubercle in T. anderssoni was described by Stormer (1930, p. 87, text-fig. 37c; pi. 11, fig. 5; pi. 13,
figs. 5-7) who noted that it bears four small pits arranged as at the corners of a square and a slightly larger
central pit which may bear a fine canal opening. Stormer (1930, pi. 12, fig. 3; pi. 13, figs. 1, 2) also illustrated
a lenticular body within the exoskeleton of the lateral tubercles of this species. On the external surface of the
glabella and genal lobes there is a weakly developed fine reticulation which is seen faintly on a few internal
moulds.

Fringe narrow, very steeply declined except laterally where a narrow brim is developed. A gentle anterior arch
is present. The details of fringe pitting are given on text-fig. 3. Two distinct sets of radii are present, arcs Ex, Ix, I2,
and In are complete and in all specimens a short E2 arc is developed posteriorly and I3 is developed
anterolaterally but never complete mesially. Arcs Ix-E) 2 share sulci which extend to between bR5 and bR14.
The limited evidence available suggests that there is no significant difference in pit development between early
and late populations of T. anderssoni.

Hypostoma and thorax unknown.

726

PALAEONTOLOGY, VOLUME 23

Pygidium similar to that of T. ceriodes angelini. Six rachial ring furrows, each with deep apodemal pits
laterally, are seen on the external surface of the rachis. On internal moulds, a seventh pair of apodemal pits lies
directly in front of the pygidial border. Pleural lobes bear four weakly developed pairs of ribs, the posterior two
barely discernible.

SLI

FS

n=8

n=8

SLu

SLI

FS

1=18

i=5

n=ll

SLuNZMn=20
SLI MZMn=l4
FS MZ^n=20

45-

40-

35.

15 16 17 18 19 20

Pits in In

15.

10.

5.

Pits along Pits in I3 Pits Missing

Post Margin Anteriorly from I3

text-fig. 3. Histograms showing the range of variation in fringe features of all
available specimens of Tretaspis anderssoni with a comparison of samples from
the Frognoya Shale (FS), from 9-28 m above the base of the overlying
Sorbakken Limestone (SLI) and from 17 m below the top of this unit (SLu).

EXPLANATION OF PLATE 90

Figs. 1-4. Tretaspis seticornis (Hisinger). 1, PMO 103953, dorsal view of internal mould of almost complete
specimen, 4-65 m above base of Lower Tretaspis Shale, Astaddammen, Asker, x 2\. 2, PMO101553, ventral
view of part of cranidium, Hogberg Member of the Solvang Formation, Frognoya, Ringerike, x4. 3,
PMO 103954, dorsal view of cast of almost complete specimen, Lower Tretaspis Shale, Ole Deviks Vei, Oslo,
x 3. 4, PMO103955, anterolateral view of cast of cranidium showing I3 developed laterally, 1-65 m above
base of Lower Tretaspis Shale, same locality as 1, x 4J.

Figs. 5-10. Tretaspis anderssoni Stormer. 5,PMO103956, dorsal view of internal mould of pygidium, 17mbelow
top of Sorbakken Limestone, Frognoya, Ringerike, x 6,8,9, holotype, PM065 1 96, dorsal, anterior, and
lateral views of internal mould of cephalon, Frognoya Shale, Frognoya, Ringerike, x 3; also figured by
Stormer (1945, pi. 1, fig. 2) and Hughes et al. (1975, pi. 4, figs. 52, 53). 7, PMO H103, posterolateral view of
cephalon, same horizon and locality as 6, x 2 \\ also figured by Stormer (1930, pi. 11, fig. 5). 10, PM080670,
frontal view of cast of cranidium, Venstop Shale, Friefjord, Skien-Langesund, x 1\.

Figs. 11-14. Tretaspis hisingeri sp. nov. 11, PMO H71, frontal view of partially exfoliated cephalon, 30-4-5 m
below top of Frognoya Shale, same locality as 6, x 3£. 12-14, PMO H75, dorsal, lateral, and frontal views of
partially exfoliated cranidium, Frognoya Shale, same locality as 6, x 3, x 3|, x 3£; also figured by Stormer
(1930, pi. 11, fig. 3; 1945, text-fig. 4).

PLATE 90

OWEN, trilobite Tretaspis

728

PALAEONTOLOGY, VOLUME 23

Discussion. T. anderssoni differs from its probable ancestor, T. seticornis in having a short I3
developed in all specimens. A broadly similar fringe development is seen in a number of described
taxa and their interrelationships are discussed below under T. hadelandica.

Stormer (1945, p. 40 1 ) considered that specimens figured by Andersson (1894) as T. seticornis from
the Lower Johnstorp Formation (Pusgillian-?Cautleyan) of Hulderstad, Oland, Sweden, probably
belong to T. anderssoni. Examination of these specimens reveals that they have pit counts at the upper
end of, or even beyond, the range of variation seen in T. anderssoni from Norway. The counts in these
Riksmuseum, Stockholm, specimens Ar21551 and Ar21553 respectively are as follows: Ej 22, 20; E2
9 (710), 78; In c. \1\, 18; I3 5, 5. Without further specimens from Oland the affinities of this material
must remain in doubt. Similarly, a specimen figured by Asklund (1936) from the Tretaspis Beds in
Jemtland has a short I3 but its affinities must await the description of further specimens.

Tretaspis hisingeri sp. nov.

Plate 90, figs. 11-14; Plate 91, figs. 1-4; text-fig. 4

71887 Trinucleus seticornis (Hisinger); Brogger, p. 24.

1 930 Tretaspis seticornis (Hisinger); Stormer {pars), pi. 11, figs. 1 , 3, 6, 7; text-figs. 33c (pars). Til a, b, 40,
47.

1934 Tretaspis seticornis ; Stormer (pars), p. 330.

1945 Tretaspis seticornis (Hisinger) forma typica; Stormer, p. 401, text-fig. 4.

1970 T. sp. [?nov.]; Ingham, p. 41.

1975 T. sp. ?nov.; Hughes et ai, p. 563.

1979 Tretaspis sp. nov.; Owen, p. 253, text-fig. 8.

Holotype. An almost complete specimen (PMO H51) from 3-5-4-0 m below the top of the Frognoya Shale on
Frognoya, Ringerike.

Material, localities, and horizons. The species has a limited stratigraphical distribution and is known from all but
the lowest part of the Frognoya Shale on Frognoya and at Hole and Norderhov, and also between 9 and 14 m
above the base of the overlying Sorbakken Limestone on Frognoya, Ringerike. The species is also known from
the upper part of the Lower Tretaspis Shale at Ole Deviks Vei and on Bygdoy and Lindoya in Oslo, the Tretaspis
Limestone at Nesbru, Asker, and the Venstop Shale in Skien-Langesund.

Diagnosis. Very narrow fringe has El5 Il5 and In complete, E2 short and a short I2 present in the vast majority of
specimens but rarely continuous anteriorly, and in some instances asymmetrically distributed about the sagittal
line. Two distinct sets of radii mesially and where I2 is developed but laterally In is in phase with L-E,^.

1^,,.

16 17 18

Pits in t

19 20

t| 10-

^T-

1

A

5 6 7
Pits along
Posterior Margin

text-fig. 4. Histograms showing the range of variation in all available specimens of
Tretaspis hisingeri sp. nov. In the case of I2, only specimens which are symmetrical
about the sagittal line or which have only one side of the fringe visible are included.
An additional five specimens are asymmetrical, and inclusion of the right or left
counts with the data shown here does not change the mean value although the left
counts increase the standard deviation to lj. The number of pits missing from I2
anteriorly from these specimens is the same for both right and left sides, and thus
are incorporated in the histogram of this feature.

OWEN: TRILOBITE TRETASPIS

729

Description. The glabella and genal lobes of T. hisingeri differ from those of T. ceriodes angelini only in having the
median node situated a little further forward, the lateral eye tubercles a little closer to the glabella, and in most of
the larger holaspids lacking any reticulation on the external surface of the exoskeleton. A specimen of meraspis
degree 4, however, has a very strong reticulation on both glabella and genae (pi. 91, fig. 4). Similar reduction in
the extent and intensity of reticulation with growth in trinucleids is well documented (Cech 1975). Genal spines
extending well beyond the pygidium. The fringe is very narrow with only El5 Ix, and In complete. A short E2 is
developed posteriorly and nearly all specimens have a few pits in I2 which is rarely continuous frontally (one
specimen out of twelve). In some of the specimens where the development of I2 can be seen on both sides of the
glabella there are up to two pits less on one side than on the other. In an extreme case the arc is absent on the left
side but contains two pits on the right (PI. 90, fig. 1 1). The range of variation in fringe features is illustrated on
text-fig. 4. Arcs In, I x , Ex , and (where present) E2 are arranged in a single set of radii laterally but I j and E l are out
of phase with In mesially and with the inner two I arcs where I2 is developed.

Hypostoma unknown.

Thorax of holaspis similar to that of T. seticornis. That of the meraspis degree 4 noted above has a narrower
rachis which occupies 25% (cf. 30%) of the segment width.

Holaspid pygidium known only from the holotype in which it is incomplete. Rachis bears at least 6 pairs of
apodemal pits. Meraspis degree 4 pygidium sub-semicircular in outline with a rachis of approximately 5 rings of
which only the anterior 2 are distinct.

Discussion. The short I2 development distinguishes T. hisingeri from all other named species. T.
hisingeri succeeds T. seticornis without overlap and probably was derived from it by neoteny. A very
similar form in which I2 is incomplete but more extensive than in T. hisingeri occurs in the Fjacka
Shale in Sweden, and it too succeeds T. seticornis (J. K. Ingham, pers. comm. 1976).

Tretaspis hadelandica hadelandica Stormer, 1945
Plate 91, figs. 5-14; Plate 92, figs. 1, 2; text-fig. 5

1923 Trinucleus sp.; Holtedahl in Holtedahl and Schetelig, p. 22.

1945 Tretaspis seticornis', Stormer, p. 384.

1945 Tretaspis seticornis var. hadelandica Stormer, pp. 384, 388, 406-407, pi. 1, figs. 3, 4.

1945 Tretaspis ceriodes (Angelin); Stormer, pp. 387, 404-405, pi. 4, fig. 16.

1945 Tretaspis kiaeri Stormer; Stormer, pp. 387, 406, pi. 1, fig. 11.

1970 Tretaspis hadelandica hadelandica Stormer; Ingham, text-fig. 17.

1973 Tretaspis seticornis', Lauritzen, p. 29.

1978 Tretaspis hadelandica hadelandica Stormer; Owen, pp. 11, 13, 14, 17.

Holotype. An incomplete cranidium (PM065187) probably from the Gagnum Limestone Formation south of
Gagnum, Hadeland.

Material, localities, and horizons. A few complete specimens and a large number of disarticulated skeletal
elements occur abundantly in the Gagnum Shale (except the lowest part in northern Hadeland) and Lunner
Kirke members of the Lunner Formation, the shales of this formation around Lunner, and in the Gagnum
Limestone and Kjorrven formations. The species is rare in the Grina Shale Member of the Lunner Formation.
Fragmentary museum material from Nittedal (precise horizon unclear), between Oslo and Hadeland, may
belong here also.

Description. Proportions of glabella and genal lobes very similar to those of T. ceriodes angelini. Specimens from
the Gagnum Shale have a well-developed reticulation on the external surface and commonly on the internal
mould, but most specimens from other units have only a subdued reticulation or are smooth. Fringe steeply
declined with a slight brim developed laterally. Genal spines long, diverging rearwards very slightly. All
specimens have two distinct sets of radii, and arcs E1; Il5 12, and In are complete. Three morphs are recognized on
the development of E2, 13, and I4 (Table 1) and the distribution of pits in each arc is shown on text-fig. 5. Arc I4 is
absent from morphs A and B which respectively have I3 incomplete and complete posteriorly. These morphs
occur in all samples, whereas morph C, which has a short I4 developed, is known only from a few populations
from the lower part of the Gagnum Shale Member, the upper part of the Lunner Formation around Lunner, and
from the Gagnum Limestone. In one specimen (PI. 92, figs. 1 , 2) I3 is complete on the right side of the cranidium
but not on the left, an asymmetry which encompasses both morph A and morph B. Samples are not large enough

730

PALAEONTOLOGY, VOLUME 23

B n=32 0,1 comP|e,e posteriorly

hh-l

text-fig. 5. Histograms showing the range of variation in fringe features of all
available specimens of Tretaspis hadelandica hadelandica with a comparison of the
range, mean, and sample standard deviation of the three morphs (A, B, and C)
present in the subspecies.

to enable detailed unit by unit comparison of the variation in each morph but no obvious stratigraphical changes
are apparent.

Hypostoma unknown. Thorax like that of T. seticornis.

Pygidium sub-semicircular in outline. Rachis crossed by 5-7 furrows each bearing apodemal pits laterally. On
the ventral surface of the pygidium there are up to ten pairs of apodemes, the posterior three of which are situated
on the steeply declined pygidial border. Pleural lobes bear three low ribs.

Discussion. When present, morph C occurs with morphs A and B which are always found together.
Their great similarity in pit distribution in arcs Ex, In, and along the posterior margin argues strongly
for these morphs being no more than broad phenotypes from the same gene pool. Their relative
abundance, however, may be ecologically controlled . T able 2 gives the relative percentages of morphs
present in the stratigraphical units in which they occur in measurable abundance.

3x2 and 2x2 contingency tests were carried out on the specimen numbers used to calculate these
percentages. The latter test was used where morph C was absent from both samples under
examination, or where expected frequencies of morph C were less than 5; Yates’s Correction was
applied in both instances. These tests show that the Gagnum Shale abundances are significantly
different from all but those of the Kjorrven Formation at the 0T% level. The Kjorrven Formation

OWEN: TRILOBITE TRETASPIS

73:

table 2. Percentages of each morph present in collections of T. hadelandica hadelandica from stratigraphical

units in Hadeland

Morph

Gagnum
Shale Member

Lunner Kirke
Member

Lunner Formation above
Lunner Kirke Member

Gagnum

Limestone

Kjorrven

Formation

A

55

17

18

13

47

B

31

83

75

77

53

C

14

0

7

10

0

Number 106

of specimens

35

44

39

19

abundances differ from those of the Gagnum Shale near the 50% level which is not significant, and
from those of the other three units at the 5% level which is considered significant. No significant
differences are present between the remaining three units where, in fact, there is a high degree of
correlation. The similarity between the Gagnum Shale and Kjorrven Formation abundances is the
product of high proportions of morph A in these units. It may be noteworthy that both units have a
much higher trilobite diversity (measured by the total number of known taxa) than the others, but
speculation on the reasons for this similarity in morph composition would be very unreliable in view
of the small sample size from the Kjorrven Formation.

T. hadelandica brachystichus Ingham, 1970, was based on samples from the Rawtheyan Stage
(Ashgill Zones 5 and 6) in the Cautley area of northern England which have I3 incomplete anteriorly
and posteriorly. Ingham also tentatively included fragments from the mid-Cautleyan Stage (Zone 3)
in this subspecies and suggested that specimens from the Gagnum Shale assigned to T. ceriodes by
Stormer may belong to the north of England form. These Gagnum Shale specimens are assigned to T.
hadelandica hadelandica morph A herein. Ingham’s material and specimens assigned to T.
hadelandica brachystichus by Price (1973, 1977) and Cocks and Price (1975) from the uppermost part
of the Sholeshook Limestone and lower part of the Slade and Redhill Mudstone (mid-Ashgill) in
south Wales, have a range of variation which overlaps that seen in the Norwegian morph A (text-fig.
6). In the case of arcs E2 and Ej and the number of pits along the posterior margin, the range and, in
the E arcs, the mean is higher than that of the Norwegian morph. The variation in number of pits in
I3, however, overlaps at the lower end of that seen in T. hadelandica hadelandica morph A and is
closer to that of T. anderssoni which, in all these characters, has a range of variation which overlaps
only at its upper end with that of T. hadelandica hadelandica morph A (text-fig. 6).

T. corner gens Dean, 1961 , was described originally from Pusgillian strata in the Cross Fell Inlier in
northern England and subsequently by Ingham (1970) from Pusgillian and lower Cautleyan (Ashgill
Zone 1) strata at Cautley and by McNamara (1979) from mid-Cautleyan (Zone 2 and lowest part of
Zone 3) strata in the English Lake District. As noted by McNamara ( 1 979, p. 62), the limited evidence
available suggests that there is a progressive reduction in I pits with the Cross Fell specimens having a
short (up to ten pits) I4 arc, some specimens lacking this arc in the Cautley material, and in all the
Lake District specimens this arc is not developed. This trend is continued in Ashgill Zone 3 in the
Lake District with T. convergens deliquus McNamara, 1979; I3 becoming incomplete anteriorly and
then laterally. The earliest examples with T. convergens deliquus morphology occur with specimens
with I3 complete anteriorly (K. J. McNamara, pers. comm. 1979).

As text-fig. 6 shows, the over-all pit distribution and the broad range of variation seen in the
subspecies of T. convergens is very similar to that of T. hadelandica hadelandica and consequently the
English forms are regarded as subspecies of T. hadelandica. It seems reasonable to suggest that
Ingham’s indeterminate specimens from Zone 3 are, in fact, T. hadelandica deliquus and that the
progressive decrease in pit number in I3 documented by McNamara continued, giving rise to T.
hadelandica brachystichus. Moreover, if the Zone 3 material from Cautley is indeed T. hadelandica
deliquus, the restriction of T. hadelandica brachystichus to Zones 5 and 6 (i.e. lower Rawtheyan) in
northern England would add weight to Ingham’s suggestion ( 1 977, p. 118) that the uppermost part of
the Sholeshook Limestone is early Rawtheyan in age. T. convergens has been recorded from lower

732 PALAEONTOLOGY, VOLUME 23

Ashgill strata at Girvan, south-west Scotland (Ingham 1970, p. 46), but the affinities of this material
are not known.

The succession of subspecies of T. hadelandica in northern England seems to represent a single
local stock and the above revision is based on this. An alternative, but more contrived hypothesis
would be the ecological replacement of T. hadelandica hadelandica morphs. Thus the Norwegian
morph C resembles early T. hadelandica convergens, morph B resembles late T. hadelandica
convergens, and early T. hadelandica deliquus and morph A resembles T. hadelandica brachystichus.

As far as morph A is concerned, the absence of E2 in some specimens, the high percentage of
individuals in which I3 is continuous frontally, and the fairly limited overlap in number of pits in I3
serves to distinguish it from T. hadelandica brachystichus. Chi-squared tests show that the ranges in I3
pits and pits missing anteriorly from this arc are distinct at the 0T% level, even when Ingham’s
samples from Zones 5 and 6 are considered together. There is only a limited amount of information
on T. hadelandica convergens and T. hadelandica hadelandica morph C from the Gagnum Shale
but this suggests that the English form commonly has more pits in E3 (19^-22^ cf. 18-1 9^) and
along the posterior margin (7-12 cf. 6-7), but fewer in E2 (6-11 cf. 10-13) and in all cases I3 is
complete whereas it is incomplete posteriorly in 33% (of nine specimens) of the Norwegian morph C.
The English form is also distinguished by its more swollen pseudofrontal lobe. Although later

n % Present n

T. hadelandica brachystichus S. Woles 23 100 23

T. hadelandica brachystichus Zone 6 19 100 19

T hadelandica brachystichus Zone 5 36 100 36

T hadelandica deliquus 1 1 100 1 1

T hadelandica convergens 7 100 7

T. hadelandica hadelandica Morph B 30 100 35

T hadelandica hadelandica totol 62 94 69

T. hadelandica hadelandica Morph A 18 82 22

T. anderssoni 38 100 76

%Complete

0

0

0

0

0

17

7

0

0

20 21 22 23

T. had. bra. S. Wales
T. had. bra Z 5 a 6
T. had. del.

T. had. conv.

T. had. had. B
T. had. had. total
T. had. had. A
T anderssoni \

VZ I 2 3

Pits Missing Anteriorly from I3
( where incomplete here ) % complete

post

T. had. bra. S. Wales 0

T. had. bra Z. 6 0

T. had. bra Z . 5 0

T. had. del. 1 1

T. had. conv. 100

T. had. had. B
T had. had total
T. had. had. A
T.t

text-fig. 6. Range, mean, and sample standard deviation of selected fringe characters of members of
the Tretaspis seticornis group in which I3 is incomplete in at least some individuals. Data for T.
hadelandica convergens, T. h. deliquus, and T. h. brachystichus based on histograms given by Ingham
(1970), McNamara (1979), and Price (1977).

OWEN: TRILOBITE TRETASPIS

733

populations of T. hadelandica convergens and early T. hadelandica deliquus lack I4, they differ from T.
hadelandica hadelandica morph B in always having I3 complete frontally. It seems most likely,
therefore, that T. hadelandica hadelandica with its broad range of variation (morphs A, B, and C) and
the British series of subspecies with, at any one level, a much narrower range of variation were at most
connected by a series of clines throughout much of the Ashgill.

T. clarkei Cooper (in Schuchert and Cooper, 1930) from Ashgill units in Quebec, Canada, has two
distinct sets of radii and thus belongs to the T. seticornis group and is not a synonym of T. ceriodes (cf.
Whittington 1941, p. 29; Lesperance 1968, p. 813; Bolton 1970, pp. 35-36). The holotype from
the Whitehead Formation and specimens figured by Bolton (1970, pi. 6, figs. 12, 15, 17, 19) from
the Vaureal Formation have I3 incomplete posteriorly at least. Of the three specimens from the
Whitehead Formation in the Hunterian Museum, two (HM A4319; 4320) have eight pits in I3 which
is incomplete anteriorly. A third specimen (HM A4321) has twelve pits in I3 which is complete
anteriorly and three pits in I4. It is not known whether the specimens are from the same horizon but
all fringe features fall within the range seen in T. hadelandica. Detailed sampling of T. clarkei
populations is needed before its affinities can be fully determined.

As is noted in the discussion of T. seticornis, specimens of Tretaspis with arc I3 developed are
known from the Fjacka Shale in Sweden. In addition to Angelin’s material of T. affinis , which has
this arc complete, other specimens in the Riksmuseum, Stockholm, have I3 incomplete but, in some
cases, extensive (J. K. Ingham, pers. comm. 1976). Dr. Ingham has also examined a specimen from the
Slandrom Limestone (probably early Pusgillian) in the Siljan district (Jaanusson and Martna 1948,
p. 187) which has a short I3 and a coarsely reticulate glabella and genal lobes. Dr. P. J. Brenchley of
Liverpool University has sent me a specimen resembling T. hadelandica hadelandica morph B from
the flank facies of the Boda Limestone (Ashgill) in the Siljan district and this is the only specimen of
Tretaspis known from these beds, and the genus is not known from the Boda Limestone itself. The
Swedish species of Tretaspis are being revised by Dr. Ingham who has taken well-localized samples
from the Fjacka Shale.

There is a great deal of other material of the T. seticornis group with an incomplete I3 and in need of
modern study. This includes specimens from Ashgill units in Poland ascribed to T. seticornis by
Kielan (1957, 1960) and Tomczyk (1962), and material from the Kraluv Dvur Formation (mid-
Ashgill) in Bohemia examined by the writer in the collections of the British Museum (Natural
History). Specimens from this latter unit were referred to T. seticornis by Havlicek and Vanek (1966)
and Pribyl and Vanek (1969). Ingham (1970, pp. 41, 49) noted that specimens which Lamont (1935,
1941) assigned to T. seticornis from the Lower Drummuck Group (Cautleyan) at Girvan has a short
I3 and was termed T. sp. by Hughes etal. (1975, p. 563). Price (1977, p. 786) noted a similarity between
an unnamed form from low in the Slade and Redhill Mudstones and this species. Dr. Ingham informs
me (pers. comm. 1976) that T. seticornis of Portlock (1843) and Fearnsides, Elies, and Smith (1907)
from the Killey Bridge Beds (low Cautleyan) in Pomeroy, Ireland, may well prove synonymous with
the broadly coeval Lower Drummuck Group form as both have a short I3, a very extensive E2, and
large lateral eye tubercles quite close to the glabella. Moreover, T. sp. probably gave rise to T.
persulcatus (Reed, 1935) from the Upper Drummuck Group (late Rawtheyan) in which E2 is
complete and the girder is indistinct posteriorly where an external pseudogirder is developed between
Ex and E2 (see Ingham, 1970, p. 44).

Schmidt (1894) assigned specimens to T. seticornis from the Lykholm Group (late Caradoc to
Ashgill) in Estonia, and Jaanusson (1956, pp. 379, 383) listed the species from the lower part of the
group, the Nabala Formation (late Caradoc). It is not known whether the material referred to by
Jaanusson is from the same beds as Schmidt’s specimens, one of which (1894, pi. 5, fig. 22) is
illustrated as having I3 complete posteriorly, but it is not clear whether two sets of radii are
developed. Assuming that existing correlations are correct, the Estonian specimens listed by
Jaanusson would prove the oldest record of the T. seticornis group should they prove correctly
ascribed to it.

734

PALAEONTOLOGY, VOLUME 23

Tretaspis latilimbus (Linnarsson, 1869) norvegicus subsp. nov.

Plate 92, figs. 3-7; text-fig. 7

1887 Trinucleus seticornis (Hisinger) (?) var.; Brogger, p. 26.

1887 Trinucleus conf. seticornis ; Brogger, p. 29.

1887 Trinucleus-, Brogger, p. 30.

1887 Trinucleus Wahlenbergi; Brogger, p. 31.

71887 Trinucleus Wahlenbergi Rouault; Brogger, p. 32.

1897 Trinucleus Wahlenbergi Rouault; Kiaer, p. 33 [Upper Isotelus Limestone, ?‘5a’].

1930 Tretaspis latilimbus (Linnarsson); Stormer (pars), pp. 67-69 [Tretaspis Limestone specimens only],
pi. 11, figs. 8, ?9, 10, 11; text-figs. 33/, ?g, non e [= T. anderssoni ], non 34 d [= T. latilimbus
latilimbus \.

1934 Tretaspis latilimbus-, Stormer, p. 330.

1945 Tretaspis latilimbus (Linnarsson); Stormer, p. 403, pi. 1, fig. 9.

Holotype. A cephalon (PMOl 1751) from the Tretaspis Limestone on Lindoya, Oslo.

Material, localities, and horizons. A great deal of very fragmentary material and rarer more complete specimens
occur at various levels in Oslo-Asker: Tretaspis Limestone on Langara, Lindoya, Ostoya, and Treneholmen;
Upper Tretaspis Shale on Hovedoya and Nakholmen; Upper Isotelus Limestone on Hovedoya, Langoyene,
Lindoya, and Skjaerholmen; all but the upper few metres of the Husbergoya Shale Formation on Hovedoya and
possibly Husbergoya, Rambergoya, and Langoyene. A specimen in limestone (?Heroya Limestone) from the
Skien-Langesund district probably belongs here also.

Diagnosis. Arcs El5 I13, and In complete. I4 short to complete, E2 present in 41% of thirty-four specimens.
Reticulation on external surface of glabella and genae subdued.

Description. Proportions of glabella and genal lobes similar to those of T. ceriodes angelini. There is a fine,
subdued reticulation on the external surface of the mesial part of the glabella and the adaxial parts of the genal
lobes, but they are smooth on internal moulds. Arcs Ex, I^, and In are complete, and I4 is developed in all
specimens, most having 3-1 1^ pits (twenty-four specimens) in this arc but one extreme specimen from the
Tretaspis Limestone has this arc complete. Two morphs (A and B) are defined on the absence or presence

EXPLANATION OF PLATE 91

Figs. 1-4. Tretaspis hisingeri sp. nov. 1, 2, holotype, PMO H51, dorsal and lateral views of partially exfoliated
almost complete specimen, 3-5-4-0 m below top of Frognoya Shale, Frognoya, Ringerike, x2£, x2; also
figured by Stormer (1930, text-fig. 47). 2, PMO80613, frontal view of partially exfoliated cranidium,
Frognoya Shale, Ringsasen, Norderhov, Ringerike, x2. 4, PMO 103957, dorsal view of cast of complete
meraspis degree 4, 7-91-7-94 m above base of Lower Tretaspis Shale, Ole Deviks Vei, Oslo, x 12^.

Figs. 5-14. Tretaspis hadelandica hadelandica Stormer. 5, 8, 11, holotype, morph B, PM065187, dorsal, frontal,
and lateral views of partially exfoliated cephalon, probably from the Gagnum Limestone Formation, south of
Gagnum, Hadeland, x 2; also figured by Stormer (1945, pi. 1, fig. 4). 6, morph B, PM098489, dorsal view of
lower lamella external to girder showing E2 complete frontally, upper part of Lunner Formation, Kjevlingen,
Hadeland, x 3-2. 7, morph C, PMO 103958, anterolateral view of partially exfoliated cranidium, Gagnum
Limestone Formation, 500 m south-east of Lunner Bakken, Hadeland, x 4 \. 9, morph A, PM099537,
oblique anterolateral view of cast of cranidium, 7- 1 -7-2 m below top of Gagnum Shale Member of the Lunner
Formation, 75 m south of Roko, Hadeland, x 5. 10, PMO103959, dorsal view of unwhitened pygidium,
lower part of Lunner Formation, 400 m east-south-east of Lunner Kirke, Hadeland, x 8. 12, PMO101483,
dorsal view of internal mould of pygidium, Gagnum Limestone Formation, Ballangrud, Hadeland, x 4. 13,
PMO 103960, dorsal view of unwhitened thorax and pygidium, lower part of Lunner Formation, Haga,
Hadeland, x4. 14, morph A, PM065193, dorsal view of partially exfoliated almost complete specimen,
Gagnum Shale Member of the Lunner Formation, Gagnum, Hadeland, x 3£; also figured by Stormer (1945,
pi. 4, fig. 16).

PLATE 91

OWEN, trilobite Tretaspis

736

PALAEONTOLOGY, VOLUME 23

respectively of E2 which occurs in 41% of the thirty-four specimens in which this feature could be determined
(Table 1). As in all species of Tretaspis the most posterior one or two E, pits always lack equivalent E2 pits. The
range of variation in fringe pitting is illustrated on text-fig. 7. Arcs Ix, El5 and, where present, E2 are out of phase
with the remaining I arcs and share sulci which extend to the anterolateral part of the fringe or even to the zone of
complication. The available samples are too small to detect differences from rock unit to rock unit and both
morphs are known from all but the lower part of the Husbergoya Formation. Personally collected material from
the Upper Tretaspis Shale shows both morphs in the same bed.

Hypostoma unknown. Thorax similar to that of T. seticornis.

Pygidium only known with certainty from a few fragments. The one figured by Stormer (1930, pi. 11, fig. 9)
may belong here or to T. sortita broeggeri (see below) as the precise horizon in the Husbergoya Formation is not
known. This specimen has ten pairs of apodemal pits, the posterior three lying on the pygidial border.

3 5 7

Pits in E2
( morph B only )

JTfl-

n R ! i i n //i— r

Pits Missing Pits along

Anteriorly from I4 Posterior Margin

text-fig. 7. Histograms showing range of variation in fringe features of all
available specimens of Tretaspis latilimbus norvegicus with a comparison, where
possible, of the range, mean, and sample standard deviation of the two morphs
(A and B) present in the subspecies. In many instances the samples are too small
for reliable standard deviations or even means to be calculated.

Discussion. Ingham (1970, p. 50, text-fig. 18a, b) chose a lectotype from Linnarsson’s original
material of T. latilimbus from the Upper Johnstorp Formation (Rawtheyan) of Vastergotland,
Sweden, and he figured a number of topotypes (1970, text-fig. 18c-/). Dr. Ingham has allowed me to
collate some of his data on topotype material of the Swedish form in the collections of the
Riksmuseum, Stockholm. All these specimens have an incomplete I4 arc (2-11 pits in forty-two
specimens) which is not continuous mesially. Where the E arc development is sufficiently well
preserved, only one specimen out of thirty-five is seen to have pits in E2 and thus the vast majority
correspond to the development seen in T. latilimbus norvegicus morph A. The range of variation seen
in the pit distribution of other arcs is similar to that of the Norwegian material and it is most likely
that the Swedish form is simply a geographical subspecies of T. latilimbus norvegicus in which morph
B has been virtually excluded. Fragments of Tretaspis from the Ulunda Formation (Rawtheyan) in
Vastergotland have a pit development similar to that of T. latilimbus norvegicus morph B (J. K.
Ingham, pers. comm. 1976).

T. Tatilimbus' distichus Ingham (1970, p. 50, pi. 7, figs. 8-16, text-figs. 14 g, 16) was based on
material from the Rawtheyan Stage (Ashgill Zone 7) in the Cautley district of northern England and
is characterized by the presence of a short I4 and seven to ten pits in E2. It thus resembles T. latilimbus
norvegicus morph B and might be regarded as being a subspecies which developed in the same way as
T. latilimbus latilimbus. However, Ingham (1970, p. 50) suggested that T. ‘ latilimbus ’ distichus may

OWEN: TRILOBITE TRETASPIS

Til

have been derived from T. hadelandica brachystichus with the completion of I3 and the development
of a short I4. Indeed, one specimen of the latter was noted by Ingham to have a pit in I4. When the
ranges in variation in E2 and I3 in T. hadelandica brachystichus are considered for Zones 5 and 6
separately (text-fig. 6; Ingham 1970, text-fig. 16) there is a suggestion of a trend towards the condition
seen in T. ' latilimbus’ distichus. Moreover, McNamara (1979, p. 63) has noted the occurrence of
specimens which he terms T. aff. latilimbus distichus from the White Limestone (top of Zone 6) in the
Lake District which he considers to be intermediate between T. hadelandica brachystichus and T.
‘‘latilimbus’ distichus. Dr. McNamara informs me (pers. comm. 1979) that the White Limestone form
has I3 complete posteriorly and in some specimens there is a single pit in I4. Thus it seems likely that
the Zone 7 form is not directly related to T. latilimbus and ultimately may best be considered a
stratigraphical subspecies of T. hadelandica.

The origins of T. latilimbus are not clear but T. hadelandica hadelandica seems to be the most likely
ancestor.

Tretaspis sortita (Reed, 1935) broeggeri Stormer, 1945
Plate 92, figs. 8-11, 13, 14; text-fig. 8
71887 Trinucleus Wahlenbergi Rouault; Brogger, p. 32.

71897 Trinucleus Wahlenbergi Rouault; Kiaer, pp. 32 {pars, ‘4d§’ specimens only), 73.

1945 Tretaspis latilimba (Linnarsson) var. broggeri Stormer, p. 403, pi. 1, fig. 10.

1979 Tretaspis sortita (Reed) broeggeri Stormer; Owen, p. 257.

Holotype. An incomplete cephalon (PMO 11957) from the upper part of the Husbergoya Shale Formation on
Skjaerholmen, Oslo.

Material, localities, and horizons. Cephala, cranidia, and rare thoracic segments are known from the upper few
metres of the Husbergoya Formation in Oslo, on the islands of Skjaerholmen, Husbergoya (upper 2 m),
Hovedoya (upper 2-5 m), South Langoyene (upper 5 m), Lindoya, and Gressholmen.

Description. Glabella and genal lobes similar to those of T. ceriodes angelini but bear a variably developed,
subdued, fine reticulation on the external surface (on the genal lobes this is restricted to the posterior parts) and
are smooth on internal moulds. Fringe steeply declined with a well-developed anterior arch and a distinct brim
laterally. Arcs E1; I1_3, and In are complete, I4 is incomplete, and three morphs are recognized on the basis of the
E2 and 1 5 development (Table 1) thus: morph A lacks E2 and I5, morph B lacks I5 but has a short E2, morph C
has both E2 and I5 present but incomplete. The range of variation in pit development is shown on text-fig. 8.
There are too few specimens to give meaningful comparisons of some of the parameters in the three morphs
separately, and in some instances morphs A and B are considered together on text-fig. 8. Arc E2 is irregularly
developed in a few specimens (e.g. PI. 92, figs. 8, 10) but in most cases where it is present it is restricted to the
posterior part of the fringe. In all morphs, arcs Ij-Ej share sulci to bR2-8 (mean bR5, sample standard deviation
1, eighteen specimens) and distinct lists are developed between most arcs over the whole fringe except Ej-I,,
where they share sulci and between Ej and E2 and also I5 and In. Two sets of radii are developed.

Hypostoma not known. Thorax and pygidium known only from a few poorly preserved fragments.

Discussion. The holotype of T. sortita broeggeri has eleven pits in E2 and one in I5, and thus is of
morph C type. Morphs A and B are indistinguishable from the two morphs constituting T. latilimbus
norvegicus although their relative abundances are very different with E2 being developed much more
commonly in T. sortita broeggeri. Stormer’s subspecies probably was derived from T. latilimbus
norvegicus with the development of 1 5 in some individuals and replaces the earlier form quite abruptly
in the Husbergoya Formation, although it is not possible to assign unequivocally isolated specimens
of morphs A and B to either form.

T. sortita sortita (Reed, 1935, pp. 3-6, pi. 1, figs. 4-10; see also Begg 1944, pp. 114, 115, pi. 5, figs.
2-7) was based on material from the Upper Drummuck Group (late Rawtheyan) at Girvan, south-
west Scotland. A complete topotype specimen was figured by Ingham (1970, pi. 8, fig. 1) who noted
(1970, p. 50) that the Scottish form has an incomplete E2, an extensive but incomplete I4, and a few
pits in I5. Dr. Ingham has informed me (pers. comm. 1976) that the vast majority of specimens from

738

PALAEONTOLOGY, VOLUME 23

Pits in E2

C

A+B

i°-|

5-

1/2 1 2 3

Pits Missing
Anteriorly from I4

Pits in I4

nb 42% of 31 specimens of morphs AH-B
and 93 % of 30 specimens of morph C
have this arc complete anteriorly

Pits in In

Pits Missing
Anteriorly from 1 5

Pits along
Posterior Margin

text-fig. 8. Histograms showing the range of variation in fringe features of
all available specimens of Tretaspis sort it a broeggeri with a comparison of the
range, mean, and sample standard deviation of the three morphs (A, B, and
C) present in the subspecies. Owing to the limited amount of data for morphs
A and B, these are considered together in most instances.

EXPLANATION OF PLATE 92

Figs. 1,2. Tretaspis hadelandica hadelandica Stormer. PMO 103961, right and left lateral views of internal mould
of cranidium showing asymmetrical I3 development, lower part of Gagnum Shale Member of the Lunner
Formation, 200 m north of Aslund, Hadeland, x 8.

Figs. 3-7. Tretaspis latilimbus (Linnarsson) norvegicus subsp. nov. 3, holotype, ?morph A, PMOl 1751, antero-
lateral view of incomplete cephalon, Tretaspis Limestone, Lindoya, Oslo, x 4; also figured by Stormer (1945,
pi. 1 , fig. 9). 4, morph B, PMO 10 1 551, anterolateral view of partially exfoliated cranidium, same horizon as 3,
west Rambergoya, Oslo, x3. 5, morph A, PMO80518, oblique anterolateral view of internal mould of
cephalon, Husbergoya Shale Formation, North Langoyene, Oslo, x 3. 6, morph A, PMO 1 03962, dorsal view
of internal mould of cephalon and part of thorax, Upper Tretaspis Shale, north Hovedoya, Oslo. 7, morph B,
PMO80573, oblique posterolateral view of cephalon, same horizon as 3, Ostoya, Baerum, x 3.

Figs. 8-11, 13, 14. Tretaspis sortita (Reed) broeggeri Stormer. 8, morph B, PMO31010, anterolateral view of
internal mould of incomplete cephalon showing irregular E2 development, upper part of Husbergoya Shale
Formation, South Langoyene, Oslo, x 2\. 9, holotype, morph C, PMOl 1957, lateral view of internal mould
of incomplete cephalon, same horizon as 8, Skjaerholmen, Oslo, x 4-); also figured by Stormer (1945, pi. 1, fig.
10). 10, PMO 100720, ventral view of cast of lower lamella, pygidium, and thorax, top of Husbergoya Shale
Formation, Rambergoya, Oslo, x 2\. 11,14, morph C, PMO 103963, frontal and anterolateral views of cast
of cephalon, same horizon as 8, Hovedoya, Oslo, x 3. 13, morph C, PMO 103964, posterior view of cast of

crushed cephalon and incomplete thorax, note weak reticulation on posteromesial parts of genal lobe, upper 2
m of Husbergoya Shale Formation, Husbergoya, Oslo, x 4.

Figs. 12, 15. Tretaspis askerensis sp. nov. 12, PM064649, frontal view of cast of crushed cranidium, middle part
of Grina Shale Member of the Lunner Formation, Grina, Hadeland, x 4; also figured by Stormer (1945, pi. 1,
fig. 1). 15, PM06376, posterolateral view of cast of incomplete cephalon, from either the lower part of the
Langara Limestone-Shale Formation or the Husbergoya Shale Formation, Hvalstad, Asker, x 6.

PLATE 92

OWEN, trilobite Tretaspis

740 PALAEONTOLOGY, VOLUME 23

Girvan are of this type and are very similar, if not identical, to the Norwegian morph C. Thus T.
sortita broeggeri differs from the Scottish form only in the proportions of constituent morphs.

Price (1977, pp. 784-785, pi. 103, figs. 1-7; text-fig. 2) assigned material to T. sortita from late
Ashgill mudstones in the Meiford area and commented on other Welsh material probably belonging
to this species. The specimens which he described have E2 developed and only one out of seven lacks
pits in I5. Unlike both the Norwegian and Scottish forms, the genal lobes are completely smooth, arcs
Ij-Ej share short sulci in only a few specimens and lists are less well developed.

Dr. Ingham informs me (pers. comm. 1976) that one specimen of Tretaspis from the type unit and
locality of T. latilimbus latilimbus, the Upper Johnstorp Formation in Vastergotland, has a short I5
developed (four pits). The E pits are not preserved but the specimen may well be of T. sortita type and
further, well-localized, collections may enable greater correlation between the Swedish, Norwegian,
and British upper Ashgill sequences.

Tretaspis askerensis sp. nov.

Plate 92, figs. 12, 15; Plate 93, figs. 1-5.

1902 Trinucleus Wahlenbergi Rouault; Kiaer, p. 78.

1945 Tretaspis seticornis (Hisinger) forma typica; Stormer, p. 406, pi. 1, fig. 6.

1978 Tretaspis aff. seticornis seticornis (Hisinger); Owen, p. 15.

Holotype. A cranidium (PMO 100657) from either the Husbergoya Shale Formation or the lower part of the
Langara Limestone-Shale Formation (i.e. ‘5a’ of Brenchley and Newall 1975) Holmenskjaeret, Holmen, Asker.

Material, localities, and horizons. Four incomplete cranidia from the type horizon and locality, a cephalon
probably from ‘5a’ at 0vre Nes badestrand, a cranidium possibly from this unit at Hvalstad, three cranidial
fragments from 2-3 m above the base of the Husbergoya Formation on Bronnoya and 1 -4 m above the base of
this unit on Langara, all Asker. Two cranidia from the lowest 13 m of the Husbergoya Formation on Kalvoya,
Baerum. Three fragmentary cranidia from a channel conglomerate in the upper part of the Langara Formation
on Ostoya, Baerum (indicating transport from the west), and one external mould of a cephalon from the Grina
Shale Member of the Lunner Formation at Grina Hadeland.

Diagnosis. Pseudofrontal lobe strongly swollen, almost circular in dorsal view. Arcs Ex, Ix, I2, and In complete. A
short I3 is present in a few specimens. f-Ej sulci deep. E2 short or absent.

Description. Most specimens of this form are noticeably smaller than those of other Norwegian species but the
material is too incomplete to quantify this adequately. Pseudofrontal lobe strongly swollen, almost circular in
dorsal view but otherwise the proportions of the glabella and genae very similar to those of T. ceriodes angelini.
The external surface of the glabella and genae bears a well-developed reticulation which is very fine except on the
posteromesial parts of the genal lobes where it is coarser. Some internal moulds fairly strongly reticulate. Fringe
steeply declined. Arcs El5 11; I2, and In complete. E2 developed in two specimens (of four) where it comprises up
to six pits. I3 present in two specimens (of seven) where it contains two or three pits, beginning at about aR3.
There are eight pits along the posterior margin of the fringe in three specimens. There are nineteen pits in In in one
topotype specimen and the Grina Shale cephalon and 1 6^ in a specimen from Kalvoya. Arcs E , and I , share deep
sulci over all but the posterior part of the fringe and are out of phase with the remaining I arcs.

Remainder of exoskeleton unknown.

Discussion. The fringe development of T. askerensis resembles that of T. seticornis and the Grina
Shale specimen was assigned to this species by Stormer (1945) and Owen (1978). T. askerensis differs
in its deep I j -Ej sulci, in having I3 developed in a few specimens, in having eight (cf. six or seven) pits
along the posterior margin, and the number of pits in In extends beyond the maximum recorded for T.
seticornis. Clearly the very limited number of specimens of both species makes objective comparison
very difficult. The more circular outline of the pseudofrontal lobe and much stronger reticulation also
distinguish the younger species although the latter character may have little taxonomic value (see
Price 1977, p. 781). T. seticornis has a very short stratigraphical range, being restricted to low
Pusgillian strata in both Norway and Sweden. T. askerensis occurs in Rawtheyan units and probably

OWEN: TRILOBITE TRETASPIS

741

was derived from, for example, T. hadelandica hadelandica or T. latilimbus norvegicus. The relatively
small size, well-developed reticulation (see Stormer 1930, p. 65) and simple fringe morphology
suggest a neotenous origin for the species.

Tretaspis sagenosus group?

Tretaspis kiaeri Stormer, 1930
Plate 93, figs. 6-15; text-fig. 9

1921 Trinucleus; Kiaer, p. 500.

1930 Tretaspis kiaeri Stormer, pp. 50-55, pi. 10, figs. 1-6; pi. 11, fig. 12; pi. 13, fig. 13; pi. 14, figs. 1-3;
text-figs. 21c, 23-26, 38.

1945 Tretaspis kiaeri Stormer; Stormer (pars), p. 403, pi. 1, fig. 12; non pp. 387, 406, pi. 1, fig. 1 1 [= T.
hadelandica hadelandica ].

1953 Tretaspis kiaeri; Stermer, p. 87.

1959 Tretaspis kiaeri Stormer; Harrington in Moore, text-fig. 70c.
non 1966 Tretaspis kiaeri Stormer; Whittington, pp. 90-92, pi. 28, figs. 1,6-12, 14.
non 1968 Tretaspis kiaeri Stormer; Whittington, p. 93, pi. 29, figs. 1, 2, 4.

1975 T. kiaeri Stormer; Hughes et al ., p. 563.
non 1975 T. aff. kiaeri; Hughes et al., p. 563.

1979 Tretaspis kiaeri Stormer; Owen, pp. 250, 251, 252, text-fig. 6.

1979 Tretaspis kiaeri Stormer; Bruton and Owen, text-fig. 6.

Holotype. An almost complete internal mould of a cephalon (PMO H197) from the Hogberg Member of the
Solvang Formation, Frognoya, Ringerike.

Material, locality, and horizon. The species is known only from the type horizon and locality from which many
hundreds of disarticulated skeletal elements are known.

Description. Proportions of glabella and genal lobes similar to those of T. ceriodes angelini except that the
glabella is a little more inflated and overhangs the fringe a little. Reticulation variable. On the glabella it is
coarsest around the median node and extends to a transverse line at the maximum width (tr.) of the occiput. The
reticulation of the genal lobes is finer and more subdued than that of the mesial part of the glabella and becomes
finer abaxially. On some internal moulds there is a faint reticulation on the genal lobes and, less commonly, the
glabella. Fringe steeply declined laterally, less so across In mesially, in front of which it is vertical. Arcs Ej_2, 1 ,_3,
and In complete mesially and posteriorly. I4 is continuous mesially but extends to the posterior margin in only 5%
of sixty-one specimens. Two morphs are defined on the absence (A) or presence (B) of I5 which occurs in 35% of
eighty-one specimens and extends mesially in 22% of the twenty-three specimens in which its frontal extent can
be determined (Table 1). The range of variation in selected fringe characters is shown on text-fig. 9. Two sets of
radii are developed and pits in the outer set, Il5 E^, share sulci to the posterior part of the fringe in some
specimens but in a few this sulcation is less extensive and L becomes discrete as far forwards as bR5. Very fine
lists are developed between all I arcs except I4 and I5.

Hypostoma unknown. Thorax similar to that of T. seticornis, although it is not known whether or not median
tubercles are present. The pygidial rachis commonly has up to six transversely directed furrows bearing deep
apodemal pits distally. These furrows are progressively less well incised rearwards along the rachis and on well-
preserved specimens (PI. 93, fig. 9; Stormer 1930, pi. 10, fig. 4) a further three to five pairs of apodemal markings
are seen, the posterior two or three pairs being situated on the anterior part of the border. Three pairs of weakly
developed pleural ribs present.

Discussion. Whittington (1966, 1968) ascribed specimens from the Ashgill of Wales to T. kiaeri. One
of these (1966, pi. 28, fig. 13) was referred to Nankinolithus Lu by Hughes et al. (1975, p. 559). The
remainder comprise at least three distinct forms of Tretaspis and have been reassessed by Price ( 1 977,
pp. 786-787) who considered specimens from the Rhiwlas Limestone (probably Rawtheyan) figured
by Whittington (1968, pi. 28, figs. 12, 16) to be similar to T. calcaria Dean, 1971, a form described
originally from the Chair of Kildare Limestone (probably Rawtheyan) in Eire. T. calcaria is almost
certainly related to T. kiaeri but differs in having I4 always complete posteriorly, I5 more extensive,
and all complete arcs have a higher pit count (e.g. 30-31 cf. 20^-27^ in Ex). As noted by Price, the

742

PALAEONTOLOGY, VOLUME 23

poorly preserved Rhiwlas Limestone material is difficult to compare with Dean’s species but
differences in fringe pitting seem slight.

Other British and Irish forms previously assigned to T. kiaeri have been reassessed by Ingham
(1970, pp. 44-57) and Price (1974, pp. 844-847; 1977, pp. 766-778). Most are clearly members of the
T. moeldenensis group and thus are distinguished from T. kiaeri primarily in having complete radial
alignment of the fringe pits. A few are T. seticornis group members and have E2 incomplete mesially.

EVOLUTION OF THE TRETASPIS SETICORNIS GROUP

The study of populations of Tretaspis from Norway indicates that the phylogenetic relationships are
more complex than was thought previously and that a purely typological approach to their taxonomy
is not possible. Nevertheless, the broad evolutionary history of the T. seticornis group is becoming
clear (text-fig. 1).

n 1 1 m n rhn fi

- 3 4 5 6 7

Pits in I5

text-fig. 9. Histograms showing the range of variation in fringe features seen
in all available specimens of Tretaspis kiaeri with a comparison of the range,
mean, and sample standard deviation of the two morphs (A and B) present in
the species.

EXPLANATION OF PLATE 93

Figs. 1-5. Tretaspis askerensis sp. nov. 1-3, holotype, PM0100657, dorsal, lateral, and anterolateral views of
partially exfoliated cranidium, Husbergoya Shale Formation, or lower part of Langara Limestone-Shale
Formation, Holmenskjaeret, Holmen, Asker, x 7. 4, PMO80463, anterolateral view of partially exfoliated
cranidium, same horizon and locality as 1-3, x 10. 5, PM0100878, cast of flattened incomplete cephalon,
probably from the type unit, 0vre Nes badestrand, Nesbru, Asker, x 6.

Figs. 6-15. Tretaspis kiaeri Stormer, Hogberg Member of the Solvang Formation, Frognoya, Ringerike. 6, 10,
holotype, morph B, PMO HI 97, dorsal and frontal views of internal mould of cephalon, x 3£; also figured by
Stormer (1930, pi. 10, fig. 1). 7, morph B, PMO H338, lateral view of internal mould of cephalon, x 3; also
figured by Stormer (1930, pi. 10, fig. 3). 8, PMO103965, dorsal view of cast of pygidium and incomplete
thorax, x4^. 9, PMO103966, dorsal view of internal mould of pygidium, x4J. 1 1, morph A, PMO 103967,
anterolateral view of internal mould of incomplete cephalon, x 2. 12, morph A, PMO H208, posterolateral
view of incomplete partially exfoliated cranidium showing pitting along the marginal band, x 4; also figured
by Stormer (1930, pi. 11, fig. 12). 13, PMO 103968, slightly oblique dorsal view of cast of glabella and left
genal lobe, note glabellar reticulation, x 14, morph B, PMO 103969, dorsal view of internal mould of
cephalon and part of thorax, same specimen as 8, x 4. 1 5, morph B, PM0354, oblique anterolateral view of
cephalon, x 3; also figured by Stormer (1945, pi. 1, fig. 12).

PLATE 93

OWEN, trilobite Tretaspis

744

PALAEONTOLOGY, VOLUME 23

The earliest known species of Tretaspis from the Anglo-Welsh and Scandinavian areas is T.
ceriodes which is restricted to latest Caradoc units in all these areas. The species is polymorphic in
Norway and almost certainly gave rise to the T. seticornis group, the replacement of the former by the
latter being geologically instantaneous and an excellent tool in recognizing the Caradoc-Ashgill
boundary (Owen 1979, p. 251). The earliest representatives of this group are distinct in different areas
with T. hadelandica in England and Hadeland and T. seticornis in Oslo-Asker, Ringerike, and
Sweden. This rapid speciation involved the development of two sets of pit radii and, with the
exception of some members of early T. hadelandica populations, the restriction of E2 to the lateral
parts of the fringe. The polymorphic nature of the ancestral T. ceriodes populations accounts for all
other fringe features of the early T. seticornis group forms. Local populations of T. hadelandica
became isolated very early on, giving rise to what are interpreted as geographical subspecies. The T.
moeldenensis group persisted into the Ashgill in Britain but not in Scandinavia.

In Britain, T. hadelandica is now interpreted as ranging from earliest Pusgillian to mid/late
Rawtheyan with a series of stratigraphical subspecies showing a progressive simplification of fringe
characters ( T . h. convergens — T. h. deliquus — T. h. brachystichus) followed by a slight reversal of this
trend within T. h. brachystichus which may have been continued with the development of T.
Tatilimbus' distichus. This reinterpretation strengthens the stratigraphical usefulness of the British
forms especially in view of the long-ranging homeomorphs present in Norway. The origins of the
Irish and Scottish T. sp. from which T. persulcatus were descended are unclear.

In Hadeland, T. hadelandica hadelandica persisted from early Pusgillian to Rawtheyan times, and
although there are differences in the percentages of constituent morphs in different units, these are
considered to reflect ecological rather than temporal controls. T. hadelandica may have given rise to a
homeomorph of T. seticornis, T. askerensis which occurs in Hadeland and Asker.

In Oslo-Asker, Ringerike, and Sweden, T. seticornis has a short stratigraphical range and gave rise
to another short ranging form, T. hisingeri. In Ringerike, T. seticornis also gave rise to T. anderssoni,
a form which has a very narrow range of variation throughout its range from mid-Pusgillian to early
Rawtheyan. In Oslo-Asker, T. hisingeri is replaced by T. latilimbus norvegicus, a polymorphic form of
uncertain origin which extends well into the Rawtheyan and which almost certainly gave rise to
T. sortita broeggeri. One of the morphs constituting T. latilimbus norvegicus is by far the dominant
form in the nominate subspecies which is a Swedish taxon developed during the Rawtheyan.
Populations of T. sortita sortita from the late Rawtheyan of Scotland differ from T. sortita broeggeri
in the proportions of constituent morphs.

There is still very little information on bed-by-bed changes in populations of Tretaspis, and the
Norwegian material is not sufficiently abundant for such a study. There is a suggestion that the
development of phenotypes in T. ceriodes angelini is to some extent progressive but as far as morphs
B, C, and D are concerned this represents no more than an increase in the upper limit of the range of
variation. Many forms have long stratigraphical ranges within which there is no directional change.
The only likely example of evolutionary trends are the zigzag evolution seen in the British T.
hadelandica subspecies and the introduction of a third morph to produce T. sortita broeggeri from T.
latilimbus norvegicus. The latter change was fairly abrupt as was the development of the T. seticornis
group itself. There is insufficient evidence to say whether or not the changes in the British subspecies
of T. hadelandica are gradual. Neoteny is thought to have produced two species, T. hisingeri and T.
askerensis and probably also T. ceriodes from the T. sagenosus group.

Acknowledgements. I am very grateful to Dr. J. K. Ingham for his considerable help and encouragement and for
his comments on an earlier draft of this paper. I have also benefited from discussions with Professor H. B.
Whittington and the late Professor L. Stormer. I thank Mr. A. Buxton and Mr. J. Smith for their help in
preparing the figures and plates, Drs. D. L. Bruton (Paleontologisk Museum, Oslo), R. A. Fortey (British
Museum (Natural History)), and V. Jaanusson (Riksmuseum, Stockholm) for access to collections in their care,
and Dr. P. J. Brenchley and his group for showing me their collections (now PMO). Most of the work was carried
out during the tenures of a N.E.R.C. studentship at Glasgow University and a N.A.T.O. fellowship at the
Paleontologisk Museum, Oslo.

OWEN: TRILOBITE TRET A S P IS

745

REFERENCES

andersson, J. G. 1894. Ueber Blocke aus dem jungeren Untersilur auf der Insel Oland varkommend. Ofv. Kongl.
Vet.-Akad. Fork. 1893, 8, 521-540.

angelin, N. p. 1854. Palaeontologica Scandinavica I: Crustacea formationis transitionis, Fasc. 2, 21-92, pis.
25-41.

asklund, b. 1936. Die Fauna in einem Geschiebe aus der Trinucleusstufe in Jamtland. Sver. geol. Unders. Afh.
(C). 400, 1-6, pis. 1-2.

bassler, R. s. 1915. Bibliographic index of American Ordovician and Silurian fossils. Bull. U.S. natn. Mus. 92,
1-1521, pis. 1-4.

begg, J. L. 1944. On the fringe of Tretaspis. Geol. Mag. 81, 113-117, pi. 5.

bolton, T. E. 1970. Subsurface Ordovician fauna, Anticosti Island, Quebec. Bull. geol. Surv. Can. 187, 31-39,
136-139, pis. 6, 7.

brenchley, p. J. and newall, G. 1975. The stratigraphy of the Upper Ordovician Stage 5 in the Oslo-Asker
district, Norway. Norsk geol. Tidsskr. 55, 243-275.
brogger, w. c. 1887. Geologisk kart over oerne ved Kristiania. Nyt. Mag. f. Naturvid. 31, 1-36.
bruton, D. L. and owen, A. w. 1 979. Late Caradoc -early Ashgill trilobite distribution in the central Oslo Region,
Norway. Norsk geol. Tidsskr. 59, 213-222.

cave, R. 1960. A new species of Tretaspis from South Wales. Geol. Mag. 97, 334-337, pi. 10.

— 1965. The Nod Glas sediments of Caradoc age in North Wales. Geol. J. 4, 279 298, pi. 12.

cech, s. 1975. Cranidial reticulation and functional morphology of the cephalic fringe in Trinucleidae
(Trilobita). Vest. Ustr. ust.geol. 50, 173-177, pis. 1-4.

cocks, L. R. M. and price, d. 1975. The biostratigraphy of the upper Ordovician and lower Silurian of south-west
Dyfed, with comments on the Hirnantia fauna. Palaeontology, 18, 703-724, pis. 81-84.
dean, w. t. 1959. The stratigraphy of the Caradoc Series in the Cross Fell Inlier. Proc. Yorks, geol. Soc. 32,
185-227.

— 1961 . Trinucleid trilobites from the higher Dufton Shales of the Caradoc Series in the Cross Fell Inlier. Ibid.
33, 119-134, pis. 7-9.

— 1962. The trilobites of the Caradoc Series in the Cross Fell Inlier of northern England. Bull. Br. Mus. nat.
Hist. {Geol.), 7, 65-134, pis. 6-18.

— 1963. The Ordovician trilobite faunas of South Shropshire, IV. Ibid. 9, 1-18, pis. 1, 2.

— 1971. The trilobites of the Chair of Kildare Limestone (Upper Ordovician) of eastern Ireland. Palaeontogr.
Soc. [ Monogr .], 1, 1-60, pis. 1-25.

fearnsides, w. G., elles, G. and smith, b. 1907. The Lower Palaeozoic rocks of Pomeroy. Proc. R. Ir. Acad. 26B,
97-128, pis. 7, 8.

havlicek, v. and vanek, J. 1966. The biostratigraphy of the Ordovician of Bohemia. Shorn. Geol. ved. [P], 8,
7-69, pis. 1-16.

hawle, I. and corda, A. J. c. 1847. Prodrom einer Monographie der bohmischen Trilobiten. 176 pp. Prague.
hayami, i. and ozawa, T. 1975. Evolutionary models of lineage zones. Lethaia, 8, 1-14.
hisinger, w. 1840. Lethaea Svecica seu Petrificata Sveciae, iconibus et characteribus illustrata. Suppl.
secundum. Holmiae.

HOLTEDAHL, o. and schetelig, J. 1923. Kartbladet Gran. Norges geol. Unders. 97, 1-46.
hughes, c. p. 1970. Statistical analysis and presentation of trinucleid (Trilobita) fringe data. Palaeontology, 13,
1-9.

— ingham, j. k. and addison, R. 1975. The morphology, classification and evolution of the Trinucleidae
(Trilobita). Phil. Trans. R. Soc. Lond. B272, 537-607, pis. 1-10.

ingham, j. k. 1970. A monograph of the Upper Ordovician trilobites from the Cautley and Dent districts of
Westmorland and Yorkshire. Palaeontogr. Soc. [Monogr], 1, 1-58, pis. 1-9.

— 1978. Geology of a continental margin 2: middle and late Ordovician transgression, Girvan.

In bowes, D. R. and leake, b. e. (eds.). Crustal evolution in northwestern Britain and adjacent regions. Geol.
J. Special Issue, 10, 163-176.

jaanusson, v. 1956. Undersuchungen uber der oberordovizischen Lykholm-Stufenkomplex in Estland. Bull,
geol. Instn. Univ. Uppsala, 36, 369-400, pi. 1 .

— and martna, J. 1948. A section from the Upper Chasmops series to the Lower Tretaspis series at Fjacka
rivulet in the Siljan Area, Dalarne. Ibid. 32, 183-193.

kiaer, j. 1897. Faunistische Uebersicht der Etage 5 des norwegischen Silursystem. Skr. Norsk Vidensk.-Akad.
Mat. Naturv. Kl. 3, 1 -76.

746

PALAEONTOLOGY, VOLUME 23

kiaer, j. 1902. Etage 5 i Asker ved Kristiania. Norges geol. Under s. 34, 1-112.

— 1921. En ny zone i Norges midtre Ordovicium. Geol. Foren. Stockh. Forh. 43, 499-502.
kielan, z. 1957. On the trilobite family Staurocephalidae. Acta. Pal. Polon. 2, 155-182.

— 1960. Upper Ordovician trilobites from Poland and some related forms from Bohemia and Scandinavia.
Palaeont. Pol. 11, i-vi, 1-198, pis. 1-36.

lamont, A. 1935. The Drummuck Group, Girvan; A stratigraphical revision with descriptions of new fossils
from the lower part of the group. Trans, geol. Soc. Glasg. 19, 288-332, pis. 7-9.

1941. Trinucleidae in Eire. Ann. Mag. nat. Hist. (1 1), 8, 438-469, pi. 5.

lauritzen, 0. 1973. The Middle Ordovician of the Oslo Region, Norway. 24. Stage 4b at Lunner, Hadeland.
Norsk geol. Tidsskr. 53, 25-40.

lesperance, p. j. 1968. Trilobite faunas of the White Head Formation, Perce Region, Quebec. J. Paleont. 42,
811-826.

— and bertrand, R. 1976. Population systematics of the Middle and Upper Ordovician trilobite Cryptolithus
from the St. Lawrence Lowlands and adjacent areas of Quebec. Ibid. 50, 598-613.

linnarsson, j. G. o. 1869. Om Vestergotlands Cambriska och Siluriska Aflagringar. K. svenska Vetensk-Akad.
Hand/. 8(2), 1-89, pis. 1,2.

loven, s. L. 1845. Svenska Trilobiter. Ofver. Kgl. Vet.-Akad. Forhandl. andra argangen (1845), 46-56, 104-1 11,
pis. 1, 2.

mccoy, f. 1 849. 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.
mcnamara, k. j. 1979. Trilobites from the Coniston Limestone Group (Ashgill Series) of the Lake District,
England. Palaeontology, 22, 53-92, pis. 7-12.
mayr, e. 1969. Principles of systematic zoology, x, 428 pp. McGraw Hill, New York.

miller, J. 1976. The sensory fields and life mode of Phacops rana (Green, 1832) (Trilobita). Trans. R. Soc. Edinb.
69, 337-367, pis. 1-4.

moore, r. c. (ed.). 1959. Treatise on invertebrate palaeontology. Part O, Arthropoda 1. xix, 560 pp. Geol. Soc.
Amer. and Univ. Kansas Press, Lawrence.

owen, A. w. 1977. Upper Ordovician stratigraphy and trilobite faunas of the Oslo region, with special reference
to Hadeland and Ringerike. Unpubl. Ph.D. thesis, Univ. of Glasgow.

— 1978. The Ordovician and Silurian stratigraphy of Central Hadeland, South Norway. Norges geol. Unders.
338, 1-23, pi. 1.

— 1979 (for 1978). The upper Ordovician succession at Norderhov and on Frognoya in Ringerike, Norway.
Norsk geol. Tidsskr. 58, 245-258.

— 1980. A new species of Cryptolithus (Trilobita) from the late Ordovician of Norway. J. Paleont. 54,
144-148.

portlock, J. E. 1843. Report on the geology of the county of Londonderry, and of parts of Tyrone and Fermanagh.
xxxi, 784 pp., pis. 1-38, A-l. Dublin and London.

pribyl, A. and vanek, J. 1969. Trilobites of the family Trinucleidae Hawle et Corda, 1847 from the Ordovician of
Bohemia. Sb. geol. ved. Paleontologie, 11, 85-137.

price, d. 1973. The age and stratigraphy of the Sholeshook Limestone of South-west Wales. Geol. J. 8, 225-246.

— 1974. Trilobites from the Sholeshook Limestone (Ashgill) of South Wales. Palaeontology, 17, 841-868, pis.
112-116.

— 1977. Species of Tretaspis (Trilobita) from the Ashgill Series in Wales. Ibid. 20, 763-792, pis. 98-103.
reed, F. r. c. 1935. The Lower Palaeozoic trilobites of Girvan. Supplement No. 3. Palaeontogr . Soc. [ Monogr .],

1-64, pis. 1-4.

Schmidt, f. 1894. Revision der Ostbaltischen Silurischen Trilobiten Abth. IV. Mem. Acad. imp. Sci. St. Petersb.
(7), 42, (5), 1-94, pis. 1-6.

schuchert, c. and cooper, G. A. 1 930. Upper Ordovician and Lower Devonian stratigraphy and palaeontology
of Perce, Quebec. Part II. New Species from the Upper Ordovician of Perce. Am. J. Sci. (5), 20, 265-288,
365-392, pis. 1-5.

stgrmer, l. 1930. Scandinavian Trinucleidae with special reference to Norwegian species and varieties. Skr.
Norsk Vidensk.-Akad. Mat. Naturv. Kl. 4, 1-111, pis. 1-14.

— 1934. Cambro-Silurian zones of the Oslo Region, with a brief correlation between British and Norwegian
sections. In holtedahl, o. et al. 1934. The geology of parts of Southern Norway. Proc. geol. Ass. 45, 307-377,
pis. 22-32.

— 1945. Remarks on theTretaspis (Trinucleus) Shales of Hadeland with description of trilobite faunas. Norsk
geol. Tidsskr. 25, 379-425, pis. 1-4.

OWEN: TRILOBITE TRETASPIS

747

stormer, l. 1953. The Middle Ordovician of the Oslo Region, Norway. 1 . Introduction to stratigraphy. Ibid. 31,
37-141, pis. 1-6.

strand, T. and henningsmoen, G. 1960. Cambro-Silurian stratigraphy. In holtedahl, o. (ed.) Geology of
Norway. Norges geol. Unders. 208, 128-169, pis. 7, 8.

tomczyk, h. 1962. Stratigraphy of Old Palaeozoic sediments from bore-holes at Vszkowce near Lubaczow. In
passendorfer, e. (ed.) Ksiega Dam. Prof. J. Samsomowicza. Polska Akad. Nauk. Geol. 123-141 [Polish],
141-148 [Russian and English summaries], pis. 25-29.

tornquist, s. L. 1883. Ofversicht ofver Bergbyggnaden inom Siljansomradet. Sver. geol. Unders. Afh. (C), 57,
1-59.

— 1884. Undersokninger ofver Siljansomradets trilobitfauna. Ibid. (C), 66, 1-101, pis. 1-3.

Whittington, h. b. 1941. Silicified Trenton trilobites. J. Paleont. 15, 492-522, pis. 72-75.

— 1966. A Monograph of the Ordovician trilobites of the Bala area, Merioneth. Palaeontogr. Soc. [ Monogr .],
3, 63-92, pis. 19-28.

— 1968. A Monograph of the Ordovician trilobites of the Bala area, Merioneth. Ibid. 4, 93-138, pis. 29-32.

A. w. OWEN

Department of Geology

Typescript received 22 August 1979 The University

Revised typescript received 7 January 1980 Dundee DD1 4HN

A TECHNIQUE FOR REVEALING THE STEREOM
STRUCTURE OF FOSSIL CRINOIDS

by G. d. sevastopulo and j. b. keegan

Abstract. The stereom of fossil crinoid ossicles preserved in an argillaceous matrix can be revealed by treating
them with hydrofluoric acid. The clay filling the stereom pores is dissolved and the skeletal calcite is faithfully
replaced by fluorite. Features discovered in selected Lower Carboniferous crinoid ossicles prepared by this
method include the following: large canals penetrating the areola in the columnals of a particular inadunate
crinoid; triple aboral nerve canals, and labyrinthic stereom in the muscle fossae of distinctive inadunate
brachials; and a regular arrangement of trabeculae forming a cubic structure in the stereom of flexible crinoid
brachials.

Over the last decade, there have been a number of studies of the detailed morphology of recent
echinoderm ossicles, using scanning electron microscopy. Of these, we pick out the surveys of
crinoid microstructure by Macurda and Meyer (1975, 1976) and the extensive investigations of
the crinoid stem by Roux (1970, 1971, 1974«, 19746, 1975) as having particular significance for fossil
crinoid studies. In all of these studies the architecture of the stereom has been shown to have a
functional significance. Unfortunately the details of the stereom are difficult to discern in most fossil
material, because carbonate cements precipitated epitaxially on the skeletal calcite occlude the
stereom pore spaces. In some examples, however, the stereom is clearly visible in thin section. The
most common cases of this are when the stereom pores are filled by an iron-rich carbonate cement
(which can be differentiated by staining), or by micritic sediment or cement, or by iron sulphides or
oxides, or by clay minerals. When such mineralogical or textural differences are exploited by natural
weathering or by controlled acid etching, the three-dimensional stereom architecture may be
revealed: Lane and Macurda (1975) established the presence of muscular articulation in naturally
weathered brachials of the Pennsylvanian cladid crinoid, Aesiocrinus\ and Lapham, Ausich, and
Lane (1976) have illustrated the structure of the stereom of Mississippian crinoid ossicles which had
been etched in weak formic acid.

Whilst trying to recover miopores from a Carboniferous marine shale, we accidentally discovered
that some crinoid ossicles treated with hydrofluoric acid (HF) showed surprisingly detailed
microstructure: clay filling the stereom pores was dissolved and the calcite of the ossicles was
faithfully replaced by fluorite, a process which has been named fluoridization by Upshaw, Todd, and
Allen (1957, p. 793). The use of hydrofluoric acid in the preparation of calcareous fossils has been
independently (and often accidentally) discovered several times (Cookson and Singleton 1954;
Grayson 1956; Wetzel 1921). Most stress has been laid on the translucent nature of fluoridized fossils
when they are immersed in liquid. Sohn (1956) was able to make visible ostracode muscle scars by
treating the valves with HF, and Upshaw et al. (1957) illustrated the internal structures of fluoridized
foraminifers. Sprinkle and Gutschick (1967) used HF to prepare blastoids preserved in a fine-grained
sandstone.

We have applied the fluoridization technique to Carboniferous crinoid material preserved in a
variety of rock types, ranging from plastic clays of the mid-western United States to indurated silty
mudstones from Ireland. Examples of the results obtained are shown in Plates 94 and 95.

(Palaeontology, Vol. 23, Part 4, 1980, pp. 749-756, pis. 94-95.|

750

PALAEONTOLOGY, VOLUME 23

METHODS

The material for which the fluoridization technique is most effective is that preserved in clay mudstones or shales
where the clay has penetrated deeply into the stereom pores. We have generally used bulk samples rather than
attempting to fluoridize particular individual specimens, because of the risk of damage. However, most of the
microcrinoids described by Lane and Sevastopulo (in press) were first picked from washed clays and then
fluoridized.

It is worth while to remove as much matrix from the sample as possible. Soft clays may be disaggregated by
being air-dried, soaked in paint thinner or paraffin, and then vigorously boiled in water with soda ash. More
indurated mudstones and shales may require simmering in Quaternary ‘O’ (Zingula 1968), but since that
detergent is weakly acid, prolonged treatment results in some etching of skeletal calcite; we prefer to treat
particularly intractable samples directly with HF.

The partially cleaned fossil material is reacted with HF; the optimum strength of the acid and length of the
reaction time vary from sample to sample. We have used 48% HF and reaction times of between 5 minutes and
1 hour for small specimens; for larger specimens weaker acid (approximately 6%) and longer reaction times (up
to 24 hours) as advocated by Grayson (1956, p. 78) lead to better results. The fluoridization can be judged to have
proceeded far enough when the surfaces of the ossicles appear bleached; it is not necessary to convert whole
ossicles to fluorite.

Two adverse effects can occur during fluoridization. Firstly, the ossicles may crack and pieces may spall off.
This can be largely avoided by reducing the reaction time to a minimum and by diluting the acid. Secondly,
a glaze-like precipitate of fluorite may form on the surface of the ossicles. This can be prevented by using a large
enough quantity of acid (we have found five times the volume of material being fluoridized a suitable amount).
When the specimens have been fluoridized, they should be thoroughly washed and dried. Specimens for study
under the scanning electron microscope should be transferred to stubs immediately, because their delicate
surfaces can be easily damaged by abrasion.

Although we have been interested principally in the preparation of crinoid material, our bulk samples have
contained many other fossils, most of which appear perfectly preserved after fluoridization. We believe that the
technique may have general application in cleaning small fossils for study under the scanning electron
microscope.

Because hydrofluoric acid is extremely dangerous, the fluoridization process should always be carried out in a
properly designed fume cupboard with an efficient extraction system, by an operator wearing protective
clothing, rubber gloves, and a face-mask. The reaction between the sample and the acid may be very vigorous,
and large amounts of carbon dioxide may be generated rapidly. It is important, therefore, to treat the sample
in an adequately large polythene vessel to prevent froth from forming and spilling out. We fluoridize
approximately 10 g of bulk sample in an 80 mm-diameter 400 ml polythene beaker.

COMMENTS ON THE SPECIMENS ILLUSTRATED

The four ossicles illustrated in Plates 94 and 95 were obtained from a bulk sample of the soft clay shale
above the Charlestown Main Limestone, collected near the bathing pool, St. Monance, Fife,
Scotland (National Grid Reference NO 536 020). The shale is of Lower Carboniferous (Brigantian)
age and has been correlated with the Neilson Shell Band (George et al. 1976, fig. 14, p. 53). The
sample was partly disaggregated by being soaked in paraffin, and then boiled in water with soda
ash. Small amounts of the disaggregated material were reacted with 48% HF for 1 hour. The
fluoridized ossicles were mounted on stubs and coated with carbon and a gold palladium
mixture, and were examined using an ETEC Autoscan, Model H-l, scanning electron microscope.
The diameters of stereom pores were measured on enlarged scanning electron micrographs and the
surface porosity by point counting along two mutually perpendicular axes as suggested by Macurda
and Meyer (1975, p. 2). The terminology used is from Ubaghs (1978, T. 58 et seq.). The illustrated
specimens and other representative material are reposited in the palaeontological collections of
Trinity College, Dublin (catalogue numbers prefixed TCD).

Pentagonal columnals (TCD 19861-3) (PI. 94, figs. 1, 3)

Columnals of this kind are moderately abundant in the sample. The longest pluricolumnal found
consists of a nodal between two pairs of internodals. The nodal is cirrus-bearing and approximately

SEVASTOPULO AND KEEGAN: STEREOM STRUCTURE OF FOSSIL CRINOIDS 751

0-8 times as long as wide; the internodals are of two orders with length to width ratios of 0-4 and 0-6.
The sides of the columnals are straight, or have a ridge or swelling around the equator, a feature
particularly well developed on the nodals. Each nodal has one or two cirral sockets positioned
between the equator and the joint surface. The sockets are comparable in some respects to cirral
facets of Mesozoic crinoids illustrated by Ubaghs (1978, T. 85, fig. 61). They are gently concave and
slope towards the joint face. The lumen of the axial canal is a vertical slit. The half of the socket closest
to the equator of the columnal is furnished with short culmina; the half closest to the joint face is
smooth. The sides of the columnal are formed of dense stereom with a surface ornament of slightly
raised granules approximately 1 5 /xm in diameter.

In facetal view (PI. 94, fig. 1) the following regions of the articulum can be differentiated:

1 . The lumen, approximately 20-25% of the width of the articulum, which appears faintly five- or
ten-lobed in well-preserved specimens.

2. An adaxially sloping concave area surrounding the lumen (the floor of the spatium),
approximately 10-12% of the width of the articulum. The degree to which this region is depressed is
variable; it is very shallow in the specimen illustrated. It is floored by open stereom (round to ovoid
pores, with diameters from 6 to 14 /xm, mostly about 12 /xm) which in broken specimens can be seen to
form a thin layer overlying denser paraxial galleried stereom like that flooring the areola.

3. In some specimens (but not the figured example) the outer margin of the floor of the spatium is
raised to form a narrow perilumen constructed of denser stereom.

4. The areola (approximately 10-15% of the width of the articulum) which is flat and floored by
paraxial galleried stereom (pore diameter 6-9 /xm; surface porosity approximately 44%). Most pores
are subrounded and bounded by four trabeculae and many are arranged in long slightly arcuate rows.

5. The crenularium (approximately 10-15% of the width of the articulum) consisting of steep-
sided culmina and crenellae (PI. 94, fig. 3). The top of the culmina and base of the crenellae are
approximately equidistant from the level of the areola. The surfaces of the culmina are dense with
conspicuously thickened trabecular intersections (pore diameters are 2-5-5-0 /xm; surface porosity
30% or less), but are underlain by paraxial galleried stereom. The crenellae are mostly floored by
galleried stereom similar to that of the areola (pore diameters 7-10 /xm), but in some the stereom is
much more open and labyrinthic.

A conspicuous feature of the articulum is the set of large tunnel-like pores (up to 35 /xm in diameter)
which in several specimens can be seen to completely penetrate the columnal. They are crudely
arranged in ten lines and extend to the outer part of the areola.

In most respects the microstructure of these Carboniferous columnals is comparable with that of
Recent and Mesozoic columnals described by Macurda and Meyer (1975, 1976) and Roux (1971).
The galleried stereom of the areola probably housed ligament fibres. The denser stereom of the
perilumen and of the crenularium served as bearing surfaces. The large pores penetrating the
columnals may have contained nerves, as suggested by Macurda and Meyer (1975, p. 3) for similar
pores in the columnals of the Recent species Isocrinus blakei. The pore diameters of the columnals
described here are consistently smaller than those reported for most Recent and Mesozoic forms.

The taxonomic affinity of the specimens is not known. They almost certainly belonged to a cladid
inadunate, possibly an ampelocrinid in view of the pentagonal stem and cirrus-bearing nodals.

Elliptical columnal of Platycrinites (TCD 19864-6) (PI. 94, figs. 2, 4)

Columnals with elliptical articular surfaces are moderately common in the sample. They vary
considerably in shape. The majority, mainly smaller specimens, are longer than wide and have a
distinct equatorial waist. Most of them bear scattered nodes or blunt spines. A few specimens are
wider than long, and some of these, possibly nodals, have conspicuous equatorial spine-bearing
flanges. The articular surfaces are also variable although a basic pattern can be observed in all of
them: a raised fulcral region along the major axis of the face separates two gently concave fields. The
lumen is small and elliptical and is surrounded by open paraxial galleried stereom (pore diameters up
to 1 3 /xm; porosity approximately 37%). The central parts of the bifascial fields are floored by paraxial

752

PALAEONTOLOGY, VOLUME 23

galleried stereom with pore diameters typically 6-10 ^ m and porosity approximately 32%. The
peripheries of the faces are slightly raised above the bifascial fields and are formed of denser stereom
(pore diameter 3-5 /xm; porosity less than 30%). The long axis of the articular surface is occupied by a
fulcral region which in many specimens consists of a broad slightly raised ridge of dense stereom
(pore diameter typically 4 ^m; surface porosity approximately 20%). In some specimens the surface of
the fulcral region is crossed by low, dense, vermiform ridges. At each end of the major axis of the
articular surface are raised culmina, generally three in number, which rise above the level of the
fulcral region (PI. 94, fig. 4). They interlock with crenellae of adjacent columnals. The culmina are
formed of dense stereom (pore diameters typically less than 4 /xm; porosity less than 20%) and the
crenellae are floored by galleried stereom (pore diameter typically 9 p.m). The major axes of opposing
faces of many of the columnals are set at 90° to each other.

The ossicles are easily identified as belonging to Platycrinites but their specific identity is not
known. In many respects their structure is comparable with that of the columnals of the Recent
millericrinid Democrinus (Macurda and Meyer 1975, pp. 4, 5) which also has synarthrial articulation.
In the Scottish Platycrinites, however, the fulcral ridge is much less dense than in Democrinus, and
the elaborate keying mechanisms of that genus are not developed. Instead, limited symplectial
articulation occurred at both ends of the fulcral ridge.

Inadunate brachial (TCD 19867-9) (PI. 95, figs. 1, 3)

This kind of brachial is the most common in the sample. All examples that have been found are
cuneate, pinnule-bearing, higher than long, and most have nodes or blunt spines on the aboral
surface, particularly along the distal margins. All the brachials were joined by oblique muscular
articulations; the fulcral ridges on the two faces of a brachial may diverge by as much as 60°. The
following regions may be differentiated on the articular surfaces (PI. 95, fig. 1):

1. The fulcral ridge, which is narrow at its mid-point and widens slightly at both ends to
approximately 75 ^m in typical specimens. The ridge is constructed of dense stereom (pore diameters
typically 4 ^ m or less; porosity approximately 20%).

2. A slightly depressed area less than 30 /xm deep bounded by the fulcral ridge and the aboral
margin. By analogy with Recent crinoids, this area in Palaeozoic inadunates has been identified as the
aboral ligament fossa which housed the extensor ligament bundles. It is floored by galleried stereom
(pore diameters typically 7 /x m; porosity approximately 35%). The pores are subrounded and
arranged in a crude rectilinear pattern. In most specimens (but not the figured example) a distinct
small deeper ligament pit occurs just aborally of the mid-point of the fulcral ridge.

3. Two wide subequal depressions, typically less than 30 /x m deep, adoral of the fulcral ridge and
on either side of its mid-point, which have been identified as interarticular ligament fossae. They
are floored by galleried stereom in which the trabeculae and pores are conspicuously wider than
elsewhere on the articular surface. Pore diameters generally range from 10 to 15 /xm; the porosity is
approximately 40%.

EXPLANATION OF PLATE 94

Figs. 1, 3. Fluoridized pentagonal columnal (TCD 19861), from the shale above the Charlestown Main
Limestone, St. Monance, Fife (Lower Carboniferous; Brigantian age). 1, slightly oblique view of the articular
surface, x 45. 3, stereopair of the crenularium and outer part of the areola, located at about 7 o’clock on
fig. 1, x 230.

Figs. 2, 4. Fluoridized Platycrinites columnal (TCD 19864), from the shale above the Charlestown Main
Limestone, St. Monance, Fife (Lower Carboniferous; Brigantian age). 2, oblique view of columnal, x 38. 4,
stereopair of part of the fulcral ridge and culmina, from the left side of fig. 2, x 1 50.

PLATE 94

imm

Mmm

l wmM

;*®CS

SEVASTOPULO and KEEGAN, Stereom structure of fossil crinoids

754

PALAEONTOLOGY, VOLUME 23

4. A slightly raised area extending from the adoral groove to the mid-point of the fulcral ridge and
separating the interarticular ligament fossae. This area bears a very weak medial groove which ends
short of the fulcral ridge. The stereom of the raised area is galleried (pore diameter 6-8 jxm) along the
margins and more open (pore diameter 10-15 pm) and less regular along the median groove.

5. Well-marked, unequal, ‘rabbit-ear’-shaped depressions on either side of the ambulacral groove,
which have been interpreted as flexor muscle scars. They are floored by distinctive dense labyrinthic
stereom (pore diameters mostly less than 4 pm; porosity approximately 20%). The surfaces of the
fossae are formed by blunt-ended trabecular rods projecting upwards (PI. 95, fig. 3).

All well-preserved specimens can be seen to have three pores 20-30 pm in diameter adoral of the
fulcral ridge on both articular surfaces. Two of them lie along a line normal to the bisectrix of the
angle of the adoral groove; the third is between the other two, closer to the fulcral ridge.

Many of the features observed are similar to those reported by Lane and Macurda (1975) for the
Pennsylvanian cladid inadunate Aesiocrinus magnificus. The upward projecting trabecular rods of
the ‘rabbit ear’ fossae were probably sheathed with a thin connective tissue layer to which the muscle
fibres were attached, as illustrated for the Recent crinoid Annacrinus by Roux (19746, pi. 1, figs. 6-7).
An unusual feature of the Scottish brachials is the presence of the three canals interpreted here as
aboral nerve canals. In Recent crinoids there is only one canal in the brachials; Lane and Macurda
(1975) showed that in Aesiocrinus a ‘double-barelled’ nerve canal was present. We have found the
‘double-barelled’ arrangement in a number of different inadunate brachials, but the triple canal has
only been found in the ossicles described here. We are unable to identify the brachials. They clearly
were from a cladid inadunate. We have recovered axillary brachials, which show that the rays were
branched and that the first dichotomy was above the first primibrachial.

Flexible brachial { TCD 19870-3) (PI. 95, figs. 2, 4)

Brachials of this kind are moderately abundant in the sample, but there is considerable variation in
the ratio of width to height; possibly more than one crinoid species is represented. The proximal
articular surfaces are extended aborally into patelloid processes and the distal surfaces each have a
fossa into which the process fits. In the specimen illustrated (PI. 95, fig. 2), the lateral margins of the
articular surface are crenulate with steep-sided culmina approximately 100 pm high. There is no
fulcral ridge, but a fulcral ridge is present on some larger specimens. The stereom of the articular
surface occurs in three different arrays. Over most of the surface, excluding the area around the
patelloid process and a narrow median area extending aborally from the adoral groove, the pores are
quadrangular to round, the diameters of 13-20 pm, and the porosity is approximately 45%. The
trabeculae on either side of the aboral/adoral axis of the brachial are oriented at similar angles to the
median line and produce a markedly rectilinear pore pattern. The stereom pores visible in side view
are approximately the same dimensions as on the articular surfaces, so that the trabeculae form
a regular cubic framework. In the median region, aboral of the adoral groove, the pores are slightly
reduced in size and the regular arrangement of the pores is lost. On the aboral part of the articular
surface around the patelloid process, the stereom is much denser (pore diameter 5-8 pm; porosity less
than 30%) and less regularly arranged.

EXPLANATION OF PLATE 95

Figs. 1, 3. Fluoridized inadunate crinoid brachial (TCD 19867), from the shale above the Charlestown Main
Limestone, St. Monance, Fife (Lower Carboniferous; Brigantian age). 1, slightly oblique view of proximal
articular surface, with pinnule facet to the left, x 48. 3, stereopair of the left side ‘rabbit ear’ fossa and
adjoining interarticular ligament fossa, x 300.

Figs. 2, 4. Fluoridized flexible brachial (TCD 19870), from the shale above the Charlestown Main Limestone, St.
Monance, Fife (Lower Carboniferous; Brigantian age). 2, view of the proximal articular surface, x48. 4,
stereopair of culmina, located near the middle of the left margin in fig. 1, x 180.

PLATE 95

SEVASTOPULO and KEEGAN, Stereom structure of fossil crinoids

756

PALAEONTOLOGY, VOLUME 23

These flexible brachials cannot be more closely identified; all the Carboniferous flexible ossicles
encountered in this study have had remarkably similar microstructure.

Most authors (for instance. Van Sant and Lane 1964, p. 51) have suggested that flexible crinoids
had only ligamentary articulations. Whether ligament fibres penetrated all the ‘cubic’ structured
stereom or were restricted to certain areas is not certain, but the former arrangement seems more
likely.

Acknowledgements. This study was completed when one of us (G. D. S.) was visiting the Department of Geology,
Indiana University, Bloomington, Indiana, U.S.A. We thank Haydn Murray, Chairman, for providing
facilities, and members of the department for discussion and other help. N. Gary Lane kindly read the
manuscript which was typed by Sandy K. Douthitt. Herschel Lentz, Department of Biology, Indiana University,
helped with the scanning electron microscopy. We thank R. B. Wilson, Institute of Geological Sciences,
Edinburgh, for information regarding the St. Monance locality. The study arose out of a micropalaeontological
project supported by the Irish National Board for Science and Technology.

REFERENCES

cookson, I. c. and singleton, o. p. 1954. The preparation of translucent fossils by treatment with hydrofluoric
acid. Geol. Soc. of Australia News Bulletin, 2, 1 -2.

GEORGE, T. N., JOHNSON, G. A. L., MITCHELL, M., PRENTICE, J. E., RAMSBOTTOM, W. H. C., SEVASTOPULO, G. D. and

wilson, R. b. 1976. A correlation of Dinantian rocks in the British Isles. Geol. Soc. Lond. Special Report No. 7,
87 pp.

Grayson, J. F. 1956. The conversion of calcite to fluorite. Micropaleontology, 2, 71-78.

lane, n. G. and macurda, D. B., Jun. 1975. New evidence for muscular articulations in Paleozoic crinoids.
Paleobiology, 1, 59-62.

— and sevastopulo, G. D. (in press). Functional morphology of a microcrinoid: Kallimorphocrinus punctatus
n. sp. J. Paleont.

lapham, K. e., ausich, w. i. and lane, N. G. 1976. A technique for developing the stereom of fossil crinoid
ossicles. Ibid. 50, 245-248.

macurda, d. b., Jun. and meyer, d. l. 1975. The microstructure of the crinoid endoskeleton. Paleont. Contrib.
Univ. Kansas, 74, 1-22, pis. 1-30.

— 1976. The morphology and life habits of the abyssal crinoid Bathycrinus aldrichianus Wyville Thomson and
its palaeontological implications. J. Paleont. 50, 647-667, pis. 1-5.

roux, m. 1970. Introduction a l’etude des microstructures des tiges de crino'ides. Geobios. 3, 79-98, pis. 14-16.

— 1971. Recherches sur la microstructure des pedoncules de crinoides post-Paleozoiques. Univ. Paris, Fac.
Sci. Orsay, Trav. Lab. Paleontol., 83 pp., 4 pi.

— 1974a. Les principaux modes d’articulation des ossicules du squelette des Crinoides pedoncules actuel.
Observation microstructurales et consequences pour f interpretation des fossiles. Acad. Sci. Paris, Comptes
Rendus, 278(D), 2015-2018.

— 19746. Observations au microscope electronique a balayage de quelque articulations entre les ossicules du
squelette des Crinoides pedoncules actuels (Bathycrinidae et Isocrinina). Univ. Paris, Fac. Sci. Orsay, Trav.
Lab. Paleontol., 9 pp., 4 pi.

— 1975. Microstructural analysis of the crinoid stem. Paleont. Contrib. Univ. Kansas, 75, 1-7, pis. 1-2.
sohn, i. G. 1956. The transformation of opaque calcium carbonate to translucent calcium fluoride in fossil

Ostracoda. /. Paleont. 30, 113-114, pi. 25.

sprinkle, j. and gutschick, r. c. 1967. Costatoblastus, a channel fill blastoid from the Sappington Formation of
Montana. Ibid. 41, 385-402, pi. 45.

ubaghs, G. 1978. Skeletal morphology of fossil crinoids. In moore, r. c. and teichert, c. (eds.). Treatise on
invertebrate paleontology. New York and Lawrence, Geol. Soc. Am., pt. T, Echinodermata 2, 1, T58-T216.
upshaw, c. F., todd, R. G. and allen, b. d. 1957. Fluoridization of microfossils. ./. Paleont. 31, 793-795, pi. 100.
van sant, J. F. and lane, N. G. 1964. Crawfordsville (Indiana) crinoid studies. Univ. Kansas, Paleont. Contrib.,
Echinodermata Art. 7, pp. 1-136, pis. 1-8.

wetzel, w. 1921 . Darstellung von Flusspat bei Zimmertemperatur. Centralbl.f. Min., Geol. u. Palaont. 444-447.
zingula, R. p. 1968. A new breakthrough in sample washing. J. Paleont. 42, 1092.

Typescript received 12 September 1979

G. D. SEVASTOPULO, J. B. KEEGAN
Department of Geology
Trinity College, Dublin 2

THE VALUE OF OUTLINE PROCESSING IN THE
BIOMETRY AND SYSTEMATICS OF FOSSILS

by G. H. SCOTT

Abstract. Widespread use of gross dimensions and similar point-to-point measurements in biometric studies of
fossils is probably due more to instrumental limitations and the influence of preceding studies than to theoretical
considerations. Are such data suitable for classificatory studies which are heavily dependent on visual
assessment of morphology? Theory suggests that the outlines of objects are particularly significant in visual
recognition because of their high information content. They provide a parsimonious description of form.
Biometry can best supplement qualitative visual processes in taxonomic studies by treating outline data in ways
that replace the information lost due to the short-term, degradable nature of visual data stored in the human
memory. Variation in the axial outlines of the foraminifer Globorotalia puncticulata (Deshayes) is examined as
an example.

Data collection is fundamental to biometry. Nevertheless, textbooks concentrate on techniques of
data reduction and analysis, and offer little guidance about the collection of data. Such limited
reference is understandable. Organisms are exceedingly diverse in form and organization. Guidelines
for the collection of quantitative data can be cited (e.g. Simpson, Roe, and Lewontin 1960) but
concepts such as ‘character’ and ‘variable’ are so context-dependent that most writers seem to
concede, at least implicitly, that their selection in biometric studies should be left to the discretion of
the student. While the literature indicates that there is considerable accord among researchers on
protozoans to vertebrates on the types of data to be collected, this does not necessarily signify
adherence to a common rationale of data collection. Precedents and instrumental constraints exert
powerful influences on the data collected in a project. Here I consider the role of biometry in
classificatory studies (broadly, recognition of taxa and allocation of specimens) in the light of theory
on the mechanisms of visual perception. It is advocated that biometry should supplement these
mechanisms by processing comparable data so that there is a parallel between qualitative and
quantitative treatments of specimens. In this way biometry can contribute to resolving the problems
of the systematist that arise from deficiencies in visual recognition.

TYPICAL PRACTICE

While it is not claimed that the measurements illustrated in text-fig. 1 portray all aspects of modern
practice in variate selection, they are sufficiently representative to indicate that biometric studies
primarily use data on the gross dimensions of structures. Point-to-point measurements of maximum
dimensions of skeletal parts form the vast majority of the data reported in the literature, and the
example of the measurement of the length of a curve (text-fig. lc) is unusual.

Instrumentation, operational convenience, and the precedents set by previous studies account for
the preference for gross dimensions. The first two, in conjunction, are fundamental. Operationally,
gross dimensions offer considerable advantages in variate selection. Much of the form of skeletal
structures consists of smooth, continuously curved surfaces. In such regions well-defined, relocatable
loci for measurement may be few, and the obvious ‘landmarks’ for the biometrician are the
extremities of the structure. Usually these are homologous within the population sampled. The
simple scales and calipers which are the stock in trade of the palaeontologist are well suited to
measurements of gross dimensions, whereas they are unsuited to determining the lengths of vectors or
curves, for example. Indeed, the widespread use of gross dimensions and of measurements between

IPalaeontology, Vol. 23, Part 4, 1980, pp. 757-768.]

758

PALAEONTOLOGY, VOLUME 23

well-defined ‘landmarks’ in biometry is probably due as much to the limitations of instruments as to
their value on purely biological grounds in morphological description and analysis.

Precedent is an ancillary influence that tends to stabilize the set of characters measured and
perhaps inhibits fresh consideration of what should be measured. Moreover, pioneering works that
use point-to-point measurements have an advantage in the selection of precedents because of the
general availability of comparable measurement devices. The measurements (partly shown in text-fig.
1a) on trilobites made by Shaw (1957) are a good example of the influence on later workers (Temple
1975) of a pioneering study.

A BASIS FOR BIOMETRY

There is very little evidence in the literature that theoretical considerations have influenced the choice
of characters for measurement. In an introduction to a major biometric study of Ostracoda, Reyment
(1963) asserted that statistical analysis would provide a comprehensive representation of variation
but made no comment on the adequacy of the measurements (carapace length, height, and breadth)
that formed the great majority of his data. It is conjectural whether the claim by Hallam and Gould
(1975, p. 517) that their nine measurements on the left valve of Gryphaea are ‘adequate to express
overall features of valve shape and the character of the sulcus’ can be substantiated. Most workers
(myself included) can be easily pilloried on the grounds of ad hoc selection of data without
justification. A relevant -example is Melville’s (1978) critique of a biometric study of leaf shape in
Ulmus. Which features should be selected for measurement?

The choice of measurements and their analysis should relate to the aims and methods of the
investigation. This is self-evident in an application of biometry to a study of functional morphology,
for example, where mechanical hypotheses are presented for testing. But it is a useful point of
departure when considering the role of biometry in the generally less-structured tasks of classifica-
tion. Here the primary activities concern the establishment of classes and the allocation of specimens.
The principal problems concern the estimation of intra-group variation and inter-group separation
or distance. Where do class limits fall? Modern evolutionary theory and research provide a cogent
account of the mechanisms of variation. The systematist, however, is presented with the end products
of various genetic, phenotypic, ontogenetic, and diagenetic processes. In a particular instance there
may be very strong reasons, a priori , to suppose that the specimens under systematic scrutiny are
samples from discrete populations. The problem is that of recognition.

Although data on distribution and ecology are significant, the primary information in the
systematics of fossils is morphological, obtained by qualitative visual examination. The immense
production of illustrations of fossils over the last two centuries attests the fundamental importance of
visual representation in systematics. Certainly, the initial phase of simple qualitative visual
assessment is followed by analysis, sometimes using quantitative data, that leads to diagnoses of taxa.
But the latter is a conscious refinement of the initial phase. The brain is an immensely fast and
powerful processor of visual imagery. Visual data are rapidly assembled, images reconstructed and
interpreted. Messages about the identity of specimens are produced almost involuntarily and are the
basis of classificatory work. The process is that used in other visual recognition tasks in day-to-day
experience, although a higher standard of recognition and discrimination is desirable. Form
variation in biological materials is often complex, with major ontogenetic and environmental sources
to be allowed for in taxonomic recognition.

What is the role of biometry in such studies? Should it supplant or supplement qualitative
perception? If only for reasons of instrumentation, the present role must be supplementary. In many
aspects the human visual system is more advanced than any similar device. It is in inter-image
discrimination that the human system is least effective, especially when sample sizes are large,
variation multidimensional, and groups ill defined. Objects are scanned and features of others
recalled in attempts to reach classificatory decisions. Here the static, long-term memories of digital
devices seem to have marked advantages over the human system. Re-recording of image information
to refresh the memory is made unnecessary. Once stored, it remains available for recall and

SCOTT: OUTLINE PROCESSING

759

text-fig. 1. Measurements of structures commonly preserved as fossils. Variate
identifications and scales are omitted in the adaptations, a, non-agnostidean
trilobite cephalon after Shaw (1957, text-fig. 11). b, gastropod, Athleta petrosa
(Conrad), after Fisher, Rodda, and Dietrich (1964, text-fig. 1). c, bivalve,
Gryphaea, after Hallam and Gould (1975, fig. 1). D, pterosaur skull,
Pterodactylus, after Mateer (1976, fig. 1). e, conodont, after Sergeyeva et al.
(1975, fig. 5). F, acritarch, after Sellberg and Kjellstrom (1975, fig. 1). G,
brachiopod, Linnarssonella girtyi Walcott, after Rowell (1966, table 4). H,
ostracod, Bairdia victrix Brady, after Cadot and Kaesler (1973, fig. 2). i,
foraminifer, Globorotalia miozea miozea Finlay, after Scott (1972, text-fig. 2). j,
ammonite, Vascoceras, after Berthou, Brower, and Reyment (1975, fig. c). K,
molar teeth of condylarth mammal, after Olson and Miller (1958, fig. 61). L,
amphibian skull, Trimerorhachis, after Olson (1953, fig. 1).

760

PALAEONTOLOGY, VOLUME 23

reprocessing without degradation. If biometry is to supplement the ‘weak’ points of visual
perception, it follows that it should process the same sort of data. The problem with ad hoc characters
is that they may record aspects of the object that are insignificant in visual processing. How does the
human system function?

Visual perception. Once the preserve of the psychologist, the mechanics of visual perception have
become an interdisciplinary subject because of their relevance in automatic pattern recognition and
allied studies. A comprehensive survey is not attempted, but there is general agreement about the
significance of the outline in object recognition. Gestalt psychologists (e.g. KofFka 1935) con-
centrated on those properties of figures that facilitated their recognition or isolation from
background data. One of their laws of organization drew attention to the importance of closure.
Closed figures tend to be perceived as units more readily than unclosed. From quite different
premises, information theorists showed that much visual data is redundant in recognition processes
because of high correlation among the data received by adjacent visual receptors. Attneave (1954)
gave a simple, convincing, example of this and suggested that early visual processing filters out much
redundant information, leaving a reduced, more economic, description of the data. Redundancy is
high in regions of an object that are homogeneous in some visual property (e.g. colour, texture,
curvature) and low in regions where such properties change rapidly. The margins of an object are
regions where redundancy is particularly low, although zones of uniform slope or curvature along the
margin have higher redundancy than those in which there are rapid changes in direction or slope.
Attneave showed that an object can be recognized readily from a simplified sketch consisting of the
points of maximum curvature of the outline linked by straight lines. Such a result is an explanation of
the verisimilitude achieved so effortlessly by the competent cartoonist or street artist. But it is also
highly suggestive to the biometrician. Marr (1976) suggested that a major element in early visual
processing is the construction of a ‘primal sketch’ from grey-level changes in the receptor data array.
Intensity changes are isolated and used to construct a description of the array. Edges are major
elements in the description.

Commentary. The review indicates the prime importance of outline data in visual recognition. There
will be many examples in which data, highly significant for recognition, lie within the outline. But, in
general, treatment of the outline is a suitable commencement for biometry in classificatory studies.
Measurement loci, as shown in text-fig. 1, show various degrees of compatibility with Attneave’s
interpretation of visual perception. Some are located on outline segments of low curvature to which
the eye gives little attention (e.g. text-fig. 1h, i). Others (e.g. text-fig. If) are on outline segments of
high curvature that are probably significant in object recognition. However, the use made of
measurement loci in most biometrical practice differs considerably from that suggested by the
foregoing theory. Biometricians have recorded distances between loci, whereas theory suggests that it
is the position of loci as well as interloci distances that is important in perception. A vectorial
approach is indicated.

Vectorial data have been collected in previous studies (text-fig. 2), although not as implementa-
tions of the rationale developed above. Examples are Anstey and Delmet (1973) and Cheetham and
Lorenz (1976) on bryozoans, Christopher and Waters (1974) on miospores, Gevirtz (1976) and
Pastiels (1953) on bivalves, Kaesler and Waters (1972) and Margerie (1977) on ostracods, Scott

text-fig. 2. Examples of outline recording, a, ostracod, Eucypris, after Margerie
(1971, fig. d). B, cheilostome bryozoan, after Cheetham and Lorenz (1976, fig. 4).
c, bivalve, Carbonicola, after Pastiels (1953, fig. 4).

SCOTT: OUTLINE PROCESSING

761

(1976) on foraminifera, and Waters (1977) on blastoids. A common aim has been to describe
accurately the form of the specimen outline. Although representative outlines were presented in
several studies, data have not usually been presented in ways that assist in the resolution of taxonomic
problems. For example, assemblies of outlines (pictograms) have, in the light of the previous
discussion, good theoretical support as effective presentations of intra-sample variation. The
problem of specimen organization within the pictogram can be readily resolved if outline coordinates
are available.

TECHNIQUE AND AN EXAMPLE

This section gives some simple representations of outline data that are useful in classificatory studies.

Data capture. Text-fig. 3 summarizes the data logging and processing system. The digitizer attached
to the stereomicroscope (Scott 1975) was built to specification and is suitable for fossils with greatest
diameters between 0 05 mm and 40 mm. It is manually guided (by movement of the travelling head)
and the x, y coordinates of loci selected by the operator are recorded in units of 5-3 jum on paper tape.

Specimens are digitized in a standard orientation. Errors in orientation are minimized when
specimens have two or more structures that are small in relation to the accuracy of the measurement
system and occur in invariant positions. Such structures are seldom available. In the example, the
axial profiles of the shell were recorded with the coiling axis aligned east-west with reference to a
cross-line in the ocular lens. The coiling axis in foraminifera and similar shells is not a physical
structure, but its position can be estimated from the location of the proloculus (initial chamber) and
umbilicus.

TRAVELLING A/D CONVERTER RAW GRAPHICS EDITED PROCESSING

MICROSCOPE P/TAPE ENCODER DATA FILE DISPLAY DATAFILE SOFTWARE

text-fig. 3. Flow diagram of data capture, editing, and processing system. The
equipment includes a custom-built digitizer, Tektronix 4006 graphics display,
Hewlett Packard 7202a graphics plotter, and Hewlett Packard 2100, Burroughs
B6700, and IBM 370/168 processors.

Editing. Errors due to mis-positioning (backlash, parallax, involuntary movement) increase in
importance as the size of the specimen or structure decreases. Graphical editing of the recorded x, y
data is highly desirable. With batch processing much can be done using lineprinter plots and editing
runs, but interactive editing with a graphics terminal is preferable. My equipment displays x, y
coordinates in order of recording and joined by straight lines (the specimen is represented as a
polygon). Coordinates may be inserted or deleted and the figure redisplayed.

Reconstruction. I record about fifty loci approximately equidistant about the periphery of the
specimen. There is no quantitative control over their position relative to the starting-point. Thus the
ith point on one specimen is not necessarily positionally equivalent to the ith point on another.
Another consideration is that only the obviously spurious coordinates can be removed by editing.
A residual of small-scale errors in positioning remains in the data. Smoothing of the data and
interpolation of points at fixed positions about the periphery are performed by fitting a Fourier Series
curve to each specimen. An angular expansion of the radius about the specimen centroid is applied
(Ehrlich and Weinberg 1970). Radii are interpolated at 10° intervals using 15 harmonics. This
produces mild smoothing. Note that this expansion is suitable only for generally convex figures in
which radii are single- valued. All subsequent processing uses the file of interpolated radii.

762

PALAEONTOLOGY, VOLUME 23

Variation in Globorotalia puncticulata sphericomiozea. Referred to this upper Miocene-lower
Pliocene planktonic foraminiferal taxon are New Zealand populations that are intermediate in
morphology and stratigraphic position between Globorotalia miozea conoidea Walters and G.
puncticulata puncticulata (Deshayes). In axial orientation G. miozea conoidea is weakly conical with
the base formed by the flattish spiral of the early whorls and the cone by the ventrally extended
chambers of the last whorl (for terminology see text-fig. 4). The keel at the shell margin is well defined
on the last-formed chamber but is usually buried by secondary calcification on earlier chambers. The
form of the shell in G. puncticulata puncticulata is globose, rather than conical. This is produced by
moderate inflation of chambers. Straight-line segments of the chamber outline are replaced by gentle
curves. There is no keel. At some horizons, some specimens of G. puncticulata sphericomiozea have
the axial form of the ancestral G. miozea conoidea (and its variant G. conomiozea Kennett). Others
anticipate the shape of G. puncticulata puncticulata. Blow (1969 p. 361) suggested that such samples
represented a mixture of two taxa on the hypothesis that keels, once evolved, are thereafter retained
in phylogeny. He rejected the idea of populations in which some specimens possessed a keel and
others did not. Although there is no theoretical support for the permanency of a structure, Blow’s
suggestion about mixed samples warrants study because Kennett (1977) showed that there was
marked deterioration in climate in the New Zealand region in the uppermost Miocene, about the
stratigraphic position of G. puncticulata sphericomiozea. Changes in the distribution of planktonic
taxa in response to shifts in watermasses and the appearance of migrants are to be expected in such a
regime. To assess Blow’s idea, the systematist needs to examine intra-sample variation. Is it
continuous? Can sub-sample clusters be detected? Here, the axial outline of the shell is examined.
This profile provides information on the shape of chambers near the location of the keel at the shell
periphery. The topics considered are the construction of a typical outline, and the pictorial
representation of within-sample variation.

text-fig. 4. Histograms show distributions of radii at 20° intervals about
centroids of fifty specimens of Globorotalia puncticulata sphericomiozea Walters
from P29/f55, Blind River, New Zealand. The polygonal outline is formed from
the mean lengths of radii spaced at 10° intervals.

SCOTT: OUTLINE PROCESSING

763

Outline representations. The distributions of radii (text-fig. 4) about the centroid of the axial outline of
fifty specimens from P29/f55 Blind River (close to sample 32 in Kennett and Watkins 1974), show
some variation in kurtosis but tend to be unimodal. The outline in the centre of text-fig. 4 is drawn
from mean values of the thirty-six radii and reflects common features in the sample outlines shown in
text-fig. 5. Gentle doming in the vicinity of the spire, rapid change in curvature of the outline of the
nth chamber at the site of the keel, and ventral extension of chambers are features of most of the
outlines in text-fig. 5 that are also apparent in the sample mean outline.

-30 | PCA Q 1

text-fig. 5. Plot of sample from P29/f55 (fifty individuals) on two largest
principal component axes (dispersion matrix, thirty-six radii as deviations from
means). PCA 1 and PCA 2 represent 81% and 6% of sample variance. Location
of axial outlines of specimens is related to their position in the plot (objectively
defined pictogram). Dotted lines show three-cluster division of sample using the
non-hierarchical clustering algorithm (sum of squares criterion) in GENSTAT
(statistical package produced by Rothamsted Experimental Station) and dashed
line is the two-cluster partition. This algorithm transfers specimens between
clusters to improve the criterion but a global optimum is not necessarily reached.

However, use of the sample mean outline as a representative form in comparisons among taxa is
contingent on negligible shape change within the sample size range. If allometry is marked, the
sample mean outline may be quite unrepresentative, not corresponding with the shape of any
specimen. Size-related changes in shape complicate taxonomic recognition and may require special
study. Brower and Veinus (1978) discussed an approach suitable for vectorial data. In the example,
mean outlines for five size-defined subsamples (text-fig. 6) are similar, and even specimens from the
extreme size classes show close resemblance, although there is a modest radial extension of the outline

764

PALAEONTOLOGY, VOLUME 23

in the vicinity of the (n-2)th chamber of the largest specimens (text-fig. 6 centre). I conclude that size-
related shape changes within the material do not greatly affect the use of the sample mean outline as a
representative form.

There is a minority of specimens (e.g. 16, 23, 33 in text-fig. 5) in which spiral and ventral segments
of the outline of the nth chamber form a rounded rather than an angular junction (70-90° radii in
text-fig. 4). In this respect they resemble G. puncticulata puncticulata. Do they form an identifiable
subsample? A quantitative or metric version of the pictogram (text-fig. 5), in which outlines are
referred to specimen positions on a principal component plot, shows that such specimens are
scattered through the sample. Thus specimens 12 and 45 lie at opposite ends of the distribution along
PCA 1 which represents much of the intra-sample variation in outline size. PCA 2 reflects variation in
the degree of ventral inflation of the outline. Again, there are specimens (e.g. 12, 16) that show
considerable difference in ventral inflation yet have rounded peripheries.

text-fig. 6. Histogram shows distribution of area enclosed by outlines (axial
profile) of fifty specimens from P29/f55. Area is taken as a natural measure of
size. The superimposed outlines used subsamples based on the histogram
intervals. Outlines at right were formed by ranking the fifty specimens by
area and dividing them into five equal subsamples.

If size can be neglected in a taxonomic judgement it is useful to examine a representation in which it
is held constant (text-fig. 7). Much of the arrangement of text-fig. 5 is preserved but there are several
displacements that clarify shape similarities. For example, specimen 1 (low spire, weak axial
inflation) lies on the periphery of the scatter in text-fig. 7 whereas in text-fig. 5 it lies between
specimens 3 and 13. Specimens 33 and 45 are dissimilar in shape but their common size causes their
close proximity in text-fig. 5. They are widely separated in text-fig. 7. A group of specimens with weak
axial inflation and a slight dome representing the early chambers (e.g. specimens 1 1, 12, 17, 32, 37, 40)
are in closer proximity in text-fig. 7 than in text- fig. 5.

Distinct clusters are not obvious in text-figs. 5 and 7. This impression is supported by the intra-
sample divisions produced by a non-hierarchical clustering algorithm in GENSTAT. Large
specimens are isolated by the procedure using raw data (text-fig. 5, 3-cluster partition) but 2-cluster
partitions using either raw or size-standardized data separate specimens that are similar in shape and
in close proximity in the principal component plots. The partitions are placed in a central location in
the scatter. This results from the fairly uniform distribution of specimens in the hyperspace.

SCOTT: OUTLINE PROCESSING

765

text-fig. 7. Principal component plot of the sample from P29/f55 using thirty-
six radii (as deviations from means) after areas of outlines were standardized.
Radii were incremented/decremented iteratively until the area of each polygonal
outline fell within 5% of an arbitrary constant, close to the mean of the enclosed
area distribution using raw data. Axes PCA 1 and PCA 2 represent 30% and 24%
of sample variance (dispersion matrix). The dashed line is the location of the
two-cluster partition produced by the non-hierarchical clustering algorithm in
GENSTAT (sum of squares criterion). Axial outlines of specimens using
standardized data are arranged according to their locations in the plot.

The data in text-figs. 4-7 indicate that a variable population was sampled, even when size is
eliminated. But the representations show gradations in form and the absence of well-defined
disjunctions in specimen distributions. A connection is not observed between the form of the
periphery of the nth chamber and the gross axial shape of the shell. These results assist the taxonomist
to assess the validity of G. puncticulata sphericomiozea in the light of Blow’s hypothesis. Inter-sample
comparisons may also be useful. Text-fig. 8, for example, indicates the changes in axial form between
G. puncticulata sphericomiozea and G. puncticulata puncticulata much more explicitly than do direct
comparisons of specimen suites. In the latter the outline of the nth chamber about the 30-70° segment
(see text-fig. 4 for locations) is raised relative to the equivalent segment in G. puncticulata
sphericomiozea. This occurs throughout the size range sampled. But in the 1 10-150° segment of the
nth chamber, inflation relative to G. puncticulata sphericomiozea is marked only in larger specimens.
The study of the transformation in shape between the taxa leads to techniques reviewed by Bookstein
(1977).

766

PALAEONTOLOGY, VOLUME 23

text-fig. 8. Inter-sample comparisons of outlines. Location of samples P29/
f55 and P29/f71 in the Blind River sequence, scanning electron micrographs
of random specimens of Globorotalia puncticulata puncticulata (Deshayes)
and G. puncticulata sphericomiozea Walters, and superimposed outlines from
the samples.

CONCLUSION

I do not contend that biometric studies using ad hoc variates should be abandoned. Rather, I suggest
that analyses with these variates usually do not integrate easily with qualitative assessment of form.
Generally, they provide an inadequate representation of the outline and may include measurement
loci not significant in visual recognition. Vector relationships between measurements are entirely
omitted yet are essential in object identification. By processing the coordinates of outlines, a
quantitative study provides information that is easily and directly related to the material posing a
classificatory problem, and amenable to statistical testing. Of course, outline data may also contain
significant functional information. For example, the form of the shell of an infaunal burrower is likely
to show adaptations to the mechanism of movement.

Representation of outlines by polar coordinates requires large sets of data that may cause
housekeeping problems on small computers. There is commonly some redundancy in the variate set
(dispersion matrices less than full rank) and a more parsimonious set is possible. However, the set
provides directly a polygonal representation of form which is easy to manipulate (magnification,
rotation, reflection) and from which image descriptors (Rink 1 976) and ad hoc variates can be derived
readily. The verbal descriptors of Riedel (1978) are less exact and less suitable for simple graphical
reconstructions and manipulations. The techniques of numerical taxonomy and automated
identification (Sneath 1979) usually operate with character states, selected by the investigator, and do
not provide shape representations at the basic population level.

Outlines are rich in information for the taxonomist. That is why they should be used in biometry.
Nevertheless, they are only a point of departure. Systems that process all pictorial information from a
specimen suite in various orientations offer the prospect of much more sophisticated assistance to the
taxonomist.

SCOTT: OUTLINE PROCESSING

767

Acknowledgements. I am grateful to A. H. Cheetham, B. W. Hayward, and N. de B. Hornibrook for reviewing
a draft of this paper.

REFERENCES

anstey, R. l. and delmet, D. A. 1973. Fourier analysis of zooecial shapes in fossil tubular bryozoans. Bull. geol.
Soc. Am. 84, 1753-1764.

attneave, F. 1954. Some informational aspects of visual perception. Psychol. Rev. 61, 183-193.
berthou, P. Y., brower, J. c. and reyment, R. A. 1975. Morphometrical study of Choffat’s vascoceratids from
Portugal. Bull. geol. Instn Univ. Uppsala N.s. 6, 73-83.

blow, w. H. 1969. Late middle Eocene to Recent planktonic foraminiferal biostratigraphy. In bronnimann, p.
and renz, H. H. (eds.). Proceedings of the First International Conference on Planktonic Microfossils, Brill,
Leiden, 1, 199-421.

bookstein, F. L. 1977. The study of shape transformation after D’Arcy Thompson. Math. Biosci. 34, 177-219.
brower, j. c. and veinus, j. 1978. Multivariate analysis of allometry using point coordinates. J. Paleont. 52,
1037-1053.

cadot, h. m. and kaesler, r. l. 1973. Variation of carapace morphology of bairdiacean and cytheracean
Ostracoda. Paleont. Contr. Univ. Kans. 61, 1-10.

cheetham, A. h. and lorenz, d. m. 1976. A vector approach to size and shape comparisons among zooids in
cheilostome bryozoans. Smithson. Contr. Paleobiol. 29, 1-55.

Christopher, r. a. and waters, j. a. 1974. Fourier series as a quantitative descriptor of miospore shape.
J. Paleont. 48, 697-709.

ehrlich, r. and weinberg, b. 1970. An exact method for characterization of grain shape. J. sedim. Petrol. 40,
205-212.

fisher, w. l., rodda, p. u. and dietrich, j. w. 1964. Evolution of the Athleta petrosa stock (Eocene, Gastropoda)
of Texas. Univ. Tex. Pubis 6413, 1-117.

gevirtz, J. L. 1976. Fourier analysis of bivalve outlines: implications on evolution and autecology. J. Int. Ass.
Math. Geol. 8, 151-163.

hallam, A. and gould, s. J. 1975. The evolution of British and American middle and upper Jurassic Gryphaea:
a biometric study. Proc. R. Soc. Lond. B 189, 511-542.

kaesler, r. l. and waters, j. a. 1972. Fourier analysis of the ostracod margin. Bull. geol. Soc. Am. 83,
1169-1178.

kennett, j. p. 1977. Cenozoic evolution of Antarctic glaciation, the circum-Antarctic Ocean, and their impact
on global paleoceanography. J. geophys. Res. 82, 3843-3860.

— and watkins, n. d. 1974. Late Miocene-early Pliocene paleomagnetic stratigraphy, paleoclimatology, and
biostratigraphy in New Zealand. Bull. geol. Soc. Am. 85, 1385-1398.

koffka, K. 1935. Principles of gestalt psychology. Routledge & Kegan Paul, London, 720 pp.
margerie, p. 1971. Essai de ‘quantification’ du contour des ostracodes a l’occasion de la description d’une
nouvelle espece de Cypridinae du Marinesien du bassin de Paris. Revue Micropaleont. 14, 227-234.

— 1977. Complement a deux articles anterieurs (1969-1972) relatifs a l’utilisation de la statistique pour la
description des ostracodes. Ibid. 20, 44-48.

marr, D. 1976. Early processing of visual information. Phil. Trans. R. Soc. Lond. B 275, 483-524.
mateer, N. J. 1976. A statistical study of the genus Pterodactylus. Bull. geol. Instn. Univ. Uppsala, n.s. 6, 97-105.
melville, R. 1978. On the discrimination of species in hybrid swarms with special reference to Ulmus and the
nomenclature of U. minor Mill, and U. carpinifolia Gled. Taxon, 27, 345-351.
olson, E. c. 1953. Integrating factors in amphibian skulls. J. Geol. 61, 557-568.

— and miller, r. l. 1958. Morphological Integration, Univ. Chicago Press, Chicago, 317 pp.

pastiels, A. 1953. Etude biometrique des Anthracosiidae du Westphalien A de la Belgique. Pubis Ass. Etude
Paleont. Stratigr. houill. 16, 1-56.

reyment, R. a. 1963. Notes on the description of post-Paleozoic fossil ostracods. J. Paleont. 37, 682-687.
riedel, w. r. 1978. Systems of morphologic descriptors in paleontology. Ibid. 52, 1-7.

rink, M. 1976. A computerized quantitative image analysis procedure for investigating features and an adapted
image process. J. Microsc. 107, 267-286.

rowell, a. J. 1966. Revision of some Cambrian and Ordovician inarticulate brachiopods. Paleont. Contr. Univ.
Kans. 7, 1-36.

scott, G. h. 1972. The relationship between the Miocene Foraminiferida Globorotalia miozea miozea and G.
praemenardii. Micropaleontology, 18, 81-93.

768

PALAEONTOLOGY, VOLUME 23

SCOTT, G. H. 1975. An automated coordinate recorder for biometry. Lethaia, 8, 49-52.

— 1976. Estimation of ancestry in planktonic foraminifera: Globoquadrina dehiscens. N.Z. J. Geol. Geophys.
19,311-325.

sellberg, B. and kjellstrom, G. 1975. Geometric characterization of acritarchs belonging to the genus
Veryhachium. Neues Jb. Geol. Palaont., Mh. 1975, 310-314.

Sergeyeva, s. p. 1975. Orientation, morphological terminology and measurements of simple conodonts. Paleont.
J. 9, 385-391.

shaw, a. b. 1957. Quantitative trilobite studies II. Measurement of the dorsal shell of non-agnostidean trilobites.
J. Paleont. 31, 193-207.

SIMPSON, G. G., ROE, a. and LEWONTIN, R. c. 1960. Quantitative Zoology, Harcourt Brace, New York, 440 pp.
sneath, p. h. a. 1979. Numerical taxonomy and automated identification: some implications for Geology.
Computers and Geosciences, 5, 41-46.

temple, j. t. 1975. Standardization of trilobite orientation and measurement. Fossils and Strata, 4, 461-467.
waters, J. A. 1977. Quantification of shape by use of Fourier analysis: the Mississippian blastoid genus
Pentremites. Paleobiology, 3, 288-299.

Typescript received 13 September 1979
Revised typescript received 26 November 1979

G. H. SCOTT

N.Z. Geological Survey
P.O. Box 30368
Lower Hutt
New Zealand

HI A TELLA — A JURASSIC BIVALVE SQUATTER?

by SIMON R. A. KELLY

Abstract. English late Jurassic (Middle Volgian) Hiatella occur in two habitats; firstly, as simple byssal nestlers
on local hard substrates and, secondly, within Gastrochaenolites- type borings penetrating hard substrates. Most
Hiatella occupy borings that they did not originally construct themselves, although ancestors as well as other
bivalve genera could have been responsible. The morphology of the Mesozoic Hiatella is compared briefly with
modern species which occur around the British Isles and which include both boring and nestling forms.
A sequence of events is postulated for the formation of the Basal phosphatized Nodule Bed of the Spilsby
Sandstone in Lincolnshire, and a palaeoenvironmental model is suggested for the East Midlands Shelf in Middle
Volgian times.

The borings made by bivalves into hard substrates have been the subject of considerable attention
from both zoologists and palaeontologists and there are many important articles in the publications
edited by Clapp and Kenk (1963), Crimes and Harper (1970, 1977), and Frey (1975). Unlike most
trace fossils, borings of bivalves may commonly contain the skeletal remains of their occupants.
However, caution is necessary in recognizing whether the occupant is primary, i.e. the organism
which originally constructed the boring, or whether it is secondary and is effectively a squatter in the
vacated domicile. There is ample evidence of modern bivalves reoccupying vacant borings, largely
those of pholads, but including (updated names) Tresus, Petricola , Macoma, and Irus (Evans 1967);
Kellia and Notirus (Stevenson 1946); Tapes , Cumingia , Kelli a, Diplodontct , Endodesma , and My ti fits
(Barrows 1917); Modiola, Scaphula, and Corbula in Mar tesla borings in brickwork (Annandale
1923); Idasola in borings of Teredo in wood (Jensen 1912). Kiihnelt (1933, 1951) recorded Ungulina,
Montacuta, Lepton , Coralliophaga, Trapezium , Venerupis , Sphenia, Perna, Lyonsia , Petricola , and
Hiatella , all of which are deformed to some degree to fit the borings in which they occur. Some
bivalves like Hiatella and Petricola (Yonge 1958; Hunter 1949) may either bore into hard substrates
or nestle epibyssally. Bivalve borings in turn may be reinfested by other phyla, e.g. hydroids and
bryozoa described by Evans (1949), surviving in the wet microenvironments of the vacant borings in
the intertidal zone. Warme (1970) noted that abandoned borings may be modified and deepened by
nestling bivalves, gastropods, polychaetes, arthropods, etc.

Records of fossil bivalves reoccupying vacant borings are much less common. Masuda (1968)
noted Barbatia, Irus, and Phlyctiderma in partially eroded Miocene borings. Itoigawa (1963)
recorded the borings of Miocene Parapholas which were subsequently infilled by sediment, and then
burrowed by Lutraria before consolidation. Kennedy and Klinger (1972) discussed a number of
encrusting and nestling organisms occupying borings constructed by a Cretaceous mytilid; these
include serpulids, a bryozoan, ostreids, and Barbatia. Jurassic Hiatella has been recognized only
rarely. Eudes-Deslonchamps (1838) ascribed two species from the Middle Jurassic of Normandy to
Saxicava (a junior synonym of Hiatella), and Chavan (1952) introduced the genus Pseudosaxicava
for a Lower Kimmeridgian species from the same area, and this name is placed as a subgenus of
Hiatella by Keen (in Moore 1969). From England, Cox (1929) described ‘ Area ' foetida from the
Portland Sand and Hartwell Clay (Middle Volgian). This latter species is conspecific with other
material described here from the Middle Volgian. The updated name of this species is Hiatella
(Pseudosaxicava) foetida (Cox 1929).

There has been little ecological information associated with these early records, though Eudes-
Deslongchamps noted that his Middle Jurassic examples were associated with borings into corals and
bivalve shells. The description here of specimens from the English Middle Volgian adds significantly

[Palaeontology, Vol. 23, Part 4, 1980, pp. 769-781, pi. 96.|

770

PALAEONTOLOGY, VOLUME 23

to the paleoecology of Hiatella. There is evidence that the shell shape is strongly controlled by the
substrate to which it is attached. There is little positive evidence for English Upper Jurassic Hiatella
having been capable of boring, while there is plenty of evidence which indicates that vacant bivalve
borings were commonly infested by Hiatella spat. Modern British Hiatella have been studied by
Hunter (1949), who described considerable variation in shell shape which is closely paralleled by the
late Jurassic forms, depending largely on whether they are boring or nestling. Strauch (1968)
suggested that the shell length of Recent Hiatella was inversely related to the winter minimum water
temperature and consequently was useful in estimation of Cenozoic palaeotemperatures. However,
this is partially doubted by Rowland and Hopkins (1971) who believe that there is a more complex
situation and that size is controlled more by mode of life in each population.

STRATIGRAPHY

The specimens used in this study are all from the Middle Volgian (equivalent to the upper part of the
Upper Kimmeridgian and the lower part of the Portlandian of England). Extensive collecting was
carried out in the Basal Spilsby Nodule Bed in a sand pit, now bulldozed, on Nettleton Hill,
Lincolnshire (TF 108989) (see text-fig. 1 for localities). Although in situ collecting is no longer
possible at this site, the hillside about 200 m to the north provides much weathered-out loose material
from the same horizon. The collections made from this horizon have been deposited with the Institute
of Geological Sciences, London (IGS). Casey (1973) referred this bed to the Titanites giganteus Zone.
The status of this zone in Lincolnshire is not clear since Wimbledon and Cope (1978) have completely
revised the zonal sequence in southern England. However, it is possible that the fauna of this bed may
represent several zones as repeated phases of phosphatization can be recognized and the ammonites
(all phosphatized) belong to the genera Crendonites, Epilaugeites, Kerberites, and Pavlovia (R. Casey
pers. comm.). The Basal Spilsby Nodule Bed rests upon eroded, plastic blue-grey Kimmeridge Clay
with occasional cementstones up to 0-2 m thick and containing Pectinatites of Lower Volgian age.
The nodule bed itself is about 0-2 m thick and is composed of brown and blackened phosphatized
concretions up to 0-2 m in diameter, but commonly 10-30 mm, together with small lyditic pebbles set
in a dark, glauconitic silty sand. Many of the concretions show compound structure and are
commonly abraded, showing signs of bioerosion, e.g. flask-shaped borings attributable to bivalves

text-fig. 1 . Sketch map of the distribution of
Middle Volgian strata in England, with loca-
tions of sites where Hiatella has been obtained.

KELLY: JURASSIC BIVALVE SQUATTER?

77:

and grazing trails probably caused by gastropods. A rich fauna, especially of bivalves, has been
obtained from this bed (Kelly 1977). The preservation of the fauna is normally as hollow
phosphatized moulds, with internal moulds of bivalves and of parts of ammonites making up a high
proportion of the nodules of the bed. Above the nodule bed lies 0-6 m of poorly consolidated
glauconitic silty sand, the base of which is pale coloured, becoming brown (ferruginous) near the
centre and grey at the top, and which contains unidentified, partly phosphatized, ammonites.

Similar phosphatized material with Hiatella occurs in the base of the Lower Greensand at Upware,
Potton, and Brickhill, and is preserved in the Sedgwick Museum, Cambridge. Although these
specimens are mixed with other phosphatized material ranging from Oxfordian to Aptian in age, they
occur with ammonites, a large proportion of which are of Middle Volgian age and they are
undoubtedly of the same age. Unphosphatized Hiatella occur in the Hartwell Clay of Buckingham-
shire and the Swindon Clay of Wiltshire (both of Pavlovia pallasioides Zone) and are preserved in the
British Museum (Natural History), the Institute of Geological Sciences, and the Sedgwick Museum,
Cambridge (e.g. PI. 96, figs. 15, 16). From an unspecified horizon in the Portland Sand of Hounstout,
Dorset (Waddington Collection, untraced), two specimens were figured as ‘ Area' foetida sp. nov. by
Cox (1929, pi. 1, figs. 2, 3). These specimens are likely to have come from the horizons recorded
by Arkell (1935, p. 310), who listed Parallelodon (Beshausenia) foetidum from the White Cementstone
and Bed 1 1 of the Emmit Hill Marls. In the latter horizon Arkell noted that another more elongate
species of the genus was also present. Hiatella has also been collected recently from borings in the
upper part of the Portland Limestone on the Isle of Portland. It is interesting to note that Woodward
(1851-1856) recorded modern Hiatella actively attacking the Portland Stone breakwater at
Plymouth, which perhaps even makes possible the reoccupation of Jurassic borings after some 135
million years.

DESCRIPTION OF BORINGS AND THE HIATELLA

In the Basal Spilsby Nodule Bed, Hiatella was collected both from within flask-shaped borings and
independently of the borings. These two types appear to be morphologically distinct and are
therefore described separately, although it is possible to find intermediate forms. As much of the
discussion in this paper centres around the occurrence of Hiatella in the borings, these structures are
described first, followed by details of the shell shape in both the boring and the non-boring habitat.

The borings. The Basal Spilsby nodules contain several types of borings of which the most
conspicuous are flask-shaped cavities or their phosphatized infillings, commonly up to 30 mm in
length (text-fig. 2 a-c). The flask is circular in cross-section (text-fig. 3c) with a maximum diameter of
13 mm. The constricted neck reaches 5 mm diameter and is circular except near the aperture, where it
becomes slightly oval and weakly flared (PI. 96, fig. 23). Oblique sections through the flask may be

text-fig. 2. Camera lucida drawings of phosphatized infillings of Gastrochaenolites borings in Basal Spilsby
Nodule Bed, Nettleton, Lincolnshire, S. R. A. Kelly Collection IGS: A, Zu2229; B, Zu2230; C, Zu2228; D,

Zu2231; E, Zu2232.

772

PALAEONTOLOGY, VOLUME 23

text-fig. 3. Camera lucida drawings of polished sections through phosphatized compound
nodules of the Basal Spilsby Nodule Bed showing Gastrochaenolites borings; some (figs. A
and b) show Hiatella in sites within the borings. Nettleton, Lincolnshire. S. R. A. Kelly
Collection, IGS: A, Zu2237; B, Zu2223; C, Zu2238; D, Zu2236.

pear-shaped (text-fig. 3a). A complete longitudinal section through an infilled boring is shown in text-
fig. 3 d. The borings are preserved as hollows penetrating the already phosphatized nodules. They may
penetrate both nodules and phosphatized matrix alike without break, which indicates that the
substrate was evenly lithified despite an apparent heterogeneous nature. Each phase of phosphatiza-
tion can be distinguished by a darkened outer margin. Absence of crushed or distorted borings also
shows that the substrate was completely lithified. Borings may not be perfectly straight but may have
bent necks (text-fig. 2b). These are presumably due to the original boring organism modifying the
direction of boring because of unsuitable substrate or of crowding by other individuals.
Interpenetrating borings also occur (text-figs. 2c, e). The first-formed boring appears to be infilled
and phosphatized before being cut across by a second boring.

Although the substrate of these borings is normally a phosphatized nodule, one particular example
shows a large piece of reptilian bone which has been attacked. The upper surface of the bone (PI. 96,
fig. 23) shows that little erosion has taken place since the original construction of the borings as the
openings are still oval. The whole flasks can be seen in Plate 96, fig. 24, together with a specimen of

KELLY: JURASSIC BIVALVE SQUATTER?

773

Hiatella in situ in one of them. The bone must have been buried before abrasion destroyed the oval
necks of the borings. Another specimen, not figured, shows a boring penetrating an icthyosaurian
vertebra. In contrast, Plate 96, fig. 18 shows part of a phosphatized nodule that has been bored and
subsequently abraded so deeply prior to final burial that only rounded bases of the deepest part of the
borings remain visible. The borings are normally found penetrating nodules; however, during phases
of reworking the nodules may become broken and the lithified boring infillings become loose. Such
infillings may be found reworked into the sediment as clasts in the manner described by Radwanski
(1977).

These borings correspond closely to the ichnogenus Gastrochaenolites Leymerie (1842), originally
described from the Calcaire a Spatangues, Neocomian, Aube, France. This name was not included in
the Treatise (Hantzschel 1975). Leymerie clearly described Gastrochaenolites as a boring in rock
which was found in association with Gastrochaena dilatata Deshayes. It is distinguished from
Teredolites Leymerie (1842) which penetrated wood and is more evenly tapered along its length.
Bromley (1972) placed both Gastrochaenolites and Teredolites with the more recent ichnotaxon
Trypanites Magdefrau (1932), which Hantzschel (1975) restricted to straight-sided tunnels of 1 -2 mm
width. The ichnogenus Gastrochaenolites is retained here for the Basal Spilsby Nodule Bed borings
until the taxonomy of these ichnogenera is clarified.

Evans (1970) showed that with increasing rock hardness the ratio of the valve length to valve
depth decreased for Penitella, and the weight of a valve of given size increased. As a consequence, the
shape of the boring also changed, becoming shorter and broader with increased hardness. It has not
yet been possible to compare in detail the borings containing Hiatella from the Portland Stone in
southern England, and therefore varied substrates cannot be compared to show whether the hardness
of the substrate affected the shape of the boring. There is also the problem of establishing without
doubt the original constructor of the boring and if several different bivalves are constructing the
borings they may each have distinctive sized and shaped borings.

Hiatella in the borings. Specimens of Hiatella found inside Gastrochaenolites borings in the Basal
Spilsby Nodule Bed range up to 12 mm in length. They are preserved as internal and external moulds
in phosphorite. The specimens illustrated on Plate 96 are largely casts made from silicone rubber. The
distinctive features of these specimens are: the tendency of the two carinae bounding the dorsal and

text-fig. 4. Sketches of Recent and Jurassic Hiatella valves to illustrate shell form in boring and non-boring
habit, a, Hiatella ( Hiatella ) arctica (Linne), non-boring habitat. Recent (after Hunter 1949); h, H. (H.) gallicana
(Lamarck), boring habitat, Recent (after Hunter 1949); c, H. ( Pseudosaxicava ) foetida (Cox), non-boring
habitat, Middle Volgian; d, H. ( P .) foetida (Cox), boring habitat. Middle Volgian.

774

PALAEONTOLOGY, VOLUME 23

ventral margins of the posterior area to be distinct only close to the umbo and to disappear gradually
towards the posterior margin (PL 96, figs. 1-6, 19; text-fig. 4 d); the posterior area tends to be weakly
inflated and the comarginal ornament is normally suppressed; the umbones are usually low; the
growth-lines may become crowded towards the commissure and there is little trace of median sulcus
on the ventral margin. All these features suggest that the shell may be becoming confined by the shape
of the boring in which it lived. Some specimens, however, are clearly too small to have constructed the
boring (text-fig. 3 a, b; Plate 96, fig. 20) and also two individuals have been found in the same boring
(PI. 96, fig. 22); such specimens lack features which indicate confining by the boring and tend to have
fully developed ornament. These features in general indicate that the Hiatella is infesting borings
which are not of its own making.

Gastrochaena in borings. About 150 specimens of Hiatella have been found in Gastrochaenolites
borings in the Basal Spilsby Nodule Bed. However, one rock specimen has two Gastrochaenolites
borings containing the bivalve Gastrochaena itself (PI. 96, fig. 17) and a single external mould of a
right valve of Gastrochaena shown as a cast in Plate 96, fig. 21 . Recent Gastrochaena sensu stricto is well
known as a borer into calcareous substrates in temperate and tropical regions. It is distinguished
from Hiatella by its large anterior pedal gape and its lack of external ornament like carinae and
lamellae. The borings associated with the Spilsby Gastrochaena fit tightly around the shells and show
weak traces of the calcareous extension tubes, which are not actually seen on any borings associated
with Hiatella. It is not clear whether Gastrochaena was a precursor to the Hiatella in the borings of the
Basal Spilsby Nodule Bed, or whether the two were contemporaneous.

Hiatella independent of borings. The best-preserved examples of Hiatella found independently of the
borings are the aragonitic examples from the Hartwell and Swindon Clays (PI. 96, figs. 15, 16). Such
specimens are normally found as disarticulated valves, while those from the Basal Spilsby Nodule
Bed are normally complete internal phosphatized moulds with valves in occlusion (steinkerns) (PI.
96, figs. 7, 8, 11-14). The independent shells commonly range up to a larger size (30 mm) than those
from the borings. Although the upper length limit of 30 mm is identical to the maximum length of the
borings, the maximum expected size of a Hiatella in a boring would be about 20 mm, because of the
constricted neck area. Presumably the destruction of further large Gastrochaenolites specimens
would provide larger Hiatella than the 12 mm recorded above. The shell is more oval in cross-section;

EXPLANATION OF PLATE 96

Figs. 1-14, 19, 20, 22. Hiatella ( Pseudosaxicava ) foetida (Cox). 1, 2, cast of complete individual, IGS Zu2216,
2217, x 1. 3, 4, cast of complete individual, IGS Zu2219, x 1. 5, 6, cast of incomplete individual, IGS
Zu2222, x 1 . 7,8, phosphatized steinkern, IGS Zu224 1 , x 1 . 9,10, cast of complete individual, IGS Zu22 1 8,
2219, 2220, x 1. 11,12, phosphatized steinkern with cast of some adhering shell, IGS Zu2242, x 1. 13, 14,
phosphatized steinkern, IGS Zu2243, x 1 . 19, phosphatized internal mould completely fitting within boring,
IGS Zu2225, x 1 . 20, cast within boring that is too small to have been made by this occupant, IGS Zu2234,
2235, x 1-5. 22, two phosphatized internal moulds of right valves representing two individuals within the
same boring; Basal Spilsby Nodule Bed, Middle Volgian, Nettleton, Lincolnshire.

Figs. 15, 16. H. ( P .) foetida (Cox). Right valve exterior, IGS Y709, Hudleston Collection, x 1; Upper
Kimmeridge Clay, Pavlovia pallasioides Zone, Middle Volgian, Swindon, Wiltshire.

Figs. 17, 21. Gastrochaena sp. 17, individuals with Gastrochaenolites- type borings, x 1-5. 21, cast (seen as
mould on fig. 17) of left valve, x2. IGS Zu2224. Basal Spilsby Nodule Bed, Middle Volgian, Nettleton,
Lincolnshire.

Figs. 18, 23, 24. Gastrochaenolites ichnosp. 18, eroded flask bases, IGS Zu2226, x 1-5. 23, reptilian bone
showing oval apertures to flask-shaped borings. 24, same specimen in broken section showing opened flasks
and an individual Hiatella steinkern in situ in one, IGS Zu2227, x 2. Basal Spilsby Nodule Bed, Middle
Volgian, Nettleton, Lincolnshire.

PLATE 96

KELLY, Jurassic boring bivalves

776

PALAEONTOLOGY, VOLUME 23

the posterior carinae are distinct throughout their length; the posterior area is gently concave;
comarginal lamellae are well developed on the posterior area, and the ventral margin is usually gently
sulcate, the latter feature giving the byssate shell greater stability in currents (PI. 96, figs. 9, 10; text-
fig. 4c). Unfortunately Oates (1974), in his palaeoecological study of the Hartwell Clay, did not
recognize Hiatella, although the collections he examined do contain them, but they tend to be
confused with species of Grammatodon. I believe that in the Hartwell and Swindon clays both the
Hiatella and Grammatodon are byssate nestlers and not shallow infauna as Oates suggested. Both
these taxa may show a weak byssal gape.

The non-boring Hiatella are believed to have been byssally attached to the exterior of local hard
substrates such as shells of ammonites and phosphatized nodules. Uninhibited growth allowed the
shells to grow to a greater size than in the borings. The large number of complete internal moulds in
the Basal Spilsby Nodule Bed, as opposed to isolated valves, probably reflects rapid burial, with the
shells still attached to the substrate. Early diagenetic phosphatization took place within the reduced
zone defined by the valves. Subsequent winnowing and destruction of the shell concentrated the
internal moulds together with other phosphatized debris. A reconstruction of a Basal Spilsby Nodule
infested with boring and non-boring Hiatella is shown in text-fig. 5.

text-fig. 5. Reconstruction of a Basal Spilsby Sandstone phosphatized nodule, partially cut away to illustrate
Hiatella ( Pseudosaxicava ) foetida (Cox) in its two ecological niches. The smaller, more constricted shelled
specimens occupy the borings, while the larger, more fully developed examples are epibyssally attached to the
exterior of the nodule. For simplification the abundant and varied associated fauna of bivalves, gastropods,
brachiopods, serpulids, etc. are omitted.

KELLY: JURASSIC BIVALVE SQUATTER?

Ill

DISCUSSION

Recent Hiatella are byssally attached to the substrate of their choice, whether living epifaunally on
hard substrates or infaunally in borings. The range in shape of the Jurassic shells is very similar to that
of the recent British species (discussed by Hunter 1949), and there is little reason to suspect that they
lived in different ways.

Hunter (1949) recognized two recent species; the first, H. gallicana (Lamarck) (text-fig. 4b) is
normally found inhabiting borings in calcareous substrates. The shell shows features akin to Hiatella
from borings in the Basal Spilsby Nodule Bed, in particular the suppression of the umbo, posterior
carinae, and lamellae. The species is accepted as a rock borer and is believed to bore with the foot
using sand grains and mucus as an abrasive. There is as yet no positive evidence for any chemical
secretion being used as in the calcium complexing compound discovered in Lithophaga by Jaccarini,
Bannister, and Micallef (1968). There is, however, one significant difference between the Jurassic and
recent species. The Jurassic species had no posterior gape, while the modern species does. H. gallicana
may frequently start its byssal life attached in the opening of an annelid boring (Parfitt 1871), which is
then enlarged and deepened into the substrate. The second species, H. arctica (Linne) (text-fig. 4a) is a
byssal nestler which is not normally associated with borings, but which may fortuitously occur there.
It is commonly found single in association with masses of byssate bivalves like Mytilus, although
Ockelmann (1958) records it occurring as monotypic clusters in Greenland. This species is similar to
the non-boring Jurassic forms described above, but is slightly more elongate and the posterior carinae
have more lamellose tuberculate ornament. H. arctica and H. gallicana are readily distinguished in
the larval stage, but the adult morphologies intergrade because of overlap in habitat, and it is not
always possible to separate them perfectly on features of the hard-part anatomy. There is therefore
little reason to attempt to separate the Yolgian ecomorphs into different species.

Bivalve borings in phosphatized hardgrounds. Although bivalves are commonly found associated with
calcareous substrates, there appear to be relatively few recorded examples of them penetrating
phosphatized hardgrounds. It is clear that the Spilsby nodules were already phosphatized at the time
of attack; any doubts that could be raised can be dispelled by the occurrence of the borings into fossil
bone, which is a primary phosphate. Carcelles (1944) recorded Lithophaga ( Diberus ) penetrating the
plates of Glyptodon and Boreske, Goldberg, and Cameron (1972) reported the occurrence of Miocene
bivalve borings in the bone of Squalodon and attributed them to Parapholas. They also recorded the
occurrence of such borings in mammoth tusks. If these borings in phosphatized substrates are
constructed by mechanical processes there are no problems. However, if chemical techniques are to
be invoked, further research along the lines of Jaccarini et al. (1968) should be investigated.

Environment of deposition of Basal Spilsby Nodule Bed. The Basal Spilsby Nodule Bed formed a
hardground not of a continuous type (e.g. type 2 of Goldring and Kazmierczak (1974, p. 957)), but of
isolated nodules surrounded by glauconitic silty matrix. The nodules show a complex depositional
history and correspond partly to the hiatus concretions of Voigt (1968), whose observations were
based on Liassic calcareous concretions. These calcareous concretions were formed by coalescence of
concretions of different age. The younger concretions envelope the older, the concretions themselves
being of early diagenetic origin. Voigt recognized the following cyclic sequence of events: 1,
formation of concretion; 2, washout; 3, corrosion, boring, and encrustation; 4, burial. The
Cenomanian phosphatized nodules described by Kennedy and Garrison (1975) correspond more
closely to the Spilsby nodules. For discussion of earlier studies on phosphatized horizons of
condensation see Bruckner (1977). Kennedy and Garrison (1975) propose the following sequence for
the formation of nodules that are largely composed of fossil moulds: 1, infilling of shell by sediment;
2, burial; 3, mould cementation (probably by high-magnesian calcite); 4, dissolution of aragonitic
shell; 5, disinterment; and 6, phosphatization, boring, and encrusting. The Basal Spilsby Nodules
appear to have formed under similar conditions, although it is believed here that phosphatization
probably took place at depth in the sediment and not on the surface of the sea floor as Kennedy and
Garrison (1975, p. 357) suggest. It is not possible to see deep burrowing bivalves like Pleuromya and

778

PALAEONTOLOGY, VOLUME 23

Lucina in life position in the Basal Spilsby Nodule Bed, although they are particularly common as
heavily darkened phosphatized internal moulds. However, in the Speeton Clay of the Yorkshire coast
(Lower Cretaceous), deep burrowers such as Thracia and Pleuromya are commonly preserved in life
position as weakly phosphatized, pink or pale-brown internal moulds with some original shell
attached. These have clearly never been exposed on the sea floor; those that have become exposed and
occur in the reworked nodule beds are usually blackened on the exterior and may show signs of
erosion.

Nl U

diagenesis

i- ■ . • ■
* « .

T- -

o

o

rm

5. Sedimentation & 6. Second Phosphatisation 7 Exhumation &

diagenesis reworking

8. Boring

9. Sedimentation, lO. Exhumation & 11. Boring 12. Final burial &

diagenesis & reworking phosphatisation

third phosphatisation

text-fig. 6. Simplified diagrammatic representation of the sequence of events leading to the formation of the

Basal Spilsby Nodule Bed.

KELLY: JURASSIC BIVALVE SQUATTER?

779

The preservation of many fossils in the Basal Spilsby Nodule Bed as phosphatized internal moulds
suggests that the confining shell walls have provided a reduced zone within the sediment. The
phosphatization occurred within this zone and appears first in deep recesses such as the umbonal
infilling in bivalves, and may appear weaker towards the commissure, especially so in forms with
commissural gapes. During a phase of winnowing these moulds would have been condensed and
concentrated in the manner described by Fursich (1978, p. 247). Once the nodules were exposed on
the sea floor they would have been open to attack by boring bivalves and grazing gastropods, etc.,
and available for encrustation by ostreids and Plicatula. During the next phase of burial, the first-
formed concretions would have been bound together by further phosphatization. Repetition of this
sequence would have increased the complexity of formation of these hiatus concretions. So far at
least three phases of phosphatization have been recognized in the Basal Spilsby Nodule Bed, as
illustrated in text-fig. 3 a-d. All these figures show light-coloured but phosphatized areas with
blackened exteriors. These are surrounded by glauconitic sand which in turn is phosphatized. Both
these earlier phases of phosphatization are cut across by borings which have then been filled with
sediment and phosphatized again. The number of phases of boring and phosphatization are likely to
be a conservative estimate, as the largest pieces of the nodule bed are small, with a maximum diameter
of 20 cm. The reconstructed series of events leading to the formation of the Basal Spilsby Nodule Bed
is shown diagrammatically in text-fig. 6.

In modern sediments phosphate formation has been described by Parker (1975) and Mannheim,
Rowe, and Jipa (1975). Parker, working on the Agulas Bank on the south coast of South Africa,
concluded that the area of phosphate formation was estuarine and undergoing regression.
Phosphatization was replacing lime mud matrix of packestones and wackestones, and sometimes
cementing conglomerates of similar reworked sediments. Deeper-water phosphate appeared to be
redeposited from shallow areas. Mannheim et al., working on Holocene sediments from the coast of
Peru, recognized that the calcareous tests of foraminifera were being replaced by phosphate. The
sediments were rich in organic debris, but occurred in an area with a low rate of terrigenous
sedimentation which allowed concentration of the phosphate. The depth at which high-
concentration phosphate occurred is between the shelf break and 1000 m. But the highest
concentration was recorded from a submarine hillock at 144 m.

The Basal Spilsby Nodule Bed is a shelf deposit; although it would be dangerous to suggest an
absolute depth, it appears to be a shallower-water deposit than the preceding Kimmeridge Clay, and
contains a much more diverse benthic macrofauna. There is no evidence for the environment being
estuarine, although it would appear that it occurs in a marine strait that crossed the East Midlands
Shelf in Middle Volgian times. To the south-east it was bounded by the Anglo- Brabant Massif, and to
the north-west by the Pennine Anticline. Despite penecontemporaneous uplift the land must have
had low water runoff and therefore low sedimentation rates in the adjacent sea. Cold currents from
the northern connection to Boreal seas could have provided the high organic content and source of
phosphate. Uplift has probably been caused by movements of axes such as the Market Weighton
structure. The Basal Spilsby Nodule Bed, which is well developed in the north of Lincolnshire,
probably represents a winnowed local topographic high on the East Midlands Shelf. Con-
temporaneous sediments like the glauconitic Hartwell Clay in Buckinghamshire probably represent
what the Basal Spilsby sediment would have been like during a phase of deposition. The Hartwell
Clay-type sediment was probably originally widespread over most of the East Midlands Shelf and
central England, from Swindon to Lincolnshire, at least in early Middle Volgian times. The bulk of it
was destroyed during phases of condensation, leaving only the phosphatized nodules.

CONCLUSIONS

Although it is clear that Hiatella is largely a squatter, reoccupying vacant Gastrochaenolites- type
borings in the Basal Spilsby Nodule Bed, it is still not established that the original borings were made
by Hiatella itself. Gastrochaena was responsible at least for some of the borings, but it seems unlikely
that these were the ones subsequently occupied by Hiatella. Certainly the necks of borings containing

780

PALAEONTOLOGY, VOLUME 23

Hiatella do not show traces of a calcareous extension tube, nor do they have figure-of-eight apertures
which are both features of Gastrochaena borings.

British Middle Volgian Hiatella has two distinctive morphological varieties. One occurs in borings
where shell features are suppressed due to the enclosure of the boring, which it was possibly unable to
modify. A larger, more elongate and fully ornamented form occurs which is not associated with
borings and was probably a simple byssate nestler. As both forms intergrade they probably represent
the same species.

The Basal Spilsby Nodule Bed represents a highly condensed and phosphatized unit once
composed of a Hartwell Clay-type sediment. It probably formed on a topographic high on the East
Midlands Shelf from which fine unlithified sediment was winnowed.

Acknowledgements. Most of this work was carried out at Queen Mary College and the material was collected
with the aid of funds from the Central Research Fund of London University. I am most grateful to the following
for stimulating discussion and for making available for study museum collections in their charge: Dr. R. Casey
and Mr. E. Smith (Institute of Geological Sciences, London), Dr. K. Kleeman (Zoological Institute, Vienna),
Dr. N. Morris (British Museum (Natural History), London), Dr. J. Wilson (Institute of Oceanographic
Sciences, Godaiming), and especially to Dr. P. F. Rawson (Queen Mary College) who critically read an early
draft of this paper.

REFERENCES

annandale, N. 1923. Bivalve molluscs injuring brickwork in the Calcutta docks. J. Proc. Asiatic Soc. Beng. 18,
555-557.

arkell, w. J. 1935. The Portland Beds of the Dorset mainland. Proc. Geol. Ass. 46, 301-347.
barrows, A. L. 1917. Geological significance of fossil rockboring animals. Bull. geol. Soc. Am. 28, 965-972.
boreske, J. r., goldberg, L. and Cameron, b. 1972. A reworked cetacean with clam borings: Miocene of North
Carolina. J. Paleont. 46, 130-139.

bromley, r. G. 1972. On some ichnotaxa in hard substrates with a redefinition of Trypanites Magdefrau.
Palaeont. Z. 46, 93-98.

bruckner, w. D. 1977. Genesis of beds rich in phosphate nodules— a historical note. Lethaia, 10, 144.
carcelles, a. 1944. Catalogo de los moluscos marinos de Puerto Quequen. Revta Mus. La Plata. Seccion
Zoologia, N.s., 3, 233-310.

casey, r. 1973. The ammonite succession at the Jurassic-Cretaceous boundary in eastern England. In CASEY, R.

and rawson, p. f. (eds.), The Boreal Lower Cretaceous. Geol. J. special issue, 5, 193-266.
chavan, a. 1952. Les pelecypodes des Sables Astartiens de Cordebugle (Calvados). Schweiz, palaeont. Abh. 69,
1-132.

clapp, w. f. and kenk, r. 1963. Marine borers, and annotated bibliography, 1136 pp. Office Naval Research
(ACR-74).

COX, L. R. 1929. Synopsis of the lamellibranchia of the Portland Beds of England. Proc. Dorset nat. Hist,
archaeol. Soc. 50, 131-202.

crimes, t. p. and harper, j. c. (eds.). 1970. Trace fossils. Geol. J. special issue, 3, 547 pp.

- 1977. Trace Fossils 2. Geol. J. special issue, 9, 351 pp.

eudes-deslongchamps, j. a. 1838. Memoire sur les coquilles fossiles lithophages des terrains secondaires du
Calvados. Mem. Soc. linn. Normandie, 6, 220-229.

evans, J. w. 1967. Relationship between Penitella penita (Conrad, 1837) and other organisms of the rocky shore.
Veliger, 10, 148-152.

- 1970. Palaeontological implications of a biological study of rock boring clams (Family Pholadidae).
In crimes, t. p. and harper, j. c. (eds.), Trace Fossils. Geol. J. special issue, 3, 127-140.

Evans, R. c. 1949. The intertidal ecology of rocky shores in south Pembrokeshire. J. Ecol. 37, 120-139.
frey, r. w. (ed.). 1975. The study of trace fossils, 562 pp. New York.

fursich, F. T. 1978. The influence of faunal condensation and mixing on the preservation of fossil benthonic
communities. Lethaia, 11, 243-250.

goldring, R. and kazmierczak, J. 1974. Ecological succession in intraformational hardground formation.
Palaeontology, 17, 949-962, pis. 125-126.

hantzschel, w. 1975. Treatise on Invertebrate Paleontology part W, supplement 1, Trace Fossils and
Problematica, 2nd edn., W1-W269. Geological Society of America and University of Kansas Press.

KELLY: JURASSIC BIVALVE SQUATTER?

781

hunter, w. R. 1949. The structure and behaviour of Hiatella gallicana (Lamarck) and H. arctica (Linne) with
special reference to the boring habit. Proc. R. Soc. Edinb., B, 63, 271-289.
itoigawa, j. 1963. Miocene rock- and wood-boring bivalves and their burrows from the Mizunami group,
Central Japan. J. Earth Sci. Nagoya Univ. 11, 101-123.

JACCARINI, v., bannister, w. h. and micallef, H. 1968. The pallial glands and rock boring in Lithophaga
lithophaga (Lamellibranchia, Mytilidae). J. Zool. , Lond. 154, 397-401.
jensen, a. s. 1912. Lamellibranchiata (pt. 1). Dan. Ingolf Exped. 2, 1-1 19.

kelly, s. R. a. 1977. The bivalves of the Spilsby Sandstone Formation and contiguous deposits, 365 pp. Ph.D. thesis
(unpubl.). University of London.

Kennedy, w. j. and garrison, R. E. 1975. Morphology and genesis of nodular phosphates in the Cenomanian
Glauconitic Marl of southeast England. Lethaia, 8, 339-360.

— and klinger, h. c. 1972. Hiatus concretions and hardground horizons in the Cretaceous of Zululand
(South Africa). Palaeontology, 15, 539-549, pis. 106-108, 3 figs.

kuhnelt, w. 1933. Bohrmuschelstudien II. Palaeobiologica, 5, 371-408.

— 1951. Contributions a la connaissance de l’endofaune des sols marins durs. Annee biol. (3), 27, 513-523.
leymerie, m. a. 1842. Suite de memoire sur le Cretace du departement de l’Aube contenant des considerations sur

le terrain Neocomien. Mem. Soc. geol. Fr., 2, (1), 34 pp.
magdefrau, k. 1932. Uber einige Bohrgange aus dem Unteren Muschelkaik von Jena. Paldont. Z. 14, 150-160.
mannheim, f., rowe, G. T. and jipa, d. 1975. Marine phosphorite formation off Peru. J. sedim. Petrol. 45,
243-251.

masuda, K. 1968. Sand-pipes penetrating igneous rocks in the environs of Sendai, Japan. Trans. Proc. palaeont.
Soc. Japan, N.s. 72, 351-362.

moore, r. c. (ed.). 1969. Treatise on Invertebrate Paleontology, part N, volumes 1 and 2, Mollusca, 6, Bivalvia,
951 pp. Geological Society of America and University of Kansas Press.
oates, m. j. 1974. The stratigraphy and palaeoecology of the Hartwell Clay (Upper Kimmeridgian) of Aylesbury,
Buckinghamshire. Proc. Geol. Ass. 85, 367-375.

ockelmann, w. K. 1958. Zoology of East Greenland marine lamellibranchiata, Meddr Gronland, 122 (4), 1-256,
3 pis., 29 figs.

parfitt, E. 1871. On the borings of molluscs, annelids and sponges into rocks, wood and shells. Rep. Trans.
Devon. Ass. Advmt Sci. 4 (2), 456-466.

parker, r. j. 1975. The petrology and origin of some glauco-conglomaratic phosphorites from the South African
Continental Margin. J. sedim. Petrol. 45, 230-240.

RADWANSKi, a. 1977. Present-day types of trace in the Neogene sequence; their problems of nomenclature and
preservation. In crimes, t. p. and harper, j. c. (eds.), Trace fossils 2. Geol. J. special issue, 9, 227-264.
Rowland, R. w. and hopkins, D. M. 197 1 . Comments on the use of Hiatella arctica for determining Cenozoic sea
temperatures. Palaeogeogr. Palaeoclimat. Palaeoecol. 9, 59-64.
stevenson, a. G. 1946. Collecting rock borers in New Zealand. Mollusca, 1 (9), 125.

strauch, f. 1968. Determination of Cenozoic sea temperatures using Hiatella arctica (Linne). Palaeogeogr.
Palaeoclimat. Palaeoecol. 5, 213-233.

voigt, E. 1968. Uber Hiatus-Koncretionen (dargestellt an Beispielen aus dem Lias). Geol. Rdsch. 60, 335-380.
warme, j. e. 1970. Traces and significance of marine rock borers. In crimes, t. p. and harper, j. c. (eds.), Trace
fossils. Geol. J. special issue, 3, 515-525.

Wimbledon, w. a. and cope, J. c. w. 1978. The ammonite faunas of the English Portlandian Beds and the zones
of the Portlandian Stage. Jl geol. Soc. Lond. 135, 183-190.
woodward, s. P. 1851-1856. A manual of the mollusca, 486 pp.

yonge, c. m. 1958. Observations on Petricola carditoides (Conrad). Proc. malac. Soc. Lond. 33, 25-31.

S. R. A. KELLY

Sedgwick Museum
Department of Earth Sciences
Downing Street
Cambridge CB2 3EQ

Manuscript received 19 March 1979
Revised manuscript received 10 December 1979

EVOLUTION OF THE SILURIAN TRILOBITE
TAPINOCALYMENE FROM THE WENLOCK
OF THE WELSH BORDERLANDS

by DEREK J. SIVETER

Abstract. Some calymenid trilobites from the Wenlock Series of the Welsh Borderland are described and
assigned to a new genus Tapinocalymene, type species T. nodulosa (Shirley 1933). An evolutionary lineage from
T. volsoriforma sp. nov. (early Wenlock) through T. vulpecula sp. nov. (late Wenlock) to T. nodulosa (late
Wenlock) is proposed, involving an increase in the length and area of the preglabellar furrow. Tapinocalymene
was probably benthic in habit, and occurs in somewhat offshore, generally deepish water, elastics. Calymene
diademata Barrande, 1846 (Wenlock, Bohemia), C. nasuta Ulrich, 1879 (Llandovery, United States) and C.
blumenbachii Brongniart, 1822 (Wenlock, England), respectively the type species of the calymenines
Diacalymene Kegel, 1927, Spathacalymene Tillman, 1960, and Calymene Brongniart, 1822 are figured and
compared with members of Tapinocalymene. S. nasuta and T. nodulosa both have a particularly long preglabellar
area, but each differs in its form and derivation. Several Ordovician and Silurian calymenid genera, some only
distantly related, evolved a long preglabellar area, the taxonomic value of which should be treated with caution.

Since the pioneer revisions made by Shirley (1933, 1936) on British Silurian calymenid trilobites,
they have occasionally been discussed in studies on non-British faunas (Campbell 1967; Haas 1968;
Schrank 1970; Whittington 19716), but except for the work of Temple (1969, 1970, 1975) on
Llandovery species, no direct attention has been paid to them. This paper is part of a wider project
undertaken by the present author to investigate north-west European Silurian (and Ordovician)
calymenids. The terminology, measurement, photographic and preparation techniques are those of
Siveter (1977, 1979), except that surface sculptural terms are now used in the sense of Miller (1975, pp.
341, 343). Specimens used in this work are housed in the following museums: British Museum
(Natural History), London (BM); Geological Museum, Institute of Geological Sciences, London
(GSM); National Museum of Wales, Cardiff (NMW); Ludlow Museum, Salop (LM); Hunterian
Museum, Glasgow (HM); Naturhistoriska Riksmuseet, Stockholm (RM); National Museum of
Natural History, Smithsonian Institution, Washington (USNM).

SYSTEMATIC PALAEONTOLOGY

Family calymenidae Milne Edwards, 1840
Subfamily calymeninae Milne Edwards, 1840

Discussion. I have advocated (Siveter 1977, p. 353) that this subfamily should contain only those
genera possessing the papillate-buttress structure, but future work may provide exceptions to this
general rule, with the possibility of buttressed forms giving rise to non-buttressed forms through the
arrested development of this feature during ontogeny (Siveter 1979, p. 373).

Genus tapinocalymene gen. nov.

Type species. Calymene nodulosa Shirley, 1933; Wenlock Series, Coalbrookdale Formation, Burrington,
Hereford, and Worcester.

Derivation of name. Greek, tapeinos, humble, alluding to the glabella which is fairly low and short relative to the
fixed cheeks.

[Palaeontology, Vol. 23, Part 4, pp. 783-802, pis. 97-101.1

784

PALAEONTOLOGY, VOLUME 23

Other species. T. volsoriforma sp. nov., T. vulpecula sp. nov.

Diagnosis. A calymenine genus which combines the following characters: Preglabellar area relatively long,
variably formed. Anterior glabellar margin normally lies behind anterior margin of fixed cheek, exceptionally
both margins are in line (tr.); dorsal glabellar surface stands just above, or anteriorly is sometimes slightly below,
fixed cheek. Glabellar lobe 2p is bridged across axial furrow to a genal buttress. Palpebral lobes are from twice to
2-5 times as wide (tr.) apart as glabellar width at 2p lobes. Hypostoma has ventral protuberance on middle of
anterior lobe; maculae well developed. Pygidial axis almost flat (tr.); inner pleural region slopes gently abaxially;
interpleural furrows are weak or obsolete. Numerous small to medium-sized granules uniformly and closely
distributed on glabella; anterior adaxial part of fixed cheek, and often anterior border, coarsely granulate.

Discussion. Shirley (1936) advocated that calymenids showing a ‘ridged’ (or ‘thickened’) preglabellar
area and a papillate 2p glabellar lobe joined to a genal buttress should be placed in his amended
concept of Diacalymene', those species with this bridge across the axial furrow but without the
‘ridged’ preglabellar area should be placed in Calymene. The separation of Diacalymene on this basis
has recently been questioned (Temple 1975; Ingham 1977; McNamara 1979) and certain British
Ashgill and Llandovery species which Shirley placed in that genus are now held in abeyance in
Calymene (sensu lato). At present I prefer to consider the type species of Diacalymene , C. diademata
Barrande, 1846 (late Wenlock age), and also several other related taxa such as D. horbingeri (Snajdr,
1975) (Llandovery of Bohemia), as generically distinct from C. blumenbachii Brongniart, 1822 and
related species. Nevertheless, assigning species to either Diacalymene or Calymene according to the
nature of the preglabellar area was impracticable in the case of Tapinocalymene nodulosa (Shirley,
1933), T. volsoriforma sp. nov. and T. vulpecula sp. nov. These species all have the 2p lobe joined to a
genal buttress, but whereas T. volsoriforma and T. vulpecula show a fairly distinct break in slope in the
preglabellar area, where the anterior side of the preglabellar furrow meets the posterior edge of the
less steeply sloping anterior border (PI. 99, fig. 2; PI. 100, fig. 2), similar to that in certain species
variously referred to Diacalymene and Calymene (s.l.), T. nodulosa (PI. 97, fig. 1 1) has a preglabellar
area similar to, though very much longer than, most Calymene species. All three species share
characters which unite them as a genus distinct from other calymenines. The preglabellar area in
Tapinocalymene is believed to have evolved quite rapidly and is useful for specific discrimination,
though not for diagnosing the genus. None of the generic characters is completely exclusive, but in
particular the low glabella which fails to protrude anteriorly beyond the fixed cheeks (PI. 98, figs. 2, 3;
PI. 99, figs. 1,2; PI. 100, figs. 1, 2), the widely separated palpebral lobes (PI. 98, fig. 6; PI. 99, fig. 13; PI.
100, fig. 4), and the style of cranidial sculpture (PI. 97, fig. 9; PI. 99, fig. 14) all combine to distinguish
Tapinocalymene. Other characteristic features include the axis, interpleural furrows, and inner
pleural region of the pygidium. All Tapinocalymene species are closely associated in time and
space.

Compared with Tapinocalymene , Diacalymene has a more raised, forwardly protruding glabella,
narrowly separated palpebral lobes, and a more pointed inner, anterior corner to the fixed cheek (cf.
PI. 98, figs. 9-11; PI. 101, figs. 5, 6, 10). Furthermore, Diacalymene lacks coarse granules on the
inner part of the fixed cheek, though both genera have small, close-set glabellar granules. ‘C.’
allportiana Salter, 1865 (Much Wenlock Limestone Formation, Dudley) has the same type of
cranidial sculpture and pygidial axis as Tapinocalymene (Shirley 1933, pp. 58, 59, pi. 1, figs. 12-14);
also the separation of its palpebral lobes falls just within the range of variation of the new genus (text-
fig. 1). ‘C.’ allportiana is certainly more closely related to Tapinocalymene and D. diademata than to C.
blumenbachii , but is excluded from Tapinocalymene as presently defined because of its more
anteriorly, and (to a lesser extent) dorsally, projecting glabella. When Diacalymene is fully reassessed
the generic position of ‘C.’ allportiana will become clearer. Most Calymene species differ from those
of Tapinocalymene in the following features: a more dorsally and anteriorly projecting glabella
having more variably sized, often larger, granules; less widely separated palpebral lobes; a steeper
slope to the inner pleural region of the pygidium; a more convex (tr.) pygidial axis; better defined
interpleural furrows, particularly distally. These differences are most obvious in the late Wenlock
species C. aspera Shirley, 1936 and C. blumenbachii (cf. PI. 97, figs. 1,3, 10, 1 1 ; PI. 100, figs. 9-12, 14,
16; text-fig. 1). Features characteristic of Tapinocalymene are occasionally exhibited, or closely

SIVETER: CALYMENID TRILOBITES

785

approached, by Calymene species. For example, C. tuberculosa Dalman, 1827, from the Wenlock of
Gotland (and Wenlock Edge) has a gently convex pygidial axis, and C. tenera Barrande, 1852 from
the Kopanina Formation (Ludlow) of Bohemia has very weak interpleural furrows. The Ludlow
species C. neointermedia R. and E. Richter, 1954 has been allied with T. nodulosa, as both have similar
scoop-like preglabellar areas (Schrank 1970, p. 122; Whittington 19716, p. 463); the same character is
present in C. puellaris Reed, 1920, also of Ludlow age. C. neointermedia and C. puellaris clearly
belong within Calymene. The similarity in the preglabellar area of the three species is believed due to
adaptive convergence.

Tomczykowa (1970) referred Tapinocalymene nodulosa to the monotypic genus Spathacalymene
from the Osgood Formation (upper Llandovery; Berry and Boucot 1970), Indiana, an assignment
made mainly because S. nasuta and T. nodulosa both have a long preglabellar area. However, it is not
unusual for calymenid genera to independently develop a long preglabellar area and, in each genus,
for it to be morphologically different. Compare, for example, that in Thelecalymene mammillata (Hall,
1861; Whittington 1971a, pi. 1, figs. 4, 5; pi. 2, fig. 1) from the upper Ordovician of the United States,
Prionocheilus foveolatus (Tornquist, 1884; Warburg 1925, pi. 4, figs. 13, 16) from the middle
Ordovician of Sweden, Reedocalymene expansa (Yi, 1957; Lu 1975, pi. 46, fig. 4) from the middle

text-fig. 1. Histogram of ratio of width be-
tween palpebral lobes to width of glabella at
lobe 2p (= variates J, and K2 of Siveter 1977,
p. 338, fig. 1). A. Spathacalymene nasuta. B.
Diacalymene diademata. c. ‘'Calymene'
allportiana. D. Calymene blumenbachii. e.
Tapinocalymene: T. volsoriforma, n = 6; T.
vulpecula, n=4; T. nodulosa , n = 10.

Ordovician of China, and Calymenesun tingi (Sun, 1931; Lu 1975, pi. 46, figs. 9-11) from the middle
Ordovician of China. There is no resemblance in the preglabellar area of Spathacalymene nasuta and
Tapinocalymene nodulosa apart from their uncommon length (see below and PI. 98, figs. 6, 9, 12; PI.
101, figs. 1,4, 8). The former differs from the latter in its very convex (sag. and tr.), more dorsally and
anteriorly projecting glabella, narrowly separated palpebral lobes (text-fig. 1), V-shaped rostral
suture, subconical inner anterior corner to the fixed cheek, narrower thoracic and pygidial pleural
region, and the lack of coarse granules on the inner, anterior part of the fixed cheek. I have no doubt
that these species are not congeneric. The three calymenids with a spatulate preglabellar area from the
Ludlow of Poland, which have been named S.flexuosa, S. brevis , and S. linguata by Tomczykowa
(1970), should also be excluded from Spathacalymene. I agree with Whittington (19716, p. 459) that
these species represent a quite separate lineage (non-calymenine; but see Siveter 1979, p. 373).
Papillicalymene Shirley, 1936 from the Ludlow of Gotland and Podolia, and Downton age glacial
erratics of the north German plain, has a very advanced type of genal buttressing (Whittington 19716,
pis. 85, 86) which easily distinguishes it from Tapinocalymene.

Occurrence. Tapinocalymene has a stratigraphic range of early to late Wenlock; that is Sheinwoodian, probably
Cyrtograptus centrifugus or C. murchisoni biozones, to Homerian, Gleedon Chronozone, G. nassa Biozone. It is
limited to the main Wenlock outcrop of the Welsh Borderland from Rushbury, Ape Dale, through the core of the
Ludlow anticline and the Wigmore Rolls area, to Dolyhir, Powys.

786

PALAEONTOLOGY, VOLUME 23

Tapinocalymene nodulosa (Shirley, 1933)

Plate 97, figs. 1-6, 8, 9, 11; Plate 98; text-fig. 2a-f

non 1827 Calymene blumenbachii var a, tuberculosa-, Dalman, p. 227.
v. 1839 Calymene blumenbachii Brongniart; Murchison {pars), p. 653, pi. 7, fig. 5 (GSM 6588),
non figs. 6, 7.

v. 1848 Calymene tuberculosa, Salter; Salter, in Phillips and Salter, p. 342, pi. 12, figs. 1, la (GSM
19642), 2, 3, 5, ?fig. 4.

v. 1849 Calymene tuberculosa Salter; Salter, p. 1, pi. 8, figs. 1, 2 (GSM 19642), 1*, 3-5, 7, ?fig. 6,
non figs. 8*, 8 ( = C. puellaris Reed; GSM 19690).
non 1851 Calymene tuberculosa (Salter); McCoy, in Sedgwick and McCoy, p. 167.

v. 1859 Calymene tuberculosa, Salter; Murchison, pi. 18, fig. 11 (GSM 6588).

v. 1865 Calymene tuberculosa, Salter; Salter, p. 91, pi. 8, figs. 1, 2, 3 (GSM 19642), 4, 5 (GSM
19646), 6.

1873 Calymene tuberculosa, Salter; Salter (pars), p. 133, non p. 166.
v. 1884 Calymene tuberculosa-. La Touche, p. 66, pi. 10, fig. 243 (GSM 6588).

1885 Calymene tuberculosa Salter non Dalman; Lindstrom, p. 66.

1888 Calymene tuberculosa Salter; Etheridge (pars), p. 46.
v. 1919 Calymene blumenbachii Brongniart; Reed, in Garwood and Goodyear, p. 20.

1925 Calymene tuberculosa Salter; Warburg, p. 158.

1927 Calymene (Diacalymene) tuberculosa Salter; Kegel, pp. 618, 620, text-fig. 2 /.
v* 1933 Calymene nodulosa nom. nov.; Shirley, p. 53, pi. 1, figs. 6-11.

1936 Calymene nodulosa Shirley; Shirley, pp. 388, 390, 393, 399, 400, text-figs. 1, 2 (pars).
1938 Calymene nodulosa Shirley 1933; Stubblefield, pp. 37, 38.

? 1953 Calymene nodulosa Shirley; Williams, pp. 199, 200 (specimens not seen),
v? 1968 Calymene nodulosa Shirley; Greig, Wright, Hains, and Mitchell, p. 354 (specimens
inadequate).

1970 Calymene nodulosa Shirley, 1933; Schrank, pp. 115, 122, 123.

1970 Spathacalymene nodulosa (Shirley); Tomczykowa, pp. 63, 70, 72, text-figs. 4 k, 5 /.

1971b Calymene nodulosa Shirley 1933; Whittington, p. 463.

Holotype. Nearly complete specimen lacking preglabellar area, with cuticle removed from abaxial part of cheeks
and abaxial pleural region of thorax, GSM 19642; figured Salter 1848, pi. 12, figs. 1, la; 1849, pi. 8, figs. 1, 2;
1865, pi. 8, figs. 2, 3; Shirley 1933, pi. 1, figs. 6-10; pi. 98, figs. 1-3.

EXPLANATION OF PLATE 97

Figs. 1-6, 8, 9, 11. Tapinocalymene nodulosa (Shirley, 1933). All specimens are from the Wenlock Series,
Homerian Stage, Coalbrookdale Formation, vicinity of Burrington, Hereford and Worcester; 2, 5, 8, 9 come
from C. lundgreni Biozone strata, sunken lane south of Burrington (SO 442 718). 1,3, 11, enrolled specimen
lacking left, and most of the right, free cheeks, HM A212/1, dorsal stereo-pair, frontal view, x 1|, left lateral
view, x 2. 2, slightly distorted cranidium, LM 4885, dorsal view, x 2. 4, incomplete thorax and pygidium,
NMW 75.35G.400, posterior view, x 1-5. 5, incomplete cranidium and rostral plate, NMW 77.31G.10,
ventral view, x2. 6, hypostoma and rostral plate, RM Ar38841, ventral stereo-pair, x3-5. 8, pygidium,
NMW 77.31G.9, posterior view, x 5. 9, cranidium, LM 4902, dorsal view, x 4.

Fig. 7. Tapinocalymene nodulosa ? (Shirley, 1933). Pygidium, NMW 77.31G.il, Wenlock Series, Homerian
Stage, Coalbrookdale Formation, base of Farley Member, G. nassa Biozone, track section 252 m at 82° from
St. Edith’s Church, Eaton, Ape Dale, Salop (SO 5023 9002; Bassett et al. 1975, loc. 25, p. 16); posterior view,
x 1-5.

Fig. 10. Calymene blumenbachii blumenbachii Brongniart, 1822. Complete enrolled specimen, GSM 19668,
Wenlock Series, Homerian Stage, Much Wenlock Limestone Formation, Dudley, West Midlands; dorsal
view, x2 figured Shirley 1933, pi. 1, figs. 4, 5.

PLATE 97

SIVETER, Calymenid trilobites

788

PALAEONTOLOGY, VOLUME 23

Type locality. Wenlock Series, Homerian Stage, Coalbrookdale Formation, Burrington, Hereford and
Worcester. The Coalbrookdale Formation in the vicinity of Burrington includes strata of the Cyrtograptus
lundgreni, G. nassa, and Monograptus ludensis biozones (Holland, Rickards, and Warren 1969). It is perhaps
most likely that lundgreni Biozone strata yielded the holotype, and I have collected T. nodulosa from this horizon
in the sunken lanes south of Burrington church.

Additional material and occurrences. At least 6 nearly complete individuals, 20 cranidia, 20 pygidia, 5
hypostomata. I have noted material in the British Museum (Natural History); Museum of the Institute of
Geological Sciences, London; National Museum of Wales, Cardiff; Hunterian Museum, Glasgow; Ludlow
Museum; Naturhistoriska Riksmuseet, Stockholm. The species occurs in the Coalbrookdale Formation of the
following localities: Calcareous concretions within the small faulted patch of shales ( = within the C. rigidus to
C. lundgreni biozones; Bassett 1974, p. 759) above the Dolyhir Limestone, quarry ‘D’ of Garwood and Goodyear
(1919, pi. 5, fig. 1, pi. 7), Dolyhir, Powys (SO 2412 5805); Birtley Lane, 6-5 km south-south-west of
Leintwardine, Hereford and Worcester (SO 3687 6888); Homerian Stage, C. lundgreni Biozone, track section at
St. Edith’s Church, Eaton, Ape Dale, Salop (SO 5001 9002). A pygidium of Tapinocalymene (PI. 97, fig. 7) from
the base of the Farley Member, Coalbrookdale Formation, the track section at Eaton (SO 5023 9002), may also
belong to T. nodulosa. Greig et al. (1968, p. 354) list the species from nearby Rushbury, Ape Dale (C. lundgreni
Biozone; Bassett et al. 1975, p. 16, fig. 2); these specimens belong to Tapinocalymene but a specific assignment
cannot be made with certainty. I cannot confirm Shirley’s (1933, p. 56) record of Wenlock-age specimens from
a quarry beside Nant Tresglen, behind Halfway Inn, 8 km east of Llandovery (SN 828 328). Williams (1953, pp.
199, 200) cited T. nodulosa from his ‘Lower’ and ‘Upper’ Wenlock groups of the Llandeilo district. This material
has not been seen but the C. cf. nodulosa figured (White, in Squirrell and White 1978, pi. 3, figs. 3, 4) from the
Wenlock of the Cennen Valley near Llandeilo is not close to T. nodulosa and does not appear to be congeneric.

Diagnosis. Preglabellar area is from about two-fifths to almost one-half as long as glabella, directed forward, and
curving progressively more steeply upward. Preglabellar furrow about three-fifths to three-quarters as long as
preglabellar area; transition in slope between steep anterior side of this furrow and relatively short (sag. and
exsag.), convex anterior border is gradual.

Description. Cranidium about twice as wide as long. Glabella slightly longer than wide with a subtrapezoidal to
bell-shaped outline; in lateral profile dorsal surface is above fixed cheek at lobe lp, is equal to or below fixed
cheek at about furrow 2p (PI. 97, fig. 11; PI. 98, figs. 11, 12). Occipital ring about one-quarter as long (sag.) as
wide, slightly wider than glabella at lobe lp, is longest medially then shortens and swings forward laterally
towards axial furrow where it is swollen. Occipital furrow longest and shallowest medially, shortens and deepens
towards axial furrow, has a more steeply inclined posterior than anterior slope. Lobe lp about one-third as wide
as glabella. Abaxial part of lp furrow deep, divides adaxially into two branches; posterior branch runs inward
and obliquely backward, shallows before finally turning inward towards median line; weaker anterior branch
directed forward and inward, not reaching as far adaxially as posterior branch. Shallow extension of posterior
branch connects with occipital furrow to separate lp lobe from frontomedian lobe. Small intermediate lobe
within fork of furrow lp (PI. 97, fig. 9). Lobe 2p is papillate, joined to adaxially directed genal buttress. Furrow
2p directed inward and slightly backward, is continued as sharply flexed shallow depression which meets anterior
branch of furrow lp, thus semi-isolating lobe 2p. Lobe 3p much smaller than 2p, slightly elongate (tr.), sited on

EXPLANATION OF PLATE 98

Figs. 1-11. Tapinocalymene nodulosa (Shirley, 1933). 1-4 are from the Wenlock Series, Homerian Stage,

Coalbrookdale Formation, vicinity of Burrington, Hereford and Worcester; 4 is from C. lundgreni Biozone
strata, sunken lane south of Burrington (SO 442 718). 1-3, holotype, partial internal mould specimen lacking
preglabellar area, GSM 19642, dorsal stereo-pair, left lateral view, x 1 - 5, dorsal view (cephalon), x 2; figured
Salter 1848, pi. 12, figs. 1, la; 1849, pi. 8, figs. 1, 2; 1865, pi. 8, figs. 2, 3; also Shirley 1933, pi. 1, figs. 6-10. 4,
cranidium, NMW 77.3 1G. 8, dorsal view, x 2. 5-12 are from calcareous concretions within faulted patch of
Coalbrookdale Formation, Dolyhir, Powys (SO 2412 5808). 5, 7, 8, partial internal mould pygidium, NMW
53.288.G1, right lateral, posterior views, x 1-5, oblique view, x2. 6, 12, cranidium, GSM Zs 195, dorsal, right
lateral views, x 2. 9-11, partial internal mould cranidium, GSM Zs 1 83, oblique view, x 6, dorsal stereo-pair,
right lateral view, x 2.

PLATE 98

SIVETER, Tapinocalymene

790

PALAEONTOLOGY, VOLUME 23

dorsolateral glabellar surface. Furrow 3p directed at about right angles to median line. Possible 4p furrow (not
observed dorsally) expressed ventrally as a ridge joined to outer, posterior end of ridge which represents furrow
3p (PI. 97, fig. 5). Frontal lobe bluntly rounded in outline, falls steeply to preglabellar furrow.

Axial furrow deep, steep-sided and narrowest around lobe lp, is at least two to three times wider anterior to
bridge of 2p lobe and genal buttress (PI. 97, fig. 9). Anterior pit deep, situated very low down on adaxial side of
axial furrow just anterior to furrow 3p; it is represented ventrally by a boss, the inner anterior slope of which is
hollowed for reception of anterior wing process of hypostoma (PI. 97, fig. 5). Some specimens show vestige of eye
ridge running down abaxial side and across base of axial furrow opposite furrow 3p (PI. 97, fig. 9; PI. 98, fig. 9). In
dorsal view anterior margin of preglabellar area is moderately (PI. 98, fig. 10) to strongly (PI. 97, fig. 2) convex
forward, in lateral profile it is raised just above frontal glabellar lobe (PI. 97, fig. 11), in frontal view it is
sometimes slightly swollen upward opposite axial furrow. Long (sag. and exsag.) preglabellar furrow passes
smoothly forward and progressively more upward on its anterior slope into short, convex (sag.) anterior border.
Outer part of anterior border slopes downward and slightly backward to rostral suture (PI. 97, figs. 6, 11).

Posterior border of cranidium lengthens (exsag.) very slightly from axial furrow to fulcrum, abaxially from
which it expands more quickly until shortening slightly and becoming less convex (exsag.) near facial suture.
Posterior border furrow has a less steeply inclined anterior than posterior slope, both slopes become more gently
inclined abaxially. Postocular part of fixed cheek slopes moderately downward to border furrow; convex
preocular part projects beyond frontal glabellar lobe, is vertical or slightly overhangs abaxial continuation of
preglabellar furrow (PI. 97, fig. 11). Mid-length of palpebral lobe is opposite some part or anterior margin of
lateral lobe 2p, initially it continues slope of fixed cheek then abaxially has a more horizontal attitude (PI. 98,
fig. 4). Posterior branch of facial suture runs outward and slightly backward then swings in broad curve to lateral
border, finally turning more posteriorly to posterior margin (PI. 98, fig. 6); anterior branches are abaxially
convex, slightly convergent. Free cheek incompletely known, slopes steeply to open U-shaped lateral border
furrow, doublure is sharply reflexed upward and outward from lateral border.

Border sector of rostral plate just greater than one- third to just less than one-half as long as wide, slightly more
than three times as long (sag.) as outer part of anterior border (PI. 97, fig. 6). Rostral suture moderately arched.
Connective sutures gently convex outwards, converge posteriorly towards angular junction of border and
doublure sectors. Inner arc of border sector about parallel to rostral suture, marked by a slight ridge (PI. 97, fig.
6). Hypostoma I T to 1-2 times as wide across anterior wings than long (sag.). Anterior margin broadly convex
forward. Anterior border flexed ventrally; border furrow shallow. Anterior wing with deep pit. Lateral margin
slightly convex abaxially between anterior wing and lateral shoulder; lateral border narrows (tr.) posteriorly;
border furrow most distinct opposite (tr.) median protuberance of anterior lobe. Posterior border flattened,
projecting into two broad spines. Faint median furrow connects two conspicuous, ovate maculae. Anterior lobe
of median body about 2-25 to 2-5 times as long as posterior lobe; a spur-like protuberance is directed ventrally
from centre of anterior lobe. Posterior lobe is crescent-shaped.

Thorax characteristically wide (tr.); anterior part of axis less wide than pleural region. Axis has thirteen rings,
each of about constant length (sag. and exsag.) and flat to gently convex in lateral profile, flexed forward
abaxially and swollen at axial furrow. Posterior band of each pleura higher than anterior, moderately convex
(exsag.), forms posterior rim to articulating facet (PI. 97, fig. 1 1; PI. 98, fig. 2). Pleural furrow moderately deep
and U-shaped at fulcrum, less well marked abaxially, dies out on articulating facet. Many specimens have
slightly sinuous course to the distal, posterior margins of thoracic pleurae due to enrolment contact of free cheek;
point of contact more dorsally positioned on posterior pleurae, becomes progressively lower on anterior pleurae,
is continued posteriorly as a cincture on the pygidium (PI. 97, figs. 4, 1 1).

Pygidial axis very gently convex (tr.), has six distinct and one indistinct axial rings and terminal axial piece.
Each ring is almost flat (sag.); anterior rings slightly inflated at axial furrow. Ring furrows shallowest medially,
become deeper towards axial furrow which becomes weaker posteriorly and scarcely present around terminal
axial piece. Inner pleural region slopes gently (tr.) to cincture, thereafter much more steeply to lateral margin.
Pleural region usually has five distinct pleural furrows to the cincture (PI. 97, figs. 3, 8), one specimen (PI. 98, figs.
5, 7, 8) has trace of a sixth. Interpleural furrows very faint. On outer pleural region pleural and interpleural
furrows very weak (PI. 97, fig. 8) or absent (PI. 98, fig. 8), leaving smooth border. Postaxial sector falls almost
vertically from terminal axial piece.

Small to medium-sized granules are evenly distributed on glabella; much larger ones on genal buttress,
anterior adaxial part of fixed cheek, and sometimes anterior border (PI. 97, fig. 9; PI. 98, fig. 9). Abaxial infla-
tions of occipital and axial rings have concentration of small granules. Small, closely spaced granules on
border sector of rostral plate and lateral border of cheek (PI. 97, fig. 6). Scattered fine granules on hypostoma,
thorax, and pygidium. Maculae and deepest part of preglabellar, pleural, axial, and articulating furrows lack
granules.

SIVETER: CALYMENID TRILOBITES

791

Discussion. Variation is present in the degree of upward curving of the preglabellar area and
impression of cincture and interpleural furrows on the pygidium, though the latter are never strongly
developed and may be almost completely absent. The largest cranidium (PI. 98, fig. 4; ?gerontic
specimen) is the only one to have the anterior glabellar margin transversely in line with the fixed
cheek.

Cranidia from Dolyhir (PI. 98, figs. 6, 9-12) have a more swollen anterior border, and thus a
relatively shorter preglabellar furrow, than in typical Burrington specimens (cf. PI. 97, fig. 2); in this
character, therefore, they approach T. vulpecula. They are placed with T. nodulosa because they lack
the more distinct break in slope between preglabellar furrow and anterior border that is diagnostic of
the new species (cf. PI. 98, figs. 6, 12; PI. 100, figs. 1, 2), and because the anterior border of other
specimens from Burrington is very similar to that in the Dolyhir material (cf. PI. 97, figs. 1, 1 1; PI. 98,
figs. 10, 11).

Tapinocalymene volsoriforma sp. nov.

Plate 99, figs. 1-15; text-fig. 2i, j

1919 Calymene blumenbachii Brongniart; Reed, in Garwood and Goodyear, p. 19.

Derivation of name. Latin, volsorium, a curved archstone, referring to the cranidial anterior border outline in
dorsal view.

Holotype. Almost complete cranidium, GSM Zs63, Garwood Collection; PI. 99, figs. 1-4.

Type locality. Wenlock Series, Sheinwoodian Stage, shale band included within the Dolyhir and Nash Scar
Limestone Formation near its base, Dolyhir Quarries near Old Radnor, Powys (see Garwood and Goodyear
1919, p. 18, pi. 7). There are no specific horizon data given for T. volsoriforma specimens in the Institute of
Geological Sciences Garwood Collection. Apart from the Pre-Cambrian, only the Dolyhir and Nash Scar
Limestone Formation and a small faulted patch of Coalbrookdale Formation outcrop in this area. The lithology
of the Coalbrookdale Formation here is different to the matrix surrounding the T. volsoriforma specimens and
this patch yields instead T. nodulosa. The matrix is also unlike the mass of pure, crystalline Dolyhir Limestone,
though Garwood and Goodyear’s ( 1 9 1 9, p. 18) description of a shale band included in the limestone near its base
fits the IGS material. I have collected a T. volsoriforma cranidium from this shale band on the north side of the
disused railway track, Dolyhir (SO 2410 5823). The exact location of the type locality amongst the Dolyhir
Quarries is unknown. The Dolyhir and Nash Scar Limestone Formation is considered to span the Cyrtograptus
centrifugus, C. murchisoni and part of the Monograptus riccartonensis biozones of the Sheinwoodian (Bassett
1974, p. 759).

Additional material and occurrences. At least three incomplete cranidia, GSM Zs62, GSM Z 19983, LM 2850; one
incomplete cephalon, GSM Zs65; one incomplete cephalon plus rostral plate GSM Zs58; one incomplete
hypostoma GSM Z 19696; seven incomplete pygidia GSM Zs22, GSM Zs24, GSM Zs57, GSM Zs59-61, GSM
Zs64. Numerous other fragments of cranidia, pygidia, and thoracic segments are present in the Garwood
Collection, Institute of Geological Sciences Museum.

Only recorded with certainty in situ from the type area, but it may also be present in the Coalbrookdale
Formation of Salop (PI. 99, fig. 16). One transported specimen (PI. 99, fig. 5) was collected from a stream bed
near English Bridge, Shrewsbury.

Diagnosis. Preglabellar area about one-third as long as glabella. Preglabellar furrow about one-sixth as long
(sag.) as preglabellar area. Anterior border relatively long, it slopes fairly gently upward and forward from the
more steeply inclined anterior side of preglabellar furrow. Axial furrow anterior to lobe 2p only slightly wider
than around lp lobe. Hypostoma with only moderately inflated subcircular protuberance on anterior lobe.

Description. Glabella essentially as in T. nodulosa but inflation within fork of lp furrow generally weaker. Axial
furrow deep and fairly narrow around lobe lp, slightly wider beside lobe 3p and frontal lobe. Anterior pit is
anterior to furrow 3p. Preglabellar furrow short and moderately deep medially, lengthens (exsag.) at
anterolateral corner of frontal lobe, continues forward and outward between fixed cheek and anterior border
into deep, narrow, lateral border furrow; steeply sloping anterior side of preglabellar furrow meets anterior
border at change of slope. Anterior border about one-quarter (PI. 99, fig. 5) to three-tenths (PI. 99, fig. 1) as long
as glabella, dorsal surface very gently convex (sag.), slopes forward and slightly upward. In lateral profile

792

PALAEONTOLOGY, VOLUME 23

anterior margin rises above anterior part of frontal lobe. Fixed cheek, palpebral lobe, and facial suture as in T.
nodulosa. Free cheek has narrow eye socle (PI. 99, figs. 7, 1 3), convex inner part of cheek falls steeply to U-shaped
lateral border furrow; junction of border furrow with lateral border is more angular (tr.) than with inner part of
cheek. Lateral border rolled under ventrally. Rostral plate (PI. 99, fig. 1 1) imperfectly preserved but apparently
similar to T. nodulosa. Hypostoma has moderate, subcircular inflation on anterior lobe (PI. 99, fig. 8). Oval
macula smooth; median furrow very faint. Posterior border expanded into two spines.

Thoracic segments (fragmentary) have axial rings inflated near axial furrow. Pygidium like that of T. nodulosa.
Axis very weakly convex (tr.), has at least seven rings plus terminal piece. Axial rings longest (sag.) and ring
furrows shallowest medially. Seventh ring furrow lacking abaxially; very faint trace medially of eighth furrow
(PI. 99, fig. 6). Inner pleural region descends gently abaxially, outer part falls more strongly. Five (possibly six)
pleural furrows are best marked on inner pleural region; interpleural furrows much weaker.

Glabella, outer part of fixed cheek and free cheek have numerous small to medium-sized granules. Inner part
of fixed cheek (especially anteriorly), genal buttress and anterior border are coarsely granulate (PI. 99, fig. 14).
Deepest parts of glabellar and preglabellar furrows lack granules; lateral and posterior border furrows have
scattered small granules. Pygidium and hypostoma are finely granulate.

Discussion. Compared to the holotype (PI. 99, fig. 1), a cranidium from outside the type area (PI. 99,
fig. 5) has a more rounded outline to the anterolateral corner of the frontal lobe, a shallower pre-
glabellar furrow, a much weaker break in slope between preglabellar furrow and anterior border, and
a relatively shorter anterior border (to the glabella). This variation is considered to be intraspecific as
Dolyhir specimens vary in a like manner (for example, PI. 99, fig. 13 in the first three of these
characters).

Two other Tapinocalymene cranidia from Salop (PI. 99, fig. 16; NMW 77.31G.13) may belong to
T. volsoriforma, but are indifferently preserved. They are associated with Monograptus flemingii
(identified by Dr. I. Strachan; pers. comm. Dr. C. N. Rodgers), indicating a post M. riccartonensis to
C. lundgreni Biozone age. Apparently T. volsoriforma, or a species close to it, existed outside the type
area at a later date.

Tapinocalymene vulpecula sp. nov.

Plate 100, figs. 1-8, 13, 15; text-fig. 2g, h

Derivation of name. Latin, diminutive of vulpes, fox, alluding to the appearence of the hypostoma.

Holotype. Incomplete cranidium, NMW 77.31G.1, collected D. J. Siveter 1971; PI. 100, figs. 1, 2.

Type locality. Wenlock Series, small disused quarry on the west side of the road from Letton to Walford, \ km
north of Letton, Hereford and Worcester (SO 3790 7080). Graptolites (SM A80317-21, SM A80387-91) from
this locality have been assigned to M. flemingii, and probably belong to the lundgreni Biozone (pers. comm.
Dr. R. B. Rickards).

EXPLANATION OF PLATE 99

Figs. 1 - 1 5. Tapinocalymene volsoriforma gen. et sp. nov. All specimens except fig. 5 are from the Wenlock Series,
Sheinwoodian Stage, included shale band in Dolyhir and Nash Scar Limestone Formation, Dolyhir, Powys.
1-4, holotype cranidium, GSM Zs63, dorsal stereo-pair, frontal, left oblique views, x 2, right lateral view,
x 2-25. 5, cranidium, silicone-rubber cast of external mould, LM 2850b, specimen from a water-transported
pebble, found near English Bridge, Shrewsbury, Salop; dorsal view, x 2. 6, 15, pygidium, GSM Zs61, dorsal,
left lateral views, x2-25. 7, cranidium and left free cheek, GSM Zs65, left oblique view, x2. 8, hypostoma,
GSM Z19696, ventral stereo-pair, x 5. 9, 10, cranidium, GSM Z19983, left lateral view, dorsal stereo-pair,
x2. 11, 13, 14, cephalon, GSM Zs58, ventral (rostral plate), dorsal views, x 2, dorsal view, x4. 12, pygidium,
GSM Zs64, dorsal view, x 2-25.

Fig. 16. Tapinocalymene cf. T. volsoriforma gen. et sp. nov. Cephalon, silicone-rubber cast of external mould,
NMW 77. 3 1G. 12b, Wenlock Series, Coalbrookdale Formation, road cutting on A489 between Horderley and
Plowden, south side of Long Mynd, Salop (SO 402 875); dorsal view, x 2.

PLATE 99

SIVETER, Tapinocalymene

794

PALAEONTOLOGY, VOLUME 23

Additional material. Only from type locality; three incomplete cranidia, NMW 77.31G.2-4; two pygidia, one
with cuticle, NMW 77.31G.6, the other an internal mould, NMW 77.31G.7; one hypostoma, internal mould
plus counterpart, NMW 77.31G.5a,b. Numerous other fragmentary cranidia, pygidia, and thoracic segments.

Diagnosis. Preglabellar area about two-fifths to one-third as long as glabella. Preglabellar furrow about as long
(sag.) as anterior border and half as long as preglabellar area. Marked break in slope where anterior side of
furrow meets posterior margin of border.

Description. Glabella similar to that of T. nodulosa. Axial furrow at least twice as wide at lobe 3p and frontal
lobe than around lobe lp. Anterior pit situated in axial furrow below lateral glabellar furrow 3p. Preglabellar
furrow deep, U-shaped (sag.), about as long as anterior border medially, shortens (exsag.) abaxially where fixed
cheek approaches anterior border; anterior side of furrow is about vertical and meets posterior part of anterior
border in a sharp break of slope, border then continues less steeply forward and upward (PI. 100, fig. 2). Anterior
border is of about constant length abaxially from median line to opposite axial furrow, thereafter shortening
(exsag.) towards facial suture. In lateral profile anterior margin is about level with or just above height of frontal
lobe. Fixed cheek, palpebral lobe, facial suture, and hypostoma essentially like that of T. nodulosa. Free cheek
and rostral plate unknown. Pygidium, showing no differences to that of T. nodulosa, has six complete, one
incomplete, very gently convex (tr. and sag.) axial rings. Anterior six ring furrows shallowest at and just either
side median line, deepen quickly abaxially; very faint seventh ring furrow does not reach axial furrow which is
weakest around terminal axial piece. Pleural region slopes very gently to cincture, more steeply abaxially. Five
pleural furrows run outward and backward to cincture, apparently absent from here to lateral margin
though this part of pygidium is imperfectly preserved. Interpleural furrows extremely faint, best seen near axial
furrow, not marked on internal mould (cf. PI. 100, figs. 8, 13). Sculpture like that of T. nodulosa and
T. volsoriforma.

Distinctions between Tapinocalymene species

T. volsoriforma differs most obviously from T. vulpecula and T. nodulosa by its much shorter
preglabellar furrow and longer anterior border (cf. PI. 97, fig. 2; PI. 99, fig. 1; PI. 100, fig. 1). Further,
T. volsoriforma has a narrower axial furrow anterior to glabellar lobe 2p (cf. PI. 97, fig. 9; PI. 99, figs.
13, 14; PI. 100, fig. 1), and a ventral protuberance on the anterior lobe of the hypostoma which seems
less well developed than that in the other two species (cf. PI. 97, fig. 6; PI. 99, fig. 8; PI. 100, figs. 7, 15).
T. vulpecula is best distinguished from T. nodulosa by its relatively shorter preglabellar furrow, longer
anterior border, and sharper break in slope between these two features. In the change from T.
volsoriforma to T. nodulosa through the Wenlock, in addition to a very marked change in the
preglabellar area, the axial furrow anterior to lobe 2p becomes wider, the hypostomal protuberance
seemingly becomes better developed (there is only one incomplete hypostoma of volsoriforma ), and
the inflation within the adaxial fork of furrow lp becomes generally stronger (cf. PI. 97, fig. 9; PI. 99,
fig. 14).

EXPLANATION OF PLATE 100

Figs. 1-8, 13, 15. Tapinocalymene vulpecula gen. et sp. nov. All specimens are from Wenlock Series,
Coalbrookdale Formation, probably Cyrtograptus lundgreni Biozone, Homerian Stage, small old quarry on
west side of road from Letton to Walford, \ km north of Letton, Hereford and Worcester (SO 3790 7080). 1 , 2,
holotype cranidium, NMW 77.3 1G.1, dorsal stereo-pair, left lateral view, x2. 3, partial internal mould
cranidium, NMW 77.31G.2, dorsal view, x 2. 4, 5, cranidium, NMW 77.31G.4, dorsal, frontal views, x 4.
6, cranidium, NMW 77.31G.3, dorsal view, x 2. 7, 15, hypostoma, silicone-rubber cast of external mould,
NMW 77.31G.5b, lateral, ventral views, x 8. 8, pygidium, NMW 77.31G.6, dorsal stereo-pair, x 2-25. 13,
internal mould pygidium, NMW 77.31G.7, dorsal view, x2-25.

Figs. 9-11. Calymene blumenbachii blumenbachii Brongniart, 1822. Complete enrolled specimen, BM 44213,
Wenlock Series, Homerian Stage, Much Wenlock Limestone Formation, Dudley, West Midlands; dorsal
stereo-pair, left lateral, posterior views, x 2.

Figs. 12, 14, 16. Calymene blumenbachii subsp. nov. Complete specimen, NMW 73.28G.3a, Wenlock Series,
Sheinwoodian Stage, Woolhope Limestone Formation, temporary trench just north of church, Woolhope,
Hereford and Worcester; dorsal, left lateral views, x 2, posterior oblique view, x 4.

PLATE 100

SIVETER, Calymenid trilobites

796

PALAEONTOLOGY, VOLUME 23

ORIGIN OF THE GENERA TAPINOCALYMENE AND SPATHACALYMENE
Tapinocalymene

T. volsoriforma and T. nodulosa show the two morphological extremes in the preglabellar area of
Tapinocalymene , that of T. vulpecula being intermediate in form. T. volsoriforma is from the lower
Wenlock; T. nodulosa is from the upper Wenlock (C. lundgreni and 1G. nassa biozones), T. vulpecula
being of probable lundgreni Biozone age. An evolutionary sequence involving an increase in the
length and area of the preglabellar furrow through the Wenlock is postulated, from the sagitally short
furrow in T. volsoriforma (text-fig. 2i, j), through the moderately long furrow of T. vulpecula (text-fig.
2g, h), to the scoop-like furrow and preglabellar area of T. nodulosa (text-fig. 2a, b). This trend is
accompanied by a loss of the angular break in slope in the preglabellar area; the intraspecific
variation attributed to T. nodulosa in the anterior border (Dolyhir specimens, text-fig. 2e, f) is taken
as further evidence of the proposed phyletic series.

The exoskeleton of T. nodulosa is wide and rather depressed suggesting a benthic habit (cf. Fortey
and Barnes 1977, p. 304 for broadly analagous conditions in certain olenids). Though the function of
the distinctive preglabellar area is problematic, it may have been used in shallow burrowing, or to
disturb superficial sediment in search of food. It may have been adapted to the carbonate mud facies
of the Coalbrookdale Formation. T. nodulosa and T. vulpecula both occur in the same Wenlock
calcareous shale facies, deposits interpreted (Bassett 1974, pp. 770-773, text-figs. 7, 8) as somewhat
offshore, deepish-water elastics; in this context my own collections show that graptolites and small
brachiopods invariably accompany these two species. The main mass of algal-rich Dolyhir and Nash
Scar Limestone Formation was formed in shallow water deposited on a local offshore topographic
high of faulted Pre-Cambrian rocks (Bassett 1 974, p. 772, text-fig. 7; Hurst, Hancock, and McKerrow
1978, p. 204); the included shale band, which yields T. volsoriforma, suggests there may have been a
temporary incursion of deeper water.

Tapinocalymene originated from Diacalymene, possibly from a stock broadly ancestral to D.
diademata, rather than from DP crassa Shirley, 1936 and allied species. It is not related to Calymene
sensu stricto. Specimens provisionally assigned to a new subspecies of C. blumenbachii are known
from the Woolhope Limestone Formation of low Wenlock age (C. centrifugus and C. murchisoni
biozones). These (PI. 100, figs. 12, 14, 16) are approximately coeval with T. volsoriforma, yet are
morphologically quite distinct (see differences between Tapinocalymene and Calymene in generic
discussion). The C. blumenbachii species group, in contrast to Tapinocalymene, is characteristic of
more onshore, generally shallower-water environments.

Spathacalymene

The preglabellar area of Spathacalymene nasuta has an inverted U-shaped, posteriorly divergent
outline in dorsal view, with a long, dorsally flattened anterior border sloping moderately steeply
posteriorly to meet a short, more steeply inclined preglabellar furrow. Apart from its length, this form
is similar to the preglabellar area of certain calymenids assigned to Diacalymene by Shirley (1936);
also similar is the pointed, forwardly and inwardly directed, anterior part of the fixed cheek. Both
similarities apply to DP crassa from the early Llandovery (Rhuddanian) of Wales which, moreover,
has a strongly convex (sag. and tr.), high glabella (relative to the fixed cheeks), as in S. nasuta (cf.
Temple 1975, pi. 25, figs. 3, 4; PI. 101, figs. 1, 4, 8 herein). The unrevised C. vogdesi Foerste (1887,
p. 95, pi. 8, figs. 12, 13; 1893, p. 526, pi. 25, fig. 25; pi. 27, figs. 12, 13; 1919, pi. 19, fig. 5) from the lower
Silurian of Ohio is a possible senior synonym of crassa, and it is also recorded (Foerste 1893, p. 527)
from Indiana, where the upper Llandovery S. nasuta occurs. Foerste (1919, p. 393) regarded vogdesi
as a ‘typical Brassfield species’; his use of the term Brassfield included strata of middle to upper
Llandovery age (Berry and Boucot 1970, p. 127).

The evolution of the preglabellar area of S. nasuta from that of ‘C’. vogdesi or a similar species
requires only the lengthening of the anterior border. This involves much less change in morphology
from ancestor to descendant than in that proposed for Tapinocalymene (cf. text-figs. 2, 3).

text-fig. 2. Proposed evolutionary lineage in Tapinocalymene gen. nov. Dorsal and
lateral outlines ofcranidia, all x 2. a, b. T. nodulosa, HM A212/1, pi. 97, figs. 1, 1 1. c, D. T.
nodulosa, GSM Zsl83, pi. 98, figs. 10, 1 1. e, f. T. nodulosa GSM Zsl95, pi. 98, figs. 6, 12.
G, h. T. vulpecula, holotype, NMW 77.31G.1, pi. 100, figs. 1, 2. i, j. T. volsoriforma,
holotype, GSM Zs63, pi. 99, figs. 1, 2.

PALAEONTOLOGY, VOLUME 23

text-fig. 3. Possible origin of Spathacalymene Tillman, 1960. A, B. Spathacalymene nasuta,
USNM 170363, dorsal and right lateral cranidial outlines, x 1, pi. 101, figs. 1, 8. c, D.
‘ Calymene ’ vogdesi, holotype, lower Silurian, Centreville, Ohio, U.S.A., dorsal and sagittal
cranidial outlines; magnification unknown, but equal to Foerste’s (1887, pi. 8, figs. 12, 13)
original illustrations.

EXPLANATION OF PLATE 101

Figs. 1, 3, 4, 7-9. Spathacalymene waswta (Ulrich, 1879). 1, 4, 8, 9, complete specimen, USNM 170363, Osgood
Limestone (late Llandovery), quarry 1 -3 km east of Napoleon, Ripley County, Indiana, U.S.A.; dorsal stereo-
pair, right lateral view, x 1, right oblique view, x 2, posterior oblique view, x 3-3; figured Tillman 1960, pi.
116, figs. 1, 4, 5, 8, 9. 3, 7, paralectotype, cephalon, rostral plate and two thoracic segments, Osgood
Formation (late Llandovery), Osgood, Indiana, U.S. A.; dorsal, ventral views, x 1 -5; figured Tillman 1960, pi.
116, figs. 10-12.

Figs. 2, 5, 6, 10. Diacalymene diademata ( Barrande, 1846). Cranidium, largely internal mould, NMW 71.8G.377,
upper part of the Liten Formation, Cyrtograptus radians- Monograptus testis biozones (late Wenlock), above
path leading from Svaty Jan pod Skalou to Vraz, south-west of Prague, Czechoslovakia; frontal, right lateral
views, dorsal stereo-pair, x 2, right oblique view, x 2-5.

PLATE 101

SIVETER, Calymenid trilobites

800

PALAEONTOLOGY, VOLUME 23

CONCLUSIONS

1. C. nodulosa Shirley, 1933, T. volsoriforma sp. nov. and T. vulpecula sp. nov. from the Wenlock
Series of the Welsh Borderland belong to a new genus, Tapinocalymene.

2. Tapinocalymene shows plasticity in the form of its preglabellar area, which links the phyletic
series T. volsoriforma, T. vulpecula, T. nodulosa.

3. Tapinocalymene was probably benthic and occurs throughout its stratigraphic range in
somewhat offshore carbonate muds. The scoop-like preglabellar area of T. nodulosa developed in
response to this mode of life and bottom conditions.

4. The possession of a long, conspicuous preglabellar area provides no basis for considering T.
nodulosa and S. nasuta congeneric; details of its morphology and origin are distinctive in both taxa,
and it is a feature that species of different calymenid lineages occasionally develop.

Acknowledgements. I thank Dr. A. W. A. Rushton for improving the manuscript, and who made available for
examination specimens in his care, as did Dr. R. A. Fortey, Mr. S. F. Morris, Dr. M. G. Bassett, Dr. R. M.
Owens, Mr. J. Norton, Dr. J. K. Ingham, Dr. V. Jaanusson, and Dr. J. E. Merida. This paper is dedicated to the
memory of the late Professor P. C. Sylvester-Bradley of Leicester University.

REFERENCES

barrande, J. 1846. Notice preliminaire sur le Systeme Silurien et les trilobites de Boheme, vi + 97 pp. Leipsic.

— 1852. Systeme Silurien du centre de la Boheme. lere partie: Recherches paleontologiques. Vol. 1. Crustaces:
Trilobites. xxx + 935 pp. Prague and Paris.

bassett, M. G. 1974. Review of the stratigraphy of the Wenlock Series in the Welsh Borderland and South Wales.
Palaeontology, 17, 745-777.

berry, w. B. N. and boucot, a. J. 1970. Correlation of the North American Silurian rocks. Spec. Pap. geol. Soc.
Am. 102, 289 pp.

brongniart, A. 1822. Histoire naturelle des Crustaces fossiles, sous les rapports zoologique et geologique, savoir
les Trilobites. Les Crustaces proprement dits, par A.-G. Desmarest. vii + 154 pp. Paris.

Campbell, K. s. w. 1967. Trilobites of the Henryhouse Formation (Silurian) in Oklahoma. Bull. Okla. geol. Surv.
115, 1-68.

dalman, j. w. 1827. Om Palaederna, eller de sa kallade Trilobiterna. K. svenska Vetensk-Akad. Handl. (for
1826), 113-152, 226-294.

etheridge, r. 1888. Fossils of the British Islands stratigraphically and zoologically arranged. 1. Palaeozoic,
comprising the Cambrian, Silurian, Devonian, Carboniferous and Permian species, i-viii, 468 pp. Oxford.
foerste, a. f. 1887. The Clinton Group of Ohio, pt. 2. Bull. Sci. Lab. Denison Unix. 2, 89-110, 140-176.

1893. Fossils of the Clinton Group in Ohio and Indiana. Rep. geol. Surv. Ohio, 7, 516-601.

— 1919. Silurian fossils from Ohio, with notes on related species from other horizons. OhioJ. Sci. 19, 367-404.
fortey, R. a. and barnes, c. R. 1977. Early Ordovician conodont and trilobite communities of Spitsbergen:

influence on biogeography. Alcheringa, 1, 297-309.

Garwood, E. j. and Goodyear, E. 1919. On the geology of the Old Radnor District, with special reference to an
algal development in the Woolhope Limestone. Q. Jl geol. Soc. Lond. 74 (for 1918), 1-30.

GREIG, d. c., WRIGHT, j. e„ hains, b. a. and Mitchell, G. H. 1968. Geology of the country around Church
Stretton, Craven Arms, Wenlock Edge and Brown Clee. Mem. geol. Surv. U.K. i-xiii, 379 pp.

HAAS, w. 1968. Trilobiten aus dem Silur und Devon von Bithynien (N.W.-Turkei). Palaeontographica, A130,
60-207.

hall, J. 1861. Report of the Superintendent of the Geological Survey, Wisconsin. 52 pp. Madison, Wisconsin.
Holland, c. h., rickards, r. b. and warren, p. t. 1969. The Wenlock graptolites of the Ludlow district,
Shropshire, and their stratigraphical significance. Palaeontology, 12, 663-683.
hurst, 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.
ingham, J. K. 1977. A monograph of the upper Ordovician trilobites from the Cautley and Dent districts of
Westmorland and Yorkshire. Palaeontogr. Soc. ( Monogr .), 3, 89-121.
kegel, w. 1927. Uber obersilurische Trilobiten aus dem Harz und dem Rheinischen Schiefergebirge. Jb. preuss.
geol. Landesanst. 48, 616-647.

la touche, J. D. 1884. + handbook of the Geology of Shropshire. 91 pp. London.

SIVETER: CALYMENID TRILOBITES

801

lindstrom, G. 1885. Forteckning pa Gotlands Siluriska Crustaceer. Of vers K. Vetensk-Akad. Fork. Stockhm. 6,
37-100.

lu yen-hao, 1975. Ordovician trilobite faunas of central and southwestern China. Palaeont. sin. n.s. (B), 152
(whole series), 1-484. [Chinese, with English text 263-463.]
mccoy, F. 1851 . In sedgwick, A. and mccoy, f. A synopsis of the classification of the British Palaeozoic rocks, with
a systematic description of the British Palaeozoic fossils in the geological museum of the University of
Cambridge. Fasc. I, 1-184. London and Cambridge.

mcnamara, K. J. 1979. Trilobites from the Coniston Limestone Group (Ashgill Series) of the Lake District,
England. Palaeontology, 22, 53-92.

miller, J. 1976. The sensory fields and life mode of Phacops rana (Green, 1832) (Trilobita). Trans. R. Soc. Edinb.
69, 337-367.

milne-edwards, h. 1840. Histoire naturelle des Crustaces, comprenant /’ anatomie, la physiologie et la
classification de ces animaux. 3, 638 pp. Paris.

Murchison, R. I. 1839. The Silurian System, founded on geological researches in the counties of Salop, Hereford,
Radnor, Montgomery, Caermarthen, Brecon, Pembroke, Monmouth, Gloucester, Worcester and Stafford; with
descriptions of the coalfields and overlying formations, i-xxxii + 768 pp. London.

— 1859. Siluria. The history of the oldest fossiliferous rocks and their foundations, with a brief sketch of the
distribution of gold over the earth. (3rd edn.). i-xx + 592 pp.

reed, F. r. c. 1920. Description of two trilobites. In gardiner, c. i. The Silurian rocks of May Hill. Proc.
Cotteswold Nat. Fid. Club, 20, 219-222.

richter, r. and richter, E. 1954. Die Trilobiten des Ebbe-Sattels. Abh. Senckenb. naturforsch. Ges. 488, 1 76.
salter, j. w. 1848. In Phillips, J. and salter, j. w. Palaeontological appendix to Professor John Phillips’
Memoir on the Malvern Hills, compared with the Palaeozoic districts of Abberley etc. Mem. geol. Surv. U.K.
2, 331-386.

— 1849. Figures and descriptions illustrative of British organic remains. Ibid., 2 Dec.

1865. A Monograph of the British trilobites from the Cambrian, Silurian, and Devonian formations.

Palaeontogr. Soc. ( Monogr .), 2, 81-128.

1 873. A catalogue of the collection of Cambrian and Silurian fossils contained in the Geological Museum of the

University of Cambridge. 204 pp. Cambridge.

schrank, e. 1970. Calymeniden (Trilobita) aus Silurischen Geschieben. Ber. Deutsch. Ges. geol. Wiss., A,
Geol.^Palaont. 15, 109-146.

Shirley, J. 1933. A redescription of the known British Silurian species of Calymene (s.l.). Mem. Proc. Manchester
lit. Phil. Soc. 75, 1-33.

— 1936. Some British trilobites of the family Calymenidae. Q. Jl geol. Soc. Lond. 92, 384-422.

siveter, d. j. 1977. The middle Ordovician of the Oslo region, Norway, 27. Trilobites of the family Calymenidae.
Norsk, geol. tiddskr. 56 (for 1976), 335-396.

— 1979. Metacalymene Kegel, 1927, a calymenid trilobite from the Kopanina Formation (Silurian) of
Bohemia. J. Paleont. 53, 367-379.

snajdr, m. 1975. New Trilobita from the Llandovery at Hyskov in the Beroun area, central Bohema. Vest,
ustred. Ust. geol. 50, 311-316.

squirrell, h. c. and white, d. e. 1978. Stratigraphy of the Silurian and Old Red Sandstone of the Cennen Valley
and adjacent areas, south-east Dyfed, Wales. Rep. Inst. Geol. Sci. 78/6, 1-45.

Stubblefield, c. J. 1938. The types and figured specimens in Phillips’ and Salter’s Palaeontological Appendix to
John Phillips’ Memoir on ‘The Malvern Hills compared with the Palaeozoic Districts of Abberley, etc.’ (Mem.
geol. Surv. U.K. 2, 1848). Summ. Progr. geol. Surv. U.K. (for 1936), 27-51.
sun, Y. c. 1931. Ordovician trilobites of central and southern China. Palaeont. sin. Ser. B, 7 (1), 47 pp.
temple, j. t. 1969. Lower Llandovery (Silurian) trilobites from Keisley, Westmorland. Bull. Br. Mus. nat. Hist.
(Geol.), 18, 197-230.

— 1970. The Lower Llandovery brachiopods and trilobites from Ffridd Mathrafal, near Meifod,
Montgomeryshire. Palaeontogr. Soc. (Monogr.), 76 pp.

— 1975. Early Llandovery trilobites from Wales with notes on British Llandovery calymenids. Palaeontology,
18, 137-159.

tillman, c. G. 1960. Spathacalymene, an unusual new Silurian trilobite genus. J. Paleont. 34, 891-895.
tomczykow a, e. 1970. Silurian Spathacalymene Tillman, 1960 (Trilobita) of Poland. Acta palaeont. pol. 15,
63-94.

tornquist, s. L. 1884. Undersokningar ofver Siljansomradets trilobitfauna. Sver. geol. Unders. Afh., Ser. C, 66,

101 pp.

802

PALAEONTOLOGY, VOLUME 23

ulrich, E. o. 1879. Description of a trilobite from the Niagara Group of Indiana. J. Cincinn. Soc. nat. Hist. 2,

warburg, E. 1925. The trilobites of the Leptaena Limestone in Dalarne. Bull. geol. Instil. Uppsala, 17, 1-162.
Whittington, H. B. 1971a. A newcalymenid trilobite from the Maquoketa Shale, Iowa. In dutro, j. t. Jr. (edit.),
Paleozoic Perspectives: A paleontological tribute to G. Arthur Cooper. Smithson. Contrib. Paleobiol. 3,
129-136.

— 19716. Silurian calymenid trilobites from the United States, Norway and Sweden. Palaeontology, 14,
455-477.

williams, a. 1953. The geology of the Llandeilo district, Carmarthenshire. Q. Jl geol. Soc. Lond. 108, 177-208.
yi, Y. e. 1957. The Caradocian trilobite fauna from the Yangtze-Gorges. Acta palaeont. sin. 5, 545-559.

131-134.

D. J. SIVETER

Typescript received 12 September 1979
Revised typescript received 20 January 1980

Department of Geology
The University of Hull
Hull HU6 7RX

SPICULE PSEUDOMORPHS IN A NEW
PALAEOZOIC CHAETETID, AND ITS
SCLEROSPONGE AFFINITIES

by DAVID I. GRAY

Abstract. A Palaeozoic chaetetid, bearing intramural spicule pseudomorphs, Chaetetes ( Boswellia ) mortoni sp.
nov., is described from the British Dinantian. Spicules are preserved as calcite, pyrite, and silica pseudomorphs.
Only silica pseudomorphs retain detail of their tylostyle form. Neomorphism locally obliterates the spicular
fabric. A primary mineralogy is suggested consisting of an aragonitic calcareous skeleton, with entrapped opal
‘A’ spicules. Comparison of morphology and microstructure with extant and fossil sclerosponges indicates
a close relationship between this chaetetid and the Ceratoporellida, and supports the sclerosponge nature of
some Palaeozoic chaetetids.

The Class Sclerospongiae Hartman and Goreau, 1972, was proposed following the rediscovery of
coralline sponges among the Jamaican coral-reef ahermatypic cryptofauna (Hartman 1969;
Hartman and Goreau 1970). Sclerosponges were defined by Hartman and Goreau (1972, p. 144) as
‘sponges secreting a compound skeleton of siliceous spicules, proteinaceous fibres and calcium
carbonate, the latter laid down as a basal mass in which the siliceous spicules may or may not be
entrapped’. The similarity of fossil chaetetids to some sclerosponges (briefly discussed by Kirkpatrick
( 1 909, 1 9 1 2a, 1 9 1 26) along with the monticuliporans), led Hartman and Goreau ( 1 972) to remove the
Chaetetida Okulitch, 1936, from the Anthozoa or Hydrozoa to the Sclerospongiae. They also erected
the Order Ceratoporellida Hartman and Goreau, 1972, to include four extant sclerosponge genera,
and added a third Order, the Tabulospongida Hartman and Goreau, 1975, following the discovery of
a tabula-bearing form from the Pacific (Hartman and Goreau 1975), of which two more extant
species have subsequently been described (Mori 1976, 1977) and a record traced back into the
Mesozoic. Stearn (1972, 1975) discussed the sclerosponge affinities of the stromatoporoids.

The recognition of the Sclerospongiae as a Class has been questioned by a number of authors. Levi
(1973) considered the sclerosponges as a Subclass of the Demospongiae, subsequently followed by
Vacelet, Vasseur, and Levi (1976) and Vacelet (1977). This classification takes into account the
organization of living sclerosponge tissue which is ‘basically similar to that of the Class
Demospongiae except that it is divided into units each of which extends down into the upper layer of
the basal calcareous skeleton’ (Hartman and Goreau 1972, pp. 144-145).

The variability in spicule form and distribution (see Table 1 ) suggests that some sclerosponges may
be related even more closely to other groups of demosponges. For example, Vacelet (1977, p. 347)
mentions that the spicule character of Tabulospongia wellsi is similar to that displayed by the
Spirastrellidae. Vacelet (1977, p. 347) also states that a basal calcareous skeleton may be a convergent
structure in many groups of demosponges. This would account for the great variability of calcareous
skeletal morphology and microstructure (see Table 1) observed in those forms classified as
sclerosponges, and would imply that ‘sclerosponge’ is a convenience term for considering groups with
a similar homeomorphic tendency.

In this paper the sclerosponges are considered as a Subclass of the Class Demospongiae Sollas,
1875.

The fossil history of sclerosponges that entrap spicules in their calcareous skeleton is represented
by a limited assortment of forms including a few ceratoporellids, one species (Kazmierczak 1974) of

[Palaeontology, Vol. 23, Part 4, 1980, pp. 803-820, pis. 102-103.]

804

PALAEONTOLOGY, VOLUME 23

the Order Muranida Kazmierczak and Hillmer, 1974, a few problematical records of stromato-
poroids and, until now, only two species of Mesozoic chaetetids. Table 1 summarizes their
distribution and variation and allows comparisons to be made with extant forms.

Chaetetids are a diverse group with separate Palaeozoic and Mesozoic histories. Differences exist
in the skeletal architecture of Palaeozoic and Mesozoic forms (Fischer 1970), and their phylogenetic
relationships are not completely understood. Scrutton (1979, p. 169) reviewed briefly their
relationships, and whilst supporting their sclerosponge affinities he emphasized the lack of
convincing spicules associated with chaetetids as ‘a major source of doubt for some workers’ to their
classification within the Porifera. Dieci, Russo, Russo, and Marchi (1977) were the first to report a
spicule-bearing ‘chaetetid’, Atrochaetetes medius Cuif and Fischer, 1974, from the Upper Triassic of
Italy, with intramural acanthostyle spicules, replaced by calcite. Atrochaetetes Cuif and Fischer,
1974, is characterized by a discontinuous backfill of fascicular fibrous carbonate extending into the
lumen (Cuif and Fischer, 1974, p. 8) rather than complete tabulae typical of the chaetetids s.s.
Continuous fascicular fibrous backfills are typical of ceratoporellids (see below). Since A. medius also
has a ceratoporellid-like spicular fabric, the genus Atrochaetetes should be regarded as an aberrant
member of the Ceratoporellida, and removed from the Chaetetida.

Kazmierczak (1979) reported intramural monaxon spicules, replaced by pyrite, within a Lower
Cretaceous (Barremian) chaetetid, Chaetetopsis favrei (Deninger 1906) from the Crimea. Like many
Mesozoic chaetetids, C. favrei increases both by intramural offset and longitudinal (pseudoseptal)
fission. The former is not known to occur in Palaeozoic chaetetids (Sokolov 1962, p. 262). The walls
of C. favrei are anhedral calcite miscospar that is possibly a neomorphic overprint (Kazmierczak
1979, p. 101), and is dissimilar from the typical fascicular fibrous microstructure of Palaeozoic
chaetetids.

This is the first report of convincing intramural spicule pseudomorphs in a Palaeozoic chaetetid,
Chaetetes ( Boswellia ) mortoni sp. nov., from the Lower Carboniferous of north Wales, northern
England, and southern Scotland. Comparison is made with other chaetetids and sclerosponges, and a
model is developed for the mode of spicule preservation. Classification of the Chaetetida within the
Sclerospongidea is supported.

OCCURRENCE AND PRESERVATION OF MATERIAL

Eight colonies of this new species have recently been collected from the Lower Asbian (Upper
Dinantian), Tynant Limestone (Somerville, 1979) (Lower Brown Limestone of Morton, 1879) of
the Llangollen area, north Wales. Although this sclerosponge is a rare element in the brachiopod-
dominated macrofauna, it has been collected from a 20-m range of cyclic strata at sites over 4 km of
outcrop, and from the underlying scree. It occurs towards minor cycle bases, in subtidally deposited
argillaceous algal-foraminiferal packstones and grainstones. The colonies were rolled and some were
fragmented prior to burial. One colony has a pronounced micritic (?endolithic algal) rim on part of its

table 1. Table of extant and fossil sclerosponges with associated spicules showing their spicule form and
relationship to the basal calcareous skeleton. Mesozoic stromatoporoids of Schnorf (1960) and Yabe and
Sugiyama (1935) are omitted owing to their uncertain spicular nature. Species of Leiospongia d’Orbigny, 1850,
and Hartmanina Dieci et al. 1974, described by Dieci, Russo, and Russo (19746) are ommitted owing to the
absence of associated spicules.

Order symbols, Cer = Ceratoporellida, Tab = Tabulospongida, Unas = Unassigned, Mur = Muranida,
Ch = Chaetetida, Stp = Stromatoporoidea; calcareous microstructure symbols, Fascic. fib. = fascicular
fibrous, Agg. spher. = aggregated spherules, Microgran. = microgranular; original mineralogy symbols,
A = aragonite, Mg-cc. = high magnesian-calcite; spicule distribution symbols, I. = intramural, E. =
extramural, m. = subparallel to microstructure fibres, s. = subparallel to growth axis of skeleton, d. =
embedded only within the distal portion of the calcareous skeleton, r. = random; spicule type symbols,
* = megasclere, ** = microsclere; spicule mineralogy symbols, cc. = calcite, pyr. = pyrite, Fe ox. = iron
oxide.

CO -£D

oi-2

O ^ ID N O

+1 +1 +1 +1 41

fO O
<D 4—

O i-

</) ZJ _C=

>»*r- -O >> r-
-*-> -O -r- 4-» -O

O > CL O >

<d o

r>> 4->
-M C

uo <o

o o
s- S-
8- Q.

qDi

spun

ra

J®D

[«®Dc

dlSt HD d)S

snoa3D40J3"i Jnfn 3jssD|jj/

DIO Z OS 3 W

qJDD‘n|qjDDi Asa

DlOZOiVIVd

806

PALAEONTOLOGY, VOLUME 23

surface (PI. 102, fig. 6). These colonies are calcific, with a variable degree of microstructure alteration
and compaction distortion. Spicule pseudomorphs have been observed in five of these colonies.

Comparison with material in the British Museum (Natural History) (repository prefix BMNH),
Royal Scottish Museum (repository prefix R.S.M.) and Merseyside County Museum (repository
prefix LIV.C.M.), led to the discovery of a further five spicule-bearing specimens of the same species.
Two of these are partly silicified with intramural spicules locally replaced by silica.

SYSTEMATIC PALAEONTOLOGY

Class demospongiae Sollas, 1875
Sub-Class sclerospongidea Hartman and Goreau, 1972
Order chaetetida Okulitch, 1936
Family chaetetidae Milne-Edwards and Haime, 1850
Subfamily chaetetinae Milne-Edwards and Haime, 1850
Genus CHAETETES Fischer von Waldheim, 1830
Subgenus boswellia Sokolov, 1939

Type species. Chaetetes {Boswellia) boswelli Heritsch, 1932, ‘Upper Dibunophyllum Zone (D2)’ of
Ivovik, Serbia, U.S.S.R.

Diagnosis. Chaetetids with thickened irregular walls and rounded corners to lumina that may be
either irregular or subpolygonal. Increase by pseudoseptal and basal fission. Incomplete fission and
separation of pseudosepta into isolated columns occurs locally. Fascicular fibrous walls. Complete
tabulae, variable in distribution. Intramural spicules (originally siliceous) present in some.

Remarks. Palaeozoic chaetetids have been subdivided generically on gross calicle morphology
(Sokolov 1939, 1962). Sokolov (1939, p. 41 1) erected the subgenus Chaetetes ( Boswellia ) to include
chaetetids with ‘thickened irregular’ calicle walls and ‘undulate rounded’ lumina, also stating that C.
{Boswellia) ‘occupies an intermediate position between . . .’ Chaetetes and the meandrine genus
Chaetetipora Struve, 1898. Species of C. {Boswellia) show this intermediate relationship very clearly,
from the more prismatic thick-walled C. {B.) uniformis, Spiro 1961, and C. {B.) heritschi Sokolov, 1950,
to the irregular calicles of C. {B.) torquis Spiro, 1961, which is very similar to some of the less
meandrine chaetetiporinids, e.g. Chaetetipora agonia Sokolov, 1950. Although this classification is
accepted provisionally here the division of the Chaetetidae at a generic level requires further
clarification, that must now be based on an understanding of poriferan growth and variation.

EXPLANATION OF PLATE 102

Chaetetes {Boswellia) mortoni sp. nov., Lower Asbian (Lower Carboniferous, Eglwyseg Escarpment, Llangollen

(Clwyd, North Wales).

Fig. 1. Paratype BMNH R49965. Compaction-fractured colony with laminar overgrowth. Negative of
longitudinal section, x 2-3.

Fig. 2. Detail of a compaction-fractured thin-wall growth zone, BMNH R44965. Longitudinal section, x 45.

Fig. 3. Holotype BMNH R49964. Transverse section illustrating the subpolygonal to slightly irregular calicle
pattern, x 12.

Fig. 4. Transverse section illustrating an irregular calicle pattern, with many pseudosepta, scalloped margins to
lumina, and isolated trabecular column (Tr), BMNH R49965, x 35.

Fig. 5. Transverse section illustrating pseudoseptal fission in the corner of a calicle, BMNH R49964, x 40 (D =
Daughter calicle).

Fig. 6. Transverse section illustrating mosaic neomorphic fabric to calicle walls with a micritic rim. Outer margin
of BMNH R49964, x 70.

Fig. 7. Fractured tabula extending into lumen, longitudinal section, BMNH R49964, x 150.

Fig. 8. Sub-spherical ‘vacuole’ (V) within the calicle wall of BMNH R50133, x 100.

PLATE 102

GRAY, Palaeozoic chaetetid

PALAEONTOLOGY, VOLUME 23

Chaetetes ( Boswellia ) mortoni sp. nov.

Plates 102, 103; text figs. 1-4; Table 2

Derivation of species name. After G. H. Morton who devoted many years of research to the Carboniferous of
north Wales in the latter part of the nineteenth century.

Holotype. BMNH R49964, Tynant Limestone (Lower Asbian), quarried face, 400 m north of Tynant Ravine,
4 km north of Llangollen, Clwyd (National Grid Ref. SJ 21964573)

Paratypes. BMNH R4429 (Morton Collection), Lower Brown Limestone (in part equivalent to Tynant
Limestone), Llangollen, Clwyd (partly silicified); BMNH R49965, Tynant Limestone (Lower Asbian), quarried
face 500 m north of Tynant Ravine (SJ 21974582).

Other material. BMNH R50134, Tynant Limestone (Lower Asbian), World’s End, 6 km north of Llangollen,
Clwyd (SJ 23314789); BMNH R50133, BMNH R50135, and BMNH R50136 loose on scree slopes near Tynant
Ravine, near the base of the Eglwyseg escarpment; BMNH R50188 and BMNH R50189, Tynant Limestone,
300 m north of Llwyn Hen-parc Gulley, Eglwyseg escarpment (SJ 22152638); LIV.C.M. 1974. 57, Eglwyseg
Escarpment, Llangollen (in scree, partly silicified); BMNH R45851, Lower Carboniferous, Ravenstonedale,
Cumbria; BMNH R46144, J. S. Baker Collection, Carboniferous Limestone (Blue Quarries), Ashfell Edge,
Ravenstonedale, Cumbria; R.S.M. 1967.66.86-89 Nicholson Collection (thin sections only; all probably from
one colony), Carboniferous Limestone, Archer Beck, Dumfriesshire.

Range. ?Holkerian (BMNH R46144 from ?Ashfell Limestone) to Lower Asbian.

Diagnosis. Chaetetes ( Boswellia ) with irregular to subpolygonal (intracolonially variable) calicles. Fascicular
fibrous walls with pseudosepta and irregular longitudinal ridges. Isolated pseudoseptal columns locally. Lumen
diameter av. c. 500 ju.m; wall thickness av. c. 110 jam. Intramural spicules (monaxon tylostyle megascleres)
subparallel the fibres, diverging distally, with their pointed (oxeote) ends directed distally. Spicule diameter c. 7
/im; spicule length variable c. 275 jam in well-preserved specimens. Tabulae well spaced. Basal and pseudoseptal
fission only.

Description

Colony form. The colonies are laminar or bulbous, rarely greater than 12 cm diameter by 8 cm high. They display
distinct growth bands in rhythms 2 to 5 mm thick, with zones of thinner calicle walls preferentially compaction
fractured (PI. 102, figs. 1, 2). On weathered and polished surfaces the thick-wall bands stand prominent, being a
paler shade of brown-grey than the compaction fractured zones. Neither epitheca, astrorhizae, nor surface
mamelons have been observed on any colony.

Calicle morphology. The calicles are irregular to subpolygonal (PI. 102, figs. 3, 4), with the mean diameter of more
polygonal lumina between 420 ^m and 535 jam (see Table 2). The walls vary greatly in mean thickness, from c. 90
ju.m to c. 140 ^m (measurements taken between ridges and pseudosepta). Pseudosepta are common, occasionally
separating from the calicle walls as isolated columns. Both pseudosepta and ridges (undeveloped pseudosepta)
longitudinally ornament the calicle walls (text-fig. 1) imparting a scalloped appearance to the lumina in

''''04-

text-fig. 1 . Block diagram of Chaetetes ( Boswellia ) mortoni sp. nov. to illustrate the
development of longitudinal ridges (L), pseudosepta (P), and isolated trabecular
columns (Tr) off the calicle walls. Diagram constructed from serial acetate peels of
BMNH R50133.

GRAY: CHAETETID SPICULE PSEUDOMORPHS

809

transverse section (PI. 102, fig. 4). Increase is by both pseudoseptal and basal fission. Pseudoseptal fission
commonly occurs in calicle corners (PI. 102, fig. 5). Incomplete pseudoseptal fission locally forms an irregular
calicle pattern. Tabulae are rarely visible, appearing well spaced ( ^ 2 per mm), although this may in part be due
to the degree of compaction fracture (PI. 102, fig. 7).

Microstructure. The walls are fascicular fibrous penicillate calcite or chalcedonic silica, with a brown ‘dusty’
appearance in thin section due to submicroscopic to micrometre-sized inclusions of ?organic material. In the
calcitic specimens these inclusions vaguely define the wall fibres and cause a variable pseudopleochroism
(between paler and darker brown) cf. Hudson (1962). One specimen, BMNH R50133 has rare subspherical
‘vacuoles’, c. 50-/xm diameter, within the calicle walls, of uncertain origin (PI. 102, fig. 8). Neomorphism has
destroyed details of the microstructure to varying degrees (PI. 102, fig. 6; text-fig. 5), resulting in inclusion-poor
areas lacking spicule relicts (especially the thin-wall growth bands). The walls rarely show a coarser fibrous
fabric, with each fibre surrounded by thin brown pellicles that are probably the remnants of the ?organic
inclusions.

Spicule form. In the calcitic specimens, spicule pseudomorphs occur within the walls subparallel to the fascicular
fibres, diverging distally as straight or slightly curved elongate rods of clearer, inclusion-deficient calcite (PI. 103,
fig. 3) up to 300 (jm long. In transverse section they appear as clear calcite circles or ellipses with a range of mean
diameters between 6-6 /xm and 8-2 /xm (PI. 103, fig. 4). Rarely they may exceed 20 /xm diameter. Although surface
detail is not visible on these pseudomorphs, their clarity varies from prominent to indistinct, reflecting variation
in neomorphism. Rarely the spicules may be preserved as pyrite pseudomorphs with aggregates of pyrite crystals
along their length (text-fig. 4c), similar to those described by Kazmierczak (1979). In contrast some calicles of
BMNH R4429 and LIV.C.M. 1974.57 are partially replaced by chalcedonic silica (PI. 103, figs. 1, 2), with perfect
intramural silica spicule pseudomorphs occurring adjacent to more vague calcitic ones. These pseudomorphs are
low-relief colourless to high-relief red-brown translucent tylostyles, with circular cross-sections (PI. 103, fig. 6),
distinct bosses at their proximal ends and distally tapering points (PL 103, figs. 2, 8, 9, 10). As with the totally
calcitic specimens these spicules diverge distally in the calicle walls (PI. 103, fig. 2), subparalleling the fascicular
fibres, but occasionally cross-cutting the wall-fibre trend at a high angle. Fossilized early corrosion features are
seen on many spicule pseudomorphs (PI. 103, figs. 7, 8, 10, 11).

Seldom, and within BMNH R4429 only, ?raphide microsclere pseudomorphs occur (PI. 103, fig. 7), as thin
rods, pointed at both ends, c. 70 /xm by 3 /xm. These are only discernible in the chalcedonic regions, and would be
too small to distinguish within the microstructure of calcitic calicle walls.

Spicule distribution. Spicule pseudomorphs have a variable distribution within the colonies. In the calcitic
specimens they are only discernible within some of the ?organic-inclusion-rich areas associated with the less-
fractured thick-wall growth bands where neomorphism has most perfectly replaced the primary fabric (see text-
fig. 2). Specimens BMNH R4429 and LIV.C.M. 1974.57 also have a variable spicule distribution (see text-fig. 3)
dependent on the replacement mineralogy. In BMNH R4429 totally and partly calcitic calicles rarely have visible
calcitic spicule pseudomorphs. Here the microstructure is masked by a dense inclusion distribution. Siliceous
spicules occur where the outer zone of these calicle walls is silicified. Only two growth bands are completely
silicified (text-fig. 3), in which the best examples of a dense spicule distribution are visible (PI. 103, figs. 1, 2).

table 2. Variation in calicle and spicule size in six specimens of Chaetetes ( Bosxvellia ) mortoni

SPECIMEN

Average lumen
diameter in /xm

Average wall
thickness in /xm

Average spicule
length in /xm

Average spicule
width in /xm

mean

s.d.

n.

mean

s.d.

n.

mean

s.d.

n.

mean

s.d.

n.

*BMNH R49964

455

130

15

115

35

15

170

70

12

6.6

1.0

20

**BMNH R4429

420

140

10

94

29

10

275

50

10

6.9

0.9

10

**BMNH R49965

485

160

25

130

40

50

155

45

15

7.5

4.0

50

BMNH R50133

520

150

15

117

23

12

154

65

10

7.2

1.8

15

BMNH R50134

423

140

10

141

34

10

207

60

10

8.2

1.1

10

BMNH R50135

535

110

10

116

28

15

(Strongly

neomorphosed microstructure)

= Holotype; ** = Paratype; s.d. = standard deviation; n. = sample number.

810

PALAEONTOLOGY, VOLUME 23

text-fig. 2. Sketch showing the relationships of the microstructural
fabrics in calcitic specimens of Chaetetes ( Boswellia ) mortoni sp. nov. in
longitudinal section. Symbols: Ta = fractured tabula; Tr = isolated
trabecular column; L = longitudinal ridge or pseudoseptum; s = calcitic
spicule pseudomorph; Fr = fracture zone of thin-wall growth bands;
I = transition from ?organic-inclusion-dense thick-wall bands to
inclusion-poor compaction-fractured zone; c = neomorphic crystal
mosaic, with varying degrees of undulose and sweeping extinction often
not discernible in the ?organic-inclusion poor regions; m = localized
neomorphic fabric of vaguely fibrous microspar-size crystals, with thin
brown pelicles.

Discussion

The over-all similarity in colony, calicle, and microstructure form confirm that the calcitic and
siliceous specimens are conspecific (see Table 2).

Species of Chaetetes are poorly defined. The size and degree of variation in calicle form, wall
thickness, and tabula density are used as species-dependent characters. Many or all of these
characters may, however, be controlled by environmental influences (e.g. Weyer 1967), and therefore
where possible care must be taken to sample as large a population as possible. Variation within the
specimens of C. ( B .) mortoni described here is mostly intracolonial. Some similarities exist between
this and previously described species. The calicle morphology is locally (PI. 102, fig. 3) similar to C.
( B .) uniformis Spiro, 1961 from the ‘Visean’ of the Moscow region, but differs by having a higher
density of pseudosepta and slightly smaller calicles. Other closely comparable species are C. (B.)
torquis Spiro, 1961, a densely tabulate form that displays many ridge swellings on its calicle walls
(Spiro 1961, pi. 3, fig. In) in a similar manner to C. ( B .) mortoni, Chaetetipora agonia Sokolov, 1950
and the larger-calicled C. dubjanskyi Sokolov, 1950. The latter two both show pseudoseptal fission in
calicle corners (Sokolov 1950, pi. 15), but differ from Chatetes ( B .) mortoni with their indistinctly
meandrine form and dense tabula distribution. Compaction-fracture of C. ( B .) mortoni colonies,
however, imparts a superficial meandrine calicle pattern which may obscure the true calicle shape.

Chaetetipora etheridgeii (Thomson, 1881) is a variably meandrine species, characterized by a
variable calicle shape, some arranged in ‘sub-stellate groups radiating irregularly around a large
central . . .’ calicle (Thomson 1881, p. 208). It is densely tabulate, with calicles commonly from 0-5 to
2-0 mm diameter, and with thin calicle walls. Chatetes ( B .) mortoni may be readily distinguished from
this chaetetid by its lack of sub-stellate calicle-groups and its rare tabulae.

Whether the presence of intramural spicule pseudomorphs is a species-dependent factor is open to
question. Extant sclerosponges have growth-variable spicule distributions (Hartman and Goreau
1972, p. 213) and Stearn (1972), remarked on a whole population of the extant Astrosclera Lister,
1900 (unassigned sclerosponge), from the Pacific without spicules. The problem is further
complicated by the variable preservation of the spicules as pseudomorphs. In three calcitic specimens

GRAY: CHAETETID SPICULE PSEUDOMORPHS

8;

CALCITIC calicle walls
and spicule pseudomorphs

CHALCEDONIC SILICA calicle
walls and siliceous
spicule pseudomorphs

CHALCEDONIC SILICA calicle
walls and siliceous
spicule pseudomorphs

CHALCEDONIC SILICA outer
zones of calicle walls,
with mainly siliceous
spicule pseudomorphs

Dominantly CALCITIC
calicle walls and
spicule pseudomorphs

Key

CALCITIC FRACTURE ZONE

finnfinnHFiRfii

FRACTURE ZONE

^Cjl AL£E DO^N K^S TJJRE^ZO NEj

113 Diagrammatic representation of a longti tudinal section through a calicle wall

| (section c. 60pm thick). Approximate number of visible spicule pseudomorphs

indicated"?

Hatching indicates extent of si licifi cation.

text-fig. 3. Schematic diagram illustrating the spicule distribution within
the calicle walls of the partly silicified BMNH R4429. Note that the higher
numbers of visible spicule pseudomorphs per section of calicle wall occur
in the chalcedonic silica zones, which are themselves in part controlled by
the growth and fracture banding within the colony.

of C. ( B .) mortoni they are undetected (BMNH R50135, BMNH R50136, and BMNH R50189).
Therefore, although the spicular character is of great significance in understanding the phylogeny
and histology of chaetetids, it must only be used with caution as a specific character in fossil forms.

MINERALOGY AND DIAGENESIS OF C. (B.) MORTONI

Basal calcareous skeleton. Extant sclerosponges secrete both aragonite (e.g. ceratoporellids) and
high-magnesian calcite (e.g. tabulosponges) in their basal skeletons. Both are possible original
mineralogies for Palaeozoic chaetetids. Fossil ceratoporellids (Viezer and Wendt 1976), probable
sclerosponges (Dieci, Russo, and Russo 1974a), and stromatoporoids (Wendt 1975), that have
retained their original aragonitic mineralogy, and have suffered little diagenetic alteration (Scherer
1977), have all been recorded from the Upper Triassic.

The in situ transformation of aragonite to calcite (Bathurst 1964; Dodd 1966), observed in
Pleistocene scleractinian corals and molluscs (James 1974; Pingitore 1976; Wardlaw, Oldershaw, and
Stout 1978), produces a secondary fabric which retains some detail of the primary microstructure.
This transformation apparently occurs via a thin solution film less than 15 nm wide (Wardlaw et al.
1978, p. 1864) or a chalky solution zone (James 1974; Pingitore 1976). In this polymorphic
transformation, relict detail of microstructure is defined by organic, and rarely aragonite, inclusions
enclosed within a coarse mosaic of brown neomorphic calcite. Each of these coarse mosaic crystals
exhibits straight (James 1974, p. 793) or undulose (Schneidermann, Sandberg, and Wunder, 1972,
p. 88) extinction under crossed polars.

The transformation of high to low-magnesian calcite involves a paramorphic incongruent
dissolution process (Plummer and Mackenzie 1974, p. 79). Fine detail is preserved in skeletal
components during this transformation (e.g. Towe and Hemleben 1976), although ultrastructural
changes may be noted (e.g. Sandberg 1975). In comparison, Lohmann and Meyers (1977, p. 1086)
described milky skeletal calcite rich in microdolomite inclusions as evidence of an original

812

PALAEONTOLOGY, VOLUME 23

high-magnesian calcite mineralogy, apparently caused by an ‘incongruent dissolution or solid-stage
exsolution’ process re-equilibrating the magnesium within coarse crystals that acted as closed or
semi-closed systems (Meyers and Lohmann 1978) during the mineralogical transformation. Richter
and Fuchtbauer (1978) used the preservation of primary structures by ferroan calcite as a criterion for
recognition of primary, high-magnesian calcite.

In calcitic specimens of C. ( B .) mortoni a relationship between the crystal form and ?organic-
inclusion distribution is observed. The paler-brown calicle walls of specimens with less included
material comprise 50 p. m to 300 /xm mosaic crystals with irregular margins, and either straight or
slightly undulose extinction under crossed-polars. Most inclusion-rich areas have sweeping or
undulose extinction and some lack a crystal mosaic. Often the wall crystals are continuous with the clear
lumen-filling spar. No distinct wall-fibre boundaries are visible but vague fibre boundaries are defined
by trains of inclusions. Similarly, the margins of calcitic spicule pseudomorphs are indistinct and
often masked by these inclusions. Compaction fractured zones lack a dense inclusion distribution
and have a neomorphic mosaic which suggests that the transformation to calcite locally destroyed
much of the original microstructure. Larger surface areas of fractured calicle walls, exposed to the
calcifying pore waters, may have induced a more rapid and destructive mineralogical transformation.

In partly silicified specimens, the chalcedonic silica cementation and replacement occurred after
compaction fracture, as indicated by the preferential silicification of their fracture zones. According
to Meyers (1977), such compaction fracture could occur as a result of overburden pressure after
burial to metres or tens of metres.

The ?organic-inclusions within the calicle walls appear to have provided a template retaining some
detail of the original microstructure. The inclusion distribution may be partly related to the
diagenetic history of the calicle walls as they are invariably less dense in compaction-fractured zones,
and partly primary, caused by a variable secretion of organic matrix within the basal calcareous
skeleton. This reliance on inclusions to define the primary microstructure, and the presence of
neomorphic crystals with irregular margins, that often continue into the lumina as clear spar suggests
an in situ transformation from primary aragonite to calcite. Schneidermann et al. (1972, p. 89) stated
that continuity of neomorphic crystal fabrics from skeletal components into surrounding spar
indicated an early aragonite void-cementation that could be ‘expected to appear only in association

EXPLANATION OF PLATE 103

Chaetetes ( Boswellia ) mortoni sp. nov.. Lower Asbian (Lower Carboniferous), Eglwyseg Escarpment,

Llangollen (Clwyd, North Wales).

Fig. 1. Paratype BMNH R4429 (Morton Collection). Longitudinal section of calicle walls replaced by
chalcedonic (length fast) silica, and zoned dolomite lumen infills. Microgranular silica spicule pseudomorphs
visible as dark streaks within the calicle walls, x 50.

Fig. 2. BMNH R4429. Detail of chalcedonic silica calicle wall in longitudinal section, showing the spicule
pseudomorphs diverging distally, x 150.

Fig. 3. Holotype BMNH R49964. Longitudinal section of calicle wall illustrating well-preserved calcitic
spicule pseudomorphs, diverging distally and subparalleling the fascicular fibres, defined by trains of
inclusions, x 120.

Fig. 4. BMNH R50134. Transverse section of calicle, with distinct and vague transverse sections through
calcitic spicule pseudomorphs (s), x 150.

Fig. 5. BMNH R4429. Microgranular silica spicule pseudomorph extending to lumen void (l) calicle wall (w)
junction, indicating original extension of distal portion of spicule into the lumen, and its subsequent
dissolution, x200.

Figs. 6-11. Microgranular silica spicule pseudomorphs of BMNH R4429. 6, transverse section, x 1000. 7,
slight surface corrosion on tylostyle (c), with adjacent ?raphide-microsclere (Mi), x 200. 8, tylostyle with
discontinuity and replacement by dolomite rhomb (r), x 200. 9, perfect tylostyle pseudomorph. Note
proximal boss and distal point, x 200. 10, discontinuous tylostyle pseudomorph (probably a dissolution
feature), x200. 11, detail of a highly corroded tylostyle pseudomorph (c), x600.

PLATE 103

GRAY, Palaeozoic chaetetid

814

PALAEONTOLOGY, VOLUME 23

with previously aragonitic skeletons’. Conversely, in silicified specimens rare microdolomite
inclusions occur (PI. 103, fig. 8) within the calicle walls suggesting a high-magnesian calcite original
mineralogy. However, intense dolomitization of the lumina of these colonies, the lack of
microdolomite inclusions within calcitic specimens, and the common association of dolomite with
chert nodules in the Tynant Limestone indicates that the magnesium may have an external source.
Staining has not revealed any obviously ‘ferroan’ calcitic specimens. It would therefore appear that
aragonite is the most probable original mineralogy of C. ( B .) mortoni.

Spicule preservation. There are significant variations in the mode of preservation of the intramural
spicules. Extant sclerosponges secrete siliceous spicules of various types (see Table 1) within or
external to a calcareous skeleton. Hartman and Goreau (1970, pp. 210, 213) and Land (1976)
described live colonies of ceratoporellans in which the opaline silica spicules (Opal ‘A’ of Jones and
Segnit (1971)) were dissolving within the basal calcareous mass.

In C. ( B .) mortoni spicules are preserved as calcite, silica, and pyrite pseudomorphs. Most
specimens contain calcitic pseudomorphs. No spicule pseudomorphs convincingly extend into the
lumina, although many spicules would have originally done so, as shown by extrapolation of spicule
microstructure where it terminates abruptly against the calicle wall edge (PI. 103, fig. 5).

The siliceous spicule pseudomorphs of BMNH R4429 and LIV.C.M. 1974.57 show dissolution
features (PI. 103, figs. 7, 11) from slight surface pitting to deep corrosion. Often the spicule
pseudomorphs are discontinuous (PI. 103, figs. 8, 10). The origin of this last feature is uncertain, but
may be a severe localized corrosion effect. Dissolved parts of spicules have been replaced by clear
chalcedonic silica indicating that a degree of dissolution occurred prior to the silicification of calicle
walls. A later diagenetic event is indicated by dolomite rhombs replacing both spicule pseudomorphs
(PI. 103, fig. 8) and chalcedonic-silica walls.

In specimen BMNH R4429 neither axial canals nor axial filaments are visible, even with scanning
electron microscopy of HF etched specimens (text-fig. 4 b) (cf. Schwab and Shore 1971), indicating
that an internal alteration of the spicule mineralogy has occurred. This is further confirmed by the
presence of sub-microscopic microgranules which impart a red-brown hue and high relief to the
spicule pseudomorphs. Deeper coloration corresponds to a more dense microgranule distribution.
They give the spicule surface a smooth but frosted appearance in transmitted light. S.E.M. with
E.D.A.X. shows that the microgranules are siliceous, and indistinguishable from the surrounding
chalcedonic-silica walls. HF etched surfaces reveal the microgranule’s form (text-fig. 4b). They vary
between 0-2 pm and 0-5 pm diameter, and have sharp edges implying an internal structural ordering.

At an early or intermediate stage of diagenesis, biogenic opal ‘A’ is either converted in situ to opal

text-fig. 4. S.E.M. photomicrographs of spicule pseudomorphs in C. ( B .) mortoni'. 4a, longitudinal
section of calicle wall with siliceous spicule pseudomorphs preferentially etched (10% HF for 3 minutes),
BMNH R4429, x 1000; 4b, transverse section of a microgranular silica spicule pseudomorph (etched in
10% HF for 3 minutes), BMNH R4429, x 5000; 4c, pyritic spicule pseudomorph, composed of pyrite
crystal aggregates (HC1 etch), BMNH R46144, x 1000.

GRAY: CHAETETID SPICULE PSEUDOMORPHS

‘CT’ (Ernst and Calvert 1969; Wise and Weaver 1974, p. 305), a disordered cristobalite-tridymite
silica polymorph (Jones and Segnit 1971), or dissolved and subsequently reprecipitated as opal ‘CT’
or quartz, filling voids and replacing carbonate grains (von Rad, Riech, and Rosch 1978). With
increasing diagenetic maturity (depth of burial; time; temperature controls) opal ‘CT’ is eventually
converted to quartz (Reich 1979). There are few records of opal ‘CT’ from pre-Cretaceous sediments.

The mineralogical composition of the spicule pseudomorphs in C. (B.) mortoni invites discussion.
Are the spicules not pseudomorphs but diagenetically aged opal ‘A’ relic spicules? Reich (1979, pp.
754-755, pi. 1, fig. 4) reported Eocene opal ‘A’ sponge spicules with a pronounced microgranular
texture similar to that observed on HF etched spicules in BMNH R4429.

Opal ‘CT’ normally occurs as bladed microspherules or lepispheres, 3 iim to 15 pm diameter.
Siliceous sponge spicules have been recorded replaced by lepispheres that subsequently have been
inverted to quartz (Reich 1979, p. 742) leaving visible relics of the precursor lepisphere. Von Rad et al.
(1978, p. 903, pi. 3, figs. 2, 3) figured lepispheres replaced by microgranular quartz, also with a similar
texture to that observed in spicule pseudomorphs of BMNH R4429 (text-fig. 4b). No lepisphere relics
are visible in these spicules. Robertson (1978, p. 25) suggested that ‘domains’ of opal ‘CT’ would
‘presumably appear’ within opal ‘A’ ‘which had escaped early diagenetic dissolution, becoming
increasingly numerous as solid state ordering proceeds’. These ‘domains’ may be analogous to the
microgranules within the spicule pseudomorphs (i.e. opal ‘CT’ without a lepisphere stage) or they
may be a microgranular quartz replacement of opal ‘CT’ microgranules.

Opal ‘CT’ is generally recognized to have a higher relief than opal ‘A’, but exceptions are known.
The high relief of microgranule-dense spicules in BMNH R4429 and LIV.C.M. 1974.57 may be
caused by internal reflections on the surfaces of the microgranules. The spicules dissolve in HF far
more readily than the surrounding chalcedonic silica (text. fig. 4a). This may be due to their fine
granular nature, but Robertson (1978, p. 22) found that opal ‘CT’ dissolved preferentially in HF with
respect to chalcedonic silica.

Thus the origin of the microgranular silica within the spicule pseudomorphs is unclear. It may be
diagenetically aged opal ‘A’; microgranular quartz replacement of opal ‘CT’ or relic microgranular
opal ‘CT’ that has not gone through a lepisphere stage (see text-fig. 5). The nature of the silica matrix

DIAGENETIC

STAGE

Opal 'A' tylostyle spicules secreted.
Spicules incorporated within the growing
?aragonitic walls.

Growth of colony.

Death.

Burial to metres or tens of metres.

Compaction fracture by pressure of overburden.

R4429 & LIV.C.M. 1974.57 /

Increase in silica concentration of pore waters.

Length-fast chalcedonic silica infills lumina, compaction
fractures and spicule corrosion voids simultaneously
replacing calicle walls and terminating spicule dissolution

Remaining opal'A' spicules '
undergo in situ transformation:

microgranular*.
quartz ^

v^opal 1 CT '

^microgranular opal'CT'
V^aged, microgranular
opal'A'

To microgranular silica

Dolomitisation of:
Lumen infill &

Growth of

outer ferroan zones
to dolomite crystals.'

Commencement

of

spicule dissolution:

Loss of lumen penetrating
oxeote ends; local corrosic

Continued
and

total spicule dissolution.

,In situ thin film transformation of taragonite
/sFeTeton to a coarse calcite mosaic, associated
wi th :

Migration of ?organic-inclusions;
variable detail retained of primary
microstructure; infill of spicule voids
with calcite and or pyrite; replacement
of earlier void-filling cements by
calcite; microsparitisation of sediments.

Further burial; final pore-filling cementation.

text-fig. 5. Chart illustrating the sequence of events in the diagenetic history of Chaetetes ( Boswellia ) mortoni sp.
nov. in relation to the preservation of microstructures and spicule pseudomorphs.

816

PALAEONTOLOGY, VOLUME 23

to the microgranules is unknown. In spicule pseudomorphs with few microgranules the matrix is
optically similar to the surrounding chalcedonic silica. The occurrence of this microgranular silica
fabric within the spicule pseudomorphs and not within the surrounding chalcedonic silica walls, and
the retention of delicate spicule corrosion features suggest that the spicules were not originally
calcareous, but probably opal ‘A’, as in extant sclerosponges.

In specimens BMNH R46144, BMNH R50188, and BMNH R45851 pyritic spicule pseudomorphs
occur, similar to those described by Kazmierczak (1979). The pyrite replacement is imperfect, with
microcrystalline pyrite (text-fig. 4c) forming discontinuous chains or aggregates of 1 ^ m to 7 ^m
crystals along the pseudomorph length. The non-pyritic parts of the pseudomorphs are replaced by
calcite. In BMNH R45851 the spicules are pseudomorphed by pyrite only towards the outer edges of
calicle walls, where they merge with highly pyritic lumen-filling sediment. In contrast, BMNH
R50188 has void-filling spar in the lumina, with pyritic spicule pseudomorphs terminating abruptly
at the calicle wall margins, indicating that pyritization occurred after at least partial spicule
dissolution. Rickard (1970) suggested that framboidal pyrite may replace organic globules or infill
gaseous vacuoles. The calcitic spicule pseudomorphs often contain ?organic-inclusions, which may
play a role in the pyrite formation.

AFFINITIES OF C. ( B .) MORTONI WITH CERATOPORELLIDS

Ceratoporellids secrete a basal calcareous skeleton of fascicular fibrous aragonite calicles. These are
subsequently infilled with fibrous aragonite by the upward-growing basal pinacoderm (Hartman and
Goreau 1970, 1972). The calicles of Ceratoporellanicholsoni are regular, rounded to polygonal, rarely
with a meandrine form (Hartman and Goreau 1972, fig. 17) similar to that of Stromatospongia
Hartman, 1969, another extant ceratoporellid genus. Unlike chaetetids, ceratoporellids do not
secrete tabulae. However, as Hartman and Goreau (1972, p. 142) point out, ‘the difference is one of
degree’, with the growth of tabulae representing periodic rather than continuous carbonate secretion
from the basal pinacoderm of the sponge animal. Atrochaetetes, a Mesozoic ceratoporellid (see
above), lacks a continuous calicle infill (Cuif and Fischer 1974) and exhibits intramural spicule
pseudomorphs in at least one species (Dieci, Russo, and Russo 1977). Its backfill may have formed by
periodic secretion, or by periodic distally directed movement of the living tissues, and may be a
character intermediate between solid backfills and tabulae.

The calicle surfaces of ceratoporellids are often ornamented with arborescent processes, rounded
knobs, and spines of aragonite. No detailed calicle surface is available on Chaetetes ( B .) mortoni for
comparison; however, longitudinal sections show the distal edges of the calicles as rounded, although
pre-burial erosion may have enhanced this. Isolated aragonitic trabeculae grow within the soft tissue
of Ceratoporella, and are subsequently incorporated within the calcareous walls (Hartman and
Goreau 1972, p. 13 5). These may be compared to the trabecular columns within C. (B.) mortoni which
remain isolated during growth. Surface mamelons and astrorhizae, evident in some specimens of
ceratoporellids as a result of differential growth-rates beneath excurrent canal systems (Stearn 1975),
are not present on studied specimens of C. ( B .) mortoni.

Ceratoporellids increase by pseudoseptal division, and Palaeozoic chaetetids by both pseudoseptal
and basal fission. Mesozoic chaetetids in contrast also increase by intramural offset, as do
tabulosponges.

Opal ‘A’ spicules are secreted from scleroblast cells within the living tissue of ceratoporellids and
are incorporated within the skeleton as the colony grows. Hartman and Goreau (1972, p. 134) state
that the spicules of Ceratoporella nicholsoni ‘entrapped in the aragonite tend to follow the orientation
of the calcareous crystalline units that surround them’. This is very like C. (B.) mortoni. The proximal
(basal) spicule heads in living ceratoporellids are embedded within an organic matrix (Hartman and
Goreau 1970). Although there is no direct evidence for organic fibres surrounding the head of
spicules in Chaetetes ( B .) mortoni the presence of ?organic-inclusions indicates an intimate
relationship between the organic, calcareous, and siliceous components of the skeleton. Hartman and
Goreau (1970, p. 213) also note that there are regions of the calcareous skeleton of Ceratoporella

GRAY: CHAETETID SPICULE PSEUDOMORPHS

817

nicholsoni devoid of siliceous spicules. Although spicule preservation is variable throughout the
colonies of Chaetetes ( B .) mortoni, the local variation in spicule distribution may in part be primary.
Ceratoporellids secrete monaxons of various forms, although they are neither known with tylostyles,
nor with microscleres. The size of C. ( B .) mortoni tylostyles does, however, fall within the size range of
known ceratoporellid spicules.

The ecology of extant sclerosponges is fundamentally different from that of fossil chaetetids.
Extant ceratoporellids are commonly found in a complex association with serpulid worms in
submarine caves and at depth on fore-reef slopes (Hartman and Goreau 1970), whereas Palaeozoic
chaetetids are common open-shelf dwellers, often associated with shallow-water carbonates.

There are significant similarities between C. ( B .) mortoni and ceratoporellids in colony, calicle,
microstructure form, and spicule character, indicating a close phylogenetic relationship. One notable
difference is the presence of true tabulae in C. ( B .) mortoni and the solid calcareous calicle infill
characteristic of extant ceratoporellids.

AFFINITIES OF C. ( B .) MORTONI WITH OTHER SCLEROSPONGES

Tabulospongids (Hartman and Goreau 1975; Mori 1976, 1977) secrete a calicular basal skeleton of
high-magnesian calcite with a lamellar microstructure. The calicles are partitioned by horizontal
tabulae, and spiny processes project into the lumina. The distribution of organic fibrils within the
skeleton of tabulospongids is documented by Hartman and Goreau ( 1 975, p. 1 67). These nanometre-
sized fibrils act both as a matrix for the calcitic skeleton, and are present within the soft tissues. In
Tabulospongia horiguchii, Mori (1976, pi. 3, fig. 3) shows that calicle wall centres are richest in organic
matrix, resembling the distribution of probable organic matter now visible in calcitic specimens of C.
(2?.) mortoni. This may be a relict primary texture in the latter.

Tabulospongids secrete siliceous spicules that are not incorporated within their basal calcareous
skeleton, but remain within the surface tissues. These spicules have a very low fossilization potential.
The spicules are a variety of complex forms, with both megascleres and microscleres secreted by the
same colony (see Table 1).

Although the calcareous skeleton is similar in design to many chaetetids, the marked differences in
microstructure and spicule form and distribution readily distinguish tabulospongids from such
Palaeozoic chaetetids as C. ( B .) mortoni.

Merlia is an unassigned extant sclerosponge that secretes a prismatic tabular basal aragonitic
skeleton,partitionedhorizontallyby incomplete tabulae. A variety ofsiliceousspiculesaresecretedwith-
in the living tissue, but not incorporated within the skeleton. Each prismatic calicle is formed by the
outgrowth and interlocking of flanges, set at 1 20°, off stout fibre fascicles which form the calicle corners
(see Stearn 1975). The architecture of Merlia therefore is subtly different from that of chaetetids. In
contrast the types of spicules secreted by Merlia are similar to those of C. ( B .) mortoni (see Table 1)
(tylostyles and raphides) indicating some histologic similarities between these sponge animals.

Some Mesozoic stromatoporoids also have similarities with C. ( B .) mortoni. Schnorf (1960) and
Yabe and Sugiyama (1935) described Lower Cretaceous and Upper Jurassic forms with clear areas
within the walls which may be sites of intramural spicules (Hartman and Goreau 1 970), or part of the
primary calcareous microstructure (Fenninger and Flajs 1974). The calcareous skeleton of these
sclerosponges resembles chaetetids in as much as they also possess calicles with tabulae and hollow
lumina. They are a diverse group, however, and show many characters atypical of Palaeozoic
chaetetids.

In addition Table 1 lists the other known forms of spicule-bearing sclerosponges. These are not
closely comparable with the present material but indicate the variety of microstructural and
morphological patterns thus far encountered within the Sclerospongidea.

CONCLUSIONS

C. ( B .) mortoni sp. nov. is a Palaeozoic (Upper Dinantian) chaetetid. The calicle morphology of this
chaetetid is variable between slightly irregular (chaetetiporinid) and subpolygonal, having irregularly

PALAEONTOLOGY, VOLUME 23

thick fascicular fibrous walls, ornamented with longitudinal ridges and pseudosepta. These factors
place it within the subgenus Boswellia Sokolov, 1949, a typical Palaeozoic chaetetid.

Comparison of the basal calcareous skeleton with extant and fossil sclerosponges, the presence of a
neomorphic mosaic, and the dependence on ?organic-inclusions to define the primary microstructure
which is variably preserved, suggest that the original calcareous mineralogy was aragonite. Text-fig. 5
summarizes the approximate sequence of diagenetic events related to the preservation of
microstructures in the skeletons of C. ( B .) mortoni.

Spicule pseudomorphs occur within the fascicular fibrous walls as long thin tylostyles, with distally
diverging oxeote ends that often would have penetrated the lumen of the sponge animal. They occur
now as calcite, pyrite, or silica pseudomorphs, their mineralogy dependent on the diagenetic history
of the basal skeleton.

In calcitic specimens, the spicule pseudomorphs are preserved as clear calcite rods in regions of the
calicle walls with a dense ?organic-inclusion distribution. They are absent from compaction-
fractured zones and are more common in thick-wall zones of growth rhythms. More rarely, the
spicules are defined by trains or aggregates of pyrite crystals.

Chalcedonic silica locally replaces the calicle walls, enveloping spicule pseudomorphs that retain
detail of their tylostyle form. Early dissolution features are fossilized within these spicules. Voids
formed by spicule dissolution are infilled with clear chalcedonic silica. Remaining spicules have
undergone alteration to microgranular silica of three possible forms, either diagenetically aged opal
‘A’, microgranular quartz replacement of opal ‘CT\ or relic microgranular opal ‘CT’ that has not
gone through a lepisphere stage.

The presence of intramural spicule pseudomorphs within an otherwise typical member of the
Palaeozoic Chaetetida further supports the sclerosponge affinities of at least some members of this
group. Comparison of this chaetetid with other sclerosponges indicates that the spicular character
and calcareous microstructure is very similar to that of the Ceratoporellida. The secretion of tabulae
in chaetetids, rather than the backfill of ceratoporellids, remains the distinguishing microstructural
feature. Of the two previously described intramural spicule pseudomorph bearing Mesozoic
‘chaetetids’ Atrochaetetes Cuif and Fischer, 1974, may be regarded as an aberrant member of the
Ceratoporellida rather than a chaetetid s.s., on account of its discontinuous backfill.

Acknowledgements. I would like to thank: Dr. C. T. Scrutton (University of Newcastle upon Tyne) and Dr. M. E.
Tucker (University of Newcastle upon Tyne) for their advice and constructive criticism through all the stages of
preparation of this paper; Dr. B. R. Rosen, Mr. R. F. Wise, Ms. J. G. Darrell (Department of Palaeontology,
British Museum (Natural History)), Dr. C. D. Waterston (Royal Scottish Museum), and Mr. P. Phillips
(Merseyside County Museum) who brought fossil material to my attention; and Mr. G. Staines (University of
Newcastle upon Tyne) for his technical assistance with the electron microscopy. This research was carried out
during the tenure of a N.E.R.C. research studentship.

REFERENCES

Bathurst, R. G. c. 1964. The replacement of aragonite by calcite in the molluscan shell wall. In imbrie, j. and
Newell, N. (eds.), Approaches to palaeoecology. New York, J. Wiley & Sons, Inc., pp. 357-376.
cuif, j.-p. and fischer, j.-c. 1974. Etude systematique sur les Chaetetida du Trias de Turquie. Ann Is Paleont.
Invertebres, 60, 3-23, pis. I-IV.

dieci, G., russo, a. and russo, f. 1974a. Nota preliminare sulla microstruttura di spugne aragonitiche del Trias
medio-superiore. Boll. Soc. paleont. ital. 13, 99-107, pis. 32-37.

— 1974 b. Revisione del genere Leiospongia d’Orbigny (Sclerospongia triassica). Ibid. 135-146,

pis. 51-53.

— and marchi, m. s. 1977. Occurrence of spicules in Triassic chaetetids and ceratoporellids.
Ibid. 16, 229-238, pis. 1-3.

dodd, j. r. 1966. Processes of conversion of aragonite to calcite with examples from the Cretaceous of Texas.
J. sedim. Petrol. 36, 733-741.

ernst, w. G. and calvert, s. e. 1 969. An experimental study of the recrystallization of porcelanite and its bearing
on the origin of some bedded cherts. Am. J. Sci. 261 A, 1 14-133.

GRAY: CHAETETID SPICULE PSEUDOMORPHS

819

fenninger, a. and flajs, G. 1974. Zur mikrostruktur rezenter und fossiler Hydrozoa. Biomineralization, 7,
69-99.

fischer, J.-c. 1970. Revision et essai de classification des Chaetetida (Cnidaria) post-Paleozoiques. Annls
Paleont. Invertebres, 56, 151-220, pis. a-f.

fischer von Waldheim, G. 1830. Oryctographie du gouvernement de Moscou, Moscow, 1st edn. 1830, 2nd edn.
1837.

hartman, w. d. 1969. New genera and species of coralline sponges (Porifera) from Jamaica. Postilla, 137, 1 -39.

— and goreau, t. f. 1970. Jamaican coralline sponges: their morphology, ecology and fossil relatives. Symp.
zool. Soc. Lond. 25, 205-243.

— 1972. Ceratoporella (Porifera: Sclerospongiae) and the chaetetid ‘corals’. Trans. Conn. Acad. Arts
Sci. 44, 133-148.

— 1975. A pacific tabulate sponge, living representative of a new order of sclerosponges. Postilla, 167,

1-14.

heritsch, F. 1932. Chaetetes und Caninia aus dem Karbon von Ivovik bei Krupanj in West-Serbien. Bull. Serv.
Geol. Yougosl. 1/2, 1-10, pis. I II.

Hudson, J. d. 1962. Pseudopleochroic calcite in recrystallized shell-limestones., Geol. Mag. 92, 492-500, pi. XXI.
james, n. p. 1974. Diagenesis of scleractinian corals in the subaerial vadose environment. J. Paleont. 48, 785-799.
jones, J. B. and segnit, e. r. 1971. The nature of opal. I. Nomenclature and constituent phases. J. geol. Soc. Aust.
18, 57-68.

kazmierczak, J. 1974. Lower Cretaceous sclerosponge from the Slovakian Tatra Mountains. Palaeontology, 17,
341-347, pis. 45-46.

— 1979. Sclerosponge nature of chaetetids evidenced by spiculated Chaetetopsis favrei (deninger 1906) from
the Barremian of Crimea. Neues Jb. Geol. Paldont. Mb. 1979, 97-108.

— and hillmer, G. 1974. Sclerosponge nature of the Lower Hauterivian ‘bryozoan’ Neuropora pustulosa
(roemer 1839) from western Germany. Acta palaeont. Pol. 19, 443-453.

kirkpatrick, r. 1909. Notes on Merlia normani Kirkp. Ann. Mag. nat. Hist. (8) 4, 42-48.

1910. On the affinities of Astrosclera willeyana Lister. Ibid. (8) 5, 380-383.

— 1912a. Merlia normani and its relation to certain Palaeozoic fossils. Nature, London. 89, 502-503.

— 1912A On the nature of stromatoporoids. Ibid. 607.

land, L. s. 1976. Early dissolution of sponge spicules from reef sediments, North Jamaica. J. sedim. Petrol. 46,
967-969.

levi, c. 1973. Systematique de la classe des Demospongia (Demosponges). In grasse, p. (ed.). Traite de
Zoologia, Masson, Paris. Vol. 3, 577-3631.

lohmann, k. c., and meyers, w. j. 1977. Microdolomite inclusions in cloudy prismatic calcites: A proposed
criterion for former high-magnesium calcites. J. sedim. Petrol. 47, 1078-1088.
meyers, w. J. 1977. Chertification in the Mississippian Lake Valley Formation, Sacramento Mountains, New
Mexico. Sedimentology, 24, 75-105.

— and lohmann, K. c. 1978. Microdolomite-rich syntaxial cements: proposed meteoric-marine mixing zone
phreatic cements from Mississippian limestones, New Mexico. J. sedim. Petrol. 48, 475-488.

milne-edwards, H. a. and haime, J. 1850-1854. A monograph of the British fossil corals. Palaeontogr. Soc.
(Monogr.) Parts 1-5, i-lxxxv, 1-322, pis. I-LXII.

mori, K. 1976. A new recent sclerosponge from Ngargol, Palau Islands and its fossil relatives. Sci. Rep. Tdhoku
Univ., 2nd ser. (Geol.), 46, 1-9, pis. 1-6.

— 1 977. A Calcitic sclerosponge from the Ishigaki-shima coast, Ryukyu Islands, Japan. Ibid. 47, 1 -5, pis. 1 -2.
morton, G. H. 1879. The Carboniferous Limestone and Cefn-y-Fedw Sandstone of the country between

Llanymynech and Minera North Wales, 140 pp. David Bogue, London.

Newell, N. D. 1935. Some mid-Pennsylvanian invertebrates from Kansas and Oklahoma: II. Stromatoporoidea,
Anthozoa and Gastropoda. J. Paleont. 9, 341-355, pis. 33-36.
okulitch, v. J. 1936. On the genera Heliolites, Tetradium and Chaetetes. Am. J. Sci. 32, 361-379.
pingitore, N. e. jr 1976. Vadose and phreatic diagenesis: Process products and their recognition in corals.
J. sedim. Petrol. 46, 978-984.

plummer, l. n. and Mackenzie, F. T. 1974. Predicting mineral solubility from rate data: application to the
dissolution of magnesian calcites. Am. J. Sci. 274, 61-83.
rad, u. VON, riech, v. and rosch, h. 1978. Silica diagenesis in continental margin sediments off Northwest
Africa. In LANCELOT, Y., seibold, e., et al. 1978. Initial reports of the deep sea drilling project. Washington (U.S.
Government Printing Office), 41, 879-905.

riech, v. 1979. Diagenesis of silica, zeolites, and phyllosilicates at sites 397 and 398. In von rad, u., ryan, w. b. f..

820 PALAEONTOLOGY, VOLUME 23

et al. 1979. Initial reports of the deep sea drilling project. Washington (U.S. Government Printing office), 47,
741-759.

richter, D. K. and fuchtbauer, H. 1978. Ferroan calcite replacement indicates former magnesian calcite
skeletons. Sedimentology, 25, 843-860.
rickard, D. T. 1970. The origin of framboids. Lithos, 3, 269-293.

robertson, a. h. F. 1977. The origin and diagenesis of cherts from Cyprus. Sedimentology, 24, 11-30.
Sandberg, p. a. 1975. Bryozoan diagenesis: bearing on the nature of the original skeleton of rugose corals.
J. Paleont. 49, 587-606.

scherer, m. 1977. Preservation, alteration and multiple cementation of aragonitic skeletons from the Cassian
Beds (U. Triassic, Southern Alps): petrographic and geochemical evidence. NeuesJb. Geol. Palaont. Abh. 154,
213-262.

schneidermann, N., Sandberg, p. a., and wunder, s. J. 1972. Recognition of early cementation of aragonitic
skeletal carbonates. Nature phys. Sci., Lond. 240, 88-89.
schnorf, a. 1 960. Parastromatoporoidae nouveaux du Jurassique superieur et du Valanginien inferieur du Jura.
Eclog. geol. Helv. 53, 729-732, pis. I III.

Schwab, d. w. and shore, r. e. 1971. Fine structure and composition of a siliceous sponge spicule. Biol. Bull,
mar. biol. Lab., Woods Hole, 140, 125-136.

scrutton, c. T. 1979. Early fossil cnidarians. In house, m. r. (ed.), Spec. Vol. Syst. 12, 161-207.

Sokolov, B. s. 1939. Stratigraphical value and types of Chaetetidae of the Carboniferous of the USSR. C.r.
(. Dokl .) Acad. Sci. URSS 23, 409-412.

— 1950. Khetetidy karbona severo-vostochnoi Ukrainy i sporedel’nykh oblastei (s opisaniem nekotorykh
tabulyat). [Carboniferous Chaetetidae of the Northeast Ukraine and adjacent areas.] Trudy vses. neft.
nauchno-issled. geol.-razv. Inst. n.s. 27, 1-101, pis. I-XX. [In Russian.]

— 1962. Gruppa Chaetetida. In orlov, y. a. (ed.), Osnovy Paleont ologii: Porifera, Archaeocyatha,
Coelenterata, Vermes. Moscow [in Russian]. Fundamentals of Paleontology, 2, 259-270. Israel Prog. sci.
Transl., Jerusalem 1971 [English translation].

Somerville, i. d. 1979. A sedimentary cyclicity in early Asbian (Lower Dl) limestones in the Llangollen district
of North Wales. Proc. Yorks, geol. Soc. 42, 397-404.

spiro, t. a. 1961. Nekotorie Vizeiskie y Namyurskie chetetidi Podmoskovbya. [Some Visean and Namurian
chaetetids from the Moscow region.] Paleont. Zh. 4, 24-28, pis. II— III. [In Russian.]
stearn, c. w. 1972. The relationship of the stromatoporoids to the sclerosponges. Lethaia, 5, 369-388.

1975. The stromatoporoid animal. Lethaia, 8, 89-100.

termier, h. and termier, G. 1976. Spongiaires hypercalcifies et ectoproctes stenolemes fossiles. C.r. Acad. Sci.
Paris, 282D, 1269-1272.

Thomson, j. 1881. On the genus Alveolites, Amplexus and Zaphrentis, from the Carboniferous System of
Scotland. Proc. Phil. Soc. Glasgow, 13, 194-237.

towe, K. M. and hemleben, c. h. 1976. Diagenesis of magnesian-calcite: evidence from miliolacean foraminifera.
Geology, 4, 337-339.

twitchell, g. b. 1929. The structure and relationships of the true stromatoporoids. Am. Midi. Nat. 11, 270-306,

pis. xvni-xxyi.

vacelet, j. 1977. Eponges Pharetronides actuelles et Sclerosponges de Polynesie francaise, de Madagascar et de
la Reunion. Bull. Mus. natn. Hist nat. 3e ser., 444, Zool. 307, 345-368.

— vasseur, p. and levi, c. 1976. Spongiares de la pente externe des recifs coralliens de Tulear (Sud-
Ouest de Madagascar). Mem. Mus. natn. Hist. nat. A44, 1 16 pp.

veizer, J. and wendt, j. 1 976. Mineralogy and chemical composition of Recent and fossil skeletons of calcareous
sponges. Neues Jb. Geol. Palaont. Mh. 1976, 558-573.

wardlaw, N., oldershaw, A. and stout, m. 1978. Transformation of aragonite to calcite in a marine
gastropod. Can. J. Earth Sci 15, 1861-1866.

wendt, J. 1975. Aragonitische Stromatoporen aus der alpinen Obertrias. Neues Jb. Geol. Palaont. Abh. 150,
111-125.

weyer, von d. 1967. Einige Chaetetidae (Coelenterata) aus dem Unterkarbon des Velberter Sattels (Rheinisches
Schiefergebirge). Geologie, 16, 1 156-1163, pis. I— II.

wise, s. w. and weaver, f. w. 1974. Chertification of oceanic sediments. In hsu, k. j. and jenkyns, h. c. (eds.).

Pelagic sediments: on land and under the sea. Spec. Pubis int. Ass. Sediment. 1, 301-306.
yabe, h. and sugiyama, T. 1935. Jurassic stromatoporoids from Japan. Sci. Rep. Tohoku imp. Univ., ser. 2
(Geol.), 14, 135-191, pis. XL-LXXI.

DAVID I. GRAY
Department of Geology

T ypescript received 1 2 September 1979 University of Newcastle Upon Tyne

Revised typescript received 21 January 1980 Newcastle Upon Tyne, NE1 7RU

ORIGIN, EVOLUTION AND SYSTEMATICS OF
THE CRETACEOUS AMMONITE SPATHITES

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

Abstract. Spathites Kummel and Decker, 1954, a predominantly early to mid-Turonian genus common in
Tethyan regions of both the Old and New Worlds, is a key genus in the early radiation of the Mammitinae (of
which the Metoicoceratinae and Fallotitinae are shown to be synonyms). Spathitoides Wiedmann, 1960 is a
strict synonym and Jeanrogericeras Wiedmann, 1960 (of which Fallotites Wiedmann, 1960 is a synonym) no
more than subgenerically distinct. The genus evolved from Metoicoceras\ a succession is demonstrated from 5.
{Jeanrogericeras) to S. {Spathites). Mammites is an early offshoot from the former subgenus, whilst the
Coniacian Buchiceras, previously referred to the Tissotiidae, is a direct descendant of the S. {S.) rioensis Powell
to S. {S.) chispaensis Kummel and Decker to S. (S.) puercoensis (Herrick and Johnson) lineage.

Lower Turonian successions, especially those of Tethyan regions, have yielded great numbers of
ammonites with a reduced acanthoceratid ornament, currently referred to Spathites and Spathitoides
of the Vascoceratinae, Fallotites of the Fallotitinae, and Jeanrogericeras of the Mammitinae. The
type specimens of the type species are mostly distinct enough, but it has proved difficult to allocate
a number of other species to one or other genus.

This is particularly true where large subsequent collections have shown that the range of
intraspecific variation in some forms spans currently accepted generic limits.

Study of the earliest described species of the group, Ammonites reveliereanus Courtiller, 1 860, as
part of our over-all revision of the ammonite fauna of the Turonian stratotype, provides a basis for
the discussion of all the above genera, which are regarded as members of the subfamily Mammitinae,
in direct lineal descent from the late Cenomanian Metoicoceras.

SYSTEMATIC PALAEONTOLOGY

Repositories of material. These are indicated as follows: OUM, University Museum, Oxford; MNHP, Museum
d’Histoire Naturelle, Paris; SP, Sorbonne Collection, now housed in the Universite Paris VI; FSR, Faculte des
Sciences, Rennes; AM, Museum d’Histoire Naturelle, Angers; CS, Chateau de Saumur; UT, Texas Memorial
Museum, Austin; WW, C. W. Wright coll.

Suture terminology. The suture terminology of Wedekind (1916) (see Kullman and Wiedmann 1970 for a recent
review) is followed here: I = Internal lobe, U = Umbilical lobe, L = Lateral lobe, E = External lobe.
Dimensions. All dimensions are given in millimetres, figures in parentheses being the percentage of the total
diameter. D = diameter, Wb = whorl breadth, Wh = whorl height, U = umbilicus.

Superfamily acanthocerataceae Grossouvre, 1894
Family acanthoceratidae Grossouvre, 1894
Subfamily mammitinae Hyatt, 1900

(= Metoicoceratinae Hyatt, 1903; Fallotitinae Wiedmann, 1960)

Discussion. The systematics of the latest Cenomanian and Turonian Acanthoceratidae are in a state
of flux, but a classification at subfamily level is emerging that will accord with phylogeny. The earliest
Acanthoceratidae known are Mantelliceras, subfamily Mantelliceratinae, which are directly
descended caenogenetically from species of Stoliczkaia (Lyelliceratidae, Stoliczkaiinae) with
bituberculate venter on the inner whorls. Mantelliceratinae, characterized by dominant, more or less

[Palaeontology, Vol. 23, Part 4, 1980, pp. 821-837, pis. 104-106.|

822

PALAEONTOLOGY, VOLUME 23

sharp ribs, appear to have persisted through the Cenomanian and seem to have produced forms with
trituberculate venters. Acanthoceratinae characterized by trituberculate venter and a tendency to
coarse rounded ribbing and tuberculation, first appeared early in the Lower Cenomanian, but later
than Mantelliceras, with Acompsoceras, whose origin is uncertain. Its strong ventrolateral and weak
siphonal tubercles, the latter normally present on the inner whorls only, suggest that its most
probable source lies in Mantelliceratinae, but trituberculate-ventered Stoliczkaia, S. ( Lamnayella )
Wright and Kennedy, persist into the Lower Cenomanian and it is conceivable that here lies the
origin of Acanthoceratinae.

During the mid Cenomanian the subfamily gave rise to the Euomphaloceratinae (Cooper 1978;
Kennedy, Wright, and Hancock 1980a), a line which ended with Romaniceras in the late Turonian.
Both Acanthoceratinae and Euomphaloceratinae are characterized by, amongst other features, a
row of siphonal tubercles. Late in the Cenomanian, Thomelites, itself probably derived from some
siphonally tuberculate member of the Mantelliceratinae, began to lose its siphonal tubercles and gave
rise to Metoicoceras, in which the venter is bituberculate in all but the early stages of early species such
as M. praecox Haas and M. latoventer Stephenson.

Hyatt (1903, p. 115) erected a family Metoicoceratidae for this genus (and placed it in his
superfamily Mantelliceratida), but he had previously (1900, p. 588) established a family Mammitidae
(in his superfamily Mammitida) for another stock with bituberculate venters. The two families were
reduced to subfamily status within the Acanthoceratidae by Wright and Wright (1951, p. 24). Re-
examination of the Lower Turonian acanthoceratids with bituberculate venters suggests an even
simpler situation with a lineal phyletic succession Thomelites -> Metoicoceras -> Spathites
{Jeanrogericeras) [ = Fallotites] -> S. ( Spathites ) -> Buchiceras and an offshoot branch Spathites
(Jeanrogericeras) — > Mammites — > Metasigaloceras. All these genera are closely related morpho-
logically and we consider that they should all be placed in one subfamily Mammitinae. We maintain
Mammitinae within the Acanthoceratidae because there are no sufficiently important distinguishing
features present throughout ontogeny to separate it further from the other subfamilies.

Genus spathites Kummel and Decker, 1954
[= Spathitoides Wiedmann, I960]

Type species. Spathites chispaensis Kummel and Decker, 1954, by original designation.

Diagnosis. Medium sized, involute ammonites with compressed to depressed whorl sections, trapezoidal when
young, but tending to the subquadrate when mature. During early to middle growth there are strong to weak
umbilical bullae giving rise to from one to three ribs, with additional shorter, intercalated ribs, each of which
bears clavate inner and outer ventrolateral tubercles on either side of a flat or concave venter. When adult, the
ornament commonly declines, leaving shells which either have blunt bullae and low ribs bearing low ventral
clavi, or are smooth with sharp ventral shoulders with or without low clavi and a concave venter crossed by low,
broad ribs corresponding to the clavi.

Suture simple with broad, asymmetrically bifid saddles and narrow lobes, pseudoceratitic in some species.

Discussion. The type species of Spathites , S. chispaensis Kummel and Decker, 1954 (p. 311, pi. 30, figs.
1, 2; pi. 31, figs. 1-15; text-fig. 1. Text-fig. 1 herein) was originally described on the basis of large
collections from the early mid-Turonian Chispa Summit Formation of northern Chihuahua, Mexico,
and Chispa Summit, Jeff Davis County, Texas. We have re-examined this material, in the collections
of the Texas Memorial Museum, Austin, Texas, and confirm the very wide intraspecific variation
described by Kummel and Decker, which is also demonstrated by new material before us (OUM KT.
859, 895, 943, etc.). There is every transition from compressed, feebly ornamented nuclei (Kummel
and Decker 1954, pi. 31, figs. 7-9) to depressed ones with strong tubercles and ribs (Kummel and
Decker 1 954, pi . 3 1 , figs. 1 0- 1 2). In all forms, however, these decorated inner whorls are followed by a
virtually smooth adult stage, where ribbing is restricted to low transverse undulations across the
venter (text-fig. 1), corresponding to long low clavi at the shoulder. These adults are close to the
holotype of Neoptychites ( Spathitoides ) sulcatus Wiedmann (1960, p. 756, pi. 7, figs. 7, 8, text-figs. 11,
12). This (text-fig. 2) comes from a somewhat lower horizon in the Turonian of northern Spain, and

KENNEDY ET AL.: SPATHITES

text-fig. la, b. Spathites ( Spathites ) chispaensis
Kummel and Decker. The holotype UT
2081 1, from the mid-Turonian Ojinaga
Formation of the Placer de Guadalupe
district, Sabaco (San Jose de Cocahuata).

Bar scale is 2 cm.

text-fig. 2a, b, c. The holotype of Spathitoid.es sulcatus Wiedmann, Geol.
Pal. Inst. Tubingen Collection no. 1162/4 from the early Turonian south
of Pedrosa, Burgos, Spain. Bar scales are 2 cm.

824

PALAEONTOLOGY, VOLUME 23

was regarded by Wiedmann as a derivative of Neoptychites, characterized by being completely
smooth with a trapezoidal whorl section, a narrow umbilicus, and a truncated, concave siphonal
region with periodic constrictions. The suture (text-fig. 2) is much subdivided, with the lateral lobe
deeper than the external, and asymmetric. The umbilical lobe is short, the incisions on the lobes are
sharp, the terminations of the saddles are rounded, and the umbilical saddle is much enlarged. As can
be seen from comparing text-fig. 1 and text-fig. 2 the species is identical with the feebly ornamented
variants of S. chispaensis, the ‘constrictions’ on the venter noted by Wiedmann corresponding to the
interspaces between the ribs on the venter of S. chispaensis. The sutures are indeed more incised than
those of S. chispaensis but no more so than S. rioensis Powell (1963, p. 1228, pi. 169, fig. 2; pi. 170, figs.
1-3, 6-7; text-figs. 5 j, 6 c-e), as can be seen from comparing text-figs. 1 b and 8c. We would, therefore,
regard Spathitoides as a synonym of Spathites.

Jeanrogericeras Wiedmann, 1960, is shown, by study at all growth stages of its type species,
Ammonites reveliereanus Courtiller, 1860, to differ from Spathites only in the shape and ornament of
the mature last whorl; no more than subgeneric separation is appropriate. Moreover,
Jeanrogericeras, with its unspecialized last whorl, intermediate between that of Metoicoceras and
typical Spathites, is clearly more primitive, and, in Europe at least, occurs earlier than S. ( Spathites ).

Wiedmann (1960, p. 741) established a genus Fallotites, with type species Vascoceras subconciliatum
Choffat (1898, p. 64), in a new subfamily Fallotitinae. The genus was characterized by inner whorls
with subquadrate or trapezoidal section, flat sides and venter and large umbilical tubercles each giving
rise to two or three weak ribs, each of which bears weakly clavate inner and outer ventrolateral
tubercles; the body chamber looses all ornament except large rounded umbilical tubercles, becomes
rounded in section and tends to uncoil. His species of Fallotites include both depressed forms and
those with a whorl section slightly higher than wide. Moreover, some of his figures show that the body
chamber retains very low coarse ribs as well as the umbilical tubercles (Wiedmann 1960, pi. 3, figs.
4-7; pi. 4, figs. 2, 3). The characters of the inner whorl are exactly those of S. ( Jeanrogericeras )
reveliereanus at the corresponding stage (compare PI. 104, figs. 6-8; PI. 105, figs. 13-15 with PI. 105,
figs. 1-12), whilst those of the outer whorl are only slightly more extreme than those of some
specimens of reveliereanus (compare PI. 106, figs. 4-5 and text-figs. 2 and 5). There is thus a
continuum that includes the variable populations of reveliereanus and Fallotites spp; indeed,
Stankievich and Pojarkova (1969) include in Fallotites species with much stronger ribbing on the
body chamber than occurs in Jeanrogericeras. We see no good reason for separating these two taxa
and regard Fallotites as a synonym of S. ( Jeanrogericeras ). Fallotites ( Ingridella ) Wiedmann, 1960,
includes species with outer whorls that resemble some Vascoceras, but they have inner whorls with
distinct but subdued inner and outer ventrolateral tubercles, very feeble ribs, and sparse, very large
rounded umbilical tubercles that persist to the outer whorl, where other ornament disappears, leaving
a depressed rounded whorl section. We would therefore regard Ingridella as a further, specialized,
subgenus of Spathites.

Occurrence. Highest Cenomanian (Wright and Kennedy, in press) to mid-Turonian of western
Europe — England; Touraine, Aquitaine, and Provence in France; Spain, Portugal, Czechoslovakia,
the U.S.S.R. (Kirgisia and the Tadzhiksian depression); southern India, north Africa; Texas, New
Mexico, and northern Mexico.

EXPLANATION OF PLATE 104

Figs. 1-5. Spathites ( Spathites ) puercoensis (Herrick and Johnson). Specimens are from the mid-Turonian part of
the Mancos Shale at USGS Mesozoic Locality D4020 I T miles south-west of Ojito Springs, San Ysidro
Quadrangle, Sandoval County, New Mexico. USGS Coll., Denver.

Figs. 6-8. Spathites ( Jeanrogericeras ) robustus robustus (Wiedmann). Inner whorls of the holotype Geol. Pal.
Inst. Tubingen Collections, Ce 1 162/12. Early Turonian of Picofrentes (Soria), Spain.

PLATE 104

KENNEDY, WRIGHT and HANCOCK, Spathites

826

PALAEONTOLOGY, VOLUME 23

Subgenus Jeanrogericeras Wiedmann, 1960
[ = Falloti tes Wiedmann, 1960]

Type species. Ammonites reveliereanus Courtiller, 1860, by original designation.

Name of type species. In 1860 Courtiller spelt the name revelieranus on p. 249 and Reveliereanus in the
explanation of plate 2, fig. 5. Where two different spellings of a name appear in the first publication, Art. 32 (b) of
the Rules of Zoological Nomenclature provides that the spelling adopted by the first reviser is to be accepted as the
correct original spelling. Courtiller himself was the first reviser and in 1867 consistently spelt the name
‘ Reveliereanus ’.

Diagnosis. Spathites in which the outer whorls retain ribs and tubercles.

Spathites ( Jeanrogericeras ) reveliereanus (Courtiller)

Plate 105, figs. 1-12; Plate 106, figs. 1-2; text-figs. 3-6

1860 Ammonites revelieranus! Reveliereanus Courtiller, p. 249, pi. 2, figs. 5-8.

1867 Ammonites Reveliereanus Courtiller; Courtiller, p. 4, pi. 3, figs. 1-4.

1894 Mammites Revellieri (Courtiller); de Grossouvre, p. 28.

text-fig. 3, a, b. Spathites ( Jeanrogericeras ) reveliereanus (Courtiller). Adult specimen from the mid-Turonian
of Loudon, France, in the Sorbonne Collections (ex de Grossouvre Collection).

KENNEDY ET AL.\ SPAT HITES

827

1896 Ammonites/ Mammites rochebruni Coquand; Peron, p. 23.

1902 Mammites binicostatus Petrascheck, p. 145, pi. 7, fig. 6 a-b\ pi. 8, figs. 1 a-b, 3a-b.

1903 Ammonites Revelieranus (Courtiller); Pervinquiere, fiche 7, la, lb.

1907 Mammites Reveliereanus Courtiller; Pervinquiere, p. 311.

1912 Mammites Revelieri Courtiller; de Grossouvre, p. 18.

1912 Mammites Reveliereanus Courtiller; Roman, p. 12, pi. 1, fig. 1, 1 a.

1928 Mammites reveller ei Courtiller; Douville, p. 1 1 .

1935 Mammites revelierei Courtiller; Faraud, p. 18, fig. 3.

1935 Mammites revelieranus Courtiller; Karrenberg, p. 131, pi. 30, figs. 2-4; pi. 33, figs. 2-3; text-fig. 2
(including vars. quadrata, globosa, and lata).

1940 Mammites revelieranus Courtiller sp.; Fabre, p. 278, pi. 10, figs. 5-6.

1960 Jeanrogericeras revelieranum (Courtiller); Wiedmann, p. 740.

1960 Jeanrogericeras binicostatum (Petrascheck); Wiedmann, p. 741, pi. 2, figs. 7-9; text-fig. 5.

1964 Jeanrogericeras revelieranum (Courtiller); Wiedmann, p. 127.

1964 Jeanrogericeras binicostatum (Petrascheck); Wiedmann, p. 126, figs. 10a-c, 11.

1967 Metoicoceras stoliczkai Sastry and Matsumoto, p. 2, pi. 1, figs. 1 a-f.

1977 Jeanrogericeras reveliereanum (Courtiller); Hancock, Kennedy, and Wright, p. 156.

Lectotype. Here designated, the original of Courtiller 1860, pi. 2, figs. 5-6, refigured by Pervinquiere 1903, figs.
T1 -T2. It was originally in the Museum of the Chateau de Saumur, but we were unable to locate it with the rest of
Courtiller’s types.

Other specimens. The original of Courtiller 1867, pi. 3, figs. 1-2, survives and is figured here as text-fig. 5; it is
probably from the Saumur region. MNHP 6777 (d’Orbigny Collection) from Saumur, one of the syntypes of A.
fleuriausianus d’Orbigny. A further individual with this number is also a J. revelieranus , but is labelled
‘Rochefort’ on the specimen. MNHP unregistered, from ‘Taillenbourg, Charente, Ligerien E’, bearing an old
label ‘ Ammonites rochebruni Coquand’ (PI. 106, figs. 1-2). SP, unregistered, de Grossouvre Collection, Loudon
(text-fig. 3). OUM KZ767-771 from the St. Cyr-en-Bourg Fossil Bed of the Champignonniere Les Rochains, 7
km south of Saumur and north-east of Montreuil-Bellay (PI. 105, figs. 1-12). FSR 1700, from Taillenbourg;
OUM KZ779-783 from the Calcaire a Cephalopodes, Cimentierie Lafarge, east of Route N 10, 5 km south-west
of Angouleme. Geol. Pal. Inst. Tubingen CE 1162/6, the original of Wiedmann 1960, pi. 2, figs. 7-9, from
Ollogoyen, Navarra, Spain. A cast of the holotype of Metoicoceras stoliczkai Sastry and Matsumoto, Geological

Survey of India Collections no. 18170, from
(Trichinopoly) district, Madras (text-fig. 4).

north of

Mungilpadi,

Perambalur

Taluk, Tiruchinapalli

Dimensions.

D

Wb

Wh

Wb: Wh

U

GPIT Ce 1162/6

132-0(100)

50-0 (380)

61-0(46)

0-82

24-0(18)

SP, Loudon

105-5(100)

37-5 (36)

43-2(41)

0-87

23-9 (23)

S5

102-5(100)

43-5 (42)

50 (49)

0-87

18-2 (17-8)

MNHP Taillenbourg

95-5(100)

45-0 (47)

40-0 (42)

1-13

18-2(19)

GSI. 18170

77-6(100)

32-2(41)

38-7 (50)

0-83

17-0 (22)

OUM KZ767

44-0(100)

19 8 (45)

22-1 (50)

0-89

7-8(17)

-(-)

14-8 (-)

16-5 (-)

0-89

- (-)

OUM KZ770

42-8(100)

26-4 (62)

19-7 (46)

1-34

7-6(18)

OUM KZ771

32-7(100)

18-6(61)

15-7 (48)

1-18

5-3(16)

OUM KZ768

-(-)

8-0 (-)

12-3 (-)

0-65

- (-)

Description. Juveniles up to 50 mm are very variable. Coiling is involute, with a small umbilicus (usually around
16% of the diameter). Our most compressed individual is OUM KZ768 (PI. 105, figs. 10-12), with a whorl
breadth to height ratio of 0-65. The whorls are high, with the greatest breadth low on the flanks, the inner flanks
being gently inflated, the outer flanks flattened, converging to the narrow, tabulate venter. Weak umbilical
bullae give rise to pairs of low flexuous ribs, which bear faint inner and well-developed outer ventrolateral clavi
on either side of the flattened venter. As inflation increases (PI. 105, figs. 7-9; text-fig. 5), the whorl section
become trapezoidal, with the greatest breadth at the umbilical bullae. In OUM KZ770 (PI. 105, figs. 4-6), with
a whorl breadth to height ratio of 1 T 8, the umbilicus is deep, with a high subvertical wall and abruptly rounded
shoulder, gently swollen inner and convergent outer flanks, and a narrow, flattened venter. There are six or seven
strong conical umbilical bullae; these give rise to groups of two or three broad, strong, straight prorsiradiate ribs,

828

PALAEONTOLOGY, VOLUME 23

d

text-fig. 4a, b, c, d. The holotype of Metoicoceras stoliczkai Sastry and Matsumoto, Geological Survey of India
Collections no. 18170, from north of Mungilpadi, Perambalur Taluk, Tiruchinapalli district, Madras, India.

with additional intercalated ribs arising below mid-flank to give a total rib-density of 23 or 24 per whorl. Each rib
bears a conical to feebly clavate inner and a strong clavate outer ventrolateral tubercle. Rib strength varies even
in individuals showing this degree of inflation, as does rib direction, from prorsiradiate to rursiradiate, as in
OUM KZ771 . The most inflated individuals, including the lectotype, have swollen sides, with a whorl breadth to
height ratio of up to 1 -4. Here the bullae are coarse, conical, and crowded, 6 to 8 per whorl, giving rise to groups
of ribs with inner and outer ventrolateral tubercles as before (PI. 105, figs. 1-3).

EXPLANATION OF PLATE 105

Figs. 1-12. Spathites {Jeanrogericeras) reveliereanus (Courtiller). 1-3, OUM KZ769; 4-6, OUM KZ770; 7-9,
OUM KZ767; 10-12, OUM KZ768. All specimens are from the mid-Turonian St. Cyr-en-Bourg Fossil Bed of
the Champignonniere Les Rochains, 7 km south of Saumur and north-east of Montreuil-Bellay, France.
Figs. 13-15. Spathites ( Jeanrogericeras ) subconciliatus hispanicus (Wiedmann). OUM Collections, early
Turonian of Pedrosa, Burgos, Spain.

PLATE 105

KENNEDY, WRIGHT and HANCOCK, Spathites

830

PALAEONTOLOGY, VOLUME 23

Adult specimens before us show a similar range of whorl inflation from compressed, as in the Loudon example
(text-fig. 3), to inflated, as in OUM KZ769. Most individuals, however, have whorl breadth to height ratios of
between 0-9 and 11. The whorl section is trapezoidal, with the greatest breadth at the umbilical bullae, the flanks
convex and the venter narrow and flattened or even sulcate. Five to seven coarse, blunt umbilical bullae each give
rise to two or three low, broad, commonly rursiradiate ribs which may bear a trace of an inner ventrolateral
tubercle (although this commonly disappears by 70-80 mm diameter), together with long, low, outer
ventrolateral clavi, linked by a broad swelling across the flattened to concave venter, giving an undulose lateral
profile (PI. 106, figs. 1, 2) to the shell. Between clavi the ventrolateral shoulders are markedly angular. The last
part of the adult body chamber may show a decline in ornament, especially of the ribs and clavi, and may
contract (text-fig. 3), so that the coiling becomes scaphitoid and the umbilicus expands to 23% of the diameter.

The sutures are moderately subdivided, with rounded incisions. E/L is massive and asymmetrically bifid; L is
deep and narrow; L/U2 small and bifid, as is U2 and the first auxiliary element (text-fig. 6).

Discussion. Courtiller introduced the name A. reveliereanus in 1860 with both a description and
figure; he illustrated additional material in 1867. Pervinquiere (1903) refigured some of these
specimens photographically. Coquand had introduced the name A. rochebruni in 1858 and de
Grossouvre (1 894) regarded rochebruni as having priority over reveliereanus ; Peron used this name in
1896 for material from Charente, Touraine, Les Jeannots, and Revest in Provence. Coquand gave no
illustrations of rochebruni and we have been unable to locate the type specimens in the Collections of
the Museum d’Histoire Naturelle or the School of Mines (now at Lyon) which contains the other
ammonite types from this work. Coquand’ s description could well be of a Jeanrogericeras, but
significant specific features mentioned by Coquand differentiate it from reveliereanus : the presence of
12/13 ventral tubercles per whorl on inner whorls, fewer tubercles on the outer, the umbilical now
becoming larger and conical and the ventral tubercles disappearing altogether. Now J. reveliereanus
has far more ventrolateral tubercles per whorl when young (18-22) and the umbilical bullae tend to
weaken with age. It thus seems unlikely, if Coquand’s description is accurate, that the two species are
the same. A. rochebruni might be a Paramammites or belong to some other genus. We would therefore
continue to use the name reveliereanus , at least until the types of rochebruni are discovered and
illustrated.

Courtiller ( 1 860) was clearly aware that individuals of his species varied greatly, describing the
‘females’ as ‘beacoup plus renflees, surtout vers l’ombilic, que les males. Leurs tubercles sont aussi
beaucoup plus developpes’. Karrenberg (1935, p. 32, text-fig. 2, pi. 30, figs. 2-4; pi. 33, figs. 2-3)
discussed this variation at length, naming three forms; (a) Typical form: the whorl section is
trapezoidal, with a variable whorl height to whorl breadth ratio. The flanks are rather flat and are
clearly differentiated from the flattened venter. The greatest breadth is at the umbilical edge, (b) Var.
quadrata. The juveniles have the whorl section of the typical form up to a diameter of 30 mm. Later,
the flanks become parallel, giving an almost quadrate section, (c) Var. globosa. The whorl section is
almost circular, with the greatest breadth at approximately mid-flank. It differs from the typical form
even when young because of the distinctive section. ( d ) Var. lata. Flanks and venter are evenly
rounded, with the whorl section significantly wider than high. The inner whorls show the typical
cross-section.

EXPLANATION OF PLATE 106

Figs. 1-2. Spathites ( Jeanrogericeras ) reveliereanus (Courtiller). Unregistered specimen in the Collections of the
Museum d’Histoire Naturelle, Paris, from Taillenbourg, Charente, France.

Fig. 3. Spathites ( Spathites ) puercoensis (Herrick and Johnson). Ventral view of the specimen illustrated as
Plate 104, fig. 1.

Figs. 4-5. Spathites ( Jeanrogericeras ) subconciliatus hispanicus (Wiedmann). OUM Collections, early
Turonian of Pedrosa, Burgos, Spain.

PLATE 106

KENNEDY, WRIGHT and HANCOCK, Spathites

832

PALAEONTOLOGY, VOLUME 23

The Touraine material shows even wider variation, as discussed above, the breadth to height ratio
ranging from 0-69 to 1 -4, with typical covariance of ornament: the compressed individuals have weak
bullae, low, flexuous ribs and scarcely detectable inner ventrolateral tubercles; the depressed
individuals have massive conical bullae, coarse ribs, and well-differentiated inner and outer
ventrolateral tubercles.

S. ( Jeanrogericeras ) binicostatum (Petrascheck) (1902, p. 145, pi. 7, fig. 6 a-b', pi. 8, figs. 1 a-b, 3a-b;
see also Wiedmann 1960, p. 741, pi. 2, figs. 7-9; text-fig. 5; 1964, p. 126, figs. 10 a-c, 11) shows
considerable variation, according to Petrascheck’s illustrations; Wiedmann (1960, 1964) dis-
tinguished it from S. ( J .) reveliereanus on the basis of the absence of intercalated ribs, eight rather
than four umbilical tubercles, sharper ventrolateral shoulders, a more marked, excavated siphonal
region, a larger umbilicus and more massive, less asymmetric lobes in the suture line. The lectotype
closely matches the individual from Loudon illustrated here as text-fig. 3, whilst sharpness of
ventrolateral shoulders appears to be in part a matter of preservation, the Bohemian material being
distorted by compaction in our experience. The alleged difference in numbers of umbilical bullae
cannot be supported: the small lectotype of J. reveliereanus very clearly has at least seven in
Pervinquiere’s figure; other specimens have eight or nine, matching the smaller specimen figured by
Petrascheck (1920, pi. 7, fig. 6 a-b). Wiedmann’s large Spanish specimen (1960, pi. 2, figs. 7-8; 1964,
fig. 1 0 a-c) has eight massive bullae, the original of Courtiller ( 1 867, pi. 3, figs. 1 -2) has seven. In many
specimens before us it is a matter of opinion whether ribs are grouped and attached to bullae or

text-fig. 5a, b. The surviving Courtiller specimen of Spathites ( Jeanrogericeras ) reveliereanus (Courtiller).
CS5, from the Saumur region, Touraine, France.

KENNEDY ET AL:. SPATHITES

833

text-fig. 6. Suture lines and whorl sections of
juvenile Spathites ( Jeanrogericeras ) reveliereanus
(Courtiller), from OUM KZ767 and 769 from the
mid-Turonian St. Cyr en-Bourg Fossil Bed of the
Champignonniere Les Rochains, 7 km south of
Saumur and north-east of Montreuil-Bellay,
France. Bar scale is 1 cm.

intercalated in some cases. None of these criteria seems sufficient to justify specific separation of two
taxa.

Metoicoceras stoliczkai Sastry and Matsumoto (1967, p. 2, pi. 1, fig. la-/) (text-fig. 4) is also a
synonym of J. reveliereanus, closely resembling the lectotype of J. binicostatum. The suture line is
clearly that of a Jeanrogericeras, rather than a Metoicoceras.

Occurrence. S. ( Jeanrogericeras ) reveliereanus has a restricted range in the mid-Turonian,
where it occurs with early Collignoniceras woollgari (Mantell), Kamerunoceras turoniense
(d’Orbigny), and other ammonites in Touraine, and at this and slightly lower levels in Aquitaine,
Provence, Spain, and Czechoslovakia. The Indian occurrence is not accurately dated.

MAMMITINE PHYLOGENY

The inferred position of S. ( Spathites ) and S. ( Jeanrogericeras ) in mammitine phylogeny is shown in
text-fig. 9. Rather than duplicate existing descriptions we refer the reader to Kennedy, Juignet and
Hancock (in press) for an account of the late Cenomanian Metoicoceras species and to Kummel and
Decker (1954) and Powell (1963) for descriptions of S'. ( Spathites ). More extensive accounts of the
late Cenomanian Thomelites and Turonian Mammites and Metasigaloceras will appear elsewhere
(Wright and Kennedy, in press).

Thomelites first appears at the base of the Upper Cenomanian, and is represented by an
undescribed form from the Chalk Basement Bed of Askerswell, Dorset (Kennedy 1970, p. 644; OUM
Collections). By the middle of the Upper Cenomanian, the genus is known from Britain, France, the
Middle East, Brazil and elsewhere, and overlaps in time the first Metoicoceras, which evolved in the
western interior and Texas regions of the United States. As already noted, the earliest species of this
genus have a siphonal tubercle when young, but later forms such as M. defordi Young (1957), M.
mosbyense Cobban and M. muelleri Cobban (1953), although endemic to this area, show a
bituberculate venter throughout. Towards the close of the Cenomanian in the American S. gracile
Zone Metoicoceras spread to the Old World and M. geslinianum (d’Orbigny) occurs in England just
below the level of the first S. ( Jeanrogericeras ) in Devon, which yields S. ( Jeanrogericeras ) cf.
subconciliatus (Choffat) (Wright’s Collection no. 25310). This occurrence can be correlated firmly
with Zone III of Wiedmann’s Iberian sequence (1960, 1964) which clearly demonstrates the
succession with S. ( Jeanrogericeras ) ['Fallotites'] in his Zones III-V and S. ( Spathites ) [' Spathitoides']
occurring only in Zone V. In Europe S. ( Jeanrogericeras ) extends upwards to overlap early C.
woollgari, R. ( Romaniceras ) kallesi Zazvorka, Neoptychites cephalotus (Courtiller), and other species
in France and Spain and a somewhat impoverished but contemporaneous assemblage in
Czechoslovakia.

This association can in turn be related to the first well-documented occurrences of S. ( Spathites ) in
the New World, where the early C. woollgari Zone fauna of northern Mexico documented by Powell
(1963) yields S. ( Spathites ) rioensis in association with C. woollgari ( — Selwynoceras mexicanum
(Bose) of Powell), Kamerunoceras isovokyense (Collignon), Neoptychites xetriformis Pervinquiere
and Mammites depressus Powell. Our present state of knowledge suggests that this occurrence is a
little below that of the Touraine assemblages. Above this a sequence can be traced through Spathites

834 PALAEONTOLOGY, VOLUME 23

( S. ) chispaensis Kummel and Decker (which occurs with Romaniceras ( Yubariceras ) ornatissimum
(Stoliczka)) at a higher level in the woollgari Zone to S. ( S .) puercoensis Herrick and Johnson (1900),
which occurs in the succeeding North American Prionocyclus hyatti Zone. Detailed discussion of
these American species is beyond the scope of this contribution, but we note that all species show as
wide a range of intraspecific variation as is shown by S. ( J .) reveliereanum, as can be seen from PI. 1 04,
figs. 1-5; PI. 106, fig. 3.

Of especial interest is the progressive change in sutural complexity in the rioensis -> chispaensis ->
puercoensis lineage, shown here in text- fig. 8. This reduction in incisions and trend towards a
pseudoceratitic form provides a clue to the evolutionary origin of the Coniacian Buchiceras Hyatt,
1875, currently classed as a tissotiid. In our view, sutural pattern, gross shell form, and ornament all
point to Buchiceras as the last member of the Mammitinae. Its evolute quadrate whorls with ribs
branching from umbilical bullae and terminating in ventral tubercles are mammitine and distinct
from Tissotiidae with siphonal keel or row of tubercles, which we now regard as an offshoot of the
Barroisiceratinae via Tissotioides. This view is confirmed by a most important specimen, now housed
at the U.S. Geological Survey at Denver, collected from the Prinocyclus hyatti Zone of Bells, Grayson
County, Texas, by the late James Conlin of Fort Worth. This has a more compressed body chamber
than typical S. ( Spathites ) but still shows a slight facet representing the outer flank between the two
rows of ventrolateral tubercles; its suture is identical with that of Buchiceras and it is clearly
intermediate between S. ( S .) chispaensis and B. bilobatum. We would also argue that the early, robust
S. ( Jeanrogericeras ) of the subconciliatusjquadratus group (e.g. PI. 104, figs. 6-8) are the origin of
Mammites. The early whorls of these species and early forms of Mammites occurring immediately
above them in southern England are identical in their general plan of decoration (text-fig. 7);

text-fig. 7. 1-3. Mammites sp. WW 19898, from the lower part of the Inoceramus labiatus Zone of White Cliff,
Seaton, Devon. The outer whorls are those of a true Mammites , the inner strongly reminiscent of Spathites

(Jeanrogericeras), x 0-56.

KENNEDY ET AL.: SPATHITES

835

text-fig. 8. Progressive modification of sutures
in the Spathites ( Spathites ) to Buchiceras
sequence: a: Spathites rioensis Powell, OUM
KT1244, low Collignoniceras woollgari Zone,
Cannonball Hill, Chihuahua, northern Mexico.
b: S. chispaensis Kummel and Decker, OUM
KT943 high C. woollgari Zone, Chispa Summit,
Texas, c: S. puercoensis (Herrick and Johnson),
USGS 15947-20, Prionocyclus hyatti Zone,
USGS Mesozoic locality D4020, IT miles
south-west of Ojito Springs, San Ysidro
Quadrangle, Sandoval County, New Mexico
(kindly supplied by W. A. Cobban), d:
Buchiceras bilobatum Hyatt, Coniacian of
Otusco, Peru. Copy of Briiggen 1910, fig. 9d. a
and c are from middle-aged specimens; b and d
from adults. Bar scale is 2 cm.

text-fig. 9. Inferred phylogy of
Buchiceras, Spathites, and
other early Mammitinae.

836

PALAEONTOLOGY, VOLUME 23

Mammites has developed by an increase in size and strengthening rather than weakening of ornament
during ontogeny.

These last observations fully confirm our initial observations on the relationships of this array of
early Turonian acanthoceratids with quadrate or trapezoidal whorls, umbilical and inner and outer
ventrolateral tubercles, and simple ribs: they are a homogeneous close-knit group. Refinements in
correlation between England, France, Spain, Portugal, and the United States permit the construction
of a detailed phylogeny and show this group to be monophyletic rather than heterochronous
homoeomorphs.

Acknowledgements. We thank J. Sornay (Museum d’Histoire Naturelle, Paris), D. Pajaud (Universite Paris VI),
J. Louail (Rennes), M. Maury (Angers), and our other French colleagues for their help on the revision of the
stratotype Turonian. Professor T. Matsumoto (Kyushu), C. Duerdon and K. Young (Austin), D. Reaser
(Arlington, Texas), W. A. Cobban and J. D. Powell (Denver), E. G. Kauffman and N. F. Sohl (Washington)
assisted with work on U.S. faunas. M. R. Cooper (Salisbury, Zimbabwe), and J. Wiedmann (Tubingen)
provided additional help and discussion. Dr. S. C. Hook of the New Mexico Bureau of Mines and Mineral
Resources, Socorro, New Mexico, kindly gave us specimens of Spathites puercoensis for study.

The technical assistance of the staff of the Geological Collections, University Museum, Oxford, is
acknowledged, as is financial assistance from the Natural Environment Research Council, Wolfson College,
Oxford, the Lindemann Trust, and the Royal Society.

REFERENCES

bruggen, H. 1910. Die Fauna des unteren Senons von Nord Peru. Neues Jb. Miner. Geol. Palaont. BeilBd. 30,
717-788, pis. 24-29.

choffat, p. 1898. Les Ammonees du Bellasien, des couches a Neolobites vibrayeanus, du Turonien et du
Senonien. Part 2 of: Recueil d’etudes paleontologiques sur la faune cretacique du Portugal. Trav. geol.
Portugal 41-86, pis. 3-22.

Cobban, w. a. 1953. Cenomanian ammonite fauna from the Mosby Sandstone of Central Montana. Prof. Pap.
U.S. geol. Surv. 243-D, 45-55, pis. 6-12.

cooper, m. r. 1978. Uppermost Cenomanian — basal Turonian ammonites from Salinas, Angola. Ann. S. Afr.
Mus. 75,51-152.

COURTiller, a. 1860. Description de trois nouvelles especes d’ ammonites du terrain cretace des environs de
Saumur et des Ammonites Carolinus et Fleuriausianus a l’etat adulte. Mem. Soc. imp. Agr. Sci. Angers, 3,
246-252, 3 pis.

— 1867. Les ammonites du Tuffeau. Ann. Soc. lin. de Maine-et- Loire, Angers, 9, 8 pp., 8 pis.

douville, H. 1928. Les Ammonites de la Craie superieure en Egypte et au Sinai'. Mem. Acad. Sci. Inst. Fr. 60, 44
pp., 7 pis.

fabre, s. 1940. Le Cretace superieur de la Basse— Provence Occidental; 1 Cenomanien et Turonien. Ann. Fac.
Sci. Marseille (2), 14, 355 pp., 10 pis.

faraud, m. 1935. Le Ligerien de la colline de Bernon, pres de Tresques (Gard). Bull. Soc. Sci. nat. Vaucluse, 1,

11 pp.

grossouvre, A. de 1894. Les ammonites de la Craie superieure. Mem. Serv. Carte Geol. Fr. 264 pp., 39 pis.

— 1912. Le Cretace de la Loire— Inferieure et de la Vendee. Bull. Soc. Sci. nat. Ouest Fr., ser 3, 2, 1-38, 3 pis.
Hancock, j. M., Kennedy, w. J. and wright, c. w. 1977. Towards a correlation of the Turonian sequences of

Japan with those of north-west Europe. Spec. Pap. Palaeont. Soc. Japan, 21, 151-168.

Herrick, c. L. and Johnson, d. w. 1900. Geology of the Albuquerque Sheet. Bull. Univ. New Mex. geol. Ser. 1,
1-67, 32 pis.

hyatt, A. 1900. Cephalopoda, in zittel, k. a. von, 1896-1900, Textbook of Palaeontology, transl. Eastman, c. r.
Pp. 502-604. London, Macmillan.

— 1903. Pseudoceratites of the Cretaceous. Monogr. U.S. Geol. Surv. 44, 351 pp., 47 pis.
karrenberg, h. 1935. Ammonitenfauna aus der nordspanischen Oberkreide. Palaeontographica, 82A, 125-161,

pis. 30-33.

Kennedy, w. J. 1970. A correlation of the uppermost Albian and the Cenomanian of south-west England.
Proc. Geol. Ass. 81, 613-677.

— and wright, c. w. 1979a. Vascoceratid ammonites from the type Turonian. Palaeontology, 22, 665-683,
pis. 82-86.

KENNEDY ET AL.\ SPATHITES

837

Kennedy, w. j. and wright, c. w. 197%. On Kamerunoceras Reyment, 1954 (Cretaceous Ammonoidea). J.
Paleont. 53, 1 165-1 178, 4 pis.

— and cooper, m. r. 1979. On Ammonites galliennei d’Orbigny, 1850. Bull. geol. Instn. Univ. Uppsala.
n.s. 8, 5-15.

— and Hancock, j. m. 1980a. The European species of the Cretaceous ammonite Romaniceras with a
revision of the genus. Palaeontology , 23, 325-362, pis. 39-50.

— 19806. Collignoniceratid ammonites from the mid-Turonian of England and northern France.
Ibid. 23, 557-603, pis. 62-67.

— juignet, p. and Hancock, j. m. Late Cenomanian ammonites from Anjou and Vendee, western France.
Ibid., in press.

kullman, ]. and wiedmann, j. 1970. Significance of sutures in phylogeny of Ammonoidea. Paleont. Contr. Univ.
Kansas , 47, 1-32.

kummel, B. and decker, j. m. 1954. Lower Turonian Ammonites from Texas and Mexico. J. Paleont. 28, 310—
319, pis. 30-33.

peron, a. 1896-1897. Les ammonites du Cretace superieur de l’Algerie. Mem. Soc. geol. Fr. 17, 88 pp., 18 pis.
pervinquiere, l. 1903. Ammonites revelieranus Courtiller, 1860. Pal. Univers. fiche 7.

— 1907. Etudes de paleontologie tunisienne. 1, Cephalopodes des terrains secondaires; systeme cretacique.
Carte geol. Tunisie, Paris, 428 pp., 27 pis.

petrascheck, w. 1902. Die Ammoniten der sachsischen Kreideformation. Beitr. Palaont. Geol. Ost-Ung. 14,
131-162, pis. 7-12.

powell, j. d. 1963. Turonian (Cretaceous) ammonites from northeastern Chihuahua, Mexico. J. Paleont. 37,
1217-1232, pis. 166-171.

roman, f. 1912. Coup d’oeil sur les zones de Cephalopodes du Turonien du Vaucluse et du Gard. In Comptes
Rendu de /’ Association Frangaise pour T Avancement des Sciences, Congres de Nimes, 1912, Geologie et
Mineralogie, 1-15, pis. 1-3.

sastry, M. v. A. and matsumoto, t. 1967. Notes on some Cretaceous Ammonites from Southern India.— Part 2,
Occurrence of Metoicoceras in Trichinopoly Cretaceous. Mem. Fac. Sci. Kyushu Univ. Ser. D. Geology, 18,
1-5, pi. 1.

stankievich, e. s. and pojarkova, z. n. 1969. Vascoceratids from the Turonian of southern Kirgisia and the
Tadzhiksian depression. In Continental Formations of eastern regions of Soviet Central Asia and Kazakhstan.
Lithology and Biostratigraphy. Izd-vo Nauk SSSR. 86-111, pis. 1-10. (In Russian.)
wedekind, R. 1916. Uber Lobus, Suturallobus und Inzision. Zentrbl. Miner. Geol. Palaont. 1916, 185-195.
wiedmann, j. 1960. Le Cretace Superieur de l’Espagne et du Portugal et ses cephalopodes. C.R. Congres des
Societes Savantes— Dijon 1959: Colloque sur le Cretace superieur frangaise, 709-764, 8 pis.

— 1964. Le Cretace Superieur de 1’ Espagne et du Portugal et ses cephalopodes. Estudios geol. Inst. Invest,
geol. Lucas Mallada, 1964, 107-148, 39 figs.

wright, c. w. and Kennedy, w. j. A monograph of the Ammonoidea of the Plenus Marls and Middle Chalk.
Palaeontogr. Soc. ( Monogr .) in press.

— and wright, e. v. 1951. A survey of the fossil Cephalopoda of the Chalk of Great Britain. Palaeontogr. Soc.
{Monogr.), 1-40.

young, K. 1957. Cretaceous ammonites from eastern Apache County, Arizona. J. Paleont. 31, 1167-1 174, pis.
149, 150.

W. J. KENNEDY
C. W. WRIGHT

Geological Collections
University Museum
and Wolfson College
Oxford

J. M. HANCOCK

Typescript submitted 1 September 1979
Revised typescript submitted 20 November 1979

Department of Geology
Kings College
Strand

London WC2

THE ORDOVICIAN TRILOBITE FAUNA OF THE
SHOLESHOOK LIMESTONE FORMATION
SOUTH WALES

by DAVID PRICE

Abstract. This paper gives a complete account of the historically important trilobite fauna of the Sholeshook
Limestone; it comprises 50 species here placed in 40 genera and representing 23 families. Three new species,
Harpidella (H.) lacrymosa, Platylichas noctua, and Primaspis llandowrorensis are included. Species of Liocnemis
and Whittingtonia occur in laterally equivalent parts of the Slade and Redhill Mudstone Formation; the former
genus is described for the first time from British strata. Proceratocephala is described for the first time from
Wales. The rostral plate of Stenopareia bowmanni (Salter) is described, and a free cheek figured. A pygidium is
described for a form tentatively referred to Panderia edita Bruton. Well-preserved topotype specimens of
Pseudosphaerexochus juvenis (Salter) give a better characterization of the species than Salter’s original syntypes.
A thorax is illustrated for Duftonia geniculata Ingham, and a distinctive hypostoma is described for a form
tentatively referred to Platylichas angulatus Warburg. Ranges and abundances are charted. The Sholeshook
Limestone was deposited in an environment intermediate between deep-slope mudstones and shelf-edge
carbonates, and has elements of its trilobite fauna in common with both facies types.

The Sholeshook Limestone Formation is a lithologically variable thickness of mudstones,
sandstones and (predominantly) siltstones of calcareous type developed between the Mydrim Shale
Formation and the Slade and Redhill Mudstone Formation at Sholeshook and Prendergast,
Haverfordwest, south-west Dyfed (type development), and in the area around Llandowror
about 30 km further east. It is present between the type Robeston Wathen Limestone and the
Slade and Redhill Mudstones at Robeston Wathen, about 13 km east of Haverfordwest. Further
details of stratigraphy were given in an earlier account (Price 1973a) which showed the formation to
have a diachronous base and to overlie the Mydrim Shales unconformably. The base of the normally
succeeding Slade and Redhill Mudstones was also diachronous, so that to the north and west of
Haverfordwest this formation contains strata laterally equivalent to the type Sholeshook Limestone.
At that time the age of the Sholeshook Limestone was considered to range from the upper part of
Zone 1 to probably Zone 3 of the Cautleyan Stage. Since then, largely as a result of re-examining the
trilobite fauna during the preparation of this paper, aided by recent descriptions of Ashgill trilobites
by other authors, it has been found necessary to revise the younger of these age limits and the
formation is now considered to range upwards into Zone 5 of the Rawtheyan Stage (Price 1980).

Both the Sholeshook Limestone and its trilobite fauna have played an important part in the
development of knowledge and concepts concerning upper Ordovician stratigraphy (see Price 1973a,
p. 226; 19736, p. 535; 1974, p. 841). The present paper deals with the trilobite fauna as a whole. The
trilobites of the basal Slade and Redhill Mudstones are also mentioned since these beds are partly
correlatives of the Sholeshook Limestone and, with the exception of two forms, all their species are
common to both formations. Occurrences of Sholeshook species in the higher Slade and Redhill
Mudstones are also noted.

The trilobite fauna of these horizons is a rich and varied one, comprising some 52 species
representing 42 genera and 23 families. Its treatment within a single relatively short paper is possible
because some of its elements have already been described in detail elsewhere (e.g. Price 1974, 1977),
and because recent descriptions of Ashgill trilobites have been given by such workers as Kielan
(1960), Whittington (1962-8), Ingham (1970-7), Dean (1971-8), and McNamara (1979a).

[Palaeontology, Vol. 23, Part 4, 1980, pp. 839-887, pis. 107-1 14.|

table 1. Ranges and abundances of trilobite species in the Sholeshook Limestone Formation. Bold solid bars where short indicate actual occurrences, where
longer show known ranges. Solid black arrow-heads show which forms range up into the overlying Slade and Redhill Mudstones at Haverfordwest and
Llandowror. Lighter bars with arrow-heads indicate the possible range of occurrence of ill-localized material.

>r. Lighter bars with

PRICE: LATE ORDOVICIAN TRILOBITES

841

Accordingly, the complete fauna is given in an annotated list below and only forms meriting further
comment are given systematic descriptions. The composition of the fauna is discussed in a final
section. The trilobites dealt with here largely come from outcrops of naturally decalcified rock and
are normally preserved as internal and external moulds; material representing the exoskeleton is only
rarely present. The vast majority of specimens are disarticulated exoskeletal parts of which a high
proportion are fragmentary or incomplete. Nevertheless a few complete or partly articulated
specimens are known, and seventeen of the Sholeshook species are represented by at least one such
specimen. Most material is somewhat distorted.

Terminology. For the purposes of description, measurement, and illustration, specimens have been viewed in the
standard orientations described by Temple (1975); for isolated exoskeletal parts the horizontal plane is generally
defined by the (maximized) sagittal length. An exception is made in the case of trinucleid cephala which,
following Hughes, Ingham, and Addison (1975) are oriented with the anterior and posterior fossulae in the
horizontal plane. All references to shape and proportions, unless otherwise stated, refer to dorsal views. The
term glabella is used to include the occipital ring and glabellar lobes and furrows are numbered from the rear.
The term ‘bullar lobes’ (Temple 1972) is used for the structures formerly designated ‘composite’ or ‘bicomposite’
lobes in the lichid genus Trochurus. Several of the forms described here have lengthy synonomies; where these are
available in earlier references they are not repeated herein.

Repositories and localities. Material is housed in the following museums: British Museum (Natural History)
(BM), Hunterian Museum, Glasgow (HM), Institute of Geological Sciences (GSM), National Museum of Wales
(NMW), and the Sedgwick Museum, Cambridge (SM). The determinations and descriptions in this paper are
based on all available material in these collections which for most forms is too numerous to be listed separately.
Maps and tables showing precise localities for specimens collected by the author and for other well-localized
material have been given previously (Price 1973a) and it is to these that locality numbers cited in the text refer.
Occurrences within the Sholeshook Limestone are shown graphically in Table 1.

ANNOTATED LIST OF COMPLETE TRILOBITE FAUNA

The forms marked with an asterisk are treated in the Systematic Account. Those based on type material from the
Sholeshook Limestone or basal Slade and Redhill Mudstones are marked thus f

Trinodus tardus (Barrande). PI. 107, figs. 1-3. See also Dean’s illustrations (1971, pi. 1, figs. 1, 2) of the syntypes
of Agnostus tardus, var. j8 convexus Salter, 1848 from the Sholeshook Limestone
Remopleurides cf. colbii Portlock*

Amphitryon radians (Barrande)? PI. 107, fig. 9; PI. 108, fig. 9. Rare; 2 cranidia from basal Slade and Redhill
Mudstones (Iocs. 3, 4), 2 from high Sholeshook Limestone (Iocs. 8b, 8c) of Prendergast
Stygina sp. indet.*

Opsimasaphus sp. indet.*

Illaenus ( Parillaenus ) cf. fallax Holm*

Stenopareia bowmanni (Salter)f. PI. 107, fig. 15; PI. 108, fig. 15. For description of topotype material see Price
1974, p. 842, pi. 1 12, figs. 1-8, ?9. Rostral plate since recognized and figured here together with free cheek not
previously figured
Pander ia edit a Bruton?*

Harpidella ( Harpidella ) lacrymosa sp. nov.*

Proetus (s.l.) cf. berwynensis (Whittington)*

Phillipsinella parabola (Barrande) aquilonia Ingham. PI. 108, fig. 16

Harpetid gen et sp. indet. Not figured. See Reed 1905a, p. 97, pi. 4, fig. 1, and Whittington 19506, p. 32
Nankinolithus cf. granulatus (Wahlenberg)*

Tretaspis moeldenensis moeldenensis Cavef
T. cf. radialis Lamont
T. aff. radialis Lamont

T. hadelandica Stormer brachystichus Ingham
Dionide sp. indet.*

Raphiophorus cf. tenellus (Barrande)*

Lonchodomas aff. pennatus (La Touche)*

See Price 1977

842

PALAEONTOLOGY, VOLUME 23

L. cf. drummuckensis (Reed)*

Ceraurinella intermedia (Kielan). PI. 1 10, figs. 7, 8.

Hadromeros cf. keisleyensis (Reed). Not figured. See Lane 1971, p. 20, pi. 3, figs. 8-11, pi. 4, figs. 1-4.

Lehua princeps (Reed)f . Not figured. Now that both cranidium (Lane 1 97 1 , p. 36, pi. 7, figs. 1 la, b) and pygidium
(Price 1974, p. 848, pi. 113, fig. 12) have been described, the species is placed firmly in genus Lehua
Pseudosphaerexochus tectus Ingham. PI. 110, figs. 9-11. Previously listed in Sholeshook faunal lists as P.

octolobatus (McCoy)

P. juvenis (Salter)f*

Sphaerocoryphe aff. thomsoni Reed*

Encrinuroides sexcostatus (Salter)f. Not figured. See Whittington 1950a, p. 535, pi. 68, figs. 7-16, text-fig. 2 and
Price 1974, p. 856, pi. 115, figs. 1-8

Atractopyge scabra Dean? PI. 110, fig. 17. See discussion of A. aff. scabra
Atractopyge aff. scabra Dean*

Cybeloides ( Paracybeloides ) girvanensis (Reed). PI. 1 10, fig. 16; PI. Ill, fig. 5
Dindymene longicaudata Kielan*

Staurocephalus cf. clavifrons Angelin*

Calymene (s.l.) cf. prolata Ingham*

Flexicalymene cavei Pricef*

Prionocheilus cf. obtusus (McCoy)*

Brongniartella cf. marocana Destombes*

Duftonia geniculata Ingham?*

Kloucekia(Kloucekia)robertsi (Reed )t 1 Not figured. See Price 1974, p. 857, pi. 1 15, figs. 9-14; pi. 116, figs.
K. (Kloucekia) extensa Rnce( ) 1, 2 and p. 862; text-fig. la-k

Liocnemis recurvus (Linnarsson)*

Calyptaulax planiformis Dean*

Toxochasmops marri (Reed). PI. 1 12, figs. 16, 17. See also Reed 1904, pi. 12, fig. 3 and Cocks and Price 1975, pi.
81, fig. 4

Platylichas noctua sp. nov.f *

P. angulatus Warburg?*

Trochurus sp. indet.*

Whittingtonia whittingtoni Kielan*

Proceratocephala cf. terribilis (Reed)*

Primaspis llandowrorensis sp. nov.f*

P. sp. indet.*

Diacanthaspisl turnbulli (Reed)f . Not figured. See Price 1974, p. 864, pi. 1 16, figs. 3-5
Glaphurella cf. harknessi (Reed)*

SYSTEMATIC DESCRIPTIONS

Family remopleuridae Hawle and Corda, 1847
Genus remopleurides Portlock, 1 843

Type species. Remopleurides colbii Portlock, 1843.

Remopleurides cf. colbii Portlock, 1 843
Plate 107, figs. 4-8

1885 Remopleurides longicostatus, Portl.; Marr and Roberts, faunal list p. 481.

1885 Remopleurides dorso-spinifer, Portl.; Marr and Roberts, faunal list p. 481 .

1905a Remopleurides Salteri, Reed, var. girvanensis ; Reed, p. 98, pi. 4, fig. 3.

1914 Remopleurides longicostatus Portl.; Strahan, Cantrill, Dixon, Thomas, and Jones, table p. 64.
1914 Remopleurides colbii Portl.; Strahan et al. (pars), list p. 71.

1914 Remopleurides salteri Reed; Strahan et al., p. 75.

1966 Remopleurides cf. colbii Portlock; Whittington (pars), p. 75, pi. 22, figs. 5, 6, 9.

?1966 Remopleurides cf. colbii Portlock; Whittington (pars), p. 75, pi. 22, fig. 7; pi. 23, figs. 1-6.
1973a Remopleurides aff. colbii Portlock; Price, tables 1-3.

1975 Remopleurides sp.; Cocks and Price, p. 705, pi. 81, fig. 5.

PRICE. LATE ORDOVICIAN TRILOBITES

843

Horizons and localities. Apart from the occurrences shown in Table 1, ranges through the Slade and Redhill
Mudstone Formation around Haverfordwest (to locality I of Cocks and Price 1975) and is known also from the
high Slade and Redhill Mudstones just south of Little Clerkenhill, 9 km further east (Grid ref. SN 045 150).

Description. Cranidium about as wide (tr.) as long (sag.) and moderately convex transversely. In lateral profile
(PI. 107, fig. 4) just less than posterior three-fifths of length straight, rest curved evenly through about 90°.
Glabellar width (tr.) immediately in front of occipital furrow slightly over half maximum width achieved just
behind mid-length. Anterior tongue of similar width (tr.) posteriorly, tapers very slightly forwards. Lateral
glabellar furrows (PI. 107, fig. 7) faint, evenly spaced (exsag.). Ip and 2p sub-parallel, gently convex anteriorly;
separated longitudinally by about 15% of sagittal cranidial length and mesially by about 30% of maximum
glabellar width; not reaching axial furrows. 3p furrows short (tr.), their adaxial ends slightly further apart than
those of the lp and 2p. Occipital furrow straight, deep, and narrow. Occipital ring strongly arched transversely,
broad (sag. and exsag.) mesially; convex posterior margin bears row of (?eighteen) prominent, backwardly
directed tubercles. Palpebral lobes narrow (tr.) anteriorly, broaden posteriorly, and indent pre-occipital part of
glabella. Librigenae (see Whittington 1966, pi. 22, figs. 5, 9) narrow (tr.) anteriorly, widening posteriorly and
produced into long, gently curved, stout genal spines; no sub-genal notch. Visual surface of eye surrounded by
broad furrow and prominent external rim. Ventral surface of doublure with strong, sub-parallel ridges separated
by broad grooves which are themselves striated. Hypostoma not known. Thorax incompletely known; axis
strongly convex, rings broad (sag. and exsag.), dorsally flat, slightly expanded abaxially, separated by broad
articulating furrows. Pleurae narrow (tr.), strongly bent-down, antero-laterally rounded, postero-laterally
drawn into short spines; inner anterior corners bear large articulating bosses fitting sockets in previous segments.
Pleural doublure bears strong longitudinal grooves. One partial thorax, SM A30963, from the high Slade and
Redhill Mudstones, bears a long median spine but the specimen does not show clearly which ring this is on.

Pygidium (PI. 107, fig. 8 and Cocks and Price 1975, pi. 81, fig. 5) sub-triangular. First axial ring narrow (sag.
and exsag.) mesially but much expanded at tips; second ring represented by pair of elongated, postero-laterally
directed lobes separated by longitudinal furrow. Faint, narrow (tr.) posteriorly tapering post-axial ridge. Margin
with two pairs of spines, anterior pair slender, thorn-like, gently curved adaxially, posterior pair broad-based
and much larger. Doublure broad with fine transverse terrace-lines arranged in posteriorly convex arc.

Discussion. The Sholeshook specimens are similar to the holotype of R. colbii as redescribed by
Whittington (1950a, p. 540, pi. 70, figs. 1, 4, 5) though that specimen is too incomplete to show either
the form of the posterior pygidial spines or whether a short 3p lateral glabellar furrow is present.
Similar also to R. colbii and the South Welsh specimens is the species described by Whittington (1966,
p. 75, pi. 22, fig. 7; pi. 23, figs. 1-6) from the Rhiwlas Limestone— though it is not well enough known
for detailed comparison. A pygidium with similarly large posterior spines to those of the South Welsh
form is seen in the species described by Ingham (1970, p. 13, pi. 1, figs. 22-25, 717-21) from the
Cautley Mudstones as Remopleurides sp. B. That form, however, has a large sub-genal notch and
appears to bear a surface Bertillon pattern of fine ridges.

Family scutelluidae Richter and Richter, 1925
Genus stygina Salter, 1853

Type species. Asaphus latifrons Portlock, 1843.

Stygina sp. indet.

Plate 107, fig. 12

1885 Stygina ; Marr and Roberts, list p. 480 {pars).

1973a Stygina'l sp. indet.; Price, table 2.

Material. SM A3 1595, internal mould of incomplete pygidium and posterior-most part of thorax (enrolled),
from the high Sholeshook Limestone Formation of Prendergast Place, Haverfordwest (locality 8b, 8c, or 8d).

Discussion. In features such as the broad, concave border, the fine sub-parallel terrace-lines on the
ventral mould of the doublure, the long, gradually tapering axis only weakly defined posteriorly, and
the post-axial ridge, the specimen shows much similarity with the pygidia of the lectotype and other
topotype specimens of S. latifrons figured by Whittington (1950, pi. 72, figs. 2, 3, 6, 9). The axial

PALAEONTOLOGY, VOLUME 23

furrows are shallow and the pygidium thus differs from that of the form figured by Whittington (1966,
pi. 21, figs. 13-15) from the Rhiwlas Limestone of North Wales where they are deeply incised. In
addition the axis of the Rhiwlas form appears to be narrower (tr.) anteriorly and to taper less rapidly
back than in either S. latifrons or the Sholeshook form.

Family asaphidae Burmeister, 1843
Subfamily asaphinae Burmeister, 1843
Genus opsimasaphus Kielan, 1960

Type species. Opsimasaphus jaanussoni Kielan, 1960.

Opsimasaphus sp. indet.

Plate 107, figs. 10-11

1914 Asaphus ?; Strahan et al ., p. 76.

1973a Opsimasaphus ? sp. indet.; Price, tables 1, 2, 7.

Material. BM It9257, internal mould of incomplete hypostoma from near middle of Sholeshook Limestone
Formation, locality 9e, Sholeshook. A poorly preserved pygidium, GSM H.T. 480, from the basal Slade and
Redhill Mudstones near Clarbeston Road Station (locality 6a) probably also belongs here.

Description. Hypostoma with convex, sub-ovoid central body. Lateral border broadest (tr.) opposite posterior
margin of central body, extended back, ornamented with faint, gently sinuous terrace-lines, and bifurcate
posteriorly where it is excavated by a deep, anteriorly expanded notch. Lateral border furrow deepest and
broadest posteriorly; transverse furrow wide (sag. and exsag.) and shallow. A small, faint median tubercle is
present at the posterior margin of the central body and maculae are developed behind the abaxial ends of the
transverse furrow.

EXPLANATION OF PLATE 107

Figs. 1-3. Trinodus tardus (Barrande). 1, SM A3 1383, internal mould of pygidium from basal Slade and Redhill
Mudstones of Pelcomb Cross (locality 2), dorsal view. 2, SM A77583a, internal mould of cephalon from
basal Sholeshook Limestone of Pentre-howell road section (locality 17), left lateral view. 3, SM A3 1378,
internal mould of cephalon from Sholeshook Limestone of Sholeshook railway cutting, dorsal view. All x 8.

Figs. 4-8. Remopleurides cf. colbii Portlock. 4-6, SM A77953, internal mould of cranidium from high
Sholeshook Limestone of Prendergast (locality 8c), left lateral, dorsal, and anterior views, x 4. 7, SM
A30721, internal mould of cranidium showing lateral glabellar furrows, Slade and Redhill Mudstones of
Upper Slade, Haverfordwest, dorsal view, x4. 8, SM A98065, internal mould of pygidium from high
Sholeshook Limestone (topmost locality 8d), Prendergast, dorsal view, x 6.

Fig. 9. Amphitryon radians (Barrande)? BM In54167a, internal mould of incomplete cranidium from basal Slade
and Redhill Mudstones of Rudbaxton (locality 4), dorsal view, x 6.

Figs. 10, 11. Opsimasaphus sp. indet. 10, BM It9257, internal mould of incomplete hypostoma, middle
Sholeshook Limestone, locality 9e, Sholeshook, ventral view. 11, GSM H.T. 480, flattened and sheared
internal mould of pygidium from basal Slade and Redhill Mudstones of Clarbeston Road Station (locality
6a), dorsal view. Both x 1 J.

Fig. 12. Stygina sp. indet. SM A3 1595, internal mould of pygidium and posterior part of thorax (enrolled) from
the high Sholeshook Limestone of Prendergast (locality 8b, c, or d), pygidium in dorsal view, x 3.

Figs. 13, 14. Illaenus ( Parillaenus ) ci.fallax Holm. 13, BM In54706, internal mould of pygidium from Craig-y-
deilo quarry, Llandowror, dorsal view, x 1. 14, BM In54464, internal mould of cranidium from high

Sholeshook Limestone (locality 8b), Prendergast, dorsal view, x \\.

Fig. 1 5. Stenopareia bowmanni (Salter). SM A99092b, cast from external mould of rostral plate, high Sholeshook
Limestone (locality 8d) of Prendergast, ventral view, x 2\.

Figs. 16, 17. Proetus (s.l.) cf. berwynensis (Whittington). 16, SM A77821, cast from external mould of
incomplete cranidium, Sholeshook Limestone horizon, Robeston Wathen (locality 10a), dorsal view, x 8.
17, SM A99093, internal mould of pygidium, same horizon and locality, dorsal view, x 6.

PLATE 107

PRICE, Sholeshook trilobites

PALAEONTOLOGY, VOLUME 23

Pygidium with long (sag.), narrow, gradually tapering axis and abaxially expanding, unfurrowed pleural ribs
separated by strong furrows. At least seven axial rings are present and there is room for several more on the
anteriorly damaged portion of the axis; there appear to be eight inter-pleural furrows. Although the ventral
mould of the doublure is not completely exposed, a broad band of concentric terrace-lines is visible postero-
laterally.

Discussion. The hypostoma is very similar to that of O. jaanussoni originally figured by Barrande
(1852, pi. 31, fig. 6; pi. 32, fig. 6) from the Kraluv Dvur Formation of Bohemia. That of O. radiatus
Salter figured by Whittington (1966, pi. 24, fig. 6) from the Crugan Mudstone Formation of north
Wales is similar in over-all form but differs in that the posterior notch is sub-angular anteriorly and
not with a rounded anterior expansion as in the south Welsh and Bohemian forms.

Family illaenidae Hawle and Corda, 1847
Subfamily illaeninae Hawle and Corda, 1 847
Genus illaenus Dalman, 1 827
Subgenus parillaenus Jaanusson, 1954

Type species. Illaenus fallax Holm, 1882.

Illaenus ( Parillaenus ) cf. fallax Holm, 1882
Plate 107, figs. 13-14; Plate 108, figs. 1-2

1885 Illaenus Bowmanni, Salt.; Marr & Roberts {pars), pp. 480-481.

1914 Illaenus davisi Salter; Strahan et al., table p. 63.

1933 Illaenus bowmanni Salter; Reed (pars), pp. 124-125.

1973a Illaenus ( Parillaenus ) cf. fallax Holm; Price, tables 1-4.

1974 Illaenus ( Parillaenus ) cf. fallax Holm; Price, p. 844 (top).

Horizons and localities as in Table 1. Not known from Slade and Redhill Mudstone Formation.

Description. Entire exoskeleton sub-oval in outline (PI. 108, fig. 1), almost If times as long (sag.) as wide (tr.);
isopygous. Cephalon sub-semicircular in outline, moderately and evenly convex (sag., tr.). Axial furrows broad
and shallow, extending forwards to about half cephalic length from posterior margin; rear sections sub-parallel
with very slight abaxially convex curvatures, then, from about mid-level of palpebral lobes, furrows curve
sigmoidally first adaxially and then outwards before dying out opposite anterior ends of palpebral lobes.
Maximum width (tr.) between axial furrows about three-fifths of maximum width of cranidium; latter achieved
on mid-level of palpebral lobes. Palpebral lobes long (exsag.) and crescentic, occupying about one-quarter of
total cephalic length (sag.) and situated at about two-thirds of their own length (exsag.) from posterior margin.
Posterior branches of facial sutures directed straight back from palpebral lobes; anterior branches curve at first
gently outwards but then converge as they approach the anterior margin and become confluent on the ventral
surface. Free cheeks thus sub-triangular in form, declined very steeply outwards, with rounded genal angles.

Thorax of ten segments, with shallow but broad and distinct axial furrows. Convex axis occupies over half
total thoracic width anteriorly, remains sub-parallel over anterior-most five or six segments then tapers
gradually back. Axial rings flat in lateral view. Pleurae simple; flat and horizontal adaxially, deflected postero-
ventrally at fulcrum in gentle, posteriorly convex curves; abaxial extremities truncated antero-laterally;
indistinctly separated articulating facets developed. Horizontal adaxial sections of pleurae become broader (tr.)
posteriorly along the thorax (PI. 108, fig. 1).

Pygidium about \\ times as broad (tr.) as long (sag.), convex transversely and in lateral profile rather flat over
most of length but dropping steeply posteriorly. Axis anteriorly occupies about two-fifths of total width (tr.), is
moderately convex, and bounded by shallow, ill-defined furrows which converge rapidly back. Anterior margins
of pleural lobes transverse adaxially for length equivalent to about one-third of axial width then deflected gently
back. Behind, shallow but broad and distinct furrows run obliquely back from anterior ends of axial furrows
diverging at about 1 30°. Doublure of even width around margin and posteriorly occupying about one-quarter of
total pygidial length (sag.); bearing rather widely spaced concentric terrace-lines.

Discussion. Holm’s material of I. fallax awaits modern redescription, but to judge from his original
figures the axial furrows appear to extend less far forward on the cranidium than in the south Welsh

PRICE: LATE ORDOVICIAN TRILOBITES

847

material. This difference also applies to cranidia figured by Ingham (1970, p. 19, pi. 2, figs. 10-20)
from the high Caradoc and the Pusgillian and Cautleyan Stages at Cautley, and here there appear to
be slight differences in cephalic proportions also— though much of the material from both south
Wales and Cautley is distorted. In those pygidia figured by Holm which are most like the Sholeshook
specimens (e.g. 1882, pi. 2, figs. 15, 18) the doublure appears to be narrower antero-laterally. I. ( P .)
davisi Salter, from the Rhiwlas Limestone of north Wales (see Whittington 1966, p. 67, pi. 20, figs.
16-23; pi. 21, figs. 1-4, 6-9) differs in having a pygidium which is strongly humped medially and has a
much broader doublure.

Subfamily panderiinae Bruton, 1968
Genus panderia Volborth, 1863

Type species. Panderia triquetra Volborth, 1863.

Panderia edita Bruton, 1968?

Plate 108, figs. 3-8

1973a Panderia aff. edita Bruton; Price, pp. 233, 243, tables 1-4.

Horizons and localities. Apart from the Sholeshook Limestone occurrences shown in Table 1, the species ranges
into the basal Slade and Redhill Mudstones at Prendergast (locality 8a) and is known also from the ‘Bala
Limestone’ outcrop near Trefanty, 4 km south-east of St. Clears (Strahan, Cantrill, Dixon, and Thomas 1909,
p. 56; Price 1973a, p. 243).

Description. Cephalon strongly convex (tr. and sag.); sub-semicircular in lateral profile (PI. 108, fig. 5). Glabella
about I4 times as long (sag.) as wide (tr.) with maximum width on mid-level of palpebral lobes. At this level also
is a small median tubercle. Broad, shallow occipital furrow present only faintly mesially but deep occipital pits
are developed where it meets the axial furrows. Latter deep and broad posteriorly where they outline small but
distinct postero-lateral swellings of glabella (PI. 108, fig. 4) then run forwards shallowing, at first straight and
slightly divergent as viewed normally to the occipital region (‘dorsally’ sensu Bruton 1968) but opposite
palpebral lobes are bowed outwards in distinctive geniculations and then deflected again abaxially to meet facial
sutures just in front of palpebral lobes. Latter broad (tr.) and crescentic, occupying almost two-fifths of total
glabellar length and situated at about two-thirds their own length from posterior margin. Short posterior
branches of facial suture postero-mesially convex. Anterior branches converge for short distance and then
gradually diverge and again converge in gentle curves to define an anterior glabellar area which is sub-
rectangular with broadly rounded antero-lateral margins and attains a maximum width (tr.) of about nine-
tenths the maximum glabellar width. Free cheeks in lateral view with broad concave embayments along antero-
lateral margins— probably to accommodate edge of pygidium during enrolment (cf. Bruton 1968, p. 26, pi. 9,
fig. 5). Eye lobes surrounded by prominent broad furrows.

Pygidium sub-semicircular in outline. Moderately convex, well-defined axis occupies one-third of total width
anteriorly and tapers only gradually back reaching to two-thirds pygidial length, its posterior end rounded.
Axial furrows broad and shallow. Anterior margins of pleural lobes transverse adaxially but about half-way out
from axial furrows deflected gently posteriorly (PI. 108, fig. 8). Broad, gently down-turned triangular facets are
developed antero-laterally. Doublure narrow, occupying only 10-15% total pygidial length at posterior margin
and narrowing antero-laterally.

Discussion. The south Welsh cranidia show much similarity with those of P. edita Bruton ( 1 968, p. 25,
pi. 9, figs. 3-8; pi. 10, figs. 1-3, 8) from the Boda Limestone (Harju Series) of the Siljan district,
Sweden. While the anterior part of the glabella in the Welsh specimens is by no means as long (sag.
and exsag.) as in one of the specimens figured by Bruton (pi. 10, fig. 2) it is of similar length to that of
the holotype (Bruton 1968, pi. 9, fig. 7). The geniculation in the course of the axial furrows in the
Welsh form, however, appears to be more prominent than in the Swedish material. No pygidium is
known for P. edita. The form described by Dean (1977, p. 108, pi. 51, figs. 7, 8; pi. 52, figs. 1-14, 16,
17) from the Chair of Kildare Limestone as P. cf. edita differs from the south Welsh species in having
anterior branches to the facial suture which converge in even curves and define a relatively shorter
(sag.), frontally narrower (tr.) anterior glabellar area, in having free cheeks which are broadest

848

PALAEONTOLOGY, VOLUME 23

further anteriorly and in having a pygidium with a relatively broader axis and much wider doublure.
P. megalophthalma Linnarsson (Bruton 1968, p. 26, pi. 10, figs. 5, 6, 9; pi. 1 1, figs. 1, 5-10) differs in
having a relatively shorter (sag.) and broader glabella lacking the postero-lateral swellings, in having
broader (tr.) free cheeks, straighter posterior branches to the facial sutures and a much wider pygidial
doublure. The north Welsh form P. lewisi (Salter), known only from the holotype (Bruton 1968, p. 28,
pi. 10, fig. 7; pi. 1 1, figs. 2-4), also has proportionally broader (tr.) free cheeks than the south Welsh
species. Examination of the holotype shows that in frontal view the anterior branches of the facial
sutures converge evenly forwards and are not sinuous.

Family aulacopleuridae Angelin, 1854
Subfamily aulacopleurinae Angelin, 1854
Genus harpidella McCoy, 1 849
Subgenus harpidella McCoy, 1 849

Type species. Harpesl megalops McCoy, 1846.

Remarks. For diagnosis of genus and subgenus, discussion, and renewed separation of Harpidella
from genus Otarion see Thomas and Owens 1978, p. 71.

Harpidella ( Harpidella ) lacrymosa sp. nov.

Plate 108, figs. 10-14

1885 Cyphaspis megalops, M’Coy; Marr and Roberts, list p. 481.

1914 Cyphaspis megalops (McCoy); Strahan et al., table p. 63.

1973a Otarion aff. tenuis Kielan; Price, tables 1-4.

Holotype. SM A3 1471 (PI. 108, fig. 10), incomplete internal mould of exoskeleton from Sholeshook Limestone
of Sholeshook railway cutting.

Horizons and localities. See Table 1.

Diagnosis. Species of Harpidella ( Harpidella ) with elongate, pear-shaped medio-frontal glabellar lobe, drop-
shaped basal lateral lobes, very small 2p lobes; large eye-lobes reaching almost to posterior border furrow; free
cheeks with shallow lateral border furrows, broad convex borders, and concave lateral margins near bases of

explanation of plate 108

Figs. 1, 2. Illaenus ( Parillaenus ) cf. fallax Holm. SM A3 1500, internal mould of entire articulated specimen from
the Sholeshook Limestone of Sholeshook, dorsal and right lateral views, x \j.

Figs. 3-8. Panderia edita Bruton? 3, 4, HM A9572a, internal mould of incomplete cephalon from the ‘Bala
Limestone’ of Trefanty, 4 km south-east of St. Clears, dorsal view as used herein and ‘dorsal’ view sensu
Bruton 1968 (normal to occipital region), x8. 5, 6, SM A3 1503, internal mould of cephalon from

Sholeshook Limestone of Sholeshook railway cutting, left lateral and anterior views, x 6. 7, BM It9290,
internal mould of pygidium from Sholeshook Limestone horizon at Robeston Wathen (locality 10a), dorsal
view, x 4. 8, SM A99096, internal mould of pygidium from high Sholeshook Limestone (locality 8c) of
Prendergast, dorsal view, x 6.

Fig. 9. Amphitryon radians (Barrande)? SM A77947, internal mould of small, incomplete cranidium from high
Sholeshook Limestone of Prendergast (locality 8b or 8c), dorsal view, x 8.

Figs. 10-14. Harpidella ( Harpidella ) lacrymosa sp. nov. 10, SM A31471, HOLOTYPE, internal mould of
incomplete articulated cephalon and thorax from Sholeshook Limestone of Sholeshook railway cutting,
dorsal view, x 8. 11, 12, SM A77536, internal mould of incomplete cranidium, middle Sholeshook

Limestone, locality 9e, Sholeshook, dorsal and left lateral views, x 8. 13, SM A77592b, cast from external
mould of left free cheek, horizon and locality as for figs. 11, 12, dorsal view, x8. 14, SM A3 1470, internal
mould of incomplete articulated exoskeleton and SM A 104833, external mould of incomplete cephalon,
Sholeshook Limestone of Sholeshook railway cutting, A31470 in dorsal view, x 8.

Fig. 15. Stenopareia bowmanni (Salter). HM A 10397, internal mould of right free cheek, middle Sholeshook
Limestone, Lan-y-gaer (locality 16b), Llandowror, dorsal view, x 2.

PLATE 108

PRICE, Sholeshook trilobites

850 PALAEONTOLOGY, VOLUME 23

short, slender genal spines; subdued cephalic ornament of fine granules, free cheeks in addition pitted, borders
and genal spines smooth.

Description. Cephalon semicircular in outline and moderately convex (tr.). Glabella sub-parabolic, broadest (tr.)
posteriorly where occupies about one-third of total cephalic width. Median and frontal lobes together form
distinctive pear-shaped unit, strongly convex (tr.) and narrowest posteriorly. Anterior part of this unit defined by
deep, broad axial furrows which anteriorly contain shallow antero-lateral pits and by the broad but shallower
pre-glabellar furrow. Rear portion defined by posteriorly convergent lp lateral glabellar furrows which are
broad and deep and form continuations of the anterior parts of the axial furrows; they run back to the occipital
furrow and isolate the basal lateral lobes from the median lobe of the glabella. Basal lateral lobes drop-shaped,
strongly convex (tr.), occupying about one-third total glabellar length. Basal lobes defined abaxially by rear
portions of axial furrows; these posteriorly divergent, narrower, and shallower than either anterior portions or
lp lateral furrows. 2p furrows never clearly seen but appear to be developed faintly on at least two specimens
(SM A77950, not figured, and the holotype) just in front of anterior ends of lp furrows and thus delimiting very
small 2p lateral lobes. Occipital furrow broad and shallow; ring rather narrow (sag. and exsag.), mesially
transverse and sub-parallel sided but abaxially narrowing and curving forwards towards axial furrows. Pre-
glabellar field drops in steep, convex slope to broad, shallow anterior border furrow. Anterior border of about
same width as furrow, moderately convex and sloping forward less steeply than pre-glabellar field (PI. 108, fig.
12). Eye-lobes large, occupying almost one-third cephalic length; their mid-lengths occur slightly in front of the
basal lateral lobes and they reach posteriorly almost to the posterior border furrow. Palpebral lobes also large
but form not clearly seen. Anterior branches of facial sutures diverge at 40-50° until reaching anterior border
where they are deflected adaxially and run obliquely across border leaving narrow, triangular anterior tongues to
free cheeks (PI. 108, fig. 13). Posterior branches curve out and gently back. Free cheeks drop steeply to shallow
lateral border furrows and broad, convex borders and are extended postero-laterally as relatively short, slender
genal spines which carry a faint median furrow. Outer margins of genal spines and lateral borders merge in
abaxially concave curves. Transverse posterior border furrows broad and deep; borders narrow and strongly
convex adaxially but much broader and more gently convex towards genal angles. Cephalic surface bears fine,
rather subdued granulation; free cheeks in addition irregularly pitted; borders and genal spines smooth.

Thorax poorly preserved and number of segments uncertain, though at least ten. Strongly convex axis
occupies just over one-third total width (tr.) anteriorly. Axial rings convex, separated by strong articulating
furrows which abaxially deepen to apodemal slots separating prominent rounded axial lobes. Axial furrows
broad but shallow. Inner portions of pleurae transverse and horizontal, then bent down and deflected posteriorly
at fulcrum; divided roughly along median line by broad, shallow pleural furrows which gradually narrow
outwards; distal extremities bluntly terminated. Pygidium only very poorly preserved; much broader (tr.) than
long (sag.), posterior margin rounded and axis well defined anteriorly.

Discussion. H. ( H .) lacrymosa sp. nov. is most similar to ‘ Otarion ’ tenue Kielan (1960, p. 63, pi. 2, figs.
1-2; text-fig. 15) from the ‘ Staurocephalus clavifrons Zone’ of Poland but differs as follows: eye-lobes
larger and set closer to posterior border furrows; genal spines slenderer, narrower-based, and not
carrying continuations of lateral and posterior border furrows; lateral border wide and prominent,
border furrow weak; anterior border furrow not wider than anterior border; cephalic ornament more
subdued, comprising smaller and more closely spaced granules, borders and genal spines not
ornamented; anterior thoracic pleurae not sharply pointed. Cranidia from the Irish Chair of Kildare
Limestone figured by Dean (1974, p. 68, pi. 28, figs. 5, 8, 10, 12, 13; pi. 29, figs. 3, 5) as ‘O’, cf. tenue
differ from both Polish and Welsh specimens in having more divergent anterior branches to the facial
sutures and in possessing broader (sag. and exsag.) anterior borders which are more strongly arched
anteriorly. The Irish specimens are more coarsely ornamented than those from Wales.

Family proetidae Salter, 1864
Genus proetus Steininger, 1831
Proetus (s.l.) cf. berwynensis (Whittington, 1966)

Plate 107, figs. 16-17

1909 Proetus cf. brachypygus Marr and Nicholson; Strahan et al., table p. 58.

1973a Astroproetus aff. berwynensis Whittington; Price, pp. 233, 243.

1973 Proetus (s.l.) cf. berwynensis Whittington; Owens, p. 20, pi. 1, figs. 2-7.

PRICE: LATE ORDOVICIAN TRILOBITES

851

Horizons and localities. Relatively abundant in the Sholeshook Limestone horizon at Robeston Wathen (locality
10a). Also occurs in the ‘Bala Limestones’ of Trewern Quarry 3 km north-west of Whitland (Strahan et al. 1914,
p. 56) and Bron-haul about 2-5 km east-south-east of Llandowror (Strahan et al. 1909, p. 56).

Discussion. Owens (1973, see synonomy), who illustrated several specimens of this form, pointed out
the close similarity to P. berwynensis (Whittington), a species known only from a single specimen
(Whittington 1966, pi. 25, figs. 14-16; Owens 1973, pi. 1, fig. 1) from the Ashgill Dolhir Beds of
Cynwyd, 3 km south-west of Corwen, Clwyd. This holotype of P. berwynensis has the genal angles
poorly preserved and it is uncertain whether or not genal spines are developed as in the South Welsh
specimens (Owens 1973, pi. 1, figs. 2, 5).

One of the specimens figured here, a cast from an incomplete external mould of the cranidium (PI.
107, fig. 16) shows the short (sag. and exsag.) pre-glabellar field, a glabellar surface ornamentation of
scattered granules of 0-075-0T mm and a prominent occipital tubercle. The other specimen is the
internal mould of a pygidium (PI. 107, fig. 17). The axis comprises, in addition to the half-ring, four
axial rings and a short terminal piece; the doublure extends mesially almost to the tip of the axis.

Family trinucleidae Hawle and Corda, 1847
Subfamily trinucleinae Hawle and Corda, 1 847
Genus nankinolithus Lu, 1954

Type species. Nankinolithus nankinensis Lu, 1954.

Remarks. Ingham (1970, p. 44) in his discussion of genus Tretaspis McCoy referred to and briefly
characterized a species-group typified by T. granulata (Wahlenberg) and T. portrainensis Lamont.
Such forms have subsequently been removed from Tretaspis and placed in genus Nankinolithus Lu
(Hughes et al., 1975, pp. 558-559, see p. 558 for diagnosis).

Nankinolithus cf. granulatus (Wahlenberg, 1818)

Plate 109, figs. 1-10

1885 Trinucleus seticornis, var. Bucklandi, Barr.; Marr and Roberts, pp. 480, 481.

1914 Trinucleus seticornis (His.); Strahan et al. (pars), table p. 64, faunal lists p. 76.

1914 Trinucleus seticornis (His.), var. bucklandi Barr.; Strahan et al., table p. 64.

1916 Trinucleus seticornis (His.), var. bucklandi Barr.; Cantrill et al., faunal list p. 50.

1973a Tretaspis cf. granulata (Wahlenberg); Price, pp. 229, 234, 241, tables 1-3, 7.

Horizons and localities. Abundant in basal 2 or 3 m of Sholeshook Limestone at Sholeshook but rarer in low
Sholeshook Limestone around Llandowror (locality 19); also abundant in basal Slade and Redhill Mudstones
between Pelcomb and Clarbeston Road Station (localities 1-6).

Description. Cephalon almost as broad (tr.) as long. Occipital ring moderately arched transversely,
longitudinally narrow, and not strongly convex, almost straight in dorsal view; furrow abaxially containing
deep, ovoid apodemal slots. Ip apodemal slots converge anteriorly at about 110°. 2p furrows in form of large
ovoid pits diverging anteriorly at 110-120°. 3p lateral furrows usually visible, even in internal moulds, as shallow
ovoid pits near mid-length of pseudofrontal lobe. Anterior fossulae only developed as very shallow depressions,
not always visible. Genal lobes moderately convex (tr. and exsag.), steeply declined antero-laterally but not
overhanging fringe; not bearing lateral tubercles or eye-ridges. Both pseudofrontal glabellar lobe and genal lobes
smooth. Posterior border furrows shallow, abaxially containing large posterior fossulae. Posterior border
narrow and only weakly convex. Upper lamella of fringe anteriorly comprises steep, slightly concave genal roll
merging into narrow horizontal brim; laterally drops outwards in smooth, gently concave curve. Genal
prolongations reach almost as far back as posterior margin of pygidium (PI. 109, fig. 2). Long slender genal spines
produced beyond these (PI. 109, fig. 10) have strong ventral ridges continuous with girder. Pits of E2, E1; and L
arcs radially in line. Frontally and antero-laterally on upper lamella exist as clearly separate pits (PI. 109, fig. 8),
but on genal prolongations Et and L tend to be contained in short radial sulci (PI. 109, fig. 10) and occasionally
these may contain of E2 also; Ej and E2 gradually merge posteriorly but usually remain present as separable pits
in all but posterior-most two or three radii. On a few specimens the E arcs are most closely merged antero-
laterally and become slightly more separated posteriorly (PI. 109, fig. 10). Lower lamella has more distinct

852

PALAEONTOLOGY, VOLUME 23

change of slope between genal roll and brim and a broad girder is present (PL 109, figs. 2, 6); E2 and E2 pits
gradually merge posteriorly and are present beyond R12 or R13 as conjunct pit-pairs (PI. 109, figs. 2, 5-6),
though merging completely only in posterior-most two or three radii. Number of pits in Ej (half-fringe) ranges
from 28 (2 specimens), through 29 to 32 (1 specimen in each case). On all specimens the pits of the innermost two
I arcs (In and that adjacent) are radially in line (PI. 109, figs. 3, 4, 8) but pits of the arcs between these and E are
arranged very irregularly and difficult to count (see PI. 109, fig. 6). Usually there are 4 I arcs frontally (E_3, In)
and antero-laterally 5 or 6 (i.e. I4 or I4 and I5 also present); at least 1 specimen (SM A77527 from Sholeshook)
appears to have 7 I arcs (Ii_6, In) antero-laterally. The number of I arcs increases by intercalation on the genal
prolongations until there are 13 (4 specimens), 14 (2 specimens), 15 (1 specimen), 16 (4 specimens), or 17 (3
specimens) pits in the posterior row. Two specimens differ from description so far given. One, a fringe fragment
shown in PI. 109, fig. 7 is unique in showing well-developed sulci containing pits of the E2, E1? and E arcs. The
other, the cephalon seen in PI. 109, figs. 3, 4, is unique in that while two E arcs are developed frontally on the
upper lamella only one E arc is present laterally from about R16.

Axis of thorax strongly convex, occupying one-fifth of total width anteriorly; rings broadest (sag. and exsag.)
mesially where gently arched forward, narrow and curved forwards abaxially; dorsal surfaces flat in lateral
profile. Axial furrows broad and prominent. Pleural lobes flat. Pleurae transverse and horizontal for most of
length but deflected ventrally and slightly posteriorly at distal ends. Pleural furrows adaxially narrow, commence
near inner anterior corners of pleurae and run slightly obliquely, broadening outwards and separating narrow
anterior and broad, strongly convex posterior pleural bands.

Pygidium about 2\ times as wide as long, with convex postero-lateral margins. Gently convex axis, anterior
occupies one-quarter of total width and tapers back at 35°. Mesially broad articulating furrows become
elongated apodemal pits abaxially; eight such pairs of pits developed. Flat pleural regions crossed by five broad
(exsag.), abaxially expanding pleural bands increasingly faintly defined posteriorly. Anterior-most bands carry
narrow, faint pleural furrows and are separated by broad interpleural furrows. Strong sub-marginal rim. Slightly
bevelled posterior margin carries faint, irregular terrace-lines.

Discussion. Material, including the holotype, of ‘Tretaspis' granulata described by Kielan (1960, p.
171, pi. 32, figs. 1-3; pi. 34, figs. 1, 2; pi. 35, figs. 1, 2; pi. 36, fig. 6; text-fig. 49) from the upper
Ordovician of Poland, Sweden, and Bohemia differs mainly in the presence (both frontally and
laterally on the fringe upper lamella) of well-developed sulci containing the E2, Ex, and E pits.

EXPLANATION OF PLATE 109

Figs. 1-10. Nankinolithus cf. granulatus (Wahlenberg). 1, 2, SM A3 1606b, a, cast from external mould and
internal mould of articulated exoskeleton from low horizon in Slade and Redhill Mudstones near Pelcomb
Cross (locality 2), dorsal views, x 3. 3, 4, GSM TCC. 1736, internal mould of large, well-preserved cephalon
from low(?) horizon in Slade and Redhill Mudstones, quarry south of Marlsborough, 7 km west-north-west of
Haverfordwest, left-lateral and dorsal views, x 3. 5, 6, SM A77526, internal mould of incomplete cephalon
from basal Sholeshook Limestone, south end of Sholeshook railway cutting, dorsal and anterior views, x 3. 7,
SM A77675a, fragment of upper lamella of fringe with E2, E1? and E pits in well-developed sulci, basal Slade
and Redhill Mudstones south-west of Knock (locality 3), oblique view, x 4. 8, SM A3 1 6 1 6, slightly distorted
internal mould of cephalon, horizon, and locality as for figs. 5, 6, anterior view, x 4. 9, GSM TCC. 1178, cast
from external mould of pygidium, horizon, and locality as figs. 5, 6, dorsal view, x 3. 10, SM A77700, left
genal area and prolongation of fringe upper lamella and genal spine, basal Slade and Redhill Mudstones of
Withy Hedge (locality 5), oblique view, x 3.

Figs. 11, 12. Dionide sp. indet. 11, SM A3 161 9, internal mould of partial cephalon, horizon, and locality as for
figs. 1, 2, dorsal view, x 6. 12, SM A77534, internal mould of incomplete cephalon from middle Sholeshook
Limestone, locality 9e, Sholeshook.

Fig. 13. Lonchodomas aff. pennatus (La Touche). SM A77629, cast from incomplete external mould of
cranidium, horizon, and locality as for fig. 7, dorsal view, x 10.

Figs. 14-16. Raphiophorus cf. tenellus (Barrande). 14, 15, BM It8101b, a, external mould showing base of
frontal spine and internal mould of cranidium, horizon, and locality as for fig. 12, dorsal views, x 10. BM
It8091a, internal mould of small incomplete articulated exoskeleton, same horizon and locality, dorsal view,
x 10.

PLATE 109

PRICE, Sholeshook trilobites

854

PALAEONTOLOGY, VOLUME 23

Although the species is described as having two E arcs present anteriorly and only one developed
laterally (Kielan 1960, p. 172), several of the illustrated specimens show separate Ex and E2 arcs
present in all but the posterior-most two or three radii of the fringe as in the south Welsh material.
The south Welsh specimens appear to have similar numbers of pit arcs frontally on the fringe but
rather more in the posterior row (cf. Kielan’s fig. 49). Kielan does not give figures for the number of
Ex pits in the half-fringe but, to judge from the illustrations and a few Polish specimens in the collec-
tions of the British Museum (Natural History), the number ranges from around twenty-seven to
around thirty-one.

Family dionididae Gurich, 1907
Genus dionide Barrande, 1847

Type species. Dionide formosa Barrande, 1846.

Dionide sp. indet.

Plate 109, figs. 11, 12

1885 Illaenus (young); Marr and Roberts, faunal list p. 481.

1973a Dionide sp. indet.; Price, tables 1, 2.

Material. Two internal moulds of incomplete cephala; SM A31619 from basal Slade and Redhill Mudstones
near Pelcomb Cross (locality 2) and SM A77534 from middle Sholeshook Limestone, locality 9e, Sholeshook.

Description. Cephalon much wider (tr.) than long (sag.). Glabella sub-quadrate, narrowest (tr.) across occipital
ring and narrowing again just in front of mid-length; moderately convex (tr.). Occipital ring narrow and convex
(sag. and exsag.), occipital furrow narrow; both gently arched forward. Deep lp lateral glabellar furrows run
forward from occipital furrow diverging slightly and reaching to about one-quarter of total glabellar length.
Genal lobes sub-quadrant shaped, apparently pitted (PI. 109, fig. 1 1), and one specimen (PI. 109, fig. 12) shows
traces of what appear to be genal caecae. This same specimen also clearly shows the inner margin of the fringe
and a ventral mould of the lower lamella with tuberculation reflecting the fringe pitting. Posterior border broad
and strongly convex (sag. and exsag.).

Discussion. The incompleteness of these specimens does not permit very useful comparison with other
dionidid species. It may be noted, however, that the narrowing (tr.) of the glabellar around its mid-
length is an unusual feature though it may be seen also in specimens of D. richardsoni Reed from the
upper Whitehouse Beds of Girvan (Reed 1903, pi. 4, fig. 3). Ingham (1974, p. 64) has noted that
specimens of Reed’s species from the upper Whitehouse Beds and those from the upper Drummuck
Group represent distinct forms.

Family raphiophoridae Angelin, 1854
Genus raphiophorus Angelin, 1854

Type species. Raphiophorus setirostris Angelin, 1 854.

Raphiophorus cf. tenellus (Barrande, 1 872)

Plate 109, figs. 14-16

1885 Ampyx tumidus Forbes; Marr and Roberts (pars), lower list p. 481 .

1973a Raphiophorus sp. indet.; Price, tables 1, 2.

Material. Internal and external moulds of small articulated exoskeleton (PI. 109, fig. 16) and of cranidium (PI.
109, figs. 14, 15) both from locality 9e, Sholeshook, and two internal moulds of cranidia (SM A31384-5) from
the basal Slade and Redhill Mudstones of Pelcomb Cross (locality 2). A partial cranidial external mould (SM
A99477) from the basal Slade and Redhill Mudstones south-west of Knock (locality 3) probably also belongs
here.

PRICE: LATE ORDOVICIAN TRILOBITES

855

Description. Ovoid glabella strongly convex transversely, moderately so longitudinally; anteriorly bluntly
pointed and projecting well beyond fixed cheeks, posteriorly contracts (tr.) rapidly between broad triangular
depressions confluent with axial and occipital furrows. Latter furrow narrow, shallow mesially. Sub-triangular
fixed cheeks strongly declined antero-laterally. Posterior border furrows strong, set slightly oblique, deepening
abaxially; borders broad and prominent. Occipital ring narrow (sag. and exsag.), gently arched dorsally.
Cranidial external moulds show position of circular-sectioned frontal spine, also apparent lack of surface
ornamentation. The articulated Sholeshook specimen appears foreshortened due to folding between the thoracic
segments and the absence of the anterior part of the glabella but shows the course of the genal spine and the
relatively large pygidium with its broad border.

Discussion. The South Welsh specimens are most like R. tenellus from the Ashgill of Sweden, Poland,
and Bohemia (Kielan 1960, p. 165, pi. 35, fig. 6; Whittington 1968, p. 94, text-fig. 6). R. tenellus is like
R. setirostris in general form but the type species appears to have a relatively shorter (sag.) and wider
glabella projecting less far beyond the fixed cheeks and with a shorter constricted posterior section.
R. acws(Troedsson)from the Ashgill of Poland (Kielan 1960, p. 168, pi. 32, fig. 4; pi. 35, fig. 7) and the
high Rawtheyan of the southern Lake District (McNamara 19796, table 2) has a glabella which
projects less far forward than in either R. tenellus or R. setirostris and which is broadly rounded
anteriorly; it also has relatively longer (exsag.) and narrower fixed cheeks which are strongly convex
antero-laterally and a smaller pygidium.

Genus lonchodomas Angelin, 1854
Type species. Ampyx rostratus Sars, 1835.

Lonchodomas aff. pennatus (La Touche, 1884)

Plate 109, fig. 13; Plate 110, figs. 1-3

1 885 Ampyx tumidus Forbes; Marr and Roberts (pars), higher list p. 48 1 .

1914 Ampyx tumidus Forbes; Strahan et al., table p. 63.

1973a Lonchodomas tumidus (Forbes); Price (pars), tables 1-3, 7.

Horizons and localities. Apart from occurrences shown in Table 1, also abundant in basal Slade and Redhill
Mudstones to north and west of Haverfordwest (localities 2, 3, 5).

Description. Cranidium broadly triangular with sagittal length (excluding frontal spine) about four-fifths the
maximum width. Glabella about twice as long (sag.) as wide (tr.) with maximum width just in front of mid-
length; strongly convex (tr.), standing high above fixigenae, frequently carinated; produced anteriorly into long,
slender frontal spine (PI. 1 10, fig. 1) which is sub-square in cross-section. Broad, shallow axial furrows separated
by one-fifth of cranidial width posteriorly, diverging forwards at about 40°, containing deep, slot-like fossulae
anteriorly. Fixigenae about as long (exsag.) as posteriorly wide (tr.); only gently convex. Occipital furrow
continuous with posterior border furrows, both broad and shallow, latter contain small, deep, round pits
abaxially. Occipital ring very narrow (sag. and exsag.), gently arched posteriorly, continuous laterally with very
narrow but convex posterior borders. External moulds show cranidial surface covered with small pits of about
0 03 mm diameter (PI. 109, fig. 13).

Pygidium more than twice as wide (tr.) as long (sag.) and broadly rounded posteriorly. Axis moderately
convex (tr.) and raised above pleural lobes, occupies one-third total width anteriorly and tapers back at 30°. On
internal moulds only articulating furrow on axis are clearly visible but there are faint indications behind of
several paired pits. A pydium with some of the exoskeleton preserved (SM A77766) shows eight pairs of raised
muscle scars on the axis. Axial furrows shallow and indistinct. Pleural lobes only gently convex, with steeply
declined wide borders; only one pair of pleural furrows is clearly visible though there appear to be faint traces of
two or three pairs of ribs behind.

Discussion. The South Welsh cranidia are similar in outline and proportions to cranidia of the high
Caradoc-low Ashgill form L. pennatus ( see Dean 1960, pi. 1 1, figs. 2, 5, 8-12; 1962, pi. 6, figs. 1, 3-5, 9,
12) but differ in the pitting of the cranidial surface and the much shallower posterior border furrows.
In these features they are like cranidia of the form described by Ingham (1974, p. 65, pi. 1 1, figs. 6-14)
as L. aff. pennatus from Zones 1 and 2 (and possibly 4) of the Cautley Mudstones. None of the south

856

PALAEONTOLOGY, VOLUME 23

Welsh pygidia, however, show clearly the development of the two or more pleural furrows seen in the
Cautley form. Ingham noted the similarity between the Cautley form he described and a specimen,
probably from the Dolhir Beds near Corwen, referred by Whittington (1968, pi. 30, figs. 13, 15,
1 8-20) to L. tumidus (Forbes). As in the south Welsh specimens, the pygidium of this form shows only
one clear pair of pleural furrows.

Lonchodomas cf. drummuckensis (Reed, 1903)

Plate 1 10, figs. 4-6

1914 Ampyx drummuckensis Reed; Strahan et al., table p. 63.

1973a Lonchodomas tumidus (Forbes); Price (pars), tables 1-3.

71974 Lonchodomas aff. portlocki (Barrande); Ingham, pp. 65-66, pi. 12, figs. 1-13.

Material. Articulated pygidium and incomplete thorax, three pygidia and one small, incomplete cranidium, all
from either the top 4 m of the Sholeshook Limestone or the base of the overlying Slade and Redhill Mudstones
(localities topmost 8d to basal 8a) at Prendergast Place. A few distorted or incomplete larger cranidia from the
same localities probably belong here too (e.g. SM A3 1 1 58), and an incomplete cranidium (HM A9656) from the
basal Slade and Redhill Mudstones of Clog-y-fran (locality 1 5) near Llandowror.

Description. Most complete and undistorted cranidium very small (PI. 110, fig. 6); broadly triangular with
sagittal length just over four-fifths maximum width. Glabella, excluding frontal spine, over twice as long (sag.) as
wide (tr.) with maximum width at about two-fifths its length from posterior margin; only moderately convex
(tr.), slightly carinate anteriorly. Axial furrows shallow and indistinct, diverging forwards at about 50°, with

EXPLANATION OF PLATE 110

Figs. 1-3. Lonchodomas aff. pennatus (La Touche). 1, SM A77701, internal mould of partial cranidium showing
form of frontal spine, basal Slade and Redhill Mudstones of Withy Hedge (locality 5), dorsal view, x 2. 2,
SM A77632, poorly preserved internal mould of pygidium from same locality, dorsal view, x 6. 3, SM
A77684a, internal mould of incomplete cranidium from basal Slade and Redhill Mudstones south-west of
Knock (locality 3), dorsal view, x 8.

Figs. 4-6. Lonchodomas cf. drummuckensis (Reed). 4, SM A 104835a, internal mould of pygidium and
incomplete thorax from basal Slade and Redhill Mudstones of Prendergast (locality 8a), dorsal view, x 2. 5,
BM It9249, distorted internal mould of pygidium from highest Sholeshook Limestone of Prendergast (locality
8b), dorsal view, x 3. 6, BM It9247, internal mould of small partial cranidium from the high Sholeshook
Limestone (topmost locality 8d) of Prendergast, dorsal view, x 1 0.

Figs. 7, 8. Ceraurinella intermedia (Kielan). 7, SM A3 1585, internal mould of incomplete pygidium from the
Sholeshook Limestone of Sholeshook railway cutting, dorsal view, x 3. 8, BM It9216, internal mould of
slightly distorted incomplete cranidium, horizon, and locality as for fig. 4, dorsal view, x 3.

Figs. 9-11. Pseudosphaerexochus tectus Ingham. 9, SM A77875, cast from external mould of pygidium from the
high Sholeshook Limestone, locality 9g, Sholeshook, dorsal view, x 5. 10,11, SM A3 1 586, internal mould of
cranidium from Sholeshook Limestone of Sholeshook, right lateral view, x 3, and dorsal view, x 2.

Figs. 12-14. Sphaerocoryphe aff. thomsoni Reed. 12, HM A9733, internal mould of flattened, incomplete
cranidium from the low Sholeshook Limestone, track south of Craig-y-deilo quarry, Llandowror (locality
18d), dorsal view, x4. 13, BM In54702, cast from external mould of partial cranidium from Sholeshook
Limestone of Craig-y-deilo quarry, Llandowror, dorsal view, x 4. 14, BM It9245, cast from external mould
of partial cranidium from middle Sholeshook Limestone, locality 9e, Sholeshook, dorsal view, x 4.

Fig. 1 5. Pseudosphaerexochus juvenis (Salter). SM A77887b, cast from external mould of small cranidium from
middle Sholeshook Limestone, locality 9e, Sholeshook, dorsal view, x 10.

Fig. 16. Cybeloides ( Paracybeloides ) girvanensis (Reed). SM A77531, cast from external mould of incomplete
cranidium from the Sholeshook Limestone of Sholeshook railway cutting, dorsal view, x 4.

Fig. 17. Atractopyge scabra Dean? SM A77971, partial internal mould of cranidium from 9^-10 m above base of
Sholeshook Limestone in Mylet road section (locality 24a), Llandowror, dorsal view, x 3.

Fig. 18. Staurocephalus cf. clavifrons Angelin. HM A9751, internal mould of cranidium, horizon, and locality as
for fig. 12, dorsal view, x 6.

PLATE 110

PRICE, Sholeshook trilobites

858

PALAEONTOLOGY, VOLUME 23

small, slot-like fossulae anteriorly. Fixed cheeks about three-quarters as long (exsag.) as wide (tr.) and gently
convex dorsally, though dropping steeply antero-laterally. Posterior border furrows shallow and indistinct
adaxially but on internal moulds deepen and widen outwards into broad slots. Occipital furrow shallow and
indistinct; ring continuous with posterior borders, gently arched posteriorly. Posterior borders convex (exsag.)
and distinct, about as broad as posterior border furrows.

Other much larger cranidia fragmentary or distorted but agree in showing shallow axial furrows, strongly
developed posterior borders, and border furrows which are indistinct adaxially but well developed laterally;
some show the free anterior part of the glabella to be sub-circular in section.

Thorax known from posterior four segments (PI. 1 10, fig. 4). Axis occupies about one-third total width; rings
flat (sag. and exsag.) dorsally, ring furrows shallow mesially and laterally but deepened between into paired slots.
Axial furrows broad and deep. Pleural furrow on anterior-most segment (second) continuous and transverse but
on those behind runs obliquely out and forwards from position about one-third pleural width from axial furrow.
Pygidium just less than twice as wide (tr.) as long (sag.) and bluntly pointed posteriorly. Axis occupying just less
than one-third total width anteriorly and tapering back at 30°; behind the articulating furrow internal moulds
show a series of six (PI. 1 10, fig. 4) or seven (PI. 1 10, fig. 5) pairs of elongated (tr.) pits. Axial furrows broad and
distinct. Pleural lobes with gently convex and steeply declined postero-lateral margins. Single pair of pleural
furrows, shallow adaxially but distinct for outer two-thirds of length and curving forwards towards antero-
lateral corners of pleural lobes.

Discussion. The South Welsh pygidia closely resemble those of the Rawtheyan form from Cautley
described by Ingham (1974) as L. aff. portlocki (Barrande) and also those of L. drummuckensis (Reed
1903, p. 18, pi. 3, figs. 1-5) from the late Rawtheyan upper Drummuck Group of Girvan. The Polish
and Bohemian Ashgill species L. portlocki (see Kielan 1960, pi. 33, fig. 8; pi. 35, fig. 4) has a pygidium
which is relatively much shorter (sag.). The specimens figured by Kielan are small but even on a much
larger Polish specimen in the Sedgwick Museum (A44183) the pygidium is three times as wide as long
and posteriorly broadly rounded. The small Sholeshook cranidium strikingly resembles a small one
figured by Ingham (1974, pi. 12, fig. 5) as L. aff. portlocki but differs from his larger cranidia. These
are said to differ from cranidia of L. drummuckensis in having a relatively shorter, less inflated, and
less well-defined glabella. The lack of large south Welsh cranidia makes similar comparisons difficult.
From the small Sholeshook cranidium alone the glabella does appear to be similar to that of L.
drummuckensis to judge from specimens of the latter in the Sedgwick Museum (e.g. A 1 09 18, A1 1 103,
A52598), though these differ slightly from the South Welsh form in possessing narrower (exsag.)
posterior borders.

Family cheiruridae Hawle and Corda, 1847
Subfamily eccoptochilinae Lane, 1971
Genus pseudosphaerexochus Schmidt, 1881

Type species. Sphaerexochus hemicranium Kutorga, 1854.

Pseudosphaerexochus juvenis (Salter, 1848)

Plate 110, fig. 15; Plate 111, figs. 8-11

1 974 Pseudosphaerexochus ( Pseudosphaerexochus ) juvenis (Salter); Price, pp. 849-850, pi. 1 1 3, figs. 5-9.
Includes full synonomy.

71974 Pseudosphaerexochus conformis (Angelin); Ingham, pp. 70-71, pi. 14, figs. 6-12.

Lectotype. Subsequently designated Whittington 1965, p. 40; GSM 24534, internal mould of cranidium from
Sholeshook Limestone of Sholeshook; figured Whittington 1965, pi. 12, figs. 2, 4, 8.

Horizons and localities. See Table 1.

Description. The lectotype and other GSM specimens used by Whittington (1965, see synonomy) in his
redescription of P. juvenis are all indifferently preserved and most are distorted. Better-preserved topotype
cranidia described here, give a much improved idea of the characters of this form. Cranidium about 1^

PRICE: LATE ORDOVICIAN TRILOBITES

859

times as wide (tr.) as long (sag.). Glabella occupies three-fifths total cranidial width; ovoid in outline with
parabolic anterior margin; pre-occipital length about four-fifths maximum width, latter on level of 2p lateral
lobes; strongly convex (tr. and sag.), in lateral profile greatest convexity is over anterior half, frontal lobe
dropping steeply forward (PI. 1 1 1, fig. 10). Basal lateral lobes in dorsal view are obliquely elongated ovoids, each
occupying one-fifth of maximum glabellar width (tr.) and separated posteriorly by about \j times this width.
Basal lateral furrows broad and deep abaxially, gently curved, dying out before reaching occipital furrow.
Exsaggital length of 2p lateral lobes about two-thirds that of basal lobes, 3p lobes slightly shorter. 2p and 3p
furrows shallower and narrower than lp and short (tr.) in dorsal view. Occipital furrow broad and deep, ring
broad and convex (sag. and exsag.), both arched posteriorly in dorsal view. Axial furrows deep and slot-like,
confluent with broad anterior border furrow, containing small round pits just in front of 3p furrows. Convex (tr.)
fixed cheeks with concave antero-lateral margins in dorsal view, genal angles broadly rounded, no genal spine
seen. Posterior border furrows broad and deep adaxially, shallowing outwards, borders narrow and convex
adaxially, broadening and flattening outwards. Palpebral lobes of similar length (exsag.) to 2p lateral glabellar
lobes, situated slightly behind them, running obliquely out and back, gently convex antero-laterally; palpebral
furrows broad and distinct. Posterior branches of facial sutures meet lateral borders in rounded curves. External
surface of glabella covered with scattered large granules or small tubercles (PI. Ill, fig. 11) which are more
prominent in small specimens (PI. 110, fig. 15). On internal moulds the glabella is sometimes finely granulated.
Fixigenal surface strongly pitted, pits usually visible on internal moulds.

Librigenae, thorax, and hypostoma unknown. Pygidium known only from incomplete specimens of which the
best have been figured previously (Whittington 1968, pi. 31, fig. 17; Price 1974, pi. 113, fig. 9).

Discussion. The ovoid glabella with its frontally parabolic outline, its more gently curved basal
furrows and relatively narrower basal lobes, and its ornament of scattered large granules clearly
differs from that of P. tectus Ingham also common in the Sholeshook Limestone (see PI. 1 10, figs. 10,
1 1). In cranidial characters. P.juvenis is more like P. octolobatus (McCoy) (see Lane 1971, pi. 8) but
that form has a glabella which is relatively broader posteriorly, with a shorter (sag. and exsag.), more
broadly rounded frontal lobe and less strongly convex anteriorly in lateral profile; also the palpebral
lobes are shorter and placed further forward and the ornament includes fine as well as scattered coarse
granules.

Cranidia from the Chair of Kildare and Kiesley Limestones figured by Dean (1971, pi. 9, figs. 3, 4,
8; pi. 10, figs. 1-3, 6, 8, 10-12; pi. 11, figs. 4-8, 11, 12) as P. conformis Angelin bear prominent genal
spines. The glabella of these forms is more broadly rounded frontally than that of P.juvenis and more
evenly convex in lateral profile with a shorter and less steeply inclined frontal lobe and more
prominent surface tubercles. In cranidia from the Chair of Kildare Limestone (e.g. pi. 9, fig. 4) the 3p
lateral glabellar lobes are longer than the 2p. Pygidia figured by Ingham (1974, pi. 14, figs. 8-10) as
P. conformis from Zones 2 and 3 of the Cautley Mudstones are strikingly similar to those from the
Sholeshook Limestone here referred to P.juvenis (cf. Price 1974, pi. 113, fig. 9), even to the surface
perforations. They are not, even the smallest of them, like the small pygidium figured by Dean (1971,
pi. 10, figs. 4, 5) from the Chair of Kildare Limestone in which the pleural regions are extremely
narrow and the spines longer, slenderer, and directed more strongly posteriorly. The fragmentary and
distorted cranidia from Cautley (Ingham 1 974, pi. 14, figs. 6, 7, 1 1 , 1 2) do not allow close comparison
with other forms.

Subfamily deiphoninae Raymond, 1913
Genus sphaerocoryphe Angelin, 1854

Type species. Sphaerocoryphe dentata Angelin, 1854.

Sphaerocoryphe aff. thomsoni Reed, 1906
Plate 110, figs. 12-14

1973a Sphaerocoryphe cf. thompsoni Reed; Price, tables 1-3.

Material, horizons, and localities. HM A9733, internal mould of flattened, incomplete cranidium, low
Sholeshook Limestone, track south of Craig-y-deilo quarry, Llandowror (locality 18d); BM In54702, external

860

PALAEONTOLOGY, VOLUME 23

mould of partial cranidium, Sholeshook Limestone, Craig-y-deilo quarry; SM A3 1 590, internal mould of partial
cranidium, Sholeshook railway cutting; BM It9245, external mould of partial cranidium, locality 9e,
Sholeshook.

Description. Anterior part of glabella sub-spherical, twice as wide (tr.) as central lobe behind and separated by
broad, mesially shallow furrow deepening laterally to pair of apodemal pits. Low posterior part of glabella short
(sag. and exsag.), broader (tr.) than long (sag.), only moderately convex (tr.). Basal lateral glabellar lobes sub-
triangular, small but distinct, strongly convex (exsag.). Occipital furrow shallow and gently arched forward
mesially, abaxially deepens to pair of apodemal pits. Occipital ring moderately broad (sag. and exsag.), arched
forward mesially and abaxially curving forwards around apodemal pits. Axial furrows broad, posteriorly sub-
parallel, containing occipital and lp apodemal pits but shallow opposite lp lateral lobes. Sub-triangular fixed
cheeks strongly convex, apically bearing pedunculate palpebral lobes; dropping steeply to broad, deep posterior
border furrows; posterior borders broad (exsag.) and convex. Broad, convex lateral borders separated from
inner parts of cheeks by prominent furrows; bear two short, broad-based pro-fixigenal spines (pi. 110, fig. 12) of
which posterior is larger. Lateral and posterior borders produced into long, stout, gradually tapering fixigenal
spines. Inflated part of glabella with ornamentation of prominent tubercles and much smaller granules densely
scattered between (PI. 1 10, fig. 14); tubercles large (0-2-0-25 mm) and widely spaced apically, smaller and more
densely packed marginally. Convex surface of cheeks pitted, though not strongly.

Discussion. Lane (1971, p. 64, pi. 13, figs. 1-4, 6-8, 10-18; pi. 15, fig. 9) selected a lectotype from
amongst Reed’s material and redescribed this and other specimens of S. thomsoni from the Starfish
Bed of Girvan. The cranidia of that species are similar in over-all form and proportions to those
described above and the only major difference appears to be the much coarser glabellar tuberculation
in the south Welsh specimens. S. kingi Ingham (1974, pp. 71-74, pi. 14, figs. 13-17; text-fig. 22) from
the Rawtheyan Stage of the Cautley Mudstones also has a more subdued glabellar ornamentation
than the Sholeshook specimens and the fixed cheeks are relatively much broader (tr.). S. punctata
(Angelin 1854, p. 77, pi. 39, fig. 6; Warburg 1925, pp. 390, 421; pi. 10, figs. 43-49) from the Boda
Limestone of Sweden is in need of redescription. A cranidium from the Chair of Kildare Limestone of
eastern Ireland referred to this species by Dean (1971, p. 33, pi. 16, figs. 1, 4, 7, 10) has a coarse
glabellar tuberculation like the south Welsh form but the inflated anterior glabellar region appears to
be proportionally smaller and the fixed cheeks are much more strongly pitted.

Family encrinuridae Angelin, 1854
Subfamily cybelinae Holliday, 1942
Genus atractopyge Hawle and Corda, 1 847

Type species. Calymene Iverrucosa Dalman, 1827.

Atractopyge aff. scabra Dean, 1962
Plate 111, figs. 1-4

1848 Cybele sexcostata, Salter, p. 343, pi. 8, figs. 9, 9a, 9b, non 10.

1853 C. ( Calym .) verrucosa, Dalman; Salter, Articla 4, p. 4.

1866 Cybele verrucosa, Dalm.; Salter, p. 324, pi. 19, fig. 7.

1885 Cybele verrucosa, Dalm.; Marr and Roberts, pp. 480, 481.

1909 Cybele verrucosa (Dalm.); Strahan et al, table p. 58.

1914 Cybele verrucosa (Dalm.); Strahan et al., table p. 63.

1973 Atractopyge scabra Dean; Price {pars), tables 1-3, list p. 233.

71974 Atractopyge sp.; Ingham, p. 82, pi. 17, figs. 1-6.

Horizons and localities. As in Table 1; not known from Slade and Redhill Mudstones other than at Prendergast
(locality 8a).

Description. Clavate glabella strongly convex (tr.); maximum width across frontal lobe less than pre-occipital
length. Occipital ring broad and strongly convex (sag. and exsag.), mesially arched forward; abaxially broadened
to form forward-curving occipital lobes. Occipital furrow broad and shallow mesially, abaxially dropping to

PRICE: LATE ORDOVICIAN TRILOBITES

861

deep, circular pits. Basal lateral lobes sub-triangular, anterior margins set strongly oblique. Ip furrows are deep
apodemal slots with triangular outlines and tendency to bifurcate adaxially. 2p and 3p lobes of approximately
equal length (exsag.) adaxially, set slightly oblique; 2p lobe narrows outwards, 3p sub-parallel sided. 2p furrows
deep, ovoid apodemal pits oblique in same manner as 2p and 3p lobes. 3p slots oblique, posteriorly divergent,
broadening (exsag.) inwards and continued adaxially as short, shallow bifurcating branches. Median lobe
narrowest on level of lp lateral furrows, widening only very gently forwards. Glabella expands rapidly in front of
3p lobes to maximum width three times that at lp furrows. Frontal lobe drops anteriorly in steep, convex slope.
Anterior border furrow narrow but distinct, transverse mesially; distally is deflected postero-laterally and
broadens; thus has form of three shortest sides of trapezium. Frontal glabellar lobe and anterior border can also
have similar outline (PI. 1 1 1, fig. 4) but usually appear more rounded. Axial furrows deep and broad; containing
deep circular pits near ends of anterior border furrow. Fixed cheeks much wider (tr.) than long (exsag.), strongly
convex (exsag.), dropping steeply to axial furrows; surmounted by palpebral lobes which form parts of long,
slender eye-stalks. These opposite 2p furrows and posterior halves of 3p lobes, separated from axial furrows by
distance equal to width of median lobe at that level. Eye-ridges run inwards and forwards to positions opposite
3p furrows. Behind these ridges are furrows which are narrow adaxially but expand outwards to form, around
the bases of the palpebral lobes, prominent depressed areas granulated and pitted but devoid of the coarse
tubercles seen on the rest of the cheek surface. Posterior border furrows broad, deep slots adaxially, borders
strongly convex over transverse inner halves then broaden as curve out and back. Free cheeks quadrant-shaped;
convex inner portions surmounted by narrow eye-stalks and separated from broad, convex borders by strong
furrows; borders produced into narrow anterior ‘tongues’ whose ends are deflected ventrally. Cranidial surface
ornamented with small, closely spaced granules (0-04-0-07 mm); in addition there are numerous much larger (up
to 0-75 mm), scattered, apically perforated tubercles, themselves granulated and absent only from the major
furrows. On mesial section of anterior border large tubercles form two alternating rows. Many of tubercles on
glabella developed in relatively constant symmetrical pattern. Surfaces of fixed cheeks irregularly pitted.

Hypostoma and rostral plate unknown; rostral suture runs along straight margin of mesial section of anterior
border, connective sutures along adaxial margins of extreme inner ends of anterior ‘tongues’ of free cheeks.

Pygidium slightly longer (sag.) than broad (tr.). Axis moderately convex (tr.), tapers posteriorly at 20°; up to
twenty rings in well-preserved material. Only first four rings continuous across axis, fifth and subsequent ring
furrows fail to reach axial furrows. Posteriorly ring furrows also become increasingly shallower mesially though
usually continuous as far back as eighth or ninth after which axis smooth mesially, ring furrows existing as paired
apodemal slots. Sharply pointed, convex (tr.) terminal piece merges anteriorly with smooth lateral borders of
axis. Four pleural ribs continuous with first four axial rings. First pair curve abaxially to mid-length of pygidium,
then gently adaxially, those behind increasingly curved until fourth pair lie sub-parallel to axial furrows. Ribs
separated by narrow, depressed anterior pleural bands, terminate in short, free, blunt points arranged en echelon
with tips of second pair lying level with axial tip.

Discussion. The form described here is closely similar in over-all form to specimens of A. scabra Dean
recently described by Ingham (1974, p. 79, pi. 16, figs. 2-14; text-fig. 24) from the Pusgillian and low
Cautleyan Stages of the Cautley Mudstones but differs in a few features. In A. scabra the 2p lateral
glabellar lobes are noticeably shorter (exsag.) than the 3p. In Sholeshook material this does appear to
be the case in a few cranidia from the basal 14 or 1 5 m of the formation around Llandowror and the
basal 2 or 3 m of Sholeshook railway cutting, and these forms have been herein tentatively referred to
A. scabra (see PI. 1 10, fig. 17). In all other specimens the 2p and 3p lobes are of sub-equal length, the
2p in some cases being slightly longer. The glabellar tuberculation in the south Welsh form differs
from that of A. scabra in that the paired tubercles are relatively less prominent, the others larger and
more numerous. On the pygidium the axial ring furrows appear to be mesially continuous further
posteriorly than in A. scabra and the pointed ends of the pleurae do not reach so far posteriorly; only
the third and fourth spines project beyond the axial tip, the first pair terminate well in front.

The Sholeshook cranidia are more like those referred by Ingham (1974) to Atractopyge sp. from
Cautleyan Zone 4, though here the 3p lateral glabellar lobes are implied to be consistently shorter
than the 2p. The second and third lateral glabellar lobes are approximately the same length in the
holotype cranidium of A. verrucosa (see Dean 1974, text-fig. 4) but the specimen is much larger than
the Sholeshook specimens and the glabella appears to have a relatively much wider (tr.) median lobe
and consequently a less clavate outline. A. verrucosa is known only from the holotype and topotype
specimens are needed before the species can be closely compared with other forms. The form from the

862

PALAEONTOLOGY, VOLUME 23

Birdshill Limestone termed A. cf. verrucosa by Dean (1974, p. 97; 1971, pis. 14, 15) has a less clavate,
less convex (tr.) glabella than the Sholeshook form and on the pygidium the pleural spines appear to
extend much further posteriorly (Dean 1971, pi. 14, figs. 8, 9).

Subfamily dindymeninae Pribyl, 1953
Genus dindymene Hawle and Corda, 1847

Type species. Dindymene fidericiaugusti Hawle and Corda, 1847.

Dindymene longicaudata Kielan, 1960
Plate 112, figs. 2-5

1973a Dindymene longicaudata Kielan; Price, tables 1, 2, 7.

19736 Dindymene longicaudata Kielan; Price, p. 538.

Holotype. Figured Kielan 1960, pi. 30, fig. 2; IG No. 2. II. 108, almost complete exoskeleton from Staurocephalus
clavifrons Zone of Brzezinki, Poland.

Material. BM In54166a, b, internal and external moulds of distorted cranidium; In54161, 54164, GSM H.T. 913,
internal moulds of incomplete pygidia, all from the basal Slade and Redhill Mudstones of Rudbaxton (locality
4); SM A77566a, b, internal and external moulds of pygidium from Sholeshook Limestone, locality 9e,
Sholeshook.

Description. Cranidium in dorsal view broader (tr.) than long (sag.); strongly convex (tr. and sag.). Occipital ring
broad and convex (sag. and exsag.), abaxially curving forwards around deep pits at ends of broad, shallow
occipital furrow. Posterior borders adaxially strongly convex (exsag.), outwards gradually narrowing to genal
angles but there broaden considerably and curve forwards; border furrows deep and slot-like adaxially. Axial
furrows deep and broad posteriorly where converge forwards slightly. Within them a second apodemal pit,
forward of the occipital pit, represents lp lateral furrow. Short lp lobe developed on median lobe between two

EXPLANATION OF PLATE 1 1 1

Figs. 1-4. Atractopyge aff. scabra Dean. 1, GSM 24546, internal mould of incomplete cranidium from
Sholeshook Limestone of Sholeshook, dorsal view, x2. 2, SM A3 1458, cast from external mould of
incomplete pygidium, Sholeshook Limestone of Sholeshook, dorsal view, x 3. 3, SM A53005b, cast from
external mould of cranidium from the high Sholeshook Limestone, locality 9h, Sholeshook, dorsal view, x 3.
4, GSM 24543, cast from external mould of cranidium from Sholeshook Limestone of Sholeshook, dorsal
view, x 3; together with counterpart internal mould GSM 24545, original of Salter 1848, pi. 8, fig. 9, 9b.

Fig. 5. Cybeloides ( Paracybeloides ) girvanensis (Reed). BM It9251, cast from external mould of incomplete
pygidium from high Sholeshook Limestone of Prendergast (locality 8c), dorsal view, x 6.

Figs. 6, 7. Prionocheilus cf. obtusus (McCoy). 6, SM A 104837, distorted internal mould of cranidium from basal
Sholeshook Limestone of Moldin (locality 25), near Llandowror, dorsal view, x 4. 7, SM A77943, internal
mould of incomplete cranidium from high Sholeshook Limestone, locality 9h, Sholeshook, right lateral view,
x 4; see also PI. 1 10, fig. 1 .

Figs. 8-11. Pseudosphaerexochus juvenis (Salter). 8, SM A3 141 7, internal mould of cranidium from Sholeshook
Limestone of Sholeshook railway cutting, dorsal view, x 2. 9, 1 0, SM A3 1432, internal mould of incomplete
cranidium from the high Sholeshook Limestone of Prendergast, dorsal and right lateral views, x 2. 11, SM
A77570, part of cast from external mould of partial cranidium from the high Sholeshook Limestone, locality
9b, Sholeshook, oblique view to show surface ornament, x 5.

Figs. 12-14. Calymene (s.l.) cf. prolata Ingham. 12, HM A9767, internal mould of incomplete cranidium from
the low Sholeshook Limestone (locality 18b) of Craig-y-deilo quarry, Llandowror, dorsal view, x 2. 13, 14,
GSM TJ. 843, internal mould of incomplete cranidium from about 24 m above the base of the Slade and
Redhill Mudstones at Robeston Wathen (locality 10c), dorsal and left lateral views, x 3.

PLATE 111

PRICE, Sholeshook trilobites

864

PALAEONTOLOGY, VOLUME 23

pits. Axial furrows narrow anteriorly as diverge round strongly convex (tr.) frontal lobe. Quadrant-shaped fixed
cheeks strongly convex, dropping steeply to posterior parts of axial furrows and posterior border furrows and
more gently anteriorly and laterally to broad, shallow border furrow. Base of right genal spine just visible on
available cranidium curving gently anteriorly. Cast from external mould (PI. 6, fig. 4) shows genal surfaces
closely pitted and bearing scattered granules, glabella with even pattern of well-spaced small granules and
scattered larger ones. Apically glabella bears prominent stout spine.

Triangular pygidial axis comprises eleven rings, first strongly convex (sag. and exsag.), those behind gradually
less so. Ring furrows deepest towards abaxial ends, leaving, behind second ring, shallow mesial portions of about
one-third axial width and narrow lateral borders; becoming fainter posteriorly so that rings behind fourth not
clearly separated in mesial region. Three pairs of pleural ribs. First two continuous with first two axial rings then
deflected posteriorly in smooth curves, separated from each other and from axis by strong furrows. Third
segments pressed close to sides of axis, separated from it posteriorly by two short, shallow furrows meeting in
acute-angled V. Pleurae terminate as stout, free spines; tips of inner two form straight line transverse to axis, first
terminate in front of this.

Discussion. The over-all similarity of the cranidium and the very close similarity of the pygidium to
those of D. longicaudata described by Kielan (1960, p. 153, pi. 26, fig. 5; pi. 28, fig. 5; pi. 29, fig. 4; pi.
30, figs. 1-3; text-fig. 43) from the Ashgill Series of Poland, Bornholm, Scania, and Vastergotland,
leave little doubt as to the specific identity. As Kielan shows in her table 6 (opposite p. 148), the
pygidia of known species of Dindymene are quite distinctive. Also distinctive of D. longicaudata are
the stout glabellar spine and the forwardly directed genal spines.

EXPLANATION OF PLATE 112

Fig. 1. Prionocheilus cf. obtusus (McCoy), SM A77943, internal mould of incomplete cranidium from high
Sholeshook Limestone, locality 9h, Sholeshook, dorsal view, x 4; see also PI. 109, fig. 7.

Figs. 2-5. Dindymene longicaudata Kielan. 2-4, BM In54166a, b, internal mould of distorted cranidium from
basal Slade and Redhill Mudstones of Rudbaxton quarry (locality 4) in left lateral and dorsal views and cast
from partial external mould in left lateral view, all x6. 5, SM A77566b, cast from external mould of
pygidium from middle Sholeshook Limestone, locality 9e, Sholeshook, dorsal view, x 6.

Figs. 6-8. Brongniartella cf. marocana Destombes. 6, SM A31174, internal mould of distorted, incomplete
cephalon from Slade and Redhill Mudstones of Redhill quarry (locality 7), dorsal view, x f . 7, SM A 1 04836,
internal mould of cranidium showing traces of lateral glabellar furrows, same horizon and locality, dorsal
view, x 1 . 8, SM A3 1172, internal mould of pygidium and partial thorax, same horizon and locality, dorsal
view, x 1.

Figs. 9-11. Duftonia geniculata Ingham? 9, GSM Pg. 134, cast from external mould of articulated thorax and
pygidium from the high Sholeshook Limestone of Prendergast (locality 8b or 8c), dorsal view, x 3. 10, BM
In54163b, cast from external mould of small, incomplete cranidium from basal Slade and Redhill Mudstones
south-west of Knock (locality 3), dorsal view, x 8. 11, BM In54162b, cast from external mould of small,
almost complete cranidium, same horizon and locality, dorsal view, x 8.

Figs. 12, 13. Liocnemis recurvus (Linnarsson). 12, SM A77634, internal mould of incomplete cranidium from
basal Slade and Redhill Mudstones south-west of Knock (locality 3), dorsal view, x 8. 13, BM In54224a,
internal mould of distorted incomplete cranidium from same horizon and locality, dorsal view, x 6.

Figs. 14, 15. Calyptaulax planiformis Dean. 14, SM A775 18, internal mould of incomplete cranidium from the
middle Sholeshook Limestone, locality 9e, Sholeshook, dorsal view, x 3. 15, GSM JM. 454, internal mould
of incomplete pygidium from about 30 m above base of Slade and Redhill Mudstones at Cilrath Fawr, 3 km
east-north-east of Robeston Wathen, dorsal view, x 4.

Figs. 16, 17. Toxochasmops marri (Reed), NMW.21.306.G.20a, internal mould of incomplete cranidium from
the Sholeshook Limestone of Craig-y-deilo quarry, Llandowror, antero-dorsal oblique view and dorsal view,
x li.

PLATE 112

11 15

PRICE, Sholeshook trilobites

866

PALAEONTOLOGY, VOLUME 23

Family staurocephalidae Prantl and Pribyl, 1948
Genus staurocephalus Barrande, 1 846

Type species. Staurocephalus murchisoni Barrande, 1 846.

Staurocephalus cf. clavifrons Angelin, 1854
Plate 110, fig. 18

1885 Staurocephalus globiceps, Portl.; Marr and Roberts, p. 481.

1973a Staurocephalus clavifrons Angelin; Price, tables 1-3, 7, p. 245.

Material, horizons, and localities. Not common but appears to range through Sholeshook Limestone Formation
(see Table 1). Known also from basal Slade and Redhill Mudstones near Pelcomb Cross (locality 2) but not from
elsewhere in that formation. Total of fifteen cranidia; thorax, pygidium, free-cheek, hypostoma, and rostral
plate not yet known.

Description. Median lobe of glabella tapers slightly forwards bounded by very broad, deep axial furrows and
indented laterally by large, shallow pits representing lp and 2p lateral furrows of which 2p are distinctly larger.
Pits representing 3p furrows small, shallow, and indistinct, situated near ends of broad, smooth furrow
separating hemispherical frontal glabellar lobe from median lobe. Lateral glabellar lobes of sub-equal exsagittal
length adaxially; basal lobes longer transversely, expanding (exsag.) slightly abaxially. Behind are small, deep,
round apodemal pits at ends of occipital furrow; furrow straight and shallow mesially. Occipital ring mesially
broad (sag. and exsag.), narrowing and curving forwards distally. Fixed cheeks strongly convex, standing higher
than median lobe and dropping to axial furrows in steep, convex slopes; apically bearing prominent palpebral
lobes surrounded by shallow furrows, their mid-lengths on the level of the 2p lateral lobes. Posterior border
furrows narrow. Posterior borders narrow and strongly convex adaxially but broadening considerably towards
genal angle. Cranidial surface tuberculated but no external moulds are available to show form of ornamentation
in detail.

Discussion. On the basis of the cranidium alone the South Welsh form cannot be distinguished either
from the holotype cranidium of S. clavifrons (Angelin 1854, pi. 24, fig. 8; Kielan 1957, pi. 4, fig. 1) or
from other material referred to the species by Kielan (1957, p. 163) or subsequently by Whittington
(1965, p. 53) or Dean (1971, p. 40). Following Ingham’s remarks (1977, p. 89) on differences between
the Polish specimens included in the species by Kielan and specimens from the Cystoid Limestone of
the Cautley area and on the occurrence of a further potentially distinguishable form from the
Swindale Limestone of Cross Fell (Ingham’s pi. 19, figs. 8-10), it appears that S. clavifrons as
previously recognized may be capable of further subdivision. The south Welsh form is therefore only
compared with Angelin’s species.

Family calymenidae Milne Edwards, 1840
Subfamily calymeninae Milne Edwards, 1840
Genus calymene Brongniart, 1822

Type species. Calymene blumenbachii Brongniart, 1822.

Calymene (s.l.) cf. prolata Ingham, 1977
Plate 111, figs. 12-14

1914 Calymene blumenbachi Brongn., var. caractaci Salter; Strahan et al. (pars), list p. 67.

1914 Calymene sp.; Strahan et al. (pars), table p. 70.

1973a Diacalymene cf. marginata Shirley; Price, p. 234.

Material, horizons, and localities. Three internal moulds of incomplete cranidia, GSM J.M. 396, GSM T.J. 483,
and HM A9767, respectively from about 18 m above the base of the Slade and Redhill Mudstones at Cilrath-
fawr, 2-25 km north-north-east of Narberth, about 24 m above the base of the same formation at Robeston
Wathen (locality 10c), and from a low horizon in the Sholeshook Limestone of Craig-y-deilo quarry,
Llandowror (locality 1 8b).

PRICE: LATE ORDOVICIAN TRILOBITES

867

Discussion. The available cranidia show a bell-shaped glabella with very prominent, convex (tr. and
exsag.) Ip lateral lobes, relatively large, rounded and convex 2p lobes with genal buttresses opposite
and small 3p lobes. The frontal lobe occupies just less than one-third the pre-occipital glabellar
length, is twice to 2\ times as wide as long, broadly rounded frontally, and extends well beyond the
fixed cheeks which are angulated antero-mesially. The pre-glabellar area is strongly upturned,
unridged and roll-like in cross-section, longer (exsag.) opposite the axial furrows, and separated from
the frontal lobe by a deep, slot-like furrow. These characteristics are shared with C. prolata Ingham
(1977, pp. 102-103, pi. 22, figs. 11-17) from Zone 3 of the Cautley Mudstones of northern England
but the south Welsh specimens appear to differ slightly from the illustrated cranidia of C. prolata in
possessing basal lateral lobes which are more quadrate in outline and in having rather less obviously
bifurcating lp lateral furrows.

Subfamily flexicalymeninae Siveter, 1977
Genus flexicalymene Shirley, 1936

Type species. Calymene blumenbachii var. caractaci Salter, 1865.

Flexicalymene cavei Price, 1974

1973 a Flexicalymene sp. nov.; Price, tables 1-4.

1974 Flexicalymene cavei Price, pp. 852-856, pi. 1 14, figs. 1-15.

Holotype. Figured Price 1974, pi. 114, figs. 1, 2; SM A57050, internal mould of cranidium from the basal
Sholeshook Limestone of Moldin (locality 25), near Llandowror.

Horizons and localities as in Table 1. Not known from Slade and Redhill Mudstones other than where ranges up
from underlying Sholeshook Limestone at localities 8a and 15.

Discussion. The form has been fully described elsewhere (Price 1974). Siveter (1977, p. 355) has
referred to the similarity between F. cavei and F. declinata (Hawle and Corda 1847) from the Kraluv
Dvur Formation of Bohemia. This similarity is not apparent from Barrande’s figures (1852, pi. 43,
figs. 53-58) but Dr. Siveter has kindly supplied the author with photographs of Bohemian specimens
of F. declinata, including the lectotype (selected Marek in Horny and Bastl 1970, p. 114). Though
these show that the two forms are closely related, there do appear to be differences. The glabella of the
Bohemian form is relatively rather broader (tr.) and shorter (sag.), with the frontal lobe, in particular,
shorter (sag. and exsag.) and less broadly rounded anteriorly. The 3p lateral glabellar lobes appear
longer (tr.) and separated from the frontal lobe by more strongly developed 3p furrows, and the lp
lateral furrows curve adaxially at their inner ends to give the median lobe a distinctive, postero-
laterally convex outline not seen in South Welsh specimens. In addition, to judge from Barrande’s
figures (1852, pi. 43, figs. 57, 58), the hypostoma may lack the prominent maculae seen on that of F.
cavei (Price 1974, pi. 1 14, fig. 8).

Subfamily pharostomatinae Hupe, 1953
Genus prionocheilus Rouault, 1 847

Type species. Prionocheilus verneuili Rouault, 1847.

Remarks. Both Siveter (1977, pp. 339, 393) and Ingham (1977, p. 103) have recently reviewed the
difficulties surrounding the choice between Pharostoma Hawle and Corda and its senior synonym
Prionocheilus. The balance of recent usage appears to be in favour of Prionocheilus.

Prionocheilus cf. obtusus (McCoy, 1846)

Plate 111, figs. 6-7; Plate 112, fig. 1
1973a Pharostoma cf. obtusum (M’Coy); Price, tables 1-4.

868

PALAEONTOLOGY, VOLUME 23

Material , horizons, and localities. Ten partial or incomplete cranidia from the following horizons and localities:
the basal Sholeshook Limestone at Moldin (locality 25) and in the Mylet road section (24a); the low Sholeshook
Limestone of Craig-y-deilo quarry (18c); the railway cutting, locality 9e and locality 9h at Sholeshook; the
Sholeshook Limestone at Lan-y-gaer (16b); and the highest Sholeshook Limestone at Prendergast (locality 8b).

Discussion. Cranidia from the Sholeshook Limestone are similar in over-all form and proportions
and in most details of morphology to the holotype of P. obtusus (McCoy) redescribed by Whittington
(1965, pp. 55-56, pi. 16, figs. 1-3, 6) from the Chair of Kildare Limestone and refigured, together with
topotype cranidia by Dean (1971, pi. 18, figs. 6, 8, 10, 12, 13, 15). They appear to differ, however, in
that the basal lateral glabellar lobes in several Sholeshook specimens are distinctly sub-quadrate in
outline (e.g. PI. 112, fig. 1) and in that small 3p lateral glabellar lobes are clearly visible in most
Sholeshook specimens. In the outline of the lp lateral glabellar lobes the South Welsh specimens
resemble P. cautleyensis Ingham (1977, pp. 104-105, pi. 22, figs. 19-23) from the Cautleyan Stage of
the Cautley Mudstones, but that form has relatively much wider (tr.) lp lateral lobes, a narrower (tr.)
frontal glabellar lobe, more strongly developed subsidiary lobes between the lp and 2p lateral lobes,
and a mesially very short (sag. and exsag.) pre-glabellar field. The pre-glabellar field in the
Sholeshook specimens appears to be of similar length and convexity to that of P. obtusus.

Family homalonotidae Chapman, 1890
Genus brongniartella Reed, 1918

Type species. Homalonotus bisulcatus McCoy, 1851.

Brongniartella cf. marocana Destombes, 1 966
Plate 11 2, figs. 6-8

1885 Homalonotus ?; Marr and Roberts, p. 480.

1885 Homalonotus bisulcatus. Salt.; Marr and Roberts, p. 482.

1914 Homalonotus rudisl Salt.; Strahan et al., table p. 74.

1973a Brongniartella sp.; Price, pp. 229-230, 242, tables 1, 2.

Material, horizons, and localities. One partial pygidium from the highest Sholeshook Limestone of Prendergast
(locality 8b); 1 pygidium from the basal Slade and Redhill Mudstones of Prendergast (8a); 1 partial pygidium, 2
pygidia with partial thoraxes, 2 incomplete cranidia, and 1 incomplete cephalon all from the Slade and Redhill
Mudstones of Redhill quarry (locality 7); all internal moulds. ?Other fragmental material from localities 7, 8a, b,
c, and 9d.

Description. Cephalon sub-semicircular. Weakly convex glabella trapezoid in outline, its lateral margins gently
convex posteriorly and concave near the mid-length; width (tr.) just in front of occipital furrow slightly less than
pre-occipital length and twice width at anterior margin. One specimen (PI. 112, fig. 7) faintly shows lateral
glabellar furrows; lp furrows commence abaxially at one- third pre-occipital glabellar length and curve inwards
for about one-quarter of glabellar width at that level, 2p furrows commence abaxially at half pre-occipital length
and curve sigmoidally inwards and back, 3p furrows very faint. Occipital furrow narrow but distinct, sinuous,
curved forward mesially and abaxially. Occipital ring occupies one-seventh glabellar length (sag.). Axial furrows
shallow, very broad posteriorly but narrow forwards. Anterior border furrow broad and shallow, anterior
border only weakly convex (sag. and exsag.). Gently convex (tr.) anterior portions of fixed cheeks each about
half anterior width of glabella. Short (exsag.) palpebral lobes on level of cranidial mid-length. Posterior border
furrows broad and shallow, borders flat (exsag.), broader (exsag.) than occipital ring. Free cheeks with indistinct
borders and border furrows.

Thoracic axis occupies over one-third total width, defined by shallow axial furrows; rings separated from half-
rings by broad, strong, articulating furrows. Pleurae bear deeply incised, curved pleural furrows. Pygidium sub-
parabolic, moderately convex (tr.). Axis anteriorly occupying less than one-third total width, defined by broad,
deep axial furrows and tapering gradually to a well-defined, bluntly rounded distal end which almost(?) reaches
posterior margin. Axis composed of nine broad, flat axial rings and narrow articulating half-ring. Pleurae show
seven abaxially broadening ribs and outwardly and backwardly curving inter-rib furrows which do not reach
lateral margins.

PRICE: LATE ORDOVICIAN TRILOBITES

869

Discussion. B. marocana Destombes (1966, p. 34, pi. 1, figs. 1-8) from the Upper Ktaoua Formation
of the Moroccan Anti-Atlas is closely similar to the south Welsh form in glabellar shape and agrees
in showing faint lateral glabellar furrows and in the posteriorly broad axial furrows and the position
of the palpebral lobes. The pygidium is similar in over-all form but the axial furrows, ring furrows,
and pleural furrows are weaker than in the south Welsh material and the axis is less well developed
posteriorly. In this latter respect the South Welsh pygidia are closer to those of the form from Zone 5
of the Rawtheyan Stage of the Cautley Mudstones referred with question by Ingham (1977, p. 109, pi.
24, figs. 1 -4) to B. robusta (Lesperance) in which the axial tip is slightly swollen. Neither that species,
however, nor the type material of B. robusta (Lesperance 1968, p. 822, pi. 106, figs. 8-13) from the
‘Upper Ashgill’ part of the Whitehead Formation of Perce, Quebec, show any sign of glabellar
furrows and both forms are said to have only eight rings on the pygidial axis. In B. platynotus
(Dalman) from the late Ashgill of Poland, Sweden, and Czechoslovakia (Kielan 1960, p. 116, pi. 19,
figs. 1-3) the glabella narrows more sharply anteriorly than in the south Welsh form and the eyes are
much further forward. As Ingham has noted (1977, p. 1 10), B. marocana is more closely related to
forms such as B. sedgwicki (Salter) and B. robusta than it is to B. platynotus.

Family dalmanitidae Vogdes, 1890
Subfamily dalmanitinae Destombes, 1972
Genus duftonia Dean, 1959

Type species. Duftonia lacunosa Dean, 1959.

Duftonia geniculata Ingham, 1977?

Plate 112, figs. 9-11

1973a Duftonia cf. lacunosa Dean; Price, tables 2, 7.

71977 Duftonia geniculata Ingham, p. 1 14, pi. 26, figs. 12-19; text-fig. 28 b.

Holotype. Figured Ingham 1977, pi. 26, figs. 12, 13; HM A5540a, b, internal and external moulds of damaged
cranidium from the mid-Rawtheyan Swindale Limestone of Cross Fell.

Material , horizons , and localities. BM In54162a, b, 54163a, b, internal and external moulds of incomplete
cranidia, In54164, internal mould of pygidium, all from basal Slade and Redhill Mudstones near Pelcomb, 4 km
north-west of Haverfordwest (locality 3); GSM Pg. 133, 134, internal and external moulds of articulated thorax
and pygidium and Pg. 123, internal mould of pygidium, from the high Sholeshook Limestone of Prendergast
(locality 8b or 8c).

Description. Two available cranidia small and poorly preserved. Both show a rather weakly developed
geniculation in the course of the 3p lateral glabellar furrows. Outer margins of 3p lateral lobes not independently
convex and postero-laterally do not project further into axial furrows than antero-lateral corners of 2p lobes.
Palpebral lobes relatively short (exsag.), extend back to level opposite anterior parts of 2p lateral lobes.
Preservation too poor to show glabellar ornamentation. Thorax of eleven segments, tapering gradually
posteriorly. Axis strongly convex, occupies over one-third total width (tr.) anteriorly. On cast of external mould
(PI. 112, fig. 9) axial rings sub-rectangular in dorsal outline, arched forward mesially and again curving gently
forwards and slightly broadening (exsag.) abaxially. Articulating furrows broad and shallow mesially but drop
abaxially into deep apodemal slots. On internal moulds these slots separate prominent, rounded axial lobes.
Axial furrows narrow and rather weak. Pleurae flat-lying over inner portions but strongly deflected ventrally at
fulcrum; divided by strong pleural furrows into broad posterior and narrow anterior convex pleural bands;
becoming much flatter towards broad, rounded abaxial extremities. Thoracic surface appears to be finely
granulated. Pygidial axis with four well-defined rings anteriorly and indications of two more behind. Axial
furrows die out at less than two-thirds pygidial length from anterior margin. Pleurae crossed by three broad
pleural ribs divided by strong pleural furrows and defined by broad interpleural furrows which extend further
laterally, though neither set reaches the lateral margins.

Discussion. The small size and poor preservation of the south Welsh cranidia preclude complete
comparison, but in all their visible features they appear close to those of D. geniculata as described by

870

PALAEONTOLOGY, VOLUME 23

Ingham (1977) from the mid-Rawtheyan Swindale Limestone of Cross Fell and from Zones 5 and 6 of
the Cautley Mudstones, as do the south Welsh pygidia. No thorax is known for D. geniculata.

Family pterygometopidae Reed, 19056
Subfamily pterygometopinae Reed, 19056
Genus liocnemis Kielan, 1960

Type species. Phacops recurvus Linnarsson, 1869.

Liocnemis recurvus (Linnarsson, 1869)

Plate 112, figs. 12, 13; Plate 113, fig. 16
1869 Phacops recurvus Linnarsson; p. 59, pi. 1, figs. 1, 2.

1885 Phacops Brongniarti, Portl.; Marr and Roberts (pars), p. 481 (lowest two faunal lists).

1960 Liocnemis recurvus (Linnarsson); Kielan, pp. 121-123, pi. 9, figs. 11, 12; pi. 21, figs. 8-11; pi. 22,
figs. 1, 2; text-fig. 32.

1973a Liocnemis cf. recurvus (Linnarsson); Price, table 7.

Type specimens. The original specimens figured by Linnarsson (1869) have not been located (see Kielan 1960,
p. 123).

Material, horizons, and localities. BM In54220a, b, 54224a, b, internal and external moulds of distorted,
incomplete cranidia; SM A3 1543, 31544, 31546, 77634, 77935, internal moulds of incomplete cranidia, mostly
distorted, and SM A77627, internal mould of pygidium; all from the basal Slade and Redhill Mudstones near
Pelcomb (localities 2, 3) 4 km north-west of Haverfordwest. SM A3 1 545, the fragmentary external mould of a
cranidium from the same horizon at Clarbeston Road Station (locality 6b) may also belong here.

Discussion. Kielan (1960) has provided a full description of this species which does not require
repetition. Distortion of the south Welsh cranidia makes it difficult to establish the exact
proportional length of the frontal glabellar lobe which on most specimens is strongly bent-down; it
does appear, however, to be slightly longer than the rest of the glabella as in the Swedish and Polish
material of L. recurvus figured by Kielan (1960).

Genus calyptaulax Cooper, 1930
Type species. Calyptaulax glabella Cooper, 1930.

Calyptaulax planiformis Dean, 1962
Plate 112, figs. 14, 15

1885 Phacops Brongniarti, Portl.; Marr and Roberts (pars), pp. 480, 482.

1962 Calyptaulax planiformis Dean, p. 98, pi. 13, figs. 1-5.

1973a Calyptaulax planiformis Dean; Price, p. 233, tables 1-4.

1975 Calyptaulax sp.; Cocks and Price, list p. 705, pi. 81, fig. 6.

Holotype. Figured Dean 1962, pi. 13, fig. 4, BM In50138, internal mould of cranidium from the Pusgillian Stage,
Swindale Beck, Cross Fell.

Horizons and localities. As in Table 1. Appears also to range through the Slade and Redhill Mudstone
Formation.

Discussion. C. norvegicus Stormer (1945, p. 417, pi. 4, figs. 2, 3) from the Gagnum Shale of Hadeland
is closely similar to C. planiformis. Dr. A. Owen informs me that in Stormer’s illustration the holotype
cranidium was tilted forward slightly thus foreshortening the frontal lobe. There thus appear to be no
important differences in cranidial proportions between the two forms and the main distinction rests
on the pygidial differences referred to by Dean (1962, p. 99). In this respect south Welsh pygidia are
like the holotype and paratype pygidia of C. planiformis, with a relatively long axis and at least the

PRICE: LATE ORDOVICIAN TRILOBITES

871

first three interpleural furrows reaching the pygidial margin (PI. 6, fig. 15). The form described by
Whittington (1962, p. 12, pi. 2, figs. 17, 18; pi. 3, figs. 15, 16) as C. aflf. norvegicus from the Rhiwlas
Limestone of North Wales differs from the south Welsh form in that the 3p lateral glabellar lobes are
narrower (tr.) anteriorly and the 3p furrows strongly geniculated and here also, as in the pygidium of
C. norvegicus figured by Stormer, the pygidial border appears to be smooth. It may be of significance
that both Whittington and Stormer refer in their descriptions to circular or sub-circular lp lateral
glabellar lobes whereas in the south Welsh specimens, and apparently in those figured by Dean ( 1 962, pi.
13, figs. 1-3), the outline is distinctly sub-quadrilateral and angular. The validity of the differences re-
ferred to here between C. planiformis and C. norvegicus must remain uncertain until more material of
the latter is illustrated. For the present the South Welsh specimens are best referred to C. planiformis.

Family lichidae Hawle and Corda, 1 847
Subfamily homolichinae Phleger, 1936
Genus platylichas Giirich, 1 90 1

Type species. Lichas margaritifera Nieszkowski, 1857.

Platylichas noctua sp. nov.

Plate 113, figs. 1-9; Plate 1 14, fig. 7.

1848 Lichas laxatus, McCoy; Salter (pars), p. 340, pi. 8, figs. 4, 4 a ( non 5, 6).

1866 Lichas laxatus, M’Coy; Salter (pars), p. 324, pi. 19, fig. 1 ( non 2, 3).

1885 Lichas laxatus, M’Coy; Marr and Roberts, lists pp. 480, 481.

1909 Lichas laxatus McCoy; Strahan et a/., table p. 58.

1914 Lichas laxatus McCoy; Strahan et al., table p. 63.

1973a Platylichas cf. laxatus M’Coy; Price, tables 1-4, 7, list p. 242.

Holotype. (PI. 113, figs. 4-6), GSM 19475, internal mould of incomplete cranidium (?together with GSM 19479,
original of Salter 1 848, pi. 8, fig. 4), Sholeshook Limestone of Sholeshook.

Diagnosis. Species of Platylichas with very wide (tr.). D-shaped composite glabellar lobes, narrow median lobe
between and relatively wide (tr.) and short frontal lobe; palpebral lobes occupy up to four-fifths length (exsag.) of
composite lobes; hypostoma with anterior lobe of median body coarsely granulated and lateral borders bearing
a few raised ridges; pygidial border spines long, only gradually tapering, first two with convex outer margins.

Horizons and localities. In addition to occurrences shown in Table 1, known also from Slade and Redhill
Mudstones of Redhill Quarry (locality 7) and basal Slade and Redhill Mudstones near Pelcomb (locality 3) and
Rudbaxton (4).

Description. Width (tr.) of cranidium greatest posteriorly where about twice sagittal length. Glabella also widest
posteriorly, width across frontal lobe being only three-quarters that at occipital ring. Latter broadest (sag. and
exsag.) mesially, narrowing and curving forwards behind ovoid occipital lobes. Median lobe narrowest just
behind mid-length of composite lobes, occupying one-sixth glabellar width at that level. Composite lobes occupy
just over two-fifths cranidial length, very wide (tr.) and prominent, separated from rest of glabellar by deep,
strongly curved longitudinal furrows; stand slightly above median lobe (PI. 1 13, fig. 5), most of their surface
horizontal but dropping steeply antero-laterally in front of anterior ends of palpebral furrows. Frontal lobe 2j-3
times as wide (tr.) as long (sag.), broadly rounded frontally, occupying about one-quarter of cranidial length
(sag.) and dropping steeply anteriorly (PI. 113, fig. 6) to strong pre-glabellar furrow and broad, flat anterior
border. Border narrows laterally where crossed at low angle by anterior branches of facial sutures (PI. 113, fig. 4).
Shallow axial furrows diverge forwards at about 55°. Palpebral lobes broad, prominent, and strongly curved,
three-quarters to four-fifths length of composite glabellar lobes and standing on same level. Separated from rest
of fixed cheeks by poorly defined palpebral furrows which at their mid-lengths are exsagittally in line with
abaxial ends of occipital ring. Fixed cheeks moderately convex (exsag.) behind palpebral lobes, separated by
deep furrows from occipital lobes. Anterior branches of facial sutures at first diverge and then converge forwards
sub-linearly; posterior branches curve sigmoidally out and back to cross posterior border at angle of about 50°.
Cranidial surface ornamented with granules of two sizes (PI. 113, fig. 8), the larger about 0-15 mm and evenly

872 PALAEONTOLOGY, VOLUME 23

distributed, the space between filled by the smaller. Occipital ring bears prominent median tubercle near
posterior margin.

Hypostoma (PI. 113, figs. 2, 3; PI. 114, fig. 7) sub-quadrate, broadest on level of posterior border furrow.
Median body moderately convex (tr.), divided by short but strong median furrows. Anterior lobe large,
rounded. Posterior lobe short (sag. and exsag.), about two-thirds width (tr.) of anterior lobe. Anterior border
narrow. Small anterior wings sub-triangular, directed dorsally. Lateral notches broad (exsag.), shallow in side
view. Lateral and posterior border furrows deep, posterior border broad and moderately convex (sag. and
exsag.); posterior margin bifurcate with broad (tr.), shallow median notch. Anterior lobe of median body evenly
covered with large (0 03 mm) granules and lateral borders bear a few raised, anastomosing lines running sub-
parallel to lateral margins.

Thorax (PI. 1 13, fig. 9) incompletely known. Axis broad (tr.), tapering back only gradually, with broad (sag.
and exsag.), sub-rectangular axial rings. Axial furrows narrow but deep. Pleurae become narrower (exsag.)
abaxially and are deflected posteriorly at fulcrum as long, backwardly directed pleural spines. Pleural furrows
commence at axial furrows near anterior margin of each segment then curve gently outwards and towards mid-
line.

Axis of pygidium moderately convex (tr.), tapering only gradually back. Four narrow and convex (sag. and
exsag.) axial rings are separated by ring furrows which gradually shallow posteriorly until fourth is only
developed laterally and does not reach axial furrows. Behind it posterior portion of axis is more convex (tr.) and
drops steeply in line with anterior ends of third interpleural furrows. Post-axial ridge at first narrows posteriorly
more rapidly than axis but then expands again towards posterior border furrow. Axial furrows broad and deep.
Pleural lobes crossed by pleural and interpleural furrows of equal prominence; three pleural ribs. First two
developed into border spines which are long, taper only gradually, and curve backwards and inwards with evenly
convex lateral margins. Gently convex border and shallow border furrow only clearly developed behind
posterior band of second pleural rib (PI. 113, fig. 7). Third pair of backwardly directed, broad-based border
spines set close together behind axis. Doublure of thorax and pygidium broad, with widely spaced terrace-lines.
External surface of both granulated in same manner as cranidium though on smaller specimens the granulation
is relatively coarser.

Discussion. In common with many other upper Ordovician species of Platylichas, the Sholeshook
form was long identified with P. laxatus (McCoy 1 846, p. 5 1 , pi. 4, fig. 9), a species erected on the basis
of a partial cranidium from strata at Ballygarvan Bridge, New Ross, Eire, the exact age of which is
uncertain. Dean (1963, pi. 43, fig. 10) refigured this holotype and a more complete topotype

EXPLANATION OF PLATE 113

Figs. 1-9. Platylichas noctua sp. nov. 1, SM A3 1531, testate, incomplete cranidium from the Sholeshook
Limestone of Sholeshook railway cutting, dorsal view, x 4. 2, 3, SM A 104839, internal mould of hypostoma
from the basal Sholeshook Limestone of Moldin (locality 25) near Llandowror, ventral and right lateral views,
x 4. 4-6, GSM 19475, internal mould of incomplete cranidium, holotype (?together with GSM 19479,
original of Salter 1848, pi. 8, fig. 4), Sholeshook Limestone of Sholeshook, dorsal, anterior, and antero-lateral
oblique views, x 4. 7, SM A3 1 5 1 3, cast from external mould of incomplete pygidium from high Sholeshook
Limestone of Prendergast, dorsal view, x6. 8, GSM 19479, cast from external mould of incomplete
cranidium (?together with GSM 19475, original of Salter 1848, pi. 8, fig. 4), Sholeshook Limestone of
Sholeshook, oblique view, x 4. 9, SM A3 1530, testate internal mould of incomplete articulated thorax and
pygidium from Sholeshook Limestone of Sholeshook railway cutting, dorsal view, x 1|.

Figs. 10, 11. Platylichas angulatus Warburg? 10, BM It9261, cast from external mould of incomplete cranidium
from the Sholeshook Limestone horizon at Robeston Wathen, flattened, dorsal view, x 5. 11, BM It9263,
internal mould of incomplete hypostoma, same horizon and locality, ventral view, x 4.

Figs. 12-14. Trochurus sp. indet., GSM Pg. 291, internal mould of poorly preserved, incomplete cranidium from
the Sholeshook Limestone, ‘middle section’ of Sholeshook railway cutting, dorsal, right lateral, and anterior
views, x 5.

Fig. 15. Glaphurella cf. harknessi (Reed), GSM Pg. 299, internal mould of cranidium from same horizon and
locality as original of figs. 12-14, left lateral view, x 4. See also PI. 1 14, figs. 17, 18.

Fig. 16. Liocnemis recurvus (Linnarsson), SM A77 627, internal mould of pygidium from the basal Slade and
Redhill Mudstones south-west of Knock (locality 3), dorsal view, x 8.

PLATE 113

PRICE, Sholeshook trilobites

874

PALAEONTOLOGY, VOLUME 23

cranidium was figured by Tripp (1958, pi. 84, fig. 4). P. laxatus differs from P. noctua in having
narrower (tr.) composite glabellar lobes with a relatively wider median lobe between; the median lobe
is narrowest further posteriorly and expands much more gradually anteriorly into a relatively longer
(sag. and exsag.) and narrower (tr.) frontal lobe. Dean (1963, pp. 235-237, pi. 43, figs. 1,2, 5,8, 11, 12)
also assigned to P. laxatus material from the Actonian Stage of south Shropshire. The cranidia
appear to be similar to the Irish specimens; the hypostoma differs from that of P. noctua in that the
maximum width is attained further anteriorly, across the shoulders, and is comparable to that on the
level of the anterior wings, also the posterior lobe of the central body is smaller.

P. nodulosus (McCoy) from the Longvillian Stage of the Bala area (Whittington 1962, pp. 25-28,
pi. 6, figs. 12, 13; pi. 7, figs. 1-14, 19; 1968, pp. 100-101, pi. 31, figs. 5, 6, 8-11, 14) differs in having a
broader median glabellar lobe with the posterior part inflated as a distinct ring, in having shorter
palpebral lobes and in having a much wider (sag. and exsag.) anterior border. In P. glenos
Whittington ( 1 962, pp. 28-3 1 , pi. 7, figs. 1 5, 1 6; pi. 8) from the Rhiwlas Limestone of north W ales and
the Chair of Kildare Limestone of eastern Ireland (Dean 1974, pp. 81-83, pi. 33, fig. 12; pi. 36, figs.
3-5, 7, 9-1 1; pi. 37, figs. 1-3, 5, 7, 10; pi. 38, figs. 3, 4, 7, 11, 12) the palpebral lobes are long as in P.
noctua but the composite lobes are less wide and tend to be posteriorly pointed in outline, the frontal
lobe is relatively longer (sag. and exsag.), less broad, and with a more strongly convex anterior margin
and the median lobe does not expand so rapidly near its posterior margin. The hypostoma is much
more finely granulated and carries many more anastomosing ridges on the lateral borders
(Whittington 1962, pi. 8, fig. 10), the posterior lobe of the central body is narrower (tr.) and the
posterior border more strongly convex (sag. and exsag.). The pygidial border spines are shorter,
much more slender, and have concave lateral margins. The hypostoma of P. crescenticus (Reed, 1935,
pp. 29-3 1 , pi. 3, figs. 13-16) from the upper Drummuck Group of Girvan is similar to that of P. glenos
in having a finer granulation and more anastomosing ridges than that of P. noctua ; in addition the
posterior lobe of the median body is relatively smaller and separated by less prominent median
furrows and the posterior margin has a far narrower (tr.) median notch. The glabella has a broader
median lobe which expands more gradually posteriorly and relatively narrower composite and
frontal lobes. If Tripp (1958, p. 579) is correct in considering P. vicinus (Reed, 1935, p. 33, pi. 3, fig. 12)
to be a synonym of P. crescenticus then the pygidium of the latter form differs from that of P. noctua
in having shorter border spines with less strongly curved outer margins.

Platylichas angulatus Warburg, 1925?

Plate 1 13, figs. 10-11; Plate 1 14, fig. 6
1973a Platylichas sp. ?nov.; Price, list p. 233.

Material. BM It9261, 9262, 9263, SM A77810, respectively external mould of incomplete, flattened cranidium
and internal moulds of three incomplete hypostomata from the Sholeshook Limestone horizon at Robeston
Wathen.

Description. Glabella equally wide (tr.) across frontal lobe and occipital ring. Latter broadest (sag. and exsag.)
mesially, behind ovoid occipital lobes narrows and curves forwards. Median lobe broad (tr.), narrowest just
behind mid-length of palpebral lobes, expanding forwards at about 40°; moderately convex (tr.), standing
slightly higher than composite lobes; posterior portion developed as distinct convex (sag. and exsag.) ring almost
twice as wide as narrowest part of median lobe and separated by broad shallow furrow. Composite lobes large,
sub-triangular, moderately convex (tr.). Strongly curved longitudinal furrows become shallow and indistinct
antero-mesially and postero-mesially. Frontal lobe short and wide, sharply angulate laterally, with only
moderately convex anterior margin; separated by narrow, distinct furrow from narrow, convex (sag. and exsag.)
anterior border. Axial furrows deep and broad posteriorly. Palpebral lobes broad (tr.), strongly curved, about
half-length (exsag.) of composite glabellar lobes and situated opposite posterior halves of these. Surface of
cranidium with exception of anterior border ornamented with variably sized, irregularly spaced granules, the
largest, on the median lobe and posterior two-thirds of composite lobes, very prominent, attaining c. 0-03 mm;
granulation markedly finer anteriorly and antero-laterally. Occipital ring bears small, posteriorly placed median
tubercle.

Hypostoma broader (tr.) than long (sag.); maximum width attained at about level of posterior border furrow;

PRICE: LATE ORDOVICIAN TRILOBITES

875

anterior margin bluntly pointed. Median body sub-pentagonal in outline, only gently convex, divided by short
(tr.) but broad and deep middle furrows. Anterior lobe about twice as broad (tr.) as long (sag.), broadest at about
mid-length, anterior margin bluntly pointed, lateral margins straight and posteriorly convergent. Posterior lobe
short (sag. and exsag.). Anterior border absent. Anterior wings small. Lateral notch shallow (tr.). Lateral
borders broaden posteriorly until about level of posterior border furrow. Lateral border furrows deep, sub-
linear, posteriorly convergent; posterior border furrow shallower, transverse. Posterior border bifurcate with
broad median notch.

Discussion. P. angulatus was described by Warburg (1925, p. 286, pi. 7, figs. 28-30) on the basis of two
cranidia from the Boda Limestone of Kallholn, Dalarne, Sweden. Her original figures are too small
to permit close comparison with other forms. More recently Dean (1974, p. 83, pi. 37, figs. 4, 6, 8, 9;
pi. 38, figs. 1, 6) referred two cranidia from the Chair of Kildare Limestone to Warburg’s species.
Although very similar in over-all form, these two cranidia show some differences from the south
Welsh specimen. The median glabellar lobe does not narrow so markedly, the transverse posterior
portion is relatively longer (sag. and exsag.) and less wide (tr.) and apparently less well separated from
the anterior part and the granulation on the median and composite lobes is less coarse than on the
south Welsh specimen. Hypostomata have not been described for either the Boda Limestone or the
Chair of Kildare Limestone form. More certain identification of the south Welsh form must await
redescription of P. angulatus from Boda Limestone material.

Subfamily ceratarginae Tripp, 1957
Genus trochurus Beyrich, 1 845

Type species. Trochurus speciosus Beyrich, 1845.

Trochurus sp. indet.

Plate 113, figs. 12-14

1914 Lichas bulbiceps Reed ?Phillips MS.; Strahan et al., table p. 63.

1973<2 Trochurus sp. indet.; Price, table 2.

Material. GSM Pg. 29 1 , internal mould of incomplete cranidium from the middle section of Sholeshook railway
cutting.

Discussion. The poorly preserved, incomplete cranidium is similar in over-all form to the holotype
cranidium of T. toernquisti (Gurich) figured by Warburg (1925, pi. 7, figs. 1, 2) from the Boda
Limestone of Boda, Dalarne, Sweden, but differs in that the lp glabellar lobes appear to be relatively
longer (exsag.) and the bullar lobes narrower (tr.) and more triangular in dorsal view. Both Warburg
(1925, p. 259) and Dean (1974, pp. 87-88, pi. 35, figs. 2, 3, 5, 8, 1 1) referred specimens from the Chair
of Kildare Limestone of eastern Ireland to T. toernquisti. To judge from the Irish specimens figured
by Dean, the Sholeshook form has both the median glabellar lobe and the bullar lobes relatively
longer (sag. and exsag.) and narrower (tr.).

Family odontopleuridae Burmeister, 1843
Subfamily miraspidinae R. & E. Richter, 1917
Genus whittingtonia Prantl and Pribyl, 1949

Type species. Acidaspis bispinosa McCoy, 1846.

Whittingtonia whittingtoni Kielan, 1960
Plate 114, figs. 1-3

1960 Whittingtonia whittingtoni Kielan, pp. 109-111, pi. 16, fig. 5; pi. 18, figs. 1-4; text-fig. 28.

1965 Whittingtonia cf. whittingtoni Kielan; Whittington, pp. 34-35, pi. 9, figs. 1 1-17.

1968 Whittingtonia whittingtoni Kielan; Whittington, p. 100, pi. 31, figs. 1-3.

1973a Whittingtonia whittingtoni Kielan; Price, p. 245, table 7.

19736 Whittingtonia whittingtoni Kielan; Price, p. 538.

876

PALAEONTOLOGY, VOLUME 23

Material. In54227, internal mould of almost complete, slightly distorted cephalon, basal Slade and Redhill
Mudstones, road-section at crossways south-west of Knock (locality 3), 4 km north-west of Haverfordwest; SM
A3 1 364, internal mould of incomplete cephalon, same horizon near Pelcomb Cross (locality 2), 4 km west-north-
west of Haverfordwest.

Discussion. The prominent, strongly convex (tr. and sag.) fronto-median glabellar lobe is similar in
both dorsal and anterior views to those of specimens figured by Kielan (see synonomy) from the
upper Ordovician of Poland and by Whittington from the Rhiwlas Limestone of north Wales,
particularly to the latter. In lateral profile this lobe on one specimen (PI. 114, fig. 3) appears to be
rather more convex and to overhang the anterior border less, but these differences may well be due to
distortion. Abaxially the median lobe drops steeply to broad, deep axial furrows which contain only
weakly developed lp and 2p lateral lobes; in this respect the cephala are more like those described by
Whittington (1965, p. 34). In all other features, the short, stout occipital spines, the elongated (tr.)
occipital node, the large eyes, the broad, deep palpebral furrows and prominent eye-ridges, the form

text-fig. 1. Proceratocephala cf. terribilis (Reed,
1914); reconstruction of cephalon approximately
x 6-5, based largely on original of Plate 1 14, fig. 4.
Details of spinose margin to anterior and lateral
border (dotted) hypothetical.

EXPLANATION OF PLATE 114

Figs. 1-3. Whittingtonia whittingtoni Kielan, BM In54227, internal mould of almost complete, slightly distorted
cephalon from the basal Slade and Redhill Mudstones south-west of Knock (locality 3), dorsal, anterior, and
right lateral views, x 10.

Figs. 4, 5. Proceratocephala cf. terribilis (Reed). 4, GSM Pg. 487, internal mould of almost complete cephalon
from the high Sholeshook Limestone of locality 9h, Sholeshook, dorsal view, x 6. 5, SM A77582a, testate,
incomplete cranidium from the basal Sholeshook Limestone of the Pentre-howell road section (locality 17),
Llandowror, dorsal view, x 6.

Fig. 6. Platylichas angulatus Warburg?, BM It9262, internal mould of incomplete hypostoma from the
Sholeshook Limestone horizon at Robeston Wathen, ventral view, x 4.

Fig. 7. Platylichas noctua sp. nov., SM A3 1537b, cast from external mould of hypostoma, Sholeshook Limestone
of Sholeshook railway cutting, ventral view, x 4.

Figs. 8-15. Primaspis llandowrorensis sp. nov. 8, 9, HM A9633, internal mould of incomplete cranidium from
the high Sholeshook Limestone of Lan-y-gaer (locality 16a), near Llandowror, anterior and dorsal views, x 4.
10, NMW.21.306.G.17, external mould of pygidium from the Slade and Redhill Mudstones ofOld Pale, near
Llandowror, dorsal view, x 5. 11-14. GSM 21053/5213, holotype, cast from internal mould of posterior
part of thorax and pygidium with ventral mould of pygidial doublure, cast from partial external mould of
cephalon and thorax and internal mould (including counterpart to above) of incomplete articulated
exoskeleton (enrolled) from Sholeshook Limestone of Craig-y-deilo quarry, Llandowror, figs. 11-13 dorsal
views, fig. 1 4 antero-lateral oblique, all x 4. 1 5, BM It9266b, cast from external mould of left free cheek from
the Sholeshook Limestone horizon at Robeston Wathen, dorsal view, x 8.

Fig. 16. Primaspis sp. indet., SM A77514b, cast from external mould of incomplete pygidium from the highest
Sholeshook Limestone of Prendergast (locality 8b), dorsal view, x 5.

Figs. 1 7, 1 8. Glaphurella cf. harknessi (Reed), GSM Pg. 296/299, cast from partial external mould of cranidium in
dorsal view and counterpart internal mould of cranidium in dorsal view, Sholeshook Limestone, ‘middle
section’ of Sholeshook railway cutting, both x 4; see also PI. 113, fig. 15.

PLATE 114

PRICE, Sholeshook trilobites

PALAEONTOLOGY, VOLUME 23

of the cheeks and surface ornamentation, the specimens resemble closely the material described by
both Kielan and Whittington. Kielan (1960, p. 1 1 1), Bruton (1966, p. 28), and Dean (1974, p. 94) have
discussed the differences between W. whittingtoni and the type species, W. bispinosa, from the Chair
of Kildare Limestone.

Genus proceratocephala Prantl and Pribyl, 1949
Type species. Acidaspis terribilis Reed, 1914.

Proceratocephala cf. terribilis { Reed, 1914)

Plate 1 14, figs. 4, 5; text-fig. 1

1914 Acidaspis sp.; Strahan et al., table p. 63.

1973a Proceratocephala cf. terribilis (Reed); Price, tables 1-4.

Material, horizons, and localities. GSM Pg. 487, internal mould of almost complete cephalon, high Sholeshook
Limestone, Sholeshook (9h); HM A9702, A9704, internal moulds of cranidia, low Sholeshook Limestone, track
south of Craig-y-deilo quarry, Llandowror (18d); SM A77582a, b, A77754a, b, counterpart moulds of cranidia,
basal Sholeshook Limestone, Pentre-howell road section (17); SM A77989, internal mould of incomplete
cranidium 9^-10 m above base of Sholeshook Limestone in Mylet road section (24a).

Description. Cranidium elliptical in outline, twice as broad (tr.) as long (sag.). Axial furrows deepest posteriorly
where strongly divergent, broad and distinct over most of length but only faintly developed forward of 2p lateral
lobes. Glabella broadest (tr.) at mid-level of lp lateral lobes. Median lobe roughly semi-cylindrical, narrowing
slightly anteriorly; strongly convex (tr.), bounded by deep, broad longitudinal furrows. Large lp lateral lobes
ovoid, strongly convex. 2p lobes of similar form but only half length (exsag.) of basal lobes; separated from them
by broad, adaxially deepening lp lateral furrows. Small 3p lateral lobes fused to median lobe to form narrow
(sag. and exsag.), transverse anterior section. Occipital furrow broad, mesially shallow; ring very broad, bearing
large, paired occipital spines, small median tubercle near anterior margin. Longitudinal furrows contain shallow
apodemal pits where they merge with the lp and 2p lateral furrows and with the longitudinal furrows. Fixed
cheeks strongly convex and steeply declined postero-laterally; antero-mesially not distinctly separated from
transverse anterior section of antero-median lobe, posteriorly only weakly separated from occipital ring.
Posterior border furrow broad, curving strongly forwards abaxially; border narrow adaxially, broadening and
curving gently forwards distally. Anterior border furrow broad; border narrow (sag. and exsag.) but form not
clearly seen. Antero-laterally on cranidium narrow (exsag.) eye-ridges run out and slightly back to small
palpebral lobes situated opposite 2p lateral furrows. Posterior branches of facial sutures run back in gently
sigmoidal curves, the anterior sections concave adaxially. Anterior branches abaxially run close to eye-ridges
before curving forwards to cross anterior border at low angle. Free cheeks crescentic. Strongly convex sutural
ridges broaden posteriorly and give rise to broad-based librigenal spines. Broad furrows, anteriorly convergent
with anterior border furrow, separate sutural ridges from remaining convex portions of cheeks. Lateral borders
broad (tr.) and spinose but number and size of spines not clear. Cranidial surface ornamented with prominent,
closely spaced tubercles of c. 01 5-0-2 mm.

Discussion. Whittington (1956, p. 515, pi. 59, fig. 13; pi. 60, figs. 2, 3, 5, 6, 10) selected a lectotype for P.
terribilis from among Reed’s syntypes and redescribed this and other material from the Rawtheyan
Starfish Bed of Girvan. In cranidial characters the south Welsh specimens do not significantly differ
from this material though the poor preservation of the free cheeks and the present lack of other parts
of the exoskeleton preclude a full comparison. The subspecies P. terribilis bituberculata Kielan (1960,
p. 107, pi. 3, fig. 3; pi. 16, fig. 1) from the upper Ordovician of Poland differs from both the Scottish
and south Welsh specimens in showing in addition to the general surface tuberculation several much
larger, regularly positioned tubercles on the cranidium.

Subfamily odontopleurinae Burmeister, 1 843
Genus primaspis R. and E. Richter, 1917

Type species. Odontopleura primordialis Barrande, 1 846.

PRICE: LATE ORDOVICIAN TRILOBITES

879

Primaspis llandowrorensis sp. nov.

Plate 114, figs. 8-15

1973a Primaspis aff. semievoluta (Reed); Price, list p. 233, table 2.

Holotype. GSM 5213, 21053 (PI. 8, figs. 11-13), internal and external moulds of incomplete articulated
exoskeleton from Craig-y-deilo quarry, Llandowror.

Paratypes. NMW 2 1. 306. G 17, external mould of pygidium and NMW 21 .306.G18, internal mould of free cheek,
both from Slade and Redhill Mudstones of Coed Old Pale, near Llandowror; HM A9633, internal mould of
cranidium, high Sholeshook Limestone, Lan-y-gaer (locality 16a); BM It9264, 9265, 9266a, b, internal mould of
partial cranidium, external mould of incomplete pygidium, and counterpart moulds of free-cheek, from
Sholeshook Limestone of Robeston Wathen; SM A77994, poor internal mould of free cheek from 9|-10 m
above base of Sholeshook Limestone in Mylet Road section (locality 24a), Llandowror. This comprises all
available material.

Diagnosis. Primaspis with relatively narrow (tr.) median glabellar lobe, small, poorly differentiated 3p lateral
lobes, centrally (not posteriorly) positioned occipital tubercle and stout genal spines, particularly broad where
they join the posterior borders; pygidium with only one pair of anterior secondary border spines and outermost
posterior pair of secondary spines not fused with bases of major spines; sculpture of small, well-spaced granules.

Description. Cranidium broadest posteriorly where width about 2£ times sagittal length. Glabella broadest at
mid-level of lp lateral lobes; width here about equal to pre-occipital length. Occipital ring strongly convex (tr.),
broad and sub-parallel sided mesially with prominent centrally positioned median tubercle, narrows sharply
behind posterior ends of lp lateral furrows and then curves forward to form prominent occipital lobes. Occipital
furrow broad and prominent. Large lp lateral lobes ovoid, 1| times as long (exsag.) as wide (tr.), strongly
convex; long axes diverge forward at c. 25°. 2p lobes about two-thirds length of lp, of similar form and
orientation, lp and 2p lateral furrows prominent, curving in and strongly back completely separating lp and 2p
lobes from median lobe; lp furrow deeper than 2p. 3p lateral lobes very small, with only slight independent
convexity; oriented antero-laterally and defined anteriorly by broad but short and shallow 3p furrows. Median
lobe strongly convex (tr.), narrowest (tr.) at mid-level of 2p lateral lobes, widest at about three-quarters length of
lp lobes. Frontal lobe about as wide (tr.) as posterior part of median lobe, roughly semicircular, dropping
steeply to broad anterior border furrow. Anterior border narrow (sag. and exsag.) and upturned (PI. 1 14, fig. 9).
Axial furrows broad and deep posteriorly. Elongated triangular strips of fixed cheeks lie outside axial furrows,
broadening and becoming more strongly convex posteriorly. Antero-laterally to these strips, separated by strong
furrows, narrow convex eye-ridges run back to palpebral lobes situated opposite posterior halves of lp lateral
lobes. Anterior branches of facial sutures curve forwards and adaxially, gradually diverging from eye-ridges;
posterior branches curve out and gradually back. Convex posterior borders narrow adaxially, broadening
rapidly outwards. Free cheeks crescentic; at inner posterior corners bear eyes on stout, elongated stalks (PI. 1 14,
fig. 14); extended postero-laterally as long, stout, broad-based librigenal spines. Lateral border furrow broad;
narrow, convex border appears to bear about thirteen slender border spines, longest posteriorly. Cephalic
surface ornamented with prominent, well-spaced granules of up to 0-175 mm.

Hypostoma unknown. Thorax of ten segments. Axis strongly convex, occupying less than one-third total
width anteriorly. Axial rings convex (sag. and exsag.), broadest mesially, abaxially narrowing then curving
forwards to form axial lobes. Articulating furrows broad and shallow mesially, outwards forming deep
apodemal slots. Axial furrows shallow. Pleurae comprise broad, strongly convex posterior bands separated by
narrow, distinct pleural furrows from lower, narrow, convex anterior bands which bear a row of about six
regularly spaced, small tubercles. Inner portions of pleurae horizontal but at fulcrum deflected ventrally and
posteriorly; posterior bands have prominent fulcral expansions (PI. 114, fig. 13). Axial rings and posterior
pleural bands with similar ornament to that of cephalon.

Pygidium, discounting border spines, sub-triangular, about three times as wide (tr.) as long (sag.). Axis
occupies one-third total width anteriorly, tapers back rapidly. First axial ring strongly convex longitudinally,
moderately so transversely, clearly defined; second ring about two-thirds as wide (tr.), lower, less convex (sag.
and exsag.), followed by low triangular terminal portion fused to convex posterior border. Convex pleural ridge
curves out and strongly back and swells slightly as it approaches posterior border; pleural areas between it the
axis and the borders depressed. Border spines long, gradually tapering. Between the major spines are four
secondary posterior spines of which the outermost are not fused with the bases of the major spines; only one pair
of anterior secondary spines is present. Ornament of well-spaced granules poorly preserved.

PALAEONTOLOGY, VOLUME 23

Discussion. In cephalic features P. llandowrorensis sp. nov. shows much over-all similarity to P.
bucculenta McNamara (1979a, p. 86, pi. 12, figs. 10-19) from Cautleyan Zone 3 in the southern Lake
District. In that form, however, the fixigenae are much broader (tr.) posteriorly, the librigenal spines
appear shorter and more rapidly tapering and the librigenal denticles are shorter. Like P. bucculenta
the south Welsh form differs from P. evoluta (Tornquist) from the upper Ordovician of Sweden (see
Bruton 1966, p. 4, pi. 1, figs. 1-9; pi. 4, fig. 9; text-fig. 2a) and north Wales (Whittington 1968, p. 98, pi.
30, figs. 25-30) in having a relatively narrower median glabellar lobe, larger 2p lateral lobes, smaller,
less prominent and less anteriorly divergent 3p lobes, a greater number of librigenal border spines,
and a central occipital tubercle.

In pygidial characters P. llandowrorensis differs from both these forms in lacking the fusion
between an outermost (third) pair of posterior secondary border spines and the bases of the
macrospines. It lacks the second anterior secondary border spines of P. evoluta and in over-all form is
relatively longer (sag.) and less broad (tr.). P. bucculenta, like P. llandowrorensis , has only one
anterior secondary spine but the macrospines in P. llandowrorensis are much longer, narrower-based,
and less strongly curved. The anterior secondary spines appear longer also and the pygidial
tuberculation is much finer. The lack of the fused border spines together with the presence of only one
pair of anterior secondary spines appear to distinguish the pygidium of P. llandowrorensis from that
of any other known species of the genus.

Primaspis sp. indet.

Plate 114, fig. 16

1973a Primaspis cf. evoluta (Tornquist); Price, tables 1, 2.

Material. Single specimen, SM A77514a, b, internal and external moulds of incomplete pygidium from the
highest Sholeshook Limestone of Prendergast Place (locality 8b), Haverfordwest.

Description. Axis strongly convex; anteriorly with well-developed articulating ring separated by broad furrow.
First axial ring strongly convex (sag. and exsag.); second lower and less wide (tr.), bearing pair of large but ill-
defined tubercles and followed by a low, rapidly tapering (tr.) terminal portion fused with the convex posterior
border. Strongly convex pleural ridge curved outwards and posteriorly. Between ridge, axis, and posterior
border pleural areas depressed. Posteriorly there are six secondary spines of which the outermost are small and
fused with the swollen bases of the major spines. All the spines but particularly the major spines, the posterior
border, the pleural ridge, and the first axial ring are ornamented with prominent granules.

Discussion. In over-all form and proportions and in the character of the ornamentation the
incomplete Prendergast pygidium resembles the corresponding portions of the pygidia of both P.
evoluta (cf. Bruton 1966, pi. 1, figs. 6-7; pi. 4, fig. 9; text-fig. 2d) and P. bucculenta (cf. McNamara
1979a, pi. 12, figs. 16, 18, 19). The border spines are relatively shorter and stouter than those of P.
bestorpensis Bruton (1966, pp. 7-9, pi. 2, figs. 1, 2, 5-6; text-fig. 2b) from the Bestorp Limestone (basal
Harju Series) of Vastergotland, and the depressed pleural areas lack the coarse granulation seen in
that form. In the absence of the antero-lateral parts of the pygidium further comparisons are not
possible.

?Family glaphuridae Hupe, 1953
Genus glaphurella Dean, 1971

Type species. Cyplnaspis ? Harknessi Reed, 1896.

Glaphurella cf. harknessi (Reed, 1896)
Plate 114, figs. 17, 18

1905a Cyphaspis cf. Harknessi, Reed; Reed, p. 98.

1914 Proetus harknessi Reed; Strahan et al., table p. 64.

1973a Glaphurella cf. harknessi (Reed); Price, tables 1, 2.

PRICE: LATE ORDOVICIAN TRILOBITES

Material, horizons, and localities. SM A 104833, internal mould of partial cranidium, Sholeshook Limestone,
Sholeshook; GSM Pg. 296, 299, internal and external moulds of cranidium, middle section of Sholeshook
railway cutting; GSM Pg. 280, internal mould of flattened partial cranidium, same horizon and locality; SM
A77948, internal mould of partial cranidium, locality 9e, Sholeshook; SM A30940, internal mould of incomplete
cranidium, basal Slade and Redhill Mudstones south-west of Knock (locality 3).

Discussion. Dean (1971, p. 44, pi. 22, figs. 3-10, 12, 13; pi. 23, fig. 1) redescribed the holotype
cranidium of G. harknessi from the Keisley Limestone of Cross Fell together with better-preserved
specimens from the Chair of Kildare Limestone of eastern Ireland. The south Welsh specimens are
similar in general form and in details of ornamentation but appear to have glabellae which are
consistently relatively longer (sag.) and less wide (tr.) than in the cranidia figured by Dean and which
do not drop so steeply anteriorly. Also, although the 2p lateral glabellar furrows are just visible on
internal moulds of some of the south Welsh cranidia, none of them show any signs of glabellar
lobation anterior to these. Such differences may simply be the effects of distortion and poor
preservation but until better specimens are available the south Welsh form is probably best only
compared with G. harknessi.

RANGES, ABUNDANCES, AND FAUNAL COMPARISONS

The known ranges or restricted occurrences of trilobite species within the three developments of the
Sholeshook Limestone Formation, together with indications of their relative abundance, are given in
Table 1. The five categories in the list of abundances are based primarily on the relative numbers of
specimens in the author’s collections and in other recent collections from the formation where an
attempt has been made to retain all potentially identifiable material (e.g. S. F. Morris collection BM,
J. K. Ingham collection HM) — though some of the forms listed as rare are known only from old
collections. Each of these categories is a generalization for the formation as a whole or for those parts
of it in which the particular taxon occurs; it has not been possible to sample in a sufficiently controlled
way throughout all developments of the Sholeshook Limestone to give a more rigorous quantitative
assessment of the abundance of each form or to chart variations in abundance at different horizons.
In comparing the Sholeshook trilobites with other Ashgill trilobite faunas it is most useful to deal
with comparisons at specific level separately from those at generic level. This is because the specific
composition of the fauna is the basis for assessing the precise age and correlation of the formation
while differences from other Ashgill faunas in generic composition probably relate to factors other
than age relationships.

The age and correlation of the Sholeshook Limestone have been extensively discussed elsewhere
(Price 1973a, b, 1980) and it is now considered that the formation ranges from high Cautleyan Zone 1
to Rawtheyan Zone 5. This conclusion is based on comparisons of the vertical distribution of
trilobite species in the formation with the stratigraphical ranges of identical or closely allied species in
the type Ashgill succession at Cautley (Ingham 1966, 1972-7). A precise correlation between the two
successions is possible because they have large numbers of species in common. Several species of
Tretaspis from Sholeshook and species of Illaenus, Stenopareia, Pseudosphaerexochus, Calymene
(s.l.), Flexicalymene, Kloucekia, Duftonia, and Toxochasmops are all considered to be conspecific and
species of Atractopyge, Lonchodomas, and Hadromeros very probably conspecific with forms of
Cautleyan or lowest Rawtheyan age at Cautley.

Accepting this correlation, the occurrence in low to middle horizons of the Sholeshook Limestone
of a few species occurring elsewhere in Rawtheyan horizons extends their stratigraphical ranges
(Price 1973a, b). This argument was initially made with reference to elements of the ‘ Phillipsinella
parabola— Staurocephalus clavifrons fauna’ where there is some confirmatory evidence for pre-
Rawtheyan occurrences from other sections (Price 19736) but it has been recently extended to apply
also to other forms such as Prionocheilus cf. obtusus and Glaphurella cf. harknessi (Price 1980). Thus
certain species in the Sholeshook trilobite fauna appear to be long-ranging and are common to
younger faunas such as those of the Rhiwlas Limestone of north Wales (Whittington 1962-1968) and
the Chair of Kildare Limestone of eastern Ireland (Dean 1971-1978). They are, however, relatively

882

PALAEONTOLOGY, VOLUME 23

few in number and when such faunas are fully compared with the Sholeshook fauna the difference in
age is reflected in the presence of different species of, for example, Illaenus, Tretaspis, Lonchodomas,
Pseudosphaerexochus, and Platylichas', over-all the number of species in common is much smaller
than in the case of the Cautleyan fauna of the Cautley Mudstones referred to earlier.

Nearly all the genera present in the faunas so far discussed and in other Ashgill trilobite faunas are
known to range through the Cautleyan and Rawtheyan Stages. Most of them are known to be
geographically wide ranging. It can reasonably be argued therefore that where these faunas differ in
generic composition it is largely as a result of differences in palaeoenvironmental factors. Although
the environmental significance of Ashgill trilobite faunas is as yet only poorly understood, there do
appear to be at least two reasonably clear associations between facies and fauna which can be referred
to here and used as a basis for comparison with the Sholeshook fauna.

The first of these associations relates to the faunas of light-coloured, relatively pure, biosparite
limestone developments often considered to be, at least in part, of ‘reef facies. These are here taken to
be relatively shallow-water accumulations probably representing shelf-edge or near shelf-edge
environments. Examples are the Boda Limestone of Sweden, the Chair of Kildare Limestone of
eastern Ireland, and the Keisley Limestone of northern England all usually considered to be
Rawtheyan in age, at least partially. These formations have trilobite faunas represented almost
entirely by disarticulated remains. Broadly their faunas appear to be characterized by the importance
in them of illaenids, cheirurids, and lichids (illaenid-cheirurid community type of Fortey, 1975). To a
much greater extent than in the Sholeshook fauna these groups are both numerically predominant
and generically diverse. When an over-all comparison is made the Sholeshook fauna does have a
number of genera in common (cf. Dean 1978, table p. Ill) such as Atractopyge, Hadromeros,
Pseudosphaerexochus , Platylichas, Illaenus, Stenopareia, Panderia, Prionocheilus, and the rarer forms
Trochurus, Sphaerocoryphe, and Glaphurella. This is as far as the similarity can be taken, however, for
many other forms apparently characteristic of the pure limestone association are not represented at
Sholeshook — Sphaerexochus, Holotrachelus, Stenoblepharum, Decoroproetus, Dicranopeltis,
Toernquistia, and isocolids are notable examples. Similarly the pure limestone faunas themselves
completely lack the following Sholeshook genera: Calyptaulax, Kloucekia, Duftonia, Liocnemis,
Lonchodomas, Raphiophorus, Flexicalymene, Brongniartella, Opsimasaphus, Encrinuroides,
Dindymene, Amphitryon, Dionide, and Nankinolithus. Chasmopines are also completely absent from
them and though Tretaspis may be present it is usually very rare.

Many of the Sholeshook genera listed above as being absent from the pure limestone association
appear to be more characteristic of a second distinct association. This relates to certain mudstone
sequences here taken to represent deeper-water, low-energy environments probably considerably
down-slope from the platform edge. Good examples are the mudstones of the ‘ Staurocephalus
clavifrons Zone1 of Poland and the Kraluv Dvur Formation of Bohemia (Kielan 1960; Havlicek and
Vanek 1966); in Wales the Ashgill mudstones of Grugan and Llanystwmdwy in the Lleyn Peninsula
(Matley 1938; Harper 1956) appear to be of similar type. Their faunas usually contain a significant
proportion of complete or almost complete articulated exoskeletons. Among trilobites which appear
to be important and characteristic elements of such faunas are several genera which occur at
Sholeshook: Liocnemis, Lonchodomas, Raphiophorus, Opsimasaphus, Dindymene, Amphitryon,
Dionide, and Nankinolithus. With the exception of Nankinolithus these genera are among the rarer
elements of the Sholeshook fauna and even Nankinolithus is only abundant at Sholeshook within a
very restricted vertical range. At Crugan Duftonia occurs in association with the genera listed above.
Kloucekia, Flexicalymene, and Brongniartella may also be less ubiquitous members of this mudstone
association. The Polish and Bohemian mudstone faunas also contain many elements, possibly
representing genera preferring even deeper water conditions— perhaps in foot-of-slope and basinal
environments, not known from Sholeshook (though present elsewhere in Wales in the mudstones of
the Abercwmeiddaw Group of the Corris-Ddinas Mawddwy area; P. M. Magor and J. K. Ingham
coll.). These include Novaspis, several cyclopygid genera and telephinids.

On general sedimentological and stratigraphical evidence the Sholeshook Limestone probably
represents an environment in the middle to upper part of the slope between platform edge and basin

PRICE: LATE ORDOVICIAN TRILOBITES

883

where deposition took place under relatively high energy conditions (skeletal material largely
disarticulated, often broken) and was dominantly clastic but with some carbonate content. In this
sense it would represent an environment intermediate between that represented by low-energy,
deeper-water mudstones on the one hand and by shallow-water carbonate accumulations on the
other. It is suggested that this ‘intermediate’ nature of the environment may be reflected in the variety
of genera in the trilobite fauna which embraces both forms more common in deeper-water mudstones
and forms occurring in pure limestones.

A few Sholeshook forms— Calyptaulax, Encrinuroides, and Toxochasmops (in fact chasmopines in
general)— do not appear to be usual members of either the pure limestone or the mudstone
associations and may be restricted to faunas of an ‘intermediate’ nature related to shallower-clastic
and impure carbonate sequences. Another characteristic of such faunas is an abundance of Tretaspis;
the genus becomes rare, as mentioned earlier, in pure limestone faunas and appears to be
progressively replaced in deeper environments by first Nankinolithus and then Novaspis. Ultimately it
may prove possible to characterize ‘intermediate’ faunas more fully as a separate third association
though clearly this would overlap with both the carbonate and mudstone associations.

Overlap of this kind might prove useful, however, in permitting within a broad ‘intermediate’
association some distinction between faunas of deeper and shallower water affinities. For instance,
when the Sholeshook trilobites are compared with those from the Cautleyan Stage of the Cautley
area (Ingham 1966, 1970-1977) one major difference is the absence of those Sholeshook genera listed
as being characteristic of the deeper-water mudstone association— Liocnemis, Raphiophorus,
Opsimasaphus, Dindymene, Amphitryon, Dionide, and Nankinolithus ( Lonchodomas is an exception).
Duftonia , Kloucekia, Brongniartella and Flexicalymene are also absent. In view of what has been
said in earlier sections these differences would suggest that the Cautleyan rocks at Cautley were
deposited under shallower water conditions than was the Sholeshook Limestone. This suggested
difference in environment might also relate to other faunal differences. For example, the Sholeshook
genus Flexicalymene is replaced at Cautley by species of Calymene (s.l.) (‘ Diacalymene') and
Gravicalymene, and Harpidella is replaced by Otarion. The odontopleurids at Cautley are represented
by Acidaspis in addition to Primaspis, and Decoroproetus is common.

Similar differences are seen when the Sholeshook fauna is compared with that of the Birdshill
Limestone, probably of Pusgillian-low Cautleyan age, of the Llandeilo area. Here again the
Harpidella of the Sholeshook fauna is replaced by Otarion, the Flexicalymene is replaced by
Gravicalymene, and both Acidaspis and Decoroproetus are present. The Birdshill Limestone is a light-
coloured, relatively pure limestone, coarse grained, sparry, and largely bioclastic; all of these
characters suggest a relatively shallow-water origin. The fauna, however, differs from that of the
Cautleyan Stage at Cautley, probably as a reflection of its closer affinity with the pure carbonate
association outlined earlier, in that species of Platylichas are important elements, Holotrachelus may
also be present and Tretaspis is very rare (information based on collections from Birdshill Limestone
in BM). The Rhiwlas Limestone of north Wales is faunally more like the Sholeshook Limestone in
containing a number of genera like Amphitryon, Dindymene, Opsimasaphus, Lonchodomas,
Raphiophorus, and Cyclopyge associated with deeper mudstone environments as well as forms like
Encrinuroides, Platylichas, Prionocheilus, Sphaerocoryphe, and Ulugtella (last named noted by Dean
1978, p. 113). Like the Sholeshook Limestone it may represent an environment towards the deeper
part of the ‘intermediate’ (broadly mid-slope) range.

REFERENCES

angelin, n. p. 1854. Palaeontologia Scandinavica I; Crustacea formationis transitionis, Fasc. 2, 21-92, pis.
25-41. Lund.

barrande, j. 1846. Notice preliminaire sur la Systeme Silurien et les Trilobites de Boheme, i-iv, 97 pp. Leipsig.

— 1 847. Uber das Hypostoma und Epistoma, zwei analogue, aber, verschiedene Organe der Trilobiten. Neues
Jb. Miner. Geol. Palaont. 1847, 385-399, pi. 8.

— 1852. Systeme Silurien du centre de la Boheme, Tre partie. Recherches paleontologiques, vol. 1, i-xxx, 935
pp., 51 pis. Prague and Paris.

PALAEONTOLOGY, VOLUME 23

barrande, j. 1972. Systeme Silurien du centre de la JBoheme , lere partie. Recherches palaeontologiques, supplement
au vol. 1, xxx + 657 pp.. Atlas of 35 plates. Prague and Paris.
beyrich, e. 1845. Ueber einige bohmische Trilobiten, 47 pp., 1 pi. Berlin.

brongniart, A. 1822. In brongniart, A. and desmarest, a. G. Histoire naturelle des Crustaces fossiles: les
trilobites, 65 pp., 4 pis. Paris and Strasbourg.

bruton, D. l. 1966. A revision of the Swedish Ordovician Odontopleuridae (Trilobita). Bull. Geol. Instn Univ.
Uppsala , 43, 1-40, pis. 1-6.

— 1968. The trilobite genus Panderia from the Ordovician of Scandinavia and the Baltic areas. Norsk geol.
Tidsskr. 48, 1-53, pis. 1-12.

burmeister, h. 1843. Die Organisation der Trilobiten aus ihren lebenden Verwandten entwickelf, nebst einer
systematischen Ubersicht aller zeither beschriebenen Arten, i-xii + 1-147 pp. Berlin.
cantrill, T. c. et al. 1916. The geology of the South Wales coalfield, part xii, the country around Milford. Mem.
geol. Surv. G.B., Sheet 227, 182 pp.

chapman, e. j. 1890. Some remarks on the classification of the trilobites as influenced by the stratigraphical
relations; with outline groupings of these forms. Trans, roy. Soc. Canada, 7, i-vii, 1 13-120.

COCKS, l. r. m. and price, D. 1975. The biostratigraphy of the upper Ordovician and lower Silurian of south-west
Dyfed, with comments on the Hirnantia fauna. Palaeontology, 18, 703-724, pis. 81-84.
cooper, G. a. 1930. Upper Ordovician and Lower Devonian Stratigraphy and Paleontology of Perce, Quebec.

Part II, new species from the Upper Ordovician of Perce. Amer. J. Sci. 20, 265-392, pis. 1 -5, 7 figs.
dalman, J. w. 1827. Om palaeaderna eller de sa kallade trilobiterna. K. Vet. Acad. Handl. Stockh. 1826, 1-40,
226-294, pis. 1-6.

dean, w. T. 1959. Duftonia, a new trilobite genus from the Ordovician of England and Wales. Palaeontology, 2,
143-149, pi. 19.

1960. The Ordovician trilobite faunas of south Shropshire, I. Bull. Br. Mus. nat. Hist. (Geol.), 4, 73- 143, pis.

11-19. 1963. The Ordovician trilobite faunas of south Shropshire, III. Ibid. 7, 215-254, pis. 37-46.

— 1971-1978. The trilobites of the Chair of Kildare Limestone (Upper Ordovician) of eastern Ireland.
Palaeontogr. Soc. ( Monogr .): (1), 1971, 1-60, pis. 1-25; (2), 1974, 61-98, pis. 26-44; (3), 1978, 99-129, pis.
45-52.

destombes, j. 1966. Quelques Calymenina (Trilobitae) de l’Ordovicien moyen et superieur de l’Anti-Atlas
(Maroc). Notes Mem. Serv. geol. Maroc. 26, 33-52, 4 pis.

— 1972. Les trilobites des sous-ordre des Phacopina de l’Ordovicien de l’Anti-Atlas (Maroc). Ibid. 240,
112 pp., 16 pis.

fortey, R. A. 1975. Early Ordovician trilobite communities. Fossils Strata, 4, 331-352.

gurich, g. 1901 . Ueber eine neue Lichas-Art aus dem Devon von Neu-Siid-Wales und fiber die Gattung Lichas
fiberhaupt. Neues Jb. Miner. Geol. Palaont. BeilBd, 14, 519-539, pis. 18, 20.

— 1907. Versuch einer Neueinteilung der Trilobiten. Zentbl. Miner. Geol. Palaont. 1907, 129-133.
harper, J. c. 1956. The Ordovician succession near Llanystwmdwy, Caernarvonshire. Lpool Manchr geol. J., 1,

385-393.

havliCek, v. and vanek, j. 1966. The biostratigraphy of the Ordovician of Bohemia. Shorn, geol. ved. pal. 8,
7-69, pis. 1-16.

hawle, I. andcORDA, A. j. c. 1847. Prodrom einer Monographic der bohmischen Trilobiten, 176 pp., 7 pis. Prague.
holliday, s. 1942. Ordovician trilobites from Nevada. J. Paleont. 16, 471-478, pis. 73, 74.
holm, G. 1882. De Svenska artena af Trilobitslaget Illaenus (Dalman). Bih. K. svensk. VetenskAkad. Handl. 7,
xiv + 148 pp., 6 pis.

horny, R. and bastl, F. 1970. Type specimens of fossils in the National Museum, Prague. Volume 1 . Trilobita, 354
pp., 20 pis. Prague.

hughes, c. p., ingham, J. K. and addison, r. 1975. The morphology, classification and evolution of the
Trinucleidae (trilobita). Phil. Trans. R. Soc. Lond. B272, 537-604, pis. 1 10.
hupe, p. 1953. Classification des trilobites. Annls Paleont. 39, 61-169 (1-110), figs. 1-92.

— 1955. Classification des trilobites. Annls Paleont. 41, 91-325, (1 11-345), figs. 93-247.

ingham, J. K. 1966. The Ordovician rocks in the Cautley and Dent districts of Westmorland and Yorkshire. Proc.
Yorks, geol. Soc. 35, 455-505.

— 1970-1977. A monograph of the upper Ordovician trilobites from the Cautley and Dent districts of
Westmorland and Yorkshire. Palaeontogr. Soc. (Monogr.)-. (1), 1970, 1-58, pis. 1-9; (2), 1974, 59-87, pis.
10-18; (3), 1977, 89-121, pis. 19-27.

jaanusson, v. 1954. Zur Morphologie und Taxonomie der Illaeniden. Ark. Miner. Geol. 1, 545-583, pis. 1-3.
Kiel an, z. 1957. On the trilobite family Staurocephalidae. Acta palaeont. pol. 2, 155-180, pis. 1-3.

PRICE: LATE ORDOVICIAN TRILOBITES

885

kielan, z. 1960. Upper Ordovician trilobites from Poland and some related forms from Bohemia and
Scandinavia. Palaeont. pol. 11, i-vi, 198 pp., pis. 1-36.

kutorga, s. 1854. Einige Sphaerexochus und Cheirurus aus den silurischen Kalksteinschichten des Gouvern-
ments von St. Petersburg. Zap. imp. miner. Obshch. 13, 105-126, pis. 1-3.
lane, p. d. 1971. British Cheiruridae (Trilobita). Palaeontogr. Soc. ( Monogr .), 95 pp., 16 pis.
la touche, J. D. 1884. A Handbook of the Geology of Shropshire , 91 pp., 22 pis. London and Shrewsbury.
lesperance, p. J. 1968. Ordovician and Silurian trilobite faunas of the Whitehead Formation, Perce region,
Quebec. J. Paleont. 42, 811-826, pi. 106.

linnarsson, J. G. o. 1869. Om Vestergotlands Cambriska och Silurischa aflagringar. K. svensk. VetenskAkad.
Hand!. 8, 1-89, pis. 1,2.

lu, yen-hao. 1954. A brief note on the upper Ordovician trilobites from Tangshan, Nanking. Proc. Pal. Soc.
China. 1.

— 1957. Index fossils of China. Invertebrates Part III. Trilobita. Peking: Geological Press.
mccoy, F. 1846. A synopsis of the Silurian fossils of Ireland , 72 pp., 5 pis. 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., ser. 2, 4, 161-179, 330-335, 392-414.

— 1851. In Sedgwick, a. and mccoy, f. A synopsis of the classification of the British Palaeozoic rocks, with a
systematic description of the British Palaeozoic fossils in the geological museum of the University of Cambridge,
Fasc. I, pp. 1-184, pis. 1a-l.

mcnamara, K. J. 1979a. Trilobites from the Coniston Limestone Group (Ashgill Series) of the Lake District,
England. Palaeontology, 22, 53-92, pis. 7-12.

— \919b. The age, stratigraphy and genesis of the Coniston Limestone Group in the southern Lake District.
Geol.J. 14,41-68, 9 figs.

marr, j. E. and Roberts, T. 1885. The Lower Palaeozoic rocks of the neighbourhood of Haverfordwest. Q. Jl
geol. Soc. bond. 41, 476-491, pi. 15.

matley, c. A. 1938. The geology of the country around Pwllheli, Llanbedrog and Madryn, south-west
Carnarvonshire. Ibid. 94, 555-606.

milne Edwards, h. 1840. Histoire naturelle des Crustaces, comprenent /’ Anatomie, la physiologie et la classifica-
tion de ces animaux, vol. 3, 1-638. Paris.

niezkowski, j. 1 857. Versuch einer Monagraphie der in den silurischen Schichten der Ostseeprovinzen vorkom-
menden Trilobiten. Archiv. Naturk. Lib.-Ehst-Kurlands, 1 (1), 517-626, pis. 1-3.

Owens, R. m. 1973. British Ordovician and Silurian Proetidae (Trilobita). Palaeontogr. Soc. (Monogr.), 98 pp.,
15 pis.

phleger, F. B. 1936. Lichadian trilobites. J. Paleont. 10, 593-615, 83 figs.

portlock, J. E. 1843. Report on the geology of the county of Londonderry, and parts of Tyrone and Fermanagh.
xxx + 784 pp., pis. 1-38a-i, Map. Dublin and London.

prantl, f. and pribyl, A. 1947. Classification of some Bohemian Cheiruridae (Trilobita). Sb. nar. Mus. Praze, 3,
Geol. (Palaeont.), 1, 1-44, pis. 1-6.

— 1948. Some new or imperfectly known Ordovician trilobites from Bohemia. Bull. int. Acad, tcheque
Sci. 49, 1-23, pis. 1-3.

— 1949. A study of the superfamily Odontopleuracea nov. superfam. (trilobites). Rozpr. st. geol. Ust. 12,
1-221, pis. 1-11.

pribyl, a. 1953. Seznam ceskych trilobitovych rodu (Index of trilobite genera in Bohemia). Knih. iitsr. Ust. geol.
25, 1-80.

price, d. 1973a. The age and stratigraphy of the Sholeshook Limestone of south-west Wales. Geol. J. 8, 225-246.

— 1973b. The Phillipsinella parabola — Staurocephalus clavifrons fauna and Upper Ordovician correlation.
Geol. Mag. 110, 535-541.

— 1974. Trilobites from the Sholeshook Limestone (Ashgill) of south Wales. Palaeontology, 17, 841-868, pis.
112-116.

— 1977. Species of Tretaspis (Trilobita) from the Ashgill Series in Wales. Palaeontology, 20, 763-792, pis.
98-103.

— 1980. A revised age and correlation for the topmost Sholeshook Limestone Formation (Ashgill) of south
Wales. Geol. Mag. 117, 485-489.

Raymond, p. e. 1913. Subclass Trilobita. In Eastman, c. r. (ed.). Text-book of Palaeontology (2nd edn.), vol. 1,
839 pp., 1594 figs. London.

— 1925. Some trilobites of the lower Middle Ordovician of eastern North America. Bull. Mus. comp. Zool.
Harv. 67, 1-180, pis. 1-10.

886

PALAEONTOLOGY, VOLUME 23

reed, f. r. c. 1896. The fauna of the Keisley Limestone— Part 1 . Q. Jl geol. Soc. Lond. 52, 407-437, pis. 20, 21.

— 1903-1906. The Lower Palaeozoic trilobites of the Girvan district, Ayrshire. Palaeontogr. Soc. ( Monogr.)-.
(1), 1903, 1-48, pis. 1-6; (2), 1904, 49-96, pis. 7-13; (3), 1906, 97-186, pis. 14-20.

— 1904. Sedgwick Museum Notes. New fossils from the Haverfordwest District. 2. Geol. Mag. Dec. 5, 1,
383-388, pi. 12.

— 1905a. Sedgwick Museum Notes. New fossils from the Haverfordwest District. 3. Ibid. 2, 97-104, pi. 4.

— 19056. The classification of the Phacopidae. Ibid. 172-178, 224-228.

— 1914. The Lower Palaeozoic trilobites of Girvan. Supplement. Palaeontogr. Soc. ( Monogr .), 56 pp., 8 pis.

— 1918. Notes on the genus Homalonotus. Geol. Mag. 5, 263-276, 314-327.

— 1933. Notes on the species Illaenus bowmanni Salter. Ibid. 70, 131-135.

— 1935. The Lower Palaeozoic trilobites of Girvan. Supplement no. 3. Palaeontogr. Soc. (Monogr.), 64 pp.,
4 pis.

richter, R. and richter, E. 1917. Palaeontologische Beobachtungen im Rheinischen Devon, I. Ueber einzehne
Arten von Acidaspis, Lichas, Cheirurus, etc. aus der Eifel. Jb. Nass. Ver.f. Naturkunde, 70, 143-161.

— 1925. Unterlagen zur Fossilium Catalogus, Trilobitae. III. Senckenbergiana, 7, 239-244.
rouault, M. 1847. Extrait du memoire sur les trilobite du department dTlle-et-Vilaine. Bull. Soc. geol. Fr. 4,
309-328, pi. 3.

salter, J. w. 1848. In Phillips, J. and salter, j. w. Palaeontological appendix to Professor John Phillips’s
Memoir on the Malvern Hills, compared with the Palaeozoic districts of Abberley, etc. Mem. geol. Surv. U.K.
2 (1), vii-xiv + 33 1-386, pis. 4-30.

— 1853. Figures and Descriptions illustrative of British Organic Remains. Mem. geol. Surv. U.K. 1 Dec., 12
pp., 2 pis.

— 1864, 1865, 1867. A monograph of the British trilobites from the Cambrian, Silurian and Devonian
formations. Palaeontogr. Soc. (Monogr.)-. (1), 1864, 1-80, pis. 1-6; (2), 1865, 81-128, pis. 7-14; (4), 1867, 177-
214, pis. 25*-30.

— 1866. Appendix on the Fossils. In ramsay, a. c. The Geology of North Wales. Mem. geol. Surv. U.K. 3,
239-363, 372-381,26 pis.

sars, m. 1835. Ueber einige neue oder unvollstandig bekannte Trilobiten. Okens Isis, 28, (4) for 1835, cols.
333-343.

Schmidt, f. 1881. Revision der ostbaltischen silurischen Trilobiten nebst geognostischer Ubersicht des
ostbaltischen Silurgebiets. Abt. 1. Phacopiden, Cheiruriden und Encrinuriden. Mem. Acad. imp. Sci. St.
Petersb., ser. 7, 30, 1-237, pis. 116.

Shirley, j. 1936. Some British trilobites of the Calymenidae. Q. Jl geol. Soc. Lond. 92, 384-422, pis. 29-31.
siveter, d. j. 1977. The Middle Ordovician of the Oslo region, Norway, 27. Trilobites of the family Calymenidae.
Norsk geol. Tidsskr. 56, 335-396, 11 pis.

steininger, j. 1831. Observations sur les fossiles du Calcaire intermediare de l’Eifel. Mem. Soc. geol. Fr. 1 (1),
331-371, pis. 21-23.

stdrmer, l. 1945. Remarks on the Tretaspis (Trinucleus) Shales of Hadeland with description of trilobite faunas.
Norsk geol. Tidsskr. 25, 379-426, pis. 1-4.

strahan, a., cantrill, t. c., dixon, E. E. l. and thomas, H. h. 1909. The geology of the South Wales coalfield.
Part 10. The country around Carmarthen. Mem. geol. Surv. U.K., Sheet 229, 177 pp.

— and jones, o. t. 1914. Idem. Part 11. The country around Haverfordwest. Ibid., Sheet 228,

262 pp.

temple, j. t. 1972. Essay Review. Ontogenies of Devonian trilobites from New South Wales. Geol. Mag. 109,
373-376.

— 1975. Standardisation of trilobite orientation and measurement. Fossils and Strata, 4, 461-467.
thomas, a. t. and owens, R. m. 1978. A review of the trilobite family Aulacopleuridae. Palaeontology, 21, 65-81,

pi. 7.

tripp, r. p. 1957. The classification and evolution of the Superfamily Lichacea (Trilobita). Geol. Mag. 94,
104-122, 8 figs.

— 1958. Stratigraphical and geographical distribution of the named species of the trilobite superfamily
Lichacea. J. Paleont. 32, 574-582, pi. 85.

volborth, A. von. 1863. Uber die mit glatten Rumpfgliedern versehenen russischen Trilobiten, nebst einem
Anhange uber die Bewegungsorgane und uber das Herz derselben. Mem. Acad. imp. Sci. St Petersb., ser. 7, 6,
1-47, pis. 1-4.

wahlenberg, g. 1818. Petrificata telluris Svecanae examinata. Nova Acta R. Soc. Scient. upsal. 8, 1-116,
293-297, pis. 1-4, 7.

PRICE: LATE ORDOVICIAN TRILOBITES

887

warburg, E. 1925. The trilobites of the Leptaena Limestone in Dalarne. Bull. geol. Instn Univ. Uppsala, 17, i-vi,
1-446, pis. 1-11.

Whittington, H. b. 1950a. Sixteen Ordovician genotype trilobites. J. Paleont. 24, 531-565, pis. 68-75.

— 19506. A monograph of the British trilobites of the family Harpidae. Palaeontogr. Soc. ( Monogr .), i, ii, 55
pp„ 7 pis.

— 1956. Type and other species of Odontopleuridae (Trilobita). J. Paleont. 30, 504-520, pis. 57-60.

— 1962-1968. A monograph of the Ordovician trilobites of the Bala area, Merioneth. Palaeontogr. Soc.
{Monogr.): (1), 1962, 1-32, pis. 1-8; (2), 1965, 33-62, pis. 9-18; (3), 1966, 63-92, pis. 19-28; (4), 1968, 93-138,
pis. 29-32.

DAVID PRICE

Department of Geology
Sedgwick Museum
Downing Street
Cambridge CB2 3EQ

Typescript received 12 September 1979
Revised typescript received 1 March 1980

PALAEOBIOLOGY OF UPPER CRETACEOUS
BELEMNITES FROM THE PHOSPHATIC CHALK
OF THE ANGLO-PARIS BASIN

by IAN JARVIS

Abstract. The phosphatic chalks of northern France exhibit a tripartite belemnite biostratigraphy, with
Actinocamax verus Miller at their base, Gonioteuthis ex gr. quadrata in their upper portions, and G. quadrata
quadrata and Belemnitella praecursor Stolley at their summit. G. granulata (Blainville) is identified from isolated
specimens collected from the base of the sequences; G. granulataquadrata (Stolley) is recognized within
‘populations’ from the summit of phosphatic chalks. Principal component factor analysis suggests that variation in
guard morphology may be attributed to differences in guard size and to the evolutionary stage reached by
individuals within the gradualistic series formed by the genus Gonioteuthis. Heterogeneity in one sample is the
result of mixing of juvenile and mature populations caused by a catastrophic event, probably a storm. The
presence of juveniles in all samples indicates a near-shore environment which was the normal habitat of
belemnites. Hardgrounds show associated concentrations of belemnites because of greater food availability and
their suitability as breeding sites.

Belemnites occur throughout the Santonian to early Campanian sequences of the Anglo-Paris
Basin, but their rarity may be measured by the observations of Rowe (1908, p. 311) who obtained
only ten specimens of Gonioteuthis from the entire Campanian section on the Isle of Wight during
more than two months’ intensive collecting. Similarly, Brydone (1914) stated that only thirteen
accurately located and identifiable specimens of Gonioteuthis had previously been recorded from the
Chalk of Hampshire and Sussex. The infrequency of belemnites in soft white chalks of this age has
been noted by other authors in sequences outside the Anglo-Paris Basin (e.g. Ernst 1964; Christensen
19766). Belemnites, in particular Gonioteuthis and Actinocamax verus Miller are, however, common
and at some levels extremely abundant in the phosphatic chalk lithofacies. Consequently, while the
sporadic occurrence of belemnites in soft white chalks has led to the inapplicability of population
analyses, material from phosphatic chalks provides a unique opportunity to examine accurately
located ‘populations’, rather than isolated specimens.

PHOSPHATIC CHALK STRATIGRAPHY

Litho stratigraphy. The phosphatic chalks of Picardy in northern France are pelletal chalks rich in
light-brown granules of phosphatized carbonate, many of which are of faecal origin (Cayeux 1939;
Willcox 1953; Tabataba'i 1977), and contain in excess of 5% P2Os. They occur in small groups of, or
isolated troughs up to 1 km in length, 250 m wide and 30 m deep in the soft white chalks of the
Santonian-early Campanian of northern France and southern England. These troughs, termed
cuvettes (e.g. de Grossouvre 1901), have an erosional origin and are floored by a well-developed
hardground, termed the basal hardground.

On top of the basal hardground there rests up to 1 5 m of phosphatic chalk which contains a prolific
and distinctive fauna at its base, including ‘ Terebella ’ phosphatica Leriche (an agglutinated worm
tube) and Diblasus arborescens Parent (a compound coral) (Jarvis 1980). The final phosphatic
chalk development is no younger than early Campanian in age and usually contains distinctive bands
of Offaster pilula Lamarck and G. quadrata quadrata (Blainville). The genesis of these deposits is
discussed elsewhere (Jarvis 1980). Despite the relative abundance of fauna at certain levels within
phosphatic chalks, little information has been published on their macrofauna, although short faunal

| Palaeontology, Vol. 23, Part 4, 1980, pp. 889-914, pis. 1 15-1 16.|

890

PALAEONTOLOGY, VOLUME 23

lists and descriptions are provided by Leriche (1908, 1911). The common occurrence of G. ex gr.
quadrata in the upper portion of phosphatic chalks has been noted previously by several authors
(Lasne 1902; Gosselet 1901; de Grossouvre 1894, 1899, 1901, 1907; Leriche 1908, 1911; Jarvis 1980).

A large ‘population’ of over 200 guards was collected by the author in 1977-1978 from the
abandoned phosphorite quarry near Hardivillers (Oise). A further 100 guards were collected from
Ribemont (Aisne) and Villers-devant-le-Thour (Ardennes) quarries, together with additional
comparative material from several other phosphatic chalk localities. All of the material utilized in the
present study (deposited in the Oxford University Museum — OUM) was collected, where possible, in
situ and carefully localized by reference to the lithostratigraphy. The location [Lambert coordinates
provide an east-west (x) and north-south (y) position ( + 50 m) on a standard map-grid, plus the
height (z) above sea-level ( + 5 m)] and lithostratigraphy (text-fig. 1) of the three quarries is described
briefly, and in addition details are provided for Beauval quarry (Somme), since this provides the most
complete extant example of the phosphatic chalk lithofacies, and is the source of much of the
comparative material.

Hardivillers quarry (Oise) x 593,22 y 213,74 z 120. A large complex of abandoned quarries lying 1-5 km north-
east of the village of Hardivillers, north of the N30 between Hardivillers and Breteuil. Although mentioned by a
number of authors (Buteux 1849; Lasne 1890, 1892; de Mercey 1887; Meunier 1891; de Grossouvre 1901, 1907;
Tabatabai 1977), little has been published on the succession. The quarries expose a 4-12-m-thick bed of
phosphatic chalk resting on top of a strongly indurated and mineralized basal hardground. The succession may
be divided conveniently into three units: a lower white chalk, a phosphatic chalk, and an upper white chalk. Two

text-fig. 1 . Location and lithostratigraphy of phosphatic chalk localities referred to in the text.
The solid lines are the limit of the Upper Cretaceous outcrop.

JARVIS: UPPER CRETACEOUS BELEMNITES

891

major biostratigraphical marker horizons are present within the phosphatic chalk: a lower 30-cm bed of
abundant O. pilula, termed the Hardivillers OfFaster Bed, and an upper 1-m-thick bed with abundant G. q.
quadrata , termed the Hardivillers Gonioteuthis Bed. The bulk of material considered in this paper originates
from the latter bed which yielded 270 specimens (OUM KZ6001-KZ6270), of which 136 were complete
guards.

Ribemont quarry (Aisne) x 193,47 y 340,63 z 90, A small, intermittently worked quarry, 2 km south-east of
Ribemont village. The quarry is situated off the minor road leading to ‘la Ferme a Chaux’, south of the D12
which links Ribemont to Villers-le-Sec. The site was described by Rabelle (1893, 1902) who noted the abundance
of G. ex gr. quadrata in the upper part of the section. The locality has never been worked for phosphorite but
exposes three thin beds of phosphatic chalk intercalated within the soft, white, flintless chalk which forms the
bulk of the sequence. The ‘population’, which originates from the uppermost phosphatic chalk, consists of 174
specimens (OUM KZ6281-KZ6455), of which seventy-three are near-complete guards. The sediment log (text-
fig. 1) illustrates the considerable relief (up to 1.5 m) on the Ribemont Gonioteuthis Hardground, a strongly
lithified and mineralized hardground which underlies the uppermost phosphatic chalk. The hardground has a
bow-shaped cross-section in the central upper portion of the quarry face, which is interpreted as a
synsedimentary depression in the surface of the hardground.

Villers-devant-le-Thour quarry (Ardennes) x 725,35 y 201,50 z 100. An intermittently worked quarry T5 km
west-south-west of Villers-devant-le-Thour, on the south side of the D 1 8 which joins the village to the N366. The
locality is referred to by de Grossouvre (1901, p. 126), Broquet (1973), and Guerin, Maucorps, Solau, and
Pomerol (1977) but no details are given. The exposure consists of 8-5 m of soft white chalk (text-fig. 1), which
includes a 1-m-thick bed of phosphatic chalk towards its top. The phosphatic chalk contains abundant oyster
and fish debris, frequent pectinids, and Gonioteuthis. The bed cuts down and thickens to nearly 2 m towards the
eastern side of the quarry. Here, at its base, a 50-cm unit of large (up to 10 cm) phosphatized intraclasts and
abundant G. ex gr. quadrata provides the source of the sixty nine guards (OUM KZ6480-KZ6549), including
twenty four near-complete examples, analysed in this paper.

Beauval quarry (Somme) x 599,89 y 266,80 z 130. A large quarry on the east side of the N16, 6 km due south of
Doullens and on the east side of Beauval village. Beauval is probably the best documented of all phosphatic
chalk localities (Buteux 1849; Meunier 1888; de Mercey 1890; Lasne 1890, 1892, 1902; de Grossouvre 1901;
Briquet 1902; Negre 1912, 1963; Tabataba'i 1977; Jarvis 1980) but despite the wealth of literature, little
stratigraphical information is available, except for Tabatabai’s (1977) foraminiferal zonation. The site displays
two major levels of phosphatic chalk (text-fig. 1) both resting on top of well-developed hard grounds. The upper
phosphatic chalk contains the Beauval OfFaster Bed and Beauval Gonioteuthis Bed, similar to those seen at
Hardivillers, towards its top. Specimens of Gonioteuthis have been collected from throughout the sequence, but
are scarce except in the phosphatic chalk which directly overlies the upper basal hardground and the Beauval
Gonioteuthis Bed itself. Insufficient well-preserved material was available for statistical analysis, but the site
provides important comparative material.

Other sites. Three other specimens are included in the comparative diagrams. One (KZ6601) originates from a
minor phosphatic chalk intercalated within the upper white chalk at Faucouzy quarry (Aisne) (x 69 1 ,75 y 233,63
z 129). The other two (KZ6788, KZ6801) come from the phosphatic chalk which directly overlies the basal
hardground at Nurlu quarry (Somme) (x 647,65 y 254,60 z 140).

Belemnite biostratigraphy. Four genera of belemnite occur in the Santonian-early Campanian
deposits of the Anglo-Paris Basin. These are Actinocamax Miller, 1829, Belemnellocamax Naidin,
1964, Gonioteuthis Bayle, 1879, and Belemnitella d’Orbigny, 1840. Of the four genera, Belemnitella is
restricted to one species ( B . praecursor Stolley), Belemnellocamax to one group ( B . ex gr. grossouvrei
(Janet)), and Actinocamax also to one species (A. verus Miller). The genus Gonioteuthis, on the other
hand, is represented by an evolutionary lineage of six species and is consequently the most
stratigraphically useful of the four genera.

Gonioteuthis has been studied in detail by Stolley (1897, 1916, 1930), Ernst (1963a, b, 1964, 1966,
1968), Ernst and Schultz (1974), and Christensen (1971, 1973, 1975a, b). The genus includes the
evolutionary lineage G. westfalica (Schliiter) (oldest), G. westfalicagranulata (Stolley), G. granulata
(Blainville), G. granulataquadrata (Stolley), G. quadrata quadrata (Blainville), and G. q. gracilis
(Stolley) (youngest), and is an outstanding example of phyletic gradualism (Christensen 19766). The

892

PALAEONTOLOGY, VOLUME 23

Gonioteuthis stock extended from the middle Coniacian to the top of the early Campanian, a period
of some 10 million years (Van Hinte 1976). The genus shows three main trends during its evolution:

(1) Progressive calcification of the anterior portion of the guard, which evolves from a convexly conical, flat,
or shallow alveolus in G. westfalica to a deep pseudoalveolus constituting up to one-third of the length of the
guard in specimens of G. q. quadrata.

(2) The development of granulation, which is non-existent or poorly developed in specimens of G. westfalica,
but is pronounced in G. granulata and stratigraphically younger species.

(3) Increasing size and stoutness of the guard, which reaches a maximum with G. granulataquadrata and early
forms of G. q. quadrata.

G. westfalica is further isolated from the other species by its greater variation in guard shape and
oval to pointed anterior cross-section, as compared to sub-rectangular to sub-quadrate in later
species. The evolution of G. q. gracilis during the latest early Campanian marks a reversal of some of
the general trends, with the return of slimmer, shorter guards and more shallow pseudoalveoli. The
species does, however, remain distinct by the continued prominence of granulation and the presence
of a notched pseudoalveolus (Ernst in Christensen 1975a, p. 37).

De Grossouvre (1894, 1899, 1901, 1907) suggested a threefold division of French phosphatic
chalks based on his observations at Hardivillers. He dated the lowest phosphatic chalk at that locality
as early Santonian ( Micraster coranguinum Zone), a conclusion which is consistent with the
foraminiferal evidence (Biozone e, Tabatabai 1977). His lower unit was typified by M. coranguinum
(Leske) and A. verus, the middle unit by G. ex gr. quadrata and O. pilula, and the upper unit by G. ex
gr. quadrata and B. ‘ mucronata ’. The overlying upper white chalk also contains the latter two species
together with M. pseudoglyphus de Grossouvre (de Grossouvre op. cit.).

I have confirmed this general classification, with some additional details. G. granulata has
previously been identified from a small number of localities (Leriche 1908) and certainly this
belemnite is present in the lowest phosphatic chalk at Beauval and Nurlu; furthermore, fragments of
Gonioteuthis have been collected from a similar level at Hardivillers. De Grossouvre’s record of B.
mucronata (Schlotheim) is regarded as a misidentification of B. praecursor (PI. 11 5, figs. 1 -3, 9), which
forms a minor element of the belemnite fauna in the Hardivillers Gonioteuthis Bed (3%) and in the
uppermost phosphatic chalk at Ribemont (4%).

A. verus (PI. 115, figs. 10-13) is the commonest belemnite in the lowest phosphatic chalk at
Beauval, Hardivillers, and Nurlu, but proportions and relative abundancies vary. At Nurlu fifty
fragments and seven complete A. verus (KZ 6733-6738, KZ6763; PI. 115, figs. 10, 11), and a well-
preserved guard of G. granulata (KZ6788; PI. 1 1 5, fig. 7) were collected from the lag on top of the basal
hardground, yet at Beauval despite the larger fauna collected, no belemnites were found in the basal
lag. In the lowest phosphatic chalk at Beauval, however, five specimens of A. verus (KZ6552-6556;
PI. 115, figs. 12, 13) and one G. granulata (KZ6551; PI. 115, figs 4-6, 8) were recovered.
Belemnellocamax ex gr. grossouvrei has been described from phosphatic chalks (de Grossouvre 1894,
1899, 1901, 1907; Leriche 1908, 1911), but despite the collection of several hundred belemnites, no
examples of this species were recovered by the author. De Grossouvre (op. cit.) suggests that this
belemnite is typical of the lower portions of phosphatic chalks.

BIOMETRY OF GONIOTEUTHIS FROM PHOSPHATIC CHALKS
Statistical methods

The variation within belemnite ‘populations’ and their identification has been based on a series of
univariate and bivariate statistics, histograms, and scattergrams similar to those applied by
Christensen (1970, 1971, 1973, 1974, 1975a, 1976a) and Christensen, Ernst, Schmid, Schulz, and
Wood (1975). Most statistical parameters were calculated utilizing an SPSS (Statistical Package for
the Social Sciences) version 7 package on an ICL 2980 computer at the University of Oxford.

Guard morphology. The following characters (text-fig. 2) were measured: total length of guard (L), depth of the
pseudoalveolus (D), dorso-ventral diameter at the alveolar end (DVDAE), lateral diameter at the alveolar end

JARVIS: UPPER CRETACEOUS BELEMNITES

893

text-fig. 2. Diagram showing the morphological
elements of the Gonioteuthis guard, a, ventral view.
B, left lateral view of a ground guard, c, cut-away
dorsal view. MLD = maximum lateral diameter;
LDAE = lateral diameter at the alveolar end; D =
depth of the pseudoalveolus; L = length of guard;
DVDAE = dorso-ventral diameter at the alveolar
end.

MLD— |

mldae;, ,ovoae.

pseudoatveolus

c

(LDAE), and maximum lateral diameter (MLD). Measurements were made with a vernier caliper to an accuracy
of 0- 1 mm. A number of other parameters, e.g. length of ventral fissure, have been measured by previous authors,
but these have been found to be of little taxonomic value and have therefore been omitted. A small number of
specimens were split to study the internal characteristics of the guard (method in Christensen 1971, p. 370), but
since internal characters are of limited diagnostic use in the genus Gonioteuthis, no measurements were made on
split material.

Univariate analysis. The following statistics were estimated: arithmetic mean (X), standard deviation (SD), and
coefficient of variation (CV). Histograms of two of the five characters (L, D) are shown in text-fig. 3. The
frequency distributions were tested for normality using the Kolmogorov-Smirnov one sample test for goodness
of fit. Clearly the univariate statistics of ‘size’ parameters can be effected by a large number of factors including
sampling bias and post-mortem sorting; furthermore, a sample often contains an indeterminate number of
juveniles (cf. Kermack 1954, p. 391) and in belemnites, as in Recent coleoids (Cousteau and Diole 1973, p. 93),
there are no criteria for determining the ontogenetic stage of an individual (Christensen 1 915a). Nevertheless, the
values may be of interpretative value, although bivariate statistics are regarded as being of greater taxonomic
significance.

Ratios. Ratios of ‘size’ parameters have been widely used in the study of belemnites. Ernst (1963a, b, 1964, 1966,
1968), in particular, characterized his samples of Gonioteuthis by mean values of various ratios, the most
diagnostic of which were the Riedel-Quotient (ratio of length of guard to depth of pseudoalveolus) and the
Schlankheits-Quotient (ratio of length of guard to dorso-ventral diameter at the alveolar end). Numerous
authors (e.g. Shaw 1956; Simpson, Roe, and Lewontin 1960; Sokal 1965; Sokal and Rohlf 1969) have criticized
the use of ratios. The main objections are that a ratio is a secondary statistic with greater variance than either of
its components, that ratios may not be normally distributed, and lastly that if the relationship between the two
characters is allometric, the ratio will change during growth. Despite these mathematical limitation, Ernst’s
Gonioteuthis stratigraphy based on mean Riedel-Quotient is generally valid, since the relationship between
length of guard and depth of pseudoalveolus is isometric in nearly all samples of Gonioteuthis (Christensen
1975a, b), and furthermore the ratio has been found to be approximately normally distributed in the present
study. Consequently, Riedel-Quotients (RQ) are reported for each sample.

text-fig. 3. Histograms of length of
guard (L) and depth of pseudo-
alveolus (D) of Gonioteuthis from the
Hardivillers Gonioteuthis Bed.

r

a

SO 60 7-0

CL

HARDIVILLERS

894

PALAEONTOLOGY, VOLUME 23

Bivariate analysis. The five ‘size’ parameters and the Riedel-Quotients were correlated utilizing Pearson
Correlation Matrices. Regression analysis was used to study the relationship between variable pairs during
growth. Regression lines were fitted by the least-squares method primarily because of the ability to compare the
slope and intercept of different samples (see Christensen 1973, pp. 115, 116 for discussion). The regression line is
written y = a + bx and the original measurements were used because of their rectilinear trend on arithmetically
scaled scatter-plots. The following statistical parameters were calculated: the percentage of variance explained
by the linear relationship ( r 2), the standard error of the y-intercept (SEa), the standard error of the slope (SE*),
the standard error of the regression line (SE^), and the value of t (ta), and the associated probability was
calculated by /-testing the intercept on the y-axis to determine whether the intercept differed significantly from
zero (Hald 1957). This final test has important biological implications since only a regression line passing
through the origin represents isometric growth (i.e. y = bx). Other possible equations (y = a + bx\ y = bxa)
represent allometric growth (the latter equation is often referred to as simple allometry, e.g. Christensen 1975a).
The regression lines were compared with each other, and with comparable ‘populations’ from Germany, by the
methods described by Hald (1957, pp. 571-579).

Results

The results of the univariate statistical analyses of the three ‘populations’ are given in Table 1.
Although histograms of the various ‘size’ parameters show a slight assymmetry (text-fig. 3), statistical
analysis utilizing the Kolmogorov-Smirnov one-sample test (Table 2) shows that all characters in all
three ‘populations’ correspond well to a normal distribution. It was noted, however, that the
probability associated with the length of guard for the Ribemont ‘population’ was considerably
lower than that for the other two samples. Although no further specimens were available for detailed
measurement, two parameters, length of guard and maximum lateral diameter, could be measured to
a lower precision (+ 0-25 and + 0-05 cm respectively) on a further forty-three guards. The resulting
histograms (text-fig. 4) are strongly bimodal, a K-S test giving D statistics of 0T958 (P = 0-03) and
0-1891 (P = 0-04) respectively, both of which are significant at the 0-05 level. The mean length of the
guards studied in detail from Ribemont is 6-30 cm, a value which lies in the trough between the two
modes of the larger sample, indicating that the ‘population’ studied in detail is a mixture of two
distinct components.

EXPLANATION OF PLATE 1 15

Figs. 1-3. Belemnitella praecursor Stolley, from the Hardivillers Gonioteuthis Bed, early Campanian, Off aster
pilula Zone. The anterior portion of the guard KZ6272 is missing. 1 —dorsal, 2— left lateral, 3— ventral views,
all x 1 .

Figs. 4-6 Gonioteuthis granulata (Blainville). Lower phosphatic chalk, Beauval, early-middle Santonian,
Micraster coranguinum Zone. The specimen is coated in a thin shiny phosphate patina, typical of basal lag
preservation. Granulation is minimal. KZ6551. 4— dorsal, 5— left lateral, 6— ventral views, all x 1.

Fig. 7. Gonioteuthis granulata (Blainville). Base of phosphatic chalk, Nurlu, early-middle Santonian, M.
coranguinum Zone. Anterior end of specimen KZ6788 with a pronounced rhombohedral anterior cross-
section. Note the concentric growth rings, x 1-5.

Fig. 8. Gonioteuthis granulata (Blainville). Anterior view of KZ6551 (figs. 4-6). The guard has a circular
cross-section. Note the shallow pseudoalveolus, x 1-5.

Fig. 9. Belemnitella praecursor Stolley. Uncoated right lateral view of KZ6272 (figs. 1 -3). The specimen is
honeycombed by Entobia cretacea Portlock, a clionid sponge boring. The fine (0T 5-0-50 mm) surface pores
can be seen in figs. 1-3, x 1.

Figs. 10-13. Actinocamax verus Miller. Base of phosphatic chalk, Nurlu (10, 11) and lower phosphatic chalk,
Beauval (12, 13). Early-middle Santonian M. coranguinum Zone. 10, KZ6738, left lateral view of a juvenile
guard, x 2. 11, KZ6736, left lateral view of an adult guard, x 2. 12, left lateral and 13, ventral views of a large
specimen, KZ6556. Note surface wrinkling and tapering, pyramidal anterior termination to the guard, x 2.

PLATE 115

JARVIS, Cretaceous belemnites

896

PALAEONTOLOGY, VOLUME 23

All three samples, like Gonioteuthis ‘populations’ collected from other facies (e.g. Christensen
1974, p. 5; 1975, p. 32), are presumably an accumulation of several generations and consist of a
growth series which both juveniles and adults (PI. 116, figs. 1-15). The approximation of the length of
guard size-distributions to normality results in mean length roughly corresponding to the size which
most specimens had reached when they died. This is not true of the Ribemont sample, which shows
two mortality peaks— around 5 0 and 6-8 cm. The second of these two maxima approximates to the
mortality peaks of the ‘populations’ from Hardivillers and Villers-devant-le-Thour (Table 1) and
from the early Campanian O. pilula Zone of south-west Miinsterland (Ernst 1964, p. 126); material
from Hover (Lower Saxony) (Ernst 1 964, p. 132) has a lower mean length, approximately 6 cm, at this
level. Comparisons with Ernst’s results must be treated with caution, however, since he sorted out
specimens of less than 4 cm length.

table 1. Univariate analyses of the ‘size’ parameters of three
‘populations’ of Gonioteuthis from phosphatic chalks.

HARDIVILLERS

Character

N

X

<j

cv

OR

L

80

6-537

0-701

10-72

5-00-8-67

D

98

1-475

0-257

17-42

0-93-2-26

DVDAE

94

1-032

0-158

15-31

0-71-1-45

LDAE

94

0-955

0-148

15-50

0-63-1-37

MLD

86

1-099

0-168

15-29

0-741-61

RQ

79

4-437

0-531

11-97

3-38-6-08

RIBEMONT

Character

N

X

a

CV

OR

L

12

6-303

0-678

10-76

5-43-7-21

D

10

1-608

0-322

20-02

1-15—2-1 1

DVDAE

5

1-162

0-134

11-53

1-05-1-38

LDAE

9

1-091

0-140

12-83

0-87-1-29

MLD

12

1-148

0-159

13-85

0-90-1-40

RQ

10

4-038

0-539

13-35

3-33-4-92

VILLERS-DEVANT-LE-THOUR

Character

N

X

a

CV

OR

L

8

6-749

0-603

8-93

5-94-7-62

D

10

1-649

0-300

18-19

1-23-2-05

DVDAE

9

1-256

0-196

15-61

0-91-1 52

LDAE

10

1-134

0-164

14-46

0-83-1-35

MLD

9

1-194

0-155

12-98

0-97-1-39

RQ

8

4-074

0-445

10-92

3-59-4-88

The smallest guard for which detailed measurements were obtained comes from Hardivillers and
has an over-all length of 5 cm (PI. 1 16, figs. 5-7); the smallest specimens from Ribemont lie within the
range 3-5-4 0 cm. A number of guards with damaged pseudoalveoli and lengths of < 5 cm were also
collected from Hardivillers, but the proportion of individuals within this size range is small. The
rarity of small guards in most samples and the total absence of guards of less than 3 -5 cm length may
be attributed to the interaction of low juvenile mortality rate and high initial growth-rate. Such an
interaction will also control the shape of the size-frequency distributions (Craig 1967; Surlyk 1972).

JARVIS: UPPER CRETACEOUS BELEMNITES

897

table 2. Results of the Kolmogorov-Smirnov test for goodness of fit to a normal
distribution for three ‘populations’ of Gonioteuthis from phosphatic chalks.

HARDIVILLERS

Character

N

D

Probability

L

80

00620

0-92

D

98

0-0783

0-59

DVDAE

94

0-0777

0-62

LDAE

94

0-0863

0-49

MLD

86

0-0559

0-95

RQ

79

0-1033

0-37

RIBEMONT

Character

N

D

Probability

L

12

0-2480

0-45

D

10

0-1384

0-99

DVDAE

5

0-2175

0-97

LDAE

9

0-1984

0-87

MLD

12

0-1075

1-00

RQ

10

0-1701

0-94

VILLERS-DEVANT-LE-THOUR

Character

N

D

Probability

L

8

0-1964

0-92

D

10

0-1324

1-00

DVDAE

9

0-1075

1-00

LDAE

10

0-0933

1-00

MLD

9

0-1395

1-00

RQ

8

0-2534

0-68

RIBEMONT

text-fig. 4. Histograms of length of guard (L) and maximum lateral diameter
(MLD) of Gonioteuthis from the phosphatic chalk overlying the Ribemont
Gonioteuthis Hardground.

898

PALAEONTOLOGY, VOLUME 23

The means of ‘size’ parameters of samples of Gonioteuthis have been used successfully for the analysis
of time trends in evolutionary lineages (Ernst 1964), and similar studies have been applied to A. verus
(Reyment and Naidin 1962), but such studies are only viable with normally distributed characters.
Since the mortality peak for early Campanian Gonioteuthis is apparently between 6-5 and 6-8 cm
length, the lower mortality peak of the Ribemont sample requires explanation. There are several
possible interpretations which include:

(1) the peaks represent different species;

(2) ‘size’ parameter distributions are normally bimodal for ‘populations’ of Gonioteuthis-,

(3) the ‘population’ has been current sorted or;

(4) the bimodality is the result of a catastrophic event.

There is no morphological evidence to support the occurrence of more than one species of
Gonioteuthis within the Ribemont ‘population’; however, stratigraphically close but specifically
distinct Gonioteuthis ‘populations’ would, once mixed, be impossible to separate on morphological
criteria alone. No Gonioteuthis ‘population’ has been described with a mean length as low as 5 cm and
with appropriate mean Riedel-Quotient, but it should be noted that the value for G. q. gracilis from
Misburg (Ernst 1964) does approach this value. Thus the possibility of specific mixing in the
Ribemont phosphatic chalk cannot be totally dismissed although it is unlikely in my opinion.

Bimodality of ‘size’ parameters may result from annual or biannual mortality peaks of a species,
but since no other ‘populations’ exhibit a bimodal distribution it must be concluded that such an
explanation is invalid. Current sorting may lead to a reduction rather than an enhancement of the
proportion of juveniles within a population, but transport of smaller individuals into the area of
deposition remains a possibility. Bimodal distributions of this type frequently result from
catastrophic events which cause the death of abnormally young ‘populations’, which are overprinted
on the normal mortality curves. A storm deposit is the commonest example although Red-Tides
provide a further possibility; the concentration of fauna into what is apparently a depression in the
hardground accords most favourably with the storm-event hypothesis.

On the basis of univariate analysis the samples of Gonioteuthis from Hardivillers and Villers-
devant-le-Thour may be considered homogenous, while the ‘population’ from Ribemont is
heterogenous. Consequently, interpretations made from the bivariate analysis of the small number of
complete specimens available from this latter locality can only be regarded as tentative.

Ratios. In general, ratios have not been utilized in this study but the validity of the Reidel-Quotient is
recognized (see above) as a method of species discrimination in specimens of Gonioteuthis. Mean
values of the Quotient for all three ‘populations’ (Table 1) lie within the range of values for
G. q. quadrata (Ernst 1964, 1968), but individual specimens from Hardivillers (PI. 156, figs. 1-15) may
be referred to G. granulata (one specimen only), G. granulataquadrata, G. q. quadrata , or G. q. gracilis.
Specimens from the other two localities may all be identified as either G. granulataquadrata or
G. q. quadrata (or possibly G. q. gracilis ). This range of possible ‘species’ within a single bed illustrates
the importance of the study of ‘populations’ rather than individuals for the reliable identification of
Gonioteuthis species.

Regression analysis. All five ‘size’ parameters and the Riedel-Quotient were compared by means of
Pearson correlation matrices (Table 3). The correlation coefficients for the Hardivillers ‘population’
were all highly significant (P < 0 001) except for the correlation between L and RQ which was less
significant (P < 0-03). Probabilities associated with matrices from the other localities were generally
higher, although all correlations remained significant at the 0 05 level in the Villers-devant-le-Thour
sample, and only probabilities associated with RQ were not significant at this level for the Ribemont
sample.

Since all five ‘size’ parameters are very closely related, four regression analyses were made for each
‘population’. The plots chosen follow the useage of Christensen (1975a):

( 1 ) length of guard (x) versus depth of pseudoalveolus (y);

(2) length of guard (x) versus dorso-ventral diameter at the alveolar end (y);

JARVIS: UPPER CRETACEOUS BELEMNITES

899

(3) dorso-ventral diameter at the alveolar end (x) versus lateral diameter at the alveolar end (y);

(4) maximum lateral diameter (x) versus lateral diameter at the alveolar end (y).

With three exceptions (Table 5) the relationships of the characters in the twelve analyses were
isometric.

Length of guard (x) versus depth of pseudoalveolus (y). A regression analysis of these two characters
provides an analogous but more comprehensive quantity than Ernst’s Riedel-Quotient and is
specifically the most diagnostic of the four regression analyses, partly because of the greater number
of comparative analyses in the literature (e.g. Christensen 1971, 1973, 1975a). Christensen (1975a)
has published regression analyses of Ernst’s (1964, 1968) original measurements of six species of
Gonioteuthis. The statistical parameters for these six ‘populations’ are given in Table 4 and the
regression lines are plotted on the relevant scattergrams (text-figs. 5-7).

Hardivillers. The statistical parameters are given in Table 5 and the values plotted on text-fig. 5a. The
‘population’ was compared to G. granulataquadrata and was found to differ significantly in slope
(0 01 >P> 0-001). A comparison with G. q. quadrata from the early Campanian Inoceramus ex gr. lingua-G. ex
gr. quadrata Zone of Hover gave a highly significant correlation between slopes (0-70 > P > 0-60), but the
position of the lines are different (P < 0-001). However, when compared with a stratigraphically younger
‘population’ of the same species (from the Echinocorys conica-Galeola papillosa Zone), the variances, slopes

table 3. Pearson correlation matrices of the ‘size’ parameters
of three ‘populations’ of Gonioteuthis from phosphatic chalks.

HARDIVILLERS

L

D

DVDAE

LDAE

MLD

RQ

L

1-000

_

_

D

0-763

1 ■()()()

_

_

_

DVDAE

0-839

0-808

1 '()()()

—

—

—

LDAE

0-825

0-810

0-980

1-000

—

MLD

0-814

0-711

0-955

0-940

1-000

RQ

-0-219

-0-786

-0-434 -

-0-426

-0-334

1 DOO

RIBEMONT

L

D

DVDAE

LDAE

MLD

RQ

L

1-000

_

_

_

_

D

0-773

1 ■()()()

—

—

—

—

DVDAE

0-870

0-923

1-000

—

—

—

LDAE

0-817

0-785

0-985

1-000

—

—

MLD

0-800

0-631

0-992

0-952

1-000

—

RQ

-0-344

-0-850

-0-877 -

-0-530

-0-287

1-000

VILLERS-DEVANT-LE-THOUR

L

D

DVDAE

LDAE

MLD

RQ

L

1-000

_

_

_

_

D

0-916

1 ■()()()

—

—

—

—

DVDAE

0-720

0-854

1-000

—

—

—

LDAE

0-794

0-900

0-970

1-000

—

—

MLD

0-916

0-900

0-946

0-959

1-000

—

RQ

-0-717

-0-931

-0-853 -

-0-943

-0-906

1-000

900

PALAEONTOLOGY, VOLUME 23

table 4. Statistical relationship between depth of pseudoalveolus and length of guard for six species of
Gonioteuthis from Germany, y = a + bx (modified from Christensen 1975a).

G. westfalica westfalica from the lower ‘westfalica beds’, Essen- Vogelheim (Ernst 1964a, p. 118; Christensen
1975a, p. 38)

D = 0-1 150 + 0-0597 L; N = 196; r = 0-2826; r2 = 7-99%; SEa = 0-0828; SE* = 0-0145; SE^ = 0-1089; ta =

1- 3882 (0-20 >P> 0-10)

G. westfalicagranulata from Biilten (Ernst 1968, p. 278; Christensen 1975a, p. 38

D = 0-0102 + 0-1106 L; N = 51; r = 0-4809; r2 =23-13%; SEa = 0-1611; SE* = 0-0279; SE^ = 0-1146; ta =
0-0633 (P> 0-90)

G. granulata from Gleidingen (Ernst 1968, p. 278; Christensen 1975a, p. 38)

D = 0-0701 +0-1299 L; N = 45; r = 0-7311; r2 = 53-45%; SEa = 0-0920; SE* = 0-0185; SE^ = 0-1083; ta =
0-7620 (0-50 >P> 0-40)

G. granulataquadrata from Weinberg (Ernst 1968, p. 278; Christensen 1975a, p. 38)

D = 0-1030 + 0-1760 L; N = 45; r = 0-7404; r2 = 54-82%; SEfl = 0-1623; SE* = 0-0244; SEr* = 0-1370; ta =
0-6342 (0-60>P>0-50)

G. quadrat a quadrata from the I. ex gr. lingua-G. ex gr. quadrata Zone of Hover (Ernst 1964a, p. 119; Christensen
1975a, p. 39)

D = -0-0888 + 0-2685 L; N = 24; r = 0-7982; r2 = 63-71%; SEa = 0-2640; SE* = 0-0432; SEyx = 0-2166; ta =
0-3364 (0-80 >P> 0-70)

G. quadrata quadrata from the E. conica-G. papillosa Zone of Hover (measurements by Ernst, statistics
after W. K. Christensen, pers. comm.)

D = -0-3776 + 0-2917 L; N = 65; r = 0-7649; r2 = 58-51%; SEa = 0-1772; SE* = 0-0309; SE^ = 0-1711; ta =

2- 0977 (0-05>P>0-02)

G. quadrata gracilis from the Germania IV quarry (north Germany) (Christensen 1975a, p. 42)

D = -0-7641+0-3620 L; N = 47; r = - ; r2 = - ; SEa = 0-1820; SE* = -0-0318; SE^ = 01 152; ta = 4-1975
(EcO-OOl)

(P > 0-90) and positions (0-40 > P > 0-30) of the lines were found to be the same, similar results were obtained
from a comparison with a ‘population’ of G. q. quadrata from the ‘Smectite’ of Hallembaye quarry in eastern
Belgium (W. K. Christensen, pers. comm.) where G. q. quadrata also occurs with B. praecursor. When compared
to G. q. gracilis, the variances were found to differ significantly (F = 2-2422 with 77 and 45 degrees of freedom;
P = < 0-01), so the test for non-equal variances was used (Hald 1957). The correspondence between the slopes of
the two lines was found to be slightly significant (0- 10 > P > 0-05 with 106 degrees of freedom) so the positions of
the lines were also tested. The test gave a probability of 0-05 > P > 0-02 with 1 22 degrees of freedom, which is not
significant. Clearly the regression line compares most closely with that for G. q. quadrata from the Hallembaye
‘Smectite’ and the German E. conica-G. papillosa Zone. It is noteworthy that a /-test on the ^-intercept for the
younger G. q. quadrata ‘population’ from Germany, G. q. quadrata from Belgium, and the Hardivillers
‘populations’ gives significant values (0-05 > P > 0-02), indicating an allometric relationship between the
characters as seen in G. q. gracilis (Christensen 1975a).

EXPLANATION OF PLATE 116

Figs. 1-15. Gonioteuthis quadrata quadrata (Blainville) from the Hardivillers Gonioteuthis Bed, early
Campanian OJfaster pilula Zone. 1-3, a medium-sized guard of average shape. KZ6040, dorsal, left lateral,
and ventral views, x 1 . 4, split anterior end of KZ6099 showing conellae, x 1 . 5-7, an adolescent guard,
KZ6033, the smallest complete specimen recovered from the bed, dorsal, left lateral, and ventral views, x 1 . 8,
anterior end of KZ6066 (figs. 13-15) showing the depth of the pseudoalveolus, x 1-5. 9, anterior end of
KZ6040 (figs. 1-3) x 1-5. 10-12, the largest specimen collected, KZ6039, dorsal, left lateral, and ventral views,
x 1. 13-15, the stoutest individual in the ‘population’, KZ6066, dorsal, left lateral, and ventral views, x 1.

PLATE 1 1 6

JARVIS, Cretaceous belemnites

902

PALAEONTOLOGY, VOLUME 23

LDAE

JARVIS: UPPER CRETACEOUS BELEMNITES

903

I'M 0N3 dVIEBAIV 3K1 IV a313WVIO 1Va31V1

t“>) QN3 aVIOSAIV 3H1 XV asiswvio iva3ivi

text-fig. 6. Scatter diagrams and regression lines for G. ex gr. quadrata from Ribemont and Villers-devant-le-Thour. The plots follow the same

format as text-fig. 5.

904

PALAEONTOLOGY, VOLUME 23

LENGTH OF GUARD (cm)

text-fig. 7. Scatter diagrams of isolated specimens of Gonioteuthis collected from
phosphatic chalks, a, length of guard (L) versus depth of the pseudoalveolus (D);
the regression lines of the six German control ‘populations’ (excluding G. q.
quadrata from the E. conica — G. papillosa Zone) and G. q. quadrata from
Hardivillers are plotted for comparison, b, maximum lateral diameter (MLD)
versus lateral diameter at the alveolar end (LDAE); the regression lines for G. q.
gracilis from Germany and G. q. quadrata from Hardivillers are plotted for
comparison.

JARVIS: UPPER CRETACEOUS BELEMNITES

905

Ribemont and Villers-devant-le-Thour. A comparison between the slopes and positions of the regression lines of
the two ‘populations’ showed that they do not differ significantly ( P > 0-40). A comparison between the slopes of
the two lines and the slope of the Hardivillers ‘population’ also revealed no significant differences
(0-60 > P > 0-50; 0-20 > P > 010), but a comparison between the positions of the lines indicated that although
the Ribemont belemnites differed significantly from those from Hardivillers (0-01 > P > 0 001), the Villers-
devant-le-Thour and Hardivillers ‘populations’ are apparently the same (0T0 > P > 0-05). The two samples
were compared to three species of Gonioteuthis from Germany (text-fig. 6a) and were found to differ significantly
from G. granulataquadrata, but could correspond with either G. q. quadrata or G. q. gracilis both in slope and
position. The small number of specimens available from these two localities makes the statistical data
inconclusive. Certainly the ‘populations’ can be assigned to G. ex gr. quadrata, but subspecific identification
remains uncertain. Intuitively the ‘populations’ correspond more closely to stratigraphically younger
‘populations’ of Gonioteuthis (text-fig. 6a) than that from Hardivillers (text-fig. 5a).

Analytical summary and comparisons

Individual specimens from all three localities have Riedel-Quotients within the range of values typical
of G. granulataquadrata and G. q. quadrata, but the mean value for all three ‘populations’ is within the
range of G. q. quadrata. The high degree of scatter of values of length of guard versus depth of
pseudoalveolus has led to occasionally ambiguous results, but nevertheless the Hardivillers sample
shows the greatest similarity to ‘populations’ of G. q. quadrata from the Hallembaye ‘Smectite’ and
the E. conica-G. papillosa Zone of Germany. Unfortunately, no further ‘size’ parameter regression-
analyses have been published for ‘populations’ of G. q. quadrata so it has been impossible to
statistically compare other parameters with control ‘populations’. However, unpublished data
(W. K. Christensen pers. comm.) indicate that the Hardivillers sample may be distinguished from
E. conica-G. papillosa Zone ‘populations’ on plots of L v. DVDAE. When this additional parameter
is considered the Hardivillers sample is seen to be similar to a ‘population’ of G. q. quadrata from the
I. ex gr. lingua-G. ex gr. quadrata Zone of Ziegelei Bremer, Bottrop-Fuhlenbrock (S. Miinsterland)
(Ernst 1964). Unfortunately, no regression data are available for this ‘population’ so a more precise
comparison is impossible.

G. q. quadrata from Hardivillers can be distinguished from German ‘populations’ of
G. granulataquadrata on L v. D and L v. DVDAE (text-fig. 5c) and from G. q. gracilis on L v. D and
MLD v. LDAE (text-fig. 5d). It is noteworthy that the relationship between L v. D, L v. DVDAE and
MLD v. LDAE must be considered allometric in the sample. The Ribemont and Villers-devant-le-
Thour ‘populations’ cannot be distinguished from one another on any character or combination of
characters. Generally the values lie closer to those of G. q. gracilis than to those of G. q. quadrata from
Hardivillers. This is apparent in the plots of L v. D, DVDAE v. LDAE (text-fig. 6b) and MLD v.
LDAE (text-fig. 6d), while in the plot of L v. DVDAE the values are closer to those of
G. granulataquadrata. These observations demonstrate that the G. ex gr. quadrata from these two
localities have a guard morphology similar to G. q. gracilis but are stouter than either German G. q.
gracilis or G. q. quadrata from Hardivillers. It is suggested, therefore, that these ‘populations’ are
stratigraphically younger than those from Hardivillers but are still G. q. quadrata, albeit more
evolved forms.

Remaining complete specimens of Gonioteuthis which have been collected from phosphatic chalks
are plotted on text-fig. 7. In the plot of D v. L (text-fig. 7a) three specimens clearly lie outside the
scatter of values for the three ‘populations’ of Gonioteuthis from the facies (text-figs. 5 a, 6a). All three
specimens are identified as G. granulata on their Riedel-Quotients. Two of the three (KZ655 1 , PI. 115,
figs. 4-6, 8; KZ6801) come from the base of phosphatic chalk sequences, while the other (KZ6601)
originates from above the main phosphatic chalk at Faucouzy. Material from Beauval (BVL) on the
plot of D v. L comes from the upper portion of the upper phosphatic chalk and it all lies very close to
the regression line for the Hardivillers ‘population’. The additional examples from Hardivillers
(HDVL) originate from the upper white chalk. It is noteworthy that both lie in the same region of the
plot as the material from Ribemont and Villers-devant-le-Thour.

The plot of LDAE v. MLD (text-fig. 7b) shows a high degree of scatter between the two reference
regression lines. All of the material from Beauval was collected from the upper phosphatic chalk

906

PALAEONTOLOGY, VOLUME 23

table 5. Statistical results of regression analyses of four ‘populations’ of Gonioteuthis.
HARDIVILLERS GONIOTEUTHIS BED

y — a + bx

N

r 2

SEa

SE b

SEj*

ta

Probability

D

-0-3784 + 0-2870 L

79

58-20%

0-1821

0-0277

0-1725

2-0775

0-05 >P> 0-02

DVDAE =

-0-1974 + 0-1933 L

75

70-32%

0-0961

0-0147

0-0882

2-0538

0-05 >P> 0-02

LDAE =

0-0188 + 0-8978 DVDAE

90

96-11%

0-0202

0-0193

0-0290

0-9331

0-40>f>>0-30

LDAE =

0-0614 + 0-8264 MLD

81

88-35%

0-0377

0-0338

0-0510

1-6316

0-20>P>0-10

PHOSPHATIC CHALK ABOVE THE RIBEMONT GONIOTEUTHIS HARDGROUND

y=a + bx

N

r2

SEa

SE b

SE^

ta

Probability

D

-0-5653 + 0-3417 L

10

59-76%

0-6343

0-0991

0-2166

0-8913

0-40>P>0-30

DVDAE =

0-1310 + 0-1604 L

5

75-54%

0-3404

0-0527

0-0768

0-3849

0-80>P>0-70

LDAE =

0-1 894 + 0-7665 DVDAE

5

96-95%

0-0918

0-0786

0-0211

2-0629

0-20 >P> 0-10

LDAE =

0-0908 + 0-8526 MLD

9

90-55%

0-1231

0-1041

0-0459

0-7376

0-50 >P> 0-40

PHOSPHATIC CHALK, VILLERS-DEVANT-LE-THOUR

y = a + bx

N

r2

SEa

SE*

SE,,

ta

Probability

D ■

- 1-3240 + 0-4455 L

8

83-90%

0-5396

0-0797

0-1271

2-4536

0-05 >P> 0-02

DVDAE =

-0-4975 + 0-2556 L

7

51-88%

0-7578

0-1101

0-1476

0-6565

0-60>P>0-50

LDAE =

0-1007 + 0-8331 DVDAE

9

94-01%

0-1009

0-0795

0-0440

0-9985

0-40>P>0-30

LDAE =

-0-0287 + 0-9533 MLD

9

92-05%

0-1167

0-1059

0-0463

0-2459

0-90>P>0-80

G. q. gracilis, GERMANIA IV QUARRY, NORTHERN GERMANY (CHRISTENSEN 1975a)

y = a + bx

N

r2

SEa

SE*

SEyx

ta

Probability

D

-0-7641 +0-3620 L

47

_

0-1820

0-0318

0-1152

4-1975

P<0-001

DVDAE =

-0-2428 + 0-2003 L

47

—

0-0693

0-0121

0-0583

3-5054 0-01>P>0-001

LDAE =

0-0250 + 0-9064 DVDAE

47

—

0-0253

0-0278

0-0290

0-9885 0-40 >P> 0-30

LDAE =

0-0506 + 0-9055 MLD

47

—

0-0299

0-0339

0-0351

1-6910 0-10>P>0-05

except KZ6551 (PI. 115, figs. 4-6, 8) which is from the base of the lower phosphatic chalk. The
Hardivillers examples are from the upper white chalk. It is clear that the specimens from Nurlu
(KZ6788, PI. 115, fig. 7; KZ6801), which are both from the base of the phosphatic chalk, lie outside
the scatter of values for the three ‘populations’.

Scatterplots of DVDAE v. L and LDAE v. DVDAE do not have sufficient resolution to
differentiate species of Gonioteuthis on the small amount of comparative material available. These
comparisons highlight the necessity for large collections in the identification of Gonioteuthis species,
but they also show that material from the base of phosphatic chalks may be attributed to G. granulata
and that this species is readily distinguished on plots of LDAE v. MLD as well as Reidel-Quotient and
D v. L plots.

Factor analysis

Since factor analysis requires listwise deletion of missing values, only the ‘population’ from
Hardivillers was regarded as being of sufficient size to enable a statistically valid application of factor
analysis. A total of sixty-nine guards were utilized in the analysis, results of which are given in Table 6.

JARVIS: UPPER CRETACEOUS BELEMNITES

907

The first principal component accounts for 87-2% of the variation and possesses a strong positive
correlation with all of the ‘size’ parameters. It is interpreted as representing guard size, which is
consistent with similar results from biological studies (Blackith and Reyment 1 97 1 , pp. 147-1 50) and
for a ‘population’ of A.plenus from England (Christensen 1974). Clearly size is intimately related to
age in marine invertebrates such as belemnites, although food availability may become an overriding
factor in Recent coleoids (Mangold-Wirz 1963).

The second principal component accounts for 6-8% of the total variance. It shows a strong positive
correlation with D, a moderate negative relationship with MLD, and negative correlations of
approximately equal magnitude with DVDAE and LDAE. The factor is consequently interpreted as
representing the evolutionary stage reached by each individual within the total ‘population’. It
reflects the increasing depth of the pseudoalveolus combined with a progressive increase in the
slenderness of the guard from the maximum robustness reached in G. granulataquadrata.

The third principal component only allows for 4-6% of the observed variation. It shows a strong
positive correlation with L and a weak negative interrelationship of approximately equal magnitude
with the remaining characters. The factor may be interpreted as a shape effect which results in the
production of elongate guards. Alternatively, it may be due to an indeterminate taphonomic control,
perhaps sorting of material prior to deposition.

table 6. Eigenvalues and Eigenvector matrix from a principal component
factor analysis of G. q. quadrata from the Hardivillers Gonioteuthis Bed.

Eigenvector:

1

2

3

Eigenvalue:

Variance:

4-361

0-342

0-228

% total

87-2

6-8

4-6

cumulative

87-2

94-0

98-6

Character

L

0-902

0-043

0-429

D

0-860

0-492

-0-125

DVDAE

0-984

-0-099

-0-097

LDAE

0-974

-0-104

-0-116

MLD

0-943

-0-279

-0-077

Calculated from a ‘population’ of 69 guards.

TAPHONOMY AND DEPOSITIONAL ENVIRONMENT

Some degree of taphonomic bias will inevitably be present in any fossil assemblage. Consequently it is
necessary to examine and if possible remove or ‘allow for’ any bias before zoological or
stratigraphical conclusions can be reached. Furthermore, an assemblages’ taphonomy, together
with its associated sedimentology, may provide valuable insights into the depositional environment
of that assemblage.

The majority of guards from the Hardivillers Gonioteuthis Bed are in excellent preservation and
complete. A summary of the encrustation is given in Table 7. Although pycnodonteine oysters,
octocorals, and serpulids are the major encrusters, a small number of cemented foraminiferids were
present on some guards. Clionid sponge borings are present in many specimens (PI. 115, figs. 1-3, 9)
but they rarely form extensive networks. A second common form of boring consists of a radiating
pattern of small ( < 0T mm) ramifying bores which occur just below the surface of the guards. These
may be attributed to algae or fungi. Oysters are the dominant encruster (PI. 116, fig. 13) but only 5%
of specimens bear more than two individuals, and where larger numbers are present they are generally
small and have identical orientations of their hinge-lines. Only four specimens exhibit more than a

908

PALAEONTOLOGY, VOLUME 23

single generation of encrustation, octocorals and serpulids following the pattern displayed by the
oysters. The small size of the majority of the epifauna (oysters only reaching a few millimetres across)
suggests that it is mostly juvenile. Examination of all specimens, including fragmentary guards,
demonstrates that few, if any, show signs of mechanical abrasion, the poor preservation of the small
number of broken guards being due predominately to the activities of boring sponges. Many
belemnites display fine, subparallel scratch marks, (PI. 115, figs. 12, 13; PI. 116, fig. 13), generally
oriented in a dorso-ventral direction and probably produced by the rasping action of a marine
organism grazing the surface of the guards. Several guards show late-stage compactional effects,
including partial crushing of their pseudoalveoli and in situ fracturing, occasionally accompanied by
recementation of the dislocated fragments. Recent solution and partial decalcification, particularly in
the area of the pseudoalveolus, has occurred in some material, but in general the guards are complete.

Specimens from the other localities display similar features to those seen in the Hardivillers sample.
Most guards from Ribemont and Villers-devant-le-Thour are severely etched and corroded by
weathering due to the proximity of the phosphatic chalk to the soil. Consequently, the data on the
encrustation (Table 7) are less reliable for these sites. The guard from the basal lag at Nurlu
(KZ6788), and hardground-associated guards from Faucouzy and Hallencourt display thin, shiny
phosphate surface veneers (PI. 115, figs. 4-6, 8) underlain by a portion of phosphatized calcite. In
contrast to the opinion expressed by Tabataba'f (1977, p. 212), extensive phosphatization of belemnite
guards was found to be uncommon.

table 7. Summary of encrustation exhibited by three ‘populations’ of Gonioteuthis from phosphatic chalks.

Locality

Number of
guards

Number encrusted (%)

Total

Oyster

Octocoral

Serpulid

> 1 species

Hardivillers

136

53

32

31

10

20

Ribemont

73

40

30

3

6

6

Villers-devant-le-Thour

24

6

6

0

0

0

The different proportions of encrusters and encrustation are not necessarily of environmental
significance since they are too readily affected by the proximity of a ‘spat’ source. The well-preserved
nature of the guards, the juvenile stage of the majority of the epifauna, and the lack of extensive
boring suggest that burial was rapid. Furthermore, the general lack of more than one generation of
encrustation indicates that re-exhumation was rare. These observations have important implications
concerning the environment of deposition. The sedimentology indicates that the phosphatic chalk
lithofacies was a relatively high-energy environment, within which the sediment was being winnowed
by current action (Jarvis 1980). Yet the taphonomy of the belemnites suggests rapid burial and
lack of re-exhumation. Clearly this removes the possibility of a low sedimentation rate due to
continual winnowing as this would result in the belemnites being exposed on the sea floor for
extended periods. Intermittent current action with rapid winnowing events followed by periods of
quiescence and burial, therefore, seems a more likely mechanism. Such a cyclic process may also
explain the apparent paradox between oxic bottom waters and sediment as demonstrated by the epi-
and infauna and the subsurface anoxia associated with contemporary phosphorites (Baturin and
Bezrukov 1979). It can be postulated that phosphatization took place during quiescent periods, and
colonization and bioturbation during times of stronger current activity.

JARVIS: UPPER CRETACEOUS BELEMNITES

909

FACIES AND ECOLOGY

The abundance of belemnites within the phosphatic chalk lithofacies vis-a-vis soft white chalks
requires some consideration. Both Gonioteuthis and A. verus occur very frequently in shallow-water
deposits situated close to ancient massifs, such as the biocalcarenites and glauconitic sands of the
Balto-Scandian area (Christensen 1976). Furthermore, Gonioteuthis from near-shore facies are
characterized by the presence of all ontogenetic stages, while ‘populations’ from offshore chalks only
contain adult specimens (Ernst 1964). This general trend also applies to other genera of belemnites,
e.g. Belemnitella and Belemnella. It must be concluded that the occurrence of juveniles in phosphatic
chalk ‘populations’ is indicative of a near-shore and therefore shallow-water environment for the
facies.

In both facies belemnites are common just below, and particularly above, hardgrounds. The
formation of Chalk hardgrounds is frequently accompanied by evidence of shallower-water condi-
tions (Bromley 1965; Kennedy 1970; Kennedy and Garrison 1975; Jarvis 1980). This explains the
occurrence of hardground associated belemnites in offshore chalks, but it does not explain their
concentration in the phosphatic chalk environment which is initially also presumably shallow-water
(Jarvis 1980).

It has been suggested (Surlyk and Birkelund 1977) that belemnites found in offshore chalks may be
considered as straying adult individuals buried outside their normal habitat. Christensen (1976) has
postulated that this apparent facies control may be causally related through the food-chain.
Belemnites probably preyed upon small fish, crustaceans, and cephalopods, as do their Recent
relatives (Naef 1922), and such prey would be more readily available in a shallow-water environment.

The association with hardgrounds results from the interaction of two independent factors. Firstly,
hardgrounds are generally regarded as representing levels of faunal condensation whether due to
omission (Kennedy and Garrison 1975) or active erosion coupled with winnowing (Jarvis 1980), or
a combination of both processes. Either process would be expected to result in above-average
concentrations of belemnites, but the taphonomy of the material indicates that the guards were not
exposed on the sea floor for extended periods. The Ribemont sample, which originates from above a
hardground, does have a greater proportion of bored fragments, but the proportion is considered
insufficient to confirm the condensation hypothesis. Furthermore, the bimodality of the ‘population’
suggests deposition following a catastrophic event rather than long-term addition of material.

The second factor is the change in ecology which would be expected after sea-floor lithification. The
new environment would embrace a mixed hard-soft substrate ecosystem, which might be expected to
lead to an increase in diversity, and probably increased abundance of organisms. The increase in the
sea-floor dependent biota would provide a food source for animals higher in the food chain, including
the fish and crustaceans which are assumed to have been the main prey of belemnites. The two factors
are not mutually exclusive and undoubtedly act in concert in the majority of cases, but I consider, on
taphonomic grounds, that the second of the two factors predominates in phosphatic chalks.

The Recent coleoid Loligo opalescens generally inhabits water depths of 120-330 m but enters
shallower water for mating and at night to feed. L. opalescens congregates in vast numbers to mate
and reproduce, after which the majority of individuals die. An estimated 20 million dead have been
observed on a small area off Baja California after such an event (Cousteau and Diole 1973). Clearly
this feature of the life cycle of a Recent coleoid, similar in size and morphology to the belemnites,
provides one mechanism for the formation of the so-called ‘Belemnitenschlachtfeld’ (Belemnite
battlefield) of the literature (e.g. Naef 1922).

The Gonioteuthis beds are around 1 m thick and contain belemnites scattered throughout,
although patches of larger numbers of individuals occur. The higher concentrations are not at a
particular level within the Beds nor are they on top of recognizable omission surfaces. It can be
postulated, therefore, that these beds are the result of a series of mass mortalities following
reproduction. The reproductive cycle of L. opalescens has another important aspect; after mating the
squids attaches its eggs to epibenthic organisms and other suitable anchorage points on the sea floor
(Cousteau and Diole 1973; Recksiek 1978). It is suggested that hardgrounds would provide a greater

910 PALAEONTOLOGY, VOLUME 23

number of potential attachment sites than a soft substrate and would therefore be preferred as a
breeding area.

It is not necessary to assume that the ‘populations’ of the Gonioteuthis beds were buried in the
immediate area of their death. During the later stages of cuvette evolution, hardgrounds would be
concentrated at the cuvette margins, where areas of lithified white chalk remained uncovered. These
areas would provide preferred living/breeding sites for belemnites but, on death, individuals would be
swept into the central portions of the cuvette and deposited. Subsequent bioturbation would remove
the identity of individual mortality events and tend to disseminate guards throughout an interval of
sediment. A further factor to be considered is the suggestion (Jarvis 1980) that phosphatic chalk
formation was in part the result of upwelling of deep ocean water in the Anglo-Paris Basin, which
provided the quantities of phosphate necessary for the deposition of such extensive phosphorites.
Such an increase in orthophosphate would undoubtedly lead to higher plankton abundances which,
as the base of the marine food-chain would result in increased numbers of higher animals including
belemnites.

The Gonioteuthis beds occur at the top of the phosphatic chalks and perhaps represent an acme
prior to the change in environment which resulted in the cessation of phosphatic chalk deposition.
Such an acme might result from any of the factors discussed when considering the range of A. verus in
phosphatic chalks (see below).

STRATIGRAPHICAL AND ENVIRONMENTAL CONTROLS

The facies model does not explain all aspects of the occurrence of belemnites in phosphatic chalks.
The best-developed hardgrounds are at the base of phosphatic chalks, yet no basal hardground
shows the concentration of Gonioteuthis guards typical of hardgrounds higher in the sequences. A.
verus ranges from the Santonian to the early Campanian in Yorkshire (Jukes-Browne and Hill 1904;
Wright and Wright 1942), Norfolk (Peake and Hancock 1961), and Germany (Schmid 1956; Ernst
1963a), but in southern England (Rowe 1901; Jukes-Browne and Hill 1904; Griffith and Brydone
1911) and northern France (de Grossouvre 1 899) the species is apparently restricted to the Santonian,
being particularly characteristic of the Uintacrinus socialis Zone. Gonioteuthis, on the other hand,
appears infrequently in the Coniacian (as G. westfalica) chalks of northern France (de Grossouvre
1899, 1901, 1907) and southern England (Rowe 1901; de Grossouvre in Rowe 1901). Rare G.
westfalicagranulata and G. granulata occur with A. verus in the mid-late Santonian of both areas
(de Grossouvre 1899; Jukes-Browne and Hill 1904; Rowe 1904; Peake and Hancock 1961) but
G. granulata is most abundant in the late Santonian Marsupites testudinarius Zone where A. verus is
uncommon. The base of the early Campanian Echinocorys depressula Subzone of the O.pilula Zone is
typified by forms intermediate between G. granulata and G. q. quadrata, i.e. G. granulataquadrata
(Jukes-Browne and Hill 1904; Griffith and Brydone 1911), and Belemnitella (probably B. praecursor)
makes its first appearance at this level (Jukes-Browne and Hill 1904). G. q. quadrata appears above
the base of the Campanian in the ‘Abundant O. pilula’ Subzone of the O. pilula Zone and continues
into the overlying G. quadrata Zone, where it becomes the index fossil (e.g. Griffith and Brydone
1911). Thus throughout the Anglo-Paris Basin Gonioteuthis is only relatively common in chalks
where A. verus is rare or absent. This relationship is clearly shown by phosphatic chalks where the
occurrence of the two genera is almost antipathetic. Explanations for this phenomenon must be
sought from environmental, evolutionary, or provincial controls since it is not due to
the relative ranges of the two fossils.

Firstly, the current regime which originally produced the cuvettes must have been of a higher order
than that which accompanied their infill. This is demonstrated by the transition from erosion to
deposition and may be reflected in the coarser and generally higher phosphate content of the lowest
phosphatic chalks. In turn, changes in regime may be reflected in substrate and biota both within the
cuvettes and elsewhere in the basin. Secondly, the major increase in the abundance of Gonioteuthis
coincides with the evolution of G. q. quadrata. It might be suggested therefore that evolutionary
changes in the Gonioteuthis stock may have enabled the genus to diversify and occupy previously

JARVIS: UPPER CRETACEOUS BELEMNITES

91

unfavourable niches. Broader environmental tolerances or changes in food requirements, for
example, would provide a mechanism. Lastly, Jarvis (1980) has suggested that the initiation of
phosphatic chalk sedimentation was in part due to major changes in oceanic circulation during the
Santonian, which accompanied the opening of the Atlantic Ocean. Later changes in the distribution
of water masses during the period of phosphatic chalk sedimentation may have enabled populations
of Gonioteuthis to enter an area which, because of oceanographical conditions, was dominated
previously by A. verus and B. ex gr. grossouvrei. Similarly, A. verus might have been excluded from
that area. Any one, or a combination of these factors, can be invoked to explain the observed changes
in belemnite distribution during the evolution of the lithofacies.

CONCLUSIONS

(1) The phosphatic chalks of northern France provide rare examples of Santonian-early
Campanian ‘populations’ of Gonioteuthis. This enables the application of statistical analysis in
the identification of species.

(2) These phosphatic chalks can be divided into three biostratigraphical subdivisions on their
belemnite assemblages:

(a) a lower unit characterised by A. verus with Micros ter coranguinum, occasional G. granulata and rare
B. ex gr. grossouvrei',

( b ) an intermediate division typified by G. q. quadrata with O. pilula;

( c ) an upper unit with G. q. quadrata and occasional B. praecursor.

(3) A combination of univariate and bivariate statistics demonstrates that the ‘populations’ from
the Hardivillers Gonioteuthis Bed and the phosphatic chalk at Villers-devant-le-Thour are
homogenous, whereas the ‘population’ from above the Ribemont Gonioteuthis Hardground is
inhomogenous. Inhomogeneity in the latter sample is the result of a catastrophic event,
probably a storm, which has resulted in the concentration of a mixed ‘population’ of young and
old individuals.

(4) Mean Riedel-Quotients for all three ‘populations’ fall within the range of G. ex gr. quadrata but
subspecies cannot be identified on Riedel-Quotient alone.

(5) G. q. quadrata has been identified as the subspecies present in the Hardivillers Gonioteuthis Bed.
This ‘population’ of the subspecies shows an allometric relationship of L v. D, L v. DVDAE
and MLD v. LDAE.

(6) The Ribemont and Villers-devant-le-Thour ‘populations’ cannot be distinguished from each
other but are distinct from that studied from Hardivillers. The subspecies at the former localities
shows affinities to both G. q. quadrata and G. q. gracilis.

(7) Principal component factor analysis illustrates that the major controls on Gonioteuthis guard
morphology are:

(a) the size of the guard, which accounts for 87-2% of the observed variation;

(, b ) the evolutionary stage of the individual within the Gonioteuthis gradualistic series, 6-8% of the
variation;

(c) a shape factor which produces elongate guards (or perhaps a taphonomic factor), 4-6%.

(8) The well-preserved nature of the guards, the juvenile nature of the epifauna, and the absence of
extensive boring suggest relatively rapid burial, while the general lack of more than one
generation of epifauna implys that re-exhumation was rare.

(9) Guard taphonomy suggests that the environment of deposition was one of intermittent current
activity with rapid winnowing, followed by periods of quiescence and burial.

(10) The occurrence of juveniles in the ‘populations’ indicates a shallow-water environment which
was the belemnites’ normal habitat because of the greater availability of food.

(11) Belemnites are concentrated at hardground levels because of:

(a) a concentration of potential prey associated with the hardgrounds;

( b ) the possibility that they may be preferred breeding sites and would therefore be the recipients of the
large numbers of dead individuals following the mass mortalities which accompany reproduction.

912

PALAEONTOLOGY, VOLUME 23

(12) The virtual mutually exclusive relationship between A. verus and Gonioteuthis in phosphatic
chalks may be controlled by:

(а) the decline in the current regime during cuvette evolution favouring Gonioteuthis]

(б) evolutionary changes in Gonioteuthis allowing the genera to occupy a previously unfavourable niche
and to oust A. verus, or

(c) changes in provincial boundaries which accompanied changes in oceanic circulation.

Acknowledgements. I thank P. Woodroof and A. S. Gale for their help in collecting the specimens described in
this paper. W. K. Christensen, W. J. Kennedy, T. J. Palmer, and R. Kirby read through and greatly improved
earlier drafts of the manuscript. W. K. Christensen kindly supplied comparative statistical data for ‘populations’
of G. q. quadrata from Belgium and Germany. The work was undertaken while the author was in receipt of a
NERC postgraduate studentship at the University of Oxford.

REFERENCES

baturin, G. n. and bezrukov p. l. 1979. Phosphorites on the sea-floor and their origin. Mar. Geol. 31, 317-332.
bayle, e. 1879. Explication de la carte geologique de la France. Atlas 1 repartie: Fossiles principaux des terrains.
blackith, R. E. and reyment, R. A. 1971. Multivariate morphometries, 412 pp. Academic Press, New York,
London.

briquet, a. 1902 Note sur les gisements de craie phosphatee exploites a Beauval. Annls Soc. geol. N. 31, 83-86.
bromley, R. G. 1965. Studies in the lithology and conditions of sedimentation of the Chalk Rock and comparable
horizons. Ph.D. thesis (unpubl.), Univ. London, 355 pp.
broquet, p. 1973. La craie phosphatee en Picardie, perspectives d’avenir de la prospection. Annls Scient. Univ.
Besanqon (3), 20, 143-152.

brydone, r. h. 1914. The Zone of Offaster pilula in the south English Chalk. Geol. Mag. 51, 359-369, 405-411,
449-459, 509-513.

buteux, c. J. 1849. Esquisse geologique du department de la Somme, 122 pp. Paris.

cayeux, L. 1939. Les phosphates de chaux sedimentaires de France. 1. France metropolitaine. Etudes des gites
mineraux de la France. Impr. Nationale, Paris, 349 pp.

Christensen, w. K. 1970. Actinocamax primus Arkhangelsky from Scania, Sweden. Stockh. Contr. Geol. 21,
69-76.

1971. Belemnitella propinqua propinqua (Moberg, 1885) from Scandinavia. Bull. Geol. Soc. Denm. 20,

369-384.

— 1973. The belemnites and their stratigraphical significance. In bergstrom, j., Christensen, w. k.,
johansson, c. and norling, e. An extension of Upper Cretaceous rocks to the Swedish west coast at Sardal.
Ibid. 22, 113-140.

1974. Morphometric analysis of Actinocamax plenus from England. Ibid. 23, 1-26.

— 1975a. Upper Cretaceous belemnites from the Kristianstad area of Scania. Fossils strata, 7, 1-69.

— 19756. Designation oflectotypes for Gonioteuthis westfalicagranulata and G. granulataquadrata. Palaont. Z.
49, 129-134.

— 1976a. First record of Belemnellocamax balsvikensis (Brotzen, 1960) from NW Germany. Neues. Jb. Geol.
Palaont. Mh. H9, 522-531.

— 19766. Palaeobiogeography of Late Cretaceous belemnites of Europe. Palaont. Z. 50, 113-129.

— ernst, G., schmid, f., schulz, m.-g. and wood, c. J. 1975. Belemnitella mucronata mucronata (Schlottheim,
1813) from the Upper Campanian: Neotype, biometry, comparisons and biostratigraphy. Geol. Jb. A28,
27-57.

Cousteau, J. and diole, p. 1973. Octopus and squid: the soft intelligence, 304 pp. Cassell, London.
craig, G. y. 1967. Size-frequency distributions of living and dead populations of Pelecypods from Bimini,
Bahamas, B.W.I. J. Geol. 75, 34-45.

ernst, G. 1963a. Stratigraphische und gesteinschemische Untersuchungen im Santon und Campan von
Lagerdorf (SW-Holstein). Mitt. geol. Stlnst. Hamb. 32, 71-127.

— 19636. Zur Feinstratigraphie und Biostratonomie des Obersanton und Campan von Misburg und Hover
bei Hannover. Ibid. 128-147.

— 1964. Ontogenie, Phylogenie und Stratigraphie der Belemnitengattung Gonioteuthis Bayle aus dem
nordwestdeutschen Santon/Campan. Fortschr. Geol. Rheinld, Westf. 7, 113-174.

JARVIS: UPPER CRETACEOUS BELEMNITES

913

ernst, G. 1966. Fauna, Okologie und Stratigraphie der Mittelsanton Schreibkreide von Lagerdorf (SW-
Holstein). Mitt. geol. Stlnst. Hamb. 35, 115-145.

— 1968. Die Oberkreide-Aufschlusse im Raume Braunschweig-Hannover und ihre stratigraphische
Gliederung mit Echinodermen und Belemniten. 1. Teil. Die jiingere Oberkreide (Santon-Maastricht). Beih.
Ber. Naturh. Ges. 5, 235-284.

and schulz, m.-g. 1974. Stratigraphie und Fauna des Coniac und Santon im Schreibkreide-Richtprofil von

Lagerdorf (Holstein). Mitt, geol-palaont. Inst. Univ. Hamburg , 43, 5-60.
gosselet, J. 1901. Observations geologiques faites dans les exploitations de phosphate de chaux en 1901. Annls
Soc. geol. N. 30, 208-243.

Griffith, c. and brydone, r. M. 1911. The zones of the Chalk of Hants , 35 pp. Dulau, London.
grossouvre, A. de 1894. Sur la craie grise. Bull. Soc. geol. Fr. (3) CR Seances, 22, 57-58.

— 1899. Quelques observations sur les Belemnitelles et en particulier sur celles des Corbieres. Ibid. 27,
129-135.

1901. Recherches sur la craie superieure 1. Stratigraphie generale. Mem. Serv. Carte geol. det. Fr ., 1013 pp.

— 1907. Sur la craie grise a Belemnitelles. Assoc. Fr. Adv. Sci. Rheims, 36 Session, Notes Mem. 403-406.
Guerin, B., maucorps, J., solau, J. and pomerol, c. 1977. Carte geologique de la France a 1/50,000, 85: Chateau-

Porcien et notice explicative. Publ. Bur. Rech. geol. Min. Orleans.
hald, A. 1957. Statistical theory with engineering applications. 783 pp. Wiley & Sons, New York, London.
jarvis, i. 1980. The initiation of phosphatic chalk sedimentation— the Senonian (Cretaceous) of the Anglo-
Paris Basin. In bentor, y. k. (ed.). Marine Phosphorites. Spec. Publ. Soc. Econ. Paleon. Min. 29.
jukes-browne, A. J. and hill, w. 1904. The Cretaceous rocks of Britain Vol. 3. The Upper Chalk of England.

Mem. Geol. Surv. UK, 566 pp. HMSO, London.

Kennedy, w. j. 1970. Trace fossils in the Chalk environment. Geol. J. Spec. Iss. 3, 263-282.

— and garrison, r. e. 1975. Morphology and genesis of nodular chalks and hardgrounds in the Upper
Cretaceous of southern England. Sedimentology, 22, 311-386.

kermack, k. a. 1954. A biometrical study of Micros ter coranguinum and M. ( Isomicr aster ) senonensis. Phil.
Trans. R. Soc. B227, 375-428.

lasne, H. 1890. Sur les terrains phosphates des environs de Doullens. Etage Senonien et terrains superposes Bull.
Soc. geol. Fr. (3), 18,441-491.

— 1892. Sur les terrains phosphates des environs de Doullens. Etage Senonien et terrains superposes (2e
Note). Ibid., 20,211-236.

— 1 902. Observations concernant le gisement de la Craie phosphatee. Lettre a M. Gosselet. Annls Soc. geol. N.,
31, 55-59.

leriche, M. 1908. Sur les fossiles de la craie phosphatee de la Picardie a Actinocamax quadratus. Assoc. Fr. Adv.
Sci. Clermont-Ferrand, 37 Session, Notes Mem. 494-503.

— 1911. Deuxieme note sur les fossiles de la craie phosphatee de la Picardie. Mem. Soc. beige. Geol. Paleont.
Hydro!. Bull., 25, 297-310.

mangold- wirz, k. 1963. Biologie des Cephalopodes benthiques et nectoniques de la Mer Catalane. ‘Vie Milieux’,
Supp. 13, 285 pp. Hermann, Paris.

mercey, N. de 1887. La craie phosphatee a B. quadratus dans le Nord de la France. Bull. Soc. geol. Fr. (3), 15,
719-725.

meunier, s. 1888. Conditions geologiques du gisements phosphate de Beauval. C.r. hebd. Seanc. Acad. Sci.,
Paris, 106, 214-217.

1891. Les phosphatieres d’Hardivillers, Oise. Naturaliste 1891, 46-49, 55-56.

miller, j. s. 1829. Observations on the genus Actinocamax. Trans. Geol. Soc. Lond. (2), 2, 63-67.
naef, a. 1922. Die fossilen Tintenfische, 322 pp. Gustav Fisher, Jena.

naidin, d. p. 1964. The Upper Cretaceous belemnites of the Russian Platform and contiguous regions,
Actinocamax, Gonioteuthis, Belemnellocamax, 190 pp. Moscow Univ. Press, Moscow. [In Russian.]
negre, G. 1912. Note sur les gisements de phosphate de Beauval et environs. Annls Soc. geol. N. 41, 235-239.

— 1963. Les craies phosphatees en France: Beauval (1886-1963). L'lnd. Chim. 50, 383-386.

peake, n. b. and Hancock, j. m. 1961. The Upper Cretaceous of Norfolk. Trans. Norfolk Norwich Nat. Soc. 19,
293-339.

rabelle, m. 1893. Quelques observations geologiques faites aux environs de Ribemont, rive gauche de l’Oise.
Annls Soc. geol. N. 21, 344-349.

— 1902. Observations geologiques aux environs de Ribemont et dans la craie phosphatee d’Etaves et Fresnoy.
Ibid., 31, 45-49.

recksiek, c. w. 1978. California’s market squid. Pacif. Discov., 31, 19-27.

914

PALAEONTOLOGY, VOLUME 23

reyment, r. a. and naidin, d. p. 1962. Biometric study of Actinocamax verus s.l. from the Upper Cretaceous of
the Russian Platform. Stockh. Contr. Geol. 9, 147-206.
rowe, a. w. 1901. The Zones of the White Chalk of the English coast. II. Dorset. Proc. Geol. Ass. 17, 1-76.

— 1904. The Zones of the White Chalk of the English coast. IV. Yorkshire. Ibid. 18, 193-296.

— 1908. The Zones of the White Chalk of the English coast. V. The Isle of Wight. Ibid. 20, 209-352.
schmid, f. 1956. Jetziger Stand der Oberkreide-Biostratigraphie in Nordwestdeutschland: Cephalopoden.

Palaont. Z. 30, 7-10.

shaw, a. b. 1956. Quantitative trilobite studies. 1. The statistical description of trilobites. J. Paleont. 30,
1209-1224.

simpson, G. G., roe, a. and lewontin, r. c. 1960. Quantitative Zoology, 440 pp. Harcourt, Brace and World Inc.,
New York, Chicago, San Francisco, Atlanta.
sokal, r. r. 1965. Statistical methods in systematics. Biol. Rev. 40, 337-391.

— and rohlf, F. j. 1969. Biometry — The principles and practise of statistics in biological research, 776 pp.
W. H. Freeman, San Francisco.

stolley, E. 1897. Uber die Gliederung des norddeutschen und baltischen Senon sowie die dasselbe
characterisierenden Belemniten. Arch. Antrop. Geol. Schlesw. Holst. 2, 216-302.

— 1916. Neue Beitrage zur Kenntnis der norddeutschen oberen Kreide, I -IV. Jber. Nieder sachs, Geol. Ver. 9,
95-104.

— 1930. Einige Bemerkungen uber die Kreide Siidskandinaviens. Geol. For. Stockh. Forh. 52, 157-190.
surlyk, f. 1972. Morphological adaptions and population structures of Danish chalk brachiopods

(Maastrichtian, Upper Cretaceous). Biol. Skr., 19, 1-57.

— and birkelund, t. 1977. An integrated study of fossil assemblages from the Maastrichtian white chalk of
northwestern Europe, pp. 257-281. In kauffman, e. g. and hazel, j. e. (eds.) Concepts and methods in
biostratigraphy, xiv+658 pp. Dowden, Hutchinson, & Ross, Stroudsburg Pennsylvania.

tabatabai, c. m. 1977. La sedimentation phosphatee (ses modialites): Petrographie et sedimentologie des craies
phosphatees du Nord du Bassin de Paris. These Dip. docteur 3e cycle, Univ. Pierre Marie Curie, Paris, 243 pp.
van hinte, J. E. 1976. A Cretaceous time scale. Bull. Am. Ass. Petrol. Geol. 60, 269-287.

WILLCOX-, N. R. 1953. Some coprolites from phosphatic chalks in SE England. Ann. Mag. nat. Hist. 6, 369-375.
wright, c. w. and wright, e. v. 1942. The Chalk of the Yorkshire Wolds. Proc. Geol. Ass. 53, 1 12-127.

I. JARVIS

Manuscript received 23 October 1979
Revised manuscript received 1 1 January 1980

Department of Geology
The University
Glasgow G12 8QQ

A NEW LABYRINTHODONT AMPHIBIAN FROM
THE CARBONIFEROUS OF SCOTLAND

by T. R. SMITHSON

Abstract. Cranial remains of a labyrinthodont amphibian Doragnathus woodi gen. et sp. nov., from localities in
the Visean and Namurian of the Scottish Carboniferous, are described. The structure of the lower jaw resembles
that of the earliest known Amphibia, but its dentition is unusual, comprising large numbers of strongly incurved,
closely spaced marginal teeth together with a row of small needle-like coronoid teeth. The relationships of
Doragnathus are discussed. A specimen of Doragnathus from Pitcorthie represents the earliest recorded
labyrinthodont in the British Carboniferous.

T he Scottish Midland Valley is one of the few areas in the world from which fossil Amphibia have
been found in Carboniferous sediments older than those equivalent in age to the British Coal
Measures (Westphalian and Stephanian). Thirteen pre-Coal Measure genera have so far been
described from a total of eleven Scottish localities. Most discoveries were made in the latter half of the
last century, but recently a diverse amphibian fauna was discovered in a bone bed at the Dora
Opencast Site, near Cowdenbeath, Fife (Andrews, Browne, Panchen, and Wood 1977; Smithson, in
press). With the exception of an almost complete skeleton of Crassigyrinus scoticus (Panchen, in
press) the Cowdenbeath fauna is represented by dissociated skeletal elements of at least six
amphibian genera. The most common of these is a hitherto undescribed labyrinthodont represented
by a large number of incomplete jaw specimens. The new form has also been found at Pitcorthie and
Niddrie (Smithson, in press) and recently by Mr. Stanley Wood and the author on the island of
Inchkeith, in the Firth of Forth.

Labyrinthodonts are rare components of the Scottish Lower Carboniferous amphibian assemb-
lage, and until recently they had been recorded at only three of the eight Lower Carboniferous
localities. The recognition that a jaw specimen from Pitcorthie in the collection of the Royal Scottish
Museum was that of a labyrinthodont and not, as had previously been thought, a lepospondyl, and
the discovery of similar material on the Island of Inchkeith, has improved this position. Although
these new specimens are incomplete and poorly preserved, material from the Upper Carboniferous
deposits at Cowdenbeath and Niddrie allows a description of a number of aspects of the cranial
anatomy of the new labyrinthodont to be given.

The fossiliferous deposits on the island of Inchkeith and at Pitcorthie occur in strata of Visean age.
The middle of the exposed sequence on Inchkeith is thought to be equivalent to the horizon of the
Burdiehouse Limestone (Davis 1 936). The fossiliferous sediments at Pitcorthie almost certainly occur
within the Anstruther Beds (Forsyth and Chisholm 1973). These lie below the Cuniger Rock Marine
Band which has been placed, on palynological evidence, well below the Burdiehouse Limestone
(Neves et al. 1973). Thus the amphibian remains from Pitcorthie are older than those from the
Burdiehouse Limestone, and the labyrinthodont remains are the earliest recorded in the British
Carboniferous.

Where necessary material was prepared with a dental mallet and industrial ‘Airbrasive’ unit, and a
solution of ‘Perspex’ dissolved in chloroform was used to repair breaks in specimens.

The following abbreviations are used for the institutions owning the material: NUZ, Department
of Zoology (University of Newcastle upon Tyne); RSM, Department of Geology (Royal Scottish
Museum).

IPalaeontology, Vol. 23, Part 4, 1980, pp. 915-923.|

916

PALAEONTOLOGY, VOLUME 23

SYSTEMATIC PALAEONTOLOGY

Class AMPHIBIA
Subclass LABYRINTHODONTIA
Order and Family Undesignated
Doragnathus gen. nov.

Type species. Doragnathus woodi gen. et sp. nov.

Etymology. The name refers to the large number of jaw specimens from the Dora Opencast Site.

Diagnosis. Labyrinthodont amphibian with a long, shallow lower jaw which terminates with a
distinct retroarticular process; small Meckelian fenestra at the mesial exposure of the splenial/post
splenial suture; dentary with room for more than eighty closely spaced, strongly incurved teeth,
labyrinthine unfolding of enamel found only below the margin of the gums; a row of slim needle-like
replaceable teeth on the coronoid series.

Doragnathus woodi gen. et sp. nov.

Text-figures 1-3

Etymology. The animal is named after Mr. Stanley Wood, who discovered the bonebed at the Dora Opencast
Site which has yielded the majority of the material attributed to Doragnathus.

Diagnosis. As for genus.

Holotype. NUZ 77.5.26 incomplete left ramus of lower jaw.

Type horizon and locality. Localized seatrock, beneath a coal seam below the Lochgelly Blackband Ionstone,
upper part of the Limestone Coal Group (Namurian A, Upper Carboniferous), Dora Opencast Site,
Cowdenbeath, Fife, Scotland.

Distribution. Scottish Midland Valley (Lothian and Fife regions).

Range. Middle Calciferous Sandstone Measures (Anstruther Beds) to Upper Limestone Group (South Parrot
Coal Shale). CjSj? zone of Visean stage (Lower Carboniferous) to E2 zone of Namurian A (Upper
Carboniferous).

Description. The description of Doragnathus is based on the most complete specimens in the collections of the
Royal Scottish Museum and the University of Newcastle upon Tyne. All have suffered post-mortem compression
and only RSM GY 1898.107.51 is preserved in any degree of completeness (text-fig. 1 a, b). A complete list of
attributed material is deposited in the Department of Geology, Royal Scottish Museum.

Lower jaw. In its over-all construction the lower jaw is similar to that of early labyrinthodonts, e.g. Ichthyostega
(Save-Soderbergh 1932; Jarvik 1952) and Metaxygnathus (Campbell and Bell 1977). Each ramus is relatively
shallow throughout its length, tapering slightly towards the synphysis in lateral view. In dorsal view each ramus
curves gently towards the midline and, when articulated, the two jaws describe a distinct U. The majority of
specimens are from animals with lower jaws approximately 12 cm long. A number of specimens, notably RSM
GY 1975.5.3, have been found at Cowdenbeath which are considerably larger; the estimated lower jaw length of
the largest of these is approximately 30 cm. The smallest recorded specimen, RSM GY 1881.43.24 from
Pitcorthie, has a lower jaw approximately 6 cm long. (All measurements included in this description have been
taken from specimens approximately 12 cm in length.)

A well-defined mandibular lateral line canal follows the ventral edge of the jaw ramus. At regular intervals it is
partially or completely bridged with bone and appears as a series of ovoid pits approximately T5 mm long along
their antero-posterior axis and approximately 1 mm wide. The adductor fossa is preserved in RSM GY
1898.107.51 and RSM GY 1975.5.3. It is a steep-sided cavity, approximately 2-8 cm long, walled laterally by the
surangular and mesially by the prearticular. The dorsal surface of the surangular is slightly convex but is not
developed into the high surangular crest of anthracosaurs and certain temnospondyls. Anteriorly the fossa
tapers and is bounded by the posterior coronoid at the level of the back of the tooth row. Half-way along the
floor of the fossa in RSM GY 1898.107.51 is a small raised rugosity which probably acted as a point of

SMITHSON: LABYRINTHODONT AMPHIBIAN FROM SCOTLAND

917

(e) (f)

0 3

1 I L_ J

cm

text-fig. 1. Doragnathus woodi lower jaw and premaxillary, natural size, (a) lateral, ( b )
mesial view of lower jaw RSM GY 1898.107.51; (c) mesial, (d) lateral view of holotype NUZ
77.8.28; ( e ) external view right premaxillary NUZ 77.5.26; (/) external view left premaxillary
NUS 78.1.26. Damaged bone surfaces, hatched; matrix, regular stipple, a, angular; amf,
anterior Meckelian fenestra; art, articular; co, coronoid; d, dentary; posl, postsplenial; pra,
prearticular; psf, postsymphysial foramen; sa, surangular; sp, splenial.

918

PALAEONTOLOGY, VOLUME 23

attachment for a mass of the adductor mandibulae musculature. In the tunnel formed between the inner and
outer walls of the ramus the cavity extends as the Meckelian space, which may have been partially occupied by
Meckel’s cartilage (Nilsson 1944).

The Meckelian fenestrae are small and restricted to the ventral margin of the mesial surface of the jaw. They
are preserved only in the holotype (text-fig. lc). The most anterior fenestra perforates the jaw at the junction of
the splenial and the postsplenial bones. A second, smaller fenestra may be present approximately 1 cm behind the
first. Unfortunately, this region is badly damaged and interpretation is difficult. The posterior Meckelian
fenestra normally found at the junction of the postsplenial and angular is absent.

The mesial surface of the symphysial region is comparatively smooth and exhibits none of the roughened areas
for ligamentous attachment normally expected. The apparent absence of strong points of ligament attachment
suggests that the symphysis was comparatively weak and a certain degree of movement of the jaw rami relative to
one another was possible. Immediately behind the symphysis the dentary is roughly triangular in section, one
side forming the lateral surface of the jaw, a second the mesial wall, and the third a tooth-bearing shelf. Directly
below the tooth-bearing shelf, the mesial wall is pierced by the postsymphysial foramen. Posterior to this, the
infradentary bones are incorporated into the mesial and lateral surfaces of the jaw and to the mesial edge of the
tooth-bearing shelf are attached the coronoids. The greatest exposure of the dentary is in the lateral wall.
Anteriorly it is approximately 5 mm deep, gently deepening posteriorly to reach its maximum depth towards the
end of the tooth row. Extending along its dorsal edge is a finely ornamented border. Below this the dentary is
almost smooth: only behind the symphysis is the dentary ornamented with irregular ridges and grooves. In
dorsal view the tooth-bearing shelf extends from the symphysis to terminate immediately in front of the adductor
fossa. Throughout its length the shelf maintains an almost constant width. It has room for more than eighty
closely spaced, strongly incurved teeth. Details of the dentition are discussed separately.

The coronoids extend from behind the symphysis to the adductor fossa and form an almost horizontal roof to
the Meckelian space. Small replaceable needle-like teeth extend along the lateral edges of the coronoids and form
a single row of teeth lying parallel to those on the dentary. The series widens posteriorly eventually forming the
anterior border of the adductor fossa. Unlike some later labyrinthodonts, the posterior coronoid is not
incorporated into the lateral wall of the fossa. It was not possible to trace the sutures between the individual
coronoids.

The Meckelian space is floored by the two splenials and the angular. The (pre-)splenial contacts the dentary
behind the symphysis. Laterally the suture between the two bones is long and straight and runs parallel with the
ventral jaw margin.

Posteriorly the splenial contacts the postsplenial along an oblique suture which passes under the ventral edge
of the jaw. Both elements are ornamented with the irregular ridges and grooves found on the dentary. In mesial
exposure it contacts the anterior coronoid under the mesial border of the coronoid shelf. Along the length of the
suture the angle between the two bones is approximately ninety degrees. Posteriorly the dorsal edge of the
splenial gently tapers away from the coronoid shelf to form a broad suture with the prearticular which contacts
the postsplenial at the jaw margin. Lying at the junction of the three elements is the anterior Meckelian fossa. The
postsplenial is a narrow strip of bone which, in lateral exposure, contacts the dentary along most of its length and
posteriorly sutures with the angular. It contacts the prearticular along a broad suture mesially and posteriorly
continues to contact the angular. Between the postsplenial and prearticular a second small Meckelian fenestra
may be present lying approximately 1 cm behind the first.

The posterior part of the lateral wall of the ramus is formed by the angular and surangular. Both are
ornamented with a system of ridges and pits which are less well defined than those on the two splenials. The
angular wraps around the posterior edge of the ramus to present a relatively narow exposure on the mesial
surface where it has a long straight suture with the prearticular.

Behind the tooth row, the lateral dorsal margin of the jaw is formed by the gently convex rim of the surangular.
Posteriorly it sheathes the articular, extending behind the glenoid to form the lateral wall of a short retroarticular
process. Passing down the posterior edge of the process is the surangular articular suture. In dorsal view, the rim
of the surangular which sheathes the articular is thickened and incorporated into the glenoid fossa. The suture
between the surangular and dentary could not be traced.

The mesial wall of the adductor fossa is formed principally by the prearticular, a long narrow element which
extends forward to fill much of the inner surface of the jaw. Posteriorly it sheathes the articular and is
incorporated into the glenoid, but unlike the surangular does not form part of the retroarticular process. It
sutures with the angular and splenial bones ventrally and its dorsal edge contacts the posterior coronoid to form
the anterior margin of the adductor fossa.

The remaining element in the lower jaw is the articular which represents the only ossification of Meckel’s
cartilage. It is embraced on its lateral and mesial surfaces by the surangular and prearticular respectively, but

SMITHSON: LAB YRINTHODONT AMPHIBIAN FROM SCOTLAND

919

text-fig. 2. Doragnathus wooz/zlower jaw. Composite restoration, natural size, (a) lateral,
(b) mesial, (c) dorsal view, (d) transverse sections of jaw at positions indicated in (c).
Abbreviations as in text-fig. 1 .

920

PALAEONTOLOGY, VOLUME 23

only on its mesial surface is it exposed where it extends posteriorly to form the internal surface of the
retroarticular process. The shape of the glenoid fossa, when viewed dorsally, resembles a distorted figure of eight.
This is clearly seen in RSM GY 1975.5.3. The articular surface is divided into two subcircular depressions by a
ridge oriented along the anterior posterior axis of the jaw. The mesial depression extends slightly further forward
than that on the lateral surface of the fossa. Both are bounded anteriorly by a well-defined precondyloid process.
The postcondyloid process is not clearly defined since the posterior margin of the fossa is incorporated into the
retroarticular process. The foramen for the chorda tympani (mandibular) branch of the seventh nerve, which in
most Amphibia pierces the lower jaw just below the glenoid fossa, could not be traced.

Premaxilla. A number of isolated, partially complete, premaxillae containing marginal teeth of the type
described from the lower jaw of Doragnathus have been found at Cowdenbeath. All are thought to be from skulls
with a lower jaw length of approximately 12 cm. In antero-lateral view the premaxilla is a narrow elongate
element ornamented with irregular ridges and grooves as found on the splenials of the lower jaw. In ventral
aspect it is gently curved and mirrors the shape of the anterior region of the mandible. There is room for more
than twenty-five teeth. The premaxillary margin of the choana is not preserved on any specimen.

The supraorbital lateral line canal passes across the exterior surface of the premaxilla. It extends over the
posterior dorsal edge onto the nasal and along the antero-lateral edge to join the second half of the supraorbital
canal on the opposite premaxilla. The canal is manifest in a variety of ways. In NUZ 77.5.26 it appears as an open
groove which is bridged over with bone at one point only. In NUZ 78.2.26, however, it is bridged over much of its
length and is visible as an irregular series of pits (text-fig. le, f). Only where it runs over onto the nasal and
opposite premaxillary does it appear in an open groove.

In palatal aspect, the premaxilla extends posteriorly as a broad shelf of bone to suture with the vomers. The
presence of a palatal fenestra is improbable.

Maxilla. An incomplete maxilla has been found at Pitcorthie (RSM GY 1881.43.24). Apart from showing that
the maxillary teeth are of the type found on the dentary and premaxilla, it yields little information.

Dentition. The marginal teeth in the upper and lower jaws of Doragnathus are identical. They are of uniform size
along most of the tooth row becoming smaller towards its posterior end. There is no parasymphysial tusk on the
lower jaw and no peaking along the dentary or maxillary. In lateral view the teeth are bullet-shaped and in
posterior aspect strongly incurved (text-fig. 36). Their bases are twice as wide as long and narrow towards the
apex of the tooth, becoming almost circular in section just below the tip.

1 mm

text-fig. 3. Doragnathus woodi teeth, (a) lateral, (6) posterior view of
marginal tooth x 12; ( c ) transverse sections through tooth as positions
indicated in (6), x 20.

SMITHSON: LABYRINTHODONT AMPHIBIAN FROM SCOTLAND

92

Labyrinthine infolding of the enamel is only found at the base of the teeth below the margin of the gums. In
section they show the characteristic infolding of the external primary dentine, but whereas in many
osteolepiform crossopterygians and most labyrinthodonts the infoldings meander and some cases are branched
(Schultze 1969), those of Doragnathus appear as straight, nonconvolute unbranched folds (text-fig. 3c). The
majority of teeth from which histological sections have been taken in fish and in other labyrinthodonts have been
the large tusk teeth on the palate and the coronoids. However, the marginal dentition of most crossopterygians
comprises teeth considerably smaller than the tusk teeth and these exhibit simple folding, e.g. Megalichthys
(Schultze 1969, p. 94), of a type very similar to that found in Doragnathus. The tortuous infolding often taken to
typify labyrinthodont teeth is undoubtedly a size-related phenomenon, a fact clearly demonstrated by Bystrow
and Efremov (1940, p. 46).

The teeth are arranged in the jaw as a series of ‘clusters and gaps’. Normally approximately seven teeth are
grouped together as a cluster and separated from a similar cluster by one or two replacement pits. However,
groups of smaller than seven occur particularly in the middle of the tooth row and the over-all pattern of tooth
replacement is unclear.

The coronoid dentition comprises a single row of small, replaceable, needle-like teeth which lie parallel to the
marginal row. The palatal dentition is unknown.

DISCUSSION

The paucity of complete or partially complete specimens of Doragnathus prevents a satisfactory
analysis of its relationships, but certain features of its lower jaw allow a number of points to be
considered. The over-all structure of the jaw resembles that of the earliest known Amphibia. It differs
from those of rhipidistian fishes in a number of respects, notably the largest teeth in the jaw are found
in the dentary, the coronoid teeth are small and there are no coronoid fangs. The prearticular fails to
reach the symphysis and the mesial surface of the jaw is perforated anteriorly by at least one
Meckelian fenestra.

The position of Doragnathus within the Amphibia seems clear. Its average skull size (represented
by specimens with lower jaws 12 cm long) is larger than that found in the majority of non-
labyrinthodont amphibians, and the largest specimens are considerably larger than any known
‘lepospondyf. The infolded internal structure of the marginal teeth suggest its inclusion within the
Labyrinthodontia, but infolded teeth have been found in the microsaur Trihecaton (Vaughn 1972),
and it is possible that infolding is primarily a function of tooth size (Thomson and Bossy 1970).
However, one additional feature which suggests affinity with the labyrinthodonts is the presence of an
anterior Meckelian fenestra. No non-labyrinthodont is known in which the mesial exposure of the
splenial/postsplenial suture is perforated.

The position of Doragnathus within the Labyrinthodontia is less certain. In most respects its lower
jaw is primitive and resembles that of Ichthyostega (Save-Soderbergh 1932; Jarvik 1952), and
M/etaxygnathus (Campbell and Bell 1977). Each has a retroarticular process and an extensive
prearticular which forms much of the mesial wall of the jaw. A single row of teeth is present on the
coronoids lying parallel to the marginal row on the dentary. However, the parasymphysial tusks
present in Ichthyostega (Campbell and Bell 1977) are absent in Metaxygnathus and Doragnathus.
Parasymphysial tusks are present in the majority of early labyrinthodonts, e.g. colosteids (Panchen
1975), loxommatids ( sensu Beaumont 1977), Crassigyrinus (as ‘ Macromerium ’ Panchen 1973; see
Panchen, in press) and Caerorhachis (Holmes and Carroll 1977), and their absence in Metaxygnathus
and Doragnathus may be significant. However, in forms in which the dentition is irregular and
the teeth numerous and of uniform size and shape, tusks are usually lost completely, e.g. the
loxammatoid Spathicephalus (Tilley 1971). This principle may also apply to Doragnathus , and the
absence of parasymphysial tusks need not necessarily indicate close relationship. In addition a
number of differences in the structure of the lower jaw suggest that the two genera are not closely
related, notably the absence in Metaxygnathus of an anterior Meckelian fenestra, the shallow lateral
extent of the dentary, and the high elevation of the articular above the tooth row.

Both Doragnathus and Spathicephalus occur at Cowdenbeath (Smithson, in press) but are easily
distinguished by differences in their dentition. The marginal teeth of Spathicephalus are chisel-shaped

PALAEONTOLOGY, VOLUME 23

922

in lateral view and are not incurved (Smithson in press, text-fig. 2) and coronoid teeth are restricted to
the anterior coronoid (E. H. Beaumont ( nee Tilley) pers. comm.). Whilst differences in the dentition
of the two forms indicate a clear generic distinction, the possibility of close relationship cannot be
ruled out. Unfortunately the lower jaw of Spathicephalus is inadequately known and little direct
comparison can be made to clarify the problem.

In a recent review of the Namurian amphibian fauna I tentatively suggested that Doragnathus be
included within the Trimerorhachoidea (Smithson, in press). This suggestion now seems untenable.
Although the lower jaws bear superficial resemblance, the presence of a posterior Meckelian fenestra
in trimerorhachids and the posterior extension of the posterior coronoid, either incorporated into the
surangular crest (e.g. in Trimerorhachis Williston 1914) or forming a distinct coronoid process (e.g.
Dvinosaurus Bystrow 1938), almost certainly precludes close relationship.

It is clear that the position of Doragnathus within the Labyrinthodontia is uncertain. Those
features which it shares with other forms are almost certainly primitive for labyrinthodonts as a
whole and, until more complete material is available, the taxonomic position of Doragnathus will
remain obscure. Only in the arrangement of the dentition, notably the uniform structure and large
number of marginal teeth, does the lower jaw of Doragnathus differ significantly from those of other
early labyrinthodonts.

Acknowledgements. I wish to thank the staff of the Royal Scottish Museum, Edinburgh, for permission to
borrow and study specimens in their care. Dr. A. L. Panchen kindly read and commented on the manuscript.
This work was completed while I was a Junior Research Associate financed by a Natural Environmental
Research Council Grant (Number GR3/2983) awarded to Dr. A. L. Panchen.

REFERENCES

Andrews, s. m., browne, M. A. e., panchen, A. L. and wood, s. p. 1977. Discovery of amphibians in the Namurian
(Upper Carboniferous) of Fife. Nature , Lond. 265, 529-532.
beaumont, e. h. 1977. Cranial morphology of the Loxommatidae (Amphibia: Labyrinthodontia). Phil. Trans.
R. Soc. B280, 29-101.

bystrow, a. p. 1938. Dvinosaurus als neotenische Form der Stegocephalen. Acta zool., Stockh. 19, 209-295.

— and efremov, J. a. 1940. Benthosuchus sushkini Efr. — A labyrinthodont from the Eotriassic of Sharjenga
River. Trans. Inst. Pal. Acad. Sci. URSS , 10, 1-152.

Campbell, k. s. w. and bell, M. w. 1977. A primitive amphibian from the Late Devonian of New South Wales.
Alcheringa, 1, 369-381.

davis, l. h. 1936. The geology of Inchkeith. Trans. R. Soc. Edinb. 58, 753-786.
forsyth, i. H. and Chisholm, j. i. 1977. The geology of east Fife. Mem. Geol. Surv. Gt. Br.
holmes, r. and carroll, r. l. 1977. A temnospondyl amphibian from the Mississippian of Scotland. Bull. Mus.
comp. Zool. Harv. 147, 489-511.

jarvik, e. 1952. On the fish-like tail in the ichthyostegid stegocephalians. Meddr. Gronland, 114, no. 12, 1-90.
neves, r., et al. 1973. Palynological correlations within the Lower Carboniferous of Scotland and northern
England. Trans. R. Soc. Edinb. 69, 23-70.

nilsson, t. 1944. On the morphology of the lower jaw of Stegocephalia with special reference to Eotriassic
stegocephalians from Spitsbergen. II, General Part. K. svenska. vetensk. Akad. Handl. (3) 21, 1-70.
panchen, a. l. 1973. On Crassigyrinus scoticus Watson, a primitive amphibian from the Lower Carboniferous of
Scotland. Palaeontology, 16, 179-193.

— 1 975. A new genus and species of anthracosaur amphibian from the Lower Carboniferous of Scotland and
the status of Pholidogaster pisciformis Huxley. Phil. Trans. R. Soc. B269, 581-640.

— In press. The origin and relationships of the anthracosaur amphibia from the Late Palaeozoic. In panchen,
a. L. (ed.), The terrestrial environment and the origin of land vertebrates, Systematics Association Special
Volume. London, Academic Press.

save-soderbergh, G. 1932. Preliminary notes on Devonian stegocephalians from East Greenland. Meddr.
Groland, 94, 1-107.

schultze, h.-p. 1969. Die faltenjahne der rhipidistiiden Crossopterygier, der Tetrapoden und der
Actinopterygier-Gattung Lepisosteus. Palaeontogr. ital. 65 (n.s. 35), 59-137.

SMITHSON: L AB Y RINTHODONT AMPHIBIAN FROM SCOTLAND

923

Smithson, t. r. In press. An early tetrapod fauna from the Namurian of Scotland. In panchen, a. l. (ed.). The
terrestrial environment and the origin of land vertebrates. Systematics Association Special Volume. London,
Academic Press.

Thomson, k. s. and bossy, K. h. 1970. Adaptive trends and relationships in early Amphibia. Forma et Functio, 3,
7-31.

tilley, E. h. 1971. Morphology and taxonomy of the Loxommatoidea (Amphibia). Ph.D. thesis, University of
Newcastle upon Tyne.

vaughn, p. p. 1972. More vertebrates, including a new microsaur from the Upper Pennsylvanian of Central
Colorado. Contr. Sci., nat. Hist. Mas. Los Angeles , 223, 1-30.

williston, s. w. 1914. The osteology of some American Permian vertebrates. J. Geol. 22, 364-419.

T. R. SMITHSON

Department of Zoology
The University

Newcastle upon Tyne NE1 7RU

Typescript received 24 September 1979
Revised typescript received 15 January 1980

A LYSOROPHID AMPHIBIAN FROM THE COAL
MEASURES OF NORTHERN ENGLAND

by M. J. BOYD

Abstract. A description is given of the presacral vertebrae and ribs of a lysorophid amphibian from the Middle
Coal Measures (Westphalian B) of Northumberland. The specimen is the earliest lysorophid yet described and is
the first certainly identifiable member of the group to be recorded from any horizon outside North America. An
isolated presacral vertebra from an unknown Coal Measures horizon at Low Moor, West Yorkshire, may
represent additional evidence of lysorophids in the British Carboniferous. Lysorophids appear to have been
present in both lacustrine and coal swamp pool environments in the Upper Carboniferous as well as surviving in
‘red bed’ environments in the Lower Permian in North America.

The Lysorophidae is a family of small ‘lepospondyl’ ( sensu Romer 1966) amphibians hitherto known
with certainty only from Upper Carboniferous and Lower Permian freshwater deposits in North
America. The lysorophids have in the past been assigned to a number of different amphibian taxa,
including the Orders Apoda (Moodie 1909) and Urodela (Sollas 1920; von Huene 1956). Romer
( 1 966) regarded the lysorophids as constituting a family of aberrant microsaurs; in a recent discussion
of lysorophid structure and relationships, however, Carroll and Gaskill (1978, p. 186) have suggested
that the members of the group are sufficiently distinct from typical microsaurs to warrant exclusion
from the Order Microsauria.

The type genus of lysorophid, Lysorophus, was first described by Cope (1877) on the basis of three
isolated vertebrae from the Upper Pennsylvanian of Danville, Illinois. Unfortunately the absence of
more diagnostic material makes it impossible to distinguish Lysorophus from other Carboniferous
lysorophids (Carroll and Gaskill 1978). One of the most fully known of described Carboniferous
lysorophids is Cocytinus Cope 1871, from the Westphalian D horizon of Linton in Ohio. An
articulated specimen of Cocytinus from Linton has recently been figured by Carroll and Gaskill
(1978, fig. 132b). A lysorophid referable to Cocytinus has also been reported by Baird (1964) from the
lower Westphalian D of Mazon Creek, Illinois. Relatively abundant lysorophid material, usually
referred to the genus Lysorophus , is known from the Lower Permian of Texas. Many of these last
specimens are preserved in a matrix which renders preparation difficult, but serial-sectioning
techniques enabled Sollas (1920) to give a detailed account of the skull and the anterior postcranial
skeleton of one specimen. Further Lower Permian lysorophid material has, more recently, been
described by Olson (1971) from the Hennessey Formation of Oklahoma.

No description has hitherto been published of a lysorophid from any locality outside North
America. Although lysorophids have been reported from the Westphalian A ox-bow lake site of
Jarrow in Co. Kilkenny, Eire (Thomson and Bossy 1970), the very poor state of preservation of most
of the specimens from this locality (Rayner 1971) makes definite identification difficult. A small
amphibian from the late Stephanian or early Autunian of Nievre in France, tentatively identified as
an ai'stopod by Thevenin (1910), may also possibly be a lysorophid (Baird 1964) but, as at Jarrow,
preservation is very poor and certain identification is not possible.

MATERIALS

The following description is of a previously undescribed lysorophid specimen which was collected,
probably during the latter half of the nineteenth century, from the Coal Measures of Northumber-
land. The specimen, registered in the collections of the Hancock Museum, Newcastle upon Tyne, as

IPalaeontology, Vol. 23, Part 4, 1980, pp. 925-929.1

926

PALAEONTOLOGY, VOLUME 23

G91.15, is from the black shale immediately overlying the Low Main Seam at the colliery of
Newsham near Blyth. This horizon lies within the Upper Modiolaris zone of the Middle Coal
Measures (Land 1974) and is Westphalian B in age. Because the specimen lacks a skull, it is
impossible to diagnose it at generic or specific level, and the specimen is therefore not named.
However, it merits description as the first certainly identifiable lysorophid to be recorded from
outside North America.

A single presacral vertebra, until recently housed in the Geology Museum of the Wigan College of
Technology in Wigan, Lancashire, but now registered as G 152.04 in the Hancock Museum
collections, may represent additional evidence of the presence of lysorophids in the British Upper
Carboniferous and is also described below. The vertebra was collected between 1880 and 1920 from
the Coal Measures of Low Moor, near Bradford in west Yorkshire. Unfortunately, its precise
horizon is not recorded. However, the holotype specimen of the large eogyrinid embolomere
Pholiderpeton scutigerum Huxley 1 869, which is also from the Low Moor area, was collected from the
shale overlying the Black Bed Coal at Toftshaw and this horizon lies in the Lower Communis zone of
the Lower Coal Measures (Westphalian A) (Panchen 1970). It is possible that vertebra G 152.04 was
collected from the same horizon.

DESCRIPTION

As preserved, specimen G91.15 (text-fig. 1 a) consists of a small slab of shale bearing an articulated series of
eighteen well-preserved presacral vertebrae, in addition to the fragmentary and incomplete neural arches of the
three preceding vertebrae. Most of the neural arches present have become detached from their respective centra
but the component parts of almost all vertebrae still lie closely adjacent to one another and exhibit few traces of
distortion or crushing. Trunk ribs are associated with the majority of the twenty-one vertebrae represented in
G9 1 . 1 5. There is no obvious variation in the structure of either vertebrae or ribs within the preserved series. This
uniformity of structure would seem to suggest that the complete animal possessed an, at least moderately,
elongated presacral vertebral column, and it is of interest to note that Cocytinus is known to possess
approximately seventy-two presacral vertebrae (Carroll and Gaskill 1978).

In a typical anterior presacral vertebra the centrum is holospondylous, spool-shaped and deeply
amphicoelous. It is probable that, as in Lysorophus (Sollas 1920), the centrum is perforated for passage of the
notochord. There is no evidence of the presence of distinct intercentra in G91.15. The external surface of the
centrum is excavated ventro-laterally to form a pair of longitudinally elongate depressions which extend almost
the full length of the element. A pair of similar depressions is present dorso-laterally but these are restricted to the
posterior half of the centrum. In the anterior one-third of the centrum is situated a pair of dorsally directed
depressed facets for articulation with the neural arch pedicels. The articulation between neural arch and centrum
is clearly sutural in all the vertebrae present in G91.15. The most notable feature of the neural arch is its
ossification in two separate halves, with the line of separation running along the length of the neural spine. The
presence of a longitudinally divided neural arch and spine in the trunk vertebrae was clearly demonstrated in
Lysorophus by Sollas (1920). Although occurring in very immature microsaurs (Carroll and Gaskill 1978) and
the adults of some ‘labyrinthodont’ amphibians, this phenomenon is not known in adult iepospondyls’ except in
the lysorophids. The neural spine is very much reduced and consists of a scarcely perceptible ridge running the
length of the dorsal surface of the neural arch. Both pre- and postzygapophyses possess horizontally orientated
articular surfaces. In the anterior one-third of the neural arch, at the level of the neurocentral articulation, is
situated a pair of elongate, antero-laterally directed diapophyses. A similar orientation of the diapophyses
is present in the anterior dorsal vertebrae of Lysorophus described by Sollas (1920, fig. 42).

The ribs of G9 1.1 5 (text-fig. la, d) are dichocephalous and possess long, curved shafts which are compressed
antero-posteriorly. None of the vertebrae show any evidence of a facet to receive the rib capitulum. The probable
original relationships of neural arch, centrum and rib are shown in text-fig. 1 d.

DISCUSSION

The vertebrae of G91.15 are typically lysorophid in structure and very closely resemble those of
Lysorophus as described by Sollas (1920). A series of Sollas’s restorations of the dorsal vertebrae
of Lysorophus, based upon serial sections, is figured (text-fig. 1 e-g) for comparison with those of

BOYD: LYSOROPHID AMPHIBIAN FROM NORTHERN ENGLAND

927

text-fig. 1. Presacral vertebrae of lysorophids. a, semi-diagrammatic representation of Hancock
Museum specimen G91.15 as preserved; b-d, restoration of an anterior trunk vertebra of specimen
G9 1 . 1 5 in b, right lateral view; c, dorsal view and d, anterior view articulated with proximal part of
trunk rib; e-g, anterior trunk vertebra of Lysorophus in e, right lateral view; /, dorsal view and g,
anterior view articulated with proximal part of trunk rib ( e-g after Sollas); h-i, Hancock Museum
specimen G1 52.04 in h, right lateral view as preserved and i, dorsal view with probable original extent
of zygapophyses restored. Cross-hatching indicates broken bone surface.

928

PALAEONTOLOGY, VOLUME 23

G91.15. Among the more significant resemblances between described lysorophids and specimen
G9 1.15 may be cited the following:

1 . Dorsal vertebrae with a neural arch ossified in separate lateral halves.

2. A sutural, rather than fused, neurocentral articulation.

3. A neural spine reduced to a low ridge on the neural arch.

4. Prominent zygapophyses with horizontally orientated articular surfaces.

5. A holospondylous and deeply amphicoelous centrum.

6. The possession of elongate, dichocephalous trunk ribs, the tuberculum of which articulates with
a diapophysis.

Whilst characters 2-6 are all parallelled in other iepospondyl’ taxa and 3-5 are present
simultaneously in aistopods (e.g. Baird 1964), character 1 and hence the combination of all six listed
characters is apparently unique, amongst described ‘lepospondyl’ amphibians, to the Lysorophidae.

Hancock Museum specimen G 152.04, from an unknown Coal Measures horizon at Low Moor
(text-fig. 1 h-i) is less certainly lysorophid. Although resembling the vertebrae of G91.15 in the
possession of a holospondylous, deeply amphicoelous centrum, and a neural arch with a much-
reduced neural spine and horizontally orientated zygapophyseal articular surfaces, G1 52.04 differs in
two respects. The neural arch is ossified as a single structure and would appear to be firmly united,
and possibly fused, to the centrum. It is possible, however, that both characters may be age-related or
may represent regional variation within the vertebral column, and G 152.04 is, therefore, here very
tentatively attributed to the Lysorophidae. The relative shortness of the diapophyses compared with
those of the vertebrae of G9 1 . 1 5 and their directly lateral, rather than antero-lateral, orientation may
indicate that vertebra G 1 52.04 derives from a more posterior region of the presacral vertebral column
than is present in the former specimen.

In addition to representing the first certain record of lysorophid amphibians outside North
America, specimen G91.15 is the earliest recognizable lysorophid yet described. The stratigraphic
range of previously described members of the Lysorophidae extends from the Upper Freeport Coal
of Linton, Ohio (Upper Allegheny or lower Westphalian D), where the group is represented by the
genera Cocytinus Cope 1871 and Molgophis Cope 1868 (Steen 1931), to the Choza Formation of the
Texas Clear Fork Group (Leonardian, Lower Permian) which has yielded specimens referable to
Lysorophus Cope 1877 (Olson 1958). The presence of specimen G91.15 in the black shale overlying
the Low Main coal seam at Newsham extends the known range of the lysorophids down into the
Upper Modiolaris zone of the Middle Coal Measures (Westphalian B).

The uncertainties as to the horizon and relationships of the amphibian represented by the isolated
vertebra G 1 52.04 must debar it from consideration in any discussion of the stratigraphic range of the
Lysorophidae. Of greater importance is the possibility of the presence of lysorophids in the small
tetrapod assemblage from Jarrow in Co. Kilkenny. The Jarrow Seam lies in the lower part of the
Communis zone (Eagar 1964) of the Lower Coal Measures (Westphalian A) and, should the reported
material prove to be diagnostically lysorophid, this would considerably antedate the Newsham
specimen described above.

The fact that the, approximately 200, amphibian specimens from Newsham in the collections of the
Hancock Museum include only the single lysorophid described in this present study suggests that
G91.15 may possibly be a transported specimen rather than a normal member of the Newsham
fauna. Romer (1930) reported only four lysorophid specimens amongst approximately 170 tetrapod
fossils from Linton, and it may be that lysorophids were atypical members of permanent water-body
communities in the Carboniferous and possibly erratics from small ponds and streams of a more
temporary nature. If, however, G91.15 is interpreted as a genuine member of the Newsham
amphibian community (previously described members of which have been listed by Land 1974, p. 61)
the nature of its environment in life is of some interest. The black shale at Newsham is usually
considered to represent the sapropel deposited in a large and deep, possibly coastal or deltaic, lake
(Panchen 1970) which was almost certainly the original environment of most of the larger fish and
amphibians known from this site. Milner (1978) has noted that Coal Measures lake deposits such as

BOYD: LYSOROPHID AMPHIBIAN FROM NORTHERN ENGLAND

929

that at Newsham appear to be characterized by a rather limited assemblage of amphibians including
eogyrinid embolomeres, loxommatids, the nectrideans Keraterpeton and Batrachiderpeton, and the
ai'stopod genus Ophiderpeton, all of which taxa are scarcely or not at all represented in the pond or
small pool faunas such as those of Linton or Nyrany in Czechoslovakia. Pointing out that, unlike the
small tetrapod assemblages of the latter sites, the member groups of the Coal Measures lacustrine
fauna appear to have no representatives in the Permian, Milner (1978) suggested that the demise of
this assemblage was due to reduction in number, and loss in continuity, of large lakes in Euramerica
as a result of the late Carboniferous Armorican orogeny. In view of this hypothesis it is interesting to
note the possibility that lysorophid amphibians inhabited both lacustrine (Newsham) and coal
swamp pool (Linton) environments in the Coal Measures, and that the group also survived into the
Lower Permian in North America. Olson (1958) has described lysorophid aestivating burrows from
the Texas Clear Fork Group, and it seems not unlikely that the acquisition of the ability to aestivate
under dry conditions may have been an important factor in the adaptation of the lysorophids to the
conditions of life prevailing in the early Permian.

Acknowledgements. My thanks are due first to Mr. A. M. Tynan, Curator of the Hancock Museum, and Dr.
Robin Grayson of the Geology Section of the Wigan College of Technology for permission to describe the
specimens in their care. I would also like to thank Miss Susan Turner (Hancock Museum) for her kindness and
assistance during the period of preparation of this work and Dr. A. R. Milner (Birkbeck College, University of
London) for his most helpful advice.

baird, D. 1964. The ai'stopod amphibians surveyed. Breviora, 206, 1-17.

carroll, R. L. and gaskill, p. 1978. The Order Microsauria. Mem. Am. Phil. Soc. 126, 1-211.

cope, E. D. 1868. Synopsis of the extinct Batrachia of North America. Proc. Acad. nat. Sci. Philad. 1868, 208-221.

— 1871. Observations on the extinct Batrachian fauna of the Carboniferous of Ohio (Linton). Proc. Am. phil.
Soc. 12, 177.

— 1 877. Descriptions of extinct vertebrata from the Permian and Triassic formations of the United States.
Ibid. 17, 183-193.

eagar, R. m. c. 1964. The succession and correlation of the Coal Measures of south-east Ireland. C.R. Congr. Int.
Carb. Stratigr. (Paris, 1963) 1, 359-374.

huene, F. von. 1956. Palaontologie und Phylogenie der Niederen Tetrapoden. Jena, Fischer.
huxley, t. h. 1869. On a new labyrinthodont from Bradford. Q. Jl. geol. Soc. Lond. 25, 309-310.
land, D. H. 1974. Geology of the Tynemouth district. Mem. geol. Surv. Gt Br. 15, 61-62.
milner, a. r. 1978. A reappraisal of the early Permian amphibians Memonomenos dyscriton and Cricotillus
br achy dens. Palaeontology, 21, 667-686.
moodie, R. L. 1909. The Lysorophidae. Amer. Nat. 43, 116-119.

OLSON, E. c. 1958. Fauna of the Vale and Choza: 14 Summary, review, and integration of the geology and the
faunas. Fieldiana, Geol. 10, 397-448.

— 1971. A skeleton of Lysorophus tricarinatus (Amphibia: Lepospondyli) from the Hennessey Formation
(Permian) of Oklahoma. J. Paleont. 45, 443-449.

panchen, A. L. 1970. Handbuch der Palaoherpetologie. Teil 5a Anthracosauria. Stuttgart, Fischer.
rayner, d. h. 1971 . Data on the environment and preservation of late Palaeozoic tetrapods. Proc. Yorks. Geol.
Soc. 38, 437-495.

romer, a. s. 1930. The Pennsylvanian tetrapods of Linton, Ohio. Bull. Am. Mus. nat. Hist. 59, 77-147.

1966. Vertebrate Paleontology, 3rd edn. Chicago, University Press.

sollas, w. j. 1920. On the structure of Lysorophus as exposed by serial sections. Phil. Trans. R. Soc. Lond. B209,
481-527.

steen, m. c. 1931. The British Museum collection of Amphibia from the Middle Coal Measures of Linton, Ohio.
Proc. Zool. Soc. Lond. 1930, 849-891.

thevenin, a. 1910. Les plus anciens quadrupedes de France. Annls Paleont. 5, 1-64.

Thomson, k. s. and bossy, k. h. 1970. Adaptive trends and relationships in early Amphibia. Forma et Functio.
3, 7-31.

REFERENCES

MICHAEL J. BOYD

Typescript received 8 November 1979
Revised typescript received 18 December 1979

Department of Natural History
Kingston-upon-Hull Museum
Queen Victoria Square, Kingston-upon-Hull
North Humberside

FLOATING ORIENTATIONS OF CEPHALOPOD
SHELL MODELS

by R. A. REYMENT

Abstract. Accurately constructed models of ammonoid shells were used in experiments on floating
orientations. These experiments show that inflated shells of the cadicone type float stably, with or without liquid
in the final chambers. Highly compressed involute shells are unstable unless the last two chambers contain liquid.
The highly evolute shell type, represented by Dactylioceras, floats on its side when empty and vertically when the
last four chambers contain liquid.

The present note is a continuation of a series of studies by the writer on the nekroplanktic properties
of cephalopod shells (Reyment 1958, 1968, 1970, 1973). The observations summarized here are based
on the behaviour of four exact models of ammonoid species; namely, the early Turonian Hoplitoides
ingens (von Koenen), Paravascoceras hartti (White), and Pseudaspidocerasl sp., and the Toarcian
Dactylioceras sp. The first three forms were selected from Brazilian specimens in the Palaeontological
Museum of Uppsala University; the Dactylioceras comes from the Jurassic of Great Britain.

In order to test the accuracy of the techniques used for making the models, as well as the structural
assumptions involved, a model of Nautilus pompilius was made. Motion pictures were made of all
experiments.

METHOD OF CONSTRUCTION OF THE MODELS
The models were made from actual specimens in the following manner. The ammonoids were dissected, and the
component parts for the models prepared by means of a commercial vacuum-moulding apparatus. The
technique of vacuum-moulding consists of quickly sucking a preheated sheet of plastic of suitable thickness
around a plaster-of-Paris mould. Vacuum-moulding is a widely used method for making children’s toys. The
required specific gravity (here taken as 2-89) was obtained by copper-plating the plastic parts until the desired
weight had been obtained.

Although the models were produced in as accurate a way as possible, it is difficult to be absolutely sure how
close to the original shell a particular replicate may be. In order to test the reliability of the method of
construction used, a shell of N. pompilius was made as a control. The resulting model is shown in text-figs. 1 a-b,
floating alongside a real shell of the pearly nautilus of about the same size.

text-fig. 1 a-b. Model of Nautilus pompilius floating alongside an actual specimen. In la, the model is to the

right, in \b d is to the left.

[Palaeontology, Vol. 23, Part 4, 1980, pp. 931-936.]

932

PALAEONTOLOGY, VOLUME 23

The experiments were made on empty and weighted shells thereby simulating the effect of the animal in the
body chamber. Salt water at a concentration of thirty-three parts per thousand was used.

INFLATED SHELLS

Two kinds of moderately evolute, inflated shells, with square to broadly oval whorl sections, were
made for studying the properties of this kind of ammonoid. Both were found to possess quite similar
buoyancy properties.

Cadicone shell

The species P. hartti (White) is a typical cadicone, with whorls in adults appreciably wider than they
are high (depressed whorl section). The specimen on which the model was based has a diameter of
20 cm. The following observations were made on the model and simulated ammonoid animal.

1 . All chambers empty: the shell floats with 20% of it above the water, measured in terms of the
diameter at right angles to the water surface. The aperture faces upwards (text-fig. 2a).

2. Three chambers liquid-filled: the shell is just in contact with the water surface but does not break
it (text-fig. 2b). The aperture is lower than for the orientation shown in text-fig. 2a.

3. Fourth chamber quarter-filled: the shell sinks to the bottom. The resting position taken up by
the cadicone is shown in text-fig. 2c.

text-fig. 2. Paravascoceras hartti (White), Early Turonian (Cretaceous), a, floating orientation of empty shell, b,
floating orientation with the last three chambers liquid-filled, c, with some liquid in the fourth last chamber, the
shell sinks, showing resting position of the model. Weight of ‘animal’ added to model (uplift compensated).

A highly ornamented evolute shell

The form here determined as Pseudaspidocerasl sp. has a diameter of 32 cm. Its whorl section is
square to rectangular and the prominent tubercles are hollow and open to the chambers. The
question of whether tubercles are hollow, open or floored, or solid, is of consequence in buoyancy
studies. The ammonoid animal was not simulated in the experiment recorded below.

REYMENT: CEPHALOPOD FLOATING ORIENTATIONS

933

1 . The empty shell: this model floats with 26% of the shell above water; the aperture faces upwards
(text-fig. 3d).

2. Three chambers liquid-filled: a small part of the shell remains above water; the aperture is still
directed upwards.

3. Four chambers liquid-filled: the shell is just buoyant; the aperture is now lower than that shown
in text-fig. 3 a. The orientation for this stage of the experiment is shown in text-fig. 2b.

4. Fifth last chamber quarter-filled: this slight increase brings about an immediate loss of
buoyancy. Prior to this, the shell floated with the body chamber directed upwards (text-fig. 3c).

text-fig. 3, Pseudaspidocerasl sp., Early Turonian (Cretaceous), a, floating orientation of empty shell, b,
floating orientation with last four chambers filled, c, loss of buoyancy occurs when a small amount of liquid is
added to the fifth last chamber.

COMPRESSED INVOLUTE SHELL

The oxynote variety of shell

This was studied by means of a model of H. ingens (von Koenen), based on a very large specimen with
a diameter of 49 cm. The ammonoid animal was not simulated in the experiment recorded below.

1 . Floating position of the empty shell: this is unstable and the shell floats at an angle to the water
surface, with 17% of the shell above water (text-fig. 4 a).

2. Three chambers liquid-filled: a small fraction of the shell remains above water; the aperture of
the body chamber is lower than for the empty shell (text-fig. 4 b).

3. Four chambers liquid-filled: the shell sinks when four chambers are entirely full of liquid, and
the resting position adopted is shown in text-fig. 4c.

SERPENTICONE SHELL

The highly evolute shell

This type was investigated by a model of a specimen of Dactylioceras sp. with a diameter of 23 cm.
The weight of the ammonoid animal was allowed for in the experiment described below (text-figs. 5c- f).

1 . The floating position of an empty shell is illustrated in text-fig. 5a.

2. Three chambers liquid-filled: the shell just breaks the surface of the water but remains
horizontally oriented (text-fig. 5b).

934

PALAEONTOLOGY, VOLUME 23

text-fig. 4. Hoplitoides ingens (von Loenen), Early Turonian (Cretaceous), a, floating position of empty shell, b ,
floating orientation with three last chambers liquid-filled, c, the shell sinks when the fourth chamber contains

liquid.

3. Last four chambers liquid-filled: there is an abrupt change in orientation, and the shell becomes
vertical. It floats upright and is reasonably stable; this is presumably the living position of the
dactylioceratid animal. About 6% of the shell remains above water.

4. Last four and a half chambers liquid-filled: the shell does not break the surface of the water and
sinks gradually to the bottom if struck sharply (text-fig. 5c).

5. If held at the depth indicated in text-fig. 5 d, the shell rises slowly to the surface.

6. If held at the depth indicated in text-fig. 5c, the shell sinks slowly. In both stages (5) and (6) the
shell contains the same amount of liquid. A motion picture is available of this part of the experiment.
Careful frame-by-frame study shows that the serpenticone type of shell, as represented by
Dactylioceras, reacts sluggishly to movement when in a state of buoyancy equilibrium. On the other
hand, it appears to be as stable as, for example, the cadicone with respect to its vertical orientation.

7. Resting position of the model on the bottom of the tank (text-fig. 5 f).

CONCLUDING REMARKS

The suite of experiments briefly reported here indicates the variability in stability shown by various
kinds of ammonoid shell. The most stable of the types studied is represented by shells with depressed
whorl sections; next, are shells with a sub-quadrate whorl section and a moderate degree of evolution.
A highly compressed and involute shell form, such as possessed by many species of Hoplitoides , does
not float in a vertical position when empty. The same observation applies for highly evolute,
serpenticone shells of dactylioceratid type, which when empty float in a horizontal position.
Serpenticones, when normally weighted with cameral liquid, appear remarkably sluggish when forces
are applied to them.

REYMENT: CEPHALOPOD FLOATING ORIENTATIONS

935

text-fig. 5. Dactylioceras sp. Toarcian (Jurassic), a, horizontal floating position of the empty shell, b , with the
last three chambers liquid-filled, the shell just breaks the surface, remaining horizontally oriented, c, with the last
four and a half chambers liquid-filled the shell is in hydrostatic equilibrium, d , held at 6 cm below the surface of
the water, the shell floats to the surface, e, held at 7-5 cm below the surface of the water, the shell sinks slowly to
the bottom./, the resting position of Dactylioceras on the bottom of the tank.

In an earlier study (Reyment 1973), the effect of varying the length of the body chamber on the
floating orientation of empty shells was the main topic of interest. In the present paper the body
chamber was allowed to remain a constant length, the experiments being directed towards studying
the relationships between the amount of liquid in the last chambers of the final whorl and the floating
orientation of the shells.

Compared with Reyment (1958, 1973) and Mutvei and Reyment (1973), the work here summarized
gives answers to several questions which could not be treated with the cruder models used in the
earlier investigations.

Acknowledgements. Very special thanks are due to Mr. Bertil Annell for his interest and skill in making the
ammonoid models. The entire method for producing them was worked out by him in consultation with Mr. Eric
Stahl. Mr. Annell also actively assisted with the experiments. My thanks also to Mr. Gustav Andersson, who
photographed the experiments, including the cinematography. All the above are at Uppsala University.

936

PALAEONTOLOGY, VOLUME 23

REFERENCES

mutvei, H. and reyment, r. a. 1973. Buoyancy control and siphuncle function in ammonoids. Palaeontology, 16,
623-636.

reyment, r. a. 1958. Factors in the distribution of fossil cephalopods. Stockh. Contr. Geol. 1, 97-184.

— 1968. Orthoconic nautiloids as indicators of shoreline surface currents. J. Sedim. Petrol. 38, 1387-1389.

— 1970. Vertically inbedded cephalopod shells. Factors in the distribution of fossil cephalopods, 2.
Palaeogeogr. Palaeoclimat. Palaeoecol. 7, 103-111.

— 1973. Factors in the distribution of fossil cephalopods, 3. Experiments with exact models of certain shell
types. Bull. Geol. Instn. Univ. Uppsala, N.s. 4, 7-41.

R. A. REYMENT

Paleontologiska Institutionen

Typescript received 13 September 1979
Revised typescript received 5 January 1980

Uppsala Universitet
Box 558

S751 22 Uppsala
Sweden

THE PALAEONTOLOGICAL ASSOCIATION

Annual Report of Council for 1979

Membership and Subscriptions. Membership totalled 1,554 on 31 December 1979, a decrease of 8 over the
previous year. There were 986 Ordinary Members (an increase of 6); 218 Student Members (a decrease of 8); and
350 Institutional Members (a decrease of 6). The number of institutions subscribing to Palaeontology via
Blackwell’s agency was 420. Subscriptions to Special Papers in Palaeontology numbered 158 individual and 1 16
Institutional members, decreases of 16 and 1 respectively since 31 December 1978. Subscriptions to Special
Papers through Blackwell’s agency rose from 109 to 131. Sales of back parts of Palaeontology to members via the
Membership Treasurer showed a sharp drop from 33 to 15 transactions in 1979. Sales of back copies of Special
Papers to members rose to 95 transactions, realizing £1,657. Thirty-two members took advantage of the special
offer of A. M. Davies’s Tertiary Faunas negotiated between the Association and the publishers.

Finance and Publications. During 1979 the Association published Volume 22 of Palaeontology in four parts at
a total cost of £37,203 (including £4,267 postage and distribution). It contained 42 papers totalling 982 pages and
130 plates. Special Paper 22 (for 1978): ‘Curation of Palaeontological Collections’ and Special Paper 23: ‘The
Devonian System’, were published in June and August 1979 respectively at a total cost of £19,679. Changes in
page format were introduced in Special Paper 23\ text area was increased by approximately 10% and the type-size
used for most of the text was slightly reduced. This has reduced the number of pages which in turn has lowered
production and distribution costs. The changes will be extended to Palaeontology in 1980. Sales of Devonian
Symposium field guides during 1979 resulted in an income of £814, which more than covers costs of production
and distribution. The Association is very grateful to all those who have made donations. However, costs
continue to rise, and subscription rates for Palaeontology and Special Papers will have to be increased in 1981.

Meetings. Eight meetings were held during 1979. The Association is indebted to the organizers and hosts, and to
those who led field excursions.

a. The Twenty-second Annual General Meeting was held in the Lecture Theatre of the Geological Society of
London on 21 March 1979. Dr. J. H. Callomon (London) delivered the Twenty-second Annual Address on
‘Jurassic Ammonites in Time and Space’.

b. A Symposium on the Terrestrial Environment and the Origin of Land Vertebrates , jointly organized with the
Systematics Association, was held on 18-19 April 1979 at the University of Newcastle upon Tyne.
Approximately seventy people attended the meeting. Dr. A. L. Panchen was local secretary.

c. A Field Meeting was organized by the Carboniferous Group on the ‘Lower Carboniferous of Dovedale,
Manifold Valley and Weaver Hills area of N.E. Staffordshire’, and led by Dr. N. Aitkenhead and Mr. J. I.
Chisholm. Eighty-six people attended the excursion which was held on 20-23 April 1979.

d. A Field Meeting on the ‘Cambrian of the Harlech Dome, North Wales’ was held on 27-29 April 1979 and
led by Dr. A. W. A. Rushton and Dr. P. M. Allen. Fifteen members attended.

e. A Field Meeting on ‘Caradoc benthic communities; Shropshire, Berwyn Hills and Bala’ was held on 14-16
September 1979 and led by Dr. P. J. Brenchley and Dr. M. J. Lockley. Ten members attended.

/. A Symposium on Evolutionary Lineages and Selection Pressures , jointly organised with the British
Micropalaeontological Society, was held on 22 September 1979 as part of the Fourth Meeting of the
Geological Societies of the British Isles at Sheffield, 19-23 September 1979. Dr. R. J. Aldridge organized
the symposium and about fifty members attended.

g. A Working Group on ‘Facies and faunas of the Tethyan Tertiary’ was held on 3 1 October 1979 at Bedford
College, London. Thirty attended the meeting and the local secretary was Dr. E. P. F. Rose.

h. The Annual Christmas Meeting, was an open meeting held at University College, Cardiff, on 17-19
December 1979 and jointly hosted by University College and the National Museum of Wales. One hundred
and twenty members attended, and the President’s Awards were presented to Dr. J. A. Crame and Dr. A. C.
Scott. Joint field excursions with the British Sedimentological Research Group visited the Carboniferous,
Triassic, and Jurassic at Ogmore-by-Sea (leaders: Professor D. V. Ager and Ms. W. Glanvill), the Triassic

938

THE PALAEONTOLOGICAL ASSOCIATION

and Jurassic of Barry and Penarth (leaders: Mr. M. Mayall and Dr. M. Tucker), and the Silurian and
Devonian of southern Powys (leaders: Dr. D. Edwards and Dr. I. Tunbridge). The local secretary was Mr.
E. W. Nield.

Council. The following members of the Association served on Council following the A.G.M. on21 March 1979:
President : Professor H. B. Whittington, F.R.S.; Vice-Presidents'. Dr. E. P. F. Rose, Dr. C. T. Scrutton;
Treasurer. Mr. R. P. Tripp; Membership Treasurer: Dr. J. C. W. Cope; Secretary: Dr. R. Riding; Editors:
Professor C. B. Cox, Dr. M. G. Bassett, Dr. K. C. Allen, Dr. R. A. Fortey; Other members: Dr. R. J. Aldridge,
Dr. M. C. Boulter (Circular Reporter), Dr. M. D. Brasier, Dr. P. J. Brenchley, Dr. D. E. G. Briggs, Dr. C. H. C.
Brunton (Institutional Membership), Dr. S. Conway Morris, Dr. M. B. Hart, Dr. P. M. Kier, Dr. S. C.
Matthews, Dr. I. E. Penn, Dr. M. Romano, Dr. D. J. Siveter, Dr. J. Watson.

Circulars. Four Circulars, 95-98, were distributed to Ordinary and Student Members and on request to over
100 Institutional Members during 1979.

Council Activities. During 1979 Council decided to institute a new category of membership for those who have
been Ordinary members of the Association for at least fifteen years, and who have retired from employment.
This ‘Retired Member’ category will commence in January 1980 and should provide a stimulus for longstanding
members to remain in the Association despite the effects of rising costs on pensions and savings.

A presentation copy of Special Paper 23 ‘The Devonian System’ was awarded to Academician D. V. Nalivkin,
to whom the volume is dedicated, at a ceremony in Leningrad.

Plans for future meetings of the Association include a symposium on ‘Life in the Precambrian’ at Leicester in
April in addition to the established programme of events. It is also proposed to introduce more one-day meetings
on specialist topics. A field meeting on Malta is planned for January 1981 and will be followed soon afterwards
by an excursion to West Germany by the Carboniferous Group.

Professor C. B. Cox completed his term as representative for the Association with the International
Palaeontological Association and his place has been taken by Dr. L. R. M. Cocks. Thanks are due to Professor
Cox.

BALANCE SHEET AND ACCOUNTS FOR THE
YEAR ENDING 31 DECEMBER 1979

1978

£

34,635

10,146

£24,489

Balance Sheet as at 31 December 1979

£ £
Current Assets

26,508

Investments at cost (see schedule). ....

36,904

-

Stock of paper ........

480

581

Devonian Symposium Guides .....

.

1,401

Sundry debtors ........

4,078

6,145

Cash at bank ........

7,768

Current Liabilities

500

Royal Society loan .......

.

1,410

Subscriptions received in advance ....

1,891

-

Provision for cost of publication of Palaeontology.

10,491

7,000

Provision for printing Special Paper No. 23 .

11,383

1,236

Sundry creditors

881

Represented by:

Publications Reserve Account

17,780 Balance brought forward 24,489

6,709 Excess of income over expenditure for the year transferred

from Income and Expenditure Account .... 95

49,230

24,646

£24,584

£24,489

£24,584

Income and Expenditure Account for the Year Ended 31 December 1979

1978

£ INCOME £ £

Subscriptions for 1979 25,468

Subscriptions for 1977/78 1,703

26,585 — 27,171

Palaeontology — Sales 12,482

— Donations 1 153

14,800 — 13,635

Special Papers— Sales 6,535

— Donations ......... 562

7,684 7,097

Profit on sale of Tertiary Faunas 216

281 Offprints— loss (414)

200 Profits on sale of investment 376

Receipt from Adelaide University 505

4,292 Investment Income (see Schedule) 5,009

£53,842 £53,595

EXPENDITURE

Cost of publication of Palaeontology.

Volume 22, Part 1 8,974

Part 2 8,723

Part 3 9,016

Part 4 10,490

34,341 37,203

Cost of publication of Special Papers:

Provision for No. 23 11,351

Under provision for No. 22 1,297

8,771 12,648

Cost of Circulars:

Preparation ........... 1,336

Postage 609

2,271 1,945

Administrative Expenses

Postage and stationery. ......... 694

Editorial expenses 253

Secretarial help

Meeting expenses 483

Membership of Societies 10

Audit fee 190

Grant and awards 25

1,750 1,655

Devonian Symposium Guides: Net loss 49

£47,133 £53,500

Excess of income over expenditure for the year transferred to Publications
£6,709 Reserve Account £95

Schedule of Investments and Investment Income as at 31 December 1979

Gross Income
Cost for Year

£ £

£12,000

134% Exchequer Stock 1987

11,520

518

£1,000

9% T reasury Stock 1 992/1 996

991

90

£1,000

9% Treasury Stock 1994

955

90

£4,000

8% Treasury Stock 2002/2006

2,192

320

£5,357

134% Treasury Stock 1997

5,000

710

£3,000

134 Exchequer Stock 1996

3,000

435

£2,000

Agricultural Mortgage Corporation Ltd. 94% Debenture 1980/1985

1,938

185

£1,500

Bootle Corporation 74% Redeemable Stock 1977/1979— (Redeemed) .

-

116

5,270

M. & G. Charifund units

4,073

785

£2,000

Imperial Group Ltd. 8% Convertible Unsecured Loan Stock 1985/1990

1,405

160

10,000

New Throgmorton Trust Ltd. 25p Income Shares

1,706

299

1,600

Commercial Union Assurance Co. Ltd. 25p Shares ....

2,157

233

600

Consolidated Gold Fields Ltd. 25p Shares

1,012

117

1,000

Clarke, Nicholls & Coombs Ltd. 25p Shares

954

33

4,091

Bank interest (net) ..........

922

Holding charges. ..........

(4)

£36,903

£5,009

Marketvalue at 31 December 1979(1978— £31,212) ....

£40,654

Report of the Auditors to the Members of The Palaeontological Association. In our opinion the accounts as set out
on pages 2 to 4, give a true and fair view of the state of the affairs of the Association at 31 December 1979 and of
its income and expenditure for the year ended on that date.

Chislehurst, Kent , February 1980

D. J. Carey & Co.
Chartered Accountants

INDEX

Pages 1 -236 are contained in Part 1 ; pages 237-474 in Part
in Bold Type indicate plate numbers.

A

Acritarchs: Jurassic of England, 151

Actinocamax verus , 115

Althaspis senniensis sp. nov., 288, 35, 36

Ambocythere campana sp. nov., 110, 11

Ammonites: Cretaceous, 821; Cretaceous, Europe, 325;

Turonian, England and France, 557
Amphibia: Carboniferous, 273; Carboniferous, England,
925; Carboniferous, Ireland, 125; Carboniferous,
Scotland, 915; Pennsylvanian, Nova Scotia, 143
Amphitryon radians , 107, 108

Andrews, P. and Tekkaya, I. A revision of the Turkish
Miocene hominoid Sivapithecus meteai, 85
Anglo-Paris Basin: Cretaceous belemnites, 889
Antia, D. D. J. See Lockley, M. G. and Antia, D. D. J.
Apatognathus cuspidatus, 299, 37; libratus , 300, 37; scal-
enus, 300, 37

Apatopygus recens, 1, 3, 5
Araucaria brownii sp. nov., 658, 83-86
Atlantic: Cretaceous brachiopods, Rockall Bank, 463
Atractopyge scabra, 860, 110; aff. scabra, 860, 111
Azolla colwellensis sp. nov., 213, 23, 24; prisca, 24; schopfi ,
23

B

Bate, R. H. See Sheppard, L. M. and Bate, R. H.
Belemnitella praecursor, 115

Benton, M. J. and Trewin, N. H. Dictyodora from the
Silurian of Peeblesshire, Scotland, 501
Biometry: outline processing, 757
Botulocyprideis simplex gen. et sp. nov., 104, 9
Boubeithyris pleta sp. nov., 532, 56, 60; tibourrensis sp.
nov., 530, 56

Boyd, M. J. The axial skeleton of the Carboniferous
amphibian Pteroplax cornutus, 273
Boyd, M. J. A lysorophid amphibian from the Coal
Measures of northern England, 925
Brachiopods: Cretaceous, Morocco, 515; Cretaceous,
Rockall Bank, North Atlantic, 463; Palaeozoic, 707
Brissopsis atlantica, 6; lyrifera, 1, 4-6
Brissus latercarinatus, 5

Britain: Eocene spores, 213. See also England; Scotland;
Wales

Brongniartella cf. marocana, 868, 112
Bryozoa: Jurassic, England, 699

C

Caddasphaera gen. nov., 164; halosa, 164, 14
Calymene blumenbachii blumenbachii, 97, 100; blumen-
bachii subsp. nov., 100; (s.l.) cf. prolata , 866, 111

2; pages 475-713 in Part 3; pages 715-947 in Part 4. Figures

Calyptaulax planiformis, 870, 112
Cambrian: trilobites, 171

Camptosaurus depressus, 438; dispar, 437, 52; prestwichii,

438, 51, 52

Carboniferous: Amphibia, 273; amphibian, England,
925; amphibian, Ireland, 125; amphibian, Scotland,
915; conodont faunas. West Malaysia, 297; Pennsyl-
vanian amphibians. Nova Scotia, 143; Pennsylvanian
fishes, North America, 315; Scottish shark, 363
Caribbean: pre-Miocene seagrass communities, 231
‘ Carneithyris ’ rockallensis sp. nov., 465
Cephalopods: body extension in locomotion, 445; orien-
tation, 931

Ceraurinella intermedia , 110
Chaetetes ( Boswellia ) mortoni sp. nov., 808, 102, 103
Chamberlain, J. A. Jun. The role of body extension in
cephalopod locomotion, 445
Chasmops odini, 29, 32, 33; sp. nov. A, 25, 27-29; sp. nov.
B, 26, 29, 32; sp. nov. C, 26, 32; sp. nov. D, 29, 32, 33;
1C. odini, 29

Clypeaster rarispina, 1; rosaceus, 1, 4
Clypeus sp., 2

Cochleosaurus florensis sp. nov., 143
Collignoniceras boreale, 586, 170; canthus, 582, 73;
carolinum, 574, 68, 76; papale, 578, 69, 70; turoniense,
584, 71, 72; woollgari, 560, 62-67, 69, 71
Collinson, M. E. A new multiple-floated Azolla from the
Eocene of Britain with a brief review of the genus, 213,
23, 24

Columbia: Plio-Pleistocene ostracods, 97
Communities: seagrass, Caribbean pre-Miocene, 23 1
Conodonts: Carboniferous, West Malaysia, 297
Coral reefs: Pleistocene, Kenya, 1
Costa, L. See Liengjarern, M., Costa, L. and Downie, C.
Crabs: Miocene, New Zealand, 471
Crame, J. A. Succession and diversity in the Pleistocene
coral reefs of the Kenya coast, 1
Cretaceous: ammonites, 821; belemnites, Anglo-Paris
Basin, 889; brachiopods, Morocco, 515; brachiopods,
Rockall Bank, North Atlantic, 463; capulid gastropod,
Saghalien and Japan, 689; European ammonites, 325;
Turonian collignoniceratid ammonites, England and
France, 557

Cretaceous-Oligocene: Caribbean seagrass communities,
231

Crinoids: stereom structure, 749

Cybeloides (Paracybeloides) girvanensis, 110, 111

Cypria aqualica sp. nov., 116, 13

Cyprideis purperi colombiaensis subsp. nov., 101, 8;
pur peri purperi subsp. nov., 99, 7, 8; purperi sp. nov.,
99

Cyrtothyris middlemissi, 536, 57
Cytheridella postornata sp. nov., 108, 10

944

INDEX

D

Dactylioceras ( Orthodactylites ) clevandicum, 652, 82; (O.)
crosbeyi, 654, 82; ( O .) semi-celatum, 646, 80-82; (O.)
tenuicostatum, 650, 82
Darwinula sp., 117, 13
Dendraster excentricus, 3, 4
Dendrerpeton, 125

Devonian: pteraspidid ostracoderm, Senni Beds For-
mation, South Wales, 287
Diacalymene diademata , 101
Dichadogonyaulax adelos sp. nov., 155, 14
Dick, J. R. F. and Maisey, J. G. The Scottish Fower
Carboniferous shark Onychoselache traquairi, 363
Dictyodora: Silurian, Scotland, 501
Dindymene longicaudata , 862, 112
Dinoflagellate cysts: Eocene-Oligocene, Isle of Wight,
475; Jurassic, England, 151
Dinoflagellates: plate patterns, 667
Dinosaurs: Jurassic, England, 41 1
Dionide sp. indet., 854, 109
Distatodinium scariosum sp. nov., 477, 54
Doragnathus gen. nov., 916; woodi gen. et sp. nov., 916
Downie, C. See Liengjarern, M., Costa, L. and Downie, C.
Duftonia geniculata, 869, 112

E

Eaton, G. L. Nomenclature and homology in peridini-
alean dinoflagellate plate patterns, 667
Eccoptochile ( Eccoptochile ) almadenensis sp. nov., 610,
78, 79; (IE.) cf. clavigera , 614, 79; (IE.) mariana , 607,
78; (IE.) cf. mariana, 79
Echinocardium cordalum, 1-6; pennatifidum, 5
Echinocyamus pusillus, 1
Echinoids: morphology and function, 39
Echinolampas crassa, 3, 5
Echinoneus cyclostomus, 2

Ecological succession: Pleistocene coral reefs, Kenya, 1
Emslandia sp., 481 , 54
Encope michelini, 4

England: Carboniferous amphibian, 925; Eocene/
Oligocene dinoflagellate cysts. Isle of Wight, 475;
Jurassic araucarian cone, 657; Jurassic Bryozoa, 699;
Jurassic dinoflagellate cysts and acritarchs, 151; Jur-
assic ornithischian dinosaur, 411; Liassic correlation,
637; Tertiary fungi, 205; Turanian collignoniceratid
ammonites, 557
Entobia cretacea, 115
Eocene: spores of Azolla, Britain, 213
Eocene/Oligocene: dinoflagellate cysts, Isle of Wight, 475
Eocladopyxis tessellata sp. nov., 481, 53
Europe: Cretaceous ammonites, 325
Eva, A. N. Pre-Miocene seagrass communities in the
Caribbean, 231

Evolution: predatory gastropods, 375; Silurian trilobites,
Welsh Borderland, 783

F

Farrow, G. E. and Owen, E. F. Shallow-water Cretaceous
brachiopods from Rockall Bank, North Atlantic, 463
Fenton, J. P. G„ Neves, R. and Piel, K. M. Dinoflagellate
cysts and acritarchs from Upper Bajocian to Middle
Bathonian strata of central and southern England, 151,
14-16

Fishes: Pennsylvanian, North America, 315

Flexicalymene cavei, 867

Forest, J. See Hyden, F. M. and Forest, J.

France: Miocene horse, 617; Turonian collignoniceratid
ammonites, 557

Functional morphology: Cambrian trilobites, 171; echi-
noids, 39

Fungi: Tertiary, England, 205

G

Galton, P. M. and Powell, H. P. The ornithischian
dinosaur Camptosaurus prestwichii from the Upper
Jurassic of England, 41 1, 51, 52
Gastrochaenolites ichnosp., 96

Gastropods: Cretaceous from Saghalien and Japan, 689;

evolution, 375
Geniculatus claviger, 302, 37
Gerdiocysta conopeum get. et sp. nov., 483, 53
Gigantocapulus giganteus, 87
Glaphurella cf. harknessi, 880, 113, 114
Glaphyrocysta paupercula sp. nov., 483, 53
Gnathodus bilineatus, 302, 38; commutatus, 304, 38; girtyi
rhodesi, 304, 38; girtyi simplex, 304, 38; nodosus, 304, 38
Gongylodinium erymnoteichos gen. et sp. nov., 158, 14;

hocneratum gen. et sp. nov., 1 59, 16
Gonioteuthis granulata, 115; quadrata quadrata, 116
Gray, D. I. Spicule pseudomorphs in a new Palaeozoic
chaetetid, and its sclerosponge affinities, 803, 102, 103

H

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

Harpidella (Harpidella) lacrymosa sp. nov., 848, 108
Hayami, I. and Kanie, Y. Mode of life of a giant capulid
gastropod from the Upper Cretaceous of Saghalien
and Japan, 689, 87
Hemiaster expurgitus, 1, 3
Hiatella (Pseudosaxicava) foetida, 96
Hibbardella acuta, 304, 37: geniculata, 305, 37; pennata,
305, 37

Hindeodella ibergensis, 305, 37; mehli, 305, 37; uncata,
305, 37; undata, 306, 37
Holectypus depressus, 2
Hominoids: Miocene, Turkey, 85
Horse: Miocene, North America and France, 617
Hourcquiceras hourcqi, 39

Howarth, M . K. The T oarcian age of the upper part of the
Marlstone Rock Bed of England, 637, 80-82
Hyden, F. M. and Forest, J. An in situ hermit crab from
the early Miocene of southern New Zealand, 471

I

Idiognathoides noduliferus inaequalis, 306, 38; noduliferus
japonicus, 306, 38; noduliferus noduliferus, 306, 38
Illaenus (Parillaenus) cf. fallax, 848, 107, 108
Ireland: Carboniferous amphibian, 125

J

Japan: Cretaceous capulid gastropod, 689
Jarvis, I. Palaeobiology of Upper Cretaceous belemnites
from the phosphatic chalk of the Anglo-Paris Basin,
889, 115, 116

INDEX

945

Jeanrogericeras reveliereanus, 826, 105, 106
Juralina ecruensis sp. nov., 549, 59
Jurassic: araucarian cone, England, 657; bivalves, 769;
Bryozoa, England, 699; dinoflagellate cysts and acri-
tarchs, England, 151; ornithischian dinosaur, England,
411

K

Kanie, Y. See Hayami, I. and Kanie, Y.

Keegan, J. B. See Sevastopulo, G. D. and Keegan, J. B.
Kelly, S. R. A. Hiatella— a Jurassic bivalve squatter?,
769, 96

Kennedy, W. J., Wright, C. W. and Hancock, J. M.
Collignoniceratid ammonites from the mid-Turonian
of England and northern France, 557, 62-77
Kennedy, W. J., Wright, C. W. and Hancock, J. M.
Origin, evolution and systematics of the Cretaceous
ammonoid Spathites, 821, 104-106
Kennedy, W. J., Wright, C. W. and Hancock, J. M. The
European species of the Cretaceous ammonite Roma-
niceras with a revision of the genus, 325, 39-50
Kenya: Pleistocene coral reefs, 1

Kutchithyris acutiplicata, 60; brivesi , 544, 58, 59; kennedyi
sp. nov., 542, 58

Kylindrocysta spinosa gen. et sp. nov., 162, 14, 15

L

Lecointriceras carinatum sp. nov., 598, 76; costatum sp.

nov., 598, T7;fleuriausianum, 590, 74, 75, 77
Liassic: Marlstone Rock Bed, Toarcian age, England, 637
Liengjarern, M., Costa, L. and Downie, C. Dinoflagellate
cysts from the Upper Eocene-Lower Oligocene of the
Isle of Wight, 475, 53, 54
Ligonodina roundyi, 307, 37
Liocnemis recurvis, 870, 112, 113
Lithodinia valensii, 15

Lockley, M. G. and Antia, D. D. J. Anomalous occur-
rences of the Lower Palaeozoic brachiopod Schizo-
crania, 707

Loeffler, E. J. and Thomas, R. G. A new pteraspidid
ostracoderm from the Devonian Senni Beds Formation
of South Wales and its stratigraphic significance, 287,
35, 36

Lonchodina bischoffi, 307, 37; ponderosa, 307, 37
Lonchodomas cf. drummuckensis , 856, 110; aff. pennatus,
855, 109, 110

Loriolithyris marocensis sp. nov., 529, 56, 60; melaitensis
sp. nov., 528, 55, 60, 61; russillensis, 522, 55, 61;
valdensis , 524, 55

M

Macfadden, B. J. The Miocene horse Hipparion from
North America and from the type locality in southern
France, 617

Maisey, J. G. See Dick, J. R. F. and Maisey, J. G.

Maretia planulata, 4

Mellita quinquiesperforata , 3, 4

Metcalfe, I. Upper Carboniferous conodont faunas of the
Panching limestone, Pahang, West Malaysia, 297, 37,

38

Middlemiss, F. A. Lower Cretaceous terebratulidae from
south-western Morocco and their biogeography, 515,

55-61

Milner, A. R. The temnospondyl amphibian Dendrer-
peton from the Upper Carboniferous of Ireland, 125
Miocene: hermit crab. New Zealand, 471; horse from
North America, and France, 617; Turkish hominoid,
85

Moira atropos, 2

Morocco: Cretaceous brachiopods, 515
Morphology: stereom structure of crinoids, 749
Morris, N. J. See Taylor, J. D., Morris, N. J. and Taylor,
C. N.

N

N ankinolithus cf. grannulatus, 851, 109
Neoprioniodus scitulusl , 307; singularis , 307, 37
Neves, R. See Fenton, J. P. G., Neves, R. and Piel, K. M.
New Zealand: Miocene hermit crab, 471
North America: Miocene horse, 617; Pennsylvanian
fishes, 315

Norway: Ordovician trilobites, Oslo region, 715
Nova Scotia: Pennsylvanian amphibians, 143

O

Olenoides serratus , 17-22
Oligocene: see Eocene/Oligocene
Onychoselache traquairi, 364
Opsimasaphus sp. indet., 844, 107
Ordovician: trilobites, Oslo region, 715; trilobites,
Portugal, 605; trilobites. South Wales, 839
Ostracoderms: Devonian Senni Beds Formation, South
Wales, 287

Ostracods: Plio-Pleistocene, Colombia and Peru, 97
Otarocyprideis elegans gen. et sp. nov., 101, 8, 9
Owen, A. W. The trilobite Tretaspis from the Upper
Ordovician of the Oslo region, Norway, 715, 89-93
Owen, E. F. See Farrow, G. E. and Owen, E. F.
Ozarkodina deliactula, 308, 37; sp., 308, 37

P

Pagurus clifdenensis sp. nov., 471

Palaeobiology: Cretaceous belemnites, Anglo-Paris

basin, 889

Palaeozoic: anomalous occurrences of brachiopods, 707;

chaetitid and sclerosponge affinities, 803
Panderia edita , 847, 108

Paraboubeithyris plicae gen. et sp. nov., 534, 56, 57, 61
Paracypris sp., 117, 13
Paramaretia peloria, 1, 5
Pelocypris sp., 108, 10; zilchi, 104, 10
Perissocytheridea formosa sp. nov., 113, 12; P.1 elongata
sp. nov., 1 16, 12

Peru: Plio-Pleistocene ostracods, 97
Phacops granulatus , 30, 31, 34

Phelodinium pachyceras sp. nov., 486, 54; pumilum sp.
nov., 487, 54

Phthanoperidinium amiculum sp. nov., 487, 53; flebile sp.
nov., 488, 54

Piel, K. M. See Fenton, J. P. G., Neves, R. and Piel, K. M.
Plants: Jurassic araucarian cone, England, 657
Platylichas angulatus , 874, 113, 114; noctua sp. nov., 871,

113, 114

Pleistocene: coral reefs, Kenya, 1

Plio-Pleistocene: ostracods from Colombia and Peru, 97

946

INDEX

Pontocyprisl sp., 118, 13
Portugal: Ordovician trilobites, 605
Pourtalesia miranda , 2

Powell, H. P. See Galton, P. M. and Powell, H. P.

Price, D. The Ordovician trilobite fauna of the Sholes-
hook Limestone Formation, South Wales, 839, 107-
114

Primaspis llandowrorensis sp. nov., 879, 114; sp. indet.,
880, 114

Prionocheilus cf. obtusus, 867, 111, 112
Proceratocephalus cf. terribilis , 878, 114
Proetus (s.l.) cf. berwynensis , 850, 107
Pseudosphaerexochus juvenis , 858, 110, 111; tectus, 110
Pteroplax cornutus, 273

R

Raphiophorus cf. tenellus, 854, 109
Recent: echinoid morphology and function, 39
Remopleurides cf. colbii, 842, 107
Reptoclausa porcata sp. nov., 703, 88
Reptomultisparsa tumida sp. nov., 700, 88
Reyment, R. A. Floating orientations of cephalopod shell
models, 93 1

Rhachistognathus muricatus, 308, 38; primus, 308, 38
Rhadinocytherura amazonensis gen. et sp. nov., 112, 11
Rieppel, O. The edopoid amphibian Cochleosaurus from
the Middle Pennsylvanian of Nova Scotia, 143
Romaniceras (Obiraceras) ornatum, 40; ( Romaniceras )
deverianum , 332, 39, 41-43; (R.) kallesi, 342, 44-46; (R.)
aff. kallesi , 47; ( Yubariceras) ornatissimum, 348, 39, 40,
45, 48-50

Romano, M. The trilobite Eccoptochile from the Ordo-
vician of northern Portugal, 605, 78, 79

S

Saghalien: Cretaceous capulid gastropod, 689
Scotland: Carboniferous amphibian, 915; Carboniferous
shark, 363; Silurian Dictyodora, Peeblesshire, 501
Scott, G. H. The value of outline processing in the
biometry and systematics of fossils, 757
Sevastopulo, G. D. and Keegan, J. B. A technique for
revealing the stereom structure of fossil crinoids, 749,
94,95

Shark: Carboniferous, Scotland, 363
Sheppard, L. M. and Bate, R. H. Plio-Pleistocene ostra-
cods from the Upper Amazon of Colombia and Peru,
97, 7-13

Silurian: evolution of trilobites from Wenlock of Welsh
Borderland, 783; trace fossil Dictyodora from Scotland,
501

Sivapithecus meteai, 85

Siveter, D. J. Evolution of the Silurian trilobite Tapino-
calymene from the Wenlock of the Welsh Borderlands,
783, 97-101

Smith, A. B. The structure, function, and evolution of
tube feet and ambulacral pores in irregular echinoids,
39, 1-6

Smith, P. H. Trichothyriaceous fungi from the early
Tertiary of southern England, 205
Smithson, T. R. A new labyrinthodont amphibian from
the Carboniferous of Scotland, 915
Spathacalymene nasuta, 101

Spathites ( Jeanrogericeras ) robustus robustus, 104; (J.)
subconciliatus hispanicus, 105, 106; ( Spathites ) puer-
coensis, 104

Spathognathodus campbelli, 308, 38; scitulus, 38
Sphaerocoryphe aff. thomsoni, 859, 110
Stahl, B. J. Non-autostylic Pennsylvanian iniopterygian
fishes, 3 1 5

Staurocephalus cf. clavifrons , 866, 110
Stenopareia bowmanni, 107, 108

Stockey, R. A. Jurassic araucarian cone from southern
England, 657, 83-86

Stormer, L. Sculpture and microstructure of the exo-
skeleton in chasmopinid and phacopid trilobites, 237,
25-34

Streptognathodus lateralis, 309, 38
Stygina sp. indet., 843, 107
Subbryantodus subaequalis, 309, 37
Synprioniodina microdenta, 309, 37
Systematics: outline processing of fossils, 757

T

Tapinocalymene nodulosa, 786, 97, 98; nodulosal 97;
volsoriforma gen. et sp. nov., 791, 99; vulpecula gen. et
sp. nov., 792, 100; cf. T. volsoriforma gen. et sp. nov., 99
Taylor, C. N. See Taylor, J. D„ Morris, N. J. and Taylor,
C. N.

Taylor, J. D., Morris, N. J. and Taylor, C. N. Food
specialization and the evolution of predatory proso-
branch gastropods, 375

Taylor, P. D. Two new Jurassic Bryozoa from southern
England, 699, 88

Tekkaya, I. See Andrews, P. and Tekkaya, I.

Tenua asymmetra sp. nov., 160, 16
Tertiary: fungi, England, 205
Thalassiphora fenestrata sp. nov., 489, 54
Thomas, R. G. See Loeffler, E. J. and Thomas, R. G.
Toxochasmops extensus, 32; extensus extensus, 25-28, 32,
33; extensus subsp. nov., 25, 27, 28, 32; marri, 112; sp.
nov., 27

‘ Toxochasmops ’ sp., 29

Trace fossils: Silurian Dictyodora, Scotland, 501
Tretaspis anderssoni, 725, 90; askerensis sp. nov., 740, 92,
93; ceroides (Angelin) angelini (Stormer), 719, 89;
hadelandica hadelandica, 729, 91, 92; hisingeri sp. nov.,
728, 90, 91; kiaeri, 741, 93; latilimbus (Linnarsson)
norvegicus subsp. nov., 734, 93; seticornis, 724, 90;
sortita (Reed) broeggeri Stormer, 737, 92
Trewin, N. H. See Benton, M. J. and Trewin, N. H.
Trichothyrites hordlensis sp. nov., 209
Trilobites: Cambrian, 171; Ordovician, Oslo region, 715;
Ordovician, Portugal, 605; Ordovician, South Wales,
839; sculpture and microstructure, 237; Silurian, Welsh
Borderland, 783
Trimerocephalus caecus, 31
Trinodus tardus, 107
Trochurus sp. indet., 875, 113
Tunesites choffati, 39; salammbo, 39
Turkey: Miocene hominoid, 85

V

Vectidinium stoveri gen. et sp. nov., 490, 54

INDEX

947

W

Wales: Ordovician trilobites. South Wales, 839; pteras-
pidid ostracoderm, Devonian Senni Bed Formation,
287

Welsh Borderland: Silurian trilobites, 783
Wenlock: trilobites, Welsh Borderland, 783
West Malaysia: Carboniferous conodont faunas, 297

Whittington, H. B. Exoskeleton, moult stage, appendage
morphology, and habits of the Middle Cambrian
trilobit e Olenoides serratus, 171, 17-22
Whittingtonia whittingtoni, 875, 114
Wright, C. W. See Kennedy, W. J„ Wright, C. W. and
Hancock, J. M.

THE PALAEONTOLOGICAL ASSOCIATION

The Association was founded in 1957 to further the study of palaeontology. It holds meetings and demonstrations as well
as publishing Palaeontology and Special Papers in Palaeontology. Membership is open to individuals and to institutions on
payment of the appropriate annual subscription. Rates for 1980 are:

Institutional membership ..... £25 (U.S. $60)

Ordinary membership ..... £12 (U.S. $29)

Student membership ..... £7-50 (U.S. $18)

There is no admission fee. Correspondence concerned with Institutional Membership should be addressed to Dr. C. H. C.
Brunton, Department of Palaeontology, British Museum (Natural History), Cromwell Road, London SW7 5BD, England.
Student members are persons receiving full-time instruction at educational institutions recognized by the Council. On first
applying for membership, an application form should be obtained from the Membership Treasurer. Subscriptions cover one
calendar year and are due each January; they should be sent to the Membership Treasurer (for name and address see
Council list below).

PALAEONTOLOGY

All members who join for 1980 will receive Volume 23, Parts 1 -4. All back numbers are still in print and may be ordered from
Marston Book Services Ltd., P.O. Box 87, Oxford 0X4 1LB, England, at £15 (U.S. $36) per part (post free).

SPECIAL PAPERS IN PALAEONTOLOGY

Members may subscribe to this by writing to the Membership Treasurer (for name and address see Council list below);
the subscription rate for 1980 is £20 (U.S. $48) for Institutional Members, and £10 (U.S. $24) 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 prices shown in the list inside the front cover. Non-members may obtain copies, at the
listed prices, from Marston Book Services Ltd., P.O. Box 87, Oxford OX4 1LB, England.

COUNCIL 1980-1981

President : Dr. W. H. C. Ramsbottom, Institute of Geological Sciences, Leeds LS15 8TQ
Vice-Presidents'. Dr. C. T. Scrutton, Department of Geology, The University, Newcastle upon Tyne NE1 7RU
Professor C. B. Cox, Department of Zoology, King’s College, Strand, London WC2R 2LS
Treasurer-. Mr. R. P. Tripp, High Wood, West Kingsdown, Sevenoaks, Kent TNI 5 6BN
Membership Treasurer'. Dr. J. C. W. Cope, Department of Geology, University College, Swansea SA2 8PP
Secretary : Dr. R. Riding, Department of Geology, University College, Cardiff CF1 1XL

Editors

Dr. M. G. Bassett, Department of Geology, National Museum of Wales, Cardiff CF1 3NP
Dr. K. C. Allen, Department of Botany, Bristol University, Bristol BS8 1UG
Dr. R. A. Fortey, Department of Palaeontology, British Museum (Natural History),
Cromwell Road, London SW7 5BD

Dr. A. L. Panchen, Department of Zoology, The University, Newcastle upon Tyne NE1 7RU

Other Members of Council

Dr. R. J. Aldridge, Nottingham

Dr. M. D. Brasier, Hull

Dr. D. E. G. Briggs, London

Dr. C. H. C. Brunton, London

Dr. S. Conway Morris, Open University

Dr. R. Harland, Leeds

Dr. M. B. Hart, Plymouth

Dr. P. M. Kier, Washington
Dr. J. Miller, Edinburgh
Dr. T. J. Palmer, Oxford
Dr. M. Romano, Sheffield
Dr. D. J. Siveter, Leicester
Dr. N. H. Trewin, Aberdeen
Dr. J. Watson, Manchester

Overseas Representatives

Australia: Professor B. D. Webby, Department of Geology, Sydney University, Sydney, N.S.W., 2006
Canada: Dr. B. S. Norford, Institute of Sedimentary and Petroleum Geology, 3303-33rd Street NW., Calgary, Alberta
New Zealand: Dr. G. R. Stevens, New Zealand Geological Survey, P.O. Box 30368, Lower Hutt
West Indies and Central America: Mr. John B. Saunders, Geological Laboratory, Texaco Trinidad, Inc., Pointe-a-Pierre,
Trinidad, West Indies

Western U.S. A. : Professor J. Wyatt Durham, Department of Paleontology, University of California, Berkeley 4, California

Eastern U.S.A.: Professor J. W. Wells, Department of Geology, Cornell University, Ithaca, New York

South America: Dr. O. A. Reig, Departmento de Ecologia, Universidad Simon Bolivar, Caracas 108, Venezuela

Palaeontology

VOLUME 23 • PART 4

CONTENTS

The trilobite Tretaspis from the upper Ordovician of the
Oslo region, Norway

a. w. OWEN 715

A technique for revealing the stereom microstructure of fossil
crinoids

G. D. SEVASTOPULO and J. B. KEEGAN 749

The value of outline processing in the biometry and
systematics of fossils

G. H. SCOTT 757

Hialella—d Jurassic bivalve squatter?

S. R. A. KELLY 769

Evolution of the Silurian trilobite Tapinocalymene from the
Wenlock of the Welsh borderlands

D. j. siveter 783

Spicule pseudomorphs in a new Palaeozoic chaetetid and its
sclerosponge affinities

d. i gray 803

Origin, evolution and systematics of the Cretaceous
ammonoid Spathites

W. J. KENNEDY, C. W. WRIGHT, and J. M. HANCOCK 821

The Ordovician trilobite fauna of the Sholeshook Limestone
Formation of South Wales

d. price; 839

Palaeobiology of Upper Cretaceous belemnites from the
phosphatic chalk of the Anglo-Paris basin

i. jarvis 889

A new labyrinthodont amphibian from the Carboniferous of
Scotland

T. R. SMITHSON 915

A lysorophid amphibian from the Coal Measures of
northern England

M. J. BOYD 925

Floating orientations of cephalopod shell models

R. A. REYMENT 931

Index to Volume 23 943

Printed in Great Britain at the Univcu\?t/ Pte\i, Oxford
by Eric Buckley, Printer to the University

O O __ O

R I E s"2 SMITHSON IAN-* INSTITUTION NOlinillSNI ^NVINOSHIIINS S3ia\/a8n_ LIBRARI ES^ SMITHSON!/

I1SNI NVIN0SH1IIMS S3iavaai1 LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHill/
z w z •, W z co z

o x MfcWk o x A?^ o 3 f/ww/ o

Wvffl; x lib. .cji.

z — 5§ N<^osji£

RIES SMITHSONIAN INSTITUTION NOlinillSNI _ NVINOSHIIINS^ S3 I a Va a I l^LI B RAR I ES^SMITHSONM
^ co „ — . z \ « _ _ ^ oo

I1SNI NVIN0SH1IIAIS S3iavaail LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHlIlf

pi -w <ji m ^2 m >SC>2*. DC,

CO \ Z ..

RIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIINS S3INVH0n LIBRARIES SMITHSONII
co z .v, co z ^ ^ co z to

< ^VN 2 < S ,-s 2

I1SNI NVIN0SH1I WS S3iavaai3 LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1II

v\ —
s\ OC
5 < lol

ca

q >\ 2. xgjiisy o vj^v. doz O xyc DQj^ ^ q

R I E S^ SMITHSON IAN-* INSTITUTION NOlinillSNI “NVINOSHIIWS Saiavaail^LIBRARI ES^SMITHSONI)
z r- z r z r- z

O /To 0 cJoX. CD X^TIT<v2X :v\ CD __ Xjo \2\ O

I IS N I NVIN0SH1IWS S3iavaail LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1II
Z co z , co z CO z

< >^uiESv 2 A S

o tel

Jjj

> XjVASV^P

R I ES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIWS^SS I a Va 8 ll^LI B RAR I ES^SMITHSONlI

| (II

Z _J z ' _1 Z

nuikiocu i IIAIC CtJ I VJ WVI a n I ( R P A R I F<R RU ITWRHN I AM IMRTITI ITIOM MnnnillRWI WWINORH1 II/*

2 5 • >-

riTUTION WN0linillSNI_NVIN0SHilWS^S3 iava8l1_ LIBRARIES SMITHSONIAN INSTITUTION

iOiOI #1(0) 4C

X^VASV^X 0 Xr/A/nc^^V X#VOSVa£/ XlVA

avaan libraries Smithsonian institution NouniiisNi nvinoshiiws saiavaan libr;

~ - z f“ . 2 r-

TITUTION NOIlfUllSNI NVIN0SH1IINS S3iavaail LIBRARIES SMITHSONIAN INSTITUTION
,v co z w z - • ^ z ,v.

-y AS 5 | I J&K I /^%1 i

o Wss,Wi x o

i 4j

> v4r 2 ’■v. > _

_ .. Z CO * Z tO * Z

i avaan libraries Smithsonian institution NouniiisNi nvinoshiiims s 3 1 avaan librj

— X CO — CO —

O d1-^ _ q — \<ouiis5i/ O

TITUTION NOliniliSN I NVINOSHIIIMS S3iavaail LI B RAR I ES^SMITHSONIAN^INSTITUTION NOIlfl
z i-»z r- z r- z