IT I ON Z NOimillSNI^NVINOSHlI WS ^S3 lyvaail^LIBRARI ES^ SMITHSON IAN-* INSTITUTION 2 NOIlfl
z r- v Z r~ z r; z
O Z 58k. o xC$t7t£>v Z O „ X'ogjvnX O
m
ugn LIBRARIES SMITHSONIAN INSTITUTION NOIinillSNI-NVINOSHllWS S3 I b \fU B 11 ~LI B R;
z in z t in z m z
\\ X
in
o x 'X o iff/:, ' x Vs'Gr. -3*7 °
> ^ > ‘ 5 >
TION NOlinillSNI NVINOSHimS</>S3iyVMail LIBRARIES^SMITHSONIAN INSTITUTION NOliO
to 5 't «o . ^ ^ co ^ to
„^v. c
o z 5 “-r 5
Z _l Z _| z -»
Jail LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINQSH1IWS S3I tflVH a H LIBR
z z r- z ""
rn \2i£S>^
CO — CO £ CO _
ITION NOlinillSNI NVIN0SH1IINS S3IHVM8n LIBRARIES SMITHSONIAN INSTITUTION NOIlf
Z _ CO z vy. CO Z
yan libraries Smithsonian institution NOlinillSNI nvinoshiiws saiavaan libr
CO ^ CO ~ CO
o _ ><0£ \^y o
JTION*- NOIinillSNI_NVINOSHllWS S3 I 8 V8 8 II “V I B R AR I E S*2 SMITHSONIAN^ I NSTITUTIO N ^NOIlfl
z I- V Z r- z i~ z
y g n LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHimS S3!8VHan LIBR
Z , CO z CO Z CO z ■
< \» S < V Z 2
:A z Ok ^ z -H
jo
c
«E '' 3> ^ «>W ■> 5 3»’ S'
to W Z CO '•’** Z CO Z CO
ITION NOlinillSNI NVIN0SH1IWS SBIBVaail LIBRARIES SMITHSONIAN INSTITUTION NOlin
co ~ in — co 5; co _
~ ^ ” jj§|5|y ^
5 W' “ zl
z
S^> - O Z ' O
-J Z _1 z -J
nan libraries Smithsonian institution NouniiisNi nvinoshiiws S3iavaan libr
z r- z r- ... z C”
2
iNOSHims ssiavaan libraries Smithsonian institution NoiinmsNi nvinoshiiws s:
,.2: r* 2: r- z _ [I > g
O O — ^<Tn7^ o ^STIT(7>V \v o
ithsonian institution NoiinmsNi nvinoshiiws saiavaan LIBRARIES SMITHSONIAN ir
2 > CO 2 "
™ a. ■» I ^ X EC CWi/ O
i vx^ >' 5 >'■ S 2
J/} 2 CO 2 CO N 2 CO
IN0SH1IIMS S3iavaan LIBRARIES SMITHSONIAN INSTITUTION NOiiniusNi_NviNOSHiiws s
CO — to — CO Z ,n
O x^osv^ O
2 _i 2
I ITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IWS S3IHVaan LIBRARIES SMITHSONIAN II
2 1- z
^ rn ^ Ngtmssx m
IN0SH1IWS SBIdVaan- LIBRARIES SMITHSONIAN ""INSTITUTION NOlinillSNI NVIN0SH1MS S
CO 2 ..... CO -p- CO 2
=]5
ITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIWS^SB I a VH a 11 LIBRARIES SMITHSONIAN II
- CO — CO X to 2
^ 'm ^ co w , co
e/ C V'£®
Z! %? _ _ _ .
O Ng*. _ X^lllS^X O v II X^UllS^^ q XON pcy O
IN0SHlllNS:ZS3iavaan"JLIBRAR I ESZ SMITHSONIAN^ INSTITUTION ^NOimillSNI^NVINOSHimS S
^2 r~ 2 z
0 O
TO
> IpI
JO x ^
rn !£ * ' ’ n ^
— CO ± CO £ — —
IITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IWS S3iavaail LIBRARIES SMITHSONIAN II
co 2 co 2 > <5 — 2
< V S < 'fc, I <
A 2
— xTog.A'ox _
^ o
co £1 _
X V5&, •3FSJ O
2 , "y / ' ' t 5
2 ' W > '' 2 >' 2
to 2 CO 2 co A 2 co
rIN0SHllWS_S3 I ava a n_LI B RAR I ES SMITHSONIANJNSTITUTION _NOIinillSNI_NVINOSHimS/rtS
O “ O cj
2 _J 2 -I z
IITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1MS S3iavaail LIBRARIES SMITHSONIAN I
VOLUME 25
Palaeontology
1982
PUBLISHED BY THE
PALAEONTOLOGICAL ASSOCIATION
LONDON
Dates of Publication of Parts of Volume 25
Part l,pp. 1-225, pis. 1-24
1 January
1982
Part 2, pp. 227-442, pis. 25-45
1 April
1982
Part 3, pp. 443-679, pis. 46-65
1 July
1982
Part 4, pp. 681-923, pis. 66-100
1 November 1982
THIS VOLUME EDITED BY M. G. BASSETT, K. C. ALLEN, R. A. FORTEY AND A. L. PANCHEN
Dates of Publication of Special Papers in Palaeontology
Special Paper No. 28 August 1982
Special Paper No. 29 October 1982
© The Palaeontological Association 1982
Printed in Great Britain
at the University Press, Oxford
by Eric Buckley
Printer to the University
CONTENTS
Part Page
Akpan, E. B., Farrow, G. E. and Morris, N. J. Limpet grazing on Cretaceous algal-bored
ammonites 2 361
Aldridge, R. J. A fused cluster of coniform conodont elements from the late Ordovician of
Washington Land, Western North Greenland 2 425
Aldridge, R. J. See Swift, A. and Aldridge, R. J.
Allen, K. C. See Marshall, J. E. A. and Allen, K. C.
Balinski, A. See Biernat, G. and Balinski, A.
Balson, P. S. and Taylor, P. D. Palaeobiology and systematics of large cyclostome bryozoans
from the Pliocene Coralline Crag of Suffolk 3 529
Barwick, R. E. See Campbell, K. S. W. and Barwick, R. E.
Biernat, G. and Balinski, A. Shell structure of the Devonian retziid brachiopod Plectospira
ferita 4 857
Blake, D. B. Somasteroidea, Asteroidea, and the affinities of Luidia ( Platasterias ) latiradiata 1 167
Boyd, M. J. Morphology and relationships of the Upper Carboniferous ai'stopod amphibian
Ophiderpeton nanum 1 209
Brenchley, P. J. and Cocks, L. R. M. Ecological associations in a regressive sequence: the latest
Ordovician of the Oslo-Asker district, Norway 4 783
Campbell, K. S. W. and Barwick, R. E. A new species of lungfish Dipnorhynchus from New
South Wales 3 509
Cocks, L. R. M. The commoner brachiopods of the latest Ordovician of the Oslo-Asker
district, Norway 4 755
Cocks, L. R. M. See Brenchley, P. J. and Cocks, L. R. M.
Coombs, W. P. Jr. Juvenile specimens of the ornithischian dinosaur Psittacosaurus 1 89
Cope, J. C. W. See Paul, C. R. C. and Cope, J. C. W.
Crame, J. A. Late Jurassic inoceramid bivalves from the Antarctic Peninsula and their
stratigraphic use 3 555
Curry, G. B. Ecology and population structure of the Recent brachiopod Terebratulina from
Scotland 2 227
De Jersey, N. J. An evolutionary sequence in Aratrisporites miospores from the Triassic of
Queensland, Australia 3 665
Donovan, D. T. and Howarth, M. K. A rare lytoceratid ammonite from the Lower Lias of
Radstock 2 439
Elliott, G. F. A new calcareous green alga from the Middle Jurassic of England: its
relationships and evolutionary position 2 43 1
Farrow, G. E. See Akpan, E. B., Farrow, G. E. and Morris, N. J.
Forsten, A. The taxonomic status of the Miocene horse genus Sinohippus 3 673
Fraser, N. C. A new rhynchocephalian from the British Upper Trias 4 709
Freeman, E. F. Fossil bone recovery from sediment residues by the ‘Interfacial Method’ 3 471
Fursich, F. T. and Palmer, T. J. The first true anomiid bivalve? 4 897
Gale, A. S. and Smith, A. B. The palaeobiology of the Cretaceous irregular echinoids Infulaster
and Hagenowia 1 1 1
Geys, J. F. Two salenoid echinoids in the Danian of the Maastricht area 2 265
Harland, R. A review of Recent and Quaternary organic-walled dinoflagellate cysts of the
genus Protoperidinium 2 369
Harland, T. L. and Torrens, H. S. A redescription of the Bathonian red alga Solenopora
jurassica from Gloucestershire, with remarks on its preservation 4 905
Henry, J.-L. See Romano, M. and Henry, J.-L.
Higgins, A. C. and Varker, W. J. Lower Carboniferous conodont faunas from Ravenstone-
dale, Cumbria 1 145
CONTENTS
iv
Part Page
Higgins, A. C. and Wagner-Gentis, C. H. T. Conodonts, goniatites and the biostratigraphy of
the earlier Carboniferous from the Cantabrian Mountains, Spain 2 313
Howarth, M. K. See Donovan, D. T. and Howarth, M. K.
Jefferson, T. H. Fossil forests from the Lower Cretaceous of Alexander Island, Antarctica 4 681
Kier, P. M. Rapid evolution in echinoids 1 1
Lund, R. and Melton, W. G. Jr. A new actinopterygian fish from the Mississippian Bear Gulch
Limestone of Montana 3 485
Marshall, J. E. A. and Allen, K. C. Devonian miospore assemblages from Fair Isle, Shetland 2 277
Mateer, N. J. Osteology of the Jurassic lizard Ardeosaurus brevipes (Meyer) 3 461
Melton, W. G. See Lund, R. and Melton, W. G.
Milner, A. R. Small temnospondyl amphibians from the Middle Pennsylvanian of Illinois 3 635
Morris, N. J. See Akpan, E. B., Farrow, G. E. and Morris, N. J.
Palmer, T. J. See Fursich, F. T. and Palmer, T. J.
Paul, C. R. C. and Cope, J. C. W. A parablastoid from the Arenig of South Wales 3 499
Riding, R. and Voronova, L. Affinity of the Cambrian alga Tubomorphophyton and its
significance for the Epiphytaceae 4 869
Rieppel, O. A new genus of shark from the Middle Triassic of Monte San Giorgio, Switzerland 2 399
Rogers, M. J. A description of the generating curve of bivalves with straight hinges 1 109
Romano, M. and Henry, J.-L. The trilobite genus Eoharpes from the Ordovician of Brittany
and Portugal 3 623
Smith, A. B. Tooth structure of the pygasteroid sea urchin Plesiechinus 4 891
Smith, A. B. See Gale, A. S. and Smith, A. B.
Stein, W. E. Jr. The Devonian plant Reimannia, with a discussion of the Class Progymno-
spermopsida 3 605
Surlyk, F. and Zakharov, V. A. Buchiid bivalves from the Upper Jurassic and Lower
Cretaceous of East Greenland 4 727
Swift, A. and Aldridge, R. J. Conodonts from the Upper Permian strata of Nottinghamshire
and North Yorkshire 4 845
Taylor, P. D. See Balson, P. S. and Taylor, P. D.
Thulborn, R. A. Liassic plesiosaur embryos reinterpreted as shrimp burrows 2 351
Torrens, H. S. See Harland, T. L. and Torrens, H. S.
Tunnicliff, S. P. A revision of late Ordovician bivalves from Pomeroy, Co. Tyrone, Ireland 1 43
Turner, R. E. Reworked acritarchs from the type section of the Ordovician Caradoc Series,
Shropshire 1 119
Turner, S. A new articulated thelodont (Agnatha) from the early Devonian of Britain 4 879
Varker, W. J. See Higgins, A. C. and Varker, W. J.
Voronova, L. See Riding, R. and Voronova, L.
Wagner-Gentis, C. H. T. See Higgins, A. C. and Wagner-Gentis, C. H. T.
Wang Lixin. See Wang Ziqiang and Wang Lixin
Wang Ziqiang and Wang Lixin. A new species of the lycopsid Pleuromeia from the early
Triassic of Shanxi, China, and its ecology 1 215
Wellstead, C. F. A Lower Carboniferous ai'stopod amphibian from Scotland 1 193
Wilson, M. V. H. A new species of the fish Amia from the Middle Eocene of British Columbia 2 413
Young, G. C. Devonian sharks from south-eastern Australia and Antarctica 4 817
Zakharov, V. A. See Surlyk, F. and Zakharov, V. A.
Zdebska, D. A new zosterophyll from the Lower Devonian of Poland 2 247
Zhang Zhongying. Upper Proterozoic microfossils from the Summer Isles, N. W. Scotland 3 443
Palaeontology
VOLUME 25 ■ PART 1 JANUARY 1982
Published by
The Palaeontological Association London
Price £1 7
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. $17-50).
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. $11).
3. (for 1968): Upper Maestrichtian Radiolaria of California, by Helen p. foreman. 82 pp., 8 plates. Price £3 (U.S. $6-50).
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. $6-50).
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. $4-50).
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. $6-50).
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. $3-50).
8. (for 1970): Cenomanian Ammonites from Southern England, by w. j. Kennedy. 272 pp., 5 tables, 64 plates. Price £8
(U.S. $17-50).
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. $11).
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. $11).
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. $22).
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. $22) (published with the Systematics Association).
13. (for 1 974) : 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. $22).
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. $17-50).
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. $12).
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. $14-50).
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. $13).
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. $44).
19. (for 1977): The Palynology of Early Tertiary Sediments, Ninetyeast Ridge, Indian Ocean, 6yE. M. kemp and w. K. Harris.
74 pp., 2 text-figs., 8 plates. Price £7 (U.S. $15-50).
20. (for 1977): Fossil Priapulid Worms, by s. c. morris. 159 pp., 99 text-figs., 30 plates. Price £16 (U.S. $35).
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. $26-50).
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. $55).
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. $66).
24. (for 1980): Dinofiagellate 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. $33).
25. (for 1980): Stereom Microstructure of the Echinoid Test, by a. b. smith. 81 pp., 20 text-figs., 24 plates. Price £15 ($33).
26. (for 1981): The fine structure of Graptolite periderm, by p. r. crowther. 1 19 pp., 37 text-figs., 20 plates. Price £25 ($55).
27. (for 1981): Late Devonian Acritarchs from the Carnarvon Basin, Western Australia, by G. playford andw.. s. dring. 78 pp.,
10 text-figs., 19 plates. Price £15 ($33).
© The Palaeontological Association, 1982
Cover: The acritarch Umbellasphaeridium saharicum Jardine et al. 1972 from the Tournaisian Bedford Shale of Ohio,
U.S.A., x 1000. I.G.S. specimen number MPK 3152 deposited in the micropalaeontological collection of the Institute of
Geological Sciences, Leeds. Photograph by S. G. Molyneux. This group of fossils was unknown when the Association was
founded, twenty-five years ago.
RAPID EVOLUTION IN ECHINOIDS
by PORTER M. KIER
Abstract. The evolution of the irregular echinoid and of the sand dollar occurred in a very short time. The first
irregular echinoid appears abruptly in the Early Jurassic (Sinemurian); and by the Toarcian, only ten million
years later, irregular echinoids possess all the features necessary to permit them to live buried in the sediment.
The first clypeasteroid appears in the Paleocene. By the middle Eocene its very specialized descendants, the sand
dollars, have a worldwide distribution. This rapid evolution and diversification seem to result from a sudden
adaptive breakthrough. The presence of so few intermediates indicates the evolutionary steps must have been
large.
Two of the most significant events in the evolution of the echinoids are the development of the
irregular echinoid and the subsequent appearance of the sand dollar. Both of these events were
believed to have occurred over a long period of time (Durham 1966, p. U289) but new evidence
suggests otherwise. The irregular echinoid was assumed to have evolved during the Triassic and
perhaps during the latter part of the Paleozoic. However, study of the Triassic faunas indicates that
no irregulars were present then and that the great changes from the regular to the irregular echinoid
occurred during the early part of the Early Jurassic. Likewise, it was believed (Durham 1966,
p. U290) that the clypeasteroids arose during the Late Cretaceous and Paleocene. New evidence
suggests that the first clypeasteroid actually appeared in the Paleocene, and that the great evolution to
the sand dollar occurred during the early Eocene.
EVOLUTION OF THE IRREGULAR ECHINOID
The first irregular echinoid, Plesiechinus hawkinsi Jesionek-Szymanska (text-fig. 1b), occurs in the
Early Jurassic (Sinemurian). This echinoid, a pygasterid, differs from all other echinoids of the same
age or older in having an asymmetrical test with small tuberculation, short and numerous spines,
differentiated pores with the adapical pores larger than the adoral ones, posteriorly eccentric
periproct, and presumably keeled teeth (although the lantern was not found with any specimens of
this species, Melville (1961) found keeled teeth in a Pygaster).
One might assume that the great changes necessary to derive this irregular echinoid from a regular
form would have taken a very long time. Many workers believe that these changes occurred during
the Triassic or possibly in the Paleozoic. They attributed the lack of intermediates found during the
Triassic to the poor fossil record of that period. However, that does not appear to be the case.
Although it is true that few echinoids are known from the Triassic, particularly from the Early and
Middle Triassic, a prolific echinoid fauna occurs in the Late Triassic St. Cassian beds of Italy. These
beds (Kier 19776) have been painstakingly searched by Rinaldo Zardini who has found many
complete tests and thousands of fragments. I have searched through all this material and have failed
to find any keeled teeth or any test fragments with the fine tuberculation of an irregular echinoid. As
shown by Kier (1977a), irregular echinoids are much more likely to be preserved than regular
echinoids. Had they been living during St. Cassian time, they should have been preserved as fossils.
Furthermore, no irregular echinoids or any form resembling them have been found in the lowermost
Jurassic (Hettangian).
I am convinced that this change from a regular echinoid to Plesiechinus hawkinsi must have
occurred extremely quickly during the latter part of the Hettangian or early Sinemurian. Evolution
then continued at a rapid rate, because by the Toarcian, still within the Early Jurassic, the cassiduloid
Galeropygus dumortieri (Paris) (text-fig. lc) has appeared with all the features of an irregular
IPalaeontology, Vol. 25, Part 1, 1982, pp. 1-9, pis. 1-2.]
2
PALAEONTOLOGY, VOLUME 25
echinoid. It is more advanced than P. hawkinsi in having a more flattened and elongate test, a more
eccentric periproct, larger adapical pores forming incipient petals, and smaller, more numerous,
tubercles and spines. Its peristome is smaller, elongate, and eccentric anteriorly. There are no gills.
The most profound differences are the lack of a lantern and lantern supports and the presence of well-
developed phyllodes.
Presumably, one of the reasons for this rapid evolution was that these changes enabled echinoids to
occupy a habitat not available to them before. They were now able to burrow (Smith 1978) into the
sediment and extract the organic material contained within. Simpson (1944, 1953), and later Stanley
(1979), proposed that a higher taxon arises rapidly through the occurrence of a sudden adaptive
breakthrough. During the Triassic all echinoids apparently lived on the surface of the sea floor and
could not burrow into the sediment. They all had jaws and grooved teeth which were used
(presumably like modern regular echinoids) to tear off and chew organic material which was then
passed into the gut. The small amount of faecal discharge could be easily carried away by water
currents. Most modern irregular echinoids lacking teeth feed differently. They burrow (text-fig. 1)
LOWER JURASSIC
HETTANGIAN SINEMURI AN TOARCI AN
ELONGATION OF THE TEST ►
DECREASE IN SIZE OF TUBERCLES AND SPINES ►
INCREASE IN NUMBER OF TUBERCLES AND SPINES ►
SHIFTING OF ANUS POSTERIORLY ►
SHIFTING OF PERISTOME ANTERIORLY ►
DECREASE IN SIZE OF PERISTOME ►
DEVELOPMENT OF PETALS ►
DEVELOPMENT OF PHYLLODES ►
GROOVED TEETH ► KEELED TEETH ► NO TEETH
text-fig. i . Evolution of the irregular echinoid showing the changes that enabled the echinoid to live buried
in the substrate, a, an Hettangian regular echinoid such as Diademopsis; b, the earliest known irregular,
Plesiechinus hawkinsi Jesionek-Szymanska; c, a cassiduloid, Galeropygus.
EXPLANATION OF PLATE 1
Figs. 1-5. Togocyamus seefriedi (Oppenheim). Paleocene, Ekekoro Formation, Ekekoro quarry, 55 km north-
west of Lagos, Nigeria. 1, accessory pore on dorsal side of USNM 312503 just beyond petal I. The large
pores on the lower right side are at the end of petal I, x 74. 2, enlarged view of accessory pore in USNM
312504 showing large neural pore, x 500. 3, side view of USNM 312505, x 19. 4, top view of USNM
312503, x 13. 5, bottom view of USNM 312504, x 15.
PLATE 1
KIER, Togocyamus
4
PALAEONTOLOGY, VOLUME 25
into the sediment, collect large amounts of sediment with their tube-feet, and pass it through the gut
while extracting the organic material. The sediment is then expelled through the anus (periproct).
Most of the differences between an irregular and regular echinoid relate to these differences in mode
of feeding.
The posterior migration of the periproct made it possible for the irregular echinoid to leave this
large volume of discharged sediment in its trail rather than over its dorsal surface. At the same time
that the periproct migrated, most irregular echinoids increased the oxygen-gathering capability of
their dorsal tube-feet by greatly broadening them. The result was the formation of the ‘petals’ so
typical of most irregulars. Specialized tube-feet were also produced around the mouth. They were
larger, more numerous, and were used to collect sediment that was then passed to the mouth.
EVOLUTION OF THE CLYPEASTEROID ECHINOID
The first clypeasteroid echinoid, Togocyamus (PI. 1, figs. 1-5), appears in Paleocene strata. By the
middle Eocene the highly specialized sand dollar had evolved. Until now it was believed that these
developments required a long time from the Cretaceous through the Eocene. New evidence indicates
that the change was much more rapid. In fact the change from a cassiduloid ancestor to a
clypeasteroid probably occurred within the Paleocene; and the change from a primitive clypeasteroid
to a sand dollar occurred during the early Eocene.
The earlier ‘Cretaceous’ origin of the clypeasteroids was based on the supposed occurrences of
three species of clypeasteroids in the Late Cretaceous. These occurrences are probably erroneous.
One of the three species is too poorly preserved to be identified, and the stratigraphic data with the
other two are inadequate. Extensive collecting has been done in the Late Cretaceous beds where these
two species supposedly were found and neither Meijer (1965) nor Ernst (1972) have found any
clypeasteroids. Ernst, in the course of his study, examined over 15000 echinoids from the Late
Cretaceous.
The chronologically later origin of the clypeasteroids is supported by their absence from the
Paleocene of the Western Hemisphere. An exhaustive search by Kier through washings of the
Paleocene Vincentown, Aquia, and Clayton Formations (from which eighteen echinoid species are
known) revealed no fragments that could be identified as clypeasteroid, although echinoid fragments
are very common. It is noteworthy that the small fibularids, like those found in the Paleocene of
Africa, were absent but normally would be expected in this material. They usually live buried in this
type of sediment and are small enough to have their tests preserved intact in these sands.
The earliest confirmed clypeasteroid, Togocyamus, occurs in the Paleocene of West Africa. It
differs so markedly from all previous echinoids that there is disagreement as to its origin. Some
workers consider the clypeasteroids to be derived from a holectypoid, but Phelan (1977, p. 419)
makes a strong case for their derivation from a juvenile stage of a cassiduloid. Its evolution must have
been extremely rapid and probably occurred within the Paleocene. The Late Cretaceous echinoid
record is not only extensive but very well studied (Ernst 1972) and no intermediates have been
found there. Current information indicates that the subsequent diversification and radiation of
the clypeasteroids from a fibularid to a sand dollar occurred very abruptly (text-fig. 2) in the early part
of the middle Eocene. Only two species, both fibularids, are known in the Paleocene: Togocyamus
EXPLANATION OF PLATE 2
Figs. 1-3. Sismondia logotheti Fraas. Early Eocene, from Siout ( = Assiout), Egypt. 1, 2, 3, top, bottom, side
views of topotype B22908, Museum National d’Histoire Naturelle, Paris, x 5.
Figs. 4-6. Periarchus lyelli (Conrad). Middle Eocene, Castle Hayne Formation, from North Carolina Lime
Company pit, adjacent to Tuckahoe Church, 3-8 miles (61 km) west of Comfort, Jones County, North
Carolina. 4, 5, 6, side, top, bottom views of USNM 312506, x 1.
PLATE 2
3
KIER, Sismondia, Periarchus
6
PALAEONTOLOGY, VOLUME 25
seefriedi (Oppenheim) and T. alloiteaui Roman and Gorodiski. During the early Eocene the
clypeasteroids (text-fig. 2) were confined to Africa and India and consist of six species of fibularids
and two species of Sismondia, the most primitive member of the laganids. By the end of the middle
Eocene, the clypeasteroids had worldwide distribution with over 62 species representing 20 genera
and 4 families.
In summary, the first clypeasteroid appeared in the Paleocene, and by the middle Eocene its far
more complex and specialized descendants were present all around the world.
MIDDLE
EOCENE
EARLY
EOCENE
PALEOCENE
text-fig. 2. Evolution and radiation of the clypeasteroid echinoid. The Paleocene
clypeasteroids are confined to West Africa and are a small, high species with incipient petals.
In the early Eocene they are more flattened, have more-developed petals, and occur in Africa
and India. By the middle Eocene they are fully developed sand dollars and are present all
around the world.
The Cretaceous origin of the clypeasteroids was suggested by previous workers not only because of j
the supposed occurrence of Cretaceous species, but also because it was believed that the Palaeocene
fibularids were not primitive enough to be ancestral to all later clypeasteroids. New information is
now available on Togocyamus. It is more primitive than previously thought and could be close to the
ancestral stock of all clypeasteroids. Although it has been assumed that its lantern supports were like
those found in later fibularids with each support composed of a single interambulacral plate, the
supports in T. seefriedi (Oppenheim) are both interambulacral and ambulacral in origin. Each ;
support ( text-fig. 3c) is formed by the extension of the primordinal interambulacral plate and the
adjacent half-ambulacral plates. This discovery is important for it has been suggested (Philip 1965, p.
58; Kier 1970, p. 105) that the clypeasteroids could be divided into two orders on the basis of the
character of the lantern supports. The suborder Clypeasterina includes all those clypeasteroids
KIER: ECHINOID EVOLUTION
7
EOCENE
text-fig. 3. The lantern supports in the clypeasteroids. c, the
oldest clypeasteroid, the Paleocene Togocyamus seefriedi
(Oppenheim) has lantern supports (indicated by a solid line)
composed of ambulacral (shaded) and interambulacral plates;
a, b. Eocene clypeasteroids have lantern supports composed of
interambulacral plates as in the suborder Scutellina (a) or
ambulacral plates as in Clypeasterina (b).
having ambulacral lantern supports (text-fig. 3b); in the suborder Scutellina the supports are
interambulacral (text-fig. 3a). The presence of supports, both interambulacral and ambulacral in
origin, in the oldest known and most primitive clypeasteroid adds weight to the supposition that
Togocyamus is close to the ancestral stock of both suborders. By simply reducing the size of the
ambulacral extensions and increasing the size of the interambulacral ones, supports could be
produced that are typical of later species of the fibularids and the rest of the Scutellina. Conversely,
the reduction in the size of the interambulacral and increase in the ambulacral extensions would
produce the typical Clypeasterina supports that first appear in the late Eocene.
A second discovery in Togocyamus is that its accessory pores are few in number and are restricted
to the border of the ambulacra (PI. 1, fig. 1). As pointed out by Durham (1966, p. U451), accessory
pores are an exclusive feature of the clypeasteroids, occurring in all species. The fact that they are less
well developed in this species than in any other is further evidence of the primitiveness of this form.
In the light of the primitive features of Togocyamus , we can now postulate the evolutionary history
of the earliest clypeasteroids:
1. The Paleocene Togocyamus (PI. 1, figs. 3-5) has a small, high test with its periproct in a primitive dorsal
position, slightly developed petals with simple nonconjugate pores, a very erect lantern with supports of
interambulacral and ambulacral origin. Its few accessory pores are confined to the borders of the ambulacra. It
has no food grooves and has a large peristome.
2. By the early Eocene, Sismondia (PI. 2, figs. 1-3) has a larger, more flattened test, a lower lantern,
interambulacral lantern supports, and a ventral periproct. The petals are better developed with conjugate pores;
the accessory pores are far more numerous.
PALAEONTOLOGY, VOLUME 25
3. The middle Eocene Protoscutella and Periarchus (PI. 2, figs. 4-6) are typical sand dollars having a large, very
flattened test, food grooves, very wide and low lantern, and a very small peristome. Accessory pores are spread
all over the ambulacra, and the test is strongly reinforced by calcareous supports that are pierced by many canals
for the water vascular system serving these pores. The adoral plate arrangement is now distinctive; there are far
fewer and larger plates than in earlier clypeasteroids. As pointed out by Durham (1966, p. U450), these changes
in the adoral plates result from the flattening of the test. In flattened species the number of plates on the adoral
surface is determined at an early ontogenetic stage and thereafter growth is only by enlargement of the plates.
The morphological changes that produced the sand dollar are specializations that enabled the
echinoid to live more efficiently in sand (Seilacher 1979). The flattened test made it easier for the
echinoid to burrow. The accessory tube-feet were used to pass sand over the top of the test and to
convey food to the food grooves. The better-developed petals increased the respiratory capability of
the petaloid tube-feet by increasing their area. The internal supports strengthened the test, enabling
the sand dollar to live in environments of higher energy. The change from the erect lantern and large
peristome in the primitive fibularid to the low lantern with horizontal teeth and small peristome in the \
sand dollar reflects a change in eating habits. According to Markel (1974, 1978; Markel, Gorny, and
Abraham 1977) and Nichols (1959), the fibularid uses its lantern to scrape organic material from sand
grains. This feeding method puts little stress on the teeth. The sand dollar, however, uses the teeth for
grinding and chewing. The great stress can be withstood because the teeth are horizontal, and the
stress is transmitted to the long axis of the teeth. The larger peristome in the fibularid permits the teeth
to extend further out of the test to grasp food; in the sand dollar the sand is passed to the teeth within
the test.
If it is true, as suggested herein, that the clypeasteroids originated in the Paleocene then all these
changes necessary to derive a sand dollar from a cassiduloid ancestor occurred within 20 million
years. Certainly there can be little question that the evolution from a fibularid to a typical sand dollar
occurred between the beginning of the early Eocene and the latter part of the middle Eocene, a period
of less than 10 million years.
CONCLUSIONS
The sudden appearance of the first irregular and the first clypeasteroid echinoids and their rapid
diversification indicate a rate of evolution much faster than previously supposed. The mechanisms
producing these great changes are uncertain, but the evolutionary steps must have been large. If each
speciation event produced only small morphological change, than a multitude of transitional species
would have resulted. The fossil record of the irregular echinoids is excellent (Kier 1977a). Even if this
rapid evolution occurred in peripherally isolated populations, somewhere in the world we should
have found more of these transitional species. I believe the absence of a large number of transitional
species is explained not because they have not been preserved as fossils, but because they never
existed. This conclusion supports Stanley’s (1979, p. 212) statement that ‘rates of evolution are
highest early in adaptive radiation, when degree of divergence per speciation event is high . . .’.
Acknowledgements. I thank Kenneth Towe and Thomas Waller (Smithsonian Institution), and Thomas Phelan
and Andrew Smith, for helpful discussions. The manuscript was read by Richard Boardman, Richard Grant,
and David Pawson. I thank them for their suggestions. Mary Hurd Lawson did the light photography; Walter
Brown did the SEM photography. Larry Isham made all the text-figures with his customary skill and
inventiveness. Jean Roman (Museum National d’Histoire Naturelle, Paris) and Richard Jefferies (British
Museum, Natural History) very kindly lent specimens.
REFERENCES
Durham, j. w. 1966 Classification, U270-U295; Clypeasteroids, U450-U491. In Moore, R. C. (ed.). Treatise on
invertebrate paleontology , Part U, Echinodermata 3, 1-2, Geol. Soc. Am. and Univ. Kansas Press.
ernst, G. 1972. Grundfragen der Stammesgeschichte bei irregularen Echiniden der nordwesteuropaischen
Oberkreide. Geol. Jb. A4, 63-175, pis. 1-7.
KIER: ECHINOID EVOLUTION
9
gould, s. J. and eldredge, n. 1977. Punctuated equilibria: The tempo and mode of evolution reconsidered.
Paleobiology , 3, 115-151.
kier, p. m. 1970. Lantern support structures in the clypeasteroid echinoids. J. Paleont. 44, 98-109, pis. 23, 24.
— 1977a. The poor fossil record of the regular echinoid. Paleobiology, 3, 168-174.
19777>.Triassic echinoids. Smithson. Contr. Paleobiol. 30, 1-88, pis. 1-21.
MARKEL, K- 1974. Morphologie der Seeigelzahne. V. Die Zahne der Clypeastroida (Echinodermata,
Echinoidea). Zh. morph. Tiere, 78, 221-256.
1978. On the teeth of the recent cassiduloid Echnolampas depressa Gray, and on some Liassic fossil teeth
nearly identical in structure (Echinodermata, Echinoidea). Zoomorphologie, 89, 125-144.
— gorny, p. and abraham, k. 1977. Microarchitecture of sea urchin teeth. Fortschr. Zool. 24, 103-1 14.
meijer, m. 1965. The stratigraphical distribution of echinoids in the chalk and tuffaceous chalk in the
neighborhood of Maastricht (Netherlands). Meded. geol. Sticht. 17, 21-25.
melville, r. v. 1961. Dentition and relationships of the echinoid genus Pygaster J. L. R. Agassiz, 1836.
Palaeontology, 4, 243-246.
nichols, d. 1959. The histology and activities of the tube-feet of Echinocyamus pusillus. Q. Jl microsc. Sci. 100,
539-555.
phelan, t. f. 1977. Comments on the water vascular system, food grooves, and ancestry of the clypeasteroid
echinoids. Bull. mar. Sci. Gulf Caribb. 27, 400-422.
philip, G. M. 1965. Classification of echinoids. J. Paleont. 39, 45-62.
seilacher, a. 1979. Constructional morphology of sand dollars. Paleobiology, 5, 191-221.
simpson, G. G. 1944. Tempo and mode in evolution, 237 pp. New York, Columbia Univ. Press.
— 1953. The major features of evolution, 434 pp. New York, Columbia Univ. Press.
smith, a. b. 1978. A comparative study of the life style of two Jurassic irregular echinoids. Lethaia, 11, 57-66.
Stanley, s. m. 1979. Macroevolution, pattern and process, 332 pp. San Francisco, Freeman.
Typescript received 28 March 1981
Revised typescript received 30 April 1981
PORTER M. KIER
Department of Paleobiology
Smithsonian Institution
Washington D.C. 20560
U.S.A.
THE PALAEOBIOLOGY OF THE
CRETACEOUS IRREGULAR ECHINOIDS
INFULASTER AND HAGENO WIA
by ANDREW S. GALE and ANDREW B. SMITH
Abstract. The taxonomy of the infaunal holasteroids Infulaster and Hagenowia is revised in the light of new
material from the Senonian Chalk of southern England. Investigation of the plating structure in Hagenowia
permits a more precise definition of its species, and a better understanding of the evolution of the rostrum.
The majority of structural modifications in the Infulaster-Hagenowia lineage were caused directly, or
indirectly, by apical elongation and size reduction. A detailed study of test and ambulacral pore morphology,
tuberculation, and spines, taken in comparison with living echinoids, allows reconstruction of the life habits
of Infulaster and Hagenowia. The evolutionary changes are related to the adoption of a specialized method
of feeding and an attempt to avoid predation.
T he echinoid genus Hagenowia , in which the apical part of the test is drawn out to form a rostrum,
is the most bizarre of a wide range of holasteroids which inhabited the Upper Cretaceous Chalk
Sea of north-west Europe. The ancestry of this genus was traced by Wright and Wright (1949) to
Infulaster, and the evolutionary story further discussed and elaborated by Ernst and Schulz (1971).
The Upper Chalk in England falls into two distinct faunal provinces, the boundary of which
ran approximately east-west through north Norfolk (Peake and Hancock 1961; Reid 1976; Rawson
et al. 1978). Infulaster and Hagenowia have been regarded as characteristic members of the northern
province fauna (where they occur commonly) and recorded hitherto only as occasional rarities in
the south. Hagenowia is the most common fossil in one interval of the Yorkshire Upper Chalk
and has been used there as a local zonal index in place of Micros ter coranguinum (Rowe 1904;
Wright and Wright 1942).
Collecting during recent years from the Senonian Chalk of south-east England has yielded
abundant material of both genera and shows that they occur frequently at certain stratigraphical
levels in at least part of the southern province. This material has provided much new information
(particularly concerning the rostrum of Hagenowia) which has allowed a reappraisal of the taxonomy
and phylogeny of Infulaster and Hagenowia.
Irregular echinoid lineages in the Chalk have provided a number of classical evolutionary stories,
of which the best known is in Micros ter (Rowe 1899). Progressive morphological changes in
Micraster were interpreted by Nichols (1959) as adaptations to an increasing depth of burial. A
similar explanation has been advanced for the Infulaster-Hagenowia story (Nichols 1959; Ernst
and Schulz 1971). Our work suggests that changes in this lineage, particularly the development and
modification of the rostrum, are related to the adoption of a specialized method of feeding,
and avoidance of predation.
METHODS OF STUDY
Although the surface details of Infulaster and Hagenowia from southern England are often
well preserved, the tests are invariably distorted by compactional crushing. For this reason,
biometrical studies on the material have not been attempted.
In a detailed study of the rostrum of H. elongata (Nielsen) Schmid (1972) used a special technique
to elucidate the plating structure. This took advantage of the naturally separate reflectivity of
IPalaeontology, Vol. 25, Part 1, 1982, pp. 11-42, pis. 3-6|
12
PALAEONTOLOGY, VOLUME 25
individual plates, accentuated by gold coating. This method proved to be effective on H. blackmorei
Wright and Wright, but unsuccessful on earlier members of the genus. To work out the plating
arrangements of these, naturally weathered rostra were stained with black ink to pick out the
sutures. Surface details of the echinoids, particularly tubercles and pores, were studied with the
scanning electron microscope
Abbreviations used for museum collections are as follows: BMNH— British Museum (Natural History);
IGS— Institute of Geological Sciences, London; MMH— Mineralogical-Geological Museum, Copenhagen.
STRATIGRAPHY
In this paper the usage of stages and zones within the Upper Chalk of England have been adopted from
Rawson et al. (1978). Text-fig. 1 gives a generalized succession of all but the lowest Senonian in east Kent,
and is based on the cliff sections between Dover and Kingsdown, and on the Isle of Thanet. The section is
abstracted from detailed measurements made by one of us (A. S. G.) and gives only selected marker horizons,
to which records of species are related.
Details of the distribution of Infulaster and Hagenowia in north-west Germany are taken from Ernst and
Schulz (1971, 1974). In connection with this, it is important to note that the Turonian-Coniacian boundary
in Germany is taken at a different level from that generally used in England, where the Holaster planus-Micraster
cortestudinarium zonal junction is taken as the stage boundary (Rawson et al. 1978). The base of the Coniacian
in the German sense probably falls within the basal few metres of the coranguinum Zone at Dover, on the
basis of evidence from Inoceramus faunas.
SYSTEMATIC PALAEONTOLOGY
Order holasteroida Durham and Melville, 1957
Family holasteridae Pictet, 1857
Genus infulaster Desor, 1858
Type Species. Cardiaster hagenowi d’Orbigny 1853 (= Spatangus excentricus Woodward 1833; see Wright and
Wright 1949, p. 455).
Diagnosis. Small to medium-sized holasterids, proportionately narrow and tall, with apex positioned
anteriorly. Anterior ambulacrum set in a sulcus which runs from the apex to the transversely
rounded peristome on the adoral surface. Paired ambulacra flush with test, non-petaloid. Apical
system uninterrupted, elongated, with four genital pores. Periproct situated at summit of steep
posterior truncation. Plastron metasternal, weakly keeled. Marginal fasciole present.
Remarks. The diagnosis given above essentially follows the description of Infulaster given by Wright
and Wright (1949), with some elaboration.
Wright and Wright (1949) regarded /. excentricus as the only valid species of Infulaster, and
placed I. krausei Desor, I. borchardi Desor, I. hagenowi (d’Orbigny), and I. tuberculatus Valette in
synonomy with it. I. tuberculatus is here considered to be a separate species, and its descendant,
Hagenowia infulasteroides Wright and Wright is transferred to Infulaster. This species displays a
number of features previously thought to be exclusive to Hagenowia, notably interruption of the
pastron and elongation of the dorsolateral plate columns. Infulaster thus differs from Hagenowia
only in its uninterrupted apical system and lack of a rostrum.
Infulaster excentricus (Woodward, 1833)
Text-fig. 2 (1)
Remarks. Large forms of Infulaster occur commonly in the Middle and Upper Turonian of the
northern province in England (north Norfolk, Lincolnshire, Yorkshire) and in the eastern extension
of the province in north-west Germany (Ernst and Schulz 1971), Poland (Nietsch 1921) and the
Caucasus (Moskveena 1959). For these, the name I. excentricus is used provisionally, and it is not
intended in this paper to revise and redescribe this material, which may include more than one species.
text-fig. 1 . Generalized succession in the Senonian Chalk (lowest beds omitted) of the Kent coast, with selected
marker bands only, showing the distribution of species of Infulaster and Hagenowia. Scale on left of column
in 5 m intervals. Flints solid black.
14
PALAEONTOLOGY, VOLUME 25
Only one specimen of I. excentricus is known from southern England, from the Coniacian M.
cortestudinarium Zone at Dover, 3-2 m below the Lower East Cliff Marl (text-fig. 2 (1); BMNH
E76832, A. S. Gale Coll.)
Infulaster tuberculatus Valette, 1913
Plate 4, figs. 2, 11; Plate 5, fig. 3; Plate 6, fig. 5; text-fig. 2 (2-4)
1913 Infulaster tuberculatus Valette, p. 5, fig. 1.
p. 1949 Infulaster excentricus (Woodward) Wright and Wright, p. 456.
Holotype. Valette’s solitary specimen came from the Coniacian zone H, of Rosoy near Sens, France. The
specimen is presumably with Valette’s collection, now in the University of Dijon.
Diagnosis. Small Infulaster in which the test is short and proportionally tall. The apex may be
acutely angled. Plastron uninterrupted by ambulacra I, V.
text-fig. 2. Infulaster spp. from east Kent: 1, Infulaster excentricus (Woodward), oral view, from
3-2 m below summit of Micraster cortestudinarium Zone, East Cliff, Dover. BMNH E76832 A. S.
Gale Coll. 2-4, 1. tuberculatus Valette. 2, group of specimens from basal M. coranguinum Zone, 0-8
m above Upper East Cliff Marl, cliffs north-east of St. Margaret’s Bay, near Dover, Kent. BMNH
E76833, A. S. Gale Coll. 3 a, b , oral and lateral views of individual without stratigraphical location,
east of Dover, Kent. BMNH E10300 Cockburn Coll. 4 a-c, dorsal, oral, and lateral views of specimen
from 0-5 m below Lower East Cliff Marl, top of M. cortestudinarium Zone, East Cliff, Dover, Kent.
BMNH E76834, A. S. Gale Coll. All specimens x 2. Coated in ammonium chloride.
GALE AND SMITH: CRETACEOUS ECHINOIDS
15
Remarks. Small Infulaster with lengths of 10-20 mm are common in the Lower coranguinum Zone
of east Kent. These are almost invariably crushed and fragmentary, but the few well-preserved
individuals show close similarities in shape with Valette’s figures of I. tuber culatus. The species
differs from I. excentricus in having a proportionally shorter test with a steep posterior slope in
addition to the consistent difference in size. Many individuals are slightly inflated just posterior to
the apex. Specimens from the higher part of the range of the species at Dover, although distorted,
have acutely angled apices, and can only be distinguished from I. infulasteroides (Wright and
Wright) by their uninterrupted plastrons.
Occurrence. At Dover the species occurs in the highest 1-5 m of the M. cortestudinarium Zone,
and the basal 18 m of the M. coranguinum Zone (text-fig. 1). The species is most common just
beneath a band of large flints about 1 m above the upper East Cliff Marl, where clusters of individuals
are found. The type specimen came from the Coniacian H of Rosoy, a level equivalent to
approximately the lower half of the coranguinum Zone.
Infulaster infulasteroides (Wright and Wright 1949)
Plate 4, fig. 2
1949 Hagenowia infulasteroides Wright and Wright, p. 470, figs. 17, 18.
1971 Hagenowia infulasteroides Wright and Wright; Ernst and Schulz, p. 138, pi. 13, figs. 1-4;
text-fig. 6.
Types. The holotype is a flint steinkern from the Haldon Gravel of Devon (Wright and Wright 1949, fig. 17;
BMNH E8403). A paratype, similarly preserved, comes from flint gravel at Lulworth, Dorset (BMNH El 709).
Both were probably derived originally from the upper part of the coranguinum Zone. A second paratype, a
crushed, incomplete test from the coranguinum Zone of the North Foreland, near Broadstairs, Kent (BMNH
E33886, Rowe Coll.) almost certainly came from a level 2-4 m below Whitaker’s 3-inch band, the only horizon
at which the species is common in east Kent.
Diagnosis. Test with acutely angled apex. Plates of interambulacra 1 and 4 on sides of test elongated.
Plastron interrupted by ambulacra I and V.
Remarks. This species is transferred from Hagenowia to Infulaster on account of its undivided
apical system, and the absence of a rostrum. The general morphology and variation is well illustrated
by Ernst and Schulz (1971).
Occurrence. In east Kent /. infulasteroides ranges from 10 m below Bed well’s Columnar Band, up
to the base of the Uintacrinus Zone (text-fig. 1). It is only common 2-4 m below Whitaker’s 3-inch
band, where it occurs in clusters. I. infulasteroides is also known from the upper coranguinum Zone
of Berkshire, Hampshire, Sussex and the Isle of Wight, and the Uintacrinus Zone of Hampshire.
Ernst and Schulz (1971) record this species from the Middle Santonian of Lagerdorf.
Genus hagenowia Duncan, 1889
(= Martinosigra Nielsen, 1942)
Type species. Cardiaster rostratus Forbes 1852 by original designation.
Diagnosis. Small holasterids in which the apical part of the test is elongated antero-dorsally to
form a tapering rostrum. Narrow, well-defined sulcus runs from apex of rostrum to circular
peristome. Apical system disjunct, with two posterior oculars at base of rostrum separated from
remainder of system (which is on rostral tip) by interambulacra 1 and 4. Two or four genital pores.
Longitudinally oval periproct is positioned at the summit of the posterior truncation. Plastron
metasternal, keeled. Marginal fasciole.
Remarks. The diagnosis of Hagenowia is amended from that of Wright and Wright (1949), to
exclude I. infulasteroides from the genus. The most important diagnostic feature of Hagenowia is
16
PALAEONTOLOGY, VOLUME 25
III
text-fig. 3. Plating structure and cross-sectional shape of the rostrum in Hagenowia rostrata (Forbes): a and
b, dorsal and lateral views of a small individual from 1 m above Bedwell’s Columnar Band, East Cliff, Dover.
BMNH E76835, A. S. Gale Coll, c, cross-section of the rostrum, taken at the 3rd plate of interambulacral
row 1 b. Specimen from 1-5 m above Bedwell’s Columnar Band, Ramsgate. BMNH E76836, A. S. Gale Coll.
d, lateral view of specimen from 2 m above the East Cliff Semitabular, East Cliff, Dover. BMNH E76837,
A. S. Gale Coll.
GALE AND SMITH: CRETACEOUS ECHINOIDS
17
the separation of the apical system. Interruption of the plastron by ambulacra I and V is seen in
all Hagenowia and is also present in I. infulasteroides.
Hagenowia rostrata (Forbes, 1852)
Plate 3, figs. 1-5; Plate 4, fig. 3; Plate 5, figs. 1,2, 5; text-figs, 3, 4
1852 Cardiaster rostratus Forbes, p. 3, pi. 10, figs. 19-24.
1858 Infulaster rostatus (Forbes); Desor, p. 348.
1881 Infulaster rostratus (Forbes); Wright, p. 307, pi. 70, figs. 2, 3.
1889 Hagenowia rostrata (Forbes); Duncan, p. 210.
1942 Martinosigra rostrata (Forbes); Nielsen, p. 163.
1949 Hagenowia rostrata (Forbes); Wright and Wright, p. 462, figs. 7, 8, 1 1, 12, non 9, 10.
1971 Hagenowia rostrata (Forbes); Ernst and Schulz, p. 138, pi. 13, fig. 5; pi. 14, fig. 1; text-fig. 7.
Lectotype. A specimen figured by Forbes (1852, pi. 10, figs 19-21; IGS 38656) was taken as lectotype by
Wright and Wright (1949). This came from the ‘Ctialk with Flints of Bostal Heath, near Plumstead’ (Forbes
1852, p. 3), in south-east London, presumably from the upper part of the coranguinum Zone.
Diagnosis. Rostrum short, sharply tapering; plates of ambulacral rows Ha, IVb contiguous, not
reduced; posterior side of rostrum broad, evenly rounded; sulcus deep; subanal protruberance
double, asymmetrical.
Remarks. The plating arrangements of the rostrum is completely known only in this species of
Hagenowia (text-figs. 3, 4). Oculars I and V are small, and positioned at the base of the rostrum.
Together with ambulacra I and V, and interambulacrum 5, they are separated from the rest of the
apical system along the dorsal margin of the rostrum by interambulacra 1 and 4 (text-fig. 3). The
a
b
c
text-fig. 4. Plating structure of the rostrum tip in Hagenowia rostrata (Forbes): a-c. Dorsal, lateral, and frontal
views of large specimen from 3 m below Whitaker’s 3-inch band. North Foreland, near Broadstairs, Kent.
BMNH E76838, A. S. Gale Coll.
18
PALAEONTOLOGY, VOLUME 25
first two plates of interambulacral rows la and 4b meet along the dorsal margin of the rostrum
above oculars I and V. The second plates are often slightly swollen in lateral profile and form a
distinct, low protuberance at the base of the rostrum (e.g. PI. 3, fig. 2b). la and 4b are occluded
from the rostrum above this level by two or three plates of the outer rows, lb and 4a.
Variations in test shape of H. rostrata is shown by a group of specimens from the upper
coranguinum Zone of the Kent coast (PI. 3, figs. 1-5). Low, depressed forms with relatively high
anterior angles (PI. 3, fig. 1) intergrade continuously through to tall individuals with steep sides
and more upright rostra (PI. 3, fig. 2). No stratigraphical separation of these forms is known to
occur. Small individuals of H. rostrata have shorter, less well demarcated rostra (PI. 3, fig. 4).
Ernst and Schulz (1971, p. 138, text-fig. 7, fig. 1) recorded a stratigraphically low ‘early form’
of H. rostrata from the (? Lower) Coniacian of Lagerdorf, north-west Germany. This has a short
posterior slope, a short rostrum, and a flat profile to the base. A single specimen (PI. 3, fig. 5)
from the lowest horizon at which Hagenowia occurs at Dover, some 2 m above the East Cliff
Semitabular flint (text-fig. 1) compares quite well in lateral profile with the figured Lagerdorf
individual. The Dover specimen differs from other English examples of H. rostrata in having a
broad, relatively shallow sulcus without narrow margins. In anterior profile, the sides of the test
do not inflect sharply as the body passes into the rostrum.
Occurrence. On the Kent coast (text-fig. 1) H. rostrata is found in a succession of discrete bands
in the coranguinum Zone, from the East Cliff Semitabular to Whitaker’s 3-inch band. It occurs in
the same zone throughout its outcrop in southern England, although precise details of horizon
are seldom recorded. In northern France a solitary example was found 3 m below the equivalent
of Whitaker’s 3-inch band at Coquelles, near Calais (A. S. G. Coll.). In Yorkshire, Rowe (1904)
used H. rostrata as a local index for the M. coranguinum Zone, which was subsequently formalized
by Wright and Wright (1942). Recent study has shown that the species of Hagenowia common in
the flintless chalk of the Yorkshire coast below the entry of Uintacrinus is, in fact, H. anterior.
The only true H. rostrata we have seen from Y orkshire are from the flinty chalk with Inoceramus
involutus of Little Weighton (Wrights’ Coll.). In north-west Germany H. rostrata occurs in the
Coniacian Chalk at Lagerdorf (Ernst and Schulz 1971, text-fig. 5; 1974, text-fig. 4 a).
EXPLANATION OF PLATE 3
Figs. 1 -5. Hagenowia rostrata (Forbes). Micraster coranguinum Zone, Kent coast. 1 a-c, oral, dorsal, and lateral
views of depressed specimen from 3 m beneath Whitaker’s 3-inch band, North Foreland near Broadstairs.
BMNH E76838, A. S. Gale Coll. 2a, b, oral and lateral views of high individual, with more vertical rostrum,
retaining radioles on the base from 2 m above Bedwell’s Columnar Band, same locality. BMNH E76848, A. S.
Gale Coll. 3, lateral view of specimen, same horizon and locality. BMNH E76849, A. S. Gale Coll. 4, small
individual with short rostrum, 1 m above Bedwell’s Columnar Band, East Cliff, Dover. BMNH E76835, A. S.
Gale Coll. 5, lateral view of specimen with poorly demarcated rostrum. 2 m above East Cliff Semitabular, East
Cliff, Dover. BMNH E76850, A. S. Gale Coll.
Fig. 6a, b. H. anterior Ernst and Schulz. Oral and lateral views of body. The rostrum has broken off, and its base
is bored. Uintacrinus Zone, Harding’s Whiting Pits, Devizes Road, Salisbury Wilts. BMNH E35788,
Blackmore Coll.
Figs. 7, 8. H. blackmorei Wright and Wright, la-c, lateral, frontal, and oral views of the holotype (body only).
The rostrum belongs to H. anterior and has been artificially attached. Lower Campanian, probably lower G. |
quadrata Zone, West Harnham, near Salisbury, Wilts. BMNH E33916, Blackmore Coll. 8a, b, lateral and
frontal views of rostrum from Hagenowia horizon, lower G. quadrata Zone, pit no. 3 of Gaster (1924), North
Lancing, near Worthing, Sussex. I.G.S. no. Zm 2907.
All specimens x 2. Coated with ammonium chloride.
PLATE 3
1
5 IIEP*
II
| -mm
k 8 ,
GALE and SMITH, Cretaceous irregular echinoids
20
PALAEONTOLOGY, VOLUME 25
Hagenowia anterior Ernst and Schulz, 1971
Plate 3, fig. 6; Plate 4, figs. 5, 10, 12; Plate 5, figs. 6, 8; Plate 6, figs. 2, 6, 8; text-fig. 5
1949 Hagenowia rostrata (Forbes); Wright and Wright, p. 462, figs. 9, 13.
1971 Hagenowia blackmorei Wright and Wright, anterior Ernst and Schulz, p. 140, pi. 13, fig. 6;
pi. 14, figs. 2, 3; text-fig. 8.
Types. The holotype is from the Middle Santonian rogalae-westfalica Zone of Lagerdorf, north-west Germany
(Ernst and Schulz 1971, text-fig. 8, fig. 1; pi. 14, figs. 2, 3). The paratypes are from the same horizon and locality.
Diagnosis. Rostrum long and slender, cross-section triangular, posterior side evenly rounded;
individual plates of ambulacral rows Ha, IVb small, separated by interambulacral rows la, 4b and
ambulacra! rows lib, IVa in the rostrum; subanal protruberance single, narrow.
Remarks. The plating arrangement of all but the basal rostrum is known in this species (text-fig.
5). Individual plate rows are proportionately longer and narrower than in H. rostrata. At least six
plates of interambulacral rows la, 4b meet along the dorsal margin of the rostrum (text-fig. 5b);
these are variable in width. Plates of ambulacral rows Ha and IVb are small, and although the
first few are usually in contact with each other, most are separated by la and 4b, and lib, IVa
(text-fig. 5a-c, e). The size and shape of these occluded plates is variable. In some individuals (e.g.
text-fig. 5e) plates of Ila and IVb are locally separated by interambulacra la, 4b and 2a, 3b.
Genitals 1 and 4 are absent.
text-fig. 5. Plating structure and cross-sectional shape of the rostrum in Hagenowia anterior Ernst and Schulz:
a , dorsal view of rostrum from 5 m above Whitaker’s 3-inch band, Kingsgate, Kent. BMNH E76839, A. S. Gale
Coll, b, dorsal view of rostrum from basal 1 m. of Marsupites Zone, Minnis Bay, Kent. BMNH E76840, A.
S. Gale Coll, c, lateral view of rostrum tip, mid Uintacrinus Zone, near Margate, Kent. BMNH E76841, A.
S. Gale Coll, d, cross-section of rostrum, taken at 2nd/3rd plate of interambulacral row 1 b, mid Uintacrinus
Zone, near Margate, e, lateral view of rostrum, base of Uintacrinus anglicus band, Foreness Point, near
Margate. BMNH E76842, A. S. Gale Coll.
GALE AND SMITH: CRETACEOUS ECHINOIDS
2
Some variation in the attitude of the rostrum occurs in M. anterior , although the body shape
is quite consistent. Lateral furrows on the sides of the test, just anterior to the periproct (Ernst
and Schulz 1971, text-fig. 8, figs. 2, 3) are sometimes present. H. anterior is raised to specific rank
on account of significant differences in rostral structure and body shape from H. blackmorei.
Occurrence. In east Kent the species first appears at 3-5 m above Whitaker’s 3-inch band in the
coranguinum Zone and ranges up into the Uintacrinus anglicus band in the lower Offaster pilula
Zone (text-fig. 1). Levels of abundance occur in the mid U. socialis Zone and in the basal Marsupites
Zone. Scattered records of this species exist from Dorset, Wiltshire, Hampshire, and the Isle of
Wight ( Uintacrinus Zone). In Yorkshire H. anterior occurs commonly in the flintless upper part
of the coranguinum Zone and ranges up into the lower Campanian I. lingua Zone. At Lagerdorf
the species ranges from the mid rogalae-westfalica Zone to the Marsupites Zone, with one doubtful
earlier record from the coranguinum-westfalica zone (Ernst and Schulz 1971, 1974).
Hagenowia blackmorei Wright and Wright, 1949
Plate 3, figs. 7, 8; Plate 4, fig. 6; Plate 6, figs. 4, 7, 9; text-fig. 6
1949 Hagenowia blackmorei Wright and Wright, pp. 467-70, figs. 14-16.
1971 Hagenowia blackmorei blackmorei Wright and Wright; Ernst and Schulz, p. 140, text-fig. 2.
Types. The specimen chosen as holotype by Wright and Wright (1949, fig. 14) is from the Lower Campanian
Chalk of East Harnham, near Salisbury, Wiltshire (BMNH E33916, Blackmore Coll.), probably from the G.
quadrata Zone. However, as pointed out by C. J. Wood (in Ernst and Schulz 1971), the body and rostrum
belong to two specimens of separate species, and have been artificially joined. The body belongs to the species
text-fig. 6. Plating structure and cross-sectional shape of the rostrum in
Hagenowia blackmorei Wright and Wright: a, b, dorsal and lateral views of
rostrum, based on specimen from 10 m. above planoconvexa bed, G. quadrata
Zone, cliffs west of Newhaven, Sussex. BMNH E76843, A. S. Gale Coll.
c, cross-section of rostrum 3 mm below tip, same horizon and locality.
22
PALAEONTOLOGY, VOLUME 25
characteristic of the lower quadrata Zone of southern England, and possesses the features of H. blackmorei
as described by Wright and Wright. The rostrum is from an individual of H. anterior from the Santonian.
The body alone is therefore selected as lectotype. The paratype is an internal and external mould in flint from
the Haldon Gravel, Devon (Wright and Wright 1949, figs. 14, 15; BMNH E8404, Vicary Coll.).
Diagnosis. Rostrum slender, upright, well demarcated from body, equilaterally triangular in
cross-section; interambulacral rows la, 4b form narrow dorsal ridge on rostrum, surface of
ambulacral rows lib, IVa slightly inset; sulcus shallow, with a gently concave floor; Ha, IVb not
present in rostrum; body squarish in lateral profile, anterior part of base vertically below peristome;
subanal protruberance single, small; posterior slope steep; fasciole noticeably oblique.
Remarks. The structure of the rostrum in this species (text-fig. 6) is similar to that in H. elongata
(Nielsen) (Schmid 1972) but individual plates are less elongated and broader than in the
Maastrichtian species. The dorsal ridge on the rostrum is broader and less sharply defined than
in H. elongata, and genitals 2, 3 are not usually separated from oculars II, IV in H. blackmorei,
as is invariably so in H. elongata. The latter species has only two, vertically situated, madreporic
pores, whereas the number and arrangement of these is variable in H. blackmorei.
A few Hagenowia rostra are known from the Upper Campanian Belemnitella mucronata Zone
of southern England (Wright and Wright 1949). These show affinities with both H. blackmorei
and H. elongata, but lack the specialized madreporic pores of the latter and are referred to H. cf.
blackmorei.
Occurrence. H. blackmorei appears to be restricted to the lower part of the G. quadrata Zone in
southern England. Gaster (1924) used the term Hagenowia horizon for this level in Sussex. Collecting
from the cliffs west of Newhaven in Sussex suggests that the species occurs only within the 12 m
of chalk above the marl pair (planoconvexa bed) taken by Brydone (1914) as the base of the G.
quadrata Zone. The species is only abundant in the interval from 9-11 m above this bed, where
approximately fifty specimens have been found. Blackmore’s material from East Harnham lacks
detailed stratigraphical location, but probably came from the correlative horizon.
Hagenowia elongata (Nielsen, 1942)
Plate 4, figs. 4, 7-9; Plate 5, figs. 7, 9; Plate 6, figs. 1, 3, 10
1942 Martinosigra elongata Nielsen, p. 163, fig. 2.
1949 Hagenowia rostrata (Forbes); Wright and Wright, p. 462.
1950 Hagenowia elongata (Nielsen); Mortensen, p. 97, fig. 100.
1971 Hagenowia elongata (Nielsen); Schmid, in Ernst and Schulz, p. 141, text-fig. 2, el.
1972 Hagenowia elongata (Nielsen); Schmid, p. 179, pis. 1-4; text-figs. 1, 2.
Diagnosis. Plate rows in rostrum very long and narrow; genitals 2 and 3 separated from oculars
II and IV by interambulacral rows 2a, 3b; madreporite with only two, vertically situated, pores;
sulcus shallow, flat.
Remarks. Now that the rostral structures of preceding species of Hagenowia are known, it is
necessary to modify the interpretation of this species from Schmid (1972). Instead of representing
both plate rows of ambulacra II and IV, the four rows of plates on the dorsal side of the rostrum
are lib, IVa (outer rows) and lb, 4a (inner rows). There is little else to add to Schmid’s excellent
description of the species. The outline of a body stated to belong to this species was shown
by Ernst and Seibertz (1977, text-fig. 6), but this new material awaits detailed description and
figuring.
Occurrence. Upper Lower Maastrichtian, Zone of Belemnella occidentalis, Denmark and north-west
Germany.
GALE AND SMITH: CRETACEOUS ECHINOIDS
23
PHYLOGENY
Wright and Wright (1949) suggested that I. infulasteroides arose from the ‘small I excentricus ’ of
the cortestudinarium Zone (here I. tuberculatus ) and subsequently gave rise to H. rostrata. They
regarded H. blackmorei as a short-lived offshoot from H. rostrata, which itself survived into the
Upper Campanian. Ernst and Schulz (1971, p. 141, text-fig. 2) modified this story only slightly,
by placing the origin of H. rostrata near the point at which an infulasteroides lineage diverged
from the main Infulaster stock. They retained separate rostrata and blackmorei lineages throughout
the Santonian and Campanian. More recently, Ernst and Seibertz (1977, p. 563, text-fig. 6) showed
the H. rostrata lineage as terminating within the Middle Santonian.
It is suggested here that only two lineages are recognizable in this group of echinoids.
1. Infulaster lineage: I. excentricus ( sensu lato) I. tuberculatus— I. infulasteroides
2. Hagenowia lineage: H. rostrata — H. anterior — H. blackmorei — H. elongata.
Since Hagenowia appeared before I. tuberculatus gave rise to I. infulasteroides, it is likely that the
origins of this genus lay in the former species, probably during the mid-Coniacian. In this case,
I. infulasteroides parallels rather than antecedes Hagenowia in the development of an acutely angled
apex and an interrupted plastron.
Although the over-all changes in each of the lineages are progressive, individual species are in
general well demarcated from both ancestor and descendant. No two species in either lineage are
known definitely to overlap in stratigraphical order.
EVOLUTIONARY TRENDS
1. Size. In the Infulaster lineage there is a dramatic decrease in size from I. excentricus, which is
characteristically between 40 and 50 mm in length, to I. tuberculatus (15-25 mm). With this change,
the test became proportionately shorter and deeper. In late I. tuberculatus and in I. infulasteroides
the apex of the test became more acutely angled resulting from an increase in the length of plate
rows on the sides of the test.
2. Rostrum. In Hagenowia the apical system is separated into two parts to form a trivium and
bivium. This was a direct consequence of the considerable elongation of the apical region.
In the earliest species, H. rostrata, oculars I and V are separated from genitals 1 and 4 by
interambulacral rows 1 and 4 along the dorsal side of the rostrum. Plate rows la and 4b are
occluded from the distal rostrum by rows lb and 4a, which meet along the dorsal mid-line. Rostral
plates are elongated, the sulcus is deep, and the body cavity within the rostrum is large. In profile,
the rostral tip is only slightly enlarged.
In H. anterior the rostrum is better demarcated from the body and is relatively narrow. Plate
rows are proportionally narrower and more plates per row are present in the rostrum. Plates of
ambulacral rows Ila and IVb are typically diminutive while genitals 1 and 4 have been lost. In
comparison with H. rostrata, the rostrum of H. anterior is more equilaterally triangular in
cross-section and the rims of the sulcus, formed by interambulacral rows 2b and 3a, are distinctly
thickened. The walls of the rostrum are also proportionally thicker and the body cavity smaller.
The sulcus is significantly shallower.
The trend towards increasing slenderness and sharper demarcation of the rostrum is continued
in H. blackmorei. Ambulacral rows Ila and IVb are lost completely from the part of the rostrum
where the plating arrangement is known. Rostral plates are more elongate than in previous species.
In cross-section, the rostrum is triangular and thickened at each corner while the body cavity is
further reduced. Ambulacral plate rows Ila and IVb are slightly depressed. The sulcus is shallower
than in H. anterior and its rims are parallel. The tip of the rostrum is broader than the shaft and
the rostrum is more or less vertical in attitude.
Rostral development is most pronounced in H. elongata. Rostral plates are narrower and more
elongate than in any of its predecessors. Interambulacral plate rows 2a and 3b always separate
24 PALAEONTOLOGY, VOLUME 25
genitals 2 and 3 from oculars II and IV. Only two vertically placed madreporic pores are
present.
Summarizing these trends, there is a progressive increase in the slenderness and clearer
demarcation of the rostrum. This is achieved through elongation and narrowing of plate rows,
increasing the number of plates in each row present, and loss of plate rows. The rostrum adopts
a more vertical attitude and is strengthened by the development of buttresses and by thickening of
the walls. The body cavity within the rostrum is progressively reduced and genitals 1 and 4 are
lost during evolution.
3. Body. The body of H. rostrata differs little from that of the ancestral Infulaster. During the
evolution of Hagenowia there were five important changes to the structure of the body. Firstly,
the plastron became interrupted in I. infulasteroides and in all species of Hagenowia. Secondly, the
base became progressively more convex. Thirdly, the peristome moved from a basal to an anterior
position with a vertical attitude. Fourthly, the posterior slope became more vertical and the fasciole
more oblique. Finally there was a change from a double, asymmetrical sub-anal protruberance in
H. rostrata to a single centrally placed protruberance in later species.
4. Pore morphology. The more important changes in pore morphology during the evolution of
Infulaster and Hagenowia include the progressive loss of phyllode and dorso-lateral pores and the
increasing differentiation of pores at the rostral tip. Three distinct regions of pores are recognizable
and will be dealt with separately. Pore terminology is taken from Smith (1980a).
(a) Phyllode pores. In Infulaster, phyllode pores are relatively large isopores, rounded in outline,
with an axially positioned neural canal in an adoral position. They are 300-350 in length in I.
excentricus but only 180-250 /z m in length in later species. Thirty such pores lie around the peristome
in I. excentricus but only 20 to 22 are present in I. tuberculatus and I. infulasteroides. Similar
phyllode pores are found in H. rostrata and H. anterior. H. rostrata has between 14 and 18 such
pores while H. anterior has slightly fewer (the exact number cannot be ascertained in any of the
specimens examined). H. blackmorei has only two small, circular unipores lacking any periporal
ornament. These lie on either side of the anterior sulcus immediately above the peristome. There
is no sign of pores in lateral and posterior ambulacra adjacent to the peristome.
( b ) Latero-dorsal pores. In Infulaster, isopores of the posterior columns of ambulacra II and IV
are twice as long as other latero-dorsal pores. These elongate isopores have no obvious neural
EXPLANATION OF PLATE 4
Bottom of micrograph adoral unless otherwise stated.
Fig. 1. Infulaster infulasteroides: latero-dorsal partitioned isopore from interambulacrum 5. BMNH E76851.
Fig. 2. I. tuberculatus: partitioned isopore with axially positioned neural canal near the apex of ambulacrum III.
BMNH E76852.
Fig. 3. Hagenowia rostrata: partitioned isopore with axially positioned neural canal near the apex of
ambulacrum III. BMNH E7684.
Fig. 4. H. elongata: simple unipore of ambulacrum III within the rostral sulcus. MM 12820.
Fig. 5. H. anterior: unipore with broad periporal area in ambulacrum III at the rostral head. Adoral to the right.
BMNH E76844.
Fig. 6. H. blackmorei: unipore with broad periporal area in ambulacrum III at the rostral head. BMNH E76846.
Fig. 7. H. elongata: unipore in ambulacrum III at the rostral head.
Fig. 8. H. elongata: tuberculation on the lateral face of the rostrum. Apex to the left, dorsal ridge at top.
Fig. 9. H. elongata: enlargement of fig. 8 showing one tubercle.
Fig. 10. H. anterior: dorsal tubercle on rostrum showing enlarged areole and radially symmetrical crenulation.
Fig. 11. /. tuberculatus: latero-dorsal tubercles.
Fig. 12. H. anterior: latero-dorsal tubercles at the base of the rostrum.
Scale bar in figs. 1-4, 6, 7, 9, 12 = 100 ^m; figs. 5, 8, 10, 1 1 = 200 Mm.
PLATE 4
GALE and SMITH, Cretaceous irregular echinoids
26
PALAEONTOLOGY, VOLUME 25
canal and the interporal partition is more or less flush with the test surface and may be very broad.
Other dorso-lateral pores are partitioned isopores with a small, laterally positioned neural canal
(PI. 4, fig. 1). All isopores are widely spaced and have no recognizable attachment area. Pores of
I. excentricus are approximately twice the size of pores in later species.
Latero-dorsal pores vary tremendously in specimens of H. rostrata. In some of the larger specimens
there are isopores, arranged as in Infulaster. One of these has elongate isopores of ambulacra lib
and IVa which are even larger than those in I. excentricus. The majority of specimens lack such
elongate isopores and have small partitioned isopores in all columns. In other specimens, all
latero-dorsal pores are reduced to unipores. Finally, one specimen has unipores in ambulacral
columns lib and IVa and small partitioned isopores in columns Ila and IVb. In H. anterior there
are no pores in rostral plates of ambulacra II and IV but there are between three and five partitioned
isopores in each column on the body above the fasciole. There are no latero-dorsal pores in H.
blackmorei.
( c ) Ambulacrum III pores. Pores in the frontal sulcus are similar in size and shape in the three
species of Infulaster and in H. rostrata. These are partitioned isopores with axially aligned neural
canals (PI. 4, figs. 2, 3). Pores near the apex are identical with those within the sulcus. In later
species of Hagenowia, pores in the sulcus are smaller than those at the apex. The 12 to 14 most
adapical pores in H. anterior and H. blackmorei are unipores, circular in outline, and with a narrow
central pore surrounded by a clear periporal area (PI. 4 figs. 5, 6). Unipores in the sulcus are
smaller with a reduced periporal area. In the apical unipores of H. elongata, the central pore is
much larger and most of the periporal area is lost (PI. 4, fig. 7). Unipores in the sulcus are minute
and lack periporal ornament (PI. 4, fig. 4).
text-fig. 7. Tubercle arrangement in Infulaster excentricus. a — anterior: b — lateral: c — posterior: d — aboral:
e— oral. Arrows on the left-hand side of diagrams show the direction of areole enlargement. Circles indicate
radially symmetrical tubercles. Stippling on the right-hand side of diagrams indicates the size and arrangement
of tubercles. The marginal fasciole appears as a black band.
GALE AND SMITH: CRETACEOUS ECHINOIDS
27
5. Tubercle arrangement. Although some marked changes occur in the detailed arrangement of
tubercles during the evolution of Hagenowia, all functionally distinct groups of tubercles were
already developed in I. ex centricus. The more important changes are outlined below.
(i) There is a slight reduction in the area of plastron tubercles in later species of Hagenowia.
This is accompanied by a corresponding increase in the tubercle-free ambulacral zones on either
side of the plastron. The plastron is extremely narrow in H. blackmorei (text-fig. 8f).
(ii) There is a relative increase in the area of ventro-lateral tubercles, which are best developed
in H. blackmorei. This is a consequence of the increasing obliquity of the fasciole (which, in turn,
arose from the progressive shift of the peristome).
(iii) Tubercles in the sub-anal region are radially arranged around two points in Infulaster and
H. rostrata (text-figs. 7, 8). In H. anterior and H. blackmorei , sub-anal tubercles are radially arranged
around a single point (text-fig. 8i).
text-fig. 8. Tubercle arrangement in Hagenowia. a-e, H. rostrata-. a— oral; b— anterior;
c — posterior; d— aboral; e— lateral, f-j, H. blackmorei. f — oral; G — anterior; h— lateral; i— posterior;
J — arrangement at the base of the sulcus and surrounding the peristome. For explanation see
text-fig. 7.
28
PALAEONTOLOGY, VOLUME 25
(iv) Latero-dorsal tubercles in Infulaster have an areole which is radially symmetrical or only
weakly enlarged adapical (PI. 4, fig. 11). In Hagenowia the areole becomes increasingly enlarged
in an adapical direction, whilst crenulation is enlarged on the opposite side (PI. 4, figs. 8-10, 12; text-
fig. 8). Latero-dorsal tubercles are least dense immediately above the marginal fasciole and become
progressively denser towards the dorsal ridge. Tubercle density over the dorsal surface increases
progressively during evolution. There are between 1 and 3 tubercles per mm2 in 7. excentricus and
I. tuber culatus and this increases to 4 or 5 tubercles per mm2 in 7. infulasteroides. A similar tubercle
density is found laterally at the base of the rostrum in species of Hagenowia. On the rostrum itself,
tubercle density increases to 8 to 10 tubercles per mm2 in 77. blackmorei and as much as 15 tubercles
per mm2 in 77. elongata where there are four closely packed rows of tubercles.
(v) The parts of interambulacra 2 and 3 which form the outer zone of lateral walls of the anterior
sulcus have relatively small and radially symmetrical tubercles (PI. 5, figs. 3-9). There are three
rather irregular rows of these tubercles above the marginal fasciole in 7. excentricus. In 7. tuberculatus
and 7. infulasteroides the interambulacral zones are narrower and there is only a single rather
irregular row of tubercles marginal to the sulcus (PI. 5, fig. 3). In 77. rostrata these tubercles are
linearly arranged along the sulcus lip and each is separated from its neighbours by a single row
of miliaries (PI. 5, fig. 5). A similar arrangement is found in later species of Hagenowia though
without the intervening miliaries (PI. 6, figs. 1-4).
(vi) Tubercle density progressively increases on the lateral and posterior edges of the peristome
in Hagenowia. They are best developed in 77. blackmorei where they form a dense U-shaped band
at the base of the sulcus (text-fig. 8j).
EXPLANATION OF PLATE 5
Bottom of micrograph adoral unless otherwise stated.
Fig. 1. Hagenowia rostrata : ventral view of rostrum apex with ambulacrum III on the right. The large
interambulacral tubercles at the apex face outwards and their spines obviously did not converge. BMNH
E76845.
Fig. 2. H. rostrata: dorsal view of rostrum apex. The large interambulacral tubercles form a collar around the
frontal sulcus. BMNH E76845.
Fig. 3. Infulaster tuberculatus-. ambulacrum III towards the apex. Large apical interambulacral tubercles set
perpendicular to the anterior ambulacrum can be seen in the top left corner. Smaller interambulacral
tubercles, adjacent to ambulacrum III, are irregularly arranged and their spines would not have formed a
protective arch above the ambulacrum. BMNH E76852.
Fig. 4. 7. tuberculatus-. enlargement of fig. 3 showing the relatively dense arrangement of variably sized tubercles
in ambulacrum III.
Fig. 5. H. rostrata: stereo view of the rostral sulcus slightly below the apex. Ambulacrum III is deeply sunken
and has denser and more uniformly sized tuberculation than 7. tuberculatus. Interambulacral tubercles
bordering the sulcus are arranged in two rather irregular rows, an inner row of small tubercles, that would
have supported spines forming a protective arch, and an outer row of more laterally facing tubercles for
excavating spines. BMNH E76845.
Fig. 6. H. anterior, ventral view of rostral sulcus about mid-length, apex to the right. Ambulacrum III is
moderately sunken and rather sparsely covered in uniformly sized miliary tubercles. There are two well-
defined rows of tubercles on the adjacent interambulacra, an inner row of small tubercles for spines of the
protective arch and an outer row of larger, laterally facing tubercles for excavating spines. BMNH E76844.
Fig. 7. H. elongata : oblique ventral view of rostral sulcus, apex to right. Ambulacrum III is only slightly sunken
and the two rows of adjacent interambulacral tubercles are well developed and set at different angles to the
sulcus. MM 12820.
Fig. 8. H. anterior, enlargement of fig. 6.
Fig. 9. 77. elongata-. as fig. 7 but viewed perpendicular to the sulcus.
Scale bar in figs. 4, 7-9 = 200 ^ m; figs. 1-3, 5, 6 = 400 ^m.
PLATE 5
GALE and SMITH, Cretaceous irregular echinoids
30
PALAEONTOLOGY, VOLUME 25
(vii) The large tubercles of the anterior interambulacra found on either side of the frontal sulcus
are arranged in two rather irregular rows with interspersed miliaries in species of Infulaster. With
the development of the rostrum in Hagenowia, the areas of interambulacra which face towards
the anterior are reduced and there is only a single row of these large tubercles on either side.
Tubercles are rather irregularly arranged in H. rostrata but in later species they abut to form a
well-defined row (PI. 5, figs. 5-9; PI. 6, figs. 1, 2).
(viii) The size and density of the miliaries which cover the floor and walls of the frontal sulcus
vary amongst species (Table 1). Miliaries are largest in I. infulaster oides and H. rostrata and become
progressively smaller in later species. Miliary density does not vary along the length of the sulcus
in I. excentricus and I. tuberculatus, but in I. infulasteroides and in Hagenowia the miliaries are
denser at the top of the sulcus, where they may reach densities equivalent to that found in the
lateral fasciole (120 to 140 miliaries per mm* 1 2). There is a corresponding increase in the density of
miliaries lying outside the sulcus near the head of the rostrum. This results in the development of
a rather diffuse band of miliary tubercles that fades towards the dorsal surface of the rostrum.
table 1. Size and density of miliary tubercles in ambulacrum III
Species
Adapical miliaries
Median miliaries
Size
Oun)
Mean
density
(mm-2)
Size
(/am)
Mean
density
(mm-2)
Infulaster excentricus
50-60
35
50-60
35
I. tuberculatus
60-70
50-55
50
50-55
I. infulasteroides
50-75
80
80-100
50-60
Hagenowia rostrata
60-75
75
80-90
40-50
H. anterior
50
100
60-70
60-70
H. blackmorei
40-50
130
50-60
70
H. elongata
40-50
140
50-60
60
FUNCTIONAL MORPHOLOGY
Size and Shape
1. Size. The marked reduction in the size of Infulaster in the early Coniacian could have had
several possible advantages. A reduction in size increases the surface area to volume ratio. This
eases the animal’s respiratory demands by increasing the percentage of oxygen that can be obtained
by direct diffusion. By reaching sexual maturity at an earlier stage, smaller species can have a
shorter generation time. This could have arisen with the development of stable and highly suitable
conditions, which would favour those species able to reproduce and multiply more rapidly. The
reduction in size may also have been brought about by predatory pressures. Small animals with
apical elongation would be much less obvious from the surface, when buried, than larger,
semi-infaunal species.
2. The anterior ambulacrum. The anterior ambulacrum provides the main, if not sole, passageway
for transferring sediment adorally in Recent spatangoids (Smith 19806). In certain species, spines
of the anterior ambulacrum are specialized for mucous string feeding (Chesher 1963; Buchanan
1966). The sunken anterior ambulacrum in Infulaster and Hagenowia must also have provided an
important pathway for sediment transportation. In Infulaster the anterior sulcus is deep but in
Hagenowia it shallows progressively. This is accompanied by increasing development of the
protective arch of spines covering the sulcus (see later, p. 36). The reduction in the depth of the
GALE AND SMITH: CRETACEOUS ECHINOIDS
31
sulcus is linked to the development of rostrum. The sulcus decreases in depth as the rostrum
increases in slenderness because of the reduction in internal volume. In Hagenoxvia the frontal
channel is maintained as the sulcus shallows by the development of a dense grill of spines arching
across ambulacrum III.
3. The rostrum. The development and elongation of the rostrum permitted Hagenowia to live
buried within the sediment while maintaining contact with surface waters (Nichols 1959; Ernst and
Schulz 1971). Via the rostrum, oxygenated water could be drawn into the burrow and the surface,
organic-rich layer of sediment passed to the mouth. Because of its small size, Hagenowia cannot
be said to have burrowed deeply. The oral surface of H. blackmorei could only have been about
3 cm below the surface, approximately the same level as postulated for I. excentricus. However,
the rostrum presumably made Hagenowia less obvious from the surface since only the rostral tip
would have disturbed the surface sediment and not the whole dorsal surface (text-fig. 11). Predation
may have been an important selection pressure.
The rostrum, which is oblique in H. rostrata, becomes progressively more vertical in later species.
Simultaneously it also develops a cross-section with thicker plates and buttressing. Early Hagenowia
presumably needed an oblique rostrum to streamline their movement through the sediment and
thus minimize the stresses applied to the thin-walled rostrum. In later species, as the rostrum
became more robust, it could be held in a more vertical position.
4. The anterior movement of the peristome. In Infulaster the peristome lies on the oral surface
near the anterior margin (text-fig. 7) but in H. rostrata it lies marginally at the base of the anterior
sulcus and in H. blackmorei it has shifted to a frontal position (PI. 3, fig. lb). This change is
thought to reflect an increasing dependence on sediment coming down the anterior sulcus. In
Recent spatangoids, phyllode tube-feet are used to collect sediment from the floor of the burrow
and transfer it to the mouth (Nichols 1959; Chesher 1969; Smith 1980o). This is also likely to
have been true in Infulaster. Particles coming down the anterior sulcus would have landed on the
floor of the burrow just in front of the peristome and within reach of the tube-feet. In H. rostrata
the mouth lies at the base of the sulcus so that tube-feet would have been able to collect and
transfer particles direct from the sulcus. In H. blackmorei the mouth is anterior and raised well
above the floor of the burrow. Peristomial tube-feet are reduced and, because of their position,
could only have been involved with sediment coming down the sulcus.
5. The plastron. The oral surface is relatively flat in Infulaster but becomes more or less keeled
in Hagenowia. An arched or keeled plastron is found in Recent spatangoids which burrow in
muddy or sandy substrata. Surface-dwelling spatangoids or spatangoids that burrow shallowly in
sands or gravels have a flat plastron. A keeled or arched plastron may provide a better arrangement
of tubercles and spines for efficient forward thrust.
6. Sub-anal protruberances. Infulaster has two clear bulges in the sub-anal region whereas H.
anterior and H. blackmorei have only a single protruberance. In H. rostrata there are two sub-anal
protruberances which are asymmetrical. Usually the right-hand bulge is larger and more prominent
than the other. Devries (1953) showed that asymmetry in the test of spatangoids relates to the way
in which the gut is coiled. This probably accounts for the asymmetry of the sub-anal region in H.
rostrata.
The change from a double to a single sub-anal protruberance is probably linked with the
development of asymmetry. Each bulge bore a tuft of spines (see p. 34). Whereas the test of I.
excentricus was broad enough to have had two distinct tufts of sub-anal spines, later species had
progressively narrower bodies. Presumably, as internal volume reduced, the gut became more
tightly packed and any asymmetry this gave to the test was enhanced. In Hagenowia this made
one of the tufts of spines larger and more posterior in position. The tuft of spines on the smaller
bulge would therefore have become less important and was lost in later species.
Pore morphology
1. Latero-dorsal pores and their tube-feet. Pore morphology suggests that specialized respiratory
tube-feet were present only in species of Infulaster and in some large forms of H. rostrata. By
32
PALAEONTOLOGY, VOLUME 25
comparison with extant species, these tube-feet were probably rather elongate with a central,
partitioned region (see Smith 1980a). Respiratory tube-feet were best developed in the posterior I
columns of ambulacra II and IV. Other columns either had less elongate respiratory tube-feet or |
had thin-walled cylindrical tube-feet. The positioning of respiratory tube-feet presumably reflects
the water-circulation pattern over the aboral surface.
The unipores and partitioned isopores in H. rostrata would have unspecialized sensory tube-feet.
In later species of Hagenowia, aboral tube-feet are reduced and finally lost. This, at first glance,
appears to be rather a strange adaptation for an infaunal echinoid. However, holasteroids, in
general, show no modifications for protecting aboral tube-feet during burial. Respiratory tube-feet
can function efficiently in infaunal spatangoids because of their sunken ambulacra and the curved
arch of protective spines above them. Respiratory tube-feet in clypeasteroids and cassidulids function
efficiently because of their extreme elongation (Smith 1980a). In Infulaster, respiratory tube-feet
were neither extremely elongate, nor associated with sunken ambulacra nor a protective arch of
spines. The tube-feet were able to play an effective part in gaseous exchange only because much
of the dorsal surface remained uncovered. In Hagenowia, all but the extreme tip of the rostrum
was covered, thus it became impossible for the tube-feet to function efficiently and they were quickly
lost.
The change in efficiency of the respiratory tube-feet may be linked with the over-all reduction
in body size that occurred. As more of the dorsal surface became covered by sediment, with apical
elongation, respiratory tube-feet presumably became less efficient. To compensate for this, a
reduction in body size took place so that a larger proportion of the total oxygen consumed could
come from direct diffusion. In Hagenowia, body size is further reduced with the complete loss of
latero-dorsal tube feet.
2. Pores and tube-feet of the anterior ambulacrum. Small isopores are found along the length of
the anterior ambulacrum in species of Infulaster and in H. rostrata. Their similarity with the ambital
isopores in other ambulacra and the fact that they are uniform in shape and size along the whole
length of the ambulacrum (excluding phyllode pores) indicates that the associated tube-feet were
sensory in function and terminated in a sensory pad rather than a disc. The pores diverge little as
they pass inwards, and the accompanying ampullae were probably cylindrical (Smith 1980a). \
These tube-feet are likely to have actively probed the sediment in front of them as well as sensing
the particles passing down the sulcus.
In later species of Hagenowia, isopores are replaced by unipores. Those in the sulcus are extremely
small and, by comparison with Recent spatangoids, are likely to have borne small, non-extensible
epithelial knobs which were chemosensory in function. Apical unipores differ both in size and ;
shape. Similar broad-rimmed unipores in Recent holasteroids and spatangoids bear relatively large
and extensible, sensory tube-feet (Smith 1980a). Such tube-feet are likely to have been important
chemical and tactile sense organs and no doubt would have extended out of the funnel to probe
the surrounding sediment.
The increase in pore size in apical unipores of H. elongata suggests that the associated tube-feet
were more extensible than those of previous species. A larger pore means that a larger volume of
coelomic fluid can pass in and out of the tube-feet rapidly.
3. Phyllode pores and their tube-feet. The isopores surrounding the peristome are much larger
than isopores of other non-respiratory tube-feet in both Infulaster and Hagenowia. Phyllode tube-feet
of Recent spatangoids and holasteroids are penicillate and work by mucous adhesion (Smith 1980a).
Their broad disc requires the support of a large diameter stem. They are therefore associated with
large, oval isopores or unipores. The arrangement of a few large pores around the peristome in
Infulaster and Hagenowia indicates that their tube-feet also collected sediment by means of a
mucous adhesive disc. However, the disc could not have been particularly broad, judging from
the size of the isopores, and may not have been penicillate.
There are only two small unipores around the peristome of H. blackmorei. These are likely to
have been associated with sensory tube-feet.
GALE AND SMITH: CRETACEOUS ECHINOIDS
33
Spine morphology
A specimen of /. infulasteroides (BMNH E35766) and a specimen of H. rostrata (BMNH E76848)
both retain a large number of different spines. Camera lucida drawings of these spines are given
in text-fig. 9.
A large number of latero-dorsal spines lie near the dorsal ridge in the specimen of I. infulasteroides.
These are about 1-75 mm in length and have a pronounced spatulate tip. The spatulate tip is set
oblique to the straight shaft. Amongst these spatulate spines is a shaft, about 5 mm in length, with
neither tip nor base preserved. The thinner, distal end tapers gently and is unspatulate. This gently
curved spine almost certainly attached to one of the large interambulacral tubercles at the apex
of the anterior sulcus.
Ventro-lateral spines, found below the marginal fasciole of I. infulasteroides, are spatulate and
2 to 2-5 mm in length. The spatulate tip is also set oblique to the shaft. On the plastron are three
spatulate spines, none of which have their base preserved. Entire spines must be greater than 1-75
mm in length. On the oral surface there are two long and slender spines, one of which is broken
1 . /. excentricus (BMNH E40770): a, latero-ventral spine; b, latero-dorsal spine.
2. I. infulasteroides (BMNH E35766): a, sub-anal spine; b, latero-ventral spine; c, latero-dorsal spine;
d, apical spine; e, plastron spine; /, spine from the protective arch across the frontal sulcus;
g, anterior excavatory spine.
3. H. anterior (BMNH E40158): spine from the protective arch across the frontal sulcus.
4. H. rostrata (BMNH E76848): a, latero-ventral spine; b, anal spine; c, plastron spine.
34
PALAEONTOLOGY, VOLUME 25
just below the milled ring, the other having no remnant of the spine base. The proximal end lies
near the large tubercles of the sub-anal protruberance. Both spines are gently curved and are about
2-75 mm in length. The shaft becomes slightly broader distally, but does not appear to be spatulate.
In the anterior sulcus there are a number of non-spatulate spines. Most of those lying on the
floor of the sulcus are small, slender, and strongly curved. They are about T5 mm in length and
do not flatten distally. These are derived from the small interambulacral tubercules lining the outer
margin of the sulcus, judging from the size and shape of the spine base. Stouter spines, about 1-75
mm in length, are found within the sulcus and on the adjacent interambulacra. The shaft is
slightly bent a little above the milled ring, and tapers distally (the extreme tip is unfortunately not
preserved). These spines are associated with the large interambulacral tubercles adjacent to the
sulcus.
Plastron and latero-ventral spines of H. rostrata are similar to those of I. infulasteroides. In the
specimen of H. rostrata, a number of stout, straight spines, up to 2-5 mm in length, are preserved
within the periproct. Their tip is slightly flattened and is set slightly oblique to the shaft. These
spines are probably associated with the large interambulacral tubercles which surround the periproct. ,
The spines in all species of Infulaster and Hagenowia are likely to be similar, judging from the
similarity in tubercle structure and arrangement. The function of each group of spines is best
discussed in conjunction with the structure of their tubercles. A reconstruction of the spine coverage
in I. infulasteroides and Hagenowia is given in text-fig. 10. Spine posture is inferred from tubercle
structure.
Tubercle structure and arrangement
The shape and arrangement of tubercles can give information on spine posture and movement
(Smith 19806). There are a number of distinct areas of tubercles in Infulaster and Hagenowia, each
of which bore spines with a specific function. Tubercle arrangement in I. excentricus, H. rostrata,
and H. blackmorei is depicted in text-figs. 7 and 8. The following groups of tubercles can be
distinguished.
1. Plastron tubercles. Plastron tubercles in both Infulaster and Hagenowia are closely spaced
with only a few interspersed miliary tubercles. Largest tubercles lie laterally with smallest tubercles
lying along the central ridge. Their areole is slightly enlarged on the meso-posterior side (text-figs.
7, 8) and platform crenulation is radially symmetrical. As in Recent spatangoids, the power stroke |
would have been towards the mid posterior, and plastron spines, with their spatulate tips, must
have provided the main thrust for locomotion. However, plastron tubercles of Recent spatangoids
have asymmetric crenulation and pronounced areole enlargement (Smith 19806). The movement
of plastron spines was probably more restricted and less effective in Infulaster and Hagenowia
compared with Recent spatangoids. Radially symmetrical crenulation indicates that plastron spines
were held more or less perpendicular to the test (Smith 19806).
2. Latero-ventral tubercles. Tubercles found below the marginal fasciole in interambulacra 1-4
have their areole enlarged adambitally and towards the fasciole. Platform crenulation is either
radially symmetrical or is slightly larger on the adoral side (opposite the direction of areole
enlargement). These tubercles are denser on the anterior half but become more uniformly distributed
in later species. They decrease in size and become more densely packed towards the fasciole.
This arrangement differs from that found in Recent spatangoids, where the areole is enlarged
in a latero-posterior direction and crenulation is asymmetrical. (Smith 19806). In spatangoids the
associated spines lie oblique to the test and excavate sediment from beneath the animal with an
oar-like action. Tubercle structure, in Infulaster and Hagenowia, indicates that, whereas oral spines
were more or less perpendicular, adambital spines were inclined slightly downwards away from
the fasciole. Areole enlargement suggests that spines pushed sediment outwards and upwards and
were principally used in excavating and burrowing.
3. Sub-anal tubercles. Tubercles are radially arranged on each sub-anal bulge with large, radially
symmetrical tubercles centrally, and smaller tubercles marginally (text-figs. 7 and 8). Marginal
tubercles often have a slight areole enlargement on the side away from the centre. A similar tubercle
GALE AND SMITH: CRETACEOUS ECHINOIDS
35
EXCAVATORY
SPINES
SEDIMENT-MOVING
SPINES
SUB-ANAL TUFT
LOCOMOTORY SPINES
text-fig. 10. Spine posture and arrangement in Infulaster and Hagenowia. A, Infulaster infulasteroides,
lateral view, b, Hagenowia blackmorei, cut-away section through the rostrum showing the anterior sulcus.
The relative size of spines is based on specimen BMNH E35766, and spine posture and function is
interpreted from tubercle structure.
arrangement in Recent spatangoids is associated with tufts of spines (Smith 19806). Infulaster and
H. rostrata must have had two sub-anal tufts of spines whereas later species of Hagenowia had
only a single tuft. The upper spines of each tuft abut against the marginal fasciole. Mucus from
the fasciole could easily have been transferred to the sub-anal spines. The sub-anal tuft of spines
is used in constructing a tunnel, in spatangoids, with the aid of specialized tube-feet (Nichols 1959;
Chesher 1968). Infulaster and Hagenowia lacked sub-anal, funnel-building tube-feet and were
unlikely to have built an extensive tunnel. However, the spines alone may have been able to
maintain a short sub-anal tunnel with the help of mucus from the fasciole. In spatangoids the
sub-anal tunnel is used to remove water from the burrow (Chesher 1968).
4. Anal tubercles. Large tubercles are found around the dorsal and lateral edges of the periproct.
They decrease in size away from the periproct and are most numerous dorsally and on either side
immediately below the periproct. The crenulate platform is radially symmetrical. There may be
slight areole enlargement on the anterior side. Beneath the periproct is a broad band of miliaries,
extending to the fasciole. A similar arrangement of tubercles is found in Recent spatangoids, where
a semicircle of longer spines surrounds the periproct. These maintain a space within the burrow
for defaecation and prevent fouling of the aboral surface.
36
PALAEONTOLOGY, VOLUME 25
5. Latero-dorsal tubercles. Tubercles above the marginal fasciole have only a weak bilateral
symmetry in I. excentricus and I. tuberculatus. This becomes much more pronounced in Hagenowia
due to a marked enlargement of the areole on the anterior or adapical side (see text-fig. 8). Numerous
miliaries occur between the tubercles. These tubercles support short, spatulate spines in /.
infulasteroides. Similar spatulate spines, attaching to bilaterally symmetrical tubercles, are found
in Recent spatangoids, where they are used for compacting the burrow walls, transporting sediment
posteriorly, and moving the dorsal mucous coat (Smith 19806). I. excentricus has a low aboral
tubercle density but later species of Infulaster and Hagenowia have progressively denser tubercula-
tion. The increased protection provided by a denser spine coverage is an adaptation for living
infaunally. Tubercles are especially dense along the rostrum, presumably for compacting and
maintaining an open passageway down into the burrow.
In H. blackmorei, tubercles just above the marginal fasciole have an adoral areole enlargement
indicating that the spatulate tips of spines were tilted in opposite directions on either side of the
fasciole, a situation common in Recent spatangoids for distributing mucus.
6. Tubercles associated with the anterior sulcus. There are two sets of interambulacral tubercles
lying adjacent to the anterior sulcus. Firstly, there are small, radially symmetrical tubercles which
form an inward-facing row along the lip of the sulcus (PI. 4, figs. 5-9; PI. 5, figs. 1-4). They
are rather irregularly arranged in early species but become organized into a well-defined line in
Hagenowia. The associated curved and non-spatulate spines would have formed an arched roofing
to the anterior sulcus (text-fig. 10b). The increasing density and organization of these spines was
partially due to the decreasing depth of the anterior sulcus and partially due to the increasing
importance of the sulcus as a pathway for transporting sediment to the mouth. In Hagenowia the
anterior shift of the peristome shows that more reliance was placed on sediment coming down the
sulcus. The arch of spines would have maintained a clear frontal pathway and would have prevented
sediment from the frontal wall of the burrow clogging or contaminating the stream of sediment
passing down the anterior sulcus. They ensured that sediment could only enter the sulcus at the apex.
EXPLANATION OF PLATE 6
Fig. 1. Hagenowia elongata: large interambulacral tubercles adjacent to the rostral sulcus that support
excavating spines. Crenulation is radially symmetrical but the areole is enlarged laterally away from
ambulacrum III. Smaller tubercles for protective arch spines can be seen towards the bottom. MM 12820.
Fig. 2. H. anterior, interambulacral tubercles, as above but with intervening miliaries. BMNH E76844.
Fig. 3. H. elongata\ row of small interambulacral tubercles immediately adjacent to the frontal sulcus (bottom of
micrograph) for spines forming a protective arch to the sulcus. MM 12820.
Fig. 4. H. elongata: enlargement of fig. 3 showing the radially symmetrical areole and the lack of crenulation.
Fig. 5. Infulaster tuberculatus : large apical interambulacral tubercles, ambulacrum III to the right. These have
radially symmetrical areoles and crenulation and are not tilted. Their spines must have been held pointing
away from ambulacrum III. BMNH E76852.
Fig. 6. H. anterior: miliary tubercles near the apex of the rostral sulcus, apex to the sulcus. These are tilted
adapically and laterally, away from the mid-line, suggesting that their spines had a meso-adoral power stroke.
BMNH E76844.
Fig. 7. H. blackmorei: adapical part of ambulacrum III showing dense, irregular arrangement of varyingly sized
tubercles at the top of the sulcus. BMNH E76846.
Fig. 8. H. anterior: large ambulacral tubercle adjacent to an apical ambulacrum III unipore. Crenulation is
slightly stronger laterally (towards the pore) and the protective spine must have been gently curved. BMNH
E76844.
Fig. 9. H. blackmorei: large apical interambulacral tubercle with a radially symmetrical areole and slightly
stronger crenulation towards ambulacrum III (top of micrograph). BMNH E76846.
Fig. 10. H. elongata: large apical interambulacral tubercles. The areole is slightly border to the posterior
(bottom of micrograph), crenulation is radially symmetrical. MM 12820.
Scale bar in figs. 4, 8 = 40 ^m; figs. 1-3, 6, 9, 10 = 100 /urn; figs. 5, 7 = 200 /u m .
PLATE 6
GALE and SMITH, Cretaceous irregular echinoids
38
PALAEONTOLOGY, VOLUME 25
The other set of interambulacral tubercles are large and outward-facing (PI. 5, figs. 5-9; PI. 6,
figs. 1-3). They form a single, well-defined row in Hagenowia. Platform crenulation is either radially
symmetrical or is enlarged slightly on the side facing the sulcus. Their areole is enlarged in the
opposite direction (text-fig. 8; PI. 6, fig. 1) showing that the power stroke of the spine was directed
away from the sulcus. These stout, pointed spines were probably used to loosen and excavate
sediment of the anterior wall of the burrow. Recent spatangoids, such as Echinocardium cordatum
(Pennant), have both an anterior arch of protective spines and adjacent excavating spines, and
their tubercles and spines are rather similar.
The apical interambulacral tubercles and spines of Infulaster and Hagenowia are the largest they
possess. There are between 16 and 20 such tubercles on the adapical plates of the anterior
interambulacra. Their areoles are either radially symmetrical or are enlarged slightly to the posterior
(PI. 5, figs. 1-3; PI. 6, figs. 5, 9, 10). Platform crenulation is more or less radially symmetrical
though crenulation may be slightly enlarged on the anterior side (PI. 6, fig. 9). Tubercles are so
positioned that, were the spines to converge adapically to form a tuft, they would have to lie at
a considerable angle to the tubercles and the plate surface. The structure of the tubercles, however,
shows that spines were not set obliquely, forming an apical tuft, but probably formed a fan-shaped
array. This fan of spines would have formed a collar to the anterior sulcus (text-fig. 10a). This
arrangement is common to all species of Infulaster and Hagenowia and is also found in certain
species of Cardiaster. As the apex is the most vulnerable part of the test in infaunal or semi-infaunal
species, these apical spines may have acted as a deterrent to predators. The fan-shaped arrangement
around the apical opening to the frontal sulcus suggests that apical spines were also involved in
collecting sediment, either acting as a funnel or helping to cascade surface sediment into the sulcus.
Within ambulacrum III, small tubercles are interspersed with miliaries near the apex (PI. 6,
fig. 7), but in the sulcus ambulacral plates are covered almost entirely by uniformly sized miliaries
(PI. 6, fig. 6). A similar arrangement is found in Micr aster but has not yet been reported in any
Recent species (Smith, 19806). In many Recent spatangoids, miliary tubercles of ambulacrum III
bear short, straight spines with fleshy tips, usually well endowed with cilia and mucous glands. A
similar type of spine was probably present in the sulcus of Infulaster and Hagenowia.
Several lines of evidence suggest that the anterior sulcus became increasingly important as a
pathway for transferring sediment to the mouth. Evolutionary changes in the size and density of
miliaries in the sulcus may therefore be related to an improved feeding mechanism. The progressive
differentiation of a distinct band of dense miliaries at the apex of the sulcus may mark the increasing
importance of ciliary currents or mucus binding for particle transportation. By concentrating
miliary spines at the apex, where sediment entered the sulcus, it is possible that copious amounts
of mucus could have been produced to enmesh particles loosely. The relatively few spines within
the sulcus could then have transported a stream of mucus-bound sediment with relative ease.
Tubercle and spine arrangement is highly distinctive in Recent, mucous string feeding spatangoids
(Chesher 1963; Buchanan 1966; Smith 19806) and very different from that in Hagenowia. Hagenowia ,
then, did not feed by means of a compact mucous string but may have had a much looser flow
of weakly bound particles.
7. Fascioles. The marginal fasciole shows little change except that it becomes more sharply
defined across the anterior interambulacra in later species of Hagenowia. The increasing obliquity
of the fasciole is a direct consequence of the changing position of the peristome. The fasciole does
not cross the anterior ambulacrum, as stated by Nichols (1959), but becomes more diffuse and
peters out towards the sulcus.
Fascioles act as pumps, drawing oxygenated water through the burrow, and as a source of
mucus. A dorsal mucous coat prevents finer particles falling between spines and clogging the
burrow. Both Infulaster and Hagenowia are likely to have lived at least partially buried and the
fasciole probably drew water down from the apex and over the central and posterior parts of
the test.
GALE AND SMITH: CRETACEOUS ECHINOIDS
39
PRESERVATION
Increased differentiation and strengthening of the rostrum in successive species of Hagenowia was
accompanied by relative thinning of the body wall. This is reflected in the increased rarity of
preservation of Hagenowia bodies in successively later species and the greater frequency of isolated
rostra.
Bodies of Hagenowia are rarely bored or encrusted by the organisms which colonized much of
the skeletal debris littering the chalk sea floor, and it is likely that once exhumed, the fragile tests
fragmented rapidly leaving only the more durable rostra. The rostra often contain sponge crypts,
and characteristic oval borings (500-1000 /xm in length) with irregularly bevelled rims. Encrusters
are extremely rare on H. rostrata, probably because they were too small to provide suitable substrata.
One preservational style occurs consistently in Hagenowia, in which the rostrum is lost, having
broken away along plate sutures near its base (e.g. PI. 3, fig. 2). The resulting irregular stump has,
in some specimens been attacked by boring organisms (e.g. PI. 3, figs. 6b, la). Such specimens
frequently retain scattered spines on the sides and base of the body. Presumably these individuals
remained in life orientation within the sediment after death, with the rostrum protruding above
the sediment-water interface, which was stable for long enough to allow colonization by borers.
Two specimens of Hagenowia have repaired injury to the rostrum. In one specimen (BMNH
E76849) the obliquely broken tip of the rostrum was sealed across, and new tubercles and genital
pores developed. Such damage was probably the result of predation.
SELECTION PRESSURES AND MODE OF LIFE
The majority of structural changes in the Infulaster I Hagenowia lineage were caused, either directly
or indirectly, by elongation of the apical region of the test, and to a lesser extent by the reduction
in size. Apical elongation necessitated disruption of the apical system, plate elongation, loss of
plate columns from the rostrum, and an increased slenderness of the rostrum. This increased
slenderness in turn reduced the internal volume of the rostrum, which caused a decrease in depth
of the sulcus, loss of two of the four gonads, and the change from extensible tube-feet with large
ampullae, to small tube-feet with reduced ampullae. The slender rostrum required buttressing at
the anterior and posterior margins, and an increasingly better organization of tubercles. The increase
in obliquity of the fasciole was a consequence of the anterior and adapical movement of the
peristome. The size reduction resulted in the change from a double to a single subanal bulge and
in the interruption of the plastron.
Various features displayed by Infulaster and Hagenowia are interpreted as adaptations to an
infaunal mode of life. The latero-ventral spines were specialized for excavating sediment from
beneath the animal. In order for the fasciole and subanal tuft of spines to have functioned, these
echinoids must have lived at least partially buried. Latero-dorsal spines modified for compacting
sediment of the burrow walls are developed to immediately below the apex, as are the excavating
spines and the protective arch of spines across the sulcus (text-fig. 10). This clearly implies that
both Infulaster and Hagenowia lived within the sediment, with their apices at or just below the
surface.
It is highly improbable that either genus could have burrowed more deeply, as funnel-building
tube-feet are lacking, and the apical spines were held in a broad fan and not a tuft. Living within
fine-grained sediments, they had to maintain direct contact with surface waters. This could only
have been achieved by keeping the apex of the test at the surface.
I. excentricus has an almost flat dorsal surface, most of which must have remained exposed on
the sea floor. Later species evolved to become less conspicuous, by apical elongation. In this way
the bulk of the animal could be removed from the surface, with only a small apical part of the
test visible. This apical elongation brought about the structural changes enumerated above. The
loss of aboral respiratory tube-feet was probably an important factor in influencing the size reduction
which occurred.
40
PALAEONTOLOGY, VOLUME 25
Other morphological changes accompany a change in feeding strategy. Infulaster used its phyllode
tube-feet to pick up sediment around the mouth. Sediment coming down the frontal sulcus was
not collected directly and probably formed only a small part of the animal’s diet. Hagenowia came
to rely much more heavily on sediment from the anterior sulcus and may have developed some
sort of mucus binding to cope with this increased volume of sediment. Adaptations linked with
this change in feeding strategy include the anterior movement of the peristome, loss of large
phyllode tube-feet, and increasingly denser arch of spines across the sulcus and their extension to
cover the peristome, the more extensible sensory tube-feet at the apex, and the increasing
differentiation of miliaries within the sulcus.
All the morphological trends and specializations discussed are directly or indirectly concerned
with adaptations for feeding or burial. The predominant pressures acting on the Infulaster -
Hagenowia lineage were for increased efficiency in feeding and for increased protection by becoming
less obvious at the surface. I. excentricus lived semi-infaunally with much of the dorsal surface
exposed, and progressed leaving a clear furrow in its wake (text-fig. 1 1). In later species, as the
dorsal surface became more oblique with apical elongation, less and less of the test remained
exposed at the surface. In H. blackmorei only the very small apical head remained near the surface,
the remainder of the body being fully covered (text-fig. 1 1). Only a small circular hole at the surface
would have marked the position of this species.
It is not entirely correct to speak of an increasing depth of burial in the Infulaster-Hagenowia
lineage, since the apex stays at the same level and the oral surface of I. excentricus lay at considerably
greater depth than in many later species, because of the reduction in size. However, with apical
elongation and the more vertical orientation of the rostrum, the oral surface was removed further
from the sea floor. The most likely cause of this change is increasing predation, with selection
text-fig. 1 1 . Mode of life of Infulaster and Hagenowia. Cut-away block sections showing the depth of burial of
Infulaster excentricus (left-hand side) and Hagenowia blackmorei (right-hand side). Further back a second
individual is illustrated to highlight the change in surface appearance with the development of apical elongation.
GALE AND SMITH: CRETACEOUS ECHINOIDS
41
against those more obvious at the surface. Predation was a threat to Hagenowia, since regenerated
rostra are known.
It is not clear whether the change in feeding habit resulted from a change in the nature of the
sediment (e.g. the development of a surface layer richer in organic material or, conversely, an
over-all reduction in organic content) or reflected a progressive adaptation to a constant sediment
type. Infulaster and Hagenowia are both found in bands in apparently uniform pelagic chalk
sequences. However, original subtle differences in the nature of the sediment may well have been
obscured by bioturbation and diagenesis.
Hagenowia provides some indirect evidence as to the nature of the chalk sea floor. The presence
of cutting and compacting spines on the rostrum implies that the sediment in which it lived was
sufficiently cohesive to both require and allow such treatment. This is in accord with the suggestion
made from other lines of evidence by Surlyk and Birkelund (1977) that the bottom was relatively
stable, with only a superficial layer (a few millimetres thick) of easily resuspendible material.
Acknowledgements. R. P. S. Jefferies (Natural History Museum, London), C. J. Wood (Geological Museum,
London), S. Floris (Geological Museum, Copenhagen), and C. W. Wright kindly made specimens in their
care available for study. A. S. G. would like to thank J. M. Hancock, K. Holdaway, C. W. Wright, and U.
Asgaard for valuable discussion and criticism. P. Howard (King’s College London) helped with the
photography. Part of this work was carried out with the support of an S.R.C. grant.
REFERENCES
brydone, r. m. 1914. The Zone of Off aster pilula in the South English Chalk. Part 1 -4. Geol. Mag. (6) 1, 395-469,
405-11, 449-57, 509-73.
buchanan, i. b. 1966. The biology of Echinocardium cor datum (Echinodermata: Spatangoida) from different
habitats. J. mar. biol. Ass. U.K., 46, 97-114.
chesher, R. H. 1963. The morphology and function of the frontal ambulacrum of Moira atropos (Spatangoida).
Bull mar. Sci. Gulf Carrib. 13, 549-573.
— 1968. The systematics of sympatric species in West Indian spatangoids: a revision of the genera Brissopsis,
Plethotaenia, Paleopneustes and Saviniaster. Stud. trop. Oceanogr. Miami , 7, viii+168 pp.
— 1969. Contributions to the biology of Meoma ventricosa (Echinoidea: Spatangoida). Bull. mar. Sci. 19,
72-110.
desor, e. 1858. Synopsis des Echinides Fossiles, vi, 321-490. Paris.
devries, a. 1953. Contribution a l’etude de la symetrie bilaterale chez les Spatangues. Bull. Serv. Carte geol.
Algerie (1) Paleont, 15, 37-69.
duncan, p. M. 1889. A revision of the genera and great groups of the Echinoidea. J. Linn. Soc. 23, 1-311.
ernst, G. and schulz, M.-G. 1971. Die Entwicklungsgeschichte der hochspezialisierten Echiniden-Reihe
Infulaster— Hagenowia in der Borealen Oberkreide. Palaont. Z. 45, 120-143.
— 1974. Stratigraphie und Fauna des Coniac und Santon im Schreibkreide-Richtprofil von Lagerdorf
(Holstein). Mitt. geol. — palaont. Inst. Univ. Hamburg, 43, 5-60.
ernst, G. and seibertz, E. 1977. Concepts and methods of echinoid biostratigraphy. Pp. 541-63. In Kauffman, E.
G., and Hazel, J. E. (eds.). Concepts and methods in biostratigraphy. Dowden, Hutchinson & Ross, Inc.,
Stroudsburg.
forbes, e. 1852. Figures and descriptions illustrative of British organic remains. Mem. geol. Surv. U.K., Dec. iv,
1-4.
gaster, c. t. a. 1924. The Chalk of the Worthing District, Sussex. Proc. Geol. Tss. 35, 89-110.
mortensen, t. 1950. A monograph of the Echinoidea V. I. Spatangoida 1. Copenhagen.
moskveena, m. m., 1959. Atlas of the Upper Cretaceous fauna of the northern Caucasus and Crimea. Trudy ,
Glav. Uprav. Gaz. Promyshal. Sov. Min. USSR. (VNIGAZ), 310 pp. [In Russian.]
nichols, d. 1959. Changes in the Chalk heart-urchin Micraster interpreted in relation to living forms. Phil.
Trans. R. Soc., Lond. B, 242, 347-437.
nielsen, k. b. 1942. Martinosigra elongata n.g. et n. sp., A new echinoid from the White Chalk of Denmark.
Meddr Dansk Geol. Foren. 10, 2, pp. 159-166.
nietsch, H. 1921. Die irregularen Echiniden der pommerschen Kreide. Abh. geol. — palaont. Inst. Greifswald, 2,
1-47.
42
PALAEONTOLOGY, VOLUME 25
peake, N. b. and Hancock, J. M. 1961. The Upper Cretaceous of Norfolk. Pp. 293-339. In Larwood, G. P. and
Funnell, B. M. (eds.). The geology of Norfolk. Trans. Norfolk Norwich Nat. Soc. 19, (6), pp. 269-375.
rawson, p. f. et al. 1978. A correlation of Cretaceous rocks in the British Isles. Geol. Soc. Lond., Special Report
No. 9, 70 pp.
reid, r. e. h. 1976. Late Cretaceous climatic trends, faunas and hydrography in Britain and Ireland. Geol. Mag.
113, 115-128.
rowe, a. w. 1 899. An analysis of the genus Micraster, as determined by rigid zonal collecting from the Zone of
Rhynchonella Cuvieri to that of Micraster coranguinum. Q. Jl geol. Soc. Lond. 55, 494-547, pis. 35-39.
— 1904. The White Chalk of the English coast: Part IV. Yorkshire. Proc. Geol. Ass. 18, 193-296.
schmid, f. 1972. Hagenowia elongata (Nielsen), ein hochspezialisierter Echinide aus dem hoheren
Untermaastricht NW-Deutschlands. Geol. Jber. 4, 177-195.
smith, a. b. 1980a. The structure, function and evolution of ambulacral tube feet and pores of irregular
echinoids. Palaeontology , 23 (1), 39-83.
— 19806. The structure and arrangement of echinoid tubercles. Phil. Trans. R. Soc., Lond. B, 289, 1-54.
surlyk, f. and birkelund, t. 1977. An integrated stratigraphical 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 of biostratigraphy. Dowden, Hutchinson & Ross Inc., Stroudsburg.
wright, t. w. 1881. Monograph of the British Fossil Echinodermata from the Cretaceous Formations. Vol. 1.
The Echinoidea. Part 9, pp. 301-324, Palaeontogr. Soc. ( Monogr .)
wright, c. w. and wright, e. v. 1942. The Chalk of the Yorkshire Wolds. Proc. Geol. Ass. 53, 112-127.
1949. The Cretaceous echinoid genera Infulaster Desor and Hagenowia Duncan. Ann. Mag. nat. Hist.
(12) 2, 454-474.
ANDREW S. GALE
Department of Geology
King’s College, London
Strand
London WC2R 2LS
ANDREW B. SMITH
Department of Geology
University of Liverpool
P.O. Box 147
Liverpool L69 38X
Typescript received 28 January 1980
A REVISION OF LATE ORDOVICIAN BIVALVES
FROM POMEROY, CO. TYRONE, IRELAND
by S. P. TUNNICLIFF
Abstract. A history of research on Ordovician bivalves from Pomeroy is given, and their stratigraphic
distribution in the Caradoc-Ashgill rocks is outlined. More than thirty taxa are described including five new
species: Praenucula dispersa, P. infirma, P. praetermissa, Ambonychia arundinea, and Cleionychia incognita.
The rostroconch Hippocardia praepristis (Reed) is illustrated.
Since Portlock (1837, 1843) first described the faunas of the Caradoc-Ashgill rocks of the Pomeroy
area of County Tyrone (text-fig. 1), some elements have been studied thoroughly, especially the
trilobites and brachiopods, which have helped to show the relationship of the Pomeroy strata to
the Ordovician of North America and southern Scotland (Mitchell, 1977, and others). Although
molluscs are abundant in the rocks of Pomeroy, they have received scant attention in recent years.
The orthoconic nautiloids were redescribed by Blake (1882) but have remained unrevised since;
the gastropods have been described by Longstaff ( = Donald) (1902, 1924) and Reed (1952). The
bivalves were largely ignored until Reed (1952) redescribed Portlock’s species and added a number
of new taxa. Reed was unaware of Portlock’s many duplicate specimens, including syntypes, now
housed in the Ulster Museum (Tunnicliff 1980), and did not have the benefit of collections made
in recent years by the Geological Survey of Northern Ireland, by Dr. W. I. Mitchell and by Mr.
R. P. Tripp and others of the British Museum (Natural History). With the aid of this new material
a critical re-examination of the bivalve fauna has been carried out and several new forms have
been recognized.
History of Pomeroy bivalve research
Portlock (1843) mentioned or described in varying detail the following species, arranged here according to
the formation from which they can now be recognized to have come.
Bardahessiagh Formation (Caradoc):
Avicula obliqua Murchison
Avicula orbicularis Murchison
Modiola Brycei Portlock
Modiola expansa Portlock
Modiola Nerei (Munster)
Modiola securiformis Portlock
Inoceramus priscus Portlock
Inoceramus transversus Portlock
Inoceramus trigonus (Munster)
Nucula radiata Portlock
Killey Bridge Formation (Ashgill, Cautleyan):
Inoceramus contortus Portlock
Cypricardia simplex Portlock
Area cylindrica Portlock
Area dissimilis Portlock
Area Eastnori (Murchison)
Area obliqua Portlock
Area regularis Portlock
Area subtruncata Portlock
Area transversa Portlock
Pectunculus ambiguus Portlock
Pectunculus Apjohnni Portlock
Pectunculus semitruncatus Portlock
Nucula acutal (Sowerby) var. imbricata Portlock
Uncites gryphus Schlotheim var.
In addition to these, Portlock had a number of specimens to which he did not refer in print but to which
he fixed labels with suggested identifications and observations. Notes on how Portlock labelled his specimens
are given by Tunnicliff (1980). His Mytilus cinctus Portlock and Posidonomyal venusta (Munster) are from
rocks of Llandovery age as is his Mytilus? semi-rugatus Portlock from Lisbellaw, Co. Fermanagh. Portlock’s
[Palaeontology, Vol. 25, Part 1, 1982, pp. 43-88, pis. 7-13.]
44
PALAEONTOLOGY, VOLUME 25
text-fig. 1. The area around Pomeroy, Co. Tyrone, based on Mitchell 1977 (taken from Tunnicliff 1980), with
Ireland inset to show the position of Pomeroy.
figures, although generally of natural size, often lack the detail which might be expected in modern illustrations.
Many of his descriptions contain useful detail such as height and length of valves and the number of teeth
visible, but others are less helpful.
Describing specimens from the collection of Sir R. Griffith, McCoy (1846) added the following species
from the Bardahessiagh Formation, Pullastra speciosa McCoy and from the Killey Bridge Formation, Nucula
protei Munster, N. subacuta McCoy, Area quadrata McCoy.
In 1878 (p. 28) Baily published a list of Portlock bivalve species giving amended generic names and gave
a list (p. 25) of the species collected by officers of the Geological Survey of Ireland during the survey of the
1870s.
Apart from some brief discussion by Hind (1910), Baily’s was the last contribution to the study of Pomeroy
Ordovician bivalves before the publication of Reed’s posthumous paper in 1952. In this Reed redescribed
most of Portlock’s species and added the following:
Ctenodonta perangulata Reed
Ctenodonta deserta Reed
Ctenodonta cf. nitida (Ulrich)
Clidophorus occultus Reed
Vanuxemia? suspecta Reed
Modiolopsis concentrica simulans Reed
Orthodesma tyronense Reed
Conocardium praepristis Reed
Paramodiolal sp.
Whiteavesia subexpansa Reed
Ambonychinia cf. volvens Isberg
Ambonychinia cf. amygdalina (Hall)
Ambonychinia cf. intermedia Isberg
Clionychia subovalis Reed
Clionychia subquadrata Reed
TUNNICLIFF: LATE ORDOVICIAN BIVALVES
45
Horizons and localities (Text-figs. 1 and 2)
The Ordovician stratigraphy of Pomeroy was discussed by Mitchell (1977, pp. 4-18). Bivalves are recorded
from the following strata:
Tirnaskea Formation (Ashgill, Hirnantian)
Killey Bridge Formation (Ashgill, Cautleyan)
Bardahessiagh Formation (Caradoc).
No bivalves are recognized as being from Mitchell’s Junction beds.
The horizon of Portlock specimens and other material from old collections is inferred from the lithologies
and from the known locality information (see Tunnicliff 1980).
The localities of many of the specimens in the old collections are uncertain, often given only as ‘Pomeroy’.
However, by comparison with material in the recent well-localized collections and with the small number of
likely sites, it is possible to suggest localities for most specimens, usually with reference to those listed by
Mitchell (1977, pp. 5-7). Portlock’s localities are discussed by Tunnicliff (1980).
Stratigraphic distribution (Table 1)
The bivalve fauna of the Bardahessiagh Formation is dominated by cyrtodontids and ambonychiids, the
former probably infaunal or semi-infaunal, the latter epifaunal and byssate. Cyrtodontal often occurs as
conjoined valves suggestive of little or no transport after death. The other infaunal genera represented in the
Bardahessiagh Formation are Praenucula, Deceptrix, Similodontal , and Lyrodesma all of which are uncommon
compared with Cyrtodontal . Similodontal and Lyrodesma both occur as conjoined or closely associated valves
Series Graptolite
& Stage zone
ASHGILL
Hirnantian
Rawtheyan
Cautleyan
D.
anceps
Pusgillian
CARADOC
Onnian
Actonian
Marshbrookian
D. complanatus
P. linearis
Woolstonian D. clingani
Longvi Ilian
Soudleyan C. wilsoni
Harnagian C. peltijer
Costonian
N. gracilis
Pomeroy Girvan Bohemia North
America
Tirnaskea
Formation
High Mains
Kosov
Formation
Killey
Bridge
Formation
Drummuck
Group
Kraluv d5
Dvur
Formation
Richmondian
Shalloch
Formation
Maysvillian
Whitehouse
Group Edenian
_ _ Bohdalec
Formation
Junction Ardwell
Beds Group
Bardahessiagh
Formation
Balclatchie
Zahorany F. y
Vinice F. ,33
Letna ^2
Formation
Shermanian
Wildernessian
text-fig. 2. Approximate correlation of the Ordovician rocks of Pomeroy based on Williams et al. (1972),
Mitchell (1977), and R. P. Tripp and J. K. Ingham (pers. comm. 1981). The approximate stratigraphic
equivalents of Barrande’s stages d2, d3, d4, and d5 in Bohemia are shown (based on Krtz and Pojeta 1974).
46
PALAEONTOLOGY, VOLUME 25
table 1. The stratigraphic distribution of bivalves from the Ordovician rocks of Pomeroy, Co. Tyrone
Bardahessiagh
Formation
Killey Bridge
Formation
Tirnaskea
Formation
Praenucula dispersa sp. nov.
X
P. praetermissa sp. nov.
X
P. aff. praetermissa sp. nov.
X
P. infirma sp. nov.
X
Deceptrix sp.
X
D. apjohnni (Portlock)
X
D. regular is (Portlock)
X
D. semitruncata (Portlock)
X
D. subtruncata (Portlock)
X
Similodonta? sp.
X
Concavodonta imbricata (Portlock)
X
Nuculites cylindricus (Portlock)
X
Nuculoid gen. and sp. nov.
X
Cyrtodonta? expansa (Portlock)
X
C.? securiformis (Portlock)
X
C.7 sp.
X
Vanuxemia contorta (Portlock)
X
Ambonychia arundinea sp. nov.
X
Cleionychia transversa (Portlock)
X
C. prisca (Portlock)
X
C. incognita sp. nov.
X
Ambonychiopsis suspecta (Reed)
X
pterineid? gen. and sp. indet.
X
? bivalve gen. and sp. indet.
X
Modiolopsis sp.
X
Corallidomus concentrica (Hall and Whitfield)
X
Corallidomus? sp.
X
Goniophora sp.
X
Colpomya simplex (Portlock)
X
Semicorallidomus? sp.
X
Cycloconcha? speciosa (McCoy)
X
Lyrodesma radiatum (Portlock)
X
Hippocardia praepristis (Reed)
?
but, like Cyrtodonta?, are never suggestive of life position but rather of having been winnowed out from the
substrate. Of the epifaunal ambonychiids, Cleionychia is best represented, and some also occur as conjoined
valves. The nature of the present-day Bardahessiagh Formation exposure (Mitchell 1977, p. 5) precludes
identifying the exact localities for most specimens.
In the Killey Bridge Formation the principal elements of the bivalve fauna are infaunal nuculoids, Praenucula,
Deceptrix, Concavodonta, and Nuculites. These are all numerous compared with other genera, most of which
are represented in the collections by only one or two specimens. Most specimens from the Killey Bridge
Formation are single valves, but conjoined valves of all nuculoid genera are present. Bivalves are recorded
from all the localities where the Killey Bridge Formation is represented by dark-grey calcareous mudstone
(e.g. localities 1-3 of Mitchell 1977) but determinable bivalves are not known from the rottenstone and
sandstone lens recorded by Mitchell (localities 4a, b of Mitchell 1977) in which brachiopods are abundant
and varied.
Determinable bivalves are also almost unknown from the Tirnaskea Formation; only Nuculoid gen. nov.
and the rostroconch Hippocardia (by Reed 1952) are recorded and the latter is more probably from the Killey
Bridge Formation. Mitchell (1977, p. 17) compared the Tirnaskea fauna with the Hirnantian Hirnantia fauna.
TUNNICLIFF: LATE ORDOVICIAN BIVALVES
47
The differences between the Bardahessiagh and Killey Bridge bivalve faunas are a reflection of their differing
palaeoenvironments; the generally sandy Bardahessiagh Formation probably represents a shallow-shelf
environment favouring the modes of life of cyrtodontids and ambonychiids, while the mudstones of the Killey
Bridge Formation were probably the product of a more stable deeper-shelf environment. The Bardahessiagh
Formation bivalve fauna has a distinctly eastern American aspect which is less obvious, but still present in
the Killey Bridge Formation.
SYSTEMATIC PALAEONTOLOGY
In general, the classification used in the Treatise of Invertebrate Palaeontology, N1 and 2 (Moore 1969) is
adopted. Where this is deviated from is stated in this text, and the author whose precedent is followed is given.
Measurements. In some cases tables of measurements are given in which the terms mean and median are
used. Mean is used to show the sum of the items divided by the number of items; median is used to convey
the mid-point between the maximum and minimum values of a group of items. All linear measurements are
given in millimetres (mm), measured with a vernier caliper or under a binocular microscope with a graticule.
Abbreviations used are L for length, H for height, AL for preumbonal length. In tables of measurements,
the term ‘angle’ is used for the angle between the anterior and posterior hinge-plates in nuculoids.
Preservation. Two terms used require a note of explanation. Composite mould is used in the sense of McAlester
1962. External cast is used to describe a naturally occurring convex specimen which shows only external
features but which has probably been produced in a similar manner to composite moulds.
This study is based on material in the following museums: Institute of Geological Sciences (IGS), British
Museum (Natural History) (BM), Sedgwick Museum, Cambridge (SM), Ulster Museum, Belfast (UM),
National Museum of Ireland, Dublin (NMI).
Class bivalvia Linnaeus, 1758
Subclass palaeotaxodonta Korobkov, 1954
Order nuculoidea Dali, 1 889
Superfamily nuculacea Gray, 1824
Family praenuculidae Pfab, 1934
Remarks. Three genera of Praenuculids are certainly recognized from Pomeroy: Praenucula and
Deceptrix from the Bardahessiagh and Killey Bridge Formations and Concavodonta from the Killey
Bridge Formation. Concavodonta is distinct from the other two genera (see discussion of
Concavodonta) but Praenucula and Deceptrix are less easily distinguished at first sight. Pojeta (1971 ,
p. 16) implied that they were synonymous (‘. . . some authors recognize Praenucula Pfab as being
distinct from Deceptrix. . .’). Pojeta, Kriz, and Berdan (1976) treated Praenucula as a subgenus of
Deceptrix. McAlester (1969, p. 229), placed Pfab’s Praeleda in synonomy with Deceptrix and both
have been used for species which show the characters of Praenucula , for example Praeleda ciae
(Sharpe) of Bradshaw 1970, and Deceptrix cf. ciae (Sharpe) of Babin and Robardet 1973. Praenucula
and Deceptrix have remained virtually unused in the British Ordovician and I can find no reference
to their occurrence in the British Isles, except Praenucula from the Arenig of Ramsey Island (Carter
1971, p. 251). Specimens from British localities which could now be placed in Praenucula or
Deceptrix have usually been recorded as Ctenodonta, often under open nomenclature.
In the species discussed here, distinction is made between Praenucula and Deceptrix principally
on the differences indicated by McAlester (1969, p. 229); that is that the posterior teeth in Deceptrix
are smaller and more numerous than the anterior teeth, while in Praenucula the anterior and
posterior teeth are similar in size and number. To this we may add that the umbones in Praenucula
lie in the posterior half while in Deceptrix they generally lie in the anterior half. Corresponding
to the greater posterior length in Deceptrix is the relatively greater length (equal to or longer than
the anterior) of its posterior hinge-plate. In Praenucula the posterior hinge-plate is equal to or
shorter in length than the anterior hinge-plate. In both genera the anterior hinge-plate is arched
towards the body of the shell and the posterior hinge-plate is straight or gently arched away from
I the body of the shell. The adductor muscle scars in both genera are equal or subequal but their
PALAEONTOLOGY, VOLUME 25
relative sizes and positions differ, being larger and more ventral in Deceptrix (Pis. 7-9). In Praenucula
the adductor muscle scars are close to, and may extend beyond, the ends of the hinge-plates, while
in Deceptrix the posterior scar lies closer to the umbo, only extending as far as the end of the
hinge-plate, and the anterior scar lies in the most anterior part of the shell, beyond the end of the
hinge-plate.
A fourth genus, Similodonta, may be represented by specimens from the Bardahessiagh
Formation.
Genus praenucula Pfab, 1934
Type species. By original designation of Pfab 1934, pp. 234-235, Praenucula dispar expansa Pfab, 1934, from
the Sarka Formation (Llanvirn) of Bohemia (Havlicek and Vanek 1966) and described and figured by
McAlester (1968, p. 46, pi. 8, figs. 3-9). See discussion under Praenuculidae above.
Praenucula dispersa sp. nov.
Plate 7, figs. 1-5; text-fig. 3 a
Derivation of name. From Latin dispersus scattered, referring to the occurrence in loose blocks.
Type specimens. Holotype IGS GU 1893, internal mould of right valve, and paratypes IGS GU 1354, 1894,
1897, 1906, 3265, UM K4203, SM A 16296, 3 left and 4 right valves; one a composite mould, the remainder
internal moulds.
Horizon and locality. All specimens are from the Bardahessiagh Formation; GU specimens from the area of
Mitchell’s Bardahessiagh Formation collecting (Mitchell 1977), the others from uncertain localities to the
south of Craigbardahessiagh.
Measurements
L
H
H/L
AL/L
Angle
Max.
13-2 mm
110 mm
0-87
0-65
155°
Min.
7-8 mm
6-8 mm
0 61
0-50
130°
Mean
10-62 mm
8-15 mm
0-77
0-50
145°
Median
10-5 mm
8-9 mm
0-74
0-58
142-5‘
Description. Ovate Praenucula with height about 0-7 of the length but variable. Umbo slightly posterior of
centre, between the posterior two-fifths and mid-point, and extending slightly above the hinge-line. Convexity
greatest slightly anterior to the umbo, maximum inflation of a single valve 2-4 mm seen in a valve 1 1 mm
a
b
c
text-fig. 3. Praenucula spp. based on internal moulds, c. x 5. ( a ) P. dispersa sp. nov., IGS GU 1906, right
valve. ( b ) P. infirma sp. nov., IGS NIL 8935, right valve, (c) P. praetermissa sp. nov., IGS NIL 8757, latex of
left valve. Abbreviations: aa— anterior adductor muscle scar, pa— posterior adductor muscle scar, a— accessory
muscle scar. Compare with PI. 7, figs. 5, 12, 21. Note the differences in size of the adductor muscle scars and the
differences in the size of the teeth and in their distribution on either side of the umbo. In each case, the finest teeth
below the umbo are too small to be represented clearly.
TUNNICLIFF: LATE ORDOVICIAN BIVALVES
49
long and 8 mm high. Angle between anterior and posterior hinge-plates about 140 to 145°. Posterior hinge-plate
straight with 12+ uniformly sized teeth. Anterior hinge-plate concave towards the body of the shell, with
13+ teeth, increasing in size forwards. Slightly chevron-shaped teeth passing beneath umbo in a continuous
series, chevrons directed towards the umbo. Anterior and posterior muscle scars clearly distinguished, equal
in size the anterior more deeply impressed than the posterior and most deeply impressed at its posterior edge.
Accessory muscles known (text-fig. 3a). Ligament and shell material unknown. Sculpture of fine concentric
lines (c. 4 per mm) (PI. 7, fig. 3).
Discussion. Nuculoid bivalves are poorly represented in the Bardahessiagh Formation, and those
that are found are not well preserved. In particular, the very soft nature of some of the rottenstone
has led to damage to, or loss of, the hinge-plate and teeth. The deformation apparent in other
species from the Bardahessiagh Formation (e.g. Cyrtodonta! spp.) appears to have produced some
variation among the Praenucula specimens; however, it is reasonable to regard them as belonging
to a single species.
Praenucula dispersa is distinguished from the Killey Bridge Formation species, Praenucula
praetermissa and P. infirma, by having smaller, less strongly chevron-shaped teeth. In the same
way it can be differentiated from P. costae (Sharpe) as described by Bradshaw (1970, pp. 630-633)
from the Llandeilo of Finistere. P. dispersa has more teeth then P. ciae (Sharpe) (of Bradshaw
1970, pp. 633-636) and Deceptrix pulchra armoricana Babin and Melou (1972, pp. 81-82, pi. 8,
figs. 4-7), and it lacks the posterior lobe seen in Hind’s Drummuck species Nuculana lobata (1910,
p. 519, pi. 4, figs. 1-3) and is less elongate than his N. curta (1910, p. 521, pi. 4, figs. 10-14) which
is also from the Drummuck Group (Ashgill).
Ctenodonta albertina Ulrich and C. filistriata Ulrich (1894, pp. 598-599, pi. 42, figs. 76-82,
text-fig. 44), both from the Cincinnatian, have a pit beneath the umbo according to Ulrich’s
descriptions. Recent illustrations (Pojeta 1971, pi. 5, figs. 14-16; 1978, pi. 2, figs. 1-2) show a
discontinuity in the dentition beneath the umbo not apparent in P. dispersa. Nucula dispar Barrande
(1881, pi. 273, VII, figs. 1-14) is more rounded than P. dispersa.
Praenucula praetermissa sp. nov.
Plate 7, figs. 16, 21, 24; text-fig. 3c
Derivation of name. From the Latin praetermittere, to overlook, referring to the fact that the species has been
overlooked by previous authors.
Type specimens. Holotype, IGS NIL 8757, composite mould of a left valve from the Killey Bridge Formation
of the Crocknagargan Stream section (locality 1 of Mitchell 1977), Pomeroy, Co. Tyrone. Paratypes: 6 left
and 6 right valves, internal, external, and composite moulds: IGS NIL 8687, 8747-8748, 8755, 8768, 8893,
8917, 8986, 8987 all from the same locality as the holotype; IGS NIG 482, Zs 2758 from locality 3 of Mitchell
1977, and IGS GU 1451 from his locality 2 (Warren Wood River). All from the Killey Bridge Formation.
Measurements
L
H
H/L
AL/L
Angle
Max.
110 mm
9-2 mm
0-83
0-70
160°
Min.
5-4 mm
3-6 mm
0-50
0-50
130°
Mean
8-6 mm
5-71 mm
0-72
0-60
143°
Median
8-2 mm
6-40 mm
0-66
0-60
145°
Description. Subovate to circular Praenucula in which the height is about 0-7 of the length. Maximum inflation
of a single valve 1-4 mm seen in a valve 110 mm long and 9-2 mm high. The umbo is at about the posterior
two-fifths. The shell is truncate behind and rounded in front. The anterior hinge-plate is broad and strongly
arched towards the body of the shell. The anterior and posterior hinge-plates meet at an angle of about 145°,
and usually bear the same number of teeth, up to nine. The teeth of the anterior hinge-plate are the larger
I and are strongly chevron-shaped; those of the posterior hinge-plate are markedly less so. The adductor muscle
scars are subequal in size, the anterior scar being the more strongly impressed and lying below and slightly
anterior to the anterior teeth with a diameter about one-quarter of the valve height; the posterior scar lies
immediately below the end of the posterior dental series. Accessory muscle scars unknown, except for one
anterior scar. Sculpture is of fine concentric lines, c. 16 per mm.
50
PALAEONTOLOGY, VOLUME 25
Discussion. This is a most distinctive form within the Pomeroy fauna, which, surprisingly, does
not seem to have been seen by previous workers (Portlock, McCoy, Reed) and yet is well represented
in recent collections, especially as fragments. Praenucula praetermissa is distinguished from P.
infirma, described herein from the Killey Bridge Formation, by its larger adductor muscle scars
and its more numerous posterior teeth, and from P. dispersa from the Bardahessiagh Formation
by its larger and strongly chevron-shaped teeth. P. praetermissa lacks the lobate posterior of Hind’s
Nuculana lobata (1910, p. 519, pi. 4, figs. 1-3) and is less elongate than his N. curta (1910, p. 521,
pi. 4, figs. 10-14), both from the Drummuck Group (Ashgill). It also lacks the pit beneath the
umbo recorded in N. curta and in the American species Ctenodonta albertina Ulrich and C.filistriata
Ulrich (1894, pp. 598-599, pi. 42, figs. 76-82, text-fig. 44).
It may be compared with other Ordovician Praenucula described from Europe: Deceptrix pulchra
armoricana Babin and Melou, 1972 ( = D . cf. ciae (Sharpe) of Babin and Robardet 1973) from the
Caradocian rocks of Normandy, has a coarser sculpture than P. praetermissa, c. 6-8 lines per mm,
grouped into ‘faisceaux’ (Babin and Melou 1972, p. 82, pi. 7, fig. 7). Praeleda pulchra Pfab, 1934,
is recorded by Havlicek and Vanek 1966, from the Llanvirn, Caradoc, and Ashgill of Bohemia,
and in shape resembles Praenucula praetermissa (Pfab 1934, p. 234, pi. 3, fig. 6) but Pfab gave no
detail of dentition or hinge, and no sculpture is evident in his figure. Nucula costae (Sharpe) from
Portugal, according to Sharpe’s description (1853, pp. 148-149, pi. 9, fig. 4a, b ) has more teeth
than P. praetermissa, about 20 (given as 25-30 at one point and as a total of about 18 elsewhere)
divided unequally between posterior and anterior, and shows the grouping of concentric striae
also described in D. armoricana. The forms described by Bradshaw (1970, pp. 630-633) as Praeleda
costae (Sharpe) from the Llandeilo of Finistere are closer in most respects to Praenucula praetermissa
EXPLANATION OF PLATE 7
Figs. 1-5. Praenucula dispersa sp. nov. 1, 2, oblique dorsal view and lateral view, internal mould of right
valve, holotype, IGS GU 1893. 3, ?composite mould of left valve, UM K4203. 4, internal mould of left
valve, paratype, IGS GU 3265. 5, internal mould of right valve, paratype IGS GU 1906. All from the
Bardahessiagh Formation (Caradoc), south of Craigbardahessiagh, Pomeroy, x 3.
Figs. 6-15, 17-19. Praenucula infirma sp. nov. 6, 7, oblique dorsal and lateral views, internal mould of left
valve, IGS NIL 8824, Crocknagargan Stream section, Pomeroy (IGR C.H721737). 8, 15, internal mould and
latex cast of external mould of right valve, IGS NIL 8682, 8682 pars, ditch exposure south of Craigbarda-
hessiagh, Pomeroy (IGR H7195 7385). 9, 10, dorsal view of external mould and dorsal view of internal
mould, conjoined valves, IGS NIL 8938, Crocknagargan Stream section, Pomeroy (IGR C.H721737). 11,
12, 13, dorsal and lateral views and latex cast of internal mould of right valve, holotype, IGS NIL 8935,
Crocknagargan Stream section, Pomeroy (IGR C.H721737). 14, internal mould of left valve, IGS NIL 8754,
Crocknagargan Stream section, Pomeroy (IGR C.H721737), x 3. 17, 18, 19, internal mould of left valve,
IGS NIL 8444, Tirnaskea Stream section, Pomeroy (IGR C.H727725); 17, latex cast, x 6; 18, oblique dorsal j
view, x3; 19, lateral view, x3. All Killey Bridge Formation (Ashgill, Cautleyan), x3.
Figs. 16, 21, 24. Praenucula praetermissa sp. nov. 16, internal mould of right valve, IGS NIL 8917,
Crocknagargan Stream section, Pomeroy (IGRc.H721737), x 3. 21, 24, left valve internal mould, holotype,
IGS NIL 8757, Crocknagargan Stream section, Pomeroy (IGR C.H721737); 21, latex cast, x 6; 24, lateral
view, x 3. All Killey Bridge Formation (Ashgill, Cautleyan).
Fig. 20. Praenucula aff. praetermissa sp. nov. internal mould of left valve, IGS NIL 8972, Killey Bridge
Formation (Ashgill, Cautleyan), Crocknagargan Stream section, Pomeroy (IGR C.H721737), x3.
Figs. 22, 23. Praenucula aff. praetermissa sp. nov. Oblique dorsal view and lateral view of crushed internal
mould of left valve, IGS NIL 8934, Killey Bridge Formation (Ashgill, Cautleyan), Crocknagargan Stream
section, Pomeroy (IGR C.H721737), x 3.
Fig. 25. Similodonta? sp., composite mould of left valve showing fine concentric ornament and anterior
adductor muscle scar, UM K4241, Bardahessiagh Formation (Caradoc), south of Craigbardahessiagh,
Pomeroy, x 3.
PLATE 7
TUNNICLIFF, Ordovician bivalves
52
PALAEONTOLOGY, VOLUME 25
but also tend to have more teeth (17-22) and no detail of the sculpture is given. N. ciae Sharpe
(1853, p. 149, pi. 9 fig. 5 a-c) from Portugal has a similar number of teeth (c. 19) and is described
as having fine concentric lines like P. praetermissa. However, P. praetermissa differs from Bradshaw’s
(1970, pp. 633-636) Praeleda ciae (Sharpe) in the posterior teeth: in Bradshaw’s P. ciae and in
Babin’s Ctenodonta ciae (Sharpe) (Babin 1966, pp. 49-52, text-fig. 1 1) the posterior chevron-shaped
teeth seem to point away from the umbo, in Praenucula praetermissa the chevrons point towards
the umbo. In this respect P. praetermissa resembles Sharpe’s N. ciae.
A single internal mould of a large left valve (IGS NIL 8972; PI. 7, fig. 20) from the Killey Bridge
Formation of the Crocknagargan Stream Section is like P. praetermissa in shape, but is 16 mm
long and 13 mm high with a maximum inflation of 4 mm. The hinge-plates meet at an angle of
145° with 13 anterior and 15 posterior teeth meeting below the umbo. Its musculature is indistinct
and sculpture unknown. Because of its large size and more central umbo, it is here recorded as P.
aff. praetermissa.
A second specimen (IGS NIL 8934; PI. 7, figs. 22, 23) from the same horizon and locality is
also recorded as P. aff. praetermissa. It is large, 1 1-8 mm long, but crushed dorso-ventrally. The
umbo is at the mid-point and the hinge-plates meet at an angle of 165° with 12 anterior and 14 +
posterior chevron-shaped teeth meeting below the umbo. The adductor muscle scars are round
and equal in size, about one-quarter of the valve height in diameter and each is close below and
slightly anterior to the end of the corresponding dental series. The sculpture is unknown. It differs
from P. praetermissa in its more central umbo and in having more teeth corresponding to its large
size.
The collections at IGS contain a specimen (IGS PT 9049) from the Corona Beds, Dufton
Shales (Caradoc, Longvillian) which is comparable with both P. praetermissa and P. pulchra
armoricana.
Praenucula infirma sp. nov.
Plate 7, figs. 6-15, 17-19; text-fig. 3 b
Derivation of name. From the Latin infirma, weak, referring to the small size of the adductor muscle
scars.
Type specimens. Holotype, IGS NIL 8935, an internal mould of a right valve, from the Killey Bridge Formation
of the Crocknagargan Stream section (locality 1 of Mitchell 1977) Pomeroy, Co. Tyrone. Paratypes: 9 left and
5 right valves, internal, external, and composite moulds: IGS NIL 8754, 8772, 8824, 8828, 8938, 8952 all
from the same locality as the holotype; IGS NIL 8444-8445 from the Tirnaskea Stream section; IGS Zs
2729, 2737-2738 from locality 3 of Mitchell 1977; IGS NIL 8682, 8682 pars from a ditch exposure at IGR
H 7195 7385. All from the Killey Bridge Formation.
Measurements
L
H
H/L
AL/L
Angle
Max.
12-4 mm
90 mm
0-81
0-64
160°
Min.
8-6 mm
4-2 mm
0-49
0-54
120°
Mean
10-4 mm
6-2 mm
0-68
0-59
145-5°
Median
10-5 mm
6-6 mm
0-65
0-59
140°
Description. Transversely subovate Praenucula in which the posterior end is slightly rounded or subtruncate
and the front is rounded or somewhat angular. The height is usually between 0-6 and 0-7 of the length. The
umbo is at about the posterior two-fifths. Maximum inflation of a single valve 2-2 mm seen in a valve 11-6 mm
long and 9-0 mm high. The highest part of the shell is a little anterior of the umbo. The hinge-plates meet
at an angle of about 140-145°; the anterior hinge-plate is broad and arched ventrally and bears up to eleven
large teeth, the posterior hinge-plate bearing up to nine which are smaller and less markedly chevron-shaped.
The dental series meet below the umbo. The adductor muscle scars are of equal size, in diameter about
one-eighth of the valve height; the anterior scar is anterior to the end of the anterior hinge-plate and lies at
the same height as the ventral edge of the plate. The posterior scar lies immediately below the end of the
posterior hinge-plate. Both scars are distinct and quite strongly impressed. At least two umbonal muscle scars
present. Sculpture of fine concentric lines, c. 16 per mm.
TUNNICLIFF: LATE ORDOVICIAN BIVALVES
53
Discussion. The small size of the adductor muscle scars separates Praenucula infirma from other
Praenucula species. In addition P. infirma is distinguished from P. praetermissa by its generally
more elongate shape, and relatively shorter posterior hinge-plate with fewer teeth, and from P.
dispersa by its larger, more strongly chevron-shaped teeth and more elongate shape. It bears some
resemblance in shape to Hind’s Nuculana curta (1910, p. 521, pi. 4, figs. 10-14, 14a, especially figs.
13, 14, 14a), but it lacks the central cartilage pit noted by Hind in that species. It differs in the
same way from Ctenodonta albertina Ulrich and C.filistriata Ulrich (1894, pp. 598-599, pi. 42,
figs. 76-82, text-fig. 44). It lacks the posterior lobe seen in N. lobata Hind (1910, p. 519, pi. 4, figs.
1-3). P. infirma shows no grouping of the growth lines as seen in Deceptrix pulchra armoricana
Babin and Melou 1972 ( = D . cf. ciae (Sharpe) of Babin and Robardet 1973) from the Caradoc of
Normandy. Like P. praetermissa , P. infirma differs from Praeleda ciae (Sharpe) of Bradshaw (1970,
pp. 633-636) and C. ciae (Sharpe) of Babin (1966, pp. 49-52, pi. 1, fig. 9) in the posterior teeth,
the chevrons apparently pointing away from the umbo in the C. ciae specimens illustrated, but
pointing towards it in infirma. This feature is, however, less marked than in Praenucula praetermissa.
In this respect P. infirma, like P. praetermissa, resembles Sharpe’s original N. ciae (1853, p. 149,
pi. 9, figs. 5 a-c).
Genus deceptrix Fuchs, 1919
Type-species. By monotypy, Deceptrix carinata Fuchs 1919, p. 79, pi. 7, figs. 1, 2. See discussion under
Praenuculidae above.
Deceptrix sp.
Plate 8, figs. 15, 18
v pars. 1843 Area regularis Portlock, p. 427, pi. 34, fig. 2 [see Deceptrix regularise IGS GSM 7805],
Material. Two left valves, IGS GU 1896 and UM K4202 and one right valve, IGS GSM 23215.
Horizon and locality. Bardahessiagh Formation; south of Craigbardahessiagh, Pomeroy, exact localities
uncertain.
Description. Shell subquadrate, truncate posteriorly and slightly rounded to subtruncate anteriorly. The height
to length ratio is between 0-62 and 0-68 and the umbo is at about the anterior two-fifths. Inflation of single
valve 1-5 to 2-0 mm. The anterior and posterior hinge-plates meet at an angle of 140 to 150°. 4+ anterior
and 10+ posterior teeth are visible in the specimens despite imperfect preservation. Anterior and posterior
muscle scars indistinct, lying close below the ends of the hinge-plates. Sculpture of fine concentric lines of
variable strength.
Discussion. Like other nuculoids, Deceptrix is uncommon in the Bardahessiagh Formation, and
they can all be described under one heading. Lacking complete information on dentition and
musculature, no specific name is proposed.
A specimen in the Sedgwick Museum (A 16306; a left valve) is said to be from the Bardahessiagh
Formation and may belong to the same species, but the anterior appears deformed and is very
slender compared with the posterior.
A specimen collected by Fearnsides, Elies and Smith, SM A 16295, also from the Bardahessiagh
Formation, was described by Reed (1952, p. 65) as Ctenodonta cf. nitida (Ulrich). It is a badly
preserved left valve with 8 + posterior and 3 + anterior teeth, the dentition being obscured in part.
Its height is 1 1 mm, and length 14 mm, giving height to length ratio 0-79. The umbo is at about
the anterior third. The angle between the anterior and posterior hinge-plates is 1 1 5° and the inflation
of the single valve is 2 mm. The musculature is unknown. It is unlike other Deceptrix from the
Bardahessiagh Formation in having a more rounded outline and the umbo closer to the anterior.
As Reed (1952, p. 65) observed, it bears a strong resemblance in form to Ctenodonta nitida (Ulrich
1894, p. 592, pi. 42, figs. 44-49), but he was clearly mistaken in suggesting that this specimen
belongs to the same species as McCoy’s Area quadrata (see Deceptrix apjohnni ) and his comparison
with Nuculites dissimilis (Portlock) is equally mistaken (see N. cylindricus).
54
PALAEONTOLOGY, VOLUME 25
Deceptrix apjohnni (Portlock 1843)
Plate 8, figs. 7, 8, 10, 11, 13, 14, 16, 17; text-fig. 4 a
v* 1843 Pectunculus Apjohnni Portlock, p. 429, pi. 34, fig. 8.
v. 1843 Pectunculus! ambiguus Portlock, p. 430, pi. 34, fig. 11.
v. 1846 Pectunculus Apjohni Portlock; McCoy, p. 19.
v. 1846 Area quadrata McCoy, p. 20, pi. 2, fig. 5 (reversed).
. 1878 Ctenodonta ambigua Portlock; Baily, p. 28.
v. 1952 Ctenodonta apjohni (Portlock); Reed, p. 60.
Type-specimens : IGS GSM 12409, Portlock’s (1843, pi. 34, fig. 8) figured specimen, is here selected as lectotype
of Pectunculus apjohnni ; paralectotype UM K4219. The only known syntype of PP. ambiguus Portlock, IGS
GSM 12410, is here selected as lectotype of P. ambiguus. The figured specimen of Area quadrata McCoy NMI
G.3. 1979, which is in the Griffith Collection, and which is the only known syntype, is here selected as lectotype
of A. quadrata. All are from the Killey Bridge Formation, Pomeroy, exact locality uncertain, but possibly
locality 3 of Mitchell, 1977.
Material'. Sixteen specimens in IGS, BM, NMI, UM.
Horizon and locality. All from the Killey Bridge Formation (Ashgill, Cautleyan), with well-localized specimens
recorded from the Crocknagargen Stream section (locality 1 of Mitchell 1977) and the Tirnaskea Stream
section. Specimens from old collections are less well localized.
Measurements
L
H
H/L
AL/L
Angle
Max.
14-0 mm
14-0 mm
1-17
0-5
160°
Min.
4-8 mm
4-4 mm
0 81
0-32
125°
Mean
11-78 mm
10-23 mm
0-93
0-36
145°
Median
9-4 mm
9-2 mm
0-99
0-41
142-5°
Description. Rounded Deceptrix, occasionally subquadrate (as in A. quadrata), usually increasing in height
towards the front. The postero-dorsal margin extends a little beyond the posterior hinge-line while the anterior
hinge-plate occupies about half the length of the well-rounded antero-dorsal margin. The umbo is situated
between the mid-point and the anterior third. Fleight occasionally greater than length but generally the height
to length ratio is about 0-9-10. Maximum inflation of a single valve 4 mm seen in the lectotype, 14 mm long
and 13 mm high. The posterior hinge-line is straight or gently curved, generally bearing up to 17 teeth
(exceptionally 26 as in A. quadrata)', the anterior hinge-plate is nearly straight or gently arched towards the
body of the shell with 10+ teeth. The anterior and posterior hinge-plates meet below the umbo at an angle
of 140-150°. The anterior adductor muscle scar is below the end of the anterior dental series, at about half
the height of the shell and is one-quarter to one-third the height in diameter. The posterior adductor muscle
scar is similar in size to the anterior scar and is close below the posterior dental series, a little anterior to its
end. The scars of three umbonal accessory muscles are present (PI. 8, fig. 17). Sculpture is of fine concentric
lines, often of variable strength towards the margin. Pallial line not seen.
Discussion. Portlock (1843, p. 429) described his Pectunculus apjohnni as a ‘very well marked species’
and its well-rounded or subquadrate appearance is certainly distinctive in the Killey Bridge
Formation fauna. It can only be confused, if fragmentary, with Deceptrix semitruncata but the
two have distinctive forms, D. apjohnni more rounded and D. semitruncata more obliquely elongate
and do not overlap in the range of their height-length ratios.
McCoy’s A. quadrata is here regarded as synonymous with D. apjohnni. A. quadrata was described
as ‘rare’ by McCoy (1846, p. 20) and only the type specimen is known and although it differs from
Portlock’s figured specimen of D. apjohnni in the manner outlined by McCoy, in particular its
apparently square form (the result of the ventral margin being obscured by matrix), it is here
considered to be an atypical specimen of D. apjohnni.
Reed (1952, pp. 60-61) likened D. apjohnni to several North American species which would now
be regarded as Similodonta, and in this he was clearly misled by the general shape of the shell, for
the dentition plainly shows D. apjohnni to be Deceptrix. In Similodonta the anterior and posterior
teeth are equal in number and size. Of the described species of Deceptrix, none matches D. apjohnni
TUNNICLIFF: LATE ORDOVICIAN BIVALVES
55
a
text-fig. 4. Deceptrix spp. based on internal moulds, c. x 5. (a) D. apjohnni (Portlock), IGS NIL 8910,
right valve. ( b ) D. semitruncata (Portlock), IGS GSM 12421, right valve, (c) D. subtruncata (Portlock), IGS
NIL 8971, right valve, (d) D. regularis (Portlock), IGS NIL 8919, left valve. Abbreviations as in text-fig. 3.
Note the differences in size and positions of the muscle scars and the differences in number and distribution
of teeth. Pedal accessory muscle scars are shown but not labelled in (a), (c), and (d).
so closely in shape as the type species D. carinata Fuchs, 1919, which has fewer anterior and more
posterior teeth.
Deceptrix semitruncata (Portlock 1843)
Plate 8, figs. 9, 12; text-fig. 4b
v* 1843 Pectunculus semi-truncatus Portlock, p. 429 (pars), pi. 34, fig. 7.
. 1878 Ctenodonta semi-truncatus Portl. (probably transversa)', Baily, p. 28.
v? 1910 Ctenodonta semitruncatus Portlock; Hind, p. 525, pi. 3, figs. 23-25.
v. 1952 Ctenodonta semitruncata (Portlock); Reed, p. 60, pi. 3, fig. 3.
Type-specimens. Lectotype selected by Reed 1952, p. 60, IGS GSM 12421, the specimen figured by Portlock;
from the Killey Bridge Formation, Pomeroy, exact locality uncertain. K4203, a syntype of Pectunculus
semitruncatus is now identified as Praenucula sp. from the Bardahessiagh Formation.
Material, horizon, and localities: Seven specimens in IGS and UM, all from the Killey Bridge Formation,
recorded from the Crocknagargan Stream section (locality 1 of Mitchell 1977) and in the Portlock Collection,
exact locality uncertain.
56
PALAEONTOLOGY, VOLUME 25
Measurements
L
H
H/L
AL/L
Angle
Max.
17 0 mm
13-0 mm
0-8
0-52
160°
Min.
110 mm
9.0 mm
0-65
0-36
140°
Mean
15-29 mm
11-14 mm
0-73
0-44
150°
Median
14-0 mm
1 1 -0 mm
0-72
0-44
150°
Description. Obliquely ovate Deceptrix, increasing in height towards the anterior, in which the height to length
ratio is between 0-6 and 0-8 and maximum inflation of a single valve seen is 3 mm. The umbo is situated
between the midpoint and the anterior third. The posterior margin is truncate and the anterior margin is
produced obliquely and is sometimes angular antero-ventrally where it is met by a slight umbonal ridge. The
posterior hinge-plate is straight or very gently curved, bearing 1 7 + teeth. The anterior hinge plate is straight
or slightly arched towards the body of the shell with 1 1 + teeth. The anterior and posterior dental series meet
beneath the umbo. The anterior and posterior hinge-plates meet at an angle of about 150°. Musculature
indistinct: anterior adductor muscle scar is at about half the height of the shell, below and a little before the end of
the anterior dental series; posterior adductor muscle scar is below the hind third of the posterior dental series.
Sculpture is of fine concentric lines of variable strength.
Discussion. D. semitruncata is compared with D. apjohnni under that species. D. semitruncata is
similar in shape to Ctenodonta britannica Babin (1966, p. 54, pi. 1 , figs. 1, 2), which may be referable
to Deceptrix but which has more numerous anterior teeth on a hinge-plate apparently arched away
from the body of the shell. C. socialis Ulrich (1894, p. 594, pi. 37, figs. 59, 60) is also similar in
shape to D. semitruncata but is a very small species with relatively fewer anterior teeth (6 out of
19, as opposed to 11 out of 17 in D. semitruncata). No other described species of Deceptrix is
readily comparable with D. semitruncata.
Deceptrix regularis (Portlock 1843)
Plate 8, figs. 1-6; text-fig. 4 d
x* pars 1843
. 1878
v pars. 1910
v. 1952
Area regularis Portlock, p. 427, pi. 34, fig. 2.
Ctenodonta regularis Portlock; Baily, p. 28.
Ctenodonta regularis Portlock; Hind, p. 524, ?pl. 3, figs. 15-17.
Ctenodonta regularis (Portlock); Reed, p. 61, pi. 3, fig. 4.
EXPLANATION OF PLATE 8
Figs. 1-6. Deceptrix regularis (Portlock 1843). 1, 4, oblique dorsal view and lateral view, internal mould of
left valve, IGS GSM 12419, Pomeroy, exact locality uncertain. 2, 5, lateral view and latex cast, internal mould
of right valve, IGS NIL 8803, Crocknagargan Stream section, Pomeroy (IGR C.H721737). 3, 6, lateral
view and latex cast, internal mould of left valve, IGS NIL 8919, Crocknagargan Stream section, Pomeroy
(IGR C.H721737). All Killey Bridge Formation (Ashgill, Cautleyan), x 3.
Figs. 7, 8, 10, 11, 13, 14, 16, 17. Deceptrix apjohnni (Portlock 1843). 7, latex cast of external mould of right
valve, IGS GSM 12410, lectotype of Pectunculus ambiguus Portlock, 1 843, Pomeroy, exact locality uncertain,
x 3. 8, internal mould of left valve, NMI G.3. 1979, lectotype of Area quadrata McCoy, 1846, Pomeroy,
exact locality uncertain, x 2. 10, internal mould of left valve, lectotype, IGS GSM 12409, Pomeroy, exact
locality uncertain, x3. 11, incomplete internal mould of left valve, UM K4253, Pomeroy, exact locality
uncertain, x 3. 13, 16, latex cast of hinge and lateral view, internal mould of left valve BM LL40005, locality
3 of Mitchell, 1977 (IGR H72977268), x 3. 14, 17, latex cast of hinge and lateral view, internal mould of
right valve, IGS NIL 8910, Crocknagargan Stream section, Pomeroy (IGR C.H721737), x 3. All Killey Bridge
Formation (Ashgill, Cautleyan).
Figs. 9, 12. Deceptrix semitruncata (Portlock 1843), oblique dorsal view and lateral view, internal mould of
right valve, lectotype, IGS GSM 12421, Killey Bridge Formation (Ashgill, Cautleyan), Pomeroy, exact
locality uncertain, x 3.
Figs. 15, 18. Deceptrix sp. 15, composite mould of right valve, IGS GSM 23215. 18, composite mould of
left valve, UM K4202. Both Bardahessiagh Formation (Caradoc), south of Craigbardahessiagh, Pomeroy,
x 3.
PLATE 8
TUNNICLIFF, Ordovician bivalves
58
PALAEONTOLOGY, VOLUME 25
Type-specimens. Lectotype (as ‘holotype’) inadvertently designated by Hind (1910, p. 524), IGS GSM 7805,
probably the specimen figured by Portlock (1843, pi. 34, fig. 2); remaining syntypes IGS GSM 12419, 23215,
UM K4202. GSM 23215 and K4202 are now identified as coming from the Bardahessiagh Formation, exact
locality uncertain, and are referred to Deceptrix sp. The lectotype and remaining syntype are from the Killey
Bridge Formation, Pomeroy, exact locality uncertain but possibly locality 3 of Mitchell 1977.
Material, horizon, and localities. Two specimens from the Portlock Collection, IGS GSM 7805, 12419 and
two from recent collections, IGS NIL 8803, 8919. All are from the Killey Bridge Formation; the locality of
the Portlock specimens is uncertain but the remaining specimens are from the Crocknagargan Stream section
(locality 1 of Mitchell 1977).
Measurements
L
H
H/L
AL/L
Angle
Max.
17-5 mm
13 0 mm
0-75
0-46
150°
Min.
10-5 mm
7.0 mm
0-64
0-36
145°
Mean
9-5 mm
13-5 mm
0-70
0-42
148-75'
Median
140 mm
10 0 mm
0-70
0-41
147-5°
Description. Transversely ovate Deceptrix, rounded at both anterior and posterior ends increasing in height
slightly towards the rear. Height to length ratio about 0-7. The umbo is at about the anterior two-fifths. The
maximum inflation of a single valve is 3 mm. The anterior and posterior hinge-plates meet below the umbo
at an angle of 150° with a continuous series of nearly twice as many smaller posterior teeth than anterior
teeth, up to 15 anterior and 25 posterior in a large specimen (PI. 8, figs. 2, 5). The posterior hinge-plate is
gently curved, the anterior is slightly arched towards the body of the shell. Adductor muscle scars vary from
quite strongly impressed (PI. 8, figs. 3, 6) to indistinct. Anterior scar lies below and a little in front of the
end of the anterior teeth; the posterior scar lies a little anterior to the end of the posterior teeth. Three
umbonal accessory muscle scars are present (PI. 8, figs. 3, 6). Sculpture is not well seen in the specimens but
appears to be of fine concentric lines.
Discussion. The lectotype of Deceptrix regularis, IGS GSM 7805, shows little other than the shape
and is not illustrated here. It is perhaps surprising that Hind (1910, p. 524) regarded this as the
type since GSM 12419 shows dentition, and there is no clear evidence which was the specimen
figured by Portlock [see Tunnicliff 1980].
EXPLANATION OF PLATE 9
Figs. 1-7, 9-11. Deceptrix subtruncata (Portlock 1843). 1-6, IGS NIL 8971, slab with internal moulds of
both valves, Crocknagargan Stream section, Pomeroy, (IGR C.H721737): 1, 2, latex cast and lateral view
of right valve; 5, 6, latex cast and lateral view of left valve (in fig. 6, difficulty in lighting the subject has
resulted in a shortened appearance of the anterior); 3, 4, dorsal views of left and right valves juxtaposed.
7, internal mould left valve, IGS GSM 12413, holotype of Ctenodonta deserta Reed 1952, Pomeroy, exact
locality uncertain. 9, oblique dorsal view and lateral view, damaged internal mould of left valve,
IGS GSM 1 2423, lectotype of Area transversa Portlock 1 843, Pomeroy, exact locality uncertain. 1 0, internal
mould right valve, IGS GSM 12422, Pomeroy, exact locality uncertain. 11, internal mould right valve,
lectotype, IGS GSM 12424, Pomeroy, exact locality uncertain. All Killey Bridge Formation (Ashgill,
Cautleyan), x 3.
Fig. 8. Hippocardia praepristis (Reed 1952), holotype, IGS GSM 24147, Killey Bridge Formation (Ashgill,
Cautleyan), Pomeroy, exact locality uncertain, x 3.
Figs. 12-20. Concavodonta imbricata (Portlock, 1843). 12, internal mould of right valve, IGS Zs 2790.
13-15, composite mould of right valve, IGS GU 1583; 15, dorsal view; 14, latex cast; 13, lateral view.
1 6, ?external cast of left valve, NMI G. 1 . 1 979, lectotype of Nucula subacuta McCoy, 1 846. 1 7, ?composite
mould of left valve, NMI G.2. 1979, cited by McCoy 1846 as Nucula protei Munster. 18-20, composite
mould of left valve, BM LL40003; 18, lateral view; 19, dorsal view; 20, latex cast. All Killey Bridge Formation
(Ashgill, Cautleyan), figs. 12-15, 18-20, from locality 3 of Mitchell 1977 (IGR H7297 7268), figs. 16, 17,
Pomeroy, exact locality uncertain. All x4.
PLATE 9
TUNNICLIFF, Ordovician bivalves
60
PALAEONTOLOGY, VOLUME 25
Although in position of umbo and in height-length ratio D. regularis closely matches the range
of D. semitruncata, it can be readily distinguished from that species by its transverse rather than
oblique elongation and its increase in height towards the posterior end, while D. semitruncata is
highest towards the anterior end. It is less easy to distinguish between D. regularis and D. subtruncata,
but the latter is always more transverse with the highest part of the shell at the umbo rather than
towards the posterior end, and has relatively larger adductor muscle scars and fewer teeth.
Pojeta’s illustrations of his Deceptrix ( D .) n. sp. 1 (1978, pi. 1 figs. 1, 2) show a close similarity
to D. regularis except that the umbo is closer to the mid-point and the shell is highest a little
anterior of the umbo and thus more closely resembles D. subtruncata. No other species of Deceptrix
is closely comparable with D. regularis.
v* 1843
vpars. 1843
v. 1843
v. 1846
v. 1846
. 1878
v non. 1910
1910
71946
v. 1952
v. 1952
v. 1952
Deceptrix subtruncata (Portlock 1843)
Plate 9, figs. 1-7, 9-11; text-fig. 4c
Area sub-truncata Portlock, p. 427, pi. 34, fig. 1.
Area transversa Portlock, p. 428, pi. 34, fig. 4 [non IGS GSM 12425 Nuculites sp.]
Area Eastnori? (Murchison); Portlock, p. 427, pi. 34, fig. 3.
Area subtruncata Portlock; McCoy, p. 20.
Area transversa Portlock; McCoy, p. 20.
Ctenodonta transversa Portl.; Baily, p. 28.
Ctenodonta aff. transversa Portlock; Hind, p. 523, pi. 3, figs. 12-14.
Ctenodonta eastnori Portlock; Hind, p. 525, pi. 3, fig. 20, ?figs. 21, 22.
Ctenodonta transversa (Portlock); Reed, p. 202.
Ctenodonta subtruncata (Portlock); Reed, p. 62, pi. 3, fig. 6.
Ctenodonta transversa (Portlock); Reed, p. 63.
Ctenodonta deserta Reed, p. 64, pi. 3, fig. 8.
Type-specimens. IGS GSM 12424, the specimen figured by Portlock (1843, pi. 34, fig. 1) is here selected as
lectotype of Area subtruncata ; paralectotypes IGS GSM 12422, TCD 14763. IGS GSM 12423, the specimen
figured by Portlock (1843, pi. 34, fig. 4) was selected as lectotype of A. transversa by Reed 1952, p. 68;
remaining syntypes IGS GSM 12425, TCD 7854. The holotype of Ctenodonta deserta Reed is IGS GSM
12413 by original designation (1952, p. 64). All from the Killey Bridge Formation, Pomeroy, exact localities
uncertain but possibly locality 3 of Mitchell 1977.
Material, horizon, and localities. Many specimens in IGS, UM, SM, TCD, in both old and recent collections,
are all from the Killey Bridge Formation. Recent collections contain specimens recorded from localities 1
and 3 of Mitchell 1977, while old collection material is less well localized.
Measurements
L
H
H/L
AL/L
Angle
Max.
19-4 mm
11-4 mm
0-64
0-51
170°
Min.
10 0 mm
5.4 mm
0-54
0-37
140°
Mean
17-3 mm
1015 mm
0-59
0-48
153-8°
Median
16-76 mm
10-26 mm
0-59
0-44
153-6°
Description. Transversely elongate, subelliptical Deceptrix, with obliquely subtruncate anterior and rounded
posterior end. Maximum height of shell below umbo. Height to length ratio about 0-6. Umbo nearly central.
Maximum inflation of a single valve is 5-0 mm in a valve 17-5 mm long and 10-0 mm high. Hinge-line straight
or slightly curved; anterior and posterior hinge-plates meet at an angle of about 155-160° in most specimens.
The posterior dental series is straight with up to twenty teeth seen in both right and left valves. The anterior
dental series is slightly arched towards the body of the shell with up to twelve teeth in the right valve and
thirteen in the left (a single, tiny subumbonal tooth in the left valve is here counted as anterior). The dental
series meet below the umbo. Adductor muscle scars are large, nearly half the height of the shell in diameter,
and distinct, each lying slightly in front of, and below the end of, its corresponding dental series. Four
umbonal accessory muscle scars are present in each valve (PI. 9, figs. 1-6), three distinct, one less so. Sculpture
of fine concentric lines of variable strength especially towards the margin. Pallial line not seen.
TUNNICLIFF: LATE ORDOVICIAN BIVALVES
61
Discussion. There is no basis on which A. transversa Portlock (IGS GSM 12423) or the specimen
which he labelled and described as A. Eastnori can be separated from Deceptrix subtruncata
(Portlock). Despite Reed’s observations (1952, p. 63) to the contrary, and Portlock’s original
description (1843, p. 428), the proportions of the shell in A. transversa are not greatly different
from those of D. subtruncata although the type specimen of A. transversa is incomplete and its
length must be estimated. Portlock’s figures of the three species (1843, pi. 34, figs. 1, 3, 4) are poor
and measurements taken from them are unreliable.
Reed (1952, p. 64) quite reasonably dissociated Portlock’s specimen of A. eastnori (GSM 12413)
from Sowerby’s species, but he failed to recognize its affinities with D. subtruncata and created a
new species, C. deserta. The second Portlock specimen of A. eastnori (GSM 12414) which Reed
compared to C. cingulata Ulrich is Nuculites sp.
The differences between D. subtruncata and D. regu/aris, with which it might be confused, are
discussed under the latter species. Of the published species of Deceptrix none compares so closely
with D. subtruncata as that figured by Pojeta (1978, pi. 1 figs. 1, 2) as D. (/).) n. sp. 1 which has
similar numbers of teeth, position of umbo, and, in the case of Pojeta’s figure 2, a closely similar
height to length ratio to the maximum seen in D. subtruncata.
Genus similodonta Soot-Ryen, 1964
Type-species. Designated by Soot-Ryen 1964, p. 498, Tellinomya similis Ulrich, 1892.
Similodonta? sp.
Plate 7, fig. 25
Material. Left valve, UM K4205 and a slab, UM K4241, bearing a left and a right valve; all composite moulds.
Horizon and locality. Bardahessiagh Formation, locality uncertain but south of Craigbardahessiagh, Pomeroy.
Measurements H L H/L AL/L Angle
K4241 (both valves) 15 0 mm 16 0 mm 0-94 0-5 c. 110°
K4205 15-6 mm 14 0 mm Ml 0-5 95-100°
Description. Triangular, equivalve, about as high as long with the umbo at the centre. Anterior and posterior
hinge-plates meeting at about 100°. Details of hinge unknown. Anterior adductor muscle scar distinct, lying
below anterior end of the anterior hinge-plate which is arched towards the body of the shell. Posterior adductor
muscle scar not known. Auxiliary musculature unknown. Sculpture of fine concentric striae (about 7 per
mm) with coarser striae developing at later growth stages.
Discussion. Reed’s description (1952, p. 63) of Ctenodonta perangulata suggests that C. perangulata
should be placed in Similodonta, but examination of his holotype (SM A 16454) has shown it to
be a Nuculites and it is here considered to be synonymous with Nuculites cylindricus (Portlock).
Genus concavodonta Babin and Melou, 1972
Type-species. Originally designated by Babin and Melou 1972, p. 83, Nucula ponderata Barrande, 1881, from
the Vinice, Zahorany, and Bohdalec Formations (Caradoc) of Bohemia (Havlicek and Vanek 1966), recorded
again from Bohemia by Pfab 1934, and also recorded from the Upper Ordovician of Normandy (Babin and
Robardet 1973), and, with doubt, the Caradoc of Brittany (Babin and Melou 1972).
Diagnosis after Babin and Melou (1972, p. 83). Genus of Praenuculidae of fairly rounded shape,
the posterior extremity being a little more slender than the anterior. Umbo prosogyrate. Imprints
of adductor muscles subequal, the anterior more strongly impressed in the shell at its posterior
edge. Pallial line complete. Teeth concavodont, that is, characterized by teeth in chevron with the
concavities towards the umbo. Ornamentation fairly strong and regular, concentric.
62
PALAEONTOLOGY, VOLUME 25
Concavodonta imbricata (Portlock 1 843)
Plate 9, figs. 12-20; text-fig. 5
* 1843
v. 1846
v. 1846
v non. 1910
v. 1952
v. 1952
Nucula acutal (Sowerby) var. imbricata Portlock, p. 430, pi. 34, fig. 10.
Nucula protei Munster; McCoy, p. 19.
Nucula subacuta McCoy, p. 19, pi. 2, fig. 3 (reversed).
Nuculana imbricata Portlock; Hind, p. 519, pi. 4, figs. 4-7, 7a [none of these specimens is
referable to Concavodonta but are referable to IPraenucula. Name imbricata re-established].
Nuculana ? imbricata (Portlock); Reed, p. 66 [affirms the use of imbricata ].
Ctenodonta cf. gibberula (Salter); Reed, p. 65.
Type specimens. Portlock’s type specimen remains untraced (Tunnicliff 1980). A lectotype for Nucula subacuta
McCoy, 1846, is here designated NMI G.l. 1979, the only known syntype.
text-fig. 5. Concavodonta imbricata (Portlock), dentition and mus-
culature based on composite mould, left valve, BM LL40003, c. x 10.
Compare with PI. 9, fig. 20. Abbreviations as in text-fig. 3.
Material. Thirty-six specimens in IGS, UM, BM, SM, mostly internal moulds of single valves, some with
counterparts, many apparently composite moulds showing the concentric ornament on the internal mould.
Horizon and localities. All specimens are from the Killey Bridge Formation, of Pomeroy, Co. Tyrone. Most
specimens are from locality 3 of Mitchell 1977. Others are from localities 1 and 2 of Mitchell and from the ji
Tirnaskea Stream section. Specimens from old collections are less precisely localized.
Measurements
L
H
H/L
AL/L
Angle
Max.
9-4 mm
6-6 mm
0-923
0-455
155°
Min.
3-8 mm
2-4 mm
0-517
0-250
100°
Mean
6-6 mm
4-5 mm
0-685
0-3367
131-1°
Median
6-6 mm
4-5 mm
0-720
0-3525
127-5°
Description. Ovate, equivalve nuculoid. Height generally about 0-7 of the length but variable. Prosogyrate
umbo at about the anterior third and extending a little above the hinge-line. Convexity greatest below umbo,
maximum inflation of a single valve 1-6 mm seen in a valve 9-4 mm long and 5-0 mm high. Angle between
anterior and posterior hinge-plates about 130°. Posterior hinge-plate straight. Teeth concavodont (see generic
diagnosis), anterior larger than posterior, apparently passing beneath umbo in a continuous series. 6 + anterior
teeth in both valves, 19+ posterior teeth seen in right valve and 16+ in left valve (29 posterior teeth were
seen in one large left valve (PI. 9, fig. 20)). Anterior and posterior adductor muscle scars clearly distinguished;
anterior slightly larger and more deeply impressed than posterior and most deeply impressed on its posterior
TUNNICLIFF: LATE ORDOVICIAN BIVALVES
63
edge. Accessory muscles unknown. Pallial line probably represented by a change in convexity close to the
ventral margin. Ligament and shell material unknown. Two forms of concentric sculpture observed; coarse
(3 to 6 striae per mm) giving the imbricate appearance (PI. 9, fig. 13), and fine (12 to 20 striae per mm) (PI.
9, fig. 12).
Discussion. This is the first record of Concavodonta from the British Isles but material under study
from Caradoc and Ashgill rocks in north Wales includes examples of Concavodonta sp. Specimens
not showing the dentition may be distinguished from Deceptrix and Praenucula by the relatively
coarse concentric sculpture, the position of the umbo and especially the straight posterior hinge-line.
Ulrich’s figures (1894, p. 589, pi. 38, figs. 25-28; pi. 42, figs. 38-40) of Ctenodonta planodorsata
(Ulrich), from the Trenton Shale, show similar straight sets of concavodont teeth, but the angle
between the anterior and posterior hinge-plates is more acute (105°) and the position of the posterior
adductor muscle scar differs, being in line with the posterior hinge-plate in C. planodorsata but
lying internally (i.e. antero-ventrally) to the teeth in Concavodonta imbricata. C. ponderata, as
described and figured by Babin and Melou (1972), Babin and Robardet (1973), and by Pfab (1934),
differs from C. imbricata only in that the posterior hinge-plate appears gently curved.
Superfamily nuculanacea H. Adams and A. Adams, 1858
Family malletiidae H. Adams and A. Adams, 1858
Genus nuculites Conrad, 1841
Type-species. Designated by Miller 1889, p. 496, Nuculites oblongatus Conrad, 1841, from the probable Middle
Devonian (McAlester 1968, p. 37) of New York.
Nuculites cylindricus (Portlock 1843)
Plate 10, figs. 1-17
71841
v* 1843
v. 1843
v. 1843
v pars. 1 843
v. 1846
v. 1846
v. 1846
1875
1878
1878
v non. 1910
v non. 1910
v non. 1910
? 1946
v. 1952
v. 1952
v. 1952
v. 1952
v. 1952
v pars. 1952
Nuculites planulatus Conrad, p. 50 [see Bretsky and Bretsky 1977, for synonymy].
Area cylindrica Portlock, pp. 428, 759, pi. 34, fig. 9.
Area dissimilis Portlock, pp. 428, 759, pi. 34, fig. 5.
Area obliqua Portlock, pp. 429, 759, pi. 34, fig. 6.
Area Eastnori (Murchison); Portlock, pp. 427, 759, non pi. 34, fig. 3.
Area cylindrica Portlock; McCoy, p. 19.
Area dissimilis Portlock; McCoy, p. 20.
Area obliqua Portlock; McCoy, p. 20.
Ctenodonta obliqua Portlock; Baily, p. 35, pi. 12, fig. 2.
Ctenodonta dissimilis Portlock; Baily, p. 28.
Ctenodonta obliqua Portlock; Baily, p. 28.
Ctenodonta dissimilis Portlock; Hind, p. 522, pi. 3, figs. 5-7.
Ctenodonta eastnori Portlock; Hind, p. 525, pi. 3, figs. 20-22.
Ctenodonta obliqua Portlock; Hind, p. 524, pi. 3, figs. 18, 19.
Clidophorus diu Lamont, p. 365, pi. 1, fig. 3.
Clidophorus cylindricus (Portlock); Reed, p. 66, pi. 3, fig. 9.
Ctenodonta dissimilis (Portlock); Reed, p. 63.
Ctenodonta obliqua (Portlock); Reed, p. 61, pi. 3, fig. 5.
Clidophorus occultus Reed, p. 67, pi. 3, fig. 10.
Ctenodonta perangulata Reed, p. 63, pi. 3, fig. 7.
Ctenodonta deserta Reed, p. 65, non pi. 3, fig. 8 [IGS GSM 12414, non holotype IGS GSM
12413],
Type-specimens. Lectotype of Nuculites cylindricus (Portlock) here selected, IGS GSM 12416, the specimen
figured and labelled by Portlock and described by Reed (1952, p. 66); paralectotypes are TCD 7876,
14761-14762, UM K4220, 4267. Lectotype of Area dissimilis Portlock, selected Reed (1952, p. 63), IGS GSM
12411; paralectotype is IGS GSM 12412. For A. obliqua Portlock, Hind, (1910, p. 524) selected IGS GSM
12415 as lectotype, leaving paralectotypes IGS GSM 12417, 12418. The holotype of Clidophorus occultus
64
PALAEONTOLOGY, VOLUME 25
Reed is IGS GSM 12415 by original designation ( occultus becomes a junior objective synonym of obliqua ).
The holotype of Ctenodonta perangulata Reed is SM A 16454 by original designation. The holotype of
Clidophorus diu Lamont is Lamont Collection No. 1 .
Material, localities, and horizon. Many specimens in IGS, TCD, UM, NMI, BM(NH), and SM, both in old
and new collections. All are from the Killey Bridge Formation of Pomeroy. Apart from recently collected
material especially from localities 1
l and 3 of Mitchell (1977),
specimens are generally poorly localized.
Measurements
L
H
H/L
AL/L
Max.
21-0 mm
16-0 mm
0-94
0-46
Min.
8-4 mm
3-8 mm
0-43
0 15
Mean
15-23 mm
91 mm
0 61
0-33
Median
14-7 mm
9-9 mm
0-68
0-30
Description. Nuculites of very variable shape, from posteriorly elongate transversely or obliquely, or rounded,
to more or less truncate posteriorly. Umbo between the mid-point and the anterior one-fifth. The height to
length ratio varies between about 0-4 and 0-95. Many specimens show a posterior fold running from the
umbo towards the postero-ventral margin (PI. 10, fig. 5). Occasionally a weak anterior fold is developed
running from the umbo towards the antero-ventral margin (PI. 10, fig. 12), anterior to the position of the
septum. Septal position indicated clearly by a strong impression on internal moulds and often evident on
external moulds. The septum extends, usually vertically, from the hinge-line to half-way to the margin (e.g.
PI. 10, figs. 2, 9) but may appear to be directed slightly posteriorly (e.g. PI. 10, figs. 1, 5) or more markedly
towards the anterior end (e.g. PI. 10, fig. 17). Posterior hinge-plate straight or slightly curved near the hind
end (PI. 10, fig. 9): anterior hinge-plate generally at an angle of about 150° to the posterior hinge-plate, but
the angle is variable between 120 and 180°. Dentition taxodont: 3+ anterior teeth and 15+ posterior teeth.
The anterior teeth are simple or slightly sigmoidal and are larger than the posterior teeth which are
chevron-shaped pointing towards the umbo and becoming simpler towards the posterior end. Dentition below
umbo obscured in all specimens. Maximum inflation of a single valve seen 3-5 mm in a valve 15-4 mm long.
Anterior adductor muscle scar distinct and occupying much of the portion of the shell anterior to the septum:
posterior adductor and other muscle scars unknown. Sculpture of fine ( c . 16-20 per mm) concentric lines
with some coarser lines interspersed, especially towards the margin in larger specimens.
Discussion. Bretsky and Bretsky (1977) have shown that N. planulatus Conrad, 1841, from the
Upper Ordovician of Quebec is a highly variable species; Watkins (1978, p. 44) has done the same
with the Silurian species N. antiquus (J. de C. Sowerby). This also appears to be the case with the
Killey Bridge Formation species. As isolated specimens, the types of Portlock’s Area cylindrica, A.
dissimilis, and A. obliqua appear to be different species but, seen in combination with the many
other specimens of Nuculites both in old and recent collections, they appear to be extreme variants
I
EXPLANATION OF PLATE 10
Figs. 1-17. Nuculites cylindricus (Portlock 1843). All preserved as internal moulds. All Killey Bridge
Formation (Ashgill, Cautleyan). 1, right valve, lectotype, IGS GSM 12416. 2, left valve, UM K4266.
3, 4, latex cast and lateral view, left valve, UM 1920-834. 5, right valve UM K4267. 6, left valve,
UM K4226. 8, right valve, IGS GSM 12412. 9, left valve, IGS GSM 12418. 10, left valve, UM K4263.
1 1 , left and right valves, probably of the same individual, UM K4264, lit from bottom right to accommodate
both valves. 12, nearly conjoined valves, UM K4233. 14, right valve, IGS GSM 12415, lectotype of
Area obliqua Portlock 1843. 15, conjoined valves, IGS GSM 1241 1, lectotype of Area dissimilis (Portlock
1843). 16, right valve, UM K4255. 17, left valve, UM K4231. All Pomeroy, exact locality uncertain,
x2i 7, right valve, IGS GU 2087, 13, left valve, IGS GU 2061. Both locality 3 of Mitchell 1977 (IGR
H7297 7268), x2$.
Figs. 18, 19. Nuculoid gen. et sp. nov., IGS NIL 9240-1. 18, both valves, showing the fracturing of both
and the greater displacement of parts of the left valve, x2. 19, concave latex cast of composite mould,
right valve showing the fine dentition, concentric ornament, and shallow sulcus, x 3. Tirnaskea Formation
(Ashgill, Hirnantian), Crocknagargan Stream section, Pomeroy (IGR H722737).
PLATE 10
TUNNICLIFF, Ordovician bivalves
66
PALAEONTOLOGY, VOLUME 25
of one form. There is no clear relationship between growth and height-length ratio that can be
related to Portlock’s species, and although many specimens may, by eye, be placed in one or other
of Portlock’s nominal species, there are others which fall between any two of the Portlock forms.
I am unable to quantify any differences between these ‘species’ and choose therefore to place them
in synonymy. There may be an argument for retaining Portlock’s obliqua and dissimilis as terms
for morphs.
Reed’s (1952) Clidophorus occultus is a junior objective synonym of Portlock’s A. obliqua and
his Ctenodonta perangulata falls within the range of the A. dissimilis form. Clidophorus diu Lamont
(1946), from the Lower Drummuck Group of Girvan, Ayrshire, appears from the figure and from
Lamont’s observations to fall within the range of the A. cylindricus form, differing only in the
absence of the posterior fold commonly seen in specimens of that form.
Unfortunately, while Bretsky and Bretsky could relate the relative abundances of the different
forms of N. planulatus to a known stratigraphic sequence, no such sequence is available in the
Killey Bridge Formation, and each collection and locality is, in effect, isolated. Many specimens
have only vague locality information.
On the basis of the description, figures, and measurements given by Bretsky and Bretsky it is
not clear that N. cylindricus is distinct from N. planulatus, but Portlock’s name is retained pending
some closer comparison.
nuculoid gen. and sp. nov.
Plate 10, figs. 18, 19; text-fig. 6
Material, horizon, and locality. A single specimen IGS NIL 9240-9241, showing nearly conjoined valves,
crushed but apparently not distorted, preserved as a composite mould. From the Tirnaskea Formation (Ashgill,
Hirnantian) of the Crocknagargan Stream section, Pomeroy, Co. Tyrone (IGR H722 737).
Description. Shell obliquely ovate and modioliform in appearance. For right valve, height 13-3 mm, length
22-9 mm giving a height-length ratio of 0-58. Inflation of the single valve at least 1 -4 mm. The small, prosogyrate
umbo is at the anterior three-tenths. A slight umbonal ridge runs to the postero-ventral angle at an angle of
text-fig. 6. Nuculoid gen. and sp. nov., based on latex cast of composite mould of right valve, IGS NIL
9240, x6-5. Compare with PI. 10, fig. 19.
TUNNICLIFF: LATE ORDOVICIAN BIVALVES
67
55° to the hinge-line, giving the shell its oblique appearance. A little posterior to the umbonal ridge is a
shallow sulcus, reflected in the sculpture and sinus in the posterior margin. This sulcus is at an angle of 35°
to the hinge-line. The posterior margin is otherwise rounded ventrally and obliquely slanted towards the
hinge-line. The anterior margin is also obliquely slanted. The ventral margin is straight. The hinge-line is
straight on both sides of the umbo, curving slightly at the posterior end. The anterior hinge-plate bears 3-4
thin blade-like parallel teeth, becoming shorter towards the anterior. The posterior hinge-plate bears 9 +
similarly thin blade-like parallel teeth extending to the postero-dorsal angle. Musculature, ligament, pallial
line unknown. Sculpture of fine regular concentric lines (about 7 per mm) more pronounced on the posterior
slope of the shell.
Discussion. Bivalves are uncommon in the Tirnaskea Formation and are usually poorly preserved.
It is unfortunate that this specimen, one of the best preserved from that horizon, should prove
enigmatic. In outline it might be taken for Modiolopsis sp., but the presence of numerous taxodont
teeth both before and behind the umbo precludes assignment to the Modiomorphidae. Suspicion
that the teeth are the result of slight deformation can be dismissed since the anterior and posterior
teeth are aligned contrariwise. This is endorsed by the way in which the posterior teeth of both
valves can be seen in the specimen. Had these structures been produced by some post-depositional
deformation, one would have expected them all to be parallel. It is possible but unlikely, that the
structures are related to the ligament.
Although generally smaller in size than the Pomeroy specimen, Silicula Jeffreys, as described
and illustrated by Allen and Sanders 1973, (pp. 263-309) and Allen 1978 (p. 394) shows many
similar features: small umbones, ovate, flattened form, fine, elongate teeth. Silicula is a modern
deep-sea protobranch of the family Siliculidae grouped by Allen (1978, p. 392) with the Malletiidae
and others in the Nuculanoida. The strong resemblance in form to Silicula suggests that the
Pomeroy specimen was a deep-water form, as would seem likely since the Tirnaskea Formation
passes into the overlying graptolitic, presumably offshore, Llandovery strata (Mitchell 1977, p. 5).
Subclass pteriomorphia Beurlen, 1944
Order arcoida Stoliczka, 1871
Superfamily cyrtodontacea Ulrich, 1894
Family cyrtodontidae Ulrich, 1894
Genus cyrtodonta Billings, 1858
Type species. Cyrtodonta rugosa Billings, 1858, by subsequent designation of Williams and Breger 1916, p. 149.
Cyrtodonta ? expansa (Portlock 1843)
Plate 11, figs. 1, 3, 4, 9
v* pars. 1843
v* pars. 1 843
v(?) 1843
v(?). 1875
1878
v. 1952
v. 1952
Modiola expansa Portlock, p. 425, pi. 33, fig. 6 [pars', longer variety],
Modiola Brycei Portlock, p. 425, pi. 33, fig. 7 [pars: see Cyrtodontal securiformis (Portlock)].
Modiola expansa Portlock; McCoy, p. 18.
Modiolopsis expansa Portlock; Baily, p. 35, pi. 12, fig. 2.
Modiolopsis expansa Portlock; Baily, 1878 [includes Brycei in expansa with doubt].
Modiodesma expansum (Portlock); Reed, p. 71, pi. 3, fig. 16.
Goniophora brycei (Portlock); Reed, p. 78 [pars: see Cyrtodontal securiformis (Portlock)].
Type-specimens. Lectotype here selected, IGS GSM 1 2445, the type of Portlock’s long variety of M. expansa
(1843, p. 425). For Portlock’s small variety see Cyrtodontal securiformis (Portlock). Lectotype of Modiola
brycei Portlock here selected IGS GSM 12443, Portlock’s figured specimen: other syntypes (paralectotypes)
IGS GSM 12444, GSM 22038 (not certainly a syntype), UM K4206-4210, 4249. All from the Bardahessiagh
Formation, Pomeroy, Co. Tyrone, exact localities uncertain, south of Craigbardahessiagh.
Material, localities , and horizon. A few specimens in IGS (GSM 12443-5) and UM (K4244, 4247, 74248,
74249) all from the Portlock Collection. Two specimens in the Griffith Collection at NMI may belong to this
species. All composite moulds from the Bardahessiagh Formation, locality as above.
68
PALAEONTOLOGY, VOLUME 25
Measurements of type-specimen. IGS GSM 12445: H 29-6 mm, L 50-6 mm, AL 8-6 mm.
Description. Elongate, obliquely ovate, becoming higher towards the posterior end, with a height-length ratio
between about 0-6 and 0-7. The straight hinge-line is about half as long as the total length of the shell.
Antero-dorsal margin almost straight or gently curved and meeting the hinge-line at an angle of about 150°
Ventral margin almost straight. Umbones prosogyrate and placed at about 0T5 of the length from the front,
projecting a little above the hinge-line. A rounded umbonal ridge extends backwards at about 35° from the
hinge-line to the ventral margin at the point furthest from the umbo. Preservation poor, but two posterior
lateral teeth can be seen at the posterior end of the hinge-line in the right valve and one in the left,
anterior teeth are not known. The anterior adductor muscle scar is small, the posterior is unknown. Maximum
inflation seen in a single undistorted valve is 4T mm in a valve 45 mm long. Sculpture of irregular concentric
striae.
Discussion. C.? expansa and C.? securiformis (Portlock) show the effects of distortion more than
other species from the Bardahessiagh Formation and this has influenced their nomenclatorial
history. Portlock (1843, p. 425) noted two varieties of M. expansa ; a longer variety, which he
figured (1843, pi. 33, fig. 6), and a smaller variety which he described briefly. The smaller variety
is now placed in C.? securiformis. Portlock’s M. brycei (1843, p. 425, pi. 33, fig. 7) was based on
specimens which he described as ‘somewhat distorted by pressure’ (1843, p. 426). Reed (1952)
removed the specimens of Portlock’s smaller variety from M. expansa and placed them in two new
species, Orthodesma tyronense (1952, p. 71) and Whiteavesia subexpansa (1952, p. 72). These are
now placed in C.? securiformis. Examination of Portlock’s type material, some of which was not
available to Reed, shows that those specimens which both Portlock and Reed placed in brycei are
either expansa or securiformis which have been, as Portlock observed, crushed dorso-ventrally. In
every case, the line of the umbonal ridge has proved the weakest and has produced, under pressure,
a sharp keel-like appearance, described by Reed as ‘acutely carinated’. It is clear that the specimens
of M. brycei which Portlock figured and described belong to C.? expansa, while those which he
described as ‘young individuals’ (1843, p. 426) are C.? securiformis.
Although the detail of the hinge is not well preserved, and the anterior teeth are not known,
the position and appearance of the posterior lateral teeth suggests that Cyrtodonta is the appropriate
genus for both expansa and securiformis but in the absence of the anterior dentition the genus
remains doubtful. The presence of the posterior teeth, which Reed (1952, p. 71) apparently
interpreted as a ligament groove, precludes the assignment of either species to the edentulous
Modiolopsis [ = Modiodesma, —Orthodesma Laroque and Newell, 1969, p. N397]. However, C.?
expansa bears a strong resemblance in form to Sphenolium sp. as figured by Pojeta (1978, pi. 8,
fig. 6) which, although previously placed in synonymy with Modiolopsis (as in the Treatise etc.,
p. N397), Pojeta (1978, p. 235) now appears to regard as a cyrtodontid. The strongly prosocline
form of C.? expansa distinguishes it from the species figured from North America by Ulrich,
Pojeta, and others which, in illustration, tend to have a more rounded outline.
v* pars. 1843
v* 1843
v* pars. 1843
v. 1846
1878
? non. 1946
v* 1952
v* 1952
v* 1952
v. 1952
v. 1952
Cyrtodonta? securiformis (Portlock 1 843)
Plate 11, figs. 5-7, 10, 12-14
Modiola expansa Portlock, p. 425 [pars: smaller variety],
Modiola securiformis Portlock, p. 425, pi. 33, fig. 8.
Modiola Brycei Portlock, p. 425 [pars: see Cyrtodonta? expansa (Portlock)].
Modiola securiformis Portlock; McCoy, p. 18.
Modiolopsis securiformis Portlock; Baily, p. 28.
Whitella cf. brycei (Portlock); Lamont, p. 366, pi. 1, figs. 1, 2.
Orthodesma tyronense Reed, p. 71, pi. 3, fig. 17.
Whiteavesia subexpansa Reed, p. 72, pi. 4, fig. 2.
Modiolopsis concentrica Hall and Whitfield var. simulans Reed, p. 69, pi. 3, fig. 14.
Modiolopsis securiformis (Portlock); Reed, p. 70, pi. 3, fig. 15.
Goniophora brycei (Portlock); Reed, p. 78 [pars: see Cyrtodonta? expansa (Portlock)].
TUNNICLIFF: LATE ORDOVICIAN BIVALVES
69
Type-specimens. Lectotype of Modiola securiformis here selected, IGS GSM 12448, Portlock’s figured
specimen: paralectotype IGS GSM 12449. Type of M. brycei: see Cyrtodonta? expansa (Portlock). Holotype
of Orthodesma tyronense IGS GSM 12447 by original designation of Reed (1952, p. 72). Holotype of Whiteavesia
subexpansa IGS GSM 12446 by original designation of Reed (1952, p. 73). Holotype of Modiolopsis concentrica
simulans Reed SM A 1 6294a, b by original designation of Reed (1952, p. 70). All from the Bardahessiagh
Formation, Pomeroy, Co. Tyrone, exact localities uncertain, south of Craigbardahessiagh.
Material , localities, and horizon. Specimens in IGS (GSM 12446-9, 22038, 24172, GU1880, Zfl021) and UM
(1920-845, K4206-4210, 4245-4246). All composite moulds, from the same locality and horizon as given
above.
Measurements. Type-specimen, GSM 12448, which is compressed dorso-ventrally; for left valve, the least
distorted: H 13-6 mm, L 30-7 mm, AL 8-2 mm. GSM 12446, undistorted, H 30 mm, L 351 mm, AL 9-7 mm.
Description. Obliquely ovate, becoming broader towards the posterior, with a height-length ratio of about
0-9, but very variable in the distorted specimens available. Maximum inflation seen in a single undistorted
valve is 3-9 mm in a valve 35-6 mm long. The hinge-line is straight or very gently curved and is about half
as long as the total length of the shell. At the posterior end of the hinge-line are 2+ posterior lateral teeth
in the right valve and 1 + in the left, but the preservation of the material prevents accurate description. The
antero-dorsal margin is curved and meets the hinge-line at an angle of about 130°. The ventral margin is
almost straight. Umbones prosogyrate, placed at about the anterior quarter projecting slightly above the hinge-
line. A rounded umbonal ridge extends backwards at about 50° from the hinge-line to the ventral margin at its
furthest point from the umbo. The anterior teeth are unknown. The musculature is unknown. Sculpture of
irregular concentric striae.
Discussion. See discussion of C.? expansa. Portlock (1843, p. 425) described his specimens of
Modiola securiformis as being ‘partly distorted by pressure’. In the lectotype (GSM 12448), the
right valve approaches the M. brycei Portlock form, with the umbonal ridge becoming sharpened;
the left valve is less distorted and has the form of the smaller variety of M. expansa. Another
specimen (UM K4245) shows two sets of distorted conjoined valves lying on the same plane,
oriented at about 45° to each other. In each the left valve is brycei- like in form, while the right
valve retains its shape with perhaps a reduction in height and an accentuation of the umbonal
ridge. Reed’s (1952) species O. tyronense and W. subexpansa are here interpreted as almost
undistorted valves of C.? securiformis. The types (GSM 12446-12447) are so alike, apart from
being opposite valves, that it is hard to understand why Reed separated them specifically, let alone
generically. Portlock’s ‘younger individuals’ of M. brycei are badly distorted C.? securiformis.
The reasons for referring securiformis to ‘ Cyrtodonta ? are discussed under C.? expansa. While
it is possible that C.? expansa and C.? securiformis may be synonymous, there are clear differences
between them: securiformis is noticeably more truncate posteriorly in appearance, its postero-dorsal
margin meeting the hinge-line at a steeper angle than in expansa , and its umbones being relatively
more posterior.
Reed’s (1952) Modiolopsis concentrica simulans seems to be a specimen of C.? securiformis which,
as a result of distortion, is higher and shorter than normal. Hind’s figure of C. transversa (1910,
pi. 4, figs. 19, 19a, ?20) from the Drummuck Group (Ashgill) closely resembles C.? securiformis,
but his specimen (BM L49860) is more inflated and slightly more prosocline than securiformis.
The specimen from the Drummuck Group figured by Lamont (1946, pi. 1, figs. 1, 2) as Whitella
cf. brycei (Portlock) should be compared with C. transversa rather than with C.? securiformis.
Cyrtodonta? sp.
Plate 13, fig. 1
v. 71843 Avicula orbicularis J. de C. Sowerby in Murchison; Portlock, pp. 425, 755 (Synoptical Table)
[pars: see Cycloconcha? speciosa (McCoy)].
Material, horizon, and locality. A single right valve, UM K4194 from the Bardahessiagh Formation, exact
locality uncertain but south of Craigbardahessiagh, Pomeroy.
70
PALAEONTOLOGY, VOLUME 25
Measurements. L 26-2 mm, H 24-5 mm, AL 2-4 mm, inflation of the single valve 3-0 mm, obliquity 40°.
Description. Subcircular with the umbo at about the anterior one-tenth. Umbo incomplete. Obliquity of valve
about 40°. The antero-dorsal margin meets the postero-dorsal margin at an angle of about 135°. Dentition
poorly preserved but two parallel posterior lateral teeth are suggested by faint grooves; anterior teeth unknown.
Anterior muscle scar faint, lying about half-way between the dorsal and ventral margins; posterior musculature
unknown. Sculpture of fine concentric striae visible close to the anterior margin, and faint traces of a coarser
concentric ornament ventrally.
Discussion. Although no Portlock label remains associated with this specimen it is almost certainly
one which he likened to Avicula orbicularis Sowerby in Murchison 1839; in shape it compares quite
well with the type of A. orbicularis (IGS Geol. Soc. Coll. 6888) but this is much larger with no
visible posterior lateral teeth. Were it not for the more rounded shape, this specimen might have
been placed in ?Cycloconcha speciosa, but in that species one would expect to see three posterior
lateral teeth in a right valve while in K4194 only two are apparent.
Genus Vanuxemia Billings, 1858
Type species. Vanuxemia inconstans Billings by subsequent designation of Miller 1889, from the Black River
and Trenton Groups.
Vanuxemia? contorta (Portlock 1843)
Plate 13, figs. 4, 8
v* 1843 Inoceramus contortus Portlock, p. 422, pi. 33, fig. 5.
v. 1952 Vanuxemia contorta (Portlock); Reed, p. 68, pi. 3, fig. 12.
Type-specimen. IGS GSM 12435, holotype by monotypy; Portlock’s brief description is of one shell.
Material, locality, and horizon. The type specimen (a fragmentary left valve) and a distorted right valve, IGS
Zf 1020, both showing external features and both from the Killey Bridge Formation, Pomeroy, exact locality
uncertain.
EXPLANATION OF PLATE 1 1
Figs. 1, 3, 4, 9. Cyrtodonta? expansa (Portlock 1843). 1, 4, conjoined valves, IGS GSM 12443, lectotype of
Modiola brycei Portlock, 1843; 1, lateral view of left valve; 4, dorsal view of both valves. 3, right valve,
UM K4247, probably the specimen figured by Baily, 1875, pi. 12, fig. 2. 9, left valve, lectotype, IGS GSM
12445. All preserved as ?composite moulds, Bardahessiagh Formation (Caradoc), south of Craigbarda-
hessiagh, Pomeroy, x 1 .
Fig. 2. Pterineid? gen. and sp. indet. ?external cast left valve IGS NIL 8981, Killey Bridge Formation
(Ashgill, Cautleyan), Crocknagargan Stream section (IGR H721737), x4.
Figs. 5-7, 10, 12-14. Cyrtodonta? securiformis (Portlock 1843). 5, right valve, IGS GSM 12447, holotype
of Orthodesma tyronense Reed, 1952, x 1. 6, left valve, IGS GSM 12446, holotype of Whiteavesia sub-
expansa Reed, 1952, x 1. 7, 10, conjoined valves lectotype, IGS GSM 12448; 7, dorsal view, x 1 10,
lateral view of left valve, x 1|. 12, external cast of left valve, SM A 16294a, holotype of Modiolopsis
concentrica var. simulans Reed, 1952, x 2. 13, 14, slab bearing two conjoined pairs of valves, UM K4245;
13, dorsal view to show the distortion of both left valves which has produced the sharp ‘umbonal ridge’,
leading to confusion in the past with Goniophora, x 1; 14, lateral view of both right valves, x 1. All except
12 preserved as ?composite moulds, Bardahessiagh Formation (Caradoc), south of Craigbardahessiagh,
Pomeroy.
Figs. 8, 11, 15. Semicorallidomus? sp. 8, latex cast of internal mould of left valve showing single small
tooth, IGS GU 1739, x4. 11, composite mould of left valve, IGS GSM 12450, x 3. 15, composite
mould of right valve, IGS GSM 12451, x3. All Bardahessiagh Formation (Caradoc), south of Craig-
bardahessiagh, Pomeroy.
PLATE 1 1
TUNNICLIFF, Ordovician bivalves
72
PALAEONTOLOGY, VOLUME 25
Measurements. Holotype incomplete. For Zf 1020 H 32-9 mm, L 25 1 mm, maximum oblique dimension
35-8 mm, apparent angle of obliquity 45 to 50°.
Description. Tall, the height being about 1-3 the length. Obliquity 45 to 50°. Length of hinge-line about half
that of valve. Dentition and musculature unknown. Umbo anterior but not terminal, prosogyrate. Valve
inflated to 8-7 mm in Zf 1020 after reduction by compression which has sharpened the umbonal ridge, especially
towards the umbo. Sculpture of fine concentric striae, 3-5 per mm, becoming stronger towards the margin.
Discussion. Neither specimen is well preserved; the type shows only the posterior region and Zf
1020 is crushed. Portlock labelled Zf 1020 as ‘indeterminate’. Reed did not use Zf 1020 when he
assigned contort a to Vanuxemia with some reservation but the more complete Zf 1020, showing the
anterior position of the umbones, supports his opinion. However, the nature of the material demands
that the generic assignment be qualified with a ?. It is here proposed that the species be restricted to
these specimens and that the species be regarded as otherwise unrecognizable.
Remarks. In describing the ambonychiids, reference is made to Pojeta 1966, in which a terminology
for certain angles and dimensions is standardized (text-fig. 7), and where a comprehensive review
of North American ambonychiids is given.
Type-species. Ambonychia radiata Hall, 1847, p. 163, by subsequent designation of Stoliczka (1871, p. 387).
Derivation of name. From the Latin arundinea, of reeds, reedy, referring both to the ornament and by allusion
to F. R. C. Reed.
v. 1843 Uncites gryphus Schlotheim var.; Portlock, p. 455, pi. 25a, fig. 8.
. 1878 Ambonychia gryphus-, Baily, p. 28.
v Ipars. 1910 Byssonychia radiata Hall; Hind, p. 487, pi. 1, figs. 20, 21, non figs. 19, 22.
v. 1952 Byssonychia gryphus (Portlock); Reed, p. 77.
Order pterioida Newell, 1965
Suborder pteriina Newell, 1965
Superfamily ambonychiacea Miller, 1877
Family ambonychiidae Miller, 1877
Genus: ambonychia Hall, 1847
Ambonychia arundinea sp. nov.
Plate 13, figs. 2, 3
Z
G
B
A
H
text-fig. 7. Linear and angular measurements used in describing Ambonychiids, after Pojeta, 1966. a— angle
alpha, b— angle beta, G— angle gamma, l— length, h— height, d— greatest dimension, z— length of hinge.
TUNNICLIFF: LATE ORDOVICIAN BIVALVES
73
Type-specimens. Holotype, right valve IGS GSM 24302. Paratypes, left valves IGS GSM 24303, BM LL
40004, right valves IGS GU 1452, 1566-1567 (counterparts), and UM 1920-835 (larger specimen). All external
moulds from the Killey Bridge Formation of Pomeroy, Co. Tyrone. GSM 24302, 24303 and UM 1920-835
are not precisely localized but probably come from locality 3 of Mitchell (1977), the Little River section (Irish
grid ref. H7297 7268), the locality for GU 1566-1567. GU 1452 and BM LL 40004 are from locality 2 of
Mitchell, Warren Wood River section (Irish grid ref. H7130 7128).
Description. Tall subrectangular Ambonychia. The greatest dimension at an angle of 50 to 55° to the hinge-line.
Length about 0-7 of the greatest dimension and height about 1-25 the length. The hinge length is 0-4 of the
length. Greatest inflation of a single valve seen is 4 mm in a valve 33 mm high (UM 1920-835). Prosocline,
with an angle gamma of 80 to 85 . Umbones not rounded but carinate in the sense of Pojeta 1962 (i.e. the
anterior portion of the valve is flattened antero-posteriorly), and projecting between 1 -4 mm and 3 0 mm above
the hinge-line. Valve surface mainly gently convex. The nature of the byssal gape is not discernible but the
byssal sinus is shallow. Dentition unknown although what may be the damaged remains of the posterior
laterals (?2 in right valve) are visible in GSM 24302. The ligament area is damaged or obscured in all the
present specimens and none shows musculature. About forty costae (range seen 33 + , 37 + , 40, 41, 43).
Discussion. Portlock briefly described his specimens (GSM 24302, 24303) as brachiopods. Hind
examined Portlock’ s specimens and assigned them, in museo to Byssonychia, as noted by Reed
(1952, p. 78) who gave the first full description, but had only these two specimens before him. He
discussed the similarity and differences between these specimens and those Byssonychia described
from Girvan by Hind (1910, p. 487, pi. 19) and other species from North America. However, he
chose to retain the specific name gryphus, and wrongly attributed its authorship to Portlock.
The North American Richmondian species Byssonychia richmondensis Ulrich, and B. robust a
(Miller), as described by Pojeta 1962, and subsequently (Pojeta 1966) transferred to Ambonychia ,
have similar numbers of costae and similar dimensions to the Pomeroy specimens, but the Irish
species has a more elongate appearance, the ratio of its length to its greatest dimension is 0-7 while for
Ambonychia richmondensis and A. robusta it is 0-5-0-6 and 0-6-0-75 respectively (based on Pojeta’s
figures). Both the American species tend more towards the acline form by about 10°.
Of Hind’s figures (1910, pi. 1, figs. 19-22) of B. radiata, which Pojeta (1962, p. 184) excluded
from B. radiata , figs. 19, 22 are comparable to the Richmondian species B. suberecta Ulrich, but
figs. 20, 21 (BM L49766-49767) are close in all respects to A. arundinea but have a slightly higher
number of costae (44 to 48).
Genus cleionychia Ulrich, 1892
Type species. By original designation Ambonychia lamellosa Hall, 1862.
Cleionychia transversa (Portlock 1843)
Plate 12, figs. 1-6
v* 1843
v pars. 1843
v. 1952
v. 1952
Inoceramus transversus Portlock, p. 423, pi. 33, fig. 11.
Inoceramus vetustus (J. de C. Sowerby) Var. priscus Portlock, p. 423, pi. 33, figs. 2, 3, non
fig. 1.
Clionychia subovalis Reed, p. 76, pi. 4, fig. 7.
Clionychia subquadrata Reed, pp. 76-77, pi. 4, fig. 8.
Type specimens. Lectotype of Cleionychia transversa IGS GSM 12439, selected by Reed 1952, p. 77; holotype
of C. subovalis Reed IGS GSM 12436 by original designation; lectotype of C. subquadrata Reed here selected,
IGS GSM 12437, the specimen figured by Reed (1952, pi. 4, fig. 8). Paralectotypes of transversa IGS GSM 12440,
UM K4199, all from the Portlock Collection; paralectotype of C. subquadrata IGS GSM 24304, from the
Wyatt-Edgell Collection.
Other material. IGS GSM 24305, from the Wyatt-Edgell Collection.
74 PALAEONTOLOGY, VOLUME 25
Horizon and locality. All from the Bardahessiagh Formation, exact locality uncertain but south of
Craigbardahessiagh, Pomeroy.
Measurements. For the lectotype IGS GSM 12439, L 67-6+ mm, H 40- 1 mm, maximum inflation of the single
valve 9-8 mm.
Description. Transversely elongate becoming relatively longer with increased size and with height about 0-6
of the length in large specimens. Inflation greatest below and slightly posterior to the umbo, gradually
diminishing towards the back. Prosogyrate umbo anterior and almost terminal and rising above the hinge-line.
Straight hinge-line about two-thirds of the valve length with duplivincular ligament area extending along
dorsal margin. Anterior end truncate, angle gamma about 90°. Anterior slope meets commissure at about
90° but may form a distinctly obtuse angle in distorted specimens. Ventral margin curving rapidly up to
anterior margin and more gently towards the posterior margin which is unclear in all the larger specimens
available, but rounded and continuous with the ventral margin in smaller specimens. Angle beta about 140°.
Musculature and dentition unknown. Sculpture of coarse concentric rugae, undulating in section.
Discussion. It is likely that the specimens on which Reed based his species C. subovalis and C.
subquadrata (1952, pp. 76-77) are juveniles of C. transversa. This is supported by measurement of
appropriate early growth stages of clear examples of C. transversa. It is found that the height to
length ratio decreases from about 0-9 at 25 mm long growth stage to 0-6 at 60 mm.
Early growth stages of C. transversa closely resemble C. undata Emmons, but the latter has a
slightly more acute angle gamma (apparently about 85° in Pojeta’s illustrations, 1966, pi. 34, figs.
1 -5) and there are no figures of late growth stages comparable to those seen in C. transversa.
Cleionychia prisca (Portlock 1843)
Plate 12, figs. 8, 12
v* pars. 1843 Inoceramus vetustus (J. de C. Sowerby) Var. priscus Portlock, p. 423, pi. 33, fig. 1, non
figs. 2, 3.
. 1878 Ambonychia undata Hall; Baily, p. 28.
v. 1952 Ambonychinia prisca (Portlock); Reed, p. 73.
EXPLANATION OF PLATE 12
Figs. 1-6. Cleionychia transversa (Portlock 1843). 1, left valve, lectotype, IGS GSM 12439. 2, left valve,
UM K4199. 3, right valve, IGS GSM 12436, holotype of Cleionychia subovalis Reed, 1952. 4, right valve,
IGS GSM 12440. 5, left valve, IGS GSM 12437, lectotype of Cleionychia subquadrata Reed, 1952.
6, left valve, IGS GSM 24304. All external casts, Bardahessiagh Formation (Caradoc), south of
Craigbardahessiagh, Pomeroy, x 1.
Figs. 7, 11, 13. Cleionychia incognita sp. nov. 7, right valve, IGS GSM 24310. 11, left valve, holotype,
IGS GSM 12441; note the bryozoan of the type referred by Portlock (1843, p. 360) to Entobia antiqua
Portlock. 13, small left valve, IGS GSM 24311. All external casts, Bardahessiagh Formation (Caradoc),
south of Craigbardahessiagh, x 1^.
Figs. 8, 12. Cleionychia prisca (Portlock 1843). 8, right valve, UM K4200. 12, right valve, lectotype,
IGS GSM 12438. Both external casts, Bardahessiagh Formation (Caradoc), south of Craigbardahessiagh,
Pomeroy, x 1.
Fig. 9. Corallidomus concentrica (Hall and Whitfield 1875), ?composite mould right valve, UM 1920-843,
Killey Bridge Formation (Ashgill, Cautleyan), Pomeroy, exact locality uncertain, x 2.
Fig. 10. Corallidomus? sp. ?external cast of right valve, NMI G.4. 1979, Killey Bridge Formation (Ashgill,
Cautleyan), Pomeroy, exact locality uncertain, x 2.
Fig. 14. Ambonychiopsis suspecta (Reed 1952), ?composite mould of left valve, holotype, IGS GSM 22103,
Bardahessiagh Formation (Caradoc), south of Craigbardahessiagh, x 1.
Fig. 15. Goniophora sp., external cast of right valve, IGS Zs 2751, Killey Bridge Formation (Ashgill,
Cautleyan), locality 3 of Mitchell 1977 (IGR H7297 7268), x4.
PLATE 12
TUNNICLIFF, Ordovician bivalves
76
PALAEONTOLOGY, VOLUME 25
Type-specimens, horizon, and localities. Lectotype IGS GSM 12438, right valve, (with a second valve only
partly visible) selected by Reed 1952, p. 74; paralectotype (not known to Reed) UM K4200. Both from the
Bardahessiagh Formation from Portlock’s locality ‘sheet 37 no. 6’ south of Craigbardahessiagh, Pomeroy.
Other material. A single specimen in the Griffith Collection, NMI; locality and horizon as for the type
specimens.
Measurements. For the lectotype, L 66-5+ mm, H 60-2+ mm, maximum inflation of the single valve 12-2 mm,
greatest dimension (oblique) 74T mm, angle alpha c. 55°, angle gamma c. 85°, angle beta c. 120°, length of
hinge uncertain, height of umbo above hinge-line 8 mm.
Description. Obliquely ovate form, narrow anteriorly and expanding rapidly towards the posterior end. Height
about 0-9 of length; precise length and detail of hinge-line unknown. Greatest dimension approximately 1-25
times length; angle alpha about 50°, angle gamma about 80 to 85°, angle beta about 120°. Maximum inflation
seen in a single valve, in the lectotype 12.2 mm. Umbo terminal. Antero-ventral slope steep, meeting the plane
of commissure at an angle of 90 to 100°. Posterior surface convex, becoming flatter towards the posterior
end. Ligament, musculature, and dentition unknown. Sculpture of coarse concentric rugae undulating in
section, about 2 per 10 mm along the plane of the greatest dimension.
Discussion. Although not well preserved, the specimens of Cleionychia prisca (Portlock) are
sufficiently distinct to be separated from C. transversa and C. incognita sp. nov. While the sculpture
in C. prisca is very like that in C. transversa, angle gamma is more acute in C. prisca and the shell
has a generally more oblique and inflated appearance. Angle alpha in C. prisca is less acute than
in C. incognita which lacks the strong concentric rugae of C. prisca.
Pojeta (1966, pp. 177, 180) pointed out that Reed’s figures were poor and that it was not clear
(despite his comments, 1952, p. 74) how Reed distinguished Cleionychia spp. from Ambonychinia
spp. None of the specimens which Reed assigned to Cleionychia is greatly inflated but they have
concentric undulating sculpture; those which he placed in Ambonychinia Isberg are more inflated,
elongate obliquely, and have prominent umbones and a more acute angle gamma, but while having
concentric ornament they apparently lack the radial costellae diagnostic of Ambonychiopsis Isberg
although in shape they resemble the latter. The specimens treated by Reed as Ambonychinia are
here assigned to Cleionychia following the suggestion inherent in Pojeta’s remarks (1966, p. 177).
Reed (1952, p. 74) pointed out the similarity of C. prisca to one of Isberg’s (1934, pi. 4, fig. 6)
figures of Ambonychinia corrugata (Lindstrom) and also noted the different appearance of the
remaining figures of A. corrugata (Isberg 1934, pi. 4, figs. 1-5, 7). The specimen in Isberg’s pi. 4,
fig. 6 appears to have stronger rugae than seen in C. prisca and each of its angles alpha, beta, and
gamma is a little more acute than seen in C. prisca.
Cleionychia incognita sp. nov.
Plate 12, figs. 7, 11, 13
Derivation of name. From the Latin incognita unknown or unrecognized, referring to the nomenclatorial
history of the type specimens.
v. 1843 Inoceramus trigonus (Munster); Portlock, p. 422, pi. 33, figs. 4, 4a.
. 1878 Ambonychia trigona Portlock; Baily, p. 28.
v. 1952 Ambonychinia cf. amygdalina (Hall); Reed, p. 75, pi. 4, fig. 5.
v. 1952 Ambonychinia cf. intermedia Isberg; Reed, p. 75, pi. 4, fig. 6.
v. 1952 Ambonychinia cf. volvens Isberg; Reed, p. 74, pi. 4, fig. 4.
Type specimens. Holotype IGS GSM 12441, a left valve, figured by Portlock 1843, pi. 33, fig. 4; paratypes
IGS GSM 24310, 24311 from the Wyatt-Edgell Collection, a right and a small left valve and IGS GSM
24306, 24307, 104237, three distorted specimens showing conjoined valves, all from the Wyatt-Edgell
Collection.
Horizon and localities. Bardahessiagh Formation, exact localities uncertain but south of Craigbardahessiagh,
Pomeroy.
TUNNICLIFF: LATE ORDOVICIAN BIVALVES
77
Measurements. For the holotype, L 41-4 mm, H 34-7 mm, greatest dimension (oblique) 49-3+ mm, angle
alpha c. 40°, angle gamma 75°, angle beta c. 120°, maximum inflation of the single valve 11-5 mm, length of
hinge 29-6+ mm.
Description. Obliquely ovate to subquadrate in form, narrow anteriorly and expanding rapidly towards the
posterior end. Height about 0-8 to 0-9 of length. Straight hinge-line about 0-7 to 0-75 length. Greatest dimension
approximately IT to 1-2 times length; angle alpha 40 to 50°. Angle gamma 70 to 80°, angle beta 115 to 135°.
Maximum inflation of a single valve seen 1 1 -5 mm in the holotype. Umbo terminal, rising above the hinge-line.
Antero- ventral slope steep, meeting the plane of commissure at an angle of about 100° but varying with size
and distortion of specimen. Posterior surface convex, becoming flatter towards the posterior end and with a
plano-concave dorsal area below the hinge-line. Ligament, musculature, and dentition unknown. Sculpture
of faint indistinct concentric lines and faint coarser rugae visible particularly on the antero-ventral surface.
Discussion. Cleionychia incognita is distinct from C. transversa and C. prisca in lacking the coarse
rugae seen over the whole surface in those species. It is less transverse than C. transversa and less
oblique than C. prisca which it closely resembles in general form but which does not show the
flattened dorsal area seen in C. incognita.
Reed (1952, p. 75) compared IGS GSM 24310 with Hall’s Ambonychia amygdalina and placed it
in Isberg’s genus Ambonychinia. Pojeta (1966, p. 163) has described amygdalina as belonging to
Ambonychiopsis Isberg. None of the present specimens of C. incognita shows the radial ornament
or anterior lobe associated with Ambonychiopsis. Ambonychinia intermedia Isberg, with which Reed
(1952, p. 75) compared IGS GSM 12441 does bear a close resemblance to C. incognita but Isberg’s
figure (1934, pi. 9, fig. 1) shows more obtuse angles gamma (c. 90°) and beta (7140°) than seen in
C. incognita. Similarly, there is a recognizable similarity between IGS GSM 24311 and A. volvens
Isberg with which Reed (1952, p. 74) compared it, but A. volvens appears to have a small anterior
lobe (Isberg 1934, pi. 2, fig. 4c) not seen in C. incognita. The flattened postero-dorsal area is not
so pronounced in other species of Cleionychia except perhaps C. intermedia. The use of the generic
name Cleionychia rather than Ambonychinia is discussed under C. prisca.
Genus ambonychiopsis Isberg, 1934
Type-species. By original designation of Isberg 1934, p. 82, Ambonychiopsis osmundsbergensis.
Ambonychiopsis suspecta (Reed 1952)
Plate 12, fig. 14
v* 1952 Vanuxemia? suspecta Reed, p. 69, pi. 3, fig. 13.
Type specimen, horizon, and locality. Holotype by monotypy IGS GSM 22103, by original designation of
Reed 1952, p. 69; a ?composite mould of a left valve, the only known specimen, from the Bardahessiagh
Formation, exact locality uncertain, but south of Craigbardahessiagh, Pomeroy.
Measurements. For the monotype L 26-9 mm, H 19-8 mm, greatest dimension (oblique) 29-1 mm, angle alpha
c. 35°, angle gamma 60°, angle beta 150°, maximum inflation of the single valve c. 3 mm, length of hinge-line
17-9 mm.
Description. The specimen is lacking in detail but shows an obliquely ovate form, very narrow anteriorly and
expanding rapidly towards the posterior end. Height is 0-7 of the length, hinge-line is 0-6 of the length and
is apparently straight. The greatest dimension, along the plane of obliquity 35° to the hinge-line, is 1-1 times
the length. Angles gamma and beta and inflation are given above. Anterior to the otherwise terminal umbo
is a small anterior lobe. The posterior surface is gently convex. The steep antero-ventral slope meets the plane
of commissure at an angle of about 60°. Ligament and musculature unknown. Dentition represented only by
two doubtful posterior lateral grooves below and anterior to the postero-dorsal angle of the margin. Sculpture
is poorly preserved and shows only irregular concentric striae.
Discussion. Portlock labelled this specimen ‘ Mytilus?—not specifically determined' . Reed described
two to three short cardinal teeth anterior to the umbo but these are not evident. The specimen
has been developed a little since Reed examined it and part of the small anterior lobe now revealed
78
PALAEONTOLOGY, VOLUME 25
may have led Reed to his conclusions and suggested to him his determination Vanuxemia? suspecta.
His use of a query (?) and the name suspecta both suggest that Reed remained doubtful as to the
nature of this specimen. Some doubt is retained in applying the generic name Ambonychiopsis to
suspecta but this assignment is justified by the ambonychiid nature of the form and sculpture, and
the presence of the anterior lobe. However, there is a lack of the radial sculpture characteristic of
Ambonychiopsis.
Ambonychinia balclatchiensis Reed (1944, p. 213, pi. 2, fig. 2) from the Balclatchie Beds is
probably an Ambonychiopsis : the illustration shows an anterior lobe, oriented such that it appears
to be dorsal. Angles alpha (c. 30°), beta (c. 150°), and gamma (c. 70°) in Reed’s figure are close
to those of A. suspecta and its greatest dimension is 103 of its length (1-1 in A. suspecta ), but the
anterior lobe in A. balclatchiensis is larger.
The preservation of the specimen is such that comparison with other species presents great
difficulty. It is here proposed that the name be restricted to the type specimen and the species be
regarded as otherwise unrecognizable.
pterineid? gen. and sp. indet.
Plate 11, fig. 2
A single, very small specimen (IGS NIL 8981) from the Killey Bridge Formation of the
Crocknagargan Stream section (Irish grid ref. H721737), is probably a pterineid. Although poorly
preserved, it can be compared in outline and size to Palaeopteria parvula Whiteaves (1897, p. 181,
pi. 20, figs. 1-3), from rocks of Black River-Trenton age (Caradoc) in the area of Lake Winnipeg.
Pterinea reticulata Hind (1910, p. 494, pi. 1, figs. 7-8) from the Drummuck Group of Girvan has
a distinctive ornament and is larger than the Irish specimen.
bivalve? gen. and sp. indet.
Plate 13, fig. 7
Material, horizon, and locality. A single specimen, IGS GSM 24308, from the Bardahessiagh Formation, exact
locality unknown, south of Craigbardahessiagh, Pomeroy.
Description. A subtriangular fragment measuring 40 0 by 30-7 mm and showing 19+ coarse concentric rugae.
Discussion. This fragmentary specimen has in the past been referred plausibly to Ambonychia undata
Hall, and indeed the coarse ornament on the specimen bears some resemblance to Hall’s figure
(1847, pi. 36, fig. la). It is considered, on the ornament alone, to resemble Cleionychia, but this
view gives rise to difficulties of interpretation and a gastropod or cephalopod affinity is more
probable. No comparable material is known from Pomeroy.
Subclass isofilibranchia Iredale, 1939
Order mytiloida Ferussac, 1822
Superfamily mytilacea Rafinesque, 1815
Family modiolopsidae Fischer, 1887
(after Pojeta and Gilbert-Tomlinson, 1978)
Genus modiolopsis Hall, 1847
Type species. Pterinea modiolaris Conrad, 1838, by original designation of Hall 1847, p. 157.
Modiolopsis sp.
Plate 13, fig. 11
Material, horizon, and locality. A single, fragmentary, small internal mould of a right valve, IGS GU 1380,
from the Killey Bridge Formation at the Crocknagargan Stream section (locality 1 of Mitchell 1977), Pomeroy,
Co. Tyrone.
TUNNICLIFF: LATE ORDOVICIAN BIVALVES
79
Description. Elongate, obliquely ovate, modioliform, with the umbo at about the anterior one-tenth to one-fifth.
Height (12-5+ mm), about half of length (26-0+ mm). Inflation of the single valve 2-5 mm. The incomplete
anterior end is narrow but the shell broadens rapidly towards the flat posterior which is also incomplete. The
ventral slope and margin show a slight sinus a little posterior to the umbo. A rounded umbonal ridge runs
towards the postero-ventral angle which is lost. An adductor muscle scar occupies much of the preumbonal
portion of the shell and two umbonal muscle scars are present. Edentulous but with a dorsal longitudinal
groove. Sculpture unknown except for a suggestion of concentric ornament on the antero-ventral slope.
Discussion. Although fragmentary, this specimen is similar in shape to two species figured by Hind
(1910): Modiolopsis exasperatus (Phillips) (Hind 1910, pi. 2, fig. 15) and M. scotica Hind (1910,
pi. 2, figs. 18-20) both recorded from the Drummuck Group (Ashgill) of Girvan. In both the
Scottish species the valves become higher very rapidly towards the posterior as in the Irish specimen,
but although the latter is incomplete, its postero-dorsal margin suggests that it is less elongate.
Genus corallidomus Whitfield, 1895
Type species. C. concentrica (Hall and Whitfield 1875), by monotypy: the only Ordovician species assigned
to the genus (Pojeta 1971, p. 6).
Corallidomus concentrica (Hall and Whitfield 1875)
Plate 12, fig. 9
* 1875 Modiolopsis concentrica Hall and Whitfield, pp. 86-87, pi. 2, fig. 18.
1893 [1895] Corallidomus concentricus Whitfield, pp. 492-493, pi. 13, fig. 2 [on p. 493], pi. 13, [as
n. gen., n. sp.].
1978 Corallidomus concentrica (Hall and Whitfield); Pojeta, pi. 12, figs. 12-14.
Material, horizon, and locality. A right valve, UM 1920-843, from the Killey Bridge Formation of Pomeroy,
Co. Tyrone, exact locality uncertain.
Description. A small modioliform shell, nearly twice as long as high (12 mm high, 22 mm long) and inflated
to nearly 2 mm at one point along the umbonal ridge. The shell broadens towards the posterior. The anterior
margin is rounded, the posterior end is obliquely truncate. The ventral margin shows a slight flexure at about
the mid-point. The umbo lies at about the anterior one-quarter. The preumbonal part of the shell is occupied
by an adductor muscle scar. Dentition uncertain but there is a suggestion of a short, thin, posterior tooth
immediately behind the umbo. The ornament is distinctive: the antero-ventral portion of the shell bears fine
concentric lines which give way at the umbonal ridge to strong, coarse, fairly regular concentric rugae. The
umbonal ridge becomes less marked away from the umbo but is defined by the change in texture and direction
of the ornament.
Discussion. The specimen closely matches the descriptions and illustrations of Hall and Whitfield
(1875, pp. 86-87, pi. 2, fig. 18) and the specimens figured by Pojeta (1978, pi. 12, figs. 12-14).
Pojeta (1978, pi. 1 2, figs. 12-14) figured Modiolopsis concentrica as Corallidomus concentrica, treating
it as synonymous with Whitfield’s type species, and is followed here; however, Whitfield’s (1893)
figures and the lack of his material (Pojeta 1971, p. 30) do give rise to some doubt as to whether
the two species are identical and therefore whether M. concentrica should be assigned to
Corallidomus. Pojeta (1971, pp. 30-33) observed that the known occurrence of Corallidomus is
restricted to the Richmondian, comparable to the Cautleyan age of the Killey Bridge Formation.
The mode of life of Corallidomus, recorded by Whitfield as boring into masses of the coral Labechia
ohioensis, is said by Pojeta (1971) to be unique among Ordovician bivalves and, although the
present specimen is not in life position, it is worth noting that masses of the tabulate coral Catenipora
tapaensis (Sokolov) [identified by Dr. D. E. White, IGS] were collected by Portlock (IGS GSM
103460, 104183, 104 184) from the Killey Bridge Formation, possibly from locality 3 of Mitchell 1977.
A left valve, NMI G.4. 1979, in the Griffith Collection is from the same horizon as UM 1920
843 but also lacks exact locality details and is here recorded as Corallidomus ? sp. (PI. 12, fig. 10).
It lacks the characteristic coarser ornament on the posterior portion of the shell and has a more
central umbo than C. concentrica.
PALAEONTOLOGY, VOLUME 25
Subclass palaeoheterodonta Newell, 1965
Order modiomorphoidea Newell, 1969
Superfamily modiomorphacea Miller, 1877
Family modiomorphidae Miller, 1877
Genus goniophora Phillips, 1848
Type species. Goniophora cymbaeformis (J. de C. Sowerby) by original designation of Phillips 1848, p. 264.
Goniophora sp.
Plate 12, fig. 15
Material, locality, and horizon. A single specimen, IGS Zs 2751, external cast of a right valve collected by
Mr. R. P. Tripp from the Killey Bridge Formation, at locality 3 of Mitchell (1977), the Little River Section
(Irish grid ref. H7297 7268).
Measurements. H 6 0 mm, L 13-8 mm, AL 2-8 mm; inflation of the single valve 1-8 mm; obliquity, measured
along umbonal ridge, 35-40°.
Description. Small Goniophora', transversely elongate with a pronounced, slightly sigmoidal umbonal ridge or
carina running from the umbo to the postero-ventral angle; height to length ratio 0-43; umbo barely protruding
above the hinge-line and situated at about the anterior one-fifth. Dentition and musculature unknown. The
surface anterior to the carina shows concentric striae of irregular strength and spacing but the posterior
surface has a well-developed concentric sculpture, regularly spaced, with 3-4 striae per mm. The posterior
surface also shows a flexure close behind the carina.
Discussion. This remains the only known Goniophora specimen from the Ordovician of Pomeroy:
Goniophora brycei (Portlock) of Reed (1952, p. 78) is a Cyrtodontal (see Cyrtodontal spp.). In the
absence of any detail of the hinge, there is some doubt as to whether Goniophora or Goniophorina
Isberg is the more appropriate genus for this specimen; but it is reasonable to place it in the more
familiar Goniophora. The lack of radial costae precludes it from being Cosmogoniophora McLearn.
Specimens of a comparable form from Caradocian rocks in north Wales are under study.
Genus colpomya Ulrich, 1894
Type species. Monotype, Colpomya constricta Ulrich by original designation of Ulrich 1894, pp. 522-523.
Colpomya simplex (Portlock 1 843)
Plate 13, fig. 14
v* 1843 Cypricardia? simplex Portlock, p. 426.
v pars. 1910 Grammysia undata J. de C. Sowerby; Hind, p. 540, pi. 5, fig. 14, Inon figs. 13, 15, 16.
v. 1952 Cuneamya simplex (Portlock); Reed, p. 79, pi. 4, fig. 10.
Type-specimen. The only known syntype, a left valve, IGS GSM 24290, from the Killey Bridge Formation,
is here selected as lectotype; locality given by Portlock as his ‘sheet 37 no. 2’, probably locality 3 of Mitchell
(1977), the Little River section, Pomeroy (Irish grid ref. H7297 7268).
Measurements. H 22-9 mm, L 39-5 mm, AL 14 00 mm, inflation of the single valve is 7-8 mm; obliquity,
measured along the umbonal ridge, 35°.
Description. Transversely elongate, rounded at posterior and anterior ends, with dorsal and ventral margins
subparallel; posterior broader than anterior. The height is about 0-6 of the length and the umbo is at about the
anterior one-third. The single valve is inflated to 7-8 mm in the type-specimen. A slightly sigmoidal rounded
umbonal ridge at an angle of 35° to the hinge-line meets the postero-ventral margin below the posterior end
of the hinge-line. The prosogyrate umbo extends above the hinge-line and is flattened, the flat area corresponding
to the very faint sulcus developed towards the margin anterior to the umbonal ridge. Lunule present but
TUNNICLIFF: LATE ORDOVICIAN BIVALVES
81
escutcheon not clearly differentiated. Dentition and musculature unknown. Sculpture of coarse, rather irregular
concentric rugae, pronounced anterior to the sulcus and becoming fainter and even indistinguishable on the
posterior surface.
Discussion. Portlock (1843, p. 426) and Hind (MS, see Reed 1952, p. 79) compared this species to
Cypricardia impressa J. de C. Sowerby (in Murchison). As Portlock observed, the sulcus in C.
simplex is much less pronounced than in C. impressa. Colpomya simplex differs from Cuneamya
in that the umbo is not terminal (cf. Cuneamya miamiensis Hall and Whitfield as figured in the
Treatise, p. N821 and Pojeta 1971, pi. 15, figs. 9, 10), but it is closely comparable to the type
species of Colpomya, C. constricta Ulrich (1894, p. 523, fig. 41) from the upper Trenton of Kentucky.
The specimen figured by Hind (1910, pi. 5, fig. 14) from the Drummuck Group (BM L49889) may
be conspecific with C. simplex.
Genus semicorallidomus Isberg, 1934
Type species. Semicorallidomus whitfieldi Isberg by original designation of Isberg 1934, pp. 175, 180 from the
Ordovician of Sweden.
Semicorallidomus? sp.
Plate 11, figs. 8, 11, 15
v. 1843 Modiola Nerei (Munster); Portlock, p. 424, pi. 33, fig. 10.
v. 1843 Mytilus? Nerei (Munster); Portlock, p. 424.
1878 Modiolopsis Nerei Munster; Baily, p. 28.
? 1952 Modiolodon speciosus (McCoy); Reed, p. 72, pi. 4, fig. 1.
v. 1952 Paramodiola? sp.; Reed, p. 73, pi. 4, fig. 3.
Material. IGS (Portlock and Mitchell collections) GSM 12450-12452, GU 1739, 1741; UM (Portlock and
Grainger collections) 4125 and counterpart, K4169, 4195, 4197-4198, 4242; NMI (Griffith collection); TCD
7871 (Portlock collection); ?SM A 16461.
Localities and horizon. All from the Bardahessiagh Formation, south of Craigbardahessiagh, Pomeroy, exact
[ localities uncertain.
Measurements
L
H
H/L
AL/L
Max.
20 0 mm
140 mm
0-85
015
Min.
100 mm
7-2 mm
0-62
0-07
Mean
14-63 mm
10-4 mm
0-71
Oil
Median
150 mm
1 0-6 mm
0-74
Oil
Description. Subovate, modiomorphoid, with a height-length ratio about 0-7. Umbo not prominent and at
about the anterior one-tenth to one-fifth. Some specimens show a slight flexure of the posterior surface.
Obliquity about 30°. Apparently isomyarian with the anterior adductor muscle scar almost below the umbo
and the posterior scar poorly defined. Single valve inflated to about 2-5 mm in the larger specimens. Concentric
ornament of varying strength, with an average of five striae per mm on the most inflated part of the shell
along the plane of obliquity. Precise nature of the hinge unknown, but the left valve apparently bears a tooth
below the umbo (IGS GU 1739, PI. 11, fig. 8).
Discussion. The name Paramodiola, used by Reed (1952) for this form, is rejected here in view of
the apparent presence of a tooth in the left valve (GU 1739); Paramodiola is said to be edentulous.
A diagnostic characteristic of Semicorallidomus is the presence in the left valve of a socket, the
reverse of what is seen in the Pomeroy form. In outline, the Irish specimens most closely resemble
the type species Semicorallidomus whitfieldi Isberg. The specimen described and figured by Reed
(1952, p. 72, pi. 4, fig. 1) as Modiolodon speciosus (McCoy) (SM A16461) was not available for
study at the time of writing, but his figure resembles the known specimens of Semicorallidomus ?:
McCoy’s specimen of M. speciosus is in the NMI (see Cycloconcha? speciosa).
82
PALAEONTOLOGY, VOLUME 25
Subclass actinodontia Douville, 1912 (after Pojeta 1978)
Family cycloconchidae Ulrich, 1884
Genus cycloconcha Miller, 1874
Type species. By original designation Cycloconcha mediocardinalis Miller, 1 874, p. 231.
Cycloconcha? speciosa (McCoy 1 846)
Plate 13, figs. 5, 6, 9, 10, 12, 13
v. 1843 Avicula obliqua J. de C. Sowerby in Murchison; Portlock, p. 425.
v. 1843 Avicula orbicularis J. de C. Sowerby in Murchison; Portlock, pp. 425, 755 (Synoptical Table)
[pars: see Cyrtodonta? sp.].
v* 1846 Pullastra speciosa McCoy, p. 17, pi. 2, fig. 2 (reversed).
Inon 1952 Modiolodon speciosus (McCoy); Reed, p. 72, pi. 4, fig. 1.
Type specimen. The original of McCoy’s figure (1846, pi. 2, fig. 2) is in the NMI Griffith Collection and as
the only known syntype is here selected as Lectotype, NMI G.5. 1979.
Material. Five right valves; GSM 21919, 21920, 104236, UM K4213, 4214. Four left valves; GSM 104235,
UM K4212, 4216 and the type specimen. All are composite moulds.
Horizon and locality. All from the Bardahessiagh Formation, exact localities uncertain but south of
Craigbardahessiagh, Pomeroy.
Measurements
L
H
H/L
AL/L
Obliquity
Max.
24-2 mm
22-0 mm
0-94
0-38
55°
Min.
16 0 mm
13-0 mm
0-72
0-29
45°
Mean
22-4 mm
18-6 mm
0-83
0-34
50-6°
Median
20- 1 mm
17-5 mm
0-83
0-34
50°
explanation of plate 13
Fig. 1. Cyrtodonta? sp. ?composite mould of right valve, UM K4194, Bardahessiagh Formation (Caradoc),
south of Craigbardahessiagh, Pomeroy, x IF
Figs. 2, 3. Ambonychia arundinea sp. nov. 2, external cast of right valve, IGS GSM 24302. 3, external cast
of left valve IGS GSM 24303. Both Killey Bridge Formation (Ashgill, Cautleyan), Pomeroy, exact locality
uncertain, x IF
Figs. 4, 8. Vanuxemia? contorta (Portlock 1843). 4, external cast of left valve, holotype, IGS GSM 12435.
8, external cast of right valve, IGS Zf 1020. Both Killey Bridge Formation (Ashgill, Cautleyan), Pomeroy,
exact locality uncertain, x 1 .
Figs. 5, 6, 9, 10, 12, 13. Cycloconcha? speciosa (McCoy 1846). 5, left valve, lectotype NMI G.5. 1979.
6, right valve, UM K42 14.' 9, left valve, UM K4212. 10, right valve, IGS GSM 21920. 12, right valve,
IGS GSM 21919. 13, left valve, UM K4216. All preserved as composite moulds, Bardahessiagh Formation
(Caradoc), south of Craigbardahessiagh, Pomeroy, x IF
Fig. 7. Bivalve? gen. and sp. indet. ?external cast IGS GSM 24308, Bardahessiagh Formation (Caradoc),
south of Craigbardahessiagh, x 1 .
Fig. 11. Modiolopsis sp. internal mould of right valve, IGS GU 1380, Killey Bridge Formation (Ashgill,
Cautleyan), Crocknagargan Stream section, Pomeroy (IGR H721737), x 2\.
Fig. 14. Colpomya simplex (Portlock 1843), external cast of left valve, lectotype, IGS GSM 24290, Killey
Bridge Formation (Ashgill, Cautleyan), Pomeroy, exact locality uncertain, x IF
Figs. 1 5 19. Lyrodesma radiatum (Portlock 1843). 15, 17, left valve, internal mould, IGS GU 3264; 15,
oblique dorsal view to show dentition, lit from bottom left, x6; 17, lateral view, lit from bottom left,
x4. 16, external cast of left valve, lectotype, IGS GSM 22178, x2. 18, external cast of right valve, IGS
GSM 22177, x2. 19, composite mould of conjoined valves, IGS GSM 104185, x2. All Bardahessiagh
Formation (Caradoc), south of Craigbardahessiagh, Pomeroy.
PLATE 13
TUNNICLIFF, Ordovician bivalves
PALAEONTOLOGY, VOLUME 25
Description. Cycloconcha? of slightly variable subovate shape, the posterior end being somewhat truncate,
with height between 0-72 and 0-94 of the length. Slight umbonal ridge is at about 50° to the hinge-line. The
anterior dorsal margin is at an angle of 145 to 155° to the posterior dorsal margin. The prosogyrate umbones
are at about the anterior one-third. Maximum inflation of a single valve seen is 2 mm in a valve 24-2 mm
long. Hinge-line shows two or three lamellar posterior lateral teeth in the left valve and three in the right
valve, subparallel to the margin, and at least one anterior lateral tooth in each. Anterior adductor muscle
scar faint, about half-way between umbo and anterior; posterior musculature unknown. Sculpture of regular
concentric striae with more prominent lines at intervals of about 2-4 mm along the umbonal ridge especially
towards the margin.
Discussion. This species is placed, with some reservation, in Cycloconcha since it possesses straight
anterior and posterior teeth, although no cardinal teeth are visible in the available specimens. I
have not seen Reed’s specimen of Modiolodon speciosus (SM A16461, 1952, p. 72, pi. 4, fig. 1),
and doubt whether it should be placed in C.? speciosa. Reed noted radial striation, stating that
this appeared to be present in McCoy’s specimen but he was working only from McCoy’s figure
and no radial striation is apparent in the type specimen. He also mentioned a deeply impressed
anterior muscle scar but in the specimens of C.? speciosa available the anterior muscle scar can at
best be described as faint. His comparison of his specimen with Modiolopsis? consimilis Ulrich,
Modiolodon obtusus Ulrich, and M. truncatus (Hall) all suggest that we should regard his specimen
as distinct from C. speciosa.
In shape, C.? speciosa resembles Cyrtodonta parva Ulrich (1894, p. 541, pi. 39, figs. 24, 25) from
the Trentonian of Minnesota but it is larger and the posterior teeth are longer. In this last respect,
comparison may be made with the upper Trenton shale form Cypricardites tenellus Ulrich (1892,
p. 237) from Minnesota. The teeth in C. tenellus, later referred to Cyrtodonta by Ulrich (1894, p.
546, pi. 40, figs. 15-19), are similar in form to those of Cycloconcha speciosa, but only two are
recorded in the right valve.
Family lyrodesmatidae Ulrich, 1894
Genus lyrodesma Conrad, 1841, p. 51
Type-species. By monotypy Lyrodesma planum Conrad, 1841, p. 51.
Lyrodesma radiatum (Portlock 1843)
Plate 13, figs. 15-19
v* 1843 Nucula? radiata Portlock, p. 430, pi. 36, fig. 11.
v. 1846 Nucula radiata Portlock; McCoy, p. 19.
1878 Ctenodonta radiata Portlock; Baily, p. 28.
v. 1952 Lyrodesma radiatum (Portlock); Reed, p. 67, pi. 3, fig. 11.
Type-specimen. Lectotype selected here IGS GSM 22178, the specimen figured by Portlock (1843, pi. 36, fig.
11) and used by Reed (1952, p. 67) in his redescription.
Material, localities, and horizon. Specimens in IGS, TCD, and NMI, including six Portlock syntypes. Mostly
external moulds, or composite moulds. All from the Bardahessiagh Formation, south of Craigbardahessiagh,
Pomeroy, exact localities uncertain except IGS GU 3264, Mitchell Collection, from Mitchell’s Bardahessiagh
collecting area (Mitchell 1977, p. 5).
Measurements Angle between antero-
and postero-dorsal
L
H
H/L
AL/L
margins
Max.
20-0 mm
12-6 mm
0-69
0-40
c. 150°
Min.
13-0 mm
8 0 mm
0-50
0-26
c. 125°
compressed
dorso-ventrally
Mean
16-24 mm
9-86 mm
0-60
0-34
Median
16-5 mm
1 1 -0 mm
0-60
0-33
TUNNICLIFF: LATE ORDOVICIAN BIVALVES
85
Description. Lyrodesma of elongate, ovate form with height-length ratio about 0-6. Maximum inflation of a
single valve is about 2 mm in a valve 20 mm long. The umbo lies at about the anterior one-third. The anterior
margin is rounded, the posterior end is subtruncate and obliquely carinate. Postero-dorsal surface bearing
17+ radial striations of variable strength and arranged somewhat irregularly. Anterior portions of the surface
show very faint concentric growth lines. Umbones rather small. Six radiating crenulate teeth of similar size
below umbo in left valve. A faint line or ridge is seen on the internal mould running from the umbo to the
ventral margin anterior to the umbonal ridge. Posterior and anterior adductor muscle scars situated dorsally,
the posterior scar being a little larger than the anterior. Accessory muscle scars poorly discerned in the present
specimen.
Discussion. As Reed noted (1952, pp. 67-68) Portlock’s description and figure of Lyrodesma radiatum
were poor, but Reed saw only two of the Portlock syntypes and had no specimens showing the
dentition. In shape and the number of teeth, L. radiatum resembles L. majus (Ulrich) and Reed
likened it to L. cincinnatiense Hall (of Ruedemann 1926) and to L. poststriatum elongatum Stewart,
1920, but until more material is available to provide further detail of the internal structure, no
close comparison can be made with other species.
Class rostroconchia Pojeta, Runnegar, Morris and Newall, 1972
Order conocardioidea Neumayr, 1891
Superfamily conocardiacea Miller, 1889
Family hippocardiidae Pojeta and Runnegar, 1976
Genus hippocardia Brown, 1843
Type species. By monotypy Cardium hibernicum Sowerby, 1815. The classification used here is that of Pojeta
and Runnegar 1976.
Hippocardia praepristis (Reed), 1952
Plate 9, fig. 8
v* 1952 Conocardium praepristis Reed, p. 80, pi. 4, fig. 11.
v. 1976 Hippocardia praepristis (Reed); Pojeta and Runnegar, p. 76.
Type specimens. Holotype by original designation, IGS GSM 24147, the only specimen known to Reed and
to the present author. The specimen is part of the Portlock Collection (Tunnicliff 1980) and bears a label
reading ‘Cypricardia— Analogous to C. cymbaeformis but differs from it and also from Cardium carpo-
morphum (J. P.)’ but Portlock (1843) published no reference to it.
Locality and horizon. The horizon of this specimen is uncertain, but Reed likened the matrix to that of the
Tirnaskea Beds (Tirnaskea Formation). If this is correct, a likely locality would be the Tirnaskea Stream
section or perhaps the Crocknagargan Stream section, both of which have exposures of Tirnaskea Formation
(Ashgill, Hirnantian) but there is no record of Portlock having specimens from the latter and there is
evidence that he had none from the Tirnaskea Stream section (Portlock 1843, p. 230 . .on the Tirnaskea
or small river there are thin calcareous layers of three or four inches thick mixed with quartzose bands,
no fossils, however, occurring in them . . .’). Lithologically, the specimen resembles known Killey Bridge
Formation specimens, and it must be assumed that the specimen is from that horizon, but the locality is
unknown.
Description. Adequate description is provided by Reed (1952), and Pojeta and Runnegar (1976)
placed praepristis in Hippocardia on this basis. The specimen is figured here to supplement Pojeta
and Runnegar.
Acknowledgements. Specimens were made available through the kindness of the following: Dr. N. J. Morris,
Mr. R. J. Cleevely, and Mr. R. P. Tripp (Brit. Mus. Nat. Hist.), Mr. P. Doughty and Mr. K. James (Ulster
Museum, Belfast), Dr. C. O’Riordan (Nat. Mus. Ireland, Dublin), Miss V. Burns (Trinity College, Dublin),
and Mr. M. Dorling (Sedgwick Museum, Cambridge). Dr. J. Kriz helped with Barrande type specimens
and discussion. Dr. Morris and Drs. D. E. Butler, A. W. A. Rushton, and D. E. White (IGS) also discussed
PALAEONTOLOGY, VOLUME 25
aspects of the work and gave advice. Drs. Butler, Rushton, and White read the manuscript critically at various
stages of preparation. Dr. J. Pojeta and Professor P. Bretsky kindly provided copies of papers. Text-fig. 1 is
reproduced by permission of the Trustees, Ulster Museum. This paper is published with the permission of the
Director, Institute of Geological Sciences.
REFERENCES
allen, j. a. 1978. Evolution of deep-sea protobranch bivalves. Phil. Trans. R. Soc. Lond. B. 284, 387-401.
— and Sanders, h. l. 1973. Studies on deep-sea protobranchia (Bivalvia); the families Siliculidae and
Lametilidae. Bull. Mus. comp. Zool. Harv. 145 (6), 263-310.
babin, c. 1966. Mollusques bivalves et cephalopodes du Paleozoique armoricain. 472 pp„ 18 pis. Brest.
— and melou, m. 1972. Mollusques bivalves et brachiopodes des ‘Schistes de Raguenez’ (Ordovicien
superieur du Finistere); consequences stratigraphiques et paleobiogeographiques. Ann. Soc. geol. N. 92 (2),
79-94, pis. 7-10.
— and robardet, M. 1973 (February). Quelques palaeotaxodontes (Mollusques bivalves) de l’Ordovicien
Superieur de Saint-Nicolas de Pierrepont (Normandie). Bull. Soc. geol. miner. Bretagne, 1972, (C), 4 (1),
25-38, pis. 1-3.
baily, w. H. 1875. Figures of characteristic British fossils, 1, Palaeozoic, lxxx. 126 pp., 42 pis. London.
— 1878. In Nolan, J., Explanatory memoire to accompany sheet 34 of the maps of the Geol. Surv. of Ireland.
Mem. geol. Surv. G.B. Dublin and London.
barrande, j. 1881. Systeme silurien de la Boheme, 4, Mollusques Acephales. 340 pp., 361 pis. Prague and
Paris.
billings, E. 1858. New genera and species of fossils from the Silurian and Devonian formations of Canada.
Can. nat. and Geol. 3, 419-444, 24 text-figs.
blake, j. f. 1882. A monograph of the British fossil cephalopoda, Pt. 1. Introduction and Silurian species,
248 pp., 31 pis.
bradshaw, M. a. 1 970. The dentition and musculature of some Middle Ordovician (Llandeilo) bivalves from
Finistere, France. Palaeontology, 13, 623-645.
bretsky, p. w. and bretsky, s. s. 1977. Morphological variability and change in the Palaeotaxodont bivalve
mollusk Nuculites planulatus (Upper Ordovician of Quebec). J. Paleont. 51, 256-271, 1 pi.
carter, r. M. 1971. Revision of Arenig Bivalvia from Ramsey Island, Pembrokeshire. Palaeontology, 14,
250-261, pis. 38-39.
conrad, T. A. 1838. Report on the Palaeontological Department of the Survey. A. Rep. geol. Surv. N.Y., 2,
107-119.
— 1841. Fifth annual report on the palaeontology of the strata of New York. Ibid. 5, 25-57.
donald, J. 1902. On some of the Proterozoic gastropoda which have been referred to Murchisonia and
Pleurotomaria Q. Jl. geol. Soc. Lond. 58, 313-339, pis. 7-9.
— ( = longstaff, j.) 1924. Descriptions of gastropoda etc. Ibid., 80, 408-446, pis. 31-38.
fuchs, A. 1919. Beitrag zur Kenntnis der Devonfauna der Verse- und der Hobracker Schichten des
sauerlandischen Faciesgebietes. Jahrbuch der Preufiischen geol. Landesanstalt (1918), 39 (1), 58-95,
pis. 5-9.
hall, j. 1847. Natural history of New York: Palaeontology Vol 1, 338 pp., 87 pis.
— and whitfield, r. p. 1875. Descriptions of Silurian Fossils. Geol. Surv. Ohio, 2 (2), 65-161, pis. 1-9.
havlicek, v. and vanek, j. 1966. The biostratigraphy of the Ordovician of Bohemia. Sbornik geologickych
ved, paleontologie, rada P. svazek 8, 7-69, pis. 1-16. With Czech summary. [In English.]
hind, w. 1910. The lamellibranchs of the Silurian rocks of Girvan. Trans. R. Soc. Edinb. 47 (3), No. 18,
479-548, pis. 1-5.
isberg, o. 1934. Studien iiber lamellibranchiaten des Leptaenakalkes in Dalarna. 493 pp., 32 pis. Lund.
kriz, j. and pojeta, j. 1974. Barrande’s colonies concept and a comparison of his stratigraphy with the
modern stratigraphy of the Middle Bohemian Lower Palaeozoic rocks (Barrandian) of Czechoslovakia.
J. Paleont. 48 (3), 489-494, 2 text-figs.
lamont, a. 1946. Lamellibranchs from the Lower Drummuck Group (Ashgillian), Girvan, Scotland. Cem.
Lime Gravel , 20 (10) (April 1946), 364-366, 1 pi., 1 text-fig.
LAROQUE, a. and NEWELL, n. D. 1969. In MOORE, r. c. 1969.
LONGSTAFF, J. 1924. See DONALD, j.
mcalester, A. l. 1962. Mode of preservation in early Paleozoic pelecypods and its morphological and ecological
significance. J. Paleont. 36 (1), 69-73, pi. 16, 1 text-fig.
TUNNICLIFF: LATE ORDOVICIAN BIVALVES
87
mcalester, a. l. 1968. Type species of Paleozoic Nuculoid bivalve genera. Mem.geol. Soc. Am. 105, i-ix, 143 pp.,
36 pis.
1969. In moore, r. c. 1969.
MCCOY, F. 1846. A synopsis of the Silurian fossils of Ireland. 72 pp. 5 pis. Dublin.
miller, s. A. 1874. Monograph of the lamellibranchiata of the Cincinnati group. Cincinnati Q. Jl. Sci. 1 (3),
211-231.
— 1877. American Palaeozoic fossils, a catalogue of the genera and species. 253 pp. Cincinnati.
— 1889. North American geology and palaeontology for the use of amateurs, students and scientists. 718 pp.
Cincinnati.
mitchell, w. i. 1977. The Ordovician Brachipoda from Pomeroy, Co. Tyrone. Palaeontogr. Soc. [ Monogr.\ ,
138 pp., 28 pis.
moore, R. c. ed., 1969. Treatise on invertebrate paleontology, pt. N. Mollusca 6, Bivalvia. 952 pp. Boulder.
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. London: John Murray, i-xxxii, 1-168, pis. 1-37.
pfab, L. 1934. Revision der Taxodonta des bohmischen Silurs. Palaeontographica, A, 80, 195-253, 3 pis.
pojeta, J. 1962. The pelecypod genus Byssonychia as it occurs in the Cincinnatian at Cincinnati, Ohio.
Palaeontogr. am. 4 (30), 169-216, pis. 22-31.
— 1966. North American Ambonychiidae (Pelecypoda). Ibid. 5 (36), 129-241, pis. 19-47.
1971. Review of Ordovician Pelecypods. Prof. Pap. U.S. geol. Surv. 695, 46 pp., 20 pis.
— 1978. The origin and early taxonomic diversification of pelecypods. Phil. Trans. R. Soc. Lond., B. 284,
225-246, pis. 1-15.
— and gilbert-tomlinson, j. 1977. Australian Ordovician pelecypod molluscs. Bull. Bur. Miner. Resour.
Geol. Geophys. Aust. 174 (6), 64 pp., 29 pis.
— kriz, j. and berdan, J. M. 1976. Silurian-Devonian Pelecypods and Palaeozoic Stratigraphy of subsurface
Rocks in Florida and Georgia and Related Silurian Pelecypods from Bolivia and Turkey. Prof. Pap. U.S.
geol. Surv. 879, 32 pp., 5 pis.
— and runnegar, b. 1976. The Palaeontology of rostroconch mollusks and the early history of the phylum
Mollusca. Ibid. 968, pp. 1-88, pis. 1-54.
portlock, j. e. 1837. In Larcom, T. Ordnance survey of the County of Londonderry, vol. 1 (with a preface
by Colby, T.). Notices i (at rear of volume), pp. 3-6, pis. 1-3.
— 1843. Report on the geology of Londonderry and parts of Tyrone and Fermanagh, i-xxxi, 784 pp., pis.
1-38, A-I, map. Dublin and London.
reed, F. R. c. 1944. Some new Ordovician lamellibranchs from Girvan. Ann. Mag. nat. Hist. Ser. 11,11 (76),
xxiii, pp. 209-215, pi. 2.
— 1952. Revision of certain Ordovician fossils from County Tyrone. Proc. R. Ir. Acad. 55 B (3), 31-136,
pis. 1-5.
sharpe, d. 1853. Description of the new species of zoophyta and mollusca. [Appendix B of ribeiro, c. On
the Carboniferous and Silurian Formations of the neighbourhood of Bussaco in Portugal.] Q. Jl. geol. Soc.
Lond. 9, 146-158, pis. 7-9.
soot-ryen, H. 1964 (1965). Nuculoid pelecypods from the Silurian of Gotland, Arkiv for miner, geol. 3 (28),
489-519, pis. 1-5.
sowerby, j. de c. 1839. In murchison, 1839.
tunnicliff, s. p. 1980. A catalogue of the Lower Palaeozoic fossils in the collection of Major General J. E.
Portlock, R.E., LL.D., F.R.S., F.G.S. etc. 112 pp., 5 figs. Ulster Museum Publication.
ulrich, e. o. 1892. New Lower Silurian lamellibranchiata chiefly from Minnesota rocks. Minnesota geol. and
nat. Hist. Surv. 19, 211-248.
— 1893 [1895]. New and little known Lamellibranchiata from the Lower Silurian rocks of Ohio and adjacent
states. Ohio Div. geol. Surv. Rept. Inv. 7 (2), 627-693, pis. 45-56.
— 1894. The Lower Silurian Lamellibranchiata of Minnesota: from vol. 3 of the Final Rept., Minnesota
geol. and nat. Hist. Surv. 475-628, pis. 35-42. (Published under separate cover prior to the entire v. 3: fide
Pojeta, 1971, p. 41.)
Watkins, R. 1978. Bivalve ecology in a Silurian shelf environment. Lethaia , 11, 41-56.
whiteaves, J. F. 1897. The fossils of the Galena-Trenton and Black River formations of Lake Winnipeg and
its vicinity. Geol. Surv. Can. Palaeozoic Fossils, 3 (3), 129-242, pis. 16-22.
whitfield, r. p. 1893 [1895], Contributions to the palaeontology of Ohio. Ohio Div. geol. Surv. 7 (2), 407-454,
pi. 13.
PALAEONTOLOGY, VOLUME 25
WILLIAMS, A., STRACHAN, I., BASSETT, D. A., DEAN, W. T., INGHAM, J. K., WRIGHT, A. D., and WHITTINGTON, H. B.
1972. A correlation of Ordovician rocks in the British Isles. Geol. Soc. Lond., Spec. Rep. 3, 74 pp.
williams, h. s. and breger, c. l. 1916. The fauna of the Chapman Sandstone of Maine, including descriptions
of some related species from the Moose River sandstone. Prof. Pap. U.S. geol. Surv. 89, 347 pp., 27 pis.
S. P. TUNNICLIFF
Palaeontology Unit
Institute of Geological Sciences
Exhibition Road
London SW7 2DE
Typescript received 26 February 1980
Revised typescript received 27 November 1980
JUVENILE SPECIMENS OF THE
ORNITHISCHIAN DINOSAUR PSITTACOSAURUS
by WALTER P. COOMBS, JR.
Abstract. Hitherto undescribed specimens of Psittacosaurus mongoliensis from the Oshih Fm., Mongolian
Peoples’ Republic, include two almost complete skulls and numerous postcranial elements. A rostral bone,
present in these and other specimens of Psittacosaurus, is a cranial element otherwise known only in
ceratopsians, and its presence indicates a sister group relationship for the Psittacosauridae and Ceratopsia.
Each of the two new specimens of Psittacosaurus is a juvenile, and are among the smallest dinosaur specimens
yet described. Parental attendance of nests was common, possibly universal among dinosaurs, but post-hatching
parental care is uncertain. Juveniles of Psittacosaurus, as well as those of other dinosaurs, may have formed
sibling groups.
Two skeletons of small bipedal ornithischians collected by the Third Asiatic Expedition (1922) of
the American Museum of Natural History were described by Osborn (1923, 1924) as the types of
Psittacosaurus mongoliensis (Oshih Fm.) and Protiguanodon mongoliense (Andai Sair Fm.), thought
to be a primitive ankylosaur and a primitive iguanodontid, respectively. To the family Psitta-
cosauridae, erected for reception of these two species (Osborn 1923), have been added several new
species of Psittacosaurus and additional fragmentary remains from various localities in Mongolia
and northern China (Young 1931, 1958; Bohlin 1953; Maleev 1954; Rozhdestvensky 1955; Chao
1962). Described below are two additional specimens of Psittacosaurus from the same locality as
the type of P. mongoliensis.
MATERIAL
AMNH 6535 (= American Museum of Natural History, New York), partial skull and jaws.
AMNH 6536, almost complete skull and jaws with numerous distarticulated postcranial elements including:
cervical, dorsal, and caudal vertebrae; ribs; scapulae; coracoid; partial ilium and ischium; humeri; femora;
tibiae; fibulae; and an almost complete left pes. The postcranial material belongs to several individuals (there
are fourteen distal ends of tibiae) of at least two different sizes (some femoral fragments are tiny and more
compatible in size with the skull AMNH 6535). Rather than allot separate numbers to every element, all of
the postcranial material is arbitrarily assigned to AMNH 6536.
Locality. Both specimens come from the Oshih Fm., Artsa Bogdo basin, western Mongolian Peoples’
Republic, and were apparently collected at the same time and in the same place as the type of Psittacosaurus
mongoliensis (Osborn 1923, 1924).
DESCRIPTION
Skull, general. In lateral view the skull is arched dorsally, truncated posteriorly, and has a short snout (text-figs.
1, 3, 4). AMNH 6536 is laterally crushed, but AMNH 6535 retains the characteristic triangular shape of
psittacosaur skulls, with the greatest breadth across the flaring jugals and a sharply pointed apex at the beak
(PI. 14). The apparent relatively large size of the brain in AMNH 6535 is a juvenile character (PI. 14,
figs. 1 and 2). The diameter of the orbit equals 33 % of cranial length in AMNH 6536; 38 % in AMNH 6535,
the relatively large orbit also being a juvenile characteristic. The nostril is clearly delineated only on the right
side of AMNH 6536, on which it is small, circular, and approximately at the level of the upper half of the orbit.
Cranial fenestrae. Both skulls are damaged posteriorly and borders of all temporal fenestrae are incomplete.
The left supratemporal fenestra of AMNH 6536 is oval with its long axis paralleling the midline. The braincase
is broad and devoid of a sagittal crest between the upper fenestrae (PI. 14). Lateral temporal fenestrae are
IPalaeontology, Vol. 25, Part 1, 1982, pp. 89-107, pi. 14.)
90
PALAEONTOLOGY, VOLUME 25
evidently tall and narrow, and considerably smaller than the orbit, unlike the condition in adult Psittacosaurus
in which the fenestrae are larger than the orbit (e.g. Osborn 1923, fig. 2). AMNH 6536 has a shallow depression,
its depth somewhat exaggerated by crushing, between nostril and orbit in the region where an antorbitai
fenestra might be located, but there is no opening into the interior of the skull (text-fig. 1). AMNH 6535 is
badly crushed in the immediate pre-orbital region, but it appears that a true antorbitai fenestra was not present.
text-fig. 1. AMNH 6536, skull, lower jaws, and anterior cervical vertebrae; right
side with key drawing. Abbreviations: at, atlas; ax, neural arch of axis; cax,
centrum of axis; d, dentary; en, external naris; f, frontal; j, jugal; l, lacrimal;
mx, maxilla; n, nasal; nax, neural spine of axis; pa, parietal; pd, predentary;
pf, prefrontal; pm, premaxilla; po, postorbital; pp, palpebral; q, quadrate; r, rostral;
ra, retroarticular process; sq, squamosal; and v, cervical vertebrae (post-atlas).
Length of reference line = 1 0 cm (approximately twice natural size).
COOMBS: JUVENILE DINOSAURS
91
Rostral. Following Maryariska and Osmolska (1975, p. 172), the most anterior element of the snout is identified
as a rostral, a bone diagnostic of ceratopsians, rather than as a premaxilla as suggested by Osborn (1923; see
also Young 1958; Chao 1962). The rostral, best preserved on the smaller skull (text-fig. 4), is roughly triangular
in lateral view with an ascending ramus that curves posteriorly to terminate near the ventral margin of the
nostril. Most of the cutting margin of the beak is composed of the rostral, but it is unclear to what extent
the rostral contributes to the anterior palatal shelf.
Premaxilla. Exact limits of the premaxillae are unclear, but the bone probably intervenes between the rostral
and the external nares and forms all the ventral border of the latter opening (text-fig. 3; Maryanska and
Osmolska 1975). Because other authors have identified the rostral as the premaxilla, the true premaxilla has
been identified as part of the maxilla. In consequence, Psittacosaurus has been reconstructed with a very large
maxilla that separates the premaxilla from the lacrimals as well as from the margin of the nostril (Osborn
1923, fig. 2a; Young 1958, fig. 51; Chao 1962, fig. 1) a pattern atypical for ornithischians. The alternative
interpretation presented here (text-fig. 3) indicates extensive contact of premaxilla and lacrimal with exclusion
of the maxilla from the border of the nostril, a pattern common among ceratopsians and ornithischians in
general (Lull 1933; Romer 1956). Ornithischians typically have a bony roof to the anterior part of the buccal
cavity composed of premaxillae, maxillae, and sometimes the vomers. In the Ceratopsia the rostral bone
forms much of the cutting edge of the beak, but forms only a small segment of the anterior palatal shelf
(Lull 1933, figs. 5 and 30; Maryanska and Osmolska 1975, fig. 9). Psittacosaurus may have a similar arrange-
ment, but sutures are unclear on the present specimens (PI. 14, figs. 3 and 4).
Maxilla. Anteriorly the maxilla may contact the posterior limit of the rostral along the edge of the mouth
in the present specimens, but contact is lost in adult Psittacosaurus. Maxillary teeth, numbering at least five in
AMNH 6535 and at least six in AMNH 6536, are withdrawn medially, so that the maxilla forms a lateral shelf
thus delineating the dorsal border of a ‘cheek-pouch’, a common feature among ornithischians (Galton 1973).
Jugal. All the ventral margin of the orbit is formed by the jugal, the suborbital bar being relatively slender
in AMNH 6535, considerably wider and giving an impression of massiveness in AMNH 6536, but in neither
skull is the jugal as wide as in adult Psittacosaurus. A ventro-laterally projecting jugal spine, located below
the lateral temporal fenestra, is present on the left jugal of AMNH 6536 (text-fig. 2), but is considerably
smaller than the prominent flange of adult Psittacosaurus skulls (Osborn 1923, 1924; Young 1958; Chao
1962). A slender ascending ramus of the jugal forms part of the postorbital bar. A quadratojugal presumably
intervened between the jugal and the ventral end of the quadrate, but the element cannot be distinguished
on either of the present specimens.
Quadrate. The elongate quadrate curves anteriorly from its dorsal, squamosal articulation and is two to three
times wider than the postorbital bar (text-figs. 2 and 3). The mandibular cotylus is compressed antero-
posteriorly and has a bulbous lateral region adjacent to a narrower medial area (PI. 14, figs. 3 and 4). The cotylus
projects below the level of the tooth row, as in most ornithischians.
Skull roof. The nasals presumably form the narrow bar between the nostrils as well as the wider region
immediately posterior to them, but sutures of the nasals with rostral, maxillae and premaxillae are not visible
on either specimen. Chao (1962) illustrates P. youngi with the nasals excluded from the narial border (text-fig.
7c), but this is not true of the type of P. mongoliensis (text-fig. 7d). A prefrontal nestles into the antero-dorsal
margin of the orbit excluding the nasal from the orbital rim (right side of AMNH 6536, text-figs. 1 and 3).
Ventrally the prefrontal contacts the lacrimal and premaxilla, but sutures are indistinct. A displaced palpebral
(‘supraorbital’ of Osborn 1923; ‘prefrontal’ of Chao 1962; see Coombs 1972) originally articulated with the
prefrontal at the anterior rim of the orbit (text-figs. 1 and 3). The palpebral has an expanded base and a
tapered posterior extension that is incompletely preserved. Posterior to the indistinct contact of nasals and
frontals above the orbits the skull roof widens and the postorbital contacts the lateral edge of the frontal. A
slender descending rod of the postorbital bone forms most of the postorbital bar. The relative contributions
of the postorbitals, frontals, and parietals to the margins of the supratemporal fenestrae cannot be determined.
Parietals form most of the arched posterior skull roof, without the sagittal crest present in adult skulls (see
also skull of Psittacosaurus osborni Young, 1931, fig. 2). The parietals bend sharply downward at the posterior
margin of the skull roof to form the most dorsal surface of the occipital region.
Occipital area. Very little of this region is exposed. The rugose distal tip of the left opisthotic of AMNH
6536 projects from matrix well below the level of the skull roof and the dorsal tip of the quadrate (text-
fig. 2). The element is roughly triangular in section, not as flattened as is typical for Ornithischia.
92
PALAEONTOLOGY, VOLUME 25
Palate. Some palatal details are visible on the smaller skull (PI. 14, figs. 3 and 4). An anterior palatal shelf
is formed partly of the rostral and partly of the premaxillae, and is roughly semicircular in shape, being less
pointed and less triangular than in adult Psittacosaurus. The vomers form a median vertical plate that
dominates the interior palatal region. The ventral keel of this plate extends below the level of the tooth rows,
a condition found in several ornithischians (e.g. the ankylosaur Panoplosaurus; Russell 1940). A narrow shelf
from the maxillae appears to contact the vomer keel anteriorly, thus shifting the internal nares posteriorly.
The internal nares are bounded laterally and anteriorly by maxillae, medially by the vomer keel, and are
bounded posteriorly by a broad plate of bone that slants obliquely anteriorly and upward toward the skull
text-fig. 2. AMNH 6536, skull, lower jaws and anterior cervical vertebrae; left
side with key drawing. Abbreviations: cat, centrum of atlas; cax, centrum of axis;
d, dentary; j, jugal; mrj, medial side of right lower jaw; mx, maxilla; o, opisthotic;
pd, predentary; pm, premaxilla; po, postorbital; q, quadrate; sp, spine of jugal;
v, cervical vertebra (post-atlas). Length of reference line = 10 cm (approximately
twice natural size).
COOMBS: JUVENILE DINOSAURS
93
text-fig. 3. Restoration of a juvenile Psittacosaurus skull based primarily on
AMNH 6536. Abbreviations: d, depression in premaxilla; f, frontal; j, jugal;
l, lacrimal; mx, maxilla; n, nasal; o, opisthotic; pd, predentary; pf, prefrontal;
pm, premaxilla; pp, palpebral; q, quadrate; r, rostral; and s, spine of jugal.
roof. The latter plate is composed of pterygoid, ectopterygoid, and palatine bones, but sutures are unclear.
The palatal structure of Psittacosaurus is similar to that of many Ornithischia, especially the quadrupedal
forms, and may be compared particularly to the palate of Bagaceratops (Maryanska and Osmolska 1975,
fig. 9).
Basicranium. The partially exposed basicranium of AMNH 6535 (PI. 14, figs. 3 and 4) is broad relative to
other skull dimensions (a juvenile feature), and has a median longitudinal depression that is flanked by
anteriorly converging ridges. A small, subspherical, posteriorly directed occipital condyle projects only slightly
below the level of the basicranial floor.
Lower jaw. The deep, massive lower jaw has a straight ventral margin and medially displaced dentary teeth
(text-fig. 1). A curved shelf marks the ventral border of the cheek pouch. The predentary and dentary bones
are fused together. Posteriorly there is a pointed coronoid process projecting upward medial to the most
anterior ventral corner of the lateral temporal fenestra. A short retroarticular process projects backward from
the articular (text-fig. 1).
Teeth. Anterior maxillary teeth have three ridges on the lateral surface that upon wear produce the ‘trilobate’
cutting margin described by Osborn (1924; text-fig. 2). Posterior maxillary teeth may have four ridges and
are generally larger but less worn than anterior teeth. All teeth that are adequately exposed on the two skulls
are worn to some degree. There are no premaxillary teeth.
Vertebrae. Cervical vertebrae attached to the skull of AMNH 6536 include the axis, which has a low, laterally
compressed, elongate neural spine as described in the type of Protiguanodon mongoliense (AMNH 6253;
Osborn 1924). A small block of bones (part of AMNH 6536) contains several vertebrae, mostly dorsals
(text-fig. 5). The centra are laterally compressed, with expanded, roughly heart-shaped articular ends. The
neural arches are not fused to the centra, a juvenile characteristic. A particularly well-preserved neural arch
of a dorsal(?) vertebra has about the same length as its centrum; the incomplete zygapophyses arch upward
and there is an almost circular diapophysis high on the side of the neural arch. A sacral vertebra in the same
block has a centrum similar in shape to dorsal centra, and has a low, broad neural arch with an elongate,
laterally compressed neural spine most of which is broken off. The sacral rib articulated on anteroposteriorly
compressed diapophyses positioned at the extreme anterior end of this sacral, which judged from the sacral
94
PALAEONTOLOGY, VOLUME 25
text-fig. 4. AMNH 6535, skull, partial lower jaw, and fragmentary anterior
cervicals, right side. Length of reference line = 10 cm (approximately twice
natural size).
series of Protoceratops (Brown and Schlaikjer 1940), is from the middle of the sacral series. Another sacral
vertebra in the block also has an elongate, compressed diapophysis but it is oriented almost horizontally,
suggesting that this vertebra belongs to the posterior end of the sacral series. A group of three caudal centra
(text-fig. 6n), all lacking the neural arch, are smaller than either dorsals or sacrals but have similar shape and
proportions. Transverse processes are not present on any of the caudals. The largest of these three caudal
centra is about 4-5 mm in length.
Ribs. Rib fragments embedded in the block with other postcranial elements have cross sections that are either
circular (proximal ends) or compressed (distal ends, text-fig. 5). One proximal end fragment has a large,
almost circular head that matches in size and shape the articular surface on the dorsal vertebra described
above, and a very small tubercle positioned a short distance from the head.
Pectoral girdle. Scapulae are represented by two almost complete elements (text-fig. 5), plus a partial scapular
blade. The scapulae are elongate and narrow, with a much wider, slightly concave area adjacent to the glenoid
and coracoid articulation. The upper end of the blade is broad but flattened. A low scapular spine arises
from the extreme anterior edge of the scapula directly opposite the glenoid. The glenoid articulation is short,
broad, slightly concave, and has a rather massive dorsal lip. The coracoid articulation is elongate and
EXPLANATION OF PLATE 14
Two views of AMNH 6535, a skull, partial lower jaw and fragments of anterior cervical vertebrae.
Fig. 1. Dorsal view, stereo pair.
Fig. 2. Key to structures visible in the dorsal view. Stippling indicates the approximate extent of a natural
endocranial cast that has been partially exposed through loss of some dorsal cranial bones. The anterior
limit of the endocranial space was determined in part by examination of AMNH 6536 which has the
anterior tip of a natural endocranial cast exposed through loss of some skull bones.
Fig. 3. Ventral view, stereo pair.
Fig. 4. Key to structures visible in ventral view. Abbreviations: at, atlas; ax, axis; bo, basioccipital; c, left
coronoid process (protruding through matrix); f, frontal; j, jugal; l, lacrimal; lj, lower jaw; m, maxilla;
n, nasal; o, opisthotic; pm, premaxilla; po, postorbital; pt, pterygoid; q, quadrate; r, rostral; t, tooth. Length
of reference line = 1-0 cm (approximately twice natural size).
PLATE 14
COOMBS, Psittacosaurus
96
PALAEONTOLOGY, VOLUME 25
text-fig. 5. AMNH 6536, three views of a block containing various postcranial elements, with key
drawings. Abbreviations: c, centrum; f, femur; il, ilium; is, ischium, n, neural arch, r, rib; sc, scapula;
sr, sacral rib; and a question mark denoting an unidentified fragment. Length of reference line =10
cm (approximately twice natural size).
text-fig. 6. AMNH 6536, various postcranial elements, a, partial left pes with distal ends of tibia and fibula;
b, posterior view of a partial left tibia with astragalus and metatarsals II, III, and IV; c, same as b in anterior
view; d, same as b in distal view; E, left humerus, flexor surface; f, right coracoid, lateral view; G, H, i, and
J, distal ends of four femora, posterior views; K, L, and m, proximal ends of three tibiae, anterior views;
N, block containing three caudal centra; o and p, proximal ends of two small femora, anterior views; Q, R, s,
and t, distal ends of four tibiae, anterior views, with distal ends of fibulae on r and T. Scale divisions in mm
(approximately twice natural size).
PALAEONTOLOGY, VOLUME 25
triangular, exactly matching an associated left coracoid (text-fig. 6f). The coracoid has a broad, slightly
concave glenoid and a foramen that is roughly equidistant from the glenoid and scapular articulations. There
is a prominent biceps tuberosity on the anterior edge of the coracoid, from which a ridge extends posteriorly
toward the glenoid. Below the biceps tuberosity the coracoid has a shallowly concave, medially inclined
surface that is the probable region of coracobrachialis origin. Clavicles have been reported for Psittacosaurus
mongoliensis (Osborn 1924) and P. sinensis (Young 1958), but there are no clavicles associated with the
scapulae of the present specimens (text-fig. 5). In view of the jumbled condition of the bones, the clavicles
could well be displaced and easily mistaken for rib fragments.
Pelvic girdle. Of the pelvic girdle there is one fragmentary piece each of ilium and ischium (text-fig. 5). The
ilial fragment is flat with a blunt, ventrally projecting ischial peduncle and a laterally compressed, rather thin
postacetabular segment. Only the base of the pubic peduncle is preserved. The ischial fragment includes an
elongate, laterally compressed pubic articulation, and an ilial articulation that appears to be more compact
and rounded, although most of the latter is hidden below the ilial fragment. The ischial blade slants posteriorly
and is narrow and laterally compressed. Most of the distal part of the ischium is missing.
Forelimb. In addition to a perfect left humerus (text-fig. 6e), there are two proximal ends of right humeri,
one about the same size as the left humerus and one considerably smaller. Proximally, the humerus is broad,
but it tapers quickly to a roughly circular shaft, then widens at the distal articulations which are rotated
relative to the head such that the extensor surface is twisted medially. A deltopectoral crest projects ventrally
and is rather close to the proximal end of the humerus. No other forelimb elements have been identified.
Hindlimb. There are numerous fragments, but unfortunately there is no association of a particular femur with
a particular tibia, nor with the pes described below, so that hindlimb proportions cannot be established. The
proximal end of a relatively large femur has a compressed, crest-like greater trochanter with an adjacent,
finger-like lesser trochanter separated from the former by a deep furrow, and an irregularly circular shaft
(text-fig. 5). Two much-smaller femora have a virtually identical set of trochanters and a medially displaced
head that is as much cylindrical as spherical (text-fig. 6o, p). The femoral shaft is curved with the convex
surface lateral, a common shape for femora of bipedal Ornithischia. None of the eight distal femoral pieces
included in AMNH 6536 can be fitted on to any of the proximal ends. Distal articulations follow the usual
bipedal ornithischian pattern of a larger lateral and smaller medial articular surface (text-fig. 6g, h, i, j).
There are seven proximal and fourteen distal ends of tibiae (text-fig. 6k, l, m, q, r, s, t). Proximally, the
tibia has an antero-posteriorly elongated articular surface that has a smoothly curved medial surface and a
notched lateral surface that gives the proximal end a trilobate shape viewed end-on. The tibial shaft tapers
to a slender rod of almost circular cross-section. Distally, the tibia is strongly compressed antero-posteriorly,
creating a wide articular surface that is a little thicker medially and has a slight depression near the centre
for reception of the astragalus. There is a slight ventral projection at the lateral corner of the distal end,
marking the fibular articulation. Fibular fragments still attached to two of the tibiae (text-fig. 6r, t) are oval
in section, only a fraction of the diameter of the tibial shaft, and fit into a shallow depression on the
antero-lateral edge of the distal end of the tibia.
Pes. The astragalus is compressed against and covers about two-thirds of the distal end of the tibia (text-fig.
6a, b, c, d) leaving a small space laterally for the calcaneum. The lateral edge of the astragalus is notched
for reception of the calcaneum, and anteriorly the astragalus curves up on to the flexor surface of the tibia.
A roughly cup-shaped calcaneum nestles against the ventrally projecting lateral corner of the tibia (text-fig.
6a), and as preserved does not contact either the astragalus or fibula. A distal row of tarsals is not present.
As is typical of bipedal ornithischians, the metatarsals are long, slender, and fit tightly together, especially
at their proximal ends (text-fig. 6a, c, d). As in the types of Psittacosaurus mongoliensis and P. sinensis,
metatarsals I through IV are present, but there is no trace of metatarsal V nor is there an articular surface
for it on metatarsal IV. Metatarsal I is strongly compressed at the proximal end and through the upper two-
thirds of the shaft. Metatarsal II is about double the width of metatarsal I proximally, but is still somewhat
compressed. Metatarsal II has a wider shaft and is longer than metatarsal I. Metatarsal III has an irregularly
square proximal articular surface and is the longest, straightest, and most massive of the metatarsals. Metatarsal
IV is about the same length as metatarsal II, but is more compressed at the proximal end, and its shaft curves
slightly away from the axis of the pes running through metatarsal III. Phalanges have the typical form found
in small bipedal ornithischians (text-fig. 6a). All are relatively long (length greater than diameter) and are
circular to subrectangular in cross-section, not dorsoventrally compressed. One almost complete ungual of
digit III is slightly longer than the adjacent phalanx and has a narrow, semi-claw shape as in Leptoceratops
(Brown 1914; Brown and Schlaikjer 1942) rather than the broad, hoof-like unguals of Protoceratops (Brown
COOMBS: JUVENILE DINOSAURS
99
and Schlaikjer 1940). The pes is narrow, long, and gracile (text-fig. 6a), suggesting a relatively fleet animal.
Of interest is the manner in which the pes is folded against the anterior surface of the tibia in one specimen
(text-fig. 6b, c, d), a post-mortem posture also found in the types of Psittacosaurus mongoliensis, Protiguanodon
mongoliense, and Stenopelix valdensis (Osborn 1923, 1924; Meyer 1857; Koken 1887). When these animals
died, they settled on to their belly with the hindlimbs folded into a sharp ‘Z’, the knee projecting forward,
ankle backward, femur atop the tibia, and pes pressed against the anterior surface of the tibia and apparently
set flat against the ground (this posture is imperfectly preserved also in specimens of Psittacosaurus sinensis'.
Young 1958, figs. 52, 53, and 54). The imprint of the pes would conform to the sitting dinosaur trackways
from the Connecticut Valley (e.g. Sauropus barrattii and Anomoepus scambus specimens described by Lull
1953). The belly-down position is not common among bipedal ornithischians. The skeletons of Iguanodon
and most hadrosaurids are typically found lying on their side with the knee and ankle both flexed rearward,
or with the pes almost aligned with the tibia. Psittacosaurs apparently had great mobility of the knee and
mesotarsal joint, with hyperextensibility of the ankle permitting the entire length of the pes to be laid flat on
the ground. The difference in joint mobility in psittacosaurs and large bipeds may be related to total body
size, the larger ornithischians sacrificing some ankle extensibility for the sake of rigidity necessary for
weight-bearing.
DISCUSSION
Taxonomy. Rozhdestvensky (1955) regarded Psittacosaurus mongoliensis Osborn (1923) as the
senior synonym of the following: Protiguanodon mongoliense Osborn (1923 = Psittacosaurus
protiguanodonensis of Young 1958); Psittacosaurus osborni Young (1931); ? Psittacosaurus tingi
Young (1931); and Protiguanodon cf. mongoliense (of Young 1931). Trivial differences among these
species are reasonably regarded as stemming from ontogenetic state and individual variation.
Psittacosaurus sinensis Young (1958) and P. youngi Chao (1962) may also be junior synonyms of
P. mongoliensis although, without having all the specimens in hand for detailed comparison of
measurements and proportions not available in extant descriptions, formal suppression of all but
one species is premature. Stenopelix valdensis Meyer (1857; see also Koken 1887; Huene 1908) from
Wealden deposits (Swinton 1936) of northern Germany was originally classified in the Hypsilopho-
dontidae (Huxley 1870), but has also been suggested as a ceratopsian (Huene 1909; Lull 1910) and
a psittacosaur (Romer 1956). The reported absence of a postpubic process and the exclusion of
the pubis from the acetabulum in Stenopelix are features unknown in psittacosaurs, but have been
reported for the pachycephalosaur Homalocephale (Maryanska and Osmolska 1974), and Stenopelix
has been classified in the Pachycephalosauridae (Galton 1976; Olshevsky 1978). Contact of the
ischium with the pubic peduncle with resultant exclusion of the pubis from the acetabulum may
also be common among ankylosaurs, although obscured in most specimens by fusion of the pelvic
elements, and ankylosaurs have a very short postpubic process. In any case, assignment of Stenopelix
to the Psittacosauridae cannot be defended on the basis of common derived characters, and the genus
is here considered as Ornithischia, incertae sedis. The family Psittacosauridae is therefore regarded
as having a single genus, Psittacosaurus , with the type species P. mongoliensis, and other species
being of questionable validity.
AMNH 6535 and 6536 come from the same stratigraphic and geographic locality as the type
of Psittacosaurus mongoliensis. While the two new skulls differ in detail from the type of P.
mongoliensis, the differences are of a kind and magnitude that are reasonably expected in such
juvenile specimens, as may be appreciated by studying the hypothesized growth series (text-fig. 7).
AMNH 6535 and 6536 are therefore assigned to the species Psittacosaurus mongoliensis.
Phylogenetic position of the Psittacosauridae. Bipedality has long been the justification for inclusion
of several families, including the Psittacosauridae, within the suborder Ornithopoda, despite
recognized peculiarities and the absence of critical features. Maryanska and Osmolska (1974) argued
that bipedality alone was insufficient evidence for such a grouping, and removed pachycephalosaurs
from the Ornithopoda (see also Olshevsky 1978). If the general dogma that bipedality is primitive
(plesiomorphic) for the order Ornithischia is true (e.g. Galton 1978), then it is unsuitable as a
100
PALAEONTOLOGY, VOLUME 25
unifying character for the suborder Ornithopoda. Moreover, psittacosaurs, like pachycephalosaurs,
have no obturator process on the ischium, a feature apparently synapomorphic for the Ornithopoda
sensu stricto. Psittacosaurs have been recognized as ceratopsian-like by several authors (Rozhdest-
vensky 1955, 1960; Romer 1956; Gregory 1957; Young 1958; Colbert 1965; Maryanska and
Osmolska 1975) but Steel (1969) noted that the absence of premaxillary teeth removed Psittacosaurus
from the immediate ancestry of Protoceratops, and the peculiar structure of the Psittacosaurus
manus (Osborn 1924) is a further barrier to placing the genus in a position ancestral to any known
ceratopsian.
Maryanska and Osmolska (1975) recognized that the substantial similarity in general cranial
morphology shared by psittacosaurs and ceratopsians outweighed any arguments on the unsuit-
ability of the former as ‘ancestors’ for the latter. They also realized that failure to identify a rostral
bone in Psittacosaurus was the single major objection to transferring the Psittacosauridae to the
suborder Ceratopsia, and they boldly suggested that the snout of Psittacosaurus had been incorrectly
interpreted by Osborn (1923, 1924) and others (Young 1958; Chao 1962). Maryanska and Osmolska
(1975) contended that a rostral hone is in fact present in Psittacosaurus , an opinion with which I
fully concur (see also Olshevsky 1978).
The rostral of ceratopsians probably originated as an epidermally induced ossification, with the
large, overlying epidermal scale eventually becoming a thick, heavily keratinized upper beak. The
rostral and upper beak thus correspond to the predentary and associated beak of the lower jaw
in origin and function. In Psittacosaurus the rostral is ill defined and in large part fused to the
premaxillae, similar to the way in which epidermal ossifications overlie and are fused to the skull
roof of ankylosaurs (e.g. Maryanska 1971, 1977). The Psittacosaurus rostral is not, therefore, a
completely separate cranial element as it is in advanced ceratopsians, but it is nevertheless present.
Revised identifications of psittacosaur skull bones (text-figs. 3 and 7c) may be compared with
identifications in original descriptions.
The family Psittacosauridae is here regarded as belonging in the suborder Ceratopsia rather
than in the Ornithopoda, but further refining of its position among ceratopsians is difficult. Many
taxa included in the Protoceratopsidae share synapomorphies with the Ceratopsidae, for example
the development of a posterior parietal-squamosal frill and at least incipient development of
horns, features not present in psittacosaurs. On the other hand, absence of premaxillary teeth in
Psittacosaurus is possibly a derived character shared with the Ceratopsidae, but not with the
Protoceratopsidae some of which have premaxillary teeth. The peculiar manus of Psittacosaurus,
and the related bipedal stance, are unique developments among ceratopsians, but whether these
are derived characters defining the basal sister group split of the suborder, or merely retention of
primitive features by Psittacosaurus, is unclear. Potentially, at least, the Psittacosauridae could be
the sister group of all other Ceratopsia, such a position leaving uncertain the question as to whether
bipedal or quadrupedal posture was the primitive condition for the suborder.
Juvenile features. Differences between adult and juvenile ceratopsian skulls (Brown and Schlaikjer
1940; Maryanska and Osmolska 1975; Dodson 1976) provide a basis for recognition of juvenile
features in the two Psittacosaurus skulls under consideration. Among these features are the following:
orbit very large; braincase relatively large; skull roof more curved than in adult; snout short; rostral
approaches and may contact maxilla, thus excluding premaxilla from border of mouth; small,
narrow lateral temporal fenestrae; no sagittal crest; suborbital bar slender; and jugal flange small
or absent. General changes in cranial proportions can be seen in the hypothetical ontogenetic series
(text-fig. 7). Quantification of allometric relationships is not justified in view of the small sample
for each age category. Consideration of juvenile features in AMNH 6535 and 6536 serves less for
establishment of growth trends than for confirmation of these specimens as young Psittacosaurus
mongoliensis and not a new taxon of small dinosaur.
Smallest dinosaur. AMNH 6535 and 6536 are each smaller than any previously reported dinosaur
specimen. A listing of the smallest dinosaurs based on basal skull length yields the following
COOMBS: JUVENILE DINOSAURS
101
text-fig. 7. A hypothetical ontogenetic series of Psittacosaurus skulls based upon a, AMNH 6535; b, AMNH
6536; c, the type of P. youngi (after Chao 1962); and d, AMNH 6254, the type of P. mongoliensis (after
Osborn 1923). Readily observed are a decrease in relative size of the orbit, a lengthening of the snout (pre-orbital
region of skull), and a shift from a rounded to a flat-topped cranial contour. Revised identifications of some
cranial elements are given in c. Abbreviations: J, jugal; l, lacrimal; mx, maxilla; n, nasal; pm, premaxilla; pp,
palpebral; and r, rostral.
102
PALAEONTOLOGY, VOLUME 25
sequence: Lesothosaurus diagnosticus, 94 mm (Galton 1978; = Fabrosaurus australis of Thulborn
1970); Protoceratops andrewsi, 76 mm and 62 mm (respectively AMNH 6419 and ZPAL MgD II/7;
Brown and Schlaikjer 1940; Dodson 1976; Maryanska and Osmolska 1975); Compsognathus
longipes, 70 to 75 mm (Huene 1925; Ostrom 1978); Bagaceratops rozhdestvenskyi, 47 mm (ZPAL
MgD 1/123; Maryanska and Osmolska 1975); juvenile prosaurauropod(?), 32 mm (Charig 1979);
Psittacosaurus mongoliensis 42 mm and 28 mm (respectively AMNH 6536 and 6535). An estimate
of total body length for the two Psittacosaurus juveniles can be obtained by scaling down the
dimensions of the two large, almost complete skeletons described by Osborn (1923, 1924), using
comparisons of median skull length. The calculated snout to tail tip length for AMNH 6536 is
about 390 mm, and that for AMNH 6535 is about 265 mm. Possible allometric changes in skull
size relative to body length were not considered in these estimates but, since the skull is
proportionately larger in juveniles, the calculated total body lengths probably err on the high side.
Scaling the complete Psittacosaurus skeleton based on a comparison of humeral length (using the
humerus shown in fig. 5) gives an estimated total length for AMNH 6536 of 340 mm. Allowing
for some allometric changes in skull size, I would estimate the total body length of AMNH 6535
at about 230 mm. For comparison with the size of these two juvenile psittacosaurs, the famous
skeleton of Compsognathus longipes , also probably a juvenile, has a length of between 750 and
810 mm (based on restorations by von Huene 1925, 1926; Ostrom 1978). The smaller Psittacosaurus
specimen described herein (AMNH 6535) belonged to a dinosaur slightly smaller than a common
pigeon ( Columba livia).
Parental care. Recently it has become fashionable to ascribe mammalian or avian behaviour patterns
to dinosaurs despite the reptilian aspect of the brains of all dinosaurs except small theropods
(Hopson 1977) and the acknowledged difficulty of deducing habits from osteology even for living
animals. Among behavioural patterns considered for dinosaurs is parental care of young (e.g.
Horner and Makela 1979). In attempting an objective analysis of parental care in dinosaurs, three
points must be kept in mind. First, there is no such thing as an ‘accepted interpretation’. The
suggestion that dinosaurs abandoned their eggs as do modern chelonians is a hypothesis that
requires proof as rigorous as that required to prove that dinosaurs cared for their young in the
manner of modern ducks or ungulates. Casting aspersions on arguments favouring parental care
does not constitute proof of the alternate hypothesis. Second, the morphologic diversity and long
evolutionary history of dinosaurs makes it exceedingly unlikely that every species practiced a similar
amount of parental care. Indeed, the egg-laying pattern of the hadrosaur Rhabdodon has been
interpreted as precluding parental care (Ginsburg 1980) while nests of the hadrosaur Maiasaurus
have been interpreted as requiring parental care (Horner and Makela 1979). Dinosaur parental
behaviour should be approached on a case-by-case basis, and generalities applicable to all dinosaurs
will necessarily be few. Finally, the selection of modern analogues to interpret dinosaur behaviour
is a subtle trap that encourages interpretations far beyond what can reasonably be concluded from
actual data. This problem has been discussed elsewhere (Coombs 1975), but the warning merits
repetition. The mental image of the relationship between adult and juvenile dinosaur is greatly
altered if described as being: (1) like elephants; (2) like opossums; (3) like ducks; (4) like crocodiles;
or (5) like mouth-breeding fishes. Use of modern analogues must be approached with great
caution.
In the absence of contrary evidence it is assumed that most dinosaurs laid eggs (Hopson 1977).
Coelophysis specimens from Ghost Ranch, New Mexico (Colbert 1961), are sometimes cited as
demonstrating dinosaurian viviparity, but are more likely a case of cannibalism in so far as jumbled,
disarticulated juvenile skeletons are enclosed in almost perfectly articulated adult skeletons (making
parental care unlikely for Coelophysis ). The supposed juvenile within the type of Compsognathus
longipes is in fact a skeleton of the lacertilian Bavarisaurus (Ostrom 1978). Dinosaur egg morphology
indicates nests buried in leaf litter or sand (Seymour 1979; Seymour and Ackerman 1980), a
construction pattern similar to that of both crocodilians and megapode birds (Frith 1956, 1962;
Seymour and Ackerman 1980). Megapode nesting habits are regarded as primitive among birds
COOMBS: JUVENILE DINOSAURS
103
(Welty 1963) and are thus an appropriate analogue for at least the primitive pattern of dinosaur
nest-building behaviour. Parental attendance of buried nests, involving protection (crocodilians),
temperature regulation (megapodes), and assistance to young at hatching (both crocodilians and
megapodes) may be a primitive archosaurian behaviour pattern that is reasonably inferred for
dinosaurs. Hatchlings of crocodilians, megapodes, and ground-nesting birds in general (excluding
those relatively free of terrestrial predator threat, e.g. most insular, Arctic, and Antarctic marine
birds) are generally precocial (mobile, self-feeding, and fast maturing). Megapode hatchlings are
notoriously precocial, at least one species being capable of flight within 24 hours of hatching (Frith
1962). The Psittacosaurus juveniles had been feeding on abrasive material, probably vegetation,
for some time prior to death, as evidenced by wear on teeth of both skulls (similar conclusion for
juvenile Maiasaurus : Horner and Makela 1979). It is therefore probable that the juveniles of
Psittacosaurus were precocial; parental offering of transported, premasticated, or regurgitated food
is unlikely, implying that parental care was unnecessary at least for feeding.
A great difference in physical size between juvenile and adult increases liability of injury to
young in species having large adults and parental care after hatching. Habits and posture also
play a part. Some crocodilians practice extended parental care after the young hatch, but crocodilians
are semi-aquatic, with young more aquatic than adults, and the posture is sprawling to semi-erect
quadrupedal with long, slender, flexible toes. The danger of injury to a juvenile crocodilian is thus
far less than to juveniles of terrestrial, bipedal dinosaurs that have a compact foot and digitigrade
stance. Among both mammals and birds that are ground dwellers and that have precocial young,
the minimum relative newborn size ( = newborn body mass expressed as a percent of adult body
mass, and hereafter abbreviated RNS) is about 0-9% (in Struthio\ Welty 1963). RNS values for
mammals with precocial young range from about 1 to 12% (calculated from data in Walker
1968) with typical values of about 3-5 % (e.g. Odocoileus, Giraffa , Rhinoceros , Antilocapra , Rangifer,
Taurotragus, and Loxodonta : Walker 1968; Case 1978). Adult body size has little influence on RNS
for mammals with precocial young. Relatively immobile, altricial young whose position and
movements are either limited or controlled by the parent have much lower RNS values (e.g. 0-33 %
in Euarctos : Walker 1968; Case 1978). The recent death of a zoo-born Ailuropoda, apparently by
smothering under a careless (inexperienced) adult female, points up the potential dangers to young
of very low RNS and terrestrial habits. RNS values of modern reptiles follow a different pattern:
values decline as adult body mass increases (Table 1). The RNS of Psittacosaurus is 0-7 to 0-8%
(calculated by cubing equivalent linear dimensions of adult and juvenile) which is low compared
to precocial young of mammals or birds, but is high compared to that of modern reptiles of similar
adult body mass (10 to 100 kg category. Table 1). RNS values for other dinosaurs include 0-24%
for Protoceratops and 0 06% for Hypselosaurus (Case 1978). Thus dinosaurs appear to follow the
reptilian pattern of declining RNS at high adult body mass. The combination of low RNS and
mobile, self-feeding habits of Psittacosaurus contrasts with the pattern for mammals and birds that
practice parental care after birth or hatching and that have active, mobile young. On the basis of
RNS values, parental care is marginally possible, but not probable for Psittacosaurus, and is very
unlikely for dinosaurs of high adult body mass.
table 1. Data from Case (1978) showing the decline in relative newborn size (RNS) at higher adult body
masses among modern reptiles.
Adult mass
Number of species
averaged
Average RNS
Less than 1 kg
6
3-8%
1-10 kg
7
1-9%
10-100 kg
5
0-42%
Over 100 kg
7
0-04%
104
PALAEONTOLOGY, VOLUME 25
Indirect support for the hypothesis that some dinosaurs did not care for their young after
hatching comes from Triassic- Jurassic trackways known as Selenichnus (text-fig. 8). The footprint
maker was a tridactyl biped with a tail sufficiently long to occasionally imprint, and the animal
was probably a juvenile of one of the many large bipedal dinosaurs that are so ubiquitous in the
Connecticut Valley (e.g. Eubrontes , Gigandipus, Anomoepus, Anchisauripus, or Sauropus; Lull 1953).
The Selenichnus trackmaker was intermediate in size between the two juvenile Psittacosaurus
described herein, although in view of the disparity in geologic age, Selenichnus trackways could
not belong to Psittacosaurus. The five slabs at the Pratt Museum (Amherst College, Amherst,
Massachusetts, U.S.A.) that contain Selenichnus tracks have the following characteristics in
common: (1) there are no instances of ‘adult’ footprints in company with Selenichnus ; and (2) there
are no instances of two or more Selenichnus tracks on a single bedding plane, headed in roughly
the same direction. Insufficient numbers of Selenichnus tracks are known to draw unequivocal
conclusions, but the evidence indicates that tiny dinosaurs, the size of small juvenile Psittacosaurus,
travelled alone, unaccompanied by either adults or fellow hatchlings, at least sometimes. This
evidence suggests that at least some dinosaurs did not practice post-hatching parental care.
text-fig. 8. Selenichnus breviusculus, trackway made by a small tridactyl, bipedal dinosaur that at least
sometimes left a sinuous tail trace (pseudo-tail traces are sometimes made by dragging the tip of one toe
through soft sediment, but that does not appear to be the case with this specimen). This particular
trackway was made on a raindrop splattered surface. According to Lull (1953) the average print length
for this species is 46 mm and the average step length is 58 mm. I estimate the size of the trackmaker
to be a little smaller than AMNH 6536, the larger Psittacosaurus specimen described herein. Lull (1953)
regarded these tracks as pertaining to a theropod. Specimen in the collection of the Pratt Museum,
Amherst College (Amherst, Massachusetts, U.S.A.).
Association of adult and juvenile skeletons might be taken as an indication of parental care.
The remarkable assemblage of Protoceratops andrewsi at Djadochta includes numerous individuals
of every age category, unhatched eggs to old adults (Brown and Schlaikjer 1940; Dodson 1976)
and association of adults and juveniles in life is thus indisputable. Horner and Makela (1979)
interpreted a group of juvenile hadrosaurs ( Maiasaurus ) as requiring parental guidance to remain
a coherent group, and an adult specimen was unearthed reasonably nearby. As noted above, the
juvenile Psittacosaurus described herein were excavated in the vicinity of the type of P. mongoliensis,
although exactly how close is difficult to ascertain from available records. Association of adults
and juveniles in fossil assemblages has little bearing on the question of parental care in so far as
such an association would sometimes occur even if the adults abandoned their eggs immediately
after laying. Only if nests were constructed far from the normal range of the adult (note arguments
of Horner and Makela 1979), and if the juveniles lived entirely outside of adult habitats would
the incidence of adult-juvenile association in fossil assemblages be reduced to nil. While such a
set of conditions may have existed for some dinosaurs, it is unlikely it was a universal situation.
COOMBS: JUVENILE DINOSAURS
105
Thus the association of adult and juvenile skeletons even in a single quarry is not sufficient proof
of extended post-hatching care.
Parental care, summary. Available evidence for parental care by dinosaurs is scant and equivocal.
Moreover, there is no clear indication of what constitutes necessary and sufficient proof of either
competing theory (i.e. that dinosaurs did, or that dinosaurs did not, care for their young). On the
basis of the data and arguments developed above, the following conclusion might be drawn:
(1) guarding of nests may be a general archosaurian pattern inherited as the primitive behaviour
of dinosaurs; (2) nests of some dinosaurs were buried and may have required parental attendance
for temperature regulation and perhaps for assisting the young at hatching; (3) egg distribution
of other dinosaurs seems to preclude parental care after laying; (4) adult-juvenile size disparity
makes parental care unlikely for very large dinosaurs, but for smaller genera the size disparity
may not have been a problem; (5) footprints indicate that at least some tiny dinosaurs, probably
juveniles, were independent of adults; and (6) for Psittacosaurus, available evidence is inconclusive.
Sibling groups. One of the two juvenile Psittacosaurus specimens is a composite of several skeletons,
with fourteen distal ends of tibiae indicating a minimum of seven individuals represented. The
jumbled, broken, incomplete condition of the skeletons suggests a post-mortem assemblage, yet
the majority of bones are from juveniles of similar, indeed almost identical, size (text-fig. 6). Such
an assemblage could result from mechanical sorting processes, or from the simultaneous deaths
of several individuals of a sibling group. Horner and Makela (1979) described an assemblage of
juvenile hadrosaurs probably derived from a single clutch, and footprints possibly belonging to
juvenile dinosaurs travelling in unison have been reported (Currie and Sargeant 1979; contra
Selenichnus evidence described above). Therefore juveniles, possibly clutch mates of some dinosaurs,
may have formed cohesive aggregates, for which the term ‘sibling group’ is proposed in preference
to ‘herd’ or ‘flock’ or ‘school’ all of which imply behavioural interactions not yet proven for
dinosaurs. A behavioural complexity similar to that of modern gregarious reptiles (e.g. Ambly-
rhynchus) is sufficient to explain sibling groups among dinosaurs. Aggregates of adult dinosaurs
might have resulted from persistence of bonds formed in juvenile sibling groups.
REFERENCES
bohlin, b. 1953. Fossil reptiles from Mongolia and Kansu. Palaeont. sin. Publ. 37, 1-113.
brown, b. 1914. Leptoceratops, a new genus of Ceratopsia from the Edmonton Cretaceous of Alberta. Bull.
Am. Mus. nat. Hist. 33, 567-580.
— and schlaikjer, e. m. 1940. The structure and relationships of Protoceratops. Ann. N.Y. Acad. Sci. 40,
133-266.
— 1942. The skeleton of Leptoceratops with the description of a new species. Am. Mus. Novit. No. 1 169,
1-15.
case, t. j. 1978. Speculations on the growth rate and reproduction of some dinosaurs. Paleobiology, 4,
320-328.
chao, s. 1962. A new species of psittacosaurs from Laiyang, Shantung. Vertebr. Palasiat. 6, 349-360. [In
Chinese.]
charig, a. 1979. A new look at the dinosaurs. Mayflower Books, Inc., New York. 160 pp.
colbert, e. h. 1961. Dinosaurs, their discovery and their world. E. P. Dutton and Co., New York, xiv + 300 pp.,
51 figs., 100 pis.
— 1965. The age of reptiles. W. W. Norton & Co., Inc., New York. 228 pp., 67 figs., 20 pis.
coombs, w. p., jr. 1972. The bony eyelid of Euoplocephalus (Reptilia, Ornithischia). J. Paleont. 46, 637-650.
— 1975. Sauropod habits and habitats. Palaeogeogr., Palaeoclimatol, Palaeoecol , 17, 1-33.
currie, p. j. and sargeant, w. a. s. 1979. Lower Cretaceous dinosaur footprints from the Peace River Canyon,
British Columbia, Canada. Ibid. 28, 103-115.
dodson, p. 1976. Quantitative aspects of relative growth and sexual dimorphism in Protoceratops. J. Paleont.
50, 929-940.
106
PALAEONTOLOGY, VOLUME 25
frith, h. j. 1956. Breeding habits in the family Megapodiidae. Ibis, 98, 620-640.
1962. The mallee fowl. Angus and Robertson, Sidney.
galton, p. m. 1973. On the cheeks of ornithopod dinosaurs. Lethaia, 6, 67-89.
— 1976. The dinosaur Vectisaurus valdensis (Ornithischia: Iguanodontidae) from the Lower Cretaceous of
England. J. Paleont. 50, 976-984.
— 1978. Fabrosauridae, the basal family of ornithischian dinosaurs (Reptilia: Ornithopoda). Paldont. Z. 52,
138-159.
ginsburg, l. 1980. Les gisements a oeufs de Dinosaures du Cretace terminal du Midi de la France et la
physiologie des Ornithopodes. Mem. Soc. geol. France, N.S., 139, 109-110.
Gregory, w. k. 1951. Evolution emerging. Macmillan Co., New York. Vol. 1, xxvi + 736 pp. (text), vol. 2,
viii + 1013 pp. (figures and plates).
hopson, j. a. 1977. Relative brain size and behaviour in archosaurian reptiles. Annu. Rev. Ecol. Syst. 8, 429-448.
horner, j. r. and makela, r. 1979. Nest of juveniles provides evidence of family structure among dinosaurs.
Nature, Lond. 282, 296-298.
huene, F. von 1908. Die Dinosaurier der Europaischen Triasformation mit Beriicksichtigung der aussereuro-
paischen Vorkomnnisse. Geol. paldont. Abh. Suppl. 1, xii + 419 pp., 351 figs., pis. 1-111.
— 1909. Skizze zu einer Systematik und Stammesgeschicte der Dinosaurier. Zentbl. Miner. Geol. Paldont.
1909, 12-22.
— 1925. Eine neue Rekonstruktion von Compsognathus. Ibid. B 5, 157-160.
— 1926. The carnivorous Saurischia in the Jura and Cretaceous formations, principally in Europe. Rev. Mus.
La Plata, 29, 35-167.
huxley, t. h. 1870. On the classification of the Dinosauria with observations on the Dinosauria of the Trias.
Q. Jl geol. Soc. Lond. 26, 32-50.
koken, e. 1887. Die Dinosaurier, Crocodiliden, und Sauropterygier des norddeutschen Wealden. Palaeont. Abh.
3, 309-419.
lull, r. s. 1910. Dinosaurian distribution. Am. J. Sci. 29, 1-39.
— 1933. A revision of the Ceratopsia or horned dinosaurs. Mem. Peabody Mus. nat. Hist. 3, 1-135.
— 1953. Triassic life of the Connecticut Valley (revised edition). Conn. Geol. Nat. hist. Surv., Bull. 81,
1-331.
maleev, e. a. 1954. (Armored dinosaurs from the Upper Cretaceous of Mongolia). Dokl. Ark. Nauk S.S.S.R.
87, 131-134. [In Russian.]
maryanska, t. 1971. New data on the skull of Pinacosaurus grangeri (Ankylosauria). Palaeont. pol. 25, 45-53.
1977. Ankylosauridae (Dinosauria) from Mongolia. Ibid. 37, 85-151.
— and osmolska, h. 1974. Pachycephalosauria, a new suborder of ornithischian dinosaurs. Ibid. 30, 45-102.
1975. Protoceratopsidae (Dinosauria) of Asia. Ibid. 33, 133-182.
meyer, h. von 1857. Stenopelix valdensis, ein reptil aus der Wealden-Formation Deutschlands. Palaeonto-
graphica, 7, 25-34.
Olshevsky, G. 1978. The archosaurian taxa (excluding the Crocodylia). Mesozoic Meanderings, 1, 1-50.
osborn, h. f. 1923. Two lower Cretaceous dinosaurs of Mongolia. Am. Mus. Novit. No. 95, 1-10.
— 1924. Psittacosaurus and Protiguanodon: two lower Cretaceous iguanodonts from Mongolia. Ibid. 127,
1-16.
ostrom, j. h. 1978. The osteology of Compsognathus longipes Wagner. Zitteliana, 4, 73-118.
romer, a. s. 1956. Osteology of the reptiles. Univ. Chicago Press, Chicago, xxi + 772 pp., 248 figs.
Rozhdestvensky, a. k. 1955. (New data concerning psittacosaurs— Cretaceous ornithopods). Vop. Geol. Asia,
2, 783-788. [In Russian.]
— 1960. (Locality of Lower Cretaceous dinosaurs in Khuzbass). Palaeont. Jour. 2, 1965. [In Russian.]
russell, L. s. 1940. Edmontonia rugosidens (Gilmore), an armored dinosaur from the Belly River series of
Alberta. Univ. Toronto Stud. geol. Ser. 43, 3-28.
Seymour, r. s. 1979. Dinosaur eggs: gas conductance through the shell, water loss during incubation and
clutch size. Paleobiology, 5, 1-11.
— and ackerman, R. A. 1980. Adaptations to underground nesting in birds and reptiles. Am. Zool. 20,
437-447.
steel, R. 1969. Ornithischia. In Handbuch der Palaoherpetologie (O. Kuhn, ed.), part 15, 84 pp., 24 figs.
swinton, w. e. 1936. Notes on the osteology of Hypsilophodon and on the family Hypsilophodontidae. Proc.
zool. Soc. Lond. 1936, 555-578.
thulborn, r. a. 1970. The skull of Fabrosaurus australis, a Triassic ornithischian dinosaur. Palaeontology,
13, 414-432.
COOMBS: JUVENILE DINOSAURS
107
walker, e. p. 1968. Mammals of the world. Johns Hopkins Press, Baltimore, xlviii + 1500 pp. (in two volumes).
welty, j. c. 1963. The life of birds. Alfred A. Knopf, New York, xiii + 546 pp.
young, c. c. 1931. On some new dinosaurs from western Suiyan, Inner Mongolia. Bull. geol. Surv. China , 11,
159-266.
— 1958. The dinosaurian remains of Laiyang, Shantung. Paleont. Sin. (n.s.), 16, 1-139. [In Chinese and
English.]
Typescript received 3 March 1980
Revised typescript received 5 December 1980
WALTER P. COOMBS, JR.
Western New England College
Springfield, Mass. 01119, U.S.A.
A DESCRIPTION OF THE GENERATING CURVE
OF BIVALVES WITH STRAIGHT HINGES
by MARGARET JENNIFER ROGERS
Abstract. A method of describing the whole of the generating curve of a lamellibranch is sought. When
lengths and angles are used to describe an outline, much of the outline remains undefined. A curve can be
fitted to an outline and the coefficients in the particular approximation employed then define the outline.
Reasons are given for fitting a Tchebychev polynominal, rather than a spline, or Fourier series containing
both sine and cosine terms. Polar coordinates r and 6 are calculated for each point on a digitized outline.
Cos 6, rather than 6 is chosen as the independent variable when the Tchebychev coefficients are calculated.
It is found that about 100 irregular spaced data points are required to produce stable coefficients, and that
adequate numerical accuracy is obtained when the outline is described by the first six coefficients. These six
coefficients can also be used as shape discriminators. As the value of c0 is a measure of size, it can be used
to standardize the other coefficients. The standardized coefficients can be used to compare the shape of the
generating curve of shells of different sizes.
Descriptions of the morphology of organisms should, ideally, be objective, reproducible, and
enable different forms to be distinguished. In this article the description of the lateral outline, or
generating curve (Raup 1966) of the Carboniferous non-marine bivalve genera Naiadites and
Curvirimula is discussed. The general form of the generating curve is elliptical. A point (the origin
of growth, or tip of the umbo) and a line (the hinge-line, which is straight, or only slightly curved
in the genera studied) define the orientation of these shells to which further descriptions can be
related. A coarse separation of shapes can be made qualitatively by simple terms such as outline
approximately triangular (text-fig. la), rectangular (text-fig. lb), or semicircular (text-fig. lc, d). The
distinction between text-fig. lc and text-fig. Id is more difficult to describe qualitatively. Measure-
ments of various lengths and angles have proved useful in describing and distinguishing shell forms
text-fig. 1 . Outlines of four shells which can be described qualitatively as
approximately (a) triangular, ( b ) rectangular, (c and d ) semicircular.
(Davies and Trueman 1927; Deleers and Pastiels 1947; Trueman and Weir 1955; Eager 1973; Hajkr,
Lukasova, Ruzicka, and Rehor 1974), but only a few points on the shell outline are defined. As
Bookstein (1978) notes, techniques which use ‘landmarks’, inter-landmark distances, and angles
ignore much of the shape of the outline. Between the landmarks the outline is undefined and cannot
be reproduced.
Papin and Khoroshev (1974) measured the radius of curvature (text-fig. 2, rc;) of successive parts
of the outline, indicating that sections of the curve should be described, rather than interpoint
distances. The value of the radius of curvature at Ph when calculated from the position of three
successive points P, 1? P;, P/+1 (text-fig. 2) on the curve, is highly sensitive to the actual position
of the points Pi_l, Pi + i. Accordingly, it is virtually impossible to obtain a reliable estimate of this
[Palaeontology, Vol. 25, Part 1, 1982, pp. 109-117]
110
PALAEONTOLOGY, VOLUME 25
text-fig. 2. Measurements on a shell outline. The hinge
OA is produced to B. Pi^1, Ph Pi+l are three consecutive
points on the outline. rc( is the radius of curvature, t; the
tangent angle, and r, the distance from 0 of point Pt.
quantity. For this reason the radius of curvature, and a related measurement, the tangent angle f,
(text-fig. 2) are not suitable as shape descriptors. Error in the measurement of the length r, (text-fig.
2), from the umbo to the point is due to the accuracy of the measuring device used and is not
related to the shape of the curve at Pt. It is essential to choose a property of the curve which can
be measured accurately as a basis of a reproducible description. The length rt satisfies this condition.
With the advent of more sophisticated devices such as a digitizing table, the (x,y) coordinates
(text-fig. 6) of many points on an outline can be recorded with ease. A large number of unequally
spaced data points will yield an accurate reproduction. If the recorded points are used to redraw
the outline (text-fig. 3), a few equally spaced data points may not reproduce the outline accurately.
Although a large number of data points produce an accurate reproduction, for the purposes of
comparison it is necessary to look for a more economical way in which to represent the outline.
A variety of curve-fitting techniques exist which allow a numerical approximation to the fossil
outline to be derived. A set of numbers which represent the coefficients in the particular
approximation employed can then define the outline. For reasons which are discussed below, the
curve-fitting method chosen here is the Tchebychev polynomial. The purpose of this article is to
determine the reliability of the Tchebychev coefficients as shape descriptors and discriminators in
the practical context of bivalves with a straight hinge.
text-fig. 3. Reproduction of shell (a) from data points taken at 20° intervals and joined by ( b ) curved and (c)
straight lines. ( d ) is reproduced from many unequally spaced data points.
CURVE FITTING TECHNIQUES
All descriptive, curve fitting techniques seek a solution to the equation
y —f(x) over a specified range of x
where x is the independent variable and y the dependent variable. A length, rt (text-figs. 2 and 6),
called the radial length below, can be measured with reasonable accuracy and is a suitable dependent
variable. The angle (text-fig. 6) at the origin between the radial length and the reference line
A OB is an appropriate independent variable since, once this angle is known, the value of the
dependent variable is precisely defined.
With these coordinates, the equation representing a circle is r = a cos 9, but the outline of a
bivalve shell is, in general, too complex to be represented by such a simple equation. Accordingly,
techniques such as splines and Fourier series have been used. A spline curve may be fitted to a
series of data points provided there are no marked inflections. If such inflections occur, the data
points are chopped up into segments at the turning-points. These turning-points are called knots.
Suitable positions for knots (x) are shown on two shell outlines in text-fig. 4. Each part of the shells
would be accurately described by the coefficients of the appropriate spline function, but these
coefficients cannot be used to compare the shapes of the shells (which may belong to the same
species) because each function relates to a different range of the independent variable.
ROGERS: BIVALVE SHELL SHAPE
text-fig. 4. Fitting spline curves to a shell
outline, x mark the position of knots.
A
o
A
o
Fourier series coefficients have been used to define the outlines of ostracodes (Kaesler and Waters
1972), Bryozoa (Anstey and Delmet 1973), blastoids (Waters 1977), and sand grains (Ehrlich and
Weinberg 1970). As yet palaeontologists cannot readily understand what outline is being described
by a particular set of coefficients. This led Scott (1980), in his study of Foraminifera, to calculate
radii from the origin to the outline at 10° intervals from the Fourier series representation of an
outline. The thirty-six measurements allow the outline to be visualized, but the description is not
economical and the outline is undefined between the radii. A Fourier series in which both sine and
cosine terms are used is periodic, the value of the independent variable ranging from 0 to 360°,
and is well suited for describing the closed curves found in the above examples, provided the origin
and orientation are defined. In the case of a sand grain (text-fig. 5a) a choice of origin and orientation
can be made on the basis of its geometry. The line of orientation could be its longest axis ( AOB )
and the origin its mid-point. In radially symmetric organisms, the centre is a point of morphologic
significance and the orientation can be defined in morphologic terms. As in the case of the sand
grain, the origin and orientation of a bivalve can be defined geometrically (text-fig. 5b), but it is
not clear that such definitions would be homologous, or have any morphologic relevance. Biological
reference points within the curve, such as muscle insertions, are rarely preserved in fossil material
and are thus unsuitable. In those bivalves having a straight hinge, an origin and orientation can
be chosen which are morphologically homologous and are usually visible in fossil material. Since
they lie on, rather than within, the outline, a periodic function may not fit the outline most
economically.
The Tchebychev series provides a function in which the independent variable lies in the range
— 1 to +1 and the NAG Library routine E02ADF enables this function to be fitted to an arbitrary
set of data points. For these reasons it was decided to use the coefficients of the polynomials of a
Tchebychev series to describe the generating curve of several bivalves. Experiments were designed
to discover how many data points are needed to ensure that the calculated coefficients are the
same wherever the data points are taken, and by whom. Further, how many coefficients are needed
to give a good representation, whether 6 or a transform of 9 as the independent variable gives
better results; how sensitive the coefficients are to a change of range in the initial data points, and
how effective they are as shape discriminators. The results of these experiments are given below.
The basic objective of this study is to create attributes (shape descriptors) which can be used to
study variation within populations and as aids to classification.
METHODS OF TAKING DATA POINTS AND CALCULATING THE
TCHEBYCHEV COEFFICIENTS
A digitizing table consists of a plane surface over which a pointer may be moved. In the device used for this
study, the pointer consists of the intersection of a pair of cross wires. When a button is pressed the x,y
coordinates of the pointer are recorded to an accuracy of 0 025 mm. The cursor must lie on the table when
the coordinates are recorded. Thus the outline to be digitized must be drawn on a piece of paper, be
photographed, or be a projected image. Material embedded in matrix cannot be digitized directly.
text-fig. 5. Geometric method of defining the origin
and orientation of an outline. AOB = longest axis,
O being the centre, (a) sand grain, ( b ) bivalve.
112
PALAEONTOLOGY, VOLUME 25
Initially, outlines of shells, scaled to be about 4 cm square, were drawn using a camera lucida. The origin
of the polar coordinates system, 0, was taken as the projection of a perpendicular from the tip of the umbo
on to the hinge (text-fig. 6b) and this was the first point digitized. When the specimen consisted of an external
representation of the shell, this origin is concealed by the umbonal swelling and its position must be inferred
from growth lines. The hinge AOB (text-fig. 6a) is the line of orientation and the posterior end of the hinge
(A) was the second point digitized. Subsequent points, i.e. Pt (text-fig. 6a) were digitized from A to B.
text-fig. 6. Diagram showing the orientation of a
shell for digitizing. AOB = hinge line, u = tip of umbo,
0 = origin of the polar coordinates, O' = origin of x, y
coordinates. xh y, are the rectangular coordinates, and
rh9t the polar coordinates of the point P ■ .
The (Xj.y,-) coordinates can be used to calculate the polar coordinates (r;,0;) of the point Pt. The shape is
then defined by the polar equation:
r = r(9)
where 6 is the independent variable and r is the single valued function of 9. An approximation to this function
may be made by a Tchebychev series of the form
r = jC0T0(v)+ £ cmTJv)
where v, the independent variable, is 9 or a transform of 9, Tmiy) is the Tchebychev polynomial of degree m,
and the cm are the coefficients whose values define the particular function r(9) in each case. A full account of
the Tchebychev polynomials can be found, for example, in Froberg (1965). The general polynomial is defined
as Tm(v ) = cos (m arccos v) and the first few polynomials are
T0(v) = 1
7» = v
T2(v) = 2v2 — 1
T3(v) = 4v3 — 3v
When v = cos 9, Tmv = cos (m9) and the expansion is effectively a Fourier cosine series, the form of which is
r = 2^0 + X COS (W^)
where b„ are the coefficients of the Fourier cosine series. In this case it is a simple matter to calculate the area
of a specimen, provided 9 ranges from 0 to n radians (0-180°).
The area = — I- Y
4 V 2
The coefficients were calculated using the NAG library routine E02ADF which computes the least-squares
approximation to an arbitrary set of data points. The introductory remarks to this routine states that more
points should be recorded where the outline changes markedly, and also at the ends of the range. The
independent variable is scaled so that its value ranges from — 1 to + 1 by the linear transform
V = (2v — umax — umin) / (umax — umin) where V = scaled independent variable
vmax — maximum value of v
vmin = minimum value of v
Thus, if 9 ranges from 0 to 180° and v = cos 9 then V = cos 9.
ROGERS: BIVALVE SHELL SHAPE
13
text-fig. 7. Root mean square residual (R.M.S.)
plotted against number of terms included.
The root mean square residual (R.M.S.) is a measure of the departure of the fitted curve from the original
data points. Initially the residuals decrease rapidly as successive terms are used (text-fig. 7); thereafter they
decrease more slowly, and indeed may increase slightly as unwanted fluctuations are produced. The number
of terms in the series required to give adequate accuracy is the number after which the residuals decrease
only slowly. In the case shown in text-fig. 7, six or seven terms give an adequate fit.
NUMBER OF COEFFICIENTS REQUIRED AND CHOICE OF THE
INDEPENDENT VARIABLE
The polar coordinates r and 9 of a digitized outline were used to calculate the coefficients of the polynomial
of the Tchebychev series from degree zero to degree eight. The plots of the R.M.S. (text-fig. 7) for the two
cases when v = 6 and v = cos 9 are typical and show that c2 has little effect on the accuracy of the fit of the
polynomial. Thereafter the accuracy of the fit improves continuously as terms are added when v = cos 9. The
R.M.S. decreases irregularly when v = 9, and it is difficult to identify where significant flattening of the curve
occurs. The values of the coefficients when v = cos 9 are given in Table 1. It can be seen that when a new
coefficient (c,) is added, the value of the preceding one, c, _ j changes considerably but the value of c, _2 is only
slightly altered. The addition of a new coefficient can cause a marked change in the value of all the preceding
coefficients when v = 9.
table 1. Value of the coefficients when polynomials of different degree are evaluated
v = cos 9 is the independent variable
Degree, m of
polynomial
Terms included
c0
Cl
c2
c3
C4
c5
c6
c7
c8
0
0-798
1
0-605
-0-251
2
0-609
-0-240
0-032
3
0-599
-0-239
0-054
0-050
4
0-606
-0-235
0052
0026
-0063
5
0-606
-0-237
0-049
0-030
-0-048
0-032
6
0-607
-0-237
0-048
0-031
-0048
0-026
-0-016
7
0-607
-0-237
0-049
0-030
-0-049
0-026
-0013
0-007
8
0-605
-0-237
0-050
0-030
-0-048
0-028
-0-013
0-003
-0-012
114
PALAEONTOLOGY, VOLUME 25
text-fig. 8. Original outline of a shell overlain by
the outline reproduced using 5 terms (a, b ) and 8
terms (c, d). The independent variable is 9 (a, c ) and
cos 9 ( b , d).
The outline was recreated from the coefficients of the polynominal, the polar coordinates of the fitted points
being calculated using the NAG Library routine E02AEF. When v = 9 the outline is somewhat more circular
than the original (text-fig. 8a, c) whereas the ends of the outline are somewhat more inaccurate when v = cos 9
(text-fig. 8 b, d). This is an unavoidable consequence of using cos 9; the calculated curve must meet the hinge
at an angle of 90° when 9 = 0 or 180°. Cos 0 is preferred as the independent variable because an accurate
portrayal of the ends of the outline is considered of less importance than the general shape which is recreated
more accurately when v = cos 9. For this reason and because of the behaviour of the R.M.S. and the stability
properties of the coefficients noted above, it was decided to adopt cos 9 as the independent variable, rather
than 9.
Inspection of the residuals suggests that the five coefficients, c0-c4, are sufficient to give a good approximation
(text-fig. 8a, b). However, it was found that the outlines recreated using the eight coefficients, c0-c7, look much
more like Naiadites (text-fig. 8c, d). The three coefficients, c5-c7, although small in value and accounting for
little of the residual error, improve the detailed representation of the outline.
NUMBER OF DATA POINTS
About 100 data points were recorded on the generating curve of one shell. A random set of thirty of these
100 points were used to calculate the coefficients of the Tchebychev polynomial to degree 8. This was done
ten times for each of 30, 40, 50, 60, and 80 data points. The variability of c0 is shown in text-fig. 9, the value
of the coefficient being plotted against the number of data points used. It is seen that when thirty data points
are used the value of the coefficient varies from 0-5 to 1-34 and is 0-94 when the full data set is used. About
100 data points yield a stable value of 0-94 for c0.
The same outline drawn and digitized by two further people gave values for c0 ranging from 0-89 to 0-94
when about 100 data points were recorded. An error of about 5 % was considered satisfactory, particularly
considering that either the material or the technique was unfamiliar to the participants.
1-5
10-
Co
0-5-
0 “I 1 1 1 1 1 1—
30 A0 50 60 80 126
No. of random data points
text-fig. 9. Value of c0 plotted against
number of random data points selected
from a set of about 100 digitized on an
outline. In each case coefficients were
calculated to degree 8. When forty or
more data points are used, the ten values
of c0 cannot be recorded separately at
this scale.
ROGERS. BIVALVE SHELL SHAPE
5
THE EFFICIENCY OF THE COEFFICIENTS AS DISCRIMINATORS OF SHAPE
Two rather different shells (text-fig. 10) were selected, and drawn and digitized by different operators.
The coefficients were calculated to degree eight for each digitized outline. When the value of the
coefficients is plotted (text-fig. 10c), it is seen that certain coefficients (c0 and c3) have distinctly
different values for the two shells, whereas other coefficients (cj are approximately the same for
both shells. These two shells are of similar size, and the zeroth coefficient is a measure of size; it
is the radius of the semicircle which fits the outline giving the smallest root mean square residual.
Using this criterion as a measure of size, c0 can be used to standardize the remaining coefficients.
Plots of the standardized coefficients ( cjc0 , c3/c0) (text-fig. 10c) show that the first coefficient cannot
be used to distinguish the two shells, but c3 becomes more effective as a discriminator. This result
is expected, because cy is a crude measure of asymmetry which is refined by successive odd-numbered
coefficients. Negative values of cy indicate that the shells are asymmetric; that is, the posterior lobe
is larger than the anterior lobe. These results indicate that shapes of Naiadites shells of any size
may be discriminated using at most seven coefficients.
text-fig. 10. (a) and (b) two shells digitized and drawn by three different people, (c) value of the
coefficients c0-c3 and the ratios cylc0 and c3/c0 • = shell a. + = shell b.
EFFECTS OF POOR PRESERVATION
As noted earlier, the origin of growth is often obscured in fossil material. However, provided the
umbo is not very large, the tests made above suggest that different operators will choose
approximately the same origin. A more serious problem is the fact that the anterior lobe is often
distorted, and thus its outline cannot be digitized with confidence. Although in Naiadites the anterior
lobe is small, it was found that very different interpretations of the anterior outline gave rise to
markedly different coefficients. Sometimes the distortion brings the anterior lobe below the level
of the hinge, thus it is not possible to measure the radial length when 6 = 180°. It was found that
provided v = cos 9 and 9 ranged from 0 to about 170° the coefficients obtained were similar to
16
PALAEONTOLOGY, VOLUME 25
those obtained when 6 ranged from 0 to 180°. In reasonably well-preserved specimens of Naiadites
it is usually possible to use a range of 6 of 0-170°.
In Curvirimula it was found that the hinge rarely extended beyond the umbo, and often the
maximum value of 6 was 120°. As noted earlier, it is impossible to compare the coefficients of two
curves if the ranges to which they fit are very different.
Further, not only should the range of 6 be the same, but the morphologic structures described
should be the same if the coefficients are to be used as discriminators. Thus, if the range of 6 chosen
is 0 to 120°, the coefficients calculated for one shell (text-fig. 11a) describe the whole of the generating
curve, whereas those for a second shell (text-fig. lib) fail to describe the anterior lobe, and the two
sets of coefficients cannot be used to compare the generating curve of the two shells. In the genera
studied the umbo and the hinge are the only structures on the outline which are homologous.
However, the anterior lobe contains the anterior adductor muscles, part of the foot and its associated
musculature, and the size of the anterior lobe is a good indicator of the mode of life of the organism.
A suitable way of comparing these shells may be to describe the posterior lobe in terms of the
coefficients which define its outline, and to describe the anterior lobe in terms of area, a parameter
which can easily be calculated from the digitized points. Such a description would still be economical,
and a comparison of shells described in this way would be justified on theoretical grounds. This
method has not yet been tested.
text-fig. 1 1. Two specimens of Curvirimula having very
different maximum value of 6.
SUMMARY AND CONCLUSION
The shape of the generating curve of those shells (including Naiadites) which have a straight hinge
extending beyond the umbo may be described economically using the coefficients of a polynomial
of the Tchebychev series of the eighth degree. Such descriptors are very stable provided that about
100 data points are used and that cos0 is used as the independent variable. C0 is a measure of
size, and tests indicate that if it is used to standardize the coefficients of higher degree, the first five
scaled coefficients can be used as shape discriminators.
Describing shells in this way has the further advantage that, because the whole of the generating
curve can be reproduced, any other features of the shell such as the longest radial length can be
calculated from these nine coefficients. If 6 ranges from 0 to 180°, other features such as the area can
be calculated directly from the coefficients.
When the hinge is not straight, or does not extend beyond the umbo, the shape can still be
described, but it is difficult to specify the range of the independent variable in such a way that
different species can be compared. In the case of Curvirimula the hinge rarely extends beyond the
umbo. For this genus it is suggested that the coefficients of a very restricted portion of the generating
curve, together with a further parameter, the area of the anterior lobe, which has functional
significance, may be used in order to compare shells. For shells which lack a straight hinge it is
possible that a periodic function may be more appropriate as a shape discriminator. The umbo
could be retained for the purpose of orientation, but it may be necessary to define a geometric origin.
Acknowledgements. The author would like to thank Professor M. H. Rogers for much assistance with the
mathematical techniques employed, Dr. G. N. Lance for critical reading of the manuscript, and Mrs. J. Bees
for drafting the illustrations. The ICL 4-75 computer at Bristol University was used for all the calculations,
and the assistance of the computer-centre staff is gratefully acknowledged.
ROGERS: BIVALVE SHELL SHAPE
117
REFERENCES
anstey, r. l. and delmet, D. A. 1973. Fourier analysis of zooecial shapes in fossil tubular bryozoans. Bull. geol.
Soc. Amer. 84, 1753-1764.
bookstein, F. L. 1978. The measurement of biological shapes and shape change. Lecture notes in biomathematics,
24, 1-191.
davies, J. h. and trueman, a. e. 1927. A revision of the non-marine lamellibranchs of the Coal Measures.
Quart. Jl geol. Soc. 83, 210-259.
deleers, c. and pastiels, A. 1947. Etude biometrique des Anthraconauta du Houiller de la Belgique. Publ.
Assoc, strat. Houilleres, 2, 1-99.
eagar, r. m. c., 1973. Variation in shape of shell in relation to palaeoecological station in some non-marine
Bivalvia of the Coal Measures of south east Kentucky and of Britain. C.R. 7th Cong. Int. Strat. Geol. Carb.,
Krefeld, 1971, 2, 387-416.
erlich, r. and weinberg, b. 1970. An exact method for characterization of grain shape. J. sed. Pet. 40,
205-212.
froberg, c. e. 1965. Introduction to numerical analysis. Reading, Mass., Addison-Wesley publ. Co., x + 340 pp.
HAJK.R, o., lukasova. A., ruzicka, b., and rehor, F. 1974. Die gattung Citothyris (Bivalvia) aus dem Karbon
und ihr statistich-geometrisches Modell. Freiberg, Fortschungsh. C306, 1-159.
kaesler, r. l. and waters, J. a. 1972. Fourier analysis of the ostracode margin. Bull. geol. Soc. Amer. 83,
1169-1178.
papin, yu. s. and khoroshev, n. g. 1974. Numerical expression of the shape of shells outlined by a convex or
convexo-concave smooth line. Zh Paleont. 8, 121-124. [In Russian.]
scott, G. h. 1980. The value of outline processing in the biometry and systematics of fossils. Palaeontology,
23, 757-768.
raup, d. m., 1966. Geometric analysis of shell coiling: general problems. J. Paleont. 40, 1178-1190.
trueman, a. e. and weir, j. 1955. The Carboniferous non-marine Lamellibranchia. Palaeontogr. Soc. Mono-
graph, 8, 207-242.
waters, J. a., 1977. Quantification of shape by use of Fourier analyses: the Mississippian Blastoid genus
Pentremites. Paleobiology, 3, 288-299.
nag library 1978. Mark 6. Numerical Algorithms Group Ltd., Oxford, U.K.
M. J. ROGERS
Typescript received 7 July 1980
Revised typescript received 30 November 1980
Department of Geology
Queen’s Building
University Walk
Bristol BS8 1TR
REWORKED ACRITARCHS FROM THE TYPE
SECTION OF THE ORDOVICIAN CARADOC
SERIES, SHROPSHIRE
by ROBERT E. TURNER
Abstract. Thirty-seven acritarch species in the type Caradoc rocks of south Shropshire are recognized as
reworked from strata of Tremadoc and Arenig/Llanvirn age, their distribution reflecting an inverted
stratigraphy. These microfossils yield valuable palaeoenvironmental data, their excellent preservation indicating
a source on, or adjacent to, the Midland Platform; erosion of the relatively unconsolidated parent sediments
occurred in a marine environment. Acritarchs were eroded and redeposited as discrete particles and wave and
current action are considered the most likely erosive agents. The rate of release of microfossils is linked to a
shallowing of the water-body; this resulted in a shift to a high energy environment with storm surges influencing
the erosion and redeposition of sedimentary particles.
Reworking of palynomorphs, the erosion and redeposition of the microfossils within younger
sediments, is a problem familiar to most palynologists (Funkhauser 1969; Richardson and Rasul
1978, 1979) and has been discussed by Wilson (1964). Palynomorphs are prone to reworking because
of their small size, abundance, and the durable nature of their complex wall. This paper attempts to
show that although reworking may present a problem in biostratigraphical studies, it can provide a
valuable insight into the depositional environment of the enclosing sediments. Murchison (1839)
introduced the term Caradoc to geology when he applied the name ‘Caradoc Sandstone’ to a group
of strata cropping out in the county of Shropshire. The clearest section was said by Murchison (p.
216) to occur in the valley of the River Onny near Horderley. Usage of the term Caradoc has been
much restricted by subsequent workers and was given status as a Series by Lapworth (1916) who
subdivided it into Groups based on lithology and faunal content. The exposures along the valley of
the River Onny are accepted as the type section on historical grounds and these still comprise the best
exposed and most complete succession of Caradoc rocks in the district. Since the recent works on the
A489 road adjacent to the river, exposure is continuous through much of the sequence and access is
better than it has been for many years. Extensive researches have been carried out on the macro-
faunas and biostratigraphy of these sediments, a comprehensive review of these works being provided
by Hurst (1979c). Palaeontological studies have concentrated on the rich shelly faunas, but Jenkins
( 1 967) published some of the earliest microfossil data with his detailed examination of the chitinozoa.
Twenty samples were collected from the Onny Valley as part of a larger study of Llandeilo and
Caradoc acritarchs from the British Isles. A further thirteen samples were obtained for comparative
purposes from north Shropshire in the vicinity of Chatwall Farm (SO 5137 9745, text-fig. 2). All
samples were treated with standard laboratory techniques for the recovery of palynomorphs and
most yielded abundant, diverse, and well-preserved acritarchs. The majority of samples examined
contain numerous acritarchs of undoubted Caradoc age (text-fig. 5); these form the bulk of the
acritarch assemblages recovered and will be described elsewhere.
All figured specimens are registered and held in the MPK series of the palynological collections at the Institute of
Geological Sciences, Leeds.
IPalaeontology, Vol. 25, Part 1, 1982, pp. 119-143, pis. 15-17.1
120
PALAEONTOLOGY, VOLUME 25
AGE OF THE REWORKED MATERIAL
Samples from the Onny Valley contain reworked acritarchs that evidently originated from different sources
since, on an age-basis, they fall into three broad categories.
Category 1. Acritarchs of Tremadoc age
Some species recovered can be identified as Tremadoc forms because their known stratigraphical ranges are
restricted to strata of this age. Their occurrences here cannot be interpreted as extensions of their ranges because
the species concerned remain unknown in strata of Arenig to Llandeilo age (see text-figs. 3, 4).
Category 2. Acritarchs of probable Tremadoc age
This category consists of taxa that are known both from the Tremadoc and the Arenig/Llanvirn but which,
other than this occurrence, have not been recorded from younger strata in the British Isles, or elsewhere (text-
figs. 3, 4). That these forms are reworked seems certain, but it is not possible to assign to them a precise
stratigraphical age. Despite this, although an Arenig/Llanvirn age is accepted as possible for some of these
individuals, it is probable that the majority were derived from rocks of Tremadoc age. This is indicated by the
large numbers involved (see text-fig. 3) which suggest derivation from particularly rich pre-existing acritarch
assemblages. It is known that parts of the Tremadoc sequence in Great Britain yield abnormally abundant
acritarch populations; a rough calculation suggested a figure of 100,000 individuals per gram of rock, for a
sample from the Shumardia pusilla Zone of the Shineton Shales of Shropshire (Downie 1958, p. 332). These
profuse numbers are reflected in the periodic reworking of Tremadoc acritarchs into other parts of the geological
column; for example, reworked Tremadoc assemblages have been identified in Devonian rocks in Oxfordshire
(Richardson and Rasul 1979) and are known from Llanvirn sediments in north-west England (author’s
unpublished data). Strata of Tremadoc age are unquestionably the commonest recognized source of reworked
acritarchs in the British Isles. In contrast, the work of Booth 1979 (unpublished Ph.D. thesis) and the present
author’s own unpublished data suggest that in Britain the Arenig/Llanvirn was a period of substantially lower
phytoplankton productivity. Thus most of the acritarchs in this group were probably redeposited from
sediments of Tremadoc age.
text-fig. 1. Location and geological setting of the Caradoc type section.
TURNER: REWORKED ACRITARCHS
121
Category 3. Acritarchs of Arenig/Llanvirn age
The species placed in this category have stratigraphical ranges restricted to strata of this age. They are not present
in the type Llandeilo of South Wales nor have they been recorded from Llandeilo or younger rocks elsewhere in
the world.
text-fig. 2. Sketch-map of the outcrop of Ordovician rocks around Church
Stretton, Shropshire, and the relative positions of exposures at Chatwall and the
River Onny.
122
PALAEONTOLOGY, VOLUME 25
text-fig. 3. The distribution and numerical abundance of reworked acritarchs in samples from the Caradoc type
section. The figure for each category of reworked acritarchs identified in every sample is given as a percentage
based on a count of 200 specimens, or 1 00 specimens where acritarchs are sparse.
TURNER: REWORKED ACRITARCHS
123
text-fig. 4. Previous records of the reworked taxa identified in the type Caradoc. Locality numbers indicate the
following references. 1. Vanguestaine 1978: 2. Timofeev 1959: 3. Rasul and Downie 1974: 4. Downie 1958: 5.
Rasul 1974: 6. Martin 1977: 7. Rasul 1979: 8. Combaz 1967: 9. Martin 1973: 10. Deunff 1961: 1 1. Gorka 1967: 12.
Timofeev 1966: 13. Martin 1969: 14. Rauscher 1974: 15. Rasul 1976: 16. Vavrdova 1972: 17. Vavrdova 1973: 18.
Booth 1979: 19. Cramer and Diez 1977: 20. Dean and Martin 1978: 21. Vavrdova 1966: 22. Vavrdova 1977: 23.
Vavrdova 1976: 24. Turner and Wadge 1979: 25. Burmann 1970: 26. unpublished data: 27. Cramer and Diez
1976: 28. Cramer, Allam, Kanes, and Diez 1974: 29. Cramer, Kanes etal. 1974: 30. Downie and Soper 1972: 31.
Burmann 1968: 32. Paris and Deunff 1970: 33. Loeblich and Tappan 1978.
124
PALAEONTOLOGY, VOLUME 25
DESCRIPTIVE PALAEONTOLOGY
Group acritarcha Evitt 1963
The system of informal ‘subgroups’ proposed by Downie, Evitt, and Sarjeant (1963) has no status under the
provisions of the International Code of Botanical Nomenclature (I.C.B.N.). As indicated by Wicander 1974
(p. 1 1), the introduction of new subgroups which reflect generic names (Staplin, Jansonius, and Pocock, 1965)
poses substantial nomenclatural problems. For these reasons it is preferred here not to organize the acritarcha
into a suprageneric classification but to simply list them in alphabetical order.
Genus acanthodiacrodium (Timofeev 1958) Deflandre and Deflandre-Rigaud 1961
The status in the literature of the ‘diacrodians’ is confused, a situation created when many of the early
species described were assigned to invalid genera by Timofeev (1959). The resultant taxonomic
confusion was exacerbated by Deflandre and Deflandre-Rigaud (1961) who revised and restricted
most of Timofeev’s original genera; unfortunately many of their emendations are either invalid or
illegitimate under various provisions of the I.C.B.N. and must be rejected. Leoblich and Tappan
1978, discussed this problem (pp. 1236-1238) and created a new genus, Actinotodissus. This appears
to be differentiated from Acanthodiacrodium on minor variations in morphology and has yet to be
widely accepted. Considerable numbers of individuals attributable to Acanthodiacrodium or possibly
to Actinotodissus were recorded here but no attempt was made to speciate them. Although the
stratigraphical distribution of the ‘diacrodians’ is not wholly understood, it is clear that in the
Tremadoc rocks of Great Britain such forms occur in abundance, while in Arenig and younger rocks
they are relatively rare.
A can thodiacrodium / Ac tino todissus spp .
Plate 16, fig. 3
Description. Central vesicle varying in outline from ovate with rounded poles to elongate-subovate. Opposite
poles bear similar processes which may be hollow or solid; the central portion of the vesicle is always without
processes. Both vesicle and process wall are usually smooth, rarely granulate; the equatorial zone may bear
longitudinal striae. No excystment structure recorded.
Genus archaeohystrichosphaeridium Timofeev 1959 ex Loeblich and Tappan 1976
The genus Archaeohystrichosphaeridium is technically invalid (see Loeblich and Tappan 1976, p. 303)
but many forms originally described under this name have not yet been transferred to other genera.
The name is retained here pending transfer of included species to other genera.
Archaeohystrichosphaeridium zalesskyi Timofeev 1959
Description. Central vesicle spherical, smooth, bearing a moderate number (15-25) of smooth, simple, hollow,
homomorphic processes which have wide bases tapering to an acuminate distal termination. The process interior
communicates freely with the vesicle cavity. No excystment structure recorded. Vesicle diameter 24-33 /un;
process length 5-10 ^m. Four specimens measured.
Remarks. The spherical forms described here probably belong to the genus Solisphaeridium Staplin,
Jansonius and Pocock (1965), but transfer should await the examination of in situ material.
Genus arkonia Burmann 1970
Arkonia tenuata Burmann 1970
Description. Central vesicle hollow, triangular in outline, compressed with each angle bearing a long, hollow,
smooth, simple process tapering gradually to an acuminate distal termination. Processes communicate freely
with the vesicle cavity and all lie in the same plane as the compression of the central body. The vesicle wall is
TURNER: REWORKED ACRITARCHS
25
ornamented with fine, closely spaced striae which are approximately parallel to the vesicle sides; these striae do
not extend on to the process wall which is smooth. No excystment structure recorded. Vesicle height 27-32 /xm;
process length 33-36 /u.m. Three specimens measured.
Remarks. A. tenuata differs from A. virgata Burmann 1970 by having more numerous, more closely
spaced, and finer striae ornamenting the vesicle wall.
Arkonia virgata Burmann 1970
Description. Similar to A. tenuata in over-all morphology but the vesicle wall is ornamented with fewer, coarse,
widely spaced striae. No excystment structure recorded. Vesicle height 28-3 1 /urn; process length 34-39 /xin. Four
specimens measured.
Genus coryphidium Vavrdova 1972
Coryphidium australe Cramer and Diez 1976
Plate 17, fig. 4
Description. Central vesicle hollow, quadrate in outline with rounded corners, strongly compressed. The vesicle
wall bears numerous (more than fifty) short, relatively thick processes but is otherwise smooth. Processes tend to
be more closely spaced towards the corners of the vesicle and sparse on the central portions. Processes are
hollow, smooth, and communicate freely with the vesicle cavity; distal terminations may be capitate, bifurcate,
or irregularly bulbous. No excystment structure recorded. Vesicle width 39-42 ^.m; process length 6-8 /im. Eight
specimens measured.
Coryphidium bohemicum Vavrdova 1972
Description. Central vesicle hollow, quadrate in outline with rounded corners, strongly compressed, the sides of
the central body may be almost straight or concave. The vesicle wall bears numerous (30-60), short, relatively
thick processes which are concentrated towards the corners, with few or none on the central portions of the
vesicle. These processes are hollow, smooth, and communicate freely with the vesicle cavity; distal terminations
may be flat-topped, bifurcate, multifurcate, capitate, or irregularly bulbous. The vesicle wall also bears well-
developed striae which are mostly restricted to those central areas having few processes; these striae are
approximately parallel to the vesicle sides around the margins but towards the centre may become strongly
concave. No excystment structure recorded. Vesicle width 22-27 /xm; process length 3-6 fim. Five specimens
measured.
Remarks. Probable occurrences of this species, recorded as ‘Indetermine forme A’ in the Caradoc of
Ombret, Belgium (Martin, Michot, and Vanguestaine, 1970) are here interpreted as reworked (see
also Martin 1977, fig. 14).
Coryphidium elegans Cramer, Allam et al. 1974
Description. Central vesicle hollow, quadrate in outline with rounded corners, strongly compressed. The vesicle
wall bears numerous (30-60), short, slender processes which tend to be concentrated towards the corners with
few or none on the central portions. Processes are smooth, apparently solid, and the distal terminations may be
rounded or capitate. The vesicle wall also bears well-developed striae which are most prominent on the central
portions but may extend into the corners; these striae are approximately parallel to the vesicle sides but tend to
become concave towards the centre. No excystment structure recorded. Vesicle width 20-21 ^m; process length
3-5 jam. Two specimens measured.
Genus cymatiogalea (Deunff 1961) Deunff, Gorka, and Rauscher 1974
Cymatiogalea cristata (Downie 1958) Deunff, Gorka, and Rauscher 1974
Plate 15, fig. 2
Description. Central vesicle spherical to sub-spherical; wall granular, divided into polygonal fields by low sutural
ridges which bear smooth, apparently solid processes dividing distally into two to four simple lateral branches;
the polygonal areas between sutural ridges lack processes. Excystment is by the development of a large round to
126
PALAEONTOLOGY, VOLUME 25
sub-polygonal polar opening, the periphery of which also bears processes. The operculum, which is commonly
preserved inside the vesicle, is devoid of processes but has a coarse granular ornament. Vesicle diameter 25-31
/xm; excystment opening 27-30 /xm; process length 15-28 fim. Eight specimens measured.
Cymatiogalea velifera (Downie 1958) Martin 1969
Plate 15, fig. 1
Description. Central vesicle spherical to sub-spherical; wall ornamented with irregular grana and divided into
polygonal fields by low sutural ridges that bear processes supporting thin membranes. The processes are smooth,
slender, hollow with a solid proximal plug separating the process interior from the vesicle cavity; distally the
processes divide into two or four simple lateral branches. Membranes are delicate and transparent although
sometimes faint striations or thickenings may be seen. Polygonal areas between sutural ridges lack processes.
Excystment is by the development of a large sub-polygonal polar opening the periphery of which also bears
processes supporting membranes. The operculum which commonly is preserved inside the vesicle lacks both
processes and membranes but has a coarse granular ornament. Vesicle diameter 30-40 /xm; excystment opening
26-35 ftm; process length 7-12 jam. Ten specimens measured.
Genus dasydiacrodium (Timofeev 1959) Deflandre and Deflandre-Rigaud 1961
Dasydiacr odium palmatilobum Timofeev 1959
Plate 15, fig. 4
Description. Central vesicle ellipsoidal, smooth with rounded poles, one pole bears approximately fifteen simple,
smooth, hollow, homomorphic processes which have wide bases and taper rapidly to an acuminate distal
termination. The opposite pole bears a larger number (about twenty-five) of much shorter but otherwise similar
processes. The intervening equatorial portion of the vesicle is without ornament. The interiors of all processes
communicate freely with the vesicle cavity. No excystment structure recorded. Length of long axis 31-33 /xm;
short axis 26-28 jam; long processes 27-30 /xm; short processes 11-14 jam. Three specimens measured.
Genus dicrodiacrodium Burmann 1968
Dicrodiacrodium normale Burmann 1968
Plate 17, fig. 5
Description. Central vesicle is heteropolar, approximately oval in outline with a broadly rounded antapical pole
and a more sharply rounded apical pole. The apical pole bears a single, smooth, hollow, cylindrical process
which has a solid proximal plug separating the process interior from the vesicle cavity. Distally this process
terminates in five to six short, simple, acuminate branches giving a grapnel-like appearance. The antapical pole
bears a dense anastomosing network of fine threadlike processes. The vesicle wall is ornamented with widely
spaced longitudinal striae. No excystment structure recorded. Vesicle height 58 /xm; vesicle width 33 /xm; length
of apical process 25 /xm. One specimen measured.
explanation of plate 15
Selected acritarchs of Tremadoc age
All figures x 1200
Fig. 1. Cymatiogalea velifera (Downie) Martin. OV/A/2b-5, MPK 2732, Onny Valley, Alternata Limestone.
2. C. cristata (Downie) Deunff, Gorka and Rauscher. OV/A/2b-l, MPK 2733, Onny Valley, Alternata
Limestone. 3. Stelliferidium stelligerum Deunff, Gorka and Rauscher. OV/A/la-1, MPK 2734, Onny
Valley, Alternata Limestone. 4. Dasydiacrodium palmatilobum Timofeev. OV/UHS/1-2, MPK 2735, Onny
Valley, Horderley Sandstone. 5. Trichosphaeridium annolovaense Timofeev. OV/A/la-1, MPK 2736, Onny
Valley, Alternata Limestone. 6. Saharidia fragile (Downie) Gombaz. OV/HS/1-1, MPK 2737, Onny Valley,
Harnage Shales, phase-contrast.
PLATE 15
TURNER, Tremadoc acritarchs
128
PALAEONTOLOGY, VOLUME 25
Genus dictyotidium (Eisenack 1955) Staplin 1961
Dictyotidium? dentatum (Vavrdova 1976) Dean and Martin 1978
Description. Central vesicle hollow, sub-polygonal in outline. The vesicle surface is divided into a small number
of polygonal fields (nine in the one individual recorded) delineated by prominent, smooth, transparent
membranes the outer edges of which carry a row of short, capitate, often flat-topped denticles. No excystment
structure recorded. Vesicle diameter 45 /urn. One specimen measured.
Genus frankea Burmann 1970
Frankea breviuscula Burmann 1970
Description. Central vesicle hollow, smooth, triangular in outline, strongly compressed. Each angle bears a single
slender, smooth process of moderate length (up to 70% of vesicle height) that tapers gradually towards the distal
end; here it divides into five or six short, simple, acuminate lateral branches that arise in a single plane normal to
the compression of the vesicle. The hollow processes all lie in the plane of the central body and communicate
freely with the vesicle cavity. No excystment structure recorded. Vesicle height 28 ^m; process length 15 /urn. One
specimen measured.
Frankea hamata Burmann 1970
Description. Similar to F. breviuscula in over-all morphology but differing in having shorter processes that
always divide distally into two long, smooth, simple, strongly recurved lateral branches with acuminate
terminations; these bifurcations occupy the same plane as the compression of the vesicle. No excystment
structure recorded. Vesicle height 24-26 /nm; process length 13-15 /u,m; length of lateral branches 11-14 pm.
Three specimens measured.
Frankea hamulata Burmann 1970
Description. Similar to F. breviuscula in over-all morphology but differing in having short slender processes. No
excystment structure recorded. Vesicle height 29 /urn; process length 9 /im; length of lateral branches 1-2 pm.
One specimen measured.
Frankea longiuscula Burmann 1970
Description. Similar to F. breviuscula in over-all morphology but with very long (up to 150% of vesicle height)
slender processes. No excystment structure recorded. Vesicle height 41 pm; process length 63 /urn; length of
lateral branches 5 /im. One specimen measured.
Frankea sartbernardense (Martin 1966) Burmann 1970
Description. Similar to F. breviuscula in over-all morphology but with very short stout processes. No excystment
structure recorded. Vesicle height 21-24 pm; process length 3-4 /um; length of lateral branches 2-3 /um. Three
specimens measured.
Remarks. Records of this species in the Silurian of Belgium (Martin 1969) are interpreted as reworked
by the present author.
Genus impluviculus (Loeblich and Tappan 1969) Martin 1977
Impluviculus cf. lenticularis Martin 1977
Plate 16, fig. 2
Description. Central vesicle hollow, compressed, polygonal in outline, apparently smooth. Each angle bears a
slender, flagelliform process which tapers gradually to a closed distal termination; these processes are hollow
proximally and communicate freely with the vesicle cavity but may become solid distally. Sometimes two
processes may be closely located forming a pair. All processes arise around the margins of the central body and
lie in the plane of compression. Vesicle diameter 9 /um; process length 24-28 /u.m. One specimen measured.
Remarks. Assignment to I. lenticularis is not certain since the processes are much longer than those in
the type material from the Tremadoc of Brabant, described by Martin (1977).
TURNER: REWORKED ACRITARCHS
29
Genus marrocanium Cramer, Kanes et al. 1974
Marrocanium simplex Cramer, Kanes et al. 1974
Plate 17, fig. 1
Description. Central vesicle hollow, smooth, quadrate in outline, strongly compressed. Each angle bears a single,
smooth, simple process which tapers to a slightly rounded distal termination. The hollow processes lie in the
same plane as the central body and communicate freely with the vesicle cavity. Thin, transparent membranes are
suspended between the processes; those in a lateral position are wide, stretching out to the process tips while the
apical and antipical membranes are seen only immediately adjacent to the vesicle. The membranes appear to
be smooth and are fragile. No excystment structure recorded. Vesicle length 30 /nm; process length 25 ^m. Three
specimens measured.
Remarks. The transparent membranes recorded here possibly envelop the entire central body, thus
forming a delicate periderm rather than being simple, single-layer structures as described by Cramer,
Kanes, Diez and Christopher (1974). More data are required to determine this point.
Genus micrhystridium (Deflandre 1937) Downie and Sarjeant 1963
Micrhystridium diornamentum Rasul 1979
Description. Central vesicle spherical, bearing two types of process. Some processes are long, hollow, relatively
few (5-10), simple, smooth, and taper gradually to a simple acuminate distal termination. The remaining
processes are much more numerous (probably more than 50), short, closely spaced, smooth, apparently solid
and hair-like. No excystment structure recorded. Vesicle diameter 16 ^m; process length 13 p.m and 3 ^m. One
specimen measured.
Remarks. This species, originally described from the Tremadoc of England has subsequently been
recorded from the Arenig/Llanvirn of North Wales (Booth 1979, p. 127; as M. robustum in part).
Genus multiplicisphaeridium (Staplin 1961) Eisenack, Cramer, and Diez 1976
Multiplicisphaeridium maroquense Cramer, Allam et al. 1974
Description. Central vesicle hollow, smooth, polygonal in outline, formed from the merging of process bases.
Processes are long, smooth, broad, and widen rapidly proximally; they divide distally in a characteristic manner
with first- and second-order lateral branches the tips of which recurve sharply to give a loosely coiled appearance.
Processes are hollow and communicate freely with the vesicle cavity. No excystment structure recorded. Vesicle
diameter 29 ju.m; process length 22 ju.ni. One specimen measured.
Multiplicisphaeridium multiradiale (Burmann 1970) Eisenack, Cramer, and Diez 1976
Plate 17, fig. 2
Description. Central vesicle hollow, smooth, polygonal in outline, formed from the merging of process bases. The
processes are long, smooth, broad, and widen rapidly proximally; they divide distally by simple bifurcation up to
the third order, forming slender acuminate lateral branches. The processes, varying in number from five to seven,
are hollow and communicate freely with the vesicle cavity. No excystment structure recorded. Vesicle diameter
22-27 fj.m; process length 14-17 /un. Ten specimens measured.
Remarks. This species resembles M. maroquense but is distinguished by the third-order branching and
the lack of sharply recurved distal tips to the process branches.
Multiplicisphaeridium rayii Cramer, Allam et al. 1974
Description. Central vesicle hollow, smooth, polygonal in outline, formed from the merging of process bases. The
processes are long, smooth, broad and widen rapidly proximally; they divide distally into four or five digitate,
generally straight, dagger-like branches with rare second-order branching. Processes are hollow and
communicate freely with the vesicle cavity. No excystment structure recorded. Vesicle diameter 38 ju.m; process
length 44 jum. One specimen measured.
130
PALAEONTOLOGY, VOLUME 25
Genus polygonium Vavrdova 1966
‘ Polygonium' spp.
Plate 16, figs. 5, 6
Description. Central vesicle hollow, polygonal to sub-polygonal, bearing numerous, long, hollow, simple
processes which communicate freely with the vesicle cavity and have acuminate distal terminations. Processes
may have a consistent concentric arrangement or may be apparently distributed at random. Vesicle wall smooth,
process wall smooth or rarely granular. Processes always have wide bases which thin rapidly to a slender stem,
tapering gradually to the distal tip. No excystment structure recorded. Vesicle diameter 26-35 ;y,m; process length
15-20 yum. Fifty specimens measured.
Remarks. The acritarchs included here under the name ‘ Polygonium ’ embrace a wider variety of
forms than is circumscribed by the genus Polygonium Vavrdova 1966. This taxon was considered by
its author to be distinguished by always having processes arranged in a consistent concentric manner
(Vavrdova 1966, p. 413). Some individuals showing this feature were recorded here (PI. 16, fig. 6) but
most specimens, otherwise indistinguishable, exhibit an apparently random process arrangement
(PI. 16, fig. 5). These forms constitute a taxonomic problem since, although they are abundant in
Tremadoc to Llanvirn strata, no valid generic name has yet been proposed for them. The term
‘ Polygonium ’ is used here to denote forms with both concentrically and non-concentrically arranged
processes.
Genus priscogalea Deunff 1961
Priscogalea distincta Rasul 1974
Plate 16, fig. 4
Description. Central vesicle spherical to sub-spherical, bearing about fifty processes which appear to be
distributed irregularly over the surface. Processes are smooth, solid, and taper towards the distal ends where they
are usually multifurcate; a few bifurcate or simple processes may be present. The vesicle wall is ornamented with
faint striae which radiate out from each process base. Excystment is by the development of a large polygonal to
sub-polygonal polar opening, the periphery of which also bears processes. The operculum is smooth and without
processes. Vesicle diameter 30-33 ^m; excystment opening 15-16 /xm; process length 6-8 /xm. Two specimens
measured.
Remarks. The striate vesicle wall justifies the transfer of this species to the genus Stelliferidium, but
this should await examination of in situ material. This species, previously known only from the
Tremadoc, has been recorded from the Arenig and Llanvirn of Britain (Booth 1979, pp. 173, 322).
EXPLANATION OF PLATE 16
Selected acritarchs of probable Tremadoc age
All figures x 1200
Fig. 1. Vulcanisphaera cirrita Rasul. OV/HS/1-1, MPK 2738, Onny Valley, Harnage Shales, phase-contrast.
2. Impluviculus cf. lenticularis Martin. OV/O/1-7, MPK 2739, Onny Valley, Onny Shales, phase-contrast.
3. Acanthodiacrodium/Actinotodissus sp. OV/UHS/1-1, MPK 2740, Onny Valley, Horderley Sandstone. 4.
Priscogalea distincta Rasul. OV/A/2b-5, MPK 2741 , Onny Valley, Alternata Limestone. 5-6. ‘ Polygonium ’
spp. 5. OV/A/la-1, MPK 2742. 6. OV/A/la-1, MPK 2743, both from the Alternata Limestone of the
Onny Valley.
PLATE 16
TURNER, probable Tremadoc acritarchs
132
PALAEONTOLOGY, VOLUME 25
Genus saharidia Combaz 1967
Saharidia fragile (Downie 1958) Combaz 1967
Plate 15, fig. 6
Description. Central vesicle circular in outline, wall thin (< 0-5 /xm), fragile, and ornamented with irregularly
sized and spaced grana. Concentric folds are developed in the wall adjacent to the periphery. Excystment is by
the development of a small central pylome but often such openings are not apparent. Vesicle diameter 35-70 jum;
pylome diameter 9-15 /u.m. Ten specimens measured.
Genus stelliferidium Deunff, Gorka, and Rauscher 1974
Stelliferidium cortinulum (Deunff 1961) Deunff, Gorka, and Rauscher 1974
Description. Central vesicle spherical to sub-spherical; the wall is thick (1-2 ^m) and bears 15-20 smooth
processes which appear to be distributed irregularly over the vesicle; these processes are hollow with a solid
proximal plug separating the process interior from the vesicle cavity; from the base they taper gradually to a
bifurcate or multifurcate distal termination. The vesicle wall is ornamented with faint striae which radiate out
from the base of each process. Excystment is by the development of a large circular to sub-polygonal polar
opening, the periphery of which always lacks processes. The operculum is smooth and is also without processes.
Vesicle diameter 29-36 ^m; excystment opening 15-23 jixm; process length 5-9 ^m. Fifteen specimens measured.
Stelliferidium stelligerum Deunlf, Gorka, and Rauscher 1974
Plate 15, fig. 3
Description. Central vesicle spherical to sub-spherical, thick-walled (1-2 /xm), and bears about sixty processes
which appear to be distributed irregularly over the surface. Processes are hollow with a solid proximal plug
separating the process interior from the vesicle cavity; they are smooth and taper gradually to a simple acuminate
or bifurcate distal termination. Processes tend to be longest at the antapex becoming progressively shorter
towards the polar opening. Excystment is by the development of a large, circular to sub-circular opening, the
periphery of which bears short, generally bifurcating processes. The operculum, which is commonly preserved in
situ, is granular and without processes. The vesicle wall is ornamented with thick prominent striae that radiate
out from each process base. Vesicle diameter 3 1 -37 ^m; excystment opening 1 8-2 1 ^m; process length 11-15 /xm.
Fifteen specimens measured.
Genus striatotheca Burmann 1970
Striatotheca frequerts Burmann 1970
Description. Central vesicle hollow, quadrate in outline, strongly compressed. Each angle bears a single long,
broad, simple process which tapers gradually to a generally rounded distal termination. The hollow processes lie
in the same plane as the central body and are in free communication with the vesicle cavity. The vesicle wall is
ornamented with fine striae that are approximately parallel to the vesicle margins around the periphery but
become concave towards the centre; these striae pass on to the processes but die away distally. No excystment
structure recorded. Dimensions of central vesicle 33-37 /xmx 31-36 jam; process length 10-16 fxm. Five
specimens measured.
EXPLANATION OF PLATE 17
Selected acritarchs of Arenig/Llanvirn age
All figures x 1200
Fig. 1. Marrocanium simplex Cramer, Kanes et al. NS/4-4, MPK 2744, Chatwall, Harnage Shales. 2. Multi-
plicisphaeridium multiradiale (Burmann) Eisenack et al. OV/A/2b-5, MPK 2745, Onny Valley, Alternata
Limestone. 3. Striatotheca quieta (Martin) Rauscher. NS/4-4, MPK 2746, Chatwall, Harnage Shales. 4.
Coryphidium australe Cramer and Diez. OV/MHS/1-3, MPK 2747, Onny Valley, Horderley Sandstone. 5.
Dicrodiacrodium normale Burmann. NS/4-1, MPK 2748, Chatwall, Harnage Shales. 6. Tunisphaeridium
eligmosum Vavrdova? OV/A/lb-1, MPK 2749, Onny Valley, Alternata Limestone.
PLATE 17
TURNER, Areniq/Llanvirn acritarchs
134
PALAEONTOLOGY, VOLUME 25
Striatotheca principalis Burmann 1970
Description. Similar to S. frequens in over-all morphology but is larger, has long, slender processes with
acuminate distal terminations and bears a vesicle ornament of coarse, widely spaced striae. No excystment
structure recorded. Vesicle dimensions 40-44 /xm x 36-38 (im; process length 20-31 /xm. Four specimens
measured.
Striatotheca principalis var. parva Burmann 1970
Description. Similar to S. principalis in over-all morphology, this variety is distinguished by its much smaller size.
It differs from S. frequens in having slender processes with acuminate distal terminations, and a vesicle ornament
of coarse, widely spaced striae. No excystment structure recorded. Vesicle dimensions 27-31 /xm x 20-23 /xm;
process length 15-20 pm. Five specimens measured.
Striatotheca quieta (Martin 1969) Rauscher 1974
Plate 17, fig. 3
Description. Central vesicle hollow, quadrate in outline, strongly compressed; each angle bears an extremely
short, simple process which tapers to a rounded distal termination. The hollow processes lie in the same plane as
the central body and communicate freely with the vesicle cavity. The vesicle wall is ornamented with line, closely
spaced striae which are approximately parallel to the vesicle margins around the periphery but become concave
towards the centre; these striations continue on to the process wall almost out to the distal tip. No excystment
structure recorded. Vesicle dimensions 29-32 ^m x 33-35 /xm; process length 3-5 /xm. Twelve specimens
measured.
Remarks. Records of this species (as Veryhachium quietum) from the Silurian of Belgium (Martin
1969) are interpreted by the present author as reworked.
Genus timofeevia Vanguestaine 1978
Timofeevia phosphoritica Vanguestaine 1978
Description. Central vesicle spherical to sub-spherical, wall smooth, divided by raised ridges into about twenty
polygonal fields; the junctions of the ridges bear smooth processes which taper gently to a bifurcate or
multifurcate distal termination. No excystment structure recorded. Vesicle diameter 32 /xm; process length 9-11
urn; diameter of polygonal fields 9-11 /xm. One specimen measured.
Genus trichosphaeridium Timofeev 1966
Trichosphaeridium annolovaense Timofeev 1966
Plate 15, fig. 5
Description. Central vesicle spherical to sub-spherical but always compressed, wall smooth, moderately thick
(about 1 pm) with compression folds developed. The vesicle bears more than 100 short, solid, smooth, simple
hairlike processes whose distal terminations may be evexate or acuminate. No excystment structure recorded.
Vesicle diameter 42-49 /xm; process length 3-4 /xm. Six specimens measured.
Genus tunisphaeridium Deunff and Evitt 1968
Tunisphaeridium eligmosum Vavrdova? 1973
Plate 17, fig. 6
Description. Central vesicle hollow, sub-polygonal in outline, wall smooth, bearing 15-20, long, cylindrical,
smooth processes which widen proximally and divide distally by means of simple bifurcation up to the fifth
order; the distal terminations of these branches are long, slender, curved, and sometimes appear to join those of
adjacent processes forming an anastomosing network of fine filaments. Processes are hollow and communicate
freely with the vesicle cavity. No excystment structure recorded. Vesicle diameter 36 /xm; process length 28 /xm.
One specimen measured.
Remarks. The specific assignment is not certain since the branching pattern of T. eligmosum is
described by Vavrdova as palmate rather than bifurcate as recorded here.
TURNER: REWORKED ACRITARCHS
35
Genus vulcanisphaera (Deunff 1961) Rasul 1976
Vulcanisphaera africana Deunff 1961
Description. Central vesicle spherical to sub-spherical, wall smooth or granular, bearing 50-100 processes which
arise from hollow conical projections having a solid tip; normally three processes arise from a common base in
this way, more rarely two or four. Processes are slender and taper gradually to a bifurcate distal termination. No
excystment structure recorded. Vesicle diameter 32-49 ^m, process length 10-14 /u.m. Three specimens measured.
Remarks. The distal bifurcations are extremely fine and delicate and are often broken off giving the
appearance of a simple acuminate termination. Published records of this species are restricted to
strata of Tremadoc age; the present author has identified the taxon in assemblages from Britain which
on their palynological content are of probable Arenig age (unpublished data). It is thus possible that
this form ranges above the top of the Tremadoc.
Vulcanisphaera cirrita Rasul 1976
Plate 16, fig. 1
Description. Central vesicle spherical to sub-spherical with 50-100 processes arising from hollow projections
which have a solid tip. The number of processes sharing a common base in this way varies from two to five.
Processes are slender and taper only slightly towards the distal tip where they branch into numerous delicate
thread-like branches. The branches of adjacent process groups may unite to form a complex anastomosing
network of fine filaments. No excystment structure recorded. Vesicle diameter 37-48 ^m; process length 9-11
/j.m. Three specimens measured.
Remarks. Originally described from the Tremadoc of Shropshire, this species has been subsequently
recorded from the Arenig/Lower Llanvirn of North Wales (Booth 1979, p. 190).
DISTRIBUTION OF THE REWORKED MATERIAL
The over-all pattern of reworking within the Onny Valley section is shown in text-fig. 5. Reworked
acritarchs first appear in the Harnage Shales which represent the onset of fine-grained sedimentation.
These acritarchs include species of both Tremadoc and Arenig/Llanvirn age, indicating that strata of
both ages were being eroded to supply sediment to the shelf during early Caradoc time. Towards the
top of the Harnage Shales and in the lower Horderley Sandstones, Arenig/Llanvirn forms dominate
the reworked portions of the assemblages; only minor elements of probable Tremadoc age are
present. A possible explanation of this is that widespread erosion of Arenig/Llanvirn sediments was
occurring but that only a small area of Tremadoc rock was exposed. From the base of the Harnage
Shales up to the middle Horderley Sandstone, reworked acritarchs consistently comprise 10-20% of
the total assemblages. Above this level the proportion of reworked specimens in the sediments
increases greatly to as much as 70%. This large and sudden increase in reworking is associated with an
increase in the percentages of both Tremadoc and probable Tremadoc forms present. The
simultaneous increase in the abundance of these two Categories tends to substantiate the Tremadoc
age suggested for most individuals placed in Category 2. A high percentage of reworked acritarchs is
evident until the upper part of the Cheney Longville Flags, always with taxa of Tremadoc and
probable Tremadoc age predominating. The maximum level of reworking occurs in the Alternata
Limestone, where up to 94% of the acritarchs are derived, the contemporaneous Caradoc forms being
swamped out. The percentage of Arenig/Llanvirn forms fluctuates throughout the middle of the
Caradoc sequence but is always small (text-fig. 5). The large numbers of reworked acritarchs of
Categories 1 and 2 that are present from the middle Horderley Sandstone through to the upper
Cheney Longville Flags suggest that an acritarch-rich source rock of Tremadoc age was extensively
breached and continued to be eroded over a substantial period of time.
In the lower Acton Scott Beds reworked acritarchs constitute a mere 1 or 2% of the total
assemblage and remain at this much-reduced level up to the top of the succession. This reduction in
reworking coincides with a return to a low-energy mudstone environment.
136
PALAEONTOLOGY, VOLUME 25
If the majority of individuals placed in Category 2 are accepted as having originated in the
Tremadoc then an interesting pattern to the reworking emerges (text-fig. 5). The distribution of taxa
is essentially inverted, reflecting successive erosion of progressively older source sediments during
Caradoc time. It should be noted that this pattern is modified by the relative abundance of Tremadoc
forms at the base of the Harnage Shales. As discussed above, this clearly indicates early erosion of a
Tremadoc source rock; however, the paucity of Tremadoc acritarchs in the overlying Harnage Shales
text-fig. 5. Showing the strati graphical distribution of samples from the Caradoc type section together with the
ages and percentages of the reworked acritarchs recorded from each horizon.
TURNER: REWORKED ACRITARCHS
137
and Horderley Sandstone shows that reworking from this source was subsequently suppressed
although not eliminated. Any explanation of this diminution in erosion of Tremadoc sediments
would be purely conjectural at present. An inversion of this type, with younger material redeposited
in the lower horizons and older forms appearing in the overlying strata, would be expected where
relatively undisturbed sediments were being eroded and quickly laid down again. It may be assumed
that the rocks discussed here would have been little altered by Caradoc time since the Ordovician was
a period of tectonic quiescence in this region (Earp and Hains 1971, p. 89).
Samples from the lower Caradoc of the Chatwall district (text-fig. 2) contain abundant reworked
acritarchs of Arenig/Llanvirn age with rare Tremadoc and probable Tremadoc forms, a distribution
similar to that in the type section. Since the Caradoc sequence is much less complete at Chatwall and
acritarchs are rare above the Harnage Shales, these occurrences are not discussed in detail here but
they demonstrate that the reworking is not a local phenomenon restricted to the Onny Valley.
PROVENANCE OF THE REWORKED MATERIAL
The Tremadoc and Arenig/Llanvirn acritarchs encountered in the Caradoc type section are
extremely well preserved suggesting that individuals were transported only short distances and
underwent rapid reburial. Possible source rocks must therefore have been located close to the site of
redeposition. It is unlikely that the reworked material was derived from the west since this was itself
an area of deposition in Caradoc time. To the east and south, the Midland Platform formed a stable
block during the Palaeozoic (text-fig. 6); Tremadoc rocks are widespread over this platform although
Arenig/Llanvirn strata are practically unknown (Richardson and Rasul 1978, p. 37). To the south-
east of Shropshire, probably close to the ancient margin of the Midland Platform, great thicknesses of
Tremadoc sediments exist. Rocks of Arenig/Llanvirn age were possibly deposited in such peripheral
areas but no traces have yet been found. Acritarch bearing Tremadoc and Arenig/Llanvirn strata are
known from North Wales (Booth 1979; author’s own unpublished data) and from the north of
England (Booth 1979; Downie and Soper 1972), but the distances involved here are considerable and
a closer source is considered more likely. No sedimentological evidence exists to indicate the direction
of transport, but it appears probable that the reworked acritarchs were derived from the Midland
Platform to the east or south-east. This is consistent with the available information on early Caradoc
palaeocurrents (Williams 1969, p. 259, fig. 8).
MECHANISM OF REWORKING
A widely accepted explanation of the mechanism for reworking of palynomorphs is that they were
eroded and transported while encapsulated within particles of pre-existing sediments and so were
protected from damage. If such recycled rock particles are present in a sediment they should be visible
under microscopic examination (Richardson and Rasul 1978, p. 37). In the Onny Valley,
lithologically diverse sediments such as the Harnage Shales, Horderley Sandstone, and Alternata
Limestone all contain numerous reworked acritarchs. Thin-sections of samples from these
formations were prepared and examined to see if such lithoclasts could be recognized but none was
observed, the sediments presenting a more or less uniformly fine-grained appearance; the dimensions
of sediment grains are between 5 and 100 /x m with the vast majority being between 30 and 60 ju.m.
Thus the grains are at most only slightly larger than reworked acritarchs recovered from the same
samples. In addition, although no acritarchs were recognized in thin section, the sediments are clearly
organic-rich and the abundant organic matter visible is trapped in the interstices. These factors make
it unlikely that acritarchs were reworked in an encapsulated state, the evidence suggesting rather that
they were eroded and redeposited as discrete sedimentary particles. This hypothesis is supported by
the distribution pattern of the reworking which is unaffected by changes in the type of sediment being
deposited (text-fig. 5). Similar reworked acritarchs are found in comparable numbers in sandstones,
limestones, and flaggy micaceous siltstones. This alone suggests that the acritarchs were being intro-
duced into the sediment/water-body system independently of the non-organic sediment particles.
138
PALAEONTOLOGY, VOLUME 25
If it is true that the reworked acritarchs were transported as individual particles then further
conclusions can be drawn. Considering the excellent state of preservation of most reworked
specimens, it is probable that dissolution of the parent rock was both easy and rapid. Since an
indurated sediment would resist erosion, the Tremadoc and Arenig/Llanvirn rocks being eroded were
probably at most only partly lithified. The retention of the most delicate morphological features on
many reworked specimens suggests that erosion and transport were not only rapid but did not take
place in a sub-aerial environment. Structures such as fine distal terminations of processes in
text-fig. 6. A suggested palaeogeographic reconstruction of the British Isles during Caradoc time (simplified,
after Williams 1969).
TURNER: REWORKED ACRITARCHS
139
HARNAGIAN
LONG VILLI AN
Developing Regression
text-fig. 7. Hypothetical diagrammatic cross-section through southern Shropshire in Caradoc time showing
the postulated sequence of events. The Caradoc succession is simplified for the sake of clarity.
140
PALAEONTOLOGY, VOLUME 25
Vulcanisphaera cirrita (PI. 16, fig. 1) and the delicate membranes of Marrocanium simplex (PI. 17,
fig. 1) would have been unlikely to survive long in turbulent conditions without the protection
provided by encapsulation. Even if mechanical damage had been avoided, such features would have
suffered rapid oxidation and disintegration. It is therefore postulated that erosion and redeposition of
these pre-existing rocks took place in a shallow marine environment; wave and current action are
considered the most likely agents for eroding and dispersing the unconsolidated sediments involved.
Under these circumstances the enclosed acritarchs would be released directly into the sea, affording
them the means both of protection and rapid dispersal and reburial (text-fig. 7). This agrees with the
limited sedimentological evidence available which suggests that the Caradoc rocks were deposited in
a shallow marine environment, possibly with off-shore barriers but with no estuaries present to have
provided a potential source of reworked acritarchs from aerially exposed sediments (Hurst 1979a, p.
36). Hurst (op. cit., 19796) has shown that in the Onny Valley the sequence from the upper part of the
Horderley Sandstone to the basal Acton Scott Beds represents a regressive phase, and that the
deposits of this interval were greatly affected by storm surge activity. The shallowing of the water
body would have led to an increasingly high-energy regime, while individual surges would have
resulted in mass sediment movement with rapid redeposition on the cessation of these geologically
ephemeral events (text-fig. 7). Text-fig. 5 shows that the high-energy environment which resulted in
the deposition of the upper Horderley Sandstone, Alternata Limestone, and Cheney Longville Flags,
coincided with the period of greatest acritarch reworking. Above the base of the Actonian the
percentage of reworked acritarchs is drastically reduced and it was at this time that storm swells
ceased to exert any significant effect (Hurst 19796, p. 196, Table 1). Unfortunately Hurst’s studies do
not extend down into the Soudleyan so it is uncertain how close the correlation is between the
increasing energy levels and the first appearance of high levels of reworking. None the less there is
clearly a link between the high-energy regime and abundant reworking, supporting the view that
unlithified Tremadoc and Arenig/Llanvirn sediments were being eroded in a shallow marine
environment.
Much has been written in recent years on the effects of storm surges on contemporaneous marine
sediments (Brenner and Davis 1973; Reineck and Singh 1972, 1973), but there appear to be few data
available on the effects of such events on soft pre-existing sediments at the water/substrate interface.
Although little consideration has been given in the literature to the possibility of reworking from such
sediments, the situation visualized here is not unique. For example. Quaternary clays in the Moray
Firth, Scotland, contain extensive assemblages of reworked late Jurassic and early Cretaceous
microfossils, particularly organic-walled microplankton (Owens and Marshall, 1978, pp. 24-26).
This unoxidized material is well preserved and is clearly derived from unconsolidated Jurassic and
Cretaceous shales and clays upon which the Quaternary rests in this area (Dr. R. Harland, pers.
comm.). The physical state of these Mesozoic sediments suggests that reworking within lithic clasts
would have been unlikely and redeposition of the microfossils as discrete particles is considered
probable.
CONCLUSIONS
The type Caradoc rocks of Shropshire yield abundant acritarch assemblages which contain Caradoc
species admixed with reworked Tremadoc and Arenig/Llanvirn taxa.
The vertical distribution of reworked forms reveals essentially an inverted stratigraphy with
Arenig/Llanvirn acritarchs predominating in the lower horizons while the older Tremadoc species
become the most abundant forms in the middle part of the sequence.
The lack of visible lithoclasts in thin sections of these rocks suggests that the microfossils were
introduced into the sediment body as individuals and were not encapsulated in redeposited fragments
of pre-existing rocks; this is substantiated by the fact that the presence and abundance of reworked
acritarchs appears to be entirely independent of lithotype.
The excellent state of preservation of the reworked acritarchs indicates that they underwent little
transportation before reburial. It also suggests that the parent sediments were relatively
TURNER: REWORKED ACRITARCHS
141
unconsolidated and that their erosion was caused by marine action in a shallow-water environ-
ment.
The high percentages of reworking in the middle part of the section are partly related to the
increasing erosion of particularly acritarch-rich Tremadoc rocks. In addition, this sequence
represents a regressive phase with storm surges having profound effects on the shallowing water
body. Such high-energy events would have greatly increased the erosion rate of the unlithified
sediments exposed at the sea bed. A state of continuous low-level erosion and acritarch reworking is
envisaged, punctuated by periodic intense turbidity associated with an upsurge in the rate of release
of pre-existing microfossils.
Acknowledgements. I am most grateful to Professor C. Downie for his encouragement during this work and for
his advice and constructive comments. I thank Dr. G. A. Booth for access to his unpublished information and for
much useful and stimulating discussion. My thanks also go to Dr. B. Owens, Dr. W. H. C. Ramsbottom, and
Dr. W. A. M. Jenkins who read the manuscript and suggested improvements.
REFERENCES
booth, G. a. 1979. Lower Ordovician acritarchs from successions in England and North Wales. University of
Sheffield, unpublished Ph.D. thesis. 383 pp.
brenner, r. l. and davies, d. k. 1973. Storm generated coquinoid sandstones: genesis of high-energy marine
sediments from the Upper Jurassic of Wyoming and Montana. Bull. geol. Soc. Am. 84, 1685-1698.
burmann, G. 1968. Diacrodien aus dem unteren Ordovizium. Paldont. Abh. B, 2, 639-652.
— 1970. Weitere organische Mikrofossilien aus dem unteren Ordovizium. Ibid. 3, 289-347.
combaz, a. 1967. Un microbios du Tremadocien dans un sondage d’Hassi-Massaoud. Act. Soc. linn. Bordeaux ,
B, 104, 1-26.
cramer, f. h., allam, b., kanes, w. h. and diez, m. d. c. R. 1974. Upper Arenigian to Lower Llanvirnian
acritarchs from the subsurface of the Tadla Basin of Morocco. Palaeontographica, B, 145, 182-190.
— and diez, m. d. c. R. 1976. Seven new Late Arenigian species of the Acritarch Genus Coryphidium
vavrdova, 1972. Paldont. Z. 50, 201-208.
— 1977. Late Arenigian (Ordovician) acritarchs from Cis-Saharan Morocco. Micropaleontology , 23,
339-360.
— kanes, w. H., diez, M. d. c. R. and Christopher, R. A. 1974. Early Ordovician Acritarchs from the
Tadla Basin of Morocco. Palaeontographica , B, 146, 57-64.
dean, w. t. and martin, f. 1978. Lower Ordovician acritarchs and Trilobites from Bell Island, Eastern
Newfoundland. Bull. geol. Surv. Can. 284, 1-35.
deflandre, G. 1937. Microfossils des silex cretaces, 1 1 . Flagelles incertae sedis. Hystrichospaeridees. Sarcodines.
Organismes divers. Annls. Paleont. 26, 51-103.
— and deflandre-rigaud, m. 1961. Nomenclature et systematique des Hystrichospheres (s.l.) Observations
et rectifications. Lab. Micropaleont. E.P.H.E. Inst. Paleontol. Mus. Paris , Multicop. 14 pp.
deunff, j. 1961. Un microplancton a Hystrichospheres dans le Tremadoc du Sahara. Revue Micropaleont.
4, 37-52.
— and evitt, w. r. 1968. Tunis phaeridium\ A new acritarch genus from the Silurian and Devonian. Standford
Univ. Pubis. Geol. Sci. 12, 1-13.
— gorka, h. and rauscher, r. 1 974. Observations nouvelles et precisions sur les Acritarches a large
ouverture polaire du Paleozoique inferieur. Geobios, 7, 5-18.
downie, c. 1958. An assemblage of microplankton from the Shineton Shales (Tremadocian). Proc. Yorks, geol.
Soc. 31, 331-349.
— evitt, w. R. and sarjeant, w. a. s. 1963. Dinoflagellates, Hystrichospheres and the classification of the
acritarchs. Stanford Univ. Pubis. Geol. Sci. 7, 2-16.
— and sarjeant, w. a. s. 1963. On the interpretation and status of some Hystrichosphere genera.
Palaeontology, 6, 83-96.
— and soper, N. J. 1972. Age of the Eycott Volcanic Group and its conformable relationship to the Skiddaw
Slates in the English Lake District. Geol. Mag. 109, 259-268.
earp, j. r. and haines, b. a. 1971. British Regional Geology: The Welsh Borderland. Inst. Geol. Sci. 1 18 pp. pub.
H.M.S.O.
142
PALAEONTOLOGY, VOLUME 25
eisenack, a. 1955. Chitinozoen, Hystrichospharen und andere mikrofossilien aus dem Beyrichia- Kalk.
Senckenberg. leth. 36, 157-188.
— cramer, F. fi. and diez, m. d. c. R. 1976. Katalog der fossilen Dinoflagellaten, Hystrichospharen und
verwandten Mikrofossilien , 4, Acritarcha part 2, Stuttgart. 863 pp.
evitt, w. r. 1963. A discussion and proposals concerning fossil dinoflagellates, hystrichospheres and acritarchs.
Proc. natn. Acad. Sci. U.S.A. 49 (2), 158-164, 298-302.
funkhouser, j. w. 1969. Factors that affect sample reliability. In tschudy, r. h. and scott, r. a. (ed.).
Aspects of Palynology, John Wiley and Sons. pp. 97-102.
gorka, h. 1967. Quelques nouveaux acritarches des silexites du Tremadocien superieur de la region de Kielce
(Montagne de Ste Croix, Pologne). Cahiers de Micropaleont. Series 1, 6, Arch. orig. Centre Documen.
C.N.R.S. 441, 1-8.
hurst, j. m. 1979a. The environment of deposition of the Caradoc Alternata Limestones and contiguous deposits
of Salop. Geol. J. 14, 15-40.
— 19796. Evolution, succession and replacement in the type Upper Caradoc (Ordovician) benthic faunas of
England. Palaeogeogr. Palaeoclimatol. Palaeoecol. 27, 189-246.
— 1979c. The stratigraphy and brachiopods of the upper part of the type Caradoc of South Salop. Bull. Br.
Mus. nat. Hist. (Geol.), 32, (4), 183-304.
jenkins, w. a. m. 1967. Ordovician chitinozoa from Shropshire. Palaeontology, 10, 436-488.
lapworth, c. 1916. In Palaeontological Work, summ. progr. Geol. Surv. Gt. Britain 1915, 36-38.
loeblich, A. R. jr. and tappan, h. 1969. Acritarch excystment and surface ultrastructure with descriptions of
some Ordovician taxa. Rev. Esp. Micropaleontol. 1, 45-57.
— 1976. Some new and revised organic-walled phytoplankton microfossil genera. J. Paleont. 50,
301-308.
— 1978. Some Middle and Late Ordovician microphytoplankton from central North America. Ibid. 52,
1233-1287.
martin, f. 1966. Les acritarches de Sart-Bernard, (Ordovicien-belge). Bull. Soc. beige Geol. Paleont. Hydrol. 74
(for 1965), 423-444.
— 1969. Les acritarches de l’Ordovicien et du Silurien beiges. Determination et valeur stratigraphique. Mem.
Inst. r. Sci. nat. Belg. 160 (for 1968), 175 pp.
— 1973. Les acritarches de l’Ordovicien inferieur de la Montagne Noire (Herault, France). Bull. Inst. r. Sci.
nat. Belg. 48 (for 1972), 1-61.
— 1977. Acritarches du Cambro-Ordovicien du Massif du Brabant, Belgique. Ibid. 51 (for 1975), 1-28.
— michot, p. and vanguestaine, m. 1970. Le Flysch Caradocien d’Ombret. Annls. Soc. geol. Beige. 93,
337-350.
murchison, r. i. 1839. The Silurian System. Murray, London (2 vols.). 768 pp.
owens, b. and marshall, j. 1978. Micropalaeontological biostratigraphy of samples from around the coasts of
Scotland. Rep. Inst. Geol. Sci. 78/20. 35 pp.
Paris, f. and deunff, j. 1970. Le paleoplancton Llanvirnien de la Roche-au- Merle, (Commune de Vieux-Vy-
Sur-Couesnon, Ille-et-Vi-laine). Bull. Soc. geol. miner. Bretagne, C, 2, 25-43.
rasul, s. M. 1974. The Lower Palaeozoic acritarchs Priscogalea and Cymatiogalea. Palaeontology, 17, 41-63.
— 1976. New species of the genus Vulcanisphaera (Acritarcha) from the Tremadocian of England.
Micropaleontology, 22, 479-484.
— 1979. Acritarch zonation of the Tremadoc series of the Shineton Shales, Wrekin, Shropshire, England.
Palynology, 3, 53-72.
— and downie, c. 1974. The stratigraphic distribution of Tremadoc acritarchs in the Shineton Shales
succession, Shropshire, England. Rev. Palaeobot. Palynol. 18, 1-19.
rauscher, r. 1974. Recherches Micropaleontologiques et Stratigraphiques dans l’Ordovicien et le Silurien en
France. Mem. Sci. geol. Univ. Louis Pasteur Strasbourg, Inst. Geol. 38 (for 1973), 209 pp.
reineck, h. e. and singh, i. b. 1972. Genesis of laminated sand and graded rhythmites in storm-sand layers of
shelf mud. Sedimentology, 18, 123-128.
— 1973. Depositional sedimentary environments-. Springer- Verlag, Berlin. 439 pp.
richardson, j. b. and rasul, s. m. 1978. Palynomorphs in Lower Devonian sediments from the Apley Barn
borehole, Southern England. Pollen, Spores, 20, 423-462.
— 1979. Palynological evidence for the age and provenance of the Lower Old Red Sandstone from
the Apley Barn Borehole, Witney, Oxfordshire. Proc. Geol. Ass. 90, 27-42.
staplin, f. l. 1961. Reef-controlled distribution of Devonian microplankton in Alberta. Palaeontology, 4,
392-424.
TURNER: REWORKED ACRITARCHS
143
staplin, F. L., jansonius, J. and pocock, s. A. j. 1965. Evaluation of some acritarchous hystrichosphere genera.
Neues Jb. Geol. Palaont. Abh. 123, 167-201.
timofeev, b. v. 1958. Uber das Alter Sachsischer Grauwacken Mikropalaophytologische Untersuchungen von
Proben aus der Weesensteiner und-Lausitzer Grauwacke. Geologie, 7, 826-845.
— 1959. Drevneyshaya flora Prebaltiki; i ee stratigraficheskoe znachenie. (The ancient flora of the Prebaltic
and its stratigraphic significance.) Trudy vses. neft. nauchno-issled. geol.-razv. Inst. 129. 319 pp. [In Russian.]
— 1966. Mikropaleofitologicheskoe isseldovanie drevnikh svit. (Micropalaeophytological research into
ancient strata). Akad. Nauk. SSSR Lab. Geol. Dokembriya. 147 pp. [In Russian.]
turner, r. e. and wadge, a. j. 1979. Acritarch dating of Arenig volcanism in the Lake District. Proc. Yorks, geol.
Soc. 42, 405-414.
vanguestaine, m. 1978. Donnees palynologiques nouvelles dans l’Ordovicien inferieur du bassin de la Senne,
Massif du Brabant, Belgique. Annls. Soc. geol. Belg. 100, 193-198.
vavrdova, m. 1966. Palaeozoic microplankton from Central Bohemia. Cas. Miner. Geol. 11, 409-414.
— 1972. Acritarchs from Klabava Shales (Arenig). Vest, ustred. Ust. geol. 47, 79-86.
— 1973. New acritarchs from Bohemian Arenig (Ordovician). Ibid. 48, 285-289.
— 1976. Excystment mechanisms of early Palaeozoic acritarchs. Cas. Miner. Geol. 21, 55-64.
— 1977. Acritarchs from the Sarka formation (Llanvirnian). Vest, ustred. Ust. geol. 52, 109-118.
wicander, E. R. 1974. Upper Devonian -Lower Mississippian acritarchs and Prasinophycean algae from Ohio,
U.S.A. Palaeontographica, B, 148, 9-43.
williams, A. 1969. Ordovician of British Isles. In marshall kay (ed.). North Atlantic— Geology and Continental
Drift, pp. 236-264.
wilson, l. R. 1964. Recycling, stratigraphic leakage and faulty techniques in palynology. Grana Palynol. 5,
425-436.
Typescript received 20 March 1980
Revised typescript received 12 November 1980
R. E. TURNER
Amoco Canada Petroleum Company Limited
Amoco Canada Building
444 Seventh Avenue S.W.
Calgary, Alberta T2P OY2
Canada
LOWER CARBONIFEROUS CONODONT FAUNAS
FROM RAVENSTONEDALE, CUMBRIA
by a. c. higgins and w. j. varker
Abstract. Conodont faunas from the Lower Carboniferous, Courceyan-Holkerian Stages, of Ravenstonedale
are described. Three zones, the Taphrognathus, Cloghergnathus, and Cavusgnathus Zones, are recognized in
the Chadian-Holkerian strata and one fauna, Fauna A, in the mid-Courceyan. These zones and the Fauna
are correlated with the standard British Stages and with the American classic sections. The faunas are typical
of the shallow water facies of the Lower Carboniferous and are dominated by the genera Taphrognathus,
Clydagnathus, Cloghergnathus, and Cavusgnathus, which characteristically occur in intertidal and shallow
subtidal environments. These shallow-water faunas are joined by deeper-water immigrants only in the Arundian
but, even at this level, they are the dominant elements, indicating only a slight water depth increase. The
absence of the deeper-water genera Gnathodus and Siphonodella adds difficulty to the correlation of this
succession with goniatite-bearing sequences, but correlation with the Avon Gorge and with shallow-water
facies in the north of England is possible. Three species, Apatognathus asymetricus, A. scandalenis, and
Cloghergnathus carinatus are proposed as new taxa.
This account of the conodont faunas of the Lower Carboniferous succession of Ravenstonedale
is a documentation of one aspect of a succession which in recent years has become increasingly
important to the stratigraphy of the Carboniferous period as a whole. The special significance of
Ravenstonedale is, however, long standing, since it falls within the area designated by Garwood
(1913, p. 451) as his type area for the Lower Carboniferous of the north of England. Garwood
(1907, 1913) had produced a zonation, based primarily upon corals and brachiopods, which was
a complement to the slightly earlier and very influential work of Vaughan (1905) in the Bristol
area. Vaughan had designated the Lower Carboniferous as ‘Avonian’ in view of ‘the completeness
of the sequence in the Avon section’ (1905, p. 264) and had further produced the well-known and
much used zonal scheme based, he supposed, upon the evolutionary lineages of corals and
brachiopods. Garwood (1913) recognized the great importance of Vaughan’s paper and conse-
quently included a correlation of his zones with those of Vaughan (Garwood 1913, p. 452).
This faunal zoning proved to be outstandingly successful. Indeed, in the opinion of Ramsbottom
(1973, p. 568) the advance of Carboniferous stratigraphy was for some time ironically hampered
by this very fact, since it appears to have induced a period when very little attention was paid to
the many other aspects of the Lower Carboniferous sequence, in particular to details of lithology
and their interpretation. Vaughan’s zonal scheme along with Garwood’s correlation for the north
of England, was the foundation upon which Dinantian stratigraphy was based for sixty years, until
it was superceded by a series of divisions based upon major cycles of deposition (Ramsbottom
1973). These cycles coincide very closely with the six new stages proposed by George, Johnston,
Mitchell, Ramsbottom, Sevastopulo, and Wilson (1976) which are designed to be applied throughout
Britain.
One consequence of the re-examination of the Dinantian rocks was that Ramsbottom (1973)
was able to show that, far from being complete, the Avon Gorge succession contains four major
non-sequences, and as a result large parts of the succession of the north of England are not able
to be directly correlated with Vaughan’s type area or zonal scheme. In addition, other biozonations
based upon the Avon Gorge succession must necessarily be as incomplete as the coral/brachiopod
zonation. One such scheme is that of Rhodes, Austin, and Druce (1969) in which they established
fourteen conodont assemblage zones, the three highest of which were actually based upon the
(Palaeontology, Vol. 25, Part 1, 1982, pp. 145-166, pis. 18-19.1
146
PALAEONTOLOGY, VOLUME 25
north crop of the South Wales coalfield rather than upon the Avon Gorge. Substantial revision
of this zonal scheme has, however, since (Austin 1973) been necessary, with the result that whilst
the Brigantian (part of three conodont assemblage zones) and Courceyan (six assemblage zones
plus a non-sequence) are usefully subdivided, the whole of the interval from just below the top of
the Asbian down to a horizon within the Chadian is occupied by the Cavusgnathus/Apatognathus
assemblage zone, non-sequences, or horizons with no conodonts (see George et al. 1976,
Table 1).
The present study in Ravenstonedale was initiated in an attempt to provide a Dinantian conodont
sequence from an area in which the succession has been demonstrated to be much more complete
than that of the Avon Gorge (Ramsbottom 1973, p. 595).
STRATIGRAPHY
At its outset the Carboniferous period presented in the North-west Province a landscape of probably some
considerable relief, dominated by rocks of Ordovician and Silurian age. During Dinantian times this landscape
was gradually overwhelmed by a major transgression, evidence for which can be seen in the progressive
overlap of Carboniferous strata on to Lower Palaeozoic and thereby giving rise to an increasing hiatus at
the base of the Carboniferous succession. Complete submergence was not achieved until Asbian times, and
as a result, within the North-west Province the local base of the Carboniferous succession varies in age from
Courceyan to Asbian.
Deposition of the relatively thick (approx. 1500 m) Dinantian succession of Ravenstonedale began with
local subsidence in a region known as the Stainmore Trough or Gulf, which opened to the east and was
bounded to the north, west, and south by the positive massifs of the northern Pennines and the Lake District.
Subsidence appears to have begun earlier in this trough than elsewhere and to have continued throughout
Dinantian times. The Ravenstonedale area thus provides the most complete Dinantian succession within the
North-west Province, with what appears to be an almost continuous record of sedimentation from Courceyan
to Namurian. Readers are referred to Johnson and Marshall (1971) for further details on the regional setting
of the Ravenstonedale area.
The present outcrop of this succession strikes north-westwards from the Dent Fault near Ravenstonedale,
towards Penrith (text-fig. 1) on the southern side of the Vale of Eden. The regional structure is relatively
simple, with gentle dips predominantly in a north to north-easterly direction. Natural stream exposure is
generally good, and when combined with other natural exposures, road-cuttings, and the numerous small
quarries, provides an almost complete record of the succession, except near its base, where much of the detail
is obscured by an extensive deposit of glacial drift. The problems associated with the lower part of the
succession were amply outlined by Holliday, Neves, and Owens (1979) in their account of a shallow borehole
programme undertaken near Ravenstonedale over the period 1975 to 1978.
Working in collaboration with Holliday et al. the present authors have recently described the conodont
faunas from the exposed part of the Pinskey Gill Beds (Varker and Higgins 1979) which form the lowest unit
of the succession. These beds, which lie unconformably upon steeply inclined Silurian Bannisdale Slates and
beneath the felspathic conglomerate, were shown by Holliday et al. (1976) to be 45-50 m in total thickness,
of which only approximately 15 m near the base are exposed. The conodont faunas indicate an age of
mid-Courceyan, i.e. late K or Z on the coral/brachiopod scheme, for the base of the Dinantian succession
in Ravenstonedale.
The overlying felspathic conglomerate, which is also poorly exposed, was calculated by Johnson and
Marshall (1971, p. 267) to be 36-4 m in thickness in Thackthwaite Gill. However, since neither of the boundaries
of the horizon are well exposed and Holliday et al. (1976, p. 349) have shown them to be gradational with
no major breaks in deposition, calculations of thickness can only be approximate. Holliday et al. (1976,
p. 348) estimated the cumulative thickness from their borehole evidence to be around 42 m and interpreted
the felspathic conglomerate as a continental interlude between the marine Pinskey Gill Beds and the overlying
marine Stone Gill Beds. Capewell (1955) considered both the Pinskey Gill Beds and the felspathic conglomerate
to be the lateral equivalents of the much-thicker basal beds which occur to the north-west, where they reach
274 m in thickness in the Lowther Valley area.
Above the felspathic conglomerate there are the Stone Gill Beds, which consist predominantly of limestone
but also include thin (usually less than 1 m) siltstone and shale/mudstone horizons, which are usually.calcareous.
The lower part of this sequence and its junction with the conglomerate is not exposed but Holliday et al.
HIGGINS AND VARKER: CONODONT FAUNAS
147
(1976, pp. 349, 350) estimated the thickness of the unexposed beds to be only slightly in excess of 42 m from
borehole evidence. Virtually all of the remaining part of the Stone Gill Beds is exposed in Stone Gill itself,
where they have been recorded and sampled, at intervals not exceeding 3 m, by the present authors (section
2, text-figs. 2, 3). The sequence includes a great variety of limestone lithologies, although calcite mudstones
predominate and large parts of the succession have suffered some dolomitization. This part of the Dinantian
succession also includes several of the marker bands described by Garwood (1913), in particular the Vaughania
Band and the Palaeechinus Bed, both of which occur near the base of the exposed succession.
|Permo-Triassic
| Westphalian
| j Namurian
! Dinantian
; Pre - Carboniferous
| ^ | Shop Granite
text-fig. 1 . Location of the Ravenstonedale Area.
The overlying Coldbeck Beds begin at the Spongiostroma Band of Turner (1950) and extend up to a nodular
algal band which outcrops beneath Coldbeck bridge at Ravenstonedale (NY 7209 0435). These beds, which
are also predominantly fine grained limestones, may be distinguished from the Stone Gill Beds by their sparser
fauna and the occurrence of several nodular algal layers. The algal layer described by Turner (1950) may be
one of these, since it appears to be slightly lower stratigraphically than the band taken by Ramsbottom ( 1 974,
p. 52) to mark the top of the unit. Exposure of the Coldbeck Beds in Stone Gill is almost complete and once
again conodont samples were taken at intervals not exceeding three metres (section 2, figs. 2, 3). It is, however,
likely that these samples do not fully represent the Coldbeck sequence since Turner (1950, pp. 30, 35) presented
evidence which suggested that approximately 30-35 m of beds had been removed from this stream section
by faulting.
The Stone Gill Beds and the Coldbeck Beds were considered by Ramsbottom (1973, p. 574) to represent
the transgressive and regressive phases of his first major cycle of the Dinantian, and they were both consequently
included in the Courceyan Stage by George et al. (1976). An inconsistancy was, however, noted by the latter
authors (1976, p. 38) in that foraminifera from the Stone Gill Beds included forms typical of the Vla
of Belgium, i.e. Chadian, not Courceyan. Holliday et al. (1979, p. 354) considered the then available
148
PALAEONTOLOGY, VOLUME 25
text-fig. 2. Geological map of the Ravenstonedale area showing the location of
the sampled sections (Based with permission upon one inch Geological Survey
Sheet No. 40)
palaeontological evidence for the position of the Chadian/Courceyan boundary to be inconclusive but
concluded on sedimentological grounds that the boundary should be lower than indicated by George et al.
and placed it, with reservation, high in the felspathic (Shap) conglomerate. On this basis, the whole of the
Stone Gill/Coldbeck succession would be of Chadian age.
The succession continues with the Scandal Beck Limestone, which was considered by George et al. (1976,
p. 38) to represent the whole of the Chadian Stage. These authors placed the top boundary of the unit, and
stage, in an unexposed part of the sequence some 1 5 m below the Brownber Pebble Bed and indicated a total
thickness for the unit of approximately 100 m. Johnson and Marshall (1971, p. 265) considered the same
interval to be closer to 125 m in thickness. The Scandal Beck Limestone was sampled for conodonts only
from approximately 10 m of partly dolomitized bioclastic limestone and calcite mudstone which occur near
the base of the unit. It is this sequence which outcrops at the confluence of Stone Gill and Scandal Beck
(section 3, text-fig. 2), which is represented in text-fig. 4.
The Brownber Pebble Bed, which forms a useful marker horizon in the region as an approximation for
the base of the Arundian Stage, is 4 to 6 m in thickness and consists of calcareous sandstone with quartz
pebbles, particularly in its upper part. Above the pebble bed the Arundian succession of Ravenstonedale may
be conveniently divided into two parts, the Michelinia grandis Beds and the overlying Ashfell Sandstone.
There are considerable variations in the published thicknesses attributed to these units. Turner (1950, p. 33)
described the M. grandis Zone as consisting of ‘nearly 150 feet (c. 46 m) of well bedded, somewhat iron-stained
HIGGINS AND VARKER: CONODONT FAUNAS
149
SG13 H I
gap0-30m
SG4 - P p p
O
LU
OQ SG23 H
gapOIOm^
s s s
s s
s s s
s s
STONE GILL
.roadbridge
> 4a fl~ A~~fpl
gapl-OOm
gap3-00m
OSG32-
LjJ
HI
o SG31-
EJ SG16-
O
gapl-OOm
gapOTOm
gap250m
— fault
krkp^n
text-fig. 3. Section in Stone Gill of the Stone Gill and Coldbeck Beds. For
legend see text-fig. 4.
limestones with a horizon of chert nodules 30 feet above the base’. Johnson and Marshall (1971, p. 265),
however, allocated 110 m to the M. grandis Zone and approximately 145 m to the whole interval between
the Brownber Pebble Bed and the base of the Ashfell Sandstone. Ramsbottom (1974, p. 55) indicated that
the beds of this same interval, which represent the transgressive phase of his third major cycle, reach about
240 m in the Ravenstonedale district. Finally, George et al. (1976, p. 38, fig. 11) allocated a thickness of
about 130 m. The ten conodont samples collected by the present authors (section 4, text-fig. 2) were collected
at regular intervals through the succession as exposed in Scandal Beck.
The Ashfell Sandstone is reported to reach 170 m in thickness in the Ravenstonedale area (Ramsbottom
1974, p. 57) but it does thin rapidly southwards. Only the top 13-25 m have been examined in the present
150
PALAEONTOLOGY, VOLUME 25
study at the well-known roadside exposures of the Ravenstonedale/Kirkby Stephen road, Ashfell Edge (sec-
tion 5, text-fig. 2). At this locality the Ashfell Sandstone is represented by massive, false-bedded sandstone
with slump structures, soft purple and green mudstones, and thin, sometimes lenticular, limestones, several
of which are highly fossiliferous, containing large, in situ, dendritic coral colonies (text-fig. 5). Turner (1950,
p. 34) considered these latter beds, the Lithostrotion martini horizon of Garwood (1913) to be confined to
the immediate vicinity of the Ashfell Edge locality.
The Arundian/Holkerian boundary is well exposed at this locality and is marked by the incoming of the
predominantly limestone sequence known as the Ashfell Limestone, which represents the whole of the Holkerian
Stage. The lowest 60 m of this unit was almost continuously exposed in the road section on Ashfell Edge
where, along with the upper part of the Ashfell Sandstone, it was measured and recorded by Dr. I. C. Burgess,
Mr. M. Mitchell, and Dr. W. H. C. Ramsbottom (Institute of Geological Sciences) in September 1972. The
sequence represented in text-fig. 5 is based upon their measured section (with their kind permission), but
unfortunately recent civil engineering work has obscured some of the higher horizons. A total of thirteen
conodont samples were taken from the horizons indicated (text-fig. 5). The upper part of the Ashfell Limestone
sequence is not continuously exposed although exposures, both natural and in quarries, are frequent enough
to allow conodont samples to be taken at fairly regular intervals. The total thickness of the Holkerian
succession on Ravenstonedale Common approaches 200 m.
SCANDAL BECK
limestone
dolomitised limestone
shale /mudstone
siltstone
sandstone
Pa/aeechinus bed
Vaughania band
Spongiostroma band
algal horizon
sample number
I.G.S. bed no. for Ashfell Edge
metres
text-fig. 4. Section of Scandal Beck Limestones seen in Scandal Beck
and Legend for text-figs. 3-5.
SB8-\C\E
SB4
SB3
gap1-90m
p p p p
V V V V
S S S S
A A A A
SG19
19
■3
HIGGINS AND VARKER: CONODONT FAUNAS
15:
The Asbian Stage is represented in this region by two main divisions, the Potts Beck Limestone (new name
George et al. 1976, p. 40) and the overlying Knipe Scar Limestone. Only the Potts Beck Limestone has been
sampled for conodonts in the present study, from a series of quarries on the north-west side of the
Ravenstonedale/Kirkby Stephen road (section 6, text-fig. 2). According to George et al. (1976) this limestone
is only found in the Stainmore Trough around Kirkby Stephen, in the I.G.S. Borehole at Raydale, and in
the River Clough near Sedbergh. Elsewhere there is a considerable non-sequence between the Holkerian and
Asbian. The Potts Beck Limestone forms the youngest stratigraphic horizon in this conodont study and was
the only limestone which did not yield faunas.
ASHFELL EDGE
X AL4 -
in
<
gap200m
36
co
§ AS3-
<
in
_l AS2-
l±J
Ll ASI-t
X
text-fig. 5. Section of Ashfell Sandstone and Ashfell Limestone seen
in road cutting at Ashfell Edge.
CONODONT FAUNAS
52
PALAEONTOLOGY, VOLUME 25
CONODONT FAUNAS
A sequence of four conodont faunas is recognizable in the Ravenstonedale succession. However,
because the element ranges of the lowest fauna are incompletely known this is not named as a
formal zone. The upper three zones are from an almost completely exposed and continuous sequence
and are named and defined with a greater degree of certainty (see text-fig. 6).
Fauna A
The elements of Fauna A have recently been described by Varker and Higgins (1979). They occur in the
outcrop section of Pinskey Gill and a series of dolomites and dolomitic limestones in the Pinskey Gill Borehole
at depths between 33 and 36 m. The fauna probably typifies the whole of the Pinskey Gill Beds, but is so
impoverished in both species and abundance that the full range of the elements is not known.
The fauna is dominated by Bispathodus aculeatus aculeatus (Branson and Mehl) and Clydagnathus unicornis
Rhodes, Austin and Druce together with hindeodellids and ozarkodinids. The association of these forms led
Varker and Higgins to conclude that the fauna was mid-Courceyan in age belonging to late K or early
Z Zones of the Avon Gorge and South Wales, and correlating with the costatus costatus / Gnathodus delicatus
Zone of Rhodes, Austin, and Druce (1969). Sevastopulo and Johnston (personal communication) would place
the fauna firmly in the Z-Zone in terms of the Irish sequence.
text-fig. 6. Stratigraphical ranges of conodont species in the Lower Carboniferous of the Ravenstonedale
area.
HIGGINS AND VARKER: CONODONT FAUNAS
53
Taphrognathus Zone
This is a partial range zone occurring in the upper part of the Stone Gill Beds and in the Coldbeck Beds.
The Shap Conglomerate sequence above the Pinskey Gill Beds and the lowest part of the Stone Gill Beds
do not contain conodonts. Faunas first appear some 40 m above the base of the exposed section in Stone
Gill and continue up to the top of the Coldbeck Beds at Coldbeck Bridge.
The fauna remains impoverished in species, but is somewhat richer in abundance than that of the Pinskey
Gill Beds. It is characterized by Taphrognathus varians Branson and Mehl, Cloghergnathus carinatus sp. nov.,
Spathognathodus scitulus (Hinde), Apatognathus cuspidatus Varker, and a variety of non-platform elements.
T. varians is probably the best-known species for it has a very wide distribution. In the Upper Mississippi
Valley and Missouri (Collinson, Rexroad, and Thompson 1972) it characterizes the middle part of the
Valmeyeran Series (Upper Osage/Lower Meramec) ranging from the Keokuk to the lower part of the Upper
St. Louis Formations. It is also known from Australia (Jenkins 1974) in strata which Jenkins concluded were
of early Visean age. Taphrognathus has also been recorded from Britain in the Main Algal Limestone of
Roxburghshire (Rhodes et al. 1969), where it is of early Visean age. Finally it occurs in Ireland (Austin and
Mitchell 1975) again in the early Visean. Cloghergnathus also occurs with Taphrognathus in Ireland as does
Spathognathodus scitulus. S. scitulus is characteristic of the St. Louis Formation, Upper Valmeyeran in the
Upper Mississippi Valley, where it overlaps the range of Taphrognathus although the closely related species
S. coalescens occurs earlier and coexists with the greater part of the range of Taphrognathus (fig. 7).
The Taphrognathus Zone correlates broadly with the Bactrognathus-Taphrognathus Zone which occurs in
the upper part of the Burlington Formation of the Upper Mississippi Valley. In Europe this fauna is uncommon
but does occur in Ireland (Austin and Mitchell 1975) with the characteristic early Visean Mestognathus
beckmani and it was given a lower Visean age. George et al. (1976) referred the Stone Gill Beds and the
Coldbeck Beds to the late Courceyan, but the presence of Vla foraminifera (Ramsbottom 1977) clearly
indicates an early Visean age for these sediments, and they are now referred to the Chadian (Ramsbottom 1977).
UPPER MISSISSIPPI VALLEY
Collinson, Rexroad * Thompson 1972
S. W. MISSOURI
Thompson ♦ Fellows
1970
BELGIUM
BRITAIN
George et al 1976
RAVENSTONEDALE
formations
conodont zones
conodont zones
assise
stages
formations
conodont zones
Apot. scalenus
V3b
Asbian
St. Louis
Cavusgnathus
V3a
v2b
Holkerian
Ashfell Limestone
Cavusgnathus
V2a
Ashfell Sandstone
—
Warsaw- Salem
Taph varians
Apatognathus
vjb
Arundian
M. grandis Beds
B
Cloghergnathus
Keokuk
Gnathodus texanus -Taphrognathus
Scandal Beck
Limestones
A
V
Chadian
G. butbosus
Coldbeck Beds
Bactrognathus
B. distortus
Stone Gill Beds
Taphrognathus
Burlington
Taphrognathus
G. cuneiformis
Bactrognathus
Bactrognathus
Tn3
Shap
Fern Glen
P. communis
Ps. muttistriatus
Courceyan
Conglomerate
G. semigtaber
Meppen
Ps. muttistriatus
/ G. semigtaber
/ P comm, carina
Pinskey Gill
Fauna A
S. isosticha
\ S. cooperi hassi
\ 6 punctotus
In2
Chouteau
S. cooperi
G. det[catus
S. cooperi cooperi
text-fig. 7. Correlation of the Ravenstonedale conodont and sedimentary sequence with the Belgian Assises
and the North American conodont sequences.
154
PALAEONTOLOGY, VOLUME 25
Taphrognathus sp. also occurs in the Lower Windsor Group of Nova Scotia (von Bitter 1976) although its
occurrence with Cavusgnathus spp. probably indicates a younger age than the Taphrognathus Zone of
Ravenstonedale.
Cloghergnathus Zone
This assemblage zone occurs in the Scandal Beck Limestone and the Michelinia grandis Beds. Every sample
has yielded a fauna although none of them are abundant.
The characteristic species present are Cloghergnathus carinatus, Spathognathodus scitulus, Apatognathus
asymmetricus, and A. scandalensis. The appearance of Magnilaterella robust a and Neoprioniodus singular is at
the base of the Michelinia grandis Beds allows the subdivision of this zone into two subzones: a lower one
with Cloghergnathus, Apatognathus asymmetricus, and A. scandalensis, and an upper one with Neoprioniodus
singularis and Magnilaterella robusta. Both of the latter species are known to have longer ranges elsewhere
and their appearance within the zone may well be due to environmental change at the base of the Michelinia
grandis Beds. The fauna of the zone is not distinctive and is merely an interregnum between the disappearance
of Taphrognathus and the appearance of Cavusgnathus.
This interval compares closely to the Taphrognathus varians- Apatognathus interval in the Upper Mississippi
Valley (Collinson et al. 1972) which occupies the Warsaw, Salem, and lower St. Louis Formations. In Ireland
(Austin and Mitchell 1975) there is a broad zone between the top of the Lower Carboniferous Shale (C^)
with T. varians, and the upper Calp Limestone (D,), with Cavusgnathus sp., of which the lower part would
correspond to the Cloghergnathus Zone.
George et al. (1976) dated the Scandal Beck Limestone (Subzone A) as Chadian and the Michelinia grandis
Beds (Subzone B) as Arundian. According to George et al. (1976) this part of the sequence is largely missing
in the Avon Gorge or is represented by non-conodont bearing strata (Austin 1973).
Cavusgnathus Zone
The Cavusgnathus Zone includes the highest beds of the Ashfell Sandstone and at least the lowest beds of the
Ashfell Limestone. The fauna includes Cavusgnathus regularis and C. unicornis, Neoprioniodus scitulus,
Spathognathodus scitulus, Magnilaterella robusta, and Apatognathus libratus.
This fauna correlates with that of the Apatognathus scalenus-Cavusgnathus Zone of the Upper Mississippi
Valley (Collinson et al. 1972) which occurs in the Upper St. Louis Formation. However, as Austin (in Austin
and Mitchell 1975) has pointed out, there is a breccia between the Upper and Lower St. Louis Formation
with some condensation of faunas. This may account for the absence of the major gap between the disappearance
of Taphrognathus and the appearance of Cavusgnathus which occurs in the Ravenstonedale sequence. This
zone compares well with the lower part of the Cavusgnathus- Apatognathus Zone of the Avon Gorge (Austin
1973) (text-fig. 8).
The Ashfell Sandstone is included in the upper part of the Arundian and the Ashfell Limestone in the
Holkerian by George et al. (1976).
Influence of environment on the conodont faunas
The influence of the depositional environment on the conodont animal in the Palaeozoic generally is
readily apparent from a perusal of the symposium volume on conodont palaeoecology (Barnes
1976), and in the Lower Carboniferous in particular from the work of Austin (1976) and von Bitter
(1976). There is a clear distinction between extremely shallow-water faunas seen in the Windsor
Group of Nova Scotia (Globensky 1967; von Bitter 1976) and Ravenstonedale and the basinal
faunas of the same age in Spain (Higgins and Wagner Gentis, in press) and in Germany (Bischoff
1957; Voges 1959). Other faunas show mixing of both shallow-water and basinal types, and Austin
(1976) has attempted to show typical associations of this type. Although these differences are
normally ascribed to a depth control on the conodont animal, there may be also an element of
provincialism as suggested by Austin (1976), for the type of fauna seen in Ravenstonedale appears
to have a restricted geographical distribution. In view of the growing evidence of environmental
control of the conodont animal it is important to know whether the faunal changes in the
Ravenstonedale sequence reflect evolutionary or environmental changes, or both.
Clydagnathus, occurring in dolomites and dolomitic limestones in Pinskey Gill is, according to
Austin (1976), found in dolomites and oolites in littoral/deltaic and offshore neritic facies only.
HIGGINS AND VARKER: CONODONT FAUNAS
.55
STAGES
George et al 1976
AVON
GORGE
RAVENSTONEDALE
conodont zones
this paper
George et al 1976
Kellaway 8 Welch 1955
Rhodes, Austin 8 Druce 1969
Austin 1973
Holkerian
Upper Clifton
Down Limestone
Cavusgnathus
Apatognathus
Cavusanathus
Arundian
Lower Clifton Down Limestone
Upper Clifton Down Mudstone
Goblin Coombe Oolite
Cavusgnathus - Apatognathus
no conodonts
B
Cloghergnathus
Lower Clifton Down Mudstone
no conodonts
Chadion
Gully Oolite
Sub -Oolite Bed
Mestognathus beckmanni
Polygnathus bischoffi
A
Taphrognathus
\\VV\VV
\\ V\\V\VVV^
G. an tet exanus
Ptacinatus
Ptacinatus - Ps.tongiposticus
Courceyon
Black Rock Limestone
Bispathodus costatus costatus
G. del i cat us
Fauna A
Lower Limestone Shale
poor exposure
S. cf. robustus
B. acuteatus
Siphonodella - Pinornatus
Shirehampton Beds
Pa. variabitis - Pinornatus
text-fig. 8. Comparison of the conodont zonation of Ravenstonedale with that of the Avon Gorge.
and is typically supratidal. Sandberg (1976), in his study of late Devonian biofacies, concluded
that Clydagnathus is abundant in offshore banks and lagoons of shallow brackish to normally
saline banks commonly occurring with algae.
Taphrognathus, according to Austin has a similar distribution to that of Clydagnathus , and von
Bitter (1976) came to a similar conclusion in relating Taphrognathus to his Biofacies II which
occurred in inner shelf and reefoid environments.
Apatognathus and Spathognathodus scitulus generally occur together and were suggested by
Austin (1976) to be an association which occurred in the littoral/delta front, offshore neritic, and
back reef lagoonal facies. Von Bitter grouped these two with his Biofacies II association.
Cavusgnathus has a similar shallow-water origin to Taphrognathus according to Austin (1976),
von Bitter (1972, 1976), and Merrill and Martin (1976).
Cloghergnathus is a relatively new genus and its distribution is poorly known, but it does occur
156
PALAEONTOLOGY, VOLUME 25
in the Windsor Group of Nova Scotia (Globensky 1967; von Bitter 1976) where it would probably
be placed in von Bitter’s biofacies II, and it also occurs in Ireland (Austin and Mitchell 1975)
where it occurs in association with Taphrognathus.
The majority of the platform conodonts in these faunas are asymmetric: Clydagnathus unicornis
and all species of Cavusgnathus are right-sided whereas Cloghergnathus carinatus is both left and
right sided and Taphrognathus varians often shows a tendency towards asymmetry. The asymmetry
appears to be associated with shallow-water environments, but is not restricted to these environ-
ments. Cavusgnathus , for example, may occur in basinal environments (Higgins 1975), and so may
Apatognathus but in small numbers only. They are only dominant in shallow-water faunas. Similarly,
the basinal faunas of this age, dominated by the symmetrical or highly ornamented platform
elements, may occur on the margins of the shallow-water zones mixed with asymmetric elements.
Von Bitter recorded the presence of simple gnathodids in his biofacies II, and more complex
gnathodids occur with cavusgnathids and apatognathids in his deep-water biofacies III. In the
Ravenstonedale sequence such mixing occurs rarely, indicating the extreme shallowness of the
water. Neoprioniodus singularis, which is a common basinal species, occurs as an uncommon
component of the Michelinia grandis Beds but its origin is known to be earlier in basinal sequences
and it must be regarded as a deeper-water immigrant into a dominantly shallow-water environ-
ment.
Thus Clydagnathus, Cloghergnathus, Taphrognathus, Cavusgnathus, and Apatognathus are
environmentally controlled to a high degree, occurring typically in littoral and lagoonal facies
but ranging into offshore infratidal facies. The faunas in which they occur are restricted in both
variety and abundance. Their persistent occurrence throughout the Ravenstonedale sequence
is a reflection of the persistence of this facies in the area, and has enabled us to study the
development of these unusual faunas through a long sequence. The proposed zones, although
fundamentally environmentally controlled, are of biostratigraphic usage within this type of
environment.
Neoprioniodus singularis, on the other hand, is clearly of only local value as a stratigraphic
marker and merely reflects the major transgression which occurred at the base of the Michelinia
grandis Beds (see also Leeder in discussion of George 1978) and brought with it immigrant faunas
into the area.
In broad outline the changes in the conodont faunas occur both at stage boundaries and at the
cycle boundaries of Ramsbottom (1973, 1977), although the two do not exactly coincide. The
boundary between the Taphrognathus -Cloghergnathus Zones is coincident with the algal horizon
at the top of the Coldbeck Beds which, according to Ramsbottom (1973, 1977), is the boundary
between two cycles. Similarly the boundary between Subzones A and B of the Cloghergnathus
Zone coincides with the Chadian-Arundian boundary although the exact horizon of the boundary
is thought to be unexposed (George et al. 1976). The beginning of the Cavusgnathus Zone coincides
with the appearance of the first limestone above the sandstone beds of the Ashfell Sandstone.
These occur in a few metres of shales immediately below the base of the Ashfell Limestone which
is taken as the boundary between the Arundian and Holkerian Stages, and the boundary between
cycles 3 and 4 of Ramsbottom (1973).
With the exception of the Subzone B fauna, all the conodont faunas are of a shallow-water type
and they occur throughout the cycles in the Chadian, through the lower part of the Arundian, and
reappear at the top of the Arundian. The general distribution of the faunas supports the attribution
of the beds to the stages of George et al. (1976), but the similarity between the faunas is probably
due to the position of the Ravenstonedale area during Lower Carboniferous times at the head of
a shallow gulf where a change of sea level to bring in sandstone or algal horizons led to the absence
of conodonts rather than a change in the type of fauna.
HIGGINS AND VARKER: CONODONT FAUNAS
.57
SYSTEMATIC DESCRIPTIONS
All type and figured material has been lodged in the micropalaeontological collection. Department
of Geology, University of Sheffield.
Genus Apatognathus Branson and Mehl
Type species. Apatognathus varians Branson and Mehl, 1934
Apatognathus asymmetricus sp. nov.
Plate 19, figs. 7, 9
Holotype. PI. 19, fig. 7. R21.
Horizon. Scandal Beck Limestone, sample SB9.
Diagnosis. Anterior process, laterally thickened, half the height of the posterior process which is
thin. Denticles on posterior process fused and twice the height of those on the anterior process.
Description. Processes unequal, diverging at about 40°. Strongly inwardly curved. Anterior process low, laterally
thickened with a shelf on the upper edge, and maintaining its height to the extremity. Up to ten denticles on
the oral surface whose inclination increases towards the cusp. The denticles are small, laterally compressed,
and are incurved. They are discrete for about half their length.
Posterior process thin, but high, up to twice the height of the anterior process and the inner face is strongly
convex. The denticles are twice as high as those of the anterior process and are fused for most of their length.
They are strongly laterally compressed and decrease in height towards the extremity of the process. The
denticles are slightly incurved.
Cusp is twice the height and width of the posterior bar denticles. The adjacent posterior bar denticle is
fused to its posterior edge, but it is largely free on its anterior edge due to the small size of the anterior
process denticles.
The cavity is triangular and is situated on the inner face of unit below the cusp. The processes are grooved
on the aboral side.
Comparison. This species most closely resembles Apatognathus cuspidatus from which it differs in
lacking lateral thickening on the posterior process and having more strongly fused denticles. It
differs from A. scandalensis in having lateral thickening on the anterior process.
Apatognathus libratus Varker
Plate 19, fig. 12
1967 Apatognathus? libratus Varker pp. 134, 135, pi. 18 figs. 3, 6, 8, 9, 12, 13.
Remarks. This is an extremely large specimen, but its major characteristics conform to the diagnosis
of Apatognathus libratus. The main difference lies in the width of the process beneath the cusp
which is twice as broad as in other figured specimens of this species.
Apatognathus scandalensis sp. nov.
Plate 19, fig. 10
Holotype. PI. 19, fig. 10, R18.
Horizon. Scandal Beck Limestone, sample SB8.
Diagnosis. A wide angled, almost symmetric, Apatognathus with thin processes and denticles on
the anterior process which are highest near, but not at, the cusp. The denticles on the anterior
process are twice the height of those on the posterior process.
158
PALAEONTOLOGY, VOLUME 25
Description. The processes diverge at about 50° and are strongly inwardly curved. The unit is slender and
thickening of the processes only occurs close to the cusp, where a slight shelf is formed.
The anterior limb is slightly longer than the posterior and is strongly curved inwardly. It is higher than
the posterior limb and bears up to thirteen laterally compressed denticles which rise to a point about one-third
from the cusp. The denticles are sharply pointed and are flatter on the outer than the inner side and are
discrete for more than half their length. The posterior process is about two-thirds the length of the anterior
process and is inwardly curved about the same amount. Only slight thickening occurs near the cusp. Up to
ten denticles occur on its oral surface and these are only half the height of the largest anterior process denticles
and are subequal.
The cusp is twice the height and width of the largest denticle on the processes, and is laterally compressed
more strongly on the outer than the inner side. It is discrete for more than half its length and is sharply
pointed. It is slightly curved inwards.
A shallow triangular pit occurs at the base of the cusp and is situated on the inner face rather than aborally.
It is continued along the processes as barely discernible grooves.
Comparison. Apatognathus scandalensis most closely resembles A. cuspidatus but differs in having
a less strongly inclined and a smaller cusp and a larger angle between the processes.
Genus Cavusgnathus Harris and Hollingworth, 1933
Type species. Cavusgnathus altus Harris and Hollingworth, 1933
Cavusgnathus regularis Youngquist and Miller
Plate 18, figs. 12, 13
1949 Cavusgnathus regularis Youngquist and Miller, p. 169, pi. 101, figs. 24, 25.
Remarks. The generally short and compact form and the regular denticulation of the blade of this
species are present in the Ravenstonedale specimens, but there are some differences from previously
described examples. The principal difference is the length of the fixed blade which extends up to
half the length of the platform and is larger than that normally found on specimens of this species.
EXPLANATION OF PLATE 18
All figures x 60
Figs. 1-11. Cloghergnathus carinatus. P. element. 1, oral view of specimen from the Stone Gill Beds, sample
SGI 9, specimen R29, transitional to Taphrognathus varians. 2, 7, oral and outer lateral views of holotype,
specimen from the Scandal Beck Limestone, sample SB2, specimen R30. 3, inner lateral view of a specimen
from the Stone Gill Beds, sample SG19, specimen R32. 4, oral view of a young specimen from the Coldbeck
Beds, sample SG32, specimen R31. 5, outer lateral view of a young specimen from the Michelinia grandis
Beds, sample MB10, specimen R35. 6, 10, outer lateral and oral views of a specimen from the Scandal
Beck Limestone, sample SB9, specimen R33. 8, 9, 11, aboral, outer lateral and oral views of a specimen
from the Coldbeck Beds, sample SG32, specimen R34, specimen transitional to Taphrognathus varians.
Figs. 12, 13. Cavusgnathus regularis. Oral and inner lateral views of a specimen from the Ashfell Sandstone,
sample AS1, specimen R36.
Fig. 14. Cavusgnathus unicornis. Inner lateral view of a specimen from the Ashfell Sandstone, sample AS1,
specimen R38.
Figs. 15, 16. Taphrognathus varians. Oral and inner lateral views of specimens from the Stone Gill Beds,
sample SGI 9, specimens R37, R39.
Fig. 17. Lonchodina sp. Inner lateral view of specimens from the Coldbeck Beds, sample SG33, specimen R27.
Fig. 18. O element of Cloghergnathus carinatus. Inner lateral view of a specimen from the Coldbeck Beds,
sample SG32, specimen R14.
Fig. 19. A, element of Cloghergnathus carinatus. Inner lateral view of a specimen from the Coldbeck Beds,
sample SG31, specimen R3.
PLATE 18
HIGGINS and VARKAR, Lower Carboniferous conodonts
160
PALAEONTOLOGY, VOLUME 25
Cavusgnathus unicornis Youngquist and Miller
Plate 18, fig. 14
1949 Cavusgnathus unicornis Youngquist and Miller, p. 619, pi. 101, figs. 18-23.
Remarks. The median carina developed only in the posterior part of the platform is prominent
only in large specimens and is often completely absent in small specimens.
Genus Cloghergnathus Austin and Mitchell
Type species. Cloghergnathus globenskii Austin and Mitchell, 1975.
Remarks. Cloghergnathus was named by Austin (in Austin and Mitchell 1975) for specimens with
a lateral blade which does not extend on to the platform as a high, crested structure as in
Cavusgnathus. The blade may occupy a sub-median position whilst remaining connected to one of
the platform sides and such forms are probably transitional to Taphrognathus. Its characteristics
are those of Clydagnathus Rhodes, Austin, and Druce an early Courceyan genus but a considerable
stratigraphic gap occurs between the last occurrence of Clydagnathus and the first appearance of
Cloghergnathus in the Chadian.
Cloghergnathus carinatus sp. nov.
P element
Plate 18, figs. 1-11
Holotype. PI. 18, figs. 2, 7. R30.
Horizon. Scandal Beck Limestone, sample SB 2.
Diagnosis. A right- and left-sided element with the anterior blade developed on the inner side of
the unit. A central trough runs the length of the platform but is occupied by a short carina in the
posterior quarter. The blade is short, one-quarter to one-third the length of the platform, and does
not extend on to the platform. It is extended above the platform and is convexly curved. The unit
is arched.
Description. The anterior blade consists of up to five denticles of which the largest is in the centre giving the
blade a convex outline. The denticles are free for about half their length. The platform is straight to strongly
curved and bears a deep median trough which is occupied by a carina for up to the last third of the platform.
The carina is very prominent in small specimens. The inner parapet is curved and commonly does not reach
the posterior end of the unit and is continued anteriorly as the blade, although the curvature of some specimens
is difficult to determine because the blade may be outwardly curved. In a few specimens the blade is slightly
offset and, whilst it originates from the inner side of the platform, it occupies a sub-central position on both
right and left forms. The oral surface of the platform is covered by nodes or transverse ridges. The platform
is widest near midlength and commonly the inner side is wider than the outer. The unit is arched in lateral
view both the oral and aboral surfaces of the platform being curved and the aboral surface of the blade being
strongly downturned. The curvature is less marked in young specimens which may be almost straight. The
cavity occupies the whole of the aboral surface of the platform and is widest slightly anterior to its mid-point
tapering to both the anterior and the posterior.
Comparisons. Cloghergnathus carinatus differs from C. globenskii Austin and Mitchell by its arched
aboral and oral surfaces, its convex blade, which originates from the inner side, and the presence
of the posterior carina. C. rhodesi is left-sided, does not possess a carina, and its blade shape is
unknown.
Remarks. There is a tendency for the blade of C. carinatus to be median in position and it may
include some of the Taphrognathus-Cavusgnathus transition forms of Rexroad and Collinson (1963;
especially fig. 24, p. 111). Austin (in Austin and Mitchell 1975) has argued that Cloghergnathus is
HIGGINS AND VARKER: CONODONT FAUNAS
61
not intermediate between Taphrognathus and Cavusgnathus but is an offshoot of the former genus
in which the blade, although lateral, does not extend on to the platform. In this respect
Cloghergnathus mirrors Adetognathus which is probably a younger homeomorph, or which may
prove to be the same genus when its full range is known. All figured specimens of Cloghergnathus
have a low blade or one which rises gradually as in C. carinatus none of them have the typical
cavusgnathoid blade. In addition, all known species of Cloghergnathus are either right- and left-sided
or only left-sided, whereas Cavusgnathus species are all right-sided. C. carinatus is also similar to
Clydagnathus darensis but differs in having a larger basal cavity.
Cloghergnathus non-platform Elements
O element
Plate 18, fig. 18; Plate 19, fig. 4
1957 Ozarkodina compressa Rexroad, p. 36, pi. 2, figs. 1, 2.
Remarks. The Ravenstonedale specimens have the prominent apical denticle of the paratype but
have a smaller number of denticles on the processes than is usual in this element. However, it is
thought to fall within the variability of the element.
N element
Plate 19, fig. 18
1941 Prioniodus varians Branson and Mehl, p. 174, pi. 5, figs. 7, 8.
1957 Prioniodina varians (Branson and Mehl), Bischoff, p. 49, pi. 5, fig. 35.
1957 Neoprioniodus varians (Branson and Mehl), Rexroad, p. 35, pi. 2, fig. 10.
Remarks. Von Bitter (1976) named Neoprioniodus camurus Rexroad as the Ne element in his
multi-element reconstruction of Cavusgnathus windsorensis. He also figured a similar form in his
reconstruction of the non-P elements of Cavusgnathus sp. from the Pennsylvanian (1972, pi. 9,
fig. 5). Baesemann (1973) figured a similar neoprioniodontid in his reconstruction of Adetognathus
(pi. 2, figs. 26, 35). No form comparable to N. camurus occurs in the Ravenstonedale samples
where the only neoprioniodontids are N. varians , N. acampylus, and N. scitu/us.
A: element
Plate 18, fig. 19; Plate 19, fig. 20
1953 Hindeodella ensis Hass, p. 81, pi. 16, figs. 19-21.
1960 Hindeodella tenuis Clarke, p. 8, pi. 1, figs. 10, 11.
Remarks. There are two variants in this element. One is typified by Hindeodella tenuis and H. ensis
and is the more common of the two. This has a long posterior process with alternating denticles
and a long, often inwardly curved anterior bar commonly with non-alternating denticles which
are as large as the large posterior bar denticles. The other variant is shorter and arched with a
cusp which is as long as the posterior process, and has non-alternating denticles of which the
posterior process denticles increase in size posteriorly.
A 3 element
Plate 19, figs. 5, 6, 8
Description. A robust unit consisting of two massive lateral processes, a large cusp, and a slender posterior
process. The cusp is large, slightly curved posteriorly, and has a triangular cross-section with sharp posterior
and lateral edges. Its surface is covered with fine discontinuous striations. The posterior bar is incomplete,
appears to be narrow, and has an oval cross-section.
The lateral processes are massive, bar-like with a square cross-section, and almost in the same plane. They
162
PALAEONTOLOGY, VOLUME 25
are as long as the main cusp and slightly inclined downwards. The denticles on the oral surface, up to five
in number, are large, increasing in size towards the extremities of the processes, and are flattened in an
antero-posterior direction. The largest denticle is approximately half the size of the cusp.
The aboral side of the unit is grooved, and the grooves meet forming a small triangular pit beneath the cusp.
Remarks. This form most closely resembles Hibbardella ortha Rexroad, the primary difference
being the shallow angle between the processes. It is also much more massive than H. ortha but
this is a characteristic shared by specimens of all species in the Ravenstonedale faunas and is most
likely due to the shallow-water environments in which the fauna occurs.
Genus Magnilaterella Rexroad and Collinson
Type species. Magnilaterella robust a Rexroad and Collinson 1963.
Magnilaterella robusta Rexroad and Collinson
1940 Lonchodina sp. Branson and Mehl, p. 171, pi. 5, fig. 10 only.
1956 Metalonchodinal sp. Elias p. 124, pi. 4, fig. 3.
1963 Magnilaterella robusta Rexroad and Collinson, pp. 14-17, pi. 1, figs. 4, 5, 9.
Remarks. Typically Magnilaterella robusta has a prominent bevel on the inner side of the lateral
process. However, Rexroad and Collinson (1965) did include specimens without the bevel which
have the characteristic denticulation on the lateral process.
At the present time the multi-element species to which M. robusta belongs is unknown.
EXPLANATION OF PLATE 19
All figures x 60
Figs. 1-3. Lonchodina sp. Inner lateral views of specimens from the Coldbeck Beds. Fig. 1 from sample SG33,
specimen Rl; fig. 2 from sample SG31, specimen R2; fig. 3 from sample SG33, specimen R40.
Fig. 4. O element of Cloghergnathus carinatus. Inner lateral view of a specimen from the Coldbeck Beds,
sample 29, specimen R26.
Figs. 5, 6, 8. A 3 element of Cloghergnathus carinatus. Posterior views of specimens from the Coldbeck Beds,
sample 29, specimens R9-11.
Figs. 7, 9. Apatognathus asymmetricus. 7, inner lateral view of holotype from the Scandal Beck Limestone,
sample SB9, specimen R21. 9, inner lateral view of a specimen from the Scandal Beck Limestone, sample
SB7, specimen R20.
Fig. 10. Apatognathus scandalensis. Inner lateral view of holotype from the Scandal Beck Limestone, sample
SB8, specimen R18.
Figs. 11, 13, 19. Apatognathus cuspidatus. Inner lateral views of specimens from the Scandal Beck Limestone,
fig. 11 (SB7), specimen R16 and fig. 19 (SB2), specimen R17, and Coldbeck Beds sample SG33, specimen
R15.
Fig. 12. Apatognathus libratus. Inner lateral view of a specimen from the Ashfell Sandstone, sample AS1,
specimen R19.
Fig. 14. Spathognathodus scitulus. Outer lateral view of a specimen from the Ashfell Sandstone, sample AS1,
specimen R5.
Fig. 15. Neoprioniodus singularis. Inner lateral view of a specimen from the Michelinia grandis Beds, sample
MB2, specimen R8.
Fig. 16. Neoprioniodus cf. acampylus. Inner lateral view of a specimen from the Michelinia grandis Beds,
sample MB2, specimen R13.
Fig. 17. Neoprioniodus sp. Inner lateral view of a specimen from the Coldbeck Beds, sample SG29, speci-
men R7.
Fig. 18. N element of Cloghergnathus carinatus. Inner lateral view of a specimen from the Scandal Beck
Limestone, sample SB9, specimen R12.
Fig. 20. A, element of Cloghergnathus carinatus. Inner lateral view of a specimen from the Coldbeck Beds,
sample SG29, specimen R24.
PLATE 19
HIGGINS and VARKAR, Lower Carboniferous conodonts
164
PALAEONTOLOGY, VOLUME 25
Genus Neoprioniodus Rhodes and Muller
Type species: Prioniodus conjunctus Gunnell 1931.
Neoprioniodus cf. acampylus Rexroad and Collinson
Plate 19, fig. 16
1965 Neoprioniodus acampylus Rexroad and Collinson, p. 11, pi. 1, figs. 25-27.
1968 Neoprioniodus acampylus Rexroad and Collinson, Thompson and Goebel, p. 37, pi. 3, fig. 2.
Remarks. This specimen conforms to the general outline and description of Neoprioniodus acampylus
but differs in possessing a more robust posterior process which is laterally thickened, and it lacks
the flaring inner lip of that species. It differs from N. camurus Rexroad in being unbowed. This
species has been recorded only from the Warsaw and Salem Formations of the Upper Mississippi
Valley (Rexroad and Collinson 1965) and the Meramacian of Indiana (Thompson and Goebel
1968). It occurs at a similar position in Ravenstonedale being confined to the Michelinia grandis Beds.
Neoprioniodus sp.
Plate 19, fig. 17
Description. Robust unit consisting of cusp and a short posterior bar. The posterior bar is laterally thickened,
approximately square in cross-section, and is about half the length of the cusp, but is not complete. Its oral
surface bears four denticles which are discrete for more than half their length and are laterally compressed
but more strongly so on the inner side. They are incurved and of varying length, the highest being the
penultimate one on the cusp. The bar is straight and is bevelled on the inner face.
The cusp is four times the length of the posterior process denticles and is strongly incurved and slightly
curved posteriorly. Its outer side is convex but the inner side has a lip along both margins which continues
down to the anticusp. The anticusp is short and appears to be pointed but the tip is broken. The aboral edge
of the cusp is strongly bevelled but the bevelling does not continue down the anticusp.
The cavity is elliptical, small, and is continued as a faint groove beneath the posterior process.
Genus Lonchodina Ulrich and Bassler
Type species: Lonchodina typicalis Ulrich and Bassler 1926.
Lonchodina sp.
Plate 18, fig. 17; Plate 19, figs. 1-3
Description. The unit includes both right and left forms; it is bowed. The cusp is large, laterally flattened
with sharp edges, strongly inwardly curved, and posteriorly inclined and curved. Typically the cusp narrows
gently towards the oral tip but in some specimens the base is very wide. The processes are the same height
and length and both are laterally thickened but the height is twice the thickness. The posterior process is
almost at right angles to the cusp and its oral surface bears three to six laterally compressed denticles of
variable size, but the two penultimate ones are the largest and one of them may be as large as the cusp. The
denticles are inwardly curved and are in the same plane as the cusp, but are more strongly posteriorly inclined.
The anterior process bears three to five denticles on its oral surface which are subequal in size and half the
size of the cusp. They are more founded than the posterior process denticles and are posteriorly and inwardly
curved. Both processes are strongly bevelled and striated on the inner side and inclined inwardly on the upper
side, giving the processes a triangular cross-section. The inner bevelled side is markedly striated on the lower
half. There is a small lip over the basal cavity. The aboral side is narrow with a small triangular pit beneath
the cusp and a narrow groove beneath the processes.
Remarks. The multi-element apparatus to which this distinctive form belongs is unknown.
Genus Spathognathus Branson and Mehl
Type species: Spathodus primus Branson and Mehl 1941.
Spathognathodus scitulus (Hinde)
Plate 19, fig. 14
1900 Polygnathus scitulus Hinde (part), p. 343, pi. 9, figs. 9, 11 only.
1949 Spathognathodus scitulus (Hinde) Youngquist and Miller, p. 622, pi. 101, fig. 4.
HIGGINS AND VARKER: CONODONT FAUNAS
165
Remarks. Austin and Rhodes recorded a fused cluster consisting of one specimen of Spathognathodus
scitulus and four specimens of Apatognathus sp. from the Lower Carboniferous of the Avon Gorge.
Undoubtedly the forms have a similar range and commonly occur in the same samples. Baesemann
(1973) described a multi-element species, Ozarkodina minuta which includes a P element, S. minutus
which is very similar to S. scitulus. Similarly, von Bitter (1976) suggested that S. cristulus, also
morphologically similar to S. scitulus, could be the P element of his Ellisonia sp. apparatus. However,
neither Baesemann’s nor von Bitters’s accompanying element are present in the Ravenstonedale
sequence, and the Austin and Rhodes model may be the more likely for the S. scitulus apparatus.
Genus Taphrognathus Branson and Mehl
Type species: Taphrognathus varians Branson and Mehl 1941.
Taphrognathus varians Branson and Mehl
Plate 18, figs. 15, 16
1941 Taphrognathus varians Branson and Mehl p. 182, p. 6, figs. 27-33, 35-40.
Remarks. A short carina occurs in the posterior quarter of the median sulcus which occurs on a
few but not the majority of previously figured specimens of this species. At the present time such
specimens are included within Taphrognathus varians.
Acknowledgements. We thank Dr. W. H. C. Ramsbottom for helpful discussion on the stratigraphy of the
Ravenstonedale area. The work was supported by a N.E.R.C. research grant which is gratefully acknowledged.
REFERENCES
Austin, R. l. 1968. Five Visean conodont horizons in the north of England. Geol. Mag. 105 (4), 367-371.
— 1973. Modification of the British Avonian conodont zonation and a reappraisal of European Dinantian
conodont zonation and correlation. Ann. Soc. geol. de Belgique, 96, 523-532.
— 1976. Evidence from Great Britain and Ireland concerning west European Dinantian conodont
paleoecology. Geol. Soc. Canada Sp. Paper No. 1 5, 202-224.
— and mitchell, m. 1975. Middle Dinantian platform conodonts from County Fermanagh and County
Tyrone, Northern Ireland. Bull. geol. Surv. Gt. Britain, 55, 43-54.
— and Rhodes, f. h. t. 1969. A conodont assemblage from the Carboniferous of the Avon Gorge.
Palaeontology, 12, 400-405.
baesemann, j. f. 1973. Missourian (Upper Pennsylvanian) conodonts of northeastern Kansas. J. Paleont. 47,
689-710.
barnes, c. R. 1976 (ed.). Conodont paleoecology. Geol. Soc. Canada Sp. Paper No. 15, 1-324.
bischoff, G. 1957. Die Conodonten-Stratigraphie des rheno-herzynischen Unterkarbons mit Berucksichtigung
der Wocklumeria-Stufe und der Devon/Karbon Grenze. Abh. hess. Landesmt. Bodenf, 19, 1-64.
capewell, J. G. 1955. The Post-Silurian pre-marine Carboniferous sedimentary rocks of the eastern side of the
English Lake District. Q. Jl geol. Soc. Lond. Ill, 23-46.
collinson, c., rexroad, c. b. and Thompson, t. l. 1971. Conodont zonation of the North American
Mississippian. In sweet, w. c. and Bergstrom, s. (eds.). Symposium on conodont biostratigraphy. Mem.
geol. Soc. Amer. 127, 353-394.
Garwood, e. J., 1907. Notes on the faunal succession in the Carboniferous Limestone of Westmorland and
neighbouring portions of Lancashire and Yorkshire. Geol. Mag. 44, 70-74.
— 1913. The Lower Carboniferous succession in the northwest of England. Q. Jlgeol. Soc. Lond. 68, 449-596.
george, t. N. 1978. Eustasy and tectonics: sedimentary rhythms and stratigraphical units in British Dinantian
correlation. Proc. Yorks, geol. Soc. 42 (13), 229-262.
JOHNSON, G. A. L., MITCHELL, M., RAMSBOTTOM, W. H. C., SEVASTOPULO, G. M. and WILSON, R. B. 1976. A
correlation of Dinantian rocks in the British Isles. Geol. Soc. London, Special Report No. 7, 1 -87.
globensky, y. 1967. Middle and Upper Mississippian conodonts from the Windsor Group of the Atlantic
Provinces of Canada. J. Paleont. 41, 432-448.
166
PALAEONTOLOGY, VOLUME 25
higgins, a. c. 1975. Conodont zonation of the late Visean — early Westphalian strata of the south and central
Pennines of northern England. Bull. geol. Surv. Gt. Britain. No. 53, 1-99.
higgins, a. c. and wagner-gentis, c. h. t. Conodont and Goniatite biostratigraphy and stratigraphic history
of the Lower Carboniferous and early Namurian of the central Cantabrian Mountains, NW Spain. (In press.)
holliday, d. w., neves, R. and Owens, b. 1979. Stratigraphy and palynology of early Dinantian (Carboniferous)
strata in shallow boreholes near Ravenstonedale, Cumbria. Proc. Yorks, geol. Soc. 42 (19), 343-356.
jenkins, t. b. h. 1974. Lower Carboniferous conodont biostratigraphy of New South Wales. Palaeontology,
17, 909-924.
Johnson, G. A. l. and marshall, a. e. 1971. Tournaisian Beds in Ravenstonedale, Westmorland. Proc. Yorks,
geol. Soc. 38, 261-280.
kellaway, G. a. and welch, f. b. a. 1955. The Upper Old Red Sandstone and Lower Carboniferous rocks of
Bristol and the Mendips compared with those of Chepstow and the Forest of Dean. Bull. geol. Surv. Gt.
Britain, 9, 1-21.
Merrill, G. k. and martin, m. d. 1976. Environmental control of conodont distribution in the Boyd and
Mattoon Formations (Pennsylvanian Missourian) Northern Illinois. Geol. Soc. Canada Sp. Paper No. 15,
243-271.
ramsbottom, w. h. c. 1973. Transgressions and regression in the Dinantian: a new synthesis of British
Dinantian stratigraphy. Proc. Yorks, geol. Soc. 39 (28), 567-607.
— 1974. Dinantian. In rayner, d. h. and Hemingway, j. e. (eds) The Geology and Mineral Resources of
Yorkshire: Occ. Publ. Yorks. Geol. Soc. 47-73.
— 1977. Major cycles of transgression and regression (mesothems) in the Nanurian. Ibid. 41 (24), 261-291.
rexroad, c. b. 1957. Conodonts from the Chester Series in the type area of southwestern Illinois. Illinois
geol. Survey Rept. Inv. 199, 1-43.
1958. Conodonts from the Glen Dean Formation (Chester) of the Illinois Basin. Ibid. 209, 1-27.
— 1965. Conodonts from the Keokuk, Warsaw and Salem Formations (Mississippian) of Illinois. Ibid. circ.
No. 388, 1-26.
— and collinson, c., 1963. Conodonts from the St. Louis Formation (Valmeyeran Series) of Illinois,
Indiana and Missouri. Illinois geol. Surv., circ. No. 355, 1-28.
Rhodes, f. h. t., Austin, r. l. and druce, e. c. 1969. British Avonian (Carboniferous) conodont faunas, and
their value in local and intercontinental correlation. Bull. British Museum (Nat. Hist.) Geol. Supplement
No. 5, 1-313.
Sandberg, c. a. 1976. Conodont biofacies of late Devonian Polygnathus styriacus zone in western United
States. Geol. Soc. of Canada, Spec. Paper No. 15, 171-186.
turner, J. s. 1950. Notes on the Carboniferous Limestone of Ravenstonedale, Westmorland. Trans. Leeds
geol. Assoc. 6 (3), 27-37.
varker, w. j. 1967. Conodonts of the genus Apatognathus Branson and Mehl from the Yoredale Series of
the North of England. Palaeontology, 10, 124-141.
— and higgins, a. c. 1979. Conodont evidence for the age of the Pinskey Gill Beds of Ravenstonedale,
north-west England. Proc. Yorks, geol. Soc. 42 (20), 357-369.
vaughan, a. 1905. The palaeontological sequence in the Carboniferous limestone of the Bristol area. Q. Jl
geol. Soc. Lond. 61, 181-307.
voges, a. 1959. Conodonten aus dem Untercarbon I und II (Gattendorfia und Pericyclus-Stufe) des
Sauerlandes. Palaont. Z. 33, 266-314.
von bitter, p. h. 1976. Palaeoecology and distribution of Windsor Group (Visean-Early Namurian) conodonts.
Port Hood Island, Nova Scotia, Canada. Geol. Soc. Canada, Sp. Paper No. 15, 225-241.
a. c. higgins
Department of Geology
University of Sheffield
Mappin Street
Sheffield SI 3JD
W. J. VARKER
Department of Earth Sciences
University of Leeds
Leeds LS2 9JT
Typescript received 1 May 1980
Revised typescript received 4 July 1980
SOMASTEROIDEA, ASTEROIDEA, AND THE
AFFINITIES OF LUIDIA (PLATASTERIAS )
LATIRADIA TA
by DANIEL B. BLAKE
Abstract. Important changes in the taxonomy and phylogenetic interpretation of stellate echinoderms were
proposed during the 1960s by H. B. Fell; certain of this author’s ideas are re-evaluated. Fell argued that
the extant west American sea star Platasterias latiradiata Gray is a surviving member of the otherwise
Palaeozoic Somasteroidea. The extant family Luidiidae was considered primitive among true asteroids and
it was included with the Palaeozoic family Palasteriscidae in the order Platyasterida. The skeletal arrangement
of Platasterias and Luidia was interpreted as having been derived with relatively limited change from a
currently unknown Cambrian crinoid ancestry. It is argued here that Platasterias is not a somasteroid but a
subgenus of Luidia, and the Luidiidae is returned to the large living order Paxillosida. The origin of the
morphology of Luidia, including Platasterias, is related to sea star behaviour and habitat rather than a crinoid
ancestry. The Luidiidae is not considered to be of major importance in delineating asteroid phylogeny.
Several papers published during the 1960s by H. B. Fell (1962a, b; 1963 a-c; 1967) developed a
series of stimulating and intriguing hypotheses on stellate echinoderm phylogeny that significantly
altered viewpoints of the history of these organisms. Certain of these ideas are reconsidered in this
paper.
Prior to Fell’s work, three classes or subclasses were recognized among stellate echinoderms;
the Asteroidea (sea stars or starfish ), Ophiuroidea (brittle stars or serpent stars and basket stars),
and the Somasteroidea, the last a taxon of primitive echinoderms then considered to be restricted
to lower and middle Palaeozoic rocks. Relationships among the groups were unclear. Because of
disparate early development, many biologists (e.g. Hyman 1955) believed the ophiuroids and
asteroids had independent origins, and were only secondarily similar in certain features. These
workers found (and still find) it undesirable to combine ophiuroids and asteroids in formal higher
taxa below the phylum level, e.g. the class Stelleroidea and subphylum Asterozoa sensu Spencer
and Wright 1966.
Other workers, stressing the fossil record, were of the opinion that the two living groups shared
a common origin, and that their ancestors or near ancestors could be recognized among the fossil
somasteroids. Further, Fell (e.g. 1948) considered that the developmental arguments for separating
the asteroids from ophiuroids were unconvincing. These workers tend to believe that the two should
| be combined in a phylogenetically unified higher taxon below the phylum level (e.g. Spencer and
Wright 1966).
In his work on phylogeny. Fell developed a number of interrelated topics: (1) the nature of the
Somasteroidea; (2) the relationships among the extant sea stars Platasterias, Luidia, the remaining
asteroids and the Palaeozoic Somasteroidea; (3) the relationship between the extant ophiuroid
Ophiocanops and the Palaeozoic ophiuroids; and (4) the phylogenetic relationships among
somasteroids, asteroids, and ophiuroids. Fell proposed a hypothesis for the origins of stelleroids
from an inferred crinoid ancestry, suggested a sequence of evolutionary events leading from the
Somasteriodea to the Ophiuroidea and Asteroidea, and transferred a number of living taxa to
groups previously considered to be of exclusively Palaeozoic occurrence. These included the family
Luidiidae from the Paxillosida to the Platyasterida, both within the Asteroidea; the genus
IPalaeontology, Vol. 25, Part 1, 1982, pp. 167-191, pis. 20-22.|
168
PALAEONTOLOGY, VOLUME 25
Platasterias from the Asteroidea (Paxillosida) to the Somasteroidea; and the ophiuroid Ophiocanops
fugiens to the Oegophiurida from the Phrynophiurida. Fell’s re-evaluations were enthusiastically
received by some workers, but a number of his ideas were challenged, including the arguments for
the derivation of asterozoans from crinoids, by Philip (1965), answered by Fell (1965); the affinities
of Ophiocanops, by Hotchkiss (1977); and the affinities of Platasterias by several workers (e.g.
Madsen 1966; Blake 1967, 1972, 1973; Pearse 1969; Algor 1971).
A survey of textbooks and papers published during the later 1960s and 1970s shows a continuing
uncertainty as to how Platasterias latiradiata should be treated, yet the question of affinities of the
species is an important one. Not only is a living fossil intriguing simply as a survivor, but its
existence raises questions on the nature of evolutionary processes. Further, once ranked as a ‘living
fossil’, a species will become the basis for reconstruction of the biology of its presumed close fossil
relatives, a fact illustrated by Fell’s (19626, p. 2) explanation of purpose for one of his papers:
‘This contribution is limited to brief discussion of the major features of somasteroid anatomy, as
illustrated by Platasterias .’ Platasterias as a somasteroid clearly will strongly influence interpreta-
tions of stellate echinoderm biology and history.
The concern of this paper is primarily with the second of the four topics cited above, the nature
of Platasterias and the Luidiidae. Other problems in the nature of relationships among stellate
echinoderm groups are in need of restudy, but these ideas are beyond the purposes of this paper.
I will argue that P. latiradiata should be considered a subgenus of Luidia, itself a genus
unequivocally included in the Asteroidea (rather than the Somasteroidea) by Fell (1963a). Assigning
Platasterias as a subgenus of Luidia does not in itself challenge Fell’s ideas on somasteroid/asteroid
phylogeny; Luidia (including Platasterias) might still be the primitive extant asteroid with many
features, as outlined by Fell, derived with little modification from the somasteroids. In re-evaluating
the taxonomic and phylogenetic position of the monogeneric Luidiidae, I will next argue that there
is no clear connection between Palaeozoic somasteroids and the luidiids (including Platasterias),
and that the Luidiidae should be transferred from the order Platyasterida to the order Paxillosida.
A functional, rather than phylogenetic, hypothesis for the origin of luidiid morphology is proposed.
SOMASTEROIDEA
The concept of the Somasteroidea originated with Spencer (1951) and evolved with the work of
Fell (1962a, b; 1963 a-c; 1967), Spencer and Wright (1966), and McKnight (1975). In spite of this
effort, the group is difficult to characterize and therefore comparisons between somasteroids and
other organisms are difficult as well. Relevant fossils are in need of restudy and reillustration. I
have based the following diagnosis on the literature (and not original study of the fossils), and I
believe it represents the fossil somasteroids as pictured by those workers who include Platasterias
in the taxon. In keeping with the arguments presented here, however, I have removed characters
derived from Platasterias only. Although such a diagnosis logically follows arguments for transfer
of Platasterias, a concept of the Somasteroidea is necessary for the comparisons that follow.
Subclass Somasteroidea Spencer, 1951
Asterozoans in which the axial skeleton consists of ambulacrals in a double series, usually paired but apparently
alternating in some species. Each ambulacral gives rise to a transverse series of ossicles; in apparently primitive
species, these ossicles are similar and rod-like elements termed virgals, but in more advanced species the
virgals are differentiated into adambulacrals, marginals, and related ossicles. A permanent ambulacral furrow
or groove is lacking so that the long axis of the ambulacral is approximately linear and horizontal, and the
ossicles essentially lie in the plane of the remaining ossicles of the oral surface. The adambulacral (or first
virgal) (different spellings of ‘virgal’ and its plural form have been used; I have followed Spencer and Wright
1966) generally abuts the lateral (abradial) margin of the ambulacral, or overlap is slight. In some species
the ambulacrals probably could be raised to form temporary ambulacral furrows. A large or small radial
channel for the radial water canal is present along the oral margin of the line of juncture of pairs of ambulacrals,
or it is more or less enclosed by the ambulacrals. The tube-feet are seated in broad basins which in some
species communicate to the body cavity. Jaw ossicles are differentiated, but odontophores are lacking. Open
BLAKE: SEA STAR PHYLOGENY
169
text-fig. 1 . Morphology of the Somasteroidea, modified slightly
after Spencer (1951) and Fell (1963a). a, abactinal arrangement
of Sturzaster marstoni (Salter), morphology considered by Spencer
to be close to that of Chinianaster; Spencer (1951, p. 95). b. ambu-
lacrals of Archegonaster pentagonus; Spencer (1951, p. 104).
c, metapinnules with cover plates, Chinianaster levyi\ Fell (1963a,
p. 394). d, interpretation of arm of Ampullaster ubaghasi, view of
oral surface; Fell (1963a, p. 394). e, reconstruction of a part of
the arm of Chinianaster, water vascular system darkened; Fell
1963a, p. 402). f, interpretation of arm, near tip, of Villebrunaster
thoralr. Fell (1963a, p. 394). Key: 1 , ambulacrals; 2, adambulacrals;
3, virgals; 4, marginals. Illustrations courtesy of H. Barraclough
Fell and the Royal Society.
buccal slits might be present at the mouth frame in at least some species. Aboral ossicles are paxilliform,
with delicate tetradiate bases. Encrusting ossicles or spinelets are present at least on some ossicles of the
ventral surface. Over all, the skeleton beyond the ambulacrals is quite delicate (text-fig. 1).
Discussion. Most of these characters are also expressed among species assigned to the Asteroidea,
but critical to the concept of the Somasteroidea is an ambulacral/adambulacral arrangement in
which an ambulacral furrow or groove is lacking (text-fig. 2). Spencer (1951, p. 88) says in reference
to somasteroids that ‘. . . these first stages show no sign of an ambulacral furrow’, and Fell (1963a,
p. 389) used the development of a furrow or groove as the basic character separating somasteroids
from asteroids. The ambulacral groove or furrow is different from the ambulacral channel for the
radial water canal; for further discussion, see under ambulacral column arrangement.
170
PALAEONTOLOGY, VOLUME 25
In somasteroids the abradial margin of the ambulacral either abuts the adradial side of the
adambulacral (first virgal), or the ambulacral can overlap the adambulacral to a very limited extent
(text-fig. 2). In his discussion of somasteroids, Fell (1963a, p. 393) says ‘Each metapinnule arises
from the abradial margin of one of the paired ambulacral (or brachial) elements . . .’. The long
axis of the ambulacral is approximately horizontal and the basin for the tube-foot lies close to the
plane of the oral surface.
In true asteroids the ambulacral rests on the aboral surface of the adambulacral, rather than
laterally adjacent to it (text-fig. 2). In addition, the axis of the adradial end of the ambulacral is
oblique to that of the general ambulacral/adambulacral articulation surfaces. These arrangements
produce a permanent furrow with the basin for the tube-foot elevated.
McKnight (1975) argued that the lack of an odontophore is an important character unifying
somasteroids. The odontophore is a small, unpaired typically T-shaped ossicle present between
members of a jaw ossicle pair. It appears to brace and thus strengthen the jaw apparatus, and its
phylogenetic development might reflect the changing food habits suggested by Spencer (1951),
away from the small particle feeding inferred for somasteroids to the large particle feeding present
in many living asteroids.
A
text-fig. 2. Stylized diagrams oriented transverse to arm axes,
showing arrangement of 1 , ambulacrals, and 2, adambulacrals
in a, somasteroids ( Chinianaster arrangement to left, Ampul-
laster to right), and b, asteroids. The overlap of the ambulacral
on the adambulacral in Ampullaster is very small; compare to
text-fig. Id. 3, radial water channel. Based on work of Spencer
(1951) and Fell (1963a), especially the latter’s fig. 7, p. 395.
COMPARISONS BETWEEN PLATASTERIAS, FOSSIL SOMASTEROIDS
AND TRUE ASTEROIDS
Platasterias is compared to fossil somasteroids and living asteroids, especially Luidia as traditionally
recognized, in the following somewhat overlapping sequence: (1) over-all body form; (2) nature of
the skeleton; (3) ambulacral column arrangement; (4) growth gradients; (5) other morphologic
features; (6) feeding. Fell’s view of Platasterias as a somasteroid was derived in large part from
his evaluation of ambulacral column arrangement and growth gradients. A discussion of
the taxonomic positions of Platasterias and Luidia sensu Fell therefore must wait until section 4.
Each section begins with a summary of the ideas of Spencer (1951), Fell (1962a et seq.), and Spencer
and Wright (1966) and continues with my comments. I have tried to include all major arguments
on Platasterias, but the discussion is not comprehensive on other topics, such as the morphology
BLAKE: SEA STAR PHYLOGENY
table 1. Changes in the distribution of important characters in the inferred somasteroid/asteroid phylogeny,
as envisaged by H. B. Fell. The diagram was prepared by the writer, based on his reading of Fell (1963a,
pp. 389 ff.; 19636) and it is intended as an aid to understanding Fell’s ideas of the major events in his inferred
phylogeny (represented by the sequence 1 through 6) rather than as an exact character distribution summary.
Dotted lines mean a character is present in only some members of a division, and minor exceptions occur.
For example, although the family is not entered separately, the Porcellanasteridae lack an anus and suctorial
tube-feet, whereas other sea stars, such as certain Astropecten and Asterias species, have subpetaloid arms.
Notes: (1) virgals similar in Division 1, dissimilar in Division 2; (2) development of actinals becomes stronger
through the sequence; (3) no entry in Fell (1963a) for this division. The terminal is the unpaired ossicle at
the distal tip of the arm.
Among somasteroids, the monogeneric Chinianasteridae was considered primitive, based largely on the
presence of undifferentiated virgals and robust ambulacrals, and the absence of communication pores between
ambulacrals for ampullae. Next in the sequence, but not separated from the Chinianasteridae in Fell 1963a,
came the Villebrunasteridae, including Villebrunaster and Ampullaster. Here, virgals are differentiated and
pores, inferred for internal ampullae, were present. Platasterias (Platasteriidae) was considered to have common
features with both the Chinianasteridae and the Villebrunasteridae but appeared to be about at the
villebrunasterid grade based on the inferred differentiation of the virgals, and the presence of internal ampullae
(19636, p. 144). The monogeneric Archegonasteridae, not included in the table, was considered to be an
advanced somasteroid, specialized in body shape and lack of interradial slits, metapinnule reduction,
and development of the robust marginals. True asteroids originated with development of a permanently erect
ambulacral furrow.
<
w Group 1
co 1 . Chinianasteridae
<
2. Platasteriidae
Group 2
< 3. Platanasteridae
L±J
Q
O 4. Luidiidae
cr
tu
H Group 3
*■* 5. Astropectinidae
6. Most other
asteroids
of the fossils. Fell’s (1963a, b) major ideas on somasteroid/asteroid phylogeny are summarized in
Table 1.
McKnight (1975) modified the concept of the Somasteroidea and transferred five Palaeozoic
genera to this taxon. His modification is included here, but I did not attempt to re-evaluate
somasteroids in light of the transferred taxa because properties of these genera were not incorporated
in, and do not appear to bear directly on. Fell’s ideas on the position of Platasterias and Luidia.
1. Over-all body form
Spencer (1951, p. 91) described somasteroids as having a large central body in which the arms
are parts of the oral surface, i.e. . . the arms are just beginning to be differentiated’. Fell (1963a)
described somasteroids as extremely flattened but, more important, as asterozoans with petaloid
arms. Fell (1963a, fig. 5) presented a sequence of diagrams illustrating the inferred transformation
of arm shape beginning with the petaloid arm of a monoserial crinoid, and ending with the triangular
172
PALAEONTOLOGY, VOLUME 25
arm of the extant astropectinid Plutonaster. The petaloid arm of Platasterias was placed by Fell
between the petaloid arms of a chiniansterid somasteroid and a somewhat weakly petaloid
platanasterid asteroid. Petaloid arms thus were considered to be subdued in the most primitive of
Asteroidea, the fossil Platyasterida, and they are absent from later groups.
Spencer (1951) did not consider the somasteroids to be flattened, rather he saw them as flexible,
noting (p. 93): ‘At one time the body is strongly compressed, at another the body of the same
species is well rounded.’
Discussion. P. latiradiata is a relatively flat species, but such a body shape is not restricted to
somasteroids, e.g. see such extremely flattened living asterinids as Anseropoda and Stegnaster.
Although the arms of Platasterias are strongly petaloid and the disc deeply notched interbrachially
(PI. 20, fig. 1), comparable shapes are known elsewhere, albeit more weakly developed. Examples
include Astropecten regalis (PI. 20, fig. 2) (Paxillosida) and Asterias forbesi (Forcipulatida) (PI. 20,
fig. 8). Inferomarginals in Astropecten regalis, like those in Platasterias, are transversely elongate
(Blake 1973, pi. 14) and the body is flattened. These similarities among taxonomically widely
separated species suggest convergence as an alternative hypothesis for origin of body shape, perhaps
under conditions such as those suggested by Madsen (1966): ‘I assume the petaloid arms in
Platasterias (brought about by the transverse elongation of the adambulcralia and inferomargin-
alia) to be secondary adjustment to a life on a shifting sandy bottom (and perhaps a primarily
ciliary method of feeding).’ Fell (1962a, p. 634) noted that the extreme narrowness of the ambulacral
furrow in Platasterias is suggestive of the fossil somasteroids, although he observed that this could
be secondary. Narrow furrows occur in other modern taxa, for example, the Ophidiasteridae
(Valvatida) and the Echinasteridae (Spinulosida).
2. Nature of the skeleton
Spencer (1951) considered Chinianaster and Villebrunaster to be the primitive somasteroids. In these
genera he recognized three types of ventral ossicles: the ambulacrals arranged in a double row, the
mouth angle ossicles, and the rod-like intermarginals or virgals arranged in transverse series
extending from each ambulacral ossicle. Fell used the term ‘metapinnules’ for the rows of virgals,
after the inferred homologous pinnules of crinoids. In more advanced somasteroids the morpho-
logically simple virgals became differentiated to form marginals, adambulacrals, and other ossicles.
EXPLANATION OF PLATE 20
Figs. 1, 5, 6, 10, 12. L. ( Platasterias ) latiradiata (Gray). 1, over-all aboral view showing petaloid
arms, alignment of abactinals, and similarity in marking to that of L. clathrata, see PI. 22, x 1. 5, lateral
view of inferomarginal, furrow left, row of processes are articular facets linking successive inferomarginals,
x 9. 6, adambulacral, proximal view, furrow right; one muscle groove (upper, light arrow) and one
articular facet (lower, dark arrow) marked; these and other structures correspond in positions with similar
structures in L. clathrata, fig. 4 (see Blake 1972, 1973; Heddle 1967 for further discussion). 10, aboral view
of cleared arm showing enlarged superomarginal row (arrow) arising at terminal, as in L. clathrata, see
fig. 9, x 3. 12, aboral view of uncleared arm showing abactinal ossicle and granule development, compare
to L. clathrata, fig. 11, x 3.
Fig. 2. Astropecten regalis Gray. Oral view showing petaloid arms and well defined marginal spines (arrow), x 1 .
Figs. 3, 4, 9, 11. Luidia clathrata (Say). 3, lateral view of inferomarginal, furrow left, x9. 4, adambulacral,
inclined proximal view, furrow right, see discussion for fig. 6, above, x 9. 9, inclined aboral view of cleared
arm, see discussion for fig. 10, above, x 3. 11, aboral view of uncleared arm, see discussion for fig. 12,
above, x 3.
Fig. 7. Luidia neozelanica Mortensen. Lateral view of inferomarginal, compare with figs. 3, 6; note outline
and lack of articular facet row in this species; see text for further discussion, x 9.
Fig. 8. Asterias sp. Aboral view showing petaloid arms in which an interbrachial notch (arrow) nearly reaches
the madreporite, x
PLATE 20
BLAKE, living sea stars
174
PALAEONTOLOGY, VOLUME 25
The change in terminology marks this differentiation. Some virgals were partially (Archegon-
asteridae) or completely lost (Archophiactinidae) (Spencer 1951).
Spencer (1951) did not describe marginals in Villebrunaster from his relatively incomplete material,
but Fell ( 1 963 6, p. 1 44) pointed out that these ossicles can be recognized in more complete specimens.
Fell (19626, p. 66; 1963a, p. 396) further noted that there are no adambulacrals, superambulacrals,
or inferomarginals in Chinianaster . The metapinnules of Chinianaster terminate in a free radiole,
as in Villebrunaster (Fell 19626, p. 16). These patterns are important in interpreting Chinianaster as
the primitive somasteroid, and Villebrunaster as a second step in somasteroid evolution.
The ambulacrals of somasteroids (text-fig. 1) were described as stout and block-like, or a lateral
wing is developed; they form a sheath or channel for the radial water vessel. The virgals are rod-like
and form the walls of channels separating successive metapinnules. The virgals of Villebrunaster
were described (Fell 19626, p. 16) as having a flattened base and a flanged keel (text-fig. lc, e)
and thus were seen as similar to those of Platasterias , but Fell (1963a, p. 422) cautioned that these
ossicles proved morphologically plastic in later evolution.
The abactinal skeleton of fossil somasteroids (text-fig. 1a) consists of an open meshwork of
paxilliform ossicles, each consisting of a slender, vertically oriented axial stalk bearing a number
of elongate basal flanges. Fell (19626, p. 14) noted that Chinianaster abactinals are quite similar
to those of Platasterias and Luidia but less similar to those of the astropectinids in that in the
latter the base is disc-like and lacks basal projections, and the stalk is relatively stout.
In living sea stars, an unpaired ossicle termed the odontophore is found between members of
an oral ossicle pair. In certain Palaeozoic asteroids, Spencer (1919 in 1914-1940) concluded that
this ossicle was derived from an occluded inferomarginal, but because of the distribution of ossicles
about the jaw region in Chinianaster and Villebrunaster, Fell (1963a, p. 401, fig. 8) suggested that
in Platasterias an analogous T-shaped ossicle was derived from an occluded, non-metapinnular
tegminal ossicle. McKnight (1975), however, argued that an odontophore (and, presumably, the
analogous T-shaped ossicle as well) is lacking from adult somasteroids. Stressing the significance of
the absence of this ossicle McKnight (1975) transferred the Flelianthasteridae and three genera of
the Taeniactinidae to the Somasteroidea from the Spinulosida sensu Spencer and Wright (1966).
The somasteroid skeleton beyond the ambulacral column was described by Spencer (1951, p. 93)
as ‘slightly built’.
Discussion. The primary similarity between Platasterias and fossil somasteroid ossicle form seems
to be in transverse elongation, and, perhaps, the development of keel- and flange-shaped ossicles,
as in Chinianaster and, to some extent, Platasterias.
Virgals differentiated as marginals are absent from Chinianaster, the inferred primitive somasteroid
(Fell, 19636, p. 144). Those of Villebrunaster (text-fig. If), representing the inferred next phylogenetic
step (Fell, 19636), are rather simple cylinders elongate parallel to the arm axis. They are not
transversely elongate as is the case in Platasterias (PI. 20, fig. 5; PI. 21, figs. 2, 5) and as presumably
would be true in a primitive somasteroid under strong influence of transverse gradients (see below).
Illustrated ambulacrals (Fell, 1963a, figs. 5, 6; text-fig. 1, herein) of fossil somasteroids are relatively
robust with a broadened adradial head and an abradial wing deflected distally, whereas those of
Platasterias (PI. 21, fig. 2) are wide and short, lacking the broadened head. In over-all outline, the
ambulacrals of Platasterias thus seem to be more primitive even than those of Chinianaster (although
ambulacrals were not considered to be derived from virgals). Chinianaster lacks differentiated
adambulacrals, whereas those of Villebrunaster (Fell 1963a, fig. 6; text-fig. 1, herein) apparently
are simple ossicles elongate parallel to the arm axis, rather than transversely elongate as is the case
in Platasterias (PI. 20, fig. 5; PI. 21, fig. 5). There are no known fossils of adambulacral shape
similar to those of Platasterias.
In contrast, ossicle morphology of Platasterias is very close to that of Luidia clathrata, differing
significantly only in degree of transverse elongation (PI. 21, figs. 1, 2; text-fig. 3). The abactinals,
marginals (PI. 20, figs. 3, 5) and adambulacrals (PI. 20, figs. 4, 6) are particularly similar. This is
true not only of the form itself, but also of the complex articulation structure arrangement of the
BLAKE: SEA STAR PHYLOGENY
.75
ambulacral column, involving both ambulacrals and adambulacrals (Blake 1972, 1973). The
abactinal surface of the adradial end of the ambulacral is truncated in Platasterias, but the basic
form of these ossicles is the same as in Luidia. Fell (19626) stressed the presence of a Y-shaped
groove on the actinal surface of the ambulacral of Platasterias, pointing out that the structure
also occurs on fossil somasteroids, and he suggested that the associated muscle is used to temporarily
elevate the ambulacral furrow. This muscle depression, however, occurs on all modern asteroids.
Fell (19626, p. 7) also noted in Platasterias that the tube-feet emerge from a broad, basin-like
depression. The transverse section of the middle of the long axis of the ambulacrals of most asteroid
species (except, for example, the compressed ossicles of certain asteriids and asterinids and the
delicate ossicles of the pterasterids) is cylindrical, hence a broad, basin-like depression is typical
of most species.
Algor (1971) pointed out that Platasterias has the ambulacral system characteristic of modern
asteroids, and he concluded that Platasterias is in no way primitive. Algor (1971) differentiated
between ancient and modern asteroids on the basis of articulation structures across the furrow,
arguing these muscles and facets were weakly developed in Palaeozoic taxa compared to the sturdy
patterns seen in living species. Algor’s sample of the Palaeozoic species was too small, for sturdy
articulation structures comparable to that of modern species do occur, e.g. in Promopalaester
magnificus (Miller) and P. dyeri Meek. Important, however, is that Platasterias is constructed in
the same pattern as Luidia, rather than in the weakly articulated manner described for the
somasteroids.
Superambulacrals are present in both Platasterias and Luidia but they have not been reported
from fossil somasteroids. Inferomarginals of both Platasterias latiradiata and L. clathrata are
step-shaped with overlapping, multifaceted contact points (PI. 20, figs. 3, 5) (Blake 1973). The
superomarginals are relatively large in Platasterias but their basic form is the same as those of L.
clathrata. Arrangement and basic morphology of mouth frame ossicles are essentially the same
between P. latiradiata and L. clathrata, although the odontophore is absent from fossil somasteroids
(McKnight 1975). Distal arm ossicles, which are not transversely elongate, are essentially identical
between P. latiradiata and L. clathrata, and much closer to each other than they are to those of
species belonging to any other genus.
Fell (19626, p. 15) suggested that the abactinals of P. latiradiata are similar to those of
L. neozelanica, however, paxilliform abactinals occur in all of the living sea star orders except the
Forcipulatida. A survey of the diagrams of Fisher (1911) shows that basal projections are present
text-fig. 3. Cross-sections of arms of A, Luidia ( Platasterias ) latiradiata-, and b, Luidia ( Pe talas ter )
clathrata. In life, arms are flexible and ossicle orientations vary with behaviour; inferomarginals can be
more steeply inclined to the horizontal than shown here or in PI. 21, fig. 1. Key: 1, abactinals; 2,
superomarginals; 3, inferomarginals; 4, superambulacrals; 5, actinals; 6, ambulacrals; 7, adambulacrals.
176
PALAEONTOLOGY, VOLUME 25
in abactinals of taxonomically diverse species and that the slender abactinals illustrated for several
of the spinulosidan families (e.g. Solasteridae, Korethrasteridae, Pterasteridae) are closer to
Spencer’s somasteroid illustration (1951, fig. 5) than are those of Platasterias or Luidia. The
abactinals of P. latiradiata and L. clathrata are very similar to each other, however.
Similarities extend to exterior morphology as well (PI. 20, figs. 1, 11, 12; PI. 22), as was
demonstrated by Madsen (1966) for P. latiradiata and L. marginata.
I have argued on the basis of skeletal morphology that P. latiradiata clearly is very close to L. clathrata , but
shows only relatively minor similarities to fossil somasteroids. Fell dissected specimens of Luidia; if the
similarities are this strong, why was he not struck by them as well? I believe the answer lies in the species of
Luidia available to Fell for dissection. For forty-three species of Luidia, Doderlein (1920) recognized ten
subgenera and four supra-subgeneric taxa he termed ‘groups’, each group named for a representative species.
Doderlein (1920, p. 223) presented a phylogenetic interpretation of Luidia in which he suggested that the
Clathrata Group was primitive and gave rise to both the Quinaria and Alternata Groups; the Ciliaris Group
was in turn derived from the Quinaria Group.
In external appearance, species of the four groups are superficially similar to one another, whereas the
ossicle morphology of the members of the Ciliaris Group is quite distinct from that of the other three (Blake
1973); I suspect that if the external morphology of the Ciliaris Group were as distinct as the skeletal morphology,
this group would be recognized as a separate genus.
In his discussions Fell placed little emphasis on the species available to him for study, but in an illustration
of abactinals (19626, p. 15) and an arm tip in cross-section (1963a, p. 386) he does note that his drawings
were based on L. neozelanica. This species had not been described at the time of Doderlein (1920), but it has
since been referred to the Ciliaris Group (Clark 1953; Fell 1963a, p. 433). Further, Fell’s (1963a, p. 395)
drawing of the cross-section of a Luidia arm shows a crescentic inferomarginal outline typical of the Ciliaris
Group, rather than the step-shape typical of the marginals of the other three groups (PI. 20, figs. 3, 5, 7).
Madsen (1966) assigned Platasterias to Petalaster, a subgenus of the Clathrata Group; although I recognized
Platasterias at the subgeneric level because of its distinctive shape, I agree with Madsen’s inclusion of the
species in the Clathrata Group.
Fell thus appears to have based his comparisons between Luidia and Platasterias on a species inferred to
be well removed in phylogenetic position (Doderlein 1920) and distinct in ossicle morphology (Blake 1973)
from those Luidia species which are suggested to be closest to P. latiradiata (Madsen 1966).
3. Ambulacral column arrangement
Although not in his diagnosis, Spencer (1951, p. 88) noted that ‘. . . these first stages . . .’ (in
reference to somasteroids in general) \ . . show no sign of an ambulacral groove . . .’. Spencer did
describe a \ . . shallow channel . . .’ (p. 102) along the midradius of Archegonaster, and ‘. . . a
deep channel . . .’ (p. 98) in Chinianaster. Spencer (1951) cited his 1914 paper for illustration of
the channel. In a drawing of an asteroid arm in cross-section, Spencer (1914 in 1914-1940) showed
that the ambulacral channel refers to the axial notch for the radial water canal. The channel is
thus a structure distinct from the ambulacral furrow or groove. Although the ambulacral channel
is deep in Chinianaster, the basins for the tube-feet are \ . . placed almost in the oral plane’. If an
ambulacral furrow is present, as in asteroids, the ambulacrals are arched and the basins for the
tube-feet are raised above the oral surface of the animal.
In Fell’s discussions, whether or not the ambulacral furrow is at most temporarily erect
(somasteroids) or permanently erect (asteroids) is essential to the notion of somasteroids and
Platasterias as a somasteroid. The change from temporarily to permanently erect was selected as the
point at which the pre-asteroid phase gave way to the asteroid phase (1963a, p. 391).
Fell described a well-developed ‘lateral wing’ on the ambulacral of Platasterias. This ‘wing’
provides attachment for the musculature extending to the first virgal (i.e. the adambulacral) that
permits temporary erection of the ambulacral ossicle to form an inverted V-shaped groove considered
homologus with the asteroid furrow. The ossicle arrangement in Platasterias was considered to be
asteroid-like and leading toward the asteroid grade of organization (19626, p. 10). Although
recognizing that in P. latiradiata, the wing extends over the adambulacral in the manner found in
asteroids, Fell (1963a, p. 393) considered ‘. . . the major [i.e. transverse] axis of the ambulacral
BLAKE: SEA STAR PHYLOGENY
177
lies almost horizontally, in the same axis as the metapinnule which it bears, as in the Villebrun-
asteridae . . In asteroids, the adambulacrals are below the abradial end of the ambulacrals and
the basins for the tube-feet permanently raised above the substrate.
The channel for the water canal in Platasterias was considered partially enclosed by the ambulacral
ossicles, much as in the Palaeozoic fossils.
Discussion. Interpretation of the ambulacral ossicles is important to the idea of a temporarily
erectable furrow (PI. 21, figs. 1, 2; text-fig. 3). In Platasterias , the impression of a linear, nearly
horizontal ambulacral ossicle is conveyed by the broad ‘lateral wing’ of the ambulacral, but
important to the nature of the furrow is the angle between the axis of the ambulacral/adambulacral
articulation and the axis of the adradial portion of the ossicle. The ambulacral/adambulacral
articulation axis is approximately horizontal in the relaxed living sea star (although capable of
broad adjustment; see Heddle 1967) and therefore the angle the adradial end of the ossicle makes
to the articulation axis reflects the relaxed, permanent furrow development. From a number of
medial arm ossicles of mature specimens that I measured on a binocular microscope stage, this
angle is approximately 20° in L. clathrata, 30° in P. latiradiata. The size of the angle depends on
the precise points selected for measurement, but the critical idea is that in Platasterias, as in Luidia,
these two axes are not parallel; a permanent furrow is present in both. Fell’s sketches (1963a,
p. 395), although presumably not intended to be precise, do accurately reflect approximate
relationships, and the presence of a furrow in both.
In a figure description. Fell (19626, p. 13) suggested that the outer, oral surfaces of the
adambulacrals, as seen in Platasterias, are erected into the furrow at the asteroid grade. Although
these faces are pulled toward one another as any sea star closes the furrow, the surfaces are not
erect in the furrow in asteroids and remain directed toward the substrate in the relaxed animal.
As is true of other surface ossicles, the adambulacrals bear spines on their surficial faces. Both
Platasterias and other sea stars have approximately vertically oriented adradial side faces on the
adambulacrals.
Equally important to ambulacral ossicle orientation is the nature of the ambulacral/adambulacral
articulation structures, for, as noted above, the same facets and muscle depressions, arranged in
approximately the same proportions, can be recognized in both Platasterias and Luidia (PI. 20,
figs. 4, 6) (Blake 1972, 1973), as well as in other asteroids. Both sea stars must be capable of
approximately the same movements; Heddle (1967) described how this musculature and articulation
can be used in locomotion and digging, and in a different genus I (Blake 1981) have argued that
these structures can be used in righting, interpretations that seem appropriate for Platasterias as
well as Luidia. The broad lateral wing, as suggested by Madsen (1966) for the inferomarginals and
adambulacrals, is a part of the morphology of Platasterias adapted to its habitat of a shifting,
sandy bottom.
As noted by Fell (19626, p. 7), the radial water canal in Platasterias occupies a channel along
the arm axis (PI. 21, fig. 1) much as in certain somasteroids. A similar channel, however, occurs
in L. clathrata (PI. 21, fig. 2) as well as in other asteroids (e.g. Asterias), hence the feature is not
of taxonomic value among living sea stars.
4. Growth gradients
Spencer (1951, p. 91) described the interambulacrals, or virgals, as arranged in linear series at an
angle to the ambulacral row with a single series arising at the abradial edge of each ambulacral.
The development of virgals provided the basis for the families recognized by Spencer: in the
Chinianasteridae, ossicles occupy the entire oral surface apart from the ambulacrals and mouth; in
the Archegonasteridae, virgals occur only distally on the arms, and virgals are lacking from the
Archophiactinidae.
Fell (1963a) developed his ideas of growth gradients in asterozoans about this transverse alignment
of oral surface ossicles. Growth gradients were considered to be parallel or weakly convergent
lines along which morphologic structures were aligned. Growth gradients were envisioned as
178
PALAEONTOLOGY, VOLUME 25
arranged in two series, one set parallel to the arm radius and a second, lateral set subperpendicular
to the first. The ambulacral ossicles, nerve ganglia, and radial water vessel were aligned along the
main longitudinal gradient, the adambulacrals along the first lateral gradient, and so on.
Each row of virgals was termed a metapinnule, and each represents a transverse gradient. The
influence of either longitudinal or transverse gradients could be stronger; where transverse gradients
dominated, metapinnule ossicles were aligned, whereas these ossicles were offset transversely but
aligned longitudinally as the longitudinal gradients became dominant. The differentation of the
typical asteroid oral surface ossicles from the morphologically similar virgals was associated with
the inferred declining influence of transverse gradients: the adambulacral from virgal 1, the
superambulacral from virgal 2, the marginal from virgal 3, and the marginal radiole, from virgal 4
(Fell, 1963a, pp. 392, 405).
Transverse gradients dominate in primitive somasteroids (e.g. Chinianaster) but gradually lose
their influence; longitudinal gradients dominated in Astropecten and phylogenetically later asteroids.
The transverse gradients were ultimately derived from the inferred crinoid ancestor, hence the
metapinnule from the crinoid pinnule.
Members of ambulacral pairs apparently alternate across the arm axis in certain fossil
somasteroids (Spencer 1951, p. 102) much as the brachials alternate in biserial crinoids. This
similarity of arrangement contributed to Fell’s view of a close relationship between crinoids and
stelleroids (Fell 1963a, p. 415). In Platasterias the transverse gradients were considered clearly
recognizable (PI. 21, fig. 5), although the ossicles are equally clearly equated with their inferred
homologues in asteroids.
Ossicle alignment was the sole criterion provided for recognition of gradient type. Most
post-Palaeozoic sea stars have few to many ossicles on the oral surface between the adambulacral
and inferomarginal series: these are the so-called actinal intermediates of Fell’s terminology, or
more simply, actinals (text-fig. 3). Actinals typically form a more or less tightly packed mosaic,
they are similar to one another in shape, and although a broad size range can be present, size
changes are gradational across the surface. The ossicles thus are perceived as forming parallel rows
of a single orientation, or, frequently, intersecting rows, one of which is oriented more or less
parallel to the arm axis, the other radiating obliquely from the furrow toward the animal margin.
Dominant orientation of rows can be consistent within taxa; Hotchkiss and Clark (1976) used the
arrangement of these ossicles to separate the Asteropseidae from the Poraniidae.
EXPLANATION OF PLATE 21
Figs. 1, 4, 6. Luidia clathrata (Say). 1, transverse view of arm, ossicles identified in text-fig. 2; see note with
text-fig. 2, x 3. 4, inclined aboral view of arm showing alignment of lateral abactinals with superomarginal
row (arrow); alignment is lost in midarm area; some spinelets are still in place, x 3. 6, oral surface of arm
showing alignment of inferomarginals (arrows), actinals, and adambulacrals. Tube-feet are visible along arm
axis, x 3.
Figs. 2, 3, 5. L. ( Platasterias ) latiradiata (Gray). 2, transverse view of arm, ossicles identified in text-fig. 2,
x 4. 3, aboral view of arm showing alignment of lateral abactinals with superomarginal (arrow) row;
alignment is lost toward midarm region, x 3. 5, oral surface of arm showing alignment of ossicles,
enlarged marginal radioles; arrow indicates inferomarginals, x 3.
Fig. 7. Asterias sp. Aboral view of interior of oral surface of arm showing alignment of ossicles in radial
series, see discussion in text; arrows indicate ambulacrals, x 3.
Fig. 8. Archaster typicus Muller and Troschel. Oral surface of arm showing well-defined fascioles (arrow)
between inferomarginals. Pits in some adambulacrals are for pedicellariae. Note similarity of lateral spines
with those of L. (P.) latiradiata and Astropecten regalis (PI. 20, fig. 2), x 3.
Fig. 9. Dermasterias imbricata (Grube). Oral surface of arm showing ambulacral spines (arrows) crossing over
furrow axis, in function forming furrow cover plates, x 3.
PLATE 21
mi
7
BLAKE, living sea stars
180
PALAEONTOLOGY, VOLUME 25
Certain authors have interpreted the oblique arrangement of actinals in true asteroids as reflecting
transverse growth gradients sensu Fell, but Fell himself equivocated on the origins and significance
of these ossicles. In reference to the astropectinid and later stages of evolution. Fell (1963a, p. 389)
said ‘Actinal intermediate plates are usually present, and are arranged in longitudinal rows; they
are sometimes also arranged in oblique series, but these latter are unrelated to the transverse
gradients which produce the ambulacrals and adambulacrals.’ The inferred phylogenetic history
of these ossicles was described as well (Fell 1963a, p. 391): ‘Actinal intermediate plates are lacking
from division 1 (i.e. Chinianasteridae, see Table 1, herein), appear as minute and irregular rudiments
in division 2 (i.e. Platasteriidae), persist as somewhat larger and more irregular elements in the
luidiids, become progressively more conspicuous in various genera of the Astropectinidae, and
extremely conspicuous in post-astropectinid groups.’ In suggesting that actinals were absent at the
chinianasterid grade but appeared at the platasteriid grade, Fell implied that these ossicles are
derived but appeared very early. In addition. Fell (19636, p. 144) stated that the virgals in Platasterias
had stabilized at four elements (identified above) not including actinals.
Fell (1963a, p. 387), however, also stated, ‘Ossicles are differentiated along the transverse gradients
in the following sequence, from within outwards: adambulacral, superambulacral (if present),
actinal intermediate plates, marginals.’ Elsewhere, Fell (1962a, p. 635) referred to actinals as
‘accessory virgalia’; in the figure description (Fell 19626, fig. 4a), the actinal was labeled ‘intercalary
virgalium (actinal intermediate)’ and again (19636, abstract) the metapinnule was referred to as
differentiated into the actinal intermediate, as well as the other ossicle types. Although he apparently
was uncertain as to how to treat these ossicles in Platasterias and Luidia, Fell’s strongest statements
seem to be that the actinals were secondary, and not derived from virgals.
In order to understand Fell’s interpretation of Platasterias and the nature of the Somasteroidea, it is necessary
to consider the several steps Fell (1963a) used to select the groups of extant sea stars he inferred to be
primitive, and the phylogenetic sequence selected to connect fossil somasteroids to living species. Among
living sea stars, Fell limited his search for the primitive group to the Luidiidae, Astropectinidae, and
Porcellanasteridae because ‘It has been generally agreed that the more primitive extant asteroids are the three
families in which the tube-feet lack suckers’ (1963a, p. 285). Of the three, the porcellanasterids were tentatively
eliminated because they lack superambulacrals, one of the ossicles inferred to be homologous with the virgals
of somasteroids.
More important to Fell’s (1963a, p. 385) phylogenetic ideas was the concept of growth gradients. Among
living sea stars, transverse rows are most clearly defined in the Luidiidae, less so (because the two systems
of gradients were seen as changing influence gradually) in the Astropectinidae and Porcellanasteridae. As
Fell had recognized the strongly developed transverse gradients in the lower Paleozoic somasteroids, their
development in the Luidiidae permitted recognition of this group as primitive among living families. The
gradients are weaker but present in the Astropectinidae (Fell, 1963c, p. 467), which was then next in the
phylogenetic sequence. The Porcellanasteridae is specialized in a number of features, and was therefore third
in the sequence.
Isolation of Platasterias as a somasteroid began with the realization that this genus ‘. . . exhibits growth
gradients identical with those of Ordovician somasteroids’ (Fell, 1963a, p. 383). Fell then obtained specimens
of Platasterias for dissections; this work led to the assignment of Platasterias to the Somasteroidea, as well
as to the refinement of a phylogenetic sequence extending from somasteroids through several steps of asteroid
evolution (see Table 1). The phylogenetic sequence was based on a variety of characters and emphasized
morphology of the fossils but not their stratigraphic position because the fossils appeared through an interval
of strata inferred to be too brief to permit recognition of phylogenetic sequence based on stratigraphic sequence
(1963a, p. 385). Fell (1967, p. 580) pointed out, however, that his phylogeny is consistent with what is known
of the stratigraphic sequence.
In summary, gradient recognition depends upon ossicle alignment. A continuing, gradual decline
in transverse gradient influence and a concomitant increase in the influence of logitudinal gradients
was suggested to have occurred during somasteroid/asteroid phylogeny.
Discussion. As pointed out by Fell, the ambulacrals, adambulacrals, superambulacrals, and
marginals are aligned in both Platasterias and Luidia (PI. 21, figs. 5, 6). Although Luidia typically
has a cluster of larger spines at the abradial ends of the marginals (PI. 22, figs. 1, 9) rather than
BLAKE: SEA STAR PHYLOGENY
18
a single larger radiole, as in Platasterias (PI. 21, fig. 5), the positions of the spines on the marginals
are the same. Although only a single large spine occurs on each Platasterias marginal, numerous
smaller spines are present, and further, number of spines apparently is not important, as is shown
by the variation found among Luidia species. Presence of a single marginal radiole in the petaloid
species Astropecten regalis (PI. 20, fig. 2) argues for the convergent evolution of marginal spine
reduction. Because there is no difference in ossicle alignment between Platasterias and Luidia, this
character cannot be used for taxonomic separation (although ossicle proportions do differ).
A more serious problem arises from the arrangement of the actinals and abactinals. These ossicles
in Luidia and Platasterias are relatively stout, and the actinals and the lateral two to four rows of
abactinals in Platasterias and the Clathrata Group species of Luidia are aligned with the four
ossicles (Fell, 1963a, pp. 392, 405) of the transverse series (PI. 20, figs. 1, 9-12; PI. 21, figs. 3-6).
Abactinal alignment with the marginals is weaker toward the middle of the abactinal surface and
among the other Luidia groups ( sensu Doderlein 1920) in general. A single row of actinals is
present in most species of Luidia, including members of the apparently primitive Clathrata Group
(Doderlein 1920), but extra series are present among the Alternata Group.
In some manner, the arrangement of the actinals and abactinals must be accommodated with
ideas on transverse gradients because the alignment satisfies the criterion of transverse gradient
recognition. Fell (1963a, p. 415) recognized the problem presented by the abactinals and noted
that these ossicles were not aligned in the primitive Chinianasteridae. He suggested (p. 416) that
in Platasterias the gradients would seem to be carried into the soft tissues in their path, and that
caused \ . . the paxillae to differentiate as if they were virgalia’. The alignment was thus considered
secondary. Fell also (1963a, p. 389) pointed out that actinal intermediates (i.e. actinals) are aligned
in transverse series in various astropectinid and phylogenetically {sensu Fell 1963a) later sea stars.
For discussion purposes, I have accepted the following hypotheses from Spencer’s and Fell’s
papers: (1) that asterozoans were derived from a crinoid ancestry, so that the primitive asterozoan
had radial rows of virgals; (2) that somasteroids are the primitive asterozoans and gave rise to the
asteroids; (3) that abactinals were present in the primitive somasteroid (because Spencer 1951,
described them in the earliest known taxa). If (1) or (2), as now understood, were rejected, then
the phylogenetic hypothesis using growth gradients could not be maintained. If (3) were rejected,
then the fossil record is misleading and where it is misleading can be interpreted only using external
criteria.
Beginning with these assumptions, abactinal alignment seen in Platasterias and Luidia can be
considered either primitive and present in the first somasteroid, or derived and appearing after the
first somasteroid grade. The actinals of Platasterias and Luidia can be considered either as primitive
and homologous with virgals, or not. If primitive, then their alignment must be primitive, i.e. they
are a part of the primitive transverse gradient. If secondary, their alignment must be secondary
as well (i.e. their alignment could not be part of a primitive transverse gradient if the ossicles
themselves did not occur in the primitive species). Where I treat the actinals as primitive, I will
treat the specific actinal arrangement seen in Platasterias and Luidia (i.e. a single actinal row) as
secondary because of the distribution of these ossicles in the accepted (see above) primitive
somasteroid.
First, if we accept both abactinal arrangement and actinal ossicles as seen in Luidia as primitive, then the
fossil record is incomplete and misleading. The abactinal pattern reported in somasteroids, including
Chinianaster, the inferred primitive genus, is an open meshwork of light paxilliform ossicles (Spencer 1951,
p. 91). If a hypothesis of declining influence of transverse gradients is to be followed, then the abactinal
arrangement seen in Platasterias and Luidia must have been derived from some unknown pre-chinianasterid
somasteroid, and retained while other characters (including ossicle differentiation) were evolving toward the
asteroid grade. This means that known fossil somasteroids had to be off the main line of somasteroid evolution
because their abactinal fields had already attained a derived, open meshwork state (text-fig. 1a; PI. 21, fig. 3),
a condition not reached in surviving asteroid evolution ( sensu Fell) until after the level of the Paxillosida.
Following the implications further because the abactinals of Luidia and Platasterias are sturdy, the primitive
abactinal condition in somasteroids likely was sturdy as well, and the relatively fragile ossicles of known
.82
PALAEONTOLOGY, VOLUME 25
fossils represent derived states. Thus, taxa with the relatively sturdy arrangement are unknown from the fossil
record, whereas the relatively fragile ones were preserved. Finally, the fossil record itself becomes unreliable
in that although most oral surface ossicle distributions in known fossils are considered to represent primitive
conditions, the abactinal arrangement is considered secondary for reasons external to the fossils themselves.
As to the actinal surface ossicles, their reduction sequence in known somasteroids began with a full oral
surface field (Chinianasteridae), continued through an intermediate step in which ossicles were present only
distally on the arms (Archegonasteridae), and ended with complete elimination of these ossicles (Archophi-
actinidae) (Spencer 1951). Thus the presence in Platasterias and most Luidia of a single row extending the
entire arm length must represent a lineage separate from known fossils because they match neither any known
fossils, nor the known reduction sequence.
The second approach is to assume, as Fell did, that abactinal alignment and actinal ossicles are secondary.
This path, however, also yields difficulties because it requires extension of the transverse gradients on to the
aboral surface at the time the influence of these gradients is hypothesized to be declining and the longitudinal
gradient influence increasing. Because the row of actinals also becomes aligned with adjacent adambulacrals
and marginals, we must hypothesize alternating rows of increasing and decreasing influence of transverse
gradients (i.e. adambulacrals, decreasing; actinals, increasing; marginals and marginal radioles, decreasing;
abactinals, increasing). Even if we were to argue that abactinals represent an ossicle system separate from
crinoids and not derived from them, and therefore not primitive in somasteroids, the problem of alignment
is not resolved. Abactinals do occur in an open meshwork pattern in the known primitive somasteroid
(Chinianaster), and Platasterias was considered to be approximately at the Villebrunaster level of organization,
at a post -Chinianaster pre-Archegonaster somasteroid grade. Transverse gradients are hypothesized to be
declining in influence during this sequence, yet in Platasterias these gradients had to be extending their
influence on to a new area of the body.
Further, because transverse alignment can be secondary among some ossicle systems, it might also be
secondary among all others. Platasterias therefore does not readily fit into a somasteroid/asteroid phylogeny
based on growth gradients. This does not challenge the hypothesis as applied only to known fossils.
Primitive and derived states could be rearranged so that abactinal alignment is considered primitive and
actinals derived, or vice versa, but this only combines inherent difficulties in different ways.
It would be possible to argue the large actinal field was not lost in the connecting links between somasteroids,
luidiids, and remaining asteroids, but this alternative removes Platasterias and Luidia from the main line of
asteroid evolution, for there is no known platasteriid or luidiid, fossil or living, with the appropriate morphology
(i.e. well-defined fascioles, and a large actinal field). Tethyaster is an astropectinid that comes quite close to
this morphology, and indeed Jangoux (1975) argued that this genus is in many ways intermediate between
Luidia and Astropecten. The marginal fascioles in Tethyaster, however, are relatively small, and the development
of fascioles also was important in Fell’s hypothesis, serving as feeding and respiration channels.
In any event beginning with the assumptions of Spencer (1951) and Fell (1962a et seq.), and
assuming either that abactinal alignment and actinal occurrence is primitive or secondary, known
fossil and modern taxa permit no application of the hypothesis of growth gradients to somasteroid/
asteroid phylogeny that does not require ignoring important aspects or morphology. In essence,
alignment is too extensive in Platasterias, more extensive than in the fossil somasteroids, and this
calls for explanation other than phylogenetic inheritance from a crinoid ancestry. Such an origin
is suggested below, under a functional explanation for alignment in Luidia.
As noted earlier here, individual ambulacral column and marginal ossicles of Platasterias are
transversely more elongate than are those of the fossil somasteroids. Again, Platasterias is showing
an inferred primitive trait more strongly than it is seen in the earliest known somasteroids, and
therefore an explanation other than inheritance of a primitive condition is demanded.
5. Other morphologic characters
The nature of the mouth frame. In modern sea stars, all ambulacral pairs are united across the
furrow by muscles and articular structures. In the somasteroids, Spencer argued that the proximal
ambulacrals (three pair in Chinianaster ) were not linked across the furrow, resulting in broad,
V-shaped openings termed buccal slits. These slits, however, are likely to have resulted from
post-mortem events. Spencer (1951, p. 100) argued that ambulacrals of certain small Chinianaster
were pulled apart after death, and Fell (1963a, p. 403) suggested that the buccal slits of Spencer
BLAKE: SEA STAR PHYLOGENY
183
resulted from such changes. The ossicles of the buccal slits in Archegonaster apparently show no
significant, distinctive morphologic characters (Spencer, 1951, fig. 9) as might be expected of ossicles
that form the margin of an opening rather than an articulated double column. Minor differences
are illustrated in these ossicles, but near-oral ambulacral column ossicles also are somewhat
distinctive in certain living sea stars (e.g. the Astropectinidae).
Arrangement of spinelets, along the fasciolar margins. In Platasterias , a series of small spinelets,
termed cover plates by Fell, line the edges of the adoral surfaces of the marginals and adambulacrals.
These spinelets were described as being held in a web and, when depressed, they cover the fasciolar
grooves between subsequent metapinnules or transverse gradient series (1963a, p. 397). The
arrangement was inferred derived from the crinoid ancestor, but differs in closing over the
interpinnular groove rather than over the pinnule surface, as in crinoids. The cover plates in
Platasterias were seen as protecting the fasciolar or feeding grooves.
A similar arrangement is found in Luidia, Astropecten, and other sea stars, although the ossicles
in these taxa typically are more slender, i.e. spine-like, and much more numerous than in Platasterias.
As in Platasterias , these ossicles can be depressed over the furrows. I was unable to recognize true
webbing in preserved specimens of Luidia (although such a covering is present in Goniopecten,
suborder Cribellina, order Paxillosida), but a mucus covering is present that in dried specimens
available to me can extend to the tips as well as between adjacent spinelets. The dense, overlapping
arrangement of many spinelets provides an effective cover for the fasciolar furrow. Various sea
stars, e.g. Dermasterias imbricata (PI. 21, fig. 9), are capable of pulling the sides of the furrow (i.e.
the adambulacrals) together and arching the furrow spines over the furrow. Functionally, these
are cover plates and such plates are widely distributed among the Asteroidea.
The arrangement found in Platasterias could be readily derived from that of Luidia by reduction
of spine number and minor changes of spine shape. Arrangement of fasciolar spinelets therefore
provides little justification for significant taxonomic separation of Luidia and Platasterias.
Super ambulacrals. Fell (1963a) argued that the superambulacrals of Platasterias and Luidia are
occluded second virgals. This explanation of origins is plausible, however, superambulacrals are
unknown in fossil somasteroids but they are present throughout the Luidiidae and Astropectinidae.
This distribution can only serve to isolate the fossil from the extant sea stars, not link them in a
single phylogenetic sequence.
Soft parts. Platasterias has a blind gut with caeca extending into the arms, an arrangement Fell
(1963a, p. 396) thought probably existed in fossil somasteroids. In addition, Platasterias has small,
simple non-suctorial tube-feet considered similar to those in Chinianaster (Fell 1962c, p. 474) and
small, double internal ampullae also inferred to be similar to those in Ordovician somasteroids.
Luidia, like Platasterias, has simple, non-suctorial tube-feet, double ampullae, and a blind gut with
caeca extending into the arms.
6. Feeding habits
Spencer (1951, p. 87) emphasized feeding behaviour in his tripartite division of stelleroids: The
grouping of starfish adopted here is based upon the activities of the arms, especially during feeding.’
The asteroids were seen as carnivores on larger organisms, primitive ophiuroids were believed
adapted for feeding on small particles on or in the bottom, and the somasteroids inferred to have
lived on planktic particles (in the suggested relatively primitive genus, Villebrunaster) or particles
from the surface sediment layers (in the suggested advanced somasteroid Archegonaster).
Fell also stressed observed and inferred feeding behaviour in his phylogenetic sequence. He
(1962a, p. 14) suggested two types of feeding are present in Platasterias, ‘microphagous ciliary
feeding’, and ‘selective detrital feeding’. In the former, particles of food were believed to be conveyed
in water currents along the fascioles to the arm radius, then proximally toward the mouth. In
specimens of Platasterias available to him. Fell observed amphipods in the mouth and in a food
groove; he suggested that the amphipods were collected by the tube-feet and passed to the mouth.
184
PALAEONTOLOGY, VOLUME 25
He inferred that this would be the limit of carnivorous feeding, because of the relatively small
mouth of Platasterias. A specimen of Platasterias in the U.S.N.M. collections has a foraminiferan,
small snail, and arthropod fragments in the mouth area; the presence of this small prey does not
appear fortuitous.
As observed by Fell (1963a), Astropecten and especially Luidia (PI. 21, fig. 6) are living species
with quite similar fasciolar furrows extending between the marginals and ambulacrals, yet these
are voracious predators of other echinoderms, molluscs, and arthropods. Although suspension
feeding has been suggested for both (Fenchel 1965, for Luidia sarsi; Gemmill 1915 and Gislen 1924
for Astropecten), Feder and Christensen (1966) doubt this behaviour occurs in these genera. The
furrows in both, as Fell (1963a, p. 391) observed for the Luidiidae, are probably respiratory in
nature. In contrast, taxonomically diverse genera in which suspension feeding has been observed
do not possess fasciolar furrows (e.g. Henricia, Spinulosida, Rassmussen 1965, ossicles are aligned
in this genus, yielding unobstructed channels between ossicle rows; Oreaster, Valvatida, Halpern
quoted in Anderson 1978; unidentified Brisingidae, Forcipulatida, Pawson 1978). Fell recognized
the differences in function between the fascioles of the modern sea stars and that inferred for the
somasteroids and assumed the function changed with the evolution of the asteroid grade.
Suspension feeding is difficult to document without direct observation. Gislen (1924) found
surface ciliary currents to be widespread in sea stars and therefore the potential for ciliary feeding
as well, although the currents flow away from the oral area in some species. Gislen (1924), however,
considered the functions of the currents to be largely for respiration and cleaning, although he did
observe some capture of particles.
The size of the mouth does not provide a useful guide to potential prey size limits, as suggested
by Fell for Platasterias (1962 b, p. 14), for both Astropecten and typical Luidia species, although
EXPLANATION OF PLATE 22
Fig. 1. Luidia alternata (Say). USNM 7528. Inclined view of disc region of specimen not distorted by food;
specimen in alcohol, x 1.
Figs. 2, 3. Luidia alternata (Say). USNM 7528. 2, inclined aboral, x and 3, oral view, x 1, of individual
containing unbroken corona of Lytechnius varegatus (Lamarck). The peristome of the echinoid (arrow,
pointing to gill slit) is centred on that of the sea star. The sea star major radius is 140 mm, relaxed minor
radius 20 mm, and height 1 5 mm. The echinoid corona diameter about 50 mm, height 25 mm. All dimensions
approximate. Spines and organic materials are missing from the echinoid but pieces of the Aristotle’s Lantern
are present. The echinoid thus apparently was dead or nearly so at time of ingestion and it had been
consumed and presumably was soon to be expelled at time of collection. In alcohol.
Figs. 4, 5. Luidia clathrata (Say). USNM E17599. 4, view into disc; and 5, lateral view of fragmentary
individual, radius about 100 mm, containing a specimen of Moira atropos (Lamarck), length about 40 mm.
The echinoid is now incomplete (arrows point to remaining coronal plates) but the aboral surface of the
sea star retains the form of the sea urchin, hence the echinoid was complete when ingested. Fine lines beyond
the echinoid plates are its spines; the echinoid apparently was ingested alive, or very recently dead; specimen
is dry, x 1 .
Fig. 6. Luidia clathrata (Say). USNM 8442. Inclined aboral view of specimen, radius about 1 10 mm, containing
Moira atropos, length about 40 mm. Spines are retained on prey, x 1.
Figs. 7, 8. Luidia clathrata (Say). USNM E3268. 7, aboral view; and 8, oral view of sea star containing a
broken but not distorted Mellita quinquiesperforatus (Leske). The spines are missing from the echinoid.
Sea star radius about 150 mm, that of the echinoid is about 50 mm. Aboral ossicles of the sea star are
distorted from life position, beyond the edge of the sea star in a relaxed state (arrow); the aboral surface
of the sea star was partially opened after collection. Compare colour marking with that of L. (P.)
latiradiata, PI. 20, fig. 1; specimen is dry, x
Fig. 9. Luidia clathrata (Say). Oral view of mouth area of specimen not distorted by food; specimen is dry,
x 1 . Luidia can consume particles of dimensions many times greater than the 2 or 3 mm mouth diameter,
and of volumes greater than that of the relaxed disc. Much caution is needed in interpreting possible food
particle size in fossil organisms with flexible bodies.
PLATE 22
1
BLAKE, sea star Luidia
PALAEONTOLOGY, VOLUME 25
effective predators, also have relatively small oral frames. The lack of importance of mouth size
is clearly seen in a series of Luidia specimens in the U.S.N.M. collections (PI. 22). In each example
illustrated, the size of the ingested echinoid exceeds that of the relaxed diameter of the disc.
Peristome size, measured in related Luidia specimens of comparable sizes, is 2 to 3 mm; oral spines
mounted around the mouth of the sea star overlap over the peristomial opening. Occurrence of
large prey in a number of specimens clearly demonstrates the behaviour is not unusual. Astropecten,
too, is capable of feeding on relatively large prey organisms (Clark 1962, pi. xiii).
Thus, Luidia has fasciolar grooves similar to those of both Platasterias and presumably of the
ancient somasteroids, but there is no clear evidence that either Platasterias or Luidia make use of
suspension feeding and therefore the presence of fascioles in Palaeozoic somasteroids provides no
morphologic evidence of suspension feeding in these organisms. The presence of very large prey
in Luidia means that mouth and body size provide no clear guide to food particle size in echinoderms
with a loosely articulated skeleton, including the Palaeozoic somasteroids.
Summary
Platasterias is very close to typical species of Luidia in all characters except those related to the
transverse elongation of ossicles, and similar to Palaeozoic somasteroids only in ways it is also
similar to Luidia. Greater differences in ossicle morphology, except for this transverse elongation,
are seen between species of the Ciliaris Group of Luidia {sensu Doderlein 1920) and Clathrata Group
( sensu Doderlein 1920) than are seen between Platasterias and the Clathrata Group of Doderlein.
I therefore consider (Blake 1972) Platasterias to be a subgenus of Luidia assignable to the Clathrata
Group, and not a somasteroid as suggested by Fell. As noted above, Madsen (1966) carried this
interpretation one step further, for he did not believe Platasterias warranted subgeneric recognition.
Summary comparison among sea star taxa is provided in Table 2, with Porania included as an
example of an asteroid well removed from the Luidiidae.
If Luidia were to be extensively subdivided, perhaps along the lines suggested by Doderlein
(1920), then Platasterias could be recognized at the generic level, but on a morphologic basis, it still
would have to be included in the same family as other Luidia species. Luidia, as noted by Fell
(1963a), clearly is a member of the Asteroidea.
THE NATURE OF LUIDIA
The affinities of the Luidiidae
In his work. Fell assumed the sea stars with nonsuctorial tube-feet, and in particular the
astropectinids and luidiids are primitive among extant sea stars. I agree with this interpretation
but, although consensus might lean toward the families cited, no general agreement has ever been
reached.
Mortensen (1922), noting that few authors had commented directly on the question of primitive
position, queried the then-active sea star workers as to their opinions. W. K. Fisher, H. L. Clark,
and R. Koehler all considered the astropectinids to be primitive, as did Mortensen himself, whereas
Doderlein thought the asterinids were primitive. Earlier, Perrier (1884) argued the forcipulates
were primitive on inferred directions of pedicellariae evolution.
Most active sea star taxonomists list the Paxillosida first in their faunal lists, seemingly thereby
implying an inferred primitive position, but as A. M. Clark pointed out to me (pers. comm.), this
is simply because they are following Fisher (1911).
In this diagnosis of the Platyasterida (including the Palasteriscidae and Luidiidae), the only
character listed by Fell (1963a, p. 392) was the development of transverse growth gradients. Spencer
and Wright (1966) expanded upon the diagnosis, but most features listed by these authors also
apply to members of the Paxillosida. They did include the presence of a single row of marginals
in the Platyasterida. Although relatively small in Luidia compared to those in, for example,
Astropecten, superomarginals are present in Luidia (PI. 20, figs. 9, 10), recognizable on the basis
of position of origin at the terminal (see criteria listed by Blake 1978), and by size among primitive
BLAKE: SEA STAR PHYLOGENY
187
( sensu Doderlein 1920) Luidia species. This leaves only the transverse gradients to unite the Luidiidae
with the Palaeozoic Palasteriscidae, but this character, as discussed above, appears unreliable.
Based on illustrations provided by Spencer (1919 in 1914-1940), however, Luidia and the
Palasteriscidae are distinct in ossicle morphology, and fascioles apparently are lacking in the fossil
family. In addition, they are separated by a Devonian to Miocene interval.
The Luidiidae, however is similar to the Astropectinidae in many soft and hard part characters,
as summarized by Fisher (1911), Fell (1963a), and other workers; included here are the nature of
the paxillae and other ossicles, and the presence of non-suctorial tube-feet. Following McKnight
(1977), I therefore have herein returned the Luidiidae to the Paxillosida.
A partial revised classification of the stellate echinoderms based on Spencer and Wright (1966),
McKnight (1975) and the arguments presented here is provided below. For comparative purposes,
the equivalent classification of Spencer and Wright (1966) is also included.
Revised classification:
Class Somasteroidea
Order Goniactinida
Family Chinianasteridae
Family Villebrunasteridae
Family Archegonasteridae
Family Archophiactinidae
Family Helianthasteridae
Class Asteroidea
Order Platyasterida
Family Palasteriscidae
Order Paxillosida
Suborder Flemizonina
Suborder Diplozonina
Family Luidiidae (includes Luidia
( Platasterias ) ladradiata)
Family Astropectinidae
Family Porcellanasteridae
Suborder Cribellina
Class Ophiuroidea
Spencer and Wright (1966):
Class Stelleroidea
Subclass Somasteroidea
Order Goniactinida
Family Chinianasteridae
Family Villebrunasteridae
Family Platasteriidae (includes Platasterias
ladradiata)
Family Archegonasteridae
Family Archophiactinidae
Subclass Asteroidea
Order Platyasterida
Family Palasteriscidae
Family Luidiidae
Order Paxillosida
Suborder Hemizonina
Suborder Diplozonina
Family Astropectinidae
Family Porcellanasteridae
Suborder Cribellina
Subclass Ophiuroidea
The gap in the fossil record of the Platyasterida sensu Fell 1963a
The fossil record does not support an inference that the lineage leading to Luidia has endured
nearly unchanged from the Palaeozoic. The order Platyasterida sensu Fell (and Spencer and Wright
1966) includes three genera assigned to two families: Platanaster (M. Ord.) and Palasteriscus (L.
Dev.), both belonging to the Palasteriscidae, and Luidia (Mio.-Rec.), belonging to the Luidiidae.
There is thus a gap from Lower Devonian to Miocene in the inferred record of the order. At first
inspection, this appears trivial because of the sketchy record of sea stars, but I believe it becomes
more important when this record is carefully considered.
Among post-Palaeozoic sea stars, Astropecten and its close allies have a relatively good fossil
record. This record seems to result from body structure, habitat, and habits. The sea stars have
relatively stout marginal frames and at least Astropecten frequently lives in shallow waters on
unconsolidated substrates, and burrows beneath the surface. The relative abundance of Astropecten
as fossils does not necessarily result from a greater, enduring abundance (although they are very
common today), but from where and how individuals live. Luidia lives in similar environments,
and it too is a burrower; although the marginal frame is not as stout as in Astropecten, the
inferomarginals of many species are comparable to those in Astropecten, and sturdier than those
of the Asteriidae, a family known from the Jurassic and whose representatives are also sometimes
found in similar habitats.
PALAEONTOLOGY, VOLUME 25
table 2. Comparison of somasteroids and asteroids. Criteria available do not permit separation of
Platasterias as a somasteroid. Porania (Poraniidae) is taxonomically distant from the somasteroids
but still has many features in common with them. Data in part from Fell (1963a).
ambulacra 1
ossicle alignment furrow
super-
ambulacral
terminal
tube feet
somasteroids
primarily transverse in not erect
some, others longitudinal
absent
present or
absent
?
L. (Platasterias) both transverse and erect
longitudinal
present
present
non-suctorial
Luidia clathrata both transverse and erect
longitudinal
present
present
non-suctorial
Porania
primarily longitudinal erect
absent
present
suctorial
other asteroids
primarily longitudinal erect
present
or
absent
present
non-suctorial
and suctorial
anus
body shape
feeding habits
somasteroids
?
petaloid to large
oral disc
inferred suspension or
small particle bottom
feeder
L. (Platasterias)
absent
petaloid
carnivore
Luidia clathrata
Porania
absent
present
small disc, strap-
shape arms
large oral disc
carnivore
suspension feeding, other
other asteroids
present
or
absent
varied
varied
Luidia is a common genus, with sixty or more species (many described since Doderlein 1920)
widely distributed in temperate to tropical seas; further, individuals commonly are abundant.
Individuals of most species of Luidia are relatively large and would not be readily overlooked in
the fossil record. Luidia superba specimens have been reported from the Galapagos Islands with
major radii up to 415 mm and an arm breadth at the disc edge of 60 mm (Downey and Wellington
1978). The absence of a fossil record for platyasterids ( sensu Fell) from Devonian to Miocene
certainly is not conclusive evidence that they were not present, but because structurally comparable
sea stars known from similar environments do have long records, the absence of Luidia or its
relatives requires explanation.
A functional explanation for the transverse alignment of ossicles in Luidia
Fell argued that the body plan of Luidia is the direct phylogenetic heritage of Cambrian crinoids;
in effect, that functional explanations of the morphology of these contemporary, mobile carnivores,
living with their oral surfaces toward the substrate, are to be sought in ancient, attached
suspension-feeding organisms living with their oral surfaces directed into the water column. 1
believe an explanation for the Luidia body plan can be developed that is more closely linked to
the life needs of sea stars.
BLAKE: SEA STAR PHYLOGENY
189
The arms of Luidia are very flexible in part because they are narrow and low and in part because
the alignment of ossicles yields essentially a segmented pattern in which each radial row of ossicles
is relatively weakly connected to adjacent rows. This arrangement weakens toward the middle of
the aboral surface of the arm, where the abactinals are too small to interfere with flexibility.
The value of flexibility to Luidia can be correlated with preferred food and habitat. Luidia is a
predator on larger solitary organisms-molluscs, arthropods, and other echinoderms— in which
long, flexible arms are useful for the manipulation of prey. Luidia is commonly an inhabitant of
shallow, often agitated, bottoms and, further, it frequently burrows into the substrate. Arm flexibility
is useful for burrowing (Heddle 1967) and also for righting (Blake 1981).
L. ( P .) latiradiata shares the ossicle arrangement of traditional Luidia species, and it, too, is an
active predator (see above), but in it the demands of environment (Madsen 1966) might have
dominated the need for maximum arm flexibility.
A superficially similar transverse alignment is seen in Asterias and other asteriids in the rows
of arm ossicles immediately lateral to the furrow columns (PI. 21, fig. 7). The ancestry of the
Asteriidae is unknown, and therefore it is not known when this alignment evolved, but asteriids,
like Luidia, are active predators with a need for relatively flexible arms.
Another hypothesis for ossicle alignment in Luidia and Astropecten can be based on the burrowing
habits of these sea stars combined with the respiratory current flow described by Gislen (1924).
Although surface currents appear typical of sea stars, deep fascioles seemingly are restricted to
burrowers. These channels, combined with their cover spinelets, should provide protection from
sediment interference for the current flow. Alignment of ossicles and their intervening channels in
turn would permit a more efficient water flow than that possible in a species with an irregular
arrangement of ossicles. Surface-dwelling species presumably would suffer less from current
disruption and, in them, furrows generally are lacking.
Archaster is an extant genus superficially very similar to Astropecten in both form and habit. It
too has relatively deep fascioles (PI. 21, fig. 8), but only between the infero-, and not the
superomarginals. Archaster, however, is an atypical member of the Valvatida rather than the
Paxillosida, as is Astropecten. It is possible that Archaster was derived directly from an
Astropecten- like source but the skeletal morphology in this genus does not seem to support such
an idea; convergence resulting from habitat similarities is a preferable hypothesis.
Thus, both the alignment of ossicles in the arm and the grooves between the aligned ossicles
seem subject to convergent evolution; use in any phylogenetic scheme must be made with great care.
SUMMARY
Platasterias latiradiata Gray is removed from the Somasteroidea and assigned at subgeneric rank
to the genus Luidia of the monogeneric asteroid family Luidiidae; Luidia ( Platasterias ) therefore
should not be singled out as a model for the reconstruction of the biology of early stellate
echinoderms. Reasons for the transfer of L. ( Platasterias ) are: (1) The single distinctive feature of
the Palaeozoic somasteroids seems to be the absence of a permanent ambulacral furrow. In contrast
the adambulacral/ambulacral arrangement in L. (P.) latiradiata is the same as that found in L.
clathrata and essentially as in all other living asteroids; a true furrow is present in L. (P.) latiradiata.
(2) Individual ossicle morphology and ossicle and muscle arrangement of L. ( Platasterias ) are
easily within a reasonable range of variation for Luidia. Although many (but not all) ossicles of
L. (P.) latiradiata are proportionately broad, they are morphologically closer to the Clathrata
Group species of Luidia than these ossicles are in turn to the Ciliaris Group species of Luidia.
(3) A reconstruction of phylogenetic history based on ossicle alignment (growth gradients) requires
important reversals of direction of evolution in order to enable fossil somasteroids and extant L.
( Platasterias ) and other luidiids to fit the hypothesis. (4) Two important ossicle types, the
odontophore and superambulacral, are present in L. ( Platasterias ) and other extant sea stars, but
they apparently are absent from the fossil somasteroids. (5) Although L. ( Platasterias ) has a petaloid
arm shape suggestive of that of certain crinoids, weakly developed petaloid arms are present in
190
PALAEONTOLOGY, VOLUME 25
other taxonomically widely separated sea star taxa; the character is subject to convergence. (6)
Known food habits of L. ( Platasterias ) suggest similarities to those of typical Luidia species and
not to those habits inferred by Spencer (1951) for the fossil somasteroids. (7) Available data suggests
soft-part morphology of L. ( Platasterias ) is essentially that of a typical Luidia.
The Luidiidae is transferred from the otherwise Paleozoic order Platyasterida to the common
post-Palaeozoic Paxillosida.
The origin of the transverse alignment of ossicles in Luidia (including Platasterias ) is ascribed
to habits and habitats of the sea star. This ossicle arrangement enhances arm flexibility, useful to
an active, burrowing predator living in shallow, often turbulent, environments. The presence of
deep furrows between aligned ossicle series also provides an open channel, unimpeded by sediment,
for the flow of respiratory water currents.
Acknowledgements. I thank Ailsa M. Clark, Maureen E. Downey, and David L. Pawson for very helpful
reviews of an earlier version of the manuscript; Porter M. Kier for identification of the echinoids; Richard
L. Turner for discussion of certain points; the authorities of the U.S. National Museum for loan of specimens;
and H. Barraclough Fell and the authorities of the Royal Society for permission to reproduce certain drawings.
REFERENCES
algor, J. r. 1971. Structure of the ambulacral system in some modern and early Paleozoic starfish. Abstr.
Prog. Geol. Soc. Am., Wash. D.C. 3, 489-490.
anderson, j. m. 1978. Studies on functional morphology in the digestive system of Oreaster reticulatus (L.)
(Asteroidea). Biol. Bull. mar. biol. Lab., Woods Hole. 154, 1-14.
blake, D. B. 1967. Skeletal elements in asteroids. Ab. 1967, Spec. Pap. Geol. Soc. Am. 115, 15-16.
— 1972. Sea star Platasterias : ossicle morphology and taxonomic position. Science, N.Y. 176, 306-307.
— 1973. Ossicle morphology of some recent asteroids and description of some west American fossil asteroids.
Univ. Calif. Pubis, geol. Sci. 104, 60 pp.
1978. The taxonomic position of the modern sea-star Cistina Gray, 1840. Proc. biol. Soc. Wash. 91,
234-241.
— 1981. The new Jurassic sea star genus Eokainaster and comments on life habits and the origins of the
modern Asteroidea. J. Paleont. 55, 33-46.
Clark, A. m. 1953. Notes on asteroids in the British Museum (Natural History) III. Luidia. Bull. Br. Mus.
nat. Hist., zool. 1, 379-412.
— 1962. Starfishes and their relations. Brit. Mus. (nat. Hist.), Pub. 377. 118 pp.
doderlein, l. 1920. Die Asteriden der Siboga-Expedition II: Die Gattung Luidia und ihre Stammesgeschichte.
Siboga-Expeditie Mono. 46b, 193-294.
downey, m. e. and Wellington, G. m. 1978. Rediscovery of the giant sea star Luidia superba A. H. Clark in
the Galapagos Islands. Bull. mar. Sci. 28, 375-376.
feder, h. m. and Christensen, a. m. 1966. Aspects of asteroid biology. In boolootian, r. a. (ed.). Physiology
of echinodermata. Interscience Pub., John Wiley & Sons, New York. 822 pp.
fell, h. b. 1948. Echinoderm embryology and the origin of chordates. Biol. Rev. 23, 81-107.
— 1962a. A surviving somasteroid from the eastern Pacific Ocean. Science, N.Y. 136, 633-636.
— 19626. A living somasteroid, Platasterias latiradiata Gray. Paleont. Contr. Univ. Kans. Artie. 6, Echino-
dermata. 16 pp.
— 1963a. The phylogeny of sea-stars. Phil. Trans. R. Soc., Lond., ser. B 246, 381-435.
— 19636. A new family and genus of Somasteroidea. Trans. R. Soc. N.Z. 3, 143-146.
— 1963c. The evolution of the echinoderms. Rep. Smithson. Instn. 4559, 457-490.
— 1965. Reply to ancestry of sea-stars by G. M. Philip. Nature, 208, 768-769.
— 1967. Echinoderm ontogeny. In moore, r. c. (ed.). Treatise on invertebrate paleontology pt. S,
Echinodermata 1, 1, S1-S296; 2, S297-S650. [Lawrence, Kansas, and Geol. Soc. Am.]
fenchel, t. 1965. Feeding biology of the sea star Luidia sarsi Diiben and Koren. Ophelia, 2, 223-236.
fisher, w. k. 1911. Asteroidea of the North Pacific and adjacent waters. Bull. U.S. natn. Mus. 76. 420 pp.
gemmill, J. F. 1915. On the ciliation of asterids, and on the question of ciliary nutrition in certain species.
Proc. zool. Soc. Lond. 1915, 1-19.
BLAKE: SEA STAR PHYLOGENY
191
gislen, t. 1924. Echinoderm studies. Zool. Bidr. Upps. 9, 1-316.
heddle, d. 1967. Versatility of movement and the origins of asteroids. In millott, n. Echinoderm Biology.
Symp. zool. Soc. Lond. 20. 240 pp.
hotchkiss, f. h. c. 1977. Ophiuroid Ophiocanops (Echinodermata) not a living fossil. J. nat. Hist. 11, 377-380.
and clark, A. M. 1976. Restriction of the family Poraniidea sensu Spencer and Wright, 1966 (Echino-
dermata: Asteroidea). Bull. Br. Mus. nat. Hist., zool. 30, 263-268.
hyman, L. H. 1955. The invertebrates: Echinodermata. McGraw-Hill, New York. 764 pp.
jangoux, M. 1975. Note on the genus Tethyaster Sladen. Rev. zool. Afr. 89, 761-768.
mcknight, d. G. 1975. Classification of somasteroids and asteroids (Asterozoa: Echinodermata). J. Roy. Soc.
N.Z. 5, 13-19.
— 1977. Classification of recent paxillosid sea-stars (Asterozoa: Echinodermata). N.Z. Ocean. Inst. Rec. 3,
113-119.
madsen, F. j. 1966. The Recent sea-star Platasterias and the fossil Somasteroidea. Nature, 209, 1367.
mortensen, t. 1922. Echinoderm larvae and their bearing on classification. Ibid. 110, 806.
— 1931. Contributions to the study of the development and larval forms of echinoderms. K. danske
Vidensk. Selsk. Skr., natur. math. 9th ser., 4, (1). 32 pp.
pawson, d. l. 1978. Some aspects of the biology of deep-sea echinoderms. Thalassia jugosl. 12, 287-293.
pearse, j. s. 1969. Treatise on invertebrate paleontology. Part S: Echinodermata 1 (a review). Q. Rev. Biol.
44, 298.
perrier, e. 1884. Memoire sur les Etoiles de mer recueillies dans la mer des Antilles et le golfe du Mexique
par M. Agassiz. Nouv. Archs Mus. Hist, nat., ser. 2, 6, 127-276.
philip, g. m. 1965. Ancestry of sea-stars. Nature 208, 766-768.
Rasmussen, H. w. 1965. Asteroids du Tertiaire inferieur de Libye (Afrique du Nord). Ann. Paleont. Invert.
52, 1-15.
spencer, w. k. 1914-1940. British Palaeozoic Asterozoa. Palaeontogr. Soc. ( Monogr .) 540 pp.
— 1951. Early Palaeozoic starfish. Phil. Trans. R. Soc., Lond., ser. B 235, 87-129.
— and wright, c. w. 1966. Asterozoans. In moore, r. c. (ed.). Treatise on invertebrate paleontology,
pt. U, Echinodermata 3, 1, Ul-U366a; 2, U367-U696. Lawrence, Kansas, and Geol. Soc. Am. Press.
D. B. BLAKE
Department of Geology
University of Illinois
245 Natural History Building
1301 W. Green St.
Urbana, Illinois 61801
U.S.A.
Typescript received 30 September 1980
Revised typescript received 10 January 1981
A LOWER CARBONIFEROUS AlSTOPOD
AMPHIBIAN FROM SCOTLAND
by CARL F. WELLSTEAD
Abstract. "Ophiderpeton' , a mid-Visean ai'stopod from the Wardie Shales near Edinburgh is described.
Although ‘ Ophiderpeton ’ possesses many of the attributes characteristic of ai'stopods and seems generally more
closely comparable to ophiderpetontids than to phlegethontiids, details of cranial anatomy such as possession of
a relatively short skull, short parietals, and the absence of a tabular-parietal contact, as well as the presence of
spinal-nerve foramina in only a portion of the vertebral column and the absence of tetra-radiate ribs,
postcranially, distinguishes ‘ Ophiderpeton ’ from both of the currently recognized ai'stopod families. On the basis
of these differences, ‘ Ophiderpeton ’ is renamed, Lethiscus stocki, and a new family, Lethiscidae, is erected to hold
the new species.
Of all Palaeozoic amphibians, the ai'stopods are the most specialized and phylogenetically isolated.
The genera known from the Upper Carboniferous are totally without limbs and have up to 230 trunk
vertebrae. The vertebrae are simple cylinders, without a trace of trunk intercentra or caudal haemal
arches. The skull has lost much of the dermal cover common to labyrinthodonts and rhipidistians.
These snake-like ai'stopod genera have been classified among the lepospondyls along with nectrideans
and microsaurs, sharing with them such features as holospondylous vertebrae, lack of an otic notch,
absence of labyrinthine infolding of teeth, and absence of palatine teeth. It is not certain that these
features indicate common ancestry.
In addition to the familiar Upper Carboniferous genera, a single specimen from the Lower
Carboniferous, originally identified by Thomas Stock (1882) as an ai'stopod, has lain undescribed for
nearly a century. The latest discussion of this specimen was that of Baird ( 1 964), who accepted Stock’s
identification, and cited a number of broad similarities it held with the Upper Carboniferous genera.
Baird also mentioned the difficulty in studying the specimen which had contributed to its neglect. To
gain more information about the early differentiation of Palaeozoic amphibians, a further attempt
has been made to study the specimen.
The specimen (text-fig. 1) is 49 cm long and is preserved in a very elongate nodule, which was
originally broken into dozens of pieces, each of which was split through the middle to give a series of
sections through the skull and vertebrae. No attempt was made by Stock to prepare the specimen
further. Preliminary efforts to remove the matrix from the bone, mechanically and chemically, have
not been successful, although X-ray photography and tomography have revealed some detail of the
skull and postcranial skeleton otherwise concealed in matrix.
Presently, enough detail can be seen in the sections to give a general description of the animal and to
provide sufficient information about its anatomy to allow discussion of its relationships to other
lepospondyls. The only further ‘preparation’ has been to clean glue and a century’s accumulation of
grime from the specimen. This cleaning revealed with startling clarity vertebrae, ribs, scales, and
sections through the skull. The bone is preserved, but is outlined with a thin coating of pyrite. The
neural and notochordal canals and large lacunae within the bones and skull are filled with calcite.
Details of bone histology are evident, and the random breakage of the nodule has produced a series of
i sections in many planes. Information provided by these sections would be lost were the bone to be
removed for production of latex casts following Baird’s technique (Baird 1955); therefore, this
otherwise very effective method of preparing specimens was not attempted. Normal external views of
the bone are rarely evident, but those available reveal sufficient detail along the column to allow
[Palaeontology, Vol. 25, Part 1, 1982, pp. 193-208.|
194
PALAEONTOLOGY, VOLUME 25
text-fig. 1. Lethiscus stocki, MCZ 2185. a, whole specimen, x^. b, skull, x 3.
WELLSTEAD: LOWER CARBONIFEROUS AMPHIBIAN
95
description of some regional differentiation. Revealed in the cleaned specimen were several
characteristics which ally it with the Ai'stopoda, but also others which distinguish it from ai'stopod
families Phlegethontiidae and Ophiderpetontidae.
MATERIALS AND METHODS
A standard medical X-ray machine and the Stratomatic tomography X-ray machine with tri-spiral movement
and 0-6 focal spot were used to produce the X-ray photographs critical to the description of this specimen. The
film used was Ilford X-ray film.
Text-fig. 3 is an X-ray of the half-nodule containing the skull less the posterior skull roof. X-rays of the half-
nodule bearing the posterior skull roof revealed no more than can be seen with the naked eye. The matrix of the
nodule is too dense for successful application of X-rays to the assembled skull-bearing nodule halves.
Abbreviations used in figures
bo— basioccipital; bs— basisphenoid; c— coronoid; cap— capitulum; ect — ectopterygoid; ep— epipterygoid;
f — frontal; it — intertemporal; j — jugal; 1 — lacrimal; m — maxilla; mnd — mandible; n — nasal; ot — otic capsules;
p — parietal; pf — postfrontal; pm — premaxilla; po — postorbital; poz — postzygapophysis; pp — postparietal;
prf — prefrontal; prz — prezygapophysis; ps — parasphenoid; pt — pterygoid; q — quadrate; qj — quadratojugal;
sq— squamosal; st— supratemporal; t— tabular; tub— tuberculum; v— vomer.
SYSTEMATIC PALAEONTOLOGY
Class AMPHIBIA
Sub-claSS LEPOSPONDYLI
Order aistopoda
Family lethiscidae fam. nov.
Diagnosis. Same as for the only known genus.
Lethiscus gen. nov.
Type species: Lethiscus stocki sp. nov.
Diagnosis. A small elongate amphibian with holospondylous vertebrae and short-snouted skull
bearing lateral temporal fenestrae. Orbits placed in the anterior third of the skull and separated from
the temporal fenestrae by the postorbital bones. The bones of the prefrontal-postfrontal-postorbital
series increase in size posteriorly. Intertemporal absent. Parietals roughly equivalent in length to the
frontals and surrounding a large parietal opening. Tabulars do not contact the parietal bones.
Postparietals relatively large. Mandibles deep and long, approximating the length of the skull.
The trunk is very long. Differentiation along the vertebral column is expressed by the presence in
the posterior portion of the column of spinal-nerve foramina, serrated neural spines and transverse
processes rising, in part, from the centra. Anteriorly, transverse processes arise solely from the neural
arches and spinal-nerve foramina are absent. Ribs are bicipital and robust.
Etymology. The generic name extends Cope’s practice of naming serpen tiform ‘lepospondyls’ for rivers in Hades.
In this case, Lethe is a stream named for the Greek god of forgetfulness.
Lethiscus stocki sp. nov.
Diagnosis. The same as for genus. Specific name honours the discoverer of the specimen.
Holotype. MCZ 2185. Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts,
U.S.A. Skull and postcranial skeleton. This is the only known specimen.
Locality and horizon. Stock (1882) discovered the specimen in the shales of the Wardie shore, north of
Edinburgh, Scotland (Wood 1977, gives further locality data). These beds, the Wardie Shales, lie in the middle of
196
PALAEONTOLOGY, VOLUME 25
the Lower Oil Shale Group (Mitchell and Mykura 1962) and have a mid-Visean age (text-fig. 2d). The base of the
Arthur’s Seat Volcanic Beds defines the base of the group in the Edinburgh vicinity, though they cannot be traced
regionally. Fitch, Miller, and Williams (1970) report a potassium-argon date of 347 ± 5 my B.P. from these
volcanics. George, Johnson, Mitchell, Prentice, Ramsbottom, Sevastopulos, and Wilson (1976) place volcanic
rocks dated at 338 + 4 my B.P. (Fitch et al. 1970) in the lower portion of the Upper Oil Shale Group. The Wardie
Shales can, therefore, be estimated as approximately 340 million to 345 million years old. George et al. (1976)
correlate the mid-Visean with the Middle Mississippian (Meramec) of North America.
DESCRIPTION
Skull
The skull (text-fig. 1) is exposed in an irregular para-frontal fracture which in one half-nodule yields an
unobscured view of a portion of the left suspensorium and of the ventral surface of the skull table posterior to the
fronto-parietal suture. In the other half-nodule only extremely irregular sections of the remainder of the skull can
be seen. A second fracture yields an oblique section of the left lateral aspect of the skull. The circum-orbital bones
have been interpreted through study of the specimen, an X-ray of the second half-nodule (text-fig. 3) and
tomographs. The remainder of the skull, including the snout, lateral and ventral margins, portions of the palate
and mandibles have been interpreted almost entirely from this X-ray and from the tomographs.
The skull is preserved in three dimensions with some distortion and fracturing of the snout, ventral skull
margins, and tabular-suspensorium regions. As is the case with the post-cranial portion of the specimen, the
text-fig. 2. a, correlation chart of Lower Carboniferous stratigraphy in the Scottish Midland Valley (after
George et al. 1976). b, range chart of the ‘lepospondyls’ (Carroll 1977; Olson 1972; Thomson and Bossy 1970).
WELLSTEAD: LOWER CARBONIFEROUS AMPHIBIAN
197
bone is well preserved, but much softer than the enclosing matrix, rendering mechanical preparation hazardous.
Bones are coated with pyrite, which accentuates sutures as well as cracks, but adheres closely to the bone. This
coating obscures any ornamentation and evidence of lateral line canals which might exist. The skull is triangular
in shape, widest at its posterior extreme and is approximately twice as wide as it is high. Openings are present in
the skull roof for external nares, orbits, the parietal opening, and temporal fenestrae. The orbits are in the
anterior third of the skull.
text-fig. 3. X-ray photograph of skull, Lethiscus stocki, MCZ2185, less the skull roof
posterior to fronto-parietal suture, x 3.
198
PALAEONTOLOGY, VOLUME 25
The relationships of the bones in the skull table can be viewed directly and are, therefore, more confidently
interpreted than are those seen in X-ray. The bones surrounding the parietal opening are assumed to be the
parietals and further homologies within the skull table follow the discussion of the skull bones in Ophiderpeton
by Thomson and Bossy (1970, p. 24).
The parietals bear dentate sutures with surrounding bones and are fused to one another posterior to the
parietal opening. The postparietals are also fused and together with the parietals comprise two-thirds of the skull
table area. The postparietals are broader posteriorly than at the parietal border and extend to the rear margin of
the skull.
The postparietals are excluded from the temporal fenestrae by the supratemporal and tabular bones.
Intertemporal bones are absent. The supratemporals are rectangular bones approximately three times as long as
they are wide. Their medial sutures with the postparietals and parietals are smoothly sinuous while their anterior
margins have dentate sutures with the parietal bones. The posterior margins are poorly defined. Ventrally, the
supratemporals appear to bear sutures with the pterygoids, although the contact is obscured by sediment and
breakage along the parafrontal fracture.
A suture-like lineation, seen in the specimen intersecting the posterior margin of the left postparietal, taken in
conjunction with the posterolateral borders of the skull table and the posterior regions of the supratemporals,
provides the evidence for the tabular bones. An element between the right supratemporal and the fused
postparietals appearing to bear a contact with the parietals may be a medial process of the right tabular.
However, the mate of this ‘tabular’ process is not found on the left side of the skull table, suggesting that the
feature is more probably an irregular fracture within the postparietals and that the tabulars have no contact with
the parietals. The ventral margins of the tabulars are obscured by sediment and breakage, but would seem to
have had a short suture with the squamosal posterior to the squamosal-pterygoid contact.
The relationships of the frontals, parietals, and the right orbital series can be observed directly. The frontals
are paired, narrow, and equal in length to the parietal and postparietal. They share an interdigitating suture with
the parietals, but have a sinuous suture with pre- and postfrontals and postorbitals.
The fragment anterolateral to the left frontal is taken questionably as the small left nasal.
The postfrontal and postorbital separate the orbit from the temporal fenestrae. The prefrontals are small
elements of indistinct shape. The postfrontals are approximately twice as long as the prefrontals and are wedge-
shaped. While the posterior portion of the right postorbital is missing, the left one seems to be complete and is
large, approximately 2-5 times the size of the postfrontal. The exact nature of the relationship of postorbital to
the lateral edge of the parietal is obscured by sediment.
The lacrimals are elongate bones which seem to form the anterior portions of the orbital margin and extend to
the external nares.
The left and right jugals can be seen directly in the specimen, although the left jugal is the better exposed along
the oblique lateral fracture surface (text-fig. 4a). The indefinite suture indicated between jugal and quadratojugal
represents a conspicuous separation between two bony elements in the specimen, but which cannot be
interpreted confidently as either a suture or fracture. The dorsal surfaces of these two bony elements appear in
text-fig. 3 as one distinct dark element. The orbital margin of the jugal is visible in X-ray as well. This portion of
the orbital margin is completed dorsally by a large anterodorsal process of the jugal.
The quadratojugal is an elongate bone which bears a long suture with the maxilla and a comparatively brief
one with the postorbital. The posterior portion of the left quadratojugal appears to have been lost through
breakage. However, some indication of this portion of the right quadratojugal is supplied by the lateral margin
of the subtemporal fossa (text-figs. 3 and 5b), which is interpreted as quadratojugal. The manner of attachment
of the quadratojugal to the suspensorium is uncertain.
An element appearing in the X-ray at the anterior extremity of the specimen to the right of the midline is
interpreted as the right premaxilla. The element has one process directed laterally toward the right maxilla and a
second directed posteriorly toward the right frontal bone. No teeth can be distinguished, however.
Both left and right maxillae appear in X-ray. They are long slim bones which extend well posterior to the orbits
and form a portion of the narial margin. The right maxilla is essentially in place and is cracked along its orbital
margin. It is expanded anteriorly into a broad process. The indentification of this process is uncertain, for
depending upon the amount of distortion in this region of the skull, it may be either a nasal process or palatal
process of the maxilla. There are 11 teeth in the left maxilla and 18 in the right.
A limited portion of the left squamosal (text-figs. 1 and 4a) is exposed. It is a rounded, rectangular bone
and is somewhat displaced to exhibit a portion of its sutural contact with the quadrate ramus of the
pterygoid.
WELLSTEAD: LOWER CARBONIFEROUS AMPHIBIAN
199
C
2 mm
text-fig. 4. Lethiscus stocki, MCZ2185. A, lateral view of skull. Composite of information from
left and right sides of skull, x3. b, left lateral view of vertebra seven. Composite, x 5. c, ventral
view of vertebra fifteen, x 5. d, left lateral views of vertebrae 57 and 58, x 5. e, notched
sarcopterygian scale and associated elongate element, x 5.
Temporal fenestrae
The temporal fenestrae (text-figs. 4a and 6a) are bounded dorsally by the parietal, supratemporal, and tabular
bones, anteriorly by the postorbital, laterally by the quadratojugal, and medially by the squamosal and by the
ascending flange of the quadrate ramus of the pterygoid. The fenestrae continue to the posterior margin of the
skull.
Palate
Elements of the palate (text-fig. 5b) are interpreted entirely from the X-rays.
Two elements, which appear in tomographs of ventral portions of the skull, may be vomers judging from their
relatively anterior position. Their relationships to each other and to surrounding bones are not known. The right
vomer appears to bear three teeth comparable in cross-sectional area to maxillary teeth. Nothing of the palatine
bones can be distinguished in the X-rays.
Portions of the right pterygoid and ectopterygoid are seen in text-fig. 5. The pterygoid bears a high dorsal
flange (text-fig. 4a) which extends from the quadrate ramus toward the supratemporal and posteriorly contacts
the squamosal. The portion of the right pterygoid identified in text-fig. 1b is probably a fragment of this flange.
The anterior portions of the pterygoids cannot be discerned in the X-ray and may be missing. Sutural contact
between the pterygoid and supratemporal is obscured by matrix and bone loss along the fracture surface of the
nodule.
200
PALAEONTOLOGY, VOLUME 25
The subtemporal fossa is defined by the pterygoid, ectopterygoid and quadratojugal. The fossa is small and is
constricted by a blunt portion of the quadratojugal. Its small size and irregular lateral margin suggest that this
region has been distorted.
The parasphenoid appears clearly in X-ray. It is displaced to the right, partially overlying the medial edge of
the right pterygoid. The cultriform process is long and has a narrow base. The basicranial processes can be
confidently interpreted on either side of the base of the cultriform process of the parasphenoid. The posterior
portions of the parasphenoid cannot be distinguished from the basisphenoid.
text-fig. 5. Lethiscus stocki, MCZ 2185. a, mandibles interpreted from X-ray and viewed dorsally,
x 3. B, portions of palate and braincase interpreted from X-ray and viewed dorsally, x 3.
Braincase
There is no evidence of braincase contact with the dermal roofing bones, suggesting that the dorsal portions of
the braincase were not ossified.
The oblong. X-ray opaque structure (text-fig. 3) centrally placed in the braincase is taken to be the matrix-filled
brain cavity. It is distorted to the right as is the entire braincase-parasphenoid unit.
A suture distinguishes the basisphenoid region of the braincase from the basioccipital. The structures lateral to
the brain cavity are assumed to be the otic capsules, although pro- and opisthotic bones cannot be individually
distinguished nor are the otic regions clearly delimited from the occipital elements.
A tubular bone is exposed in the calcite filling of the inter-orbital space (text-fig. 4a). It tilts posteriorly
as it rises toward the skull roof and is broken at its ventral extremity. From its structure the bone appears to
be the left epipterygoid, although displaced anteriorly from its normal position dorsal to the basicranial
articulation.
WELLSTEAD: LOWER CARBONIFEROUS AMPHIBIAN
201
text-fig. 6. Comparison of skulls, a, reconstruction of Lethiscus stocki, MCZ 2185, x3.
B, Ophiderpeton (Thomson and Bossy 1970), drawn to length of Lethiscus skull.
Mandibles
With the exception of tooth alveoli revealed in a fracture section of the left mandible, the lower jaws are visible
only as outlines in X-ray. The left mandible lies with its medial surface turned dorsally. The right mandible has its
lateral surface upward. In the posterior margin of each mandible, U-shaped structures corresponding to the
positions of surangular, articular, and angular can be differentiated, but no sutures can be discerned between
these elements.
An elongate, rectangular element lies across the right mandible approximately one-third of its length from the
symphysis. It bears a pointed process and is tentatively identified as a coronoid, probably of the right mandible.
Fourteen teeth are apparent in the left mandible. Nine can be counted in the right. The teeth of the mandibles
seem to be short and peg-like as are those of the maxillae, though the nature of their crowns is not certain.
Hyoid elements
Two elongate elements revealed in the X-ray (text-fig. 3; also 1b) are hyoid elements, perhaps epibranchials.
Fragments lateral to vertebrae 4 and 5 (text-fig. 7a) may be additional hyoid elements or possibly ribs, but no
positive identification can be made. There is no evidence of gill rakers, internal or external gills, nor of the sickle-
shaped hyoid noted in other ai'stopods (Baird 1964).
Vertebrae
Seventy-eight vertebrae are visible in sequence (text-figs. 7, 8, 9). An additional vertebra can be distinguished in
X-rays dorsal to vertebra 65. Vertebrae 1 through 5, 9 through 12, 15 through 28, and 42 through 46 are viewed
ventrally. Alternating with these series are vertebrae exposed in lateral view and vertebrae 49, 50, and 51 which
are seen end-on.
The vertebrae are clearly holospondylous. Bony elements seen between centra of the sixth, seventh, and eighth
vertebrae are found nowhere else in the column and may be mineralized intercentral cartilages similar to those
described in salamanders (Wake 1970; Wake and Lawson 1973) or merely displaced fragments of the adjacent
centra.
202
PALAEONTOLOGY, VOLUME 25
The centra are hour-glass-shaped and deeply amphicoelous. Sections along the column show the notochord to
be severely constricted and possibly discontinuous at midcentrum. The lateral and ventral surfaces of the centra
are smooth except for slightly concave, round facets near the anterior rim which are the articular surfaces for the
rib capitulum (text-fig. 4b). The centra bear no pits, grooves, or accessory processes.
Vertebral length increases antero-posteriorly along most of the column. The average length for vertebrae 6, 7,
and 8 is 4-5 mm. Vertebrae 34, 35, and 37 average 5-2 mm in length, while vertebrae 52 through 78 average
approximately 6 mm in length. The untapered nature of these vertebrae, as well as the presence of ribs along the
column, suggests that this portion of the skeleton represents the trunk of Lethiscus. The isolated vertebra is only
4 mm long and, by virtue of its small size, is the only indication of a tail in the specimen.
Neural arches of the vertebrae are swollen, unpaired, and are fused to their centra with no trace of suture The
pedicel length is approximately two-thirds that of the centrum. The neural spines anterior to vertebra 37 are not
well exposed. However, a transverse section of vertebra 9 (text-fig. 10a) shows the spine to be relatively high.
More posteriorly, the neural arches can be seen to extend the length of the neural arch and to be tall, rising
antero-posteriorly (text-fig. 9a and b). The spines bear jagged edges and have faint grooves between the teeth of
the serrations, suggesting a crinkled appearance. It is not clear whether the jagged appearance is natural or due to
poor ossification or breakage. Transverse sections through the vertebral column show that the neural spines
bifurcate at their posterior extremities and bear a deep medial groove (text-fig. 10 c and d).
Neural-arch processes bearing zygapophyses project at approximately 30° from the sagittal plane, but extend
little laterally beyond the centrum (text-fig. 10c and d). Zygapophyses are oblique to the sagittal plane.
All post-atlantal vertebrae bear stout transverse processes which, in the more anterior vertebrae, are located
on the neural arches (text-figs. 7a and 10a), rather than on the centra as in other a'fstopods. The processes project
laterally, but breakage obscures the surface of rib articulation. In contrast to the transverse processes of the
anterior vertebrae, the transverse processes of vertebrae 44, 57, and 68 through 78 can be seen to arise, in part,
from the centra (text-figs. 9 and 10b).
text-fig. 7. Lethiscus stocki , MCZ 2185, postcranial skeleton, x 1-4. a, vertebrae l through 15.
B, vertebrae 15 through 30.
4U\\
204
PALAEONTOLOGY, VOLUME 25
Where the lateral surfaces of the posterior vertebrae are exposed, as in vertebrae 57, 58, and 68, a
fossa is seen immediately posterior to the transverse processes (text-fig. 9a). In the floor of this fossa
are foramina, presumably for exit of spinal nerves. Such foramina cannot be identified in vertebra 7,
the only anterior vertebra suitably exposed.
Ribs
Fragments of ribs are exposed along the length of the vertebral column, but are best revealed posterior to
vertebra 36. Here, ribs are seen to have a dog-legged bend just posterior to the rib head. Distal to this bend the
ribs are straight. An accessory rib process is suggested by the sharp angle of the bend, but no substantial evidence
of the characteristic K-shaped ai'stopod rib is exhibited in the specimen. The ribs are bicipital anterior to vertebra
44 (text-fig. 1 0 a and b), but the nature of their vertebral articulation is not exposed more posteriorly. The rib head
is set at a sharp angle to its body as can be seen in the vicinity of vertebrae 19, 20, 43, and 60.
text-fig. 10. Selected sections of vertebrae, x 5. Roman numerals indicate approximate position of section in
reference vertebra, a, vertebra 9. b, vertebra 44. c, vertebra 15 and prezygapophyses of vertebra 16.
D, vertebra 23.
Ventral armour
Numerous, spindle-shaped dermal elements comprise the ventral armour, which would seem to have been
continuous posterior to the level of the sixth vertebra. The pattern of the ventral armour has been severely
disrupted. The ventral armour elements in Lethiscus appear to be stouter and to have blunter ends than those in
Ophiderpeton , but this appearance may be due to the irregular sections in which the elements are exposed.
Dorsal osteoderms
Baird (1964) noted dorsal osteoderms in this specimen, but the only elements which might be interpreted as
osteoderms are actually mineralized gas bubbles (text-fig. 8a). Sections of these elements show them not to be
bone, but calcite cores coated with pyrite. Gas bubbles preserved in this manner have been reported in specimens
from the Wardie Shales previously (Wood 1977).
Enterospira
In the vicinity of the forty-first vertebra is what appears to be a coprolite. Its segmented appearance is likely due
to a spiral valve in the intestine of Lethiscus. As the coprolite has not been excreted, the term enterospira may be
more correct (Williams 1972).
WELLSTEAD: LOWER CARBONIFEROUS AMPHIBIAN
205
Sarcopterygian scales and possible pectoral girdle
Dorsal to vertebrae 6, 7, and 8 are many bony elements. Most of these are the spindle-shaped armour seen
elsewhere in the specimen. Four others are probably slim fragments of rib. Most interesting are the three largest
elongate bones and three irregularly round elements near them.
The round elements bear faint concentric rings and compare favourably with scales of sarcopterygian fishes. It
should be noted, however, that concentric growth structures similar to these rings occur in endochondral limb
and girdle elements of tetrapods as well ( Mesosaurus , de Ricqles 1974). The middle ‘scale’ has a notch in its rim
which is similar to a glenoid fossa (text-fig. 4e). In the notch is a tiny fragment of bone, but whether the fragment
has actually come to rest in a notch-like glenoid or whether compressional forces merely created the notch by
forcing the fragment into the rim is uncertain.
The three large elongate elements could be posteriorly displaced hyoid elements or possibly remnants of the
dermal pectoral girdle.
Although these six bones occur where a pectoral girdle would be expected, their lamentably poor exposure
allows no confident identification. There is no other evidence of girdles or limbs in the specimen.
COMPARISONS
Skull
The skulls of both Lethiscus and Ophiderpeton (text-fig. 6) are relatively high-sided and possess
temporal fenestrae bordered by identical elements of the skull roof. The orbits are anterior in
position. Maxillae are elongate, slim, and extend far posterior to the orbit. The frontal bones have a
similar proportional length relationship with the orbits. In Lethiscus, Ophiderpeton, and nectrideans
(Thomson and Bossy 1970) the skull table is comprised primarily by the parietal and postparietal
bones. Lethiscus, however, appears to represent a more primitive pattern in that its parietal-
postparietal suture is anterior to the supratemporal-tabular suture (Panchen 1970). As a result,
Lethiscus lacks the tabular-parietal contact, the presence of which has been suggested as linking
a'istopods and nectrideans (Thomson and Bossy 1 970). The later establishment of the tabular parietal
contact in Ophiderpeton may then have come as the result of the posterior movement of the parietal-
postparietal suture as suggested in anthracosaurs (Panchen 1970). Concomitant with the movement
of this suture may have been the proportional increase in length of the parietal bones seen in
Ophiderpeton.
In contrast to Ophiderpeton the supratemporals of Lethiscus are large and the skull table is much
shorter. There is no indication of the intertemporal in Lethiscus. This bone may have been
incorporated into the large postorbital. However, elongate pustulated bones have been identified as
intertemporals in Ophiderpeton (Thomson and Bossy 1970).
In contrast to Ophiderpeton the jugal and quadratojugal in Lethiscus contact one another. The
quadratojugal in Lethiscus also bears contacts with the postorbital and maxilla, perhaps
strengthening the connection of the lateral skull margin with the skull table in response to bone loss
resulting from fenestration of the skull roof. In Lethiscus the maxilla contacts the narial margin, but is
excluded from it by the lacrimal in Ophiderpeton as reconstructed by Thomson and Bossy.
Other possible differences, such as the presence of nasals and the extent of the lacrimals and
prefrontals cannot be determined unequivocally. The mandibles in Lethiscus and the palate and
braincase in both Lethiscus and Ophiderpeton are also too poorly known to allow useful comparison.
Vertebrae
The vertebrae of Lethiscus differ from those of other a’istopods in lacking basipophyseal accessory
articulations and median ventral ridges (as seen in Ophiderpeton nanum, Steen 1931) and in possessing
high neural spines. Absence of incontestable limb girdles makes distinction of trunk and caudal
regions difficult. Lethiscus does, however, exhibit antero-posterior differentiation of the vertebral
column in the position of transverse processes and by the possession of spinal-nerve foramina only in
the posterior portion of the column. Similar differentiation has not been found in Ophiderpeton,
although McGinnis (1967) reported that the two anterior-most vertebrae in Phlegethontia do lack
spinal-nerve foramina.
206
PALAEONTOLOGY, VOLUME 25
The anterior vertebrae of Lethiscus are remarkable in their similarity to those of some microsaurs,
as well as to those of early reptiles. These in each case have smooth, unpitted surfaces, and are hour-
glass-shaped. The neural-arch pedicels also bear the transverse processes. In contrast, neuro-central
sutures are present consistently within the microbrachiomorph microsaurs and variably within the
tuditanomorphs, but are absent in Lethiscus. Lethiscus also lacks trunk intercentra, which occur in
several microsaur genera.
The vertebrae of adelogyrinids and ‘lysorophids’ (including both the Molgophidae and
Lysorophidae) differ from those of Lethiscus in the consistent presence of neuro-central sutures, but
are similar in lacking accessory articulations and in bearing the transverse processes on neural arches.
The lysorophids further differ in the paired nature of their neural arches. The paired status of neural
arches in the adelogyrinids is equivocal (Carrol 1 967; Brough and Brough 1 967; Watson 1921-1 923).
The trunk vertebrae of nectrideans are similar to those of Lethiscus in lacking intercentral elements
and in ossifying as single units. However, the only described nectridean in which intravertebral spinal
nerve foramina can actually be seen is the urocordylid Crossotelos, keraterpetontids do not possess
such spinal nerve foramina (Milner, A. C., pers. comm.). Furthermore, nectridean neural spines are
specialized in their possession of accessory articulations, and rugose ornamentation (as in
Diploceraspis and Diplocaulus), or in being flat-topped, fan-shaped structures with crenulated edges
(as in Sauropleura and Keraterpeton , Baird, 1965; Steen 1938). Some faint suggestion of neural-spine
crenulation is present in Lethiscus , but, as noted, the neural spines are otherwise serrated and inclined
antero-posteriorly to the frontal plane.
The high number of trunk vertebrae found in Lethiscus is seen elsewhere only in a'fstopods and in
the tiny-limbed Lysorophus (Olson 1971) among the iepospondyl’ amphibians. Nectrideans, in
contrast, characteristically have short trunks (and long tails).
Limbs and girdles
Baird (1964) wrote that there is nothing interpretable as pectoral or pelvic girdle in any aistopod.
Although there is no contradictory evidence in Lethiscus, a recent study (see Boyd, M. J. F., this
volume) has discovered an interclavicle in Ophiderpeton nanum.
Ventral armour
Spindle-shaped ventral armour like that in Lethiscus is seen in other ai'stopods and nectrideans
(Fritsch 1879; Huxley 1867) and with sculptured surfaces in the adelogyrinid Adelospondylus (Carroll
1967). Ventral armour in the microsaurs is variable (Carroll and Gaskill 1978), but never consists of
spindle-shaped elements. Such dermal armour is unknown in the lysorophids.
DISCUSSION
Lethiscus possesses nearly all the aistopod characteristics compiled by Baird (1964). Aistopod
character states which can not be confidently identified in Lethiscus (e.g. hypapophyseal flanges of
caudal vertebrae, K-shaped ribs, and sickle-shaped hyoid) are those in portions of the specimen not
preserved or which are poorly exposed.
While Lethiscus is more closely comparable to Ophiderpeton than to Phlegethontia in the relatively j
primitive nature of its skull, robust ribs, and heavy ventral armour, it is distinct in further details of its
skull and post-cranial anatomy. Its short skull and the absence of a tabular-parietal contact indicate i
a less derived state than that of Ophiderpeton , while the absence of accessory vertebral processes and
K-shaped ribs and the presence of tall neural spines indicate that Lethiscus possessed a differently
specialized post-cranial skeleton and trunk musculature than either the Ophiderpetontidae or the j
Phlegethontiidae.
The presence of spinal-nerve foramina in a portion of the vertebral column of Lethiscus is j
especially interesting, for such foramina are known to occur only in urodelan lissamphibians, the j
A'istopoda, and the nectridean Crossotelos. The presence of these foramina is considered to be a
derived state in salamanders (Edwards 1 976; Hecht and Edwards 1977) and would appear to be so in
WELLSTEAD: LOWER CARBONIFEROUS AMPHIBIAN
207
aistopods and Crossotelos. Particularly significant is the observation that the spinal-nerve foramina
of salamanders are expressed in patterns characteristic of the various families (Edwards 1976).
Although the pattern of spinal-nerve foramina is not well known in the vertebral column of
Ophiderpeton, the contrast in patterns between Phlegethontia and Lethiscus suggests that the spinal-
nerve foramina pattern may allow distinction of aistopod families also.
The occurrence of an aistopod in mid-Visean rocks provides some limited confirmation of the
estimated several million or tens of millions of years required to accomplish limb loss in tetrapods
(Lande 1977). Although Lethiscus cannot be described with absolute certainty as lacking limbs or
girdles, the rudiments of a possible pectoral girdle demonstrate the degenerate nature of any limbs it
may have possessed. Assuming tetrapod monophyly, this limb reduction was achieved within a
period of at least 30 million to 40 million years elapsing between Late Devonian tetrapod origins,
represented by Metaxygnathus (Campbell and Bell 1977) and Ichthyostega, and the occurrence of
Lethiscus in the mid-Visean.
Lethiscus is the earliest known member of a group of small Palaeozoic amphibians known as
‘lepospondyls’, but unfortunately it reveals little about the evolution of any of these animals or of
tetrapods in general because of the specializations of the skull and post-cranial anatomy already
attained in the Lower Carboniferous. Lethiscus seems to confuse the issue somewhat, for, although
Thomson and Bossy (1970) linked nectrideans and aistopods through the shared possession of the
tabular-parietal contact, the absence of such a contact in Lethiscus suggests that it was either achieved
independently in the two orders or was lost in Lethiscus subsequent to a nectridean-ai'stopod
dichotomy. Similarly, the extremely limited occurrence of intravertebral spinal-nerve foramina in
nectrideans suggests that these foramina were probably developed separately in nectrideans and
aistopods, arguing against consideration of the presence of the foramina as a shared derived-
character state.
Acknowledgements. I am pleased to acknowledge Dr. Robert L. Carroll for giving me the opportunity to study
this specimen. This report benefitted critically from comments received from Dr. Carroll, Dr. Donald Baird, Dr.
A. R. Milner, Dr. A. C. Milner, Mr. Timothy Smithson, and Mr. Robert Holmes, to all of whom I am grateful.
X-ray photography has been indispensible to the description of portions of this specimen. I thank Dr. Robert
Hanson and Sandra and Andre Hamelin of the Radiology Unit of the Royal Victoria Hospital, Montreal, for
their kind and patient assistance.
REFERENCES
baird, d. 1955. Latex micro-molding and latex-plaster molding mixture. Science, 122, 202.
1964. The aistopod amphibians surveyed. Breviora, No. 206, 1-17.
1965. Paleozoic lepospondyl amphibians. Am. Zool. 5, 287-294.
brough, m. c. and brough, j. 1967. Studies of early tetrapods. I. The Lower Carboniferous microsaurs. II.
Microbrachis, the type microsaur. III. The genus Gephyrostegus. Phil. Trans. R. Soc. 252, 107-165.
Campbell, k. s. w. and bell, m. w. 1977. A primitive amphibian from the late Devonian of New South Wales.
Alcheringa, 1, 369-381.
carroll, r. l. 1967. An adelogyrinid lepospondyl amphibian from the Upper Carboniferous. Can. J. Zool. 45,
1-16.
— 1977. Patterns of amphibian evolution: an extended example of the incompleteness of the fossil record. In
hallam, a. (ed.). Patterns of evolution as illustrated by the fossil record. Elsevier Sci. Publ. Co., Amsterdam.
591 pp.
— and gaskill, p. 1978. The order Microsauria. Mem. Am. Phil. Soc. 126, 211 pp.
edwards, j. l. 1976. Spinal nerves and their bearing on salamander phylogeny. J. Morph. 148, 305-328.
fitch, f. j., miller, j. a. and williams, s. c. 1970. Isotopic ages of British Carboniferous rocks. In Sixieme
Congress International de Stratigraphie et de Geologie du Carbonifere, Sheffield, 1967. 2, 771-789.
fritsch, A. 1 879. Fauna der Gaskohle und der Kalksteine der Permformation Bohmens. 1 . Prague. 1 82 pp., 48 pis.
GEORGE, T. N., JOHNSON, G. A. L., MITCHELL, M., PRENTICE, J. E., RAMSBOTTOM, W. H. C., SEVASTOPULO, G. D.,
wilson, r. b. 1976. A correlation of Dinantion rocks in the British Isles. Spec. Rept. 7, Geol. Soc. Lond. 87 pp.
208
PALAEONTOLOGY, VOLUME 25
hecht, m. k. and Edwards, J. l. 1977. The methodology of phylogenetic inference above the species level. In
hecht, m. k., goody, p. c. and hecht, B. M. (eds.). Major patterns in vertebrate evolution. Plenum, New York,
pp. 3-51.
huxley, t. h. 1 867. Description of vertebrate remains from Jarrow colliery. Pt. 1 . Trans. R. Ir. Acad. 24 (Science),
353-369.
lande, r. 1977. Evolutionary mechanisms of limb loss in tetrapods. Evolution, 32, 73-92.
mcginnis, h. j. 1967. The osteology of Phlegethontia, a Carboniferous and Permian ai'stopod amphibian. Univ.
Calif. Pubis, geol. Sci. 71, 1-46.
mitchell, f. h. and mykura, w. 1962. The geology of the neighbourhood of Edinburgh. Mem. geol. Surv. U.K.,
3rd edn., 159 pp.
olson, e. c. 1971. A skeleton of Lysorophus tricarinatus (Amphibia: Lepospondyli) from the Hennessey
Formation (Permian) of Oklahoma. J. Paleont. 45, 443-449.
— 1972. Diplocaulus parvus n.sp. (Amphibia: Nectridea) from the Chikasha formation (Permian:
Guadalupian) of Oklahoma. Ibid. 46, 656-659.
panchen, a. l. 1970. Batrachosauria: Anthracosauria. Handb. Palaoherp. 5/A, 84 pp.
ricqles, A. j. de 1974. Recherches paleohistologiques sur les os longs des tetrapodes. 5. Cotylosaurs et
Mesosaures. Annls. Paleont. Vertebres, 60, 13-48, 7 pis.
steen, m. c. 1931. The British Museum collection of Amphibia from the Middle Coal Measures of Linton, Ohio.
Proc. cool. Soc. Loud. (1930), 849-891.
1 938. On the fossil Amphibia from the Gas Coal of Nyrany and other deposits in Czechoslovakia. Ibid. B,
108, 205-283.
stock, t. 1882. Notice of some discoveries recently made in Carboniferous vertebrate paleontology. Nature, 27,
22.
Thomson, k. s. and bossy, k. h. 1970. Adaptive trends and relationships in early Amphibia. Forma et Functio, 3,
7-31.
wake, d. b. 1970. Aspects of vertebral evolution in the modern Amphibia. Ibid. 33-60.
— and lawson, r. 1973. Development and adult morphology of the vertebral column in the plethodontid
salamander Euryces bislineata, with comments on vertebral evolution in the Amphibia. J. Morph. 139,
251-300.
watson, D. m. s. 1921-23. The Carboniferous Amphibia of Scotland. Palaeont. hung. 1, 221-251.
williams, m. e. 1972. The origin of spiral coprolites. Paleont. Contr. Univ. Kans. 59, 1-19.
wood, s. p. 1977. Recent discoveries of Carboniferous fishes in Edinburgh. Scott. J. Geol. 2, 251-258.
CARL F. WELLSTEAD
Redpath Museum
McGill University
859 Sherbrook Street West
Montreal, P.Q., Canada H3A 2K6
Typescript received 8 July 1980
Revised typescript received 24 October 1980
MORPHOLOGY AND RELATIONSHIPS OF THE
UPPER CARBONIFEROUS AISTOPOD AMPHIBIAN
OPHIDERPETON NANUM
by M. J. BOYD
Abstract. The holotype and only recorded specimen of the Carboniferous ai'stopod amphibian Ophiderpeton
nanum Hancock and Atthey 1868 is described in detail and figured for the first time. The vertebrae, ribs,
dermal squamation, and premaxilla are characteristic of Ophiderpeton and confirm that O. nanum is a member
of that genus. The relationship of O. nanum to other described Ophiderpeton species is obscured by the absence
of most of the skull in the holotype and by the apparently sub-adult nature of the specimen. The ventral
osteoderms are, however, unusually filamentous for Ophiderpeton and it is suggested that O. nanum be retained
as a distinct species, pending revision of the Ophiderpetontidae. A small isolated bone in the holotype may
be an interclavicle, suggesting the retention of a vestigial pectoral girdle in ophiderpetontid a'istopods.
Ophiderpeton nanum Hancock and Atthey 1868, is a small ‘lepospondyl’ amphibian of the
order Aistopoda from the Upper Carboniferous of Great Britain. The holotype and only described
specimen of Ophiderpeton nanum was collected around the middle of the nineteenth century and
is from the black shale immediately overlying the Low Main coal seam at Newsham, near Blyth,
in Northumberland. This horizon lies within the Upper Modiolaris zone of the Middle Coal
Measures (Land 1974) and is Westphalian B in age. O. nanum forms part of a large and well-known
amphibian assemblage from Newsham, whose other members (listed by Land 1974, p. 61) include
eogyrinid embolomeres, loxommatid temnospondyls, and keraterpetontid nectrideans. The above
taxa appear to be characteristic of Upper Carboniferous coal-swamp lake environments (Milner
1978).
Although O. nanum was one of the first a'istopods to be described, no detailed account of the
structure or relationships of this species has hitherto been published and the holotype specimen
has never been figured. Neither the original, necessarily superficial, description of O. nanum by
Hancock and Atthey (1868) nor a subsequent brief account given by Steen (1938), are detailed
enough to allow adequate comparison of this form with other described a'istopods. The present
paper is intended to remedy this deficiency.
The holotype of O. nanum consists of an incomplete skull and anterior postcranial skeleton
preserved in counterpart on two small slabs of shale. The two halves of the specimen are registered
in the collections of the Hancock Museum, Newcastle upon Tyne, as G25.34 and G25.35. Because
of the small size and fragile nature of the holotype, the only preparation attempted has been the
removal of small quantities of matrix from the immediate vicinity of the skull by means of mounted
needles.
DESCRIPTION
The skull and most of the preserved postcranial skeleton are situated on slab G25.34 (text-fig. la). Slab G25.35,
henceforth referred to as the counterpart, bears impressions of all but the most anterior seven vertebrae of
G25.34, in addition to a small number of actual ribs and isolated patches of the ventral squamation. At some
point in the past, small areas of the surface of slab G25.34, bearing sections of vertebral column and associated
structures, have flaked away and been lost. However, as the resultant gaps in the vertebral series are all
posterior to the seventh vertebra, the number of vertebrae originally preserved may be estimated by reference
to the unbroken series of impressions on the counterpart. The skull and skeleton of G25.34 are preserved
IPalaeontology, Vol. 25, Part 1, 1982, pp. 209-214.]
text-fig. 1. Ophiderpeton nanum H & A. a-e, Hancock Museum specimen G25.34: a, semi-diagrammatic
representation of specimen as preserved. Stippling indicates extent of ventral squamation; b, isolated (?)skull
element. Stippling represents area preserved as impression only; c , interclavicle as preserved; d, interclavicle
restored; e, detail of ventral squamation. /, Hancock Museum G25.35. Three isolated vertebrae as preserved.
All scale lines represent 5 mm. n.sp., neural spine.
BOYD: CARBONIFEROUS AMPHIBIAN OPHIDERPETON
2:
with their ventral surfaces uppermost. A large number of ribs is present. Although most of the vertebrae and
ribs are overlain by a thin ‘mat’ of closely packed, filamentous gastralia, this is often so closely applied that
the contours of the structures which it conceals may be clearly seen.
Skull
The skull apparently became disarticulated and scattered prior to preservation and is largely absent. Three
distinct masses of poorly preserved bone lie immediately anterior to the first preserved vertebra. It is possible
that they represent elements of the palate or neurocranium. A slight depression in the matrix anterior to these
may mark the site of the isolated premaxilla noted by Steen (1938, p. 223). No trace of the bone itself remains;
it has presumably become detached from the slab and lost.
A relatively large isolated bone, situated to the right of the most anterior four vertebrae in text-figure 1 a,
may, doubtfully, also belong to the dermal skull. Although much of the bone itself has been lost, the clear
impression remaining in the matrix allows no question as to its original form. Its identity, however, is obscure.
The element (text-fig. 1 b) does not appear to correspond to any bone known in the skulls of either Ophiderpeton
(A. C. Milner, pers. comm.) or phlegethontiid a'fstopods (e.g. Gregory 1948; McGinnis 1967). The triradiate
form of the element makes it unlikely that it represents part of the dermal pectoral girdle. Possibly the bone
in question does not, in fact, pertain to O. nanum.
A second isolated element, lying to the left of the vertebral column at the level of the third and fourth
vertebrae (text-fig. la), is of some interest. Approximately one-quarter of the bone is missing but it is clear
that the complete element was of roughly rhomboidal shape (text-fig. 1 c-d). It is suggested that this element
represents a displaced interclavicle. This hypothesis receives support from the presence of a slight (?parasternal)
process at one angle of the bone and the fact that the margin of the element directly opposite the above
process is noticeably fimbriated. Interclavicles with fimbriated anterior margins have been described in a
number of both ‘labyrinthodont’ (e.g. Milner 1980, fig. 5) and ‘lepospondyf (e.g. Carroll and Gaskill 1978,
fig. 120g) amphibians. No clearly defined ornament or areas for clavicular overlap are visible on the suggested
interclavicle but it is possibly preserved with its dorsal surface uppermost.
Vertebrae
Although several vertebrae have been lost from the articulated series of G25.34 (text-fig. la), it is apparent
from the impressions on the counterpart that the first forty-three vertebrae were originally present. The
presence of a single median ventral ridge in all the vertebrae preserved indicates that all are precaudal; the
centra of ai'stopod caudal vertebrae are characterized by a pair of hypapophyseal flanges demarcating the
haemal canal (e.g. Zidek and Baird 1978). The vertebrae exhibit a gradual increase in size from anterior to
posterior of the series, the most anterior centra measuring 2 mm in length and the most posterior approximately
3 mm. Well-developed parapophyses are borne by all but the most anterior two vertebrae, in which they are
represented only by scarcely perceptible lateral projections from the centrum. In vertebrae 3-10 the
parapophyses, as preserved, have a posterolateral orientation. In the remaining vertebrae (1 1-43) they project
laterally. In the absence of any indication of the original position of the pectoral girdle it is impossible to
determine how many of the vertebrae are cervicals. However, Baird (1964) has noted that the parapophyses
of ai'stopod dorsal vertebrae change their orientation from posterolateral to lateral to anterolateral, from
anterior to posterior of the column. It may therefore be concluded that, with the exception of an unknown
number of cervicals, the vertebrae of G25.34 are all anterior- or mid-dorsal in position. The vertebral count
in an entire specimen of O. nanum is unknown. However, a figure of 100+ has been given by Baird (1964,
p. 6) for a juvenile of O. granulosum Fritsch.
The structure of a typical mid-dorsal vertebra of O. nanum will be apparent from text-fig. 2b- d. The centrum
is holospondylous and probably deeply amphicoelous. The latter is the ai'stopod condition and is suggested
in G25.34 by the high degree of compression undergone by most centra. Ventrally, the centrum bears a
median ridge running the length of the element and broadening at its anterior and posterior ends. The ventral
ridge, which appears to have been mistaken for a low neural spine by Hancock and Atthey (1868, p. 277),
is flanked by a pair of elongate depressions. The parapophyses are dorsoventrally compressed structures,
slightly expanded distally, and have their origins low on the lateral surface of the centrum at approximately
the mid-point of its length. None of the vertebrae shows any evidence of the presence of basapophyseal
processes.
The vertebrae of G25.34 are preserved with only their ventral surfaces exposed and yield no information
on the structure of the neural arch. However, the counterpart slab bears, in addition to impressions of most
of the vertebrae of G25.34, three isolated vertebrae from a more posterior region of the trunk (text-fig. 1/).
Each of the three is incomplete but together they allow a restoration of most of the neural arch. Two are
212
PALAEONTOLOGY, VOLUME 25
almost entire vertebrae with their dorsal surfaces exposed, and the third is a very clear impression of the
dorsal surface of the neural arch with the neural spine itself remaining in position. The neural arch is a long
and low structure extending the full length of the centrum; whether the two are fused or merely sutured
together cannot be determined. The neural spine is very weakly developed and forms a low and narrow ridge
running the length of the, otherwise almost horizontal, dorsal surface of the neural arch. The zygapophyses
are widely spaced and possess horizontally orientated articular surfaces. Unfortunately, the nature of the
available material makes it impossible to determine whether the neural arch is pierced laterally by foramina
for the spinal nerves, as is known to be the case in Phlegethontia (McGinnis 1967) and at least one Ophiderpeton
species (Baird 1964).
Ribs
Numerous ribs are preserved along the length of vertebral column of G25.34 (text-fig. la). Almost all are
incomplete and most are at least partially obscured by overlying areas of the ventral squamation. However,
a composite restoration of rib structure in O. nanum may be made with confidence. Free ribs appear to have
been borne by all but the most anterior two of the forty-three vertebrae originally present. The first two
vertebrae are also known to be without free ribs in Phlegethontia (McGinnis 1967). There is no apparent
variation in rib form within the preserved series. The ribs do, however, show a slight increase in size from
anterior to posterior.
The structure of a typical precaudal rib of O. nanum is depicted in text-fig. 2a. The rib is tetraradiate, in
the manner characteristic of the Ai'stopoda (Baird 1964). The capitulum is short and stout, and appears to
possess a slightly recessed head for articulation with the parapophysis of the vertebra. In all ribs in which
it is preserved, the costal process is approximately twice the length of the capitulum and is connected
to the latter for most of its length by a thin ‘web’ of bone. Whether the head of the costal process was
recessed, as has been stated to be the case in Ophiderpeton by Baird (1964, p. 7), cannot be determined. The
shaft, which contributes about two-thirds of the over-all length of the rib, is a robust, stiletto-like structure
which tapers distally to a fine point. A well-developed posteromedial process is present. In most of the
preserved ribs the posteromedial process exhibits a slight lateral curvature, towards the medial side of the shaft.
Scales
A well-developed ventral squamation is present and, as preserved, takes the form of a ‘mat’ of closely packed
gastralia extending from the tenth vertebra to the forty-third (text-fig. la). The squamation has been displaced
text-fig. 2. Restoration of mid-dorsal vertebra and rib of Ophiderpeton nanum H & A. a, rib of left side in
dorsal view; b-d, vertebra in b, dorsal view; c, left lateral view and d, ventral view. p-m. proc., posteromedial
process.
BOYD: CARBONIFEROUS AMPHIBIAN OPHIDERPETON
213
to one side of the vertebral column in the anterior part of the trunk, but more posteriorly it overlies the
vertebrae and ribs. The individual gastralia are extremely elongate, almost filamentous, structures (text-fig.
\e). In their form they somewhat resemble the hair-like gastralia of the phlegethontiid Aornerpeton mazonense
(Gregory) (Gregory 1948; Lund 1978), but their tightly packed arrangement and their extent beyond the
anterior thoracic region are characteristic of Ophiderpeton (Baird 1964). As preserved, the orientation of the
gastralia is very variable and Hancock and Atthey (1868, p. 277) suggested that the scales originally lay with
their long axes at 90° to the vertebral column. It would seem more likely, however, they were originally
arranged en chevron, as in many other Palaeozoic amphibians. Such an arrangement is still preserved in some
areas of the ventral squamation. The absence of ventral osteoderms in the ‘cervical’ region and beneath the
most anterior dorsal vertebrae is of some interest and may indicate, as Steen (1938, pp. 223, 224) suggested,
that the specimen is a sub-adult individual.
Although not altogether absent, as had previously been thought (Steen 1938), the dorsal squamation is
represented in G25.34 only by a small number of rounded, pebble-like osteoderms scattered at intervals along
the length of the vertebral column (text-fig. la). The dorsal squamation of Ophiderpeton species usually
consists of numerous such scales covering the dorsal and lateral surfaces of the trunk region and tail (Baird
1964). Its almost complete absence in the holotype of O. nanum may, like the restricted ventral squamation,
indicate that the specimen is a juvenile animal. Such an assumption also offers an explanation of the ready
dissociation of the skull elements subsequent to the death of the animal.
DISCUSSION
Despite the incompleteness of the only available specimen of O. nanum , it is apparent that this
species is a member of the family Ophiderpetontidae, as defined by Baird (1964). Reference to the
Ophiderpetontidae is suggested by the following characters of the holotype:
1. The structure of the ribs, in which the shaft is stout and stiletto-like and a well-developed
posteromedial process present. In the phlegethontiids Phlegethontia and Aornerpeton the shaft is
usually slender and flexible and the posteromedial process weakly developed or absent (Baird
1964; Lund 1978).
2. The presence of a well-developed ventral squamation composed of numerous, closely packed
gastralia forming a continuous plastron. The phlegethontiid ventral squamation consists of a
series of widely spaced, filamentous gastralia restricted to the anterior trunk region (e.g. Gregory
1948, pi. 1, fig. 4).
3. The apparent presence of an, at least partial, dorsal or lateral squamation of pebble-shaped
osteoderms; no dorsal and lateral armour is known to occur in the Phlegethontiidae (Baird 1964).
4. The form of the premaxilla, now lost, described by Steen (1938, p. 223). This was stated exactly
to resemble the premaxilla of O. amphiuminum , described by Steen in 1931 (fig. 17b-c). The
premaxilla of O. amphiuminum and other Ophiderpeton species differs markedly from that of
Phlegethontia in possessing an anterior ascending process (McGinnis 1967, p. 38).
As presently constituted, the Ophiderpetontidae contains only the type genus, Ophiderpeton
Huxley 1867, and Coloraderpeton Vaughn 1969. The latter is a monotypic genus, established by
Vaughn (1969) for a number of vertebrae, ventral osteoderms, and an attributed rib from the
Upper Pennsylvanian of the Sangre de Cristo Formation in central Colorado. The vertebrae of
Coloraderpeton are distinguished from those of both Ophiderpeton and Phlegethontia principally
by the form of the neural spine. This, although relatively somewhat higher than in the last two
genera, is restricted in its anterior-posterior extent to the middle one-third of the neural arch and
possesses a crenulated dorsal edge. The dorsal vertebrae otherwise appear typically a'istopod and
possess foramina for exit of the spinal nerves. Coloraderpeton was placed in the Ophiderpetontidae
by Vaughn on the basis of the form of the rib, which has a stout shaft and apparently also a
posteromedial process, and of the closely packed, fusiform ventral osteoderms. It is clear from the
structure of the vertebrae that the holotype specimen of O. nanum cannot be placed in the genus
Coloraderpeton. On the other hand, the similarity of the ribs, vertebrae, and ventral squamation
to those of other described Ophiderpeton species appears to confirm the correctness of Hancock
and Atthey’s (1868) reference of the Newsham a'istopod to this genus.
214
PALAEONTOLOGY, VOLUME 25
O. nanum was distinguished by Hancock and Atthey (1868) from the only previously described
Ophiderpeton species, O. brownriggii Huxley 1867, by its relatively small size and the almost
filamentous nature of the ventral osteoderms. The former is clearly unreliable as a taxonomic
criterion, especially in view of the probability, noted above, that the O. nanum holotype represents
a sub-adult individual. The significance of the second character is uncertain. None the less, in view
of the almost complete absence of a skull in the holotype specimen and the lack of adequately
detailed descriptions of other Ophiderpeton species, it would seem advisable to retain O. nanum as
a distinct species, at least pending thorough revision of the Ophiderpetontidae.
The possible presence of an interclavicle in O. nanum is a feature of some importance. Steen
(1931, p. 876) identified two bones in a specimen of O. amphiuminum from the Westphalian D of
Linton, Ohio, as clavicle and cleithrum, but these were subsequently reinterpreted by Baird (1964)
as possibly representing hyoid elements. Most recent workers have concurred in regarding described
a'istopod species as lacking both limbs and limb girdles. However, Goin and Goin (1971, p. 67)
have pointed out that forelimbs are not known for most fossil members of the urodele family
Sirenidae, despite their presence in the three extant species of this group. They suggested, therefore,
that the A'istopoda may possibly also have possessed vestigial forelimbs which, together with the
pectoral girdle, were usually lost prior to fossilization. If the element tentatively identified as an
interclavicle in the holotype of O. nanum is correctly so interpreted, its presence at least partially
confirms the above hypothesis.
Acknowledgements. My thanks are due first to Mr. A. M. Tynan, Curator of the Hancock Museum, for
permission to describe the specimens in his care. I also wish to thank Miss Susan Turner (Hancock Museum),
Dr. A. L. Panchen (University of Newcastle upon Tyne), and Dr. A. C. Milner (British Museum (Natural
History)) for helpful comments and discussion.
REFERENCES
baird, d. 1964. The a'istopod amphibians surveyed. Breviora, No. 206, 1-17.
CARROLL, r. l. and GASKILL, p. 1978. The Order Microsauria. Mem. Am. phil. Soc. 126, 1-211.
goin, c. J. and coin, o. b. 1971. Introduction to herpetology , 2nd edn. W. H. Freeman, San Francisco. 353 pp.
Gregory, J. t. 1948. A new limbless vertebrate from the Pennsylvanian of Mazon Creek, Illinois. Am. J. Sci. 246,
636-663.
HANCOCK, A. and ATTHEY, t. 1868. Notes on the remains of some reptiles and fishes from the shales of the
Northumberland coal field. Ann. Mag. nat. Hist. (4), 1, 266-278, 346-378.
HUXLEY, t. h. 1867. In HUXLEY, t. h. and WRIGHT, E. p. On a collection of fossil Vertebrata from the Jarrow
Colliery, County of Kilkenny, Ireland. Trans. R. Ir. Acad. 24, 351-368.
land, d. h. 1974. Geology of the Tynemouth district. Mem. geol. Surv. Gt Br. 15, 61-62.
lund, R. 1978. Anatomy and relationships of the Family Phlegethontiidae (Amphibia, A'istopoda). Ann.
Carnegie Mus. 47 (4), 53-79.
mcginnis, h. j. 1967. The osteology of Phlegethontia, a Carboniferous and Permian a'istopod amphibian. Univ.
Calif. Pubis geol. Sci. 71, 1-46.
MILNER, a. r. 1978. A reappraisal of the early Permian amphibians Memonomenos dyscriton and Cricotillus
brachydens. Palaeontology , 21, 667-686.
— 1980. The temnospondyl amphibian Dendrerpeton from the Upper Carboniferous of Ireland. Ibid. 23,
i25- I41 . r .
steen, m. c. 1931. The British Museum collection of Amphibia from the Middle Coal Measures of Linton, Ohio.
Proc. zool. Soc. Lond. 1930, 849-892. .
1938. On the fossil Amphibia from the Gas Coal of Nyrany and other deposits in Czechoslovakia. Ibid. (B)
108, 205-284.
vaughn, p. p. 1969. Upper Pennsylvanian vertebrates from the Sangre de Cristo Formation of central Colorado.
Contr. Sci. Los Angeles Mus. 164, 1-28.
zidek, j. and baird, d. 1978. Cercariomorphus Cope, 1885, identified as the a’istopod amphibian Ophiderpeton.
J. Paleont. 52, 561-564.
MICHAEL J. BOYD
Department of Natural History
Kingston upon Hull Museum
Typescript received 3 August 1980 Queen victoria Square
Revised typescript received 15 October 1980 Kingston upon Hull, North Humberside
A NEW SPECIES OF THE LYCOPSID PLEUROMEIA
FROM THE EARLY TRIASSIC OF SHANXI,
CHINA, AND ITS ECOLOGY
by wang ziqiang and WANG lixin
Abstract. A new species, Pleuromeia jiaochengensis, is recorded from the early Triassic of Jiaocheng district
in Shanxi (Shansi) Province, China. Its small size, morphological features of the strobilus and sporophylls,
abortive leaves, and the undeveloping rhizophore separate this from all other species. The succulent sporophylls
may be a major area of photosynthesis. Based on lithology and distance from known marine strata, an inland
desert environment is suggested for this new species. Its stratigraphic significance is discussed.
It is well known that Palaeozoic floras flourished during Upper Carboniferous and Permian time,
but suffered from a worldwide arid climate at the end of the Permian. Only a few relics survived
into the early Triassic where they faced severe climatic conditions. One such genus is Pleuromeia
which is regarded as an important early Triassic index fossil, and thought to be a link between
the Palaeozoic Sigillaria and modern Isoetes.
Pleuromeia was first recorded from the Bunter Sandstone in Germany nearly a century ago.
Prior to 1960, apart from a few records from France, Spain, and the eastern U.S.S.R. all the
records were from Germany. In particular, Magdefrau (1930, 1931) provided much information
about the genus from a large number of German specimens.
Since the 1960s new material has been recorded from the early Triassic in the Soviet Union,
Japan, and China. Pleuromeia rossica was described by Neuberg (1960u) from the Russian Platform,
whilst Krassilov and Zakharov (1975) added further information on the genus from material in
the far eastern part of the Soviet Union. Kon’no (1973) described P. hatai from the Scythian in
north-eastern Japan, and in China, Wang, Xie, and Wang (1978) described well-preserved specimens
of P. sternbergi (Munster) Corda and P. rossica Neuberg from the early Triassic of the Qinshui
Basin in Shanxi. The genus Pleuromeia therefore has a widespread distribution. It is worth noting
that from the early Triassic of eastern Australia, Cylostrobus a lycopsid-like strobilus which shows
some similarity to Pleuromeia of the Northern Hemisphere, has recently been regarded by Retallack
(1975) as Pleuromeia longicaulis (Burgess).
After the discovery of P. sternbergi and P. rossica from the early Triassic Heshankou Formation
in Qinshui Basin, Shanxi Province, the present writers found many well-preserved specimens
belonging to a new species of Pleuromeia in the Luijiakou Formation beneath the Heshankou
Formation in Jiaocheng district (text-fig. 1), not far from the famous Xuan-zhong Temple of Shanxi
Province. This paper describes these specimens.
FOSSIL-BEARING STRATA
The Luijiakou Formation in Jiaocheng district is the western extension of the sandstone beds of the
‘Shischienfeng Series’ at West Hill in Taiyuan, and consists mainly of reddish-purple, fine-grained sandstones.
According to data supplied by the Regional Geological Survey of Shanxi Province, its thickness in Jiaocheng
district exceeds 460 m, the Formation being subdivided into three parts. The upper part, consists of 108 m
of grey and reddish-purple fine-grained feldspathic sandstones intercalated with reddish-purple siltstones and
sandy shales. The middle part is composed of 215 m of red, grey to greyish-purple fine-grained feldspathic
sandstones interbedded with reddish-purple silty shales and shales. The shales contain abundant ripple marks.
IPalaeontology, Vol. 25, Part 1, 1982, pp. 215-225, pis. 23-24.|
216
PALAEONTOLOGY, VOLUME 25
sun-cracks, and cross-bedded structures sometimes intercalated with a few lenses of greyish-white sandstones
containing fossil plants and greyish-green shales with fossil estherians. The lower part, consists of 138 m of
reddish-grey and purplish-grey fine-grained, cross-bedded feldspathic sandstones containing abundant
laminations of magnetite, intercalated with lenses of greyish-white feldspathic sandstones. The fossil plants,
which include Pleuromeia, are within the middle part of the Formation approximately 130 m above its base
at four localities (numbered Z01-Z04) near Jaoertou village, in the north-western mountainous area of
text-fig. 1. Map showing region of the fossiliferous localities. Inset shows eastern half of China.
EXPLANATION OF PLATE 23
All figures x 1 unless otherwise stated. All specimens from the Luijiakou Formation, Shanxi Province.
Figs. 1-13. Pleuromeia jiaochengensis sp. nov. 1-2; 6-11. Various strobili; 1, Z01-021; 2, showing the awl-like
leaves on the part of stem beneath the strobilus, Z01-01; 6, Z01-024; 7, Z01-027; 8, Z01-019; 9, Z01-022; 10,
Z01-019; 11, syntype, a strobilus containing megaspores (at a), Z01-061. 3-5, complete young plants; 3,
syntype, Z0 1-020; 4, the same x4, 5, Z0 1-061. 12. Shows a decorated stem, with short surface ridges,
Z0 1-061. 13. Syntype, stem surface showing sparse, faint leaf-scars, Z0 1-206, x2.
PLATE 23
WANG and WANG, Pleuromeia
218
PALAEONTOLOGY, VOLUME 25
Jiaocheng district. At locality Z01, there are many well-preserved specimens of strobili, stems, rhizophores,
and even some complete plants in a lens of greyish-white sandstone, 0-2 m thick and 2 m across. At Z04,
there are many sporophylls, stems, and rhizophores occurring in larger lenses. In addition, at localities Z02
and Z03, there are specimens tentatively included within the genera Crematopteris sp. Phyllotheca sp.
Taeniopteris sp. Neocalamites sp. by the present authors. All specimens illustrated in this paper, are reposited
in the Tianjin Institute of Geology and Mineral Resources, Ministry of Geology.
SYSTEMATIC PALAEOBOTANY
Family pleuromeiaceae
Genus pleuromeia Corda, 1852
Pleuromeia jiaochengensis sp. nov.
Plates 23, 24, text-figs. 2-3
Syntypes. Z0 1-020 (small complete plant); Z0 1-061 (showing strobilus and megaspores; Z0 1-206 (showing
stem surface); Z04-159 (showing rhizophore).
Diagnosis. A lycopsid of small shrub size, probably dioecious, generally 20-30 cm high but may
reach almost 50 cm (see text-fig. 2, the reconstruction of a whole plant). Stem erect, unbranched,
maximum diameter 1-5 cm, terminating in a large strobilus measuring one-quarter to one-third of
the plant height (PI. 23, figs. 3-4). Surface of stem covered with sparse, faint, at times barely visible
leaf-scars, usually smooth, with small obscure pits where the leaves were originally attached.
Decorticated stems with short ridges (PI. 23, fig. 12). Leaf-scar lens-shaped, obscure in outline,
with a central pit (PI. 23, fig. 13). Leaves awl or spine-like, 2-3 mm in length, about 1 cm apart
on the upper portion of the stem (PI. 24, fig. 6), having dehisced from the lower part. Rhizophores
tuberous when mature, covered with oblong to oval, well-separated appendage-scars; the younger
rhizophores only slightly swollen at the base of the stem and covered with rather thick appendages,
maximum length 3 cm, usually on the lower part. Strobilus probably unisexual, narrowly
spike-shaped, maximum length 20 cm. Sporophylls spirally arranged, imbricate, 4-5 in each spiral
(PI. 23, fig. 6), the spiral from 60° to quite an acute angle to the rachilla, giving a total of more
than 100 sporophylls in a mature strobilus. Sporophylls spatulate with a sagittate apex, longer
than wide, ovate to oblong in outline, adaxially concave where the sporangium lies. The border
of the sporophyll is wider anteriorly than laterally, and in mature strobili the apex of the anterior
border of larger sporophylls is slightly reflexed upwards (PI. 24, fig. 3; text-fig. 3d left), and is
probably visible on the outside of the strobilus. Where the upward reflexed part of a sporophyll is
truncated, it clearly shows a retuse, small anterior border (PI. 24, fig. 5). Sporangium discoid, oval or
orbicular in outline, attached to the rachilla only at its proximal end. Surface of sporangium with
many more or less parallel lines (PI. 24, fig. 4) which converge slightly at both ends. Thickness of
EXPLANATION OF PLATE 24
All figures x 1 unless otherwise stated. All specimens from the Luijiakou Formation, Shanxi Province.
Figs. 1-15. Pleuromeia jiaochengensis sp. nov. 1 -2. Part of strobili containing megaspores (at a). 1 , Z0 1 —0 1 ,
x 2. 2, an enlargement of syntype, PI. 21, fig. 11, showing megaspores associated with nearly all the
sporophylls. 3-5. Sporophylls and sporangia. 3, two sporophylls, the right one with upward reflexed apex,
Z01-08 x 2; 4, sporangia near base of a strobilus, showing the parallel lines on their surface, Z0 1-042; 5, a
larger sporophyll having lost its reflexed apex, Z0 1-061. 6. Part of a stem, showing spine-like leaves,
Z0i-00. 7. Enlargement of a sporangium containing megaspores (a), from the strobilus of PI. 21, fig. 11,
Z01-019, x 10. 8. A megaspore, Z01-022, x 50. 9-15. Various rhizophores; 9, 10, showing appendage
scars, Z01-060, Z01-135; 1 1, on the right, a young rhizophore showing a few appendages on the lower part,
Z0 1-054; 12, the same x 2; 13-15, showing appendages on the lower part, and appendage scars on the upper
part; 13, Z01-027; 14, Z04-151; 15, syntype, Z04-159.
PLATE 24
WANG and WANG, Pleuromeia
220
PALAEONTOLOGY, VOLUME 25
TEXT-FI
sis sp. I
..■C A
r:;---w.v*’ iS v v
i®S#
->y:
'•• '■< 5 " .>■'
d e c ortic ate d
stem
,^>:c &
11
Fife
#15 fefc
G. 2. Pleuromeia jiaochengen-
iov. Reconstruction of a whole
plant, x 1.
sporangium decreases gradually from proximal to distal end,
resulting in an aggregation of spores at the proximal end near
the rachilla. Megaspores tetrad, spherical or triangular, laevi-
gate or granulate, 300-500 p.m in diameter. Trilete mark
distinct, shorter than spore radius. Microspores unknown.
Discussion and comparison. The single, erect, terminal strobilus
terminating on unbranched woody stem, are characters asso-
ciated with the genus Pleuromeia rather than Selaginella or
Isoetes. Though the sporophylls of this new species have an
elongate anterior border similar to those in Selaginella , they
differ from it in lacking elongate acuminate distal tips
(Retallack 1975). Also, the thick, spatulate shape is quite unlike
the sporophyll of Selaginella. In addition, Selaginella is
herbaceous with a dichotomous stem and dimorphic leaves
often in four rows (Chaloner 1967; Smith 1955), quite unlike
those of P. jiaochengensis.
Isoetes, which is usually considered a close relative of
Pleuromeia, lacks a definite stem from the lobed rhizophore,
and also lacks a distinct strobilus. Unlike Pleuromeia, the
sporophylls of Isoetes are placed within the centre of a cluster
of leaves.
In the past, when identifying species of Pleuromeia, authors
have placed considerable emphasis on features of the leaf-scars
on the stem surface. During the past two decades, more recent
finds have greatly increased knowledge of the reproductive
organs; for example, the strobilus, sporophylls, sporangia,
megaspores, and microspores. These are now used much more
in identification. Also, Dobruskina (1974) pointed out that the
leaf-scars on specimens assigned to Pleuromeia and Pleuro-
mopsis by Brick (1936) and Sixtel (1962) respectively, from the
late Triassic of Central Asia, are not from these genera, and
that the leaf-scars of Pleuromeia vary in shape depending
on the degree of decortication prior to burial. Some specimens
from the late Permian of the Petchora Basin in the Soviet
Union included by Neuberg (19606, pi. 5, fig. 2; pi. 14,
right) in the genus Viatscheslavia have similar leaf-scars
to those in Pleuromeia. Hence leaf-scar features are not
now considered as important as in the past in identifying
Pleuromeia.
Of the six previously described species of Pleuromeia from
the Northern Hemisphere, only three, namely P. sternbergi,
P. rossica, and P. hatai, have been found in sufficient numbers
and are well preserved enough to give knowledge of the
complete plant. Of the rest, there is not enough real evidence
for their assignation as distinct species. P. oculina (Blancken-
horn, 1886) lacks reproductive organs, P. olenekensis
(Krassilov and Zakhalov, 1975) is too incomplete to compare
with other species, and the designation of P. obrutschewi Elias is
in dispute (Krassilov and Zakhalov 1975).
P. jiaochengensis differs from all previously described species,
by its small size, relatively large strobilus with elongate
WANG AND WANG: PLEUROME1A JIAOCHENGENSIS
221
sporophylls, and its undeveloping rhizophore with appendage-scars which are less in number, and
larger. P. jiaochengensis is usually 20-30 cm high, with a maximum height of under 50
cm. This compares with 2 m in P. sternbergi and 1 m in P. rossica. The smallest specimens of P.
hatai may be of similar size.
The strobilus of P. jiaochengensis is larger in proportion to the whole plant than in other species,
except P. hatai. The strobilus of P. hatai, however, differs from P . jiaochengensis in being cylindrical
rather than spike-shaped. Also, in P. hatai there are more sporophylls in a strobilus, and they
are smaller in size.
The sagittate shape of the sporophyll of P. jaiochengensis is slightly different from that in all
text-fig. 3. Pleuromeia jiaochengensis sp. nov. a, complete young plant, x 2; b, a strobilus with
megaspores, x 2; c, a strobilus, x 1; d, two sporophylls, the left one with an upward reflexed apex,
x2; e, a sporangium containing megaspores, x 10; /, a megaspore, x 50; g, a rhizophore covered
with appendage-scars, x 1 ; h, a stem surface with a few leaf-scars, x 2.
222
PALAEONTOLOGY, VOLUME 25
other species. Plate 24, fig. 5, right, shows a larger sporophyll of P . jiaochengensis, its upward reflexed
tip having been truncated, exposing a retuse anterior margin which differs from P. rossica, in having a
notched margin (PI. 24, fig. 5). In P. sternbergi the mature sporophylls also have an elongate outline,
but unlike P . jaiochengensis the lateral and anterior margins are of approximately equal width.
Although the nature of the sporangium of P. jaiochengensis is not as yet clear, the uniform
megaspores are different from those in P. rossica, but similar to those in P. sternbergi and P. hatai.
However, unlike these two species, the sporangium has a markedly thickened proximal end (PI.
24, fig. 7; text-fig. 3e), which contains a mass of megaspores.
The awl or spine-like leaves of P. jaiochengensis is one of the most important characters in
comparing this with other species. No clear leaf-scars are present after the leaves have dehisced. The
leaves of P. sternbergi and P. hatai are linear in outline and, after dehiscence, leave obvious leaf-scars
on the outer stem surface. The leaves of P. rossica are probably scale-like (Dobruskina 1974), and
rather similar to those in P. jaiochengensis, but when dehisced they leave densely covered and
distinctive leaf-scars.
An undeveloping rhizophore is also regarded by the present writers as an important character
for distinguishing the new species. Its appendage scars are larger and less in number than other
species. In addition, unlike the other species, the rhizophore has neither horns nor suture lines
(kliifte) on the lower surface.
When describing Pleuromeia from the Qinshui basin of Shanxi, Wang et al. (1978) noted that
the difference between the strobili of P. sternbergi and P. rossica was so great that, as pointed out
by Kon’no (1974), they might represent different genera. With regard to this, P. jaiochengensis
throws more light on P. sternbergi based on its unisexual strobilus without a pedicel, and on the
character of the sporophylls.
Retallack (1975) transferred the lycopsid-like strobilus Cylostrobus (originally designated by
Helby and Martin 1965) which had previously been considered more like the palaeozoic lycopsids, to
Pleuromeia longicaulis (Burgess). The specimens are from the early Triassic of eastern Australia, and
is the first record of this genus in the Southern Hemisphere. This type of strobilus shows some
important features, for example, it is bisexual (monoecious), the sporophylls have a remarkable
ribbed keel-like apex on their dorsal surface, and the microspores are monolete and covered with
dense spines. All these features are strikingly different from Northern Hemisphere Pleuromeia
species, including P. jaiochengensis. The basis for the designation of the Australian species given by
Retallack, mainly results from an analysis of their environment and a few associated fossil stems. In
fact, there are two types of stems associated with Cylostrobus sydneyensis from the upper Narrabeen
Group; one bearing a stigmaria-like stem-base, and the other with branchlets (Helby and Martin
1965, p. 399, pi. 1 , fig. 6) more like a palaeozoic lycopsid than Pleuromeia. The present authors are not
as yet convinced that this Australian species should be included within the genus Pleuromeia.
Recently, Ash (1979) has described Skilliostrobus, a new type of lycopsid strobilus also from the early
Triassic of Australia, and this is similar to Cylostrobus in having bisexual, monoecious features.
ECOLOGICAL PROBLEMS CONCERNING P. JAIOCHENGENSIS
Pleuromeia was originally regarded as a desert xerophyte similar to the present-day Cactaceae.
More recently, linking the plant-bearing beds with associated sediments containing dolomite and
halite, together with marine fossil animals, a marginal marine habitat was suggested. Magdefrau
(1931) was first to suggest an halophytic habitat for Pleuromeia. Hirmer (1933), Neuberg (19606),
and Kon’no (1973) have supported this view, and more recently Krassilov and Zakharov (1975)
and Retallack (1975) have suggested a mangrove-type environment along the sea-shore, or
marginally on deltas. This type of habitat is not envisaged for P. jiaochengensis. Unlike the records
from western Europe, Siberia, and Japan, no marine animal fossils have been recorded from the
strata containing P. jiaochengensis in Jiaocheng district. Also, there is no evidence for lagoonal
sediments, with the absence of dolomite and halite, and only a few doubtful reports of gypsum.
WANG AND WANG: PLEUROMEIA JIAOCHENGENSIS
223
According to the geological exploratory reports of the Shanxi Regional Geological Survey Team
the Formation in this area contains a lot of unweathered minerals such as feldspar and magnetite.
Cross-bedding, ripple marks, and sun-cracks are relatively common, and indicate an arid climate.
The ‘Shischienfeng Series’ with marine intercalated beds are known only from Linyuao, and
Tongchuan districts of north Shanxi (Shengsi), and are at least 300 km from the Jiaocheng district
(see text-fig. 1 and Yin and Lin 1979). We therefore consider P. jiaochengsis to be a small plant
growing near desert oases. As all specimens are found in greyish-green sandstones, the new species
might grow very near or even partly in the water bodies. Its small size also suggests arid conditions;
the undeveloped and abortive leaves probably resulted from extreme transpiration in an arid climate.
The appendages are stout and strong, probably to withstand desert conditions. Further evidence
for an arid environment comes from the Estheridae which are recorded from the shales intercalated
with the plant-bearing sandstones. This type of small fauna is often found in temporary inland
bodies of water.
Plate 23, figs. 3-4, show a small but complete plant which is only 3 cm high. The stem is almost
smooth, the large strobilus is well developed at the apex, and has a few distinctive appendages at its
base. The appendages are plano-concave and fleshy, and were probably succulent. They are more
developed than the sterile leaves, and may well have carried out much of the photosynthesis.
TABLE 1. The stratigraphic sequence of the ‘Shischienfeng Series’ in Shanxi.
Early Triassic
Induan Olenekian
‘Shischienfeng Series’
Heshankou Formation
Pleuromeia sternbergi
P. rossica
Luijiakou Formation
P. jiaochengensis
Later Permian
Tartarian
Sunjiakou Formation
Shihtienfenia permica
224
PALAEONTOLOGY, VOLUME 25
STRATIGRAPHIC OCCURRENCE OF THE GENUS PLEUROMEIA,
AND THE AGE OF P. JIAOCHENGENSIS
In Germany the Pleuromeia- bearing beds have a rather short vertical range, extending from Middle to Upper
Bunter (i.e. from Bausandstein to Chirotheriumsandstein). The age according to Lozovsky, Movschovich,
and Mimich (1973) is considered to be equivalent to the Tirolites-Columbites ammonite zone of Middle and
Upper Olenekian age, despite the lack of the ammonite evidence. At Russian Island near Vladivostock and
north-east Japan, Pleuromeia occurs in rocks of early Triassic age dated on the ammonites. On the Russian
Platform, P. rossica occurs in the Ribinsk Formation together with the amphibian Benthosuchus, and is dated
as early Triassic. According to Lozovsky et al. (1973), this Formation is equivalent to the Owenites ammonite
zone, and may be referred to the early Olenekian.
At Mangeschlack near the Caspian sea, P. sternbergi occurs with marine lamellibranchs which also occur
in the Upper Bunter Sandstone in Germany (Dobruskina 1974) and so are dated as late Olenekian. The
Pleuromeia localities in the Vosges, at Mangeschlack in Central Asia, and on Russian Island in the Far East,
are suggested by some authors as being as late as Middle Triassic. However, in the present writers’ opinion,
the exact age is in doubt, through lack of detailed information.
The Luijiakou Formation in central Shanxi is thought to be equivalent to the middle part of the so-called
‘Shischienfeng Series’. For many years, no fossils were found, and the age was in doubt, but thought to be late
Permian to early Triassic, depending on the exact age of the ?late Permian reptile Shihtienfenia permica Young,
found in the underlying Sunjiakou Formation (see Table 1).
Recently Wang, Xie, and Wang (1978) described P. sternbergi and P. rossica from the overlying Heshankou
Formation and suggested an early-middle Olenekian age for that Formation. The discovery of P.jiaochengensis
in the Luijiakou Formation lends further support to the above age determination. In addition, another of
the fossil plants associated with P. jiaochengensis, is Crematopteris sp. This species is confined to the Bunter
Sandstone in western Europe. If the age of the Heshankou Formation containing P. rossica is referred to the
Olenekian stage, or is equivalent to the Ribinsk Formation on the Russian Platform, then the age of the
Luijiakou Formation can accurately be assigned to the Induan of early Triassic age.
Acknowledgements. Sincere thanks are due to Professor X. X. Lee of Nanjing Institute of Geology and
Palaeontology, Academia Sinica, for his encouragement.
REFERENCES
ash, s. R. 1979. Skilliostrobus gen. nov. a new Lycopsid cone from the Early Triassic of Australia. Alcheringa, 3,
73-79.
blanckenhorn, m. 1886. Die fossilen Flora des Buntsandsteins und des Muschelkalke der Umgegend von
Commern. Palaeontographica, 32, 117-154.
brick, m. M. 1936. The first discovery of a Lower Triassic flora in Central Asia. Trans. Geol. Inst. Acad. Sci. 5,
461-474. [In Russian.]
chaloner, w. g. 1967. In Traite de paleobotanique. Tome II. Bryophyta, Psilophyta, Lycophyta. Ed. boureau, e.
Masson et Cie, Paris. Pp. 1-802.
dobruskina, I. a. 1974. Triassic lycopods. Palaeont. Z. 111-124. [In Russian.]
helby, r. and martin, a. r. h. 1965. Cylostrobus gen. nov. cones of Lycopsidean plants from the Narrabeen
Group (Triassic) of New South Wales. Aust. J. Bot. 13, 384-404.
hirmer, m. 1933. Rekonstruktion von Pleuromeia sternbergi (Muster) Corda, Nebst Bermerkungen zur
morphologie der Lycopodiales. Palaeontographica , B 78, 47-56.
kon’no, e. 1973. New species of Pleuromeia and Neocalamites from the upper Scythian bed in the Kitakami
Massif. Sci. Rep. Res. Insts To hoku, 43, 2nd ser. (Geol.), 97-115.
krassilov, v. a. and zakharov, d. 1975. Pleuromeia from the Lower Triassic of the Far East of the USSR. Rev.
Palaeobot. Palynol. 19, 221-232.
lozovsky, v. r., mavschovich, e. v. and mimich, m. t. 1973. On the stratigraphy of Early Triassic sediments on
the Russian Platform. Proc. Acad. Sci. USSR, ser. Geol. 111-117. [In Russian.]
magdefrau, K. 1930. Beitrage zur Kenntnis der Thuringischen Buntsandsteinzeit. Geol. Thtir. 2, 285-289.
— 1931. Zur Morphologie und Phylogenetische Bedeutung der fossilen Pflanzengattung Pleuromeia. Beih.
Bot. 61. 48, 119-140.
WANG AND WANG: P LEU RO M El A JIAOCHENGENSIS
225
neuberg, M. f. 1960a. Permian Flora of Petchora Basin, Part I. Trans. Geol. Inst. Acad. Nauk. SSSR, 43, 1 64.
[In Russian.]
— 19606. Pleuromeia Corda from the Lower Triassic deposits of the Russian platform. Ibid. 65-90.
retallack, G. J. 1975. The life and times of a Triassic Lycopod. Alcheringa, 1, 3-29.
sixtel, t. a. 1962. Later Permian and Early Triassic floras in South Fergana. Stratigr. Palaeont. Uzbekstana and
its neighbouring Region. [In Russian.]
smith, G. M. 1955. Cryptogamic Botany, Vol. II: Bryophytes and Pteridophytes. Mcgraw-Hill Co. 1959. Chinese
translation. Pp. 1-287.
WANG LIXIN, XIE zhimin, and WANG ziqiang, 1978. On the occurrence of Pleuromeia from Qinshui Basin in
Shanxi Province. Acta Palaeont. Sin. 17, 195-212. [In Chinese, English summary.]
yin hong-fu and lin he-mao, 1979. Triassic marine sediments from northern Shanxi with note on the age of the
Shischienfeng Group. Acta Stratigr. Sin. 3, 233-241. [In Chinese.]
WANG ZIQIANG
Tianjin Institute of Geology and Mineral Resources
Ministry of Geology
No. 26, Jintang Highway, Tianjin (Tientsin)
People’s Republic of China
Typescript received 13 June 1980
Revised typescript received 19 December 1980
WANG LIXIN
The Regional Geological Survey Team of Shanxi
Yuci, Shanxi Province
People’s Republic of China
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 1981 are:
Institutional membership. . £32 (U.S. S71)
Ordinary membership. £15 (U.S. $33)
Student membership. . £9-50 (U.S. $21)
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 1982 will receive Volume 25, Parts 1-4. All back numbers are still in print and may be ordered from
Marston Book Services, P.O. Box 87, Oxford 0X4 1LB, England, at £15 (U.S. $33) 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 1982 is £25 (U.S. $55) for Institutional Members, and £12-50 (U.S. $27-50) 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, P.O. Box 87, Oxford 0X4 1LB, England.
COUNCIL 1981-1982
President: Dr. W. H. C. Ramsbottom, Institute of Geological Sciences, Leeds LS15 8TQ
Vice-Presidents: Professor C. B. Cox, Department of Zoology, King’s College, Strand, London WC2R 2LS
Professor J. W. Murray, Department of Geology, The University, Exeter EX4 4QE
Treasurer: Mr. R. P. Tripp, High Wood, Botsom Lane, West Kings Down, 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
Marketing Manager: Dr. R. J. Aldridge, Department of Geology, The University, Nottingham, NG7 2RD
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
Dr L. R. M. Cocks, Department of Palaeontology. British Museum (Natural History). Cromwell Road. London SW7 5BD
Other Members of Council
Dr. D. E. G. Briggs, London
Dr. C. H. C. Brunton, London
Dr. R. Harland, Leeds
Dr. A. R. Lord, London
Dr. J. Miller, Edinburgh
Dr. T. J. Palmer, Oxford
Dr. M. Romano, Sheffield
Dr. D. J. Siveter, Leicester
Dr. P. W. Skelton, Open University
Dr. P. D. Taylor, London
Dr. A. T. Thomas, Birmingham
Dr. H. S. Torrens, Keele
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
IVest Indies and Central America: Mr. John B. Saunders, Geological Laboratory, Texaco Trinidad, Inc., Pointe-a-Pierre,
Trinidad, West Indies
U.S. A: Dr R. Cuffey, Department of Geology, Pennsylvania State University, Pennsylvania, U.S. A.
South America: Dr. O. A. Reig, Departmento de Ecologia, Universidad Simon Bolivar, Caracas 108, Venezuela
Palaeontology
VOLUME 25 ■ PART 1
CONTENTS
Rapid evolution in echinoids
P. M. KIER 1
The palaeobiology of the Cretaceous irregular echinoids
Infulaster and Hagenowia
A. S. GALE and A. B. SMITH 1 1
A revision of late Ordovician bivalves from Pomeroy,
Co. Tyrone, Ireland.
S. P. TUNNICLIFF 43
Juvenile specimens of the ornithischian dinosaur Psittaco-
saurus
W. P. COOMBS, JR. 89
A description of the generating curve of bivalves with straight
hinges
M. J. ROGERS
Reworked acritarchs from the type section of the Ordovician
Caradoc Series, Shropshire
R. E. TURNER 1
Lower Carboniferous conodont faunas from Ravenstonedale,
Cumbria
A. C. HIGGINS and W. J. VARKER 145
Somasteroidea, Asteroidea, and the affinities of Luidia
{Platasterias) latiradiata
D. B. BLAKE 167
A Lower Carboniferous aistopod amphibian from Scotland
C. F. WELLSTEAD 193
Morphology and relationships of the Upper Carboniferous
aistopod amphibian Ophiderpeton nanum
M. J. BOYD 209
A new species of the lycopsid Pleuromeia from the early Tri-
assic of Shanxi, China, and its ecology
WANG ZIQIANG and WANG LIXIN 2 1 5
Printed in Great Britain at the University Press, Oxford
by Eric Buckley, Printer to the University
issn 0031-0239
Palaeontology
VOLUME 25 • PART 2 APRIL 1982
Published by
The Palaeontological Association London
Price £1 7
THE PALAEONTOLOGICAL ASSOCIATION
The Association publishes Palaeontology and Special Papers in Palaeontology. Details of membership and subscription rate;
may be found inside the back cover.
The journal Palaeontology is devoted to the publication of papers on all aspects of palaeontology. Review articles arc
particularly welcome, and short papers can often be published rapidly. A high standard of illustration is a feature of th<(
journal. Four parts are published each year and are sent free to all members of the Association. Typescripts on all.aspects o
palaeontology and stratigraphical palaeontology are invited. They should conform in style to those already published in thi;l
journal, and should be sent to Dr. R. A. Fortey, Palaeontological Association, Department of Palaeontology, British Museun i
(Natural History), Cromwell Road, London SW7 5BD, England, who will supply detailed instructions for authors on reques
(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. $17-50).
2. (for 1968): Evolution of the Shell Structure of Articulate Brachiopods, by A. williams. 55 pp., 27 text-figs., 24 plateA
Price £5 (U.S. $11).
3. (for 1968): Upper Maestrichtian Radiolaria of California, by Helen p. foreman. 82 pp., 8 plates. Price £3 (U.S. $6-50
4. (for 1969): Lower Turonian Ammonites from Israel, by R. freund and m. raab. 83 pp., 15 text-figs., 10 plates. Price £j
(U.S. $6-50).
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. $4-50).
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. $6-50).
7. (for 1970): Shell Structure of the Craniacea and other Calcareous Inarticulate Brachiopoda, by A. williams an\
a. d. Wright. 51 pp., 17 text-figs., 15 plates. Price £1-50 (U.S. $3-50).
8. (for 1970): Cenomanian Ammonites from Southern England, by w. J. Kennedy. 272 pp., 5 tables, 64 plates. Price £
(U.S. $17-50).
9. (for 1971): Fish from the Freshwater Lower Cretaceous of Victoria, Australia, with Comments on the Palaeoj
environment, by M. waldman. 130 pp., 37 text-figs., 18 plates. Price £5 (U.S. $11).
10. (for 1971): Upper Cretaceous Ostracoda from the Carnarvon Basin, Western Australia, by r. h. bate. 148 pp., 43 tex \
figs., 27 plates. Price £5 (U.S. $11).
11. (for 1972): Stromatolites and the Biostratigraphy of the Australian Precambrian and Cambrian, by M. R. waltei
268 pp., 55 text-figs., 34 plates. Price £10 (U.S. $22).
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. $22) (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. hughe]
261 pp., 26 plates. Price £10 (U.S. $22).
14. (for 1974): Palaeogene Foraminiferida and Palaeoecology, Hampshire and Paris Basins and the English Channel, h
j. w. Murray and c. a. wright. 171 pp., 45 text-figs., 20 plates. Price £8 (U.S. $17-50).
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. $12).
16. (for 1975): The Ostracod Fauna from the Santonian Chalk (Upper Cretaceous) of Gingin, Western Australia, /|
j. w. neale. 131 pp., 40 text-figs., 22 plates. Price £6-50 (U.S. $14-50).
17. (for 1976): Aspects of Ammonite Biology, Biogeography, and Biostratigraphy, by w. J. Kennedy and w. a. cobba
94 pp., 24 text-figs., 11 plates. Price £6 (U.S. $13).
18. (for 1976): Ostracoderm Faunas of the Delorme and Associated Siluro-Devonian Formations, North West Terri torie
Canada, by d. l. dineley and e. j. loeffler. 218 pp., 78 text-figs., 33 plates. Price £20 (U.S. $44).
19. (for 1977): The Palynology of Early Tertiary Sediments, Ninetyeast Ridge, Indian Ocean, byE. m. kemp and w. k. harr
74 pp., 2 text-figs., 8 plates. Price £7 (U.S. $15-50).
20. (for 1977): Fossil Priapulid Worms, by s. c. morris. 159 pp., 99 text-figs., 30 plates. Price £16 (U.S. $35).
21. (for 1978): Devonian Ammonoids from the Appalachians and their bearing on International Zonation and Correlt
tion, by M. r. house. 70 pp., 12 text-figs., 10 plates. Price £12 (U.S. $26-50).
22. (for 1978, published 1979): Curation of palaeontological collections. A joint colloquium of The Palaeontological Associ
tion and Geological Curators’ Group. Edited by m. g. bassett. 280 pp., 53 text-figs. Price £25 (U.S. $55).
23. (for 1979): The Devonian System. A Palaeontological Association International Symposium. Edited by M. R. house, cJ
scrutton, and M. G. bassett. 353 pp., 102 text-figs., 1 plate. Price £30 (U.S. $66).
24. (for 1980): Dinofiagellate Cysts and Acritarchs from the Eocene of southern England, by J. b. bujak, c. downie, G.
eaton, and G. l. williams. 104 pp., 24 text-figs., 22 plates. Price £15 (U.S. $33).
25. (for 1980): Stereom Microstructure of the Echinoid Test, by a. b. smith. 81 pp., 20 text-figs ., 24 plates. Price £15 ($31;
26. (for 1981): The fine structure of Graptolite periderm, by p. r. crowther. 1 19 pp., 37 text-figs., 20 plates. Price £25 ($51
27. (for 1981): Late Devonian Acritarchs from the Carnarvon Basin, Western Australia, by G. playford andR. s. dring. 78 p
10 text-figs., 19 plates. Price £15 ($33).
© The Palaeontological Association , 1982
Cover: An acritarch of the genus Multiplicisphaeridium from the Tournaisian Bedford Shale of Ohio, U.S. A. Institute;
Geological Sciences specimen number MPK 3569. S.E.M. x 1000. Within the last twenty-five years the number of descrit
acritarch taxa has increased dramatically, and these organisms are now used extensively to date rocks of late Prccambr
to Devonian age.
ECOLOGY AND POPULATION STRUCTURE OF
THE RECENT BRACHIOPOD TEREBRATU LIN A
FROM SCOTLAND
by GORDON B. CURRY
Abstract. The ecology and population structure of the Recent articulate brachiopod Terebratulina retusa
(Linnaeus) are described. The population studied occurs around the margins of a depression of more than 220 m
in the Firth of Lome, Scotland, and is predominantly attached to the horse-mussel Modiolus modiolus
(Linnaeus). Spawning occurs regularly in late spring and late autumn, and is initiated at temperatures of
10-1 1°C. The highly synchronized reproductive cycle, from spawning to spatfall, occurs within 3 weeks in
nature. Length-frequency histograms prepared from large representative samples collected at regular inter-
vals during 1977-1979 are unimodal and right-skewed due to the predominance of juveniles. Regularly spaced
subsidiary peaks in the histograms correspond to biannual settlement cohorts; in later life successive peaks
merge to form a single annual peak. This pattern is identical to that predicted by computer-based simulations.
Recently settled specimens grow rapidly to an average length of 2-75 mm within 3 months during both spring and
autumn; thereafter the animals grow (initially by 4 mm per year) throughout life, although at a progressively
reducing rate from the third year of life onwards. Growth slows or ceases in winter in all but recently settled
specimens. The maximum life span is 7 years. The mortality rate remains constant, although the causes of death
are not apparent. The growth-lines form biannually, at times of pronounced environmental and physiological
disturbance.
It is not generally realized that the Recent brachiopod fauna of the British Isles (21 species of 17
genera, Brunton and Curry 1979) is significantly more diverse than that of New Zealand (12 species
of 9 genera). Considering the relative diversity of the two faunas, it may appear strange that New
Zealand has become the classic area for Recent brachiopod research. However, this apparent
anomaly is readily explained since the New Zealand species are far more accessible, with 25% of the
species abundant intertidally and a further 30% common in nearshore shallow subtidal habitats. By
contrast, British brachiopods are very rarely found intertidally and the majority of species have only
been collected in small numbers from widely dispersed and often inaccessible localities.
Intermittent research during the last 100 years has demonstrated that several of the British species
are locally abundant, especially off the west coast of Scotland and in the Western Approaches (Atkins
1959a, b, 1960 a, b, c, 1961; Davidson 1886-1888; Chumley 1918; see summary maps in Brunton and
Curry 1979), although dredging is the only practicable method of sampling these predominantly
deeper-water populations. The recent upsurge of interest in palaeoecology has emphasized the need
for precise data on the life-habits and ecology of Recent representatives of fossil phyla, and more data
are needed to augment the patchy and often contradictory information available to the brachiopod
palaeoecologist. The present study was initiated in the light of such inadequacies, and when it became
clear that the comprehensive dredging facilities available at Dunstaffnage Marine Research
Laboratory, nr. Oban, Scotland, could ensure access to the abundant brachiopod populations off the
west coast of Scotland.
Although dealing entirely with living animals, this study was conducted from a palaeontological
viewpoint. As such, the main interest was the interpretation of features on the shell, in particular the
analysis of population structure and dynamics based on size-frequency histograms prepared from
large samples of brachiopods collected at regular intervals throughout 1977-1979. The main
advantage of working with living, as opposed to fossil, populations is that it is possible to seek
IPalaeontology, Vol. 25, Part 2, 1982, pp. 227-246.|
228
PALAEONTOLOGY, VOLUME 25
confirmatory evidence for conclusions reached by direct observations on the living population. This
paper first describes aspects of the biology and ecology of the animal, and goes on to apply the results
to the interpretation of size-frequency histograms.
MATERIAL
The species under investigation was Terebratulina retusa (Linnaeus), which is the most abundant of the Recent
British brachiopods. Initially the intention was to collect regular bulk samples from the large populations of T.
retusa which were known to occur in the Sound of Mull (text-fig. 1). However, during an exploratory cruise in
March 1977 an extremely abundant population, ideally situated for regular sampling, was discovered near a
depression in the nearby Firth of Lome (text-fig. 1). This depression, which reaches a maximum depth in excess
of 220 m, is mid-way between the Island of Kerrera and the south-west coast of the Island of Mull (Grid. ref. NM
745 265; Stn. 1 in text-fig. 1), and is thought to lie along the Firth of Lome Fault (a subsidiary splay of the Great
Glen Fault, Barber et al. 1979). At this locality the brachiopods are predominantly attached to the horse-mussel
Modiolus modiolus (Linnaeus) which occurs in dense beds around the smoothly sloping margins of the
depression.
The samples were collected, by the R/V Calanus or Seol Mara, using a conventional ‘clam-dredge’ (1-2 m
wide x 2 m long), the body of which consists of an outer framework of interlocking iron chain and an inner
nylon meshwork with a maximum aperture of 15 mm. This dredge is designed to collect material resting on, or
partially buried within, the substrate, and proved to be extremely efficient at sampling the Firth of Lome mussel
beds. For each sample the dredge was trawled slowly along the substrate for approximately 5-10 minutes, in a
text-figure 1. Map showing locations of R/V Calanus
sample stations off the west coast of Scotland; station
numbers refer to Table 1 (see opposite).
CURRY: B R ACHIOPOD TEREB RATE LIN A FROM SCOTLAND
229
north-easterly or south-westerly direction. Once landed on deck, the sample was washed with sea-water to
remove any adherent sediment, and then placed in plastic baths supplied with flowing fresh sea-water.
The disturbance caused by the dredging was minimal, and the majority of the sample had become
‘acclimatized’ to the plastic baths and commenced feeding within half an hour of arriving on deck. Mechanical
damage during dredging was not significant, and only seven specimens out of the 818 brachiopods collected in
March 1977 (i.e. 0-85%) proved unmeasurable because of shell damage. Due to the efficiency of the ‘clam-
dredge’, and the density and abundance of both mussels and brachiopods, all attempts to sample the Firth of
Lome populations were successful. In addition, the Firth of Lome depression is a prominent submarine feature,
easily recognizable using the ship’s depth-sounding equipment, and it was therefore possible to ensure that all
samples were collected from the same population.
table 1. Results of the dredging operations of the R/V Calanus, west coast of Scotland, 1977-1979
(a = abundant, c = common, r = rare, x = absent).
STATION No
LOCATION.
DEPTH.
SEDIMENT .
abund
T.retusa
. j a t ta : tuna n f. .
abund
anomala
.] attachment .
ASSOCIATED FAUNA.
1
Firth of Lome
146-I83m.
fine mud
a
1
1 Modiolus
a
1
1 Modiolus
see text.
2
Rabbit Island
13m.
--
c
1 rocks
r
, ri :ks
Pecten
Garbh Reisa
20m.
—
X
! --
A
1 --
Pecten
4
Sound of Jura
(north-east )
?3-128m.
--
X
1
X
1
1
Modi plus and
hydrozoans
5
Sound of Jura
110m.
-
r
1
1 vesicular
, basalt
X
large no. of dis-
articulated Modi plus
6
Sound of Jura
(south)
37m.
--
1
x
_J
Pecten. Balanus.
?
E. of Colonsay
37m.
sand
X
' --
X
seaweed, crabs.
8
North-east of
Colonsay
55m.
—
0
1 rocks and
1 clinker
c
1 rocks and
1 clinker
--
9
Torran Rooks
91-llPm.
—
X
1
|
A
1 ..
—
10
North-west of
Iona
?3-110m
--
r
1 rocks
1
X
— 1
1
—
11
North of Iona
73-123m.
sand
1
1
X
1 —
—
12
Treshnish Is.
29-55m.
-
x
1
X
“
1
Balanus . serpulids
on rocks
13
Lunga
37m.
—
X
! -
X
1 --
Balanus on rocks
14
North-west of
Mull
37-46m.
X
1
Pecten. eohinoids
15
East of Coll
146-183m.
—
X
1 ..
0
1 rooks
--
16
Ardmore Point,
Mull
73-213m.
mud
X
1
X
1
1?
Mingary Bay,
Ardnamurchan
37m.
mud
X
1
=
rocks
1
13
Tobermory Bay,
Mull
73m.
mud
c
1 clinker
1
a
clinker and
1 rocks
Modiolus
19
Sound of Mull
(north)
110m.
mud
c
' rocks
a
* rocks
l
20
Sound of Mull
(north)
Qlm.
mud
a
1 modiolus
1
c
1 Modiolus
1
21
Sound of Mull
(north )
91m.
mud
x
1
1
1
c
1 rocks
--
22
Sound of Mull
(north )
91m.
mud
a
1 Modiolus
1
a
. Modiolus
_J
23
Sound of Mull
(north )
91-l46m.
mud
r
1
r
1 Modiolus
|
?4
Sound of Mull
(north)
18- 37m.
-
0
1 shell
1 fragments
X
|
25
Sound of Mull
( south )
110- 128m.
mud
a
Modiolus
l
a
Modiolus
26
Sound of Mull
( south )
123-146m.
mud
c
1 clinker
1
r
1 clinker
1
2?
Sound of Mull
(south)
l8-37m.
mud
c
1 clinker
1
r
1 clinker
_l
2 3
Li smore Island
37m.
-
r
1 vesicular
1 basalt
r
vesicular
1 basalt
29
Lismore Island
15-2 6m.
-
r
1 rocks
1 -
230
PALAEONTOLOGY, VOLUME 25
The samples were transported back to the laboratory and placed in an outside aquarium through which fresh
sea-water was continually being circulated. Because of the difficulty in seeing very small brachiopods (recently
settled post-larvae are less than 0-5 mm in length and their shells are transparent), the surfaces of all mussels and
other potential brachiopod substrates were examined using a binocular microscope. Once detected, each
brachiopod was removed from its substrate by severing its pedicle with a sharp scalpel, measured to the nearest
01 mm using ‘MITUTOYO’ dial calipers, and then preserved in either 10% formalin or 70% alcohol. The
samples have been deposited in the Department of Palaeontology, British Museum (Natural History), London,
and the registration numbers quoted in the text (with the prefix ZB) refer to the Recent brachiopod collections in
that museum.
ECOLOGY OF TEREBRATULINA RETUSA
Distribution. The precise geographic limits of the distribution of T. retusa are unknown, as there are
two morphologically similar and often confused species of Terebratulina in the North Atlantic.
Positively identified T. retusa have been collected from as far north as Norway (the type area) and as
far south as Spain and the Mediterranean. The species would appear to be confined to the north-
eastern North Atlantic, although it has been recorded from the east coast of Greenland (Wesenberg-
Lund 1940). The possibility that the two named species are members of a Terebratulina cline has been
considered by several authors (e.g. Wesenberg-Lund 1941), and certainly the presence of individuals
with intermediate morphological characteristics in the mid North Atlantic could explain the
confusion over the geographic distribution of the two species, which are quite distinct.
T. retusa is relatively common off the west coast of Scotland, and has been recorded from many of
the sea-lochs (Davidson 1886-1888; Chumley 1918). On a more local scale the comprehensive
dredging equipment available on the R/V Calanus provided an opportunity to investigate the pattern
of distribution of T. retusa around the Islands of Mull and Jura (text-fig. 1). The results of these
dredging operations provide strong evidence of an essentially patchy distribution, which is illustrated
by the results of a comprehensive sampling programme in the Sound of Mull (Stns. 18-27 in text-fig. 1
and Table 1). In most of these samples the horse-mussel M. modiolus formed the bulk of the sample,
whilst T. retusa varied from being the numerically dominant constituent (e.g. Stn. 22), to being a
minor constituent (e.g. Stn. 23), or being totally absent (e.g. Stn. 21). Suitable substrate for
brachiopod attachment was present at the latter station, and the fact that the habitat occupied by T.
retusa is not occupied by other sessile epibenthos suggests that competition for space is not an
important factor in these deeper water localities. It is likely, therefore, that the patchy distribution is
an inherent feature of the T. retusa population, reflecting the short duration, and hence limited
dispersal range, of the pelagic larval stage (see below). Many other Recent brachiopod populations
appear to be patchily distributed (Rudwick 1970, p. 156).
It is appropriate to sound a cautionary note on the use of benthic dredging to determine the
distribution pattern of sessile epibenthos such as T. retusa. Benthic dredges will only collect free-lying
or loosely attached material which falls within a restricted size range, and it is clear that the substrates
used by brachiopods are very variable both in grain-size and in the extent to which they are anchored
to the sea bed. Therefore the absence of brachiopods in a particular sample may simply be due to the
inability of the dredge to sample substrates such as large boulders or rock-faces. This feature was
evident during the present study, as dredged samples from the Sound of Jura yielded no brachiopods
(Stn. 4), and yet divers have reported quantities of T. retusa attached to large boulders in this region
(R. Harvey, pers. comm. 1978). Benthic dredging is only a satisfactory method of sampling
populations in which the predominant substrate is of a size and nature that can readily be collected, as
is the case with the mussel beds in the Firth of Lome.
Depth. Within the study area the most abundant populations of T. retusa occur at depths of 100-200
m, although a few specimens have been collected from an estimated depth of 3 m in the Sound of
Raasay (M. K. Howarth, pers. comm. 1978), and a reasonably abundant population is known to
occur at a depth of 13-20 m off the coast of Rabbit Island (Stn. 2 in text-fig. 1 and Table 1). Despite
these occurrences, T. retusa is not a common constituent of the shallow subtidal ecosystem along the
Scottish west coast, and there is no record of any dead shells being washed up on beaches in this
CURRY: B RACHIOPOD TEREB RATE LIN A FROM SCOTLAND
23
region. The known depth range of T. retusa is from 3-1478 m, although it is most commonly found
between 100 and 500 m. Environmental conditions vary considerably with depth, and it is the
tolerance of an organism to such variations which determines its depth range; depth in itself is not
considered to be a controlling factor (Moore 1958). Presumably, therefore, T. retusa is prevented
from colonizing intertidal or shallow tidal habitats by the combined effects of biological and physio-
chemical factors characteristic of such habitats, such as competition with other organisms, rapid
daily temperature fluctuations, the possibility of increased predation intensity, periodic desiccation,
salinity fluctuations, wave turbulence, and intensity of incident light. Competition for space is
certainly an inhibiting factor, as was illustrated during the present study when brachiopods collected
from the Sound of Mull were reintroduced into the shallow marine environment of Dunstaflfnage Bay
in recoverable cages. These brachiopods were killed by a dense settlement of fast-growing barnacles
which engulfed and smothered them within 3 months. Prolific swarms of barnacles are common in
shallow-water habitats along the Scottish coast, but in the deep-water localities where T. retusa is
abundant they are rare and solitary (in fact a different species). Clearly the relatively slow-growing
T. retusa is at a considerable disadvantage in areas where suitable substrates are in short supply.
Temperature. Temperature is one of the most important environmental parameters for cold-blooded
marine invertebrates (Moore 1958), limiting their geographic range, and affecting all aspects of life by
virtue of its controlling effect on the rates of metabolic processes (i.e. Van’t Hoff’s Law— see Prosser
1973). The general pattern of temperature fluctuation experienced by T. retusa in the Firth of Lome
depression was determined (sea-water samples were collected from the vicinity of the brachiopod
population using insulated sampling bottles), and has been compared with surface water
temperatures in text-fig. 2. The temperature of bottom waters in the Firth of Lome range from a
February minimum of 6-5 °C to a maximum of 13°C in August, an annual range of 6-5 °C which is
smaller than would be expected at this latitude because of the warming effect of the North Atlantic
Drift Current. The over-all pattern and range is similar in surface waters, (minimum in January of
6°C and a maximum of 14°C in August— annual range of 8°C), although it is apparent from text-fig.
2 that the bottom and surface waters are virtually never isothermal and can differ by as much as
2-5 °C. This vertical stratification of the water column, the thermocline, is a common feature of
Scottish sea-lochs and, although complex, can be considered as reflecting the greater susceptibility of
text-figure 2. Annual temperature curves for the surface and
bottom (approx. 150 m) waters in the Firth of Lome, Scotland.
232
PALAEONTOLOGY, VOLUME 25
surface waters to prevailing air temperatures (e.g. being warmer in summer and cooler in winter).
Other factors which enhance this vertical stratification are density and salinity gradients produced by
the high levels of surface run-off of fresh water from the surrounding land. The thermocline breaks
down during two short periods in spring and autumn when the water column becomes thoroughly
mixed, isothermal, and uniformly saturated with oxygen (i.e. the points at which the curves in text-fig.
2 cross).
Substrate. The predominant utilization of M. modiolus as substrate by the brachiopods in the Firth of
Lome depression is well illustrated by the analysis of the substrate of attachment of the 786 specimens
of T. re tusa collected on 5 May 1977 (Table 2). The sample contained 204 living mussels, 166 of which
No. of
No. of
°/. °f
table 2. Substrate of attachment of the 786
Substrate
Occurrences
Brachiopods
Total
specimens of T. retusa collected on 5 May 1977
from the Firth of Lome.
Living mussel
166
591
75
Dead mussel
27
53
7
Mussel frags.
33
63
9
Hydrozoan
10
13
2
Dead gastropod
5
14
2
Dead brachiopod
4
10
1
Living & dead
4
8
!
Sponge/ascidian
4
16
2
Unattached
4
8
1
Rock
2
9
1
Tube-worm
1
1
0.1
(i.e. 81%) had been utilized as substrate by T. retusa. A total of 707 brachiopods (91% of the sample)
were attached to living or dead mussels or to fragments of mussel shell. Mussels are ideal substrate for
brachiopods, being sessile and of much longer life span (at least 20 years (Comley 1978) as compared
with a maximum of 7 years for T. retusa). The relatively large surface area of the mussels is readily
bored by the anchoring pedicle rootlets of T. retusa, and is free of obstructions which would hinder
the rotation of brachiopods around their pedicles, a procedure which enables them to move away
from localized disturbances and to take up preferred feeding orientations. In addition, the majority of
brachiopods are attached anteriorly, close to the inhalant and exhalant feeding currents of the
mussels, and are likely to benefit from the enhanced flow of nutrient-rich sea-water in such regions.
The selective sampling bias of the benthic dredge makes it impossible to determine to what extent the
data in Table 2 can be considered representative. However, the thin covering of fine-grained
glauconitic mud in this area is likely to preclude dense settlements directly on to rock surfaces,
because of the high risk of being smothered or choked by suspensions of sediment stirred up by
bottom currents. Under such circumstances the elevated position resulting from attachment to
mussels would be all the more advantageous for brachiopods, especially as the mussels’ ability
for limited reorientation will ensure that their anterior region remains above the substrate in the event
of sediment accumulations. Nevertheless, direct attachment to substrates such as boulders or rock-
faces is to be expected in the Firth of Lome depression when suitable surfaces are free of sediment due
to inclination or current scour.
Current. Little can be said about the currents in the vicinity of the Firth of Lome population, as direct
measurements of velocity was beyond the scope of this study. However, indirect evidence of strong
bottom currents comes from the fact that the great majority of dead shells are moved away from
the living population and have presumably accumulated in the deepest regions of the depression. This
phenomenon is being investigated with the aim of assessing the fossilization potential of the Firth of
Lome brachiopod population. A strong and constantly flowing current is a necessity for sessile
benthic invertebrates such as T. retusa, which feeds on material carried in suspension in the
surrounding sea-water.
CURRY: BRACHIOPOD TEREBRATULINA FROM SCOTLAND
233
Associated fauna. Apart from T. retusa the surfaces of the Firth of Lome mussels are used as substrate
by a wide variety of sessile organisms and plants, the most numerous of which being the cemented
inarticulate brachiopod Crania anomala (Muller), hydroids, sponges, chitons, spirorbid worms,
foraminifera, the bivalve Anomia, and gastropods (e.g. Cappula). The dredged samples also contained
abundant ophiuroids ( Ophiothrix and Ophiura), and asteroids ( Crossaster , Solaster, and Asterias ) are
often present but in much smaller numbers. Occasionally echinoids ( Echinus ) and decapods ( Munida
and Galathea) have been recovered. As compared with shallow-water localities in the Firth of Lome,
this fauna is impoverished, especially in sessile epibenthic organisms which would be in direct
competition with T. retusa for space and nutrients. Burrowing animals, such as bivalves and
polychaetes, are rare, presumably because of the thin sediment cover. The external shell surfaces of T.
retusa are themselves used as substrate by a wide variety of animals and plants, in particular sponges,
hydroids, bryozoans, and spirorbid worms. Shallow borings, almost entirely confined to the primary
shell layer of T. retusa, are thought to be the work of phoronids and fungi (B. Akpan, pers. comm.
1980). These borings are not predatory, and none of the major predators in this fauna appears to feed
on T. retusa, although it is impossible to assess the effect of all potential predators (e.g. fish). Some
small worms have been found in the brachial cavity of a few specimens of T. retusa, but it is not clear if
these are parasitic or have merely been drawn in by the feeding currents.
REPRODUCTION
It was of great importance to accurately determine the timing and frequency of spawning in T. retusa,
as such information is necessary to establish the population structure and dynamics. In T. retusa the
sexes are separate and can be distinguished on the basis of colour (males are whitish, whilst the tissues
of females are orange-red), and by the examination of the gonads. Males and females are
approximately equally represented (342 adult specimens could be sexed in the sample collected on 26
October 1977, 181 of which were male, and 161 female). One pair of gonads is present in each valve,
developing within the mantle canals between the inner and outer mantle epithelium and extending
posteriorly into the body cavity. Individual gametes develop along narrow interconnecting genital
canals, which are anchored to the outer epithelium by linear membranes (Hancock 1859). Columns of
tissue, situated in the interstices between the genital canals, prevent the gonads from being damaged
or crushed between the two layers of tissue. It seems likely that the gonadal pits which have been
recognized on the internal surfaces of some fossil brachiopods represent the points of attachment of
the supportive columns rather than, as Rudwick (1970) suggests, the points at which the gonad itself
was attached. Rigid calcareous spicules are present in many regions of the body tissues of
T. retusa, but are particularly densely developed above the gonads thereby providing further
protection.
The periodicity of larval recruitment was determined by monitoring the state of development of
gonadal tissues throughout the year (by comparing at least twenty adult specimens from each
sample), and by microscopically examining the surfaces of all potential brachiopod substrate for
recently settled post-larvae. It became clear that the gonads of T. retusa pass through two full
development cycles per year, culminating in the release of gametes in late spring and late autumn.
Recently settled post-larvae have only been observed at these times, and the spawning event must be
highly synchronized, of short duration, and affecting the vast majority of sexually mature
individuals. For example, the entire reproductive cycle, from spawning through to spatfall, was
completed within the 3-week period between samples collected in early and late May 1977. The
precise timing of the natural spawning event must be related to temperature, as T. retusa spawns at a
temperature of 10-1 1 °C in both autumn and spring. A similar temperature control on the timing of
spawning activity has been recognized in both articulate (Rickwood 1968) and inarticulate
brachiopods (Paine 1963; Yatsu 1902).
To investigate the spawning behaviour in greater detail, ripe individuals collected immediately
prior to one of the autumn spawning periods were induced to spawn in the laboratory by rapid
controlled fluctuations of temperature and the addition of mobile mature sperm teased from males.
234
PALAEONTOLOGY, VOLUME 25
Temperature variation is a common method of inducing spawning in marine invertebrates
(P. Redfern, pers. comm. 1978) because the resulting stress stimulates the release of gametes. This
method had previously been used by the author and Dr. J. Richardson in a successful attempt to
induce spawning in the Recent New Zealand brachiopod Terebratella inconspicua (Sowerby). It
seems that synchronization of spawning in neighbouring specimens depends upon the presence of
male sperm, or perhaps hormones secreted along with it, in the surrounding sea-water; it was
noticeable, both with T. inconspicua and Terebratulina retusa, that spawning commenced shortly
after suspensions of sperm had been drawn into the brachial cavity by feeding current. Once
spawning is initiated the ova and sperm are moved from the gonads to the brachial cavity via the
metanephridia, and then extruded from the valves by a series of snapping movements similar to those
used during the ejection of faecal or pseudofaecal pellets (Rudwick 1970).
Fertilization occurs both in the brachial cavity of females and on the surrounding substrate, and a
high degree of synchronization is obviously vital as the sperm remain in suspension and is rapidly
carried away from the breeding population by bottom currents. The ova, by contrast, settle in dense
clusters around the margin of the parent. Only a small proportion of ova ejected during the
laboratory experiment were fertilized, although in nature the proportion may be greater. Both
fertilized and non-fertilized ova appeared to be particularly susceptible to bacteriological attack,
although once again this may have been due to atypical laboratory conditions, in particular
unavoidable increases in sea-water temperature due to the heat from the microscope lamps. The free-
swimming larvae which developed from ova fertilized in the laboratory appeared to be very similar in
size, shape, and activity to those of T. septentrionalis (Couthony) from the coast of Maine (Morse
1873) although no attempt was made to study them in detail. The laboratory-reared larvae of T.
retusa had reached an advanced stage of development (with a third peduncular segment indicating
settlement was imminent) within 5 days of spawning, which is in keeping with the observed maximum
duration of 3 weeks in nature.
During one of the natural spawning periods, attempts were made to collect pelagic larvae from the
Firth of Lome using a plankton net trawled at the surface and at an estimated depth of 100 m. No
brachiopod larvae were recovered, indicating perhaps that the larvae remain close to the sea-floor
during their pelagic stage. However, the evidence for this is not conclusive, especially as the pelagic
stage appears to be of at most a few days’ or perhaps hours’ (see below) duration, and it was
impossible to conduct a comprehensive plankton study for the expected duration of the spawning
period. Some pelagic larvae are known to actively swim towards light sources during their early
development stages (e.g. Paine 1963), and certainly the larvae of the abyssal inarticulate Pelagodiscus
atlanticus (King) has been recovered at the surface (Ashworth 1915). An alternative explanation for
the lack of brachiopod larvae in the plankton samples is that the larvae are brooded to an advanced
stage of development within the brachial cavity of adult females, thereby further curtailing the
duration of the free-swimming stage. Such a phenomenon has been observed in T. septentrionalis
from the Bay of Fundy, Canada (Webb et al. 1976), and is of particular significance as no special
brood pouches are developed. Instead the brooded larvae are simply held between the filaments of the
lophophore and the body wall. In the Firth of Lome the disturbance caused by the dredging process
would almost certainly cause the release of any brooded larvae within T. retusa and hence it will be
difficult to determine if such a phenomenon is widespread. However, the space available for brooded
larvae within the brachial cavity of T. retusa, assuming a mechanism similar to that of T.
septentrionalis, would be insufficient to hold all ova shed by one individual during a spawning period,
and therefore a proportion must still be extruded from the brachial cavity.
The density of larval settlement on the Firth of Lome mussel beds following each spawning event is
remarkable. The number and size of brachiopods attached to each mussel collected on 24 May 1977
was recorded, and the percentage of mussels with attached brachiopods from each year-class was
plotted (text-fig. 3). Representatives of the ‘0’ year-class (which in this sample had settled within the 3
weeks prior to collection) and the 1st year-class (the two 1976 cohorts) were both present on more
than 50% of the mussels in the sample, which represents a considerably greater proportion of the
available substrate, as the mussels live in dense clusters with the lowermost specimens being less
CURRY: BRACHIOPOD TEREBRATU LIN A FROM SCOTLAND
235
text-figure 3. Percentage of the total number of
specimens of Modiolus modiolus (Linnaeus)
collected on 24 May 1 977 which had brachiopods
from each of the seven year classes of brachipods
attached. The sample yielded 127 mussels and 554
brachiopods; a total of 21 mussels ( = 16-5%) had
no attached brachiopods.
YEAR-CLASS
accessible for brachiopod larvae. The fact that fewer mussels had ‘0’ year-class than 1st year-class
brachiopods attached is probably due to the difficulty in picking out the former, and the trend of the
graph would suggest that more than 50% of the mussels are covered at a spawning event. As many as
thirty brachiopods of different ages have been found on a single mussel, and on average each mussel
in a sample has 3-4 brachiopods attached. The density of settlement provides further indirect
evidence of a short pelagic larval stage, as the degree of dispersion increases with the length of the
free-swimming stage. Both autumn and spring spawnings appeared to be equally successful, although
there was no practicable method of comparing the numbers of larvae produced.
POPULATION STRUCTURE AND DYNAMICS
Problems such as biased sampling, selective preservation, and post-mortem transportation have
bedevilled attempts to interpret fossil population structure and dynamics by means of size-frequency
distributions. Size-frequency distribution does, however, have considerable potential as an analytical
tool in palaeontology, provided the fundamental problem of obtaining a representative sample can
be overcome. The simplest and most effective method of checking for ‘representativeness’ is to collect
a series of large samples from the population in question and to compare the resulting size-frequency
distributions. Providing adequate care has been taken to eradicate recurring sampling inadequacies,
similar results from a series of samples can be considered as good evidence that such results are an
accurate reflection of the situation in the parent population. The ability to rationalize the recurring
pattern of size-frequency distribution within the limits of theoretical population biology can be
regarded as a further useful check on the validity of such results. In this study both approaches
provided strong evidence that the regular samples from the Firth of Lome population were
representative, and certainly the favourable situation of the T. retusa population makes sampling
bias unlikely, and obviously the distortion resulting from preservational bias is not pertinent to
studies of living animals. The number of specimens in a sample is certainly a critical factor, however,
and the necessity of having a large sample has been stressed by several authors (see Hallam, in
discussion at end of Craig and Oertel 1966). The density and abundance of brachiopods at the Firth
of Lome locality, combined with their favourable situation for sampling, facilitated the collection of
acceptably large samples.
Further problems have arisen because of confusion over the descriptive terminology applied to
size-frequency diagrams, which warrants mention for the sake of clarity. In conventional statistical
236
PALAEONTOLOGY, VOLUME 25
usage the term ‘skewed’ refers to the gradually sloping ‘tail’ of an assymetrical distribution (Simpson
et al. 1960), and therefore a unimodal skewed distribution which slopes away gradually on the right-
hand side (e.g. text-figs. 4, 6) would be described as right-skewed, or positively skewed. A left-skewed,
or negatively skewed, distribution slopes gradually from the left-hand side to a prominent mode on
the right-hand side of the diagram. This convention has been reversed by some authors, who relate
‘skewed’ to the mode rather than the ‘tail’ (e.g. Raup and Stanley 1978; Thayer 1978). Throughout
this study the more conventional usage has been adopted.
The over-all shape of a size-frequency distribution reflects the relative proportion of animals of
different age-groups, and is an important diagnostic feature. All samples analysed during this study
yielded a unimodal right-skewed length-frequency histogram (e.g. text-fig. 4) with sexually immature
juveniles forming considerably more than 50% of the sample at all seasons. For example, 71% of the
81 1 specimens collected in March 1977 were less than 9-6 mm in length (i.e. not more than 2 years
old — see below). This predominance of juveniles is not unexpected, as the long-term success of any
brachiopod population depends upon the regular influx of large numbers of larvae. The population
structure of T. septentrionalis from the Bay of Fundy, Canada, is also unimodal and right-skewed at
all seasons (Noble et al. 1976), although, just as in T. retusa, the position of the main mode varied
slightly depending on the average size of the most recently settled cohort.
Significantly there is no over-all bimodality discernible in any of the length-frequency histograms
of T. retusa. A bimodal size-frequency distribution is characteristic of some living brachiopod
populations (e.g. Rudwick 1962; Doherty 1976, 1979), and it is now generally accepted that the
secondary (right-hand) mode forms as a result of the merging of older age-groups because of the
gerontic slowing or cessation of growth. Such a growth strategy (known as determinate growth
(Doherty 1976) because adults continue to survive without growing after attaining their maximum,
or determinate, size) is very different to that of T. retusa , which continues to grow throughout life.
When examined in greater detail, it is apparent that the length-frequency histograms of T. retusa
are characterized by regularly spaced subsidiary peaks (e.g. text-fig. 4). Once the biannual spawning
season and its timing had been determined by direct observation, it was then possible to interpret
these peaks in terms of settlement cohorts. The underlying principle on which this analysis is based is
that each of the peaks represents a cohort of brachiopods which settled on the mussel beds after one
table 3. Analysis of the March 1977 length-
frequency histogram (Fig. 4). All measure-
ments are in mm.
Annual Biannual Date of Year-
Increment Peak Increment Settlement Class
Autumn 1976 la
Spring 19 76 1 b
Autumn 1975 2 a
Spring 1975 2 b
Autumn 1974 - 3a
Spring 1974 3 b
...1973 4
---1972 5
---1971 6
---1970 7
^^2 .75-
4<<%4 25'
4<%>.75<
5<%J8 25'
;.5<%10.25-
3 %^.75,
*<2L570 '
text-figure 4. Length-frequency histogram of T. retusa,
March 1977 sample (ZB3727-ZB3736). la, b, etc., refer to
settlement cohorts (see text).
CURRY: BRACHIOPOD TEREBRATULINA FROM SCOTLAND
237
text-figure 5. Width-frequency (a) and height-frequency (b) histo-
gram of T. retusa, March 1977 sample (ZB3727-ZB3736). 1 a, b, etc.,
refer to settlement cohorts (see text).
of the biannual reproductive events. Such a technique has been widely used, and is a logical
assumption provided that the animals have a sharply defined breeding season, and that the average
growth-rate of individual cohorts is such that successive peaks do not merge (at least in the early
stages of life). The majority of these subsidiary peaks can be recognized in all histograms, but the
March 1977 length-frequency histogram (text-fig. 4) has been selected as a standard and will be
analysed in detail. It can be demonstrated that the width-frequency histogram (text-fig. 5a), and to a
lesser extent the height-frequency histogram (text-fig. 5b), yield essentially the same results as the
length-frequency histogram, but in T. retusa the maximum incremental increase in shell dimensions
occurs anteriorly (i.e. in length) throughout ontogeny, and hence the measured shell width and height
dimensions are of smaller absolute value with the result that the resolution between peaks in the
frequency distribution is reduced. An alternative method of analysing length-frequency data, in
which the data is plotted on probability paper, is described in the Appendix.
The March 1977 sample was collected approximately 2 months prior to the spring 1977 spawning
period, and therefore the first two peaks on the left-hand side of text-fig. 4 (at 2-75 mm and 4-25 mm)
correspond to the autumn and spring cohorts in 1976, and are therefore labelled 1(a) and \(b)
respectively in text-fig. 4 and Table 3. Similarly the paired peaks at 6-75 mm (2a) and 8-25 mm (2b)
represent the two cohorts which settled in autumn and spring 1975 respectively; the autumn and
spring cohorts in 1974 are likewise clearly represented by twin peaks at 10-25 mm (3a) and 1 1 -75 mm
(3b). Each of these pairs of peaks have identical spacing (i.e. 1 -5 mm— ‘Biannual’ Increment in Table
3), and the interval between cohorts which settled 1 year apart remains remarkably constant (i.e.
approx. 4 mm— Annual Increment in Table 3). Clearly the rate of growth of T. retusa remains
constant throughout the first 3 years of life, at least within the limits of resolution of this method of
mensuration. Another striking feature of this analysis is that each of the biannual spawning periods
in the previous 3 years has been highly successful, indicating that this relatively deep-water
population is not subject to environmental disturbances known to cause aperiodic disruption to the
238
PALAEONTOLOGY, VOLUME 25
text-figure 6. Length-frequency histograms of T. retusa a, 5 May 1977 sample
(ZB3737-ZB3746). B, 24 May 1977 sample (ZB3747-ZB3756). c, January 1979 sample
(ZB3717-ZB3726). D, July 1977 sample (ZB3757-ZB3766). E, August 1977
sample (ZB3767-ZB3776).
CURRY: BRACHIOPOD TEREBRA TULIN A FROM SCOTLAND
239
reproductive cycle of shallow-water marine invertebrates. As mentioned above, the spawning events
observed subsequently during 1977-1979 appeared to be equally successful.
The position and spacing of the remaining peaks in text-fig. 4, indicates that the growth-rate
progressively decreases from the third year of life onwards, resulting in the merging of the paired
peaks to form a single broad peak representing the biannual cohorts in 1973 (14-73 mm— labelled 4 in
text-fig. 4 and Table 3) and in 1972 (17-25 mm— labelled 5). There is no peak corresponding to the
1971 cohorts in text-fig. 4, but such a peak is present at 19-75 mm in the 5 May 1977 sample (text-fig.
6a) and such a value is included in Table 3 for the sake of completeness (as described below there is
very little movement of peaks between March and May). Ambiguity over the position of the peaks
corresponding to the older age-groups is not surprising considering the rarity of specimens— the
grouping tentatively identified as representing the 1970 cohorts in text-fig. 4 (i.e. 21-75 mm— labelled
7) includes a mere two specimens out of a total sample of 81 1 . However, similar groupings of slightly
more specimens do occur in other samples, and its interpretation as a 7th year-class can also be
justified on the basis of the spacing between previous peaks in text-fig. 4 (see Table 3).
table 4. T. retusa age-groups in the March 1977
sample; size range refers to maximum shell length
measured in mm.
Age Year of Size
Group Settlement Range
1 1976 up to 5.5
2 1975 5.6 - 9.5
3 1974 9.6- 13-5
4 1973 13.6 - 16.0
5 1972 16.1- 18.5
6 1971 18.6 - 20.0
7 1970 over 20.0
No. of °/0 of °/ o Rate of
Specimens Total Mortality
365 45
213 26 42
121 15 43
63 8 45
40 5 40
7 09 83
2 0.2 71
Following on from this analysis it was possible to divide the March 1977 sample into age-groups
(= year-classes) on the basis of a range of lengths (Table 4). The boundaries of the age-groups are
somewhat arbitrary, and the higher and lower values in each grouping include specimens which
rightly belong in the age-groups on either side. Nevertheless, the exercise is useful as it quantifies the
relative proportions of age-groups and provides an estimation of mortality-rates.
Comparison with theoretical model. One of the most significant contributions to the study of
population structure and dynamics in recent years has been the comprehensive theoretical
simulations of Craig and Oertel (1966). Having identified five main factors which influence the over-
all shape of a size-frequency distribution, Craig and Oertel programmed a computer to produce an
exhaustive series of histograms using various combinations of these five factors. Having established
the characteristic population structure of T. retusa in nature, it was obviously of interest to identify
which of Craig and Oertel’s experiments was conducted using a combination of factors similar to
those prevailing in the Firth of Lome population, and to compare the theoretical simulation with the
actual population structure.
Craig and Oertel’s five factors are (1) recruitment strategy, (2) growth-rate, (3) coefficient of
variation of growth-rate, (4) mortality-rate, (5) cessation of growth. T. retusa clearly falls within the
‘boreal’ recruitment strategy of Craig and Oertel, which was defined as two short spawning periods in
autumn and spring. The growth-curve of T. retusa is intermediatory between the ‘linear’ and the
‘high-to-low’ curves used by Craig and Oertel, but is closer to the former than the latter. The third
factor, the coefficient of variation of growth-rate, was an attempt to allow for the effects of the
varying growth-rates of individuals in a single cohort. There is no indication of the extent of such
variability in T. retusa, but Craig and Oertel mostly used a coefficient of 2, which they believed to be
an acceptable average value. The mortality-rate of T. retusa remains constant, at least during the first
5 years of life (see below). The final factor related to the timing and duration of any cessation of
growth during the year, and there is good evidence that T. retusa falls within Craig and Oertel’s
category (a), namely a winter stoppage of 3 months’ duration (see below).
This permutation of factors was combined in Craig and Oertel’s experiment number 34, and the
resulting size-frequency histogram (Craig and Oertel, 1966, fig. 17, p. 346) is strikingly similar to the
length-frequency histograms of T. retusa. The over-all shape of all these distributions is identical.
240
PALAEONTOLOGY, VOLUME 25
being unimodal and right-skewed. The computer simulation had evenly spaced peaks corresponding
to the biannual cohorts, which merge into a single annual peak in later life, exactly as in the natural
population. Craig and Oertel described this experiment as follows:
Boreal recruitment, three winter months cessation of growth with doubling of mortality, coefficient of
variation 2 . . . growth-rate linear . . . mortality constant.
Boreal recruitment consists of two equal waves in late spring and early autumn, separated by a short summer
interval and a long winter interval. This forms twin peaks in the living population ... the groups of twin peaks are
equidistant, and the twins have identical intervals. This peak spacing is diagnostic for linear growth.
Obviously there are differences in detail between the theoretical and the actual, especially as the
critical factors in nature tend to vary slightly with increasing age rather than remain constant.
Nevertheless, the results of this comparison are very encouraging, and demonstrate the potential of a
combined theoretical and empirical approach to the study of population structure and dynamics.
Annual Biannual Year-
Increment Peak Increment Class
table 5. Analysis of the peak spacing in the width-
frequency histogram of Rostricellula rostrata,
Ulrich and Cooper (data, in mm, are from Walker
and Parker 1976).
Relevance for fossil populations. The analysis of the population structure and dynamics of fossil
brachiopods is best conducted using both size-frequency histograms and growth-line counts, as the
inherent deficiences of one method are compensated for by the strengths of the other. However, under
ideal conditions of preservation, a frequency histogram prepared from a sample of fossil brachiopods
can, on its own, yield precise data on the autecology of that species. In preparing a width-frequency
histogram from a sample of the Middle Ordovician species Rostricellula rostrata, Ulrich and Cooper
from Tennessee, Walker and Parker (1976) used the same grouping adopted for the T. retusa
histograms (i.e. 0-5 mm), and hence the prominent peaks in the R. rostrata histogram can be analysed
in the usual tabular form (i.e. Table 5) allowing direct comparison with the living population (i.e.
Table 3).
This method of analysis emphasizes the remarkable similarity in the population structure between
the living and fossil population, especially in the pattern of peak spacing. Both histograms have
prominent regularly spaced twin peaks, with the separation between each peak in a twin (i.e. the
‘biannual increment’— see Tables 5 and 3) being identical in both populations (i.e. 1 -5 mm). Just as in
the T. retusa histogram, these twin peaks are not discernible on the right-hand side of the diagram,
because the slowing of the growth-rate in older specimens has resulted in the merging of the twins to
form a single peak representing the total annual recruitment (e.g. at 14-25 mm in Table 5). There are
differences in the absolute values of comparable increments (e.g. annual increments— compare
column 1 in Tables 5 and 3) and the fossil species had a shorter life span and a more rapidly decreasing
growth-rate, but such differences are of minor significance compared to the similarity in the over-all
pattern of peak spacing. Both theoretical and empirical data indicate that this pattern of regularly
spaced twin peaks is characteristic of animals inhabiting temperate latitudes, and the fossil
population can reasonably be assumed to have experienced broadly similar seasonal cycles of
temperature, food supply, etc., to those prevalent today in the Firth of Lome or other temperate
habitats. Apart from comparison with equivalent living populations, the reconstruction of
population structure and dynamics in fossil brachiopods would, under ideal circumstances, be based
on the analysis of several large samples rather than just one, and would also take into consideration
pertinent localized environmental and biological factors; as mentioned above, the interpretation of
growth-lines can further refine the resulting data. Nevertheless, the success of this preliminary
comparison between the population structure of living and fossil brachiopods augers well for this
technique.
CURRY: B R ACHIOPOD TEREBRATU LIN A FROM SCOTLAND
241
Seasonal growth pattern. To obtain a more precise picture of the pattern of seasonal growth in T.
retusa, the position of peaks corresponding to individual cohorts was recorded in samples collected
throughout the year. The results (Table 6) suggest that the growth-rate of adults decreases
significantly in winter months, as the peaks appear to remain stationary between January and May.
table 6. Movements of peaks corresponding to the 4th
and 5th year-classes of T. retusa (measurements, in mm,
refer to maximum shell length).
Age Expected
Group Mar May Aug Oct Jan in Mar
4 14.75 14.75 16.25 17.25 1725 17.25
5 17.25 17.25 18.25 18.75 19.75 19.75
The evidence for this is based on the assumption that the mid-point of a particular peak is an
acceptable measure of the average length of specimens in a cohort; although the growth-history of
individuals within a cohort will vary depending on localized environmental conditions. The slowing
of growth in winter is an expected and readily explained phenomenon, bearing in mind the direct
relationship between decreasing temperature and a reduction of the rate of metabolic activity in cold-
blooded invertebrates. It seems likely that the rate of growth of T. retusa slows progressively during
the autumn and winter, and reaches a minimum, or perhaps ceases altogether, in mid-winter when
temperatures are lowest and food supplies minimal. There is some evidence that sexually-mature
females have a slightly slower rate of growth than mature males (text-fig. 7a), perhaps indicating that
the production of ova is more demanding in terms of available nutrients. However, the pattern of
modes in text-fig. 7a is rather confused and far from unequivocal, and this facet of sexual dimorphism
does not appear to be noticeable in the combined length-frequency distributions. The increments
involved, like those produced by variable growth-rates in a single cohort, are so small as to be outside
the limits of resolution of the length-frequency histogram.
In contrast, the growth of recently settled post-larvae appears to proceed rapidly during the first 3
months of settled life, and to be independent of prevailing environmental conditions. For example,
the cohort which settled in late autumn 1976 had attained an average length of 2-75 mm by the
following March (i.e. \{a) in text-fig. 4), a rate of growth greater than at any other stage of life.
Similarly, the cohort which settled in May 1977 had grown to an average length of 2-75 m by the
following August (compare text-figs. 6b, 6e). The growth-history of T. retusa, therefore, has three
distinct phases: (1) approximately 3 months of rapid growth immediately following settlement in
both autumn and spring; (2) the remainder of the first 3 years of life, when the annual growth-rate
remains constant, but varies seasonally depending upon ambient sea-water temperatures; (3) the
remainder of life, during which annual growth-rates decrease progressively.
The initial period of rapid post-settlement growth is a prudent feature, as the early development of
a rigid protective exo-skeleton will reduce mortality rates during this most vulnerable stage of settled
life. The ability of the autumn cohorts to grow in winter months indicates perhaps that their modest
nutrient requirements are adequately satisfied by even the reduced winter food supplies. It may also
be of significance that the volume of the feeding ‘chambers’ in juveniles is considerably greater than
that enclosed by the valves of the shell, as the filaments of the lophophore extend a considerable
distance beyond the shell margin when feeding. In adults the filaments are almost entirely contained
within the brachial cavity when feeding; in juveniles the increased risk of predation resulting from the
exposed filaments and widely gaping shell (as much as 45° as compared to « 15° in adults) may be
offset by the advantages gained by the relatively rapid shell growth due to the proportionately high
levels of nutrient intake. The reduction in the rate of annual growth-rate at the end of 3 years
coincides with the onset of sexual maturity, and probably reflects a fundamental change, with the
nutritional requirements of the developing gonads receiving precedence over other metabolic
processes.
Growth-line analysis. The analysis of the pattern of growth-line formation in brachiopods is best
conducted on a cumulative basis using a large number of individuals from a single population. The
main difficulty of attempting a growth-line analysis of an individual is that the pattern is often
incomplete, and hence difficult to interpret. Nevertheless, a sample analysis of a single adult specimen
242
PALAEONTOLOGY, VOLUME 25
text-figure 7. A, sexual dimorphism in shell growth as determined
from 342 sexually mature specimens of T. retusa collected 26 October
1977 from the Firth of Lome population, b, c, age pyramids for T.
retusa population in the Firth of Lome before (b, from March 1977
sample) and after (c, from 24 May 1977 sample) spatfall. D, Growth
curve for T. retusa as determined from the analysis of the March 1977
sample (i.e. Table 3).
of T. retusa with a complete record of growth-line formation is included at this stage as it contributes
further to our knowledge of the growth history of the population. The specimen (ZB 3717) was
collected during January 1979 at an approximate depth of 165 m from the Firth of Lome, and was
15-5 mm in length. Having measured the distance to each growth-line from the posterior margin of
the shell along the medial axis of the pedicle valve (Table 7), and determined the spacing between
individual growth-lines (column 2 in Table 7), it was then apparent that growth-lines form biannually.
The age-group analysis in Table 4 indicates that a specimen of length 15-5 mm should be 4 years old; a
conclusion which is confirmed by growth-line analysis which indicates that the 1st growth-line on this
specimen was formed in autumn 1975 (Table 7) and therefore it must have settled in spring 1975.
The spacing between growth-lines (column 2 in Table 7) on this specimen indicates that
approximately two-thirds of the annual growth occurs during the ‘summer’ period, whilst the
CURRY: BRACHIOPOD TEREBRATULINA FROM SCOTLAND
243
remaining one-third occurs during ‘winter’. The exact timing of growth-line formation has not been
determined, but is assumed to occur in autumn and spring, and at times of pronounced
environmental and physiological stress. If so, the ‘summer’ and ‘winter’ periods would therefore be of
approximately 6 months’ duration, and would probably be more accurately designated as ‘summer/
autumn’ and ‘winter/spring’ respectively.
table 7. Analysis of growth lines on single specimen of T.
retusa (ZB3717) collected from the Firth of Lome on 21
January 1979 (data in mm).
Increment Growth-line Formation
-Autumn 1975
-Spring 1976
-Autumn 1976
-Spring 1977
- Autumn 1977
-Spring 1978
- Autumn 1978
-Spring 1979
Mortality. The data in Table 4 indicates that roughly 40% of the brachiopods in each age-group die
per year. This estimation applies only to specimens more than 1 year old, as there is no viable method
of determining the mortality rate during the periods prior to and immediately following settlement.
Significantly higher mortality-rates are likely to occur during these early stages of life, as has been
determined in other living brachiopod populations (e.g. Doherty 1976, 1979). As the number of
specimens in the older age-groups is so small, the significant increase in mortality-rates amongst the
6th and 7th age-group (Table 4) may not necessarily be representative, although older specimens may
indeed be more susceptible to disease, stress, etc.
The causes of mortality in the T. retusa population are not apparent, although it was not possible to
examine large numbers of dead shells for signs of predation. There is, however, very little evidence of
repaired shell damage in living specimens, which suggests that the level of predation is low. Probably
a large proportion of deaths occur in winter because of the stress caused by less favourable
environmental conditions. It may be significant that Craig and Oertel (1966) incorporated a doubling
of the mortality rate in winter in the experiment which yielded a size-frequency distribution similar to
that of T. retusa. The fact that a few adult specimens did not develop gonads during spawning periods
may be symptomatic of diseases or infections which may account for a small proportion of the annual
mortality. However, as gonad development is discernible during winter months, it would appear that
starvation is not a major cause of death.
The small tissue content of brachiopods probably partially explains the apparent lack of predators
on T. retusa , although predatory gastropod borings have been found in other species (e.g. Logan
1979). Because of their low nutritional value, it may indeed be more reasonable to look for potential
predators at the ‘micro’ rather than the ‘macro’ level, and certainly carnivorous micro-organisms
could readily gain access to the brachial, and perhaps body, cavity via the feeding currents. Some
nematoid worms do appear to feed on brachiopod lophophoral filaments (McCammon 1971)
although there has been no detailed study of the extent and effect of such predation/parasitism.
CONCLUSIONS
The large population of T. retusa in the Firth of Lome is well established, and the species is relatively
abundant in deeper waters off the west coast of Scotland. The main reasons for this success are the
recurring efficiency of the reproductive activity and the absence of competing organisms or readily
apparent predators. Representatives of the genus Terebratulina have a long history, and have been
present in the North Atlantic since its inception. It survives as one of the most cosmopolitan and
abundant of living brachiopod genera, occurring in all oceans. The longevity and success of this
genus are probably due to an adaptability of pedicle morphology which greatly increases its range of
substrates, combined with an ability to colonize habitats (such as the Firth of Lome depression)
inimical to other more ‘successful’ epibenthonic organisms. Critical factors in this latter capability may
244
PALAEONTOLOGY, VOLUME 25
be a generally low level of nutrient requirements, and the ability to survive prolonged periods of
adverse conditions by a virtual cessation of metabolic processes, as demonstrated by the specimens
which survive in the outside aquarium at Dunstaffnage despite being subjected to rapid daily
temperature fluctuations. Clearly another important factor in the longevity and success of the genus
is its remarkable morphological conservatism, with Cretaceous and Recent species being almost
indistinguishable. T. retusa is unlikely, however, to reassume the role of its ancestors as a common
constituent of shallow marine ecosystems because of its inability to compete for available space in
such environments.
Acknowledgements. I am indebted to my joint supervisors Dr. C. H. C. Brunton and Dr. P. Wallace for their
advice, guidance, and encouragement throughout the course of this study. Other staff of the British Museum
(Natural History) also contributed greatly; in particular I would like to acknowledge Dr. L. R. M. Cocks,
Mr. E. F. Owen, Mr. A. Rissone, Dr. M. K. Howarth, Mrs. H. Brunton, Miss L. Cody, and Mrs. P. P.
Hamilton-Waters. Dr. A. Ansell kindly supervised my work at the Oban Laboratory, and I would also like to
thank Mr. C. Comley, Mrs. L. Robb, Mr. and Mrs. R. Harvey, Mr. S. Knight, and the Captains and crews of the
R/V Calanus and Seol Mara for their willing co-operation. I am also grateful to Dr. A. Williams for his interest
and support. The work was carried out during the tenure of a Department of Education (Northern Ireland)
Postgraduate Research Studentship, which is gratefully acknowledged.
REFERENCES
ashworth, j. h. 1915. On the larvae of Lingula and Pelagodiscus (Discinisca). Trans. R. Soc. Edin. 51, 45-69.
atkins, d. 1959a. The growth stages of the lophophore of the brachiopods Platidia davidsoni (Eudes
Deslongchamps) and P. anomioides (Phillippi) with notes on the feeding mechanism. J. mar. biol. Ass. U.K. 38,
103-132.
19596. The early growth stages and adult structure of the lophophore of Macandrevia cranium (Muller),
(Brachiopoda, Dallinidae). Ibid. 335-350.
— 1960a. A new species of Brachiopoda from the Western Approaches, and the growth stages of the
lophophore. Ibid. 39, 71-89.
— 19606. A note on Dallina septigera (Loven), (Brachiopoda, Dallinidae). Ibid. 91-99.
— 1 960c. The ciliary feeding mechanism of the Megathyridae (Brachiopoda) and the growth stages of the
lophophore. Ibid. 459-479.
— 1961. The growth stages and adult structure of the lophophore of the brachiopods Megerlia truncata (L.)
and M. echinata (Fischer and Oehlert). Ibid. 41, 95-111.
barber, p. l., dobson, m. r. and Whittington, R. J. 1979. The geology of the Firth of Lome, as determined by
seismic and live sampling methods. Scott. J. Geol. 15, 217-230.
brunton, c. h. c. and curry, G. b. 1979. British brachiopods Synopses of the British Fauna (New Series), No. 17,
64 pp. Linnean Society and Academic Press, London.
cerrato, m. 1980. In rhoads, D. c. and lutz, R. a. (eds.). Skeletal growth of aquatic organisms, pp. 417-465,
Plenum Press.
chumley, j. 1918. The fauna of the Clyde sea area. Glasgow Univ. Press. 200 pp.
comley, c. A. 1978. Modiolus modiolus (L.) from the Scottish west coast. I. Biology. Ophelia , 17, 167-193.
craig, G. Y. and oertel, G. 1966. Deterministic models of living and fossil populations of animals. Q. Jl geol.
Soc. Lond. 122, 315-355.
davidson, t. 1886-1888. A monograph of Recent Brachiopoda. Trans. Linn. Soc. Ser. 2, 4, pts. I— III.
doherty, p. j. 1976. Aspects of the feeding ecology of the brachiopod Terebratella inconspicua (Sowerby).
Unpublished M.Sc. thesis, Univ. of Auckland.
— 1979. A demographic study of a subtidal population of the New Zealand articulate brachiopod
Terebratella inconspicua. Mar. Biol. 52, 331-342.
Hancock, a. 1859. On the organisation of the Brachiopoda. Phil. Trans. R. Soc. (for 1858), 148, 791-869.
Harding, J. p. 1949. The use of probability paper for the graphical analysis of polymodal frequency
distributions. J. mar. Biol. Ass. U.K. 28, 141-153.
logan, a. 1979. The Recent Brachiopoda of the Mediterranean Sea. Bull, de Tlnstitut oceanograph., Monaco, 72,
no. 1434. 112 pp.
mccammon, h. m. 1971 . Behaviour in the brachiopod Terebratulina septentrionalis (Couthouy). J. Exp. mar. biol.
Ecol. 6, 35-45.
CURRY: BRACHIOPOD TEREBRATULINA FROM SCOTLAND
245
moore, h. b. 1958. Marine ecology. John Wiley & Sons, New York.
morse, E. s. 1873. Embryology of Terebratulina, Mem. Bos. Soc. nat. Hist. II, 249-264.
noble, j. p. A., logan, A. and webb, G. r. 1976. The Recent Terebratulina community in the rocky subtidal zone
of the Bay of Fundy. Lethaia, 9, 1-17.
paine, R. t. 1963. Ecology of the brachiopod Glottidia pyramidata. Ecol. Monog. 33, 255-280.
prosser, c. l. 1973. Comparative animal physiology (3rd edn.). W. B. Saunders, Philadelphia.
raup, d. m. and Stanley, s. m. 1978. Principles of paleontology (2nd edn.). Freeman & Company, San Francisco.
481 pp.
rickwood, a. e. 1968. A contribution to the life-history and biology of the brachiopod Pumilus antiquatus ,
Atkins. Trans. R. Soc. N.Z. (Zool.), 10, 163-182.
rudwick, m. j. s. 1962. Notes on the ecology of brachiopods in New Zealand. Ibid. 1, 327-335.
1970. Living and fossil brachiopods. Hutchinson, London. 199 pp.
simpson, G. G., roe, A. and lewontin, r. c. 1960. Quantitative zoology (revised edn.). Harcourt, Brace & Co.,
New York. 440 pp.
thayer, c. w. 1975. Size-frequency and population structure of brachiopods. Palaeogeog., Palaeoclim .,
Palaeoecol. 17, 139-148.
walker, K. R., and Parker, w. c. 1976. Population structure of a pioneer and a later stage species in an
Ordovician ecological succession. Paleobiol. 2, 191-201.
webb, G. R., logan, a. and noble, j. p. a. 1976. Occurrence and significance of brooded larvae in a Recent
brachiopod. Bay of Fundy, Canada. J. Paleont. 50, 869-870.
wesenberg-lund, e. 1940. Brachiopoda in the zoology of east Greenland. Meddr Gron. 121 (5), 12 pp.
1941. Brachiopoda. Danish Ingolf expedition, 4, (12), 17 pp.
yatsu, n. 1902. On the habits of the Japanese Lingula. Annot. Zool. Japan, 4, 61-67.
G. B. CURRY
Department of Geology
Original typescript received 14 October 1980 ^he University
Glasgow
Final typescript received 10 January 1981 G12 8QQ
APPENDIX. The use of probability graph paper for the analysis of polymodal length-frequency histograms
By way of an additional check on the interpretation of the population structure of T. retusa, the length-frequency
data used to construct text-fig. 4 was also plotted on probability graph paper (text-fig. 8). By this means it is
0.01 0.1 1 10 50 90 99 99 9 99 99
CUMULATIVE PERCENTAGE — >
text-fig. 8. Length-frequency data collected 22 March 1977 (« = 811)
plotted on probability graph paper.
246
PALAEONTOLOGY, VOLUME 25
possible to check that text-fig. 4 is indeed composed of a series of more or less overlapping normal distributions
corresponding to individual settlement cohorts. Harding (1949) was the first to realize that the ability of
probability paper to pick out individual normal distributions in a polymodal distribution would be of great use
to population biologists, but few workers have made use of this method (Cerrato 1980). The methodology of this
method of analysis is straightforward. The data is plotted as a cumulative percentage (i.e. % of total sample less
than x mm) on a non-linear horizontal scale, which is so arranged that a plot of points corresponding to a normal
distribution yields a straight line. As Harding (1949) demonstrated, a polymodal distribution plots out as a series
of straight lines corresponding to each of the constituent normal distributions. Although there is no information
as to the size distribution of animals in each settlement cohort in the Firth of Lome, it is a reasonable and widely
accepted assumption that each cohort would plot out as an essentially normal distribution. By this method of
analysis, therefore, it is possible to check that each peak in text-fig. 4 does represent a settlement cohort (or
amalgamation of two cohorts in the adult specimens) as suggested in the main text.
When the March 1977 data is plotted in this manner the results (text-fig. 8) confirm the interpretation of
population structure outlined in the text. Short discrete lines are clearly distinguishable amongst juvenile
specimens, and correspond to biannual cohorts. The degree of differentiation is less amongst 3- and 4-year-old
specimens, which is consistent with growth-rates decreasing following the onset of sexual maturity. The
suggestion that biannual cohorts coalesce in later life to form annual peaks is confirmed by text-fig. 8, and the
pattern of growth amongst adults is more discernible than in the original length-frequency histogram. For
example, a 6th year-class is clearly marked in text-fig. 8, but could only be inferred from text-fig. 4 (i.e. Table 3).
However, the clearly differentiated 7th year-class in text-fig. 8 must be considered as artificial, as the gap in the
histogram before the last two specimens automatically results in a strong differentiation from the preceding
groups of specimens.
A NEW ZOSTEROPHYLL FROM THE LOWER
DEVONIAN OF POLAND
by DANUTA ZDEBSKA
Abstract. A new genus and species Konioria andrychoviensis assigned to the Zosterophyllophytina is described
from the Lower Devonian (Emsian) of two boreholes in the Bielsko-Andrychow area of the Polish Western
Carpathians. K. andrychoviensis possesses dichotomous axes covered on their lower part with long subulate
spines and on their upper part with short triangular spines. Apices of axes form hooks. In addition to spines, the
axes show 1-4 longitudinal wings. The reniform to rounded sporangia are borne singly at dichotomies. The
structure of pyritized axes shows a central exarch strand with scalariform tracheids and a hypodermis. In
connection with the unusual position of the sporangia the problem of the evolution of the lycopod sporophyll
is discussed. Konioria appears to suggest that the lycopod sporophyll originated from ends of fertile axes,
in accordance with the Telome Theory of Zimmermann (1930). The Telome Theory, however, is based on
the Rhynia- type of organization, while other evidence suggests that the lycopods originated from the
Zosterophyllophytina.
This paper describes a single species obtained from two bore-holes, Andrychow 2 and Andrychow 4
in the Bielsko-Andrychow area of the Polish Western Carpathians (see Turnau 1974). The depth
in Andrychow 2 is between 2300-8 and 2306-6 m and in Andrychow 4 between 2245-5 and 2250-8 m.
The age of these rocks was determined as Lower Devonian (Emsian) on the basis of their lithology
by Konior (1965, 1966, 1968, 1969) and this was confirmed by the miospores (Konior and Turnau
1973; Turnau 1974).
Fragments of Drepanophycus spinaeformis Goepp. and Dawsonites sp. have been recovered from
the same cores, but have still to be described. Preliminary investigations of the plant material were
carried out by Maria Reymanowna who gave a short description of Konioria under the name of
Psilophyton sp. referred to in Konior (1965). Well-preserved fragments of the plant allow conclusions
to be drawn on its systematic position and on the probable course of evolution of the lycopod
sporophyll.
MATERIAL AND METHODS
The plant axes are preserved as coaly compressions in a dark-grey siltstone. These remain intact when removed
from the matrix with hydrofluoric acid. When macerated in Schulze’s solution, most disintegrated and showed
no cells. Several axes yielded tracheid fragments with circular bordered pits. A few cuticle fragments of axes
are in a state of natural maceration. These are translucent enough to show both a central dark vascular strand
and stomata. Certain other axes were pyritized and were not very compressed. From these, thin sections were
prepared for reflected light microscopy, using a modified method described by Edwards (1968). To prevent
the crumbling of the pyritized axes during sectioning, they were embedded in dental plaster of paris, and
were sectioned using a small dental saw. These sections were fixed to glass slides with Canada balsam and
ground by hand in the usual way.
The photographs on PI. 25, figs, 5, 6; PI. 26, figs. 5, 7, 8; PI. 27, figs. 3, 5, 6, 7, 9, and PI. 28, figs. 2, 3, 8, 9
were taken with the Zeiss Photomicroscope III Stand, and the micrographs on PI. 25, fig. 3; PI. 27, figs. 1, 2, 4;
and PI. 25, fig. 4 with the Cambridge S 600 SEM in the Department of Botany, Birkbeck College, University
of London.
The remaining photographs were taken with an Exacta camera and a Zeiss-Jena lightmicroscope in the
Botanical Institute of the Jagiellonian University, Krakow.
(Palaeontology, Vol. 25, Part 2, 1982, pp. 247-263, pis. 25-28.|
248
PALAEONTOLOGY, VOLUME 25
SYSTEMATIC PALAEONTOLOGY
Order zosterophyllales
Family zosterophyllaceae
Genus konioria gen. nov.
Konioria andrychoviensis sp. nov.
Plant fragments extracted from the rock samples include axes with both long and short spines, axes with only
short or long spines, sterile and fertile apices, and pyritized axes (Table 1).
table 1 . Characters found on separate parts of the plant which indicate that they belong to the same plant (cross
indicates presence of character)
__PLANT PASTS
CHARACTERS —
Axis with
long and
short spines
Axis with
long spines
only
Axis with
short spines
only
Sterile
apex
Fertile
apex
Pyritised
axis
Circinatelly
coiled axis
long spines
with
minute teeth
+
-
-
-
Short spines
-
-
-
*
Wing
+
-
-
+
-
Stomata
-
-
*
-
-
-
Axes with spines. The Konioria axes are preserved on the bedding planes in a dark-grey siltstone. The core is 9 cm
wide which limits the length of the specimens; originally they were longer. There are also many pieces which were
broken into short lengths before deposition.
The axes dichotomize into unequal or almost equal parts diverging at about 30° (PI. 25, figs. 1, 4). The
distance between successive dichotomies is about 4 cm towards the top of the plant. Below dichotomies the axes
widen and gradually change into two branches. One axis was 2 mm wide at 1 cm below its first dichotomy and 1-5
mm at 1 cm below its second dichotomy. The length of this axis is over 6 cm. Other specimens suggest a possible
length of about 15 cm for the largest axes which are 4 mm wide. The length of the plant is unknown but
some estimate is possible by relating the extent of tapering along a single axis to the extremes of axis
diameter (0-3-4 0 mm).
The longest unbranched axis available is 7 cm but most branch at closer intervals than this and near the apex
branching is very close together. Successive dichotomies are not all in the same place but the precise angle
between them in the lower part cannot be given. Near the apex it is about 90°. A characteristic and remarkable
feature of the axes is that they have one to four longitudinal wings. These may lie in the plane of compression (PI.
25, fig. 4) where they appear as borders to the axis, or they may be present on the surface of the axis (PI. 25, figs. 5,
EXPLANATION OF PLATE 25
Fig. 1. Para type, axes with two branches, S/98/ 14; x 1-2.
Fig. 2. Axes with long spines, S/98/12b; x 3.
Fig. 3. SEM. Stem with long and short spines; x 120.
Fig. 4. Axes with ‘compression margin’ (wings), S/98/9a; x 3.
Fig. 5. Surface of axis showing three wings (arrowed) (specimen destroyed); x 126.
Fig. 6. Compressed axis showing wing and a long spine at an acute angle (specimen destroyed); x 6.
PLATE 25
ZDEBSKA, Konioria
250
PALAEONTOLOGY, VOLUME 25
6; PI. 26, figs. 7, 8; PI. 27, fig. 6; PI. 28, figs. 2, 9). These wings are present on tall axes. Occasionally, the wings take
a spiral course (PI. 28, fig. 9) which is not the result of twisting of the axis. Twisted axes are seen on PI. 25, fig. 6.
On wings as well as on the axes, the spines may be dense (PI. 26, fig. 6). The author considered the possibility that
the wings were mere compression borders, but decided they were not, on the evidence given below:
1 . In transverse sections of pyritized axes the wing is continuous with the thick, carbonized layer (pi. 26, fig. 6).
2. Cells of the hypodermal layer below the carbonized layer do not enter the wing (PI. 26, fig. 6).
3. In a transverse section of a flattened pyritized axis the wing is not a continuation of the longer axis of the
section. If the wing were formed as a result of flattening of the plant, it would occur in the same plane
(PI. 26, fig. 6).
4. Often the wing forms a spiral on the axis (PI. 28, fig. 9).
Axes bear spines of varying size and frequency. The spines range from being short and triangular to long and
subulate (PI. 25, fig. 3). On the rock the spines are visible at the sides of the axes, while on their surface only their
bases are seen as large and small dots. The spines are distributed irregularly; rarely are they very frequent on one
surface of the axis and not on the other (PI. 26, figs. 1, 2). The length of the spines is very unequal on different
parts of the axis and range from 0T to 4 mm. The longest spines, up to 4 mm long and 0-3 mm wide at the base are
on the thickest axes, but occasionally there may be fairly long spines on the narrower axes (PI. 26, fig. 3), and
occasionally just one long spine is present (PI. 28, fig. 3). Usually the narrower axes bear shorter spines (PI. 26, fig.
5; text-fig. 1) or occasionally none at all. On some axes (PI. 26, fig. 2) both long and short spines occur. Most
spines are at about 90° to the axis. The longer spines show longitudinal ridges (PI. 27, fig. 2) with occasional
minute teeth 12-16 /xm long (PI. 27, figs. 2, 4). Teeth are not obvious on the shorter spines (PI. 25, fig. 3). The long
spines with teeth are most numerous on some circinately coiled axes (PI. 28, fig. 8) which the author thinks
possibly represent early stages in the development of a shoot (these are not included in the reconstruction). Some
spines show a dark core of unknown cellular structure (PI. 28, fig. 3) suggesting a vascular strand. Such structures
are spines and not sporangial stalks, because they are quite long and situated at the side of the axis and not below
the dichotomies.
Cuticle preparations of axis showed no clearly marked epidermal cells (PI. 25, fig. 3) apart from dark stomata.
The stomata are visible as irregularly distributed elliptical dark dots, because they are usually covered with a
dark substance (PI. 28, figs. 1, 7). They are oval and orientated longitudinally, and sometimes show the guard
cells. The spines have no stomata.
Apex of axis. All apices whether sterile or fertile are much branched and curved to form hooks (text-fig. 1 ; PI. 26,
fig. 5; PI. 27, fig. 6; PI. 28, figs. 1, 2). These illustrations show all the variation observed in both sterile and fertile
apices. The spines on the ends of axes are never long, and are sometimes infrequent or absent (text-fig. 1). The
vascular strand may be visible by scanning electron microscopy but not after maceration. The dark strand in PI.
28, figs. 1, 7, shows an untreated specimen photographed by transmitted light.
Fertile axes. The sporangia cannot be distinguished on axes still in the rock, probably because the diameter of a
sporangium is not much larger than the width of the axis below the dichotomy. They are visible only on axes
removed with hydrofluoric acid. Most of the distal branches are sterile, but in some instances a single sporangium
occurs at the final or penultimate dichotomy, and perhaps also further down. The sporangia are mostly situated
slightly below the angle of the dichotomy and any stalk they possess must be very short and is concealed (PI. 26,
figs. 7, 8; PI. 28, fig. 2; text-figs. 2, 3). It appears that the small mound visible on the surface of most sporangia is
EXPLANATION OF PLATE 26
Figs. 1, 2. Two sides of a pyritized axis, S/98/23; x 17-5. 1, surface with numerous short spines. 2, surface with
bases of two long spines and a few short spines.
Fig. 3. Fragment of axis with long spines and wing, S/98/26; x 20.
Fig. 4. Pyritized axis just below dichotomy (specimen destroyed); x 30.
Fig. 5. Apex showing short spines, S/98/26; x 20.
Fig. 6. Transverse section of pyritized axes. On the left-hand side is a wing with spines, S/98/79; x 25.
Figs. 7, 8. Holotype, fragment of axis with sporangium situated at the level of a dichotomy (both sides),
S/98/31; x 30.
PLATE 26
ZDEBSKA, Konioria
252
PALAEONTOLOGY, VOLUME 25
EXPLANATION OF PLATE 27
Fig. 1. SEM. Axis with two flattened branches and a stalked sporangium (arrows); x 30.
Fig. 2. SEM. Surface of spine showing longitudinal ridges with minute teeth; x 300.
Fig. 3. Paratype, transverse section of uncompressed axis showing elliptical xylem strand, S/98/ 15; x 50.
Fig. 4. SEM. Long spine with round base and minute teeth; x 25.
Fig. 5. Cells of hypodermis in longitudinal section, S/98/18; x 125.
Fig. 6. Forked apex showing wing on under side (specimen destroyed); x 20.
Fig. 7. Protoxylem and metaxylem in longitudinal section showing scalariform tracheids, S/98/17; x 1 10.
Fig. 8. Pyritized axis split longitudinally and showing vascular strand with circular bordered pits,
S/98/49; x 375.
Fig. 9. Metaxylem and protoxylem from fig. 3; x 167.
PLATE 27
ZDEBSKA, Konioria
1 V
254 PALAEONTOLOGY, VOLUME 25
caused by the stalk (PI. 26, fig. 8). Some, however, have a longer stalk which is clearly inserted in the angle of the
dichotomy (PI. 27, fig. 1).
The sporangia are flattened and composed of two equal valves which separate along their entire margin. The
largest are oval or slightly reniform (PI. 26, figs. 7, 8; text-fig. 2) but some are round (PI. 28, fig. 2; text-fig. 2). They
are typically about 2-5 mm wide, but the round ones are smaller and one was only 0-5 mm wide. The surface of
the outer valve may show diverging cells under scanning electron microscopy. Often the surface of a sporangium
bears minute spines (text-fig. 2). Nothing is known of the deeper layers of the sporangial wall. Along the line of
dehiscence the wall is flat and appears thin (PI. 28, fig. 4). Although the two valves are always pressed together the
sporangia seem to have dehisced and shed all their spores. Maceration yielded no spores nor did it yield a central
mass which might represent compacted spores.
Pyritized axes. Ten pyritized axes were studied. They are rather narrow, 0-5-2 mm wide. Some show long spines
(PI. 26, fig. 3), some short spines and long broken spines (PI. 26, fig. 2), and some all short spines (PI. 26, fig. 1).
These specimens were not sectioned, the sections being prepared from unfigured specimens which did show short
spines or bases of long spines.
text-fig. 2. Sporangia attached below the dichotomy. Small spines present
on the surfaces of sporangia.
EXPLANATION OF PLATE 28
Fig. 1. Paratype, naturally cleared apex showing dark core and stomata as dark spots, S/98/29; x 25.
Fig. 2. Apex with sporangium (specimen destroyed); x 30.
Fig. 3. Fragment of axis with spines; large spine apparently with ?vascular strand coming from the axis,
S/98/27; x 20.
Fig. 4. SEM. Fragment of sporangium valve; margin appears to be thin and smooth, but the wall shows
?papillae; x 300.
Fig. 5, 6. Tracheids isolated by Schulze maceration from carbonized axis. Tracheids show circular bordered
pits. 5, S/98/75; x 600; 6; S/98/76; x 546.
Fig. 7. Naturally cleared apices showing stomata as dark spots on the epidermis, S/98/30; x 24.
Fig. 8. Young coiled axis densely covered with long spines, S/98/25; x 20.
Fig. 9. One branch of the dichotomy (the other broken off") showing spiral wing (specimen destroyed); x 19*5.
PLATE 28
ZDEBSKA, Konioria
256
PALAEONTOLOGY, VOLUME 25
The transverse section on PI. 27, fig. 3, shows a black, coalified outer layer which can be regarded as the
epidermis, inside this are about four layers of thick-walled cells which represent the hypodermis. Occasionally
the outer coalified layer is thicker, suggesting that the hypodermal cells are also coalified. In longitudinal section
the hypodermal cells are elongated and show no obvious pits (PI. 27, fig. 5). Length of cells from 1 50 to 450 pm,
width from 30 to 85 pm.
Inside the hypodermis, there is a region in which the cells are scarcely visible, and these surround a circular (PI.
26, fig. 4) or elliptical (PI. 27, fig. 3) xylem strand, with small tracheids (regarded as protoxylem) to the outside
and larger tracheids (regarded as metaxylem) to the inside (PI. 27. fig. 9). The metaxylem shows well-developed
scalariform tracheids (PI. 27, fig. 7). Three different stems showed exactly similar scalariform tracheids with no
round pits. However, a further specimen with the surface fractured longitudinally showed round pits in what
appears to be a metaxylem tracheid (PI. 27, fig. 8), but when this stem was polished, no pits were seen. The author
believes that the round pits do not represent pyrite crystals, which usually have an angular outline. Yet another
specimen (showing small spines), which was not pyritized, was unusual when macerated in Schulze’s solution, in
that it yielded excellent tracheids in longitudinal view, which showed round bordered pits in two rows in some of
the tracheids (PI. 28, figs. 5, 6). The variety of xylem pitting is large, but such evidence does suggest that all the
specimens belong to one species.
Reconstruction of Konioria
There are no large specimens of this plant available, but the separate fragments are linked by
common characters (Table 1). The reconstruction (text-fig. 4) is made from the drawings of all those
separate parts.
Konioria gen. nov.
Type species. Konioria andrychoviensis sp. nov.
Diagnosis. Erect axes slender, branching by more or less equal dichotomy in different planes. Lower
dichotomies further apart than upper ones. Apices narrow to points, and are curved to form hooks.
Axis usually with 1 -4 narrow longitudinal wings. Surface of axes with spines, which range from long
and subulate to short and triangular. Epidermis of axis with longitudinally orientated stomata. Inside
the epidermis is a hypodermis of up to four layers of thick-walled cells. The hypodermal cells are
rounded in transverse section but elongated longitudinally. The area of the cortex and phloem, is not
ZDEBSKA: LOWER DEVONIAN ZOSTEROPHYLL KONIORIA
257
preserved. A central xylem strand has the smallest tracheids to the outside. Metaxylem tracheids have
scalariform thickening. Sporangia are borne singly below dichotomies, usually on short stalks. The
sporangia are round or reniform in outline, and composed of two equal valves, with the dehiscence
line along the whole margin.
text-fig. 4. Reconstruction of Konioria andrychoviensis, x 3 (approx.).
Konioria andrychoviensis sp. nov.
Plates 25-28
1965 Psilophyton sp. Reymanowna in Konior, p. 217, figs. 1, 3.
Diagnosis. Axes 0-3-4 mm wide, branching frequently, with short to long spines. Lower part of axis
bearing subulate spines up to 4 mm long usually with minute lateral teeth, the upper part of the axis
has shorter and triangular spines. Sporangia up to 2-5 mm wide but often smaller, frequently bearing
minute spines near their bases.
Horizon. Lower Devonian, Emsian.
Locality. Two bore-holes Andrychow 2 and Andrychow 4 in the Bielsko-Andrychow area of the Western
Carpathians, Poland.
258
PALAEONTOLOGY, VOLUME 25
Type specimens. All specimens are deposited in the Palaeobotanical Museum of Institute of Botany,
Jagiellonian University, Krakow, S/98. Holotype; S/98/31, PI. 26, figs. 7, 8. Paratypes: S/98/4, PI. 25, fig. 1; S/98/
35, text-fig. 4; S/98/ 1 5, PI. 27, fig. 3; S/98/29, PI. 28, fig. 1 .
Derivation of name. The generic name Konioria is after Professor Konrad Konior who found the material and
was kind enough to give it to the author for investigation. The specific name is derived from the name of the
locality Andrychow.
DISCUSSION
Comparison with Psilophyton goldschmidtii and P. arcticum
Konioria might be confused with P. goldschmidtii (Halle 1916) and P. arcticum (Hoeg 1942), because
of similar external morphology of the axes as seen when on the rock surface. A character common to
both Konioria and P. goldschmidtii is the ‘compression margin’ (wing), which is distinctly visible on
specimens seen on the rock surface. Halle (1916) gives no explanation of its nature. In Konioria the
‘margins’ are in fact wings running along both sides of the axis. A second similar character are the
subulate spines of variable length up to 4 mm. The spines of P. goldschmidtii, however, are less
numerous. The two plants differ in their branching. In Konioria the axes dichotomize, whilst in
P. goldschmidtii the branching is sympodial and dichotomous. The axes of Konioria are straight while
in P. goldschmidtii they are zigzag shaped (Nathorst 1913; Halle 1916; Hoeg 1967). In P. goldschmidtii
the anatomical structure of the axes and the sporangia are unknown. Although these two plants show
several common characters, the differences between them and the different mode of preservation
suggest that the new plant should not be included in P. goldschmidtii. When comparing these two
plants, the author was able to study the figured specimen described by Halle (1916). The similarities
and differences between the two plants mentioned above were confirmed by this material.
In 1959 Ananiev described P. goldschmidtii from the Devonian of south-eastern Siberia. Ananiev
gives no description of the plant, but his photographs (PI. 7, figs. 1,2, 4; PI. 8, fig. 4; PI. 12, fig. 1;P1. 14,
figs. 1, 2; PI. 24, fig. 2d) clearly show the differences between the branching of Konioria and
P. goldschmidtii.
In 1932 Lang described P. princeps and P. goldschmidtii from the Strathmore Beds in Scotland
under the name Psilophyton. The Psilophyton figured by Lang on PI. 2, figs. 24, 25, 27, is similar to
Konioria in that it shows axes with a ‘compression margin’ (wing). His PI. 2, figs. 24, 33, shows the
presence of short spines and bases of broken long spines similar to Konioria. The differences in
branching between the two plants are clearly seen when comparing Lang’s PI. 2, figs. 24, 25, 27, with
Konioria. Lang describes the branching as pseudomonopodial, which is clearly different from the
dichotomous branching in Konioria. An investigation of Lang’s material in the British Museum of
Natural History in London, confirms the similarities and differences between the two taxa.
Konioria shows some external similarity to P. arcticum Hoeg 1942, which also possesses a distinct
margin (see his PI. 9, fig. 3; PI. 12, fig. 1). Hoeg does not explain the nature of this margin. Unlike
Konioria, however, P. arcticum has pseudomonopodial branching and hair-like spines from 4 to
6 mm long.
Comparison with genera of the subdivision Zoster ophyllophytina
Attributing Konioria to the Zosterophyllophytina, the author uses the classification of the
‘psilophytes’ given by Banks (1968, 1975). Konioria shows the characters of the subdivision
Zosterophyllophytina, i.e. the lateral arrangement of sporangia and the exarch xylem strand.
Comparison with Crenaticaulis and Gosslingia. Konioria possesses a greater number of characters
in common with Crenaticaulis (Banks and Davis 1969) and Gosslingia (Edwards and Banks 1965;
Edwards 1970) than with other genera. All three plants show dichotomous branching, although
Crenaticaulis and Gosslingia also have pseudomonopodial branching. Crenaticaulis and Gosslingia
also show scars below dichotomies. In Crenaticaulis axillary branches are present which were
compared by Banks and Davis (1969) with the rhizophores of Selaginella. Also, in Konioria, the
sporangia are borne in the same position. Another character in common in these three genera is that
the sporangia do not form spikes.
ZDEBSKA: LOWER DEVONIAN ZOSTEROPHYLL KONIORIA
259
Konioria differs from Gosslingia in possessing spines. Crenaticaulis shows characteristic short
tooth-like spines arranged in one or two rows, while Konioria has irregularly arranged spines ranging
from long and subulate to short and triangular. The sporangia of Gosslingia and Konioria split into
two equal valves whilst in Crenaticaulis there is a large abaxial and a small adaxial valve.
Comparison with Sawdonia and Euthursophyton. Konioria shows a certain similarity with Sawdonia
ornata (Dawson) Hueber, see for example, Dawson 1871; Hueber 1964, 1971; Hueber and Banks
1967; Ananiev and Stepanov 1968. The two plants have the same dichotomous mode of branching
and the lateral arrangement of sporangia which split into two equal valves. Sawdonia differs from
Konioria in having glandular spines and sporangia distributed along the axis.
Konioria and Euthursophyton hamperbachense Mustafa (Mustafa 1978) are similar in their
dichotomously branching axes covered with long spines, in occasionally showing circinately coiled
axes and in possessing an exarch strand. In Euthursophyton , however, the sporangia are unknown.
Konioria, unlike Euthursophyton, possesses differentiated spines, from long and subulate with minute
teeth to short and triangular. In addition, wings on the surface of the axes are present in Konioria but
absent in Euthursophyton. In the Euthursophyton axes no hypodermis has been described.
Comparison with Zosterophyllum and Rebuchia. Konioria differs from both Zosterophyllum (Croft
and Lang 1942; Edwards 1969a, b; Lele and Walton 1961) and Rebuchia (Dorf 1933; Hueber 1970,
1972a, b ) in not having the sporangia arranged in spikes. In addition, these two genera have smooth
axes, which in Zosterophyllum show in their lower parts H -shaped branching. Common to
Zosterophyllum, Rebuchia, and Konioria are lateral, reniform sporangia splitting into two equal
valves. Edwards (1969) demonstrated exarch xylem in Zosterophyllum Hanover anum. In Rebuchia the
xylem strand is not well known, and is mentioned only by Lepekhina, Petrosian, and Radchenko
(1962). It is possible that this is also a common character of all three genera.
The systematic position of Konioria
The author accepts the classification of the early land plants into Rhyniophytina, Trimerophytina,
and Zosterophyllophytina (see Banks 1968, 1975). The lateral position of sporangia and the exarch
xylem strand of Konioria suggest its affinity with genera of the subdivision Zosterophyllophytina.
Comparisons show that Konioria differs from other Zosterophyllophytina in having sporangia
attached below a dichotomy, in showing wings on the axis, and in possessing spines ranging from
long and subulate to short and triangular. These are the characters of the new genus and species
Konioria andrychoviensis.
In his classification. Banks (1968) divided the order Zosterophyllales into two families,
Zosterophyllaceae and Gosslingiaceae, which differ in the grouping of sporangia. In a later paper
Banks (1975) distinguishes only one family, the Zosterophyllaceae. If this later publication were
followed, Konioria would become a member of the Zosterophyllaceae, fitting well the diagnosis of
this family. However, it would appear that a return to the former division of the Zosterophyllo-
phytina into two families would be better, with the Zosterophyllaceae having sporangia in spikes, and
the Gosslingiaceae having sporangia scattered along the axis. These two families were also recognized
by Kasper, Andrews, and Forbes (1974). Perhaps Konioria would even merit the establishing of a
third family, characterized by having sporangia attached below the dichotomies.
The morphology and the anatomical structure of Konioria provide additional data to support
Banks’s (1968) classification of the early land plants, in that the morphological characters discussed
above are further evidence that the Zosterophyllophytina forms a natural group. This group,
according to the hypothesis of Banks (1968; Chaloner and Sheerin 1979). gave rise to the
Lycophytina.
Konioria and the evolution of the lycopod sporophyll
The unusual position of the sporangium of Konioria appears to be related to that in certain early
lycopods. According to Zimmermann (1930) the sporophylls of the lycopods developed from
260
PALAEONTOLOGY, VOLUME 25
ultimate branchlets with sporangia of the Rhynia- type. However, according to Banks (1968), the
Rhyniophytina gave rise to all groups of plants, except the lycopods which developed from the
Zosterophyllophytina. This opinion is confirmed by the anatomical structure and the type and
arrangement of sporangia in the Zosterophyllophytina. Accordingly, the intermediate forms leading
to the sporophyll of lycopods have been looked for among the Zosterophyllophytina. In accordance
with Zimmermann’s Telome Theory, the processes of Konioria suggest an evolutionary sequence of
the lycopod sporophyll from the Zosterophyllophytina. Text-fig. 5 is an attempt to arrange such a
text-fig. 5. Proposed evolutionary sequence of the lycopod sporophyll. 1, Gosslingia\ 2. Konioria ;
3, Colpodexylon', 4. Leclercqia', 5. Cyclostigma.
sequence of existing fossil plants, which could lead to the sporophylls of lycopods. The starting-point
is Gosslingia (Edwards 1970), which bears lateral sporangia scattered along the axis. The next stage is
Konioria with lateral sporangia borne below dichotomies. Therefore, Konioria would appear to be the
link between the Zosterophyllophytina and such lycopods as Colpodexylon (Banks 1944) and
Leclercqia (Banks, Bonamo, and Grierson 1972) which have a sporophyll divided into three or more
segments. The unequally dichotomizing axes of Konioria can be regarded as the initial form from
which the sporophylls of those lycopods originated. It can be assumed that the wider branch of the
unequal dichotomy of Konioria would change into the main axis, as a result of the process of over-
topping. The other narrower branch of Konioria consisting of ends of axes with a sporangium
attached at the basis of the dichotomy, could be transformed into the lycopod sporophyll by the
processes of planation and reduction. In Konioria there are unequal dichotomies of the hook-like
apices under which the sporangia are attached. Therefore, it is possible to derive the sporophylls of
Colpodexylon directly from Konioria, because in Konioria there are apices with three or four branches
above the sporangium. It is more difficult to derive the sporophyll of the lycopod Leclercqia
complexa, which ends with five segments, from Konioria, because five times divided apices of
Konioria were not found. In theory, however, such branching is also possible.
ZDEBSKA: LOWER DEVONIAN ZOSTEROPHYLL KONIORIA
261
There is also the theoretical possibility of deriving the undivided sporophylls of Cyclostigma
(Chaloner 1968) and similar lycopods by a reduction from sporophylls showing more than one
division.
As a result of those considerations, the explanation by Zimmermann (1930) of the origin of the
lycopod sporophyll from sterile and fertile axes by overtopping and reduction appears justified, and
can be envisaged directly from the Zosterophyllophytina, but not from the Rhyniophytina. The
sequence proposed in the present paper appears to confirm the axial origin of the lycopod sporophyll.
In the light of new facts established about the position of sporangia in Drepanophycus and
Protolepidodendron, these two genera are excluded from the sequence. Drepanophycus ( Protolycopo -
dites) devonicus is not placed here, because Schweitzer and Giesen (1980) established that sporangia
of this plant do not occur on sporophylls, but on the axis among microphylls. According to
Schweitzer and Giesen, the sporangia in D. spinaeformis Goepp. also have a similar position, and do
not occur on the sporophyll as described by Krausel and Weyland (1930). Also, Protolepidodendron
scharianum Potonie and Bernard which according to Krausel and Weyland (1932) possessed
bifurcating sporophylls is not shown in the sequence. According to Schweitzer and Giesen (1980), its
sporophylls show a double dichotomy with a sporangium present below each dichotomy. Schweitzer
and Giesen think that P. wahnbachense Krausel and Weyland also shows this type of sporophyll (see
their reconstruction (text-fig. 12, p. 15)). However, there exists a controversy about this reconstruc-
tion, because according to Fairon-Demaret (1979) these sporophylls bear several sporangia, and she
tentatively attributes this plant to the Sphenopsida and not to the Lycopsida.
In summary, the above considerations demonstrate the possibility that the lycopod sporophyll
may have originated through a change in the ends of axes into sporophylls divided into segments, and
to a reduction of the number of these segments (text-fig. 5).
Acknowledgements. I would like to express my sincere thanks to Professor J. Dyakowska from the Botanical
Institute of the Jagiellonian University for her kind care during the preparation of my doctor’s thesis and to the
late Professor M. Kostyniuk for his comments. To Dr. M. Reymanowna from the Botanical Institute of the
Polish Academy of Sciences my thanks for consultation and critical remarks. My gratitude is due to Professor
K. Konior from the Geological Institute (Carpathian Section) in Krakow for supplying the material and for his
stimulating remarks. I wish to thank Professor W. G. Chaloner for my stay during 1973 in his Department in
Birkbeck College of the University of London, for discussion, and the opportunity to take microphotographs
and SEM micrographs. I thank Professor T. M. Harris for his remarks on the paper and Professor H. P. Banks
and Dr. D. Edwards for consultation. To Professor B. Lundblad from the Swedish Museum of Natural History I
am indebted for the loan of specimens of Psilophyton goldschmidtii. I thank the British Museum (Natural
History) for the opportunity to study the W. H. Lang collection.
REFERENCES
ananiev, a. r. 1959. Die wichtigsten Fundstellen von Devonftora im Ssajan-Altaj Berggebiete, Tomsk. [In Russian:
German summary.]
— and stepanov, s. a. 1968. Nachodki organov sporonosenia u Psilophyton princeps Dawson emend. Halle v
niznem devonie Yuzno-Minusinskoi Kotloviny (Zapadnia Sibir). Trudy tomsk. gos. Univ. 202, 31-46. [In
Russian.]
banks, h. p. 1944. A new Devonian lycopod genus from southeastern New York. Am. J. Bot. 31, 649-659.
— 1968. The early history of land plants. Pp. 73-107. In drake, e. t. (ed.) Evolution and environment , Yale
University Press, New Haven.
1975. Reclassification of psilophyta. Taxon , 24, 401-413.
— bonamo, p. M. and Grierson, j. d. 1972. Leclercqia complexa gen. et sp. nov., a new lycopod from the late
Middle Devonian of eastern New York. Rev. Palaeobot. Palynol. 14, 19-40.
— and davis, m. r. 1969. Crenaticaulis, a new genus of Devonian plants allied to Zosterophyllum, and its
bearing on the classification of early land plants. Am. J. Bot. 56, 436-449.
chaloner, w. g. 1968. The cone of Cyclostigma kiltorkense Haughton, from the Upper Devonian of Ireland. J.
Linn. Soc. 61, 25-36.
262
PALAEONTOLOGY, VOLUME 25
chaloner, w. G. and sheerin, a. 1979. Devonian macrofloras. Special papers in Palaeontology, 23, 145-161.
croft, w. N. and lang, w. h. 1942. The Lower Devonian flora of the Senni Beds of Monmouthshire and
Breconshire. Phil. Trans. R. Soc. 231 B, 131-163.
dawson, j. w. 1871. The fossil plants of the Devonian and Upper Silurian formations of Canada. Geol. Surv.
Can. 1-92.
dorf, e. 1933. A new occurrence of the oldest known terrestrial vegetation, from Beartooth Butte, Wyoming.
Bot. Gaz. 95, 240-257.
edwards, d. 1968. A new plant from the Lower Old Red Sandstone of South Wales. Palaeontology, 11, 683-
690.
— 1969a. Further observations on Zoster ophy llum llanoveranum from the Lower Devonian of South Wales.
Am. J. Bot. 56, 201-210.
— 1969ft. Zosterophyllum from the Lower Old Red Sandstone of South Wales. New Phytol. 68, 923-931.
— 1970. Further observations on the Lower Devonian plant. Gosslingia breconensis Heard. Phil. Trans. R.
Soc. 258 B, 225-243.
— and banks, H. p. 1965. Branching in Gosslingia breconensis. Am. J. Bot. 52, 636.
fairon-demaret, m. 1979. Estinnophyton wahnbachense (Krausel and Weyland) comb, nov., une plante
remarquable du Siegenien d’Allemagne. Rev. Palaeobot. Palynol. 28, 145-160.
halle, t. G. 1916. Lower Devonian plants from Roragen in Norway. K. svenska Vetens. Akad. Handl. 57,
1-46.
hoeg, o. A. 1942. The Devonian and Downtonian flora of Spitsbergen. Norges Svalb. Ishars-Unders. Skrifter,
83, 1-228.
— 1967. Psilophyta. Pp. 191-352. In boureau, e. (ed.), Traite de Paleobotanique. II. Masson et Cie, Paris.
hueber, F. M. 1964. The Psilophytes and their relationship to the origin of ferns. Bull. Torrey bot. Club,
21, 5-9.
1970. Rebuchia: a new name for Bucheria Dorf. Taxon, 19, 822.
— 1971. Sawdonia ornata: a new name for Psilophyton princeps var. ornatum. Ibid. 20, 641-642.
— 1972a. Rebuchia ovata, it vegetative morphology and classification with the Zosterophyllophytina. Rev.
Palaeobot. Palynol. 14, 113-127.
1972ft. Early Devonian land plants from Bathurst Island, District of Franklin. Geol. Surv. Pap. Can.
71-28, 1-17.
and banks, H. p. 1967. Psilophyton princeps'. the search for organic connection. Taxon, 16, 81-85.
kasper, a. e., Andrews, h. n. and forbes, w. h. 1974. New fertile species of Psilophyton from the Devonian of
Maine. Am. J. Bot. 63, 339-359.
konior, k. 1965. Le Devonien inferieur dans la base des Karpates bordurales de la region Cieszyn-Andrychow.
Bull. Acad. Pol. Sc. ser. geol. geogr. 13, 215-219.
— 1966. Remarques sur le developpement et Page du Devonien inferieur du substratum de la region
Bielsko-Andrychow. Ibid. 15, 231-232.
— 1968. Lower Devonian in borehole Andrychow 4. Kwart. geol. 12, 827-842. [In Polish: English summary.]
— 1969. The Lower Devonian from boreholes in the Bielsko-Andrychow region. Acta geol. pol. 19, 177-220.
[In Polish: English summary.]
— and turnau, e. 1973. Preliminary study of microflora from Lower Devonian deposits in the area of
Bielsko-Wadowice. Ann. Soc. geol. Pol. 43, 273-282.
krausel, r. and weyland, h. 1930. Die Flora des deutschen Unterdevons. Abh. preuss. geol. Landesanst. n.f.
131, 1-92.
— 1932. Pflanzenreste aus dem Devon. IV. Protolepidodendron Krejci. Senckenbergiana, 14, 391-403.
lang, w. H. 1932. Contributions to the study of the Old Red Sandstone flora of Scotland. VIII. On Arthrostigma.
Psilophyton and some associated plant-remains from the Strathmore Beds of the Caledonian Lower Old Red
Sandstone. Trans. R. Soc. Edinb. 57, 491-521.
— and cookson, i. c. 1935. On a flora, including vascular land plants, associated with Monograptus , in rocks
of Silurian age, from Victoria, Australia. Phil. Trans. R. Soc. 224 B, 421-449.
lele, k. m. and walton, j. 1961. Contribution to the knowledge of ‘ Zosterophyllum myretonianum' Penhallow
from the Lower Old Red Sandstone of Angus. Trans. R. Soc. Edinb. 64, 469-475.
lepekhina, v. G., Petrosian, n. M. and Radchenko, G. p. 1962. The most important Devonian plants of the Altai-
Sayan mountain region. Trudy vses. nauchno-issled. geol. Inst. 70, 61-189. [In Russian.]
mustafa, h. 1978. Beitrage zur Devonflora, III. Argumenta palaeobot. 5, 91-132.
nathorst, a. G. 1913. Die Pflanzenreste der Roragen Ablagerung. In Goldschmidt, v. m. (ed.), Das Devongebiet
am Roragen bei Roros, Kristiania. Pp. 25-27.
ZDEBSKA: LOWER DEVONIAN ZOSTEROPHYLL KONIORIA
263
Schweitzer, h. j. and geesen, p. 1980. Uber Taeniophyton inopinatum, Protolycopodites devonicus und
Cladoxylon scoparium aus dem Mitteldevon von Wuppertal. Palaeontographica, 173 B, 1-25.
turnau, E. 1974. Microflora from core samples of some Paleozoic sediments from beneath the flysch
Carpathians (Bielsko-Wadowice area, southern Poland). Ann. Soc. geol. Pol. 64, 143-169.
zimmermann, w. 1930. Die Phylogenie der Pflanzen. 1. Gustav Fischer. Jena.
DANUTA ZDEBSKA
Typescript received 21 June 1980
Revised typescript received 29 April 1981
Institute of Botany
Jagiellonian University
ul. Lubicz 46
31-512 Krakow, Poland
TWO S ALENIOID ECHINOIDS
IN THE DANIAN OF THE
MAASTRICHT AREA
by J. F. GEYS
Abstract. Two species of salenioid echinoids, Salenia minima Agassiz and Desor, 1846 and Hyposalenia
heliophora (Agassiz and Desor 1846) from the Danian strata in the Maastricht area, are described and discussed.
Biometrical parameters are statistically treated and compared with those of some other salenioids. It is shown
that both species are probably good index fossils for strata of Danian age.
Previous systematic treatments of the Danian echinoderm fauna in the Maastricht area were made
exclusively by investigators studying the Cretaceous Maastricht Chalk. Lambert (1911) was the first
to study the echinoid fauna of the Maastricht Cretaceous as a whole. The fauna was systematically
revised by Smiser (1935). Both these authors essentially studied the important collections of the
Koninklijk Belgisch Instituut voor Natuurwetenschappen in Brussels (K.B.I.N.). The so-called post-
Maastrichtian (or Danian) echinoids from the Maastricht area, are less well represented in these
collections, which may explain why the characteristic Danian species were not revised by both these
authors. The collections of the K.B.I.N., as well as those of the Natuurhistorisch Museum at
Maastricht (N.H.M.M.), were revised by M. Meijer. Unfortunately, his results were only partly
published, but Meijer (1965) was the first to draw attention to the important differences between the
echinoid faunas of the Houthem Formation (post-Maastrichtian) and the underlying chalk beds (Ma
to Md), belonging to the Maastricht Chalk. As a result he separated his ‘zone III' from the echinoid
assemblages of the Maastricht Chalk, and he presumed that this zone was of Dano-Montian age.
Meijer’s view proved to be correct when Moorkens (1972) correlated the post-Maastrichtian
Geulhem Chalk (Houthem Formation) with the Ciply Tuffaceous Chalk, which had been previously
correlated with the type Danian at Fakse (Denmark) by Rasmussen (1965).
Two of the most characteristic species from the post-Maastrichtian Houthem Formation in the
Maastricht area are discussed in this paper. They occur in the Ciply Tuffaceous Chalk, in the vicinity
of Mons (Belgium), but they are absent in overlying and in underlying strata. In the Maastricht area,
both species seem to be confined to strata which Felder (1975) termed the Geulhem Chalk of the
Houthem Formation. A correlation between the Geulhem Chalk and the Ciply Tuffaceous Chalk is
thus confirmed.
The specimens studied belong to the collections of the Natuurhistorisch Museum at Maastricht (N.H.M.). In
some specimens, a few dimensions were measured by means of calipers (absolute error: OT mm). Statistical
calculations were carried out, using the formulas and symbols proposed by Till (1974).
The following abbreviations are used:
D: ambital diameter of the test III-5;
h: total height of the test;
ds: diameter of the apical system, taken between the centres of the distal borders of ocular III and genital 5;
dp: diameter of the peristome III-5.
IPalaeontology, Vol. 25, Part 2, 1982, pp. 265-276, pi. 29.|
266
PALAEONTOLOGY, VOLUME 25
SYSTEMATIC DESCRIPTIONS
Class echinoidea Leske, 1778
Subclass euechinoidea Bronn, 1860
Superorder echinacea Claus, 1876
Order salenioida Delage and Herouard, 1903
Family saleniidae Agassiz, 1838
Subfamily saleniinae Agassiz, 1838
Genus salenia Gray, 1835
(= Cidarelle Desmoulins, 1835; = Bathy salenia Pomel, 1838)
Type species. Cidarites scutigera Munster in Goldfuss 1826, by original designation.
Salenia minima Agassiz and Desor, 1 846
Plate 29, figs. 1-4
*.1846 Salenia minima, Agassiz and Desor, p. 342.
.1850 Salenia minima, d’Orbigny, p. 273.
.1857 Salenia minima, Desor, p. 151.
1857 Salenia minima. Bosquet, no. 839.
1859 Salenia minima, Binkhorst van den Binkhorst, p. 120.
.1864 Salenia minima, Cotteau, pp. 171-173, pi. 1040, figs. 1-10.
1879 Salenia minima, Ubaghs, p. 228.
1881 Salenia minima, Mourlon, p. 125.
.1910 Salenia minima, Lambert and Thiery, p. 21 1.
.1935 Salenia minima, Mortensen, p. 369, fig. 195 d.
v.1965 Salenia minima, Meijer, p. 23.
.1979 Salenia minima, Geys, p. 320.
non 1935 Salenia minima, Smiser, p. 28, pi. 2, fig. 6 a-d.
Type locality. Ciply, Hainaut, Belgium. Ciply Tuffaceous Chalk, Danian.
Studied specimens from the Maastricht area {Geulhem Chalk). Vroenhoven, Belgian Limburg: 88 specimens;
Geulhem, Dutch Limburg: 18 specimens.
Dimensions. D: 1-7-9-2 mm; h: 0-9-6 0 mm; h/D ratio: 0-48-0-70; ds: 1-7-6-8 mm; ds/D ratio: 0-60-1-00; dp:
0-8-4-8 mm; dp/D ratio: 0-36-0-63.
Description. The adoral side of the test is flat; the peristome is circular to subpentagonal, not sunken. Gill slits are
very small. The perignatic girdle consists of small, spatulate auricles, not in contact with each other, and low
arcuate apophyses.
The apical system is large and covers most of the aboral side of the test in young specimens. It is slightly convex
and consists of eleven smooth plates, separated by distinct sutures. Sutural depressions are very small or absent.
The ocular plates are triangular and a little concave. The genital plates and the suranal plate are nearly flat. The
genital pores have an excentric position in the genital plates, being nearer to the apex. The madreporite can easily
be distinguished by its irregular, more or less triangular poriferous depression. The periproct is oval or
subtrigonal.
Each ambulacral series shows ten to twelve crenulate, non-perforate primary tubercles. The ambulacra are
straight. The bigeminate character of its plates is very regular. The tubercles are close together and
extrascrobicular granulation is almost completely absent. The axes of the pore pairs have an inclination of
about 45°.
Interambulacral tubercles are crenulate, non-perforate. Five of them make a series. The areoles are smooth,
large, and close together. Sometimes they are confluent, in the vicinity of the ambitus. A ring of six to eight
scrobicular tubercles surrounds them on all but the adradial sides. Interradial extrascrobicular surfaces are
coarsely granulated.
Variability, dp, ds, and h were plotted against D. Some parameters and the reduced major axis lines (rmal) were
computed for each of these graphs, using the formulas proposed by Till (1974).
GEYS: SALENIOID ECHINOIDS
267
dp (
5-
4-
3-
2
1
mm)
Salenia minima
r = °-97 .
slope = 25°30'
rmal : dp= 0,06 + 0,48 D
123456789 D(mm)
text-fig. 1. dp-D plot of Salenia minima, with reduced major axis line (rmal).
a. The rmal of the dp-D plot is given by
dp = 0-06 + 048 D (1) (Text-fig. 1).
The 95% confidence intervals of intercept and slope are respectively
1 -96 sa(D/dp) = +0-16 mm (2)
1 96 sb(D/dp)= ±0-04 (3).
From (1) and (2) one can conclude that the origin of the plot is included in the confidence band of the rmal, and
that the dp/D ratio can be considered constant. Moreover, the slope (b = 0-48 ± 0-04) does not differ significantly
from the mean dp/ratio (= 0-50), as follows from (1) and (3). The relative size of the peristome is not influenced
by over-all size or age of the specimen.
b. The rmal of the h-D plot is given by
h = - 0-3 1 + 0-68 D (4) (Text-fig. 2)
The 95% confidence intervals of intercept and slope are respectively
1-96 sa(D/h) = +0-13 mm (5)
1-96 sb(D/h)= ±0-03 (6).
Obviously the origin is not included in the confidence band of the rmal, as follows from (4) and (5). (6) indicates
that the slope (b = 0-68 ± 0-03) differs significantly from the mean h/D ratio ( = 0-59). One can thus conclude that
the h/D ratio is not constant, though the relationship between h and D is linear. Larger specimens tend to be less
flattened than smaller individuals.
c. The rmal of the ds-D plot is given by
ds = 0-62 + 0-64 D (7) (Text-fig. 3),
with 95% confidence intervals of intercept and slope respectively
1 -96 sa(D/ds) = + 0T 8 mm (8)
1-96 sb(D/ds)= ±0-02 (9).
On one hand, one can conclude from (7) and (8) that the origin is not included in the confidence band of the rmal;
on the other, (7) and (9) indicate that the slope (b = 0-64 + 0-02) differs significantly from the mean ds/D ratio
(= 0-82). Larger individuals are shown to have a relatively smaller apical system than smaller individuals. This
could be put as follows: the apical system seems to grow less fast than the entire animal.
The relationship is clearly linear for specimens with D > 4-5 mm and for specimens with D < 4-5 mm
268
PALAEONTOLOGY, VOLUME 25
h (mm)
text-fig. 3. ds-D plot of Salenia minima , with reduced major axis line (rmal).
GEYS: SALENIOID ECHINOIDS
269
separately, but not for the population as a whole. This is clearly demonstrated by computing the rmal for both
partial populations separately.
For large specimens (D>4-5 mm) one obtains
ds= -1-18 + 0-87 D
with 95% confidence intervals
1-96 sa(D/ds) = ±0-61 mm
1 -96 sb(D/ds) = ±0-04.
For small specimens (D<4-5 mm) one obtains analogously
ds = 0-39 + 0-73 D
with 95% confidence intervals
1 96 sa(D/ds) = ±0-23 mm
1 96 sb(D/ds)= ±0-07.
From these equations we can conclude that the intercept is significantly different in large and small individuals.
The difference in slope between the rmals of both populations is much less important. It seems as if the rate of
growth of the apical system is drastically reduced when the specimens attain 4 to 5 mm over-all diameter. After
crossing the D = 5 mm threshold, the rate of growth of the apical system seems to be more or less restored.
Discussion. Salenia minima is superficially similar to other small salenioids, such as Salenidia pygmaea
(Hagenow 1940) and Salenidia maestrichtensis (Schliiter 1892). The lack of sutural depressions in its
smooth apical system makes Salenia minima easy to recognize.
Smiser (1935) confused Salenia minima and Goniopygus minor. What he figured as Salenia minima
is in fact a Goniopygus minor Sorignet, 1850. This error is discussed in a forthcoming paper of mine
(Geys 1981).
A very similar species to Salenia minima has been described from the Danian of Denmark:
Salenidia danica Ravn, 1928. Both species probably differ in the structure of their ambulacra and in
the shape of their ocular plates.
For both Salenidia pygmaea and Salenidia maestrichtensis h-D, ds-D, and dp-D plots were
statistically analysed, respectively by Nestler (1965) and by Geys (1979). A comparison with Salenia
minima is instructive. The flattening in shape of larger, hence older specimens, as demonstrated in
Salenia minima, exists in Salenidia pygmaea but not in Salenidia maestrichtensis. The relative
shrinking of the peristome, demonstrated in Salenidia pygmaea and Salenidia maestrichtensis, is not
indicated in Salenia minima. The growth of the apical system, however, shows similar patterns in all
three of the species. Very small, very young specimens have apical systems covering the complete
adapical surface. In Salenidia maestrichtensis a non-linear relationship between D and ds was
demonstrated. For Salenidia maestrichtensis it was established that the beginning of the final linear
segments in its d-D plot (D >4-5 mm) coincides with the attainment of sexual maturity. It is not clear
whether the same is true for Salenia minima.
Subfamily hyposaleniinae Mortensen, 1934
Genus hyposalenia Desor, 1856
( = Peltastes Agassiz, 1838, non Peltastes Rossi, 1807; Peltosalenia Quenstedt, 1874)
Type species. Echinus acanthoides Desmoulins 1837, by subsequent designation of Mortensen (1935).
Hyposalenia heliophora (Agassiz and Desor, 1 846)
Plate 29, figs. 5-8
v*.1846 Salenia heliophora, Agassiz and Desor, p. 342.
v.1850 Salenia heliophora, d’Orbigny, p. 273.
v.1856 Hyposalenia heliophora, Desor, p. 148.
1857 Hyposalenia heliophora. Bosquet, no. 842.
v. 1 864 Peltastes heliophorus, Cotteau, pp. 122-124, pi. 1029, figs. 1-7.
270
PALAEONTOLOGY, VOLUME 25
1874 Peltastes heliophorus, Cotteau, p. 642.
1879 Hyposalenia heliophora, Ubaghs, p. 228.
1881 Hyposalenia heliophora, Mourlon, p. 125.
.1892 Peltastes cf. heliophorus, Schliiter, pp. 152-154.
.1910 Peltastes heliophorus, Lambert and Thiery, p. 209.
v.1928 Hyposalenia heliophora, Lambert and Jeannet, p. 203.
.1935 Peltastes cf. heliophorus, Kongiel, p. 31, pi. 2, fig. 5 a-c.
.1935 Hyposalenia heliophora, Mortensen, p. 344, fig. 188g.
.1939 Hyposalenia heliophora, Kongiel, p. 20, pi. 3, figs. 19-21.
.1966 Hyposalenia heliophora. Fell and Pawson, p. U379, fig. 277-1/
v.1966 Hyposalenia heliophora, Meijer, p. 23.
.1979 Hyposalenia heliophora, Geys, p. 320.
Type Locality. Maastricht, Dutch Limburg, the Netherlands. Geulhem Chalk, Houthem Formation, Danian.
Other occurrences. Ciply, Hainaut, Belgium: Ciply Tuffaceous Chalk, Danian (Agassiz and Desor 1846). Ciply,
Hainaut, Belgium: Malogne Gravel, Danian (Cotteau 1874). Berlin, Germany: ‘Geschiebe’, reworked in
Pleistocene (Schliiter 1892). Gora Pulawska, Poland: Siwak, Lower Danian (Kongiel 1935, 1939).
Studied specimens from the Maastricht area ( Geulhem Chalk). Vroenhoven, Belgian Limburg: 41 specimens;
Geulhem, Dutch Limburg: 225 specimens.
Dimensions. Forty specimens, chosen at random, were measured. D: 1-7-12-0 mm; h: 1 -0-7-5 mm; h/D ratio:
0-5-0-7; ds: 1 -6-7-7 mm; ds/D ratio: 0-62-1 -00; dp: 0-6-5-2 mm; dp/D ratio: 0-26-0-58.
Description. The adoral side is slightly convex, the peristome a little sunken. The peristome is distinctly
decagonal and shows rather large gill slits. These slits are surrounded by low, blunt ridges. The auricles are short,
rectangular, and spatulate. They are not in contact over the perradial suture.
The apical system is large and pentagonal, covering most of the adapical surface. It is conically convex and
consists of eleven plates, separated by very fine, hardly visible sutures. Real sutural depressions are absent. The
ocular plates are pentagonal, with more or less straight distal margins. The genital plates are hexagonal, the
suranal plate is pentagonal. All the plates are covered by a conspicuous sculpture of ridges and furrows, which
radiate from the centres of the plates, and which are perpendicular to the sutures. The genital pores have a central
position in the plates. The periproct is oval or subtrigonal. It is situated between the apex and genital plate 5, so
that the test is symmetrical with respect to the III-5 plane.
Ambulacral series consist of twelve or thirteen crenulate, non-perforate primary tubercles. The tubercles just
below the ambitus are the largest in size. From there upwards, their size diminishes abruptly. The ambulacra are
straight. The plates are very regularly bigeminate. The poriferous zones are uniserial throughout. The pores of
each pair are separated by a low ridge. The axes of the pore pairs have an inclination of 45° below, and less than
45° above, the ambitus. Very small scrobicular tubercles, or granules, surround the primaries.
The interambulacral tubercles are crenulate, non-perforate. There are four or five in each series. The areoles
are smooth, large, and confluent. Scrobicular rings are hence discontinuous. Granulation is coarse on the
interradial miliary surfaces.
Variability, dp, ds, and h were plotted against D. Reduced major axis lines (rmal) and some parameters were
computed for each of these graphs.
EXPLANATION OF PLATE 29
Figs. 1 -4. Salenia minima Agassiz & Desor; NHMM-MM 895; Vroenhoven, Belgian Limburg; Geulhem Chalk.
1 , adapical view, x 4. 2, adoral view, x 4. 3, lateral view, x 4. 4, detail of ambulacrum at the ambitus, x 12.
Figs. 5-8. Hyposalenia heliophora (Agassiz & Desor); NHMM-MM 899; Geulhem, Dutch Limburg; Geulhem
Chalk. 5, detail of ambulacrum at the ambitus, x 12. 6, adapical view, x4. 7, adoral view, x4. 8, lateral
view, x 4.
PLATE 29
GEYS, salenioid echinoids
272 PALAEONTOLOGY, VOLUME 25
a. The rmal of the dp-D plot is given by
dp = — 0 06 + 0-38 D (text-fig. 4)
with 95% confidence intervals of intercept and slope, respectively
1 -96 sa(dp/D) = + 0-1 9 mm
1 -96 sb(dp/D) = ±0-04.
From these equations can be concluded that the origin of the plot is included in the confidence band of the
rmal, and that the mean dp/D ratio ( = 0-37) does not differ significantly from the slope (b = 0-38 + 0 04). Hence,
there is no difference in relative size of the peristome, between young (small) and old (large) specimens. The dp/D
ratio is constant.
b. The rmal of the h-D plot is given by
h = 0-16 + 0-57 D (text-fig. 5)
with 95% confidence intervals
1 -96 sa(h/D) = ±0-11 mm
1 96 sb(h/D) = 0-01.
The origin of the plot is clearly not included in the confidence band of the rmal. Moreover, the mean h/D
ratio ( = 0-62) differs significantly from the slope (b = 0-57±0-01). Though the relationship between h and
D is linear, the h/D ratio is not constant. Older (larger) specimens are slightly more flattened than young
(small) ones.
c. The rmal of the ds-D plot is given by
ds = 0-60 + 0-62 D (Text-fig. 6)
with 95% confidence intervals
1 -96 sa(ds/D) = ±0-12 mm
1 -96 sb(ds/D) = +0-01.
Obviously the origin of the graph is not included in the confidence band of the rmal. The mean ds/D ratio
(= 0-78) differs significantly from the slope (b = 0-62 + 0-01). The ds/D ratio is thus not constant. The apical
system, covering the entire adapical surface in very young individuals, grows relatively smaller in older
specimens. One could say that the apical system has a rate of growth less fast than that of the animal as a whole.
Unlike the other salenioids, discussed above, the relationship between ds and D is linear in Hyposalenia
heliophora.
h(mm)
Hyposalenia heliophora
rmal : h = 0,16 ♦ 0,57 D
slope = 30°30'
r = 0,99
1 23456789 10 11 D(mm)
text-fig. 5. h-D plot of Hyposalenia heliophora with reduced major axis line (rmal).
ds(mm)
Hyposalenia heliophora
rmal : ds = 0,60 + 0,62 D
slope = 32°
r = 0,99
D(mm)
text-fig. 6. ds-D plot of Hyposalenia heliophora with reduced major axis line (rmal).
274
PALAEONTOLOGY, VOLUME 25
Discussion. It is clear from the synonymy list that little confusion has arisen with repect to H.
heliophora. Cotteau (1864) describes this species as being easily recognized, owing to the character-
istic structure of its ambulacra and of its apical system. Little can be added to that statement. Like
Salenia minima Agassiz and Desor, 1 846, and Salenidia pygmaea (Hagenow 1 840), H. heliophora shows
a slight tendency to flatten in shape with age. Like Salenia minima, but unlike Salenidia pygmaea and
Salenidia maestrichtensis (Schliiter 1892), H. heliophora has a constant dp/D ratio: the relative size of
its peristome does not change with over-all size or age. The growth pattern of the apical system in H.
heliophora shows some characteristics, common to most, if not all, Salenioids: a strong reduction in
relative size with age or with over-all size. Unlike the other species which were statistically analysed,
H. heliophora has a constant rate of growth in its apical system.
text-fig. 7. Detail of the apical system, with periproct, suranal plate and surrounding
genital plates in Salenia minima (a) and Hyposalenia heliophora ( b ). Enlarged, x 8.
CONCLUSIONS
On various occasions Salenia minima and H. heliophora have been reported from the Upper
Cretaceous in Belgium and in the Netherlands (Bosquet 1857; Binkhorst 1859; Ubaghs 1879;
Mourlon 1881). These reports are not really erroneous, because the Houthem Formation, as well as
the Ciply Tuffaceous Chalk, were included within the Maastrichtian or ‘Senonian’ at that time. The
strati graphical ranges of both species were precisely documented by Meijer (1966) in Dutch Limburg:
the ‘uppermost part of the Maastricht Chalk’ (now called Houthem Formation), which he called
‘zone III’ and was then considered to be of Dano-Montian age. Among Cretaceous salenioids
in various Dutch and Belgian collections (Geys 1979), I could not trace any specimen of either of the
species described here, which was of true Cretaceous age.
Specimens of Salenia minima and of H. heliophora, in the Natuurhistorisch Museum at Maastricht,
were collected exclusively in the so-called ‘post-Maastrichtian’. These strata, included in zone III of
Meijer (1966), and called Geulhem Chalk of the Houthem Formation by Felder (1975), were
considered of Danian age by Meijer (1959). In the Mons Basin the same two species seem to be
restricted to the Ciply Tuffaceous Chalk and the underlying Malogne Gravel. The Ciply Tuffaceous
Chalk was shown to be of Danian age by Rasmussen (1965).
The debate whether the Mesozoic-Cenozoic boundary is situated at the base or at the top of the
Danian is not yet completely resolved. Some authors still include the Danian with the Upper
Cretaceous (e.g. Davies 1975). By far the most widespread point of view is to include the Danian into
the Lower Tertiary (Berggren 1963; Moorkens 1972). The evidence is mainly based on various
GEYS: SALENIOID ECHINOIDS
275
palaeontological arguments: foraminifera and calcareous nannoplankton (Cepek and Moorkens
1979), ostracods (Deroo 1966), brachiopods (Kruytzer and Meijer 1958). Both Salenia minima and
H. heliophora seem to be index fossils for Danian strata. Their presence in Maastrichtian or in
Montian deposits has not yet been demonstrated.
Acknowlegements. I am indebted to Dr. D. G. Montagne (Natuurhistorisch Museum, Maastricht, the
Netherlands), to Drs. X. Misonne and P. Sartenaer (Koninklijk Belgisch Instituut voor Natuurweten-
schappen, Brussels, Belgium), to Mrs. D. Gaspard (Universite de Paris-Sud, Orsay, France) and to Professor
Dr. J. Remane (Universite de Neuchatel, Switzerland) for permission and facilities to study the collections in
their care. I also wish to thank Ing. P. J. Felder (Natuurhistorisch Museum, Maastricht, the Netherlands) and
Dr. T. Moorkens (Deutsche Texaco A.G., Wietze, Federal Republic of Germany) for critically reading the
manuscript and for suggesting some improvements.
REFERENCES
agassiz, l. and desor, e. 1846. Catalogue raisonne des families, des genres et des especes de la classe des
echinodermes I. Annls. Sci. nat. (5) Zoologie, 6, 305-374.
berggren, w. A. 1963. Problems of Paleocene stratigraphic correlation. Revue Inst. fr. Petrole, 18, 1448-1457.
binkhorst van de binkhorst, j. t. 1859. Esquisse geologique et paleontologique des couches Cretacees du
Limbourg, et plus specialement de la craie tuffeau, avec carte geologique, coupes, plan horizontal des carrieres de
St. Pierre, etc., 268 pp. Maastricht: Van Osch-America.
bosquet, j. 1857. Fossiele fauna en flora van het Krijt van Limburg. In w. staring, De Bodem van Nederland,
376-388. Haarlem: Kruseman.
Cepek, p. and moorkens, t. 1969. Cretaceous-Tertiary Boundary and Maastrichtian-Danian Biostratigraphy
(Coccoliths and Foraminifera) in the Maastrichtian type area. In w. k. Christensen and t. birkelund (eds.),
Cretaceous-Tertiary Boundary Events Symposium II. Proceedings, 137-142. Kobenhavn.
cotteau, G. 1862-1867. Paleontologie franqaise. Description des animaux invertebres commencee par Alcide
d'Orbigny. Terrain Cretace, VIII. Echinides, 892 pp., pis. 1007-1204. Paris: Masson.
1874. Note sur les echinides cretaces de la province du Hainaut. Bull. Soc. geol. Fr. (3) 2, 368-660, pi. 19-20.
davies, A. m. 1975. Tertiary faunas. A textbook for oilfield Paleontologists and students of Geology. II. The
sequence of Tertiary faunas, 447 pp. London: Allen & Unwin.
deroo, G. 1966. Cytherea (Ostracodes) du Maastrichtien de Maastricht et des regions voisines. Resultats
stratigraphiques et paleontologiques de leur etude. Meded. Geol. St. (C, V, 2) 2, 197 pp.
desor, e. 1855-1859. Synopsis des Echinides fossiles, 64 + 490 pp., 44 pis. Paris: Reinwald.
felder, w. m. 1975. Litostratigrafie van het Boven Krijt en het Dano-Montiaan in Zuid Limburg en het
aangrenzend gebied. Toelichtingen Geologische Overzichtskaart van Nederland, 63-71. Haarlem: Rijksgeolo-
gische Dienst.
geys, J. f. 1979. Salenioid Echinoids from the Maastrichtian (Upper Cretaceous) of Belgium and the
Netherlands. Palaont. Z. 53, 296-322, 6 pis.
— 1981. Arbacioid Echinoids from the Maastrichtian (Upper Cretaceous) of Belgium and the Netherlands.
Ibid. 55, 257-270.
kongiel, r. 1935. W sprawie wieku ‘siwaka’ w okolicach Pulaw. Prace Tow. Przyj. Nauk Wilnie, 9, 57 pp., 8 pis.
1939. Materjaly do znajomosci polskich jezowcow kredowych. I. Jezowcow regularne. Ibid. 13, 54 pp., 3 pi.
kruytzer, e. m. and meijer, m. 1958. On the occurrence of Crania brattenburgica (v. Schlotheim, 1820) in the
region of Maastricht (the Netherlands). Natuurhist, Maandbl. 11958) 11-12, 135-141.
Lambert, j. 1911. Description des echinides cretaces de la Belgique. II. Echinides de l’etage Senonien. Mem. Mus.
r. Hist. Nat. Belg. 16, 81 pp., 3 pis.
— and jeannet, a. 1928. Nouveau catalogue des moules d’echinides fossiles du Musee d’Histoire Naturelle de
Neuchatel, executes sous la direction de L. Agassiz et E. Desor. Denkschr. Schweiz. Naturf. Gesell. 64,
83-233, 2 pis.
and thiery, p. 1909-1925. Essai de nomenclature raisonnee des echinides, 607 pp., 15 pi. Chaumont:
Ferriere.
meijer, m. 1959. Sur la limite superieure de l’etage Maastrichtien dans la region type. Bull. Acad. roy. Beige
Cl. Sci. 45, 316-338.
— 1965. The stratigraphical distribution of echinoids in the Chalk and Tuffaceous Chalk in the
neighbourhood of Maastricht (Netherlands). Meded. Geol. St. ( N.S. ), 17, 21-25.
moorkens, t. 1972. Foraminiferen uit het stratotype van het Montiaan en uit onderliggende lagen van de boring
276
PALAEONTOLOGY, VOLUME 25
te Obourg (Met een overzicht van de stratigrafie van het Paleoceen in Belgie). Natuurwet. Tijdschr.
54, 117-127.
mortensen, t. 1935. A monograph on the Echinoidea. II. Bothriocidaroida, Melonechinoida, Lepidocentroida and
Stirodonta, 647 pp., 89 pis. Kobenhavn: Reitzel.
mourlon, M. 1881. Geologie de la Belgique, 317 + 392 pp. Bruxelles: Hayez.
orbigny a. d’ 1850. Prodrome de Paleontologie stratigraphique universelle des animaux mollusques et rayonnes,
faisant suite au Corns elementaire de Paleontologie et de Geologie stratigraphique, II, 427 pp. Paris: Masson.
ravn, J. p. J. 1928. De regulaere echinider i Danmarks Kridtaflejringer. Kgl. Danske Vidensk. Selsk. Skrifter,
Naturv. og Math. Afd. 9 (7) 1, 5-62, pis. 1-6.
Rasmussen, H. w. 1965. The Danian affinities of the Tuffeau de Ciply in Belgium and the ‘post-Maastrichtian’ in
the Netherlands. Meded. Geol. St. ( N.S. ), 17, 33-38, pis. 8-9.
schluter, c. 1892. Die regularen Echiniden der norddeutschen Kreide. II. Cidaridae. Saleniidae. Abh. Kon.
Preuss. geol. Landesamt ( N.F .) 5, 243 pp., 21 pis.
smiser, j. 1935. A monograph of the Belgian Cretaceous Echinoids. Mem. Mus. roy. Hist. Nat. Belg.
68, 98 pp., 9 pi.
till, r. 1974. Statistical methods for the Earth Scientist, 154 pp. London: Macmillan.
ubaghs, c. 1879. Description geologique et paleontologique du sol du Limbourg, 275 pp. Roermond: Romer.
J. F. GEYS
Typescript received 7 July 1980
Revised typescript received 8 December 1980
Laboratory for Mineralogy, Geology and Physical Geography
State University Centre (RUCA)
Groenenborgerlaan 171, B-2020 Antwerpen
Belgium
DEVONIAN MIOSPORE ASSEMBLAGES
FROM FAIR ISLE, SHETLAND
by J. E. A. MARSHALL and K. C. ALLEN
Abstract. Miospore assemblages have been isolated from a Devonian sequence of Old Red Sandstone facies, on
Fair Isle, Shetland. The special problems encountered in processing these palynomorphs with their high
carbonization levels and subsequent darkening are mentioned. Thirty miospore species are recorded and their
taxonomic problems and stratigraphical significance are discussed. Comparisons with similar assemblages from
the northern hemisphere indicate a Givetian (in parts specifically late Givetian) age for the Fair Isle material. The
genus Rhabdosporites Richardson 1960 is emended, and R. langii Richardson 1960 and R. parvulus Richardson
1965 are combined. The ecological significance of Geminspora (Balme) Owens 1971 is discussed.
Shetland contains one of the most complex set of continental Devonian rocks in Britain, with
different rock sequences juxtaposed by major transcurrent faults. These complex structural
relationships have proved very difficult to elucidate, and this has not been helped by the poor
biostratigraphic control between the different basins. It is hoped that palynological contributions will
give the biostratigraphic basis for a comparison of the time relations in these sedimentary basins, and
will yield information on the timing of the movements along the major transcurrent faults, about
which there is still much controversy (Smith 1977). This paper deals with the sequence found in the
small (5 km x 3 km), isolated Devonian outlier of Fair Isle, which lies 39 km south-west of the
southern tip of mainland Shetland (text-fig. 1). Emphasis is placed on stratigraphic palynology; little
attempt has been made to formally modify, or add to existing taxonomy. This is because of serious
preservational problems encountered in the palynological studies, which are discussed later.
THE DEVONIAN SUCCESSION ON FAIR ISLE
Although Fair Isle has a relatively small Devonian outcrop, its geographical position (text-fig. 1) is
of importance because it is generally considered (Mykura and Young 1969; Mykura 1972a) to lie just
to the east of the Walls Boundary Fault, which is a possible extension of the Great Glen Fault. It has
also been suggested (Mykura 1976; Donovan, Archer, Turner, and Tarling 1976) that it is part of the
same sedimentary basin as that in which the Walls Sandstone was deposited, and movements along
this fault have placed it in its present position. Any indication therefore of the precise age could help
clarify these palaeogeographic relationships.
The most recent and complete accounts of the geology of Fair Isle are those of Mykura (1972a,
19726, 1976), and this brief synopsis is drawn from them. The Fair Isle sedimentary sequence (for
which neither the top nor the base is seen), is composed of over 3000 m of dominantly clastic
terrigenous sediments, all steeply dipping east-south-east. The rocks have been subdivided (Mykura’s
nomenclature and notation followed here) into four stratigraphic subdivisions (see text-figs. 1, 2), on
the basis of lithological differences. These units can be traced with varying degrees of accuracy across
the island, but the presence of two major east-north-east trending faults create local correlation
difficulties which are only partially solved by lithological mapping.
The lowest unit is the Ward Hill Group (la and lb in Mykura’s notation), composed of over
1600 m of conglomerates and sandstones. This is succeeded by the Observatory Group (2a, 2b, and
2c) with over 900 m of sandstones and finer sediments including dolomitic siltstones and mudstones.
Above this, the Vaasetter Group (3a and 3b) which is only exposed in the central fault block, is
composed of about 600 m of conglomerates and sandstones, most of which are inaccessible. The top
IPalaeontology, Vol. 25, Part 2, 1982, pp. 277-312, pis. 30-33.|
278
PALAEONTOLOGY, VOLUME 25
» 36 Sample Location
3b Stratigraphic Unit ( See column) 0 Macro- Plants (Chaloner '72 )
text-fig. 1. Geological map of Fair Isle (after Mykura), showing location of palynological samples
and plant macrofossil sites.
MARSHALL AND ALLEN: DEVONIAN MIOSPORES
279
NORTHERN
SUCCESSION
SAMPLE
LOCATION
Svalbardia »cotica0 24i
D. roskiliensisO
c.f. Th. milleri O
Q co <
q o ^
o
= 1 a 0
o
O ° o
0o°o ^
O o
CENTRAL
SUCCESSION
BU NESS
GROUP
» S
r
3b
Sheep Craig Sndst.
^ o o
If
o - J
2
Huni
1b
to
VAASETTER
GROUP
OBSERVATORY
GROUP
□
□
1^ ol
Fine-grained Sediment
Pebbly Sandstone
Conglomerate
O Macro- Plants
1 Sample Location
text-fig. 2. Stratigraphical section of Fair Isle (after Mykura) showing position of
palynological samples.
280
PALAEONTOLOGY, VOLUME 25
unit, the Bu Ness Group (4a and 4b), consists of approximately 600 m of conglomerates, sandstones,
and siltstones; is seen only in the northern succession of the island where it is faulted against the
Observatory Group (2a, 2b, and 2c).
Increasing tectonic disturbance is seen in the rocks towards the southern end of the island where
they exhibit prominent cleavage, minor folding, mineralization, and the emplacement of small dyke
intrusions. The cause of this disturbance has been discussed by Mykura (1972a) who suggested a
possible granitic intrusion lying off the south-east coast of the island.
Previous palaeontological records are very restricted, and largely limited to the plant macrofossil
remains described by Chaloner (1972), for which the localities are shown on text-fig. 1. The fossil
plants clearly indicate a Devonian age. Chaloner tentatively suggests an age not older than middle
Siegenian for the Observatory Group (Unit 2a) based on the presence of Dawsonites roskiliensis
Chaloner 1972, and an age not older than Eifelian, but more likely Middle Devonian, for the Bu
Ness Group (Unit 4a), based on the occurrence of Svalbardia scotica Chaloner 1972. Other plants
include cf. Thursophyton milleri (Salter) Nathorst 191 5 in the Observatory Group and hostinellid axes
and cf. Prototaxites Dawson 1859 in the Bu Ness Group.
The Bu Ness Group yields the only animal fossils yet found, which include dipnoan scales, an
arthrodire plate (' ICoccosteus Agassiz 1844) and the branchiopod Asmussia Pacht, 1849. These also
favour a Middle Devonian age for this group.
MATERIAL AND METHODS
Four samples were initially provided by the Institute of Geological Sciences (Edinburgh), and when
two of these were found to contain miospores a field trip was made to the island, and seventy-eight
samples were collected from fine-grained dark-grey to black clastic rocks for palynological analysis.
Of these, sixty-one were processed, but only ten gave assemblages of sufficient quantity and
preservation to merit further study. The best-preserved assemblages were from the Observatory
Group; followed by the Bu Ness and Ward Hill Groups. Very poor preservation was found in the
argillaceous units (2a) of the Furse Argillaceous Beds where there was some cleavage development.
The lack of miospores in the southern part of the island is of interest, as it is in accord with the
increased deformation reported by Mykura (1972a). Megaspores were also rare and usually
fragmentary. This can be related to the orthogonal crack sets seen on some of the larger specimens,
which may result from an incipient cleavage (or shrinkage), causing their fragmentation during
deformation or in the processing (Burmann 1969).
The samples were firstly demineralized with hydrochloric and hydrofluoric acids, then screen-
washed through a twenty micron nylon sieve, before being cleaned of insoluble fluorides with
repeated hot HC1 treatments to give a kerogen concentrate. The miospore assemblages were very
highly carbonized (black in colour), and the only oxidizing medium found capable of clearing them
was a fuming Schulze mixture. The oxidations were carried out in a porosity-2 sinter funnel/buchner
flask system linked with a low-pressure air line to give continuous aeration (Neves and Dale 1963).
The Schulze mixture was made up with fuming nitric acid either at full strength or freshly diluted with
water, but always at a concentration greater than 70%. The time taken for the miospores to be
oxidized to a level suitable for transmitted light microscopy varied from 5 to 30 minutes. In addition,
a marked preferential clearing was seen, with some of the thicker walled (e.g. Hystricosporites spp.)
miospores never attaining more than a low level of translucency. Samples from different horizons
showed very different oxidation characteristics. Those from the Furse Argillaceous Beds (2a),
required only 5 minutes’ oxidation time and a relatively weak fuming Schulze, to reveal a poorly
preserved assemblage; whilst samples from the Bu Ness Group needed stronger and longer oxidation,
but gave much better-preserved assemblages.
After oxidation the miospores showed the phenomenon of redarkening (to opacity) in 1 to 3 days,
with the thicker-walled miospores deteriorating much more rapidly. This degradation was
accelerated by heating and water removal, so that an inert non-water miscible plastic mounting
medium could not be used, as dehydration was impossible under normal conditions. Silicone oil was
MARSHALL AND ALLEN: DEVONIAN MIOSPORES
281
also tried, but the dehydration step involving tertiary butyl alcohol, produced a rapid darkening
reaction in the cleared sample. Eventually, glycerol jelly was used as the mounting medium, and this
gave assemblages which could be studied for up to 3 days. The breakdown and darkening of the
exines when mounted in glycerol jelly was noticeable, with degradational products imparting a yellow
stain to the mounting medium. Miospores with thicker exines (e.g. Hystricosporites, Geminospora)
showed a more rapid return to opacity, and had to be studied during the first day.
A subsequent study of palynological assemblages from the Devonian of the south-east mainland
of Shetland also showed this oxidation problem. There, it was much more serious, with viability times
of about 5 minutes in glycerol jelly. During this time the miospores could be seen to redarken and
start to dissolve. It was then that a special technique was developed which involved a matching of the
oxidizing media strength to a miospore assemblage preservation, and the use of a rapid drying
technique which dehydrated the miospores faster than they could redarken, so that an inert plastic
mountant could be used (Marshall 1980).
The assemblages from Fair Isle were all studied from repeated oxidation, with fresh assemblage
slides being made up every 2 to 3 days. The study was therefore limited by the amount of organic
residue per sample; some giving enough material for seventy to eighty slides, others only having
sufficient for four or five. All the slides have been kept, and co-ordinates given for illustrated
specimens in the event that advances in techniques, such as infra-red microscopy, will enable
palynomorphs to be studied without further recourse to oxidative clearing.
One difficulty in dealing with such highly carbonized assemblages (‘vitrinite’ reflectance
measurements from Fair Isle give values of approx. 4 to 5%, pers. comm. Dr. J. M. Jones), is that the
identification of reworked components by colour differences is not possible. Three species are
described ( Emphanisporites rotatus, Camptozonotriletes aliquantus, and Grandispora Inaumovii)
which occur in a very low proportion, and it is possible that these may represent reworked
components (see Clayton, Higgs, and Keegan 1977 for a discussion of Emphanisporites). However,
until one is better able to recognize features of reworking such as breakage and erosion (as in Birks
1970), both this possibility and that of their continued presence as minor and rare elements in the
flora must be considered.
PREVIOUS PALYNOLOGICAL STUDIES FROM SHETLAND AND
ADJACENT AREAS
The pioneer work on Old Red Sandstone palynology in Britain was carried out by Lang (1925) as part
of his study of the Orcadian Basin flora. Later, Richardson produced a series of papers (1960, 1962,
1965) on the classical Orcadian area. Since then, very little work has been published except for brief
taxonomic lists as, for example, Donovan, Collins, Rowlands, and Archer (1978), who give an
account of a miospore assemblage from the island of Foula (western Shetland). Two doctoral
theses have included a certain amount of Orcadian palynology. Fannin (1970), in conjunction with
Richardson, has tabulated information on the palaeoecology and stratigraphic distribution of
miospore assemblages from Orkney, whilst Fletcher (1976) studied the megaspores of the Melby
Fish Beds. Although a useful amount of data has accumulated on the taxonomy of Devonian
spores from the Orcadian Basin, very little is known of their detailed stratigraphic distribution,
and this has been a major handicap to any precise correlation.
STATISTICS AND DATA HANDLING
The treatment of the simple bivariate statistical data, such as the exoexine and intexine diameters of
miospores, presents problems for data handling and statistical testing. Classical regression analysis is
not applicable, as there are no dependent and independent variates; furthermore, there is no easy
solution to fit a best line to this type of two-error data. Various iterative methods such as those described
by Williamson ( 1 968), Y ork ( 1 966), and comparatively reviewed by Brooks, Hart, and Wendt ( 1 972),
282
PALAEONTOLOGY, VOLUME 25
demand statistical estimates for the variances of each data point. These are certainly not worth the
extra time involved for the quality of data produced on the Fair Isle material. An approximate
method is the Reduced Major Axis line as described by Kermack and Haldane (1950), and further
documented by Till ( 1 974) and others. The statistical validity of the line produced is in doubt, but as it
appears correct (i.e. passes through the centre of the experimental scatter), it is a useful means of
comparing populations, providing not too much reliance is placed on any subsequent statistical tests
based upon it.
The data in the bivariate plots were in fact computed for logarithmic and linear fits, using both
classical regression and the reduced major axis line to find the best fit. The latter was found to be more
successful in describing a line through a set of data points. However, the logarithmic fit did not give
much advantage over the linear model. When data values were taken from published graphical plots
(Richardson 1965), an electronic digitizing machine (D-MAC) was used to provide more accurate
results and known estimates of error.
NOMENCLATURE AND SYSTEMATICS
The morphological terminology used is that of Smith and Butterworth ( 1 967), and their classification
system is followed except for the inclusion of Hystricosporites, Ancyrospora, and Geminospora in an
incertae sedis group, following the practice of Streel (in Becker, Bless, Streel, and Thorez 1974). No
categories higher than infraturmae are used, and the retusoid miospores are retained in the Laevigati
and Apiculati, because haptotypic features are not here regarded as being of major classificatory
importance.
Figured material is housed in the palaeobotanical collection of the Department of Botany, Bristol
University. Co-ordinates given refer to a Leitz Orthoplan microscope no. 715334. A ringed reference
slide is also provided. Each sample number (i.e. Fair 66) is followed by a strew slide number and then
the appropriate co-ordinates.
Infraturma laevigati Bennie and Kidston emend. Potonie and Kremp 1954
Genus trileites Erdtman ex Potonie 1956
Type species. Trileites spurius Dijkstra emend. Potonie 1956
Trileites langii Richardson 1965
Plate 30, fig. 1
Dimensions (two specimens). Maximum equatorial diameters 141 and 160 p.m.
Lithostratigraphic Range. Ward Hill Sandstone (lb); sample Fair 66 (see text-fig. 2).
EXPLANATION OF PLATE 30
All figures x 400 unless otherwise stated.
Fig. 1. Trileites langii Richardson 1965. Fair 66.2, 48.9, 98.7
Figs. 2, 5. Acinosporites lindlarensis Riegel 1968 var. minor McGregor and Camfield 1976. Fair 37.80, 37.4, 104.8,
x 1000 Distal and proximal surfaces respectively.
Figs. 3, 4. Camptozonotriletes aliquantus Allen 1965. 3, Distal view. 4, x 1000 Distal surface showing
sculpture. Both Fair 66.2, 17.5, 109.8
Fig. 6. Retusotriletes rotundus (Streel) Lele and Streel 1969. Fair 37.79, 19.9, 106.0.
Fig. 7. Calamospora atava (Naumova) McGregor 1964. Fair 37.79, 26.9, 96.5
Fig. 8. Emphanisporites rotatus (McGregor) McGregor 1973. Fair 28.32, 1 14.6, 27.2
Fig. 9. Chelinospora concinna Allen 1965. x 1000. Fair 37.1, 25.9, 1 15.3
Fig. 10. Convolutispora disparalis Allen 1965. x 1000 Fair 24.40, 50.0, 100.3
PLATE 30
284
PALAEONTOLOGY, VOLUME 25
Remarks. Similar to the population described by Richardson (1965), but smaller in over-all diameter.
However, it is comparable with the miospores described by Chi and Hills (1976), and is here referred
to Trileites langii.
Genus retusotriletes Naumova 1953 emend. Streel 1964
Type species. Retusotriletes simplex Naumova 1953
Retusotriletes rotundus (Streel) Lele and Streel 1969
Plate 30, fig. 6
Synonymy, see McGregor (1973, p. 20).
Dimensions (eighteen specimens). Maximum equatorial diameter 56-80 /urn (mean 67 /nm).
Lithostratigraphic range. Ward Hill, Observatory, and Bu Ness Groups (lb to 4a). Found in all samples
examined.
Remarks. This is an example from the wide variety of retusoid forms which occur in the Fair Isle
assemblages.
Genus calamospora Schopf, Wilson, and Bentall 1944
Type species. Calamospora hartungiana Schopf, Wilson, and Bentall 1944
Calamospora atava (Naumova) McGregor 1964
Plate 30, fig. 7
Dimensions (twenty-four specimens). Maximum equatorial diameter 27-71 fan (mean 56 /j.m).
Lithostratigraphic range. Ward Hill Group to Bu Ness Group (lb to 4a). In all samples examined except Fair 33
(see text-fig. 2).
Remarks. Compares well with the emended species described by McGregor (1964), but exhibits a
slightly greater size range.
Infraturma murornati Potonie and Kremp 1954
Genus convolutispora Hoffmeister, Staplin, and Malloy 1955
Type species. Convolutispora florida Hoffmeister, Staplin, and Malloy 1955
Convolutispora dispar alis Allen 1965
Plate 30, fig. 10
Dimensions (six specimens). Maximum equatorial diameter 36-54 jun (mean 46 ^m).
Lithostratigraphic range. Bu Ness Group (4a); samples Fair 24 and 28.
Remarks. As stated in Allen (1965) it is thought likely that sculptural elements of this type result from
corrosion of the exine. It is noticeable that a wide range of sculpture is present, with transitional
forms resembling Raistrickia Potonie and Kremp 1954.
Genus emphanisporites McGregor 1961
Type species. Emphanisporites rotatus McGregor 1961
MARSHALL AND ALLEN: DEVONIAN MIOSPORES
285
Emphanisporites rotatus McGregor emend. McGregor 1973
Plate 30, fig. 8
Synonymy, see McGregor (1973, p. 46).
Dimensions (three specimens). Maximum equatorial diameter 33, 38, and 50 /xm.
Lithostratigraphic range. Observatory and Bu Ness Groups (4a and 2c); samples Fair 28 and 37.
Remarks. Clayton et al. (1977) have documented a series of sporadic occurrences of Emphanisporites
spp. from the later Devonian and early Carboniferous of southern Ireland. Similar sporadic
occurrences were also noted by Richardson (1965) from the Eday Flags of the Orcadian Basin. It
seems likely that the Fair Isle occurrences are only rare examples of a minor but persistent element in
the flora. The decision when to regard elements of an assemblage as reworked or rare is often difficult
(see Captozonotrilites aliquantus), and should be based on features such as stratigraphic persistence,
association with other possible reworked elements, and obvious signs of physical reworking (see
Birks 1970).
Genus acinosporites Richardson 1965
Type species. Acinosporites acanthomammillatus Richardson 1965
Acinosporites lindlarensis Riegel 1968 var. minor McGregor and Camfield 1976
Plate 30, figs. 2, 5
Dimensions (one specimen). Maximum equatorial diameter 48 /xm.
Lithostratigraphical range. Observatory Group (2c); sample Fair 37.
Remarks. The Fair Isle miospore closely resembles specimens described by McGregor and Camfield
(1976) from sediments of Emsian to Givetian age from the Moose River Basin, Ontario. Although
this species has a Geminospora organization, the authors follow McGregor and Camfield (1976), by
placing it in Acinosporites.
Infraturma crassiti Bharadwaj and Venkatachala 1962
Genus aneurospora (Streel) Streel 1967
Type species. Aneurospora goensis Steel 1964
Aneurospora greggsii (McGregor) Streel in Becker et al. 1974
Plate 31, fig. 1
Synonymy, see Streel in Becker et al. (1974, p. 24).
Dimensions (eight specimens). Maximum equatorial diameter 68-116 /xm (mean 84 /xm)
Lithostratigraphic range, restricted to the Bu Ness Group (4a); samples Fair 24, 28, 31, and 33.
Comparisons. Similar problems to those noted by Lele and Streel (1969) and Streel (1972) were
encountered in separating this genus from Geminospora. A possible synonymy is with Archaeo-
zonotriletes nalivkinii Naumova as figured by Chibrikova (1977, pi. XIX, fig. 11), but since no
description was given, it is impossible to make a detailed comparison. Certain specimens also show
similarities with Geminospora svalbardiae Allen 1965.
Infraturma cingulicavati Smith and Butterworth 1967
Genus camptozonotriletes Staplin 1960
Type species. Camptozonotriletes vermiculatus Staplin 1960
286
PALAEONTOLOGY, VOLUME 25
Camptozonotriletes aliquantus Allen 1965
Plate 30, figs. 3, 4
Dimensions (one specimen). Maximum exoexine diameter 77 /un, maximum intexine diameter 57^m.
Lithostratigraphic range. Ward Hill Group (lb), sample Fair 66.
Remarks. The known stratigraphic occurrences for this species are Siegenian to Lower Eifelian (Allen
1967); Upper Siegenian to Lower Emsian (Massa and Moreau-Benoit 1976), and lower Eifelian
(Riegel, 1973, 1974). These occurrences are significantly different from the Givetian age assigned to
the Fair Isle succession, and its appearance may be the result of reworking.
Genus denosporites Berry emend., Potonie and Kremp 1954
Type species. Densosporites covensis Berry 1937
Remarks. Butterworth et coll. 1964 (and in Staplin and Jansonius 1964) in an emendation of the
densospore group of miospores, provided unified limits (based largely on Carboniferous material) for
the subdivision of generic groups. These genera have not proved applicable for the Devonian
densospores from Fair Isle, but until a more unified revision is carried out, we feel they should be
retained.
Densosporites devonicus Richardson 1960
Plate 31, fig. 12
Dimensions (thirty-six specimens). Maximum equatorial diameter 55-120 /xm (mean 85 //.m).
Lithostratigraphic range. Ward Hill to Bu Ness Group (lb to 4a). In all samples examined.
Remarks. Richardson (1965), in his study of Middle Devonian miospores from the Orcadian Basin,
gave the sculptural details, and the relative widths of the light and dark zones of the cingulum as
criteria for distinguishing Densosporites devonicus from D. orcadensis. These same criteria were used
in Fair Isle in an attempt to substantiate the differences between the two species, but no systematic
variation of these characters, as claimed by Richardson (1965, p. 581), was found. McGregor and
Camfield (1976) and McGregor (19796) also had difficulty in distinguishing between the two species.
This raises doubts as to the significance of the stratigraphic distribution of the two species as recorded
by Richardson (1965).
Genus samarisporites Richardson 1965
Type species. Samarisporites orcadensis (Richardson) Richardson 1965.
EXPLANATION OF PLATE 31
All figures x 400 unless otherwise stated.
Fig. 1. Aneurospora greggsii (McGregor) Streel 1974. Fair 31.8, 28.5. 100.6
Fig. 2. Cirratriradites avius Allen 1965. Fair 66.2, 12.8, 95.2
Fig. 3. Samarisporites orcadensis (Richardson) Richardson 1965. Fair 31.5, 12.1, 104.5
Figs. 4, 7. Cirratriradites sp. A. 4, x 1000. Detail of distal sculpture. 7, x400. Fair 66.9, 13.9, 110.6
Figs. 5, 6. Samarisporites mediconus (Richardson) Richardson 1965. 5, Proximal view. 6, Distal view. Fair
37.71, 13.5, 101.2
Figs. 8, 9. Samarisporites conannulatus (Richardson) Richardson 1965. 8, Distal view. 9, Proximal view. Fair
37.80, 5.1, 100.3
Fig. 10. Auroraspora macromanifestus (Hacquebard) Richardson 1960. Fair 37.82, 14.8, 1 12.4
Fig. 1 1 . Auroraspora micromanifestus (Hacquebard) Richardson 1960. Fair 37.82, 7.1, 111.0
Fig. 12. Densosporites devonicus Richardson 1960. Fair 66.3, 4.0, 95.7
PLATE 31
MARSHALL and ALLEN, Devonian miospores
288
PALAEONTOLOGY, VOLUME 25
Remarks. The generic status of Samarisporites is similar to that of the Devonian Densosporites, in
being an element of the densospore complex. The genus Samarisporites, as originally proposed by
Richardson (1965), was to accommodate zonate spores with a variety of solely distal sculptural
elements. The justification for the erection of this genus was that its initial assignation to
Cristatisporites (Potonie and Kremp 1954) was invalid, because the latter has both proximal and
distal sculpture. However, it was suggested by Playford (1971), that with the emendation of
Cristatisporites (Butterworth et coll. 1964) as part of a general reorganization of the densospore
group, that the use of Samarisporites as a distinct generic category could be abandoned. The
emendation, however, still includes the possible presence of a proximal sculpture in the form of a ring
of setae. It also restricts the distal sculpture to being dominantly mammoid in type, and showing no
differentiation in form. The present usage of Samarisporites includes forms with a wide variety of
distal sculpture (e.g. coni, cristae, verrucae) which cannot be accommodated within Cristatisporites
(sensu Butterworth et coll. 1964). It is proposed therefore to use Samarisporites for these species, until
a more unified set of limits for the densospore group is proposed, which accommodates both
Devonian and Carboniferous representatives.
Richardson (1960, 1965) erected three species of Samarisporites based largely on sculptural
differences and their distribution on the distal surface (e.g. central packing in S. mediconus , ring
development in S. conannulatus). The Fair Isle populations contain intermediate forms, and there is
a complete morphological transition series between Richardson’s S. mediconus, S. orcadensis, and S.
conannulatus, which can be considered as occupying distinct positions on the various trends. It would
be interesting to speculate whether the continuous variation this plexus of species exhibits, could be
treated in the same way as in the morphon concept recently outlined by Van der Zwan (1979, 1980).
However, not enough individuals have yet been found to systematically describe the variation both in
a morphological and stratigraphical sense, either to record separate species or varieties in a morphon,
or to combine them as a single species.
Samarisporites mediconus (Richardson) Richardson 1965
Plate 31, figs. 5,6
Dimensions (eight specimens). Maximum equatorial diameter 73-130 ^ m (mean 108 ^m), cingulum 10-26 ^m
wide (mean 17^m).
Lithostratigraphic range. Ward Hill to Bu Ness Group (lb to 4a); samples Fair 31, 37, and 66.
Remarks. Closely resembles the type species, except that there is a greater variety of labra
morphology.
Samarisporites orcadensis (Richardson) Richardson 1965
Plate 31, fig. 3
Dimensions (three specimens). Maximum equatorial diameter 88, 97, and 146 /j,m. Cingulum width 15-20 /xm.
Lithostratigraphic range. Observatory and Bu Ness Groups (2c and 4a); samples Fair 31, 33, and 36.
Remarks. The three individuals found compare closely with Richardson’s holotype. However, they
show greater variety of labra morphology in both length and height.
Samarisporites conannulatus (Richardson) Richardson 1965
Plate 31, figs. 8, 9
Dimensions (five specimens). Maximum equatorial diameter 104-120 (mean 113 ^m). Cingulum 14-21
wide, maximum width interradially.
Lithostratigraphic range. Ward Hill to Bu Ness Groups (lb to 4a); samples Fair 31, 36, 37, and 64.
MARSHALL AND ALLEN: DEVONIAN MIOSPORES
289
Remarks. Compares closely with the type species except that the labra show a greater amount of
variation in length and height.
Genus cirratriradites Wilson and Coe 1940.
Type species. Cirratriradites saturnii (Ibrahim) Schopf, Wilson, and Bentall 1944
Cirratriradites avius Allen 1965
Plate 31, fig. 2
Dimensions. Maximum equatorial diameter 80-130 /xm (mean 105 /xm), twenty-eight species measured.
Maximum intexine diameter 58-78 /xm (mean 67 /xm), eight specimens measured.
Lithostratigraphic range. Bu Ness and Ward Hill Groups (4a and lb); samples Fair 24, 28, 33, 64, and 66.
Remarks. Although the type species of this genus is characterized by having distinctive fovea, the
formal designation of the genus considered this feature to be of little importance. The assignment by
Allen (1965) of Cirratriradites avius to this genus, was made on the presence of the reduced sculpture
and apparently thin equatorial flange. As it may seem desirable to restrict the use of Cirratriradites to
miospores with distinctive fovea, it may in the future be necessary to refer this, and other species of
similar organization, to a new genus within the densospore complex. The intexine is only seen in
overmacerated specimens. Hymenozonotriletes punctomonogrammos Arkhangelskaya (in Filiminova
and Arkhangelskaya 1963), from the Mosolovian of the Central Devonian Field is clearly similar.
Cirratriradites sp. A
Plate 31, figs. 4, 7
Description. Miospores trilete; camerate; amb triangular. Suturae indistinct. Exine two-layered, intexine thin
and closely appressed to the exoexine. Exoexine infrapunctuate, distally sculptured with muri (2-6 /xm high)
fused into a cristoreticulate pattern, more dense in central area. Cingulum with an apparent thin margin, but
possessing a thicker inner zone.
Dimensions (two specimens). Maximum equatorial diameter 125 and 135 /xm.
Lithostratigraphic range. Ward Hill Sandstone (lb); sample Fair 66 only.
Remarks. Differs from Cirratriradites avius in having a dense distal sculpture of muri. Hymenozono-
triletes monogrammos Arkhangelskaya (1963) recorded from the Vorobyevskian, Starskoolian,
Chernoyarian, and Mosolovian beds (Eifelian to Givetian) of the Russian Platform, Volga-Urals,
and Karatau (Filimonova and Arkhangelskaya 1963, Arkhangelskaya 1974, Raskatova 1969,
Chibrikova 1977) is very similar, having a distal network of muri and may prove to be synonymous.
Infraturma patinati Butterworth and Williams emend. Smith and Butterworth 1967
Genus chelinospora Allen 1 965
Type species. Chelinospora concinna Allen 1965
Chelinospora concinna Allen 1965
Plate 30, fig. 9
Dimensions (eight specimens). Maximum equatorial diameter 36-65 /xm (mean 45 /xm).
Lithostratigraphic range. Observatory Group (3c); samples Fair 36 ad 37.
Infraturma monopseudosacciti Smith and Butterworth 1967
Genus auroraspora Hoffmeister, Staplin, and Malloy 1955
Type species. Auroraspora solisortus Hoffmeister, Staplin, and Malloy 1955.
290
PALAEONTOLOGY, VOLUME 25
Population variation in Auroraspora. Four species of Auroraspora have been described from the
Middle Devonian of the Orcadian Basin, separated on features such as size, shape, and the relative
dimensions of the intexine and exoexine layers. The exoexine and intexing diameters for the type-
species populations of Auroraspora micromanifestus Richardson 1960, A. macromanifestus Richard-
son 1960, A. minuta Richardson 1965, and A. aurora Richardson 1960 (with holotype positions
40 60 80 100 120 140 160 180 200 220 240
DIAMETER OF EXOEXINE (pm)
text-fig. 3. Graphical plot of exoexine and intexine diameters for Auroraspora spp.
from Fair Isle. The stars mark the holotype positions of species recorded by
Richardson 1960, 1965. The regression line is for all Auroraspora from Fair Isle.
marked), are plotted in text-fig. 3, and as can be readily seen, the degree of separation is not sufficient
to delimit the species purely by this method. The differences could be related to ontogenetic variation
in a sporangium, with a changing ratio of exoexine to intexine diameters as the miospores increase in
size. Other characters cited as important in delimiting species, such as shape and eccentricity, are
probably not mutually exclusive, and can be related to the presence or absence of labra and their
degree of development, which are again themselves a poor classificatory character. Prominent labra
on a small miospore will give a triangular shape, whereas small labra on a large miospore often results
in a circular shape which, when compressed, can increase the eccentricity of the two exine layers. A
histogram of the exoexine size distribution (text-fig. 4) also provides valuable information, as a
skewed population is seen, which is reminiscent of the pattern produced during the development of
heterospory (see Chaloner 1967). Size distribution is also given for individual collection samples, and
this shows the effects of some sedimentary sorting. However, bimodal populations are apparent in
three of the four assemblages, as well as in the five point moving average line.
It is believed that the species differences are ontogenetic and perhaps reinforced by an incipient
heterospory, but more information is needed on miospore populations found both dispersed and in
situ from single sporangia, before combining certain miospore species. A note of caution is also
necessary in case the lumping of species causes the loss of potentially valuable morphological data
which may be of stratigraphic value as populations of Auroraspora succeed each other, with slightly
different degrees of size distribution (viz. the possible biostratigraphic value of changing over-all
diameters in Retusotriletes from the Russian Platform as shown by Naumova 1953, p. 18, text-fig. 6).
A sensible course in handling species such as those of Auroraspora , is to give precise descriptions and
details of dimensions as illustrated in text-figs. 3 and 4.
MARSHALL AND ALLEN: DEVONIAN MIOSPORES
291
text-fig. 4. Frequency distribution plot for Auroraspora. Note skewed distribution present in different samples
and remains on five point moving average lines.
Auroraspora macromanifestus (Hacquebard) Richardson 1960
Plate 31, fig. 10
Dimensions (seventeen specimens). Maximum exoexine diameters 120-200 ^m (mean 153 /xm), maximum
intexine diameters 60-115 /xm (mean 83 /xm). See text-fig. 3 for size distributions and ratios of central-body
diameter to whole-body diameter (mean 1 : 4).
Lithostratigraphic range. Ward Hill Group to Bu Ness Group (lb to 4a); samples Fair 8, 31, 37, and 66.
Remarks. Differs from specimens described by Richardson (1960) from the Middle Devonian of the
Cromarty area (Scotland) only in size range.
Auroraspora micromanifestus (Hacquebard) Richardson 1960
Plate 31, fig. 11
Dimensions (one hundred and two specimens). Maximum exoexine diameter 65-120 /xm (mean 89 /xm),
maximum intexene diameter 35-85 /im (mean 56 /x m). See text-fig. 3 for size distribution and ratio of central-
body diameter to whole-body diameter (mean 1 : 2).
Lithostratigraphic range. Ward Hill to Bu Ness Group (lb to 4a); samples Fair 8, 24, 28, 31, 33, 36, 37, 64, 66, and
67.
Genus grandispora (Hoffmeister, Staplin, and Malloy) Neves and Owens 1966 sensu Playford 1971
Type species. Grandispora spinosa Hoffmeister, Staplin, and Malloy 1955.
Grandispora Inaumovii (Kedo) McGregor 1973
Plate 32, fig. 5
Description. Miospore trilete, camerate, amb ovate. Suturae accompanied by wavy labra up to 6-5 /xm in height
and individually 2 /xm wide. Exine two-layered, intexine laevigate, indistinct, exoexine 1 /xm thick, shagreenate,
distally sculptured with sparse, gently tapering spines. There are eleven spines around the equatorial periphery
which measure up to 20 /xm in length.
292
PALAEONTOLOGY, VOLUME 25
Dimensions (one specimen). Exoexine diameter 130 x 106 /xm, intexine diameter 80 x 62 /xm.
Lithostratigraphic range. Observatory Group (2c); sample Fair 37.
Comparisons. This miospore closely resembles Grandispora Inaumovii described by McGregor and
Camfield (1976) from the Middle Devonian of the Hudson Bay area (Canada) and by McGregor
(1973) from the Middle Devonian of Gaspe (Canada).
Grandispora velata (Eisenack) Playford 1971
Plate 32, fig. 2
Synonymy, see Owens 1971, p. 46
Dimensions (nine specimens). Maximum exoexine diameter 94-123 /xm (mean 111 /xm), maximum intexine
diameter 57-80 /xm (mean 70 /xm).
Lithostratigraphic range. Ward Hill to Bu Ness Groups (lb to 4a); samples Fair 24, 36, 37, and 66.
Comparisons. The Fair Isle population differs from the type population in having a slightly smaller
sculptural size range.
Grandispora protea (Naumova) Moreau-Benoit 1980
Plate 32, figs. 3, 4
Dimensions (ten specimens). Maximum exoexine diameter 84-184 /xm (mean 136 /xm), maximum intexine
diameter 39-97 /xm (mean 70 /xm).
Lithostratigraphic range. Ward Hill to Bu Ness Groups (lb to 4a); samples Fair 24, 36, 37, and 66.
Genus velamisporites Bharadwaj and Venkatachala 1962
Type species. Velamisporites rugosus Bharadwaj and Venkatachala 1962.
Remarks. Evans (1970) has interpreted the organization of the genus Perotrilites Erdtman ex Couper
1953 as being zonate and not perinate as previously supposed (see Playford 1971). Prior to this,
Palaeozoic perinate miospores were placed in Perotrilites, but these are now removed to
Velamisporites, which was originally thought to be a junior synonym, but now forms a satisfactory
genus for miospores with such an organization.
Velamisporites sp. A
Plate 32, fig. 1
Description. Miospores trilete, camerate, amb triangular to rounded. Suturae accompanied by labra commonly
1 -5-2-5 /xm high (maximum 4 /xm) which continue as folds on to the perine layer. Exine three-layered; intexine
EXPLANATION OF PLATE 32
All figures x 400 unless otherwise stated.
Fig. 1. Velamisporites sp. A. x 1000. Fair 66.2, 19.5, 102.2
Fig. 2. Grandispora velata (Eisenack) Playford 1971. Fair 37.64, 14.3, 100.7
Figs. 3, 4. Grandispora protea (Naumova) Moreau-Benoit 1980. 3, detail of distal sculpture x 1000. 4, x400.
Fair 37.71,49.5, 108.1
Fig. 5. Grandispora Inaumovii (Kedo) McGregor 1973. Fair 37.80, 12.5, 93.2
Figs. 6, 7. Rhabdosporites sp. A. 6, x400. 7, x 1000: detail showing three wall layers. Fair 37.71, 44.4, 95.1
Figs. 8-10. Rhabdosporites langii (Eisenack) comb. nov. 8, ‘ parvulus ’ type Fair 37.62, 19.4, 94.60. 9, 7 angii'
type Fair 37.80, 40.2, 105.5. 10, ‘ parvulus ’ type, compare with Geminospora. Fair 37.82, 40.2, 105.5
PLATE 32
MARSHALL and ALLEN, Devonian miospores
294
PALAEONTOLOGY, VOLUME 25
laevigate, often indistinct, closely appressed to the exoexine. Exoexine commonly infrapunctate, 1-5-4 /xm thick,
with an interradial maximum. Perine wrinkled (especially on the contact areas, where it forms muroid folds)
0-25-0-5 (im thick, sparsely sculptured with parallel sided, flat-topped rods (up to 1 /xm high) mixed with grana
and coni (up to 0-5 /xm high).
Dimensions (seven specimens). Maximum perine diameter 61-76 /xm (mean 66 /xm), maximum exoexine diameter
46-67 /xm (mean 55 /xm). Separation between these two layers 5-17 /xm (mean 1 1 /xm).
Lithostratigraphic range. Ward Hill to Bu Ness Groups (lb to 4a); samples 31, 37, and 66.
Remarks. Many of the described species of Devonian Perotrilites resemble the Fair Isle material.
However, all differ in size and minor sculptural details. Perotrilites pannosus Allen 1965 has a thick
perine with more folds, and coni with bifurcating tips. P. conatus Richardson 1965 has a denser
sculpture of cones and a different ratio of wall-layer dimensions. P. aculeatus Owens 1971 has a
thinner perine and a sculpture of cones. P. selectus (Arkhangelskaya) McGregor and Camfield 1976
is larger and has cones.
Genus rhabdosporites Richardson emend.
Type species. Rhabdosporites langii (Eisenack) Richardson 1 960.
Emended diagnosis. Miospores radial, trilete, amb circular to triangular. Camerate, with separate
exoexine and intexine attached at the proximal pole. Exoexine with or without limbus, sculptured
with low rods, coni, and grana. Intexine laevigate.
Discussion. In 1960 the genus Rhabdosporites was erected to accommodate certain distinctive
pseudosaccate miospores previously described by Eisenack (1944), Lang (1925), and others. Its main
distinguishing features according to Richardson (1960) were the proximally attached non-limbate
bladder and the evenly distributed over-all sculpture of closely packed rods with parallel sides and flat
tops. Subsequently, the genus, and particularly the type species Rhabdosporites langii, has proved
to be one of the most common Devonian miospores, showing a wide stratigraphic range and geo-
graphic distribution. Later workers, however, have not kept to the original diagnosis, placing in
Rhabdosporites limbate miospores as well as those with sculptural elements containing grana and
coni. There have been some good arguments for this; Owens (1971) pointed out a discernible limbus
in the illustration of the type species, and Lele and Streel (1969, p. 102) commented on some well-
preserved specimens from the same horizons as Richardson’s material, with coni often as a dominant
element, as well as rods with truncated tips. It would seem sensible therefore to widen the scope of the
genus to include similar spores, with both a limbus and a more variable ornament of reduced size.
Camerate spores with coarser sculpture are included in other genera (e.g. Grandispora). In widening
the generic concept of Rhabdosporites, we are aware of the close similarity between small specimens
of Rhabdosporites and Geminospora (Balme) Owens 1971, which has been mentioned by previous
authors (e.g. Lele and Streel 1969, p. 103). Geminospora is typified by a thin-walled intexine
either closely appressed to, or showing a variable degree of separation from, a sculptured exoexine
with a thickened distal surface. This organization is difficult to distinguish (without recourse to
microtome sections) from a limbate pseudosaccate miospore (see PI. 32, fig. 8) with a high
intexine -exoexine cavity ratio as seen in small Rhabdosporites. Various indirect criteria were used in
attempts to distinguish the forms, based on the presence of the thickened distal surface which is a
characteristic of the Geminospora group. It might be expected that this would hold the miospore rigid,
with the intexine and proximal exoexine positioned in the shaped distal exoexine ‘cup’, thus limiting
the eccentricity between the two bodies when compressed. The thickened distal exoexine would also
deflect and restrict exoexine folding, to give different folding characteristics on the proximal and
distal surfaces. Further evidence for the similarity (or inability to easily distinguish) between
Rhabdosporites and Geminospora comes from the study of in situ spores (see Allen 1980), where
spores assignable to Rhabdosporites and Geminospora are recorded from closely related progymno-
sperms.
MARSHALL AND ALLEN: DEVONIAN MIOSPORES
295
Rhabdosporites langii (Eisenack) comb. nov.
Plate 32, figs. 8-10
1925 Type B Lang, p. 256, pi. 1, figs. 3-6.
1944 Triletes langii Eisenack, p. 112, pi. 12, fig. 4.
1960 Rhabdosporites langii (Eisenack) Richardson, p. 54, pi. 14, figs. 8, 9.
1963 Rhabdosporites firmus Guennel, p. 256, fig. 12.
1965 Rhabdosporites parvulus Richardson, p. 588, pi. 93, figs. 5-7.
1971 Rhabdosporites micropaxillus Owens, p. 49, pi. 15, figs. 3-7.
1973 Rhabdosporites sp. Hamid, p. 202, pi. 10, no. 1.
For additional synonymies, see Moreau-Benoit, 1980, pp. 29 and 30.
Dimensions (one hundred and sixty-five specimens). Maximum exoexine diameter 45-166 (mean 90 /un),
maximum intexine diameter 39-146 ^m (mean 65 pm).
Lithostratigraphic range. Ward Hill to Bu Ness Group (lb to 4a)— samples Fair 8, 24, 28, 31, 33, 36, 37, 64, 66,
and 67.
Remarks. Rhabdosporites parvulus Richardson 1965 was associated with R. langii in all the samples
studied, and attempts were made to discriminate between the two species as did Richardson (1965,
p. 588), using a graphical plot of exoexine and intexine diameters. A graph of the data from Fair Isle
(text-fig. 5) shows no obvious separation into two populations, but a gradual change in the ratio of
160
140
El 20
1
wlOO
40
40 60 80 100 120 140 160 180
DIAMETER OF EXOEXINE (pm )
text-fig. 5. Size variation in Rhabdosporites langii from Fair Isle compared with published
data from the Orcadian Basin. Reduced major axis lines are seen to be closely comparable
between Fair Isle and a combined R. langii and R. parvulus population from the Orcadian
Basin. Separate R.M.A. lines for R. langii Richardson 1960 and R. parvulus Richardson 1965
only apply to parts of the population which is clearly seen to be continuous. The boxes show
the limits of the type-species population, and the stars show the holotype position. Larger
points refer to two or three individuals.
296
PALAEONTOLOGY, VOLUME 25
text-fig. 6. Frequency plots for Rhabdosporites exoexine diameters.
the exoexine and intexine diameters as the spore size increases. Further, the data from Richardson’s
graph were replotted as a single population, and reduced major axis lines calculated for the combined
data. It can be seen from this result that there is no obvious difference between regression lines for [
combined populations of R. langii and R. parvulus and similar lines calculated for the separate species
as defined by Richardson. A regression line calculated for the Fair Isle population shows a strong j
similarity to the combined Richardson plot. A frequency distribution plot of the exoexine diameters
(see text-fig. 6) gives a smooth population curve and no sign of bimodality, which might be expected if i
two distinct species were present. The stratigraphic range chart of Richardson (1965, opposite p. 590) j
shows a disjunct appearance for R. parvulus only in the Eday Group, whilst the text (p. 588) does
indicate its appearance further down the sequence, such that it is coincident with R. langii. \
Rhabdosporites has also been identified as an in situ miospore in both Tetraxylopteris schmidtii (Beck)
Bonamo and Banks 1967 and in Rellimia thompsonii Leclercq and Bonamo 1973 (in Leclercq and
Bonamo 1971). These occurrences also provide further information on the similarity between i
Rhabdosporites langii and R. parvulus , as both of these descriptions also comment on the presence of
miospores with a size range including both R. langii and R. pavulus in the same sporangium. These
authors also consider that the differences described by Richardson are ontogenetic, and as miospores S
matured inside a sporangium, the exoexine expanded more rapidly than the intexine, showing an
increasing ratio between the two wall diameters as well as increased sculpture size. It is suggested that
the continua seen in the dispersed species can be attributed to this ontogenetic variation, and that
with the size distribution data collected for these in situ miospores to confirm this similarity there is
now no reason to consider langii and parvulus as separate species. If the enlarged concept of the genus
to include limbate spores with sculptural elements of grana and coni as well as rods is accepted, then
R. micropaxillus Owens 1 97 1 is intermediate between R. langii and R. parvulus and can be regarded as
synonymous; as are R. firmus Guennel 1963 and Rhabdosporites sp. Hamid 1973, which similarly
differ only in the presence of grana and coni amongst the sculptural elements.
MARSHALL AND ALLEN: DEVONIAN MIOSPORES
297
Rhabdosporites sp. A
Plate 32, figs. 6, 7
Description. Miospores trilete; amb circular; exine three-layered. Suturae length equal to inner body radius,
accompanied by simple labra (each 1 /xm wide). The two inner wall layers laevigate, homogeneous; the outer wall
infrapunctate, sculptured with coni, rounded-tipped rods, and grana (0-25-1 high). The three walls are
variable in thickness and carry folds which show them to be separated except on the proximal surface. Outer wall
appears limbate.
Dimensions (three specimens). Maximum diameters: outer wall 90 /xin, 101 /urn, 96 jim; middle wall 74 (im, 67 /im,
80 ^m; inner wall 58 jum, 43 ^m, 61 ^m. Optically discernible wall thicknesses: outer wall 2-0 /u.m, 2-5 /xm, 0-5 /xm;
middle wall 0-25 /xm, TO /xm, 0-5 /xm; inner wall 2-0 /xm, 1-5 /xm, 2-0 /im.
Lithostradgraphic range. Observatory Group (2c); sample Fair 37.
Remarks. Rhabdosporites sp. A of Richardson 1965 is similar in having three layers, and differs only
on size and minor sculptural features. ICalyptosporites sp. A Mortimer and Chaloner 1972 appears to
be conspecific, and these authors also suggest a possible further synonymy with a miospore figured by
Ozolin'a (1960a, pi. 1, fig. 29) and identified as Archaeozonotriletes micromanifesus var. minor
Naumova. Hemer and Nygreen (1967) illustrate a triwalled miospore as Rhabdosporites sp. but as no
description was provided, further comparison is impossible. Massa and Moreau-Benoit (1976)
record the presence of Rhabdosporites sp. A Richardson, from their palynozones 6 and 7, which are
dated as late Givetian and early Frasnian.
Similar in situ miospores have been recovered by Tetraxylopteris schmidtii Bonamo and Banks
1967, where they occurred in subordinate numbers to specimens similar to Rhabdosporites langii and
R. parvulus. Bonamo and Banks (1967) included a personal communication from Richardson, which
expressed the opinion that the extra layer could have arisen by a splitting of one of the two wall layers
of Rhabdosporites. The middle wall of the Fair Isle specimens appears to be a more definite layer than
in similar miospores illustrated from T. schmidtii. A possible explanation for this difference is that the
third layer may be a teratological feature, and a possible way of showing this, as opposed to the wall
splitting hypothesis, is to sum the various thicknesses of the wall layers. It might be expected that any
wall splitting would provide a total equal to the mean intexine plus exoexine value of a normal
Rhabdosporites. Unfortunately the data are not available to show this, as no wall-thickness
measurements are provided for the Cromarty populations of these spores. The presence of a limbate
wall feature would also create difficulties and uncertainty in measurement.
INCERTAE SEDIS
Genus hystricosporites McGregor 1 960
Type species. Hystricosporites delectabilis McGregor 1960.
Hystricosporites cf. corystus Richardson 1960
Plate 33, figs. 7-9
Description. Miospore trilete; amb circular to triangular, ovate in lateral compression. Suturae accompanied by
prominent labra (19-36 /xm high, 56-116 /xm in length as seen in lateral compression). Exine two-layered,
intexine laevigate, closely appressed to the exoexine. Exoexine with paired radial muri on the contact areas
(length 20-28 /xm, width 4-7 /xm), proximo-equatorially and distally sculptured with 5-14 grapnel- tipped spines
(20-52 /xm long).
Dimensions. Maximum equatorial diameter 60-130 /xm (mean 96 /xm), eleven specimens measured. Maximum
height (excluding labra) in lateral compression 65-122 pm, six specimens measured.
Lithostradgraphic range. Ward Hill to Bu Ness Groups (lb to 4d); samples Fair 8, 24, 31, 33, 36, 37, 64, 66,
and 67.
298
PALAEONTOLOGY, VOLUME 25
Remarks. The Fair Isle population is very similar to Hystricosporites corystus Richardson 1962,
differing only in gross size. Richardson’s Hystricosporites cf. corystus should not be confused with the
Fair Isle miospores.
The proximal radial muri and intexine are seen only in overmacerated specimens. Frequently only
the bases of the spines are complete, but when intact the grapnel-tips are of the laterally extended
form (see Owens 1971, p. 87). Difficulties were encountered in relating spores with intact spines to
those showing proximal lip and intexine detail, because the high maceration levels needed to clear the
miospores resulted in severe erosion of the grapnel-tips.
Genus ancyrospora Richardson emend. Richardson 1962
Type species. Ancyrospora grandispinosa Richardson 1960.
Ancyrospora ancyrea (Eisenack) Richardson 1 962
Plate 33, figs. 3, 10
Dimensions (seventeen specimens). Maximum equatorial diameter 81-165 /un (mean 120 (xm).
Lithostratigraphic range. Ward Hill to Bu Ness Groups (lb to 4a), samples 8, 24, 28, 31, 33, 36, 37, 64, 66,
and 67.
Remarks. Placed in Ancyrospora ancyrea and not A. ancyrea var. ancyrea Richardson 1962 because it
does not show the wide flange development which characterizes the variety. It conforms to text-fig. 5
(p. 178) of Richardson (1962), and is morphologically average for the three described varieties.
The same maceration problems were encountered in studying this species as with Hystricosporites
cf. corystus , in that it was rare to find completely cleared spores in which the grapnel-tipped spines
were preserved.
Ancyrospora ancyrea cf. var. brevispinosa Richardson 1962
Plate 33, figs. 4-6
Description. Miospores trilete; amb triangular to subtriangular. Suturae accompanied by labra (1 ^m high).
Exine two-layered, intexine laevigate, thin, closely appressed to the exoexine. Exoexine shagreenate, with a dark
halo, or triangular darkening in the proximal polar area, contact areas often with sinuous folds. Proximo-
equatorially and distally sculptured with grapnel-tipped spines (1-10 /un long), the tips 1 -2 across have little
discernible detail (PI. 33, fig. 6). Flange development variable (0-12 ^ m wide) with some very narrow forms
(PI. 33, fig. 5) and others interradially.
Dimensions (one hundred and fifty specimens). Maximum equatorial diameter 39-95 ^m (mean 65 /urn).
Lithostratigraphic range. Ward Hill to Bu Ness Group (lb to 4a); samples 8, 24, 28, 31, 33, 36, 37, 64, 66, and 67.
EXPLANATION OF PLATE 33
All figures x 400 unless otherwise stated.
Fig. 1. Geminospora sp. B. Fair 31.5, 6.3, 101.1
Fig. 2. Geminospora sp. A. Fair 31.5, 36.6, 101.7
Figs. 3, 10. Ancyrospora ancyrea (Eisenack) Richardson 1962. 3, Fair 37.64, 29.3, 107.1. 10, detail of spine
process, x 1000.
Figs. 4-6. Ancyrospora ancyrea cf. var. brevispinosa Richardson 1962. 4, wide flange form. Fair 8.1, 11.4, 106.1.
5, narrow flange form. Fair 37.64, 14.4, 96.9. 6, detail of spine process from 5, x 1800.
Figs. 7-9. Hystricosporites cf. corystus Richardson 1960. 7, detail of spine process from 9. 8, well oxidized
specimen. Fair 67.30, 100.6, 33.4. 9, dark specimen, lateral compression. Fair 37.63, 9.1, 96.9.
Fig. 11. Geminospora tuberculata (Kedo) Allen 1965. x 1000. Fair 37.64, 22.4, 99.2
Fig. 12. Geminospora svalbardiae (Vigran) Allen 1965. x 1000. Fair 28.20, 50.2, 1 14.3
PLATE 33
300
PALAEONTOLOGY, VOLUME 25
Remarks. This spore is closely similar to Ancyrospora ancyrea var. brevispinosa Richardson 1962
except in gross diameter, which seems to reflect a genuine difference, as opposed to modifications in
the population based in sedimentary sorting processes. The evidence for this (see text-fig. 7) is that
whilst the size frequency distributions for Ancyrospora ancyrea from both Fair Isle and the Orcadian
Basin (see Richardson 1962, p. 185) clearly parallel each other, those for var. brevispinosa (this paper)
are quite different. It is postulated on a simple model that any sorting process which could modify the
size-frequency distributions of the variety brevispinosa and cf. brevispinosa to this extent, would also
show a difference in the Ancyrospora ancyrea populations, which is not so.
2 50
_A. ancyrea var. brevispinosa^
type pop. Rich. ’62
N> “
8 5
70 90
MIOSPORE DIAMETER (pm)
text-fig. 7. Frequency plot for Ancyrospora ancyrea cf. var. brevispinosa. There is a difference in
population size range compared with that of Richardson 1 962, which is not reflected in A . ancyrea.
The much-reduced flange on the exoexine (see PI. 33, fig. 5) produces some problems in using the
genus Ancyrospora, which, as defined by Richardson (1962) in his emended diagnosis, has an
extended equatorial flange of pseudoflange. The evidence from the Fair Isle sequence suggests that a
continua exists from obviously flanged miospores to almost flangeless ones, and therefore in part
outside the generic diagnosis. The presence of sinuous surface folds on the exoexine and the
occasional appearance of a membranous top exine layer, suggests the possibility that a third layer is
developed, similar to A. fallacia Urban (1970), and frequently lost on oxidation. Another point of
interest is a possible relationship with Perotrilites bifur catus Richardson 1962, which could be
interpreted as an overmacerated Ancyrospora with the spalling off of a third outer exine layer
normally closely appressed to the exoexine.
geminospora (Balme) Owens 1971
Type species. Geminospora lemur ata Balme 1962
Discussion. As well as the difficulties discussed earlier in separating Rhabdosporites from Gemino-
spora, there is an apparent morphological transition series between Geminospora and Grandispora in
our material. The problem has previously been described in other assemblages by Neves and Owens
(1966) and Playford (1971). Where one draws the dividing line between truly camerate miospores
such as Grandispora which show a clear separation of intexine and exoexine (sensu Playford 1976
and not McGregor 1973), and Geminospora which has both widely separated and closely appressed
MARSHALL AND ALLEN: DEVONIAN MIOSPORES
301
wall layers in the same population, is at present arbitrary. A further difficulty is that increased
separation of wall layers may result from the oxidative processes used in clearing the highly
carbonized Fair Isle miospore assemblages.
Geminospora tuberculata (Kedo) Allen 1 965
Plate 33, fig. 1 1
Dimensions (eleven specimens). Maximum equatorial diameter 42-67 /xm (mean 56 /am).
Lithostradgraphic range. Ward Hill to Bu Ness Groups (lb to 4a); samples Fair 24, 28, 31, 33, 36, 37, 66, and 67.
Remarks. The Fair Isle material differs only from that described by Allen (1965) in having slightly
smaller sculptural elements and greater variation of haptotypic features. The emendation of
Archaeozonotrilites tuberculatus Kedo by Allen (1965) to accommodate the Spitsbergen population
of Geminospora is open to some modification as more Soviet miospores of closely related species have
now been described, into which parts of the population could be accommodated. We believe that
there are problems of synonymy between Geminospora and Archaeozonotriletes (sensu Naumova)
and a thorough separate revision is preferable to minor amendments made here.
Geminospora svalbardiae (Vigran) Allen 1965
Plate 33, fig. 12
Dimensions (nine specimens). Maximum equatorial diameter 47-72 (im (mean 58 jam).
Lithostradgraphic range. Ward Hill to Bu Ness Groups (lb to 4a); samples Fair 8, 31, 33, 36, 37, and 66.
Remarks. The population described by Allen (1965) is very similar to the Fair Isle material, differing
only in being of larger over-all size and having a more reduced sculpture. Possible synonymies include
Geminospora maculata Taugourdeau-Lantz 1967 and Geminospora plicata Clayton and Graham
1 974 (nomen illeg. —junior homonym for G. plicata Owens 1971). Both these forms, as described, are
similar, and differ only in minor features of size and type of sculpture.
Geminospora sp. A
Plate 33, fig. 2
Description. Miospore trilete; amb circular to triangular. Suturae straight and extending to the intexine margin.
Exine two-layered; intexine laevigate, thin with conspicuous folds, and of variable separation from the exoexine.
Exoexine 5 /xm thick, with a sculpture of coni (2 jam high), some of which are arcuate or biform, typical
height 2 jam.
Dimensions (one specimen). Maximum equatorial diameter 70 /x m.
Lithostradgraphic range. Observatory Group (2c); sample Fair 37.
Remarks. Similar miospores have been described from the Soviet Union as Archaeozonotriletes
visendus Chibrikova, A. visendus var. echinatus Chibrikova, A. egregius Naumova, A. tuberculatus
var. aculeatus Raskatova, A. acutus Raskatova, A. pensus Kedo, A. lasius var. minor Naumova.
Unfortunately its single occurrence in the Fair Isle sequence limits its value at present.
Geminospora sp. B
Plate 33, fig. 1
Description. Miospores trilete; amb roughly triangular. Suturae simple or accompanied by thin labra commonly
1 -2 jam high (maximum 5 /am). Exine two-layered with exoexine and intexine layers separate, attached only at the
proximal pole. Intexine laevigate, thin, showing a variation in degree of separation. Exoexine 2-5—1 4 /am thick
(mean 6-5 /xm) with individuals showing an interradial maximum (up to 2-5 /am thicker). Exoexine sculpture
302
PALAEONTOLOGY, VOLUME 25
(except for contact areas) with cones and fine tapering spines 1 -5 /xm long (mean 2-5 jim), denser and of greater
length distally than equatorially or proximo-equatorially.
Dimensions (twenty-seven specimens). Maximum exoexine diameter 65-140 fxm (mean 89 /xm).
Lithostratigraphic range. Ward Hill and Observatory Groups (lb to 2c); samples. Fair 37 and 66.
Remarks. Archaeozonotriletes comptus Naumova (1953), A. comptus var. expletivus Chibrikova
(1959) (see also Raskatova 1969 and Chibrikova 1962, 1977), A. tuber culatus Kedo var. minor Kedo
(1976), and A. tuber culatus Kedo var. triangulatus Kedo 1976 are all miospores of a similar
construction, sculpture, and size. However, their descriptions are not detailed enough to confidently
place Geminospora sp. B into any one of these species.
SEQUENCE AND STRATIGRAPHIC SIGNIFICANCE OF THE
MIOSPORE ASSEMBLAGES
The number of miospore-bearing horizons has proved to be small, but fortunately of wide
distribution. This is shown for selected miospore taxa in text-fig. 8. There is no major palynostrati-
graphic change in the succession until the Bu Ness Group, which is characterized by the incoming of
Aneurospora greggsii and Convolutispora disparalis.
The general problems of Devonian palynostratigraphy have been discussed at length by several
authors (Owens and Richardson 1972; Richardson 1974; and McGregor 1979a) so will only be briefly
alluded to in this stratigraphical synthesis. The total stratigraphical ranges for some of the taxa from
Fair Isle are plotted in text-fig. 9 and show long ranging distributions. This type of compilation
suffers from the lack of miospore assemblages securely dated by independent means (notably
Ammonoids, Conodonts), which can be compared accurately with the type sections of the traditional
stages. The data suggests a Givetian age for the Fair Isle assemblages but does not place reliance on
the restricted ranges of several poorly known species. A more precise correlation is possible by
drawing comparisons with selected areas containing well-dated assemblages.
Comparisons with European Assemblages ( excluding the Soviet Union but including Spitsbergen)
The spore sequences described by Richardson (1965) for the Orcadian Basin are now geographically
the closest described Devonian palynofloras to Fair Isle. The abundance in Fair Isle of Ancyrospora
ancyrea cf. var. brevispinosa with the first appearance of Grandispora Inaumovii suggest possible
equivalence with the Eday Beds. An interesting comparison, and possibly significant, is the very rare
occurrences of Emphanisporites spp. in both Fair Isle and the Orcadian Basin, which are not thought
to be the result of reworking (Clayton et al. 1977). A major difference between the Fair Isle beds and
the Eday Group on Orkney is the complete absence of Geminospora spp. in the latter, compared with
its relative abundance and diversity in Fair Isle. It could be argued that this is of stratigraphic
importance, except that Geminospora spp. is known from independently dated Givetian beds from
over a wide area including Illinois (Sanders 1968), Spitsbergen (Allen 1965), and southern England
(Mortimer and Chaloner 1972). It has been suggested (Richardson 1967) that the control is
ecological, and this will be discussed later.
The spore assemblages described from borehole material of southern England by Mortimer and
Chaloner (1972) compare quite closely, but add little to the specific age assignment apart from
confirming it as Givetian. Detailed comparisons of the Fair Isle spore assemblages can also be made
with the sequences described by Streel and associates (reviewed in Streel 1972) from Belgium and
northern France. Streel in a discussion of marker spores for the Givetian-Frasnian boundary
considers the first appearances of Aneurospora greggsii, Ancyrospora langii, and Samarisporites
hesperus to be of major significance. The occurrence of Aneurospora greggsii in the Bu Ness Group
does suggest that this higher part of the Fair Isle sequence is late Givetian in age and close to the
Givetian-Frasnian boundary. Streel (1972) also gave the range of Grandispora velata and
IRhabdosporites langii (closely comparable with the Fair Isle form) from the Tournai borehole as
MARSHALL AND ALLEN: DEVONIAN MIOSPORES
1
1'
O W > O
3 3 3 § 3
2 5' 5' & 2
(/)(/)(/></)
3 2.
text-fig. 8. Stratigraphical distribution of selected miospores in Fair Isle.
overlapping the range of Aneurospora greggsii, but the information was not sufficient to suggest that
this was of any widespread stratigraphic value. Loboziak and Streel (1980), in a study of Givetian and
Frasnian rocks from Boulonnais (France), considered several species to be of importance in
delimiting beds of late Givetian age, including Chelinospora concinna, Ancyrospora ancyrea var.
brevispinosa, and Convolutispora disparalis, all found in the Fair Isle section.
In Spitsbergen, Allen ( 1 967) described a zonal scheme of three assemblages for the Devonian rocks.
The Fair Isle sequence has many similarities to the triangulatus assemblage with Geminospora
tuberculata, Cirratriradites avius, Convolutispora disparalis, Densosporites devonicus, Chelinospora
concinna, and Grandispora protea in common. The triangulatus assemblage was assigned to the
Sampless sS 8S
Gp. Obs. Gp.
Hill Gp.
Convolutispora disparalis
Aneurospora greggsii
Samarisporites orcadensis
Grandispora? naumovii
-• Velamisporites sp. A
-• Grandispora velata
"• Samarisporites mediconus
Grandispora protea
"• Geminospora svalbardiae
Geminospora tuberculata
— • Samarisporites conannulatus
Auroraspora macromanifestus
Retusotrilites rotundus
Calamospora atava
Hystricosporites cf. corystus
Oensosporites devonicus
Cirratriradites avius
-+-+ Ancyrospora ancyrea cf. var. brevispinosa
Ancyrospora ancyrea
Geminospoa sp. B
Rhabdosporites langii
-+-• Auroraspora micromanifestus
i
I
.3
1
1
2
304
PALAEONTOLOGY, VOLUME 25
text-fig. 9. Stratigraphical distribution of selected taxa which occur on Fair Isle. * Refers to comparison record.
Givetian by Allen (1967) although the lower part is without Densoporites devonicus and is considered
to be of possible Eifelian age (McGregor and Camfield 1976).
A detailed study has been made by Beju (1972) on rocks of Devonian age from the Moesian
platform of Romania. The Eifelian subzone (D2a) has the long-ranging Grandispora protea and
Calamospora atava in common with Fair Isle. The Givetian subzone (D2b) is more similar, with
Ancyrospora ancyrea, Rhabdosporites langii, Grandispora velata, Auroraspora macromanifestus, and
Densosporites devonicus in common. These last three are considered by Beju to be specifically
characteristic of this subzone. The Frasnian zone (D3) contains Geminospora svalbardiae, G.
tuberculata, Rhabdosporites parvulus, and Auroraspora micromanifestus.
Comparisons with North American Assemblages
The Frasnian Ghost River spore assemblage described from Alberta by McGregor (1964) compares
with the Bu Ness Group of the Fair Isle sequence only in the presence of Aneurospora greggsii.
McGregor (in McGregor and Uyeno 1972) described a sequence of spore assemblages from the
Canadian Arctic Archipelago, and of these the closest to the Fair Isle material is assemblage D of late
Givetian age, characterized by the appearance of Grandispora protea and Chelinospora concinna. The
disappearance of Densosporites devonicus is not held to be of major significance as its replacement by
D. orcadensis in the Orcadian sequence may be anomalous in view of the apparent synonymy which
exists between the two species (see McGregor and Camfield 1976; McGregor 1979a and this paper).
Devonian spores described by Owens (1971) from Middle and Upper Devonian rocks of the
Canadian Arctic Archipelago, suggest that broad similarities can be drawn, but his work is largely
based on new species from a few horizons, and is of limited value for detailed comparisons.
MARSHALL AND ALLEN: DEVONIAN MIOSPORES
305
Sanders (1968) described the miospores from an acritarch-dominated assemblage of Givetian age,
with Rhabdosporites langii (and R. parvulus), Geminospora tuberculata , and examples of Ancyrospora
spp. and Velamisporites spp. ( Perotrilites ).
Comparisons with North African and the Middle-East Assemblages
Massa and Moreau-Benoit (1976) described a series of miospore zones covering the Devonian strata
of western Libya, and limited comparisons can be drawn with the Fair Isle succession. Palynozones
4 and 5 are the most similar, having Auroraspora micromanifestus, Emphanisporites rotatus,
Rhabdosporites parvulus (R. langii), Grandispora velata, Geminospora svalbardiae, and Samarisporites
mediconus in common, and these are dated as Couvinian and lower Givetian respectively.
Comparisons with miospore assemblages from the Russian Platform
General comparisons. Table 1 lists some miospore taxa from the Russian Platform together with their
stratigraphical ranges, which are considered to be broadly comparable with taxa from Fair Isle.
Comparability is variable, ranging from likely direct equivalence to part synonymy or closely
similar. Some species, for example Hymenozonotriletes protea, can be referred to several western
species.
In the compilation of Table 1 it was hoped to keep the stratigraphic information in the form of the
local units used over the Russian Platform, but correlation difficulties made it impossible to construct
a single table for the data, so that a more generalized scheme was necessarily produced. The placing
of the Eifelian-Givetian and the Givetian-Frasnian boundaries is open to question, with some
amendments having been made in certain areas. Hymenozonotriletes punctomonogrammos and H.
monogrammos (Arkhangelskaya 1974), for example, are found in the Mosolovian and Chernoyarian
beds of the Central and Eastern Russian Platform, and are part of the polymorphos-monogrammos
zone which has many taxa similar to the Fair Isle material, but is placed in the Eifelian by
Arkhangelskaya. A similar problem is seen in the Givetian-Frasnian beds of the Baltic region, where
the Soviet workers draw the boundary at the base of the Gauja formation (e.g. Ozolin'a 1963),
whereas recently compiled data (Westoll 1979 and pers. comm.), places it at the base of the Snetogor,
which gives this succession great importance as being continuous across the Middle/Upper Devonian
boundary and containing good vertebrate faunas. Unfortunately although this section appears
to be well documented (e.g. Ozolin'a 1955, 1960a, 1960 b, 1961, 1963; Kedo 1966) the miospore
illustrations and descriptions are poor, making comparisons difficult. The compilation of data
presented in Table 1 shows a Givetian age for the Fair Isle assemblages but it is impossible to give
greater precision with this type of synthesis.
Comparisons with zonal schemes from the Russian Platform
Several zonal schemes have been published which cover the relevant stratigraphic interval, and
specific comparisons are made in an attempt to give a more detailed correlation. Kedo (1966)
produced a zonal scheme for the western part of the Russian Platform and although there are some
taxa in common, it is very difficult to make a precise correlation, except to record an influx of
Archaeozonotriletes (sensu Naumova = Geminospora spp.) in the upper part of the Givetian, which
compares with Fair Isle, but contrasts with the Orcadian Givetian which does not contain
Geminospora. This indicates either younger Givetian rocks in Fair Isle than in the Orcadian Basin, or
the possibility of an ecological control (see below).
Raskatova (1969, 1974) has produced a more compatible scheme which again shows an influx
of Archaeozonotriletes in the upper Givetian (zones IV to VIII) Vorbyevkian to Staryi’ oskolian.
Hymenozonotriletes punctomonogrammos Arkhangelskaya and H. monogrammos Arkhangelskaya
are also present, but occur lower down in the uppermost Eifelian as well as the Givetian, so that the
age assignment is not clear cut.
From the same horizons, Arkhangelskaya (1974) has also recorded these two species of
Hymenozonotriletes as well as taxonomic equivalents of Rhabdosporites langii, Densosporites
devonicus, Grandispora protea, and Samarisporites spp. These were compared with the Weatherall
306
PALAEONTOLOGY, VOLUME 25
table 1. Comparative table of selected Soviet and western taxa occurring on Fair Isle. Records include likely
synonymies, part synonymies, and close comparisons. Sources include Ozolin'a 1955, 1960a, 19606, 1961, 1963,
Kedo 1955, 1960, 1966, 1967, Chibrikova 1959, 1962, 1972, 1977, Chibrikova and Naumova 1974,
Arkhangelskya 1963, 1974, Naumova 1953, Andreeva 1962, Raskatova 1969, 1974, and Pokrovskaya 1966.
WESTERN TAXA
SOVIET COMPARISONS
EIFELIAN
GIVETIAN
FRASNIAN
Archaeotri letes sincerus Kedo
•
•
Ancyrospora ancyrea
Archaeotriletes splendidus Kedo
•
Archaeotri letes hamulus Naumova
•
•
Retusotri letes verrucosus (Naumova in litt.) Kedo
•
Aneurospora greggsii
Re tusotri letes punctatus Chibrikova
•
•
Archaeozonotriletes nalivkinii Naumova
•
•
Cirratriradites avius
Hymenozonotri letes punctomonogrammos Arkhangelskaya
•
Cirratriradites sp. A
Hymenozonotri letes monogranmos Arkhangelskaya
•
Densosporites devonicus
Hymenozonotri letes meonacanthus Naumova
•
•
•
Archaeozonotriletes meonacanthus Nomen nudum
•
Geminospora tuberculata
Archaeozonotriletes tuberculatus Kedo
•
Archaeozonotriletes plicata (Naumova in litt.) Kedo
•
Retusotri letes parvimammatus Kedo
•
•
Geminospora svalbardiae
Archaeozonotriletes notatus Naumova
•
Archaeozonotriletes rugosus Naumova
•
•
Archaeozonotriletes visendus Chibrikova
•
•
Archaeozonotriletes egregius Naumova
•
•
Archaeozonotriletes tuberculatus var. aculeatus Raskatova
•
Geminospora sp. A
Archaeozonotri letes acutus Raskatova
•
Archaeozonotriletes visendus var. echinatus Chibrikova
•
Archaeozonotriletes pensus Kedo
•
Archaeozonotriletes lasius var minor Naumova
#
Archaeozonotriletes comptus Naumova
#
Geminospora sp. B
Archaeozonotriletes comptus var. expletivus Chibrikova
•
Archaeozonotriletes tuberculatus var. minor Kedo
•
Archaeozonotriletes tuberculatus var. triangulatus Kedo
•
Grandispora naumovii
Archaeozonotriletes naumovii Kedo
•
•
Hymenozonotri letes curticonicus Kedo
Hymenozonotri letes ventosus Kedo
•
Grandispcra protea
Hymenozonotri letes proteus Naumova
Hymenozonotri letes verus Naumova
•
•
Hymenozonotri letes endemicus Chibrikova
•
Hymenozonotri letes echini formis Naumova
•
•
Hymenozonotri letes longus Arkhangelskaya
•
•
Grandispora velata
Hymenozonotri letes tener var. concinna Chibrikova
•
Hymenozonotri letes proteus Naumova
•
•
Grandispora velata (Richardson) Playford
•
Hymenozonotri letes facetus Arkhangelskaya
•
•
Hymenozonotri letes polymorphus (Naumova) Kedo
•
•
Rhabdosporites langii
Archaeozonotriletes micromanifestus Naumova
•
•
Hymenozonotri letes varius Naumova
•
•
Rhabdosporites parvulus Richardson
•
•
Formation from Arctic Canada (Givetian), the Thurso and Eday Beds from the Orcadian Basin
(Givetian), and the Achanarras Fish Bed (Eifelian-Givetian). The recent zonal scheme produced by
Chibrikova and Naumova (1974) is not very detailed with regard to specific taxa, and is largely
restricted to generic distribution. The appearance of Archaeozonotriletes is considered significant,
and is used as a marker for the Givetian stage. The zonal scheme produced by Naumova (1953) is
difficult to apply, as the taxa are not often placed in a detailed zonal scheme, but in broad categories
MARSHALL AND ALLEN: DEVONIAN MIOSPORES
307
such as Middle or Upper Devonian. The Fair Isle assemblages fit best into the Givetian and Frasnian
groups.
A compilation of these occurrences has been presented by Andreeva (in Pokrovskaya 1966) and,
as indicated, uses Chibrikova’s (1962) and Naumova’s (1953) work. The significant occurrence of
Archaeozonotriletes ( = Geminospora) with subordinate Hymenozonotriletes is used as a marker for
the Givetian stage. The spores recovered by Chibrikova (1962) from the Vorobyevkian of Western-
Bashkiria contain several species which are similar to the Fair Isle assemblages, with examples such as
Archaeozonotriletes comptus var. expletivus, A. meoncanthus, A. visendus, and A. egregius.
00
3
3
*2.
N
o
3
3
$
rt
c
D.
3 $
a §
^ S'
,3
Other
fT
CD
O
S'
2
z
o
sS
o
o
3
^ £
2 ^
a
3
O
"3
o
2
ft
St
a *p
Fair
24
6
14
39
2
20
19
Bu Ness Group Fair
28
4
39
3
0-5
30
23-5
Fair
33
2
19
15
3
53
8
Observatory Group Fair
36
8
47
11
2
13
19
Fair
67
5
37
2
27
27
2
Ward Hill Group Fair
66
6
33
36
-
21
1
Fair
64
19
35
2
5
38
1
table 2. Relative proportions of major miospore types from Fair Isle. The figures
are percentages of numbers of each group from 150 total counts.
Ecological considerations
An ecological control on the distribution of some Devonian miospores has been speculated by several
authors (e.g. Streel 1967; Richardson 1965, 1967, 1969), and the example cited is usually that of
Geminospora spp. {Archaeozonotriletes of Soviet palynologists). These miospores are thought to be
dispersed from plants growing on and around the lower flood plain and marginal marine areas, whilst
miospores from the bifurcate and pseudosaccate zonate groups are more continental in aspect and
related to upper flood plain and possibly lacustrine deposits. The assemblages from Fair Isle are
relevant to this discussion as they often contain abundant Geminospora spp. (fluctuating from 39% to
2%, see Table 2), in what appears to be a true internal basin facies. The presence of Svalbardia scotica
(Chaloner 1972) in the Bu Ness Group is also significant, because other species are known to contain
in situ spores of Geminospora and Aneurospora (see Allen 1980), suggesting that their origin is local.
The occurrences are also not restricted in the Fair Isle succession (see text-fig. 6), being found in the
minor argillite horizons (?lacustrine deposits) of the Ward Hill Group (alluvial fan deposits) as well
as the overbank deposits of the Bu Ness Group. This distribution contrasts strongly with that of
Richardson ( 1 965, 1 967, 1 969), who noted an absence of Geminospora from the lacustrine and fluvial
deposits from the Orcadian basin lying to the south. A possible explanation is that the time difference
between the Orcadian and Fair Isle deposits is sufficient to allow the migration of an ‘ Archaeozono-
triletes ’ microflora from the Russian Platform, where it occurs in lower Givetian deposits (Raskatova
1974) contemporaneous with the Orcadian Basin. A less appealing hypothesis is that the Fair Isle
lacustrine and fluvial deposits were sustained by permanent water flow, giving more equable
conditions than those in the Orcadian Basin, and similar to conditions seen on delta tops or marine
margins, where Geminospora is known to thrive (see Becker et al. 1974). These authors describe in
detail the distribution of Aneurospora greggsii, and suggest it is not restricted by ecological factors in
their environment studied, which unfortunately do not include lacustrine or alluvial fan facies. It
308
PALAEONTOLOGY, VOLUME 25
seems that this conflict between an edaphic or stratigraphic control will not be resolved without
further detailed palaeoecological studies of this type.
Acknowledgements. J. E. A. Marshall is indebted to the Natural Environment Research Council for the receipt of
one of their C.A.S.E. studentships and also to the Institute of Geological Sciences (Leeds) for the use of their
facilities. The support of Newcastle University is also acknowledged for the latter part of the work. Both K. C.
Allen and J. E. A. Marshall acknowledge the use of research facilities and support of the Botany Department,
Bristol University. Useful and informative discussions have also been of great assistance from the following: Dr.
B. Owens (I.G.S. Leeds), Professor T. S. Westoll, F.R.S. (Newcastle University), and Dr. W. Mykura (I.G.S.
Edinburgh). The receipt of some preliminary samples from the I.G.S. Edinburgh (courtesy of Mr. P. J. Brand) is
also gratefully acknowledged. The help of the warden and staff at the Fair Isle Bird Observatory during the field-
work is also acknowledged.
REFERENCES
AGASSIZ, L. 1844-5. Monographie des Poissons fossiles du Vieux Gres Rouge ou Systeme Devonien ( Old Red
Sandstone) des lies Britanniques et de Russie. Neuchatel.
allen, k. c. 1965. Lower and Middle Devonian spores of North and Central Vestspitsbergen. Palaeontology , 8,
687-748.
— 1967. Spore assemblages and their stratigraphical application in the Lower and Middle Devonian of North
and Central Vestspitsbergen. Ibid. 10, 280-297.
— 1980. A review of in situ late Silurian and Devonian spores. Rev. Palaeobot. Palynol. 29, 253-270.
andreeva, e. m. 1962. Important Middle Devonian spores from the Kuznetz Basin. Contr. Phytostratig. Dev.
Sed. Altae-Sayan Mountain Region. Trudy. VSEGEI, n.s. 70, 191-213. [In Russian.]
Arkhangelskaya, a. d. 1963. New species of spores from Devonian deposits of the Russian Platform. Trudy
Vnigri , 37, 18-30. [In Russian.]
— 1974. Spore based zonation of the lower part of the Middle Devonian and its inter-regional correlation in
central and eastern European regions of the U.S.S.R. In Palynology of Proterophy te and Palaeophyte. Proc.
3rd Int. Palynol. Conf., Novosibirsk, 1971. 56-59. [In Russian.]
balme, b. e. 1962. Upper Devonian (Frasnian) spores from the Carnarvon Basin, Western Australia.
Palaeobotanist, 9, 1-10.
beck, c. b. 1957. Tetraxylopteris schmidtii gen. et sp. nov., a probable pteridosperm precursor from the
Devonian of New York. Am J. Bot. 44, 350-367.
becker, G., bless, m. h. m., streel, m. and thorez, j. 1974. Palynology and ostracode distribution in the Upper
Devonian and basal Dinantian of Belgium and their dependance on sedimentary facies. Meded. Rijks. geol.
Dienst, n.s. 25, 9-99.
beju, d. 1972. Zoning and correlation of the Palaeozoic of the Moesian Platform on the basis of
palynoprotistological assemblages. Petrol, si Gaze, 23, 714-722. [In Romanian.]
berry, w. 1937. Spores from the Pennington Coal, Rhea County, Tennessee. Amer. Midi. Nat. 18, 155-160.
bharadwaj, d. c. and venkatachala, b. s. 1962. Spore assemblages out of a Lower Carboniferous shale from
Spitzbergen. Palaeobotanist, 10, 18-47.
birks, h. j. b. 1970. Inwashed pollen spectra at Loch Fada, Isle of Skye. New Phytol. 69, 807-820.
bonamo, p. m. and banks, h. p. 1967. Tetraxylopteris schmidtii. its fertile parts and its relationships within the
Aneurophytales. Am. J. Bot. 54, 755-768.
brooks, c., hart, s. R. and wendt, I. 1972. Realistic use of two error regression treatments as applied to
Rubidium-Strontium data. Rev. Geophys. Space Phys. 10, 551-557.
burmann, g. 1969. Inkchlungundmechanishe deformation. Zeitzh. fur angewandte Geologic, Bd. 15, 7, 355-363.
butter worth, m. a. and collaborators. 1964. Densosporites (Berry) Potonie and Kremp and related genera. C.r.
5iemeC0ng. Strat. Geol. carbonif. Paris (1963), 1, 1049 1057.
— and williams, r. w. 1958. The small spore floras of coals in the Limestone Coal Group and Upper
Limestone Group of the Lower Carboniferous of Scotland. Trans. R. Soc. Edinb. 63, 353-392.
chaloner, w. G. 1967. Spores and land plant evolution. Rev. Palaeobot. Palynol. 1, 83-93.
1972. Devonian plants from Fair Isle, Scotland. Ibid. 14, 49-61.
chi, b. and hills, l. v. 1976. Biostratigraphy and taxonomy of Devonian megaspores, Arctic Canada. Bull. Can.
Petrol. Geol. 24, 640-818.
chibrikova, E. v. 1959. Spores from Devonian and older deposits of Bashkir. Akad. Nauk. S.S.S.R., Materials
on Palaeont. and Strat. of Devonian and older deposits of Baskir, 3-116. [In Russian.]
MARSHALL AND ALLEN: DEVONIAN MIOSPORES
309
CfflBRiKOVA, e. v. 1962. Spores of Devonian terrigenous deposits of western Bashkiria and the western slopes of
the southern Urals. Acad. Sci. U.S.S.R., Bashkir. Branch, Inst. Mining Geol., Brachiopods, Ostracods and
Spores Middle and Upper Dev. Bashkiria. 353-476. [In Russian.]
1972. Plant microfossils from the southern Urals and Cisuralia. Acad. Sci. U.S.S.R., Bashkir. Branch, Inst.
Geol. 1-222. [In Russian.]
— 1977. The stratigraphy of the Devonian and the older Palaeozoic from the Ural and Preural territories of the
U.S.S.R. Ibid. 1-160. [In Russian.]
— and naumova, s. n. 1974. Zonal complexes of Devonian spores and pollen in the European part of the
U.S.S.R. and their analogues abroad. In Palynology of Propterophyte and Palaeophyte, Proc. 3rd. Int.
Palynol. Conf., Novosibirsk, 1971, 39-47. [In Russian.]
clayton, G. and graham, j. r. 1974. Miospore assemblages from the Devonian Sherkin Formation of south-
west County Cork, Republic of Ireland. Pollen Spores, 16, 565-588.
— higgs, K. t. and keegan, j. b. 1977. Late Devonian and early Carboniferous occurrences of the miospore
genus Emphanisporites McGregor in southern Ireland. Ibid. 19, 415-426.
couper, R. A. 1953. Upper Mesozoic and Cainozoic spores and pollen grains from New Zealand. N.Z. Geol.
Surv. Palaeont. Bull. 22, 1-73.
dawson, j. w. 1859. On fossil plants from the Devonian rocks of Canada. Q. Jl geol. Soc. London, 15, 477-488.
donovan, R. n., archer, r., turner, p. and tarling, d. h. 1976. Devonian palaeogeography of the Orcadian
Basin and the Great Glen Fault. Nature, Lond. 259, 550-551.
— Collins, A., Rowlands, M. a. and archer, R. 1978. The age of sediments on Foula, Shetland. Scott. J. Geol.
14, 87-88.
— foster, r. j. and westoll, t. s. 1974. A stratigraphical revision of the Old Red Sandstone of North-eastern
Caithness. Trans. R. Soc. Edinb. 69, 171-205.
eisenack, a. 1944. Uber einige pflanzliche Funde in Geschieben, nebst Bemerkungen zum Hystricho-
sphaerideen-Problem. Z. Geschiebeforsch. 19, 103-124.
evans, p. h. 1970. Revision of the miospore genera Perotrilites Erdtm. ex Couper 1953 and Diaphanospora Balme
and Hassell 1962. Bull. Bur. Miner. Resour. Geol. Geophys. Aust. 116, 65-74.
fannin, N. G. T. 1970. The sedimentary environment of the Old Red Sandstone of western Orkney. Ph.D. thesis,
University of Reading.
filiminova, a. b. and Arkhangelskaya, A. d. 1963. Spore assemblage of the Vorbyev horizon of the Middle
Devonian of the Russian Platform. Trans. Vnigri, 37, 31-36.
fletcher, a. j. 1976. Studies on Devonian megaspores. Ph.D. thesis. University of Bristol.
flinn, d. 1961. Continuation of the Great Glen Fault beyond the Moray Firth. Nature, Lond. 191, 589-591.
guennel, G. k. 1963. Devonian spores in a Middle Silurian reef. Grana palynol. 4, 245-261.
hamid, m. e. p. 1974. Sporenvergesellschaftungen aus dem unteren Mitteldevon (Eifel-Stufe) des siidlichen
Bergischen Landes (Rheinisches Schiefergebirge). Neues Jahrb. Geol. Palaeont., Abh. 147, 163-217.
hemer, d. o. and nygreen, p. w. 1967. Devonian palynology of Saudi Arabia. Rev. Palaeobot. Palynol. 5, 51-61.
hoffmeister, w. s., staplin, F. L. and malloy, r. e. 1955. Mississippian plant spores from the Hardinsburg
Formation of Illinois and Kentucky. J. Paleont. 29, 372-399.
kedo, G. i. 1955. Spores of the Middle Devonian of the northeastern Byelorussian S.S.R. Acad. Sci. B.S.S.R.,
Inst. Geol. Sci., Palaeont. Stratigr. B. S.S.R. 1, 5-59. [In Russian.]
— 1960. Spore pollen complexes of the Paleozoic deposits of Byelorussia. Inter. Geol. Congr., 21st Sess.,
Rept. Sov. Geol., Prob. 6, 196-200. [In Russian.]
— 1966. Spore of Middle Devonian deposits from the western part of the Russian Platform. Inst. Geol.
Riga, Palynol. in Geol. Research in the Baltic Sov. Republ. 7, 7-14. [In Russian.]
— 1976. New species of spores from the Lower and Middle Devonian of the western Russian Platform. Acad.
Nauk. B. S.S.R. 1976, 108-171. [In Russian.]
kermack, k. a. and haldane, J. b. s. 1950. Organic correlation and allometry. Biometrika, 37, 30-41.
lang, w. h. 1925. Contributions to the study of the Old Red Sandstone flora of Scotland. 1. On plant remains
from the fish beds of Cromarty. 2. On a sporangium bearing branch system from the Stromness Beds. Trans.
R. Soc. Edinb. 54, 253-280.
leclercq, s. and bonamo, p. m. 1971. A study of the fructification of Milleria ( Protopteridium ) thomsonii Lang
from the Middle Devonian of Belgium. Palaeontographica, Abt. B, 136, 83-1 14.
— 1973. Rellimia thomsonii a new name for Milleria ( Protopteridium ) thomsonii Lang 1926 emend.
Leclercq and Bonamo 1971. Taxon, 22, 435-437.
lele, k. m. and streel, m. 1969. Middle Devonian (Givetian) plant microfossils from Goe (Belgium). Soc. Geol.
Belg. Ann. 92, 89-121.
310
PALAEONTOLOGY, VOLUME 25
loboziak, s. and streel, m. 1980. Miospores from Givetian to Lower Frasnian sediments dated with conodonts
in the Boulonnais (France). Rev. Palaeobot. Palynol. 29, 285-299.
mcgregor, d. c. 1960. Devonian spores from Melville Island, Canadian Arctic Archipelago. Palaeontology, 3,
26-44.
— 1961. Spores with proximal radial pattern from the Devonian of Canada. Bull. geol. Surv. Can. 76, 1-11.
— 1964. Devonian miospores from the Ghost River Formation, Alberta. Ibid. 109, 1-31.
— 1973. Lower and Middle Devonian spores of eastern Gaspe, Canada. I. Systematics. Palaeontographica,
Abt. B, 144, 1-77.
— 1977. Lower and Middle Devonian spores of eastern Gaspe, Canada. II. Biostratigraphy. Ibid. 163,
111-142.
— 1979a. Spores in Devonian Stratigraphical Correlation. In house, m. r., scrutton, c. t. and bassett, m. g.
(eds.). The Devonian System. Spec. Pap. Palaeont. 23, 163-184.
— 19796. Devonian miospores of North America. Palynology, 3, 31-52.
— and camfield, M. 1976. Upper Silurian? to Middle Devonian spores of the Moose River Basin, Ontario.
Bull. geol. Surv. Can. 263, 1-63.
— and uyeno, T. T. 1972. Devonian spores and conodonts of Melville and Bathurst Islands, District of
Franklin. Geol. Surv. Pap. Can. 71-13, 1 -37.
marshall, j. e. A. 1980. A method for the successful oxidation and subsequent stabilisation of high rank, poorly
preserved spore assemblages. Rev. Palaeobot. Palynol. 29, 313-319.
massa, D. and moreau-benoit, a. 1976. Essai de synthese stratigraphique et palynologie du systeme Devonien en
Libye occidentale. Revue Inst. fr. Petrole, 31, 287-333.
moreau-benoit, a. 1979. Les spores du Devonien de Libye, Premiere partie. Cahiers de Micropaleontologie, 1979
(4), 1-58.
— 1980. Les spores du Devonien de Libye. Deuxieme partie. Ibid. 1980 (1), 1-53.
Mortimer, m. G. and chaloner, w. G. 1972. The palynology of concealed Devonian rocks in southern England.
Bull. geol. Surv. Gr. Br. 39, 1-56.
mykura, w. 1972a. The Old Red Sandstone sediments of Fair Isle, Shetland Islands. Ibid. 41, 1-31.
19726. Igneous intrusions and mineralisation in Fair Isle, Shetland Islands. Ibid. 33-53.
1976. Orkney and Shetland. British Regional Geology. Institute Geological Sciences, Edinburgh.
— and young, b. r. 1969. Sodic Scapolite (Dipyre) in the Shetland Islands. Rep. Inst. geol. Sci., No.
69 (4), 1-8.
nathorst, A. G. 1915. Zur Devonflora des westlichen Norwegens. Bergens Mus. Aarbok, 1914-15, No. 9, 12-34.
naumova, s. N. 1953. Spore-pollen assembages of the Upper Devonian of the Russian Platform and their
statigraphic significance. Trans. Inst. Geol. Sci., Acad. Sci. U.S.S.R. 143 (Geol. Ser. 60), 1-204. [In Russian.]
neves, R. and dale, b. 1963. Modified filtration system for palynological preparations. Nature, Lond. 198,
775-776.
and Owens, b. 1966. Some Namurian camerate miospores from the English Pennines. Pollen Spores, 8,
337-360.
owens, b. 1971. Miospores from the Middle and early Upper Devonian rocks of the western Queen Elizabeth
Islands, Arctic Archipelago. Geol. Surv. Pap. Can. 70-38, 1-157.
— and richardson, j. b. 1972. Some recent advances in Devonian palynolygy— a review. Report of C.I.M.P.
Working Group No. 13B. 7e Cong. Inter. Stratig. Geol. Carbonifere, 1971, 1, 325-343.
ozolin'a, v. 1955. Upper Devonian (Dze) spore-pollen complexes from the Latvian S.S.R. Latvijas P.S.R.
Zinatnu Akademijas izdevums, Riga. 2 (91), 53-61. [In Latvian.]
— 1960a. Data on investigation of spores and pollen of the Frasnian stage of the Upper Devonian of the
Latvian S.S.R. Latv. Acad. Sci., Geologijas un Derigo Izraktenu Instituta Raksti, 5, 171-187. [In Latvian.]
— 19606. Middle Devonian spores from the Latvian S.S.R. Ibid. 189-205. [In Latvian.]
— 1961. Spore pollen spectra of the Frasnian Stage of the Upper Devonian of the Latvian S.S.R. Ibid. 7,
129-139. [In Russian.]
— 1963. Spore pollen spectra of the Frasnian stage in the Alanda Well. Ibid. 10, 129-139. [In Russian.]
pacht, r. 1849. Der devonische kalk in Livland, Ein Beitrag zur Geognoise de Ostee provinzen. Dorpat.
peppers, R. a. and damberger, h. h. 1969. Palynology and Petrography of a Middle Devonian coal in Illinois.
Circ. III. St. geol. Surv. 445, 1-36.
pettitt, j. m. 1965. Two heterosporous plants from the Upper Devonian of North America. Bull. Br. Mus. Nat.
Hist. 10, 81-92.
playford, g. 1971. Lower Carboniferous spores from the Bonaparte Gulf Basin, Western Australia and
Northern Territory. Bull. Bur. Miner. Resour. Geol. Geophys. Aust. 115, 1-104.
MARSHALL AND ALLEN: DEVONIAN MIOSPORES
311
— 1976. Plant microfossils from the Upper Devonian and Lower Carboniferous of the Canning Basin,
Western Australia. Palaeontographica, Abt. B, 158, 1-71.
Pokrovskaya, i. m. (ed.) 1966. Palaeopalynology. Proc. all Union Sci. Res. Geol. Inst. n.s. 141. 3 vols. [In
Russian.]
potonie, r. 1956. Synopsis der Gattungen der Sporae dispersae. I. Teil: Sporites. Beih. Geol. Jahrb. 23, 1-103.
— and kremp, G. 1 954. Die Gattungen der palaozoischen Sporae dispersaeund ihre Stratigraphie. Geol. Jahrb.
69, 111-194.
raskatova, L. G. 1969. Spore complexes from the Middle and Upper Devonian from the south eastern part of
the Central Devonian field. Voronezh University Press , Voronezh, 1969, 1-168. [In Russian.]
— 1974. Detailed stratigraphic subdivision of the Devonian of the Central Devonian field in the light of
palynological evidence. In Palynology of Proterophyte and Palaeophyte, Proc. 3rd Int. Palynol. Conf.,
Novosibirsk, 1971, 67-71. [In Russian.]
richardson, j. b. 1960. Spores from the Middle Old Red Sandstone of Cromarty, Scotland. Palaeontology, 3,
45-63.
— 1962. Spores with bifurcate processes from the Middle Old Red Sandstone of Scotland. Ibid. 5, 171-
194.
— 1965. Middle Old Red Sandstone spore assemblages from the Orcadian Basin, north-east Scotland. Ibid. 7,
559-605.
— 1967. A reconnaissance of some Upper Devonian and Lower Carboniferous spores from New York State
and Pennsylvania (U.S.A.). Rev. Palaeobot. Palynol. 1, 63-64. (Abs. only.)
— 1969. Devonian spores. In tschudy, r. h. and scott, r. a. (eds.), Aspects of Palynology, 193-222. Wiley-
Interscience, New Y ork.
— 1974. The stratigraphic utilization of some Silurian and Devonian miospore species in the northern
hemisphere: an attempt at a synthesis. Int. Symp. Belgian Micropalaeont. Limits, Namur, 1974. Publ.
No. 9, 1-13.
riegel, w. 1968. Die Mitteldevon-Flora von Lindlar (Rheinland). 2. Sporae dispersae. Palaeontographica, Abt.
B, 123, 76-96.
— 1973. Sporenformen aus den Heisdorf-, Lauch- und Nohn-Schichten (Emsium und Eifelium) der Eifel,
Rheinland. Ibid. 142, 78-104.
— 1974. Spore floras across the Lower/Middle Devonian boundary in the Rhineland (G.F.R.). Palynology of
Proterophyte and Palaeophyte, Proc. 3rd Int. Palynol. Conf., Novosibirsk, 1973, 47-53.
Sanders, r. B. 1968. Devonian spores of the Cedar Valley coal of Iowa, U.S.A. J. Palynology 2 and 3, 17-32.
schopf, J. M., wilson, L. R. and bentall, r. 1944. An annotated synopsis of Palaeozoic fossil spores and the
definition of generic groups. Rep. Inv. III. State Geol. Surv. 91, 1-72.
smith, a. h. v. and butter worth, m. a. 1967. Miospores in the coal seams of the Carboniferous of Great Britain.
Spec. Pap. Palaeont. 1, 1-324.
smith, D. i. 1977. The Great Glen Fault. In The Moray Firth Area Geological Studies, 46-59. Inverness Field
Club.
staplin, F. l. 1960. Upper Mississippian plant spores from the Golata Formation, Alberta, Canada.
Palaeontographica, Abt. B 107, 1-40.
and jansonius, j. 1964. Elucidation of some Palaeozoic Densospores. Ibid. 114, 95-1 1 7.
streel, M. 1964. Une association de spores du Givetian inferieur de la Vesdre, a Goe (Belgique). Ann. Soc. Geol.
Belg. 87, 233-262.
— 1967. Associations de spores du Devonien inferieur Beige et leur signification stratigraphique. Ibid. 90,
11-53.
— 1972. Dispersed spores associated with Leclercqia complexa Banks, Bonamo and Grierson from the late
Middle Devonian of New York State. (U.S.A.). Rev. Palaeobot. Palynol. 14, 205-215.
taugourdeau-lantz, j. 1967. Spores nouvelles du Frasnien du Bas Boulonnais (France). Revue micropaleont.
10, 48-60.
till, R. 1974. Statistical methods for the earth scientist, pp. 1-154. Macmillan.
urban, j. b. 1970. Ancyrospora fallacia, a new sporomorph exhibiting deceptive variations in preservation.
Micropaleontology, 16, 221-226.
van der zwan, c. j. 1979. Aspects of Late Devonian and Early Carboniferous palynology of southern Ireland. 1 .
The Cyrtospora cristifer morphon. Rev. Palaeobot. Palynol. 28, 1 -20.
— 1980. Aspects of Late Devonian and Early Carboniferous palynology of southern Ireland. 2. The
Auroraspora macra morphon. Ibid. 30, 133-155.
vigran, j. o. 1964. Spores from Devonian deposits, Mimerdale, Spitsbergen. Skr. nor. Polarinst. 132, 1-32.
312
PALAEONTOLOGY, VOLUME 25
westoll, t. s. 1979. Devonian fish biostratigraphy. In house, m. r., scrutton, c. t. and bassett, m. g. (eds.). The
Devonian System. Spec. Pap. Palaeont. 23, 341-353.
Williamson, j. h. 1968. Least-squares fitting of a straight line. Can. J. Phys. 46, 1845-1847.
wilson, l. r. and coe, e. z. 1 940. Descriptions of some unassigned plant microfossils from the Des Moines Series
of Iowa. Amer. Midi. Nat. 23, 182-186.
york, d. 1966. Least-squares fitting of a straight line. Can. J. Phys. 44, 1079-1086.
J. E. A. MARSHALL
Gearhart Geodata Services Ltd
Howe Moss Drive
Kirkhill Industrial Estate
Dyce, Aberdeen AB2 OGL
k. c. ALLEN
Department of Botany
The University
Bristol
BS8 1UG
Typescript received 10 June 1980
Revised typescript received 12 December 1980
CONODONTS, GONIATITES AND THE
BIOSTRATIGRAPHY OF THE EARLIER
CARBONIFEROUS FROM THE CANTABRIAN
MOUNTAINS, SPAIN
by A. C. HIGGINS and C. H. t. wagner-gentis
Abstract. Six conodont zones, including one new one, the Paragnathodus multinodosus Zone, are represented in
the Lower Carboniferous of the Cantabrian Mountains. The absence of conodont zones known to occur in
north-west Europe points to four major stratigraphic gaps in the Spanish sequence, three in the Tournaisian and
one in the Visean. High conodont abundances, indicating a low rate of sedimentation in the Tournaisian rocks,
and the numerous gaps in the stratigraphic record point to a slow initial Carboniferous transgression. The
patchy distribution of earliest Tournaisian sediments contrasts with the widespread distribution of the latest
(anchor alis Zone) Tournaisian sediments when the transgression reached its culmination. Lower conodont
abundances and thicker successions coincide with a more complete Visean/early Namurian sequence in which
only one major break is detectable. A goniatite fauna from the lower Visean is described including one new
genus, Pseudogirtyoceras and two new species P. villabellacoi and Winchelloceras palentinus. This fauna confirms
the early Visean age of the typicus Zone. One new species of conodont, Scaliognathus angustilateralis sp. nov., is
also named.
The Dinantian and early Namurian sedimentary sequence in the Cantabrian Mountains of north-
west Spain consists of less than 50 m of condensed deposits formed on the Cantabrian Block. This
was a stable platform uplifted during the late Famennian and the recipient of renewed sedimentation
from latest Famennian onwards. The age and nature of these sediments is not always the same all
over the area and this is true particularly of the deposits formed during the latest Famennian and
Tournaisian. Several formations are involved (compare Wagner, Winkler Prins, and Riding 1971).
The earliest deposits belong to the Ermita Formation (Gres de l’Ermitage of Comte 1959) which was
laid down after the Famennian uplift and which appears to have a diachronous base, older at the
flanks of the uplift and younger in its central part. A calcareous sandstone lithology is the most
common for the Ermita Formation but limestone has been recorded in the top part. Samples in the
highest Ermita Formation yielded conodonts attributed to the costatus Zone (Higgins, Wagner-
Gentis, and Wagner 1964), but a few localities have now shown the presence of a Protognathodus
fauna of earliest Carboniferous age. The Ermita Formation thus appears to span the latest
Famennian and the earliest Tournaisian.
In the south-eastern part of the Cantabrian Mountains, corresponding to the northern part of the
province of Palencia, there is an area which shows many resemblances to the western Pyrenees and
which is characterized by a nodular limestone, the Vidrieros Formation of Van Veen (1965),
spanning the Famennian -Tournaisian boundary (Van Adrichem Boogaert 1967). This region does
not show the effects of Famennian uplift and contains a different development of Devonian strata, i.e.
the Palentine facies of Brouwer (1964). It occupies an area between two major faults, viz. the Ruesga
Fault in the south and the southern boundary fault of the Picos de Europa to the north. Appreciable
tectonic shortening is a feature of both faults and the abrupt contacts between areas of different facies
development is apparently due to extreme telescoping as a result of the tightening of the arcuate fold
belt in north-west Spain (R. H. Wagner, pers. comm.).
The Ruesga Fault is apparently continued westwards by the Leon Fault of Marcos (1967)
IPalaeontology, Vol. 25, Part 2, 1982, pp. 313-350, pis. 34-36-1
314
PALAEONTOLOGY, VOLUME 25
(originally described as the Leon Line by De Sitter 1962). Although this important fracture zone has
been widely credited as a syn-sedimentary control, it seems to be due to post-sedimentary tectonics
with a telescoping effect which gradually diminishes westwards (R. H. Wagner, pers. comm.). The
sediments investigated for the present paper were all laid down south of the Leon-Ruesga tectonic
fault line (text-fig. 1) even though some of the localities are presently north of this fault.
text-fig. 1 . Structural palaeogeographic areas. Palaeozoic, Cantabrian Mountains.
In this area the Ermita Formation is generally present. It is followed by either the Vegamian
Formation of black shales or the Baleas Formation of crinoidal limestone. These two formations
(Wagner et al. 1971) seem to be mutually exclusive. Since they both seem to correspond to the same
time interval within the Tournaisian (cooperi-communis Zone of the present paper) it may be that the
Baleas Limestone was formed on ridges in the basin which received the Vegamian Shales (compare
Higgins et al. 1964). The Baleas Formation, described by Wagner et al. (1971) is only a few metres
thick. The Vegamian Formation (Comte 1959; Wagner etal. 1971) is generally less than 10mthick.lt
is characterized by phosphatic and chert nodules and black shales which are often cherty. Its
macrofauna includes the goniatite Muensteroceras arkansanum Gordon, as recorded by Wagner-
Gentis (in Wagner et al. 1971). This species was described originally from the late Kinderhookian of
Arkansas, U.S.A.
The most widespread formation in the Carboniferous of north-west Spain is the succeeding
Genicera Formation of Wagner et al. (1971) which is the ‘Marbre Griotte’ of Barrois (1882) (also
called Villabellaco Formation— Wagner and Wagner-Gentis 1963— and Alba Formation— Van
Ginkel 1965). It is characterized by nodular and wavy-bedded limestones and up to 25 m thick. The
basal unit (Gorgera Member of Wagner et al. 1 97 1 ) is marly and usually a vivid red colour, although a
slightly reddish-grey colour has also been found. Over a large area of northern Leon, including the
Genicera type section, an interval of red shales and cherts (Lavandera Member) separates the basal,
HIGGINS AND WAGNER-GENTIS: CONODONTS AND GONI ATITES
315
marly unit of nodular limestones from the main part of generally wavy-bedded to nodular limestones
(Canalon Member of Wagner et al. 1971). The chert unit is absent in the southernmost exposures of
northern Leon and in the Revilla Nappe structure of northern Palencia. Conodont work reported in
the present paper shows that the chert unit corresponds to a sizable time gap which is equally
apparent where the cherts are absent. The basal unit, which is only a few metres thick, has yielded
goniatites of the late Pericyclus (II j 8/y) Zone corresponding to late Tournaisian. Its conodont fauna
corresponds to the anchoralis Zone. The Gorgera Member may be occasionally absent and although
the rare occurrences where this seems to be the case ought to be checked for thrusting (shearing in the
Age
SSW of Genicera (Leon)
Olleros de Alba
(Leon)
Formation
Member
Formation
Namurian A/B
Barcaliente
Barcaliente
Olleros
Genicera
Canalon
Olaja Beds
Genicera
Visean
Lavandera
Gorgera
Tournaisian
Vegamian
Vegamian
Ermita
Ermita
Famennian
table 1. Formations in the Porma-Bernesga area, based on Wagner et al. (1971).
flanks of isoclinal synclines is of common occurrence in the Cantabrian Mountains), it appears likely
that the base of the Genicera Formation may not always be of exactly the same age.
The top of the formation reaches into the lower Namurian (E2b Zone) and is undoubtedly
diachronous. Wagner et al. (1971) described the Olaja Beds, a condensed mudstone sequence of only
a few metres, as the lateral equivalent of the highest part of the Canalon Member. The lower
Namurian Olaja Beds occur only in the southernmost exposures of the Cantabrian Mountains in
northern Leon, where they form the base of a thick terrigenous sequence with turbidites (Olleros
Formation of Wagner et al. 1971). In the more northerly exposures, the more complete Canalon
Member is followed with gradual transition by the Barcaliente Limestone Formation. The latter was
formed on a carbonate platform (the same Cantabrian Block as accumulated the Genicera Limestone
and Chert Formation) which was sufficiently shallow to give rise to evaporitic sediments. The
presence of a turbidite basin to the south indicates the diminished area of the Cantabrian Block in
mid-Namurian times, whilst the terrigenous facies of the Olaja Beds provides an early indication of
the northwards withdrawal of the southern margin of the block in early Namurian times.
316
PALAEONTOLOGY, VOLUME 25
text-fig. 2. Location of the sampled sections in the southern part of the Cantabrian Mountains.
CONODONT ZONES AND LOWER VISEAN GONIATITE OCCURRENCE
Six conodont zones and a Protognathodus fauna are represented in the sequence but the boundaries of the fauna
and the lower three zones are non-sequences and the sequence is probably incomplete. Five of these concurrent
range zones were defined by Higgins (1974) but are here modified and one zone is new.
Protognathodus fauna. The base of the Ermita Formation is a transgressive horizon of late Devonian age,
belonging to the costatus Zone of Ziegler (1962). In only two of the sections, the Aviados and the Villabellaco
sections in the Porma-Bernesga area, does the base of the Carboniferous rest on pr e-costatus rocks. On the other
hand, only at Santiago de las Villas and Olleros de Alba is there a transition between the Devonian and
Carboniferous where a thin limestone at the top of the Ermita Formation yields a conodont fauna belonging to
the Protognathodus fauna. This fauna includes Polygnathus communis communis Branson and Mehl,
Polygnathus inornatus sensu Branson and Mehl, Protognathodus kockeli (Bischoff ), Protognathodus meischneri
Ziegler, Protognathodus kuehni Ziegler and Leuteritz, and one specimen doubtfully referred to Pseudo-
polygnathus dentilineatus. This fauna was referred to the kockeli-dentilineatus Zone by Higgins et al. (1964) as
was a similar fauna described by Adrichem Boogaert (1967) from the Triollo area. The presence of
Protognathodus kuehni suggests that this fauna belongs to the younger Protognathodus fauna of Ziegler (1969).
The presence of a newly discovered siphonodellid in this fauna also clearly indicates closer affinity with the
Upper rather than the Lower Protognathodus fauna. This siphonodellid was referred by Higgins (1974) to
Siphonodella sulcata (Huddle) but it is poorly preserved and of very doubtful determination. The assignment of
these samples to the Upper Protognathodus fauna places them in the Carboniferous rather than at the top of the
Devonian (Sandberg, Ziegler, Leuteritz, and Brill 1978).
Siphonodella cooperi-polygnathus communis Zone. This zone, defined by Higgins (1974), is widespread in the
Porma-Bernesga area where it occurs in the Baleas and Vegamian Formations. Key species in the rich and
abundant faunas include Siphonodella obsoleta Hass, S. duplicata Branson and Mehl morphotype 2, and S.
cooperi Hass morphotype 2 of Sandberg et al. (1978). These range from the Upper Duplicata Zone to the Upper
Crenulata Zone according to Sandberg et al. (1978). However, Gnathodus delicatus Branson and Mehl, G.
punctatus (Cooper), and G. semiglaber Bischoff, which are also present in the zone, overlap the range of species of
Siphonodella in the Belgian Tournaisian (Groessens \911b) in late Tn2. Also present in the zone is Polygnathus
communis carina Hass which first appears in Tn3 in the Belgian sequence. A late Tn2/early Tn3 age indicates the
span of the cooperi-communis Zone. This age probably indicates at least partial equivalence to the German
crenulata Zone (see text-fig. 7).
Anchoralis Zone. The anchoralis Zone is easily recognized and is a widespread zone in the Cantabrian
Mountains. Scaliognathus anchoralis Branson and Mehl, Doliognathus latus Branson and Mehl, Pseudo-
polygnathus triangulus pinnatus, Gnathodus texanus (Roundy) pseudo semiglaber Thompson, and G. delicatus are
HIGGINS AND WAGNER-GENTIS: CONODONTS AND GONI ATITES 317
text-fig. 3. Ranges of the stratigraphically important conodont species in the Cantabrian Mountains.
318
PALAEONTOLOGY, VOLUME 25
table 2. Distribution and abundance of conodonts from the Cantabrian Mountains.
among the key indices in a rich and varied fauna. D. latus is used as a basal subzonal index in the Missouri
(Thompson and Fellows 1970) and Belgian (Groessens \911b) sections. In the Porma-Bernesga area, D. latus is
present throughout the zone and in most samples of this age, which implies that the upper part of the zone is
missing in the southern part of the Cantabrian Mountains, and that the zone is of early Tn3c age.
Gnathodus typicus Zone. This zone was defined by Marks and Wensink from the Pyrenees (1970). The
characteristic species G. typicus Cooper, G. antetexanus, and G. texanus pseudosemiglaber mark the beginning of
HIGGINS AND WAGNER-GENTIS: CONODONTS AND GONI ATITES
319
\ species
sample \
Pinornotus
Bi. stabilis
S. obsoleta
G de/icatus
G semig/ober
Ps triangu/us pinna t us
D lotus
Sc anchora/is
Sc. angusti/atera/is
Pbischoffi
G ant et exanus
G. texanus pseudosemig/aber
G typicus
G symmutatus
G homopunctatus
Pa commutatus
G bi/ineafus
Sp. Campbell i
G girtyi
Pa nodosus
Pa. mu/tinodosus
total per kilo
1197B
2
3 14
19
Robledo
1
3 2
6
de Caldas 1194
8
9
1193
4
i ■’
18
Viadangos 1^58
10 20
30
de Arbos 1357
5 1
0 |0
48
3520
2
i
2
9
N
20 i
15
3
40
14
4
25
25 -
2
41
1
1
12
i
2
47
F
6 30 !
20
83
E
2 K) 1
D
2 40 ;
44
c
2
0 15
2 1
40
B
1 2
38
A
6 2
2
3077
1069T
i
S
2
R
4
18
L
4
5
C t' K
2
2
Getino j
6 16
2!
5 25
4 3
H
10 50 <
82
G
F
4
E
2 10
D
7 25 ;
C
6 20
B
7 21
28
A
10 60 (
9 7
1072 N
l
1
M
1 12
13
L
2
2
ii
4
Venta de G
3
c
8
5
4
1340 1
>
11698
1
II
3
Genicera ,,66c -
1
i
20
5
27
23
4 10
91 1
158
1338
4
\ species
sample \
Pr kocke/i
Pr meischneri
Ps. primus
P communis communis
Pinornotus
Bi stabilis
Ps. mu/tistriafus
G. de/icatus
Ps. triangu/us pinna t us
D. lotus
S anchora/is
G. antetexanus
G texanus semig/ober
G typicus
G symmutatus
G. homopunctatus
Pa commutatus
G bi/ineatus
Sp. campbelli
Pa. nodosus
Pa. mu/tinodosus
M bipluti
total per kilo
1120
1119
1118
1116
1115
Santiago 1114
de las Villas 1113
1112
1110
1314
1311
1310
1
1
1
1
2
in
i.
1
1
2
40
36
20
1
8
• 1
1 42
3
58
1149T
Entrago h
0
B
1 1
2
1
1
1
14
1 ;
9
127
1
II
■ 2
2 2 ;
12 144
9
165
14 <
1 2 15
3
38
10 12
1
23
4 25
29
1 2
'
2916
2915
2913
2912
2911
2910
2909
Villabellaco ““
2905
2904
2903
2902
2901
138C
1 2
l
4
6
8
4 4
8
:■ 18 21
3
48
14 36
53
1 21
1
.. 1
53
Revilla V|
25
25
133
31
31
\ 9 51
2
1 28
5
5 55
3 3 25
i'
4o
5 23153 f
18
7 30
3 '
1
12
7
Sta.Olajad. Varga 372
_ _ , 1103m
S.E. of lv
Genicera y
6
6
90 -
10
104
70
table 3. Distribution and abundance of conodonts from the Cantabrian Mountains.
the dominance of species of Gnathodus which characterizes much of the Visean and early Namurian. G.
homopunctatus Bischoflf appears at the base of the zone and this occurrence together with the other species would
suggest an equivalence of this zone to the G. homopunctatus Subzone of Vla age in Belgium (Groessens 19776).
The goniatites from the basal marly unit ( typicus Zone) of the Villabellaco Limestone in northern Palencia
have been found in the following sequence:
138 D, 5 0 m above the base Pseudogirtyoceras villabellacoi sp. nov.
Merocanites marshallensis (Winchell)
Merocanites subhenslowi Wagner-Gentis
Ammonellipsites kayseri (Schmidt)
320
PALAEONTOLOGY, VOLUME 25
138 c, 2-3 m above the base
1 - 1 - 2 m above the base
138 b, 0-5-0-6 m above the base
136 B, east of locality 138, at
the same level as 138 c
Merocanites marshallensis (Winchell)
Merocanites subhenslowi Wagner-Gentis
Ammonellipsites kayseri (Schmidt)
N autellipsites hispanicus (Foord and Crick)
Merocanites marshallensis (Winchell)
Muensteroceras cf. crassum Foord
Winchelloceras palentinus sp. nov.
Muensteroceras parallelum (Hall)
Merocanites marshallensis and Winchelloceras are known from the Marshall Sandstone in Michigan, which is
considered to belong to the Osagean (Manger 1979, pp. 214, 215). Popov (1968) recorded Winchelloceras from
the Fascipericyclus 3 Zone (Lower Visean) of the Tien Shan in Central Asia and (1975) from the CjV, unit in the
Urals.
Ammonellipsites kayseri has been described from Liebstein (Erdbach and Breitscheid section). West Germany
in the Pey Zone, in Aragon (Spain), Schmidt (1931) mentioned it with a Lower Visean fauna and in south-west
England it occurs at Tawstock and Coddon Hill (Prentice and Thomas 1960, p. 6).
Muensteroceras parallelum was described from the Rockford Limestone of Indiana. A conodont fauna below
the ammonoid layer in the type locality belongs to the Osagean (Rexroad and Scott 1964) and Manger (1979,
p. 213) attributes the goniatite bed also to the Osagean. The species is also known from the Sj unit of the Hassi
Sguilma Stage of Pareyn (1961, p. 50 and table IV) which contains a basal Visean fauna. Librovitch (1927, p. 42)
mentioned this species from the Tien Shan, where it occurs with a fauna of the middle and upper 1 Pericyclus
Stufe’ (Schmidt 1925).
Revilla
Conodont Zones
Villabellaco
mu/tinodosus
nodosus
bi/ineatus
typ/cus
2916
2915
2914
2913 H
2912
2911
2910
2909 H
2908-
2907-
2906-
2905-
2904-
2903.
2902-
2901-
138D-
138C-
138B-
2899-
2898 H
134X1-
X-
IX-
VIII-
VII-
— VI-
V-
IV-
— Ill-
Goniatites
Pseudogirtyoceras villabellacoi
Nauteiiipsites hispanicus
Muensteroceras cf. crassum
Merocanites marshallensis
Merocanites subhenslowi
Ammonellipsites kayseri
Winchelloceras palentinus
Grey, nodular
and wavy bedded
limestone
metres
text-fig. 4. Conodont zones and lower Visean goniatite occurrence in the Villabellaco area, Revilla Nappe.
HIGGINS AND WAGNER-GENTIS: CONODONTS AND GONIATITES
321
Muensteroceras cf. crassum is known from the Lower Carboniferous Limestone of Ballinacarriga, Co.
Limerick, Eire.
Merocanites subhenslowi and Nautillepsites hispanicus have previously been described from Olleros de Alba
(Leon) and Villabellaco (Palencia) (Higgins et al. 1964; Wagner-Gentis 1960). In both papers it was suggested
that they occurred in an equivalent to the lower B-zone, but this is now seen to be incorrect. An earlier horizon is
involved. The lower part of the Villabellaco Limestone yielded conodonts of the typicus Zone (Higgins, this
paper) which corresponds to Vla of the Belgian sequence. An early Visean age is also indicated by the majority of
the goniatite species found.
The goniatite fauna from the lower part of the Villabellaco Limestone has Muensteroceras parallelum in
common with the Rockford Limestone of Indiana, and Merocanites marshallensis and Winchelloceras (a
different but comparable species, Winchelloceras allei) with the Marshall Formation of Michigan. The latter has
been assigned an early Visean age by Miller and Garner (1955, pp. 118-119) (see also Brenckle, Lane, and
Collinson 1974) but a late Tournaisian age by Manger (1979, p. 221, fig. 3). The lower part of the Villabellaco
Limestone containing the goniatites comparable to those of the Marshall Formation, has also yielded conodonts
of the typicus Zone which belongs to Vla of the Belgian sequence. It thus appears that the Marshall Formation,
may be lower Visean rather than upper Tournaisian. On the other hand, Kazakhstania, another element of the
Marshall Fauna, is regarded as a late Tournaisian index (Librovitch 1940, pp. 324-325; Manger 1971). This
goniatite does not occur in the Villabellaco Limestone.
Gnathodus bilineatus Zone. The marked reduction of the texanus group and its disappearance a little way into
the zone marks the beginning of the Gnathodus bilineatus Zone. This change coincides with the appearance of the
important G. bilineatus (Roundy) and Paragnathodus commutatus (Branson and Mehl). This abrupt change
is due to the absence of several faunas, notably the late appearance of P. commutatus which appears before
G. bilineatus in northern Europe, and the absence of the G. texanus fauna which is important in Missouri
(Thompson and Fellows 1970).
Paragnathodus nodosus Zone. The increase in the complexity of P. commutatus by the addition of platform
nodes marks the appearance of P. nodosus (Bischoff). This widespread species appears in the zone of
Neoglyphioceras spirale, Go in Germany (Meischner 1970) and in V3b in Belgium (Groessens 19776). G.
bilineatus is still a very important zonal constituent.
Paragnathodus multinodosus Zone. The increasing complexity of platform ornamentation marks the
appearance of P. multinodosus (Wirth). Again G. bilineatus is an important element of the fauna. Budinger and
Kullmann (1964) related the appearance of P. multinodosus to the granosus Zone (Go y). The consistent later
appearance of P. multinodosus relative to P. nodosus, often only 1 m higher, is only detected in closely sampled
sections. The P. multinodosus Zone continues into the overlying Namurian where in the absence of the later form
of G. bilineatus, G. bilineatus bollandensis and the rarity of G. girtyi Hass it has not proved possible to subdivide
the early Namurian.
Consideration of the Faunas. The faunas are dominated by platformed conodonts with wide basal
cavities and flaring platforms typified by species of Gnathodus and Paragnathodus. In the Visean these
are characteristic of deep-water cephalopod limestones in many parts of the world and the deep-
water nature of the Visean griotte limestones and the presence of goniatites confirms this as a general
observation. Nevertheless, both in the Tournaisian and the Visean there are peculiarities in the
faunas which are worthy of note.
The Tournaisian faunas are of two types: one, occurring in the crystalline limestone of the Baleas
Formation and the basal unit of the Genicera Formation, is dominated by gnathodids whereas the
other, occurring in the fine grained limestone of the Ermita Formation and in the conglomerates and
shales of the Vegamian Formation, is dominated by siphonodellids and polygnathids. Meischner
(1970) suggested that the gnathodid faunas of the German anchor alis Zone were characteristic of
schwellen environments whereas the polygnathids dominate in the basin environments. Matthews,
Sadler, and Sellwood (1972) suggested that the contrast is between faunas with large basal cavities
and flaring platforms, the gnathodids, and those with small basal cavities, the polygnathids. The
distribution of the Baleas Formation in the Porma-Bernesga region is in the form of an east-west
trending area within the Vegamian shale outcrop and it was suggested by Higgins et al. (1964) that it
was formed on a ridge, in a deeper-water sea. The conodont faunas would confirm this view, because
the Vegamian faunas, although few, are dominated by polygnathid conodonts. Whether the
322
PALAEONTOLOGY, VOLUME 25
text-fig. 5. Conodont zones in the sections at Entrago and Robledo de Caldas.
HIGGINS AND WAGNER-GENTIS: CONODONTS AND GON I ATITES
323
text-fig. 6. Generalized geological map of part of northern Leon after Wagner et al. (1971) with structural units after de Sitter (1962).
324
PALAEONTOLOGY, VOLUME 25
crystalline limestone faunas of the Baleas were formed in the same depth of water as the griotte
limestone is unknown but it seems likely in view of the similarity of conodont type.
The other anomaly occurs in the Visean where the rarity of G. girtyi contrasts with its abundance in
the basin limestones and shales of northern Europe. This species is occasionally found in the G.
bilineatus Zone but principally occurs in the P. nodosus Zone. In the Pyrenees (Marks and Wensink
1970) it appears to occur only in the P. nodosus Zone. There is no record of G. girtyi in the Namurian
of Spain. Similarly, the absence of G. bilineatus bollandensis, the high abundance of Spathognathodus
campbelli Rexroad, and the almost total restriction of P. multinodosus to the Cantabrian Mountains
and the Pyrenees, point to the partial geographic isolation of this area during the early Namurian
(Ej, E2).
IMPLICATIONS OF THE CONODONT DATINGS FOR THE HISTORY OF
THE AREA
Comparison of the conodont sequence with more complete ones in Belgium, Germany, and Great Britain
indicates the presence of considerable breaks in the Lower Carboniferous succession in the Cantabrian
Mountains. However, there is little evidence of extensive erosion and the lack of faunal reworking and the high
faunal abundance implies slow but continuous deposition separated by periods of non-deposition. The sequence
of events is as follows:
NAMURIAN (El5 E2)
VISEAN
multinodosus Zone
nodosus Zone
bilineatus Zone
typicus Zone
anchoralis Zone
tournaisian cooperi-communis Zone
Upper Protognathodus f.
DEVONIAN
costatus Zone
Average number of
specimensjKg
Continuous deposition ] 39
l 27
Widespread non-deposition
Widespread deposition 19
Local non-deposition
Major transgression 97
Widespread non-deposition,
some erosion. Continuation
of swell topography during early
part of the zone
Development of swells 89
Widespread non-deposition
Patchy distribution
Erosion surface in most areas
Widespread transgression
Late Devonian history. Adrichem Boogaert (1967) clearly demonstrated the transgressive nature of
the late Devonian rocks which, in the form of the Ermita Formation, commonly rest unconformably
on the underlying rocks.
Tournaisian history. There are undoubtedly sections in the Cantabrian Mountains which show a
transition from the Devonian into the Lower Carboniferous. The sections at Santiago de las Villas
and Olleros de Alba in the Porma-Bernesga area yield the earliest Carboniferous, Upper
Protognathodus, fauna in a sequence which is uninterrupted from the Devonian. Unfortunately, the
latest Devonian rocks do not yield conodont faunas. A transition probably also exists in the Vidrieros
Formation in the northern part of the Palentine Basin (Adrichem Boogaert 1967) where the costatus
Zone is followed by the Protognathodus fauna, but a full sequence of zones has not yet been
demonstrated. Similarly, there may well be a transition in the Candamo Formation in Loredo, but
again a full sequence of zones has not been proved (Del Rio and Menendez Alvarez 1978). These are
HIGGINS AND WAGNER-GENTIS: CONODONTS AND GONI ATITES 325
text-fig. 7. Correlation of the conodont sequence in the Cantabrian Mountains with those of Belgium and Germany. The suggested faunal and
stratigraphic breaks in the Spanish sequence are indicated by diagonal lines.
CANTABRIAN MOUNTAINS
(after Groessens 1975,
Higgins & Bouckaert 1968)
GERMANY
(after Meischner 1970,
Conil & Paproth 1968)
SIPHONODELLA
ZONATION
(after Sandberg et al. 1978)
Pa. multinodosus
Pa. nodo8u8
e2
G. bilineatus bollandensis
S2
G. bilineatus schmidti
isosticha-U . crenulata
\\\\\\\\\\\\\N
E!
Pa. nodosus
v3c
v3bY
Pa. nodoeu8
Go
Gos
G. bilineatus bilineatue
G. typicus
V3ba
G. bilineatus bilineatus
%
G. bilineatus bilineatus
VV3«
- “
-
anchoralis-bilineatus
Vlb
Via
M. beckmanni
Pa. cormrutatus
G. homopunctatus
mm
S. anchoralis
-3
S. anchoralis
S. anchoralis
E. burling tonensis
D. lotus
S. anchoralis
S. crenulata
mms.
S. oooperi-P. communis
mm
P. communis carina
Do. bouckaerti
Sp. bultyncki
Siphonodella
G,
Si. triangulus triangulue
sandbergi
Si. tTiawulus ivaequalis
u. duplicata
L. duplicata
kocke li-denti lineatua
sulcata
u. Pro tognathodu8 fauna
U. Protognathodus fauna
ccstatu*
Pro togna thodu 8 fauna
L. Protognathodus fauna
M. & u. costatus
praesulcata
text-fig. 7. Correlation of the conodont sequence in the Cantabrian Mountains with those of Belgium and Germany. The suggested faunal and
stratigraphic breaks in the Spanish sequence are indicated by diagonal lines.
326
PALAEONTOLOGY, VOLUME 25
the exceptions, however, for the Devonian-Carboniferous boundary is usually an erosion surface
representing a regression of early Carboniferous age. In the Porma-Bernesga area the gap at the base
of the Carboniferous increases in size in a northerly direction from Santiago de las Villas to
Viadangos de Arbas where the Visean rests directly on the late Devonian (see text-fig. 8). This trend is
also seen in the pre-Ermita unconformity in this area (Adrichem Boogaert 1967) where it was inter-
preted as part of a general trend of increasing hiatus in a northerly and easterly direction towards the
‘Asturian Geanticline’ which occupies the central part of the Cantabrians. The same trend towards
this platform area is evident in the Lower Carboniferous, but is one of increasing disconformity
rather than unconformity and is an expression of the Bretonic Movements.
These early Tournaisian (Tnlb) deposits represent the termination of the late Devonian
transgression and were followed by a long period of regression. Evidence of the upper part of the
sulcata Zone and the following zones up to the upper part of the sandbergi Zone of Sandberg et al.
(1978) is lacking. Comparison with Belgium (Groessens 19776) reveals the absence of all but the
highest part of the Siphonodella Zone or lower part of the Polygnathus communis carina Zone which
implies a gap ranging from Tnlb to Tn3a. This regression is represented in all sections of the
Porma-Bernesga area by an erosion surface. The subsequent transgression resulted in a different
pattern of sedimentation due in part to the emergence of positive areas in a generally subsiding basin.
Dominating the facies pattern is the Vegamian Formation, a thin sequence of black, often cherty,
shales with thin conglomeratic and phosphatic nodule horizons. In the Porma-Bernesga area a
conglomeratic sandstone at the base of the Vegamian Formation at Santiago de las Villas and
Genicera yields a fauna of polygnathids, siphonodellids, and pseudopolygnathids belonging to the
cooperi-communis Zone. The upper limit of the Formation is more difficult to define. The highest beds
of the Genicera section yield Pseudopoly gnathus triangulus pinnata Voges, which ranges through Tn3
in Belgium and in other sections it is overlain by beds of anchoralis (Tn3c) age. A pr e-anchoralis age
for the major part of the Vegamian seems likely, although an early anchoralis age for the highest part
cannot be ruled out. To the east and north-east of the Porma-Bernesga area, the base of the
Vegamian appears to be of cooperi-communis age (Adrichem Boogaert 1967) where the faunas were
referred to the lower anchoralis zone. The upper surface appears to be diachronous since Budinger
and Kullmann (1964) recovered late Visean (GoIII a/j3) goniatites from the highest part of the
Formation, although Wagner et al. (1971) question the attribution of these faunas to the Vegamian
Formation.
In the Porma-Bernesga area, along a zone extending from Aviados to Pola de Gordon, the
Vegamian black shales are absent and their place is taken by the crystalline limestones of the Baleas
Formation. This thin unit, 12 m thick, is rich in conodonts with an average of eighty-nine conodonts
per kilogram, indicating slow sedimentation. The faunas are of the swell type and indicate the
presence of a swell extending east-west through the Porma-Bernesga area, which is here named the
Pola de Gordon Swell. In the area to the east, north of Sabero, a similar crystalline limestone again
replaces the Vegamian Shales and is also of cooperi-communis age (Adrichem Boogaert 1967). It may
represent the easterly continuation of the Pola de Gordon Swell. At Entrago, Budinger and
Kullmann ( 1 964) recorded from the lower part of the sequence a crystalline limestone of ‘Baleas’ type
also yielding a mid-late Tournaisian conodont fauna. As in the area to the south, the Vegamian is
absent and it seems probable that this marks the position of another swell. These swells are close to,
and parallel with, the early Carboniferous coastline (see Adrichem Boogaert 1967) and represent
crinoidal banks forming on ridges close to the shore-line.
The anchoralis Zone, although extremely thin (c. 2 m) is the most widespread horizon in the
Tournaisian of the Cantabrian Mountains and represents a considerable extension of the sea. In the
majority of sections the base of the zone is an erosion surface, but one section, at Genicera, exhibits a
transition from the Vegamian Formation into the overlying Genicera Formation. The earliest
deposits of the zone included the top 1 m of the Baleas Formation in the Pola de Gordon area where
erosion surfaces at the base and top separate the unit from cooperi-communis Zone below and the
later part of the anchoralis Zone above. The most widespread part of the zone occurs at the base of the
Genicera Formation where it occupies the basal metre of the red and grey nodular limestones
HIGGINS AND WAGNER-GENTIS: CONODONTS AND GONIATITES
327
text-fig. 8. Geographical location and conodont sequences of the sections on the Porma-Bernesga area. The units
are structural ones, usually isoclinal synclines separated by thrusts. The inset ribbon diagram illustrates the
distribution of the zones.
328
PALAEONTOLOGY, VOLUME 25
(griotte). The fauna commonly occurs in a fine grained non-nodular limestone and is of the ‘swell’
type being composed of gnathodids together with some locally abundant polygnathids and
pseudopolygnathids. However, there are no swells identifiable in this zone, merely a general
shallowing over the whole region except in the extreme east where the Vegamian Formation persists.
In the south-east, the Revilla Nappe, the Vegamian Formation is absent and the oldest
Carboniferous is a thin sandy limestone of anchor alis age representing the southerly extension of the
Lower Carboniferous sea.
Visean history. As already stated, it is believed that the anchoralis Zone is incomplete, only the lower
part being present. There is, however, no obvious physical expression of this break in the rock
successions except the occasional absence of the anchoralis faunas as at Viadangos de Arbas and
possibly at Gildar-Monto where Budinger and Kullmann (1964) date the base of the Carboniferous
as late Visean.
The Visean begins with the typicus Zone, less than 5 m thick, and with a much-reduced conodont
abundance of nineteen specimens per kilogram compared to ninety-seven per kilogram in the
anchoralis Zone, indicating more rapid deposition. At the top of the zone the nodular limestones give
way to horizons and nodules of radiolarian cherts in red shales, which are extensively developed in
the Porma-Bernesga-Esla area.
Immediately above the chert the fauna changes markedly with the incoming of Gnathodus
bilineatus and the change reflects the absence of at least one fauna. In comparison with Belgium and
Ireland Paragnathodus commutatus appears later in Spain, both in the Cantabrian Mountains and in
the Pyrenees (Marks and Wensink 1970), by at least one zone. Also absent is the G. bulbosus fauna of
mid-Osage age (Thompson and Fellows 1970). It is possible that these zones are represented by the
chert horizon but in chert-free sections the faunal break is still present. This non-sequence is clearly of
more than local significance.
Above the chert the sequence appears to be unbroken, although still condensed, and both the
goniatite and the conodont records are complete.
SYSTEMATIC PALAEONTOLOGY
(A. C. Higgins)
CONODONTS
Where specimens can be related to natural assemblages they are referred to multi-element species such as
Gnathodus bilineatus and Idioprioniodus conjunctus. Where this is not possible specimens are referred to disjunct
species. The non-platform elements in the Spanish sections are too disproportionately represented, either due to
breakage or selective sorting, to allow the construction of multi-element species by means of similarity of ratios
and ranges.
Type and figured specimens are housed in the micropalaeontology collection, Department of Geology,
University of Sheffield.
Multi-element species
Genus gnathodus Pander, 1856
Type species. Gnathodus mosquensis Pander 1856.
Gnathodus bilineatus bilineatus (Roundy) 1926
1926 Polygnathus bilineata Roundy, p. 13, pi. 3, fig. 10 a-c.
1964 Gnathodus bilineatus (Roundy), Schmidt and Muller 1964, p. 1 14, figs. 7, 8.
P element
Plate 34, figs. 1, 3
For a recent synonymy see Higgins 1975, p. 28.
HIGGINS AND WAGNER-GENTIS: CONODONTS AND GONI ATITES
329
Remarks. In the Gnathodus bilineatus Zone two morphotypes of this species occur. One is the typical
form of G. bilineatus with an inner platform consisting of a transversely ribbed ridge extending to the
posterior end of the platform whereas the other has a short inner platform which does not extend to
the posterior end. The latter morphotype was referred to G. delicatus by Adrichem Boogaert (1967)
and to G. bilineatus subsp. nov. by Higgins (1974). Although it bears some similarities to G. delicatus
it would considerably extend the range of that species, and its occurrence is separated from that
species by a stratigraphic gap. It is probably better regarded as a variant of G. bilineatus rather than a
relative of G. delicatus.
O Element
Plate 34, fig. 19
1932 Bryantodus delicatus Stauffer and Plummer, p. 29, pi. 2, fig. 27.
1941 Ozarkodina delicatula (Stauffer and Plummer), Ellison, p. 120, pi. 20, figs. 40-42, 47.
For a recent synonymy see Higgins 1975, p. 69.
Remarks. The denticles of the O element and the other non-platform elements of Gnathodus bilineatus
are ornamented with marked striations similar to those of the Oz element of Idiognathodus delicatus
illustrated by von Bitter (1972). They consist of striations, 1 to 2 microns thick, rounded and
sometimes continuous but sinuous and in other instances thickening and thinning down the length of
the denticle. This pattern has been described by Norby (1975).
N Element
Plate 34, fig. 25
1933 Synprioniodina sp. Gunnell, p. 296, pi. 31, fig. 6.
1941 Synprioniodina microdenta Ellison, pp. 108-111, 119, pi. 120, figs. 43-46.
For a recent synonymy see Higgins 1975, pp. 38, 39.
Ala+b Elements
Plate 34, figs. 20, 22
1957 Hindeodella ibergensis Bischoff, p. 28, pi. 8, figs. 33, 37, 39.
For a recent synonymy see Higgins 1975, pp. 38, 39.
Remarks. Baesemann (1973) provided a complete description of these two elements. The Ala
element, typically with forwardly inclined denticles on the anterior process is illustrated extensively
by Higgins (1975, p. 39, fig. 8a, c, d,f). The Alb element, with strongly inwardly flexed denticles on the
anterior process is illustrated by Higgins (1975, p. 39, fig. 8 b, e).
Alc Element
1959 Hindeodina uncata Hass, p. 383, pi. 47, fig. 6.
A recent synonymy is given in Higgins (1975, p. 44).
Remarks. Although this element was not included by Baesemann (1973) or von Bitter (1972) in
Idiognathodus delicatus it is commonly present in Lower and Upper Carboniferous faunas in
association with gnathodid and idiognathodid elements and it was recorded from such assemblages
by Schmidt and Muller (1964, p. 1 17, fig. 9 (6)).
330
PALAEONTOLOGY, VOLUME 25
A 2 Element
Plate 34, fig. 24
1957 Angulodus walrathi Bischoff, p. 17, pi. 5, figs. 44, 45.
1975 Hindeodella simplex (Higgins and Bouckaert), Higgins 1975, pi. 5, figs. 10, 12, 13.
A recent synonymy is given in Higgins 1975, pi. 42.
Remarks. A complete description is given in Higgins and Bouckaert (1968, pp. 28, 29) and Baesemann
(1973, p. 704).
A 3 Element
Plate 34, fig. 26
1965 Hibbardella acuta Murray and Chronic, p. 598, pi. 73, figs. 3-5.
A recent synonymy is given in Higgins 1975, p. 34.
EXPLANATION OF PLATE 34
All specimens x 40
Gnathodus bilineatus bilineatus (Roundy)
Figs. 1, 3. P Element, Villabellaco Section, sample 2902, oral views of 2902 (1 and 2).
Fig. 19. O Element, Olleros de Alba Section, sample OLI, inner lateral views of OLI (1).
Fig. 20. Aib Element, Entrago Section, sample 1149C, inner lateral view of 1 149C (1).
Fig. 22. Ala Element, Entrago Section, sample 1 149K, inner lateral view of 1 149K (1).
Fig. 24. A 2 Element, Matallana Section, sample 1060A, inner lateral view of 1060A (1).
Fig. 25. N Element, Entrago Section, sample 1 149C, inner lateral view of 1 149C (2).
Fig. 26. A3 Element, Entrago Section, sample 1 149K, inner lateral view of 1 149K (2).
Fig. 2. Gnathodus delicatus Branson and Mehl, Pola de Gordon Section, sample 3085, oral view of 3085 (1).
Figs. 4, 14. Gnathodus cuneiformis Mehl and Thomas. 4, Baleas Quarry, Sample 1264, oral view of 1264 (18).
14, Olleros de Alba Section, sample OLV, oral view of OLV (1).
Fig. 5. Gnathodus typicus Cooper. Olleros de Alba Section, sample OLIV, oral view of OLIV (1).
Fig. 6. Gnathodus semiglaber Bischoff. Baleas Quarry, sample 1264, oral view of 1264 (19).
Fig. 7. Gnathodus homopunctatus Bischoff. Revilla Section, sample 134V oral view of 134V (1).
Fig. 8. Protognathodus collinsoni Ziegler. Santiago de las Villas Section, sample 1310, oral view of 1310 (1).
Fig. 9. Gnathodus girtyi girtyi Hass. Revilla Section, sample 134IX, oral view of 134X (1).
Fig. 10. Protognathodus meischneri Ziegler 1969. Santiago de las Villas Section, sample 1310, oral view of 1310.
Fig. 11. Gnathodus texanus pseudosemiglaber Thompson and Fellows 1970, Matallana Section, sample 1060A,
oral view of 1060A (2).
Figs. 12, 13, 15. Paragnathodus multinodosus (Wirth). Revilla Section, sample 134IX, oral view of 134IX (2-4).
Figs. 16, 17. Mestognathus beckmanni Bischoff. Villabellaco Section, sample 2910, oral and aboral views of
2910(1).
Idioprioniodus conjunctus (Gunnell)
Fig. 18. B3a Element, Entrago Section, sample 1 149K, posterior view of 1 149K (3).
Fig. 21. Nj Element, Olleros de Alba Section, sample OLI, outer lateral view of OLI (2).
Fig. 23. Blb Element, Entrago Section, sample 1 149C, inner lateral view of 1 149C (3).
Fig. 27. N2 Element, Matallana Section, sample 1060A, outer lateral view of 1060A (3).
Fig. 28. Bla Element, Matallana Section, sample 1060A, inner lateral view of 1060A (4).
Fig. 29. B3b Element, Matallana Section, sample 1060A, outer lateral view of 1060A (5).
Fig. 30. Scaliognathus sp. nov. Olleros de Alba section, sample 1340, oral view of specimen 1340 (1), x60.
Fig. 31. Scaliognathus sp. nov. Olleros de Alba section, sample OLI, oral view of specimen OLI (5), x 60.
Fig. 32. Scaliognathus sp. nov. Olleros de Alba section, sample OLI, oral view of specimen OLI (4), x 60.
PLATE 34
HIGGINS and WAGNER-GENTIS, Earlier Carboniferous conodonts
332
PALAEONTOLOGY, VOLUME 25
Genus idioprioniodus Gunnell 1933
1933 Idioprioniodus Gunnell, p. 265.
1952 Duboisella Rhodes, p. 895.
1972 Neoprioniodus Rhodes and Muller, von Bitter, p. 68.
1973 Idioprioniodus Gunnell, Baesemann, p. 703.
1974 Idioprioniodus Gunnell, Merrill and Merrill, pp. 1 19-130.
Type species. Idioprioniodus typus Gunnell, 1933.
Idioprioniodus conjunctus (Gunnell) 1931
1931 Prioniodus conjunctus Gunnell, p. 247, pi. 29, fig. 7.
1974 Idioprioniodus conjunctus (Gunnell), Merrill and Merrill, p. 120.
Remarks. Merrill and Merrill (1974) proposed this species for pre-Missourian representatives of
Idioprioniodus which would be assigned to Idioprioniodus typus (Gunnell) except for the presence of
an N2 (metalonchodinid) element.
Nx Element
Plate 34, fig. 21
1931 Prioniodus conjunctus Gunnell, p. 247, pi. 29, fig. 7.
A complete synonymy is given in Higgins 1975, p. 66.
N2 Element
Plate 34, fig. 27
1931 Prioniodus bidentatus Gunnell, p. 247, pi. 29, fig. 6.
1941 Metalonchodina bidentata (Gunnell), Branson and Mehl, p. 106, pi. 19, fig. 34.
A complete synonymy is given in Higgins 1975, p. 63.
Bla Element
Plate 34, fig. 28
1933 Idioprioniodus typus Gunnell, p. 265, pi. 31, fig. 47.
1941 Ligonodina typa (Gunnell), Ellison, p. 1 14, pi. 20, figs. 8-11.
1953 Ligonodina roundyi Hass, p. 82, pi. 15, figs. 5-9.
1972 Neoprioniodus conjunctus (Gunnell), Von Bitter, p. 69, pi. 12, fig. 3, Hi element.
1973 Idioprioniodus lexingtonensis (Gunnell), Baesemann, p. 703, pi. 3, fig. 1.
Remarks. Higgins (1975) regarded Ligonodina roundyi and L. typa as separate species because the
former species has discrete anterior process denticles. However, both species have the same range and
there are transitional specimens throughout this range.
Blt) Element
Plate 34, fig. 23
1931 Prioniodus clarki Gunnell, p. 247, pi. 29, fig. 8.
1941 Lonchodina clarki (Gunnell), Ellison, p. 116, pi. 20, figs. 21, 27, 30, 31.
1957 Lonchodina cf. projecta Ulrich and Bassler, Bischoff, p. 34, pi. 1, fig. 20.
1961 Lonchodina cf. projecta Ulrich and Bassler, Higgins, pi. 11, fig. 10.
1968 Lonchodina bischoffi Higgins and Bouckaert, p. 43.
1972 Neoprioniodus conjunctus (Gunnell), Von Bitter, p. 69, pi. 12, fig. 4 a-c, PI. element.
HIGGINS AND WAGNER-GENTIS: CONODONTS AND GONIATITES
333
1973 Idioprioniodus lexingtonensis (Gunnell), Baesemann, p. 704, pi. 3, fig. 2, BU) element.
1975 Lonchodina bischoffi Higgins and Bouckaert, Higgins, p. 59, pi. 2, figs. 1-4, 8.
For a complete description see Higgins and Bouckaert (1968), p. 43.
Remarks. None of the illustrated specimens of Lonchodina clarki from the Pennsylvanian are as
complete as the specimens assigned to L. bischoffi which have been described from the Visean and
Namurian. Both the anterior process and the posterior process of the former species are incomplete.
None the less, in all other respects the two species are identical and there seems little justification in
retaining them as separate species.
B2 Element
1931 Prioniodus lexingtonensis Gunnell, p. 246, pi. 29, fig. 4.
1941 Ligonodina lexingtonensis (Gunnell), Ellison, p. 115, pi. 20, figs. 13-15.
1973 Idioprioniodus lexingtonensis (Gunnell), Baesemann, p. 704, pi. 3, figs. 3, 8.
For a more complete synonymy and a description see Higgins (1975), p. 58.
B3a Element
Plate 34, fig. 18
1931 Prioniodus subacodus Gunnell, p. 246, pi. 29, fig. 5.
1941 Hibbardella subacoda (Gunnell), Ellison, p. 118, pi. 20, figs. 22, 26.
1953 Roundya barnettana Hass, p. 89, pi. 16, figs. 8, 9.
1958 Roundya costata Rexroad, p. 26, pi. 2, figs. 5-8.
1972 Neoprioniodus conjunctus (Gunnell), Von Bitter, p. 70, pi. 16, figs. 2a, b, Tr element.
1973 Idioprioniodus lexingtonensis (Gunnell), Baesemann, p. 704, pi. 3, fig. 9,B3a element.
Remarks. Specimens from the Missourian illustrated by Baesemann (1973) and the late Pennsyl-
vanian ilustrated by Von Bitter (1972) and referred to Roundya subacoda are indistinguishable from
early Carboniferous specimens of R. barnettana. The distinction between the two species is the
massivity of the unit of R. subacoda compared to R. barnettana. Comparison between them is made
difficult by the incomplete nature of the later Pennsylvanian specimens but the distinction would
seem to be of superficial importance.
B3b Element
Plate 34, fig. 29
1941 Lonchodina ?ponderosa Ellison, p. 116, pi. 20, figs. 37-39.
1958 Lonchodina paraclaviger Rexroad, p. 22, pi. 4, figs. 7-10.
1973 Idioprioniodus lexingtonensis (Gunnell), Baesemann, p. 704, pi. 3, figs. 4, 5, B3b element.
1975 Lonchodina paraclaviger Rexroad, Higgins, p. 60, pi. 2, fig. 9.
Remarks. The similarity between Lonchodina ponderosa and L. paraclaviger is illustrated by Higgins
1975, pi. 2, figs. 9, 1 1 . Both species have processes almost in the same plane, subequal denticles on the
anterior process which are approximately equal in size to the cusp, and a similar basal cavity. The
subsymmetrical nature of the unit suggests that this is a B3 element as suggested by Baesemann rather
than a B2 element.
Disjunct elements
Genus doliognathus Branson and Mehl, 1941
Type species. Doliognathus latus Branson and Mehl, 1941.
334 PALAEONTOLOGY, VOLUME 25
Doliognathus latus Branson and Mehl, 1941
For synonymy up to 1967, see Thompson (1967).
1970 Doliognathus latus Branson and Mehl, Thompson and Fellows, p. 45.
1971 Doliognathus latus Branson and Mehl, Higgins, pi. 2, figs. 1, 3-6, 8.
1971 Doliognathus cf. latus Branson and Mehl, Higgins, pi. 2, figs. 2, 7.
Remarks. The considerable variation in this species was illustrated by Voges (1959, p. 274) and
Higgins (1971, pi. 2). The typical form, illustrated by fig. 1 of pi. 2 (Higgins 1971) has a transversely
ribbed or noded platform, smoothly curved margins, and the outer lateral process is approximately at
right angles to the main axis of the unit. The main variant (pi. 2, figs. 2 and 7) has more pronounced
platform ribs, strongly irregular platform margins, and the outer lateral process is more strongly
curved or directed posteriorly. This latter form was referred to Doliognathus cf. latus by Higgins
(1971).
Genus gnathodus Pander, 1956
Type species. Gnathodus mosquensis Pander, 1956.
Gnathodus cuneiformis Mehl and Thomas
Plate 34, figs. 4, 14
1947 Gnathodus cuneiformis Mehl and Thomas, p. 10, pi. 1, fig. 2.
1971 Gnathodus cf. girtyi Hass, Higgins, pi. 5, fig. 3.
A more complete synonymy is given in Thompson and Fellows 1970, pp. 45, 46.
Remarks. Thompson and Fellows (1970) pointed out that Gnathodus cuneiformis is homeomorphic
with G. girtyi in the United States where the former species occurs in the Osagean and the latter in the
Chesterian Stage. The same pattern can be observed in Spain where G. cuneiformis occurs in the
anchoralis zone of late Tournaisian age and G. girtyi occurs in the bilineatus Zone of late Visean age.
The range of variation of G. girtyi is much greater than that of G. cuneiformis but symmetrically
platformed specimens of both species are indistinguishable.
Gnathodus delicatus Branson and Mehl
Plate 34, fig. 2
1938 Gnathodus delicatus Branson and Mehl, p. 145, pi. 34, figs. 25-27.
1971 Gnathodus delicatus Branson and Mehl, Higgins, pi. 5, figs. 5, 7, 8, 11, 13.
Recent synonymies have been given by Butler (1973, p. 497) and Matthews et al. (1972, p. 559).
Remarks. Butler (1973) commented that specimens from the upper part of the range of this species
show the development of a distinct parapet at the anterior end of the inner side of the platform. This
form (pi. 5, figs. 5, 7 of Higgins 1971) also occurs in the Spanish sections where it overlaps the range of
Gnathodus antetexanus, a species which, as pointed out by Butler, has a similar feature.
Gnathodus girtyi girtyi Hass, 1953
Plate 34, fig. 9
1953 Gnathodus girtyi Hass, p. 80, pi. 14, figs. 22-24.
1975 Gnathodus girtyi girtyi Hass, Higgins, p. 31, pi. 10, figs. 5, 6.
A more complete synonymy is given in Higgins (1975), p. 31.
Remarks. Only the weakly ornamented subspecies of Gnathodus girtyi is represented in the Spanish
succession and it is a rare species.
HIGGINS AND WAGNER-GENTIS: CONODONTS AND GONI ATITES 335
Gnathodus typicus Cooper, 1939
Plate 34, fig. 5
1939 Gnathodus typicus Cooper, p. 388, pi. 42, figs. 77, 78.
1964 Gnathodus typicus Cooper, Rexroad and Scott, p. 31, pi. 2, fig. 3.
1970 Gnathodus typicus Cooper, Thompson and Fellows, pp. 89, 90, pi. 3, figs. 3, 13.
Remarks. Specimens with a short anteriorly pointing inner platform and a weakly ornamented outer,
wide, and rounded platform are referred to this species. Expansion of the carina would allow it to be
placed in Gnathodus semiglaber with which species it has much in common.
Genus paragnathodus Higgins, 1975
Type species. Spathognathodus commutatus Branson and Mehl, 1941.
Remarks. Paragnathodus commutatus and P. nodosus (Bischoff) 1957, have recently been described
by Higgins 1975, pp. 70-72. The composition of the multi-element genus is unknown but it is likely to
correspond to the natural assemblage Lochreia of Scott 1942.
Paragnathodus multinodosus (Wirth), 1967
Plate 34, figs. 12, 13, 15; text-fig. 106
1962 Gnathodus commutatus var. multinodosus Higgins, pp. 8, 9, pi. 2, figs. 13-18.
1967 Gnathodus commutatus multinodosus n.ssp. Wirth, p. 208, pi. 19, figs. 19, 20.
1974 Gnathodus commutatus multinodosus Higgins, Austin and Husri, pi. 2, fig. 13.
Discussion. Variation in this species is mainly seen in the shape of the cup and its ornamentation. The
cup shape can vary from being subsymmetrical to asymmetrical where the inner side is strongly
folded both anteriorly and posteriorly and does not gradually taper to the posterior. The nodes
typically evenly cover the platform surface where they are situated on a slightly raised shelf. PI. 34, fig.
12, illustrates a specimen in which the nodes are raised on two anteriorly directed ridges identical to
those found in Paragnathodus nodosus and it may be a variant of this species. Scanning photographs
reveal the presence of micronodes on the major nodes of P. multinodosus (PI. 35, fig. 2).
Remarks. The distribution of this species is very restricted. Apart from its widespread presence in the
Cantabrian Mountains it occurs in the Pyrenees (Wirth 1967; Marks and Wensink 1970; Perret 1974)
and possibly in Ireland (Austin and Husri 1974) but is unknown elsewhere despite the large
numbers of faunas of this age which have been described. The Irish occurrence is abnormal because it
apparently occurs before P. nodosus , whereas in Spain it appears slightly later.
Genus scaliognathus Branson and Mehl, 1941
Type species. Scaliognathus anchoralis Branson and Mehl, 1941.
Discussion. Despite the widespread nature and stratigraphical importance of this distinctive anchor-
shaped genus it remains monospecific. The Spanish successions, being condensed and broken by non-
sequences, do not give a clear pattern of evolution of the genus. Nevertheless, the material is well
preserved and abundant, and it is possible to identify the main variations in the genus. Three main
forms are recognized:
1. Forms with plate-like lateral and anterior processes. These are referred to Scaliognathus
anchoralis.
2. Forms with a plate-like anterior process, but blade-like subequal lateral processes which are
curved anteriorly. These are referred to S. angustilateralis sp. nov.
3. Forms with straight unequal lateral limbs which project at right angles to the anterior process.
The posterior limbs are blade-like. These are referred to Scaliognathus sp. nov.
336
PALAEONTOLOGY, VOLUME 25
Groessens (1977a) suggested an origin for Scaliognathus from Dollymae bouckaerti (Groessens) in
which the first scaliognathid was a slender, highly arched form with poor plate development. This
form would resemble form 2 above.
Scaliognathus anchoralis Branson and Mehl, 1941
Text-fig. 9a, b , d, g
1941 Scaliognathus anchoralis Branson and Mehl, p. 102, pi. 19, figs. 29-32.
1964 Scaliognathus anchoralis Branson and Mehl, Higgins et al. pi. iv, fig. 17.
1967 Scaliognathus anchoralis Branson and Mehl, Adrichem Boogaert, p. 50, pi. 5, figs. 2-4, 8, 9.
1969a Scaliognathus anchoralis Branson and Mehl, Matthews, pp. 272, 273, pi. 49, figs. 2, 4, 8, and 9
only.
19696 Scaliognathus anchoralis Branson and Mehl, Matthews, pi. 51, figs. 1, 2.
1971 Scaliognathus anchoralis Branson and Mehl, Groessens, pi. 1, fig. 10 only.
1971 Scaliognathus anchoralis Branson and Mehl, Higgins, pi. 3, figs. 3, 5-7, 9; pi. 4, fig. 2.
1974 Scaliognathus anchoralis Branson and Mehl, Matthews and Thomas, pi. 50, figs. 8, 9.
1974 Scaliognathus anchoralis Branson and Mehl, Jenkins, pi. 1 19, figs. 6-9.
text-fig. 9. a, Scaliognathus anchoralis Branson and Mehl. Olleros de Alba section, sample 1340, oral view of
specimen 1340 (2), x 60. b , Scaliognathus anchoralis Branson and Mehl. Baleas Quarry, sample 1264, oral view
of specimen 1 264 (23), x 60. c,/, Scaliognathus angustilateralis sp. nov. Baleas Quarry, sample 1264; fig. c aboral
view of cusp, x 1080; fig. / aboral view of specimen 1264 (20), x 60. d, g, Scaliognathus anchoralis Branson and
Mehl. Olleros de Alba section, sample OLI; fig. 5 oral view of specimen OLI (3), x 60; fig. g detail of main
denticle, x 360. e, Scaliognathus angustilateralis sp. nov. Baleas Quarry, sample 1264, oral view of holotype,
specimen 1264(22), x60.
HIGGINS AND WAGNER-GENTIS: CONODONTS AND GONI ATITES
337
Diagnosis. Paired conodonts with an anchor-like shape consisting of three processes. The anterior
process is wide, tapering to its anterior extremity, the two posterior lateral processes are also wide and
plate-like with a row of posteriorly inclined denticles on a flat and horizontal upper surface, which is
in the same plane as the surface of the lateral process.
Description. The anterior limb is broad at the posterior tapering anteriorly with a convex outer and a convex,
straight or slightly concave, inner margin. There is a prominent median carina consisting of large discrete
denticles at the anterior becoming fused into a low ridge towards the posterior. At the posterior the carina is
continued beyond the extremity of the platform as a horn-like denticle which is commonly large and may have
a triangular cross-section where the carinal ridge continues up its oral face, but there is no sign of this in the
holotype which has a small posterior denticle. There is commonly a sulcus on either side of the carina and
bordering the sulci is a row of nodes which may develop into transverse ridges. The lateral extension of the
anterior limb is marked, giving rise to a plate-like structure with sharp margins and a flat to concave upper
surface.
The lateral processes are as long or slightly shorter than the anterior limb either in length, denticulation, or
curvatures. They are projected or curved anteriorly at approximately 70° to the anterior limb and, if curved,
curvature is stronger on the outer limb. They are as wide or wider than the anterior limb in the median area of the
unit but taper to a point. The oral surface is flat and in the same plane as the oral surface of the anterior limb with
which it is continuous. There is commonly a row of low nodes along its anterior margin. Near, or at the posterior
margin, is a row of posteriorly inclined denticles which may be subequal in size or increase slightly in height
towards the extremity of the limbs. The holotype and para type have a prominent shelf, with a row of the
transverse ridges posterior to the denticle row, but this is atypical for the species in general and more commonly
the shelf, although often present, is insignificant. The denticles may be discrete, but are more commonly in
contact in the lower third of their length.
The aboral surface of both the anterior and the lateral limbs is convex and smooth except for prominent split
keels which are grooved along their length. The keels meet at the centre of the lateral limbs and a triangular open
pit which is open in small specimens but becomes more closed in adult specimens. The anterior face of the lateral
process is almost at right angles to the oral face, whereas the posterior face is at an angle of approximately
45 degrees.
Discussion. The holotype of Branson and Mehl (1941, pi. 19, figs. 30, 32) has very broad plate-like
lateral processes and a denticular row which originates near the midline of these processes. The
specimen figured by Jenkins (1974, pi. 1 19, fig. 6) is close to the holotype but few others, mainly from
Europe, are exactly of this type. The description and diagnosis given above retains the plate-like
nature of the processes but the typical specimen has a denticular row originating from the posterior
edge of the processes which typifies the majority of figured specimens. This redefinition of the species
would include form 1 and possibly form 2 of Matthews (1969a).
Range. Scaliognathus anchoralis Zone.
No. of specimens fifty.
Scaliognathus angustilateralis sp. nov.
Text-fig. 9c, e,f
1967 Scaliognathus anchoralis Branson and Mehl, Adrichem Boogaert, p. 185, pi. 3, fig. 11.
1969 Scaliognathus anchoralis Branson and Mehl, Matthews, pp. 272, 273, pi. 49, figs. 1, 6 only.
1971 Scaliognathus anchoralis Branson and Mehl, Groessens, pi. 1, fig. 9 only.
1971 Scaliognathus anchoralis Branson and Mehl, Higgins, pi. 3, figs. 1, 2, 4, 8, 10 only.
1973 Scaliognathus anchoralis Branson and Mehl, Butler, p. 510, pi. 58, figs. 6 and 7 only.
Derivation of name. Refers to the narrowness of the lateral process.
Holotype. Text-fig. 9 d, from the Baleas Formation, Baleas Quarry, Pola de Gordon. Slide 1264(22).
Diagnosis. A species of Scaliognathus with narrow lateral processes which are posteriorly in-
clined, curved anteriorly, and in the same plane as the row of denticles which are developed on its
surface.
338
PALAEONTOLOGY, VOLUME 25
Description. The anterior limb is slender, plate-like, with a convex outer and convex to concave inner margin
which are approximately parallel in the posterior half but taper sharply in the anterior half usually terminating
before the end of the process. The oral surface of the limb is bisected by a median carina consisting of partially
fused, large denticles in the anterior half which becomes a fused, low, nodular ridge in the posterior half
extending up the oral face of the horn-like terminal denticle. The margins of the limb are slightly crenulate
because of the development of a row of nodes and transverse ridges on each side. These are separated from the
carina by a shallow sulcus.
The lateral limbs are subequal and are curved anteriorly, often strongly, and in extreme variants the anterior
half of the outer limb may be parallel to the anterior limb. There is no development of a plate on the limbs and
they consist of narrow, strongly posteriorly inclined and often curved, blade-like processes. There may be a row
of small nodes along the anterior margin or the oral face may be smooth. The denticles originate from the
posterior margin and are in the same plane as the processes. They are long, discrete for more than half their
length, and laterally compressed.
The aboral side is convex and has a raised split keel which meets in a triangular pit.
The oral surface of the platform adjacent to the main denticle and the carina is covered by an irregular to
hexagonal pattern of furrows. These cross and interrupt a pattern of branching fine striae. The oral surface of the
main denticle also has a pattern of coarse striations but raised rather than sunken and always semiregular. These
run approximately parallel to the margin of the denticle and meet at the ridge which bisects it. There are no fine
striae on the oral surface. The aboral surface has an irregular pattern of coarse striations, again ridges, overlying
a finer one, but in this instance the finer one may branch off the coarser ones. These patterns are also found in
Scaliognathus anchoralis and Scaliognathus sp. nov.
Range. Lower part of the anchoralis Zone.
No. of specimens sixty.
Scaliognathus sp. nov.
Plate 34, figs. 30-32; text-fig. 10a
1969 Scaliognathus anchoralis Branson and Mehl, Matthews, pp. 272, 273, pi. 49, figs. 5 and 7 only.
Description. Anterior limb is slender and strongly arched with convex margins to a weakly developed plate. The
oral surface of the plate is unornamented and smooth except for the median row of discrete pointed denticles
which are inclined posteriorly.
The lateral limbs are unequal, the outer being up to twice the length of the inner. Both limbs are slender, being
only slightly thickened adjacent to the junction with the anterior limb. They are steeply inclined, almost vertical,
with a slight inclination towards the posterior. Their oral edge is ornamented with discrete, long denticles
which are laterally flattened, with up to five on the outer and two or three on the inner side. The denticles are in
the same plane as the limbs. The largest denticle is at the end of the carina and is triangular being bisected by
a low ridge.
text-fig. 10. a, Scaliognathus sp. nov. Detail of platform area adjacent to main denticle of specimen 1340 (1),
xl080. b, Paragnathodus multinodosus (Wirth). Detail of carinal node of specimen 1341X, x 550. c,
Protognathodus meischneri Ziegler 1969. Detail of carinal node of specimen 1310 (2), x 550.
HIGGINS AND WAGNER-GENTIS: CONODONTS AND GONI ATITES
339
The aboral side of the anterior limb is convex but that of the lateral limb is sharp edged except adjacent to their
junction with the anterior limb. All the limbs have split keels which meet in a triangular cavity beneath the base of
the triangular denticle.
Range. Upper part of the anchoralis Zone.
GONIATITES
(C. H. T. Wagner-Gentis)
Family prolecanitidae Hyatt, 1884
Genus merocanites Schindewolf, 1922
Type species. Ellipsolithes compressus Sowerby, 1813.
Merocanites marshallensis (Winchell)
Plate 35, figs. 1, 2, 6; text-fig. 11a, b
1862 Goniatites Marshallensis Winchell, pp. 362, 363.
1955 Merocanites marshallensis Winchell, Miller and Garner, pp. 154-157, pi. VII, figs. 5-9, text-figs.
13c and 16.
Material. Two specimens and a large number of whorl-sections from localities 138, 138d, 1548, from the
Villabellaco Limestone in Palencia.
Description. The shell is a serpenticone with a rectangular cross-section of the whorl. The venter and ventro-
lateral edge are both rounded, the umbilical wall is flat and perpendicular to the lateral side (see text-fig. 1 1 A).
Neither ornament nor constrictions have been observed. The suture consists of an inflated ventral lobe of which
the siphonal point can be rather long. The ventro-lateral saddle is rounded, constricted, and low. The first lateral
lobe is pointed, constricted, and not as deep as the ventral lobe, but it is wider than the ventral or second lateral
lobe. The following lateral saddle is rounded, constricted, and higher than the ventro-lateral saddle. The second
lateral lobe is pointed, constricted, and longer than the first lateral lobe. The last lateral saddle is rounded,
constricted, and smaller than the preceding saddles. The third lateral lobe is pointed, considerably smaller than
the previous lateral lobes, and slightly asymmetrical. The suture crosses the umbilical wall in a straight line,
sloping downwards. It forms a lobe on the dorsal side, which is followed by a narrow, rounded saddle. The
median, dorsal lobe is narrow, long, and rounded (see text-fig. 1 In).
The dimensions of one of the shells (loc. 138d) are: diameter 38 mm; width 10 mm; umbilicus 14 mm; height of
whorl 15 mm; opening 13 mm.
Remarks. Merocanites marshallensis europaeus Kullmann, 1963, pp. 276-278, from the Esla area and
Puente de Alba, province of Leon, is quite different from M. marshallensis (Winchell), in that its
whorl cross-section is too circular, its ventral lobe, although inflated, is considerably slimmer, and its
third lateral lobe is too developed.
Comparisons. The specimens described here differ from M. applanatus bicarinatus Pareyn in that the
first lateral lobe is wider; the whorls are slightly more indented and, above all, no remnants are found
of any ventro-lateral ridges. One specimen from Olleros de Alba (Leon), which is identical to
bicarinatus, shows very clearly the differences mentioned above.
Occurrence. M. marshallensis (Winchell) is known from the Marshall Sandstone in Michigan, U.S.A., where it
occurs together with Winchelloceras allei, Muensteroceras oweni, Kazakhstania karagandaensis, Gatlendorfia
stummi, and Imitoceras romingeri which indicate a basal Osagean age (Furnish and Manger 1973, p. 3). In the
Villabellaco Limestone (Palencia), it occurs together with Merocanites subhenslowi, Nautellipsites hispanicus,
Ammonellipsites kayseri, and Pseudogirtyoceras villabellacoi at 1-5 to 2-5 m above the base of the limestone,
which indicate a lowest Visean age.
Family muensteroceratidae Librovitch, 1957
Genus muensteroceras Hyatt, 1 883
Type species. Goniatites oweni var. parallela Hall, 1860.
340
PALAEONTOLOGY, VOLUME 25
Merocanites morshollensis (Winchell)
a x2 b xl
text-fig. 1 1. a, Suture of specimen BM(NH) C.82316. b, Cross-section of same
specimen.
Muensteroceras parallelum (Hall)
Plate 35, figs. 4, 5; text-fig. 12
1860 Goniatites oweni var. parallela Hall, p. 100, figs. 13-14.
1903 Muensteroceras parallelum Hall, J. P. Smith, pp. 121, 122, pi. XVI, fig. 3; pi. XIX, figs. 1, 2.
1927 Muensteroceras aff. parallelum Hall, Librovitch, pp. 32, 33, pi. V, figs. 8, 9; pi. VI, fig. 1.
1951 Munsteroceras parallelum Hall, Miller and Collinson, p. 471, fig. 7.
1961 Munsteroceras parallelum Hall, Pareyn, pp. 96, 97, pi. VII, figs. 1-3.
Material. One solid specimen from locality 136b from the Villabellaco Limestone (Palencia).
Description. The shell is a very flat, involute platycone, with rounded venter and flat sides. At a diameter of
60 mm the width measures 1 5 mm. The umbilicus is poorly preserved, but is probably one-sixth of the diameter.
The previous whorl, however, at a diameter of 22 mm has a width of 10 mm, which is twice as wide as the last
whorl. It also has more rounded lateral sides than those of the last whorl. Neither constrictions nor ornament
are preserved. The sutures consist of ventral lobes with parallel sides which touch each other, and thus form
two parallel lines on the venter. The ventro-lateral saddles are fairly narrow and rounded; the lateral lobes are
V-shaped and their points reach lower than the ventral lobes; they do not, however, touch the preceding ventro-
lateral saddles. The lateral lobes are wide and rounded. There is a small, narrow, pointed umbilical lobe on the
umbilical edge. The dimensions of the shells are: diameter 60 mm; width 15 mm; umbilicus 10-12 mm; height
of whorl 25 mm.
Comparisons. Muensteroceras parallelum is very similar to M. rotella de Koninck but the latter
appears to differ by having the ventral lobes encasing each other, whereas in M. parallelum they only
touch each other.
EXPLANATION OF PLATE 35
Fig. 1. Merocanites marshallensis (Winchell). BM(NH) C.82315. Lateral view, showing the rather wide ventro-
lateral lobe, x 3.
Fig. 2. Merocanites marshallensis (Winchell). BM(NH) C.82316. Lateral view, showing shape of shell, x 3.
Fig. 3. Muensteroceras cf. crassum (Foord). BM(NH) C.82318. Ventro-lateral view, showing sutures, x 3.
Fig. 4. Muensteroceras parallelum (Hall). BM(NH) C.82317. Showing lateral and part of ventral sutures, x 1.
Fig. 5. Muensteroceras parallelum (Hall). Same specimen as fig. 4. Lateral view, showing the different shapes of
the last and penultimate whorl, x 1 .
Fig. 6. Merocanites marshallensis (Winchell). BM(NH) C.82315. Ventral view, showing inflated ventral lobe, x 1.
PLATE 35
HIGGINS and WAGNER-GENTIS, Earlier Carboniferous goniatites
342
PALAEONTOLOGY, VOLUME 25
Occurrence. Villabellaco Limestone, locality 1 36b, at approximately the same horizon as 138c, but further east in
the outcrop. M. parallelum was found originally in the Rockford Limestone of Indiana, U.S.A. It is also known
from Hassi Sguilma in Algeria, in the S! unit of Pareyn 1961, and from the Tien Shan in Central Asia (Librovitch
1927).
Muensteroceros parallelum (Hall)
text-fig. 12. Suture of specimen BM(NH)
C. 82317.
Muensteroceras cf. crassum Foord
Plate 35, fig. 3; text-fig. 13
1903 Glyphioceras ( Muensteroceras ) crassum Foord, pp. 193-194, pi. XLIII, fig. 10 a-c.
1927 Muensteroceras crassum Foord; Librovitch, pp. 34-35, pi. VI, fig. 6 a-c.
1941 Muensteroceras crassum Foord; Delepine, p. 58, pi. II, figs. 4-6.
1961 Muensteroceras crassum Foord; Pareyn, pp. 100-101, pi. VIII, figs. 11-16.
1964 Muensteroceras cf. crassum Foord; Wagner-Gentis in Higgins et al.
Material. One solid specimen from the Villabellaco Limestone, found between localities 138b and c.
Description. The shell is an involute ellipsocone. Width about half the diameter. The whorls have a rounded
venter and rounded sides, with the greatest width near the umbilicus. The umbilicus is approximately one-
quarter of the diameter. Umbilical edges are rounded and the umbilical wall is almost perpendicular and sloping
towards the centre of the umbilicus (see text-fig. 1 3 b).
No ornament has been preserved. There is a faint, shallow impression of a constriction which crosses three-
quarters of the sides more or less in a straight line, and then forms a rounded sinus across the venter.
The suture line consists of a fairly narrow, parallel-sided ventral lobe with a low median saddle. The ventro-
lateral saddles are rounded and narrow; the lateral lobes are pointed and as deep as the ventral lobe. The ventral
side of the lateral lobe is straight, whereas the umbilical side is curved. The second lateral saddle is low, wide, and
rounded. The umbilical lobe is small and pointed, and situated just past the umbilical edge, on the umbilical wall
(see text-fig. \2>a). The suture lines do not encase each other.
Dimensions. Diameter 23 mm; width 12 mm; umbilicus 6 mm; height of whorl 9 mm.
Comparisons. The specimen compares with Muensteroceras subglobosum Librovitch, 1927,
pp. 35-36, text-fig. 17; pi. VI, fig. 7; pi. VII, figs. 1, 2, in having a similar suture line, but differs in
having a larger umbilicus.
Occurrence. Villabellaco Limestone (Palencia) between the localities 1 38b and c. It has also been described from
Olleros de Alba (Leon). M. crassum was first recorded from the Lower Carboniferous Limestone of
Ballinacarriga, Co. Limerick, Eire. It is also known from Hassi Sguilma, in Algeria, in the unit of Pareyn
(1961), and from the Tien Shan in Central Asia (Librovitch 1927).
Family pericyclidae hyatt, 1900
Genus ammonellipsites Parkinson, 1822
Type species. Ellipsolithes funatus Sowerby, 1814.
HIGGINS AND WAGNER-GENTIS: CONODONTS AND GONI ATITES
343
Ammonellipsites kayseri (Schmidt)
Plate 36, figs. 2, 3, 5-7; text-fig. 14a, b
1889 Pericyclus virgatus Holzapfel, p. 34, Taf. Ill, figs. 8, 9.
1925 Pericyclus kayseri Schmidt, pp. 554, 555, Taf. 20, fig. 10.
Material. Ten specimens from the Villabellaco Limestone (Palencia) and one specimen from Olleros de Alba
(Leon). All show sutures and a number show the ornament fairly clearly or at least traces of the ornament. A
plaster cast of Holzapfel’s specimen (1889, III, figs. 8, 9) has been used in the description.
Description. The shell is discoidal, involute with a small umbilicus. The venter and the sides are rounded. The
early whorls are wider than high, but gradually become higher than wide. The greatest width is half-way down
the lateral side of the whorl. The umbilicus, about one-fifth of the diameter, has rounded shoulders and rounded
sides, which are perpendicular to the lateral sides (see text-fig. 14 b).
The ornamentation consists of fairly fine undivided ribs, which lean slightly forward over the sides, and form a
very shallow sinus over the venter. The ribs are sharp edged and on the venter the distance between the ribs is
more than the width of the rib itself. A specimen with a diameter of about 20 mm has six ribs per 5 mm, on the
venter. No constrictions are observed.
The suture consists of a medium saddle, which reaches to one-third of the ventro-lateral saddles. These are
slightly spatulate with rounded points and have a tendency to diverge. The rounded points are due to the easy
erosion of the sharp point. The lateral lobe is spatulate and pointed. The second lateral saddle is spatulate with a
rather blunted point and reaches four-fifths of the height of the first lateral saddle. The umbilical lobe is situated
on the umbilical wall and is pointed and wide.
Villabellaco Limestone Plastercastofholotype
B.M. (N.H.) numbers C82324 C82321 C82322 Holzapfel’s specimen
Diameter
20 mm
19 mm
—
38-5 mm
Width
10 mm
1 1 mm
17 mm
19
mm
Umbilicus
—
4 mm
—
8
mm
Height of whorl
8 mm
8 mm
20 mm
17
mm
Height of opening
—
6 mm
12 mm
12
mm
Comparison. The specimens look very similar to Pericyclus virgatus de Koninck, and Holzapfel
identified his specimens from Liebstein with this species. Unfortunately, Pericyclus virgatus does not
show any sutures and it is therefore impossible to decide whether kayseri is the same as virgatus.
Occurrence. In the Erdbach and Breitscheid cephalopod limestone at Liebstein, Germany (Holzapfel 1889, pp.
34 and 35; Schmidt 1925, p. 494), which according to Schmidt (1925) belongs to the Ily zone of the Visean, which
344
PALAEONTOLOGY, VOLUME 25
Ammonellipsites kayseri (Schmidt)
t
specimen BM(NH) C.82322.
is also known as the Pey of the Erdbachium, the very base of the Visean. From Spain, Schmidt (1931, p. 1035)
recorded it together with P. kochi, Imiteroceras sp., Merocanites applanatus, Muensteroceras aff. inconstans, M.
cf. spheroidale, in Palencia at 1 -3 to 2-8 m above the base of the Villabellaco Limestone, and in Leon at Olleros de
Alba. In Great Britain it is mentioned by Prentice and Thomas (1960, p. 6) from Tawstock and Codden Hill
where it occurs with Prolecanites aff. similis.
Family girtyoceratidae Wedekind, 1918
Genus winchelloceras Ruzhencev, 1965
Type species. Beyrichoceras allei Miller and Garner, 1955.
Winchelloceras palentinus sp. nov.
Plate 36, fig. 1; text-fig. 15a-c
Type material. One solid specimen preserved in a marly limestone, showing part of the living chamber, sutures,
and constrictions.
Repository of holotype. British Museum (Nat. Hist.), No. C82317.
Diagnosis. Shell involute, platyconiform with an umbilicus less than one-sixth of the diameter.
Constrictions strongly marked, with a deep, narrowly rounded sinus on the venter, ventro-lateral
EXPLANATION OF PLATE 36
Fig. 1. Winchelloceras palentinus sp. nov. BM(NH) C.82319. Ventro-lateral view, showing sutures and
constriction, x 3.
Fig. 2. Ammonellipsites kayseri (Schmidt). BM(NH) C. 82324. Ventro-lateral view, showing ornament, x 3.
Fig. 3. Ammonellipsites kayseri (Schmidt). BM(NH) C.82320. Ventro-lateral view, showing ornament and
suture, x 3.
Fig. 4. Pseudogirtyoceras villabellacoi sp. nov. BM(NH) C.82323. Ventro-lateral view, showing sutures and
keeled venter, x 3.
Fig. 5. Ammonellipsites kayseri (Schmidt). BM(NH) C.82322. Lateral view, showing traces of ornament and
sutures, x 3.
Fig. 6. Ammonellipsites kayseri (Schmidt). BM(NH) C.82321. Ventral view showing sutures, x 2.
Fig. 7. Ammonellipsites kayseri (Schmidt). Plastercast of holotype. Lateral view, x 1-5.
PLATE 36
HIGGINS and WAGNER-GENTIS, Earlier Carboniferous goniatites
346
PALAEONTOLOGY, VOLUME 25
salient, shallow sinus, and low salient on the lateral sides. Suture with a ventral lobe, of which the
spikes are pointed outwards and the cheeks strongly diverging apically. At a diameter of 38 mm the
median saddle is still at the very bottom of the ventral lobe. Ventro-lateral saddle rounded, lateral
lobe pointed, second lateral saddle wide and rounded. A small V-shaped umbilical lobe is positioned
on the umbilical wall.
Description. The shell is a nearly flat, involute platycone. The fairly flat, only slightly rounded, venter is
perpendicular to the almost flat lateral sides. The ventro-lateral edge is rounded and there is a suggestion of a
groove just below the edge. From there, the lateral side gently curves to become flat near the umbilical region,
where the shell has its greatest width. The umbilicus is small, and stepped with rounded edges and narrow walls,
which are perpendicular to the sides (see text-fig. 156).
text-fig. 15. a , Suture of specimen BM(NH) C. 82319. b. Cross-section,
same specimen, c. Constriction, same specimen.
No ornament is preserved. The shell shows four deep, narrow constrictions per whorl. They form a deep,
narrowly rounded sinus over the venter and a ventro-lateral salient, which may create the impression of a ventro-
lateral groove. On the lateral side the constrictions form a shallow sinus and low salient. They then fade out into
the umbilicus (see text-fig. 15c).
The sutures have a ventral lobe with an extremely low median saddle, flanked by spikes that are pointing
outwards. The cheeks of the ventral lobe are sinuous, first bending outwards and then straightening, in a manner
which is opposite to the sinuous cheeks of the ventral lobe in the beyrichoceratids. The ventro-lateral saddles are
rounded. The pointed lateral lobes are wide and slightly inflated and the second lateral saddles are wide and
rounded, ending in a small V-shaped lobe on the umbilical wall (see text-fig. 1 5a). There are twelve sutures visible
on the last whorl, half of the whorl forming the living chamber.
The dimensions of the shell are: diameter 38 mm; width 10 mm; umbilicus 6 mm; height of whorl 19 mm; height
of opening 7 mm.
Comparisons. Differs from Winchelloceras allei in having a lower median saddle, a flatter venter, and
generally a slimmer outline.
Occurrence. Found in the basal 50 cm of the Villabellaco Limestone (Palencia) section. The genus is known from
the U.S.A., where it occurs in the Coldwater shale/Marshall sandstone of Michigan, which is of basal Osagean
age (Furnish and Manger 1973). It is also recorded from Tien Shan (Popov 1968) and the Urals (Popov 1975) in
rocks of CjVj age.
pseudogirtyoceras gen. nov.
Type species. Pseudogirtyoceras villabellacoi sp. nov.
HIGGINS AND WAGNER-GENTIS: CONODONTS AND GONI ATITES
347
Diagnosis. P seudogirtyoceras closely fits the description of Girtyoceras, but is distinguished by having
a narrow ventral lobe.
Repository of type species. British Museum (Nat. Hist.), No. C82323.
P seudogirtyoceras villabellacoi sp. nov.
Plate 36, fig. 4; text-fig. 16a, b
Description. The shell is an involute oxycone at a diameter of 20 mm. At a diameter of approximately 1 5 mm the
venter is still rounded but at 20 mm it starts to form a keel. The greatest width (10 mm) is at the umbilicus from
where the sides curve gently to the keel. The umbilicus is two-sevenths of the diameter, its walls being nearly
perpendicular to the sides with which it makes a rounded edge (see text-fig. 16 b). Two constrictions are visible,
and these form a wide, shallow sinus on the lateral sides and a sinus on the ventral side. No ornament is
preserved.
The suture consists of an extremely narrow ventral lobe, with its cheeks diverging apically. The two secondary
lobes on both sides of the median saddle are narrow and pointed. The median saddle reaches over half the height
of the ventral lobe. The siphonal notch consists of a deep loop. The first lateral saddles are rounded and directed
towards the umbilicus. The first lateral lobes are V-shaped and pointed, whereby the ventral cheek is convex and
the umbilical cheek is concave. The second lateral saddle is wide and rounded and at the umbilical edge there is a
small, pointed umbilical lobe (see text-fig. 16a).
text-fig. 16. a, Suture of specimen BM(NH) C. 82323. b , Cross-section of same
specimen.
Comparison. This specimen is similar to Girtyoceras form H of Moore (1946, p. 403) but differs in
having a much narrower ventral lobe.
Occurrence. It occurs in the Villabellaco Limestone (Palencia) at locality 138d, 3 m above the base, together with
Nautellipsites hispanicus, Ammonellipsites kayseri, Merocanites marshallensis, and M. subhenslowi.
Merocanites subhenslowi Wagner-Gentis
See Higgins et al. 1964, pp. 238-245, pi. Ill, figs. 1 1-13; pi. IV, fig. 14; text-figs. A-c, for the description of this
species.
Nautellipsites hispanicus (Foord and Crick)
See Wagner-Gentis 1960, pp. 43-51, figs. 1-3, for the description of this species.
Acknowledgements. We wish to thank Dr. R. H. Wagner for collecting many of the samples, and for considerable
discussion during the writing of the paper.
348
PALAEONTOLOGY, VOLUME 25
REFERENCES
adrichem boogaert, h. a. van. 1967. Devonian and Lower Carboniferous conodonts of the Cantabrian
Mountains (Spain) and their stratigraphic application. Leid. geol. Meded. 39, 129-192.
Austin, R. l. 1974. The biostratigraphic distribution of conodonts in Great Britain and the Republic of Ireland.
Belgian Geol. Survey Intern. Symposium on Belgian Micropalaeontological Limits, Namur 1974, publ. No. 3,
1-17.
— and husri, s. 1974. Dinantian conodont faunas of County Clare, County Limerick and County Leitrim. An
appendix. Ibid. 18-69.
baesemann, J. f. 1973. Missourian (Upper Pennsylvanian) conodonts of north-eastern Kansas. J. Paleont. 47, 4,
689-710.
barrois, c. 1882. Recherches sur les terrains anciens des Asturies et de la Galice. Mem. Soc. Geol. du Nord, I,
1-630.
bischoff, G. 1957. Die Conodonten-Stratigraphie des rheno-herzynischen Unterkarbons mit Berucksichtigung
der fVocklumeria-Stufe und der Devon/Karbon-Grenze. Abh. hess Landesamt Bodenforsch. 19, 1-64.
branson, e. b. and mehl, m. G. 1938. Conodonts from the Lower Mississippian of Missouri. Univ. Mo. Stud. 13,
128-148.
— 1941. New and little known Carboniferous conodont genera. J. Paleont. 15, 97-106.
brenckle, p., lane, r. and collinson, c. 1974. Progress towards reconciliation of Lower Mississippian
Conodont and Foraminiferal Zonations. Geology, 433-436.
brouwer, a. 1964. Deux facies dans le Devonien des Montagnes Cantabriques meridionales. Breviora geol.
astur. 8, 1-4, 3-10.
budinger, p. and kullmann, j. 1964. Zur Frage von sedimentations brechungen im goniatiten-und conodonten
fiihrenden oberdevon und Karbon de Kantarbischen Gebirges (Nordspanien). Neues Jb. Miner. Geol.
Paldont. 7, 414-429.
butler, m. 1973. Lower Carboniferous conodont faunas from the Eastern Mendips, England. Palaeontology,
16, 3,477-517.
comte, p. 1959. Recherches sur les terrains anciens de la Cordillere Cantabrique. Mem. Inst. Geol. Minero
Espana, 60, 1 -440.
conil, r. and paproth, e. 1968. Mit Foraminiferen gegliederte Profile aus dem nordwest-deutschen Kohlenkalk
und Kulm. Decheniana, 119, 51-94.
cooper, c. l. 1939. Conodonts from a Bushberg-Hannibal horizon in Oklahoma. J. Paleont. 13, 379-442.
delepine, g. 141. Les Goniatites du Carbonifere du Maroc et des confins algero-marocains du Sud (Dinantien-
Westphalien). Not. Mem. Serv. Geol. Maroc., No. 56, pp. 1-110.
del Rio, p. and menendez alvarez, j. r. 1978. Estudio lito y biostratigrafico de la Caliza de Candamo (Loredo,
zona central de Asturias). Trabajos de Geol. 379-387.
foord, a. h. 1903. Monograph of the Carboniferous Cephalopoda of Ireland. Palaeontogr. Soc. of London, V,
147-234.
furnish, w. m. and manger, w. L. 1973. Kinderhook Ammonoids. Proc. Iowa Acad. Sci. 80, 1-10.
ginkel, a. c. van, 1965. Spanish Carboniferous fusulinids and their significance for correlation purposes. Leid.
geol. Meded. 34, 175-225.
groessens, e. 1977a. Hypothese concernant revolution de conodontes utiles a la biostratigraphie du Dinantien.
Belgian Geol. Survey Intern. Symposium on Belgian Micropaleontological Limits, Namur 1974, publ. No. 16,
1-16.
— 19776. Distribution de conodontes dans le Dinantien de la Belgique. Ibid. publ. no. 17, 1-193.
hass, w. h. 1953. Conodont of the Barnett Formation of Texas. U.S. Geol. Surv. Prof. Paper, 243-F, 69-94.
higgins, a. c. 1962. Conodonts from the ‘Griotte’ Limestone of North West Spain. Notas Comun. Inst. geol.
min. Esp. 65, 5-22.
— 1971. Conodont biostratigraphy of the late Devonian-early Carboniferous rocks of the south central
Cantabrian Cordillera. Trabajos de Geol. 3, 179-192.
— 1974. Conodont zonation of the Lower Carboniferous of Spain and Portugal. Belgian Geol. Survey Intern.
Symposium on Belgian Micropaleontological Limits, Namur 1974, publ. no. 4, 1-17.
1975. Conodont zonation of the late Visean-early Westphalian strata of the south and central Pennines of
northern England. Bull. geol. Surv. Gr. Br. 53, 1 -90.
— and bouckaert, j. 1968. Conodont stratigraphy and palaeontology of the Namurian of Belgium. Mem.
Expl. Cartes Geol. et Min. Belgique, 10, 1-64.
— wagner-gentis, c. h. t. and wagner, r. h. 1964. Basal Carboniferous strata in part of Northern Leon,
HIGGINS AND WAGNER-GENTIS: CONODONTS AND GONIATITES
349
NW Spain: Stratigraphy, Conodont and Goniatite Faunas. Bull. Soc. belg. Geol. Paleont. Hydrol. LXXII,
205-248.
holzapfel, E. 1889. Die Cephalopoden-fuhrenden Kalke des unteren Carbon von Erdbach-Breitscheid bei
Herborn. Palaont. Abh., n.f. Bd. I, Heft 1, pp. 1-73.
jenkins, T. b. h. 1974. Lower Carboniferous conodont biostratigraphy of New South Wales. Palaeontology ,
17, 909-924.
kullmann, J. 1963. Die Goniatiten des Unterkarbons im Kantabrischen Gebirge (Nordspanien). II.
Palaontologie der U.O. Prolecanitina Miller & Furnish. Die Altersstellung der Faunen. N. Jb. Geol. Palaont.
116, 3, 269-324.
librovitch, l. s. 1927. Lower Carboniferous Cephalopods from the Son Kul Region (Tian-Shan Mountains).
Com. Geol. Materiaux Gener. App/., Livr. 74, 1-57.
— 1940. Carboniferous ammonoids of North Kazakhstan. Acad. Sci. USSR, Palaeontol. Inst., Palaeontology
of USSR, IV, Fasc. 1, 1-343.
lys, M. and serre, b. 1958. Contributions a la connaissance des microfaune du Paleozoique. Etudes
micropaleontologiques dans le Carbonifere marin des Asturies (Espagne). Revue Inst.fr. Petrole, 13, 879-916.
manger, w. l. 1979. Lower Carboniferous ammonoid assemblages from North America. Huiteme Congr.
Avanc Etud. Stratigr. carb., c.r. 3 (1975), 211-221.
marcos, a. 1967. Estudio geologica deo reborde NW de los Picos de Europa (region de Onis-Cabrales,
Cordillera Cantabrica). Trabajos de Geol. 1, 39-46.
marks, p. and wensink, h. 1970. Conodonts and the age of the ‘Griotte’ Limestone Formation in the Upper
Aragon Valley (Huesca, Spain). Proc. K. ned. Akad. Wet., Ser. B, 73, 238-275.
Matthews, s. c. 1969a. A Lower Carboniferous conodont fauna from East Cornwall. Palaeontology, 12,
262-275.
— 1969 b. Two conodont faunas from the Lower Carboniferous of Chudleigh, South Devon. Ibid.
276-280.
— sadler, p. m. and selwood, e. b. 1972. A Lower Carboniferous conodont fauna from Chillaton, south-west
Devonshire. Ibid. 15, 550-568.
— and thomas, J. m. 1974. Lower Carboniferous conodont faunas from north-east Devonshire. Ibid. 17,
371-385.
mehl, M. G. and thomas, L. A. 1947. Conodonts from the Fern Glen of Missouri. J. Sclent. Labs. Denison Univ.
40, 3-19.
meischner, d. 1970. Conodonten-Chronologie des Deutschen Karbons. Sixieme Congr. Avanc. Etud. Stratigr.
Carb. Sheffield 1967, III, 1169-1180.
Merrill, G. K. and Merrill, s. m. 1974. Pennsylvanian non-platform conodonts. Ha: The dimorphic apparatus
of Idioprioniodus. Geol. Palaeontol. 8, 119-130.
miller, a. K. and collinson, c. 1951. Lower Mississippian ammonoids of Missouri. J. Paleont. 25, No. 4,
454-487.
— and garner, H. F. 1955. Lower Mississippian Cephalopods of Michigan. Part III. Ammonoids and
Summary. Contr. Mus. Palaeont. Univ. Mich. XII, 8, 113-173.
moore, e. w. J. 1946. The Carboniferous goniatite genera Girtyoceras and Eumorphoceras. Proc. Yorks, geol.
Soc. 25, 387-445.
norby, R. d. 1975. Micromorphology of natural conodont apparatuses of Chesterian age (abs). Geol. Soc. Amer.
Abs. with programmes, 7, 829.
pareyn, c. 1961. Les Massifs Carboniferes du Sahara Sud-Oranais. Publ. Centre Recherches Sahariennes (ser.
Geol.), 1, tome II, Paleontologie, 1-244.
perrit, m. F. 1974. Biostratigraphie par Conodontes du Carbonifere inferieur des Pyrenees Bearnaises. C.r.
Acad. Sci. Paris, 279, 110, 791-794.
POPOV, a. v. 1968. Visean ammonoids of the northern Tien Shan and their stratigraphic significance. Akad
Nauk. Kirgisian SSR, Geol. Inst. 1-116.
— 1975. Ammonoidea. Palaeontological Atlas. Carboniferous Strata of the Urals. M-vo geologii SSSR.
Vsecoyuz nef. naytz-issled. geol-ravz. in-t. Trudi. 383, 111-130, Tab. 34-51.
prentice, j. E. and thomas, J. m. 1960. The Carboniferous goniatites of North Devon. Abstr. Proc. Third.
Conf. Geol. Geomorphol. SW England, 6-8.
rexroad, c. B. 1958. Conodonts from the Glen Dean Formation (Chester) of the Illinois Basin. Illinois Geol.
Surv. Rept. Invest. 209, 1-27.
and scott, a. j. 1964. Conodont zones of the Rockford Limestones and New Providence Shale
(Mississippian) in Indiana. Bull. Indiana Dep. Conserv. Geol. Surv. 30, 1-54.
350 PALAEONTOLOGY, VOLUME 25
roundy, p. v. 1926. The microfauna in Mississippian formations on San Saba County, Texas. U.S. Geol. Surv.
Prof. Paper , 146, 1-63.
Sandberg, c. A., ziegler, w., leuteritz, K. and brill, s. M. 1978. Phylogeny, speciation and zonation of
Siphonodella (conodonta. Upper Devonian and Lower Carboniferous). Newsl. Stratigr. 7 (2), 102-120.
schmidt, H. 1925. Die carbonischen Goniatiten Deutschlands. Jb.preuss. geol. Landesanst. 45 (1924), 489-609.
— 1931 . Das Palaozoikum der spanischen Pyrenaen. Abh. Ges. Wiss. Gottingen, Math.-physik., Kl. (3), 5, 85.
— and muller, k. j. 1964. Weitere Funde von Conodonten-Gruppen aus dem oberen Karbon des
Sauerlandes. Palaont. Z. 38, 105-135.
scott, h. w. 1942. Conodont assemblages from the Heath Formation, Montana. J. Paleont. 16, 293-300.
sitter, l. u. de. 1962. The structure of the southern slope of the Cantabrian Mountains: Explanation of a
geological map with sections, scale 1 : 100,000. Leid. geol. Meded. 26, 255-264.
Thompson, T. L. 1967. Conodont zonation of lower Osagean rocks (Lower Mississippian) of southwestern
Missouri. Missouri geol. Surv. Water Resour. Kept. Inv. 39, 1-84.
— and fellows, L. d. 1970. Stratigraphy and conodont biostratigraphy of Kinderhookian and Osagean rocks
of southwestern Missouri and adjacent areas. Ibid. Rept. Inv. 45, 1-263.
veen, j. van. 1965. The tectonic and stratigraphic history of the Cardano area, Cantabrian Mountains, north-
west Spain, Leid. geol. Meded. 35, 45-103.
voges, a. 1959. Conodonten aus dem Untercarbon I und II (Gattendorfia und Pericyclus-Stufe) des Sauerlandes.
Palaont. Z. 33, 4, 266-314.
von bitter, p. h. 1972. Environmental control of conodont distribution in the Shawnee Group (Upper
Pennsylvanian) of eastern Kansas. Paleont. Contr. Univ. Kansas, Art. 59, 1-105.
wagner, R. h. 1963. A general account of the Palaeozoic rocks between the Rivers Porma and Bernesga (Leon,
NW Spain). Bol. Inst. geol. min. Espaiia, LXXIV, 171-331.
— and wagner-gentis, c. h. t. 1963. Summary of the stratigraphy of Upper Palaeozoic Rocks in NE
Palencia, Spain. Proc. Kon. ned. Akad. Wet., Series B, 66, 149-163.
— winkler prins, c. F. and riding, r. e. 1971 . Lithostratigraphic units of the lower part of the Carboniferous
in northern Leon, Spain. Trabajos de Geol. 4, 603-633.
wagner-gentis, c. h. t. 1960. On Nautellipsites hispanicus (Foord and Crick) Estudios geol. Inst. Invest, geol.
Lucas Mallada, XVI, 43-51.
wirth, M. 1967. Gliederung des hoheren Palaozoikums (Givet-Namur) im Gebiet des Quinto Real (West-
pyrenaen) mit hilfe von Conodonten. Abh. Neues Jahrb. Geol., Pal. 127, 179-240.
ziegler, w. 1962. Taxonomie und Phylogenie Oberdevonischer Conodonten und ihre stratigraphische
Bedeutung. Abh. Hess. Landesamt. Bodenf. 38, 1-166.
— 1969. Eine neue Conodontenfauna aus dem hochsten Oberdevon. Fortschr. Geol. Rheinld. u. Westf. 17,
343-360.
— and leuteritz, k. 1970. Palaontologischer Anhang: Conodonten. In koch, m., Leuteritz, k. and ziegler,
w., Fortschr. Geol. Rheinld. u. Westf. 17, 712-715.
A. C. HIGGINS
Department of Geology
University of Sheffield
Sheffield SI 3JD
C. H. T. WAGNER-GENTIS
Mayfield
Cross Lane
Typescript received 19 May 1980 Calver
Revised typescript received 23 October 1980 Sheffield S30 1XS
LIASSIC PLESIOSAUR EMBRYOS
REINTERPRETED AS SHRIMP BURROWS
by RICHARD A. THULBORN
Abstract. A peculiar nodule from the Upper Liassic (Toarcian) shales of Whitby, Yorkshire, is interpreted
as an infilled burrow system which was probably excavated by a shrimp-like crustacean (possibly Glyphea
sp.). This interpretation is supported by comparisons with fossil crustacean burrows in the ichnogenus
Thalassinoides. The nodule had formerly been regarded as a cluster of fossil embryos from the aquatic reptile
Plesiosaurus.
In September 1887 H. G. Seeley delivered four reports on fossil reptiles to the British Association
meeting at Manchester. Three of those reports dealt with anatomy and systematics, but the fourth
concerned the more unusual subject of fossil embryos. Seeley described and exhibited a peculiar
nodule showing supposed embryos of a Jurassic plesiosaur— an aquatic reptile of the suborder
Plesiosauria (order Sauropterygia). His account of the embryos was summarized the following year
(Seeley 1888), though a full description did not appear until 1896.
Seeley’s plesiosaur embryos have attracted little attention: they have never been figured, and it is
difficult to find more than passing mention of them in the literature of vertebrate palaeontology. They
were briefly noticed by Woodward (1898, p. 161), Williston (1914, p. 94), and de Saint-Seine (1955,
p. 422), but seem never to have been re-examined in detail. Abel’s classic work Vorzeitliche
Lebensspuren (1935) devoted much attention to embryos and neonates of ichthyosaurs (reptile order
Ichthyosauria), but made no mention at all of plesiosaur embryos. There has often been speculation
about the mode of reproduction in plesiosaurs (see, for example, Robinson 1975), but the embryos
described by Seeley seem to have been overlooked.
This paper provides the first illustrations of the supposed plesiosaur embryos, and offers a
reinterpretation of these curious fossils.
MATERIAL
The material described by Seeley was acquired in 1909 by the British Museum (Natural History), where it is
now catalogued as R 3585. It comprises: an irregular nodule of pyritic mudstone and shale, approxi-
mately 11 x 8x8 cm (text-fig. 1); a small fragment, broken from the main nodule; five slides, each with a
thin section; a slide with twenty-three stained serial sections from the neck region of a modern lizard
embryo. The small fragment was detached in Seeley’s search for internal structure, and it provided material for
the thin sections; the serial sections of the lizard embryo had been obtained for comparative purposes (Seeley
1896, p. 21).
The nodule almost certainly came from coastal exposures of the Upper Lias (Whitbian sub-stage of the
Toarcian) near Whitby, Yorkshire. Seeley received the nodule in 1887 from J. F. Walker (then curator of the
Yorkshire Natural History Society), who had obtained it from a dealer in Whitby. The Whitby fossil-dealers
were not averse to importing their wares from the Lower Liassic of Lyme Regis, Dorset, but probably did so only
in the case of exceptionally fine and valuable specimens (notably fossil fishes; see Blake 1876, pp. 257-259). The
Liassic shales near Whitby are rife with oddly shaped concretions and nodules (Hallam 1962; Howarth 1962),
and it seems unlikely that a dealer would have imported an example from Dorset when so many were to hand
locally.
[Palaeontology, Vol. 25, Part 2, 1982, pp. 351 359. |
352
PALAEONTOLOGY, VOLUME 25
The Whitby Lias is about 76 m thick, and may be summarized as follows (incorporating data from Hallam
1962; Howarth 1962, 1973; Hemingway 1974):
Formation Thickness (m) Ammonite Zone
Stage
Dogger (sandstones and ironstones)
- unconformity
Cement Shales
5-5]
Main Alum Shales
22-0
Hard Shales
4-5 J
Bituminous Shales
23-5 1
Jet Rock Shales
8-5 )
Grey Shales
13-5}
Hildoceras bifrons
Harpoceras falciferum
Dactylioceras tenuicostatum
Bajocian
Toarcian
The succession consists of shales, with minor argillaceous limestones and rare seams of siderite mudstone.
Nodules and concretions are extremely abundant; some are scattered at random through the shales, while others
occur in constant bands which serve as useful marker horizons (see Howarth 1962). Concretions in these marker
horizons often have highly distinctive shapes, and have been named accordingly— ‘cannon balls’, ‘cheese
doggers’, ‘curling stones’, ‘pseudovertebrae’, and so on. Unfortunately the nodule described by Seeley cannot be
referred with certainty to any of these marker horizons. Seeley’s specimen has a matrix of soft, grey, flaky, and
non-bituminous shale, and is unlikely, for that reason, to have been obtained from the zone of Harpoceras
falciferum— where, the shales are usually brown in colour and often have a characteristic ‘oily’ smell. Nor is it
very likely that Seeley’s specimen came from the Hard Shales or the Main Alum Shales: in many places these
shales weather to a brown or reddish colour, and they sometimes produce efflorescent alum. The specimen
probably originated from the Grey Shales or the Cement Shales.
DESCRIPTION
The supposed plesiosaur embryos are rounded masses of grey-brown mudstone protruding from a core of flaky
grey shale (text-fig. 1). The shale is very soft, and contains tiny blebs and veins of white calcite, together with
occasional flecks of black plant material. The bodies of the embryos are slightly harder than the shale matrix,
extremely fine-grained, and rich in finely divided pyrite. Seeley originally believed the embryos to be phosphatic
(1888), but later determined that they were in fact pyritic (1896). The embryos, and some areas of intervening
shale, have a greasy lustre which has probably been enhanced, if not produced, by repeated handling of the
specimen.
Seeley identified and numbered four principal embryos, with ‘indications of three or four others’ (1896, p. 20).
His numbering is still visible on the specimen, in faded red ink, and will be followed here (see text-fig. 2). For the
sake of brevity I will also adhere to Seeley’s descriptive terminology (e.g. the ‘head’ of embryo 1, the ‘limbs’ of
embryo 3, and so on); and to assist my discussion the specimen will be oriented as a spheroid with the ‘head’ of
embryo 1 directed to the ‘North Pole’.
Seeley’s description (1896) is both detailed and accurate and need not be repeated here. However, the specimen
has certain features that were not mentioned by Seeley. First, it is badly damaged in the region of the ‘North
Pole’, where it seems that several large flakes have been removed by hammer blows (text-figs. 1a, 2a). The
fragment detached to provide thin sections was clearly the last piece to be removed, for it can still be fitted on to
the main nodule, where it forms part of an older fracture-scar. In some places the specimen appears to have been
subjected to rather rough mechanical preparation; there are, for example, very deep needle-marks along the right
side of the ‘neck’ in embryo 1 .
The embryos are clustered in a partly overlapping arrangement; the ‘neck’ of embryo 3, for example, is largely
concealed by the ‘body’ of embryo 1 . Elsewhere one embryo may be joined to another without the slightest trace
of a dividing line (see text-fig. 1b). The arrangement of the supposed embryos is not entirely random: they extend
in a radiating pattern, along meridians, when viewed from the ‘South Pole’ (text-fig. 2b).
Seeley could find no definite trace of organic structure in the specimen, and his identification of ‘embryonic
plesiosaurs’ rested almost entirely on the criterion of shape. He considered that embryo 1 showed ‘the head,
neck, body, tail and limbs of such a shape as a plesiosaur would show’, and pointed out that the other embryos
were basically similar in form; differences in shape between one embryo and another were taken to reflect
‘various stages of development’ (Seeley 1896, p. 23). Nevertheless, Seeley did attempt to identify anatomical
THULBORN: ‘PLESIOSAUR EMBRYOS’
353
features such as bones and teeth within the embryos — though he did so with extreme circumspection. These
possible organic structures merit close inspection.
Placenta. A ‘smooth film of hard clay’ over parts of the specimen was considered by Seeley to have ‘much the
aspect of a defining membrane of a placenta-like character’ (1896, p. 20). This apparently refers to the greasy
lustre on the embryos and on some intervening areas of shale. The lustre is almost certainly artificial, for it is
easily reproduced by gently rubbing a finger-tip over a fresh patch of the shale.
Median dorsal ridge. ‘Each of the four principal specimens is characterized by a median longitudinal ridge or
blunt angle which extends down . . . what I regard as the dorsal surface; corresponding in position to the neural
spines of the vertebrae’ (Seeley 1896, p. 21). The ridge is visible in unweathered and undamaged portions of the
embryos, though it is sometimes very faint. Nowhere is there any trace of organic structure in, or beneath, the
ridge; and there is no evidence to confirm that the ridge is ‘dorsal’ in position.
Sclerotic ring. On the right side of the ‘head’ in embryo 1 is a shallow oval pit which was tentatively identified by
Seeley as the orbit. This pit originally contained ‘a scale with a radiated structure, which had the appearance of
being a sclerotic circle of bones about the eye’, but this was subsequently lost from the specimen (Seeley 1896,
pp. 22, 24). The supposed orbit is now floored by a paper-thin layer of finely crystalline pyrite, which is overlain
near the anterior margin by a patch of shale containing flecks of white calcite. The floor of the orbit appears to be
featureless — though Seeley maintained that there was possibly ‘some evidence of a radiating structure’ (1896,
p. 25). The ‘radiated scale’ described by Seeley is unlikely to have been a sclerotic ring, because it was originally
applied to the concave floor of the ‘orbit’: sclerotic plates are usually embedded in the lateral half of the eyeball,
and therefore tend to be arched outwards from the orbit (Walls 1942). If the ‘radiated scale’ were a series of
sclerotic plates it would be necessary to suppose that the entire eyeball had been squashed flat in the embryo; and if
the eyeball had been squashed or ruptured it would be reasonable to expect evidence of similar distortion in other
regions of the soft-bodied embryos. In any case, it seems improbable that delicate and superficial ossifications
would have been preserved, while deeper and presumably more robust bones were obliterated. I suspect that the
‘radiated scale’ was no more than a thin vein of calcite, resembling similar veins in other parts of the nodule.
Scapula. The ‘left forelimb’ of embryo 2 is truncated by a fracture, behind which ‘the dorsal aspect of the limb
appears to include a surface bone in the position of the ascending process of the plesiosaurian scapula’ (Seeley
1896, p. 23). I can find no trace of bone in this area; there is a slight surface irregularity postero-dorsal to the
‘limb’, but this is not particularly reminiscent of the outline of a scapula. Moreover, the ascending ramus of the
plesiosaurian scapula is situated in front of the glenoid cavity, and not behind it.
Muscle segments. According to Seeley, ‘there appear to be some faint indications of transverse segmentation like
that of muscles, in the region of the neck in the specimen No. 1 , and in the dorsal region in specimen No. 2’ ( 1 896,
p. 23). There seems to be no trace whatsoever of segmental structure in embryo 2. Much of the ‘neck’ of embryo 1
has an irregular, flaky, and pitted surface. In its posterior third the right side of the neck carries a series of deep
needle-marks which extend into the adjoining shale; it is improbable that these would be taken as evidence of
segmentation, even at a casual glance. Above this line of needle-marks is a smooth area which proves, on close
inspection ( x 50), to carry an ornament of extremely fine, straight, and parallel scratches. These microscopic
scratches were most probably produced with a mild abrasive: they are so regular in depth, spacing, and orienta-
tion (antero-dorsal to postero-ventral) that they cannot have been produced individually, and they are so
unvaryingly straight and parallel that they do not seem to be a natural feature. Over all, there is no clear evidence
of segmentation in any of the embryos.
Mandible and teeth. ‘It is possible . . . that the lower jaw may be indicated in the film of clay, which is imperfectly
preserved beneath the head [of embryo 1], and that some small badly preserved white spots arranged in linear
succession are indications of teeth’ (Seeley 1896, p. 24). The ‘head’ of embryo 1 meets the shale matrix without
any indication of a separate mandible. Alongside the right surface of the ‘head’ is a string of tiny white spots,
each about 0T5 mm in diameter; these are blebs of calcite similar to those scattered throughout the specimen,
and their roughly linear arrangement seems to be fortuitous.
Skull roofing bones, external nostrils, and parietal foramen. All these features were tentatively identified by Seeley
on the ‘head’ of embryo 1 . The surface of the ‘head’ is marked with a random assortment of pits, furrows, and
scratches, all of which seem to be preservational and weathering effects comparable to those in other parts of the
embryos. None of the grooves or scratches can justifiably be regarded as a mid-line suture between skull roofing
bones, and none of the pits can be matched up into a pair that could convincingly represent the external nostrils.
354
PALAEONTOLOGY, VOLUME 25
text-fig. 1. Four views of nodule from the Upper Liassic of Whitby, Yorkshire, showing supposed plesiosaur
embryos. BM(NH) R.3585; scale indicates 2 cm.
Other anatomical features of the embryos (heads, necks, bodies, tails, and limbs) were identified by Seeley
purely on the basis of their shape and their position. In summary, there seems to be no evidence of organic
structure in any of the supposed embryos.
DISCUSSION
There are several reasons why the specimen is unlikely to be a cluster of plesiosaur embryos. First, it
lacks any definite organic structure. Some of the anatomical features identified by Seeley (1896) are
no more than surface irregularities, scratches, pits, and similar weathering effects (e.g. ‘scapula’,
‘nostrils’, and ‘skull bones’); other such features are patches of calcite (e.g. ‘teeth’) or artifacts (e.g.
‘placenta’). Second, the supposed embryos show differences in shape and proportions. Seeley
accounted for this variation by suggesting that the embryos were in different stages of development,
but this is not a particularly convincing explanation. Where embryos, neonates, or juveniles of
reptiles are preserved as natural groups of fossils the siblings are invariably in the same state of
development— as, for example, in ichthyosaurs (Hauff 1921; Hoffmann 1958; Urlichs, Wild, and
Ziegler 1979), in nothosaurs (Halstead 1969), and in ornithopod dinosaurs (Horner and Makela
1979). Next, and most important, it is highly improbable that soft-bodied embryos could have been
preserved ‘in the round’ in a shale matrix. Sediment compaction usually ensures that soft organisms
are preserved in a partly or completely flattened state, yet the supposed plesiosaur embryos form a
THULBORN: PLESIOSAUR EMBRYOS’
355
text-fig. 2. Diagrammatic interpretation of text-fig. 1. Shale matrix unshaded, and supposed plesiosaur
embryos (here regarded as sediment-filled crustacean burrows) indicated by solid shading. Fracture surfaces
indicated by oblique shading. Supposed embryos (= burrows) are numbered following Seeley (1896), and each
has its ‘dorsal ridge’ (= groove in burrow floor) shown schematically. Abbreviations: c— cut surface; f— area
from which a fragment was detached to provide thin sections for Seeley (1896); h ‘head’ of supposed embryo
(=part of shaft leading to burrow chamber); m— needle-marks (shown schematically) along the side of
supposed embryo 1 (= burrow 1); p— spots indicating location of possible faecal pellets; t— ‘tail’ of supposed
embryo (= burrow termination). Scale indicates 2 cm.
spheroidal mass which is not obviously flattened or distorted. If this specimen were a group of
embryos it would be necessary to suppose that it had been enclosed in a syngenetic or diagenetic
concretion (following terminology of Pantin 1958), or that the embryos had been mineralized before
compaction of the surrounding sediment. The specimen was clearly not enclosed in a concretion, for
the supposed embryos are separated and overlain by soft shale; and it seems unlikely that embryos
could have been mineralized in utero or in an unconsolidated shale matrix. In rare circumstances
human embryos are known to undergo spontaneous abortion and calcification— to become ‘stone
babies’ or ‘lithopedia’ (Halstead and Middleton 1972)— but there are no reports of comparable
abnormalities among fossil vertebrates. It seems a very remote possibility that the supposed
plesiosaur embryos could be reptilian equivalents of ‘lithopedia’.
What, then, is the most probable identity of the specimen? It is quite unlike a concretion, or even an
aggregate of small concretions. Some parts of the supposed embryos bear a resemblance to
coprolites, but the specimen as a whole seems too complicated in structure to be a mass of coprolites
356
PALAEONTOLOGY, VOLUME 25
or of intestinal fillings. Each of the supposed embryos is a multi-lobed object (with several lobes
representing the ‘tail’ and the ‘limbs’), whereas individual coprolites are normally sausage-shaped,
spiral, fusiform, or pellet-like objects which are not developed into lobes (see Amstutz 1958, and
many references therein). Nor does the specimen resemble any known endocranial cast of a verte-
brate (Edinger 1929; Jerison 1973).
From a review of literature dealing with trace fossils and problematica I strongly suspect that the
supposed plesiosaur embryos are actually the sediment-filled burrows of a crustacean— most
probably a shrimp resembling those of the decapod superfamily Thalassinoidea. The thalassinoid
shrimp Callianassa excavates distinctive burrow systems in modern marine sediments (see
Braithwaite and Talbot 1972), and comparable burrows have been reported as far back as the
Triassic (Fiege 1944; Ireland, Pollard, Steel, and Thompson 1977) and even the Upper Carboniferous
(Warme and Olson 1971; Chamberlain and Clark 1973). Fossil thalassinoid burrows are often
referred to the ichnogenus Thalassinoides (reviewed by Bromley and Frey 1974); they vary
considerably in size and architecture, according to the specific identity of the burrower (e.g. see
Braithwaite and Talbot 1972) and to the nature of the sediments (e.g. see Bromley 1967).
Nevertheless, it is possible to give a brief generalized description (based on Weimer and Hoyt 1964;
Glaessner 1969; and sources mentioned above). A thalassinoid burrow system comprises one or more
chambers (or tunnels) opening to the sea floor by a steeply inclined shaft which is wide enough to
permit free passage of the burrower. A second shaft, which is often narrower and distinctly tapered,
extends to the surface from another part of the chamber system. The shrimp occupies one of the
chambers and creates a current by movements of its swimmerets; water enters the wider shaft and is
expelled through the narrower shaft, often serving to flush out faecal pellets. The chamber occupied
by the shrimp is big enough for the animal to turn round completely in somersault fashion. The
various chambers are connected by constricted passages, and the shrimp may excavate several short
galleries from the sides of a chamber. In some cases the chambers occur at random, while in others
they may form a spiral or radial pattern. In soft sediments the shafts may be deep, and there may be
many chambers excavated at several horizons; in more resistant sediments the burrow system may
comprise little more than ramifying tunnels confined to a preferred horizon. When such burrow
systems are encountered as fossils they are commonly filled with minerals or sediments which differ
from the surrounding rock.
The supposed plesiosaur embryos are readily interpreted as thalassinoid burrow fillings (see text-
fig. 2). Their ovoid ‘bodies’ seem to represent a series of chambers arranged in a roughly radial
pattern, while their ‘necks’ may be identified as steeply inclined shafts connecting chambers at
different levels or leading towards the former sea floor. Some of the ‘necks’ have a distinctly tapered
tip and may be regarded as excurrent shafts. There is no trace of a broader incurrent shaft, but this
would most likely have been situated in the region of the badly damaged ‘North Pole’. The ‘tails’ are
probably burrow-ends, which are ‘usually somewhat conical, coming to a blunt point’ in thalassinoid
burrows from the English Chalk (Bromley 1967, p. 162). The ‘limbs’ of the supposed embryos may be
identified as short galleries developed from the walls of the main chambers.
Nearly all major features of the supposed embryos can be matched in thalassinoid burrows of
one sort or another. And several other facts would seem to confirm that Seeley’s specimen is a
thalassinoid burrow system. The shale matrix of the specimen contains a few tiny spheroidal objects
(about 0-5 mm in diameter) which may possibly be faecal pellets flushed out from the interior of the
burrow; several of these occur in the shale around embryo 4. Next, the invertebrate fauna of the
Whitby Lias includes a variety of small decapod crustaceans, and some of these ( Glyphea spp. and
Eryma sp.) are similar in size and configuration to modern burrowing shrimps. Blake listed seven
species of crustaceans from the Whitby Lias, and mentioned that ‘numerous crustacean claws occur
in many zones’ (1876, p. 429). Glyphea and Eryma are not close relatives of the living thalassinoid
shrimps, but ‘it must be remembered that similar burrows indicate possibly similar body shape and
behaviour, not systematic identity of the burrower’ (Glaessner 1969, p. 430). Bromley and Frey
(1974) listed many crustaceans know to construct burrows of Thalassinoides type, and among these
were Jurassic shrimps of the superfamily Glypheoidea (Sellwood 1971; Bromley and Asgaard 1972).
THULBORN: PLESIOSAUR EMBRYOS’
357
It is possible, then, that one or more of the crustaceans reported from the Whitby Lias (most probably
Glyphea spp.) may have constructed thalassinoid burrows. Remains of crustaceans are rarely found
inside thalassinoid burrows, probably because the animals left their burrows to moult (see Glaessner
1969, p. 434). Finally, burrows are abundant at some horizons in the Whitby Lias. The distribution
of these burrows was studied by Morris (1979), who did not specify the occurrence of Thalassinoides
but mentioned that the trace fossil assemblages were dominated by Chondrites. In other sedimentary
settings Chondrites may be associated with another trace fossil, Gyrolithes, which is sometimes linked
with Thalassinoides to form a single burrow system (Bromley and Frey 1974). This very indirect
evidence may hint at the existence of thalassinoid burrows in the Upper Lias at Whitby; such an
occurrence would not be surprising, since Thalassinoides is widespread in the British Lower Lias
(Sellwood 1970; Sellwood, Durkin, and Kennedy 1970). It is worth noting that Morris found
Chondrites restricted to a ‘normal’ shale facies — ‘a homogeneous bioturbated sediment often
containing sideritic nodules or horizons’ (1979, p. 117). Morris studied only the lower part of the
Toarcian succession at Whitby (as high as the Hard Shales), so that his ‘normal’ shales are in fact
represented by the Grey Shales. It has already been deduced, on other grounds, that the supposed
plesiosaur embryos came from the Grey Shales or from the Cement Shales (which were not
considered by Morris). In other words, it seems probable that Seeley’s specimen came from a
‘normal’ shale horizon in which burrows are abundant.
In summary, the nodule interpreted by Seeley (1896) as a group of plesiosaur embryos bears a
strong resemblance to a system of thalassinoid burrow fillings. Crustaceans known from the Whitby
Lias may well have included forms capable of excavating such burrows; and burrows (though not
specifically identified as thalassinoid) are abundant in the ‘normal’ shale facies from which the nodule
is likely to have come. The specimen has only one structural feature that I have been unable to match
in any other thalassinoid burrow: this is the distinct ‘dorsal ridge’ noted by Seeley. If the specimen is
interpreted as a burrow system this ‘dorsal ridge’ will actually represent a longitudinal groove in the
floor of each burrow chamber. It is reasonable to suppose that this groove was produced by the
appendages or the telson of the crustacean inhabitant.
Thalassinoid burrow fillings have commonly been mistaken for quite different objects; Bromley
reported (1967, p. 158) that thalassinoid burrows of Upper Cretaceous age have been interpreted as
benthonic algae, plant roots, sponges, solution channels, concretions, and trace fossils of unidentified
marine organisms. Seeley’s misinterpretation is all the more understandable when one considers that
his specimen had been ‘improved’ by artificial means. Parts of the supposed embryos seem to have
been smoothed off and polished with a mild abrasive, leaving a microscopic pattern of unnaturally
straight and parallel scratches. And in places it appears that the main burrow chambers (or embryo
‘bodies’) have been trimmed to a more suggestive shape by removal of some side-galleries (which
would otherwise have remained as rather puzzling supernumerary ‘limbs’). Such trimming is
apparent along the right side of embryo 1 , where there is a series of deep needle-marks. These marks
extend from the fossil into the shale matrix, whereas needle-marks produced in normal preparation
would extend in the opposite direction. Either this work with a needle was intended to remove
portions of the fossil, or it was an unbelievably clumsy attempt at preparing the specimen. Next, the
suitably constricted appearance of the ‘neck’ in embryo 2 (text-fig. 2b) is due partly to the fact that it
has been cut down with a knife-blade or similar instrument. The cut surface is easily visible; it is flat
and sharp-edged, in contrast to the irregular and broadly rounded surfaces elsewhere. Finally there is
evidence of a much coarser, but none the less effective, ‘improvement’: several hammer blows in the
region of the ‘North Pole’ removed any trace that might have existed of an incurrent shaft leading to
the burrow system. It is not altogether surprising to discover these alterations to Seeley’s specimen,
for many such ‘improved’ or ‘repaired’ fossils are known to have come from the dealers at Whitby
(see, for example, Blake 1876, p. 428; Dance 1976, p. 103).
Acknowledgements. I thank Dr. Angela Milner and Dr. Alan Charig for providing help and facilities at the
British Museum (Natural History). Text-fig. 1 was prepared by Mandy Cilento (Zoology Department,
University of Queensland), and Susan Turner (Queensland Museum) helped me in searching the literature.
358
PALAEONTOLOGY, VOLUME 25
REFERENCES
abel, o. 1935. Vorzeitliche Lebensspuren, Gustav Fischer, Jena. 644 pp.
amstutz, G. c. 1958. Coprolites: a review of the literature and a study of specimens from southern Washington.
J. sediment, petrol. 28, 498-508.
blake, j. f. 1876. Palaeontology. In tate, r. and blake, j. f. The Yorkshire Lias. Van Voorst, London, pp. 243-
475.
braithwaite, c. j. r. and talbot, m. r. 1972. Crustacean burrows in the Seychelles, Indian Ocean. Palaeogeogr.,
Palaeoclimatol., Palaeoecol. 11, 265-285.
bromley, r. G. 1967. Some observations on burrows of thalassinidean Crustacea in chalk hardgrounds. Q. Jl
geol. Soc. Lond. 123, 157-182.
— and asgaard, u. 1972. Notes on Greenland trace fossils. II. The burrows and microcoprolites of Glyphea
rosenkrantzi, a Lower Jurassic palinuran crustacean from Jameson Land, East Greenland. Rept geol. Surv.
Greenland , 49, 15-21.
— and frey, r. w. 1974. Redescription of the trace fossil Gyrolithes and taxonomic evaluation of
Thalassinoides, Ophiomorpha and Spongeliomorpha. Bull. geol. Soc. Denmark, 23, 311-335.
chamberlain, c. k. and clark, d. l. 1973. Trace fossils and conodonts as evidence for deep-water deposits in the
Oquirrh Basin of central Utah. J. Paleont. 47, 663-682.
dance, p. 1976. Animal fakes and frauds. Sampson Low, Maidenhead. 128 pp.
edinger, t. 1929. Die fossilen Gehirne. Ergebn. d. Anat. 28, 1-249.
fiege, k. 1944. Lebensspuren aus dem Muschelkalk Nordwestdeutschlands. N. Jb. Miner. Geol. Palaont. Abh.
B88, 401-426.
GLAESSNER, M. F. 1969. Decapoda. In moore, r. c. (ed.), Treatise on Invertebrate Paleontology, Part R,
Arthropoda ( Crustacea exclusive of Ostracoda, Myriapoda, Hexapoda ). Geol. Soc. Amer. and Kansas Univ.,
pp. 400-533.
hallam, a. 1962. A band of extraordinary calcareous concretions in the Upper Lias of Yorkshire, England.
J. sediment, petrol. 32, 840-847.
halstead, L. b. 1969. The pattern of vertebrate evolution. Oliver and Boyd, Edinburgh. 209 pp.
— and Middleton, j. 1972. Bare bones, an exploration in art and science. Oliver and Boyd, Edinburgh. 1 19 pp.
hauff, b. 1921. Untersuchung der Fossilfundstatten von Holzmaden im Posidonienschiefer des oberen Lias
Wiirttemburgs. Palaeontographica, 64, 1 -42.
hemingway, j. e. 1974. Jurassic. In rayner, d. h. and Hemingway, j. e., The geology and mineral resources of
Yorkshire. Yorks, geol. Soc., Leeds, pp. 161-223.
hoffmann, j. 1958. Einbettung und Zerfall der Ichthyosaurier im Lias von Holzmaden. Meyniana, 6, 10-55.
horner, j. r. and makela, r. 1979. Nest of juveniles provides evidence of family structure among dinosaurs.
Nature , Lond. 282, 296-298.
howarth, m. k. 1962. The Jet Rock Series and Alum Shale Series of the Yorkshire Coast. Proc. Yorks, geol. Soc.
33,381-422.
1973. The stratigraphy and ammonite fauna of the Upper Liassic Grey Shales of the Yorkshire coast. Bull.
Brit. Mus. ( nat . Hist.), Geol. 24, 238-277.
Ireland, r. j., pollard, j. e., steel, R. J. and Thompson, d. b. 1977. Intertidal sediments and trace fossils from the
Waterstones (Scythian -Anisian?) at Daresbury, Cheshire. Proc. Yorks, geol. Soc. 41, 399-436.
jerison, h. j. 1973. Evolution of the brain and intelligence. Academic Press, New York and London.
morris, K. a. 1979. A classification of Jurassic marine shale sequences: an example from the Toarcian (Lower
Jurassic) of Great Britain. Palaeogeogr., Palaeoclimatol., Palaeoecol. 26, 117-126.
pantin, H. m. 1958. Rate of formation of a diagenetic calcareous concretion. J. sediment, petrol. 28, 366-371 .
robinson, J. a. 1975. The locomotion of plesiosaurs. N. Jb. Geol. Palaont. Abh. 149, 286-332.
saint-seine, p. de. 1955. Sauropterygia. In piveteau, j. (ed.), Traite de Paleontologie, V ( Amphibiens , Reptiles,
Oiseaux). Masson, Paris, pp. 420-458.
seeley, h. g. 1888. On the mode of development of the young in Plesiosaurus. Rept. Brit. Assn Adv. Sci.,
Manchester 1887, Section C, 697-698. [Abstract; a second, edited, version is in Geol. Mag. (3), 4, 562-
563.]
— 1 896. On a pyritious concretion from the Lias of Whitby, which appears to show the external form of the
body of embryos of a species of Plesiosaurus. Ann. Rept Yorks, philos. Soc. 1895, 20-29.
sellwood, b. w. 1970. The relation of trace fossils to small scale sedimentary cycles in the British Lias. In crimes,
t. p. and harper, j. c. (eds.). Trace Fossils (Geol. Soc. London, Special Issues, No. 3). Seel House Press,
Liverpool, pp. 489-504.
THULBORN: PLESIOSAUR EMBRYOS’
359
sellwood, B. w. 1971. A Thalassinoides burrow containing the crustacean Glyphea udressieri (Meyer) from the
Bathonian of Oxfordshire. Palaeontology , 14, 589-591.
— durkin, m. K. and Kennedy, w. j. 1970. Field meeting on the Jurassic and Cretaceous rocks of Wessex:
report by the directors. Proc. geol. Assn ., 81, 715-732.
urlichs, m., wild, r. and ziegler, b. 1979. Fossilien aus Holzmaden. Stuttgarter Beit, zur Naturk. C 11, 1-34.
walls, G. l. 1942. The vertebrate eye and its adaptive radiation. Bull. Cranbrook Inst. Sci. 19, 1-785.
warme, J. E. and olson, r. w. 1 97 1 . Lake Brown wood spillway. In perkins, b. (ed.), Trace fossils: a field guide.
Louisiana State Univ. School Geosci., misc. Publns No. 71. pp. 27-43.
weimer, r. j. and hoyt, j. h. 1964. Burrows of Callianassa major Say, geologic indicators of littoral and shallow
neritic environments. J. Paleont. 38, 761-767.
williston, s. w. 1914. Water reptiles of the past and present. Chicago Univ. Press, Chicago. 251 pp.
woodward, A. s. 1898. Outlines of vertebrate palaeontology for students of zoology. Cambridge Univ. Press,
Cambridge. 470 pp.
RICHARD A. THULBORN
Original typescript received 7 January 1981
Revised typescript received 14 April 1981
Department of Zoology
University of Queensland
St. Lucia
Queensland 4067
Australia
LIMPET GRAZING ON CRETACEOUS
ALGAL-BORED AMMONITES
by ETIE BEN AKPAN, GEORGE E. FARROW, and NOEL MORRIS
Abstract. Phosphatized internal moulds of Anahoplites sp. and Euhoplites sp. reveal sets of six-toothed radula
marks closely comparable to those scratched on to Recent molluscs by Acmaea ( Tectura ) virginea in water
depths of up to 25 m on the muddy inshore shelf of western Scotland. Both the Recent and Cretaceous acmaeid
marks truncate algal borings. Depth of the Gault sea during the Spathi subzone (M. Albian) was probably
between 8 and 30 m, i.e. within the euphotic zone. Subsequent to lithification and phosphatization the moulds
were bored by phoronids and scraped by regular echinoids.
Farrow and Clokie (1979) have recently drawn attention to the close association between the
Recent limpet Acmaea ( Tectura ) virginea (Muller 1776) and the shell-boring alga ‘conchocelis’ on
which it feeds. ‘Conchocelis’ is the resting phase of red seaweeds of the family Bangiaceae: use of the
term is as in Farrow and Clokie ( 1 979). In this paper we describe a similar association from the Lower
Cretaceous which gives an indication of the depth of deposition of part of the Gault Clay and
additionally demonstrates the persistence of acamaeid radula design. The fossil material consists of
both shells and ammonite steinkerns bearing grazing traces: nearly all come from the Spathi Subzone
of the Middle Albian, only one specimen from Dunton Green being of uncertain horizon. The
grazing traces were collected from Copt Point, Folkestone, and are found on ten out of fifty-eight
phosphatic internal moulds of hoplitid ammonites.
We outline the evidence for our interpretation of these traces, and document their occurrence
alongside the only known species of Acmaeidae from the British Gault.
CHARACTERISTICS OF THE ASSOCIATED FAUNA
The form of the grazing traces from the Gault with their six parallel incised grooves preserved as
negatives has led us to conclude that they probably belonged to an acmaeid. Patelliform shells are
uncommon in the north-west European Albian, and particularly so in the Gault facies. There is one
exception, however, Patella tenuistriata Michelin, 1838. This species is recorded from Folkestone by
Price (1879) from his beds 2 II, 2 V, 2 VI, 2 VII, and 7 VIII, that is, scattered through the Middle
Albian. We have examined nine specimens from the Gault at Folkestone with no further locality
information. Of these, two are preserved on ammonite or nautiloid shells. Their protoconchs appear
to have been lost during life and no epifauna or epiphyte is apparent. Further specimens of particular
interest are two specimens from the Hampden Park borehole, Sussex, from the Spathi Subzone, the
same subzone as our grazing trails. Another specimen, from an uncertain horizon of the Gault at
Dunton Green, Kent (IGS Z5013), has a well-preserved muscle scar which confirms their
identification as Acmaeidae. The only other limpet-like forms from the British Gault are attributed to
the Calyptraeacea. Since these are suspension feeders, they are unlikely to have produced the incised
traces we describe.
The clays of the Spathi Subzone occur widely across south-east England and have an extremely
restricted benthonic fauna preserved. We are indebted to Adrian Morter of the I.G.S. for information
concerning the faunas of this horizon, put together from information from several boreholes in East
Anglia. Morter’s Bed 3 contains common Birostrina concentrica and 'O steed cf. incurva, and less
commonly the brachiopod Moutonithyris, the gastropods Anticonulus and Rissoina, the bivalves
Entolium, Inoceramus, Pycnodonte, Ludbrookia , Neithea, Pectinucula , and the echinoid Hemiaster.
(Palaeontology, Vol. 25, Part 2, pp. 361-367.]
362
PALAEONTOLOGY, VOLUME 25
Morter’s Bed 2, a grey clay bioturbated with Chondrites and other traces, contains the bivalves
Birostrina, Inoceramus, Mesosacella, Pectinucula, Pseudolimea, Pycnodonte , Rastellum, the
brachiopods Kingena , Moutonithyris, and Tamarella, and the solitary coral Cyclocyathus. These
faunas are typical of the Spathi Subzone benthos. One or two additions are known from Folkestone,
including Oistotrigonia fittoni (Deshayes), which seems to be restricted to the clay facies.
We suggest that the fauna was limited by the fine-grained sediments. These were, however, not
always oxygen-deficient, as nuculoid infaunal deposit feeders form a significant part of the fauna.
Shells of nektonic cephalopods are present in sufficient numbers to have provided a substrate for
epifaunal grazers and this would have been the case for the acmaeids. The trochaceans and
rissoaceans may also have browsed on shell fragments, or storm-derived algae or sponges.
We feel that the coincidence of the occurrence of Acmaea-Mke grazing traces and the records of the
only known species of Acmaeidae from the British Middle Albian is sufficient to justify our
interpretation as one being the product of the other.
SYSTEMATIC DESCRIPTION
Acmaea tenuistriata (Michelin, 1838)
Material. Shells: nine specimens from the Gault of Folkestone (M.-U. Albian) BM(NH) GG. 19971 -19978; two
specimens from the Spathi Subzone, M. Albian, Hampden Park borehole, Sussex, IGS, BDL 6727 and 6731;
one specimen from an unspecified nodule bed in the Gault, Dunton Green, Kent, IGS Z5013.
Description. Broad, oval, bilaterally symmetrical, limpet-like shells with the apex approximately half-way
between the mid-point and the anterior margin. The height is approximately one-third of the length. The largest
specimen is about 15 mm in length. There is a radiating ornament of fine ribs with wide inter-spaces. No
protoconchs are preserved. The shell structure is partly preserved in two specimens, BM(NH) GG. 19977-19978
(text-fig. 1 d), and consists of concentrically arranged cross lamellae which appear like vertical prisms in radial
section. The shell is preserved entirely in a pinkish form, typical of Gault shells originally formed entirely of
aragonite. The body and foot attachment scar is well preserved in IGS Z5013 and is illustrated in text-fig. la-c. It
consists of a broad horseshoe-shaped area of scar joined at its anterior by a thinner line.
text-fig. 1 . Acmaea tenuistriata (Michelin). a-c steinkern with muscle attachment scar inked in:
IGS Z5013, unspecified nodule bed, Gault, Dunton Green, x4. ( a ) view of left side, ( b ) dorsal,
(c) view of anterior. ( d ) specimen with part of the shell preserved: BM(NH) GG. 19977, Middle-
Upper Albian, Copt Point, Folkestone, x 3, dorsal view.
AKDAN ET AL.: LIMPET GRAZING
363
Discussion. The shape of this shell and the form of its muscle scar are characteristic of the Family
Acmaeidae. The general shape of this species, with its apex placed anterior to the mid-point, and its
fine radial striae, are reminiscent of the subgenus Acmaea ( Tectura ) Gray, type species A. virginea
(Muller), quoted as Patella parva da Costa (jun. sub. syn.) rather than Acmaea s.s., type species
A. mitra (von Eschscholtz), a Californian species with a taller shell and central apex. However,
without a thorough revision of the fossil members of the family we feel unable to place this species
in a particular subgenus.
Grazing traces
Seven specimens are in the BM(NH); four collected by EBA and GEF, GG 19979-19982, and three
collected by H. Lister, GG 19983-19985. The grazing traces are preserved as exquisitely detailed
protruberances on the surfaces of the phosphatic moulds (PI. 37, figs. 2-4). Individual markings, up
to 200 ju.m wide by 400 jum long when fully developed, consist of six relatively broad ridges separated
by narrow furrows. Overlapping of adjacent marks, however, rarely permits the full complement of
ridges to be appreciated, and sets of sub-parallel ridges are the norm, their orientation being more
consistent on some specimens (PI. 37, fig. 4) than on others (PI. 37, fig. 3). Successive truncation of sets
of ridges suggests at least three phases of trace activity, the lowest being poorly preserved and having
a smoothed appearance.
Borings
Three types seem to be present. The first consists of a widespread fine granulation on the surface of
moulds where radula traces occur, in the form of small upstanding mounds of circular cross-section
(PI. 37, fig. 4). These represent the infilling of borings into the original ammonite shell. Having an
approximate diameter of 6 fxm they are probably algal in origin.
The second and third types represent unfilled borings into the lithified phosphatic moulds. The
former are oval in cross-section, 500 jum in diameter, and penetrate deeply into the moulds: the latter
are tunnels very close to the surface of the moulds which clearly truncate grazing traces in places
(PI. 37, fig. 3). The tunnels measure about 100 in depth and between 180 and 300 /urn in maximum
diameter, and are visible for a few millimetres before disappearing, only to reappear. These are
probably the work of phoronids (Bromley 1970).
COMPARISON WITH RECENT TRACES
The Gault traces are similar to Recent A. ( Tectura ) virginea radula marks incised into mollusc shells
(PI. 37, fig. 1) but are preserved in a negative form. The comparison extends from over-all similarity of
shape to details of six broadly incised elements, demonstrating a closer affinity in radula design
between the presumed Gault acmaeid and the Recent A. (77) virginea than between living A. virginea
and A. tessulata (Muller 1776), whose radula is of four-pronged construction.
However, the Gault grazing traces are wider and less consistent in orientation than are those of the
Recent A. ( T .) virginea. This size difference is a reflection of the sizes of the causative limpets. For
example, the longest Gault acmaeid measures 15 mm whereas the Recent British A. ( T .) virginea is
scarcely more than 7 mm long. Variation in the size of Recent Acmaea dictates the width of the
grazing traces produced, ‘scoops’ ranging from 80 to 100 /urn wide.
DEPTH OF DEPOSITION OF THE GAULT CLAY
The similarity between the fossil and Recent grazing traces is the more remarkable on account of the
further association between the grazing marks and the fine granulation seen on the Gault
specimens — the negative of the algal borings seen pitting the Recent examples. It is this association
that sheds light on the depositional environment of the Gault during the Spathi Subzone, and in order
to assess the probable depth of water it is necessary to consider something of the ecology of living
Acmaea.
364
PALAEONTOLOGY, VOLUME 25
The genus Acmaea occurs widely in intertidal and shallow water-shelf habitats and there is much
evidence for an algal grazing habit. Craig (1968) has described A. pelta living on micro- and macro-
algae. Black (1976) described A. incessa living exclusively on the kelp Egregia, and many intertidal
species graze upon epilithic algae (Kozloff 1973; Carefoot 1976). Fretter and Graham (1976)
recorded A. virginea on algal-coated shell debris off the British coast. While these observations
suggest that the depth was within the photic zone, it is possible to make a more refined estimate by
looking in more detail at a single area.
Living Acmaea virginea on the Scottish continental shelf
Farrow and Clokie (1979, Table 1) working in the Firth of Clyde, found A. virginea down to only
1 8 m, coincident with the local limit of the euphotic zone as measured biologically by the last record of
‘conchocelis’ (Clokie, Boney, and Farrow 1979). This correlation is well established regionally,
Acmaea extending into deeper water where substrates are coarser, environmental energy higher, and
water clearer, as off the Orkney Islands where the euphotic limit is 38 m (Farrow, Allen, Akpan, and
Brown 1 98 1 ). A muddier, Firth-of-Clyde or Firth-of-Lorne-type environment is more appropriate as
a Gault analogy, although the present latitude of 56° N is further north than that of Folkestone in the
Albian, which was more like 33° N (Smith, Briden, and Drewry 1973, fig. 7, map 2).
Evidence on A. virginea distribution in one area of the Firth of Clyde is summarized on text-fig. 2. It
is clear that the limpet extends down to within a few metres of the euphotic limit. Since, however,
Acmaea grazing is intensive only where algal infestation is heavy we may conclude that deposition of
the Gault Clay during this part of the Spathi Subzone was well within the euphotic zone. Depth was
unlikely to have been greater than 30 m and could even have been as shallow as 8 m.
TAPHONOMY
The evidence is clear that these Gault ammonites were infested with shell-boring algae to such an
extent that the full thickness of shell must have been riddled with them, for the limpet grazing marks
were made on the inside of the body chambers: any on the outside would have had negligible
fossilization potential. It remains to be established, however, whether the algae infested dead shells
drifting near the ocean surface, maybe for several years (House 1973, p. 309; though cf. Reyment
1958), or whether they bored into the shells when the latter were on the sea bed. Since the
environmental interpretation for this part of the Gault hinges on this point, further consideration is
necessary. There is some supporting circumstantial evidence.
First, there is the occurrence of probable regular echinoid scratches both on the phosphate moulds
and on associated shell debris: this scratching is commonest today within the euphotic zone.
However, the occurrence of these traces on the ammonite moulds post-dates not only the grazing
marks but also lithification of the moulds, and could be ruled misleading. Second, it is to be doubted
whether the ‘conchocelis’ phase of members of the family Bangiaceae (coastal red algae) would be
sufficiently dispersed to infest a multiplicity of widely scattered floating cephalopods. It seems easier
EXPLANATION OF PLATE 37
Fig. 1. Grazing traces of Recent Acmaea ( Tectura ) virginea on algal bored Dosinia shell: 9 m: Isle of Bute, Firth
of Clyde, Scotland. For locality see text-fig. 2.
Fig. 2. Grazing traces of Lower Cretaceous acmaeid on surface of hoplitid ammonite skeinkern: note the grazing
front (cf. fig. 1): BM(NH) GG.19983: Spathi Subzone, Middle Albian, Copt Point, Folkestone.
Fig. 3. Phoronid boring cutting through acmaeid grazing traces on surface of hoplitid ammonite steinkern:
BM(NH) GG. 19980: Spathi Subzone, Middle Albian, Copt Point, Folkestone.
Fig. 4. Grazing traces of Lower Cretaceous acmaeid on surface of hoplitid ammonite steinkern: note their
relatively consistent orientation (cf. fig. 3): BM(NH) GG. 19979: Spathi Subzone, Middle Albian, Copt Point,
Folkestone.
PLATE 37
AKPAN et al.. Limpet grazing
366
PALAEONTOLOGY, VOLUME 25
• with Acmaea GRAZING MARKS
text-fig. 2. Depth of occurrence of Recent shells showing algal
borings and Acmaea grazing marks: Firth of Clyde, Scotland.
Note that the limit of grazing is slightly shallower than the limit
of algal boring by ‘conchocelis’ (taken as a biological indication
of the limits of the euphotic zone).
to envisage punctured shells being concentrated on the sea bed by winnowing and wave action, thus
providing a typical habitat for the acmaeids, which are unknown pseudo-planktonically.
The following taphonomic sequence is therefore suggested:
1 . Damaged ammonite shells sink on to muddy sea floor and are colonized by boring algae on their inner and
outer surfaces (damage is supported by incomplete preservation: Reyment 1958; Seilacher 1971).
2. Acmaeid limpets colonize the dead shells, feeding on the endolithic algae in ‘sweeping’ fashion where
infestation is high, but with more clearly defined ‘scoops’ where infestation is lower.
3. Draught filling of very fine mud replicates the body chamber before the limpet grazing has reached such an
advanced stage that the weakened shell falls apart: many ammonites may have been grazed to destruction in the
absence of an intervening smothering of mud. (Attempts artificially to replicate modem grazing traces with
Epotex resin have been markedly less successful than Nature’s amazingly high-fidelity moulds in the Gault!)
4. Burial, lithification, and phosphatization (with dissolution of aragonite?).
5. Re-exhumation by winnowing of mud: concentration of nodules without transportation.
6. Boring by polychaetes? and phoronids: biting by regular echinoids.
CONCLUSIONS: BORING ALGAE AND MOLLUSCAN GRAZING TRACES
IN THE FOSSIL RECORD
The long geological record of both the boring algae (e.g. Palaeoconchocelis starmachii from the
Silurian of Poland, Campbell 1980) and the patellaceans (Trias-Rec.) suggests that earlier radula
marks may soon come to light. Boekschoten (1967) and Voigt (1977) have recorded a range of
AKPAN ET AL.: LIMPET GRAZING
367
molluscan traces as far back as the upper Jurassic, with chiton marks more frequently encountered
than those of limpets. Taylor (1981, pi. 2/3) has figured what are clearly acmaeid radula marks on an
Isognomon from the Portlandian, kerberus Zone. Again they seem to be associated with an algal-
bored shell.
The implications of the kind of approach reported in this paper are twofold. First, it shows that
when reasonably stringent sampling programmes are put into effect on present-day continental shelves
it is possible to refine the determination of palaeoecological parameters, in this instance water depth.
As this parameter is a difficult one to assess in clay sequences, this is a step forward, and the result for
the Spathi Subzone of the Gault of 8-30 m may come as something of a surprise. The second
implication lies in the evidence which is afforded for a remarkable lack of change in radula design or
pattern of feeding in the acmaeids since the Lower Cretaceous.
black, r. 1976. The effect of grazing by limpet Acmaea insessa on the kelp, Egregia laevigata in the intertidal
zone. Ecology, 57, 265-277.
boekschoten, G. s. 1967. Palaeoecology of some Mollusca from the Tielrode sands (Pliocene, Belgium).
Palaeogeogr., Palaeoclimat., Palaeoecol., 3, 311-362.
bromley, R. G. 1970. Borings as trace fossils and Entobia cretacea Portlock, as an example. In T. p. crimes and
j. c. harper (eds.) Trace fossils. Geol. J. Spec. Issue, 3, 49-90.
Campbell, s. e. 1980. Palaeoconchocelis starmachii, a carbonate boring micro-fossil from the Upper Silurian of
Poland (435 million years old): implications for the evolution of Bangiaceae (Rhodophyta). Phycologia, 19,
25-36.
carefoot, t. 1976. Pacific seashores. J. J. Douglas. Vancouver.
clokie, J., boney, a. d. and farrow, G. e. 1979. Use of Conchocelis as an indicator organism: data from Firth of
Clyde and N.W. Shelf. Br. Phycol. J. 14, 120-121.
craig, p. a. 1968. The activity pattern and food habits of the limpet Acmaea pelta. Veliger, 11 ( Suppl .), 13-19.
farrow, G. e. and clokie, j. 1979. Molluscan grazing of sub-littoral algal-bored shells and the production of
carbonate mud in the Firth of Clyde, Scotland. Trans. R. Soc. Edinb. 70, 139-148.
— allen, n. h., akpan, e. b. and brown, b. j. 1981. Bioclastic carbonate sedimentation and biofacies on a
high-latitude, tide-dominated shelf: NE Orkney Islands, Scotland. [In prep.]
fretter, v. and graham, a. 1976. The Prosobranch molluscs of Britain and Denmark. Pleurotomariacea,
Fissurellacea and Patellacea. J. Mollusc. Stud., Suppl. 1, 36 pp.
house, M. r. 1973. An analysis of Devonian goniatite distribution. Spec. Pap. Palaeont. 12, 305-317.
kozloff, e. m. 1973. Seashore life of Puget Sound, the Strait of Georgia, and the San Juan Archipelago. Seattle and
London: Univ. Washington Press. 282 pp. +28 pis.
michelin, h. 1838. Sur une argile dependant du Gault observee au Gaty, Communede Gerodot, Departement de
L’Aube. Mem. Soc. Geol., Paris, 3, 97-103.
price, f. G. h. 1879. The Gault, viii + 81 pp. London.
reyment, r. a. 1958. Some factors in the distribution of fossil cephalopods. Stockh. Contr. Geol. 1, 98-185.
seilacher, a. 1971. Preservational history of ceratite shells. Palaeontology, 14, 16-21.
smith, a. g., briden, j. c. and drewry, G. e. 1973. Phanerozoic world maps. Spec. Pap. Palaeont. 12, 1-42.
taylor, p. d. 1981. Bryozoa of British Portland beds (Upper Jurassic). Palaeontology, 24, 863-875, pis. 121-122.
voigt, e. 1977. On grazing traces produced by the radula of fossil and Recent gastropods and chitons. In t. p.
crimes and j. c. harper (eds.) Trace fossils 2, Geol. J. Spec. Issue, 9, 335-346.
REFERENCES
ETIE B. AKPAN
GEORGE E. FARROW
Department of Geology
University of Glasgow
Glasgow G12 8QQ, Scotland
Typescript received 9 October 1980
Revised typescript received 10 February 1981
NOEL MORRIS
Department of Palaeontology
British Museum (Natural History)
Cromwell Road
London SW7 5BD, England
'
A REVIEW OF RECENT AND QUATERNARY
ORGANIC-WALLED DINOFLAGELLATE CYSTS
OF THE GENUS PROTO PERIDINIUM
by REX HARLAND
Abstract. A review is given of Recent and Quaternary organic-walled dinoflagellate cysts belonging to the
genus Protoperidinium Bergh emend. Balech 1974. There is a similarity between the taxonomies based upon
thecal and cyst morphologies, the major difference being one of hierarchy. A scheme is presented which attempts
to amalgamate the two taxonomic systems, and which shows the range of cyst morphology attributable to a
single living genus, the general non-conservative nature of certain cyst morphologies, and the use of common
cyst forms amongst different dinoflagellates. Two new names, Fuscusasphaeridium and Asymmetropedinium, are
introduced, and Quinquecuspis Harland is emended and diagnosed in Latin.
Recently a number of important reviews have appeared describing the taxonomy of peridiniacean
dinoflagellates and their cysts. These include Lentin and Williams (1975) and Stover and Evitt (1978),
both of which were mainly concerned with dinoflagellate cysts, their morphology, and particularly
their archeopyle shape and structure. Unfortunately little or no attention was paid to the taxonomy
of extant dinoflagellates and their cysts, or to the recently fossilized cysts found in Quaternary
sediments. Similarly, in the field of phycology a major revision of a part of the genus Peridinium
Ehrenberg 1832 was published by Balech (1974), but with little reference to dinoflagellate cysts. This
paper attempts to look at the classification of living peridiniacean dinoflagellates, particularly those
associated with the genus Peridinium, in relation to Recent and Quaternary dinoflagellate cysts.
TAXONOMY OF THE GENUS PERIDINIUM
The genus Peridinium was erected in 1 832 by Ehrenberg with Peridinium cinctum (O. F. Muller) as the
type species. In the early part of the twentieth century, with the advent of major oceanographical
expeditions and the increasing number of species attributed to the genus (now well over 200) a
number of schemes to subdivide the genus were published. Jorgensen (1913) comprehensively
reclassified the genus, and used the number of plates that border the first apical as the principal
character in the formation of subgenera, and the pattern of the dorsal plates on the epitheca, in
particular the relationship with the precingular plate series, for the division of the subgenera into
sections.
In the first instance the genus was divided into two subgenera: Orthoperidinium [T contacts four plates 1", 2', 4',
and 7”] and Metaperidinium [1 ' contacts five plates 2', 1 ", 2", 4', and 7"]. The third condition Paraperidinium [T
contacts six plates 2', 1 ", 2", 4', 6", and 7"] was not given subgeneric status because of the considerable variation
in the length of the sutures between the plates. The subgenus Orthoperidinium was then divided into the
following three sections:
Tabulata 2a plate contacts 3" and 4" or 4" and 5" plates
Conica 2a plate contacts 3", 4", and 5"
Oceanica 2a plate contacts 4" only,
Similarly the subgenus Metaperidinium-.
Pyriformia 2a plate contacts 3" and 4" or 4" and 5"
Paraperidinium see discussion above
Humilia 2a plate contacts 4" only; with solid antapical horns
Divergens as above with hollow antapical horns.
| Palaeontology, Vol. 25, Part 2, 1982, pp. 369-397, pis. 38-42.|
370
PALAEONTOLOGY, VOLUME 25
This system was revised by Paulsen ( 1 93 1 ) to take into account a greater number of characters, and he
used the terms ‘ortho’-contacts 1", 2', 3', 4', and 7", ‘meta’ -contacts l", + 2”, 2', 3', 4', + 6", and 7",
and ‘para’-contacts 1 ", 2", 2', 3', 4', 6", and 7" to characterize the first apical plate (see text-fig. 1); and
‘quadra’-contacts 4" only, ‘hexa’-contacts 3", 4", and 5" and ‘penta’-contacts 3" and 4" to
characterize the dorsal epithecal configuration, especially the contacts at the posterior border (see
text-fig. 2). Paulsen included in his treatment of Peridinium, forms with two intercalary plates as
text-fig. 1 . Species of Protoperidinium illustrating the three different types of
first apical plates (after Graham 1942). The first apical plate is stippled. The
ortho species also shows growth bands.
text-fig. 2. Examples of the three different styles of dorsal epithecal tabulation (after Graham
1942). The second dorsal intercalary plate is hachured.
opposed to the normal three referring them to the subgenus Archaeperidinium ( = genus
Archaeoperidinium of Jorgensen (1913)). These taxonomic schemes all rely upon the ‘stability’ of
plate patterns in the region of the first apical and dorsal epithecal plates and in general on the
possession of three dorsal intercalary plates. Graham (1942) noted, however, that the number of
plates in the cingulum could be important in the taxonomy of Peridinium species and that their study
had been neglected despite the fact that the dissection of the plates was not difficult.
Balech (1974), in his major revision of a part of the genus Peridinium , used the nature of the
cingular plates to differentiate between freshwater Peridinium sensu stricto with five cingular plates,
and marine Peridinium species with four cingular plates. The generic name Protoperidinium Bergh
was used to accommodate these marine species. Protoperidinium was then subdivided into three
subgenera based upon the number of precingular and intercalary plates. Further subdivisions were
recognized using the criteria first employed by Jorgensen (1913) and Paulsen (1931).
Although the basis of this taxonomic subdivision is the number, shape, and size of the plates
together with the tabulation pattern, I have never seen a reasoned account explaining why the number
of cingular plates and the number of intercalary plates should be given hierarchical priority over the
configuration of the first apical plate, or why the configuration of the first apical plate should be given
HARLAND: DINOFLAGELLATE CYSTS
37:
priority over the mutual relationship of the second intercalary plate and its neighbours. Indeed,
evidence from the cyst morphology suggests that the shape and relative position of the second
intercalary plate is possibly more significant, because it is through this site that excystment occurs.
Recent work by Gocht and Netzel (1974, 1976) on the overlap system in peridiniacean and
gonyaulacacean dinoflagellates (Durr and Netzel 1974) has suggested that this second intercalary site
is the keystone for both overlap or imbrication (Dorhofer and Davies 1980), and in archeopyle
formation, and that it must be genetically determined. This is in contrast to the first apical plate which
does not appear to play such a major role in thecal or cyst function.
It is, however, Balech’s (1974) scheme that is used herein as a basis for discussing the contribution
of cyst morphology. We are obliged to use thecal morphology as the starting-point since the
information is potentially complete and only a small proportion of the dinoflagellates produce
fossilizable cysts as a part of their life-cycle (Dale 1976). Cysts of the genus Protoperidinium are
known to have a varied morphology ranging from simple spheres to quite complex cysts with
processes and horns. Often the common unifying feature is the style of archeopyle formation, which
always involves the use of an intercalary paraplate with or without its adjacent paraplates. Only in
rare cases is it possible to distinguish a clear paratabulation. In this study emphasis is necessarily
placed upon gross morphology and archeopyle formation in investigating the relationships of the
various Protoperidinium cysts to their respective thecal stages.
SYSTEMATIC DESCRIPTIONS
In each of the subsequent discussions of genera, subgenera, sections, and species, the relevant cyst
morphologies will be particularly noted, with further comment reserved for later sections. Cysts
known to the author are described in some detail, otherwise the reader is referred to the best published
description. The accompanying thecal tabulation diagrams have been standardized, do not
necessarily correspond to details of the particular taxa in nature, and should therefore be treated as
diagrammatic.
All the cysts illustrated here are registered in the MPK series and are housed in the Palynological
Collections of the Institute of Geological Sciences (IGS), Leeds.
The taxonomic changes that have resulted from this review, including the erection of two new taxa
and the change in status of several others, are all handled within the Appendix. The taxonomic system
used herein is, however, summarized in Table 1.
Division pyrrhophyta Pascher 1914
Class DINOPHYCEAE Fritsch 1929
Order peridiniales Haeckel 1 894
Family peridiniaceae Ehrenberg 1832
Genus protoperidinium Bergh emend Balech 1974
Type species. Protoperidinium pellucidum Bergh 1881; S.D. by Loeblich, Jr. and Loeblich, III, 1966.
Remarks. This genus accommodates, for the most part, the marine species formerly belonging to the
genus Peridinium Ehrenberg. They have four cingular plates in contrast to the five in Peridinium sensu
stricto. An exception is P. faeroense Paulsen which has five cingular plates (Dale 1978). The type
species possesses a para first apical plate arrangement and a hexa dorsal epithecal configuration (text-
fig. 3). Cysts capable of fossilization have not been recorded for the type species.
Subgenus Minusculum (Lebour) Balech 1974
Type species. Protoperidinium ( Minusculum ) bipes (Paulsen) Balech 1974; O.D.
Remarks. This subgenus is characterized by six precingular plates and three intercalary plates. The
unique ‘boomerang’ shape of plate 6" (Balech 1974) is also a significant feature (text-fig. 4). No
fossilizable cysts have been observed from species attributable to this subgenus.
table 1. Taxonomy of Protoperidinium species that produce fossilizable cysts
GENUS
SUBGENUS
SECTION
SPECIES
Protoperidinium
Minus culum
A rchaeperidinium
A rchaeperidinium stat. nov.
minuium
Stelladinium stat. nov.
compressum
Fuscusasphaeridium nomen nov.
ave liana
denticulatum
excentricum
Protoperidinium
Para
Quadra
Penta
Hexa
Protoperidinium stat. nov.
latissimum
Meta
Quadra
Penta
Hexa
-
-
Ortho
Quadra
Votadinium stat. nov.
oblongum
Penta
Asymmetropedinium nomen nov.
punctulatum
Hexa
Brigantedinium stat. nov.
conicoides
Selenopemphix stat. nov.
nudum
subinerme
Quinquecuspis stat. nov.
leonis
Trinovantedinium stat. nov.
pentagonum
table 1. Taxonomy of Protoperidinium species that produce fossilizable cysts
HARLAND: DINOFLAGELLATE CYSTS
373
text-fig. 3 {left). Epithecal tabulation of Protoperidinium
{ Protoperidinium ) pellucidum Bergh ex Loeblich, Jr. and
Loeblich, III (after Lebour 1925).
text-fig. 4 {right). Epithecal tabulation of Protoperidinium
{Minusculum) bipes (Paulsen) Balech (after Balech 1974).
Subgenus Archaeperidinium (Jorgensen) Balech 1974
Type species. Protoperidinium {Archaeperidinium) minutum (Kofoid) Loeblich, III, 1969; O.D.
Remarks. Archaeperidinium has seven precingular plates, an ortho first apical and two intercalary
plates, and contains at least five species that are known to produce fossilizable cysts. These are
described and discussed herein.
Protoperidinium {Archaeperidinium) minutum (Kofoid) Loeblich, III 1969
Text-fig. 5
Remarks. The thecal plate configuration of this the type species shows the distinct ortho first apical
and two symmetrically placed intercalary plates. The cysts produced by this species are described in
Wall and Dale (1968) and illustrated on their pi. 4, fig. 7 but they have not been seen by the present
author. The archeopyle, however, is reported to be intercalary (? la or 2a) but is rarely seen as an
aperture; Wall and Dale’s (1968) illustration appears to show an attached operculum. The cyst is
otherwise characterized by short, hollow processes with a flat-topped distal extremity that bears
spinules around the circumference. The cysts are known from Recent sediments. Incubation
experiments (Wall and Dale 1968) have clearly established the link between this theca and cyst.
Protoperidinium {Archaeperidinium) avel/ana (Meunier) Balech 1974
Text-fig. 6, 7b; Plate 38, figs. 4-9
text-fig. 5 {left). Epithecal tabulation of Protoperidinium
{Archaeperidinium) minutum (Kofoid) Loeblich, III (after
Wall and Dale 1968).
text-fig. 6 {right). Epithecal tabulation of Proto-
peridinium {Archaeperidinium) avellana (Meunier) Balech
(after Wall and Dale 1968).
374
PALAEONTOLOGY, VOLUME 25
Remarks. The thecal tabulation illustrated shows the salient features but it is interesting to note that
there is some asymmetry in the position of the intercalary plates and especially plate 4". The cysts of
this species have been described and illustrated by Wall and Dale (1968, pi. 4, fig. 2) and consist of
spherical, smooth-walled brown bodies with a single dorsal intercalary archeopyle. The shape of the
operculum is seen in Wall and Dale (1968), Reid (1977), Harland (1977) and in the present
illustrations. The degree of reflection of the detailed thecal plate shape is often quite remarkable.
There can be some confusion between these cysts and those of P. (A.) denticulatum (Gran and
Braarud) Balech 1974 although the archeopyles are distinct (text-fig. 7). In palynological
preparations many such cysts are often so crumpled that a proper analysis of the archeopyle shape is
not always possible. Also, the archeopyle shape can be somewhat modified and need not be a perfect
reflection of the thecal plate counterpart. Incubation experiments (Wall and Dale 1968) have
established the link between this theca and this cyst.
text-fig. 7. Apical view of the cysts A — Protoperidinium
( Archaeperidinium ) denticulatum (Gran and Braarud)
Balech (after Wall and Dale 1968), and B— P. (A.)
avellana (Meunier) Balech showing the distinct archeo-
pyle shapes, both symmetrical.
The cyst was originally described as Chytroeisphaeridia cariacoensis Wall 1965, but may be placed
in Reid’s (1977) genus Brigantedinium recently validated by Harland and Reid (1980) in Harland,
Reid, Dobell, and Norris (1980). However, the type species of Brigantedinium , B. simplex , is the cyst
of P. ( Protoperidinium ) conicoides (Paulsen) Balech, a member of a separate subgenus. The placement
of P. (A.) avellana in Brigantedinium may, therefore, be inappropriate and perhaps a separate
designation needs consideration (see later discussion).
Cyst description. Simple spheroidal brown cyst made up of autophragm or with two very closely adpressed wall
layers. Paratabulation not present although the archeopyle shape does hint at the configuration of the 2a, 4', 3',
and la contacts (text- fig. 7b). Cyst wall generally smooth but may appear somewhat shagreenate; the possible
indications of two indentations on the ? sulcal area may be a reflection of flagellar pores. Archeopyle intercalary
formed by the loss of paraplate 2a which is transversely elongate and symmetrical (see text-fig. 7b). Operculum
free. Cyst diameters fall generally within the range 50-0-55-0 ^m.
EXPLANATION OF PLATE 38
All figures are illustrated at a magnification of x 500 and were photographed in plain transmitted light unless
otherwise indicated.
Figs. 1-3. Protoperidinium ( Protoperidinium sect. Brigantedinium) conicoides (Paulsen) Balech, Nomarski
interference contrast. 1, dorsal view of MPK 1232 showing simple morphology and standard hexa
archeopyle. 2, ventral view of MPK 1232 with the two flagellar indentations. 3, dorsal view with the
archeopyle and operculum, specimen MPK 1243.
Figs. 4-9. Protoperidinium ( Archaeperidinium sect. Fuscusasphaeridium) avellana (Meunier) Balech, figs. 4 and 5
by Nomarski interference contrast. 4, apical view showing symmetrical archeopyle with operculum in place,
specimenMPK 1236. 5, ditto, specimen MPK 1238. 6, oblique view with operculum coming free, specimen
MPK 1235. Figs. 7-9, various levels of focus showing operculum free and more crumpled nature of the cyst,
specimen MPK 2768.
Figs. 10-12. ?Indet. Protoperidinium cyst, various levels of focus to show the general morphology and the zigzag
archeopyle split, specimen MPK 2769.
PLATE 38
HARLAND, Protoperidinium
376
PALAEONTOLOGY, VOLUME 25
Protoperidinium ( Archaeperidinium ) compressum (Abe) Balech 1974
Text-figs. 8, 9; Plate 39, fig. 12
Remarks. The thecal tabulation shows the features of this subgenus, but in particular the asymmetry
in the disposition of the intercalary and 4" plates is worth noting. Text-fig. 8 shows the theca as
expanded whereas in life the species is compressed dorso-ventrally, hence the name. The cyst of this
species is particularly striking in its stellate morphology, and has been described by Wall and Dale
(1968), Bradford (1975), Reid (1977), and Harland (1977). The archeopyle is intercalary and formed
by the displacement of two paraplates which remain attached laterally (text-fig. 9). They are
text-fig. 8 (left). Epithecal tabulation of Protoperidinium
(Archaeperidinium) compressum (Abe) Balech (after Wall
and Dale 1968).
text-fig. 9 (right). Archeopyle formation in Protoperidinium
(Archaeperidinium) compressum (Abe) Balech, A — Speci-
men MPK 1256 (see PI. 39, fig. 12), B— Idealized scheme.
symmetrically placed and fold back to open a large archeopyle. It is thought that both la and 2a
paraplates are involved. They do not appear to reflect the asymmetry of the theca. These cysts are
unique and to date at least two species (herein PI. 39, fig. 12, and in Wall and Dale 1968, pi. 2, figs.
15-17) are known, only one of which can be definitely assigned to P. (A.) compressum. Incubation
experiments (Wall and Dale 1968) have established the link between this theca and the illustrated cyst
described below.
Cyst description. Stellate, peridinioid cysts, compressed dorso-ventrally made up of autophragm. Wall smooth,
and epitract smaller than hypotract. Cyst carries one apical, two antapical, and two lateral horns that are long,
solid, and acicular. Thickening of the horn bases is sometimes apparent. Paratabulation not present. Archeopyle
intercalary formed by the opening of two intercalary paraplates that remain attached laterally. Cysts range in
length, excluding the processes, from 26-0-37-0 ^m.
Protoperidinium (Archaeperidinium) denticulatum (Gran and Braarud) Balech 1974
Text-figs. 7a, 10
Remarks. The tabulation of this species shows a remarkable similarity to that of P. (A.) avellana j
especially with respect to the slight asymmetry of the two intercalary plates; best seen in relation to the ]
position of plate 4". It is, therefore, perhaps not surprising that the cyst illustrated and described by
Wall and Dale (1968, pi. 3 fig. 30) is very similar to that produced by P. (A.) avellana (see Wall and
Dale 1968, p. 277) although the identification is admittedly not positive. Archeopyle formation is by |
loss of a single transversely elongate intercalary paraplate whose configuration is symmetrical, but
unlike that of P. (A.) avellana (text-fig. 7). In palynological preparations confusion between the two
forms is common. The form illustrated here has a provisional assignment to this species. Incubation ]
experiments by Wall and Dale (1968) failed to positively identify the theca therefore a clearly '
established link between this theca and cyst is not possible.
HARLAND: DINOFLAGELL ATE CYSTS
377
Protoperidinium ( Archaeperidinium ) excentricum (Paulsen) Balech 1974
Text-fig. 1 1
Remarks. The thecal tabulation of this species is very similar to both those of P. (^4.) avellana and P.
(A.) denticulatum. The cysts are brown oval bodies that Wall and Dale (1968, p. 278) describe as
being flattened in a polar direction. The archeopyle has not been described. Cysts attributable to this
species have not been seen by the present author. Although cysts enclosed by thecae were recorded by
Wall and Dale (1968) it is not clear if any thecae were obtained by incubation and positively identified.
text-fig. 10 {left). Epithecal tabulation of Proto-
peridinium ( Archaeperidinium ) denticulatum (Gran
and Braarud) Balech (after Wall and Dale 1968).
text-fig. 1 1 (right). Epithecal tabulation of Proto-
peridinium (Archaeperidinium) excentricum (Paulsen)
Balech (after Lebour 1925).
Discussion of the subgenus Archaeperidinium
Dinoflagellate cysts of the subgenus Archaeperidinium fall into three distinct categories. The first is
exemplified by the type species, a cyst with short processes and a single intercalary archeopyle; the
second a stellate cyst with a two paraplate intercalary archeopyle; and finally the third with brown
spherical/spheroidal cysts with single intercalary transversely elongate archeopyle. These differences
in gross cyst morphology underline differences in the tabulation pattern; the type species is
symmetrical; the second is very markedly asymmetrical and divided laterally into a clearly marked
dorsal and ventral epitheca, and finally those that are only slightly asymmetrical especially in regard to
the position of plate 4". I believe this is a sufficiently natural genotypic division to warrant the
application of particular names. The names chosen are derived from both biological and
palaeontological literature and include Stelladinium Bradford 1975, a name coined to apply to the
unique stellate morphology of P. compressum cysts. Herein I shall treat them as separate sections such
that:
1. Protoperidinium (Archaeperidinium sect. Archaeperidinium stat. nov.) includes minutum
2. Protoperidinium (Archaeperidinium sect. Stelladinium stat. nov.) includes compressum
3. Protoperidinium (Archaeperidinium sect. Fuscusasphaeridium nomen nov.) includes avellana,
denticulatum , and excentricum
The only new name, Fuscusasphaeridium, is formally erected in the appendix together with the change
of status of the other two taxa. This approach hopefully embodies the general ideas expressed by Wall
and Dale (1968), Dale (1976 and 1978), and Reid and Harland (1977). An amalgamation of systems
should be possible since we are dealing with common organisms, albeit at different stages of their life-
cycle.
Some difficulties are still apparent, for instance P. (P.) punctulatum (Paulsen) also has brown
spherical cysts with transversely elongate archeopyles but with a very different thecal tabulation.
Possible differences in cyst archeopyle morphology, i.e. ? attached and an asymmetrical opercula
(see later discussion) may serve to differentiate them from cysts of P. punctulatum.
378
PALAEONTOLOGY, VOLUME 25
Subgenus Protoperidinium (Bergh) Balech 1974
Type species. Protoperidinium ( Protoperidinium ) pellucidum Bergh 1881, S. D. by Loeblich, Jr. and Loeblich, III,
1966.
Remarks. The thecal tabulation of P. (P.) pellucidum has already been discussed, see text-fig. 3. Balech
(1974) does, however, subdivide the subgenus into a number of units based upon the pattern of the
first apical plate and its adjacent plates, and then again on the dorsal intercalary configuration. Since
the type species has a para first apical plate that group will be discussed first.
A. Para-species
i. Quadra. No fossilizable cysts attributable to dinoflagellates in this group are known.
ii. Penta. No fossilizable cysts attributable to dinoflagellates in this group are known.
iii. Hexa. Only one species within this unit is known to produce fossilizable cysts and that is
described below. The type species, P. ( P .) pellucidum, also belongs in this group.
Protoperidinium { Protoperidinium ) latissimum (Kofoid) Balech 1974
Remarks. The thecal tabulation of this species clearly demonstrates the para first apical plate and the
hexa dorsal intercalary arrangement and the symmetry of the pattern. The fossilizable cyst for this
species has been described by Wall and Dale (1968) as large, pentagonal, and dorso-ventrally
compressed. A paracingulum is represented by broad weakly excavated lateral lobes. The epitract is
triangular in outline and large, while the hypotract is much smaller and has two small antapical horns.
The archeopyle is intercalary, has a hexa shape but is asymmetrical {see Wall and Dale 1968, pi. 2, fig.
7), is large, taking up most of the dorsal surface of the cyst. The operculum is free. Incubation
experiments (Wall and Dale 1968) have firmly linked this theca with this cyst.
The cyst genus Leipokatium Bradford 1975 was erected to accommodate cysts very similar to that
described above. Since, however, the type species for the subgenus also is assigned to this group it
might be prudent to suppress Leipokatium in favour of Protoperidinium {Protoperidinium sect.
EXPLANATION OF PLATE 39
All figures are illustrated at a magnification of x 500 and were photographed in plain transmitted light unless
otherwise indicated.
Figs. 1-3. Protoperidinium {Protoperidinium sect. Selenopemphix) conicum (Gran) Balech. 1, apical view to
show the reniform ambitus and offset standard hexa archeopyle, specimen MPK 2770. 2, ditto, specimen
MPK 2771. 3, ditto, specimen MPK 2772.
Figs. 4, 5. Indet. ?Protoperidinium cyst, Xandarodinium xanthum Reid. 4, ?apical view showing over-all
morphology, Nomarski interference contrast, specimen MPK 1261. 5, ditto, specimen MPK 2773.
Fig. 6. Protoperidinium {Protoperidinium sect. Selenopemphix) subinerme (Paulsen) Loeblich, III, apical view
showing broad paracingular zone, lack of ornament, and offset archeopyle with operculum, specimen MPK
1634.
Figs. 7-11. Protoperidinium {Protoperidinium sect. Trinovantedinium ) pentagonum (Gran) Balech. 7, dorsal
view showing the broad hexa archeopyle, hyaline nature of the cyst, and the paracingulum. 8, ventral view
with parasulcus and planar paracingulum. 9, optical section illustrating the nature of the apical boss,
specimen MPK 1240. 10, dorsal view with broad hexa operculum and nature of the parasutural and
intratabular spines. 1 1, ventral view and deeply indented parasulcus, specimen MPK 2774, all figures with
Nomarski interference contrast.
Fig. 12. Protoperidinium {Archaeperidinium sect. Stelladinium) compressum (Abe) Balech 1974, dorsal view
showing stellate morphology and the 21 archeopyle with the two opercular paraplates attached laterally,
Nomarski interference contrast, specimen MPK 1256.
PLATE 39
HARLAND, Protoperidinium
380
PALAEONTOLOGY, VOLUME 25
text-fig. 12 {left). Epithecal tabulation of Proto-
peridinium {Protoperidinium) latissimum (Kofoid) Balech
(after Wall and Dale 1968).
text-fig. 13 {right). Epithecal tabulation of Proto-
peridinium {Protoperidinium) claudicans (Paulsen) Balech
(after Wall and Dale 1968).
Protoperidinium stat. nov.). However, no fossilizable cysts are known from the type species and
therefore the thecal and cyst morphologies cannot be compared. In fact certain thecal tabulation
details are different between P. pellucidum and P. latissimum, but it is difficult to know if these are
taxonomically significant at this level (compare text-figs. 3 and 12).
B. Meta-species
Fossilizable cysts have not been described from Modern species attributable to the quadra, penta, or
hexa subdivisions of this unit. However, it is of interest that a new species of Phthanoperidinium
described by Edwards and Bebout (1981) has a meta configuration.
C. Ortho-species
This group contains the most numerous fossilizable Protoperidinium cysts and like those above can be
subdivided as follows:
i. Quadra. This includes the species P. {P.) claudicans (Paulsen) Balech 1974 and P. {P.)
oblongum (Aurivillius) Balech 1974. Balech (1974) considers the latter to be a part of P. {P.)
oceanicum Vanhoffen. Both species produce fossilizable cysts.
Protoperidinium {Protoperidinium) claudicans (Paulsen) Balech 1974
Text-fig. 13
Remarks. The roughly symmetrical thecal tabulation clearly shows the ortho first apical plate and a
penta dorsal configuration. This perhaps indicates that the dorsal configuration is subject to some
phenotypic variation as this species is usually placed in the quadra group (Balech 1974) although it is
admitted that some possess the penta pattern. The cyst of this species has been described by Wall and
Dale (1968) and Reid (1977). In dorsal view it is a chordate cyst that bears numerous short-pointed
spines. Paratabulation not present but the parasulcus is deep and separates the two broad
asymmetrical antapical lobes. Paracingulum not observed. Archeopyle formation is reported to be by
loss of the 2a intercalary paraplate which is subapical in position and has a tendency to truncate the
apex. Its shape tends to be pentagonal and if indeed only the one paraplate is involved then it is surely
an enlarged archeopyle. Incubation experiments have confidently established the theca/cyst
relationship in this species (Wall and Dale 1968). These cysts have in the past been included in Reid’s
(1977) genus Votadinium.
Protoperidinium {Protoperidinium) oblongum (Aurivillius) Balech 1974
Text-fig. 14; Plate 40, figs. 10-12
Remarks. The thecal tabulation as illustrated clearly demonstrates the ortho first apical plate and
quadra 2a plate and the symmetrical nature of the over-all pattern. The cyst of this species has been
HARLAND: DINOFLAGELL ATE CYSTS
38:
described by Wall and Dale (1968), Reid (1977), and Harland (1977), with the last two authors refer-
ring it to Reid’s (1977) genus Votadinium. It may be that this species also exhibits variation in the
dorsal thecal configuration, as in P. claudicans, since there appears to be some variation in cyst
morphology (see Wall and Dale 1 968, pi. 1 , figs. 23-29). Incubation experiments have established the
theca/cyst relationship within the species, but some experiments have shown that relationships exist
with the varieties latidorsale Dangeard, inaequale Dangeard, and symmetricum Dangeard and this
may in part explain some, if not all, of the cyst morphological variations.
text-fig. 14. Epithecal
tabulation of Protoperi-
dinium ( Protoperidinium )
oblongum (Aurivillius)
Balech (after Wall and
Dale 1968).
Cyst description. The cysts are chordate and smooth, formed of autophragm. They possess a clearly defined
deeply indented parasulcus which effectively divides the two rounded antapical lobes. Faint traces of probable
reflected flagellar pores may sometimes be seen in the sulcus. Paracingulum not seen. The archeopyle is broad
and quite large and appears to reach the apex of the cyst on the dorsal surface, giving the cyst a truncated
appearance similar to P. claudicans. Whether this archeopyle is an enlarged intercalary involving the loss of
paraplate 2a or whether it includes paraplates la, 2a, 3a, and 3' is not known. Cyst diameters range from 54-0 to
75-0 /xm.
Discussion of ortho quadra Protoperidinium species
The two species of this subdivision of the subgenus are clearly related in terms of both their thecal
tabulation and cyst morphology. The cysts have in common their over-all chordate morphology and
archeopyle style despite the uncertainty as to which paraplates may be involved. The use of a
particular name to identify the condition is, therefore, clearly necessary. In this instance Votadinium is
available for use as a section of the subgenus; hence Protoperidinium ( Protoperidinium sect.
Votadinium stat. nov.).
The variation in the dorsal tabulation of P. claudicans and the variation in the cysts of P. oblongum
may point to a deficiency in the taxonomy of these species, with perhaps a lack of sufficient evidence
on the varieties that exist and their respective cysts. Further incubation experiments are essential to
more closely define the apparent variation.
ii. Penta. Only one species Protoperidinium ( Protoperidinium ) punctulatum (Paulsen) Balech
1974 produces fossilizable cysts and is included here.
Protoperidinium ( Protoperidinium ) punctulatum (Paulsen) Balech 1974
Text-figs. 15-17; Plate 42, figs. 3-6
Remarks. This dinoflagellate, although usually penta, can have a hexa dorsal configuration as seen in
the illustrations, and can appear slightly asymmetrical. The cyst has been described by Wall and Dale
(1968) and is spherical, brown, with a large intercalary archeopyle formed by the loss of paraplate 2a.
Operculum may be attached (Wall and Dale, pi. 2, fig. 27). The aperture is transversely elongate but is
asymmetrical and this may be a significant difference compared with the cysts of P. (A.) avellana and
P. ( A .) denticulatum (compare text-figs. 7 and 17). Confusion between all these cysts can occur if well-
preserved, three-dimensional orientated specimens are not available. However, the nature of the
archeopyle and thecal tabulation does serve to distinguish this dinoflagellate from other species. Wall
and Dale (1968) have established by incubation the link between this theca and cyst.
382
PALAEONTOLOGY, VOLUME 25
text-fig. 15 (left). Epithecal tabulation of a hexa Protoperidinium (Protoperidinium)
punctulatum (Paulsen) Balech (after Wall and Dale 1968).
text-fig. 16 (centre). Epithecal tabulation of a penta Protoperidinium (Protoperidinium)
punctulatum (Paulsen) Balech (after Lebour 1925).
text-fig. 17 (right). Apical view of the cyst of Protoperidinium (Protoperidinium) punctulatum
(Paulsen) Balech showing the laterally elongate, asymmetrical archeopyle.
A name is required to accommodate these dinoflagellates and to differentiate them from section
Fuscusasphaeridium which is erected herein to hold the two Archaeperidinium species discussed
previously. The name proposed is Asymmetropedinium nom. nov. (see appendix), hence
Protoperidinium (Protoperidinium sect. Asymmetropedinium) punctulatum for this species.
Cyst description. Spheroidal brown cyst made up of autophragm or two closely adpressed wall layers. Surface
smooth, shagreenate, or slightly granulate. Paratabulation only revealed by archeopyle formation. Archeopyle
intercalary formed by the loss of paraplate 2a. Operculum large, transversely elongate but asymmetrical. Cyst
diameter ranges from 40 0 to 60 fim.
iii. Hexa. This group contains the following species that produce fossilizable cysts.
Protoperidinium (Protoperidinium) conicoides (Paulsen) Balech 1974
Text-fig. 18; Plate 38, figs. 1-3
Remarks. The epithecal tabulation of this species illustrates the nature of the ortho first apical plate
and the hexa dorsal intercalary arrangement together with the symmetry. The cyst of this species has
EXPLANATION OF PLATE 40
All figures are illustrated at a magnification of x 500 and were photographed in plain transmitted light unless
otherwise indicated.
Figs. 1-8. Protoperidinium cyst, Lejeunia paratenella Benedek. 1, low-focus dorsal view with attenuated hexa
archeopyle. 2, high-focus dorsal view showing the paracingulum and nature of aciculate antapical
horns. 3, ventral view with slightly displaced paracingulum, specimen MPK 2775. 4, dorsal view with
archeopyle and denticulate cingular parasutures. 5, ditto in phase contrast, specimen MPK 2776. 6, dorsal
view showing attenuated hexa archeopyle with operculum in place, Nomarski interference contrast. 7,
ventral view, Nomarski interference contrast, specimen MPK 1247. 8, dorsal view with attenuated hexa
archeopyle, Nomarski interference contrast, specimen MPK 1245.
Fig. 9. Protoperidinium (Protoperidinium sect. Quinquecuspis) leonis (Pavillard) Balech, dorsal view with
standard hexa archeopyle exhibiting slight apical tongue. Difference in archeopyle style between this and
previous species is marked, specimen MPK 2777.
Figs. 10-12. Protoperidinium (Protoperidinium sect. Votadinium) oblongum (Aurivillius) Balech. 10, ventral
view showing overall chordate morphology. 11, dorsal view with archeopyle truncating apex. 12, ditto,
Nomarski interference contrast, specimen MPK 2778.
PLATE 40
HARLAND, Protoperidinium
384
PALAEONTOLOGY, VOLUME 25
been described by Wall and Dale (1968), Reid (1977), and Harland (1977), and was originally
placed in the cyst genus Chytroeisphaeridia as C. simplicia by Wall (1965). More recently it has been
chosen as the type species for the cyst genus Brigantedinium Reid which has been validated by
Harland and Reid (1980) in Harland et al. (1980). The cyst morphology is, like many other
Protoperidinium cysts, a simple brown ball. This cyst has not been successfully incubated but the
link between theca and cyst is almost certain (Wall and Dale 1968).
text-fig. 18. (left) Epithecal tabulation of Proto-
peridinium (Protoperidinium) conicoides (Paulsen)
Balech (after Wall and Dale 1968).
text-fig. 19 (right). Epithecal tabulation of Proto-
peridinium (Protoperidinium) conicum (Gran) Balech
(after Wall and Dale 1968).
Cyst description. A spherical brown cyst made up of autophragm which may be smooth, shagreenate, or loosely
reticulate. May exhibit a slightly indented paracingular and parasulcal area, the latter possessing indications of
the flagellar pores. Archeopyle formed by loss of a single intercalary paraplate (2a) of standard hexa shape which
may show interesting detail around the margin (PI. 38, fig. 1). This detail, especially, toward the apex, perhaps
suggests the loss of more than just a single paraplate. Operculum free. Cyst diameters range from 30-0 to 500 jum.
This cyst can be distinguished from other similar cysts by the nature of its archeopyle (see earlier sections) but
it is often difficult to recognize if the cysts are crushed.
Protoperidinium (Protoperidinium) conicum (Gran) Balech 1974
Text-fig. 19; Plate 39, figs. 1-3; Plate 42, figs. 1,10
Remarks. The thecal epitabulation shows the main features of this group and is distinctly
symmetrical. An interesting feature is, however, the almost rectangular shape of plates 2' and 4'. The
cysts of this species have been described by Wall and Dale (1968), Bradford (1975), Reid (1977), and
Harland (1977). The cyst genus Multispinula Bradford has been applied to them although this is a
junior synonym of Selenopemphix Benedek which was originally erected for cysts showing polar
compression and an offset archeopyle (Stover and Evitt 1978). Bujak (1980) formally emended
Selenopemphix to draw attention to this offset archeopyle and also to include spinose forms. Wall and
Dale (1968) established the theca/cyst link by incubation studies.
Cyst description. Ovoidal to reniform cysts probably made up of two wall layers, the outer making up the solid
aciculate processes. Cyst usually has a polar compression due to its general cyst morphology of low epicystal and
hypocystal cones (see Wall and Dale 1968, pi. 2, figs. 4, 5). This compression may also have the effect of slightly
rotating the epicyst to accentuate the offset archeopyle. PI. 39, figs. 1 -3 demonstrates the variable amount of
offset of the archeopyle. The sulcal area is exhibited as an indentation in the ventral side of the ambitus and the
paracingulum is displayed as a circumferential band. Processes are generally most common at the apex and
circumference although they also occur all over the cyst. Archeopyle intercalary formed by the loss of paraplate
2a which exhibits a standard hexa shape and is asymmetrically placed. Operculum free. The offset archeopyle
may indicate the utilization of an additional paraplate to the 2a in archeopyle formation. Cyst diameter 40 0 to
60 0 ^m.
HARLAND: DINOFLAGELL ATE CYSTS
385
There is some confusion between these cysts and those of P. { P .) nudum that have a similar morphology but are
smaller. Bradford (1975) and Reid (1977) argued that a continuous size gradation existed between the two such
that only the one cyst species can be recognized in the absence of knowledge of the thecal morphology.
Protoperidinium ( Protoperidinium ) leonis (Pavillard) Balech 1974
Text-fig. 20; Plate 41, figs. 1-14; Plate 42, figs. 7, 9.
Remarks. The epithecal tabulation of this species indicates a possible assignment to the penta group
and not to the hexa as Balech (1974) suggests, but Lebour (1925) has figured a specimen from
Plymouth Sound with a clear hexa dorsal arrangement. Again some slight variation in the dorsal
tabulation pattern is apparent. A symmetrical arrangement of plates on the epitheca is normal. The
cysts of this species have been described by Evitt and Davidson (1964), Wall and Dale (1968),
Bradford (1977), Reid (1977), and Harland (1977) using such generic taxa as Lejeunia Gerlach,
Trinovantedinium Reid and Quinquecuspis Harland (see Harland 1977 for some synonymies). Wall
and Dale (1968) established the theca/cyst link by incubation but had difficulty in distinguishing
P. leonis from such species as P. marielebourae (Paulsen) and P. obtusum (Karsten). The range of cyst
morphologies place within P. (P.) leonis reinforces the need for further study as it is perfectly possible
that there are a number of separate species involved, including those represented in PI. 40, fig.
9— species 1, PI. 41, figs. 1, 2, 7, 8— species 2, PI. 41, figs. 3-6— species 3, PI. 41, figs. 9, 10— species 4
and PI. 41, figs. 11-14 — species 5.
All these forms can be attributed to the genus Lejeunia Gerlach emend Stover and Evitt 1978 now
Lejeunecysta Artzner and Dorhofer 1978 but I believe there is a justification in keeping them within a
separate taxon, Quinquecuspis. This is based upon their brown colour, thick wall, deeply indented
parasulcus, often discontinuous paracingulum, and the standard hexa archeopyle, and an archeopyle
index of c. 0-35-0-45. By contrast, Lejeunia cysts are often paler in colour, thin walled, have a planar
continuous paracingulum with an attenuated hexa archeopyle, and an archeopyle index of c. 0-2-0-3.
Cyst description. Peridinioid acavate brown cysts made up of autophragm which is often thickened at the apex
and antapex. The cyst surface is generally smooth but may be somewhat shagreenate. Epitract may be conical or
have shoulders, whilst the hypotract carries two asymmetrical horns. Paratabulation may be represented by
distinctly indented paracingulum and sulcus, the former delimited by low discontinuous or continuous ridges.
Archeopyle intercalary by loss of paraplate 2a. Archeopyle shape can vary from standard hexa (PI. 41 , figs. 3, 14)
to standard hexa with an apical tongue (PI. 41, fig. 1). The antapical margin of the archeopyle is often at, or very
close to, the paracingulum. Operculum free. Cyst diameter varies from 60 0 to 75 0 (im.
Protoperidinium ( Protoperidinium ) nudum (Meunier) Balech 1974
Text-fig. 21
text-fig. 20 {left). Eptithecal tabulation of Proto-
peridinium {Protoperidinium) leonis (Pavillard) Balech
(after Wall and Dale 1968).
text-fig. 21 {right). Epithecal tabulation of Proto-
peridinium ( Protoperidinium ) nudum (Meunier) Balech
(after Wall and Dale 1968).
386
PALAEONTOLOGY, VOLUME 25
Remarks. The epithecal tabulation pattern shows all the usual features of this group; the ortho first
apical and the hexa dorsal intercalary arrangement. The cyst has been described by Wall and Dale
( 1 968) as being similar to those of P. ( P .) conicum but smaller (31.0 /x m, excluding spines as compared
to 50.0 /xm for P. (P.) conicum ), and possessing relatively longer spines with conical bases. The
archeopyle is formed by the displacement of the dorsal intercalary paraplate (2a). The attached
operculum carries two spines. The confusion in distinguishing this cyst from those of conicum in
palynological preparations has already been noted. Wall and Dale (1968) incubated two thecae from
cysts but some doubt in identification of the thecae is apparent from their Inudum designation.
Protoperidinium ( Protoperidinium ) pentagonum (Gran) Balech 1974
Text-fig. 22; Plate 39, figs. 7-11; Plate 42, fig. 8
Remarks. This species fits well into this group by virtue of its epithecal tabulation whereas its cyst,
described by Wall and Dale (1968), Bradford (1977), Reid (1977), and Harland (1977), is somewhat
different from those described above. It has in fact been described under the names Lejeunia
applanata by Bradford (1977) and Trinovantedinium capitatum by Reid (1977) and Harland (1977).
This cyst is quite distinct from others described within this group by virtue of its hyaline wall,
possession of short sutural and intratabular spines, and a broad hexa archeopyle. The cyst of P. (P.)
pentagonum was chosen by Reid (1977) as the type for his genus Trinovantedinium. This genus was
used to accommodate other cysts, especially those having thick brown walls, a peridinioid outline,
and standard hexa archeopyle. I prefer to restrict the genus Trinovantedinium to include the cyst of
P. ( P .) pentagonum and any similar forms (see Wall and Dale 1968, pi. 2, figs. 9-10 and 11-12 and
Bradford 1977), but not to include brown cysts with standard hexa archeopyles. The cyst
Sumatradinium hispidum (Drugg) Lentin and Williams could be related. Wall and Dale (1968)
incubated a cyst to establish the cyst/theca relationship for P. (P.) pentagonum. However, other
similar cysts produced thecae with minor differences such that a number of separate species or
varieties may be involved.
Cyst description. Pentagonal peridinioid cyst made up of autophragm, epicyst with apical boss, and sometimes
shoulders. Hypocyst has two slightly asymmetrical antapical horns or bulges. Paratabulation not easily
recognized except for the planar non-indented paracingulum marked by sutural line of processes, the deeply
indented parasulcus, and the pattern of intratabular processes defining paraplate areas. Processes are short, rigid
with hollow bulbous bases, fine, aciculate with bifid, capitate, or infundibular solid tips. Processes are both
sutural and intratabular but their dispositions are not sufficiently clear to reveal the paratabulation. Archeopyle
intercalary formed by the loss of 2a paraplate, operculum free, archeopyle is a broad hexa style (PI. 39, fig. 10).
Cyst maximum length ranges from 65-0 to 70-0 /im.
EXPLANATION OF PLATE 41
All figures are illustrated at a magnification of x 500 and were photographed in plain transmitted light unless
otherwise indicated.
Figs. 1-14. Protoperidinium ( Protoperidinium sect. Quinquecuspis) leonis (Pavillard) Balech. 1 , dorsal view with
archeopyle and operculum showing a clear apical tongue. 2, ventral view with deeply indented parasulcus
and somewhat discontinuous paracingulum, specimen MPK 1230. 3, ventral view with standard hexa
archeopyle, apical margin by transparency. 4, dorsal view, specimen MPK 2779. 5, ventral view with
deeply indented parasulcus and discontinuous paracingulum. 6, dorsal view, specimen MPK 2780. 7,
dorsal view showing posterior archeopyle margin almost affecting the paracingulum. 8, ventral view,
specimen MPK 278 1 . 9, ventral view. 1 0, dorsal view, specimen MPK 2782. 1 1 , ventral view with deep
parasulcus and discontinuous paracingulum. 12, dorsal view, specimen MPK 2783. 13, ventral view with
two flagellar parapores. 14, dorsal view showing standard hexa archeopyle with no apical tongue, specimen
MPK 2784.
PLATE 41
2
3
4
10
14
HARLAND, Protoperidinium
388
PALAEONTOLOGY, VOLUME 25
Protoperidinium ( Protoperidinium ) subinerme (Paulsen) Loeblich, III, 1969
Text-fig. 23; Plate 39, fig. 6
Remarks. The thecal tabulation reveals the ortho first apical plate and the dorsal epithecal style. The
cyst of this species has been described by Wall and Dale (1968) as being characterized by a relatively
wide and deeply indented cingulum. The cyst illustrated here (PI. 39, fig. 6) is believed to be
attributable to P. (P.) subinerme. Wall and Dale (1968) established the theca/cyst relationship of this
species although they recognized that the resultant theca from incubation was more elongate in the
polar direction than other thecae of this species.
Cyst description. Ovoidal to reniform cysts, usually with a polar compression due to the low cone morphology of
the epi- and hypocyst. Marked paracingular zone characterized by a broad circumferential band. Cyst smooth,
light brown in colour, and possessing an offset standard hexa intercalary archeopyle formed by the displacement
of ?2a paraplate. Operculum may be attached. Cyst diameter ranges from 45 0 to 60-0 ^m.
Discussion of the subgenus Protoperidinium
Within the para species of this subgenus the cyst morphology is of a peridinioid form with a reduced
hypocyst and an intercalary standard hexa archeopyle. Unfortunately only one cyst type is known
text-fig. 22 (left). Epithecal tabulation of Proto-
peridinium ( Protoperidinium ) pentagonum (Gran) Balech
(after Wall and Dale 1968).
text-fig. 23 (right). Epithecal tabulation of Proto-
peridinium (Protoperidinium) subinerme (Paulsen)
Loeblich, III (after Wall and Dale 1968).
EXPLANATION OF PLATE 42
All figures, except the stereoscan photomicrographs, are illustrated at a magnification of x 500 and were photo-
graphed in Nomarski interference contrast.
Fig. 1. Protoperidinium (Protoperidinium sect. Selenopemphix) conicum (Gran) Balech, specimen MPK 2949. A
small specimen at the lower end of the size range that could be confused with P. (P. sect. Selenopemphix) nudum
(Meunier) Balech.
Fig. 2. Indet. Protoperidinium cyst, specimen MPK 2950, with a well developed zigzag split archeopyle.
Figs. 3-6. Protoperidinium (Protoperidinium sect. Asymmetropedinium) punctulatum (Paulsen) Balech, specimen
MPK 2951, 2952 and 2953) respectively. Specimens showing the over-all cyst morphology and nature of
operculum, particularly its asymmetrical morphology.
Fig. 7. Protoperidinium (Protoperidinium sect. Quinquecuspis) leonis (Pavillard) Balech, specimen MPK 2954
showing cyst morphology and archeopyle, together with a continuous paracingulum, x e. 1000.
Fig. 8. Protoperidinium (Protoperidinium sect. Trinovantedinium) pentagonum (Gran) Balech, specimen MPK
2956, showing cyst morphology and the broad hexa archeopyle, x c. 1000.
Fig. 9. Indet. Protoperidinium cyst, specimen MPK 2957. Could be part of the cyst variation as currently
attributed to P. (P. sect. Quinquecuspis) leonis (Pavillard) Balech, x c. 1000.
Fig. 10. Protoperidinium (Protoperidinium sect. Selenopemphix) conicum (Gran) Balech, specimen MPK 2958,
showing over-all cyst morphology, x c. 1000.
PLATE 42
HARLAND, Protoperidinium
390
PALAEONTOLOGY, VOLUME 25
from this group at the present time. The designation Protoperidinium ( Protoperidinium sect.
Protoperidinium ) may be used for thecae with a para first apical and hexa dorsal intercalary
arrangement and for peridinioid cysts with reduced hypocysts. However, it must be admitted that the
type species does not appear to have fossilizable cysts.
In the ortho species we have most of the living cyst species including, in the quadra subsection, the
chordate cysts, with or without spines and having an archeopyle that truncates the apex, i.e. the
species P. (P.) claudicans and P. (P.) oblongum respectively. The designation Protoperidinium
{Protoperidinium sect. Votadinium) can be used for these ortho, quadra species.
Amongst the ortho penta species only the cyst of P. (P.) punctulatum is known, and like so many
Protoperidinium cysts it is a brown sphere. It is, however, characterized by its archeopyle which is
transversely elongate and asymmetrical. The name Asymmetropedinium is proposed at section level to
accommodate this species.
In the ortho hexa species a fourfold division based upon cyst morphology is possible. The thecal
tabulation does not in itself support this division but perhaps other factors need to be considered. The
first division is represented by P. (P.) conicoides which has a spherical brown cyst with a standard
hexa archeopyle; the second by P. (P.) conicum, P. (P.) nudum, and P. (P.) subinerme that have cysts
commonly showing polar compression because of the low conate shape of the epi- and hypocyst, and
an offset standard hexa archeopyle. A third type represented by P. (P.) leonis contains brown
peridinioid cysts with standard hexa archeopyles and finally the fourth is characterized by P. (P.)
pentagonum with a hyaline peridinioid cyst with a broad hexa archeopyle. The names Brigantedinium
Reid, Selenopemphix Benedek, Quinquecuspis Harland, and Trinovantedinium Reid respectively are
used here to designate these dinoflagellates at section level and are names that were first established in
the palaeontological literature.
Unattributed ? Protoperidinium cysts
Remarks. There are many ? Protoperidinium cysts described in the literature that have been observed
in palynological assemblages of Recent sediments, but whose attribution to the thecate stage is
unknown. A number of such forms are illustrated and commented upon here. The first (PI. 38, figs.
10-12, and PI. 42, fig. 2) is in essence a brown spherical cyst that opens by means of a zigzag split.
Reid (1972) described such cysts from intertidal sediments around the British coast. It is interesting to
note that the brown spherical morphology is again apparent, but that the archeopyle morphology is
sufficiently different to differentiate these cysts from the other forms described earlier. It is, however,
quite possible that these cysts are from glenodiniacean dinoflagellates such as Diplopsalis (see Wall
and Dale 1986, pi. 4, fig. 20).
A second type, that has been described under the name of Xandarodinium xanthum Reid 1977, is
illustrated on PI. 39, figs. 4-5. This cyst has a unique morphology with hollow processes carrying
distal furcate and bifid tips. Reid (1977) describes these cysts as having a possible single paraplate
intercalary archeopyle, but an archeopyle has not been seen by me. The morphology in regard to the
ovoidal/reniform ambitus suggests a polar compression and hints at an attribution to sect.
Selenopemphix, but confirmation and details of archeopyle formation must await further study.
Artemisiocysta cladodichotoma Benedek may be related to the cyst form described here. This cyst
might be a representative of a gymnodinialean dinoflagellate (see Wall and Dale 1968, pi. 4, fig. 29).
The final type (PI. 40, figs. 1-8) has been referred to as Lejeunia paratenella Benedek 1972 by
Harland (1977) and Trinovantedinium olivum Reid 1977 by Reid (1977). It is characterized by its
peridinioid outline, pale brown colour, thickened apex and antapex, the latter often developed into
acicular horns, planar paracingulum delimited by denticulate sutures, and an attenuated hexa
archeopyle (Lentin and Williams 1975). This cyst differs, therefore, from those referred to as
Quinquecuspis, and I prefer to see it accommodated in Lejeunecysta Artzner and Dorhofer 1978.
Lentin and Williams (1975) have attributed a standard hexa archeopyle to Lejeunecysta (as Lejeunia),
but the nature of the archeopyle in the type species has not been clearly demonstrated. The specimen
figured by Benedek (1972) does, however, appear to have a standard hexa archeopyle which may
HARLAND: DINOFLAGELLATE CYSTS
391
indicate that Lejeunecysta is a senior of Quinquecuspis and that a new name needs to be erected for
these cysts. Until the thecate form is identified further speculation is unwarranted.
GENERAL DISCUSSION
A taxonomic system capable of amalgamating both cyst and thecal data must be one of the aims of
dinoflagellate research and would inevitably lead to a rationalization of much of our fossil data. The
exercise, outlined above, tests that possibility and shows this, or a similar system, is practical. It is
clear that both the major and minor divisions of the genus Protoperidinium Bergh are largely upheld
by differing cyst morphologies (text-fig. 24). The genus is basically characterized by cysts of spherical
to peridinioid shape, possessing an intercalary archeopyle. The subgenus Archaeperidinium is
characterized by such unique cysts as those of P. (A.) minutum, P. (A.) compressum, and simple brown
spherical cysts distinguished by broad hexa, transversely elongate, symmetrical archeopyles. Cyst
morphology can also be used to further subdivide the subgenus at section level. These subdivisions
need, however, further testing in relation to total dinoflagellate morphology.
Similarly the subdivisions of the subgenus Protoperidinium based upon thecal morphology appear
to be largely substantiated by differences in cyst construction. The para species may have cysts
with reduced hypocysts and no paracingulum (but the evidence is based upon one species only).
The different divisions of the ortho forms have the following morphologies; the quadra cysts are
■ chordate in shape with an archeopyle that is basically intercalary but truncates the apex and may
involve additional paraplates; the penta cysts are brown and spherical, with a broad hexa archeo-
pyle that is asymmetrical; and the hexa cysts are both spherical and peridinioid, with standard or
broad hexa archeopyles. In this last case the cyst morphologies may serve to further subdivide the
group.
Dale (1978) has pointed out that dinoflagellate cysts are often non-conservative, with large
differences in cyst morphology being reflected by small ‘minor’ differences in thecae. This is supported
by the present study; indeed in the ortho hexa cysts large differences in cyst morphology are seemingly
not reflected in thecal morphology at all. In my view, major and minor differences in morphology are
relative terms. What should be remembered is that the genetic information is common to both stages
of the life cycle, and that all morphological differences must be critically evaluated. Dinoflagellate
taxonomy should be based upon the holomorph where possible or on as much information as is
available such that a common taxonomy can evolve. In this way it is necessary to evaluate both the
cyst-based taxonomy, often ‘overclassified’ and the thecate-based taxonomy, often ‘underclassified’
(Dale 1978), in order to arrive at the best amalgamation. Whilst agreeing with Dale (1978) that new
cyst-based nomenclature should not be created to artificially maintain the cyst-based taxonomy,
much useful information can be derived from good taxonomic work based upon cysts, thecae, or on
both.
From a review of this nature the major difference between the thecal-based taxonomy and the cyst-
based system is one of hierarchy (see text-fig. 24). The palynologist would regard his cyst taxa as
separate genera whereas the phycologist might be prepared to accept them at section level. Once this
hierarchical difference is recognized, the conflict between the two systems largely vanishes and both
theca and cyst data can point to areas where further study is needed to help resolve or enlarge upon
taxonomic decisions. In addition, it should assist the palynologist to view his fossil taxa in a better
perspective. It is clear, however, that the species is the common denominator, is constant, and should
be the basis of all our mutual research.
This study has also highlighted the ‘variation’ in thecal and cyst morphologies, especially in the
ortho quadra species, which appear to have both quadra and penta dorsal configurations as in P. (P.)
claudicans, the ortho hexa species such as P. (P.) leonis with both penta and hexa configurations, and
the ortho penta species such as P. (P.) punctulatum that have both penta and hexa dorsal epithecal
patterns. In the ortho quadra group the data also suggest variation in cyst morphologies, from
chordate cysts with archeopyles truncating the apex to cysts with a peridinioid shape and standard
hexa archeopyles. Recent studies have tended to show that the dorsal epithecal and epicystal pattern
392
PALAEONTOLOGY, VOLUME 25
Genus PROTOPERIDINIUM
text-fig. 24. Recent and Quaternary Protoperidinium cyst types and their respective positions in the taxonomic
subdivision of the genus.
is of considerable importance in dinoflagellate organization as both the keystone plate and overlap
system, together with archeopyle position, all relate to that pattern, and suggest a common genetic
control.
Some of the ‘variation’ discussed may in fact be more artificial than real, because much of it is
inextricably linked to taxonomic difficulties. The examples of P. ( P .) oblongum and P. (P.) leonis are
cases in point. Much of these difficulties will need detailed research on both thecal and cyst
morphologies, together with population studies, before any satisfactory conclusions can be reached.
There is a range of cyst types associated with a single living genus, i.e. from simple brown spheres to
peridinoid and quite complex stellate morphologies. The cyst types are, however, dominated by two
cyst morphologies — the simple brown spheres and the peridinoid cysts, the intercalary archeopyle is a
common factor throughout, although its shape, style, and involvement of adjacent paraplates is
different in the separate groups. The occurrence of the simple brown spherical cyst morphology in the
subgenera Archaeperidinium and Protoperidinium (and in both ortho penta and ortho hexa groups)
may be cited as a good example of the development of a common cyst form by different dinoflagellates
(homoeomorphy). Indeed this morphology is also seen in cysts of unknown affinity (see PI. 38, figs.
10-12) and in the glenodiniacean dinoflagellates (Reid 1977). There is difficulty in recognizing species
among these brown spheres because of a usual lack of well-preserved and well-orientated specimens.
In addition, brown peridinioid cysts also occur in more than one dinoflagellate group.
HARLAND: DINOFLAGELLATE CYSTS
393
It is interesting to observe the lack of correspondence between some dorsal epithecal tabulations
and the partial paratabulations revealed by archeopyle formation. In particular, the stellate cyst of P.
(A.) compressum appears to have a symmetrically placed type 21 archeopyle but not a symmetrical
dorsal epithecal tabulation pattern. Similarly P. (P.) conicum, P. (P.) nudum , and P. (P.) subinerme
have symmetrically placed 2a thecal plates, but offset type I archeopyles. Although the over-all
genetic control is apparent there does not seem to be a strict template mechanism. Whether or not
additional paraplates are involved in these cases in archeopyle formation, there is an obvious need for
caution. The mechanism and control of archeopyle formation is a fascinating question and is clearly
more complex than at first appears.
Unfortunately there are still a number of cysts, described in the literature, of possible
Protoperidinium affinity, whose ‘parental’ thecae are not known. One of these is the cyst referred to as
Lejeunia paratenella Benedek by Harland (1977). It would be of special interest to know the thecal
affinity of L. paratenella in order to compare it with other Protoperidinium species known to produce
Quinquecuspis type cysts. This should serve to help in understanding the nature of Lejeunecystaj
Quinquecuspis species in the fossil record.
Other fossil cysts may be underrepresented within this scheme because of their susceptibility to
oxidation in nature, in the palynological preparation technique, and to treatment with strong acids
(Dale 1976; Reid 1972). It is apparent from Lentin and Williams (1975) that the fossil record of the
genus Protoperidinium includes only those cysts attributable to the subgenus Protoperidinium,
including ortho forms, such as the fossil genera Rhombodinium Gocht and Wetzeliella Eisenack,
which have a geological record from the Palaeocene, and ortho forms such as Alterbia Lentin and
Williams and Deflandrea Eisenack, with records from at least the Albian. Other peridiniacean cysts
with more complex compound archeopyles are not represented here. The cysts portrayed within the
sections Selenopemphix and Quinquecuspis \Lejeunecysta have fossil records at least as far back as the
Late Eocene and Late Cretaceous respectively.
CONCLUSIONS
This study should have demonstrated the feasibility of combining cyst and thecate morphological
data into a sensible taxonomic scheme which might be of some application to both Quaternary and
some Tertiary peridiniacean dinoflagellates. It is, of course, not complete, because much more thecal
data are needed before a fully comprehensive taxonomy can be attempted. I hope it does suggest a
possible way forward and one that could be equally applied to gonyaulacacean, glenodiniacean,
ceratiacean, or pyrophacacaean dinoflagellates where and when sufficient evidence is available.
Notable points of interest have been the morphological range of cysts from a single living genus, and
the hierarchical difference between the sections erected herein and what would be undoubtedly
regarded as separate genera by palynologists. The species concept does, however, appear to be a
mutually acceptable base.
Acknowledgements. The author would like to acknowledge the various discussions with Dr J. P. Bujak,
Geological Survey of Canada, Mr B. Dale, University of Oslo, Dr P. C. Reid, Institute for Marine
Environment Research, Professor W. A. S. Sarjeant, University of Saskatchewan, and Dr L. E. Stover, Exxon
Production Research, on the nature and classification of Protoperidinium cysts. However, the views expressed
herein are solely his responsibility. I thank Barrie Dale, and Drs Evitt and Reid for their critical reading. This
paper was presented at the V International Palynological Conference in Cambridge, June 1980. Thanks are also
due to Mrs Jane Sharp for her excellent preparations and single grain mounting, to Mrs Sue Crook for her
draughting skills, and to Mrs Margaret Metcalfe for her typing. Latin translation of the various diagnoses are
by Mr John A. Rooney. This paper is published with permission from the Director, Institute of Geological
Sciences (N.E.R.C.).
REFERENCES
artzner, d. g. and dorhofer, G. 1978. Taxonomic note: Lejeunecysta nom. nov. pro. Lejeunia Gerlach 1961
emend. Lentin and Williams 1976 — dinoflagellate cyst genus. Can. J. Bot. 56, 1381-1382.
394 PALAEONTOLOGY, VOLUME 25
balech, E. 1974. El genero ‘Protoperidinium’ Bergh, 1881 (‘Peridinium’ Ehrenberg, 1831, partim.). Rev. Mus.
Cienc. natur., Hidrobiol. 4, 1-79.
benedek, p. n. 1972. Phytoplanktonten aus dem Mittel- und Oberologozan von Tonisberg (Niederrheingebeit).
Palaeontographica, Abt. B, 137, 1-71.
Bradford, m. r. 1975. New dinoflagellate cyst genera from the recent sediments of the Persian Gulf. Can. J. Bot.
53, 3064-3074.
— 1977. New species attributable to the dinoflagellate cyst genus Lejeunia Gerlach, 1961 emend. Lentin and
Williams 1975. Grana, 16, 45-59.
bujak, j. p. 1980. Dinoflagellate cysts and acritarchs from the Eocene Barton Beds of southern England. In
bujak, j. p., downie, c., Eaton, G. l. and williams, G. l. Dinoflagellate cysts and acritarchs from the Eocene of
southern England. Palaeontology , Spec. Pap., 24, 36-91.
dale, B. 1976. Cyst formation, sedimentation, and preservation: factors affecting dinoflagellate assemblages in
Recent sediments from Trondheimsfjord, Norway. Rev. Palaeobot. Palynol. 22, 39-60.
— 1978. Acritarchous cysts of Peridinium faeroense Paulsen: implications for dinoflagellate systematics.
Palynology, 2, 187-193.
dorhofer, g. and davies, E. h. 1980. Evolution of archeopyle and tabulation in Rhaetogonyaulacinean
dinoflagellate cysts. Life Sciences Miscell. Publ., Royal Ontario Museum, 1-91.
durr, G. and netzel, h. 1974. The fine structure of the cell surface in Gonyaulax polyedra (Dinoflagellata). Cell
Tiss. Res. 150, 21-41.
edwards, L. E. and bebout, j. w. 1981. Emendation of Phthanoperidinium Drugg and Loeblich 1967, and a
description of P. brooksii sp. nov. from the Eocene of the mid-Atlantic outer continental shelf. Palynology,
5, 29-41.
enrenberg, c. G. 1832 [separate 1830]. Beitrage zur Kenntniss der Infusorien und ihrer geographischen
Verbrietung besonders in Sibirien. Abh. preuss. Akad. Wiss. 1830, 1-88.
evitt, w. r. and davidson, s. e. 1964. Dinoflagellate Studies I. Dinoflagellate cysts and thecae. Stanford Univ.
Pubis., Geol. Sci. 10 (1), 1-12.
gocht, h. and netzel, h. 1974. Rasterelektronenmikroskipische Untersuchungen am Panzer von Peridinium
(Dinoflagellata). Arch. Protistenk. 116, 381-410.
— 1976. Reliefstrukturen des Kreide— Dinoflagellaten Palaeoperidinium pyrophorum (Ehr.) in Ver-
gleich mit Panzer-Merkmalen rezenter Peridinium- Arten. Neues Jb. Geol. Palaont., Abh. 153, 380-
413.
graham, H. w. 1942. Studies in the morphology, taxonomy, and ecology of the Peridiniales. Carnegie Institution
of Washington, 542, 1-129.
harland, r. 1977. Recent and Late Quaternary (Flandrian and Devensian) dinoflagellate cysts from marine
continental shelf sediments around the British Isles. Palaeontographica, Abt. B, 164, 87-127.
— reid, p. c., dobell, p. and norris, G. 1980. Recent and sub-Recent dinoflagellate cysts from the Beaufort
Sea, Canadian Arctic. Grana, 19, 211-225.
jorgensen, e. (1912) 1913. Bericht fiber die von der schwedischen Hydrographisch — Biologischen Kommission
in den schwedischen Gewassern in den Jahren 1909-1910 eingesammelten Planktonproben. Svenska
hydrogr. — biol. komm. Skr. 4, 1-20.
lebour, m. v. 1925. The dinoflagellates of northern seas. Mar. Biol. Assoc. U.K., Plymouth, 1-250.
lentin, J. k. and williams, G. l. 1975. A monograph of fossil peridinioid dinoflagellate cysts. Bedford Inst, of
Oceanogr. Rept. Bl-R-75-16, 1-237.
paulsen, o. 1931. Etudes sur le microplankton de la Mer d’Alboran. Trab. Inst, esp Oceanogr. 4, 5-
108.
reid, p. c. 1972. The distribution of dinoflagellate cysts, pollen and spores in Recent marine sediments from the
coast of the British Isles. Unpubl. Ph.D. thesis, Univ. of Sheffield, 1-273.
— 1977. Peridiniacean and glenodiniacean dinoflagellate cysts from the British Isles. Nova. Hedwigia, 24,
429-463.
— and harland, R. 1977. Studies of Quaternary dinoflagellate cysts from the North Atlantic. Am. Ass.
stratigr. Palynol., Contr. Ser. 5A, 147-169.
stafleu, F. A. et al. 1978. International Code of Botanical Nomenclature. Bohn, Scheltema & Holkema, Utricht,
1-457.
stover, l. E. and evitt, w. r. 1978. Analyses of pre-Pleistocene organic-walled dinoflagellates. Stanford Univ.
Pubis. Geol Sci. 15, 1-298.
wall, D. 1965. Modern hystrichospheres and dinoflagellate cysts from the Woods Hole region. Grana palynol. 6,
297-314.
HARLAND: DINOFLAGELLATE CYSTS
395
wall and dale, 1968. Modern dinoflagellate cysts and evolution of the Peridiniales. Micropaleontology, 14,
265-304.
Typescript received 12 January 1981
Revised typescript received 29 March 1981
REX HARLAND
Institute of Geological Sciences
Ring Road Halton
Leeds LSI 5 8TQ
APPENDIX
The appendix is used here to indicate new names, and names whose status has been altered. The order
adopted follows that presented in the main text.
Subgenus Archaeperidinium (Jorgensen) Balech 1974
Section Archaeperidinium Jorgensen stat. nov.
Type species. Protoperidinium ( Archaeperidinium sect. Archaeperidinium) minutum (Kofoid) Loeblich, III, 1969
Remarks. This name has been used at generic level by Jorgensen ( 1 9 1 3), at sub-generic level by Balech
(1974), and here is used at section level as one of the sections of the subgenus Archaeperidinium. It is
characterized by cysts with unique process structure and a symmetrical epithecal tabulation pattern.
Section Stelladinium Bradford 1975 ex Harland and Reid 1980 stat. nov.
Type species. Stelladinium reidii Bradford 197 5 = Protoperidinium ( Archaeperidinium sect. Stelladinium) cf.
compressum (Abe) Balech 1974.
Remarks. This name, originally erected as a genus, is altered in status to become a section of the
subgenus Archaeperidinium. The type species reidii is based upon a cyst holotype and not a thecate
form but the parent must be a species similar to P. compressum. It is characterized by stellate cysts and
an asymmetrical thecal tabulation.
Section Fuscusasphaeridium nom. nov.
Derivation of name. Latin: fuscus, brown, sphaera, ball.
Diagnosis. Spherical to spheroidal brown cyst of autophragm with a laterally elongate, symmetrical,
archeopyle formed by the loss of a single intercalary paraplate (2a). Margin of archeopyle often
shows its configuration with paraplates 3', 4', la, and 6". Operculum free.
Cista autophragmatis aurea quae forman habet sphericalem vel spheroidalem et latus elongatum,
symmetricalis est. Archaeopyla facta est uno paraplato intercalari amisso (? 2a). Margo archaeopylae saepe
configurationem suam ostendit cum paraplatis 3', 4' la et 6". Operculem liberum est.
Type species. Protoperidinium ( Archaeperidinium sect. Fuscusasphaeridium) avellana (Meunier) Balech
197 4 = Protoperidinium avellana Meunier 1919, pp. 56, pi. 18, figs. 37-41. Recent.
Remarks. The new section erected here accommodates those dinoflagellates possessing brown
spherical cysts with symmetrical, transversely elongate archeopyles and with slightly asymmetric
dorsal epithecal tabulations.
Subgenus Protoperidinium (Bergh) Balech 1974
Section Protoperidinium Bergh 1881 stat. nov.
Type species. Protoperidinium ( Protoperidinium sect. Protoperidinium) pellucidum Bergh 1881.
396
PALAEONTOLOGY, VOLUME 25
Remarks. Protoperidinium is used here at section level as a part of the subgenus Protoperidinium. It
may be characterized by cysts with inflated epicysts but this is as yet unproven since the type species
does not appear to have fossilizable cysts.
Section Votadinium Reid 1977 stat. nov.
Type species. Votadinium calvum Reid 1977 = Protoperidinium ( Protoperidinium sect. Votadinium ) oblongum
(Aurivillius) Balech 1974.
Remarks. This name is given the new status of section within the subgenus Protoperidinium, and is
characterized by chordate cysts with large dorsal intercalary archeopyles that truncate the apex
together with ortho quadra epithecal tabulation patterns.
Section Asymmetropedinium nov.
Derivation of name. Greek: asymmetro, without symmetry; ope, opening; and dinium of dinoflagellate affinity,
with reference to the asymmetrical archeopyle.
Diagnosis. Spherical to spheroidal, brown cyst of autophragm that possesses a laterally elongate
archeopyle which is asymmetrical about the dorso-ventral plane, formed by the loss of a single
intercalary paraplate (?2a). Margin of archeopyle may show its configuration with adjacent para-
plates. Operculum ?attached.
Cista aurea autophragmatis quae habet et formam sphericalem vel spheroidalem et latus elongatum
asymmetricalis est. Archaeopyla facta est uno paraplato amisso (?2a). Fieri potest ut margo archaeopylae
ostendat configurationem cum paraplatis quae proximae sunt. Operculum ? adiunctum.
Type species. Protoperidinium ( Protoperidinium sect. Asymmetropedinium) punctulatum (Paulsen) Balech
\91A = Peridinium punctulatum Paulsen 1907, p. 19, fig. 28. Recent.
Remarks. Asymmetropedinium is used here for ortho penta dinoflagellates that have brown spherical
cysts with transversely elongate archeopyles that are asymmetrical.
Section Brigantedinium Reid 1977 ex Harland and Reid 1980 stat. nov.
Type species. Brigantedinium simplex (Wall) Reid 1977 = Protoperidinium ( Protoperidinium sect. Brigantedinium )
conicoides (Paulsen) Balech 1974.
Remarks. This section is characterized by an ortho hexa epithecal configuration and brown spherical
cysts with a standard hexa archeopyle. Brigantedinium was originally coined at generic level by Reid
(1977) to accommodate all brown spherical cysts.
Section Selenopemphix Benedek 1972 stat. nov.
Type species. Selenopemphix nephroides Benedek 1972 = ? Protoperidinium ( Protoperidinium sect.
Selenopemphix) subinerme (Paulsen) Loeblich, III, 1969.
Remarks. Selenopemphix, originally erected as a genus by Benedek (1972), is here altered in status to a
section of the subgenus Protoperidinium to hold ortho hexa dinoflagellates whose cysts show polar
compression and offset archeopyles. The type species, the fossil S. nephroides Benedek, is very similar,
if not identical to the cyst of P. ( P .) subinerme (Paulsen).
Section Quinquecuspis Harland ex et emend. Harland stat. nov.
Type species. Quinquecuspis concretum (Reid) Harland 1977 =1 Protoperidinium ( Protoperidinium sect.
Quinquecuspis) leonis (Pavillard) Balech 1974.
Emended diagnosis. Pentagonal, peridinioid, acavate, brown cyst, made up of autophragm which
thickens towards the apex and antapex. Epitract conical or with shoulders, hypotract with well or
HARLAND: DINOFL AGELL ATE CYSTS
397
poorly developed asymmetrical horns. Paratabulation absent, paracingulum conspicuous, planar,
and delimited by low continuous or broken ridges. Deeply indented parasulcus sometimes with
flagellar scars. Archeopyle intercalary, standard hexa, formed by the loss of paraplate 2a, with or
without an apical tongue. Operculum free.
Cista aurea pentagonalis peridinoidalis et acavata, ex autophragmate facta quod densum fit versus apicem et
antapicem. Epitractum habet vel formam conicalem vel humeros, hypotractum habet cornua asymmetricalia
quae bene vel male formata sunt. Paratabulatio nulla est sed paracingulum et videri potest et planare finitum est
rugis quae vel imae vel fractae sunt. Parasulcus alte dentatum cum cicatricibus flagellaribus. Archaeopyla
intercalaris, hexa normalis quae facta est amisso paraplato 2a, cum vel sine lingua in apice. Operculum liberum.
Remarks. This name was originally erected at generic level (Harland 1977) but did not comply with
Article 36 of the International Code of Botanical Nomenclature (Stafleu et al. 1978) which requires
that Recent algal taxa be furnished with a Latin diagnosis. The type species had a Latin diagnosis in
its original publication (Reid 1977) and was valid from that date. As a section of the subgenus
Protoperidinium it is characterized by an ortho hexa epithecal tabulation and peridinioid brown cysts
with a standard or modified standard hexa archeopyle.
These cysts are noted in particular for their thick brown walls, and archeopyles often having an
apical tongue and deeply indented sulcal areas. During archeopyle formation a break often occurs
down across the paracingulum (see PI. 41, figs. 4, 7) from the posterior archeopyle margin.
Section Trinovantedinium Reid 1977 stat. nov.
Type species. Trinovantedinium capitatum Reid 1977 = Protoperidinium ( Protoperidinium sect. Trinovantedinium)
pentagonum (Gran) Balech 1974.
Remarks. Trinovantedinium , a genus erected by Reid (1977), is herein given the new status of section
within the subgenus Protoperidinium. It accommodates dinoflagellates with ortho hexa epithecal
tabulations and hyaline peridinioid cysts with short spines and a broad hexa archeopyle.
A NEW GENUS OF SHARK FROM THE
MIDDLE TRIASSIC OF MONTE SAN GIORGIO,
SWITZERLAND
by O. RIEPPEL
Abstract. Associated material from the Middle Triassic of Monte San Giorgio, Kt. Tessin, Switzerland,
demonstrates that the sharks Nemacanthus tuberculatus and Acrodus bicarinatus constitute a single taxon that
must be included in a new genus, Acronemus tuberculatus. Finspine structure of Acronemus is distinctly
different from that of Nemacanthus monilifer, but does correspond to the general ctenacanthiform pattern.
Other features of A. tuberculatus such as tooth structure and placoid scales have previously been reported
for Triassic hybodontiform sharks only. A discussion of the orders Ctenacanthiformes and Hybodontiformes
concludes the study.
The Grenzbitumenzone of Monte San Giorgio, Kt. Tessin, Switzerland, has so far yielded four
hybodontiform shark genera, viz. Hybodus, Acrodus , Aster acanthus, and Palaeobates. A fifth
selachian taxon, the most frequently found one in the deposits at Monte San Giorgio, will be
described in the present contribution. This shark is important as it highlights some problems in
the distinction of the Mesozoic orders Hybodontiformes and Ctenacanthiformes as defined by
Maisey (1975).
Kuhn (1945) first described the genus Acrodus from the Middle Triassic of Monte San Giorgio.
Kuhn’s (1945) specimen d was not fully prepared at the time. The radiograph showed, however,
that the proportions of the finspines of this shark were very different from those of Acrodus, and
that the palatoquadrate was cleaver-shaped (Kuhn, 1945, fig. 4), a feature which is otherwise not
known in the genus Acrodus. Peyer (1957) mentioned an Acrodus from Monte San Giorgio with a
tuberculate finspine ornamentation. Acrodus finspines always show a costate ornament.
The misidentifications by Kuhn (1945) and Peyer (1957) are based on the fact that in the specimens
mentioned by these authors typical Acrodus teeth are associated with stout and tuberculate finspines.
The teeth and the finspines have been named and referred to different genera by earlier authors.
Bellotti, in an unpublished manuscript on the fossil fishes at the Museo Civico di Milano (1873),
had named the teeth Acrodus bicarinatus and the spines Nemacanthus tuberculatus. The first author
to use these names in a formal publication was Bassani (1886). He used the names in connection
with a valid diagnosis, referring to Bellotti’s specimens and manuscript. According to the kind
information provided by Professor G. Pinna, Museo Civico di Storia Naturale Milano, the entire
old collection on which Bellotti’s research was based has been destroyed during the Second World
War. No type material has been preserved. Likewise, Bellotti’s (1873) manuscript can no longer
be located. The only material still available is unpublished drawings by Bellotti including those of
N. tuberculatus.
The associated material from Monte San Giorgio demonstrates that A. bicarinatus and N.
tuberculatus are a single taxon. Since Bassani (1886) is the formal author of both names, and since
in his publication N. tuberculatus has page priority, the correct species name for this shark must
be tuberculatus. The description of the Monte San Giorgio material will make it clear that this
shark species cannot be included in the genus Nemacanthus, nor in the genus Acrodus. A new genus
must consequently be erected. Since no original type material is preserved, it is appropriate to
select a neotype for that species.
IPalaeontology, Vol. 25, Part 2, 1982, pp. 399-412, pi. 43.|
400
PALAEONTOLOGY, VOLUME 25
SYSTEMATIC PALAEONTOLOGY
Class CHONDRICHTHYES
Subclass ELASMOBRANCHII
Order ctenacanthiformes incertae familiae
Genus acronemus n.gen.
Type and only known species. Acronemus tuberculatus (Bassani 1886).
Revised diagnosis. Small (30-35 cm long) ctenacanthiform shark; teeth Acrodus-Wke, with a single
and blunt main cusp, crown ornamented with prominent longitudinal and transverse carinae, root
without lingual torus; palatoquadrate cleaver-shaped; finspines short and stout, heavily tuberculated
with a longitudinal ridge along the leading edge of the crown, not or only slightly recurved,
posterior wall concave without denticulate ornamentation, central cavity displaced posteriorly;
placoid scales with a lanceolate crown ornamented with three or five longitudinal striae.
Acronemus tuberculatus (Bassani 1886)
Selected synonymy
1886 Nemacanthus tuberculatus, Bassani, p. 30.
1886 Acrodus bicarinatus, Bassani, p. 31.
1891 Nemacanthus tuberculatus. Woodward, p. 117.
1910 Acrodus bicarenatus, Alessandri, p. 34, pi. 7, figs. 6-9.
1910 Nemacanthus tuber colatus, Alessandri, p. 36, pi. 7, fig. 10.
Neotype. Palaontologisches Institut und Museum der Universitat Zurich T 1548, Monte San Giorgio, point
902, layer 118, collected 20.9.1957.
Diagnosis. Same as for genus.
Referred material. All specimens with T-numbers belong to the Tessin collection, Palaontologisches Institut
und Museum der Universitat Zurich. T 1177, dentition and two finspines; T 1178, dentition, two finspines,
skull fragments; T 1 181, isolated tooth; T 1289, dentition, anterior finspine; T 1291, dentition, anterior finspine;
T 1292, isolated tooth; T 1426, isolated tooth; T 1427, isolated tooth; T 1431, sectioned finspine; T 1434,
isolated tooth; T 1448, dentition; T 1457, teeth, sectioned finspine; T 1487, dentition; T 1531, dentition and
two finspines; T 1551, incomplete finspine; T 2465, dentition, upper and lower jaws, two finspines (see Kuhn
1945, fig. 4); T 3297, isolated tooth; T 3812, dentition; T 3818, two finspines; T 3819, dentition; T 3820,
dentition; T 3821, dentition and jaw fragments; T 3822, dentition and anterior finspine; T 3827, dentition;
T 3825, dentition and two finspines; T 3826, dentition; T 3828, posterior finspine; T 3829, dentition and anterior
finspine; T 3831, dentition; T 3833, dentition; T 3834, dentition; T 3835, dentition and finspine; T 3836,
dentition; T 3837, dentition; T 3840, sectioned finspine; T 3841, anterior finspine; T 3843, sectioned finspine;
T 3844, finspine; T 3845, finspine; T 3846, sectioned finspine; T 3847, sectioned finspine; T 3849, fifty-two
isolated teeth; British Museum (Natural History) P 19450, complete dentition.
Distribution. Middle Triassic of Southern Alps (Grenzbitumenzone of Monte San Giorgio, Kt. Tessin,
Switzerland, and the same beds near Besano, Lombardy, Italy).
Description
Neurocranium and jaws. Virtually nothing is known of the neurocranium. Remains are observed in T 1548
(lateral view, text-fig. 1) and in T 2465 (dorsal view, text-fig. 2). There is a very prominent postorbital process
which lies just in front of the otic process of the palatoquadrate. The rostrum appears to have been short
and blunt, projecting little beyond the suborbital ramus of the palatoquadrate.
The palatoquadrate (text-figs. 1 and 2) shows a large postorbital ramus (slightly more than half the total
length of the palatoquadrate) with a prominent otic process. The latter articulates with the postorbital process
of the neurocranium. The anterior edge of the otic process is steeply inclined. The posterior edge is thickened
and thus forms a rim which limits the area of origin of the m. adductor mandibulae externus. The lower end
of the thickened posterior edge forms the articular condyle of the palatoquadrate which fits into a facet on
the posterior end of Meckel’s cartilage. The narrow, tapering suborbital ramus makes up slightly less than
half of the total length of the palatoquadrate. A weak elevation of its dorsal edge in its anterior portion
(text-fig. 2) forms the orbital process which articulates with the suborbital shelf of the neurocranium.
RIEPPEL: NEW TRIASSIC SHARK
401
text-fig. 1. Skull remains in Acronemus tuberculatus T 1548. Abbreviations: crt,
ceratohyal; hy, hyomandibula; lc, labial cartilages; Me, Meckel’s cartilage; nc, neuro-
cranium; po, postorbital process; pq, palatoquadrate. Scale equals 10 mm.
Meckel’s cartilages (text-figs. 1 and 2) are elongated and moderately deep. The ventral margin is convex
and slightly thickened. The thickened ventral rim limits the site of insertion of the m. adductor mandibulae
externus. The dorsal edge is concave except for the posterior portion where the dorsal edge is straight.
All the Monte San Giorgio material is strongly compressed, but from a thickening it appears that both
the dorsal edge of Meckel’s cartilage as well as the ventral edge of the palatoquadrate formed an outwardly
turned shelf for the support of the teeth.
Labial cartilages. Only broken fragments are observed in T 1548 (text-fig. 1) along the ventral edge of Meckel’s
cartilage.
Hyoid arch. The articulation of the hyomandibula with the ceratohyal is preserved just posterior to the jaw
articulation in T 1548 (text-fig. 1). The jaw suspension in a shark with a cleaver-shaped palatoquadrate may
be assumed to have been amphistylic (Schaeffer 1967).
Cephalic spines. These are not recorded in any specimen.
Teeth. The teeth of Acronemus tuberculatus have long been known under the name of Acrodus bicarinatus
(Bassani 1886; Alessandri 1910). The teeth are characterized by a single, blunt main cusp. No accessory lateral
cusps are developed (PI. 43, fig. 1). There is a marked longitudinal and a marked transverse ridge across the
crown. The ridges meet at right angles on the apex of the main cusp (text-fig. 3 and PI. 43, fig. 3). From the
longitudinal and transverse ridges fine striae radiate towards the edges of the crown. The main cusp forms
a bulbous lingual projection which overlaps the next inner tooth of the same tooth family. This results in a
supporting mechanism for the outer functional tooth very similar to the one observed in A. lateralis.
The enameloid layer covering the crown is of the single-crystallite type (PI. 43, fig. 4) as defined by Reif
(1973), without a superficial shiny layer. The root bears no expanded lingual torus. Root foramina were very
difficult to observe because the bituminous matrix does not allow chemical preparation to expose these foramina.
402
PALAEONTOLOGY, VOLUME 25
The large lateral teeth are symmetrical, with the main cusp in a fairly central position. The crown is high
and the lower edge of the root is distinctly concave (text-fig. 3). Towards the symphysis the teeth become
progressively smaller and the crown lower and asymmetrical in that the main cusp is shifted towards the
distal side of the crown. The lower edge of the root becomes less concave or even straight (text-fig. 3). Distal
to the large lateral teeth there is a series of very low but very elongated and distinctly asymmetrical teeth
(text-fig. 3) with the main cusp displaced towards the mesial side of the tooth. The lower edge of the crown
is almost straight. Again, the teeth diminish in size towards the distal end of the jaw.
Pectoral girdle. The left pectoral girdle is preserved in T 1 548 (text-fig. 4). Its structure clearly resembles that
of Acrodus and Hybodus (Koken 1907). The suprascapular portion is an elongated, spinous structure curved
in an anterior direction. The scapulocoracoid and the ventral coracoid bar are poorly preserved. Below the
pectoral girdle, a single basal element of the pectoral fin is preserved (text-fig. 4). The incompleteness of the
specimen does not allow the reconstruction of the basal skeleton of the pectoral fin.
Finspines. The finspines of Acronemus (syn. Nemacanthus tuberculatus) are relatively short and broad which
results in characteristic stout proportions. The anterior finspines are relatively longer and relatively broader
than the posterior finspines (text-fig. 5; PI. 43, figs. 6, 7). The crown of the anterior finspine is frequently of
a broad-based, upright triangular shape (text-fig. 6a). However, it may become somewhat elongated and
recurved to a variable degree (text-fig. 5; PI. 43, figs. 6, 8). The posterior finspines are relatively smaller and
narrower with an upright crown that is never distinctly curved in a posterior direction.
The most characteristic feature of the spines is their conspicuous tuberculation. Large and rounded tubercles
are arranged in a regular pattern on which it is possible to superimpose straight longitudinal lines. It is equally
possible to superimpose on the pattern of tuberculation curved and obliquely oriented lines which run in a
posterodorsal direction across the lateral surface of the crown (text-fig. 6a). There is in some spines the tendency
of the tubercles to fuse into segments of ridges running in a curved posterodorsal direction across the lateral
surface of the crown (text-fig. 6; PI. 43, figs. 6, 7). The arrangement of the tubercles along curved, obliquely
RIEPPEL: NEW TRIASSIC SHARK
403
text-fig. 3. The teeth of Acronemus tuberculatus. a: T 3821, articulated part of dentition (surface abraded).
Scale equals 5 mm. b-g: T 3849, isolated teeth. B-c, large lateral teeth in lingual, lateral, and occlusal views;
d-e, successive mesial teeth in lingual view; f-g, successive distal teeth in lingual view.
oriented lines reflects the growth pattern of the spine. The completely closed enameloid mantle of euselachian
finspines frequently shows similarly oriented growth lines (Maisey 1977).
In some specimens the tubercles are small and scattered irregularly (text-fig. 6b). The scattered arrangement
is probably due to wear, but the smaller size of the tubercles results in a larger number of longitudinal rows
counted across the base of the crown. Also, the anterior finspines have a larger number of longitudinal rows
of tubercles than the posterior finspines since the anterior spines are relatively broader. On the anterior
finspines there are from 10 to 23 longitudinal rows of tubercles counted at the base of the crown, the average
ranging from 13 to 15 rows. On the posterior finspines there are usually 13 rows of tubercles, variation ranging
from 9 to 13 rows.
Along the leading edge of the crown the finspines bear a more or less conspicuously developed longitudinal
ridge. In T 1289 this ridge, as far as preserved, retains an enameloid covering. The anterior ridge results from
the fusion of a single anteromesial row of tubercles. This is demonstrated by the finspines of T 3825 in which
this fusion is incomplete. The anterior ridge usually extends ventrally along the leading edge of the finspine
beyond the lowermost anterior tubercles up to close to an anterior projection formed by the upper part of
the root. This anterior projection indicates the level to which the finspine was inserted in the epaxial trunk
musculature.
The posterior wall of the finspine is concave. The posterior opening of the central cavity extends upwards
to the level of the lowermost posterior tubercles. A longitudinal row of tubercles runs along the posterolateral
edge of the crown, but these tubercles do not differ in size or shape from those covering the lateral surface
of the crown.
The line which connects the lowermost tubercles along the anterior and posterior edges of the finspine is
chosen to represent the base of the crown. The angle 6 between the base of the crown and the normal on
the longitudinal axis of the finspine (text-fig. 5) ranges from 8 to 24 in the available spines.
Five finspines of Acronemus were sectioned at successive levels along their long axis. A composite and
slightly diagrammatic representation of finspine histology shown in text-fig. 7a may be compared to the
section figured in text-fig. 7b. There is a lamellar inner trunk layer which is only apically distributed, above
the posterior opening of the central cavity. The central cavity is generally small at this level, sometimes even
(ontogenetically?) reduced to the size of an ordinary vascular canal (T 1431, section Nr. 8). The lamellar
inner trunk layer is surrounded by a well vascularized outer trunk layer which also forms the entire root of
the finspine. The central cavity is displaced backwards which results in a thick anterior wall largely made up
by the vascularized outer trunk layer. The vascular canals and their surrounding denteons are rather small
and widely spaced which gives the outer trunk layer a relatively compact appearance. Towards the leading
edge of the spine a distinctly larger, longitudinally running vascular canal is observed lying straight in front
of the central cavity (text-fig. 7). The tubercles are formed by the mantle component, but no clearcut boundary
separates the mantle component from the outer trunk layer (for the terminology of finspine histology see
Maisey 1979).
404
PALAEONTOLOGY, VOLUME 25
text-fig. 4. Tentative reconstruction of Acronemus tuberculatus, based on T 1548. Approx. xO-4.
text-fig. 5. Variation in size and shape of the anterior (left) and posterior (right) finspines of Acronemus
tuberculatus. a, T 1178; B, T 2465; c, T 1548. Scale equals 10 mm. For further explanations see text.
text-fig. 6. Ornamentation of the anterior (left) and posterior (right) finspines of Acronemus tuberculatus.
a, T 3818; b, T 2465. Scale equals 10 mm.
Scales. Flank scales of the specimen T 1548 are shown on PI. 43, fig. 5. They are of the non-growing, placoid
type. They bear a recurved, lanceolate crown which is ornamented with three or five widely spaced longitudinal
striae.
COMPARISON WITH THE GENUS NEM ACANTHUS
The finspines of Acronemus tuberculatus were originally referred to the genus Nemacanthus Ag.
(Bellotti 1873, in Bassani 1886). Type species of the latter genus is Nemacanthus monilifer Ag., with
which comparison thus must proceed. N. monilifer is represented by finspines from the Rhaetic of
England which differ in several respects from those of A. tuberculatus. Assuming a relatively
406
PALAEONTOLOGY, VOLUME 25
text-fig. 7. Finspine histology of Acronemus tuberculatus. a. Semidiagrammatic representation of
a transverse section through the crown of a finspine, based mainly on section T 3847/7. Abbreviations:
cc, central cavity; lvc, longitudinal vascular canal, b. Ground section Nr. T 3847/7. Approx, x 30.
constant structure of the finspines within shark genera, this justifies the erection of a new genus
to include the spines of tuberculatus (Maisey, pers. comm.).
The finspines of N. monilifer (comparative material used for this study is housed in the British
Museum (Natural History) and is listed in the Appendix) are long and slender in comparison to
Acronemus. They are regularly curved in a posterior direction. If the total height of the finspines
is divided by their maximum width, the values obtained for two complete specimens of N. monilifer
are 8-56 (BM(NH) P.8328) and 9-3 (BM(NH) P.46830). Nine anterior finspines of A. tuberculatus
have a corresponding mean value of 3T5, while six posterior finspines have a mean value of 4T.
In both genera the finspines bear an anterior enamelled ridge as well as a tuberculate mantle
ornament, but the tubercles are fewer in number in N. monilifer. In the latter genus one usually
RIEPPEL: NEW TRIASSIC SHARK
407
text-fig. 8. Semidiagrammatic sections through the finspines of chimaeroids and selachians, a, Leptacanthus
longissimus, BM(NH) Egerton coll.; b, Ctenacanthus angustus, BM(NH) P.9581; c, Nemacanthus monilifer,
BM(NH) P.15497, section 10; d, Palaeobates keuperinus, BM(NH) P.7604; e, Hybodus lawsoni, BM(NH)
P.2174a; f, Hybodus acutus, BM(NH) P.6157. Scale equals 5 mm.
counts less than ten rows of tubercles across the base of the crown, variation ranging from 7 rows
in BM(NH) P.46830 to 1 1 rows in BM(NH) P.2852, whereas in A. tuberculatus one usually counts
more than 10 rows (9-23) of tubercles across the base of the crown. The connection of the lowermost
tubercles at the anterior and posterior margins of the finspine results in a line which slants much
more steeply in a posterodorsal direction in Nemacanthus as compared to Acronemus. This line
intersects the normal on the long axis of the spine at an angle 9 (text-fig. 5) which in N. monilifer
ranges from 60° (BM(NH) P.46830) to 78° (BM(NH) P.51433), but from 8-24° in Acronemus. A
function of this angle is the fact that the tubercles approach the posterior edge of the spine only
towards its apex in N. monilifer. The extent of tuberculation in N. monilifer may either be due to
ontogenetic changes or to wear. In the specimen BM(NH) P.8328 isolated rudiments of a few worn
or resorbed tubercles can be observed on the posteroventral part of the lateral surface of the crown,
an area which lies well below the regularly tuberculated part of the crown but above the ventral
end of the enamelled anterior ridge. The angle 9 might be an indication of the degree of posterior
inclination of the finspine insertion relative to the long axis of the body. Specimen BM(NH) P.8328
suggests that the degree of posterior inclination of finspine insertion increased during ontogeny in
N. monilifer.
In both genera the posterior wall of the finspine is concave. In Acronemus longitudinal rows of
unmodified tubercles run along the posterolateral edges of the crown. In some specimens of N.
monilifer rows of small but pointed denticles are observed running along the posterolateral edges
of the crown (Maisey 1977, PI. 1. fig. d, and specimens BM(NH) P.1882, P.51433). Such
posterolateral rows of denticles are also observed in some specimens of the genus Ctenacanthus
(BM(NH) P.2525, P.2529).
The histology of the finspines is basically similar in Acronemus and Nemacanthus , except for the
detailed histology of the thick anterior wall. In N. monilifer the vascular canals are large and closely
juxtaposed, which results in the characteristic open spongy texture of the anterior wall (Maisey
1977, pi. 2, fig. c; Stromer 1927, text-fig. 12). This contrasts with the much more compact texture
of the anterior finspine wall in Acronemus.
DISCUSSION: THE CLASSIFICATION OF ACRONEMUS
Maisey (1975) subdivided the phalacanthous sharks ( sensu Zangerl 1973) into three groups of
ordinal rank, the Hybodontiformes, the Ctenacanthiformes, and the Euselachiformes. The
distinction of these three groups is based mainly on a detailed study of finspine structure. On the
basis of finspine structure alone, Acronemus clearly has to be classified with the Ctenacanthiformes
as defined by Maisey (1975). The structure of the palatoquadrate ties in well with such a conclusion.
408
PALAEONTOLOGY, VOLUME 25
It is cleaver-shaped in Acronemus as in some Palaeozoic ctenacanthiformes such as Goodrichthys
(Moy-Thomas 1936), whereas Hybodus and its allies show a reduction of the otic process. The
cleaver-shaped palatoquadrate probably is a primitive feature, however (Schaeffer 1967). Cephalic
spines are not reported in ctenacanthiform sharks or in Acronemus.
However, other features do not support the classification of Acronemus with the Ctenacanthi-
formes. The teeth of Acronemus are very similar to those of Acrodus, and they lack an expanded
lingual torus on the root. The teeth of Palaeozoic ctenacanthiforms are of the multicuspid cladodont
type with a lingual torus on the root (Maisey 1975). Which type of tooth was possessed by the
Lower Triassic (Stensio 1921, 1932) or Rhaetic Nemacanthus is controversial (Maisey 1977). No
associated Nemacanthus specimen has yet been found.
All Palaeozoic ctenacanthiform sharks have scales of a composite, growing type (Reif 1978).
Acronemus has placoid scales which are typical of pre-Rhaetian hybodontids (Reif 1978, p. 126),
as well as of euselachians.
These latter features of Acronemus might indicate that the ctenacanthiform and hybodontiform
sharks can only be recognized by their finspine structure. However, they might also indicate that
the distinction between the two orders is not as clearcut as it would appear (Schaeffer and Williams
1977). In fact, some mixture of hybodontiform and ctenacanthiform features was noted by Dick
(1978) in his description of Tristychius.
Among all those finspines examined by me (see Appendix) there is enough variability to
substantiate this point. The only ctenacanthiform features which are really constant throughout
the material examined are a flat or concave posterior wall and a posteriorly displaced central cavity
which results in a relatively thick anterior wall. In hybodontiform finspines the posterior wall may
be flat or convex to a variable degree, and the anterior wall of the finspine is never much thicker
than the posterior wall. This latter feature is a function of the posteriorly displaced central cavity
of ctenacanthiform spines, but the position of the central cavity may again be correlated at least
to some extent with the degree of convexity of concavity of the posterior wall. If the central cavity
is held at a constant distance from the leading edge of the spine it will appear in a central position
if the posterior wall is strongly convex, but it will appear in a posteriorly displaced position if the
posterior wall is strongly concave. The crown of the Hybodus finspine BM(NH) P.57794 is broken
at successive levels. Examination of the four levels of breakage showed that the anterior wall gets
relatively thicker towards the apex of the spine as the diameter of the central cavity diminishes.
A finspine of Nemacanthus monilifer (BM(NH) P. 1 5497) which was sectioned at successive levels
from top to bottom shows that the relative thickness of the anterior wall remains constant throughout
the length of the spine.
The anterior wall of the ctenacanthiform finspine is said frequently to be of an open spongy
texture (Maisey 1975). This in fact is only true for N. monilifer (and for the euselachian genus
Breviacanthus Maisey 1976). The anterior wall of the Ctenacanthus finspine has exactly the same
EXPLANATION OF PLATE 43
Acronemus tuberculatus
Fig. 1. Tooth T 3849 in lingual view, x 10.
Fig. 2. Tooth T 3849 in lateral view, x 10.
Fig. 3. Tooth T 3849 in occludal view, x 10.
Fig. 4. Tooth T 3849, cross-section etched with 2n-HCl for 5 sec. to show the single-cristallite enamel, approx,
x 150.
Fig. 5. Placoid scale from specimen T 1548, approx, x 130.
Fig. 6. Anterior finspine from specimen T 3818, x 1-8.
Fig. 7. Posterior finspine from specimen T 3818, x 1-8.
Fig. 8. Anterior finspine T 1289, x 1-25.
PLATE 43
RIEPPEL, Middle Triassic shark
410
PALAEONTOLOGY, VOLUME 25
vermiculate texture produced by relatively small vascular canals as is observed in typical
hybodontiform finspines. Acronemus has an even more compact finspine histology.
Hybodontiform finspines are said to show a concentric ring of longitudinally running vascular
canals around the central cavity (Maisey 1975, 1978). This is not always very clearly differentiated.
In addition, a comparison with chimaeroid finspines (see Appendix) as well as with acanthodian
finspines (Krebs 1960) indicates that concentrically arranged vascular canals are a primitive
character-state in selachians. However, in all the hybodontiform and ctenacanthiform finspines
examined by me there is a conspicuously larger longitudinal vascular canal which lies straight in
front of the central cavity.
An important difference between hybodontiform and ctenacanthiform finspines is the ornamenta-
tion of the posterior wall in hybodontiforms, with one or two longitudinal rows of denticles, said
to be absent in ctenacanthiforms (Maisey 1975). Hybodontiform finspines typically carry a double
or a single row of rather large denticles close to, or on, the midline of the posterior wall. Such
denticles indeed are absent in ctenacanthiforms. But in genera such as Ctenacanthus and
Nemacanthus (Maisey 1977, PI. 1, fig. d, and quotations above) rows of small denticles are arranged
along the posterolateral edges of the crown which has a concave posterior wall. Maisey (1975)
considers the denticles on the posterior wall of hybodontiform finspines as either highly modified
mantle components or as specialized scales secondarily fused to the finspine. The posterolateral
denticles of ctenacanthiform finspines are merely considered to represent ‘posterolateral rows of
pointed tubercles’ (Maisey 1977, p. 265). This distinction appears meaningless since it is impossible
to determine the morphogenetic status of the hybodontiform versus the ctenacanthiform denticles.
If the denticles are considered homologous, it is possible to claim that posterolaterally placed
denticles are a primitive character state while denticles which have shifted to a posteromesial
position represent an advanced character state. Outgroup comparison with chimaeroid finspines
(see Appendix) demonstrates that in this group there are usually two longitudinal rows of
posterolaterally placed denticles, arranged along the edges of a concave posterior wall (text-fig.
8a). One undescribed chimaeroid finspine (BM(NH) P.2850) shows a single row of denticles along
the midline of the concave posterior wall. This single row of denticles may well have originated
from double rows of posterolaterally placed denticles that have shifted towards the midline of the
posterior wall. A similar phenomenon has been noted for Hybodus finspines (Maisey 1978). A
single row of posteromesially arranged denticles appears to originate from the superposition of
two collateral rows of denticles on the midline of the posterior wall (text-fig. 8e-f). This point is
further corroborated by a consideration of the finspines of Palaeobates keuperinus (text-fig. 8d). In
fact, the structure of these finspines is perfectly intermediate between the ctenacanthiform and
hybodontiform types and thus illustrates the potential difficulties in recognizing the two orders.
In P. keuperinus the posterior wall of the finspine is weakly convex (as in some Hybodus finspines),
but it carries two posterolaterally placed rows of denticles (as in ctenacanthiforms). The central
cavity is small and somewhat displaced backwards as in ctenacanthiforms, but the posterior wall
still retains a considerable thickness although it is thinner than the anterior wall.
In summary it can be stated that hybodontiform and ctenacanthiform sharks together form a
group of selachians with a finspine structure involving an apically distributed inner trunk layer
surrounded by a trabecular outer trunk layer. The details of the histology of the outer trunk layer
are quite variable and only characteristic at the generic level. There is no clearcut boundary between I
the outer trunk layer and the mantle component. The mantle ornamentation may be tuberculate
or costate. As evidenced by the genus Acronemus, the orders Hybodontiformes and Ctenacanthi-
formes can at present be recognized on the basis of finspine structure alone. Ctenacanthiform
finspines show a flat or concave posterior wall with two rows of posterolaterally placed denticles
(primitive) and a posteriorly displaced central cavity. Hybodontiform finspines show a flat or
convex posterior wall with one or two rows of denticles close to or on the midline of the posterior
wall (derived). The central cavity is rather centrally positioned. However, intermediate forms such
as P. keuperinus and Tristy chius (Dick 1978) do occur.
On the basis of finspine structure, Acronemus can be classified as a ctenacanthiform shark. The
RIEPPEL: NEW TRIASSIC SHARK
411
posterior wall of the finspine is concave, and the central cavity displaced backwards. If such an
arrangement is accepted it must be concluded that neither the ‘hybodontiform’ tooth structure nor
the presence of placoid scales can be used to distinguish hybodontiform and ctenacanthiform
sharks from the Triassic.
Acknowledgements. I thank Dr. C. Patterson who received me at the British Museum (Natural History),
granting me free access to the collection of fossil fishes. Drs. C. Patterson, Chr. Duffin, and W.-E. Reif all
provided some opportunity to discuss the material presented in this study. Preparative and photographic
work was done by H. Lanz. The SEM-micrographs were taken at the Institut fur Pflanzenbiologie der
Universitat Zurich by U. Jauch. Text-fig. 5, 6, and 7a were prepared by O. Garraux.
REFERENCES
alessandri, d. de. 1910. Studii sui pesci triasici della Lombardia. Mem. Soc. ital. Sci. Nat., Milano, 7, 1-145.
bassani, fr. 1886. Sui fossili e sulfeta degli schisti bituminosi triasici di Besano in Lombardia. Atti Soc. ital. Sci.
Nat., Milano, 29, 15-72.
dick, J. R. F. 1978. On the Carboniferous shark Tristychius arcuatus Agassiz from Scotland. Trans. R. Soc.
Edinburgh, 70, 63-109.
koken, e. 1907. Ueber Hybodus. Geol. Palaeont. Abh. (n.f.), 5, 261-276.
krebs, b. 1960. Ueber einen Flossenstachel von Gyracanthus (Acanthodii) aus dem Oberkarbon Englands.
Eclogue geol. Helv. 53, 811-827.
kuhn, E. 1945. Ueber Acrodus- Funde aus dem Grenzbitumenhorizont der anisischen Stufe der Trias des Monte
San Giorgio (Kt. Tessin). Eclogae geol. Helv. 38, 662-673.
maisey, j. G. 1975. The interrelationships of phalacanthous selachians. N. Jb. Geol. Palaeont. Mh. 1975 (9),
553-567.
— 1976. The Middle Jurassic selachian fish Breviacanthus n.g. N. Jb. Geol. Palaeont. Mh. 1976 (7), 432-438.
— 1977. The fossil selachian fishes Palaeospinax Egerton, 1872 and Nemacanthus Agassiz, 1837. Zool. J. Linn.
Soc. 60, 259-273.
1978. Growth and form of finspines in hybodont sharks. Palaeontology, 21, 657-666.
— 1979. Finspine morphogenesis in squalid and heterodontid sharks. Zool. J. Linn. Soc. 66, 161-183.
moy-thomas, J. a. 1936. The structure and affinities of the fossil elasmobranch fishes from the Lower
Carboniferous rocks of Glencartholm Eskdale. Proc. zool. Soc., Lond. 1936, 761-788.
peyer, b. 1957. Ueber die morphologische Deutung der Flossenstacheln einiger Haifische. Mitt, naturf. Ges. Bern
(n.f.), 14, 159-176.
reif, w.-e. 1973. Morphologie und Ultrastruktur des Hai-‘Schmelzes’. Zoologica Scripta, 2, 231-250.
— 1978. Types of morphogenesis of the dermal skeleton in fossil sharks. Palaeont. Z. 52, 110-128.
Schaeffer, b. 1967. Comments on elasmobranch evolution. In gilbert, p. w., mathewson, r. f. and rall, d. p.
(eds.). Sharks, Skates and Rays. John Hopkins Press, Baltimore. Pp. 3-35.
— and williams, m. 1977. Relationships of fossil and living elasmobranchs. Amer. Zool. 17, 293-302.
stensio, e. 1921. Triassic fishes from Spitzbergen, Pt. 1. Adolf Holzhausen, Vienna, i-xxviii, 1-307, pis. 1-35.
1932. Triassic fishes from East Greenland, collected by the Danish expeditions in 1929-1931. Medd. om
Grenl. 83, 1-305.
stromer, e. 1927. Ergebnisse der Forschungsreisen Prof. E. Stromers in den Wiisten Aegyptens. 2. Wirbeltier-
Reste der Baharije-Stufe (unterstes Cenoman). 9. Die Plagiostomen, mit einem Anhang fiber kano- und
mesozoische Rfickenflossenstacheln von Elasmobranchiern. Abh. Bayr. Akad. Wiss., math.-natw. Abt. 31,
1-64.
woodward, a. s. 1891. Catalogue of the fossil fishes in the British Museum ( Natural History), Vol. 3. British
Museum (Natural History), London. Pp. i-xliv, 1-567, pis. 1-16.
zangerl, r. 1973. Interrelationships of early chondrichthyans. In greenwood, p. h., miles, r. s. and Patterson,
c. (eds.). Interrelationships of fishes. Zool. J. Linn. Soc. 53, Suppl. 1, 1 14.
O. RIEPPEL
Palaeontologisches Institut u. Museum
der Universitat Zfirich
Kfinstlergasse 16
CH-8006 Zfirich
Typescript received 7 January 1981 Switzerland
412
PALAEONTOLOGY, VOLUME 25
APPENDIX
All the comparative material used to study finspine structure is housed in the British Museum (Natural
History). The specimen numbers are as follows.
Sharks: Hybodus sp., P.47145, P.57794; acutus , P.6157; ensatus , P.2162; lawsoni , P. 2174a; minor , P.51443;
reticulatus , one uncatalogued specimen; Ctenacanthus sp., P.2525, P.28928-9. P.34865; angustus, P.9581,
P.9262; pustulatus, P.2529; Nemacanthus monilifer, P.1882, P.2852, P.2854, P.8328, P. 1 5497, P.46830, P.51433;
Palaeobates keuperinus, P.2161, P.7604.
Chimaeroids: Edaphodon sp., P.3097; Leptacanthus longissimus , Egerton coll.; Myriacanthus paradoxus,
P. 1736a, one uncatalogued specimen; undescribed genus, P.2850.
Ground sections — Sharks: Breviacanthus brevis, P.2851; Ctenacanthus angustus, P.9581; Lonchidion sp.,
CP. 12.9.64; Nemacanthus monilifer, P.2217, P.15497; Wracanthus, P.45768.1.
Chimaeroids: Callorhynchus callorhynchus, Zool. Dept. 1935.4.23.17; Deltoptychius armigurus, P.11358;
Edaphodon sp., P.11807; ? Erismacanthus sp., P.6257; Metopacanthus granulatus, P.43065; Myriacanthus
paradoxus, P.1736.
A NEW SPECIES OF THE FISH AMIA FROM THE
MIDDLE EOCENE OF BRITISH COLUMBIA
by MARK V. H. WILSON
Abstract. A new species of amiid fish is described from a semi-articulated skeleton found in Middle Eocene
freshwater shales of the Allenby Formation, Princeton Group, south-central British Columbia. The new
species is assigned to Amia because it lacks Kindleia specializations such as styliform teeth, and it shares skull
specializations with A. calva, A. scutata, and A. uintaensis. The new species is reconstructed as a deep-bodied
piscivore with large jaws and strong, sharp teeth. The holotype is the first identifiable skeleton to be found
among many amiid scales recovered from numerous fossil-fish assemblages in southern British Columbia and
northern Washington State.
This paper presents a description and partial reconstruction of a new species of amiid fish, based
on a single partially articulated skeleton and several disarticulated skull bones from Middle Eocene
freshwater shales in British Columbia. The significance of the discovery lies in the relatively complete
information obtainable about the anatomy of this fish, and the resulting implications for the
taxonomy of fossils of Amia elsewhere in North America.
Amia calva, the only living species of amiid, is confined to the fresh waters of south-eastern
North America, but Late Cretaceous and Tertiary records of amiids are widespread in central and
western parts of the continent (text-fig. 1), as well as in Europe and Asia. North American fossil
amiids were reviewed by Boreske (1974), who recognized three valid fossil species in a single genus
from the twenty-three species and seven genera previously described. Many of the rejected names
were based on poorly preserved material or on single vertebrae, and were considered to be junior
synonyms or nomina dubia. The three fossil species considered valid by Boreske are the Late
Cretaceous to Middle Eocene Kindleia fragosa, the Palaeocene to Early Oligocene A. uitaensis, and
the Oligocene A. scutata.
None of the previously known Eocene occurrences of Amia from British Columbia, Washington,
and Oregon had been identified to species, because they consisted primarily of undiagnostic scales
(Cavender 1968; Wilson 1977a, 1978a, 1979, 1980) which were nevertheless very common fossils.
These westernmost records seemed to reflect habitats similar to those preferred by other Amia
species including A. calva (Scott and Crossman 1973): warm, shallow, swampy conditions, as
evidenced by lithology and the associated fishes, insects, and plants (Wilson 1980).
The discovery of the Amia specimens described here resulted from work by palaeobotanists on
silicified plant fossils, preserved in an outcrop of alternating chert and coal layers in the Allenby
Formation (Boneham 1968; Miller 1973; Robison and Person 1973; Basinger 1976; Basinger and
Rothwell 1977). James Basinger discovered a trionychid turtle in shales immediately overlying the
chert. When I revisited the site in 1977 and 1978, I obtained the fish specimens described here,
along with coprolites containing fish bones, and disarticulated remains of small suckers
(Catostomidae: Amyzon sp.), and trout-perches (Percopsidae: Libotonius sp.). The fish occur in the
shales with carbonized plant fossils which include stems and twigs, dicotyledonous leaves,
taxodiaceous leafy shoots, seeds, ferns, and amber (Wilson 1980).
GEOLOGY AND AGE
The specimens were found in a hard black siliceous shale immediately overlying a 10-m-thick
outcrop of interbedded carbonaceous chert and coal which extends into the Similkameen River
IPalaeontology, Vol. 25, Part 2, 1982, pp. 413-424.|
414
PALAEONTOLOGY, VOLUME 25
text-fig. 1. Map of North America showing the fossil locality. Also
shown are Late Cretaceous through Oligocene sites from which
remains of Amia and/or Kindleia have been reported (modified after
Boreske 1974), together with the distribution of Recent Amia calva
(stipple, after Scott and Crossman 1973).
from its east bank, 8 -4 km south of Princeton, British Columbia (text-fig. 1). According to Boneham
(1968) the fossiliferous layers are 550 m above the Princeton -Black Coal of the Allenby Formation,
Princeton Group. The Allenby Formation consists of sedimentary rocks, deposited in fresh water,
and yields an abundant assemblage of fishes and insects at several localities in the Princeton Basin
(Wilson 1977a, 19776, 19786, 1980).
The age of the fossils is Middle Eocene, based on potassium -argon ages of approximately 47-50
million years for the Allenby Formation (Hills and Baadsgaard 1967), and on the occurrence of
the mammalian genus Trogosus elsewhere in the formation (Russell 1935). Other occurrences of
fossil amiid scales in British Columbia and in the Klondike Mountain Formation of Washington
State are of a similar geologic age (Wilson 1977a, 1978a, 1979).
MATERIALS AND METHODS
The Amia fossils occur as fractured bone, spread parallel to bedding planes. The hard and tough
matrix contains a faithful imprint of the original shape and ornamentation of the bones and,
therefore, the specimens were prepared by removing the fossilized bone with mechanical tools and
an ultrasonic probe. Casts of the resulting impressions were made in black latex. Measurements
were taken from the original shale moulds. For photographic purposes the black-latex casts were
coated with ammonium chloride. The reconstructions were made by drawing the outlines of the
bones at uniform scale, using a camera lucida attached to a Wild M8 stereomicroscope, and then
graphically assembling the bone outlines with regard to perspective distortions and suture outlines.
For dermal roofing bones of the skull, the external surfaces of the bones correspond to their
WILSON: NEW SPECIES OF THE FISH AMIA
415
ornamented portions. Osteological terminology follows Boreske (1974) except that, following Janot
(1967), the term ‘posttemporal’ is used in place of ‘suprascapular’ (Table 1). The abbreviation
UAVP designates that specimens are deposited in the Vertebrate Paleontology Collections,
Department of Geology, The University of Alberta.
table 1 . Abbreviations used in the figures
a
angular
es
extrascapular
P
pectoral fin
sc
supracleithrum
ao
antorbital
fr
frontal
pa
parietal
sm
supramaxilla
br
branchiostegal
h
hyomandibular
pop
preopercle
so
subopercle
c
cleithrum
io
infraorbital
ps
parasphenoid
sy
symplectic
ch
ceratohyal
la
lachrymal
pst
parasphenoid teeth
V
vomer
cot
coronoid teeth
m
maxilla
Pt
(dermo)pterotic
Vt
vomerine teeth
d
dentary
me
metacleithrum
r
rostral
vtb
vomerine tooth base
ds
dermosphenotic
n
nasal
s
posttemporal
e
endopterygoid
op
opercle
sa
surangular
SYSTEMATIC DESCRIPTION
Class OSTEICHTHYES
Order amiiformes
Family amiidae Bonaparte, 1837
Genus amia Linnaeus, 1766
Type species. Amia calva Linnaeus, 1766.
Amia hesperia sp. nov.
Text-figs. 2-6
Diagnosis. Deep-bodied Amia having square parietals; long frontals with shallow orbital excavation;
large nasals without anterior notch; large fourth infraorbital; deep maxilla, mandible, and opercle;
dentary with large teeth; lachrymal with posterior notch; parasphenoid with long posterior ramus
and short tooth patch; vomers with short tooth patch; vomers and coronoids with sharp, conical
teeth; and cleithrum with arms at obtuse angle.
Holotype. UAVP 14758 (text-figs. 2-5 and 6a, b), an almost complete, partially articulated fish in part and
counterpart with estimated total length about 55 cm, preserved as a mould in hard siliceous shale, and
collected by the author’s party in 1977.
Etymology. The specific epithet is from the Latin hesperius meaning ‘western’.
Locality and age. Ashnola chert site, 8-4 km S of Princeton, British Columbia (U.T.M. Grid Reference
10UFK783724), from the Middle Eocene Allenby Formation, Princeton Group.
Non-type material. UAVP 13801, a patch of scales and two branchial tooth plates; UAVP 13804 (text-fig. 6c),
a right dentary and maxilla; UAVP 13805, a right extrascapular (text-fig. 6f), right fourth infraorbital, and
left fifth infraorbital; UAVP 13806 (text-fig. 6d), a branchial toothplate; and UAVP 13812 (text-fig. 6e), a
right opercle. All of the above specimens were collected at the type locality in 1977, within a few centimetres
of the holotype.
Description. Unless otherwise indicated, the following description is based on the holotype, which shows the
dermal investing bones of the skull in the part (text-fig. 3) and many ventral and internal bones of the skull
in the counterpart (text-fig. 4). The postcranial skeleton is present but not well preserved (text-fig. 2). Non-type
material is limited, but corroborates conclusions based on the holotype.
Parietals are approximately square (text-fig. 3), and frontals are relatively long and narrow (width to length
ratio 0-41), with shallow orbital excavations (‘orbital concavity ratio’ of Boreske 1974 is 0117; ‘dermosphenotic
angle’ is 136°). The ‘parietal/frontal’ ratio is 0-37. Extrascapulars are decidedly wider laterally than medially
416
PALAEONTOLOGY, VOLUME 25
(text-figs. 3, 6f), dermopterotics substantially overlap the frontals laterally, and are tapered both anteriorly
and posteriorly. Nasals are large, not notched anteriorly, and fit the anterior outline of the frontals posteriorly.
The rostral is stout, and the ornamented area of the antorbitals is small.
Laterally, the lachrymal has a prominent posterior notch which fits the anterior end of the small second
infraorbital (text-fig. 3). The large fourth infraorbital is deep posteriorly, where it has an angular margin. It
tapers to the orbit, but forms less of the orbital rim than does the fifth infraorbital. The latter is about as
deep anteriorly as posteriorly, and is considerably smaller than the fourth infraorbital.
text-fig. 2. Amia hesperia sp. nov., holotype, part, UAVP 14758a, latex peel of whole fish.
The premaxilla is not preserved. The maxilla is a stout, deep bone with a marginal row of numerous, very
small teeth (text-figs. 3, 4). The posterodorsal margin of the maxilla has an elongate excavation which fits a
long, deep supramaxilla.
The mandible is deep and stout, with a steeply inclined posterior border formed by the angular and
surangular (text-figs. 3, 4), a broad coronoid process, and a dentary with a long row of large, pointed teeth
(text-figs. 3, 4, 6b). Proportions and shape of the dentary are more clearly shown in UAVP 13804 (text-fig.
6c), where, toward the front of the bone, the ventral margin is apparently angled ventrally. This is interpreted
as a medial deflection of the mandibular ramus as seen most noticeably in Kindleia fragosa (Boreske 1974,
figs. 16b, 18). The coronoids are not visible, but small, sharp teeth located internal to the dentary teeth, where
coronoid teeth would be expected, are visible in the holotype (text-fig. 6b).
The parasphenoid (text-fig. 4) is partially obscured by a branchiostegal, a ceratohyal, an angular, and the
vomers, but its proportions are evident. It has a relatively long posterior ramus, 0-84 times the length of its
anterior ramus. The ascending process of the right side is partially covered in the holotype by an angular, i
but is approximately perpendicular to the long axis of the parasphenoid. The posterior portion of the
parasphenoid tooth patch consists of many tiny denticles. Where it is not obscured by the ceratohyal (text-fig.
4), the anterior ramus of the parasphenoid appears devoid of denticles.
Vomers are elongate (text-fig. 4) but have short-toothed portions. Most teeth are broken, but unbroken
teeth together with some broken tooth tips indicate that the vomerine teeth are conical and sharply pointed
(text-fig. 6a). The endopterygoid (text-figs. 3, 4) is plate-like and bears numerous tiny denticles.
The opercular bones are deep relative to their length. The opercle in the holotype has a length to height
ratio of 0-90, and these proportions are also seen in UAVP 13812 (text-fig. 6e). The subopercle has an elongate
anterodorsal ramus. The preopercle is not well preserved (text-fig. 4), but seems similar to those of other
species of Amia.
The hyomandibular, preserved in lateral view (text-fig. 3), has a posteroventrally directed opercular process
and a moderately developed posterodorsal notch. The symplectic is partially visible in the holotype (text-fig.
WILSON: NEW SPECIES OF THE FISH AMIA
417
text-fig. 3. Amia hesperia sp. nov., holotype, part, UAVP 14758a, latex peel of skull. Abbreviations Table 1.
3). The ceratohyal (text-fig. 4) is typical of the genus, and bears branchiostegals which vary from narrow to
broad. Branchiostegal ornamentation is extensive and posterior margins of preserved branchiostegals are
rounded. Branchial arches are not preserved, but several tiny bones bearing numerous sharp, hollow, conical,
and slightly curved teeth (text-fig. 6d) were found in the shales close to the holotype. In view of the similarity
418
PALAEONTOLOGY, VOLUME 25
text-fig. 4. Amia hesperia sp. nov., holotype, counterpart, UAVP 14758b, latex peel of skull. Abbreviations
Table 1. The two areas outlined in white are enlarged in text-fig. 6a, b.
of teeth of these bones to teeth of the holotype, the bones are interpreted as branchial tooth plates of the
same species.
The dermal pectoral girdle is well preserved in the holotype (text-figs. 3-5). The cleithrum is like that of
other Amia species, but the ornamented area is more extensive than in A. calva. As well, the posterodorsal
WILSON: NEW SPECIES OF THE FISH A MIA
text-fig. 5. Amia hesperia sp. nov., holotype UAVP 14758. a. Reconstruction of para-
sphenoid and vomers in central view. b. Left hyomandibular, lateral view. c. Left ceratohyal.
D. Left dermal pectoral girdle in lateral view. Abbreviations Table 1.
ramus of the cleithrum forms an obtuse angle of about 120° with the an tero ventral ramus, as in K. fragosa
but not in A. calva or A. uintaensis where the rami are at right angles to each other (Boreske 1974, fig. 21).
The plate-like posterior portion of the cleithrum is broadly rounded. The posttemporal (text-fig. 3) is elongate,
and the posterolateral portion is ornamented. The pectoral fin has about sixteen rays and originates medial
to the postero ventral angle of the cleithrum (text-fig. 3). Only the general position of the pelvic fin is apparent
in the holotype.
The postcranial portion of the axial skeleton is poorly preserved (text-fig. 2), but a few general statements
are possible. Anterior trunk vertebrae are about four times as broad as they are long, and bear dorsal neural
facets and ventral aortal facets as in other Amia species (Boreske 1974, fig. 11). Ribs, neural and haemal
spines, hypurals, and pterygiophores are all slender, rod-like elements. Posterior caudal vertebrae are
diplospondylous. The dorsal fin originates approximately above the seventeenth trunk centrum, and the anal
fin consists of about nine rays.
Scales are thinner than those of Cretaceous and Palaeocene K. fragosa from Alberta (O’Brien 1969), thicker
than scales of A. calva, but similar to other Eocene Amia scales from British Columbia (Wilson 1977a). They
are mostly rounded apically and truncate basally, about two-thirds as wide as long, with an apical (posterior)
focus, a thick, horseshoe-shaped rim of smooth lamellar bone around the lateral and apical margins, and a
central area of somewhat thinner bone ornamented on the internal surface by small bumps. Externally the
circuli or ridges radiate to the margins from the focus, which is approximately triangular and has its apex
directed basally (anteriorly). In the focal area the circuli form a vermiculate pattern.
Reconstruction
The parasphenoid and vomers of A. hesperia are reconstructed in text-fig. 5a. Notable features
are the inferred short parasphenoid tooth patch, the ascending processes perpendicular to the
CO
420
PALAEONTOLOGY, VOLUME 25
cot
1 cm
text-fig. 6. Amia hesperia sp. nov. a. Detail of anterior portion of vomers of holotype, UAVP 14758b,
showing pointed vomerine teeth, b. Detail of dentary tooth row of holotype, UAVP 14758b, showing pointed
coronoid teeth, c. Right dentary and maxilla, lateral view, UAVP 13804. D. Branchial tooth plate, UAVP
13806. e. Right opercle, lateral view, UAVP 13812b. f. Right extrascapular, dorsal view, UAVP 13805c.
Abbreviations Table 1.
long axis of the parasphenoid, and the elongate vomers with their anterior one-third occupied by
sharp teeth.
Text-fig. 7 shows the skull, reconstructed in dorsal view, compared with partial reconstructions
of the skulls of A. calva, K. fragosa , A. scutata, and A. uintaensis. Notable are the elongate frontals
with shallow orbital excavations, relatively short parietals, tapered extrascapulars, long
WILSON: NEW SPECIES OF THE FISH AMIA
421
text-fig. 7. Comparison of dorsal skull reconstructions of North American amiids. a. Amia hesperia, sp.
nov., based on the holotype, UAVP 14758. b. A. calva. c. Kindleia fragosa. d. A. scutata. e. A. uintaensis. b-e
after Boreske (1974), not to uniform scale. Abbreviations Table 1.
dermopterotics, large nasals shown interdigitating with the frontals, and the over-all shape of the
skull roof, which is elongate but only slightly narrower at the rear of the orbit than it is at the
rear of the extrascapulars.
The skull of A. hesperia is reconstructed in lateral view in text-fig. 8. In this view the skull
appears relatively deep and large-jawed. Features that contribute to the impression of skull depth
are the large lachrymal, deep opercle and subopercle, deep maxilla and supramaxilla, deep mandible,
and large teeth in the dentary. The obtusely angled cleithrum is additional evidence of a deep
body. In view of the poor preservation of the postcranial skeleton, a reconstruction of the entire
fish is not presented here. However, the evidence of the dorsal fin apparently originating over the
seventeenth trunk centrum suggests that A. hesperia was long-bodied, rather than short and stout
as was K. fragosa (Boreske 1974, fig. 21).
DISCUSSION
In addition to the one Recent and three fossil species of Amia from North America that were
recognized by Boreske (1974), there are a number of nominal European and Asian fossil species
(Boreske 1974; Janot 1967). Fairly complete information is available about several of these, including
some which are most similar to the North American K. fragosa (A. kehreri Andreae, A. valenciennesi
Agassiz, A. munieri Priem, and A. russelli Janot), and others which are similar to A. uintaensis (A.
robusta Priem and possibly A. mongoliensis Hussakof). Janot (1967), Estes and Berberian (1969),
422
PALAEONTOLOGY, VOLUME 25
text-fig. 8. Comparison of lateral skull reconstructions of North American amiids. A. Amia hesperia, sp.
nov., based on the holotype, UAVP 14758. B. A. calva. c. Kindleia fragosa. D. A. scutata. E. A. uintaensis. b-e j
after Boreske (1974), not to uniform scale. Abbreviations Table 1.
WILSON: NEW SPECIES OF THE FISH AMIA
423
Boreske (1974), and Gaudant (1980) advocate synonymizing Kindleia Jordan with Amia Linnaeus.
The main distinguishing features of Kindleia are the presence of styliform teeth on the coronoids,
dermopalatines, and vomers; differences in proportions of some skull bones; and a shorter body
with less separation between the skull and the origin of the dorsal fin and between the insertion
of the dorsal fin and the caudal fin, and about twelve fewer trunk centra and eight fewer
monospondylous caudal centra (Boreske 1974). These authors cited intra- and interspecific
variability in Amia, and found the above differences insufficient grounds for separation into two
genera, especially considering that Estes and Berberian (1969) believe most features of Kindleia to
be primitive. Gaudant (1980), however, treats Kindleia as a subgenus of Amia, and cites several
derived features shared by K. fragosa and A. kehreri in support of this view. Gaudant’s list of
shared derived characters is very similar to the list of features (above) that Estes and Berberian
(1969) interpreted as mostly primitive. For example, Gaudant cites a relatively short, deep body;
an enlarged fourth infraorbital; short parietals; and styliform coronoid, palatal, and vomerine teeth.
I agree with Gaudant that these characters are mostly derived, but go further in favouring the
recognition of Kindleia as a genus distinct from Amia. To Gaudant’s list of shared derived characters
can probably be added the relatively short, wide frontal: the frontals of the geologically older but
closely related genera Urocles (Lange 1968) and Enneles (Santos 1960) are long and narrow. An
enlarged fourth infraorbital found also in A. hesperia and in Enneles (Santos 1960) should probably
be removed from the list of derived features of Kindleia.
Retention of Kindleia as a separate genus is also useful because most species share a number of
other similarities, some or all of which might be primitive. These include the deep orbital notch
in the frontal; the short vomers; the small supramaxilla; the narrow maxilla and mandible; the
short, truncated gular plate; and fewer than seventy-five vertebrae.
Boreske (1974) suggested that molluscs were a more important part of the diet for Kindleia than
for Amia. At some localities in the Palaeocene Paskapoo Formation of Alberta, Kindleia remains
are found with finely crushed mollusc shells concentrated in patches on bedding planes. To my
knowledge gut contents have not been observed in Kindleia specimens, but the explanation of the
styliform teeth as an adaptation to molluscivorous habits is a reasonable one. Perhaps the fish did
not completely swallow the shells with their contents, or perhaps the patches of broken shells
represent regurgitated gastric residues.
The new species described here clearly belongs with Amia. A. hesperia has none of the Kindleia
specializations listed above. It shares with some or all of the other well-known species of Amia
derived features such as a shallow orbital excavation in the frontal, a deep maxilla with elongate
supramaxilla, the lack of an anterior notch in the nasal, the presence of a posterior notch in the
lachrymal, and a tooth patch restricted to the front end of the vomer.
A. hesperia differs from valid North American species of Amia (Boreske 1974) in the following
features. Compared with A. uintaensis it has slightly shorter frontals, shorter parietals, unnotched
nasals, interdigitating nasals and frontals, posteriorly notched lachrymals, more angular fourth
infraorbitals, a relatively longer posterior parasphenoid ramus, parasphenoid ascending processes
more nearly at right angles to the long axis, shorter parasphenoid and vomerine tooth patches,
and more obtusely angled cleithra. Compared with A. scutata and A. calva it has longer frontals,
shorter parietals, interdigitating nasals and frontals, a shorter parasphenoid tooth patch, more
rectangular opercles, and more obtusely angled cleithra. It differs further from A. calva in having
longer pterotics, larger fourth infraorbitals, a relatively shorter posterior parasphenoid ramus,
ascending processes more nearly at right angles to the long axis, and more extensive ornament on
the cleithra.
A. robusta, a Palaeocene to Oligocene European species, is very similar to A. uintaensis. Compared
with A. hesperia it has more elongate frontals, more rectangular parietals, a longer parasphenoid
tooth patch, and more right-angled cleithra, judging by bones illustrated in Janot (1967).
Piscivorous habits for A. hesperia are implied by the large mouth and the large, sharp teeth.
This supports previous conclusions (Wilson 1980) based on association of amiid scales with
coprolites containing fish bones at many Eocene localities in British Columbia and Washington
424
PALAEONTOLOGY, VOLUME 25
State. I would predict that the amiid(s) occurring at these other localities will prove to be closely
related to or conspecific with A . hesperia. The geographic distribution of Eocene amiid species in
North America, with A. hesperia in the extreme west and A. uintaensis and K. fragosa in the
mid-west, is suggestive of geographic ranges separated by a north-south barrier such as a continental
divide similar to the one separating western and eastern species of fishes today.
Acknowledgements. I thank Drs. J. F. Basinger and W. N. Stewart for bringing the locality to my attention,
and L. A. Lindoe for field and laboratory assistance. Dr. B. G. Naylor made suggestions that resulted in
improvements in the manuscript. This research was supported by Natural Sciences and Engineering Research
Council of Canada operating grant A9180.
REFERENCES
basinger, j. f. 1976. Paleorosa similkameenensis, gen. et sp. nov., permineralized flowers (Rosaceae) from the
Eocene of British Columbia. Can. J. Bot. 54, 2293-2305.
and rothwell, G. w. 1977. Anatomically preserved plants from the Middle Eocene (Allenby Formation) of
British Columbia. Ibid. 55, 1984-1990.
boneham, R. F. 1968. Palynology of three Tertiary coal basins in south-central British Columbia. Ph.D. thesis,
University of Michigan, 105 pp.
bores ke, j. r. 1974. A review of the North American fossil amiid fishes. Bull. Mus. Comp. Zool. Harv. 146, 1-87.
CA vender, t. m. 1968. Freshwater fish remains from the Clarno Formation Ochoco Mountains of north-central
Oregon. The Ore Bin, 30, 125-141.
estes, R. and berberian, p. 1969. Amia (= Kindleia) fragosa (Jordan), a Cretaceous amiid fish, with notes on
related European forms. Breviora, 329, 1-13.
gaudant, j. 1980. Sur Amia kehreri Andreae (poisson Amiidae du Lutetien de Messel, Allemagne) et sa
signification paleogeographique. C.R. Acad. Sc. Paris, 290D, 1107-1110.
hills, l. v. and baadsgaard, h. 1967. Potassium-argon dating of some Lower Tertiary strata in British
Columbia. Bull. Can. Petrol. Geol. 15, 138-149.
janot, c. 1967. A propos des amiides actuels et fossiles. Colloques Int. Centr. Nat. Rech. Scient. Paris, 163,
139-153.
lange, s. p. 1968. Zur Morphologie und Taxonomie der Fischgattung Urocles aus Jura und Kreide Europas.
Palaeontographica, ser. A, 131a, 1-78.
miller, c. n. 1973. Silicified cones and vegetative remains of Pinus from the Eocene of British Columbia. Contr.
Mus. Paleontol. Univ. Mich. 24, 101-118.
o’brien, d. e. 1969. Osteology of Kindleia fragosa Jordan (Holostei: Amiidae), from the Edmonton Formation
(Maestrichtian) of Alberta. M.Sc. thesis. University of Alberta, 118 pp.
robison, c. r. and person, c. p. 1973. A silicified semiaquatic dicotyledon from the Eocene Allenby Formation of
British Columbia. Can. J. Bot. 51, 1373-1377.
russell, l. s. 1935. A Middle Eocene mammal from British Columbia. Am. J. Sci. 29, 54-55.
scott, w. b. and crossman, e. j. 1973. Freshwater fishes of Canada. Fish Res. Bd. Can., Bull. 184.
santos, r. da s. 1960. A posujao sistematica de Enneles audax Jordan e Branner da Chapada do Araripe, Brasil.
Monografias Div. Geol. Miner. Bras. 17, 1-25.
wilson, m. v. h. 1977a. Middle Eocene freshwater fishes from British Columbia. Life Sci. Contr., Roy. Ont. Mus.
113, 1-61.
— 1977ft. New records of insect families from the freshwater Middle Eocene of British Columbia. Can. J.
Earth Sci. 14, 1139-1155.
— 1978a. Eohiodon woodruffi n.sp. (Teleostei, Hiodontidae), from the Middle Eocene Klondike Mountain
Formation near Republic, Washington. Ibid. 15, 679-686.
— 1978ft. Paleogene insect faunas of western North America. Quaest. Entomol. 14, 13-34.
— 1979. A second species of Libotonius (Pisces: Percopsidae) from the Eocene of Washington State. Copeia,
1979, 400-405.
— 1980. Eocene lake environments: depth and distance-from-shore variation in fish, insect, and plant
assemblages. Palaeogeogr., Palaeoclimat., Palaeoecol. 32, 21-44.
MARK V. H. WILSON
Department of Zoology
The University of Alberta
Edmonton, Alberta T6G 2E9 Canada
Typescript received 19 February 1981
A FUSED CLUSTER OF CONIFORM CONODONT
ELEMENTS FROM THE LATE ORDOVICIAN
OF WASHINGTON LAND, WESTERN
NORTH GREENLAND
by R. J. ALDRIDGE
Abstract. A fused conodont cluster, comprising six distacodontiform elements and one oistodontiform
element, from upper Ordovician limestones of the Aleqatsiaq Fjord Formation of Washington Land, north
Greenland, is described. A notable microstructural feature of all the elements is the presence of oblique
striations along the anterior margins of the lateral faces. The distacodontiform elements are similar to elements
included in Acodusl mutatus (Branson and Mehl) and Dapsilodus obliquicostatus (Branson and Mehl), but
the apparatus structure appears to be different and the cluster is assigned to Besselodus arcticus gen. et sp. nov.
Conodont elements are normally found as isolated, discrete specimens, and until the mid 1960s
conodont taxonomy and nomenclature were almost entirely based on the morphology of single
elements. A multielement concept of conodont species was first proposed over a hundred years
ago, by Hinde (1879), and the discovery of ‘natural assemblages’ of conodonts on shale bedding
surfaces by Schmidt (1934) and Scott (1934) provided direct evidence that each animal possessed
a skeletal apparatus consisting of different element types. Subsequently, several hundred similar
associations, each representing the skeletal remains of an individual animal, have been found,
mostly in Carboniferous shales (e.g. Scott 1942; DuBois 1943; Rhodes 1952). Descriptions of many
of these naturally occurring apparatuses have yet to be published. Additional evidence of apparatus
structures is provided by fused conodont clusters recovered from acid-insoluble residues (e.g.
Rexroad and Nicoll 1964; Austin and Rhodes 1969; Pollock 1969; Ramovs 1977, 1978), most of
which represent complete or partial apparatuses.
Once the multielement structure of conodont apparatuses was appreciated it was only a matter
of time before workers began to reconstruct apparatuses from collections of isolated specimens.
Walliser (1964), for example, suggested reconstructions of nine Silurian conodont apparatuses
(informally named ‘Conodonten-Apparat’ A-J), although no naturally occurring apparatuses or
clusters were at that time known from the Silurian. A transition of emphasis in conodont taxonomy
from a single element to a multielement basis has followed as more and more reconstructions,
based on statistical, distributional, and morphological evidence, have been published. Huddle (1972)
has documented the development of the early phases of this transition. Of importance in this
taxonomic revolution was the realization that apparatuses conform to a limited number of basic
plans (Klapper and Philip 1971 ; Barnes, Kennedy, McCracken, Nowlan, and Tarrant 1979). Hence,
the structures exhibited by naturally occurring apparatuses and clusters, together with those of
well-established reconstructions, can serve as models in the analysis of new collections.
Naturally occurring apparatuses and clusters are rare, and it is probable that the apparatuses
of most species will be recognized entirely from collections of isolated elements. However, the
direct evidence furnished by natural associations is of paramount importance to multielement
taxonomy; not only do they provide templates for reconstructions, but new finds allow testing
and evaluation of current hypotheses. Additionally, it is only from these associations that we can
confidently assess the relative numbers and dispositions of the component elements.
The only natural associations of late Ordovician conodonts recorded to date are three clusters
[Palaeontology, Vol. 25, Part 2, 1982, pp. 425-430, pi. 44.|
426
PALAEONTOLOGY, VOLUME 25
of Belodina compressa (Branson and Mehl) from Canada (Barnes 1967; Nowlan 1979). Another
cluster, also of coniform elements, has recently been isolated from late Ordovician limestones of
northern Greenland. Although the material is limited, the apparatus differs in structure from those
reconstructed for similar Ordovician and Silurian elements, and the cluster is assigned to a new
taxon, Besselodus arcticus.
SAMPLE LOCALITY AND HORIZON
The cluster was recovered from Geological Survey of Greenland sample no. GGU 242821, collected by Dr.
J. M. Hurst from the Aleqatsiaq Fjord Formation of Aleqatsiaq Fjord, Washington Land, western North
Greenland (see Peel and Hurst 1980; Hurst 1980). The sample of fissile, argillaceous calcilutite is from 132 m
above the base of the formation and approximately 140 m below the top. The conodont fauna includes
Amorphognathus aff. ordovicicus Branson and Mehl, which is also present in samples GGU 242820 and GGU
242822, from 20 m below and above the horizon. A late Ordovician, Ashgill, age is indicated.
THE CONODONT FAUNA
The 730 gm sample available for processing yielded only a small number of conodont specimens. Some of
these are fragmentary, but the following are identifiable:
Amorphognathus aff. ordovicicus (Branson and Mehl)
Pa element 1
Sa element 2
Belodina compressa (Branson and Mehl)
compressed rastrate element 1
eobelodiniform element 1?
Besselodus arcticus gen. et sp. nov.
in cluster: distacodontiform element 6
oistodontiform element 1
isolated: distacodontiform elements 8?
oistodontiform elements 5?
Coelocerodontus trigonius Ethington 1
Ouludusl sp. Sc element 4
Pander odus sp. 2
Pseudobelodindl sp. rastrate element 1
Pseudooneotodus sp. 1
Isolated specimens similar to those of Besselodus arcticus occur sporadically and in small numbers throughout
the lower and middle Aleqatsiaq Fjord Formation. Small acodontiform elements are occasionally found in
the same samples, but it is not at present possible to ascertain whether these specimens belong in Besselodus
or represent another apparatus.
SYSTEMATIC DESCRIPTION
Genus besselodus gen. nov.
Type species. Besselodus arcticus sp. nov. ; from the Aleqatsiaq Fjord Formation of Washington Land.
Diagnosis. The apparatus contains at least two members, distacodontiform and oistodontiform.
All known elements are laterally compressed with sharp anterior and posterior edges.
EXPLANATION OF PLATE 44
Figs. 1-8. Besselodus arcticus gen. et sp. nov. 1-2, lateral views of sub-cluster ‘a’, partial holotype, MGUH
15071, x 250. 3-4, lateral views of sub-cluster ‘b’, partial holotype, MGUH 15072, x 250. 5-6, lateral
views of isolated specimen, MGUH 15073, x 250. 7, detail of partial holotype MGUH 15071, x 660. 8,
detail of partial holotype MGUH 15072, x 660.
PLATE 44
ALDRIDGE, fused conodont cluster
428
PALAEONTOLOGY, VOLUME 25
Discussion. Distacodontiform ( = acontiodontiform) elements similar to those of Besselodus are
included in the reconstructions of the Silurian genus Dapsilodus (Cooper 1976, p. 211) and of the
Ordovician species Acodusl mutatus (Branson and Mehl). The generic assignment of A.? mutatus
was discussed fully by Lofgren (1978, pp. 43-45), who pointed out that the apparatuses of the
type species of Acodus, Distacodus, and Acontiodus are all unknown.
Dapsilodus has an apparatus of distacodontiform, modified distacodontiform and acodontiform
elements (Barrick 1977, p. 50); no oistodontiform element is included, nor are there any
oistodontiform or modified oistodontiform elements known from the Silurian that might be
considered as homologous to the oistodontiform of Besselodus. A A mutatus also has an apparatus
that includes distacodontiform and acodontiform elements (Bergstrom and Sweet 1966, pp.
303-305), although there have been suggestions that an oistodontiform element should also be
included here (Barnes and Poplawski 1973, p. 779; Sweet, in Cooper 1976, p. 211). This has been
contested by Lofgren (1978, pp. 45, 57), who found no distributional relationship between
oistodontiforms and elements of A.? mutatus in her early Ordovician samples from northern Sweden.
The oistodontiform included by Barnes and Poplawski (1973) is Oistodus venustus Stauffer, which
is not closely similar to the element in the cluster of Besselodus.
There is currently little evidence for the inclusion of an acodontiform element in Besselodus.
None occurs in the cluster, nor are there any isolated specimens in the remainder of the, admittedly
small, sample from the same horizon.
Besselodus arcticus sp. nov.
Plate 44, figs. 1-8
Diagnosis. The apparatus contains at least two members, distacodontiform and oistodontiform.
All known elements are laterally compressed with sharp anterior and posterior edges; at the anterior
margins, the lateral faces of all elements display prominent oblique striations.
Description. The material to hand comprises a cluster and a few isolated specimens, all from the same sample.
The cluster consists of very small, delicate coniform elements, all of which are broken at the tips. The elements
are all very thin and translucent and white matter cannot be distinguished. Unfortunately, the cluster fell
apart into two sub-clusters on initial picking, and subsequent handling has been kept to a minimum.
Sub-cluster ‘a’ (PI. 44, figs. 1, 2, 7) is 0-25 mm in maximum dimension and consists of four distacodontiform
elements in subparallel orientation. Each component measures about 0-2 mm from base to broken tip and is
laterally compressed with sharp anterior and posterior edges. The basal outline is a highly compressed ellipse.
Each lateral face bears a longitudinal costa, slightly to the posterior of the mid-line; on those specimens
where the basal portion is visible, the costae terminate sharply a short distance from the basal margin. The
posterior edge of each element is gently curved; the anterior edge is more sharply curved near the basal
margin, producing a slight geniculation. As far as can be ascertained all the elements are bilaterally symmetrical,
or very nearly so, and there is no apparent gradation in size or curvature. Each element displays well-developed
fine, parallel striae on the lateral faces at the anterior margin; at the broken tips these striae are at an angle
of less than 10° to the anterior edge, towards the base this angle steadily increases to greater than 20°. The
striae fade towards the base and terminate at the point of geniculation (PI. 44, fig. 7).
Eight isolated distacodontiform specimens from the same sample show similar features, with some variation
in curvature apparent. Preservation is variable; the best-preserved is illustrated in PI. 44, figs. 5, 6. This
specimen differs from those in the cluster in lacking the geniculation of the anterior edge, producing a more
triangular outline for the basal portion of the unit.
Sub-cluster ‘b’ (PI. 44, figs. 3, 4, 8) is 0-24 mm in maximum dimension. Two distacodontiform elements
comparable with those in sub-cluster ‘a’ are fused in subparallel orientation; a third, broken and cracked,
oistodontiform element is orientated so that its cusp is parallel to those of the distacodontiform elements.
The strong geniculation of the oistodontiform results in the basal margin of the unit lying at an angle of 30°
to the basal margins of the distacodontiforms. The visible lateral face of the oistodontiform shows a longitudinal
costa on the posterior portion of the cusp; oblique striae at the anterior margin of the cusp are also apparent,
ranging in angle from 12° at the broken tip to 20° near the geniculation, where they terminate. These striae
compare closely with those displayed by the distacodontiform elements.
There are five small, isolated oistodontiform specimens in the sample, but those examined under the scanning
ALDRIDGE: FUSED CONODONT CLUSTER
429
electron microscope do not exhibit clear oblique marginal striations; they may or may not be referable to
the same species as the cluster.
Holotype. MGUH 15071 and MGUH 15072, separated portions of a single cluster from sample GGU 242821;
deposited in the type collection of the Geologisk Museum, Copenhagen.
Discussion. The cluster probably represents a partial, rather than a complete, apparatus. The
presence of a single oistodontiform may indicate that it is a partial or complete half-apparatus.
In addition to the general similarity of the morphology of the distacodontiform elements, the
oblique striae on the anterior margins of the elements may indicate a relationship to Dapsilodus.
Similar striae are apparent on specimens of D. obliquicostatus figured by Serpagli (1970, pi. 23,
figs. 1-10, pi. 24, figs. 1-6), Cooper (1976, pi. 2, figs. 11-12, 18-20), and Barrick (1977, pi. 2, figs.
6, 10). Serpagli (1970) noted the absence of similar striations on specimens referred to the older,
possibly ancestral, species Acodus ? mutatus. In the material referred to that species by Lofgren
(1978, p. 44, pi. 2, figs. 9-21) longitudinal striations are developed and a single specimen (pi. 2,
fig. 11a, b) also possesses anterior striae at an angle of 5-10° to the anterior margin. The presence
of anterior striae may not be an important character in determining relationships, as they are
displayed by other specimens, and are apparent in the clusters of Belodina compressa figured by
Nowlan (1979, especially fig. 35.2).
Acknowledgements. This contribution arises from a major co-operative project with the Geological Survey of
Greenland; I am grateful to them for providing the sample, especially to Drs. J. M. Hurst and J. S. Peel for
collecting the material and arranging the loan. Mr. A. Swift processed the sample in the laboratory and first
recognized the cluster. I also thank Mr. L. Green and Mr. D. J. Jones for photographic assistance. Dr. Anita
Lofgren (University of Lund) kindly read and commented on the typescript. This paper is published with
the permission of the Director of the Geological Survey of Greenland.
REFERENCES
Austin, R. l. and Rhodes, f. h. t. 1969. A conodont assemblage from the Carboniferous of the Avon Gorge,
Bristol. Palaeontology, 12, 400-405.
barnes, c. R. 1967. A questionable natural conodont assemblage from Middle Ordovician limestone, Ottawa,
Canada. J. Paleont. 41, 1557-1560.
— Kennedy, D. J., mccracken, a. d., nowlan, G. s. and tarrant, g. a. 1979. The structure and evolution
of Ordovician conodont apparatuses. Lethaia, 12, 125-151.
— and poplawski, m. l. s. 1973. Lower and Middle Ordovician conodonts from the Mystic Formation,
Quebec, Canada. J. Paleont. 47, 760-790, pis. 1-5.
barrick, j. e. 1977. Multielement simple-cone conodonts from the Clarita Formation (Silurian), Arbuckle
Mountains, Oklahoma. Geol. and Palaeontol. 11, 47-68, pis. 1-3.
Bergstrom, s. M. and sweet, w. c. 1966. Conodonts from the Lexington Limestone (Middle Ordovician) of
Kentucky and its lateral equivalents in Ohio and Indiana. Bull. Am. Paleont. 50, 269-441, pis. 28-35.
cooper, b. j. 1976. Multielement conodonts from the St. Clair Limestone (Silurian) of Southern Illinois.
J. Paleont. 50, 205-217, pis. 1-2.
dubois, E. p. 1943. Evidence on the nature of conodonts. Ibid. 17, 155-159, pi. 25.
hinde, G. J. 1879. On conodonts from the Chazy and Cincinnati Group of the Cambro-Silurian, and from the
Hamilton and Genesee-Shale divisions of the Devonian, in Canada and the United States. Q. Jl geol. Soc.
Lond. 35, 351-369, pis. 15-17.
huddle, j. w. 1972. Historical introduction to the problem of conodont taxonomy. Geol. and Palaeontol. SB 1,
3-16, pi. 1.
hurst, j. m. 1980. Silurian stratigraphy and facies distribution in Washington Land and Western Hall Land,
North Greenland. Bull. Gronlands geol. Unders. 138, 1-95.
klapper, g. and philip, G. m. 1971. Devonian conodont apparatuses and their vicarious skeletal elements.
Lethaia, 4, 429-452.
lofgren, a. 1978. Arenigian and Llanvirnian conodonts from Jamtland, northern Sweden. Foss, and Strata,
13, 1-129, pis. 1-16.
nowlan, G. s. 1979. Fused clusters of the conodont genus Belodina Ethington from the Thumb Mountain
430
PALAEONTOLOGY, VOLUME 25
Formation (Ordovician), Ellesmere Island, District of Franklin. Curr. Res. Part A, Geol. Surv. Can., Pap.
79-1A, 213-218.
peel, j. s. and hurst, J. M. 1980. Late Ordovician and early Silurian stratigraphy of Washington Land, western
North Greenland. Rapp. Gronlands geol. Unders. 100, 18-24.
pollock, c. a. 1969. Fused Silurian conodont clusters from Indiana. J. Paleont. 43, 929-935, pis. 110-112.
ramovs, a. 1 977. Skelettapparat von Pseudo furnishius murcianus (Conodontophorida) im Mitteltrias Sloweniens
(NW Jugoslawien). Neues Jb. Geol. Palaont. Abh. 153, 361-399.
— 1978. Mitteltriassische conodonten-clusters in Slowenien, NW Jugoslawien. Palaont. Z. 52, 129-137.
rexroad, c. b. and nicoll, R. s. 1964. A Silurian conodont with tetanus? J. Paleont. 38, 771-773.
Rhodes, F. h. T. 1952. A classification of Pennsylvanian conodont assemblages. Ibid. 26, 886-901, pis. 126-129.
schmidt, H. 1934. Conodonten-Funde in urspriinglichen Zusammenhang. Palaont. Z. 16, 76-85, pi. 6.
scott, h. w. 1934. The zoological relationships of the conodonts. J. Paleont. 8, 448-455, pis. 58-59.
— 1942. Conodont assemblages from the Heath formation, Montana. Ibid. 16, 293-300, pis. 37-40.
serpagli, E. 1970. Uppermost Wenlockian-upper Ludlovian (Silurian) conodonts from Western Sardinia.
Boll. Soc. Paleontol. Ital. 9, 76-96, pis. 21-24.
walliser, o. h. 1964. Conodonten des Silurs. Abh. hess. Landesamt Bodenforsch. 41, 1-106, pis. 1-32.
Typescript received 27 November 1980
Revised typescript received 26 January 1981
R. J. ALDRIDGE
Department of Geology
University of Nottingham
Nottingham NG7 2RD
A NEW CALCAREOUS GREEN ALGA FROM THE
MIDDLE JURASSIC OF ENGLAND: ITS
RELATIONSHIPS AND EVOLUTIONARY
POSITION
by GRAHAM F. ELLIOTT
Abstract. Leckhamptonella llewellyae gen. et sp. nov., is described from the Middle Jurassic (Inferior Oolite;
Aalenian) of the Cotswold district, England. Although known only from fragmentary material, it is recognizable
as the remains of a new serial-segmented member of the Udoteaceae (Chlorophyta). It is compared with Boueina
and Arabicodium (both Mesozoic) and Halimeda (Cretaceous-Recent), and shows certain affinities with the
latter especially. The evolution of serial-segmented Udoteaceae from Lower Palaeozoic to Recent is briefly
reviewed.
The fossil described below occurs as pieces and fragments of original small calcareous-walled hollow
ovoids, near-circular in transverse section and curved-elongate in vertical section. The walls show a
very distinctive structure formed by irregular but characteristic branching pores, with possible
vestiges of a structure interior to the wall. The material studied occurs in an oomicrite with
subordinate echinoderm, molluscan, brachiopod, and other organic fragments, often worn and
encrusted and presumed current-swept and not on the site of growth.
The new fossil does not show the optical extinction of echinoderm-calcite pieces, nor the laminar
structure of molluscs or brachiopods. The wall-structure is not that of stromatoporoids or hydrozoa,
and has not the spicular mesh of any known sponge. It does, however, have the characteristics of
certain green calcareous algae, notably the Order Dasycladales and the family Udoteaceae of the
Order Caulerpales. The resemblance is very much with the cortical structure of the ‘serial-segmented’
members of the Udoteaceae ( Halimeda , and various extinct genera discussed below). It possesses the
marked irregularity in detail which is so characteristic, and the general pore-plan is that of yet another
variant of the branching utricles of the family. It does not, in most specimens, show clearly the
longitudinal medullary threads appropriate to this interpretation, but two well-preserved pieces show
traces of this structure, which has to be carefully distinguished from peripheral diagenesis and
staining of the clasts. The absence is probably due to the rolled state of the fragments and the usual
originally weak medullary calcification in the Udoteaceae. In the Permian Tauridium, normally found
fragmented, very much more abundant material than is available with the present fossil still leaves
reconstruction a difficult task.
If comparison is made with the Udoteacean Ovulites (Cretaceous, Cenozoic), the external
dimensions and shape of the Cotswold fossil as so far known are comparable with those of O.
margaritula (Lmk.) Munier-Chalmas (cf. Massieux 1966). In these, however, and other Ovulites spp.,
the calcareous wall is thinner, the pores near-uniformly straight, thin, and radial, and no trace
survives of medullary structures. Ovulites is usually regarded as remains of a plant comparable with
the living Penicillus, the ‘Neptune’s Shaving Brush’, where the thallus is of different morphology to
that of Halimeda. Ovulites and Halimeda both appear in the Cretaceous, and it is interesting that the
new Jurassic fossil brings both to mind, even if apparently much more similar to the latter in
structure.
If the present remains are envisaged as those of a dasycladalean, they are anomalous in showing
udoteacean irregularity and not a verticillate arrangement. The thin calcification of the swollen heads
of the Jurassic Petrascula and Coniporella belongs to very much larger individuals.
IPalaeontology, Vol. 25, Part 2, 1982, pp. 431-437, pi. 45.|
432
PALAEONTOLOGY, VOLUME 25
It would thus seem that the fragments can be interpreted as those of ovoid serial-segments of a new
udoteacean, the medullary part originally weakly calcified and now largely missing in the worn
fragments available.
SYSTEMATIC PALAEONTOLOGY
Division chlorophyta (Green Algae)
Order caulerpales Feldmann 1946
Family udoteaceae Feldmann 1946
Genus leckhamptonella gen. nov.
Udoteacean segments with calcified cortical zone showing swollen branching utricles each dividing into several
thinner outer parallel utricles which again divide peripherally. Type-species Leckhamptonella llewellyae sp. nov.,
dedicated to Dr. Llewellya Hillis-Colinvaux in recognition of her extensive studies of the living Halimeda.
Leckhamptonella llewellyae sp. nov.
Plate 45, figs. 1-6
Description. Leckhamptonella with presumed ovoid segments of observed length up to 2-70 mm and estimated
matching diameter of approximately 1 -80 mm (but a section of 2-40 diameter may indicate larger segments).
Medullary zone mostly missing in fossil pieces: thickness of presumed medullary filaments where preserved
about 0 030 mm. Thickness of calcified cortical zone about 0-36 mm; inner half of this occupied by a zone of
outwardly directed waisted and swollen branching utricles of 0-030-0 090 mm diameter. These each divide into
several straight, thin, near-parallel outer utricles of about 0-015 mm diameter. These are at right angles to the
longitudinal axis of the segment, and divide again just below the segment-surface, to short adjacent peripheral
utricles of 0-010 mm or less diameter, which expand up to 0-020 mm terminal diameter, almost touching.
Holotype. The specimen figured in PI. 45, fig. 1; British Museum (Natural History), Dept. Palaeontology
registered number V.60703. Middle Jurassic, Lower Inferior Oolite (Aalenian bradfordensis subzone); Upper
Freestone facies of Scotsquar Hill Limestone (Mudge 1978): Leckhampton, Cheltenham, Gloucestershire.
Paratypes. The specimens figured in PI. 45, figs. 2-6; registered numbers V.60704-60708 inch; same locality and
horizon.
Other material. About thirty-five fragments in thin-sections: same locality and horizon.
THE EVOLUTION OF THE ‘SERIAL-SEGMENTED’ UDOTEACEAE
The members of the Udoteaceae, to which Leckhamptonella is referred, show a structure of repeatedly
branching threads, much intertangled and interwoven. In Udotea itself the whole thallus is a single
EXPLANATION OF PLATE 45
Figs. 1-6. Leckhamptonella llewellyae gen. et sp. nov. All pieces in thin-section from the Middle Jurassic,
Aalenian bradfordensis subzone, Lower Inferior Oolite; Leckhampton, Gloucestershire, England. 1 , portion
of vertical section of calcareous wall, showing presumed medullary thread at base, swollen branching cortical
utricles, radial utricles, and terminal peripheral utricles (shown a little left of top centre); x 80; Holotype,
British Museum (Natural History), Department of Palaeontology, registered number V.60703. 2, portion of
transverse section of large unit, with external dark crust, x 30; V.60704. 3, portion of vertical section,
showing two presumed medullary threads (bottom right), swollen utricles and radial utricles, x 80;
V. 60705. 4, tangential subsurface section, showing close-set peripheral utricles, x 80; V.60706. 5, slightly
oblique vertical section of whole side of unit, x 40; V. 60707. 6, tangential-longitudinal section of wall
showing swollen and radial utricles, heavy dark crusting externally, x 40; V.60708.
Fig. 7. Oblique cut, piece of Tauridium sp. for comparison; Upper Permian, Southern Tunisia, x 40; V. 54052.
PLATE 45
ELLIOTT, calcareous green alga
434
PALAEONTOLOGY, VOLUME 25
fan-shaped structure, but another common growth-form, of which Halimeda is the modern
representative, is shown by what may be termed ‘the serial-segmented udoteaceans’. In these, known
since the Lower Palaeozoic, the plant is (or is presumed to have been) a clump of branching
successions of numerous individual calcified segments, connected by uncalcified threads or filaments.
Each segment shows internal coarse longitudinal medullary filaments from end to end, and an outer
cortical zone of lateral, branching, and often swollen utricles which terminate in a surface layer of
closely packed peripheral utricles. The medullary filaments give rise distally during growth to new
segments beyond the old, the connecting filaments being either separate or partially or completely
fused at the nodes, so uniting the whole segmented plant as a flexible structure. The peripheral utricles
are concerned with the assimilatory functions, and the swollen inner cortical utricles with eventual
transference of their content to deciduous reproductive outgrowths (gametangia) at the times of
sexual reproduction (Hillis-Colinvaux 1980). These last are only really known in the living Halimeda
(one fossil record in Pfender 1940, p. 245) and our knowledge is incomplete.
Calcification of the segments is usually heavy cortically (interutricular), but much less so in the
medullary zone. In the fossils the segments occur dissociated, mixed, and often broken. They are
preserved because of the calcification. The behaviour of the medullary filaments at the nodes, so
important in classification of the living Halimeda (Hillis-Colinvaux 1980) is thus not normally known
in the extinct genera. A solitary example in the collections of the British Museum (Natural History) of
an undescribed udoteacean from the Upper Permian of Tunisia, reg. no. V. 54065, shows the nodal
filaments parallel and in contact along their length, but apparently not fused (if this is the correct
interpretation of their slightly different mineralizations). The variation in detail and degree of
structure and calcification between older and younger segments of individual plants, between those of
different individuals of the same species, and between those typical of different species, as known in
living Halimeda (Hillis 1959; Hillis-Colinvaux 1980) can make precise specific evaluation of the
fragmentary resorted fossil remains difficult, though often occurrences seem to be of a single species.
Ideally, one should have a real abundance of material to describe (Conard and Rioult 1977), but this
is not always available. The state of geological preservation may add to these difficulties, explaining
such records as Boueina/ Arabicodium, Boueina/ Halimeda (Bismuth, Bonnefous, and Dufaure 1967).
The evolution of serial-segmented udoteaceans, as preserved fossil, appears to have consisted of
variation and different combinations of the basic structures outlined above, from the Lower
Palaeozoic onwards (text-fig. 1). Thus Dimorphosiphon (Ordovician) shows a coarse thread-structure
and Palaeoporella (Ordovician-Devonian) a fine one, while Maslovina (Silurian) has the dense layer
of small peripheral utricles typical of some later genera (Obrhel 1968), but not present in the other
two. Text-fig. 1 shows a selection of patterns in different genera from Dimorphosiphon to Halimeda. Is
this evolution more or less random, or does it show some progression in time with the surviving
Halimeda as the most advanced as well as the latest of its kind? And does the new Jurassic
Leckhamptonella throw any light on this problem?
It can be seen from text-fig. 1 that the central medullary filaments can be thick or thin, straight or
tangled in varying degree in different genera. Similarly, the cortical utricles, while often branching,
vary much in spacing or crowding, and in degree of swelling between genera. The examples figured
show various combinations, typical of those genera. On mechanical grounds, strong straight
medullary filaments, with some fusion and flexibility at the nodes, would equip such a plant to
withstand a moderate degree of water-movement and enable it to colonize moderate-energy
environments, other conditions being suitable. In addition, swollen cortical utricles and a layer of
close-set peripheral utricles would provide for quick segment-growth and eventually for rapid
production of gametangia at times of sexual reproduction. This combination, familiar in Halimeda , is
apparently first achieved in the Permian Tauridium, though with different proportions in the utricles,
but these were extremely fragile plants post-mortem, almost invariably found as debris or
comminuted. Apparently calcification was thin and mostly outer-cortical, and in life the plant was
probably confined to quiet waters. In Halimeda, however, calcification is heavy, and its success as
witnessed by abundance, wide distribution, and occurrence in a range of low to moderate energy-
environments (Hillis-Colinvaux 1980) is well known.
ELLIOTT: CALCAREOUS GREEN ALGA
435
text-fig. 1 . Diagrammatic representations of filament and utricle structure in various
udoteacean genera, mostly based on materials in the collections of the British Museum
(Natural History): (a) Dimorphosiphon (Ordovician); ( b ) Aphroditicodium (Permian;
BM(NH) Dept. Palaeont. reg. no. V.59461); (c) Tauridium (Permian); (d) Arabicodium
(Jurassic-Cretaceous); (e) Boueina (Triassic-Cretaceous); (/) Halimeda incrassata
(Ellis) Lamx (Recent) (Hillis 1 959); (g) Leckhamptonella llewellyae Elliott (Jurassic);
( h ) diagrammatic growth-plan of serial-segmented udoteacean. See also comparisons
for some other Palaeozoic genera in Obrhel (1968, fig. 1) and Guilbault and Mamet
(1976, fig. 2).
436
PALAEONTOLOGY, VOLUME 25
Leckhamptonella llewellyae shows similarities in cortical utricle-structure to typical modern
Halimeda spp. It differs in that the third layer of cortical branches are straight, thin, and parallel
before dividing into peripheral utricles, whereas in the modern Halimeda spp. swollen branches and
branchlets usually continue outwards to the peripheral utricles. In the light of the functional
reasoning above this difference would be a primitive character is Leckhamptonella. The general
appearance of the medullary zone is not known, and so cannot be compared with those of other
genera.
Halimeda is itself known rarely from the Lower Cretaceous (Dragastan and Bucur 1979; possibly
Wells 1944), becomes more common in the Upper Cretaceous, and is abundant throughout the
Cenozoic to the present day. It has been considered as arising by hybridization from the earlier
Mesozoic Boueina and Arabicodium (Elliott 1965) or as more closely related to Boueina as evidenced
by comparable intrageneric species groupings (Conard and Rioult 1977). The fragmentary condition
of the Middle Jurassic Leckhamptonella , as described above, precludes a detailed comparison of these
four genera, but the cortical structure of Leckhamptonella appears closer to that of the later Halimeda
than to those of the earlier genera.
Arabicodium , Boueina , and Halimeda were all Tethyan in origin: Boueina appears in the Upper
Triassic of Central Europe and of Thailand (Flrigel 1975; Kemper, Maronde, and Stoppel 1976) and
its pan-tropical distribution in the Lower Cretaceous has been plotted by Elliott (1981). Halimeda
may have appeared in the Lower Cretaceous of both hemispheres (Dragastan and Bucur 1979; Wells
1944) and was certainly widely distributed in the Upper Cretaceous (Elliott 1981). Arabicodium from
the Jurassic-Cretaceous of the western Tethys (Mediterranean-Middle East) had a Jurassic straggler
as far north as southern England (Elliott 1975) in the same area where Leckhamptonella occurs.
Acknowledgements. My thanks are due to Mr. Mark Crawley and Mr. P. V. York, both British Museum
(Natural History), for drawings and photography respectively.
REFERENCES
bismuth, h., bonnefous, j. and dufaure, ph. 1967. Mesozoic microfacies of Tunisia. In Guidebk Ann. Fid Conf.
Petroleum Expl. Soc. Libya: 9th Guidebk to Geology: A history of Tunisia. Pp. 159-214.
conard, m. and rioult, m. 1977. Halimeda elliotti nov. sp., algue calcaire (Chlorophyceae) du Turonien des
Alpes-maritimes (SE France). Geol. mediterran. 4, 83-98.
dragastan, o. and bucur, I. 1979. Upper Aptian Microfossils from the Camenita Valley-Sasca Romana
(Resita-Moldova Nova Zone, Banat), Revue roum. Geol. Geophys. Geogr. ( ser . Geol), 23, 111-115.
elliott, g. f. 1965. The interrelationships of some Cretaceous Codiaceae (calcareous algae). Palaeontology, 8,
199-203.
— 1975. Transported algae as indicators of different marine habitats in the English middle Jurassic. Ibid. 18,
351-366.
— 1981. The Tethyan dispersal of some chlorophyte algae subsequent to the Palaeozoic. Palaeogeogr.,
Palaeoclimat., Palaeoecol., 32, 341-358.
flugel, E. 1975. Kalkalgen aus Riffkomplexen der alpin-mediterranen Obertrias. Verh. Geol. B.-A. Wien, Jhg.
1974, h. 2-3, 297-346.
guilbault, j. p. and mamet, b. l. 1976. Codiacees (Algues) ordoviciennes des Basses-Terres du Saint-Laurent.
Can. J. Earth Sci. 13, 636-660.
hillis, l. w. 1959. A revision of the genus Halimeda (Order Siphonales). Publ. Inst. mar. Sci. Univ. Tex. 6,
321-403.
hillis-colinvaux, l. 1980. Ecology and taxonomy of Halimeda-, primary producer of coral reefs. Adv. mar. Biol.
17, 1-327.
kemper, e., maronde, h. d. and stoppel, d. 1976. Triassic and Jurassic Limestone in the region Northwest and
West of Si Sawat (Kanchaburi Province, Western Thailand). Geol. Jb. B21, 93-127.
massieux, m. 1966. Presence d 'Ovulites dans le ‘Calcaire Ypresien’ des Corbieres septentrionales et discussion sur
la nature de l’algue Griphoporel/a arabica Pfender. Rev. Micropaleont. 8, 240-248.
mudge, d. c. 1978. Stratigraphy and sedimentation of the Lower Inferior Oolite of the Cotswolds. Jl geol. Soc.
Lond. 135, 611-627.
437
ELLIOTT: CALCAREOUS GREEN ALGA
obrhel, j. 1968. Maslovina meyenii n.g. et sp.— neue Codiacea aus dem Silur Bohmens. Vest, ustred Ust seo!
43, 367-370. 5
PFENDER, j. 1940. Les algues du Nummulitique egyptien et des terrains Cretaces- Eocenes de quelques regions
mesogeennes. Bull. Inst. Egypte, 22, 225-250.
WELLS, J. w. 1944. Cretaceous, Tertiary, and Recent Corals, a sponge, and an alga from Venezuela. J. Paleont
18, 429-447.
After completion of this paper, I read the memoir of Dr. D. Vachard (Doc. Trav. IGAL Paris no. 2,
dated 1980, issued 1981) in which he suggests that various genera, including Aphroditicodium Elliott
and Tauridium Guveng should be considered as synonyms of Permocalculus (Vachard 1981, 382).
Aphroditicodium he considers to be a well-preserved Permocalculus fragilis. The point is of some
importance because it would transfer these two genera from Udoteaceae (green algae) to
Gymnocodiaceae (red algae). I have examined material from Burma, much better preserved than the
type of Aphroditicodium , and more extensive material of Tauridium from north Italy, in both of which
the medullary structures and their relations to the cortical structures are well shown, better than in the
types. However, in none of these specimens are reproductive structures to be seen (an absence to be
expected in Udoteaceae), whereas in Permocalculus they are conspicuous in many individuals, even
when preservation is poor. Whilst agreeing with Dr. Vachard as to the similarities of structural plan in
these fossil genera, and as to the various changes in taxonomic allocation (to which I contributed) in
the past, I consider that Aphroditicodium and Tauridium are Udoteacean genera, however wide a
variation the Gymnocodiacean Permocalculus shows.
Typescript received 20 February 1981
Revised typescript received 8 May 1981
G. F. ELLIOTT
Department of Palaeontology
British Museum (Natural History)
Cromwell Road
London SW7 5BD
ADDENDUM
A RARE LYTOCERATID AMMONITE FROM THE
LOWER LIAS OF RADSTOCK
by D. T. DONOVAN and M. K. howarth
Abstract. A single lytoceratid ammonite from the Jamesoni or Ibex Zone of the Lower Lias of Clandown
Colliery Quarry, Radstock, Avon, is made the holotype of Derolytoceras (£>.) radstockense sp. nov. It is the
first record in Britain of a genus that is largely restricted to Tethyan areas of south and central Europe.
The partial or complete restriction of certain groups of ammonites to the Tethys has long been
known, and was documented in some detail by Donovan (1967) for the Lower Jurassic. At this
time, the Suborders Phylloceratina and Lytoceratina are so restricted except for certain limited
spans of time when certain genera are found in northern Europe. Thus, Phylloceratina are
represented in Britain by Tragophylloceras in the Pliensbachian and Phylloceras in the Upper
Pliensbachian and the Toarcian, and Lytoceratina by Lytoceras in the Upper Pliensbachian and
the Toarcian and by genera of the Alocolytoceratinae in the Upper Toarcian. A second class of
occurrence comprises rare finds, usually made from a single horizon. The following fall into this
category:
? a member of Juraphyllitidae from the Planorbis Zone of the Stowell Park Borehole,
Gloucestershire (Spath 1956, p. 158). One of us (DTD) has re-examined this specimen, which
was preserved in soft mudstone, but it is now in such an abraded state that further comment is
not possible.
Galaticeras jacksoni Howarth and Donovan, a member of Juraphyllitidae, known from nine
examples from the Flatstones (Obtusum Subzone) of the Dorset coast; Galaticeras sp. has been
recorded from two boreholes in southern England (Howarth and Donovan 1964).
Meneghiniceras lariense (Meneghini), another juraphyllitid, a single specimen from the Grey
Shales (Semicelatum Subzone) in the Toarcian near Whitby, Yorkshire (Howarth 1976).
Aegolytoceras rotundicosta (Tutcher and Trueman) from the Jamesoni Limestone of Radstock,
discussed in detail below.
The present paper records an example of a lytoceratid genus new to Britain, from the Lower Lias
of Radstock, collected sixty years ago but only recently recognized for what it is.
SYSTEMATIC PALAEONTOLOGY
Superfamily lytocerataceae Neumayr 1875
Family derolytoceratidae Spath 1927
Genus Derolytoceras Rosenberg 1909
Type species. Ammonites lineatus tortus Quenstedt 1885, subsequently designated by Spath 1924 (p. 4).
Subgenus derolytoceras
Derolytoceras ( Derolytoceras ) radstockense sp. nov.
Text-fig. 1
Diagnosis. Derolytoceras attaining at least 80 mm diameter, with evolute, slowly expanding whorls,
and an evenly rounded elliptical whorl section. Regular radial ribs are strong from 13 mm diameter,
(Palaeontology, Vol. 25, Part 2, 1982, pp. 439-442.]
440
PALAEONTOLOGY, VOLUME 25
text-fig. 1. Derolytoceras ( D .) radstockense sp. nov. Holotype. Jamesoni Limestone, Lower Pliensbachian,
Jamesoni or Ibex Zones, Clandown Colliery Quarry, Radstock, Avon. Bristol University Geology Museum
no. 2877, x 0-95.
and curve forwards to the middle of the venter. Three or four constrictions per whorl are similar
in shape to the ribs, and are accentuated by an enlarged rib immediately in front. The suture-line
is Lytoceratid with highly indented and undercut saddles.
Material. The holotype only, Bristol University Geology Museum no. 2877, from the Jamesoni Limestone,
Lower Lias (Lower Pliensbachian: Jamesoni or Ibex Zone) at Clandown Colliery Quarry (ST 679 558), near
Radstock, Avon. For the section at this quarry see Tutcher and Trueman 1925, p. 600. Collected in 1922 or
1923 by Mr. T. R. Fry.
Measurements. Maximum diameter 80-5 mm. At 78T mm diameter, whorl height = 23-8 mm (0-30), whorl
breadth =181 mm (0-23), umbilical width = 38 0 mm (0-49). There are thirty-three ribs and four constrictions
on the outer whorl ending at 80 mm diameter.
Description. The specimen consists of about four visible whorls ending at a diameter of 80 mm. The mouth
border is not present, for at least the final one-eighth of a whorl is missing. Suture-lines can be clearly seen
just over half a whorl behind the aperture at a diameter of 57 mm. From the rough state of preservation it
is not possible to see whether there are more suture-lines at a larger size, but the phragmocone may have
extended about one-eighth of a whorl further forwards. This makes the length of the body chamber that is
preserved between three-eighths and half a whorl in length. The whorls are evolute but not completely so,
for each whorl overlaps about 20% of the next inner whorl. The whorl section is compressed elliptical and
has an evenly rounded venter and umbilical wall. Capricorn ribs are well developed on all visible whorls from
the smallest size seen, about 13 mm diameter, up to the aperture, though they diminish in strength on the
last half whorl. They are approximately radial on the side of the whorl, then curve forwards towards the
middle of the venter. There are four well-marked constrictions on the outer whorl that follow the line of the
ribs exactly. They are made more prominent by an enlarged rib immediately in front of each one. Three
constrictions can be seen on the next inner whorl, but at smaller sizes the preservation is not good enough
for constrictions to be visible.
The suture-line (text-fig. 2) is typically Lytoceratid. The large first and second lateral saddles are deeply
indented and undercut, and have complicated moss-like endings. The large first lateral lobe is divided into I
three prongs.
Remarks. Ammonites of the superfamily Lytocerataceae, other than Lytoceras itself, are almost
unknown in the Sinemurian and Pliensbachian in Britain. The only one known previously is the
DONOVAN AND HOWARTH: LYTOCERATID AMMONITE
441
single specimen from the Jamesoni Limestone at Radstock that was the holotype and sole basis
of the name ‘ Peripleuroceras' rotundicosta Tutcher and Trueman (1925, p. 646, pi. 41, fig. 1)
(BM(NH) C. 41 760). That specimen is smooth up to about 30 mm diameter, and then develops
capricorn ribs mainly on the venter, that have become fairly strong by its maximum size of 49
mm diameter. It is wholly septate. It belongs to the Derolyoceratidae genus Aegolytoceras Spath
1924, of which Peripleuroceras Tutcher and Trueman 1925 is a junior synonym. Other species of
Aegolytoceras develop the characteristic coarse capricorn ribs immediately behind constrictions,
reverting to fine ribbing in front of constrictions. Derolytoceras ( D .) radstockensis differs in having
regularly strong capricorn ribs that develop at an earlier growth stage, and the only irregularity
on the phragmocone is the single slightly stronger rib that follows in front of each constriction.
It is difficult to tell whether the decrease in rib strength on the last half whorl is due to poor
preservation or is the onset of adult ornamentation. These are features of the genus Derolytoceras ,
of which, at 80 mm diameter, it is one of the largest known examples. The subgenus D.
(Derolytoceras) has ribs from an early growth stage, certainly by 13 mm diameter which is the
smallest size visible in this Radstock specimen, while the subgenus D. ( Tragolytoceras ) remains
smooth up to about 25 mm diameter. So the Radstock specimen belongs to the nominal subgenus.
The type species is D. (D.) tortum (Quenstedt), of which the lectotype (Quenstedt, 1885, p. 309,
pi. 39, fig. 15), refigured by Wiedmann (1970, p. 995, text-fig. 8e, pi. 6, fig. 3), is from the Upper
Pliensbachian of south-west Germany. It has quickly expanding whorls and very strong ribs
commence at about 11 mm diameter. Another Sinemurian or Pliensbachian species is D. (D.)
haueri Rosenberg (1909, p. 251, pi. 11, figs. 31, 32), which is based on specimens of up to 15 mm
diameter that have slowly expanding and finely ribbed evolute whorls. Though it is difficult to
compare the much larger Radstock specimen with either of these species, it appears to differ from
both of them in its slowly expanding whorls and coarse ribs.
At first sight the Radstock specimen seems to have considerable resemblances to the Polymorphitid
genus Platypleuroceras, and many examples of P. brevispina (J. de C. Sowerby) or closely allied
species have been found in the Jamesoni Limestone at Radstock (e.g. Tutcher and Trueman 1925,
p. 650, pi. 39, fig. 3; pi. 40, fig. 2). The Derolytoceras differs, however, in lacking the ventro-lateral
tubercle that occurs in all the specimens of Platypleuroceras , and in having more strongly projected
ribs on the venter. It also has constrictions that are only rarely present in Radstock Platypleuroceras
(though constrictions are better developed in similar Polymorphitidae from Germany, e.g. Quenstedt
1885, pi. 32, fig. 6; pi. 33, figs. 11, 12). Finally, the suture-line of the Derolytoceratid is distinctively
Lytoceratid, and suture-lines in Platypleuroceras do not have such deeply indented or undercut
saddles.
The family and generic classification used here is that of the Treatise (Arkell, 1957, p. LI 94).
Two more recently proposed generic names are Adnethiceras Wiedmann (1970, p. 997) for species
that have ventro-lateral tubercles, and Lytoconites Wiedmann (1970, p. 1004) which is considered
here to be a synonym of Aegolytoceras. Peripleuroceras Tutcher and Trueman 1925, is also a
synonym of the latter genus. Wiedmann (1970, p. 988) did not admit a separate family
Derolytoceratidae, but placed it and its included genera in the Lytoceratidae. The distinctive
442
PALAEONTOLOGY: VOLUME 25
capricorn ornament of Derolytoceratidae seems to us, however, to be sufficiently different from
the ornament of normal Lytoceratidae to warrant separation of this Sinemurian to Toarcian group
as a family. Most genera of Derolytoceratidae were also discussed by Fantini Sestini (1973) who
reinterpreted the genus Audaxlytoceras Fucini 1923 (type species Ammonites audax Meneghini 1881
(non Oppel, 1863)) on the basis of specimens that were larger than the very small originals, and
concluded that it was the oldest name for the group of species that have usually been referred to
Aegolytoceras Spath 1924.
The best collection of Derolytoceratidae, which was not discussed by either Wiedmann (1970)
or Fantini Sestini (1973), comes from the Pliensbachian at Monte di Cetona, central Italy. Nine
species of "Deroceras' were described by Fucini (1903, pp. 166-185): all of them have Lytoceratid
suture-lines, and amongst them are the largest known Derolytoceratidae, some specimens attaining
165 mm diameter. Several develop ventro-lateral tubercles and belong to Adnethiceras (e.g. A.
olenoptychum Fucini, A. mutans Fucini and A. instabile Fucini), and about eight examples were
figured of Aegolytoceras pecchiolii (Meneghini), which is the most highly developed species of its
genus. Nothing in the collection is exactly like the Radstock specimen, though the rather poorly
preserved examples of Derolytoceras connexum Fucini (1903, p. 176, pi. 26, figs. 7, 8) are the
nearest, differing mainly in lacking clear constrictions.
REFERENCES
arkell, w. J. 1957. In R. c. MOORE, Treatise on invertebrate paleontology. Part L, Mollusca 4, Cephalopoda,
Ammonoidea. Geol. Soc. Amer.
donovan, d. t. 1967. The geographical distribution of Lower Jurassic ammonites in Europe and adjacent areas.
Pubis Syst. Ass. 7, 111-134.
fantini sestini, n. 1973. Revisione del genere Audaxlytoceras Fucini, 1923 (Ammonoidea). Riv. ital. Paleont.
Stratigr. 79, 479-502, pi. 49.
fucini, a. 1903. Cefalopodi liassici del Monte di Cetona. Part 3. Palaeontogr. ital. 9, 125-185, pis. 19-27.
howarth, m. k. 1976. An occurrence of the Tethyan ammonite Meneghiniceras in the Upper Lias of the
Yorkshire coast. Palaeontology , 19, 773-777.
— and donovan, d. t. 1964. Ammonites of the Liassic family Juraphyllitidae in Britain. Ibid., 7, 286-305, pis.
48, 49.
quenstedt, F. A. 1885. Die Ammoniten des Schwdbischen Jura. 1, der Schwarze Jura (Lias), 241-440, pis. 31-54.
Tubingen.
rosenberg, p. 1909. Die Liasische Cephalopodenfauna der Kratzalpe im Hagengebirge. Beitr. Palaont. Geol.
Ost.-Ung. 22, 193-345, pis. 10-16.
spath, l. f. 1924. On the Blake Collection of Ammonites from Kachh, India. Mem. geol. Surv. India Palaeont.
indica, 9, mem. 1, 29 pp.
— 1956. The Liassic ammonite faunas of the Stowell Park Borehole. Bull. geol. Surv. Gt. Br. 11, 140-164, pis.
9, 10.
tutcher, j. w. and trueman, a. e. 1925. The Liassic rocks of the Radstock district, Somerset. Q. Jl. geol. Soc.
Lond. 81, 595-666, pis. 38-41.
wiedmann, j. 1970. Uber den Ursprung der Neoammonoideen— Das Problem einer Typogenese. Eclog. geol.
Helv. 63, 923-1020, pis. 1-10.
D. T. DONOVAN
Department of Geology
University College
Gower Street
London WC1E 6BT
M. K. HOWARTH
Department of Palaeontology
British Museum (Natural History)
Cromwell Road
Typescript received 13 October 1980 London SW7 5BD
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 l or 1982 arc:
Institutional membership . . . £32 (U.S. $71)
Ordinary membership. . . . £15 (U.S. $33)
Student membership. . £9-50 (U.S. $21)
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 1982 will receive Volume 25, Parts 1-4. All back numbers are still in print and may be ordered from
Marstori Book Services, P.O. Box 87, Oxford 0X4 1LB, England, at £15 (U.S. $33) 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 1982 is £25 (U.S. $55) for Institutional Members, and £12-50 (U.S. $27-50) 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, P.O. Box 87, Oxford 0X4 1LB, England.
COUNCIL 1982-1983
President : Professor A. Hallam, Department of Geological Sciences, The University, Birmingham B15 2TT
Vice-Presidents : Professor J. W. Murray, Department of Geology, The University, Exeter EX4 4QE
Dr. R. A. Fortey, Department of Palaeontology, British Museum (Natural History), Cromwell Road, London SW7 5BD
Treasurer. Dr. M. Romano, Department of Geology, The University, Sheffield SI 3JD
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
Marketing Manager: Dr. R. J. Aldridge, Department of Geology, The University, Nottingham NG7 2RD
Editors
Dr. K. C. Allen, Department of Botany, The University, Bristol BS8 1UG
Dr. M. G. Bassett, Department of Geology, National Museum of Wales, Cardiff CF1 3NP
Dr. D. E. G. Briggs, Department of Geology, Goldsmiths’ College, London SE8 3BU
Dr. L. B. Halstead, Department of Geology, The University, Reading RG6 2AB
Other Members of Council
Dr. C. H. C. Brunton, London
Dr. E. N. K. Clarkson, Edinburgh
Dr. D. Edwards, Cardiff
Dr. R. Harland, Leeds
Dr. P. D. Lane, Keele
Dr. A. R. Lord, London
Dr. J. Miller, Edinburgh
Dr. T. J. Palmer, Aberystwyth
Dr. D. J. Siveter, Hull
Dr. P. W. Skelton, Open University
Dr. P. D. Taylor, London
Dr. A. Thomas, Birmingham
Dr. H. S. Torrens, Keele
Dr. N. H. Trewin, Aberdeen
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
U.S. A: Dr R. Cuffey, Department of Geology, Pennsylvania State University,Pennsylvania, U.S. A.
South America: Dr. O. A. Reig, Departmento de Ecologia, Universidad Simon Bolivar, Caracas 108, Venezuela
Palaeontology
VOLUME 25 • PART 2
CONTENTS
Ecology and population structure of the Recent brachiopod
Terebratulina from Scotland
GORDON B. CURRY 227
A new zosterophyll from the Lower Devonian of Poland
DANUTA Z DEB SKA 247
Two salenioid echinoids in the Danian of the Maastricht area
j. F. geys 265
Devonian miospore assemblages from Fair Isle, Shetland
J. E. A. MARSHALL and K. C. ALLEN 277
Conodonts, goniatites and biostratigraphy of the earlier Carboni-
ferous from the Cantabrian Mountains, Spain
A. C. HIGGINS and C. H. T. WAGNER-GENTIS 313
Liassic plesiosaur embryos reinterpreted as shrimp burrows
RICHARD A. THULBORN 351
Limpet grazing on Cretaceous algal-bored ammonites
ETIE BEN AKPAN, GEORGE E. FARROW, and NOEL MORRIS 361
A review of Recent and Quaternary organic-walled dinoflagellate
cysts of the genus Protoperidinium
REX HARLAND 369
A new genus of shark from the Middle Triassic of Monte San
Giorgio, Switzerland
o. rieppel 399
A new species of the fish Amia from the Middle Eocene of British
Columbia
MARK V. H. WILSON 413
A fused cluster of coniform conodont elements from the late
Ordovician of Washington Land, Western North Greenland
R. J. ALDRIDGE 425
A new calcareous green alga from the Middle Jurassic of England:
its relationships and evolutionary position
GRAHAM F. ELLIOTT 431
A rare lytoceratid ammonite from the Lower Lias of Radstock
D. T. DONOVAN and M. K. HOWARTH 439
Printed in Great Britain at the University Press, Oxford
by Eric Buckley, Printer to the University
issn 003 1-0239
ITUTtON NOliniliSNi NVINOSHJLI1NS S3IHVHen LIBRARIES SMITHSONIAN INSTITUTION NOI1
z ~ \« z. r- z r~ 2
m 2 m ~ * ' m ^ wj-J^ *2 >s
CO f , t ,
~ W '* E tO ~ CO ”
ivaan libraries Smithsonian institution NoiiniusNi nvinoshjjws ssiavaan libi
z w 2 CO 2 w . « V w a I I
< ^ S S S <
tution NoiinniSNi_ nvinoshiiws saiavaan libraries Smithsonian institution noii
<2 „ — ^ z \ 40 ^ ^ 5 oo ... = oo
jvaan libraries Smithsonian institution NoiiniusNi nvinoshliws saiavaan lib
“ z r- . z r- z “
ITUTION NOlinilJLSNI NVINOSHJLIWS S3iavaail LIBRARIES SMITHSONIAN INSTITUTION MOIJ
in 2 to 2 .... to 2 ... co
O
z
> "4^ S v. >
avaan libraries Smithsonian institution NoiiniusNi NviNOSHiiws^saiavaan lib
5 ” v 5 NSjypcjy " 5 “ ^oIu%^ O
itution2 NoiiniusNi "VjviNOSHims^sa lavaail^LIBRARI eszsmithsonian-,institution2noij
Z r- Z r- 2 f“_z
yosy^X E? m ^ m S£ * m to
jvaan i
(fe I
7 ' 1 ' \V rr VL,rt3L>^ I ' I -r ' * ' ~r
~ CO £ CO ? CO £
jvaan libraries Smithsonian institution NoiiniusNi nvinoshiiws S3iavaan libi
^ <£ ^ . z > to z to z
s i s i •* ' ^ ^ i I
OO A- 2 ^ •-•« 2 to Z w
ITUTION NOliniliSNi _ NVIN0SH1IWS S3iaV08n LIBRARIES SMITHSONIAN INSTITUTION NOII
CO ~ CO — CO
&JP s iOP I s I1&JP I lOp IImP « if
O _ O
z — _i z
jvaan libraries Smithsonian institution NoiiniusNi nvinoshiiws saiavaan lib
z •” — z r- z r_
Oin s2iz m ^ m *2 m
sjviNOSHims S3iavaan libraries Smithsonian institution NOimillSNI nvinoshuins.
n, Z r~ Z r- z ^ > ■
O — Z-iau7N O O
SMITHSONIAN INSTITUTION NOimillSNI NVIN0SH1MS S3IBVHan LI B R AR 1 ES SMITHSONIAN
z c/> z , <2 z
7 z
>' S x-vosj-ix 2 ‘".W > 2
(« ••• 2 CO Z CO - Z </)
nvinoshiiws saiavaan libraries Smithsonian institution NouniiiSNi nvinoshhws
- w 5 '
J <
;Qva5*S>''/ O 5 ’ O ~ '^viVA5J^/
z
SMITHSONIAN INSTITUTION NOIlfUllSNI NVIN0SH1I1NS S3 I U VH 3 H_LI B R AR I ES SMITHSONIAN^
P o I
rn «; rn I" m xg\ pcz ^ 'f^p" m
co _ c/> \ fE co — w
NVIN0SH1IINS S3iavnan LIBRARIES SMITHSONIAN INSTITUTION NOIlfUllSNI nvinoshiiws
CO Z CO 2 v.„ CO z
5 v > 2 ^ >
Z CO * Z £/) v z
SMITHSONIAN INSTITUTION NOIlfUllSNI NVINOSH1IIMS S3IHVyail LIBRARIES SMITHSONIAN
O X^AI pi>- _ O ~ Q _ •*??«" O
NVIN0SHimSZS3 I y vy a n^LI B RAR I ESZSMITHSONIAN~,|NSTiTUTIONZNOIlfUllSNl“VlNOSHimS.
z r- Z r- 2 r-
o zCstit/T^x ~ O _ /oa VnX O
SMITHSONIAN INSTITUTION NOIlfUllSNI NVIN0SH1MS S3iyvyail LI B RAR I ES SMITHSONIAN
z > co z co z > 52 ^ — . z
2 >• 2 >’ 2 s
CO ' Z CO z to z </
nvinoshiiws S3iyvyan libraries smithsonian institution noiidiiisni_nvinoshiiws
W =: co - CO 2 '
Kk h am - /#w\ g "%% D
o
-1 z _J z
SMITHcnNIAN INSTITUTION NOimillSNI NVIN0SH1IWS S3iyvyaiT LIBRARIES SMITHSONIAN
Z - Z _ H g [
a
Live Music Archive
Librivox Free Audio
Metropolitan Museum
Cleveland Museum of Art
Internet Arcade
Console Living Room
Books to Borrow
Open Library
TV News
Understanding 9/11