PA ou

‘ j

BO in git Legis Ae?

tier a

‘(ie aa oP j

THE: rr 20 VE ace tt ig Bee, as: ; Bear FURAL— + i SR a) ooh rtheaenaahrr va pice aihvieedih” a ag

' a ie Ec dy

has a

S74 \

3 it ae ¥)

BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY)

GEOLOGY VOL, 21 1972-1973

TRUSTEES OF THE BRITISH MUSEUM (NATURAL HISTORY) LONDON: 1974

DATES OF PUBLICATION No. No. No. No. No. No.

30 June 29 December I4 June . 31 October 1 November 29 December

PUR wDH

PRINTED IN GREAT BRITAIN BY JOHN WRIGHT & SONS LIMITED, AT

OF DHE PARTS

1972 1972 1973 1972 1972 1972

THE STONEBRIDGE PRESS, BRISTOL BS45NU

i

CONTENTS

GEOLOGY VOLUME 21

PAGE The shell structure of Chonetacean brachiopods and their ancestors. C. H. C. BRUNTON I

Postcanine occlusion in Cynodonts and Tritylodontids. A. W. CROMPTON 27

The Lower Miocene ruminants of Gebel Zelten, Libya. W. R. HAMILTON 73

The affinities of Halcyornis from the Lower Eocene. C. J. O. Harrison & C. A. WALKER I51

Dinoflagellate cysts and Acritarchs from the Kimmeridgian (Upper Jurassic) of England, Scotland, and France.

G. U. Girmez & W. A. S. SARJEANT 7 [See also Bull. Br. Mus. nat. Hist. (Geol.), 18 (7) 1970 : 231-331] Mid-Tertiary Cytherettinae of north-west Europe. M. C. KEEN 259

Index to Volume 21 351

IETACEAN BRACHIOPODS

: { (NATURAL HISTORY) _ : Vol. 21 No. E

ee

eel a uel at es 01

eh

We

aie SHELL STRUCTURE OF CHONETACEAN BRACHIOPODS AND THEIR ANCESTORS

BY

HOWARD BRUNTON

Pp. 1-26; 9 Plates, 8 Text-figures

BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) GEOLOGY Vol. 21 No. 1 LONDON : 1972

THE BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY), instituted im 1949, ts issued in five series corresponding to the Departments of the Museum, and an Historical series.

Parts will appear at irregular intervals as they become veady. Volumes will contain about three or four hundred pages, and will not necessarily be completed within one calendar year.

In 1965 a separate supplementary series of longer papers was instituted, numbered serially for each Department.

This paper is Vol. 21, No. 1 of the Geological (Palaeontological) series. The abbreviated titles of periodicals cited follow those of the World List of Scientific Periodicals.

World List abbreviation Bull. Br. Mus. nat. Hist. (Geol.).

© Trustees of the British Museum (Natural History), 1972

TRUSTEES OF THE BRITISH MUSEUM (NATURAL HISTORY)

Issued 30 June, 1972 Price £2-00

BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) GEOLOGY err Vol. 21 No. 1

\ Wi

ADDENDA ET CORRIGENDA

p. 6, caption to Fig. 2 The inhalant current arrow should have a flight on its tail, the exhalant current arrow no flight.

p. 7, paragraph 3, line I For “‘teleolae’’ read “‘taleolae’’.

p. 9, caption to Fig. 3B, line 3 An arrow without a flight should be inserted before “‘exhalant current’’.

p. 24 For “GRAnrT, R. E. (in press)” read “GRANT, R. E. 1972. The lophophore and feeding mechanism of the Productina (Brachiopoda). J. Paleont., Tulsa, 46:

213-48, pls 1-9”.

P. 25 The reference after WiLiiams, A. 1968, should be WiLLiAMs, A. 1969. The book was published by the University of Wales Press not “‘aloes’’. The reference following should be WiLLiAMs, A. 1970.

P. 25 Under “Locality details of figured specimens”, the second name should read “Leptelloidea leptelloides”’ not ‘‘Leptelloides leptelloides’’.

by ae : , - 2 aa AY oe & - pat \ i

tf ce 7 ets Orgs)... oe Brio tA

_ ‘

o ae ¥ we

‘+ eh YD ' c

i : a6

ta _

; . ae — - i

ad

eu 7 —— a A _— ' ® =, & fiona) a 7 i

pi

|,

i

mae SHELL STRUCTURE OF CHONETACEAN BRACHIOPODS AND THEIR ANCESTORS

By C. H. C. BRUNTON

CONTENTS Page I. INTRODUCTION 5 : : : ¢ : 9 ° : 3 Il. MorpHoLocy . : c : : . ; ; : 4 Ill. ANCESTRAL STOCKS : : : : : 2 i : 9 IV. PLECTAMBONITACEA 6 : : : : 6 5 : 10 V. SHELL STRUCTURE . : E : : : . : ¢ 13 Plectambonitacea é é : 5 : : . 13 Chonetacea. ° : é : : : é 5 16 VI. DISCUSSION AND CONCLUSIONS . ¢ : 4 : “ . 20 VII. ACKNOWLEDGEMENTS : : ; : . : . c 23 VIII. REFERENCES . : : : : . ° F . : 24 SYNOPSIS

The criterion of shell microstructure has been added to general morphological considerations in this study of possible ancestors for the Chonetacea. The results support Havlicek’s suggestion that the sub-family Aegiromeninae was the plectambonitacean stock from which the first chonetacean evolved in the late Ordovician. Ordovician to Permian chonetaceans have been studied and the observed changes in their shell structure are discussed. Functional morphological interpretations are presented and consideration is given to the significance of spines in these brachiopods.

I. INTRODUCTION

Dur1nG the past ten years or so there have been several notable publications on the Chonetacea. The origin and phylogeny of the group has always been problematical, for, as Muir-Wood wrote in 1962, ‘More research requires to be done . . . if the relationship of these very numerous and wide-spread forms is to be established.’ She briefly discussed chonetacean origin indicating that she favoured the Plectam- bonitacea as ancestors, but reminded readers of Paeckelmann’s suggestion (1930) that the chonetids were derived from a strophomenid stock.

More recently Boucot & Harper (1968) have revised Silurian and Lower Devonian Chonetacea, but restricted themselves to phylogeny within the group and made no comment upon ancestry. However, Havli¢ek (1967) while discussing the evolution of the Sowerbyellidae ‘. . . assumes Chonetoidea to be incontestibly the direct precursor of the superfamily Chonetacea...’. Havliéek derives Chonetoidea from an Aegivomena stock whilst Eochonetes is derived from the Sowerbyella stock. In his view, therefore, Eochonetes is not in the group ancestral to the chonetaceans as I had implied in 1968.

Thanks to the stimulus provided by the work of Williams, especially his shell structure study of 1968 which formulated a framework for future research, we are now in a better position to apply detailed examinations of shell microstructure to

4 SHELE STRUCTURE

the problems of phylogeny in a meaningful fashion. Williams interpreted the shell microstructure of the Chonetidina, Productidina and Strophomenacea as being fundamentally the same, (i.e. entirely laminar) while that of the Plectambonitacea differs in retaining a fibrous secondary layer. In his phylogenetic chart Williams (1968) derived the chonetids, productids and oldhaminids from the Strophomenacea, which in turn, along with the Davidsoniacea and ‘probably the triplesiidines’ he derived from the Plectambonitacea by a process of neoteny: the laminar shell of the strophomenides being homologized with the laminar primary layer of the plectambonitaceans.

We are left, therefore, with the need to investigate the possibilities of chonetids being derived either from plectambonitaceans involving a change of shell structure, or from the strophomenaceans, with which they appear to have a common shell structure.

Clearly, in any such study the more factors investigated the better. I believe the microstructure of the brachiopod shell to be so intimately bound up with the metabolism of the living organism as to be of profound importance systematically ; nevertheless, it is vitally important to consider closely the gross morphology of the valves so as not to suggest evolutionary relationships involving highly improbable morphological changes. The purpose of this study is the detailed investigation of the chonetacean shell microstructure, together with that of their possible ancestors, in the hope of resolving the early phylogeny of this group, the evolution of which took place during Upper Ordovician times.

I retain here the view expressed in 1968 that the Chonetacea should be classified more closely to Productacea than, for instance, to the Strophomenacea. The removal of the Cadomellacea from Muir-Wood’s suborder Chonetidina (Cowen & Rudwick 1966) leaves only the Chonetacea, and, at present, I am in favour of leaving the situation fluid to the extent of retaining the Chonetacea as a superfamily within the Strophomenida and not using the term Chonetidina. The Productidina usefully unites several superfamilies which have reasonably established morphological characteristics in common. The Strophomenida as a whole is an order within which there is room for phylogenetic reappraisals, e.g. the removal of Thecospira and Cadomella (Brunton & MacKinnon, in press), and introduction of the Triplesiacea and Thecideacea, problems upon which several palaeontologists are engaged and have already commented (Cowen & Rudwick 1966, Williams 1968, Rudwick 1968, Wright 1970, Mr. D. McKinnon personal communication January 1971).

II. MORPHOLOGY

The most recent and complete description of chonetaceans is that of Muir-Wood (1962). Several aspects of their morphology have been discussed (Brunton 1968) but further work now necessitates some revision.

The main point of departure from the views expressed in 1968 concerns the possible reconstruction of the lophophore (1968, text-fig. 50). At that time the traditional idea of a spirolophe for chonetaceans was followed. However, more recent studies on this group, productaceans, strophalosiaceans and thecideaceans

CHONETACEAN BRACHIOPODS 5

leads to the opinion that the lophophore in these fossil groups consisted of a schizolophe or variously lobed forms of a ptycholophe. In the sense of Williams & Rowell (¢w Williams et al 1965 : H37-8) it is a lobed trocholophe because the lophophore is believed to have had only a single series of filaments, as in Recent Thecidellina, rather than the more complex double filaments characteristic of most other Recent lophophores from the schizolophous to plectolophous stages. Rudwick (1968) reminds us that the terminology employed to describe lophophores was based originally upon the arrangement of the brachial axis without consideration for the number of filament rows. Thus whilst phylogenetically there is logic in Williams and Rowell’s terminology (1965, H38), it is less confusing to retain the previously existing definitions qualified, where necessary, to indicate if the filament series is double or single. Believing that the thecideaceans may be remnants of the Strophomenida, derived from a productidinid stock not far removed from Cooperina Termier, Termier & Pajaud 1967 (but see Cooper & Grant I969), it is logical to interpret the chonetacean lophophore as having been only a single row of filaments (Text-fig. 1). In his study of the Triassic brachiopods Thecospiva and Bactrynium

Fic. 1. Diagrammatic dorsal valve interior of a chonetacean showing, on the left, general morphology and, on the right, the inferred anatomy and suggested water currents associated with the lophophore, a. — anderidium; a.ad. — anterior adductor muscle scar; ad.v. — adductor muscle, ventral attachment area; b.p. — brachial platform; b.w. — inner epithelium of the body wall separating the brachial cavity from the visceral cavity. The epithelium of the visceral cavity has been cut ventro-medianly to expose the muscles; did. — diductor muscle; did.v. — ventral attachment area of the diductor muscle; m. — mouth; m.e. — mantle epithelium; m.s. — median septum; n. — possible position of the nephrostome in the body wall, excreting into the posterior exhalant current; p.ad. — posterior adductor muscle scar; s. —socket; s.r. —socket ridge ;3-> inhalant and — exhalant feeding/respiratory water currents.

6 SHELL STRUCTURE

Rudwick (1968) discussed their affinities with the Thecideacea and the general phylogeny of that superfamily. He accepted the Termiers’ & Pajaud’s (1967) assignment of the Permian genus Cooperina to the Thecideacea, so that it became the earliest representative of the superfamily, which he considered as derived from a Lower Permian or even earlier davidsoniacean. Study of Cooperina shows that Cooper & Grant (1969) are correct in placing the genus in the Strophalosiacea and it seems that the morphological evidence presented by Rudwick for a derivation from the Davidsoniacea is better met by the Strophalosiacea, a group believed to have diversified widely and to have given rise to both the Richthofeniacea and Lyttoniacea. Rudwick (1968, 1970) related the Lyttoniacea and Thecideacea to the Davidsoniacea.

While preparing this paper I was privileged to be shown the script of a paper by Grant (in press) in which he describes a calcified ‘ptycholophous brachiophore’ in Permian Productidina (ptycholophous here used as by Grant in the morphological sense of being several lobed). The interpretation of this structure supports the contention that the brachial ridges of Productidina do indeed mark the positions to which the lophophore was attached in the living animal and that the ‘ptycholophous’ (or multilobed trocholophous) lophophore described by Grant evolved as a specialization from the more generalized schizolophe or simple ptycholophe characteristic of many Strophomenida.

A study of particularly well preserved Pennsylvanian and Permian chonetacean interiors in the collections of the National Museum of Natural History, Smithsonian Institution, Washington, shows that in some rugosochonetids and species of Dyoros the antero-median tuberculation of the brachial ridges merges into the anterior end of the median septum. In some thickened (? old) specimens of Dyorvos the anterior end of the median septum is winged laterally in such a way as almost to bridge the gap between it and the brachial ridges (Text-fig. 2). The exaggerated development of tubercules and prolonged deposition of secondary shell in the regions of the brachial

P br. ae Aas ici may Aa

5mm

Fic. 2. Anterior view of Dyorvos sp. (Chonetacea) dorsal valve from the Permian of Texas illustrating the highly tuberculate brachial ridges and anteriorly winged median septum. The inferred position of the lophophore and feeding water currents are added on the right. > inhalant current; 3+ exhalant current; b.p. — brachial platform; b.r. brachial ridge, m.s. — median septum.

CHONETACEAN BRACHIOPODS 7

ridges and median septum probably resulted from the stresses imposed by the lophophore upon the secretory mantle epithelium in these areas. It may have been simply the weight of the lophophore hanging from the dorsal valve that induced this additional shell secretion in those specimens living in a position with their dorsal valves uppermost.

The chonetacean lophophore is envisaged as having been suspended from the dorsal inner epithelium with the postero-median mouth segment attached to the body wall between and probably ventral to its support by the anterior tips of the anderidia (= lateral septa of Muir-Wood 1962). The strong tuberculation of the brachial ridges probably reflects the strong attachment of the mantle to the shell where it supports the lophophore (Text-fig. 3). Antero-medianly the generative tips of the lophophore probably recurved posteriorly onto the median septum (Text-fig. 1); however, it is impossible to say whether they remained separated by the septum or united in the median plane as in living Megathiris.

The relationship between pseudopunctae and teleolae requires clarification. In the brachiopod volume of the Treatise (1965 : H420) Muir-Wood incorrectly redefined the term taleolae, specifically as applied to chonetids, saying that they had a ‘central cavity’. In observing many taleolae under the scanning electron microscope a central cavity has never been observed. It may be that the impression of such a structure resulted from optical effects or that the cavities were those of rib apertures.

Taleolae, as defined by Williams (1956), are rods of calcite in the axial position of many pseudopunctae (Pl. 1 figs 1, 2). Taleolae are found neither in all pseudo- punctae nor necessarily forming the core to the complete pseudopunctum in which they occur. When absent the shell layers or fibres can be traced into the centre of the structure, which is composed of inwardly and commonly anteriorly directed conical flexures producing tubercules on the inner surfaces of the valves (PI. I, fig. 4). This structure contrasts with the outward flexures of shell fabric surrounding endopunctae (Brunton 1969, fig. 11, 1971; pl. 11, figs 8, 10). The pseudopunctae of upper Palaeozoic chonetaceans have taleolae while those of geologically older specimens are without, or with only weakly developed taleolae. Baker (1970) describes Moorellina specimens within which the pseudopunctae differ in character; those of the dorsal valve have cores of primary shell resembling taleolae, whilst those of the ventral valve are without and composed entirely of radially disposed bundles of secondary fibres.

In 1968 the author’s text-figure 45 indicated that he interpreted the sub-median ridges in the ventral valves of some chonetids as being the traces of mantle canals. It is now believed that in some narrow-bodied species with thickened shell these ridges developed in the regions bordering the ventral edge of the dorsal median septum. In such specimens the brachial cavity of the closed shell would have been divided almost completely into two chambers. The significance of this is not clear and the relationship of these ridges to mantle canal traces is in doubt; the effect, however, is similar to that of mantle canals, viz. the pressure of a structure onto the internal epithelial surface inhibiting shell deposition with increased shell thickening on either side.

8 SHELL STRUCTURE

a.ad.

p.ad.

Fic. 3A. Ventro-lateral perspective view of the dorsal valve interior of Dyovos sp. from the Permian of Texas illustrating the surface morphology. a. — anderidia; a.ad. — anterior adductor scar; b.p. — brachial platform; b.r. — brachial ridge; c.p. — cardinal process; dm.f. — dorso-median fold; m.s. — median septum (here

strongly tuberculate) ; p.ad. — posterior adductor scar; p.m. — posterior margin of valve; s. — socket.

CHONETACEAN BRACHIOPODS 9

If one ignores the Daviesiellidae, in particular Daviesiella and Airtonia, and Chonostrophia (which may not be a chonetacean) the Chonetacea are morphologically conservative from their origins in the Upper Ordovician to the topmost Permian. Normally they share a gently concavo-convex profile, multicostellate ribbing, low interareas set at a wide angle from each other and several pairs of more or less posteriorly directed hollow spines on the posterior margin of the ventral valve. Teeth, sockets and socket ridges are developed, and in the dorsal valve the adductor scars commonly are divided by a ridge (anderidia). The cardinal process is low, internally bilobed in early forms but tending to become knob-like and commonly trifid externally; the median septum is only high anteriorly and developed late in ontogeny. The pseudopunctate shell manifests itself internally by tubercules. At least in early stages of ontogeny several genera were attached to the substrate by a thin apical pedicle protected by a shelly sheath.

Til. ANCESTRAL STOCKS

At first sight certain strophomenaceans seem suitable as ancestors to the Chonetacea; general shape and size of some stropheodontids seem correct, as are features such as the dorsal valve protegulal node and development of a pedicle sheath on such genera as Pholidostrophia. Williams’ 1968 investigations indicated that the shell structure is comparable in both groups, 1.e., essentially laminar. However, it is shown below that the shell structures differ and the above character- istics are common to the Strophomenida.

Previous morphological studies have shown that many features are shared by the Plectambonitacea and Chonetacea with the result that several palaeontologists (e.g. Chao 1928, Sarycheva & Sokolskaya 1959, Muir-Wood 1962, Havlicek 1967) have thought these superfamilies to be phylogenetically related.

Strophomenacea, Plectambonitacea and Chonetacea obviously share strophomenid characteristics, but whilst doing so certain features of the Chonetacea tend to be more akin to those of the Plectambonitacea, especially the Sowerbyellidae, than to the Strophomenacea. Thus the Sowerbyellidae and Chonetacea share similar shell profiles and outlines and do not include the same elaboration of shell shapes met

Fic. 3B. Inferred principal anatomy added to the above specimen (in red) and possible ciliary induced water-current through the lophophore (in blue). 3+ Inhalant current; exhalant current; ad. — adductor muscle, divided dorsally by anderidium; b.w. — body wall with left side removed to expose visceral cavity; did. — diductor muscle; f.r.l. — filaments of right lophophore lobe; g.t. — generative tip of lophophore; 1.1.1. — left lobe of lophophore; m.l. — mouth (median) segment of lophophore, cut on left to expose muscle bases; r.1.1. — right lobe of lophophore; v.a. — left visceral region; v.v.c. — position of the interior of the ventral valve when the shell was closed; v.v.o. — inferred relative position of the ventral valve when the shell was open by about 20°. N.B. This reconstruction is not intended to imply a ventral movement of the ventral valve when the shell opened. Other than a portion of the body wall, epithelial layers, such as the mantles, are not depicted and it should be remembered that epithelia would have covered all the internal shell surfaces illustrated.

10 SIENA IEAL, Sb ee OCC 7G (Ol 1

within the Strophomenacea. Similarly with external ornamentation, save that accentuated costae common to several sowerbyellid genera are only hinted at in the oldest chonetacean genus, Stvophochonetes.

Early and mid-Ordovician Plectambonitacea are more varied in shell shape and it seems that the Strophomenacea inherited this tendency to variety. By upper Ordovican times shell shape was more stable in Plectambonitacea and from such a stock the Chonetacea inherited their conservative outline and profile.

Within the dorsal valves of sowerbyellids are paired ridges lateral to the mid-line or median septum, which is low or absent posteriorly leaving a cavity at the base of the cardinal process; a cavity interpreted as that of the brephic valve. These features are more common to the chonetaceans than to strophomenaceans. In Plectambonitacea the ridges (‘inner’ and ‘outer side septa’ and ‘bema’ of Cocks 1970) may be homologized with the anderidia, accessory septa and brachial ridges of chonetaceans. The pit at the base of the cardinal process is the alveolus of chonetaceans.

Whilst some Strophomenacea share some of these features, as well as denticulate hinge lines, their general combination is more in keeping with Sowerbyellidae and Chonetacea. Furthermore, some mid- and upper Ordovician Sowerbyellidae, e.g. Eochonetes, Chonetoidea, possibly Sentolunia, have hollow canals in the ventral valve posterior margin which are closely comparable to the spine canals of chonetaceans. Havlicek (1967 : 38) suggested that these canals may have accommodated hold-fasts used to attach these plectambonitaceans to seaweed.

On looking at mid- to upper Ordovician faunas for possible chonetacean ancestors it seems that the morphological requirements may best be met by the Sowerbyellidae.

IV. PLECTAMBONITACEA

Since this study was started Cocks has published on Silurian Plectambonitacea (1970). His paper contains useful and interesting discussion on functional morphology, but not all his suggested reconstructions of the musculature and feeding mechanisms are accepted here. These operations have important implications upon the way in which the plectambonitacean/chonetacean shells are envisaged as having been organized. Whilst agreeing to the possibility of the shells being able to snap shut as a defence mechanism and possibly also as a repositioning mechanism, it is difficult to envisage a brachiopod habitually feeding by pumping water through its brachial cavity by means of a flapping valve system such as proposed by Rudwick (1961) for Richthofenacea and invoked by Cocks. A ciliary induced water current seems to be a well tried, stable and energy conserving system widely used in invertebrates and invariably used in extant brachiopods. By whatever means a water current is produced its function is to provide for respiration and feeding. Evidence suggests that a lophophore is required for these purposes, even allowing for the possibility of feeding on dissolved nutrients as proposed by McCammon (1969) and it seems likely that the plectambonitaceans, even the structurally specialized Eoplectodonta, retained a lophophore capable of a normal ciliary beat.

It may be unwise to compare a living genus so widely separated from the

CHONETACEAN BRACHIOPODS Il

Plectambonitacea taxonomically, but the dorsal internal morphology of Megathiris has, what are believed to be, analogous structures. In Megathiris the quadrilobed ptycholophe, with a single series of filaments, is supported by a calcareous loop partially fused to three ridge-like pillars which are prominent anteriorly. The loop, and thus the lophophore, is a few millimetres behind the anterior faces of these ridges (see Treatise 1965, H 836 for fig.). The lophophore does not project pos- teriorly along these ridges to the extent that might be supposed from Atkins’ figure 6 (1960, —her figure 7 gives a clearer impression of the true situation) which is reproduced in the Treatise (Williams et al 1965, fig. 41). The body wall, behind which the muscles and viscera are situated, extends antero-dorsally between these ridges as a dissected plane at approximately 45° to the commissural plane and following the postero-dorsal side of the loop. In this way the anteriorly exaggerated median septum and pair of ridge-like pillars lift the lophophore from the dorsal valve floor allowing the antero-median extension of the body cavity, including dorsal adductor muscles. It is thought that the plectambonitacean morphology, exempli- fied by Eoplectodonta, achieved similar results.

The socket ridges, of Williams, or clavicular plates, of Cocks (1970) are considered to have functioned as postero-lateral supports to the body wall in the region of the mouth segment of the lophophore, much as were the opinions of Kozlowski (1929) and Opik (1933), (As the principal points of pivot in these shells occurred at the posteromedian surfaces of these structures the term socket ridge is favoured). In this respect Eoplectodonta displays a feature common to many of the articulate brachiopods, that of a close relationship between articulation and support of the body wall in the region of the lophophore.

The plectambonitacean lophophore probably followed the lateral edges of the bema (Cocks), or lophophore platform (Williams), so that a variously modified ptycholophe, in which the generative zone (or zones) recurved postero-medianly, was suspended from the dorsal mantle. Assuming a ciliary induced water current from the brachial lip across the filamentous area of the lophophore, a circulation may have been achieved in which water entered ventrally, perhaps particularly medianly, and passed out dorsally, close to the dorsal valve and especially laterally (Text-fig. 4). There seems little good reason why many of the later plectambonita- ceans, those that had reduced their teeth, could not have had a wide gape while feeding. A wide gape might be advanced as the reason for very large, anteriorly extended, dorsal adductor muscle scars (Text-fig. 5), rather as is the situation in Megathiris today, which opens to about 45° or Thecidellina opening more than 60°. Whether or not the dorsal adductor scars of Eoplectodonta covered the bema, as suggested by Cocks (1970), it seems clear that these areas and those between the two pairs of septa on the dorsal valve, accommodated body tissues and that the lophop- hore was elevated on these septa. However, as in Megathiris, the lophophore probably did not follow the septal crests because in heavily thickened shells the septa may touch the interior of the ventral valve when the shell is closed, leaving little or no space for the brachial axis.

The morphology of the geologically older (Ordovician) plectambonitaceans is indicative of a schizolophe and perhaps only in geologically younger members of

12 SHELL STRUCTURE

the Sowerbyellidae did the lophophore evolve into a quadrilobed structure.

In various Ordovician species of Anoptambomites and Bimuria there is muscle scar evidence for the dorsal adductors being restricted posteriorly, in a more traditional position, and these scars are divided by raised areas which may be homologized with Cocks’ ‘outer side septa’ and possibly with the anderidia of chonetaceans.

It is believed that the Sowerbyellidae, particularly Eoplectodonta and Plectodonta, were specialized plectambonitaceans which nevertheless retained essentially normal systems of feeding. Their ‘cousins’, the Aegiromeninae remained more generalized and are morphologically more suitable to have provided the stock from which chonetaceans evolved.

The Aegiromeninae tend to be small-sized shells, commonly about Io mm. wide, gently concavo-convex with shorter interareas and more regular ribbing than the Sowerbyellinae. Socket ridges are reduced and the dorsal median septum does not extend posteriorly to the cardinal process but appears to be flanked by the adductor

Fic. 4. Stylized illustration, based on Bimuria siphonata Cooper, from the mid-Ordovician of Pratt Ferry, Alabama, showing the internal dorsal valve morphology on the left with the inferred lophophore and main muscles on the right (red). The blue arrows indicate the main circulation of water through the lophophore, and this is further illustrated by the small diagram of an open shell, viewed posteriorly, on the right. (The dorsal valve is uppermost and water enters from the front) a.ad. — anterior adductor scar; ad.v. — ventral attachment area of adductor muscle; b.p. — brachial platform; b.r. — brachial ridge; did. — diductor muscle; m. — mouth; p.ad. — posterior adductor scar; s.r. — socket ridge.

CHONETACEAN BRACHIOPODS 13

muscle scars and variously placed elongate tubercules which may have assisted in the support of the lophophore. There is a complete lack of the strong dorsal internal ridging typical of the Sowerbyellinae and Leptellinidae and this morphology is entirely suitable as being ancestral to the sparsely featured early chonetacean dorsal interiors.

Woe GSISNZILIL, Gi wires

Plectambonitacea

Morphologically the Sowerbyellinae form a closely knit subfamily. The shell structure of the genera investigated (Sowerbyella, Viruella, Eoplectodonta, Thaer- odonta, Plectodonta and Eochonetes) supports this unity. In these genera the secondary shell layer is standard in that the fibres show an internal mosaic (PI. 1, fig. 4) and the typical cross-sectional stacking (PI. 2, figs 1, 2) familiar within Recent terebratulids and rhynchonellids. The outer primary layer appears to be more variable, thin and commonly poorly preserved, and may be differentiated simply as a layer of much smaller ‘fibres’ (Pl. 2; figs 3,6). These outer elements do not seem to show the brick-like cross-section or lateral fusion that would be expected in a lamellose fabric. Taleolae are not strongly developed in the pseudopunctae of these shells.

Itc. 5. Median longitudinal section of Eoplectodonta showing the inferred adductor and diductor muscles as they might have been when the shell was shut (5A) and open with a gape of 45° (5B). The ventral attachment area of the adductor muscle (coarse stippling) is close to the median plane whilst the dorsal attachment area is between the submedian septum (or inner side septum), here omitted, and the outer side septum, seen beyond the muscle.

14 SHEN ICAL, Sab IR WG Ae RIS,

An unexpected structure shows in the shell of particularly well preserved Eoplectodonta transversalis (Dalman) specimens from the uppermost Llandovery of Gotland. The pseudopunctate secondary layer also has small endopuncta-like canals of about 3 ym diameter surrounded by small outward deflections of the fibres producing a cone-in-cone structure (Pl. 3, figs 2, 3) contrasting with that of the pseudopunctae. It is not yet known to what extent these small endopuncta-like structures pervade the shell and it has only been possible to trace any one of them over a distance of about 80um through the secondary layer. They run subparallel to the pseudopunctae and it seems, therefore, that they were controlled by anteriorly migrating points of outer epithelium as distinct from the fixed positions of caeca around which the epithelium moved.

The subfamily Aegiromeninae seems rather more varied in its shell structure, as judged by evidence from Aegivomena, Aegivia and Sericoidea. Within this subfamily the shell structure differs from other Sowerbyellidae.

In Aegivomena aquila (Barrande), from the middle Ordovician of Czechoslovakia, the secondary shell is not entirely composed of standard fibres. Whilst retaining a well-separated, independent appearance, as if having been encased within organic sheaths during life, the typical fibre cross-sectional shape has almost been lost; only in rare instances (Pl. 3, fig. 5) can such fibres be distinguished and these tend to be towards the outer surfaces of the valves i.e. they were formed at early stages in the growth of the shell. Normally the fibres are about 14um wide and elliptical in cross-section, their edges overlapping adjacent fibres to various extents (PI. 3, fig. 5). The shell fabric is strongly pseudopunctate and these normally have taleolae (Pl. 4, figs I, 2). Towards the external surface of valves the fibres are of a slightly smaller dimensions and tend to be thinner. A strongly differentiated primary layer has not been recognized, if indeed it ever existed, but these smaller external fibres may indicate a gradation from a thin laminar primary layer to the fibrous secondary layer.

Aegiria grayi (Davidson) from the Wenlock Shales of Dudley, England, is sparsely pseudopunctate and the fibres of the secondary layer retain a rather more standard appearance (PI. 4, fig. 3). In these respects the species is somewhat more akin to the Sowerbyellinae, but the general morphology would not warrant a change to this subfamily. Primary shell was not distinguished in the material studied.

In Sericoidea restricta (Hadding) from the Caradoc of Girvan, Scotland, the sparsely pseudopunctate secondary layer shows virtually no sign of retaining standard fibres. The ‘fibre’ units within the shell appear to be well separated, as if formed in the standard manner within organic sheaths, and are of comparable dimensions (25-30 um wide and 3-4 um thick). Orientation of the ‘fibres’ remains ~ subparallel from layer to layer (Pl. 4, fig. 4, Pl. 5, fig. 1), thus retaining the organization of the standard regime rather than the marked alteration in the orientation of blades in adjacent sheets typical of many Strophomenida (see Armstrong 1969). Again, a well-differentiated primary layer has not been discovered unequivocally. Recrystallization is most common at the shell surfaces and pressure solution of the enclosing sediments interferes with the external shell fabrics. However, over certain areas of the valve’s exterior a layer of small laminae

CHONETACEAN BRACHIOPODS 15

can be seen (Pl. 5, fig. 2). These are only 3-4 um wide, appear to grade within one or two layers into the full-sized fibres, and probably constitute a remnant primary layer.

We have, therefore, within the Aegiromeninae a differentiation of at least the secondary shell layer away from the standard parallel fibrous fabric typical of geologically older plectambonitaceans (Ajhtiella, Inversella, Toquimia, Leptestia, Leptelloidea, and Bilobia), genera studied to demonstrate the standard nature of the early plectambonitaceans (PI. 5, figs. 3, 4) and which continued on within the Sowerbyellinae. Ptychoglyptus and Xenambonites have not been studied.

It is Havli¢ek’s opinion (1969 : 38) that the sub-family Aegiromeninae is the group from which chonetaceans evolved. Further, he assumed ‘Chonetoidea to be incontest- ably the direct precursor of the superfamily Chonetacea in which canals extended posteriorly into long hollow spines’. Unfortunately it has been impossible to find Chonetoidea specimens suitably preserved for the study of their shell, but morphological considerations support Havli¢ek’s opinion that Chonetoidea evolved from a Sericoidea-like ancestor.

It is necessary, therefore, to test this suggested phylogeny against the shell structure of the oldest known chonetaceans. The oldest undoubted species is Chonetes (Eochonetes) primigenius Twenhofel (1914) from Anticosti Island, Canada. Twenhofel recorded the species from four formations, the Charleton (= Vauréal) and Ellis Bay Formations of Richmond (high Ordovician) age, and the Gun River and Jupiter River Formations of Lower Silurian age. The holotype was figured from the Gun River Formation and a ventral valve exterior figured from the Charleton (Vauréal) Formation. It is still generally agreed that the Ellis Bay Formation is uppermost Ordovician in age. Dr. O. A. Dixon has been kind enough to send rock samples with this species from Mile 5, Juniper River and Mile 2 47 Mile Road, Anticosti, from the Ellis Bay Formation. Amongst these are several examples of dorsal valve interiors (Pl. 5, figs 5-8), figured for the first time here, and the shell substance is reasonably well preserved in the ventral valves. The species was assigned by Muir-Wood (1962) to her new genus Stvophochonetes. Boucot & Harper (1968) called into question the validity of both Stvophochonetes and Proto- chonetes of Muir-Wood (1963). A study of Lindstrém’s specimens of Stvophochonetes cingulatus (in the BM(NH) collections and used by Muir-Wood in defining the genus) and of Protochonetes ludloviensis Muir-Wood, type species of that genus, together with specimens of P. striatellus (Dalman) from the Wenlock of Gotland, a species very close to /udloviensis, convinces me of the separate identity of the two genera. In the author’s experience unabraded S. cingulatus and S. primigenius specimens always have a ventral median accentuated rib. The outline is relatively less wide than in Protochonetes ludloviensis or P. striatellus, and whilst spines may be abundant on Stvophochonetes (up to at least seven pairs) they extend more or less perpendicularly from the valve margin. Those of Protochonetes extend postero- laterally. A divided ventral median septum in Strophochonetes has never been observed.

It is perhaps significant that a collection, as yet undescribed, made by Dr. Cocks from the low Wenlock Knockgardner Beds of Girvan, Scotland, includes many

16 SHELE STRUCTURE

chonetacean specimens which appear to show characteristics intermediate between those of the types of Strophochonetes and Protochonetes. The specimens are small, approximately Io mm. wide, and in outline resemble Protochonetes; the ribbing is even but the ventral median septum is ill defined posteriorly. Spines seem to be variably disposed, some nearly perpendicular, others at an angle to the hinge line.

Chonetacea

The shell of S. primigenius is characterized by its parallel to subparallel arrange- ment of fibre-like elements (PI. 6, fig. 1, 2) which both overlap adjacent fibres laterally and, in other parts of the shell, abut to their neighbouring fibres with a more or less perpendicular plane of separation. Whilst the packing of these fibres is tight, they retain a discreteness and do not show signs of having fused laterally with adjacent units, as is the situation in the bladed and sheet fabrics of Armstrong (1969) or truly laminar fabrics of Williams (1968, 1970). These lath-like fibres are 6 to 10 ym wide and of variable thickness, but commonly between 2 and 4 ym thick. Pseudopunctae are sparsely developed in ventral valves (Pl. 6, fig. 1), but judging from the dorsal valve internal tuberculation pseudopunctae are more common in this valve.

A clearly differentiated primary layer has not been recognized, but towards the exterior of the valves the fibres have the appearance of Williams’ ‘crested lamellae’ (1968 Pl. 21, figs 2, 4). In Strophochonetes this structure may result from slight recrystalization of the outermost shell layers.

The seemingly separate nature of the shell ‘fibres’ and lack of sheet fabrics leads to the conclusion that the shell was laid down essentially in the standard way as proposed by Williams (1956, 1968), that is by individual outer epithelial cells secreting the calcite for individual fibres which were separated from one another by organic sheaths.

Other chonetacean records from Ordovician rocks are unsatisfactory. Study of the specimens recently referred to by Lister, Cocks & Rushton (1970) from upper Ordovician rocks of the Bobbing Bore, Kent indicates that they probably are chonetaceans. However, preservation is poor and the shell material is lacking or altered. Reed (1944) described a new species, Chonetes (Eochonetes) celtica, from the Upper Ordovician Balclatchie Beds of Ayrshire, Scotland, which was assigned to Strophochonetes by Muir-Wood (1962). Neither spines nor spine bases can be seen on the holotype in the Hunterian Museum, Glasgow and the acutely angular relationship of the interareas is much more suggestive of a plectambonitacean than a chonetacean.

From Middle Llandovery rocks of Newlands, Girvan, Scotland Dr. Cocks has found two ventral valves of a Strophochonetes species showing spines, but no shell is preserved. By mid-Silurian times chonetaceans were becoming more abundant, but their main diversification did not take place until the Upper Palaeozoic during which first the Plectambonitacea and then the Strophomenacea died out.

In addition to Strophochonetes the following chonetaceans have been studied for shell structure: Protochonetes striatellus (Dalman) from the mid Silurian of Gotland,

CHONETACEAN BRACHIOPODS 17

from where also comes a small chonetacean species, possibly Eoplicanoplia Boucot & Harper 1968; P. ludloviensis Muir-Wood from Upper Ludlow rocks of Eastnor, Hertfordshire; Dawsonelloides canadensis (Billings) from Lower Devonian rocks of Gaspé, Quebec; Retichonetes vicinus (Castelnau) from mid-Devonian Arkona shale of Ontario; Rugosochonetes species from Lower Carboniferous strata of County Fermanagh, N. Ireland; Mississippian of Oklahoma, and basal Namurian of Northumberland; Neochonetes from the Permian of Texas, USA, and specimens from the Permian of Russia.

In general the shell fabric of these later chonetaceans supports that seen in S. primigenius. Pseudopunctation, including well differentiated taleolae, became more strongly developed by the lower Devonian (PI. 7, figs. 1, 3) and continued within the stock. The greater part of the shell thickness retained a lath-like fibrous nature (PI. 7, figs 1, 2), although each ‘fibre’ was only from 2-4 ym wide and up to about Ium thick, until the early Devonian when there are clear signs of lateral fusion of ‘fibres’ (Pl. 7, fig. 3) into units 8-10 ym wide. In lower Carboniferous specimens, such as R. si/leest Brunton, while lath-like units are readily distinguishable throughout much of the shell their orientation from layer to layer is variable (PI. 7, fig. 4) and towards external surfaces sheets of blades are developed (PI. 8, figs 1, 2) into what approaches a true cross-bladed fabric (PI. 8, fig. 3).

Thus it seems that a trend away from the typical fibrous secondary shell of many Ordovician Plectambonitacea can be traced through members of the Aegiromeninae into the earliest known chonetaceans of the Lower Palaeozoic and on into the Upper Palaeozoic when chonetaceans were at their most abundant and diverse (Text-fig. 6). It seems, therefore, that within the Chonetacea the laminar shell fabric, like that of the Strophomenida other than the Plectambonitacea, developed indepen- dently from that in the Strophomenacea which, in Williams’ (1970) view, arose from a Cambro-Ordovician plectambonitacean-like ancestor derived from the nisusiid Billingsellacea (Text-fig. 7). This change in shell structure involved a reduction in the size of fibres indicating a reduction in the size of the secretory outer epithelial cells. This trend continued in the early chonetaceans, along with a loss in regularity and consistency in growth direction of the fibres at any one time or at different times during ontogeny. This may have resulted from the increased development of pseudopunctz to which small areas of specialized epithelium became fixed. In this way local areas of epithelium may have been retarded in their general anterior growth, so distorting the uniformity of calcite secretion in adjacent areas. Furthermore, an increasingly mobile epithelium, in terms of periodic retraction from the valve edges, would have resulted in the likelihood of renewed forward growth taking place in slightly altered directions and consequently the non-alignment of new fibres.

If the development of all laminar shell is as inferred by Williams (1968) for Juresamia then a continued reduction in epithelial cell size did not continue. In Willams’ view a single epithelial cell (implied by his text-fig. 25 as being about I2 um wide) secreted several blades, each to some extent separated by impersistent proteinous strands and abutting laterally to form more or less continuous sheets. The alternative is for each blade to have been secreted from single epithelial cells, in

18 SHELL STRUCTURE

the case of Devonian and Carboniferous chonetaceans between 2 and 5 um wide, which progressively ceased to produce the protein sheets which separate normal fibres. But in whatever way laminar shell was deposited it is clear that the epithelium was unusually mobile by modern standards (Brunton 1969), and that the proteinous strands and old cell boundaries were ruptured at times of mantle regression. At such times the regressing epithelial cells probably laid down a proteinous sheet continuous with the periostracum. During transgressive renewed

were PLECTAMBONITACEA CHONETACEA STROPHALOSIACEA|PRODUCTACEA STROPHOMENACEA

[PERMIAN |

J Intermediate Laminar

Standard Cross-bladed

Laminar

Craspedalosia

CHONETIDAE ANOPLIIDAE EMESIS Helaspis ©: :

Devonalosia Spinulicosta

Transitional == Fibrous =

Leptaenisca

Plectodonta

Ludlow

Protochonetes

Wenlock

\

Z < a =) = 7)

Aegiria

Llandovery

,

il

. Strophochonetes

Ashgill

Thaerodonta ‘YEochonetes

Caradoc

Aegiromenas

Llandeilo

ORDOVICIAN

Virvella

Llanvirn Arenig Sowerbyellinae Aegiromeninae

. 4 ’ ‘ 5 , chonetids productids SOWERBYELLIDAE

Fic. 6. Inferred phylogenetic relationships between those genera of the Plectambonitacea and Chonetacea in which shell microstructure has been studied. Those taxa in which the name is horizontal have not been studied in detail. Principal features of the secondary layer shell fabric are differentiated and labelled in italic script. Five productid genera are included to indicate the results of preliminary investigations on their shell structure and relationships. It is suggested that the plectambonitacean to chonetacean changes in shell structure may have continued and given rise to the productids. Leptaenisca, commonly cited as ancestral to the Productacea, would seem not to have a typically strophomenacean shell of cross-bladed laminae.

CHONETACEAN BRACHIOPODS 19

calcite deposition these proteinous layers would have become entombed within the shell fabric and consequently separated one skeletal sheet from another so accentuating the lamination typical of this type of brachiopod shell fabric. Preliminary results from the investigation of mid-Devonian productacean and strophalosiacean shell microstructures shows them to be composed of semi-parallel lath-like units 2-3 wm wide with little development of laminar sheets (PI. 8, fig. 4, Pl. 9, fig. 1), whilst Carboniferous and Permian productaceans have typical cross- bladed fabrics (Pl. 9, figs 2-4). Such fabrics may be explained as a continuation of the evolutionary trend outlined above, but further study is in progress on this

Recent t

Cal

r aad ee xe (x Thecospira & Cadomella ——. Koninckinacea, Spiriferida ) SINCE

<<

Jurassic S

Ja x Triassic : a) | Lp eR er = LYTTONIACEA cooperininid Permian RICHTHOFENIACEA | | | : Merah caeicennee eee a 4

DAVIDSONIACEA

STROPHALOSIACEA

Carboniferous SRODUETAGEA

CHONETACEA | Devonian STROPHOMENACEA bee Se I, ce | Silurian

7 ae een Le | |

BILLINGSELLACEA

PLECTAMBONITACEA

Cambrian

Fic. 7. Speculative phylogeny of certain superfamilies of the Strophomenida, together with their ancestral stock, the Billingsellacea. The strophalosiacean Cooperina-like group may be close to the stock from which the Thecideacea arose. Pseudopunctation was developed within the Davidsoniacea and at the start of the Plectambonitacea. Endopunctation developed in the Thecideacea, possibly early in the Jurassic. (*Thecospiva and Cadomella have been placed in the Davidsoniacea and Chonetacea respectively. It is thought likely that they belong to the spiriferide Koninckinacea).

20 SHELE STRUCLIURE

question and the more traditional derivation of these stocks, via Leptaenisca, from the Strophomenacea may yet prove possible. Study of two imperfectly preserved specimens from the Haragan Shale of Oklahoma shows that the shell fabric of Leptaenisca is not truly laminar. The genus can not, therefore, be excluded from possible productidine ancestral stocks by reason of its shell alone. The shell of Permian strophalosiaceans appear to have retained a less laminar shell than productaceans.

VI. DISCUSSION AND CONCLUSIONS

This study, based upon shell microstructure, supports Havli¢ek’s conclusions, based upon morphology and stratigraphy, that the family Chonetacea was derived from aegiromeninid Plectambonitacea.

Ordovician plectambonitaceans have a shell structure with small ‘fibres’ (possibly equivalent to the ‘laminae’ of Williams 1968) about 6 wm wide in the outer layer, which grade rapidly into a normal parallel-fibrous shell fabric similar to that of Recent brachiopods. This gradational change may simply be a reflection of the increase in size of epithelial cells away from the mantle edges; a possibility which cannot be tested without studying well preserved and undamaged shell margins. Within the mid-Ordovician to Silurian aegiromeninid Plectambonitacea a progressive change occurred which links the shell structure of this subfamily to that of the earliest known chonetaceans in the uppermost Ordovician.

Like some aegiromeninids, the lower Palaeozoic chonetaceans have a shell composed of small lath-like fibres which retain their individuality, in contrast to the sheet structures that began to develop in Upper Palaeozoic specimens.

The internal morphology of aegiromeninids, particularly that of the dorsal valve, is simpler than that of most other plectambonitaceans. Within the subfamily various morphological features were ‘tried’, some of which may be homologous to chonetacean characteristics, and Havlitek (1967) suggested that some genera altered their way of life from benthonic to epiplanktonic, being attached to floating algae. Thus it was a group undergoing much evolutionary change.

The socket ridges of Sowerbyellinae extend antero-laterally and probably assisted in the support of the body wall. In the Aegiromeninae socket ridges are commonly reduced, whilst in the Chonetacea they functioned only as socket bounding ridges and the role of body-wall support was filled by the anderidia. The anderidia probably developed from the outer side septa of the Sowerbyellinae and the low ridges dividing the dorsal adductor muscle scars of, for example, Aegivomena. An anteriorly prominent dorsal median septum is common to Aegiromeninae and Chonetacea and in both taxa it is believed to have been involved in the support of a simple schizolophe, more or less fused to the dorsal mantle. From the Sowerbyellinae, through the Aegiromeninae and into the Chonetacea there is a reduction in the skeletal support for the teeth. Dental plates are reduced and all

CHONETACEAN BRACHIOPODS 21

but lost in Sericoidea, Sentolunia and Chonetoidea and are lacking in the Chonetacea. In the ventral interareas of the last two genera Havlicek (1967) has recorded fine canals penetrating the shell substance, as in contemporaneous Eochonetes, and these structures are essentially the same as the canals leading from the valve interior into the spines of chonetaceans. All that is required is for the plectambonitacean epithelial evaginations, responsible for the canals, to have retained generative buds at their tips so as to have grown posteriorly beyond the posterior margin. Being generative, in the same way as the rest of the mantle margins, implies the sequential secretion of a protective periostracum followed by mineral deposition around the epithelial cells to form a hollow spine. It is rather as if the epithelial cells of an endopunctum retained a generative tip so that growth, restricted to that local area, continued more or less perpendicular to the valve surface. (This is not to say that I believe in a direct relationship between endopunctae and spines.)

The weakly concavo-convex profile, the outline and external ornamentation of Sentolumia and Chonetoidea are in accord with the morphology of the first chonetaceans, Stvophochonetes, and it may be that the strong ventral median rib characteristic of this genus (Pl. 6, figs. 3, 4) is a remnant feature of the Plectam- bonitacea. In contrast to Boucot & Harper (1968) the present study indicates that Protochonetes evolved from Strophochonetes. Shell structure studies on the Anopliidae suggest that their origin was in common with other chonetaceans and that this family evolved in the lower to mid-Silurian by morphological differentiation.

In considering the distribution of ancestral stocks and general evolution of the chonetaceans it should be remembered that the present wide geographical separation between the European Chonetoidea-like stock and North American Strophochonetes would have been less in Upper Ordovician times, if current theories of continental drift and the degree of crustal shortening in the North Atlantic region during the Caledonian orogeny are accepted. In discussing Ordovician faunal provinces Williams (1969) suggested a Caradocian palaeogeography in which oceanic currents would have distributed marine organisms (other factors permitting) in the European and North American provinces. Within the Ashgill of Bohemia, Havlicek (1967) and Havlicek & Vanék (1966) record several aegiromeninid species morphologically close to the chonetacean ancestor, but no chonetaceans. In the Richmond Series of Anticosti Island, Canada, Twenhofel (1914) only recorded Plectambonites sericeus (presumably Sowerbyella) with Chonetes primigenius. Thus, unless more recent faunal work on Anticosti proves the presence of Aegiromeninae in rocks older than those from which the first Strvophochonetes are recorded it seems this was not the area in which the evolutionary change took place. Both Aegiro- meninae and Chonetacea are found in Girvan, Ayrshire, but the Chonetacea postdate those at Anticosti. It seems possible, therefore, that the evolutionary change took place in the Upper Ordovician within the southern region of Williams’ palaeogeo- graphical model.

Using Williams’ (1969) model it is suggested that some Bohemian Aegiromeninae, possibly Chonetoidea itself, became widely distributed along the southeast margin of the Caradocian seas, perhaps helped by having become epiplanktonic through

SHELL STRUCTURE

Nn N

their ability to fix to marine algae (Bergstrom 1968). Within this stock posterior ‘hold fasts’ retained the ability to secrete shell material, so evolving tissue-filled posteriorly directed spines. Like their ancestors the spat would have been attached by their pedicle to hard material on the sea-floor, or perhaps to seaweeds. However, at an early age the pedicle atrophied and the development of the spines would have helped stabilize benthonic specimens residing in areas subject to marine currents, particularly those specimens facing into the current which were consequently more susceptible to being overturned posteriorly when the shell opened. In a low velocity unidirectional flow from front to back a ‘dead water’ zone behind the raised dorsal valve might have prevented the burial or erosion of the spines spread out more or less at the sediment to water interface. These adaptations contributed to the evolution of the Chonetacea in the Lower Palaeozoic fine-grained sedimentary environments in which they are commonly found. During Upper Palaeozoic times chonetaceans spread into regions of coarse shelly detritus as well as living in silt and mud environments.

<< Marine flow

SS

°

° ° ° Cie Cm cy

ex) ° OGl. =

° Po 5 9 0° °

Fic. 8. Hypothetical chonetacean adult community on a soft-bottomed sea floor. The two shells at the top right are dead; one overturned (seen in transverse section), the other part buried. The other three specimens are living (with marginal setae). The two front specimens are cut in longitudinal section; on the left parallel to and on the right along the median line. In these specimens musculature, body wall and lophophore are represented and the arrows indicate the possible flow of water within the brachial cavity.

CHONETACEAN BRACHIOPODS 23

The spread of chonetaceans to Britain and Europe would have been achieved by Williams’ northeasterly oceanic flow, together with the more general break-down in provinciality which started at the close of the Ordovician and became marked during the mid- and upper Silurian.

Representatives of the Chonetacea were the first brachiopods to have developed long tubular spines. Ontogenetic studies of Carboniferous species indicate that these spines normally grew posteriorly at the time of their origin. Thus, in relation to the commissural plane the lateral spines at any particular growth stage were directed posteriorly and were well suited for the support of shells on the substrate. If there was a directional water flow in the environment and if the young shell was able to choose its orientation on settlement it is likely that the water circulatory system outlined above would best have been served by facing into that flow. In this situation posteriorly directed spines are well adapted to the stabilization of the shell (Text-fig. 8).

This demonstration of a gradual change in the shell microstructure from certain Plectambonitacea species to Chonetacea species supports the contention of Williams & Wright (1967) and others that we have here an evolutionary sequence; one which ranges across a subordinal division of the classification in the Treatise. Furthermore, the greater complexity discovered within the skeletal fabrics of these strophomenids allows wider speculation upon phylogenetic relationships and the modification of the relationships suggested by Williams in 1968 and 1970. The phylogenies of the superfamilies presented here (Text-fig. 7) are poorly understood at the points of origin of the Productacea and Thecideacea. Williams (1970) derived the Triplesiacea from the Davidsoniacea which arose from the Billingsellidae. In his view the nisusiid Billingsellacea gave rise to the Orthacea, Clitambonitacea, Gonambonitacea, Strophomenacea and Plectambonitacea.

VII. ACKNOWLEDGEMENTS

This study would not have been possible without the generous donation or loan of specimens from various sources, in particular I want to thank my colleague Dr. L. R. M. Cocks, Dr. G. A. Cooper of the National Museum of Natural History, Washington D.C., Dr. O. A. Dixon of the University of Ottawa, Dr. V. Havlicek of the Geological Institute, Prague, Dr. V. Jaanusson of the Natural History Museum, Stockholm, Dr. W. D. I. Rolfe of the Hunterian Museum, Glasgow and Dr. A. Roomusoks of Tartu, Estonia SSR. I appreciate the stimulating and helpful discussions held with several colleagues especially Dr. Cooper (Washington), Professor A. Williams ot Queens University, Belfast and Dr. Cocks who were good enough to comment upon the draft script, and Mr P. Minton of the Civil Engineering Department, Imperial College, London, and am grateful to the Director of the British Museum (Nat. Hist.) who granted leave of absence allowing me to visit the National Museum of Natural History, Washington. I have received valuable assistance from the staff of the Electron Microscope Unit and Photographic Department of this Museum.

24 SHELL STRUCTURE

VIII. REFERENCES

ARMSTRONG, J. 1969. The cross-bladed fabrics of the shells of Tevvakea solida (Etheridge and Dun) and Stveptorhynchus pelicanensis Fletcher. Palaeontology, London, 12, (2) : 310-320, pls 57-60.

Atkins, D. 1960. The ciliary feeding mechanism of the Megathyridae (Brachiopoda), and the growth stages of the lophophore. J. mar. biol. Ass. U.K., Plymouth, 39 : 459-479.

Baker, P. G. 1970. The growth and shell microstructure of the thecideacean brachiopod Moovrellina granulosa (Moore) from the Middle Jurassic of England. Palaeontology, London, 13 (1) : 76-99, pls 18-21.

BERGSTROM, J. 1968. Some Ordovician and Silurian brachiopod assemblages. Lethaia, Oslo, 10 (3) : 230-237, figs I-9.

Bovucot, A. J. & Harper, C. W. 1968. Silurian and Lower Middle Devonian Chonetacea. J. Paleont., Tulsa, 42 (1) : 143-176, pls 27-30.

Brunton, C. H.C. 1968. Silicified brachiopods from the Viséan of County Fermanagh (II). Bull. By. Mus. nat. Hist. (Geol.), London, 16 (1) : 1-70, pls 1-9.

1969. Electron microscopic studies of growth margins of articulate brachiopods. Z. Zellforsch., Berlin, 100 : 189-200, 13 figs. 1971. An endopunctate rhynchonellid brachiopod from the Viséan of Belgium and

Britain. Palaeontology, London, 14 (1) : 95-106, pls 11, 12.

Brunton, C. H. C. & MacKinnon, D. I. (In press) The systematic position of the Jurassic brachiopod Cadomella. Palaeontology, London, 15.

Cuao, Y. T. 1928. Productidae of China. II. Chonetinae, Productinae and Richthof- eniinae. Palacont. sinica, Peking, B.5 (3) : 1-103, pls 1-6.

Cocks, L.R.M. 1970. Silurian brachiopods of the Superfamily Plectambonitacea. Bull. Br. Mus. nat. Hist. (Geol.), London, 19 (4) : 141-203, pls I-17.

Cooper, G. A. & Grant, R.E. 1969. New Permian brachiopods from West Texas. Smithson. Cont. Paleobiol., Washington, 1 : 1-20, pls 1-5.

Cowen, R. & Rupwick, M. J. S. 1966. A spiral brachidium in the Jurassic Chonetoid brachiopods Cadomella. Geol. Mag., London, 103 (5) : 403-406.

GRaAnT, R. E. (in press).

Haviicex, V. 1967. Brachiopoda of the suborder Strophomenidina in Czechoslovakia. Rozpr. Ustred. Ustav. Geol., Praha, 33 : 1-235, pls 1-52.

Haviicek, V. & VANEK, J. 1966. The biostratigraphy of the Ordovician of Bohemia. Sbornik geol. Ved., paleont., Praha, 8 : 7-69, pls 1-16.

Koztowski, R. 1929. Les Brachiopodes gothlandiens de la Podolie polonaise. Palaeont. Polon., Warsaw, 1 : 1-254, pls 1-12.

Lister, T. R., Cocks, L. R. M. & Rusuton, A. W. A. 1970 (for 1969). The basement beds of the Bobbing borehole, Kent. Geol. Mag., London, 106 (6) : 601-603.

McCammon, H. M. 1969. The food of articulate brachiopods. J. Paleont., Tulsa, 43 (4) : 976-985.

Murr-Woop, H.M. 1962. On the Morphology and Classification of the Brachiopod Suborder Chonetoidea. Bry. Mus. nat. Hist., London, VIII + 132 pp., 16 pls.

Opix, A. A. 1933. Uber Plectamboniten. Acta comment. Univ. tartu geol., Dorpat, 24: I-79, pls. 1-12.

PAECKELMANN, W. 1930. Die Fauna des deutschen Unterkarbons, Die Brachiopoden, 1 Teil. Preuss. geol. Landesanst Abh., 122 : 144-326, pls 9-24.

REED, F. R.C. 1944. Notes on some new Ordovician brachiopods from Girvan. Ann. Mag. nat. Hist., London, 11 : 215-222, pl. 3.

Rupwick, M. J.S. 1961. The feeding mechanism of the Permian brachiopod Provichthofenia Palaeontology, London, 3 (4) : 450-471, pls 72-74.

—— 1968. The feeding mechanisms and affinities of the Triassic brachiopods Thecospiva

Zugmayer, and Bactrynium Emmrich. Palaeontology, London, 11 (3) : 329-360, pls 65-68.

1970. Living and fossil bvachiopods. 199 pp. Hutchinson Univ. Lib., London.

CHONETACEAN BRACHIOPODS 25

SARYCHEVA, T. G. & SoxotsKayaA, A. N. 1959. ‘The Classification of the Pseudopunctate Brachiopods’. Doklady Akad. Nauk SSR, Leningrad, 125, 1 : 181-184 (in Russian). TERMIER, H., TERMIER, G. & Payaup, D. 1967. Découverte d’une Thécidée dans le Permien

du Texas. Comp. Rend. Acad. Sci, Paris, 263 : 332-335. Wiitiams, A. 1956. The calcareous shell of the Brachiopoda and its importance in their classification. Biol. Rev., Cambridge. 31 : 243-287. 1968. Evolution of the shell structure of articulate brachiopods. Spec. Papers Palaeontology, Londen, 2 : 1-55, pls. 1-24. W 1969. Ordovician faunal provinces with reference to brachiopod distribution, in The Pre-Cambrian and Lower Palaeozoic Rocks of Wales. (Ed. Wood A.) : 117-154, Univ. aloes Press, Cardiff. t9Z- Origin of laminar-shelled articulate brachiopods. Lethaia, Oslo, 3 : 329-340, figs I-10. es A. et al. 1965. Tveatise on Invertebvate Paleontology. Ed. Moore, R. C., Pt. H. Brachiopoda. 927 pp., 746 figs, Kansas. WitiiaMs, A. & Wricut, A. D. 1967 in Harland, W. B. e¢ al (Eds.). The Fossil Record Brachiopoda : 397-421, Geol. Soc., London. Wricut, A.D. 1970. A note on the shell structure of the triplesiacean brachiopods. Lethaia, Oslo, 3 : 423-426, 2 figs.

Locality details of figured specimens.

PLECTAMBONITACEA

Leptestia musculosa Bekker, Uhaku (Cic) [Upper Llandeilo] Lower Ordovician

of Uhaku, Estonia . ; : 0 3 : : 0 : c Plate 5 Leptelloides leptelloides (Bekker), Kukruse oe) [Low Caradoc] Upper

Ordovician of Kuttejou, Estonia . é 3 c : Plate 5 Sowerbyella (Viruella) liliifera Opik, enteecee (Cu) {Low Caradoc] Upper

Ordovician of Estonia 6 5 : : : Plate 2 Eoplectodonta transversalis (ivanieabers)s Lone? ees, Marl, Llandovery,

Lower Silurian of Nyhamn, Gotland, Sweden 5 Plates 1-3 Aegivomena aquila (Barrande), Zahorany Formation [Mid- Cede] nee:

Ordovician of central Bohemia . Plates, 3, 4 Aegivia grvayt (Davidson), Upper Wenlock: Sladen, vei Dynes, eneester

shire, England . : Plate 4 Sevicoidea vestricta (Hadding), Reh Acdyvell Gaon (Upper Caradoc) Upper

Ordovician of Craighead, Girvan, Scotland . Plates 4, 5

CHONETACEA

Strvophochonetes primigenius (Twenhofel), Ellis Bay Formation [Ashgill] Upper Ordovician of Mile 5, Jupiter River and Mile 2, 47 Mile Road, Anticosti Island, Canada. (Mile 2 locality is 15-20’ above the base of the Ellis Bay Formation: Mile 5 is close to the top junction of the Ellis Bay Formation with the Becscie.) —im lit. TE Bolton, Geological Survey of

Canada, Ottawa) : : 9 : : ‘ 5 : 5 ° Plates 5, 6 Dawsonelloides canadensis (Billings), Grande Greve Limestone, Siegenian, Lower Devonian of Gaspé, Quebec, Canada . : 5 5 : : Plate 7

Retichonetes vicinus (Castelnau), Arkona Shale, Hamilton Group. Mid

Devonian, 4 ml. upstream from Hungry Hollow Br., 2 ml. E of Arkona,

Ontario, Canada : : 5 : Plates 1, 7 Rugosochonetes silleesi Beaten, high Glener Limestone Low D zone

Viséan, Lower Carboniferous, of Sillees R, nr. Bunnahone Lough, 2 ml. NW

of Derrygonnelly, Co. Fermanagh, N. Ireland : - : : : Plates 7, 8

26 SHELE STRUCTURE

STROPHALOSIACEA

Devonalosia wrightorum Muir-Wood & Cooper, Lower Ferron Point shale, Hamilton Group, Mid-Devonian of abandoned Ps Portland Cement Co. pit, Alpena, Michigan, USA é 5 : ‘

PRODUCTACEA

Helaspis luma Imbrie, Genshaw Formation, Hamilton Group, Mid-Devonian of Long Lake, 74 ml. NNE of Alpena, Michigan, USA . : : c

Eomarginifera lobata (J. de C. Sowerby), Great Limestone, E2 Lower Namurian of Greenleighton, Northumberland, England . :

‘Dictyoclostus’ sp, Carwood, Lower Mississippian of 2 ml. SW of Borden, Indiana, USA

Horridonia horrida (J. Sorcyl lower Teen Pemmere e rsbuiby, Gera, Germany . :

Howarp Brunton, Ph.D. Department of Palaeontology

BritisH Museum (NATURAL History) Lonpon, SW7 5BD

Plate 8

Plate 9 Plate 9 Plate 9

Plate 9

S

PLATE 1 Pseudopunctae

Fic. 1. Fracture through a pseudopunctum, with taleola, close to the anterior margin of the dorsal valve of Rugosochonetes silleesi Brunton, from Viséan shales of Co. Fermanagh, N. Ireland. Eroded internal surface is to the bottom, viewed posteriorly. SEM (Scanning electron microscope), 1150.

Fic. 2. Deeply exfoliated dorsal valve exterior of Retichonetes vicinus (Castelnau) from the middle Devonian Arkona Shale of Ontario, Canada, showing ribbing and a taleola within a pseudopunctum. The exterior of the shell is uppermost and the anterior is to the top. SEM, x 440.

Fic. 3. Deeply exfoliated dorsal valve interior of Aegiromena aquila (Barrande) from the Caradoc of Czechoslovakia, showing a completely ‘fibrous’ pseudopunctum. The valve interior is to the top. SEM, x1tIoo.

Fic. 4. The internal mosaic surrounding a pseudopunctum of Eoplectodonta transversalis (Wahlenberg), from the Llandovery of Gotland, Sweden. The anterior margin of the valve is to the right. SEM, x550.

Bull. By. Mus. nat. Hist. (Geol.) 21, 1 PLATE 1

PEATE 2 Sowerbyellinae, standard shell fabric

Fics 1-3. Sowerbyella (Viruella) liliifera Opik from Low Caradoc of Estonia. Valve interiors to the top. 1 -— Cut and lightly etched transverse section of the ventral valve showing typical secondary fibres. SEM, x1o0oo. 2-— Fractured dorsal valve interior viewed posteriorly , showing the three dimensional aspect of typical secondary fibres. SEM, tooo. 3 — Trans- verse fracture, close to dorsal valve margin, showing the transition from the sediment and primary shell, near the bottom, to secondary fibres at the top. SEM, x 1000.

Fics 4-6. Eoplectodonta transversalis (Wahlenberg) from the Llandovery of Gotland, Sweden. Valve interiors to the top. 4 — Latex impression of a ventral valve interior, i.e. an internal mould, showing the mantle canal impressions extending from deeply impressed lanceolate diductor muscle scars. 3. 5 — Detail of the umbonal region of fig. 4 showing the small medianly placed adductor muscle scars. x6. 6 — Transverse fracture through the external region of the valve showing, from bottom up, a thin layer of micrite, small lath-like primary lamellae and the start of standard secondary fibres. SEM, 1200.

Bull. By. Mus. nat. Hist. (Geol.) 21, 1 PLATE 2

PLATE 3 Eoplectodonta and Aegiromena

Fics 1-4. Eoplectodonta transversalis (Wahlenberg) from the Llandovery of Gotland, Sweden. Interior surfaces to the top. 1 — General view of the dorsal valve interior showing the prong-like elongations, despite broken tips, of the socket ridges. 3. 2 — Exfoliated dorsal valve interior, about 1.5 mm. behind the anterior margin, showing pseudopunctae and several small endopunctum-like flexures of the secondary fibres (arrowed). SEM, 250. 3 — Detail of endopunctum-like flexures, seen in section, from centre of fig. 2. SEM, x 3000. 4 — Detail of the internal surface, slightly exfoliated, showing the same flexures. SEM, 1200.

Fic. 5. Aegiromena aquila (Barrande) from the Caradoc of Czechoslovakia. Deeply exfoliated dorsal valve interior (to the top right) near the antero-lateral margin and close to the external surface showing one of the rarely occurring almost typical secondary fibres. The antero-lateral margin is to the bottom. SEM, 1200.

ATE 3

IPL,

Mus. nat. Hist. (Geol.) 21, 1

Bull. Br.

4 P

>

om

owt yo

PLATE 4 Aegiromeninae

Fics 1, 2. Aegiromena aquila (Barrande) from the Caradoc of Czechoslovakia. _1— Exfoliated dorsal valve, about half way through the shell and towards the antero-median margin, showing a pseudopunctum with taleola. Anterior is to the top. SEM, x1100. 2 — Exfohated ventral valve exterior showing ribbing, distribution of pseudopunctae, mainly within the rib interspaces, and parallel fibrous nature of the shell. Anterior is to the top. SEM, x120.

Fic. 3. Deeply exfoliated ventral valve interior, close to the external surface, of Aegiria grayi (Davidson) from the mid-Silurian of Dudley, England, showing almost standard parallel fibres of the secondary layer. Interior is to the top left. SEM, x 1800.

Fic. 4. Exfoliated ventral valve exterior, near the antero-lateral margin, of Sericoidea restricta (Hadding) from the Caradoc of Girvan, Scotland, showing intermediate, atypical ‘fibres’. Anterior is to the top. SEM, x 1000.

4

IPL NATE

nat. Hist. (Geol.) 21, 1

Mus.

Bull. Br.

PLATE 5

Fics 1, 2. Sericoidea restricta (Hadding) from the Caradoc of Girvan, Scotland. Ex- teriors to the top. 1 — Exfoliated ventral valve exterior showing subparallel ‘fibres’. Antero- lateral margin to the left. SEM, «1250. 2 —Slghtly exfoliated ventral valve exterior close to its antero-median margin showing a single layer of small primary shell lamellae with sub- parallel ‘fibres’ below, including one near-normal transverse section (arrowed). Anterior is to the bottom. SEM, x 1000.

Fic. 3. Broken section near the anterior margin of the dorsal valve of Leptelloidea leptel- loides (Bekker), from low Caradoc of Estonia, showing the typical secondary layer fibres of the non-aegiromeninid Plectambonitacea. Exterior to the top. SEM, 1250.

Fic. 4. Deeply exfoliated ventral valve exterior, near the posterior margin, of Leptestia musculosa Bekker from the Upper Llandeilo of Estonia, showing typical secondary shell of the early Plectambonitacea. Interior uppermost. SEM, x 1ooo.

Fics 5-8. Strophochonetes primigenius (Twenhofel) from the Ellis Bay Formation, Ashgill, of Anticosti Island, Canada. 5 — Interior of a young dorsal valve. 4. 6— Posterior view of a dorsal valve showing the quadrifid myophore of the cardinal process and low flanking chilidial plates (arrowed on one side). 9g. 7 — Ventral valve exterior and part of a dorsal valve interior. x3. 8 — Detail of the dorsal valve cardinalia, note the strongly bilobed cardinal process. 4.

itp 2

2

nat. Hist. (Geol.)

Bull. Br. Mus.

PEATE 6

Strophochonetes primigenius (Twenhofel) from the Ellis Bay Formation of Anticosti Island, Canada.

Fies 1, 2. Deeply exfoliated ventral valve exterior, close to the antero-lateral margin (to the top) showing pseudopunctae and the disposition of secondary transitional ‘fibres’. SEMs, «550 and x 1000.

Fic. 3. Well preserved ventral valve exterior showing the bases of spines and the accentuated

median rib. x3.

Fic. 4. Part of one of the fossiliferous slabs of limestone showing many ventral valve ex- teriors and one dorsal valve interior (figured PI. 5, fig. 5). The accentuated median rib, typical of Strophochonetes and spine bases show on most specimens. » 2.

Bull. By. Mus. nat. Hist. (Geol.) 21, 1 IPILIN W138, ©

PLATE 7 Devonian and Lower Carboniferous Chonetacea

Fics 1, 2. Dawsonelloides canadensis (Billings) from the Lower Devonian of Gaspé, Ouebec, Canada. Valve exteriors uppermost. 1 — Deeply exfoliated exterior at a pseudo- punctum with taleola. SEM, x960. 2 — Small fibre-like units close to the external surface showing some lateral fusion. SEM, x 4000.

Fic. 3. Retichonetes vicinus (Castelnau) from the Mid-Devonian of Ontario, Canada. Slightly exfoliated ventral valve interior, at a pseudopunctum with taleola, showing the fusion of lath-like elements to give impersistent sheets. Antero-lateral margin to the bottom left. SEM, x 1100.

Fic. 4. Rugosochonetes silleesi Brunton from the Viséan of Co. Fermanagh, N. Ireland. Slightly exfoliated internal surface of a dorsal valve postero-medianly. Anterior is to the right. The individual shell lamellae are thin and do not retain a parallel orientation from layer to layer. SEM, x 2100.

Bull. By. Mus. nat. Hist. (Geol.) 21, 1 PLATE 7

PAGES

Fics 1-3. Rugosochonetes silleesi Brunton from the Viséan of Co. Fermanagh, N. Ireland. Dorsal valve exterior uppermost. 1 — Somewhat eroded external surface of the valve showing thin laminae forming wide sheets. SEM, 1100. 2 — Detail from the centre of fig. 1 showing surfaces within the laminae like those onto which shell growth may have occurred in the living animal. SEM, 5500. 3 — Fracture surface a little below the valve exterior (to the top) and close to the valve margin (to the bottom right) showing sheets of thin blade-like laminae. The sheet surface is broken by persistent and impersistent grooves which in life accommodated organic material separating individual blades. The ridges, at 20° to 30° from the grooves, mark the interblade boundaries of the adjacent removed sheet. SEM, x 2600.

Fic. 4. The strophalosiacean Devonalosia wrightorum Muir-Wood & Cooper, from the Mid Devonian of Michigan, U.S.A. Fracture surface near the anterior margin of the dorsal valve. The exterior is just off the top of the micrograph. Well-differentiated units resembling crested laminae.SEM, x 2200.

PLATE 8

Bull. By Mus. nat. Hist. (Geol.) 21, 1

PLATE 9 Productacea

Fic. 1. Helaspis luma Imbrie, from the Mid-Devonian of Michigan, U.S.A. showing external surfaces of crested laminae with subparallel orientation. SEM, x 2800.

Fic. 2. Eomarginifera lobata (J. deC Sowerby), from the Lower Namurianof Northumber- land, England, showing cross-bladed structure near the centre of the base of the trail. Exterior to the top, anterior to the left. SEM, 2000.

Fic. 3. ‘Dictyoclostus’ sp. from the Low Mississippian of Indiana, U.S.A. showing cross- bladed fabric on an exfoliated interior from the ventral valve trail. SEM, x 4000.

Fic. 4. Horridonia horrida (J. Sowerby), from the Permian of Germany, showing typical cross-bladed fabric close to the external surface of the ventral valve near the postero-lateral margin. Exterior to the top right. SEM, x 4000.

Bull. By. Mus. nat. Hist. (Geol.) 21, 1 PLATE 9

“A LIST OF SUPPLEMENTS ==

TO THE GEOLOGICAL SERIES

OF THE BULLETIN OF oe.

THE BRITISH MUSEUM (NATURAL HISTORY). :

Pp. 213: 30 Plates; 2 Text-figures. 1965. £6, . Et-Naccar, Z. R. Stratigraphy and Planktonic Rotaniinieee 0! Cretaceous—Lower Tertiary Succession in the Esna-Idfu Region, Egypt, U.A.R. Pp. 291; 23 Plates; 18 Text-figures. 1966. {10. . Davey, R. J., Downie, C., SARGEANT, W. A. S. & WiLztams, G. L. Mesozoic and Cainozoic ‘Dinoflagellate Cysts. Pp. 248; 28 Plat figures. 1966. {7. : . APPENDIX. Davey, R. J., Downie, C: SARGEANT, W. As S. & W _ Appendix to Studies on Mesozoic and Cainozoic Dinoflagellate C 1969. 8op. ; ;-Evviortr, G. F. Permian to Palaeocene Calcareous Algae @asydad uC Middle East. Pp. 111; 24 Plates; 17 Text-figures. 1968. £5.12}. . Ruopes, F. H. T., Austin, R. L. & Druce, E. C. British Avo: ferous) Conodont Ems; and their value in local and continental 01 Pp. 315; 31 Plates; 92 Text-figures. 1969. 11. — é ; . CuiLps, A. Upper Jurassic Rhynchonellid’ Brachiopods from Europe. Pp. 119; 12 Plates; 40 Text-figures. 1969. £4.75. . Goopy, P. C. The relationships of certain Upper Cretaceous special reference to the Myctophorids. Pp. 255; 102 Text-figures. . OwEN, H. G. Middle Albian Stratigraphy in the ‘Paris ‘Ba 3 Plates; 52 Text-figures. 1971. {6.. . Sippigur, Q. A. Early Tertiary Ostracoda of the: family ‘Tr from West Pakistan. Pp. 98; 42 Plates; 7 Text-figures. 1971.

| a oo 3 oO 3 shes zi ¥ a Q a oe e a es

__ TRITYLODONTIDS CROMPTON

z

5

SEUM

-

ULLETIN OF (N

* 1

SH M

NINE OCCLUSION I

vA

‘

POSTCANINE OCCLUSION IN CYNODONTS

AND TRITYLODONTIDS pe $0 JAN 1973

BY

ALFRED WALTER CROMPTON

Museum of Comparative Zoology, Harvard University, Cambridge, Mass.

Pp. 27-71; 7 Plates, 14 Text-figures

BOLLE TIN’ OF THE BRITISH MUSEUM (NATURAL HISTORY) GEOLOGY Vol. 21 No. 2 LONDON : 1972

THE BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY), instituted im 1949, 1s issued in five series, corresponding to the Departments of the Museum, and an Historical serves.

Parts will appear at irregular intervals as they become veady. Volumes will contain about three or four hundred pages, and will not necessarily be completed within one calendar year.

In 1965 a separate supplementary series of longer papers was instituted, numbered serially for each Department.

This paper ts Vol. 21, No. 2 of the Geological series. The abbreviated titles of periodicals cited follow those of the World List of Scientific Periodicals.

World List abbreviation Bull. Br. Mus. nat. Hist. (Geol.).

© Trustees of the British Museum (Natural History), 1972

TRUSTEES OF THE BRITISH MUSEUM (NATURAL HISTORY)

Issued 29th December, 1972 Price £3.40

POSTCANINE OCCLUSION IN CYNODONTS AND TRITYLODONTIDS

By A. W. CROMPTON

CONTENTS

I. INTRODUCTION : : ; : : 30

II. PosSTCANINE DENTITION OF THE , PROCYNOSUCHIDAE AND GALESAURIDAE : c : . : 32

III. PostcANINE DENTITION OF THE DIADEMODONTIDAE AND TRIRACHODONTIDAE : : : : : 35 IV. PostcANINE DENTITION OF THE “TRAVERSODONTIDAE é : : 40 1. Description of three new species of Scalenodon 3 41 2. Occlusion and jaw movements in Scalenodon Pisa POS 44 3. Occlusion and jaw movements in Scalenodon hirschson . 49 4. Postcanine dentition of Scalenodon attridger . b : 53 5. Postcanine dentition of Scalenodon charigi . : : 54

6. Occlusion and jaw movements in a species of Massetognathus . é : 56 7. Postcanines of Exaevetodon ed Capp nod onioen sie é 58 V. DiIsTRIBUTION OF THE TRAVERSODONTIDAE A 60 VI. Occtustion In Tritylodon AND THE ORIGIN OF THE Stee eaon gaan 61 VII. SUMMARY AND DISCUSSION 5 - < : : é : 64. VIII. AcKNOWLEDGEMENTS Z : : : : 5 : 5 68 IX. REFERENCES . ‘ ; : : ‘ : : : : 69 SYNOPSIS

Within the gomphodont cynodonts and their descendants, the tritylodontids, it is possible to trace the initiation and progressive improvement of postcanine occlusion and complex masti- catory movements, beginning with primitive cynodonts in which mastication did no¢ involve actual contact between matching upper and lower postcanine teeth. Occlusion in advanced cynodonts was functionally similar to that of primitive mammals with tribosphenic molars, and the mechanisms by which occlusion evolved in the two groups also appear to have been similar. In primitive gomphodont cynodonts and primitive mammals the crowns of occluding teeth had to be moulded by wear to produce accurately matching shearing surfaces; major features of the crown were thereby obliterated. In advanced members of both groups the topography of the crowns was modified so that only a little wear was needed to produce matching shearing planes.

A clear correlation appears to have existed between the occlusal relationships of the teeth of cynodonts and their replacement patterns. The enamel of cynodonts and tritylodontids was thin and apparently worn through rapidly, so that the structure of the crowns was soon destroyed; in order to compensate for this, worn gomphodont teeth were lost from the front of the row and new ones added behind.

The tritylodontids were probably derived from traversodont cynodonts. The longitudinally orientated shearing planes on the postcanine teeth became more numerous and the relative extent of the backward movement of the lower jaw during the final stages of mastication was progressively increased.

Three new species of traversodont cynodonts are named (Scalenodon hirschsoni, S. attridget, S. charigi).

30 POSTCANINE OCCLUSION

INTRODUCTION

ACCURATE occlusion between cheek teeth with complex crown patterns is a mammal- ian character. It involves complicated relationships between the cusps, ridges and basins of occluding teeth and also mandibular movements that are seldom directly orthal during the final stages of the masticatory cycle (power stroke, Crompton & Hiiemae, 1969a & 6), but are also partially transverse and forward. The relative amount of upward, forward and sideways movement during this phase of occlusion differs widely in the various mammalian orders. Many of the advanced cynodonts and tritylodontids independently developed occlusal patterns which in terms of function closely parallel those of later mammals. The purpose of this paper is to describe and discuss the development of postcanine occlusion in several groups of cynodonts, which are the most mammalian of the therapsid reptiles and the group from which mammals almost certainly arose. Although the cynodonts which had dental occlusion and which are discussed in this paper were not ancestral to mammals this study does throw some light on the mechanism involved in developing dental occlusion of the mammalian type.

Numerous authors have described and discussed the morphology of the teeth of therapsid reptiles; but few have described occlusal relationships, and except for one or two cases (Watson 1911, Parrington 1946) no attempt has been made to determine jaw movements during mastication or dynamic occlusion of the cheek teeth of this group. The functional aspects of mammalian occlusion also have been neglected, but recent papers on wear facets on the molars of living and extinct mammals (Butler, 1961; Mills, 1964, 1966, 1967; Kermack, Lees & Mussett, 1965; Crompton & Jenkins, 1967, 1968) and cineradiographic studies of mastication in a primitive mammal (Crompton & Hiiemae, 1969, a, b & c) have provided a model with which to compare the dynamic occlusal relationships and possible jaw movements in cynodonts.

The infraorder Cynodontia (Fig. 1) arose in the late Permian, reached its greatest diversity in the Middle Trias and became extinct in the early part of the Late Trias. As will be shown below, the Tritylodontidae which survived until the Middle Jurassic can be considered as late survivors of the cynodonts. The Ictidosauria (Diarthro- gnathus, Pachygenelus and Trithelodon) were probably also late survivors of the Cynodontia and a case can perhaps be made for including them within the cynodonts. The cynodonts are usually divided into the following families: the Procynosuchidae (I am including genera which have been placed in separate families by some authors, e.g. Silphedestidae, Dviniidae, Cynosauridae) ; the Galesauridae; the Cynognathidae; the Chiniquodontidae (this family probably includes most of the South American carnivorous cynodonts which have not yet been adequately described but which are at present being studied by Prof. A. S. Romer) ; the Trirachodontidae; the Diademo- dontidae; and the Traversodontidae. The interrelationships and time-spans of these families are shown in Figure 1. The last three families are commonly referred to as the gomphodont cynodonts and it is only in them that complex occlusion between upper and lower postcanine teeth occurred. Postcanine occlusion is present in the Ictidosauria, but it is not complex. The Traversodontidae were the most varied and abundant of the cynodonts and their remains have been discovered in the

ee EEE

31

CYNODONTS AND TRITYLODONTIDS

“UMOUY SIV SUIIOF 0} eIpoUT -19}UT OU 3nq syuOpouAD ATIe9 0} P9zeToI 9IOM oso} ATJUeIedde ‘aeptyUOposjsoz0g 9y} pure seplLey}OouYyon y oy} ‘(YOe]q) sorprurey [PUIUIe 4SOTLIV9 OM OY} Oe UMOYS OSTY “(peyoreU Ayyeorz19A) sorprurey yuopouds isofeur oy} Fo sdrysuorejeiiezur pue sueds-ouny “1 ‘DI

mM >~ N

VIENVSOCGILD! ~

AVGILNOGOYLSOZO3 RF. ee ee

ITT 3VGIYSHLOANHSN yp” onol

TUTE NIBO HD} — — — — — — | HI

TTT LNO

AVAILNOGOTALIYNI

MET fe ee Ser Lees rol. i. oer = oIssvuner AS Seaeul NVINYSd

32 POSTCANINE OCCLUSION

Middle Trias of East Africa, Late Trias of South Africa and in great abundance in the Trias of South America (see Crompton, 1955; Crompton & Ellenberger, 1957; Bonaparte, 1967a&b; Romer, 1967; and Sill, 1969 for pertinent literature). Numerous representatives of the remaining cynodont families have been found in the sedi- mentary rocks of the Karoo system of South Africa (Haughton & Brink, 1954 and Lehman, 1961 for pertinent literature), although they are also known from South America (Bonaparte, 1967a & b), China (Young, 1959 & 1961) and Russia (Tatarinov, 1968). Dental occlusion (i.e. toorh-to-tooth contact) is absent in the Procyno- suchidae and Galesauridae but, as members of these families are ancestral to the later gomphodont cynodonts, their postcanine dentitions will be briefly considered.

II. POSTCANINE DENTITION OF THE PROCYNOSUCHIDAE AND GALESAURIDAE

The postcanine dentition of the Procynosuchidae is best known in the South African species Leavachia duvenhagei. Ina recent review of procynosuchid literature Anderson (1968) suggested that Procynosuchus, Galecranium and Galeophrys were probably synonymous with Leavachia; the postcanine row of Leavachia duvenhaget consisted of 7 to 11 teeth according to the age of the individual. Alternate tooth replacement was observed in all specimens studied.

Mendrez (1967) figured the postcanines of an unidentified species of Leavachia. The crowns of postcanines from the middle of the row were circular in cross-section while those further back were longitudinally ovate (Fig. 2A). A large external cusp was present. On the internal edge of the crowns of both upper and lower teeth there was a series of small cusps, the most anterior and posterior of which were visible in external view; in this view therefore, the crown appeared to be tricuspid. In lateral view it could be seen that upper and lower postcanines alternated with one another; the centre of the lower tooth lay medial to the space between two upper teeth. The lower postcanines bit medial to the uppers but because there was no contact between upper and lower teeth matching shearing surfaces did not develop. The postcanines of these early cynodonts were therefore capable of gripping, puncturing and possibly crushing food but could not shear food efficiently.

The postcanine dentition of the aberrant procynosuchid Dvinia prima from the Russian Permian was described in detail by Tatarinov (1968), who placed the genus in a new family Dviniidae. The postcanine teeth were considerably more complex than those of Leavachia. The crowns of the upper postcanines (“‘molars’’) (Fig. 2B). were transversely ovate with a central cusp completely surrounded by several small cusps on the edge of the crown. Unlike those of Leavachia these cusps were present on the external edge of the crown as well as on the internal edge and the external cusps were slightly larger than the internals. The lower molars of Duinia prima were similar to the uppers except that additional cusps were present in the centre of the crown medial to the main cusp. The postcanines of Dvinia prima could have been derived from those of Leavachia by the addition of a series of small external cusps. In Dvinia upper and lower canines did not occlude, i.e. there was no tooth- to-tooth contact, and the lower postcanines bit medial to the uppers. In these respects the teeth of Dvinia were superficially similar to those of Leavachia. Because

CYNODONTS AND TRITYLODONTIDS

Fic. 2. Galesaurids and procynosuchids. Postcanine teeth. (In these and all other text-figures lower teeth are stippled, upper teeth are plain). A. Leavachia. Crown view of upper from middle and back of row. After Mendrez, 1967. B. Dvinia prima. Crown views of upper and lower. After Tatarinov, 1968. C. Thrinaxodon. Crown views of upper and lower. D. Thrinaxodon. Crown views of last four uppers and last four lowers. E. Thrinaxodon. Internal views of last four uppers and last four lowers. F. Glochinodontoides gracilis. External view of first three uppers. G. Glochinodontoides gracilis. Crown view of first three uppers. H. Cynidiognathus. External view of posterior lower postcanines. After Fourie, 1964.

33

34 POSTCANINE OCCLUSION

the teeth were transversely ovate they appear to be similar to those of gomphodont cynodonts, but in the latter group crown-to-crown occlusion was present.

Several small Permian cynodont skulls have been described (e.g. Protocynodon, Nanictosuchus, Nanictosaurus, Silphedestes ; see Haughton & Brink, 1954 for pertinent literature). These appear to have had postcanine teeth similar to those of the galesaurid Thrinaxodon liorhinus but until they have been adequately prepared and studied their taxonomic position will remain in doubt. It is possible that the fore- runners of the Galesauridae will be found among these little-known cynodonts.

The best-known member of the Galesauridae is Thrinaxodon liorhinus. Several undistorted skulls of this species have been prepared with acetic acid and the crown patterns and relationships of upper and lower postcanines studied in detail. The postcanine teeth did not occlude, replacement was alternate, their crowns were longitudinally ovate and both upper and lower crowns (especially the lowers) had a well developed internal cingulum which supported several cusps. In Fig. 2C, D& E and Plate 1 the postcanines are shown in crown and internal view. It will be seen that the upper and lower teeth tended to alternate with one another in position and that when the jaws were closed a considerable space separated the outer surface of the lower teeth and the inner surtace of the upper teeth. All the specimens studied showed this. Asin procynosuchids, the postcanines ot Thrinaxodon did not develop matching wear facets on the internal surface of the uppers or external surface of the lowers, thus indicating that shearing did not take place; i.e., upper and lower teeth did not occlude with one another. Lateral movement of the lower jaws sufficient to enable upper and lower teeth to come into contact appears to have been prevented in both Galesauridae and Procynosuchidae by the strong transverse processes of the pterygoids and the large pits in the palate which accommodated the lower canines. Tribolodon from the younger Cynognathus zone had postcanine teeth similar to those of Thrinaxodon but the internal cingulum of the lower postcanines was reduced and, in the material available for study, I never observed more than two subsidiary cusps on the internal surface of either upper or lower postcanines.

The postcanine teeth of the remaining galesaurid genera, i.e. Galesaurus, Glochi- nodon, Glochinodontoides, Platycraniel’us and Cynosuchoides, were distinct from those of Thrinaxodon. Upper and lower teeth (Fig. 2F, G; Plate 2) were apparently identical and were characterized by a strongly recurved main cusp with a sharp — blade-like cusp behind and by a total absence of subsidiary cusps on the anterior or internal surfaces. The tip of the main cusp was worn but, as in Thyinaxodon, there were no matching shearing surfaces on upper and lower teeth. In a few specimens available for study teeth were being replaced, but there appears to have been a “break-down” of the alternate tooth-replacement pattern that characterized Thrinaxodon and the procynosuchids. The way in which these peculiar teeth functioned is not understood.

Galesauridae with teeth of this type may have been ancestral to the Cynognathidae or Chiniquodontidae. In Cynognathus and Cynidiognathus (Fourie, 1964) the post- canines usually lacked internal cingula; the main cusp was slightly recurved (Fig. 2H); and, depending upon the position of a postcanine in the tooth row, one, two or three anterior and posterior accessory cusps may have been present. In Cyno-

CYNODONTS AND TRITYLODONTIDS 35

gnathus there was no obvious alternate replacement of the postcanines but the presence of matching wear facets on the external surface of the lowers and internal surface of the uppers shows that the teeth were used for shearing.

Ill. POSTCANINE DENTITION OF THE DIADEMODONTIDAE AND TRIRACHODONTIDAE

The earliest known cynodont to develop postcanines with a complex occlusal pattern was the gomphodont cynodont Diademodon from the early Trias of South Africa. The dentition has been described by Seeley (1895), Watson (1911), Broili & Schréder (1935) Brink (1955, 1957), Crompton (1955, 1963), Fourie (1963, 1964), Hopson (1971) and Ziegler (1969). The postcanine row was differentiated into two to six conical teeth in front; three to nine transversely widened gomphodont teeth in the middle; and two to five posterior teeth which range in crown structure from semi- gomphodont to fully sectorial (Fig. 3B). The number of teeth present depended upon the age of the individual. The upper gomphodont teeth were transversely ovate in crown view and wider than the corresponding lower teeth (Fig. 3A—C and Plate 2B) so that, when the teeth occluded, the inner and outer margins of the upper teeth overlapped the corresponding margins of the lowers (Fig. 14C). The alternate tooth replacement characteristic of Thrinaxodon did not occur in Diademodon; the details of replacement in the latter genus are not yet fully understood, but it was complex, not alternate, and apparently confined to the anterior and posterior regions of the postcanine row. At the front of the row sharp conical teeth replaced worn gomphodont teeth; at the back gomphodont teeth replaced either semi-gomphodont or sectorial teeth. Consequently the occlusal relationships between upper and lower gomphodont teeth were never disturbed by replacement in the middle of the row as would have been the case if replacement had been of the alternate pattern found in the earlier Galesauridae.

Ziegler (1969) concluded that there was probably no replacement at the back of the postcanine row of Diademodon, the pattern being essentially mammalian in that only the incisors, canines and anterior postcanines were replaced. His paper was based entirely upon published accounts of the dentition. Ziegler correctly pointed out inconsistencies both in my own and in other authors’ attempts to interpret the tooth replacement pattern that was present in Diademodon, but his conclusions too are open to criticism and do not appear to be substantiated by the known material.

Dr. J. Hopson of the University of Chicago is at present preparing a manuscript on tooth replacement in Diademodon and Dr. J. Osborn of the Dept. of Oral Anatomy, Guys Hospital, London, is completing a manuscript on the order of tooth eruption and replacement in Diademodon and some early cynodonts. Because of this I do not wish to enter here into a full discussion of tooth replacement in Diademodon but would like to point out some problems raised by Dr. Ziegler’s interpretation.

In the smallest known specimen of Diademodon (Kitching, private communication) the postcanine row contains two or three gomphodont teeth and 1s terminated by two or three teeth which appear to have been either semi-gomphodont or sectorial in structure. In slightly larger specimens (Brink, 1963b) the postcanine row contains five gomphodont teeth with three semi-gomphodont or sectorial teeth behind. In

36 POSTCANINE OCCLUSION

still larger specimens (Crompton 1955, 1963b) up to seven gomphodont teeth are present with four or five semi-gomphodont or sectorial teeth behind. The Diade- modon skulls of different sizes might be considered as different species but this is an extremely unlikely interpretation as many were found at one locality (Brink, 1963b) ; it is more reasonable to suppose that they represent growth stages of a single species (Kitching, 1968). Ziegler appears to have overlooked the size range of the available material of Diademodon and the fact that the postcanine row always ends in a series of teeth which become progressively more sectorial. It would be impossible to obtain the growth stages listed above without replacement at the back of the row. In the large series of jaws studied by Fourie (1964) there is clear evidence of the loss of some of the anterior conical postcanine teeth, and a specimen of Diademodon in East Berlin (Crompton, 1963b) clearly documents the replacement of the most anterior gomphodont tooth by aconical tooth. In 1955 I interpreted a longitudinally ovate opening below an unerupted posterior postcanine tooth of Diademodon as an alveolus, but it seems that Ziegler does not accept this as evidence of replacement of a sectorial tooth by gomphodont tooth at the back of the postcanine row.

Diademodon is closely related to other gomphodont cynodonts such as Tvivachodon, Cricodon, Scalenodon and a new and as yet undescribed and unnamed reptile from the Ntawere Formation of the Luangwa Valley (Zambia). In a Tvivachodon skull prepared by Dr. F. R. Parrington and figured by myself (1963a) there is clear evi- dence of the replacement of the sectorial teeth at the end of the postcanine row by gomphodont teeth. Ina late survivor of the Trirachodontidae, Cricodon (Crompton, 1955), a longitudinally ovate foramen containing the root of a sectorial tooth lies below an unerupted gomphodont tooth; this foramen is similar in shape and position to that found in Diademodon but was clearly an alveolus and not a gubernacular canal. In some species of Scalenodon the postcanine row is terminated by small gomphodont teeth; unerupted larger gomphodont teeth lay above the small teeth and would presumably have replaced them later. This evidence, derived from growth stages of Diademodon and from related forms that have been figured or studied in South Africa, suggests that the posterior postcanine teeth were replaced during growth. Dziademodon may eventually have reached a mature stage when replace- ment ceased in the postcanine series and this may explain why in several of the larger individuals there is no clear evidence of replacement at the back of the row. Ziegler is perhaps unaware that most of the available Diademodon material is poorly pre- served and has undergone little or no development; careful preparation of the existing material will probably show teeth that were in the process of being replaced at the time of death.

The important points concerning Diademodon are that gomphodont teeth were lost in front (or were replaced by conical teeth) ; that they were added behind (by the replacement of existing semi-gomphodont or sectorial teeth); and that there is no evidence of replacement of gomphodont teeth by gomphodont teeth. Consequently, a series of gomphodont teeth always shows a progressive increase in wear towards the front; this is hecause of the occlusal relationships of the postcanine teeth. The replacement patterns of teeth in gomphodont cynodonts ensured that in Diademodon a fresh supply of cusped gomphodont teeth were added during growth and that in

CYNODONTS AND TRITYLODONTIDS 37

the later forms a fresh supply of teeth with deep shearing planes were added. In gomphodonts the enamel covering the crowns of the teeth appears to have been thin and perhaps soft, so that the structure of the crown was soon worn away and the teeth were perhaps replaced more rapidly than in mammals.

The exact order of tooth replacement occurring in the ontogeny of Diademodon is not fully understood; it is hoped that further work on material already available will clarify this process and meet the objections raised by Dr. Ziegler.

Crown views of unworn upper and lower gomphodont teeth of Diademodon are given in Fig. 3A and Plate 2. The upper crowns were transversely ovate whereas the lower crowns tended to be more circular. The upper crown was dominated by an external main cusp from which a well-defined ridge ran directly forwards, bearing two or three subsidiary cups which became smaller towards the front ; another ridge, ill-defined and slightly crenulate, ran backwards from the tip of the external main

EXTERNAL

ae

lcm

Fic. 3. Diademodontids and trirachodontids. Postcanine teeth. A. Diademodon. Crown view of typical upper and lower gomphodont teeth. B. Diademodon. Crown views of last five uppers and last four lowers to show shearing surfaces produced by wear. C. Asin B, but superimposed as when occluded. D. As in C, occluded, but in internal view. E. Cricodon. Crown views of typical upper and lower. For key to abbreviations see p. 69.

38 POSTCANINE OCCLUSION

cusp. The internal edge of the crown was dominated by two cusps, the anterior and posterior internal cusps, which, although smaller than the external cusp, were almost as high. The anterior edge of the crown bore a series of four or five small cusps which tended to become bigger towards the mid-line. The posterior edge supported one or two well-defined cusps of unequal size and a series ot small cuspules or crenulations. A series of crenulated ridges radiated from the centre of the crown towards both the anterior and the internal cusps, with shallow valleys or basins lying between. A series of three or four smaller ridges radiated from the base of the external main cusp towards the centre of the crown to terminate in an ill-defined high area which supported a series of small cusps and which, together with the ridges radiating from the external and internal cusps, formed an ill-defined transverse ridge. A shallow occlusal basin was formed between the transverse ridges of two consecutive teeth, the ridge running backwards from the main cusp of the anterior tooth and the ridge running forwards from the main cusp of the posterior tooth.

The crowns of the lower gomphodont teeth of Diademodon were similar to those of the upper. Each had an external main cusp from which a ridge ran backwards, supporting one or two cusps which became smaller towards the back. A weak ridge ran forwards from the main cusp to terminate in a prominent cusp on the antero- external edge of the crown. A prominent cusp was present on the middle of the internal edge of the crown connected with the base of the main cusp by a ridge running transversely across the crown. Ridges supporting smaller cuspules radiated medially from the main cusp. A series of cusps was present on the anterior edge of the crown, from each of which a ridge ran backwards; they became bigger towards the mid-line. The central region of the crown was relatively high and flat.

The unworn crowns appear to have been capable of crushing and puncturing food and it seems that the details of the crown pattern were quickly worn away in most Diademodon specimens. The enamel of Diademodon postcanines was extremely thin. The external main cusp of the lowers pounded into the basin formed between two adjoining upper postcanines (Figs 3C & 14) and, as wear proceeded, insignificant vertical shearing planes (Sh.s.) were developed between the external surface of the lower cusp and the internal surfaces of the ridges running forwards and backwards from the external main cusps of two adjoining upper teeth (Figs 3B & 14). As the external main cusps were worn down, these planes were lost and the anterior gom- phodont teeth were reduced to featureless nubbins of dentine. In several primitive mammals, e.g. Eozostrodon and Tinodon (Crompton & Jenkins, 1967, 1968), occlusal planes were established by destroying major features of the crown.

The gomphodont postcanine teeth of Diademodon were followed by a series of three or four teeth which became progressively more sectorial towards the back, that is, the inner extension of the crown became progressively smaller. During occlusion the main cusp of the first semi-gomphodont lower tooth met the crown of the matching semi-gomphodont upper tooth internal to the main cusp of the latter and slightly behind the gap between the last gomphodont tooth and first semi- gomphodont tooth. The occlusal details of the posterior postcanines are illustrated in crown view in Fig. 3C and in internal view in Fig. 3D. As the lower jaw closed, shearing was possible between the outer surface of the main cusp of the lower semi-

CYNODONTS AND TRITYLODONTIDS 39

gomphodont tooth and the inner surface of the main cusp of the upper. However, as the tip of the main cusp of the lower tooth abutted against the occlusal surface of the internal extension of the corresponding upper cusp, it was rapidly worn down and its shearing function reduced. In the more advanced traversodonts the lower jaw was pulled backwards as the teeth came into occlusion; this postero-dorsally directed power stroke may have been initiated in Diademodon.

The postcanine dentition of Diademodon showed a significant advance beyond the galesaurid condition.! The lower teeth lay directly below the crowns of the upper rather than internal to them. The cusps of upper and lower postcanines could therefore be used more effectively for puncturing and crushing food. In addition, matching vertical shearing surfaces were present on both the gomphodont series and the sectorial series of postcanine teeth; these were rapidly obliterated by wear but this was compensated for by the type of tooth replacement present in Diademodon which ensured that new gomphodont and sectorial teeth either replaced existing teeth at the back of the row or were added thereto. The gomphodont postcanines of Diademodon may be derived from those of early galesaurids or procynosuchids by widening the teeth in a lingual direction. The Lower Triassic cynodonts Pascual- gnathus and Andescynodon recently discovered in South America by Bonaparte (1967b) may throw considerable light on the evolution ot the Diadeimodon type of postcanine tooth from those of earlier cynodonts.

The postcanine teeth of Tvivachodon of the South African Lower Trias and of Cricodon of the East African Middle Trias were slightly different from those of Diademodon. The postcanine row consisted of six or seven transversely ovate gomphodont teeth with two or three sectorial teeth behind. There was a sharp break between these two series; the transitional zone present in Diademodon, where the teeth become progressively more sectorial towards the back, was absent. The upper and lower gomphodont teeth in Tvivachodon (Fig. 14) and Cricodon (Fig. 3E and Plate 3) had three main cusps, an external, a central and an internal arranged to form a prominent transverse ridge across the crown. The anterior and posterior margins of the crown each bore a row of small cusps. The faint ridges which ran forwards and backwards from the external and internal main cusps of the uppers were not worn by the lowers to produce the vertical shearing surface seen in Diade- modon ; some degree of shearing may have been present between the sectorial teeth, but these were so small in comparison with the gomphodont teeth that it could not have been significant. The gomphodont teeth were presumably used to puncture and crush. The transverse ridges of the upper and lower teeth alternated with one another during occlusion but the wear facets suggest that they could not have sheared effectively.

1} have assumed that Diademodon was derived from a galesaurid similar to Thvinavodon. This assumption is based upon the similarity of the sectorial teeth of Diademodon to the teeth of Thrinaxadon and upon the anteroposterior alignment of the cusps on the external edge of the upper teeth of Diademo- don. It is based also on the fact that galesaurids were the dominant cynodonts of the Lystvosaurus zone and that as a family they represent a distinct advance over the procynosuchids of the earlier Kistecephalus zone. However, some or all of the gomphodont cynodonts may have arisen directly from the Procynosuchidae (Bonaparte, 1963); this would explain the single external cusp on the gomphodont teeth of the Trirachodontidae.

40 POSTCANINE OCCLUSION

In the Trirachodontidae there was no alternate tooth replacement and, as in Diademodon, new gomphodont teeth were added behind to replace sectorial teeth.

IV. POSTCANINE DENTITION OF THE TRAVERSODONTIDAE

Until recently relatively little was known about Middle to Late Triassic therapsids (Crompton 1955, 1963B); recent work in South America, South Africa and East Africa, however, has led to the discovery of numerous therapsid remains, most of which have still to be described. It is now clear that the gomphodont cynodonts of the family Traversodontidae were the dominant cynodonts during the Middle Trias and early part of the Late Trias. The basic pattern of the crowns of the postcanine teeth of the traversodontids is characteristic of the group and separates it very clearly from the earlier gomphodont cynodonts. Relatively minor differences in the dentition and occlusal relations distinguish the genera of traversodontid cynodonts from each other, the structure of the postcanines providing a useful key for identify- ing genera and species and for determining the relationships and evolutionary history of the group. Like those of diademodontids, the postcanines of traversodontids were subjected to severe wear which in most cases eventually obliterated all details of the crown pattern. It will be shown below that the wear of the crowns of traversodontid postcanines resulted from a combination of complex occlusal relation- ships and jaw movements. The occlusion of traversodontid postcanines closely paralleled that of primitive mammals with tribosphenic molars. The abundance and diversity of traversodontids during Middle Triassic times may have been partly due to their highly evolved masticatory apparatus.

In an earlier paper (Crompton 1955) three new genera of cynodonts (Scalenodon, Cricodon and Aleodon) and several indeterminate remains were described from the Manda Formation of Tanzania. No attempt was made to classify them above the generic level. Romer (1967) has recently revised the classification of gomphodont cynodonts; it is now apparent that Scalenodon angustifrons and the isolated maxilla with two teeth which I compared with the South American genus Gomphodontosuchus should both be included in the family Traversodontidae, while Aleodon brachyrham- phus is clearly not a traversodontid and Cvicodon metabolus appears to have been a surviving member of the Trirachodontidae. This accords with a view expressed by Bonaparte (1963). Subsequently Brink (1963a) described a gomphodont, Luangwa drysdalli, from the Ntawere Formation of Zambia; it is not well preserved and the occlusal aspects of the teeth are not known, but it appears to have been closely related to Scalenodon angustifrons, if not actually identical. In 1963 the British Museum (Natural History)—University of London Joint Palaeontological Expedi- tion (Attridge, Ball, Charig & Cox, 1964) collected additional material from the Ntawere Formation of Zambia and the Manda Formation of Tanzania; at least one new genus and two new species of traversodontid cynodonts were discovered as well | as additional specimens of Scalenodon angustifrons and Aleodon brachyrhamphus. I hope to give a full description of this new material in a later publication. Mean- while, because the postcanine teeth of all the East African traversodontids (including the new material) are discussed below, the new specimens have been named and briefly described in order to avoid future confusion.

CYNODONTS AND TRITYLODONTIDS 41

1. Description of three new species of Scalenodon from the Middle Trias of East Africa

Family TRAVERSODONTIDAE von Huene 1936 Genus SCALENODON Crompton 1955

TypeE-sPecies. Trivachodon angustifrons Parrington 1946.

Since I first described Scalenodon (Crompton 1955 : 647) other genera have been discovered in East Africa, South Africa and South America which are closely related to it. Because of this the original diagnosis given for Scalenodon is no longer applicable, several of the supposedly diagnostic features mentioned having sub- sequently proved to be diagnostic of the family Traversodontidae as a whole rather than of the genus Scalenodon in particular. A revised diagnosis for Scalenodon is therefore given below.

This diagnosis is based entirely upon the characters of the upper postcanine teeth; it does not include features of the lower postcanine dentition because the latter is not known in all species of the genus. A detailed description of the rest of the skulls and skeletons of the several species of Scalenodon is now projected; this may necessitate some revision of the classification suggested in this paper.

DiacGnosis. Small to medium-sized traversodontid cynodonts in which the upper postcanine teeth are transversely ovate; the external margin of the crown is gently convex; two main cusps are present (external and internal), the latter lying at the internal end of a prominent transverse ridge which usually supports an additional (central) cusp; small antero-external and antero-internal cusps are occasionally present; the inner surfaces of the external cusps form a vertical, antero-posteriorly aligned shearing surface, towards which the accessory cusp does not contribute substantially (contrast South American traversodontids); anterior and posterior cingula are present ; during the power stroke of occlusion the backward movement of the lower jaw was limited so that matching transverse ridges, upper and lower, were never drawn across one another.

ComMENts. The upper postcanines of the four species of Scalenodon show some striking similarities to those of various genera of traversodontid cynodonts from South America. The type-material from both continents is at present being studied in order to determine, if possible, whether those similarities are due merely to con- vergence or to a closer phylogenetic relationship than is indicated by the present classification.

S. angustifrons (Parrington). Material of this species includes not only the holo- type (Cambridge University Museum of Zoology, Ruhuhu Field Catalogue no. 120B) but also a number of other specimens which I referred to it when proposing the genus (Crompton 1955). Details of horizon and localities are given and discussed in the same work. Now that three new species of Scalenodon have been recognized (see below) it is possible to give a specific diagnosis for S. angustifrons. The diagnostic characters of all four species are compared in Table 1. Note also that in S. angirsti- Jrons there is a ridge of small cuspules on the outer surface of the main cusp of the upper postcanines.

POSTCANINE OCCLUSION

42

pedojeaep Aj100d

[[@M Ior1ezue o}epidsno-u0u surioy ‘podojeaep [jam

[7eM Ior1ozue YSsy Aq peyoeuuod ‘a81e] yA0q

yuesqe oq 0} sieodde dsno je1}u99 ‘UMOIO JO O8pa

Ior1e}sod jo y1ed sur0F jerered

A]Ysnor suIsieur Iotio}sod pure rIoreyue

anbrjqo Aaa

uMouyun

ge1

UsIADYI “S

MmosyIsSAry *S Ul UeYy} yUoUTUIOId sso] nq pedojeasp |[JemM

souturojsod

Ior19}sod ur [yea

Joliejue 9}eprdsno sul1oy podoyeAep [jam

as81e] [eUIS}UI ‘]]eUuIS [PUIE}x9

sdsno

[e1}ue9 pure [eure}xe

sjtoddns ‘uorjztsod UI Io110}sod 910 [14S

Jo}no ueyy JopIA UOTSeI IoUUT

anbijqo

Ajiorsieysod our|-prur wio1y APMP SUTAINO

yol

1aspUyD °S

pedoyjeasp [Jom

pedoyjeaesp [jam

jyuesoi1d

sdsno ye1}Ue0 pue [eure

-xo szioddns ‘1aMoy

‘SUOA{YSNSUD *S Ul ULY} uorisod ur ror1e3sod o10ul

SUOALYSNSUD “S$ UI uey} IepnsuezeI s10Ur

enbijqo jou

oul[-prut 0} JeTvered Ayreou

vl

WMoOsYyIsAYy *S

I alavy

pedojeasp Ay100d

pedojasep Aj1ood

yuesqe

sdsno [e1} W989 pue [eUuIe}

-xo sjioddns ‘sjjem

Ior1eysod pur ror1ezue

deozs YIM YstY “y}00} FO a[ppruz puryeq

Iojno uey} IopIM I9AOU WOTSEI IOUUT

enbiyqo jou

Ajiori1e}sod out[-prut wor Aeme SUTAINO

ar

SUOALYSNSUD *S

:wInpNsuro Io1103sod

:UIN[NSUIO IOLIeZUe

:sdsno Aiossao0e IOTIE}Ue [eUrIe}UI pue [euIE}xo

!OSPII OSIOASUeIL

1 MOTA UMOIO UI 4490} yuopoyduros jo odeys :e][IXeUL UL 399} JO SurzZesS

: MOI-Y}.004 jo uorztsod

:oInsy-}xoL

CYNODONTS AND TRITYLODONTIDS 43

Scalenodon hirschsoni sp. nov. Plate 5; Text-figs 7-9

DERIVATION OF NAME. In honour of Dr B. Hirschson, who was a member of the British Museum (Natural History)—-University of London Joint Palaeontological Expedition to Northern Rhodesia and Tanganyika, 1963, and who has done so much to help vertebrate palaeontologists and archaeologists working in southern Africa.

MaTERIAL. Only the holotype in the British Museum (Natural History), Pal. Dept. regd. no. R. 8577. Field catalogue no. U12/3/26. Partial skull with well- preserved postcanine teeth.

Horizon. Manda Formation (Middle Trias, probably Anisian).

Locatity. Ruhuhu Valley, $.W. Tanzania: Locality U1r2 of the B.M.(N.H.)— University of London Joint Expedition, 1963. Between the Hiasi and Nijalila streams, just south of the Rutukira River; the most northerly of the Expedition’s localities west of the Njalila.

DEscripTIon. See Table 1 for diagnostic characters. Note also that the incisors and canines, both upper and lower, are procumbent. The postcanines are described in greater detail on p. 49.

Scalenodon attridgei sp. nov. Plate 6; Text-figs 10A, B

DERIVATION OF NAME. In honour of Mr John Attridge, of Birkbeck College, London, who was a member of the British Museum (Natural History)—University of London Joint Palaeontological Expedition to Northern Rhodesia and Tanganyika, 1963, and who made many of the discoveries of that expedition, including the unique specimen of this species.

MATERIAL. Only the holotype in the British Museum (Natural History), Pal. Dept. regd. no. R. 8578. Field catalogue no. U2/z. A beautifully preserved snout with complete upper dentition; probably a young individual.

Horizon. Manda Formation (Middle Trias, probably Anisian).

Locatity. Ruhuhu Valley, S.W. Tanzania; Locality U2 of the B.M.(N.H.)— University of London Joint Expedition, 1963. Immediately next to the Peramiho- Litumba dirt road, on its left (south-western) side; low in the K8, before the Expedi- tion’s turn-off to Njalila and Mkongoleko, and probably only about a mile before the K7 boundary.

Description. See Table 1 for diagnostic characters. Note also that the anterior edge of the unworn crown of the upper postcanines bears a row of five distinct cusps, of which the innermost is the largest. The postcanines are described in greater detail on p. 53.

B

44 POSTCANINE OCCLUSION

Scalenodon charigi sp. nov. Text-fig. 10C

DERIVATION OF NAME. In honour of Dr Alan J. Charig, Curator of Fossil Reptiles in the British Museum (Natural History), who initiated and participated in the B.M.(N.H.)—University of London Joint Palaeontological Expedition to Northern Rhodesia and Tanganyika in 1963.

MATERIAL. Only the holotype in the Cambridge University Museum of Zoology, Ruhuhu Field Catalogue no. 136. Part of a left maxilla in which two postcanine teeth are preserved.

Horizon. Manda Formation (Middle Trias, probably Anisian).

Locatity. Ruhuhu Valley, S.W. Tanzania; Locality B26 of Stockley (1932 : 620). Gingama, south of the Ruhuhu River.

DESCRIPTION. See Table 1 for diagnostic characters. Note also that the central main cusp of the upper postcanines appears to be absent (this part of the tooth is damaged).

ComMMENTS. This specimen was described and figured by Crompton (1955: 659-660, fig. 14E), who compared it with the Brazilian species Gomphodontosuchus brasiliensis. It is now clear that its postcanine teeth are distinct from those of the other Ruhuhu traversodontids; the antero-external and internal accessory cusps are better developed than in the other East African forms, the antero-external forming part of a vertical shearing surface. It is therefore desirable, despite the smallness of the preserved portion, to base a new species on this specimen.

S. charigi is similar in some respects to Gomphodontosuchus brasiliensis but is nevertheless distinct from that too. The postcanine teeth, though smaller, closely resemble those of the South American genera Exaeretodon and Gomphodontosuchus; the matching transverse ridges, however, did not cross one another during the power stroke of occlusion as they did in Exaeretodon and Massetognathus.

2. Occlusion and jaw movements in the primitive traversodontid Scalendon angustifrons

The most primitive traversodontid postcanines known are those of Scalenodon angustifrons.2 Although the dentition of this species has been described in detail (Crompton 1955), the new material discovered by the 1963 Joint Expedition has revealed many features of the postcanines and their occlusion which were not dealt with before.

Fig. 4 shows three views of typical upper and lower postcanines of S. angustifrons; stereo-photographs are given in Plate 4.

2Tf Andescynodon and Pascualgnathus described by Bonaparte (1966, 19676) are considered as traver- sodontids, then these should be regarded as the most primittve members of the family. I, however, should prefer to place them in the Diademodontidae.

CYNODONTS AND TRITYLODONTIDS 45

Upper teeth. These consisted essentially of three cusps arranged to form a transverse row, the central cusp being nearer the inner side of the crown. The internal (i.c.u.) and central (c.c.u.) cusps formed a high, prominent transverse ridge (t.r.) with anterior and posterior surfaces nearly vertical; a deep embayment (e.) separated this ridge from the external cusp (e.c.u.). A sharp ridge (a.r.) running forwards from the apex of the external cusp had a nearly vertical inner face which formed the outer border of a deep valley (a.v.) in the anterior surface of the tooth,

ANTERIOR

lcm EXTERNAL

Fic. 4. Scalenodon angustifrons. Postcanine teeth. A. Posterior view of upper. B. Posterior view of lower. C. Crown view of upper. D. Anterior view of upper. E. Anterior view of lower. F. Crown view of lower. For key to abbreviations see p.69

46 POSTCANINE OCCLUSION

immediately in front of the embayment between the external and central cusp (see Fig. 4D). Asmaller, less well defined valley (p.v.) occupied a corresponding position behind the transverse ridge. The anterior valley of one postcanine and the posterior valley of the preceding tooth together formed a deep occlusal basin with high antero- posteriorly aligned shearing surfaces.

Lower teeth. The crown of a lower postcanine of S. angustifrons consisted of two high anterior cusps, the external (e.c.l.) being higher than the internal (i.c.l.); a high ridge ran backwards along the outer surface of the crown (p.r.) from the apex of the external cusp, its outer surface being more or less vertical. The external and internal cusps were connected by a transverse ridge which dipped down to a saddle between them; the anterior surface of this ridge was nearly vertical. A deep basin (b.1.) lay behind the two main cusps, rimmed posteriorly by a row of small cuspules of which the outermost (p.a.c.) was the largest. A small anterior accessory cusp (a.a.c.l.) was present on the anterior surface of the crown.

OccLusion. In Fig. 6 several opposing postcanines are shown in crown view and oblique internal view at the beginning of dynamic occlusion (A, B, respectively) and at the end (C, D). At the beginning of dynamic occlusion the anterior surface of the transverse ridge formed by the two main cusps of the lower tooth sheared past the posterior surface of the transverse ridge formed by the central and internal main cusps of the preceding upper tooth. (In Fig. 6B lower postcanine 6 is shearing

Fic. 5. Scalenodon angustifrons. Lateral views of postcanine teeth, showing the postero- dorsal direction of the power stroke. A. At the beginning of dynamic occlusion. B. At the end of dynamic occlusion.

CYNODONTS AND TRITYLODONTIDS 47

against upper postcanine 5). This produced wear facets on the front of the trans- verse ridge of the lowers and on the back of the transverse ridge of the uppers. The central cusp of the upper tooth fitted into a groove on the anterior surface of the lower tooth between the two main cusps. The outer surface of the external main cusp of the lower tooth sheared past the posterior portion of the vertical internal surface of the external main cusp of the preceding upper postcanine; the position of the postcanines at the beginning of dynamic occlusion as seen in lateral view is given in Fig. 5A, which shows the resulting striations (w.f.) on the outer surface of the external main cusp of the lowers. As the jaws continued to close (Figs 5B and 6D) the lower jaw moved slightly backwards rs well as upwards so that the external surface of the lowers sheared past the anterior portion of the vertical internal surface of the external cusp of the corresponding upper postcanine tooth. The arrows in

EXTERNAL S 6 i

ANTERIOR

lcm

Fic. 6. Scalenodon angustifrons. Details of tooth occlusion. The numbers 4 to 7 indicate the positions of the teeth in the postcanine series. A. Crown views of the postcanines at the beginning of dynamic occlusion, with upper and lowers superimposed. B. Oblique internal view of the same. The heavy arrow indicates the passage of the tip of the internal main cusp of the lower teeth during dynamic occlusion. C. Crown views of the postcanines at the end of dynamic occlusion, with uppers and lowers super- imposed. The main transverse and longitudinal shearing surfaces are drawn in heavy lines. D. Oblique internal view of the same. For key to abbreviations see p. 69.

48 POSTCANINE OCCLUSION

Figs 6B and 5A indicate the direction of movement of a single lower postcanine during the final stages (‘“power stroke’’) of occlusion. The posterior surface of the external cusp of the lower tooth and the anterior surface of the external cusp of the corresponding upper formed two opposing crescents when seen in lateral view (Fig. 5A). Such an arrangement is ideal for cutting provided that the lower jaws moved slightly backwards during occlusion. An analogous arrangement typified the shearing surfaces of tribosphenic molars (Crompton & Hiiemae, 1969b). When the postcanines of S. angustifrons were in tight occlusion (Fig. 6, C and D) the external main cusp of a lower postcanine lay in the valley (Fig. 4D, a.v.) in the anterior surface of the corresponding upper postcanine and the central cusp of the upper lay above the posterior basin of the lower. The dentition of S. angustifrons was characterized also by the addition of new gomphodont teeth with shearing surfaces behind and by the loss of worn postcanines in front. The replacement pattern was similar to but simpler than that of Diademodon.

The occlusal pattern of S. angustifrons represented a distinct advance over that of Diademodon. Not only were transversely orientated shearing planes added, but the occlusal basin of the upper postcanines was considerably deepened by the develop- ment of deep valleys immediately internal to the external main cusp. Consequently the tip of the main cusp of the lower postcanine was not worn down as rapidly as it was in Diademodon, where it abutted directly against the crown surface of the occluding tooth. Deepening the occlusal basin also increased the height of the shearing surfaces.

The postcanines of S. angustifrons therefore consisted essentially of a series of transversely and longitudinally orientated shearing planes. The positions of these shearing planes on upper and lower postcanines are indicated by heavy lines on Fig. 6C. Occlusion in other traversodontid cynodonts and tritylodontids was basically a modification of the arrangement that was present in Scalenodon angusti- frons. The tips of the cusps were used for puncturing, the sides of several of the cusps for shearing and the posterior heel of the lower posterior postcanines provided a firm basin for crushing, analogous to the talonid basin of the tribosphenic molar. The potscanines of S. angustifrons were therefore functionally similar to the tribo- sphenic molars of primitive mammals. In mammals with tribosphenic molars and in some of the insectivores and herbivores with more specialized molars the jaw moves not only vertically during occlusion but also transversely and forwards in order to utilize a series of shearing plates. In S. angustifrons the mandible moved vertically and slightly posteriorly during occlusion and thereby also utilized a series of shearing planes.

It is generally assumed that in primitive cynodonts the tympanic membrane was partially attached to the posterior surface of the quadrate (see Hopson, 1966 for a complete review of this problem). The posterior movement of the jaw during dynamic occlusion in S. angustifrons was apparently too great to be accommodated within the available space between the glenoid of the articular and the condyle of the quadrate. Parrington (1946) suggested that in Thvinaxodon and later cynodonts the quadrate itself must have been capable of antero-posterior movement but recognized that such movement would have torn or stretched the small tympanic membrane

CYNODONTS AND TRITYLODONTIDS 49

because the stapes rested against the quadrate. In a new skull of S. angustifrons (B.M.(N.H.) R. 8579), discovered in 1963 in Tanzania, it can be seen that the quad- rate was held in a groove in the squamosal and could slide both downwards and forwards; the amount of movement of which it was capable appears to have been sufficient to have allowed the mandible to be pulled backwards a little during the final stages of dental occlusion. Kemp (1969), following on the earlier work of Parrington (1955), has shown that the quadrate was extremely mobile in gorgon- opsians too so that, despite the firm junction between the articular and the quadrate, the mandible was capable of antero-posterior movements during mastication; a mobile quadrate was presumably present in all cynodonts and therocephalians. In S. angustifrons the external auditory meatus presumably lay in a groove of the squamosal and, as Parrington (1946) has shown, the groove was terminated by a semicircular lip which supported the tympanic membrane without involving the posterior surface of the quadrate. Movement of the quadrate would therefore not have involved the tympanic membrane directly, but this does not solve the problem completely as the stapes was presumably in contact with both tympanic membrane and quadrate, and stapes and quadrate may have moved together. Unfortunately the relationship between the stapes and the quadrate of advanced cynodonts is not well known, but the removal of the tympanic membrane contact from the quadrate to the squamosal in primitive traversodontids may be related to the antero-posterior movements of the mandible which appear to have taken place during occlusion.

The postcanine tooth rows of S. angustifrons diverged backwards. Consequently antero-posteriorly aligned cutting surfaces of opposing teeth would have tended to separate during extensive backward movement of the lower jaw and for this reason the amount of antero-posterior movement during occlusion in S. angustifrons was probably small. It is doubtful whether the jaw could have swung far enough laterally to retain contact on one side. A mobile mandibular symphysis would have overcome this limitation, but the nature of the fossil material suggests that the two rami were firmly united. The fossilized remains of S. angustifrons consist of numerous fragments, indicating that the skeletons of this animal were usually scattered and broken before fossilization; despite this the mandibular rami are usually found fused at the symphysis, as would not be expected had the symphysis been mobile during life. By contrast, the mandibular rami of tritylodontids and early mammals are seldom if ever preserved fused at the symphysis, which suggests that the latter was mobile. Szalay (1969) has argued that primitive primates too had a mobile symphysis because Palaeocene primate mandibles are usually preserved separated.

3. Occlusion and jaw movements in Scalenodon hirschsoni

The only known specimen of S. hivschsont is ideal for the study of occlusion because it yielded to preparation with acetic acid; the lower jaw was thereby freed from the remainder of the skull. It was possible to study details of the structure of the teeth, the wear facets and occlusal relationships. Among the features of this species which clearly separate it from S. angustifrons are that the upper incisors and lower canines

50 POSTCANINE OCCLUSION

are slightly procumbent and that the postcanine rows are nearly parallel to one another, the last postcanine lying immediately in front of the transverse process of the pterygoid. The basic structure of the crowns of the postcanines (Figs 7A, B, 8; Plate 5) is essentially the same as that of S. angustifrons. The crowns of the upper postcanines are relatively longer antero-posteriorly than those of S. angustifrons and the portion of the crown lying in front of the transverse ridge is considerably wider

1.€.U.

ANTERIOR

EXTERNAL

Icom

Fic. 7. Scalenodon hirschsoni sp. nov. Postcanine teeth. A. Crown view of upper. B. Crown view of lower. C. Crown views of uppers and lowers superimposed to show the relative positions at the beginning of dynamic occlusion. D. The same, at the end of dynamic occlusion.

| |

|

CYNODONTS AND TRITYLODONTIDS 51

than the portion behind; this is probably due more to addition to the anterior part of the crown than to a posterior migration of the transverse ridge because small additional cusps (a.a.c.u.) are present on the ridges running forwards from both external and internal main cusps. The embayment between the external and central main cusps (Fig. 8A) is deeper than that of S. angustifrons and the transverse ridge is not as high as in the latter species. A fairly prominent posterior cingulum (Fig. 7A, p.c.) is present.

In the lower postcanines (Fig. 7B) the transverse ridge formed by the two maiu cusps is not as high as that of S. angustifrons and the anterior accessory cusp (a.a.c.].) islarger. Details of occlusion and of the amount of longitudinal movement during mastication are illustrated in Figs 7C—D, 8A—D and 9A-E. Because the transverse ridge of the uppers also is lower than in S. angustifrons and because it was apparently worn down fairly rapidly, its posterior surface does not form a high wall (see internal views of beginning and end of dynamic occlusion, Fig. 8C—D); nevertheless small matching wear facets on that surface and on the anterior surface of the transverse ridge of the lower postcanines indicate that some shearing took place in this position. The mandibular movements which probably took place during occlusion have been reconstructed by manipulating the opposing jaws and by studying the striations on the wear facets of opposing teeth. These movements are illustrated in Fig. 9 in external view by showing several positions of the lower postcanines 5 and 6 relative to the upper postcanines 4 and 5. As the jaws closed, the anterior part of the external surface of the main cusps of lowers 5 and 6 sheared past the internal surfaces of the external main cusps of uppers 4.and 5 (Fig. gA—B). As the mandible proceeded backwards (Fig. 9B—C—D) the external surface of the external main cusp of lower postcanine 5 sheared past the internal surface of the external main cusp of upper postcanine 5. This shear, as it would have appeared in internal view, is illustrated in Fig. 8C-D. The wear facets on the external surface of the lower postcanines resulting from this backward movement are shown in Fig. 9F. The important point is that at the beginning of dynamic occlusion the transverse ridge of the 5th lower postcanine lay behind the transverse ridge of the 4th upper postcanine (Fig. 7C). In essence, therefore, occlusion resulted from a posterior and a vertical jaw movement, just as in S. angustifrons. However, a lightly worn 5th lower postcanine of S. hirschsom shows a distinct wear facet (Fig. 7B, p.w.f.) on the posterior surface

_ of the main cusp; this matches a wear facet on the anterior surface of the transverse

ridge of the 5th postcanine. It is difficult to account for these facets if the power stroke of the lower jaw was directed dorso-posteriorly. Admittedly they could have resulted from the postero-dorsal surface of the transverse ridge of the lower tooth being drawn backwards and downwards across the antero-ventral surface of the transverse ridge of the upper. This movement, as it would appear in external view, is shown in Fig. gD. Although this would have involved crushing between the opposing transverse ridges, it would have required that the lower postcanines be

_ dragged down an inclined plane. The same wear facets, however, would have been

formed if the mandible had moved forwards and upwards during dynamic occlusion as shown in Fig. 9E so that the leading edge of the transverse ridge of the lowers sheared past the trailing edge of the transverse ridge of the uppers. This movement

POSTCANINE OCCLUSION

lom

Fic. 8. Scalenodon hivschsoni sp. nov. Postcanine teeth. A. Posterior view at the beginning of dynamic occlusion. B. Posterior view at the end of dynamic occlusion. C. Oblique internal view at the beginning of dynamic occlusion. D. Oblique internal view at the end of dynamic occlusion. E. Sagittal section through part of opposing postcanines to show matching shearing surfaces resulting from anteriorly directed power stroke. F. Internal view of lower postcanine to show wear facet resulting from anteriorly directed power stroke.

CYNODONTS AND TRITYLODONTIDS 53

as it would appear if a longitudinal section were cut through the teeth, is shown in Fig. 8E; the enamel-like material (en.) is considerably thicker on the anterior surface of the transverse ridge of the lower tooth than on the occlusal surface or in the posterior basin, just as would be expected if this species were capable of a forwardly directed power stroke as well as the usual backwardly directed stroke.

4. Postcanine dentition of Scalenodon attridgei

This species is known only from an isolated snout, with the upper teeth well preserved on both sides. The postcanine row (Fig. 10A, B, Plate 6) of eight teeth ends behind in three teeth which become progressively smaller, as in many specimens of S. angustifrons. The first five teeth are so worn that most of the details of crown structure are lost, but the 6th and 7th postcanines are only slightly worn and the 8th not at all. Although the last two teeth are smaller than the more anterior ones and would presumably have been replaced later in life by larger gomphodont teeth, they are of great interest. The basic pattern of the postcanines of S. attridgei is similar to that of S. angustifrons and S. hirschsoni. They are, however, set obliquely in the maxilla, and the internal surface is slightly wider than the external surface. The

Fic. 9. Scalenodon hirschsoni sp. nov. Postcanine teeth. A-D. External views to show successive stages of the posteriorly directed power stroke. Heavy arrow in A indicates total extent of backward movement involved. E.D. The same, but to show anteriorly

directed power stroke. F. External view of lower teeth to show wear facets on the external surfaces.

54 POSTCANINE OCCLUSION

important feature ot the upper postcanines of S. aftridgei is the presence of a row of well developed cuspules along the anterior border of the crown of postcanines 7 and 8; in postcanines 5 and 6 they have been obliterated by wear. The antero-medial cuspule (a.a.c.u.) is the largest of these cuspules and they tend to become smaller towards the exterior (Fig. 10B). A high ridge joins the external main cusps to the most external cuspule of the anterior row. The central and internal main cusps (c.c.u. and i.c.u.) are both large and well differentiated and form the transverse ridge. A basin (b.u.) is present in the occlusal surface of the crown; this is bordered behind by the transverse ridge, in front by the anterior row of cuspules and externally by the vertical inner wall of the external main cusp. A deep valley separates the internal main cusp and the most internal cuspule of the anterior row (a.a.c.u.) so that the basin is completely surrounded except for this narrow valley opening internally. In the 7th postcanine the tips of the central and internal main cusps and the tips of the cuspule forming the anterior ridge are worn away. In the more anterior teeth wear has tended to obliterate the original details of the crown pattern and especially the anterior row of cuspules. The tips of the central and internal main cusps and the crest of the transverse ridge have been worn away so that the occlusal surface of the crown in front of the transverse ridge is a plane sloping slightly upwards in an antero- external direction; this is best seen in the anterior view of the upper postcanines (Fig. 10B). Except for two internal cuspules, most of the cuspules of the anterior row of the 6th postcanine have been worn away and only a low wall remains. A feature of the crown of the 5th and 6th postcanines which is not present in the smaller 7th and 8th is a faint posterior ridge or cingulum (p.c.) close to the posterior margin ot the crown. A shallow valley which widens slightly towards the external side lies anterior to it; this will be referred to as the posterior basin (p.b.). The wide forward and upward sloping flat surfaces of the crowns of the upper postcanines of S. attridget could have resulted from both antero-dorsal and postero-dorsal power strokes. The postcanines of the Brazilian species Tvaversodon stahleckeri (von Huene 1944 : 48) are extremely worn, but from what remains of the upper teeth they appear to have been similar to those of S. attridgez.

5. Postcanine dentition of Scalenodon charigt

This specimen, consisting of an isolated maxilla with two posterior postcanines, was previously compared (Crompton, 1955) with the Brazilian species Gomphodonto- suchus brasiliensis. Although the teeth were badly damaged several details can still be seen, and, now that more traversodontids from South America and East Africa have been described and figured, additional comments may be made.

The teeth (Fig. 10C) are set obliquely in the maxilla. The external anterior

accessory cuspule (a.a.c.u.) is larger than in S. hivschsoni and the internal surfaces of ©

the two external cusps form a high shearing surface. The main transverse ridge is situated near the posterior edge of the crown. The central cusp appears to be absent. The anterior wall is high and is terminated internally by a high rounded cuspule (a.a.c.u.) ; consequently the anterior basin (b.u.) is deep and occupies most of the occlusal surface of the crown. The shearing surface on the internal face of the

CYNODONTS AND TRITYLODONTIDS 55

external cusp is continuous with that on the posterior surface of the anterior wall of the same tooth; this suggests that the anterior surface of the transverse ridge of the corresponding lower postcanine sheared up the anterior wall of the upper postcanine rather than across it. Although the anterior wall was present in S. attridgei, it was

| cm

lcm

Fic. 10. A. Scalenodon attridgei sp. nov. Crown view of last four upper postcanines. B. Scalenodon attridgei sp. nov. Anterior views of the same. C. Scalenodon charigi sp.nov. Crown view of last two upper postcanines. For key to abbreviations see p. 69.

56 POSTCANINE OCCLUSION

rapidly worn away and apparently did not form a significant vertical shearing sur- face. The structure of the anterior wall, the absence of a central cusp and the oblique position of the crown relative to the longitudinal axis of the palate are reminiscent of the South American genus Exaeretodon.

6. Occlusion and jaw movements in a species of Massetognathus

The postcanines of the Argentine traversodontid Massetognathus pascuali have been described in detail by Romer (1967), but he does not discuss occlusion in detail. Dr. Bonaparte presented the Peabody Museum with a jaw fragment that has been tentatively assigned to the genus Massetognathus, and in which upper and lower postcanines were preserved in tight occlusion; the jaws have been carefully separated by Mr C. Schaff. Distinct wear facets are preserved on most of the teeth, and by matching upper and lower wear facets it has been possible to determine the jaw movements that must have taken place in Massetognathus during the final stages of the masticatory cycle. Although the postcanines are almost identical to those of M. pascuali the wear pattern is slightly different and the fragment should perhaps be placed in a different species or genus. This may be possible when the entire fauna of gomphodont cynodonts from Chafares is fully described.

The postcanines (Fig. 11 & Plate 7) are very similar to those of S. angustifrons except that two accessory cuspules rather than one are present on the ridge leading forwards from the external main cusp, which latter lies further back than in S. angustifrons. A characteristic feature of Massetognathus and Exaeretodon not present in the African traversodontids is that the external margin of the upper post- canines as seen in crown view is drawn outwards and backwards (Fig. 11B) to forma distinct lobe. The transverse ridge is high and forms the posterior border of the crown, the posterior cingulum being but poorly developed. The inner surface of the external main cusp and of the two anterior accessory cuspules forms a high vertical wall which is more pronounced than in S. angustifrons. The point of junction between the transverse ridge and the external main cusp lies near the posterior border of the crown. The lower postcanines are similar in structure to those of S. hirschsont but lack the anterior accessory cuspules. As in all gomphodont cyno- donts the amount of wear increases progressively towards the front of the jaw. Matching shearing planes are shown in Fig. 11D & E. As the teeth came into occlusion the outer surface of the 4th lower postcanine (Fig. 11D) sheared past the internal surface of the external main cusp of the 3rd upper postcanine (wear facets 2 in Fig. 11C, D & E); then, as the jaws continued to close, it also sheared past the inner surface of the two anterior accessory cuspules of the 4th upper postcanine (wear facets 1). The two facets, 2 and 1, on the 3rd and 4th upper postcanines respectively, are therefore continuous. This action was similar to that in Scalenodon. At the beginning of occlusion the tips of the two main cusps of the 4th lower postcanine lay anterior to the tips of the central and external main cusps of the 3rd upper post- canine, i.e. the transverse ridge of the lower lay in front of the transverse ridge of the upper (Fig. 14). The central cusp of the upper lay directly behind the valley separating the two lower cusps. Because of this, as the mandible was drawn back-

Sif

CYNODONTS AND TRITYLODONTIDS

‘wOIsN{N00 ormeUAp jo Suruurseq oy} ye Ired Sutsoddo jo morta so110}Ss0g “YY

‘soovjins Sureoys f sioddn omy Jo MorA uMOID “|

Suryoyeur Moys 0} “IQMOT JO MOIA UMOID “CG

‘ayO4s Iomod sy} Jo sasejs [euy oy} Sutmnp sioddn yj 0} oatyeIor

Y}00} IoMo] & JO (Moxre Aavay) JUSUIEAOUI oY} PU 4003 10ddn 943 Jo sdsno [euIE}xe

a4} JO SOP} [EUIN}UI 9Y} UO SBdvJINSs SuIIveYys oy} Jo yYZdep oy} MOYS 0} MOIA [eUIOZUT

‘9 «xoddn uiom Aysis JO MoIA UMOID “q “IOMO] UIOM APYSI[S JO MorA UMOID “VW "y}09} euTUeOJsSOg + ‘(UInesnyY Apoqeeg efexX ut ueutoeds) ‘ds snyzwusojassppy “11 “DI

fos hE Vv

YOINSLNV

58 POSTCANINE OCCLUSION

wards and the lower cusps were dragged across the transverse ridge, pronounced wear facets were formed on either side of the central cusp of the uppers and on the tips of the main cusps of the lowers. These matching facets have been numbered 4 and 5 in both uppers and lowers (Fig. 11, D & E). As the mandible was drawn further back the main cusps of each lower tooth carved grooves into the anterior wall of the succeeding upper tooth. These grooves have been numbered 6! and 7! on the upper tooth (Fig. rrE) and the matching shearing surfaces 6 and 7 on the lower tooth (Fig. 11D). It is clear from the orientation of these matching facets that they were produced by the backward and upward movement of the mandible during the final stages of the masticatory cycle. A study of the wear facets in this species of Massetognathus therefore suggests that, as in S. angustifrons, the jaws were drawn backwards during dynamic occlusion; but those facets which suggest an anteriorly directed power stroke in S. hirschsoni and S. attridget are absent in Massetognathus. Perhaps the most important feature of occlusion in this species of Massetognathus is that the transverse ridge of the lowers was drawn across the transverse ridge of the uppers. This was not the case in Massetognathus pascuali, S. angustifrons or S. hirschsoni, where the anterior surface of the transverse ridge of the lowers sheared up the posterior surface of the transverse ridge of the uppers. It is possible to derive the situation in the new species of Massetognathus from that in S. angustifrons simply by increasing the extent of the antero-posterior movement during occlusion. This is important when considering the ancestry of the tritylodontids.

7. Postcanines of Exaeretodon and Gomphodontosuchus

Bonaparte (1962) has described the teeth of Exaeretodon frenguelli; 1 was able to study postcanine teeth of this species in the Museum of Comparative Zoology at Harvard and at the Instituto Miguel Lillo in Tucuman. The upper postcanines (Fig. 12A, B, C) are similar to those of Massetognathus; and the postero-external extension of the external region is so marked that the tooth as seen in crown view may be divided into two lobes, a lateral and a medial. Well developed external (e.c.u.), anterior accessory (a.a.c.u.) and posterior accessory (p.a.c.) cusps are present

on the external margin of the tooth; their internal surfaces form a continuous antero- |

posteriorly aligned shearing surface which extends right along the tooth from front to back. The transverse ridge terminates short of the base of the external main cusp; its central cusp is absent. Two anterior accessory cusps (a.a.c.u.) are present, one internal and one external. In posterior view (Fig. 12C) it can be seen that the crest of the transverse ridge rises very sharply ventrally towards the tip of the internal main cusp. The anterior wall (a.w.) of Exaeretodon is a prominent feature; it can be seen in anterior view (Fig. 12B) that it too rises very sharply ventrally to

terminate in the internal anterior accessory cusp. A shallow basin separates this:

anterior wall from the transverse ridge. The cutting surface on the inner face of the external main cusp of each tooth is continuous with those in front and behind. Because the anterior wall and the transverse ridge of each upper postcanine lie obliquely the internal main cusp of each lower postcanine is considerably further back than the external (Fig. 12D). Mandibular movements during dynamic

CYNODONTS AND TRITYLODONTIDS 59

occlusion were probably similar to those of Massetognathus pascuali and Scalenodon angustifrons. Wear facets indicate that the transverse ridges of the upper and lower postcanines were not drawn across one another from front to back but it appears that, instead, the primitive transverse shear still took place between the anterior surface of the transverse ridge of the lower tooth and the posterior surface of the transverse ridge of the upper. There is no indication that there was a for- wardly directed power stroke. However it will not be possible to discuss jaw move- ments in Exaeretodon with any degree of confidence until the wear facets on the abundant postcanines have been studied.

The lower postcanines of the southern African traversodontid Scalenodontotdes macrodontes Crompton & Ellenberger 1957 are almost identical in size and structure with those of Exaeretodon. Scalenodontoides was found in association with melanoro- saurid (prosauropod) dinosaurs and is therefore younger than the East African traversodontids.

I have been unable to study the type of Gomphodontosuchus brasiliensis (von Huene 1944-48) and the following remarks are based upon stereophotographs of the specimen taken by Dr J. Hopson. This genus has exaggerated some of the features of the postcanines of Exaeretodon. The teeth are set in the jaw more

qa.d.c.u.

e.C.u. a.a.c.u.

Fic. 12. Exaeretodon frenguelli. Typical postcanine teeth. A. Crown view of upper. B. Anterior view of upper. C. Posterior view of upper. D. Crown view of lower. For key to abbreviations see p. 69.

60 POSTCANINE OCCLUSION

obliquely so that the angle between the transverse ridge and the inner shearing surface of the external cusp is smaller than in Exaeretodon; the obliquity is reflected also in the outline of the crowns of the lower postcanines. In Gomphodontosuchus the anterior wall of the upper postcanine is higher than in Exaeretodon, the crown basin therefore deeper; this presumably indicates an increase in the cutting function of the anterior wall.

V. DISTRIBUTION OF THE TRAVERSODONTIDAE

Traversodontids are known to have existed for a long period of time and their dentitions indicate that they occupied several different ecological niches. At present they are known with certainty only from Africa and South America; indeed, this is true of all gomphodont cynodonts, and it has therefore been suggested that they were restricted to the southern continents. However, a large lower jaw of what may prove to be traversodontid cynodont was discovered in the Upper Triassic Wolfville Formation of the Newark Group in Nova Scotia by Dr R. L. Carroll and Dr D. Baird (Romer, 1967) ; unfortunately no postcanine teeth were preserved in situ, but the size of the jaw and the structure of the symphysis showed close similarity to Scalenodontoides macrodontes from southern Africa.

The postcanine teeth of traversodontids from several horizons and numerous localities in Africa are remarkably similar to those of traversodontids from various localities in South America. Scalenodon angustifrons closely resembles Masseto- gnathus pascuali; S. charigi closely resembles Exaeretodon frenguelli and Gompho- dontosuchus brasiliensis; and S. attvidger closely resembles Tvaversodon stahleckeri. The traversodontids of the Chanares Formation (Massetognathus) are comparable in size to the various species of the East African Scalenodon, but, while no East African form approaches the gigantic size of the traversodontids from Ischigualasto (Exaere- todon, Proexaeretodon and Ischignathus), Scalenodontoides from southern Africa indicates that large gomphodont cynodonts were present on that continent during Late Triassic times. Unfortunately the traversodontids collected in Brazil (Colbert, 1963) have not yet been described. No African site has been discovered which has | yielded traversodontids in anything like the abundance ot those of South America. |

The similarity of the terrestrial Early Triassic faunas from Argentina recently described by Bonaparte (1967) and of terrestrial Middle Triassic faunas from various parts of South America to African faunas of corresponding age, including the cynodonts, may indicate a close connection—or at least easy migration routes— between Africa and South America during Triassic times. Recent views (Bullard 1969, Menard 1969) on the relative positions of the continental masses of Africa and South America before the end of the Cretaceous suggest that they were extremely close, if not united, during the Trias.

Early cynodonts (Late Permian to Early Triassic) are known from South America (Bonaparte, 1967a), South Africa (Haughton & Brink, 1954), East Africa (Parrington, 1936), China (Young, 1961), and Russia (Tatarinov, 1968). A carnivorous cynodont of Early to Middle Triassic age probably occurs in China (Young, 1959). The descen- dants of the cynodonts, the tritylodontids, have been discovered in Late Triassic de-

CYNODONTS AND TRITYLODONTIDS 61

positsin South America (Sill, 1969), Africa (Fourie, 1968), China (Young, 1947), North America (Colbert, pers. comm.) and Europe (Kermack, 1965 and Kihne, 1956). Early mammals, also the descendants of cynodonts, have been discovered in the Late Triassic of southern Africa (Crompton, 1964), China (Rigney, 1963) and Europe (Kermack, 1965). In view of the world-wide distribution of these related groups it would not be expected that cynodonts should be totally absent from northern continents during Middle Triassic times. In particular, the gomphodont cynodonts may therefore have enjoyed a world-wide distribution; it may just be that their northern representatives (other than the jaw found by Carroll and Baird in Nova Scotia) have not been discovered as yet, and their apparent absence from northern continents may be due to the lack of suitable continental deposits of Middle Triassic age rather than to the absence of the animals themselves (Colbert, 1963).

The distribution of other Middle Triassic groups tends to support this view. For example, several Middle Triassic archosaurs are known from southern continents (Charig, 1967), but only a few, almost accidental finds are all that is known of the archosaurs of this age from northern continents (Krebs, 1965).

VI. OCCLUSION IN TRITYLODON AND THE ORIGIN OF THE TRITYLODONTIDAE

Several features of the skull and dentition of the traversodontids suggest that a member of this family may have been ancestral to the tritylodontids. This sugges- tion, made originally by Crompton & Ellenberger in 1957, is supported by the new material The crowns of the upper cheek teeth of Tvitylodon (Fig. 13J) consist essentially of three longitudinal rows of crescent-shaped cusps, three cusps each in the internal and central rows and two cusps in the externalrow. The corresponding lowers (Fig. 13K) consist of two rows each of three crescent-shaped cusps. The crescent of the upper cusps are concave forwards, whereas the crescents of the lower cusps are concave backwards (Fig. 13L). The two rows of lower cusps occluded between the three rows of upper cusps. Wear facets bearing parallel striations are developed on both surfaces of all the cusps except the outer surface of the external cusps of the uppers and the inner surface of the internal cusps of the uppers, which do not face any other surface; they indicate that during mastication the jaws moved horizontally. The extent of movement during occlusion is shown in Figs 13L and 14. The lower postcanine commenced dynamic occlusion by making contact with the upper tooth one position further forward in the upper jaw. The anterior edges of the

upper cusps and the posterior edges of the lower cusps formed a multiple cutting

mechanism which would have been effective only if the power stroke during mastica- tion were directed backwards. Occlusion therefore involved the dragging of the tips of the central row of cusps of the upper teeth and of the tips of both rows of cusps of the lowers across the floors of the valleys between the longitudinal rows of cusps of the opposing teeth. Asa result these crescentic cusps were rapidly worn down, there- by decreasing the shearing action of their near-vertical surfaces. In many specimens of tritylodontids the crowns of the postcanine teeth consist of almost flat surfaces scarred by longitudinal grooves. The backward jaw movements during occlusion, the rapid wearing down of teeth, the eruption of new teeth at the back to provide

POSTCANINE OCCLUSION

62

“IBMOT FO MOTA UMOID “ST —e jo Mor

‘goIsn[o00 orureudp jo soSvys [eu Jo MoTA yeroyey

‘roddn Jo MorA UMOID “4 “Y}00} OUTUROJSOg

~ we STEED ae x

aie I ‘zemoy Fo AOTA UMOID “ET

‘(uinesnyy Apoqeeg sex ul wourrseds)

‘ds snyjousojassppy “uorsnjo00 o1ureuAp jo sosejs [eu JO MOIA [eIoJeT “YJ “IOMOT JO MOIA UMOID ‘“q ‘oddn Jo MoIA UMOID “q “YI90} OUTUeOJSOg “AOU ‘ds mmosyasy uopouajvIS “UOISH[IIO o1uIeUAP Jo sosvjs [PUY JO MOIA [eIO}ET “DQ “IOMOT FO MITA UMOID ‘q -roddn fo MoIA UMOID “YW “Y}99} OUTURDJSOg ‘suodsfysnsuD UopouajvIg “EI “DIY

TWNYSALXS

a YOIYSLNV

CYNODONTS AND TRITYLODONTIDS 63

new shearing surfaces and the loss of worn teeth in front, as well as numerous skull and skeletal features, suggest a close relationship between the tritylodontids and traversodontids. In addition there are several other similarities between the denti- tions of individual traversodontids and of Tritylodon. These are most marked in Scalenodon hirschsom and the new species of Massetognathus described above. In Tritylodon and S. hirschsom the rows of postcanine teeth are parallel to the longi- tudinal axis of the skull and are not arranged obliquely along the edge of the maxilla as in S. attridget, S. charigi and earlier cynodonts. The incisors and lower canines of S. hirschsom are procumbent and this, taken together with the parallel postcanine rows, may indicate that there was increased posterior jaw movement during mastica- tion. In Massetognathus the postcanine rows are not parallel but there is neverthe- less a marked tendency for the rows to be directed away from the edge of the face towards the midline of the skull (Romer, 1967). The nature of the shearing planes between the external surface of the lower postcanines and the internal surface of the external cusps of the upper postcanines was essentially the same in Tvitylodon and the two traversodontid genera (cf. Fig. 13C, F, I & L). In both Tvitylodon and the traversodontids lower postcanines occluded with two upper postcanines because of the extensive backward movement of the lower jaw during occlusion. The main difference lies in the presence of the additional cusps of the postcanines in Tvitylodon. However, cusps that were not present in the more primitive traversodontid S. angustifrons (Fig. 13A) were added to the crown of S. irschsoni (13D) and Masseto- gnathus (13G).

In these latter animals cusps have been added to the uppers in front of the external and internal main cusps, i.e. in positions which suggest the initial steps in the forma- tion of the external and internal rows of cusps of the tritylodontid upper postcanines. It is significant that the largest cusps of tritylodontid upper postcanines are situated posteriorly and that the anterior cusps decrease progressively in size. The posterior accessory cusps on the external surface of the lower postcanines of traversodontids may also indicate the initial step in the greater development of the external row of cusps typical of tritylodontid lower postcanines. Although the postcanines of S. hurschsoni tended to resemble those of tritylodontids it is unlikely that the former reptile was ancestral to the latter; the power stroke was directed forwards in S. Jurschsoni, whereas it is the backwardly directed power stroke which appears to have characterized Tvitylodon occlusion. In Tritylodon the external and internal rows of cusps of the lower postcanines sheared between the external, central and internal cusps of the corresponding upper postcanines; therefore, if the ancestor of Tritylodon were to be found amongst the traversodontids, some indication of this occlusal pattern would be expected in one of the latter. In all the species of Scalenodon the transverse ridge of each lower tooth always occluded with the posterior surface of the transverse ridge of the preceding upper tooth; this shows that the lower jaw was not thrust far enough forwards before the power stroke for the transverse ridge of the lowers to be drawn backwards over the transverse ridge of the uppers. In the new species of Massetognathus however, the transverse ridges were drawn across one another during occlusion; this appears to be the only known traversodontid where this happened. Neither this species of Massetognathus nor S. hirschsoni appears to

64 POSTCANINE OCCLUSION

be directly ancestral to the tritylodontids, but, taken together, they show that some traversodontids were developing a postcanine dentition and mandibular movements which were very close to those of tritylodontids. Once two transverse ridges shearing past one another had been developed, it was a relatively simple matter to add more cusps in front of the uppers and behind the lowers, and thus to obtain postcanines of the tritylodontid type. In Fig. 13G & J an attempt has been made to homologise the cusps and regions of the crowns of the upper postcanines of the new species of Massetognathus and of Tritylodon. The posterior transverse row of cusps has been labelled A, B and C; the second row A!, B! and C! and the third row B” and C”. In Fig. 13H & K the same has been attempted for the lower teeth. It is clear that forms such as Exaeretodon, Gomphodontosuchus and S. charigi, which had lost the central cusp of the upper postcanines, could not have been ancestral to the tritylodontids; neither could forms such as S. attridgei with strongly developed anterior masticatory movements.

VII. SUMMARY AND DISCUSSION

The postcanine teeth of all the major groups of cynodonts have been briefly described and figured.

Three new species of traversodontid cynodonts, Scalenodon hirschsoni, S. attridget and S. charigi have been named and briefly described from their postcanine teeth.

An attempt has been made to trace the evolution of postcanine occlusion in advanced cynodonts. This is shown diagrammatically in Fig. 14. The term occlusion implies that there was tooth-to-tooth contact during the masticatory cycle; the teeth were constructed so that shearing, puncturing and crushing were possible between corresponding upper and lower teeth. Occlusal relationships

Fic. 14. Origin of the postcanines of Tyvitylodon. For each species the superimposed crown views of upper and lower postcanines are shown next to a posterior view of an occluding pair. Heavy lines indicate the orientation of the shearing planes. White circles indicate the principal cusps of the upper postcanines and black circles the principal cusps of the lower postcanines. Arrows indicate the extent of backward movement, from the beginning of tooth contact to its tightly closed completion. A. Thrinaxodon sp. Tooth-to-tooth contact did not occur. B. Tvivachodon sp. C. Diademodon sp. Lowers occluded directly with uppers and small shearing surfaces were produced by wear. D. Scalenodon angustifrons. Crowns of uppers and lowers were modified to reduce destruction of the tips of the cusps, and transverse elongated shearing surfaces were present. These surfaces came into use as the lower jaw was drawn upwards and backwards. G. Ewaeretodon sp. A modification of the S. angustifrons pattern. H. Scalenodon hirschsoni sp. nov. Occlusion involved both an anteriorly directed power stroke and a backwardly directed one. E. Massetognathus sp. The transverse shearing surface of the lower postcanines was drawn across the transverse shearing surface of the corresponding uppers (this did not happen in D, G & H above); the transverse shearing surfaces were modified in that small longitudinal shearing surfaces were formed by wear. F. Tyvitylodonsp. ‘The addition of extra cusps, in front of the upper teeth and behind the lowers, increased the length of the longitudinal shearing surfaces (which were comparable to those formed by wear in Massetognathus). There were no transverse shearing surfaces. The series from Diademodon to Tritylodon shows a progressive increase in the extent of the backward movement of the lower jaw during the power stroke.

—

CYNODONTS AND TRITYLODONTIDS 65

become progressively more complex within the cynodonts.

In the early cynodonts, the Galesauridae and Procynosuchidae, the lower post- canines bit internal to the uppers, tooth-to-tooth contact between opposing post- canine teeth was not possible and matching shearing planes are therefore absent (Fig. 14A). In both these families replacement of the postcanine teeth was alternate.

In the carnivorous cynodonts, the Cynognathidae and Chiniquodontidae, the lower postcanines still bit internal to the uppers, but matching shearing surfaces are occasionally found on the external surface of the lowers and internal surface of the uppers; this suggests that some form of shearing was possible, but the shearing

AX THRINAXODON sp

B TRIRACHODON sp Ay

C DIADEMODON sp

G EXAERETODON sp D SCALENODON ANGUSTIFRONS H scALeNopon HIRSCHSONI

a

66 POSTCANINE OCCLUSION

mechanism was primitive in comparison with mammalian carnassials and presum- ably played only a minor role in mastication. Clear-cut alternate replacement of the postcanine teeth was lost.

In the gomphodont cynodonts (Diademodontidae, Trirachodontidae and Traverso- dontidae) the lower postcanine teeth bit directly against the upper postcanines rather than internal to them. In the Trirachodontidae (Fig. 14B) the transverse ridges of the upper and lower postcanines alternated with one another but complex occlusal patterns were not developed. The crowns of the unworn postcanines of the Diademodontidae were characterized by one or two major cusps and an intricate pattern of smaller cusps and ridges; these were rapidly obliterated by wear so that small matching shearing planes were produced on the outer surface of the main cusp of the lower teeth and the inner surfaces of the main cusp of the uppers (Fig. 14C). However, the main cusp of the lower tooth, because it bit directly against the occlusal surface of the uppers (Fig. 14C), was rapidly worn down; the shearing surfaces can have been effective only for a short time. The continued possession of teeth with vertical shearing surfaces was nevertheless ensured by the addition of new gomphodont and sectorial teeth behind, worn teeth being lost from the front of the postcanine row. As would be expected, there was no alternate tooth replacement; occluding teeth were added sequentially at the end of the row during growth.

The occlusal patterns characterizing the Diademodontidae and closely related forms were refined in the Traversodontidae, where high ridges and deep basins provided effective shearing surfaces. Because the tips of the cusps occluded either opposite basins or externally to matching teeth, they were not worn down as rapidly as in the Diademodontidae. In primitive traversodontids the postcanines did not erupt with accurately matching upper and lower shearing surfaces, but, unlike the teeth of Diademodontidae, they needed relatively little wear of the crown surface to produce them; in Scalenodon angustifrons these shearing surfaces were aligned both transversely and longitudinally (Fig. 14D). In order that both these planes could be used effectively during the power stroke of mastication, jaw movements during this phase must have been both upwards and backwards. The transverse ridge which connected the two main cusps of the lower postcanine sheared past the trans- verse ridge formed by the central and internal cusps of the upper postcanine. The backwardly directed power stroke ended when the transverse ridge of the lower postcanine abutted against the anterior surface of the transverse ridge of the follow- ing upper. The arrow in Fig. 14D indicates the extent of this backward movement of a lower postcanine relative to the matching upper postcanine teeth during the power stroke of occlusion.

In a specimen belonging to a new species of Massetognathus (Fig. 14E) the basic Scalenodon type of postcanine was slightly modified. External cusps were added in front of the external main cusp of the upper postcanines and the longitudinal shearing plane was consequently increased in length. This was apparently coupled with an increase in the length of the backwardly directed component of the power stroke. The twomain cusps of the lowers commenced shearing in front of, rather than behind, the transverse ridge of the uppers. As the lower jaw was drawn backwards the lower cusps wore grooves between the internal and central cusps and between the

CYNODONTS AND TRITYLODONTIDS 67

central and external cusps of the uppers; these continued on to the anterior wall of the succeeding upper postcanine. Consequently in the new species of Masseto- gnathus what corresponded to the transverse shearing surface of Scalenodon angusti- frons was broken down into longitudinally orientated shearing surfaces on the sides of the main cusps. As in Diademodon, these grooves or shearing surfaces were produced by wearing away a substantial portion of the crown.

In Tritylodon (Fig. 14F) these modifications observed in the postcanines of the new species of Massetognathus were taken a stage further. The length of the backwardly directed component of the power stroke was increased and cusps were added in front of the upper teeth and behind the lowers. These additional cusps were smaller than the main cusps and they formed two additional transverse rows. Longitudinally orientated shearing planes similar to those which resulted from wear in the new species of Massetognathus were present on freshly erupted hardly worn teeth in Tnitylodon. In Oligokyphus an additional row of cusps was added, increasing the length of longitudinally orientated shearing surfaces still further. The entire trend is towards a lengthening of the backward component of the power stroke.

The large South American traversodontids, Exaeretodon (Fig. 14G), Proexaeretodon and Ischignathus, increased the length and height of the vertical shearing surfaces on the internal surfaces of the external main cusps of the upper teeth. The transverse shearing surfaces are obliquely orientated, the central cusp has been lost and a large external cusp added in front of the external main cusp of the uppers. For these reasons it is unlikely that these traversodontids could have been ancestral to the tritylodontids.

In Scalenodon liyschsom (Fig. 14H) and S. attvidge: wear facets indicate that both longitudinally and transversely orientated shearing planes were present, but, unlike other traversodontids, they could make both forwardly and backwardly directed power strokes during mastication.

The postcanine teeth of traversodontids and the tribosphenic molars of primitive mammals functioned in similar ways. Both have shearing surfaces on the vertical faces of the main cusps and both had jaw movements during the final stages of the masticatory cycle which were not directly orthal; in primitive therian mammals the power stroke had a marked transverse component (Crompton & Hiiemée 1969a & 0), while in traversodontids it had a strong posterior component. These movements in the horizontal plane permitted several shearing surfaces to be used as the jaws were closed.

In traversodontids but not in the tritylodontids the mandibular symphysis was massive and presumably immobile during life. The left and right lower postcanine teeth were slightly further apart than the corresponding upper postcanines; the opposite is true of primitive mammals. These two characters of traversodontids suggest that during the final stages of mastication both mandibular rami were drawn directly backwards and occlusion had to occur on both sides simultaneously; the greater the extent of the backward movement the more nearly parallel the rows of postcanine teeth. This is not possible in a primitive mammal (Crompton & Hiiemae, 1969a & 6). Significant transverse mandibular movements in cynodonts and tritylodontids were prevented by the massive transverse processes of the pterygoid

68 POSTCANINE OCCLUSION

bones and the structure of the postcanine teeth. The structure of the lower jaw of primitive mammals (Crompton 1963, Krebs 1969) suggests that remnants of the transverse processes of the pterygoids may have been present in some cases; it is possible that their reduction in early mammals was coupled with changes in jaw musculature permitting the introduction and strengthening of transverse mandibular movements.

The mechanism involved in developing occlusion in gomphodont cynodonts and in mammals (Crompton & Jenkins, 1968) appear to have been similar. In the early forms of both groups the crowns of corresponding upper and lower teeth were shaped by wear to produce matching shearing planes; in both a complex series of cusps and ridges had to be obliterated by wear before the teeth could function efficiently. This is particularly true of the cynodont Diademodon, of the new species of Masseto- gnathus and of the mammal Eozostrodon (=Morganucodon). In the later travers- odontids, tritylodontids and mammals the postcanines and molars lacked superfluous crown structures that had to be worn down and the crowns already possessed shearing planes that were genetically determined rather than produced by wear. The numerous Middle Triassic cynodonts from South America and the abundant early mammal teeth should be studied in detail in order to clarify further this aspect of evolutionary change.

The distribution of traversodontid cynodonts is briefly discussed; it suggests ready access between the continental masses of Africa and of South America during the Trias. The Traversodontidae may nevertheless have had a worldwide distribution.

ACKNOWLEDGEMENTS

I wish to thank Dr F. R. Parrington for the loan of the Tanzanian material in the University Museum of Zoology, Cambridge; Dr J. F. Bonaparte for presenting the Massetognathus jaw fragments described in this paper and for allowing me to study the magnificent collection of Triassic vertebrates at the Instituto Lillo in Tucuman, Argentina; Dr A. S. Romer for the opportunity to study the South American cynodonts in the collections of the Museum of Comparative Zoology, Harvard University; Dr A. J. Charig for his permission to study the cynodont material collected on the British Museum (Natural History)—University of London Joint Palaeontological Expedition to Northern Rhodesia & Tanganyika, 1963; and Drs A. J. Charig, K. Hiiemae, J. A. Hopson, J. Osborne and F. R. Parrington for reading the manuscript and for making many useful suggestions. The drawings were prepared by Mrs R. Rowen, the photographs taken by Mr A. Coleman and several drafts of the manuscript patiently typed by Miss M. Newton and Mrs I. Copeland.

This work has been supported by grants from the United States National Institutes of Health (RO1-DE-02648) and the National Science Foundation (GB 4435).

CYNODONTS AND TRITYLODONTIDS 69

ABBREVIATIONS USED IN TEXT-FIGURES

a.a.c.l. anterior accessory cusp of lower elle internal cusp of lower postcanine postcanine Gre internal cusp of upper postcanine a.a.c.u. anterior accessory cusp of upper p.a.c. posterior accessory cusp postcanine p.b. posterior basin af. anterior ridge p.c. posterior cingulum av. anterior valley ptr. posterior ridge a.w. anterior wall Pave posterior valley b.l. basin in lower postcanine p.w.f. postcanine wear facet b.u. basin in upper postcanine S. saddle CC. central cusp of upper postcanine Sh.s. shearing surface ie: embayment Unies transverse ridge e.cing. external cingulum tate transverse ridge of lower postcanine ec: external cusp of lower postcanine t.r.u. transverse ridge of upper postcanine Sci, external cusp of upper postcanine w.f. wear facet. en. enamel REFERENCES

AnpbErsSON, J. M. 1968. The confused state of classification within the family Procynosuchidae. Palaeont. afy., Johannesburg, 11 : 77-84, 1 fig.

ATTRIDGE, J., BALL, H. W., Cuaria, A. J., & Cox,C.B. 1964. The British Museum (N.H.) University of London Joint Palaeontological Expedition to Northern Rhodesia and Tangan- yika, 1963. Nature, London, 201 : 445-449.

Bonaparte, J. F. 1962. Descripcion del craneo y mandibula de Evxaeretodon frenguellit,

Cabrera y su comparacién con Diademodontidae, Tritylodontidae y los cinodontés

sudamericanos. Pubines. Mus. munic. Cienc. nat. tvadic., Mar del Plata, 1 : 135-402,

4 pls., 16 figs.

1963. La Familia Traversodontidae. Acta geol. lilloana, Tucuman, 4 : 163-194, Io figs. 1966. Una nueva “‘fauna’”’ tridsica de Argentina (Therapsida: Cynodontia Dicynodontia), Ameghiniana, Buenos Aires, 4 : 243-295, 2 pls., 29 figs.

1967a. New vertebrate evidence for a southern transatlantic connexion during the

Lower or Middle Triassic. Palaeontology, London, 10 : 554-563, 7 figs.

——1967b. Dos nuevas “‘faunas”’ de veptiles Tvidsicos de Argentina. 1 Simposio internacional sobre estratigvafia y Palaeontologia del Gondwana. Resumenes de Trabajos presentados a las sesiones. p. 7-8.

Brink, A. S. 1955. A study of the skeleton of Diademodon. Palaeont. afy., Johannesburg,

3 : 3-39, I0 figs.

1957. Speculations on some advanced mammalian characteristics in the higher mammal- like reptiles. Palaeont. afy., Johannesburg, 4 : 77-96, 5 figs.

—— 1963a. Two cynodonts from the Ntawere Formation in the Luangwa valley of Northern Rhodesia. Palaeont. afy., Johannesburg, 8 : 77-96, 4 figs.

— 1963b. Notes on new Diademodon specimens in the collection of the Bernard Price Insti- tute. Palaeont. afy., Johannesburg, 8 : 97-111, 7 figs.

Bro, F., & ScHRODER, J. 1935. Beobachtungen an Wirbeltieren der Karrooformation. IX. Uber den Schadel von Gomphognathus Seeley. S.B. bayer. Akad. Wiss., 1 : 115-182, 42 figs.

BULLARD, E. 1969. The origin of the oceans. Scient. Am., New York, 221 : 66-75, 6 figs.

Butter, P. M. 1961. Relationships between upper and lower molar patterns. Internat. Colloq. on the Evolution of Mammals. Kon. Vlaamse Acad. Wetensch. Lett. Sch. Kunsten Belgie, Brussels, pp. 117-126, 4 figs.

Cuaric, A. J. 1967. Subclass Archosauria. Jn Harianp, W. B. e al. (Eds.) The Fossil Record. Geological Society of London, 708-718, 725-731.

CoLpert, E. H. 1963. Relationships of the Triassic reptilian faunas of Brazil and South Africa, S. Afr. J. Sci., Cape Town, 59 : 248-253, 1 fig.

70 POSTCANINE OCCLUSION

Crompton, A. W. 1955. On some Triassic cynodonts from Tanganyika, Pyoc. zool. Soc. Lond., 125 : 617-669, 15 figs.

1963a. On the lower jaw of Diarthrognathus and the origin of the mammalian lower jaw.

Proc. zool. Soc. Lond., 140 : 697-750, pls. 1-2, 17 figs.

1963b. Tooth replacement in the cynodont Thrinaxodon liorhinus Seeley. Ann. S. Afr. Mus., Cape Town, 46 : 479-521, 17 figs.

1964. A preliminary description of a new mammal from the Upper Triassic of South Africa. Proc. zool. Soc. Lond., 142 : 441-452, 4 figs.

Crompton, A. W. & ELLENBERGER, F. 1957. On a new cynodont from the Molteno Beds and the origin of the tritylodontids. Ann. S. Afr. Mus., Cape Town, 44 : 1-14, pl. 1, 5 figs.

Crompton, A. W. & HilEMAE, K. 1969a. Functional occlusion in tribosphenic molars.

Nature, London, 222 : 678-679.

1969b. Molar occlusion and mandibular movements during occlusion in the American opossum, Didelphis marsupials. J. Linn. Soc. (Zool.), London, 49 : 21-47, pl. 1, to figs.

1969c. How mammalian teeth work. Discovery, New Haven, 5 : 23-34, 9 figs. Crompton, A. W. & JENKINS, F. A. Jr. 1967. American Jurassic symmetrodonts and Rhaetic ‘““Pantotheres’”’. Science, New York, 155 : 1006-1009, 3 figs.

1968. Molar occlusion in late Triassic mammals. Biol. Rev., Cambridge, 43 : 427-458, pls. 3, 11 figs.

Fourig, S. 1963. Tooth replacement in the gomphodont cynodont, Diademodon. S. Afr. J. Sci., Cape Town, 59 : 211-213, 1 fig.

1964. Cynodonts of the Lystvosauvus and Cynognathus zones. D.Sc. thesis. Uni- versity of Stellenbosch.

1968. The jaw articulation of Tvitylodontoides maximus. S. Afr. J. Sci., Cape Town, 64 : 255-265, 4 figs.

Haueuton, S. H. & Brink, A. S. 1954. A bibliographical list of Reptilia from the Karroo Beds of Africa. Palaeont. afy., Johannesburg, 2 : 1-187.

Hopson, J. A. 1966. The origin of the mammalian middle ear., Am. Zoologist, Utica, N.Y.,

6 : 437-450, I1 figs.

1969. The systematic position of the cynodont Sysphinctostoma snuthi and on a recon- sideration of postcanine succession in Diademodon (Reptilia, Therapsida). In press. Kemp, T. S. 1969. On the functional morphology of the gorgonopsid skull. Phil. Tvans.

R. Soc., London, 256B : 1-83, 25 figs, pl. 1.

KERMACK, K. A. 1965. The origin of mammals. Science Journal, London, 1 : 66-72, 5 figs. —— Lees, P. M. & Mussett, F. 1965. Aegialodon dawsom, a new trituberculosectorial

tooth from the lower Wealden. Pyvoc. R. Soc. B, London, 162 : 535-554, 6 figs.

Kitcuinc, J. W. 1968. On the Lystrosaurus zone and its fauna with special reference to

some immature Lystrosauridae. Palaeont. afy., Johannesburg, 11 : 61-76, 5 figs. Kress, B. 1965. Ticinosuchus feyox nov. gen. nov. sp. Ein Pseudosuchier aus der Trias des Monte San Giorgio. Schweiz. palaeont. Abh., Basel, 81 : 1-140, pls. 3, 68 figs. 1969. Nachweis eines rudimentaren Coronoids im Unterkiefer der Pantotheria (Mam- malia). Paldont. Z., Berlin, 43 : 57—63, pls. 4, 4 figs. Ktune, W. G. 1956. The Liassic thevapsid Oligokyphus. x xX 149, 12 pls. 66 figs. British Museum (Nat. Hist.), London.

LeumMan, J. P. 1961. Cynodontia, pp. 140-191, 37 figs, in Tvaité de Paleontologie. Tome VI. Ppp. 1-1135. Masson et Cie., Paris.

Menarp, H.W. 1969. The deep-ocean floor. Scient. Am. New York, 221 : 126-145, 11 figs.

MENDREZ, C. H. 1967. Sur quelques critéres de distinction entre Thérocéphales et Cyno- dontes, pp. 429-437, 7 figs in Problémes actuels de Paléontologie (Evolution des Vertébrés) Editions du Centre National de la Recherche Scientifique, Paris, pp. 1-474.

Mitts, J. R. E. 1964. The dentitions of Pevamus and Amphitherium. Proc. Linn. Soc. Lond., 175 : 117-133, pls. 2, 6 figs.

CYNODONTS AND TRITYLODONTIDS 71

Mitts, J. R. E. 1966. The functional occlusion of the teeth of Insectivora. J. Linn. Soc.

(Zool.), London 47 : 1-25, pls.

1967. A comparison of lateral jaw movements in some mammals from wear facets on

the teeth. Avrchs. oval Biol. London 12 : 645-661, 14 figs.

PaRRINGTON, F. R. 1936. On tooth replacement in theriodont reptiles. Phil. Tyans. R. Soc., London 226B : 121-142, 11 figs.

— 1946. On the cranial anatomy of cynodonts. Proc. zool. Soc. Lond. 116 : 181-197,

Io figs.

1955. On the cranial anatomy of some gorgonopsids and the synapsid middle ear. Proc. zool. Soc. Lond. 125 : 1-40, 14 figs.

Riegney, H. W. 1963. A specimen of Morganucodon from Yunnan. Nature, Lond. 197 : 1122, 1 fig.

Romer, A. S. 1967. The Chanares (Argentina) Triassic reptile fauna. III. Two new gomphodonts, Massetognathus pascuali and M. teruggii. Breviorva, Cambridge, Mass., 264 : 1-25, 10 figs.

SEELEY, H.G. 1895. Mesearches on the structure organization and classification of the fossil Reptilia. Part IX, section 3. On Diademodon. Phil. Tvans. R. Soc. B, London, 185 : 1029-1041, pl. 89.

Sit, W. D. 1969. The tetrapod-bearing continental Triassic sediments of South America. Am. J. Sci,, New Haven, 267 : 805-821, 3 figs,

StockLEy, G. M. 1932. The geology of the Ruhuhu coal fields, Tanganyika Territory. Q. Jl geol. Soc. Lond., 88 : 610-622, pls. 1.

Szatay, F. S. 1969. Mixodectidae, Microsyopidae and the Insectivore-Primate transition. Bull. Am. Mus. nat. Hist., New York, 140 : 195-330, pls 40, 28 figs

Tatarinov, L. P. 1968. Morphology and systematics of the northern Dvina cynodonts (Reptilia, Therapsida; Upper Permian). Postilla, New Haven, 126 : 1-51, 8 figs.

Watson, D. M. S. to11. The skull of Diademodon with notes on some of those of other cynodonts. Ann. Mag. nat. Hist., London, (8) 8 : 293-330, 9 figs.

Youne, C.-C. 1959. Note on the first cynodont from the Sinokannemeyeria faunas in Shansi, China. Vertebr. palasiat., Peking, 3 : 124-131, 4 figs., 3 pls.

—— 1961. On a new cynodont from N.W. Shansi. Verteby. palasiat., Peking, 5 : 109-114, pls. 1.

ZiEGLER, A. C. 1969. A theoretical determination of tooth succession in the therapsid Diademodon. J. Palaeont., Tulsa, 43 : 771-778, 2 figs.

A. W. Crompton, M.Sc., D.Sc., F.Z.S. Museum of Comparative Zoology HARVARD UNIVERSITY

CAMBRIDGE

Mass.

S.A.

PLATE

Thrinaxodon liorhinus

A. Oblique internal view of last three left lower postcanines. B. Oblique internal view of last five left upper postcanines.

PEATE 1

Bull. Br. Mus. nat. Hist. (Geol.) 21, 2

A

I——_ 0.5 cm——+ |

PLATE 2 A. Glochinodontoides gracilis (holotype, American Museum of- Natural History, no. 2223). Crown view of first four left upper postcanines.

B. Diademodon sp. (Bernard Price Institute, Johannesburg, no. 1675). Crown view of isolated postcanines; all except centre right are uppers.

Bull. Br. Mus. nat. Hist. (Geol.) 21, 2 PEALE 2

18.8)

4 ’

—

PLATE 3

Cricodon metabolus (holotype, Cambridge University Museum of Zoology, Ruhuhu Field Catalogue no. 74). A. Crown view of 7th and 8th right lower postcanines.

B. Crown view of 7th right upper postcanine.

Bull. By. Mus. nat. Hist. (Geol.) 21, 2 PEALE 3

I-00. 5cm~+>I

-

isc

TEA NTDID A

Scalenodon angustifrons (holotype, Cambridge University Museum of Zoology, Ruhuhu Field Catalogue no. 120B). A. Crown view of last two left lower postcanines.

B. Crown view of last three right upper postcanines.

Bull. By. Mus. nat. Hist. (Geol.) 21, 2 P LATE 4

J ==(0)5) fil

PLATE 5

Scalenodon hirschsoni sp. nov. (holotype, B.M.(N.H.) no. R. 8577). A. Crown view of 5th and 6th right lower postcanines. B. Crown view of 5th right upper postcanine.

Bull. Br. Mus. nat. Hist. (Geol.) 21, 2 PAGES 15

Ii=0.5cm—+ I

PLATE 6 Scalenodon attridgei sp. nov. (holotype, B.M.(N.H.) no. R.8578). - A. Entire palate. B. Crown view of right upper postcanines.

PLATE 6

Bull. By. Mus. nat. Hist. (Geol.) 21, 2

ek Ii=— 0.5 cm—+ I

PLATE 7

Massetognathus sp. (Yale Peabody Museum). Matching upper and lower postcanines. A. Crown views of 3rd, 4th and 5th left lower postcanines. B. Crown views of 3rd, 4th and 5th left upper postcanines.

Bull. Br. Mus. nat. Hist. (Geol.) 21, 2 PEALE 7,

i<— 0.5 cm—+ I

4

— >

A LIST OF SUPPLEMENTS TO THE GEOLOGICAL. ‘SERIES - OF THE BULLETIN OF

Pp. 213; 40. Plates; 2 Text figutes: ae pee ; . Et-Nacear, Z. R. Stratigraphy and Planktonic Poiuniniies “of Cretaceous—Lower Tertiary Succession in the Esna-Idfu Region, li Egypt, U.A.R. Pp. 201; 23 Plates; 18 Text-figures. 1066. (£10. ;

. Davey, R. J., Downie, C., SARGEANT, W. A. S. & Wirttams, G. L Mesozoic and Cainozoic Dinoflagellate Cysts oe 248; 28 figures,” 1960." £70.07 : . APPENDIX. Davey, R. J., DowniE, C., SaRbRane, W. A. s. &W. Appendix to Studies on Mesozoic and. Cainozoic Dinoflagellate C 1969. 8op. - Ere . Etziott, G. F. Permian to Palaeocene Calas Algae Daay isy¢ Middle East. Be I11; 24 Plates; 1g Keaton os Re 12:

Pp. 315; 31 Plates; 92 ievhgiea! ee a a ee . Cuitps, A. Upper Jurassic Rhynchonellid Brachiopods ae Salles re. = 12 Plates; 40 ee ge oe 5: es

pPle ae nga a rete E Shi ae: slay); Printed in England by Staples Printers Limited at their Kettering, Northants, establishment

rs

THE LOWER MIOCENE RUMINANTS OF GEBEL ZELTEN, LIBYA

BY

WILLIAM ROGER HAMILTON |

Pp. 73-150; 14 Plates, 13 Text-figures

BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) GEOLOGY Vol. 21 No. 3 LONDON: 1973

THE BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY), instituted im 1949, 15 issued in five series corresponding to the Departments of the Museum, and an Historical serves.

Parts will appear at irregular intervals as they become ready. Volumes will contain about three or four hundred pages, and will not necessarily be completed within one calendar year.

In 1965 a separate supplementary series of longer papers was instituted, numbered serially for each Department.

This paper is Vol. 21, No. 3 of the Geological (Palaeontological) series. The abbreviated titles of periodicals cited follow those of the World List of Scientific Periodicals.

World List abbreviation Bull. Br. Mus. nat. Hist. (Geol.).

© Trustees of the British Museum (Natural History), 1973

TRUSTEES OF THE BRITISH MUSEUM (NATURAL HISTORY)

Issued 14 June, 1973 Price £4.30

THE LOWER MIOCENE RUMINANTS OF GEBEL ZELITEN, LIBYA

By W. R. HAMILTON

CONTENTS Page I. INTRODUCTION : : 6 P 4 : : : c 76 II. SySTEMATIC DESCRIPTIONS : . é 6 : : é 79 Tragulidae : : : : 0 : : : : 79 Palaeomerycidae : : j ; : : : : 81 Giraffidae 4 : : . ‘ j : : : 85 Sivatheriidae . : : : : é : ; : 103 Bovidae . : : : 2 . P 125 Ill. THE EVOLUTION OF PRIMITIVE GIRAFFOIDS . : é 6 131 IV. THE EFFECT OF THE OSSICONES ON GIRAFFOID EVOLUTION . : 134 V. A CLASSIFICATION OF THE GIRAFFOIDEA . 3 é é 5 136 VI. East AFRICAN RUMINANTS : : : F ‘ : i 139 Tragulidae 5 : : : 5 c . : - 139 Gelocidae ‘ b : : : : 3 : : 140 Palaeomerycidae é 5 5 5 : ‘ 5 5 142 Bovidae . : : : : : : : : : 145 VII. REFERENCES . ; é : 3 : : : ; ; 148 SYNOPSIS

Ruminants from the Lower Miocene (Burdigalian) deposits of Gebel Zelten, Libya, are described. The skull and dentition of a new giraffoid—Zarafa zelteni— are described in detail. This genus exhibits dental characteristics which ally it to the palaeomerycids; however details of the cranial anatomy indicate a close relationship to the palaeotragines. Zarafa is classified as a palaeotragine and its palaeomerycid features are interpreted as evidence of a common ancestry for the two groups. The Palaeotraginae and Giraffinae are grouped in the Giraffidae.

The skull and dentition of Prohbytherium magnieri indicate that it is a member of the Sivatheriidae and primitive features of the skull suggest the divergence of the two families soon after the origin of the Giraffoidea in the late Oligocene.

The different evolutionary trends exhibited by the Giraffidae and the Sivatheriidae are related to the different fighting methods used in intra-specific combat. In the Giraffidae an elongation of the neck and limbs was possible while in the Sivatheriidae selective pressures existed tending to maintain the short neck and limbs.

76 LOWER MIOCENE RUMINANTS

The other Gebel Zelten ruminants are poorly represented but a new genus and species— Canthumeryx sivtensis—is described and identified asa palaeomerycid. A new tragulid species— Dorcatherium libiensis and three bovid species—Gazella sp., Protvagocerus sp. and Eotragus sp. are present in the fauna.

A further study of the ruminants from the Miocene of East Africa, enables the species Palaeomeryx africanus and Walangania gracilis to be synonymized as Walangania africanus and the upper dentition of Propalaeoryx nyanzae is described. A new species—Gelocus whitworthi— is described, this is the first recorded occurrence of the Gelocidae in Africa.

The influence of these new discoveries upon the accepted classification of the giraffoids is discussed.

I. INTRODUCTION

The fossil vertebrate fauna of Gebel Zelten, Libya was discovered by Arambourg (19g61a and b) who published preliminary notes on some new elements in the fauna. Further collections were made by R. J. G. Savage of Bristol University between 1964 and 1968. Many crania of Prolibytherium magnierit were collected and as some of these were of superior quality to the holotype, described by Arambourg (1961a); a full and detailed study of this species has been made. The presence of a second giraffoid—Zarafa zelteni—was not known until 1968 when M. White of Bristol University, completed the preparation of a skull which had been discovered, enclosed in a sandstone block. These two species form the basis of this work but as the study proceeded its scope was expanded to include the other ruminants of the area and finally a revision of some elements of the East African Miocene fauna was made.

Gebel Zelten

Gebel Zelten lies about 200 km south of the Gulf of Sirte, Libya. It consists of an elongate mesa running northwest-southeast for about 140 km at 19° 30’—20° 30’E. 28°-29°N. The ESSO Company oil camp of Zelten lies to the north of the gebel and the Oasis Oil Company of Libya camp lies to the south; the road between these camps crosses the gebel at its narrowest point where it is only 8 km. wide. In the west the gebel rises 40-60 m above the Zelten Rambla; it dips gently to the east and blends with the Calenscio Serir at its eastern end. The plateau is capped by marine sandstone which is Lower Miocene in age and the edges are dissected by steep walled wadis up to 3 km in length (Savage and White 1965). It is in these wadis that many of the vertebrate remains are found though some of the sites are on areas of washout from the wadis.

Detailed geological studies of the area have been published by Magnier (1962) and Selley (1968 and 1969); a detailed study of the geology with reference to the vertebrate sites is in preparation (Savage pers. comm.). The vertebrate remains are found in fluviatile deposits which probably originated in a coastal, alluvial flood plain (Selley 1969). The conditions at the time of deposition are interpreted as those of the savannah by Desio (1935) and again by Savage and White (1965).

Desio (1935) indicated that the Zelten deposits are Burdigalian and Helvetian in age and he stated that the deposits on the south side of Gebel Zelten are of Aquitanian

GEBEL ZELTEN, LIBYA 77

age. Arambourg and Magnier (1961) and Arambourg (1961a and b, 1963a and b.) have consistently placed the deposits in the Burdigalian, and Arambourg (1963b) states that the Gebel Zelten deposits rest on Oligocene marine beds and are overlain by Helvetian marine beds. Savage and White (1965) indicate a Burdigalian age and this was later refined (Savage in Selley 1969) to Early Burdigalian or Late Aquitanian.

Terminology

In most anatomical details the terminology used follows that of Sisson and Grossman (1953). The nomenclature applied to the dentition (Text fig. 1) 1s mainly after Arambourg (1947). In the upper molars the postero-lingual cusp, termed the hypocone by Arambourg (1947) is here termed the metaconule after Weber (1928). In the lower molars I have treated the antero-lingual corner as if the paraconid has been entirely lost, thus the antero-lingual cuspid is the metaconid with an anterior mesostylid. The ‘Palaeomeryx fold’ is used to define any fold of enamel which runs vertically down the postero-labial face of the protoconid into the median valley.

a b ¢c d

Fic. 1. Ruminant molar and premolar cusp nomenclature. (A) Third upper premolar. (B) First upper molar. (c) Second, third and fourth lower premolars. (Dp) Second and third lower molars. Upper dentition. a: parastyle. b: paracone. c: metacone. d: metastyle. e: mesostyle. f: protocone. g: accessory crest. h: metaconule. i: ento- style. Lowery dentition. j: mesostylid. k: metaconid. 1: metastylid. m: entoconid. n: entostylid. o: hypoconid. p: ectostylid. q: protoconid. r: parastylid. s: para- conid. +t: hypoconulid. u: entoconulid. v: anterior fossette. w: posterior fossette.

78 LOWER MIOCENE RUMINANTS

The abbreviations used in reference to the dentition follow the convention in general use in Britain and America. I, C, P and M represent incisor, canine, pre- molar and molar respectively; the tooth is then defined by a number added above or below the line to indicate presence in the upper or lower jaw; thus Pz, is the lower, fourth premolar which is in contact with My, the first lower molar. The deciduous cheek teeth are referred to as D1, D; etc. The external side of the tooth is labial and the internal side is lingual. Anterior and posterior with reference to the dentition, indicate those directions which apply if the mandible or maxilla is in the horizontal position.

Frontal Appendages

In current usage the term ‘horn’ can refer to any cranial appendage but in this work a restriction of the term is applied. In the ruminants the type of frontal appendage is some times the main criterion on which the classification of a genus is based as stated by Pilgrim (1941) :

‘“, . . the varying types of frontal appendage—horns—constitute one of the most important distinctions between the different families of the Pecora.’ In this situation it is clearly desirable that separate terms be applied to the different types of frontal appendage. Voorhies (1969) listed four types of frontal appendages in the living artiodactyls as: ‘I. the unshed true horns growing on the bony core in the Bovidae, 2. the deciduous antlers of the Cervidae, 3. the annually-shed horny sheath growing over a permanent, vascular bony core in the Antilocaprinae, and 4. the bony core permanently covered by skin (‘velvet’) in the Giraffidae.’

The first three of these appendages are referred to as; ‘horns’, ‘antlers’ and ‘horn- cores’ respectively but no widely accepted term exists for the giraffid appendage. Lankester (1907) used the term ‘ossicone’ to refer to this appendage and defined ossicones as:

“. . . independently ossifying bony cores which are found in Okapia and Giraffa on the frontal and parietal areas and in the giraffe also in the median position.’ A slight widening of the application of this term was made by Ginsburg and Heintz (1966) who applied the term to the palaeomerycid appendage and its application to all giraffoid cranial appendages is desirable.

Abbreviations

The prefix ‘M’ refers to specimens in the collections of the British Museum of Natural History, London: ‘B.U.’ in the Department of Geology, University of Bristol and ‘P’ in the Institut de Paléontologie, Paris. Specimens described in chapter 4 are the property of the National Museum, Nairobi, Kenya and are defined by the prefix ‘K’.

Classification

The system of classification used in this work differs slightly from that of Simpson (1945), this is mainly due to alterations within the Giraffoidea and primitive Cervoidea. The Palaeomerycidae is treated as a family of the Giraffoidea and follow-

GEBEL ZELTEN, LIBYA 79

ing the suggestion of Ginsburg and Heintz (1966), those genera of the family Palaeomerycidae which lack frontal appendages, are removed to the separate family Dremotheriidae. The family Blastomerycidae is grouped with the Dremotheriidae in the Dremotherioidea. The position of the Dromomerycidae is uncertain and requires further study, but in this work the family is treated as a group probably originating in the nearctic region and having no direct relationship to the Palaeo- merycidae.

The family Giraffidae has been split by the establishment of the Sivatheriidae as a separate family.

ACKNOWLEDGEMENTS

I would like to express my deepest thanks to all those who have helped in making this work possible. Dr. R. J. G. Savage suggested the topic and by his encourage- ment, advice and friendship, he has contributed much towards its completion. The technical staff of the Department of Geology, University of Bristol, have all helped and I would particularly like to thank Mr. M. White for his work in preparation of the material and Mr. R. Godwin who prepared the plates.

I would like to thank the staffs of the museums that I have visited. Professor M. Crusafont Pairo of Sabadelle, Spain and Professor Lehman of |’Institut de Paléontologie, Paris; have allowed me access to the collections and provided facilities forstudy. Dr. A. Sutcliffe of the British Museum of Natural History, allowed me to use the collections and facilities and the Keepers of palaeontology and zoology extended facilities on which I borrowed material.

Dr. L. S. B. Leakey allowed me to study and redescribe the ruminant material from Kenya and Dr. A. W. Gentry gave advice and comment on the intricacies of bovid classification. Dr. Churcher provided up to date information on his study of the East African giraffids which helped to avoid any overlap in our studies.

This study was carried out under a N.E.R.C. Research Studentship and the Univer- sity of Bristol provided me with facilities during the tenure of this studentship from 1967 to 1970.

Il. SYSTEMATIC DESCRIPTIONS Family TRAGULIDAE Milne-Edwards 1864

DiaGnosis: Small primitive ruminants; lacking frontal appendages. Dentition primitive but with upper incisors reduced or absent. Upper canines large, especially inthe male. Molars bunodont but showing selenodonty in advanced forms. Limbs showing features of advanced ruminants but with varying degrees of fusion exhibited by the metacarpals and metatarsals. Navicular and cuboid fused. (After Milne- Edwards 1864).

Genus DORCATHERIUM Kaup 1833

Diacnosis: This genus was defined by Whitworth (1958 p. 3) whose diagnosis is followed here.

TYPE SPECIES: Dorcatherium naw Kaup 1833.

80 LOWER MIOCENE RUMINANTS

Dorcatherium libiensis sp. nov.

Diacnosis: A small species of Dorcatherium. Length of lower tooth row P3—M3 about 39 mm.

REMARKS: This species is established on a size basis and further work in the Gebel Zelten area may reveal wide variation in the species; however as the species is very rare at Gebel Zelten and as collecting in the area has been terminated it was decided, with some hesitancy, to establish the species on the basis of a single specimen.

HoLotypPe: M.26684. A fragmentary right mandible with Ps to Me and the alveoli of P; and P2 preserved.

Locatity: The Lower Miocene (Burdigalian) deposits of Gebel Zelten, Libya.

Lower Dentition. The molars are heavily worn but the main details of their anatomy are visible (pl. 1, fig. 1). Mi is complete though badly cracked posterior to the median valley and the postero-lingual corner of Mz is missing. The molars are bunodont as in D. chappuist and D. nawi and strong anterior and posterior cingula were present on M; and Mz. The posterior face of the protoconid bears a strong fold similar to a ‘Palaeomeryx fold’ but more lingually situated and similar to that of D. chappuist. The postero-lingual region of the metaconid carries a deep vertical groove on both molars, as in D. naw and D. chappuisi this fold seems to be characteristic of the genus Dorcatherium. A small ectostylid is present in the median valley as in D. naut, this stylid is usually absent in D. chappuisz.

The anterior end of P3 is missing but P4is complete. The premolars are elongate and similar to those of D. chappuisi and D. naw. The dentition of D. libiensis is thus similar to that of D. chappuwisi and is distinguished from this species and D. pigottt on a size basis only.

The presence of this mandible in the Gebel Zelten fauna serves to establish the presence of the genus in North Africa at this time but gives no indication of the relative abundance of tragulids at Gebel Zelten, as specimens of this size and smaller, are relatively rare in the collection (Savage and White 1965).

TABLE I D. hbiensis D. chappuisi D. pigott M.26684 (Whitworth) (Whitworth) Length Width Length Width Length Width P3 10:0 mm 36mm 13:55mm 5‘I mm 8-4 mm 3:1 mm Pa 96mm 40mm 12:3 mm 6-0 mm 7-9 mm 3°38 mm Mi 9-3 mm 58mm 11:7mm 7:2 mm 8-2 mm 4°38 mm Moe IIl-o mm 66mm 12:9 mm 9:0 mm 8-9 mm 5°3 mm

Superfamily GIRAFFOIDEA Simpson 1931

DiaGnosis: Medium to large sized ruminants. Cheek teeth brachyodont or occasionally hypso-brachyodont. Enamel of cheek teeth usually rugose. Ossicones present in male and sometimes female. Metapodials fully fused.

GEBEL ZELTEN, LIBYA 81

ComMENTS: Thisis a well defined group of ruminants which is probably related more closely to the Cervoidea than the Bovoidea and is in many respects more primitive than either group.

Family PALAEOMERYCIDAE Lydekker 1883

Diacnosis: A group of primitive giraffoids exhibiting features that may indicate a close relationship to the cervoids. Ossicones are present and in some species these have a long proximal region and a whorl of short tines distally. The mandible is shallow with brachyodont selenodont molars which have lightly rugose enamel. P may be present. Metastylid and entostylid usually very prominent and strong cingula are present anteriorly and posteriorly. A ‘Palaeomeryx fold’ is often present. Labial ribs of the upper molars prominent. Accessory crests often present in the fossettes and entostyle usually present in the median valley.

Genus CANTHUMERYYX nov.

Diacnosis: A medium sized ruminant, about as large as the fallow deer, Dama dama. Lower dentition similar to Propalaeoryx but lacking Py. Lower molars relatively high with weaker mesostylid and a stronger more flexed metastylid than is usual in the palaeomerycids. Strong anterior and weak posterior cingula on M, and Mz. Premolars elongate similar to usual palaeomerycid pattern but strong entostylid on P4.

TYPE SPECIES: Canthumeryx sirtensis sp. nov.

Diacnosis: As for genus.

DERIVATION OF NAME: The generic name is from Canthus a character of Greek mythology: according to Appollonius: ‘Fate had decreed that he and the great seer Mopsus should wander to the ends of Libya to be destroyed.’ (Rieu 19509). The trivial name refers to the Sirte basin in which the specimens were discovered. Hototyre: A mandibular fragment with D3 to Ms, the last molar being only partially erupted. Ms, P4 and Pe have been dissected out. Locatity: The Lower Miocene (Burdigalian) of Gebel Zelten, Libya. MATERIAL: M.26682 Holotype. M.26683 A right mandibular fragment with D4 and M; erupted. P3 and P4 dissected out. B.U.zo111 An isolated lower right M3 showing moderate wear. Lower Dentition. The lower dentition forms a closed series from P2 to M3 and P;

is absent as in Palaeomeryx but in contrast to Propalaeoryx. The enamel of the molars is finely rugose and the molars are higher and more elongate than in Palaeo-

82 LOWER MIOCENE RUMINANTS

meryx but lower and slightly wider than in Propalaeoryx. The metaconid of My, is transversely flattened with a weak mesostylid, a strong lingual mb and a strong metastylid which is flexed more lingually than in Palaeomeryx and is longer and more slender than in Propalaeoryx. The entoconid has a strong lingual rib, the cuspid is almost parallel to the axis of the molar as in Propalaeoryx whereas in Palaeomeryx it has a more diagonal orientation. The posterior end of the entoconid is reduced and rounded in Mj, (pl. 1, fig. 2). The hypoconid is isolated until very late in wear and the anterior fossette would be entirely worn away before the hypoconid wear trace joined that of the protoconid. The hypoconid is much lower than the protoconid, but due to the increased overall height of the tooth, this difference is less marked than in Palaeomeryx. The posterior end of the hypoconid is produced lingually forming the posterior end of the tooth and causing the posterior fossette to open lingually in M, (pl. 1, fig. 3).

Mg is very similar to My but the posterior end of the entoconid bears a strong keel, causing it to extend further posteriorly and tending to close the posterior fossette which opens at the postero-lingual corner of the tooth; this contrasts with Pro- palaeoryx in which the entoconid of Me is rounded posteriorly and the posterior fossette opens lingually asin My. In Palaeomeryx the posterior end of the entoconid usually bears a strong crest in M; and M2.

Mz has a feeble mesostylid and a strong metastylid. The entoconid is transversely flattened and its anterior region consists of a strong crest of enamel which meets the anterior face of the protoconid. The posterior region of the entoconid consists of a strong crest and from its posterior end the long mentoconulid curves posterolabially, to blend into the middle of the lingual face of the hypoconulid. The protoconid joins the posterior end of the metaconid closing the fossette even in the unerupted condition ; this region is very variable in Palaeomeryx but in Propalaeoryx it is similar to Canthumeryx. The posterior end of the hypoconid is short, it meets the hypoconulid but fails to reach the entoconulid. The crescentic hypoconulid is lower than the hypoconid. The molars each have a strong ectostylid in the median valley (pl. 1, fig. 4) and Mg has a small stylid in the posterior valley. A strong anterior cingulum is present on each molar and feeble posterior cingula are present on M, and Mo.

The Pg of Canthumeryx is generally similar to that of Palaeomeryx or Propalaeoryx. The metaconid is the highest part of the tooth, it is strongly swollen anteriorly (pl. 1, fig. 5) and produces a strong wing posteriorly, these features are similar in Palaeomeryx but in Propalaeoryx the anterior and posterior projections are absent. The pro- toconid is joined to the metaconid by a strong ridge of enamel as in Palaeomeryx. The entostylid is very strong in Canthumeryx (pl. I, fig. 5), curving across the whole posterior face of the tooth as in Propalaeoryx and in contrast to Palaeomeryx in which it is reduced lingually. The entoconid of Palaeomeryx usually curves postero- lingually at its lingual end but there is no indication of such a curvature in Canthu- meryx or Propalaeoryx. The paraconid and parastylid are very strong and widely divided in Canthumeryx (pl. 1, fig. 5) and Propalaeoryx whereas in Palaeomeryx they are usually weaker and less widely divided. The labial face of the P, is swollen labially in Palaecomeryx but in Canthumeryx and Propalaeoryx this swelling is absent

GEBEL ZELTEN, LIBYA 83

and the wall is vertical. A strong groove on the labial wall separates the hypoconid from the protoconid, this groove is very strong in Palaeomeryx weaker in Canthumeryx and weaker again in Propalaeoryx.

Ps is more elongate than Py; the metaconid is lower than in Py and it has no posterior fold or anterior swelling, it is also situated further posteriorly than on the Py (pl. 1, fig. 5). The posterior region consists of a hypoconid and an entoconid as in the P, but the entostylid is joined to the posterior face of the entoconid from which it curves postero-lingually and does not join the hypoconid (pl. 1, fig. 5); this is probably an individual variation which is also found in Palaeomeryx and Walangania and in these genera it is more usual for the posterior region of the P3 to resemble that of the P,.

The Pz of Canthumeryx is relatively simple with a single, conical, primary cuspid in the central region from which an unforked anterior crest is produced; this curves lingually at its anterior end. A posterior swelling and a postero-lingual crest are produced from the primary cuspid. The posterior region is much lower than the primary cuspid, it consists of a single transverse crest which falls away posteriorly. Strong anterior and posterior cingula are present on all the premolars.

The Dy, is heavily worn in both specimens. It is elongate and trilobed; its anterior end is narrower than the posterior end and the anterior, median and posterior fossettes are joined (pl. 1, fig. 2). The anterior fossette is bounded by a strong labial cuspid and a narrow lingual cuspid, the anterior end is closed by a small stylid. In both specimens the anterior region is very heavily worn and very little surface detail is visible. The metaconid is high and selenodont as in the molars (pl. 1, fig. 2), it has a strong metastylid which communicates with the high entoconid.

The posterior region of the entoconid is shortened as in the Mj and the posterior fossette opens lingually (pl. 1, fig. 2). The protoconid is stout and crescentic, in the heavily worn condition its wear trace is joined to the antero-labial cuspid. The hypoconid is higher than the protoconid but this may be a wear factor. The Dghasa very strong ectostylid in the median valley. The posterior cingulum is strong and the anterior cingulum continues along the labial face as far as the anterior valley, in which there is a weak cingulum.

The D3 is elongate narrowing anteriorly. The primary cuspid lies in the middle of the tooth and has a feeble lingual cuspid. The anterior region is similar to that of Ps with a long crest giving rise to antero-lingual and lingual branches. The posterior region has a single central hypoconid which is joined by a crest to the primary cuspid. The hypoconid produces a posterior branch which curves lingually at its posterior end and a lingual branch which curves posteriorly, a large enamel island is produced between these branches.

The Dg is known from the alveoli only (pl. 1, fig. 4); these are single, anterior and posterior and indicate that D2 was slightly less elongate than the Ds. The absence of a D, indicates that P; was also absent.

The dentition of Canthumeryx resembles that of Propalaeoryx more closely than any other ruminant, and both resemble Palaeomeryx. Differences of the metastylid, entoconid, height and width serve to distinguish the molars of Canthumeryx from those of Palaeomeryx and Propalaeoryx and many details of the fourth premolars

84 LOWER MIOCENE RUMINANTS

present differences between these genera. The absence of a Pj in Canthumeryx is an important difference distinguishing it from Propalaeoryx.

TABLE 2 The Lower Dentition of Canthumeryx M.26682 M.26683 B.U.20111 Length Width Length Width Length Width D3 16-3 mm 5°7 mm -— — — — D4 22:2 mm — 230mm =I1'I mm — — P2 13-6 mm 5°3 mm = — — — P3 — — 18-7 mm 8-6 mm — — P4 15-8 mm 75mm 19-0 mm 9:8 mm — = Mi 195mm I10mm 201mm 14:0mm —_— — Me 203mm 13°6mm — — — — M3 292mm 13:3 mm -— — 312mm 13'7mm

Palaeomeryx sp.

The presence of a species of the genus Palaeomeryx in the Gebel Zelten fauna is indicated by two fragmentary molars; M.26691 and B.U.20112. The specimens are both third molars of which the former is the more complete.

The metaconid is broken off but its postero-labial region indicates that it was joined to the protoconid and entoconid. The entoconid is transversely compressed and its posterior region is shortened, thus the posterior fossette opens lingually in contrast to Canthumeryx. The protoconidis crescentic and very stout, a “‘Palaeomeryx fold’ may have been present but the posterior face of the protoconid is very heavily worn. The anterior end of the hypoconid is produced anteriorly and its posterior end which joins the entostylid is much longer than in Canthumeryx. The hypo- conulid is stout and curves around the posterior end of the tooth, joining the entostylid and enclosing a posterior enamel island in contrast to Canthumeryx in which the posterior enamel island would not be formed in this position. A strong ectostylid is present in the median valley and a feeble stylid is present in the posterior valley of BaUi2onr2:

These specimens are distinguished from Canthumeryx by details of their dental anatomy and also by their smaller size; they resemble Prolibytherium closely in size but are more brachyodont and differ in anatomical details from this genus.

_ TABLE 3

The Lower Molars of Palaeomeryx sp M.26691 B.U.2z0112 K.R.442.51

Mg

Length 24°2 mm — 28-I mm Width of anterior lobe I1-5 mm — 12°5 mm Width of posterior lobe I1-o mm II-o mm 1370 mm

Width of accessory lobe 6-6 mm 6-7 mm 6-7 mm

GEBEL ZELTEN, LIBYA 85

Palaeomerycidae Indet.

A single pair of ossicones M.266g90 (pl. 1, fig. 6), cannot be definitely assigned to any group of the Giraffoidea. These ossicones diverge at an angle of 40° and slope posteriorly at an angle of about 50° from the vertical. The bone surface has many fine vertical striations which fade out well above the base and it is unlikely that a horny sheath could have been present. The ossicones were very centrally positioned on the cranium and the region between them curves smoothly with no sign of a median suture. The small area of cranium that is preserved, indicates that the animal was slightly larger than Prolibytherium.

The cervid genus Dicrocerus has long pedicles which are comparable in form with this specimen but in Dicrocerus the pedicles were supra-orbitally situated and were less divergent. The ossicones of Climacoceras diverged at an angle of about 60° (MacInnes 1936), their internal structure is similar to ordinary bone with a core of vesicular structure; this agrees with M.26690. MaclInnes (1936) also states that the shaft of Climacoceras was nearly straight throughout its length. The ossicones of Climacoceras may represent a condition derived from ossicones similar to M.26690.

Family GIRAFFIDAE Gray 1821

Diacnosis: Giraffoids in which the neck and limbs are usually lengthened. Ossicones small, consisting of a single tine. Degree of facial flexion small. Cheek teeth brachyodont; upper molars with strong mesostyle. Paracone and metacone having a diagonal orientation on the molar. - Lower premolars exhibiting molariza- tion. On the fourth lower premolar, the metaconid is strong and the hypoconid and entoconid are separated from the protoconid. ‘Palaeomeryx fold’ usually absent.

ComMENTs: This group previously included the Sivatheriidae which is here treated as a separate family of the Giraffoidea.

Subfamily PALAEOTRAGINAE Pilgrim rort

D1aGnosis: Primitive, medium sized giraffids, usually with one pair of supra- orbital, frontal ossicones. A second pair of ossicones may be present on the anterior extremities of the frontals. Skull usually elongate. Cheek teeth brachyodont. Limbs and neck slightly elongate. (After Colbert 1935a.)

Genus ZARAFA nov.

Dracnosis: A very primitive palaeotragine with flattened, laterally expanded frontals and frontal sinuses in the supraorbital region. Supraorbital ossicones present. Paired lacrymal foramina present on the anterior edge of the orbit. Basicranial and basipalatal planes almost parallel. Maxilla very shallow. Cheek teeth primitive and very brachyodont. A strong accessory crest present on the posterior region of the metaconule.

86 LOWER MIOCENE RUMINANTS

Zarafa zelteni sp. nov.

DiaGnosis: As for genus.

DERIVATION OF NAME: The generic name is from the Arabic for ‘giraffe’. The trivial name is from Gebel Zelten, the area from which the type specimen was collected.

Ho.otyre: An almost complete but edentulous skull (M.26670). The premaxilla and anterior region of the maxilla are missing and only the proximal region of the nasals is preserved. The lingual wall of the third molar is the only dental fragment preserved.

LocaLity: The material is all collected from the Lower Miocene (Burdigalian) deposits of Gebel Zelten, Libya.

MATERIAL: M.26670 Holotype. An almost complete skull of an adult individual. M.26671 A right maxillary fragment with P4 to M%. The dentition exhibits medium wear. M.26672 A left maxillary fragment with D2 to D4. The first two permanent molars are dissected out. M.26673 A cranial fragment consisting of the supra-occipital and parietal region. M.26674 A cranial fragment with the frontal-parietal suture and the anterior part of the frontals preserved. M.26675 A fragment of right mandible with Mg showing light wear. The ascending ramus and condyle are preserved, though badly shattered. M.26676 A heavily worn M3. M.26677_ A lightly worn Mg.

Skull. The skull was found enclosed in a large sandstone nodule from which it was removed by the standard acetic acid preparation method. The postorbital region is in an excellent state of preservation but much of the preorbital region is missing. Slight crushing has occurred in the preorbital region. The specimen is from a mature individual but the acid preparation has opened and defined the sutures; these are not visible in areas prepared by hand.

Maxilla. It is likely that the maxilla was very shallow and probably resembled that of the juvenile giraffe. The postero-lateral region of the maxilla is missing but its shape is indicated as an internal cast (pl. 2). The facial tuberosity lies above M2. The palatine process of the maxilla is badly broken and interpretation of the surface features is difficult. The surface is shallowly convex, its posterior edge extends to the maxillary tuberosity and into the orbit where it contributes to the lacrymal bulla. The anterior palatine foramen lies on the palatine-maxillary suture opposite theanterior end ofthe M2; more posteriorly than in Okapiabut similarto Givaffa. The anterior part of the bone is missing from the level of the anterior edge of P?. The maxilla extends behind the third molar, forming a large maxillary tuberosity which is badly broken in the specimen.

Nasal. A small part of the nasal is preserved on the antero-dorsal edge of the

GEBEL ZELTEN, LIBYA 87

prelacrymal vacuity. The bone is very flattened in the same plane as the frontal, as in Palaeotragus microdon.

Lacrymal. The large lacrymal forms the posterior edge of the prelacrymal vacuity and extends into the orbit. The antero-lateral face of the bone is concave in contrast to Okapia in which it is plane or Givaffa in which it is convex. The concavityin Zarafa results from the lateral expansion of the frontal bones. Thelacry- mal is expanded dorsally between the orbit and the prelacrymal vacuity. This dorsal expansion is more marked than in the other giraffids and may be due to the expansion of the frontals but a large lacrymal is also found in Dremotherium. A small lacrymal tubercle stands on the antero-dorsal edge of the orbit as in Okapia and paired lacrymal foramina are present behind the edge of the orbit. The maxillary foramen lies posterior to the lacrymal tubercle in the same position as in Okapia. The lacrymal bulla is badly broken posteriorly.

Jugal. The jugal forms the ventral and postero-ventral edges of the orbit and contributes about half of the postorbital bar; as in Okapia a strong ridge forms the ventral edge of the orbit. The lateral face of the jugal is concave and the ventro- lateral region of the bone is badly eroded. The facial region is large and elongate resembling the cervids rather than Okapza.

Palatine. Two parallel depressions run antero-posteriorly along the palatine. The bone is badly eroded posteriorly but it is preserved as a vertical plate in the pterygo-palatine fossa which is shallower but otherwise similar to that of Okapia.

Frontal. The orbital region of the frontal is concave with the orbital opening of the supraorbital canal lying in the most dorsal part. The ethmoid foramen lies ventro-medial to the supraorbital canal. The frontal-parietal suture runs dorsally from the alisphenoid to the top of the skull and medially across the dorsal face to the median suture (pl. 3). The temporal region of the frontal is very small its concave ventro-lateral face resulting from the great expansion of the posterior supraorbital crest. A strong postorbital ridge runs transversely between the supraorbital process of the frontal and the squamosal bone. The frontal forms the dorsal edge of the preorbital vacuity and dorso-lateral to this the bone thickens forming a high supraorbital crest. As in Okapia the supraorbital foramen is directly above the orbital opening of the supraorbital canal; a shallow concavity lies anterior to the fora- men but there is no supraorbital groove in Zarafa. Lateral to the foramen the bone rises sharply to a peak and a process has been broken off revealing an extensive supraorbital frontal sinus which indicates the presence of a supraorbital ossicone. Ossicones are found in a similar position in Palaeotragus microdon and Samotherium sinense (Bohlin 1926) and this is probably the primitive position of ossicones in the Giraffidae. Postero-medial to this region is a strong lateral ridge which continues on the parietal.

Parietal. The temporal face of the parietal is dorso-laterally inclined with a convex anterior region and a concave posterior region. The parietal crest crosses the dorsal part of the temporal region and probably continued as far as the edge of the nuchal crest but the posterior region is missing. The dorsal face of the parietal is shallowly concave with raised lateral and medial ridges.

Occipital. The mastoid foramen lies on the postero-lateral face of the supra-

88 LOWER MIOCENE RUMINANTS

occipital region and the occipital forms its ventro-medial wall (pl. 4, fig. 1). The shape of the nuchal crest was probably similar to that of Okapia but the supra- occipital region is broken dorsally (pl. 4, fig. 1) and most of the nuchal crest is missing. The occipital condyles and foramen magnum are large in Zavafa and dorso-medial to the edge of the condyle the bone surface is concave but it is produced as a large swelling over the foramen magnum, a similar swelling is present in Palaeotragus microdon but is less pronounced in Okapia and absent in Givaffa and Prolibytherium. Dorsal to the swelling the bone is excavated as insertion for the semispinalis capitis muscle; these excavations are shallower than in Okapia. A weak median occipital crest stands between the excavations but the external occipital protuberance is not preserved.

The paroccipital process projects ventrally to the level of the ventral edge of the condyles (pl. 4, fig. 1) asin Okapia. The basioccipital region has a pair of anterior occipital swellings between which the bone surface is concave, a median keel begins at the anterior end of this concavity and continues on the basisphenoid (pl. 4, fig. 2). The lateral face of the basioccipital is concave, this concavity giving way anteriorly to the paired basilar tubercles. The occipital-basisphenoid suture is closed. In Okapia the anterior swellings of the occipital condyles are much stronger than in Zarvafa and the median keel is lost ; in contrast a median depression runs posteriorly from the basioccipital—basisphenoid suture. The basilar tubercles are stronger and more elongate in Okapia than in Zarafa. The tubercles of the basioccipital provide insertion for the rectus capitis ventralis muscle which acts to flex the head downwards.

Sphenoid. The posterior region of the orbitosphenoid is concave with the optic foramen lying at its posteriorend. Behind the optic foramen is a large foramen at the base of the alisphenoid. This foramen results from the fusion of the foramen rotundum and the foramen lacerum anterius (Colbert 1933) and through it emerge cranial nerves III, IV, VI and part of V, it is here referred to as the foramen rotundum. Behind the foramen rotundum the pterygosphenoid is produced as a ventral process (pl. 2) with convex lateral and concave medial faces. The ridge runs posteriorly from the posterior edge of the pterygosphenoid and forms the medial wall of the formen ovale through which the mandibular branch of cranial nerve V emerges. The foramen ovale is elongate in Zarafa as in Okapia and Guiraffokeryx but in contrast to the giraffines and sivatheriids in which the foramen is circular (Colbert 1935b). Postero-lateral to the foramen ovale a groove runs along the edge of the basisphenoid and dorsal to the auditory bulla as far as the eustachian canal and the foramen lacerum medius. The basisphenoid is transversely convex with a strong median keel fading out anteriorly. The basisphenoid of Okapia does not bear a median keel but is otherwise very similar to that of Zarafa.

Squamosal. In Zarafa the squamosal surface is concave lateral to the foramen ovale. The temporal condyle of the glenoid gives way laterally and posteriorly to a glenoid cavity. The post-glenoid process is a strong, high, transverse ridge extending further laterally than in Okapia or Giraffa. The anterior edge of the glenoid region is formed by a strong ridge which begins at the alisphenoid tuberosity and continues laterally as far as the zygomatic arch. The squamosal is convex dorsally and con- tributes about half of the temporal wall of the skull, it is also produced laterally as

7

GEBEL ZELTEN, LIBYA 89

part of the zygomatic arch. Over the ear region the squamosal forms a strong lateral tuberosity which contributes the postero-dorsal third of the external auditory meatus. The anterior edge of the tuberosity is produced into the temporal crest which runs anteriorly as far as the post-glenoid process. The squamosals of Zarafa and Okapia are similar.

Ear Region. In Zarafa the external auditory meatus is very ventrally situated, it is a postero-laterally directed tube formed by the petrosal and squamosal bones. The antero-lateral face of the external auditory meatus is concave with a strong hyoid process lying lateral to the bulla.

The bulla has a relatively thick wall, it is almost spherical in shape and is larger than that of Okapia. A large bulla is a primitive feature of the giraffids (Colbert 1938). The inner ear was partly exposed on the left side of the specimen; it is more elongate than that of Prolibytheriwm but appears to be similar in the main features of its ventro-lateral face. The fossa tensor tympani is very deep as in Prolibytherium. The bone surface is swollen postero-ventrally and above the swelling is a shallow concavity which leads anteriorly to the hiatus falloppii. These are the only parts of the inner ear visible.

The temporal canal opens between the ear region and the paroccipital process. Lateral to this the petromastoid suture runs dorsally between the squamosal and occipital regions. The petromastoid forms the postero-lateral part of the nuchal crest and has a concave posterior face.

Mandible. Only the posterior part of the mandible is known in Zarafa. The mandibular foramen is very large (pl. 4, fig. 3) and is situated more anteriorly in Zarafa than in Okapia. A deep depression runs from the foramen towards the condyle and from the ventral edge of the mandibular foramen a shallow groove Tuns antero-ventrally, this groove accommodates the lingual nerve and below it the face of the mandible is concave as far as the angle which was probably weak. Ventro- medial to Mg the surface of the mandible is swollen and convex. The condyle is expanded laterally and medially with a slightly curved articular surface as in Okapia. The lateral face of the mandible is slightly convex and in the region behind M3 the bone surface rises medially, forming a sharp medial ridge.

The Skull as a whole. The dorsal view of the skull (pl. 3) is dominated by the laterally expanded frontals but the extent, to which the lateral expansion of the frontals has effected the individual bones of the facial region, is difficult to assess. It is likely that the ancestral giraffids possessed a lacrymal fossa similar to that of the cervids; this assumption is made more probable by the condition of Proliby- theriwm and the concavity in Zavafa may represent a stage in the reduction of the fossa, however it is more likely that the expansion of the anterior supraorbital crest formed the concavity of the lacrymal. The dorsal expansion of the lacrymal in Zavafa may also be a primitive feature as Dremotherium has a large lacrymal bone but the lateral expansion of the frontals has probably influenced the lacrymal causing it to expand dorsally. The facial region ot Zarafa exhibits some features that are certainly primitive; the paired lacrymal foramina, elongate jugal bones, shallow maxilla and large maxillary tuberosity are all features found in the cervids and palaeomerycids but absent in adult giraffids.

B

90 LOWER MIOCENE RUMINANTS

The main parameters of the gross cranial anatomy in the giraffids were studied by Colbert (1938) who used seven features of the skull in an attempt to establish the primitive features of Okapia. Owing to the broken nature of the skull of Zarafa only four of Colbert’s parameters can be used (table 5). The degree of facial flexion is given as the angle between the basipalatal and basicranial axes; this angle is similar in Dremotherium, Okapia, Zarafa and Palaeotrvagus microdon and in these genera it is less than in advanced giraffids. The degree of facial flexion is usually treated as a measure of the evolutionary stage reached by the genus under con- sideration and on this basis Zarafa is as primitive as Dremotheriwm or Palaeotragus mucrodon. The angle between the basipalatal axis and the line joining the anterior edge of the orbit to the anterior border of the M1, indicates that the orbit of Zarafa is more anteriorly situated than in the other giraffids except Okapza in which the orbit is in a similar position to that of Dremotherium. An anteriorly situated orbit is usually regarded as a primitive feature. The nasals are flat in Zavafa and Palaeotragus microdon but in the other giraffids and in Dremotheriuwm the nasals are more flexed relative to the skull roof and the basipalatal axis. The condition of the primitive giraffid skull is discussed below and it is postulated that a wide, flattened skull roof is probably primitive for the giraffids.

In general shape of the skull, position of the ossicones and features of the basi-

TABLE 4 Zarafa zeltent. Measurements of the skull and dentition The Skull M.26670 M.26673 Height above M3 88 mm Postorbital length. (From anterior edge of orbit to occipital condyle) 187 mm Maximum width of frontals 179 mm Maximum width across occipital region 103 mm Width of occipital condyles 57 mm Maximum height of occipital region 78 mm 77 mm The Upper Dentition M.26671 M.26672 Length Width Length Width P4 I5 mm 20 mm — + M1 19 mm — 21 mm I9 mm M2 22 mm 24 mm 23 mm 22 mm M3 22 mm 23 mm — -- D2 — = 17 mm 8 mm Ds — — 18 mm 13 mm D4 — — 18 mm I5 mm The Lower Dentition M.26675 M.26676 M.26677 Length Width Length Width Length Width M2 25 mm 17 mm M3 33 mm 17 mm 32 mm I5 mm

(All dental measurements are maximum.)

oI

GEBEL ZELTEN, LIBYA

ofl xhAayof{vAry

(‘oor = y38u9] [W!—uq)

off +6 Sor 6S 89 yjS8ue] Iejour-1ejoureid 0} P[[XeUI JO YSIOY Jo O17 eYy ails Ol oll opt o8l s[eseu JO Iopi0q Jotiodns pur sixe [eyeTediseq usemzoq a[Suy ols 009 028 ofl 028 “TIN JO Iopiog Ior1ezue 94} OF FIG10 04} Jo espa Joe} oY} Sututol oul] pue sixe [e}eledIseq useMmyoq a[suy BOL Xe) AY, of of Soxe [elueioIseq pue [eze[ediseq usemyoq o[suy off{vsixy UOposaUM *q vrdvyoQ D{VADZY wNAaYoMaA

(g€6r yraqjog sazye ‘xAayoffysry pure vffony ‘uoposorm snsvmoanjog ‘v1dvyQ 10} sjusWeInseay)) Wmayjez v{pAvZ JO SeInyeoy [eIUWeIO ssoIxy

¢ alavy

92 LOWER MIOCENE RUMINANTS

cranial region Zarafa resembles the genus Palaeotragus and in particular the species P. microdon.

Upper Dentition. The deciduous dentition of Zavafa forms a closed series it is therefore likely that the permanent cheek teeth also formed a closed series from P2 to M8. The molars are more brachyodont than those of Palaeotragus, resembling in degree of brachyodonty the molars of Palaeomeryx. The enamel is finely rugose and the molars are four rooted with the lingual roots fused.

M! is almost square with the posterior half of the tooth displaced labially relative to the anterior half. The parastyle is strong and its labial rib is stronger than that of Palaeotragus. The paracone has a strong labial rib, similar to that of Palaeomeryx and stronger than that of Palaeotragus. The posterior end of the paracone lies lingual to the mesostyle which is the most labial part of the tooth. The metacone of M! is the same height as the paracone, its selene has a diagonal orientation on the tooth (pl. 5, fig. 1) and owing to the reduction of the labial rib it is more smoothly curved than the paracone. A weak metastyle forms the postero-labial corner of the tooth. The protocone is stout with a weak labial swelling which lends slight angularity to its crescentic shape, the posterior extension of the protocone ter- minates in the median valley without meeting the anterior face of the metaconule (pl. 5, fig. 1). The metaconule is higher than the protocone, its anterior region curves labially between the paracone and the metacone and from the anterior face of this region a small crest is produced into the median valley. A conule in the an terior fossette joins this crest and also joins the protocone and metacone giving an‘h’ shaped wear trace (pl. 6, fig. 1). The metaconule has a stronger labial swelling than the protocone and behind this the height of the cusp is rapidly reduced. A long narrow accessory crest is produced from the metaconule and runs antero- labially to meet the base of the metacone (pl. 5, fig. 1). M1 has strong anterior and posterior cingula. The M! of M.26672 (pl. 5, fig. 1) has a strong entostyle in the median valley, this arises from the base of the metaconule and does not join the protocone. M.26671 (pl. 6, fig. 1) has strong cingula in the median valley region but an entostyle is not developed. M?and M8 are similar in the main features of their anatomy to M!.

P4 is three rooted and brachyodont with finely rugose enamel and anterior and posterior cingula which do not join as a lingual cingulum (pl. 6, fig. 1). The parastyle has a strong labial rib (pl. 6, fig. 2) as in Palaeomeryx but in contrast to Palaeotragus in which the labial rib of P4 is feeble. The strong labial rib of the paracone curves anteriorly and carries a deep groove on its anterior face. The paracone of the Palaeomeryx P4is similar to that of Zavafa but in Palaeotragus the labial rib is reduced and the cusp is more compressed transversely than in the other two genera. The metacone has a weak swelling posterior to the paracone rib in Zarafa whereas in Palaeomeryx the labial rib of the paracone is clearly fused to the metacone rib, a groove indicating the line of fusion and in Palaeotragus a single rib is present with no posterior swelling. The protocone of the P4is stout and more regularly crescentic than in the molars; the anterior region joins the parastyle at an early stage of wear. On the posterior region of the protocone a weak accessory crest joins the base of the metacone and anterior to this crest a weaker parallel crest runs into the fossette

GEBEL ZELTEN, LIBYA 93

but does not join the base of the metacone (pl. 6, fig. 1). This region of the P4 differs from both Palaeomeryx and Palaeotragus but this crest is probably an individual variation of Zarafa.

D4 is molariform (pl. 5, fig. 1), the parastyle is stronger than in the permanent molars with a strong labial rib. The paracone is thicker transversely than in the molars and in the anterior fossette there are a number of small conules (pl. 5, fig. 1). The mesostyle is stronger than in the molars, remaining as an independent style until late in wear. The metacone has a weak labial rib and the metastyle is stronger than in M1 or M2. The metaconule of D4 is similar to that of the molars, a strong fold is produced from its anterior region into the median valley producing a forked anterior wear trace asin the molars. A strong accessory crest is produced from the posterior region of the metaconule and runs antero-posteriorly in contrast to the molars in which it runs antero-labially. A small entostyle is produced from the antero-lingual face of the metaconule into the median valley asin the molars. A cingulum stretches over the antero-lingual region of the protocone and a small cingulum covers the postero-lingual corner of the tooth.

D3 is triangular with a single anterior root and a posterior pair. The enamel is smooth except on the lingual faces of the paracone and metacone where it is rugose. The parastyle is strong, existing as a separate style until late in wear, it lies near the anterior end of the tooth and the wear trace of the paracone branches labially to meet it. The paracone has a very strong labial swelling with an anterior groove. The metacone and its anterior and posterior styles are similar to those of D4 or the molars. The protocone is elongate, its anterior region joins the protostyle which forms the anterior end of the tooth. A small crest is produced into the fossette opposite the parastyle and opposite the paracone a stronger crest is produced (pl. 5, fig. 1). The posterior end of the protocone joins the lingual part of the metaconule and there is no median valley. The metaconule of D? is similar to that of D4, it is crescentic and in the anterior region it is produced between the paracone and metacone (pl. 5, fig. 1). A strong crest is produced from the anterior region of the metaconule into the posterior fossette, this crest runs posteriorly across the base of the metacone terminating near the middle of the fossette. An accessory crest is produced from the posterior region of the metaconule, this also runs to the middle of the fossette but does not join the anterior crest. A lingual cingulum runs along the base of the protocone ending in the middle of the tooth and a short cingulum is produced at the postero-lingual corner.

D? is very heavily worn and few surface features are visible. The parastyle is strong as in D2 and joins a weak protostyle anteriorly. The paracone is high and more anteriorly situated than in D8, it has a strong labial swelling. The metacone is elongate with a weak labial swelling. The protocone is small and its anterior region joins the protostyle and the parastyle, behind this is a weak crest asin D3. There are three fine accessory crests on the posterior region of the metaconule (pl. 5, fig. 1). D2 has a weak lingual cingulum which runs around the base of the metaconule and the posterior region of the protocone. The anterior region of D2 resembles that of D3 but the posterior region does not show any molariform features.

Lower Dentition. The lower dentition of Zarafa is known from a single Mz and

94 LOWER MIOCENE RUMINANTS

two third molars. These teeth are identified as Zavafa on the basis of their size, giraffoid features, height and enamel features.

Mg resembles Canthumeryx in height and is lower than Palaeotragus rouenit. The mesostylid is weaker than in Canthumeryx or Palaeotragus and the metaconid rib is weak (pl. 5, fig. 3). In Palaeomeryx and Canthumeryx this rib is strong but in Palaeotragus it consists of a slight swelling of the surface. The entoconid has a weaker lingual swelling than in Canthumeryx or Palaeomeryx and resembles Palaeotragus. The posterior region of the entoconid is shortened but expands as a crest at some distance above the crown (pl. 5, fig. 3); a similar expansion is present in P. rouenit (M.8367) but in Canthumeryx the posterior crest is very strong throughout itsheight. The posterior fossette opens lingually (pl. 5, fig. 2). The anterior fossette is very shallow and widens at its anterior end (pl. 5, fig. 2) ; a similar widening occurs in Palaeotragus but not in Canthumeryx. The anterior cingulum is weaker in Zarafa than in Canthumeryx and the posterior cingulum is very short and small (pl. 5, fig. 2). Mg has no ectostylid.

The metaconid of Msg (pl. 5, fig. 4) is similar to that of Mg (pl. 5, fig. 2) and is more nearly parallel to the axis of the tooth than in Palaeotragus or Canthumeryx. The metaconid rib is weak as in Mz and the metastylid is also weak. The entoconid is similar to that of Mg but in its posterior region it expands backwards to join the strong entoconulid (pl. 5, figs. 4 and 5) thus closing the posterior fossette. The protoconid and anterior end of the hypoconid are similar to those of Mz but the posterior end of the hypoconid is flattened.

The accessory column consists of a stout hypoconulid which curves around the posterior end of the tooth, its antero-labial end abuts on the posterior end of the hypoconid and its lingual end curves anteriorly and joins the entoconulid (pl. 5, fig. 5). A strong ectostylid is present in the median valley and a weaker stylid stands in the posterior valley.

The cheek teeth of Zavafa show resemblances to both Palaeomeryx and Palaeotragus so that in some respects Zarafa may be regarded as intermediate between these genera. The labial ribs of the paracone and parastyle are weaker than is usual in Palaeomeryx though slightly stronger than in Palaeotragus. The mesostyle is strong in Zarafa as in Palaeomeryx but the smoothly curved general shape of the metacone and reduction of the labial metacone rib are very similar to Palaeotragus and contrast with Palaeomeryx. The protocones are similar in Zarafa and Palaeotragus and lack the strong postero-lingual spur which is present in Palaeomeryx. The cingula are more reduced in Zarafa than in Palaeomeryx but less reduced than in Palaeotragus. The upper molars of Zavafa are more hypsodont than in Palaeomeryx but are still much lower than in Palaeotragus. On the P4 of Zarafa the labial ribs of the paracone and parastyle are stronger than in Palaeotragus though weaker than in Palaeomeryx, with this exception the P*is very similar in all three genera though a lingual cingulum is sometimes present in Palaeomeryx only.

The lower molars of Palaeomeryx are low crowned and wider relative to their length than in Zarafa, Canthumeryx or Palaeotragus. The lingual faces of the lower molars carry weak ribs in Zavafa and Palaeotragus whereas in Palaeomeryx these ribs are much stronger. A very weak ‘Palacomeryx fold’ is present in M.26675, this fold

GEBEL ZELTEN, LIBYA 95

rarely occurs in Palaeotragus but it is usual and often strong in Palacomeryx. The lower molars are higher crowned in Zarafa than in Palaeomeryx though lower than in Palaeotragus.

The upper and lower cheek teeth of Zavafa show features in which they resemble Palaeotragus and as these features are more advanced than the primitive ruminant condition they are interpreted as indicative of true relationship between the genera.

Post-cranial material. The Gebel Zelten collection includes a considerable amount of post-cranial material but owing to the conditions of preservation none of this material was found in association with the cranial elements. Specimens may be assigned to one or other of the ruminant genera with varying degrees of confidence, highest in the case of Zavafa which is the largest ruminant found at Gebel Zelten. In cranial features Zarafa is sufficiently like the other palaeotragines for the assumption to be made that the post-cranial elements must also show affinities with this group.

MaTERIAL: B.U.20115—An almost complete right femur. B.U.z0116—A complete right tibia. B.U.zo117—-A complete right metatarsal. B.U.zo118—A complete left caleaneum. B.U.20119—A right calcaneum. B.U.20120—A right astragalus. B.U.20121—A left astragalus. B.U.20122—A left astragalus. B.U.20123—A left scapular fragment. B.U.20124—The distal end of a left humerus. B.U.20125—A right olecranon region. B.U.20126—The proximal end of a left radius. B.U.20127 —The distal end of a left radius. B.U.z0128—A fragment of a right metacarpal. B.U.20129—An anterior phalange. B.U.20130—A posterior phalange. B.U.20143 —An anterior phalange. B.U.20144—A posterior phalange. B.U.20145—A pos- terior phalange. B.U.20146—The proximal end of a left tibia. B.U.20147—-An axis. B.U.20148—An axis. B.U.20149—A seventh cervical vertebra. B.U.20150 —A thoracic vertebra, probably the eighth. B.U.20151—A fourth lumbar vertebra. B.U.20152—A sixth lumbar vertebra.

The pectoral girdle Scapula. The glenoid is shallowly concave and almost elliptical with a deep glenoid notch lying postero-lateral to the coracoid process (text fig. 7c). The coracoid process is strongly developed and projects further from the body of the bone than the tuber scapulae. In Okapia and Giraffa the tuber scapulae is very strongly developed and has grown over the coracoid process whereas in Zarafa the tuber scapulae is weak, consisting of a swelling lateral to the coracoid process. The surface of the tuber scapulae is heavily sculptured as the origin for the biceps brachii muscle. The spine of the scapula is shifted anteriorly and the supraspinous region is very narrow as in Okapia. The base of the spine indicates that it rises smoothly and an acromion process was probably present as in Palaeotragus and in contrast to Okapia and Givaffa in which this process is entirely absent. The absence of the acromion process appears to be related to the development of the tuber scapulae as in Equus, ‘Camelus and large species of Bos the acromion process may be lost, in which case the tuber scapulae is of a size comparable to that of Givaffa. A strong tuber scapulae presumably indicates a greater development of the biceps brachii muscle and similarly reduction or absence of the acromion process indicates a reduction in the importance of the acromial part of the deltoideus muscle. The infraspinatus region

96 LOWER MIOCENE RUMINANTS

is wide with a thickened posterior edge resulting in the concavity of the infraspinous fossa, the posterior edge is also concave as in Givaffa. The subscapular fossa is concave distally but the subscapular face of the neck is slightly convex. The shape of the neck of the scapula indicates that the blade was as elongate as that of Okapza. The anterior shift of the spine is also similar to Okapia.

Humerus. The distal end of the humerus is known from a single badly shattered specimen. The coronoid fossa is shallow with a deeply pitted surface and lateral to this the bone is expanded as a very strong lateral epicondyle similar to Okapia and Giraffa. The lateral condyle is wide with a concave surface and the medial condyle is also wide. The olecranon fossa is very deep. The distal end of the humerus is similar to that of Okapia and Giraffa.

Radius. The lateral tuberosity of the radius is weaker than in Okapza or Giraffa and the medial tuberosity is concave. These tuberosities provide attachment for the ligaments of the elbow and the smaller size of the lateral tuberosity indicates a weaker ligament; this is possible in a small relatively light animal such as Zarafa. The distal end of the radius has the usual three articular facets. The scaphoid facet is relatively wide; a concavity at the anterior end of the facet acts as a stop preventing over extension; this concavity is about the same depth in Okapia and shallower in Giraffa. The lunar facet has deep anterior and posterior depressions which also act as stops. The cuneiform facet is more oblique than the other two facets and its postero-medial region consists ofa vertical concavity. The dorsal face of the bone has two strong ridges asin Okapia. The shaft of the bone, although incompletely known, appears to have been more slender than in Okapia.

Ulna. The olecranon process of the ulna is the only region preserved. The semilunar notch is transversely narrow and the edges of the facet are rounded giving it a transversely convex shape which agrees closely with that of Givaffa. In Okapia the semilunar facet is transversely plane which greatly restricts the lateral mobility of the elbow. The semilunar region is produced disto-laterally as a strong process articulating partly with the lateral condyle of the humerus and partly with the posterior face of the radius; this facet is similarly developed in Givaffa. The olecranon process is short and transversely flattened with a concave medial face. The distal end is heavily sculptured for the insertion of the triceps muscle.

Metacarpal. ‘The facets of the proximal end are similar to those of Okapia and on the postero-medial face of the proximal end the surface of the bone is heavily sculptured as in Okapia but there is no indication of metacarpal V. The presence of this metacarpal is variable in Givaffa (Fraser 1951) and may also have been variable in Zarafa. The shaft of the metacarpal is more slender than that of Okapia; its cross-section is flattened as is usual in the ruminants, with a deep channel on the posterior face to accommodate the flexor tendons. The distal head has the usual paired condyles converging slightly as in Givaffa. The keel of the condyle is very strong and in the posterior region it extends proximally as a strong ridge over the articular face and beyond this onto the body of the bone. The strength of this keel resembles the cervid condition. In both Okapia and Guiraffa the condyle is ex- panded in the interdigital region which reduces the apparent strength of the keel. The distal head of the metacarpal is much wider than the shaft in Zarafa, Givaffa

GEBEL ZELTEN, LIBYA 97

and the cervids but in Okapia it is only slightly wider as the shaft is thickened.

Phalanges. The phalanges are very slender. Their postero-proximal epicondyles are very short whereas in Okapia and Giraffa they are elongate, covering about one third of the posterior face of the phalange. The interdigital face has a weak tuberos- ity in the distal region to which the interdigital ligament attaches. The distal articular face consists of a central depression flanked by two expanded areas which extend over the end of the bone; this facet is slightly oblique which causes the second phalange and hooves to move towards each other when weight is placed on the foot.

Five phalanges are identified with Zarafa, these include two larger and three small ones. The difference in size between these groups is approximately the same as the size relationship between the anterior and posterior phalanges of Okapia; for this reason the larger are identified as anterior and the smaller as posterior. The smaller phalanges will not be described separately with the description of the pelvic limb.

The Pelvic Limb

Femur. The head of the femur is small relative to the length of the bone; its articular surface extends onto the neck and around the lower side of the head as in Okapia. The neck of the femur is long and the head is displaced further medially than in Okapia or Givaffa. In both Okapia and Giraffa certain tendencies towards a graviportal condition of the limbs are evident, these tendencies are most apparent in the rotation of the articulations into the line of the shaft. The greater displace- ment of the articulations or curvature of the shaft of the limb bones in Zarafa indicates a lighter animal. The trochanter minor is weak and is not displaced as far medially as it is in Okapia or Giraffa; as a result the trochanteric ridge appears stronger in Zarafa than in the extant giraffids.

The distal region of the femur is of the usual ruminant pattern. On its anterior face the medial ridge is high and expanded but is less swollen than that of Okapia or Givaffa, this indicates that the ‘stifle joint’ was less effective than in Okapia or Givaffa as expected in a smaller lighter animal (Shuttleworth 1943). The medial and lateral epicondyles are very strong but are less transversely expanded than those of Okapia. The shaft of the femur is slender and slightly curved anteriorly (posterior face concave) as in the medium sized cervids and in contrast to Okapia and Giraffa in which the shaft is straight and relatively stout. The supracondyloid fossa is more elongate than in Okapza, this is probably due to the generally narrower nature of the distal region in Zarafa.

The femur of Zavafa is much more slender than that of Okapia and both extremities are narrower. The curvature of the shaft indicates a lightly built animal and features such as the positionof thetrochanter minorare similarto the cervidcondition.

Tibia. The proximal head of the tibia is narrow transversely as in the cervids and in contrast to Okapia or Giraffa in which the head is relatively wide. The cnemial crest is high and blends gradually into the shaft over the proximal quarter of its length; this crest is similar in the medium sized cervids but in Okapia it occupies the proximal third of the anterior face. The antero-proximal region of the cnemial crest forms a large triangular tuberosity bounded laterally by the sulcus muscularis

98 LOWER MIOCENE RUMINANTS

and medially by a shallow depression through which the middle patellar ligament passes. A small tubercle is produced posterior to the spine, this is similar to that of Okapia and provides attachment for the posterior cruciate ligament.

The distal end of the tibia is much narrower in Zarafa than in Okapia and in general features it resembles the cervid tibia. The shaft is slightly curved posteriorly as in the cervids and in contrast to Okapia in which the shaft is straight. The tibia of Zarafa is relatively shorter than that of Capreolus but longer than that of Okapia or Giraffa. Itissimilarinits main features to the tibia of Givaffa whichis more primitive than that of Okapia.

Calcaneum. The tuber calcis is very long and similar to Capreolus or Palaeomeryx whereas in Palaeotragus and Samotherium it is slightly shortened and in Okapia and Giraffa it is extremely shortened. The tuber calcis is more flattened than in Okapia. The posterior face of the sustentacular process is plane whereas in Okapia and Samotherium it is concave. The lateral face of the calcaneum bears an area of heavy sculpturing occupying most of the antero-proximal region, posterior to this is a small oval facet to which the lateral ligament of the ankle attaches. The fibular facet is of the usual primitive form with a raised convex posterior region and a concave anterior region as in Palaeomeryx and in contrast to Okapia in which this facet is specialized by the loss of the anterior region. The facet in Zarafa indicates that the fibula was probably similar to that of Capreolus. The sustentacular facet is wide and transversely convex, it does not show any reduction of the proximo-medial corner, such as is found in Okapia and Giraffa as a specialization facilitating greater flexion of the ankle. The calcaneum is generally more like that of the cervids than the extant giraffids, this is however partly due to the specialization of the ankle in Okapia and Giraffa.

Astragalus. This is more elongate than in Okapia or Givaffa and compares very closely in all its main features with the astragalus of Palaeomeryx.

Metatarsal. The proximal facets of the metatarsal are similar to those of Okapia. A deep notch lies on the lateral edge of the bone between the facets, this probably housed metatarsal V which appears to have been well developed in contrast with the extant giraffids in which it is represented as a thin ribbon of bone, entirely fused to the lateral face (Fraser 1951). A strong tubercle lies between and medial to the facets, this is metatarsal II and is stronger than in Givaffa or Okapia; it continues on the medial face as a wide ribbon of bone fused to the shaft and continuing much further distally than in Givaffa.

The shaft of the metatarsal is very long and slender in Zarafa, it has a deep cross- section as in Givaffa, Samotherium and Palaeotragus, this is narrower transversely than in Okapia. The two condyles of the distal extremity are similar to those of Givaffa but the ridges of the condyles are stronger as on the metacarpal. A deep channel runs down the anterior face of the bone, this channel is open for its whole length as in Okapia, Givaffa and the bovids whereas in the cervids the channel is roofed over in the distal region.

Vertebrae Axis. The axis of Zarafa is smaller and more elongate than that of Okapia. The

GEBEL ZELTEN, LIBYA 99

neural spine has a high posterior tubercle (text fig. 2a) from which it slopes down- wards anteriorly and projects over the odontoid process as in Okapia. The odontoid process is similar to that of Givaffa and the anterior articular facet has a depression near to and around the base of the odontoid process as in Givaffa but in contrast to Okapia in which the facet blends into the odontoid process. The anterior articular facet slopes almost vertically as in Giraffa. The intervertebral foramen (text fig. 2a) is in the same position as that of Okapia, it has a deep postero-ventral channel which joins it to the anterior channel of the vertebrarterial canal. This canal begins near the middle of the bone and emerges on the posterior face, dorso-lateral to the articulation. The posterior articular process is less pronounced than in Okapia, its articular facet is almost circular and faces postero-ventrally. The transverse

Fic. 2. The vertebrae of Zavafa. (Half natural size.). (A) Lateral view of the axis (B.U.20148). (Bs) Lateral view of the seventh cervical vertebra (B.U.20149). (c) Lateral view of the fourth lumbar vertebra (B.U.20151). (bp) The anterior face of the fourth lumbar vertebra (B.U.20151). (E). Anterior face of the sixth lumbar vertebra (B.U.20152). (F) Lateral view of the eighth thoracic vertebra (B.U.20150). a: neural spine. b: posterior tubercle. c: posterior articular process. d: odontoid process. e: inter- vertebral foramen. f: vertebrarterial canal. g: transverse process. h: anterior process. j: nutrient foramen. k: tubercular facet. 1: posterior facet for the head of the rib. m: anterior facet for the head of the rib. n: tuberosity.

100 LOWER MIOCENE RUMINANTS

process is more slender than that of Okapia and is more dorsally situated; lying at the same level as the top of the centrum (text-fig. 2a). The posterior articular face of the centrum is concave as in Okapia; and in Zarafa it is more nearly vertical than in Okapia.

Cervical vertebra. The seventh cervical vertebra (text fig. 2b) is similar to that of Okapia with no signs of the specialization found in Givaffa (Lankester 1908). The centrum is thicker than in Okapia and both anterior and posterior articulating faces of the centrum are more nearly vertical in Zavafa than in Okapia; in this feature the vertebra is similar to that of Givaffa. The anterior articulating processes are more widely divergent than in Okapia but the articulating facets are orientated at the same angle in both genera. The posterior articular processes are lateral only with no indication of the medial articulation such as occurs in Okapia (Lankester Ig10). The facets are elongate which indicates that there was considerable freedom of move- ment of the lower neck region.

Thoracic region. A single thoracic vertebra (B.U.20150) agrees closely with the eighth thoracic vertebra of Okapia. The anterior face of the centrum is shallowly convex and the facet extends ventrally as it does in Okapia. The facet for the head of the rib (text fig. 2f) is more closely blended with the centrum face than in Okapza. The ventral ridge is strong with a tubercle in the postero-dorsal corner of the concave lateral face (text fig. 2f), this is equally well developed in Okapia. The posterior face of the centrum is concave with the facets for articulation with the rib situated high up at the dorso-lateral corners of the face (text fig. 2f). The transverse process is well developed with a tubercular facet running antero-dorsally from its ventro-lateral region. The mamillary process is only weakly developed and is more laterally situated than in Okapfia. Both anterior and posterior articular processes are deformed as only the left side is developed in each case and the facets on this side have undergone compensatory size increases. The neutral spine is very slender and its posterior region is only slightly thickened; the spine was probably higher than in Okapia and more posteriorly inclined.

Lumbar vertebrae. The fourth lumbar vertebra (B.U.20151) is more elongate than in Okapia or Giraffa with a strong ventral ridge and ventro-lateral concavities as in Okapia. A large nutrient foramen lies lateral to the ventral ridge, just anterior to the middle of the centrum (text fig. 2c); in Okapia and Givraffa many small foramina are found in this region. The anterior face of the centrum is shaped as shown (text fig. 2d) and the posterior face is expanded slightly transversely as in the other giraffids. The anterior articular process is strong and high with a medially concave articular facet as is usual in the lumbar region of the ruminants. The posterior articular process is short and resembles Giraffa as it faces ventro-laterally, rather than laterally as it does in Okapia. The posterior articular process is not produced as far posteriorly as it is in Okapia. The neural spine is stout and high; it is expanded anteriorly in the distal region which is also thickened and has an excavated surface for the insertion of the longissimus dorsi muscle.

The sixth lumbar vertebra (B.U.20152) is free with a transversely expanded centrum (text fig. 2e) bearing a strong ventral ridge in the anterior region only. The anterior articular process has an elongate articular facet directed postero-

GEBEL ZELTEN, LIBYA IOI

ventrally as in Givaffa and not ventrally as in Okapia which indicates a greater freedom of movement at this joint in Zavafa and Giraffa than in Okapia. The transverse process (text fig. 2e) is slender and slopes slightly ventrally as in Givaffa. The facet of the posterior articular process is elongated antero-posteriorly but is not as long as that of Givaffa. The neural spine is more anteriorly situated in Zavafa than in Givaffa and the posterior part of the centrum is slightly elongated.

Functional Interpretations. The forelimb and vertebral column of Zarafa are poorly known and it is not possible to make any meaningful functional interpreta- tions of these regions. Fortunately the hind-limb is almost completely known and it has been used to assess the degree of cursorial adaptation attained.

The main recent study of functional features of the limbs of mammals is that of Smith and Savage (1956) who in their section on the hind limb deal only with the muscles originating on the pelvis. In Zarvafa the pelvis is not known and therefore this group of muscles cannot be studied but the mechanical advantage of the biceps femoris and semi-tendinous muscles, which insert on the cnemial crest, can be estimated using the formula:

Mechanical advantage of biceps __ Height of cnemial crest femoris and semitendinous muscles ————‘ Total length of lower leg

In this equation it is assumed that the femur and pelvis are rigid which they are not; but the accuracy of the figures obtained is sufficiently high as the formula is used for comparative purposes rather than to obtain absolute measurements. The height of the cnemial crest of the tibia is measured from the posterior face of the tibial spine and the length of the lower leg is measured vertically from the knee to the ground.

If the mechanical advantage of the locomotory muscles is high then a slow power- ful action is indicated as found in graviportal animals. A low mechanical advantage of the locomotory muscles indicates a weak but rapid action and is characteristic of cursorial animals, thus the mechanical advantage of a group of muscles may be used to measure the degree of cursorial adaptation attained. Table 7 indicates that Givaffa has a lower mechanical advantage for the biceps femoris than the other genera and Givaffa is more cursorially adapted than these genera. In degree of cursorial adaptation, as measured using the muscles inserting on the cnemial crest Zarafa lies between Okapia and Capreolus.

The mechanical advantage of the gastrocnemius muscle which inserts on the end of the tuber calcis can be estimated using the formula:

Mechanical advantage of —— Length of tuber calcis

gastrocnemius muscle Length of ankle and pes

As may be expected Givaffa again exhibits considerably greater cursorial adaptation than the other genera and Zarafa again lies between Okapia and Capreolus (table 7).

The osteological features of Zavafa indicate a lightly built animal and this is borne out by features of the limbs which taper rapidly and are very slender, thus the low degree of cursorial adaptation, as measured from the mechanical advantages of the muscles, must be interpreted as a primitive feature of Zarafa.

102 LOWER MIOCENE RUMINANTS

TABLE 6

Measurements of Zavafa zelteni post-cranial elements Vertebrae B.U.20147 B.U.20148 B.U.20149 B.U.20150 B.U.20151 B.U.20152

Length of centrum 69 mm 65 mm 32 mm 34 mm 47 mm 37 mm Total height of vertebra 63 mm 68 mm 72mm 131 mm oI mm 47 mm Depth of centrum 20 mm 19 mm 23 mm 24 mm 23 mm 21 mm Width of anterior end of centrum 52 mm 49 mm I9 mm 28 mm 35 mm 37 mm Width of posterior end of centrum 35 mm 34 mm 34 mm 36 mm 39 mm 48 mm Scapula B.U.20123 Width ot neck of scapula (Min.) 36 mm Depth of neck of scapula (Min.) 19 mm Width of glenoid (Ant-post) 47 mm Depth of glenoid (Transverse) 39 mm Humerus B.U.20124 Distal end Width across epicondyles 59 mm Antero-posterior depth of distal articular surface 27 mm Radius B.U.20126 B.U.20127 Width of proximal articular surface 49 mm Depth of proximal articular surface: Minimum 29 mm Maximum 18 mm Width of distal articular surface 44 mm Depth of distal articular surface 29 mm Metacarpal B.U.20128 Length 317 mm Width of distal end 46 mm Depth of distal end 25 mm Phalanges B.U.20129 B.U.20143. B.U.20144 B.U.20145 B.U.20130 Length 70 mm 68 mm 63 mm 62 mm 58 mm Width of proximal end 21 mm 21 mm 20 mm 19 mm 20 mm Depth of proximal end 23 mm 25 mm 23 mm 24 mm 23 mm Width of distal end 16 mm 18 mm 16 mm 16 mm 17 mm Depth of distal end 14 mm 15 mm I2 mm 13 mm 12mm Femur B.U.20115 Total length (Maximum) 341 mm Width of proximal end 84 mm Depth of articular head 36 mm Width of distal end 66 mm Depth of distal articular head: Maximum 83 mm Minimum 58 mm

GEBEL ZELTEN, LIBYA 103

TABLE 6 (cont.)

Tibia B.U.20116 B.U.20146 Length 349 mm — Width of proximal end 68 mm 69 mm Depth of proximal end 76 mm 72 mm Width of distal end 46 mm — Depth of distal end 35 mm — Calcaneum B.U.20118 B.U.20119 Total length 108 mm 115 mm Length of tuber calcis 74 mm 76 mm Width across sustentaculum 34 mm 32 mm Depth of tuber calcis 28 mm 27 mm Maximum depth of bone 41 mm 42 mm Astragalus B.U.20120 B.U.2012t B.U.20122 Maximum length 49 mm 41 mm 53 mm Minimum length 40 mm 40 mm 41 mm Width proximally 29 mm 30 mm 32 mm Width distally 29 mm 29 mm 30 mm Maximum depth 23 mm 23 mm 25 mm Metatarsal B.U.20117 Length 349 mm Width of proximal end 39 mm Depth of proximal end 41 mm Width of distal end 42 mm Depth of distal end 29 mm TABLE 7 Measurements used for functional interpretation of the Zavafa hind limb Givaffa Okapia Zraafa Capreolus Height of cnemial crest of tibia 10-5 mm 5°7 mm 6-5 mm 3°5 mm Total length of lower leg 179°0 mm 87:5 mm 87-0 mm 45°5 mm Length of tuber calcis 15-9 mm 8-6 mm 8-7 mm 5-0 mm Length of ankle and pes II0‘o mm 51-0 mm 50°0 mm 26:0 mm Mechanical advantage of biceps femoris muscle 0-059 0:065 0:075 0:077 Mechanical advantage of gastrocnemius muscle 0-163 0-166 0-174 o-192

Family SIVATHERIIDAE nov.

Diacnosis: Large giraffoids with strongly expanded frontal bones and pneumatized skull roof. On the frontal and parietal bones a pair of large, flattened, variably branched ossicones is produced. A great increase in overall body size occurs in advanced genera but elongation of the limbs and neck does not occur.

104 LOWER MIOCENE RUMINANTS

REMARKS: This group has previously been treated as a subfamily of the Giraffidae but the presence of Prolibytherium and Zarafa in deposits of early Miocene age indicates a division which extends back into the Oligocene. A more natural grouping results if the Palaeotraginae and Giraffinae are grouped together in the Giraffidae and the Sivatheres are placed in a separate family the Sivatheriidae.

Genus PROLIBYTHERIUM Arambourg 1961

Diacnosis: A primitive sivatheriid of small size. The cranium is narrow and exhibits very little facial flexion. The frontals support large, aliform ossicones which extend anteriorly in the supraorbital region and posteriorly over the parietal and occipital region. The occipital condyles are large with very thickened bone. A lacrymal fossa and paired lacrymal foramina are present.

Prolibytherium magnieri Arambourg 1961

Diacnosis: As for genus.

Ho.otype: A cranium with badly shattered ossicones, described by Arambourg (196Ia).

LOCALITY AND HORIZON: The material is all collected from the Lower Miocene (Burdigalian) deposits of Gebel Zelten, Libya.

REMARKS: Specimens in l'Institut de Paléontologie, Paris, include dentitions of Prolibytherium but these were not found with the type material and were not described by Arambourg.

MATERIAL: Specimens in l'Institut de Paléontologie are defined by capital letters.

M.21901 An almost complete skull; the ossicones, cranium and left maxilla

are complete. The upper dentition is heavily worn and lacks P3.

M.21899 A right mandible with heavily worn dentition.

M.26678 An endocranial cast that was dissected out of a skull.

M.26679 A fragment of cranium, with ossicones and bearing the only large

lacrymal fragment.

M.26680 An edentulous mandible which has a complete diastema.

M.26681 An isolated lower third molar, ightly worn.

ipeie AS An almost complete mandible with Pz to Mg exhibiting slight wear.

JPME, 183. A mandibular fragment with almost unworn M; to M3.

P.M. C. A mandibular fragment with heavily worn M;. The P, was dis- placed to the side during life and is thus virtually unworn.

P.M. D. A mandibular fragment with heavily worn M, to M3. PME: Pg and P4, both lightly worn. PME: A badly shattered M3. The paracone and antero-labial region is

missing but the metacone is complete. B.U.20175 A right mandible with heavily worn P3 to M3. B.U.20176 A cranium with the external auditory meatus preserved.

GEBEL ZELTEN, LIBYA 105

THE SKULL AND DENTITION

The Skull. The skull roof is entirely covered by large ossicones (pl. 7) which are completely fused to the frontal and parietal bones without any visible suture. The facial region is incompletely known as the premaxillary and nasal bones are not preserved and the jugalis known from a single small fragment. A complete lacrymal has not been discovered but several large fragments of the lacrymal are known. With these exceptions the cranial elements are almost completely known from well preserved specimens.

Maxilla. The maxilla of M.2tgor is broken near the postero-labial alveolus of M3, the break continuing antero-dorsally along the maxillary jugal suture (text fig. 3a). The maxilla is complete dorsally as far as the maxillary-nasal suture and anteriorly it is broken along what is probably the maxillary-premaxillary suture (pl. 8). The lateral surface of the maxilla is convex with a feeble facial crest which joins the weak facial tuberosity at its posterior end. The infraorbital foramen is slightly larger than the lingual alveolus of P%; it is deeply inset below the curve of the maxilla above P? as in Sivatherium and Capreolus (text fig. 3b). In Okapia and Givaffa the infraorbital foramen is more anteriorly situated than in Prolibytherium; this difference may be due to the expansion and increased height of the maxilla in the extant giraffids. The molars and P4 have a diagonal orientation relative to the main contours of the maxilla which results in strong juga alveolaria on the surface of the maxilla (pl. 8). Above the facial tuberosity the surface of the maxilla is deeply concave indicating the presence of a lacrymal fossa (pl. 8), this region agrees closely with that of Capreolus and differs from Okapia and Giraffa. Above this region the maxilla forms the antero-ventral border of the preorbital vacuity (text fig. 3a), this edge is similar to that of Capreolus (text fig. 3b).

The palatine process of the maxilla is shallowly concave with the median suture raised above the bonesurface. The anterior palatine foramen is level with the antero- lingual corner of M2 and the palatine groove runs anteriorly from it (pl. 9). Behind M3 is a large maxillary tuberosity as in Palaeomeryx and Zarafa.

The maxilla of Prolibytherium is very shallow in the region under the orbit, as in Zarafa. The maxillae of Prolibytherium and Capreolus agree closely in shape (text fig. 3b) ; this agreement is less close in the posterior region where the maxilla is reduced in Capreolus and the jugal is expanded ventrally.

Lacrymal. ‘The orbital face of the lacrymal is concave and the border of the orbit is sharp. Paired lacrymal foramina are present, the dorsal foramen lies behind the lacrymal tubercle and in Prolibytheriwm both foramina lie behind the edge of the orbit whereas in Capreolus they are situated on the edge of the orbit. The anterior face of the lacrymal is very concave and the bone is thin ventrally. Although the ventral part of the lacrymal is not known, the concavity of the dorsal region and the reduction in thickness of the bone agree with the concavity of the dorsal part of the maxilla and indicate the presence of a deep lacrymal fossa. A lacrymal fossa is absent from all giraffids and the other sivatheriids but is present in cervids and in some bovids. In the cervids this fossa houses the facial gland which produces a secretion used in territorial marking. The establishment and maintenance of a territory involves a certain amount of intraspecific combat and the occurrence of this

Cc

106 LOWER MIOCENE RUMINANTS

in Prolibytherium is also indicated by the large ossicones. The presence of a lacrymal fossa in Prolibytheriwm the cervids and Dremotherium could result from parallel development but in this instance it is more likely that its presence is a primitive feature.

Jugal. A fragment of the jugal is preserved attached to the maxilla (text fig. 3a). The dorsal edge of this fragment runs postero-ventrally and is probably the suture with the lacrymal. The flattened antero-dorsal face of the jugal is continuous with the lacrymal fossa.

Palatine. The palatine is transversely concave and the depth of this concavity increases posteriorly (pl. 9). The anterior palatine foramen opens on the anterior suture. The median suture is slightly raised as on the maxilla and the posterior end of the palatine is deeply indented forming the edge of the pterygo-palatine fossa (pl. 9); here the bone is vertical with a lateral face which is continuous with the maxillary tuberosity. The lateral face of the palatine is penetrated by the posterior palatine foramen which is similarly situated in Capreolus but is higher in Okapia.

b

Fic. 3. The maxilla of Prolibytherium (B.M.21901). (Half naturalsize). (a) Lateral view of the maxilla. a: maxillary-premaxillary suture. b: maxillary-nasal suture. c: border of preorbital vacuity. d:lacrymalfossa. e:fragmentofjugalbone. f:maxillary- jugal suture. g: facial tuberosity. h: infraorbital foramen. (B) Comparative outlines of the maxilla in Pvolibytherium and Capreolus. Prolibytherium —————————: Capreolus ----------------

GEBEL ZELTEN, LIBYA 107

Frontal. The orbital region of the frontal is concave and the orbital opening of the supraorbital canal is in the dorsal region of this concavity; probably as a result of the thickening of the bone, this opening is situated more medially than in Okapia or Givaffa. From its orbital opening the supraorbital canal runs dorso-medially for a short distance before turning sharply anteriorly, it continues in this direction as far as the dorsal face, thus opening much more anteriorly than is usual in the ruminants. The medial edge of the supraorbital process forms a strong ridge which continues medially as far as the alisphenoid tuberosity. The temporal region of the frontal is convex dorso-ventrally and blends into the ossicone in its dorsal region. The supraorbital foramen is anterior to the highest part of the orbit and the deep supraorbital groove runs anteriorly from the foramen to the lateral end of the frontal-nasal suture (text fig. 4). The supraorbital crest is formed by the thickening of the frontal along the edge of the orbit. Behind the supraorbital foramen the bone forms a triangular depression bounded antero-laterally by the supraorbital crest and medially by the median suture which is raised into a crest running posteriorly as far as the frontal eminence (text fig. 4).

Parietal. The lateral face of the parietal is convex and the dorsal region fuses with the ossicone. In front of the occipito-parietal suture the dorso-lateral face of the bone is concave with a strong parietal crest which blends into the ossicone anteriorly. The opening of the large temporal canal is visible on the broken edge of the bone (pl. 7).

Ossicones. The ossicones fuse completely with the expanded frontal and parietal bones, they are large and aliform and have been arbitrarily divided into anterior and posterior palmations for the purposes of description (text fig. 4). From the frontal eminence a strong crest runs antero-laterally forming the posterior edge of the triangular supraorbital concavity of the frontal (text fig. 4); this crest continues laterally as the leading edge of the anterior palmation. Another crest runs postero- laterally from the frontal eminence and forms the thick posterior edge of the posterior palmation. Lateral to these crests the ossicones become much thinner with concave dorso-lateral faces.

The dorsal and ventral surfaces of the ossicone bear a large number of fine, radiating grooves (pl. 7). Immediately dorsal to the supraorbital process is a small foramen on the leading edge of the anterior palmation, a deep groove originates from this foramen and runs postero-laterally across the lateral concavity of the ossicone, giving off many smaller branches over the surface.

The ventral surface of the ossicone is convex antero-posteriorly and shallowly concave transversely. A large groove begins postero-lateral to the supraorbital process and runs postero-laterally nearly to the edge of the ossicone where it forks, one branch follows the lateral edge of the anterior palmation and the other follows the edge of the posterior palmation. Above the glenoid cavity a large foramen, at the base of the ossicone, gives rise to a deep groove which divides into many fine grooves radiating over the lateral surface of the ossicone. A medial foramen is situated above the occipital region, at the base of the ossicone, lateral to this is a stout column beyond which is a larger lateral foramen (pl. 10, fig. 2) many fine grooves radiate from this foramen.

108 LOWER MIOCENE RUMINANTS

Occipital. The dorsal surface of the supraoccipital region is convex as far as the nuchal crest. In the palaeotragines the supraoccipital region is reflected over the condyles but in Prolibytherium it 1s vertical or slopes slightly anteriorly as in Capreolus. The mastoid process forms the lateral wings of the nuchal crest and the supra- occipital forms the dorsal part of the crest (pl. 10, fig. 2). The medial edge of the mastoid foramen is formed by the occipital bone. On the posterodorsal face of the supraoccipital region are paired depressions providing insertion for the semispinalis capitis muscles—the chief extensors of the head and neck. The external occipital protuberance lies medio-ventral to these pits (pl. Io, fig. 2), its surface is deeply sculptured (pl. 10, fig. 2) for attachment of the nuchal ligament. The ventral part of the external occipital protuberance is produced ventrally as a strong median crest. The occipital condyles have sharp keels on their lateral faces (pl. Io, fig. 1). The condyles are set very close to the skull; they are relatively large and the bone is

Fic. 4. Dorsal view of the ossicones of Prolibytherium (B.M.21g901) (One sixth natural size.). a:posteriorpalmation. b:anteriorpalmation. c—d: cross-section of the ossicone. e: position of frontal eminence. f: supraorbital groove. g: fronto-nasal suture. h: dorsal opening of supraorbital canal. x-y: cross-section of the ossicone.

GEBEL ZELTEN, LIBYA 109

greatly thickened. The foramen magnum is small relative to Okapia or Giraffa and the incisura intercondyloidea, which is strong is the giraffids, is reduced to a small notch on the ventral edge of the foramen magnum.

The basioccipital region is extremely thickened with a pair of large anterior swellings produced from the anterior region of the occipital condyles (pl. 10, fig. 1). Weaker swellings are present in Okapia and Givaffa. Mead (1906) found that these anterior swellings are present in all horned ruminants, in which they act to strengthen the atlanto-occipital joint during combat and according to Webb (1965) they also act to prevent over flexion of the joint during grazing. The great weight of the ossicones in Prolibytheriwm would increase the requirement for such swellings for both functions. The median keel originates at the anterior end of the basioccipital and runs posteriorly, merging into the condyles between the swellings (pl. 10, fig. 1). Paired basilar tubercles lie in front of the swellings (pl. Io, fig. 1) ; they have sculptured surfaces and fade anteriorly into slight concavities. The tubercles provide insertion for the rectus capitis ventralis muscles which act to flex the head and to incline it laterally. The eustachian canal follows the lateral edge of the basioccipital region.

The paroccipital process is stout, with a thickened anterior face providing origin for the digastric muscle. The posterior face of the process is concave and is the area of insertion of the rectus capitis lateralis muscle. The anterior face of the process is sculptured for attachment to the meatus region of the squamosal.

Sphenoid. The optic and ethmoid foramina are more ventrally situated in Prolibytherium than they are in Givaffa or Okapia and the ethmoid foramen is more anteriorly situated in Prolibytherium. Behind the optic foramen is the large foramen rotundum (pl. ro, fig. 1) and behind this is the foramen ovale which is circular in Prolibytherium, as in Giraffa and Sivatherium and in contrast to Zarafa and the other palaeotragines in which it is oval. The orbitosphenoid is concave in front of the optic foramen. The alisphenoid has a plane lateral face and a concave ventral face, with a strong alisphenoid tuberosity at the junction of these faces. The pterygoid groove follows the edge of the basisphenoid ventro-medial to the foramen rotundum and the foramen ovale, it is roofed over at its posterior end by the temporal process of the sphenoid; this process has two longitudinal grooves on its surface, the medial one of these carries the eustachian tube and the other is continuous with the pterygoid groove. Only the posterior end of the pterygo-sphenoid is preserved. The basisphenoid is cylindrical widening slightly at the posterior end, near the suture with the basioccipital (pl. 10, fig. 1).

Squamosal. The cranial surface of the squamosal is overlapped by a downward process of the parietal and is not visible. The lateral face is convex and the post- glenoid foramen is directed dorso-laterally. In front of the postglenoid foramen the bone surface is convex as far as the squamosal-alisphenoid suture. The glenoid cavity is transversely convex and the postglenoid process is slightly concave on its anterior face, the process is elongate transversely as in Zarafa (pl. Io, fig. 1). The temporal condyle is shallowly convex and its anterior edge forms a strong crest as far as the alisphenoid. The bone is concave medial to the glenoid cavity and a groove runs postero-laterally around the postglenoid process. The zygomatic arch is missing as the bone is broken immediately lateral to the glenoid cavity. The lateral

110 LOWER MIOCENE RUMINANTS

tuberosity of the squamosal is stronger than in Okapzia, it has a concave posterior face which is deeply sculptured. The antero-ventral face has a deep groove running laterally across it, this is the postero-dorsal part of the tube of the external auditory meatus.

Petrosal. The tympanic bulla is not preserved but the region which surrounded the bulla indicates that it was about the same relative size as that of Okapia or Zarafa. The petrosal contributes most of the meatus, with a small contribution from the squamosal in the postero-dorsal region. The meatus opens laterally as in Okapia or Givaffa and in contrast to Capreolus in which it slopes dorso-laterally from the bulla. The stylomastoid foramen lies postero-lateral to the meatus. Antero- ventral to the meatus is a strong hyoid process.

The inner ear is figured in ventro-lateral and medial views (text figs 5a and b). In ventro-lateral view (pl. 11, fig. 3) the most prominent region is the fossa tensor tympani, a large deep depression housing the tensor tympani muscle which inserts on the malleus. Anterior to this fossa is a deep, narrow channel, the hiatus falloppii, which carries the superficial petrosal nerve, a branch of cranial nerve VII. Posterior to the fossa tensor tympani is a small foramen which is continuous with a deep, ventrally running groove—the aquaeduct falloppii, this in turn is continuous with the stylomastoid foramen (text fig. 5a). Cranial nerve VII enters the petrosal via the

A

Fic. 5. The right petrosal of Pyrolibytherium (1:25 x natural size). (A) Ventro-lateral view. (B) Dorso-medial view. a: lateral tuberosity of the squamosal bone. b: aquae- duct falloppii. c: fossa tensor tympani. d: hiatus falloppii. e: fenestra ovalis. f: fenestra rotunda. g:aquaeduct cochlea. h:chordatympani. i: groove of the foramen lacerum posterius. j: petrosal wall of the stylomastoid foramen. k: squamosal-petrosal suture. 1: squamosal region of the ear. m: floccular fossa. n: wall of the temporal canal. o: aquaeduct vestibularis. q: crus commune. r: groove of the sulcus venus basilaris cranii. s: position of the semicircular canals. wu: internal auditory meatus.

GEBEL ZELTEN, LIBYA III

internal auditory meatus and after running along the aquaeduct falloppii leaves by way of the stylomastoid foramen. Postero-ventral to the fossa tensor tympani is a large opening—the fenestra rotunda, above which is the smaller fenestra ovalis (text fig. 5a). Posterior to and partly overhanging the fenestra ovalis is the stapedial promontory which provides origin for the stapedial muscle. The whole ventro- lateral face of the petrosal is very similar in Prolibytherium and Capreolus.

The medial view of the petrosal (pl. 11, fig. 4) is dominated by the internal auditory meatus which has the shape shown (text fig. 5b). Cranial nerve VII enters the petrosal posteriorly and cranial nerve VIII enters anterior to this; separate openings for the nerves are visible inside the meatus. Above and slightly behind the meatus is the wide, shallow floccular fossa which houses the floccular lobe of the cerebellum. Between the floccular fossa and the internal auditory meatus the bone is swollen by the crus commune. The semi-circular canals lie ventral to the internal auditory meatus, these were located by X-raying the petrosal but owing to the size of the specimen this technique could not be used to discover any further details of the internal anatomy.

The aquaeduct vestibularis and aquaeduct cochlea carry small veins. The medial region of the petrosal in Prolibytherium differs only slightly from that of Capreolus, the main difference being that the region around the floccular fossa and the fossa itself are more concave and larger in Capreolus. The petrosal of Okapia differs from Prolibytherium as the ventral part of the petrosal is expanded and the floccular fossa is relatively smaller in Okapia.

Brain. Asingle endocranial cast of Prolibytherium was available (pl. 12, fig. I and 2); this was formed from large calcite crystals and was dissected out of an almost complete cranium. The rhinal sulcus is situated very high on the lateral face of the brain (text fig. 6a) ; this reflects the greater importance of the rhinencephalon relative to the neopallium and is a primitive feature which is also found in Dremotherium, while in Capreolus and Okapia the rhinal sulcus is more ventrally situated. The anterior and posterior rhinal sulci meet each other at an angle of about 150°; this agrees with the angle given by Sigogneau (1968) for Dvemotherium. In both Prolibytherium and Dremotherium the two regions of the rhinal sulcus join with a smoother curve whereas in Capreolus the anterior and posterior regions bend sharply upwards before joining. The anterior rhinal sulcus runs slightly ventrally and its anterior end is lost, due to an imperfection of the cast.

The olfactory lobes are produced anteriorly and a part of the olfactory peduncle is visible in dorsal view (text fig. 6b); this is similar to Amphitragulus and is more primitive than Dremotherium or Capreolus in which the olfactory peduncle is shorter and the lobes are flexed ventrally, lying under the neopallium. Posterior to the lateral fissure the brain is expanded laterally which results in a great widening of the posterior region; this is similar to Dremotherium but in Capreolus the posterior region is not expanded and in Okapia the brain is uniformly very wide.

The cerebellum and vermis are relatively large and lie at the same level as the neopallium ; this is a primitive feature also found in Dremotheriwm. The neopallium is flattened in the vertical plane and its dorsal surface is almost flat as in Okapia whereas in Capreolus this region is curved antero-posteriorly and transversely.

112 LOWER MIOCENE RUMINANTS

The parts of the suprasylvian sulcus are clearly defined (text fig. 6a). The anterior branch curves medially with a very small descendant branch, whereas in Capreolus the inward curvature is more marked and a large descendant branch is produced running just in front of the sylvian sulcus. The central part of the suprasylvian sulcus is shallowly convex dorsally, this convexity is similar to Dremotherium but is less marked than in Capreolus. In Okapia (Clark 1939) the posterior and central parts of the suprasylvian sulcus are separated and overlap extensively. In Proliby- therium the posterior branch of the suprasylvian sulcus extends posteriorly almost as far as the cerebellum, it does not bifurcate whereas in Dremotherium a weak descendant branch is produced. A long posterior extension of the suprasylvian sulcus indicates an advance over the tragulid condition in which the posterior

Fic. 6. The brain of Prolibytherium (B.M.26678) (Natural size x 0-7). (A) Right lateral view of the endocranial cast. (B) Dorsal view of the endocranial cast. a: posterior suprasylvian sulcus. b: lateral groove. c: anterior suprasylvian sulcus. d: sylvian sulcus. e: gamma sulcus. f: olfactory lobe g: anterior rhinal sulcus h: anterior ectosylvian sulcus i:lateralfissure j: posterior rhinal sulcus and posterior ectosylvian sulcus k: piriform lobe 1: delta sulcus m: olfactory peduncle.

GEBEL ZELTEN, LIBYA 113

region of the sulcus is very short. The presence of a well marked descending limb of the suprasylvian sulcus, is stated by Black (1915) to be a giraffid feature and its absence in Prolibytherium is therefore a primitive feature.

The ectosylvian and rhinal sulci are separate near the lateral fissure but fuse completely at a short distance from the fissure both anteriorly and posteriorly, as in Dremotherium. This feature is primitive and in Capreolus and Okapia the rhinal and ectosylvian sulci are separate over their whole length. The lateral fissure is triangular and large (text fig. 6a) as in Dremotheriwm and Capreolus. The sylvian sulcus is produced from the dorsal region of the lateral fissure, this sulcus runs antero-dorsally and almost reaches the suprasylvian sulcus (text fig. 6b) ; this feature is similar to the cervids rather than the giraffids Clark (1939). Behind the sylvian sulcus is a deep groove which is more nearly vertical in Capreolus but is absent in Dremotherium. In Dremotherium the sylvian sulcus is more vertical than in Capreolus or Prolibytherium. The posterior region, between the suprasylvian and ectosylvian sulci, is occupied by the ‘delta’ sulcus (Sigogneau 1968) ; this sulcus is forked anteriorly and curves ventrally in its posterior region. The ‘gamma’ sulcus is very deep and much shorter in Prolibytherium than in Dremotherium.

The brain as a whole compares closely with that of Dremotherium but details of the sulci and olfactory lobes indicate that it was more primitive in Prolibytherium. The long sylvian sulcus is a cervid rather than a giraffid feature; however in Amphutragulus the sylvian sulcus is very short, this casts some doubt upon the importance of this feature in establishing relationships in early ruminants. As Prolibytherium is an early sivatheriid it is very unfortunate that no brain casts of sivatheriids were available. Agreement between the brain casts of Prolibytherium and the sivatheriids would probably be greater than the agreement between Prolibytherium and the giraffids.

Mandible. The ascending ramus is high with a concave medial and a plane lateral face in the dorsal region. The mandibular foramen is at the same level as the tooth row, it is elongate antero-posteriorly and a shallow channel runs anteriorly from it. The lower part of the ascending ramus is concave laterally and provides insertion for the masseter muscle. The horizontal ramus is convex dorso-ventrally on both sides. The ventral edge of the ramus is curved with the deepest region below Mz (pl. 11, fig. 2). The bone surface slopes dorso-medially behind the Mg and a sharp ridge is formed which continues dorsally as the leading edge of the ascending ramus. A small posterior mental foramen lies ventral to Pe (pl. 11, fig. 2) in both specimens in which it is preserved. The anterior mental foramen is double in M.26680, it lies at the level of the posterior end of the symphysis and the two foramina communicate internally. The diastema of Prolibytherium is about the same length as the distance Pz to Mz, which is the same as in Capreolus but is much shorter than in Giraffa or Okapia in which the diastema is longer than the tooth row.

Upper Dentition. The molars are generally similar in form; they are brachyodont with finely rugose enamel. M.z21gor isso heavily worn that the crown of M1 is almost completely worn away and the only remaining feature is the posterior fossette (pl. 9). The antero-labial corner of M2 bears a strong parastyle and both the paracone and parastyle have strong labial ribs. Posterior to the paracone the wear trace is very

114 LOWER MIOCENE RUMINANTS

wide but the mesostyle appears to have been originally part of the posterior lobe joining the paracone at a late stage in wear. The metacone is very stout and a slight postero-labial swelling indicates that a weak metastyle may have been present. The protocone is crescentic with the wear trace of its anterior region joining the parastyle and closing the fossette anteriorly (pl. 9). The posterior end of the pro- tocone joins the metaconule in the median valley. The metaconule is also crescentic and is posterior end joins the metacone or possibly the metastyle thus closing the posterior fossette. At the anterior end of the metaconule is a small crest which is produced into the median valley on the lingual side of the main region of fusion between the wear traces, this crest joins the protocone enclosing a small median enamel island in M! and M2 but this is not developed on M3 (pl. 9). Each molar has a strong entostyle in the median valley; in each case this is produced from the antero- lingual region of the metaconule but on M$ it joins a strong lingual cingulum pro- duced from the postero-lingual face of the protocone (pl. 9). Anterior cingula are present on each molar and posterior cingula were present on M! and M2.

A single badly shattered molar showing very little wear is known (P.M.F.). The mesostyle has a strong labial rib which is joined to the metacone. The metacone has a weak labial swelling and the metastyle has a strong labial rib. The protocone and metaconule are angularly crescentic and the latter is smaller than the protocone. A strong accessory crest is present in the posterior fossette.

The molars of Prolibytherium are smaller than those of Zavafa, detailed comparison is difficult owing to the heavily worn condition of the Prolibytherium dentition but the details of the median valley region differ between the two genera; this difference is due to the absence of the small conule in the anterior fossette of Prolibytherium resulting in a more simple wear pattern at the posterior end of the protocone.

P4 is three rooted and brachyodont, it is smaller, relative to the molars and its lingual region is smaller than in Zarafa. The parastyle is strong with a large labial swelling and the paracone is very stout with lingual and labial ribs; the lingual rib is weaker than in Zarvafa but it has an anterior groove as in Zarafa and Palaeomeryx. Posterior to the paracone rib, the labial wall is almost flat as far as the metastyle. The metacone is stout and almost completely fused with the paracone, its labial swelling is weak though slightly stronger than that of Zarafa. The protocone is stout and joins the labial lobe at both ends thus closing the fossette. The fossette of P4is much deeper than that of Zarafa, this is also true of the fossettes of the molars indicating that the upper dentition of Prolibytherium is much higher than that of Zarafa, though still brachyodont relative to the advanced giraffids or sivatheriids. At the posterior end of the fossette of P4is a weak accessory crest (pl.g). The antero- lingual cingulum of the P4 is weak relative to that of Zarafa.

P? is stout with a heavily worn crown which is concave antero-posteriorly. There are three strong ribs on the labial face which correspond to the parastyle, paracone and metacone. Between the parastyle and paracone ribs is a very deep groove (pl. 8) and a shallower groove separates the paracone and metacone ribs. No surface features are visible except a small, transversely elongate enamel island at the posterior end of the tooth.

Lower Dentition. The lower dentition forms a closed series from Pg to Mg. The

GEBEL ZELTEN, LIBYA 115

enamel of the molars is finely rugose. P.M.B. has almost unworn molars and M.26681 exhibits signs of only slight wear, these molars show that the cheek teeth are higher in Prolibytherium than in Zarafa.

M3 has an accessory column but the anterior region differs only slightly from the other molars. The mesostylid has a strong lingual rib which runs postero-ventrally from the antero-lingual corner of the tooth (pl. 10, fig. 4). The metaconid is high and transversely compressed (pl. Io, fig. 3), its posterior end lies lingual to the anterior end of the entoconid which is lower than the metaconid and has a more diagonal orientation on the tooth (pl. to, fig. 3). The protoconid is crescentic and slightly angular, its anterior end joins the mesostylid early in wear. The posterior end of the protoconid joins the anterior end of the entoconid early in wear (pl. Io, fig. 3) and a little later the metaconid joins the entoconid thus isolating the hypoconid until very late in wear; plate 11, fig. 1 shows a heavily worn dentition in which the hypoconid of the Mg has just joined the protoconid. The hypoconid is lower than the protoconid, its anterior end stops in the median valley near the entoconid. The hypoconid is not crescentic in the unworn condition as the posterior region stops near the middle of the cuspid and in this region the cuspid falls away and is continued labially a few millimetres below the top of the crown; thus after moderate wear the cuspid becomes crescentic but the posterior region is always shorter and appears narrower even in the heavily worn condition. This feature is restricted to the Mg and is related to the presence of the accessory column. The accessory column is more simple in Prolibytheriwm than in Zarafa; it consists of a stout hypoconulid which is as high as the hypoconid and is similar in shape to the anterior half of that cuspid. The posterior half of the hypoconulid has a weak lingual twist but there is no approach to the crescentic shape found in the giraffids. The anterior half of the hypoconulid meets the postero-labial face of the entoconid and on the lingual side of this junction is a small fold of enamel which is probably a weak ectostylid. Mg has a feeble anterior cingulum and usually an ectostylid in the median valley.

Mj and M¢ are similar to the anterior region of M3. At the postero-lingual corner of Mg is a small entostylid which joins the postero-lingual extension of the hypoconid ; in all specimens the entoconid is joined to the entostylid in this region.

P, is long and narrow (pl. 11, fig. 1). The protoconid is the highest part of the tooth but the metaconid is almost as high, these two cuspids are joined by a narrow ridge which has a deep groove on its antero-lingual face (pl. 11, fig. 1). The anterior part of the tooth curves lingually and from the lingual face the lingual and antero- lingual paraconid and parastylid are produced; these branches are separated by a shallow lingual valley which is quickly lost during wear, leaving a single anterior region (pl. 11, fig. 1). Between the paraconid and metaconid is a deep, wide lingual valley at the lingual end of which is a strong stylid. The posterior region of P4 consists of a stout labial hypoconid to which a postero-lingual entostylid and a lingual entoconid are joined. These two cuspids are separated by a shallow lingual valley which is lost early in wear but may persist for some time as a small enamel island. Between the protononid and the posterior region is a deep valley separating the metaconid and entoconid lingually and labially a deep groove separates the protoconid and hypoconid (pl. 11, fig. 1). A single unworn Py is known; on this the posterior

116 LOWER MIOCENE RUMINANTS

region is entirely separated from the anterior region as the lingual and labial valleys join over the crown of the tooth. The separation of these two regions is found in the giraffids and in some palaeomerycids.

P3 has a high protoconid with a strong postero-lingual wing—the incipient meta- conid. The anterior region consists of a single wear trace derived from a paraconid and parastylid asin Py. A lingual stylid closes the anterior valley but this is much weaker thanin Py. The posterior region is separated from the protoconid by a strong lingual valley and a weaker labial valley. The hypoconid is feeble in P3. The entoconid is strong but the entostylid is very slender and forms the posterior face of the tooth.

The Pz of Prolibytherium is a high conical tooth (pl. 11, fig. 2) of the primitive ruminant premolar form. The protoconid is high and a weak postero-lingual fold represents the metaconid. Anteriorly the paraconid consists of an unbranched crest which runs anteriorly from the face of the protoconid (pl. 11, fig. 1) and twists lingually at the distal end. The posterior region is transversely wide with a weak entoconid and entostylid. This tooth is similar that of Palaeotragus roueni but in this species the anterior crest is divided into a parastylid and a paraconid which is a more advanced condition.

The Skull as a Whole. The dorsal view of the skull is dominated by the enormous ossicones which are completely fused to the frontal and parietal bones. The great thickening of the basioccipital region and the occipital condyles is directly related to the presence of the ossicones. Though the facial region is poorly known it is of great interest as a lacrymal fossa and paired lacrymal foramina are present, these are primitive features previously unrecorded in the Giraffoidea but similar to the cervoid or dremotherioid condition. The maxilla is shallow as in Capreolus and it is likely that the basicranial and basifacial regions were widely divergent as in Zavafa. The endocranial cast shows similarities to the dremotherioids. In most features the skull of Prolibytheriwm is similar to the early pecorans but this resemblance is decreased by the presence of the ossicones in Prolibytherium.

TABLE 8

Prolibytherium magniert. Measurements of the skull and dentition

The Skull B.M.21901 Postorbital length. (From anterior edge of orbit to occipital condyle) : 132 mm Width of ossicones above the orbit 320 mm Maximum length of the ossicones 425 mm Width of condyles 51 mm Height of occipital region 71 mm Height of occipital region plus ossicones 114 mm Maximum width of occipital region 93 mm The Upper Dentition B.M.21901 Length Width

p2 12-5 mm 9:0 mm

aie oe eee

GEBEL ZELTEN, LIBYA 117 TABLE 8 (cont.) p4 II1-o mm I4°;0 mm M} 15:0 mm 17°55 mm M2 18-0 mm 20:0 mm M3 20:0 mm 20:0 mm The Lower Dentition

M.21899 M.26681 B.U.20175 P.M. A. P.M. B. P.M. C P.M. D. P2 Width 5°3 mm as -— — — + — Length 10-0 mm — — — — -- — Ps Width 6-5 mm = — 7-0 mm — as — Length 11-0 mm — Izomm 12:0mm --- — — Pa, Width 8-5 mm — 9-0 mm 775 mm a 7°5 mm — Length 130 mm — 145mm 12°8mm — 13-5 mm — Mi Width 12.0 mm — Iz0mm 12;°0mm — — 11-5 mm Length 1370 mm — I5omm 135mm 140mm — 16°5 mm Me Width 12°5 mm — I3z0mm 13°0mm — — 130 mm Length 16°5 mm — — 185mm 185 mm — 17-5 mm M3 Width izomm 130mm 13:°mm 12°55 mm — 130 mm Length 235mm 265mm 265mm 24:°5mm 25:0mm — 25-0 mm

Post-cranial material. Details of the skull indicate that Prolibytherium was probably similar to the more primitive cervids and it is therefore likely that its post-cranial material resembled that of the cervids. On this basis the post-cranial material of Prolibytheriwm was identified by comparison with a skeleton of Capreolus.

MATERIAL B.U.20153 A proximal fragment of a left scapula. B.U.20154a The distal end of a left humerus.

b A complete left radius.

c The proximal end of a left ulna. B.U.20155 The distal end of a metacarpal. B.U.20156 _ A first phalange.

B.U.20157__—CA first phalange.

B.U.20158 __ A first phalange; posterior? B.U.20159 The proximal end of a right tibia. B.U.20160 The proximal end of a left tibia. B.U.20161 =‘ The distal end of a right tibia. B.U.20162 The distal end of a left tibia. B.U.20164 =A right calcaneum.

B.U.20165_ ~—A right astragalus.

B.U.20166 __—A right astragalus. B.U.20167__—A left astragalus.

B.U.20168 A right astragalus.

118 LOWER MIOCENE RUMINANTS

B.U.20169 A distal, metatarsal fragment.

B.U.20170 An almost complete atlas vertebra. B.U.20171 _ A slightly broken seventh cervical vertebra. B.U.20172 A complete first thoracic vertebra. B.U.20173, An almost complete thoracic vertebra.

Scapula. The glenoid region of the scapula has an almost circular outline with a small glenoid notch (text fig. 7a). The coracoid process is strong but the tuber scapulae is weak. The spine begins near the glenoid and rises steeply indicating the presence of a strong acromion process. The neck of the scapula is wider than in Zarafa or Okapia; it has a deep vascular groove on the medial surface and a strong tuberosity on the posterior edge which is the area of origin for the teres minor muscle. The outlines of the glenoids in four giraffoid genera (text fig. 7) demonstrate the dif- ferences which exist in the development of the coracoid process and the tuber scapulae. The tuber scapulae of Prolibytherium (text fig. 7a) is very feeble, in Zarafa it is stronger (text fig. 7c) and in the extant giraffids it is very strong (text fig. 7b and d). The coracoid process is strongly developed in Prolibytherium and weakly developed in the giraffids indicating that in Prolibytheriwm the coraco-brachialis muscle was more highly developed than in the giraffids. The neck of the scapula indicates that the bone was shorter and wider than that of Okapia and probably resembled the scapula of Ovis.

Humerus. The coronoid fossa is much deeper than in Okapia or Giraffa; this fossa provides origin for the extensor carpi radialis and the common digital extensor muscles, both of which were strongly developed in Prolibytherium. The lateral condyle is narrow with a transversely concave face, this concavity is caused by very strong lateral and intermediate ridges. The intermediate ridge separates the

Fic. 7. The glenoid cavity and surrounding region of the left scapula. (a) Prolibytherium (B.U.20153) Natural size. (B) Okapia 0-5 x Natural size. (c) Zarafa (B.U.20123) 0:5 x Natural size. (p) Givaffa, 0-25 x Naturalsize. a: glenoid cavity. b: glenoid notch. c:coracoid process. d: tuber scapulae.

GEBEL ZELTEN, LIBYA 119g

condyle from the synovial fossa and all the ridges of the distal head of the humerus are much stronger than in Okapia. The olecranon fossa is very deep and the articular region extends further into the fossa than in Okapia. The medial epicondyle is very strong, it is produced distally as a large process overhanging the medial condyle (text fig. 8a). The medial condyle of Okapia is more weakly developed than that of Prolibytherium and slopes proximally from the condyle (text fig. 8b). The medial, distal and lateral faces of this epicondyle are very heavily sculptured in Prolibytherium and the main digital flexor muscles have areas of origin on this epicondyle. The lateral epicondyle of Prolibytherium is slightly stronger than that of Okapia, this region provides origin for the ulnaris lateralis muscle.

Radius. The central part of the proximal face of the radius is deeply excavated posteriorly for the insertion of the interosseous ligament. This excavation is about the same size in Prolibytherium and Okapia (text fig. 9). The postero-proximal region of the bone is sculptured for attachment to the ulna and this region bears two wide concave facets which are continuous with the proximal facets and articulate with the ulna. The lateral part of the proximal region forms a strong lateral tuberosity providing insertion for the lateral ligament of the elbow and also origin for the common lateral digital extensor muscles. The radial tuberosity, about the same size in Prolibytherium and Okapia, provides insertion for the biceps brachii muscle. In Prolibytherium it is situated at some distance distal to the articulation in contrast to Okapia in which the tuberosity is very near the articulation (text fig. 9).

The shaft of the radius is slender and curved anteriorly as in Okapia and Capreolus. The distal end is transversely narrower than in Okapza; it has the usual articulations

Fic. 8. Medial view of the distal end of the humerus. (a) Prolibytherium (B.U.20154a). Natural size. (B) Okapiao-5 x Naturalsize. a:medialepicondyle. b: medial condyle. c: shaft of humerus.

120 LOWER MIOCENE RUMINANTS

forthe carpals. The area lateral to the cuneiform facet is sculptured indicating a very close attachment of the ulna in this region. In Prolibytherium the posterior face of the radius is convex in the region immediately proximal to the distal facet; this contrasts with Okapia in which this face is concave. The lateral tuberosity of the distal end is strong and the medial tuberosity is stronger in Prolibytherium than in Okapia; these tuberosities provide insertion for the ligaments of the carpal joint.

Ulna. The olecranon process is flattened and though it is broken off distally it was probably aslong as that of Okapia. Theshaft of the ulna is also flattened transversely and bends medially at its distal end asin Okapia. The interosseous space is short and relatively narrow. In Prolibytherium the articular facet is convex transversely, in contrast to Okapia in which the facet is flat transversely. The greater convexity of the facet reflects the deeper synovial fossa and stronger ridges of the distal end of the humerus. The facet of the olecranon in Prolibytherium extends further dorsally than in Okapia resulting in a more elongate facet and at full extension the olecranon fits into the deep olecranon fossa and locks the elbow.

Metacarpal. The distal end of the metacarpal has a large nutrient foramen in the mid-line proximal to the condyles asin Zarafa. The distalend of the boneis wideand indicates that the metacarpal was relatively stout. The condyles are of the usual ruminant pattern with strong keels as in the cervids.

Tibia. In Prolibytherium the proximal head of the tibia is narrower than that of Okapia but is otherwise similar in general anatomy. The cnemial crest is higher in Prolibytherium than in Okapia or Zarafa (text fig. 10) resembling that of Capreolus. The medial face of the cnemial crest is plane and the anterior region of the crest is

Fic. 9. The radius of Prolibytherium. (A) The anterior face of the proximal region of the radius in Prolibytherium (B.U.20154b). Natural size. (B) The anterior face of the proximal region of the radius in Okapia. 0-5 x Natural size. (c) The proximal articular facet of the left radius in Pyolibytherium. Natural size. (p) The proximal articular facet of the left radius in Okapia. 0:5 x Natural size.

GEBEL ZELTEN, LIBYA 121

heavily sculptured. The anterior tuberosity of the crest is wide and similar to that of Capreolus. The lateral face of the crest forms a deep elongate concavity which is deeper than that of Okapia or Capreolus.

The distal end of the tibia is narrow transversely. The medial articular groove is very deep and the intermediate ridge is high resembling that of Zarafa or the cervids. The anterior edge of this ridge continues as a high anterior process. The synovial fossa is large and extends laterally and medially into the articular grooves. The articulation for the lateral malleolus consists of a wide posterior facet and a smaller anterior facet asin Zarafa. The anterior and posterior facets are separated by a deep groove for the tendons of the lateral extensor muscles of the foot. The medial malleolus is heavily sculptured and as strong as in Okapia.

Calcaneum. The calcaneum differs in minor details from that of Palaeomeryx but in major features it is more similar to Palaeomeryx than to any other giraffoid.

Astragalus. In agreement with the calcaneum the astragalus of Prolibytherium

Fic. 10. The proximal head of the right tibia. (a) Prolibytherium. (B.U.20159 and B.U.20160). Natural size. (B) Okapia 0-5 x Natural size. (c) Zavafa (B.U.20146) 0-5 x Naturalsize. (D) Capreolus Natural size. a:medialcondyle. b: lateral condyle. c: tibial spine. d: cnemial crest.

122 LOWER MIOCENE RUMINANTS

resembles that of Palaeomeryx in all its main features. Both the astragalus and calcaneum contrast strongly with those of Okapia but this is mainly due to the specialized nature of the ankle in Okapia.

Metatarsal. This bone is larger than the metacarpal; the reverse of the condition in Okapia and Giraffa but similar to the cervid condition. The distal end is more flattened antero-ventrally than in Okapia or Zarafa. The anterior face of the bone carries a deep channel which is open and passes between the condyles, this channel is very restricted in the region just proximal to the condyles and the walls of the channel are sculptured as in Zavafa. The condyles are similar to the cervids or Zarafa.

Phalanges. The phalanges identified as Prolibytherium are smaller than those of Zarafa but are otherwise similar in all details and reference should be made to the description of Zarafa (p. 97).

THE VERTEBRAL COLUMN

Allas. The atlas of Prolibytherium (text fig. 11c and d) is identified by its size which agrees with the size of the occipital condyles. The general form of the atlas is similar to that of other ruminants. The anterior articular facets are very deep and their lateral curvature is sharper than in Okapza, agreeing with the condition of the occipital condyles. The wing of the atlas is broken off. The intervertebral and alar foramina are closely associated ventrally and dorsally they have a common opening (text fig. IIc) as in Capreolus but in contrast to Okapia where they are widely separated. The ventral face of the atlas is similar to that of Okapia and the ventral tubercle is very strong (text fig. 11d). The posterior articular facets are swollen medially and concave laterally, differing from Okapza in which the facets are more nearly plane. In Prolibytherium the posterior articular facet extends further laterally than that of Okapia and as in Giraffa, it is continued onto the postero-lateral process of the wing. Although this process is broken off it appears to have been as strong as that of Givaffa, indicating a strengthening of the atlanto-axis articulation in Prolibytherium. The bone of the atlas is very thick.

Cervical vertebrae. The centrum of the seventh cervical vertebra (text fig. I1a) is stouter than that of Okapiza and has a strong ventral ridge. The anterior end of the centrum is displaced dorsally relative to the posterior end but less so than in Okapia. On the posterior face the lateral facet for the head of the rib is wider and more clearly defined than in Okapfia. The posterior articular process is lateral only and the articular facet is more elongate in Prolibytherium than in Okapia (text fig. I1a). The anterior articular process is similar to that of Okapza with similarly shaped facets. The transverse processes and neural spine are broken off.

Thoracic vertebrae. The centrum of the first thoracic vertebra is very short and stout. The anterior facet of the centrum (text fig. 11g) is shallowly convex and the facets for the head of the rib (text fig. 11f) face much more anteriorly than in Okapza. The posterior facet of the centrum is shallowly concave with wide lateral facets for the heads of the ribs (text fig. 11e). The ventral face of the centrum is not greatly swollen. A small tubercle is produced at the anterior end of the ventral face and the paired posterior tubercles are very laterally situated. The transverse process is

GEBEL ZELTEN, LIBYA 123

strong and the saddle shaped articulation for the tubercle faces ventrally (text fig. 11f) whereas in Okapia it faces antero-ventrally. The anterior articular process is similar to that of Okapia, with wide lateral articulations extending further medially thanin Okapia. The articulation of the posterior process is median in Prolibytherium and the paired facets are well defined (text fig. 11e); they are elongate, narrow and opposed at an acute angle as in Givaffa. The neural spine slopes more posteriorly than in Okapia, it is the same relative length in Prolibytheriwm and Okapia. The

Fic. 11. The vertebrae of Prolibytherium. (a and B) Lateral and posterior views of the seventh cervical vertebra. (B.U.20171). (c and D) Lateral and posterior views of the atlas (B.U.20170). (£, F and G) Posterior, lateral and anterior views of the first thoracic vertebra. (B.U.20172). All 0-5 x Natural size. Fine stippling indicates articular regions. a: neural spine. e: common dorsal opening of the intervertebral and alar foramina. g: transverse process. h: anterior articular process. k: tubercular facet. 1: posterior facet for the head of the rib. m: posterior articular facet. n: anterior face of centrum. o: posterior face of centrum. p: ventral tubercle.

124 LOWER MIOCENE RUMINANTS

cross-section of the neural spine is triangular with thickening of the posterior edge and heavy sculpturing in the region just dorsal to the posterior articular process.

A single vertebra is known from the posterior thoracic region of Prolibytheriwm, this is probably the tenth or eleventh thoracic vertebra. The neural spine is stronger than in Okapia and it slopes more steeply posteriorly. Both faces of the centrum and the anterior and posterior articular processes of Prolibytherium are similar to those of Giraffa.

Functional interpretations. Features of the fore-limb bones of Prolibytherium indicate that the locomotory muscles were very strong and a requirement for strength rather than speed is indicated. The animal had relatively short limbs probably not unlike those of Ovis. The presence of large ossicones would have necessitated great strength in the forelimbs and if the ossicones were used in intraspecific combat, as seems likely, then this requirement for strength would be exaggerated.

Details of the skull and atlas vertebra indicate that great forces were encountered by the head of Prolibytherium. The lower cervical region is represented by the seventh cervical and first thoracic vertebrae. If the neural spine of the seventh cervical vertebra was vertical during life, then the neck extended almost horizontally from the shoulders and it is likely that the head did not rise much above the level of the shoulders. The stoutness of the centra of the lower neck vertebrae and the shortness of the atlas indicate that the neck was short and probably very stout.

TABLE 9 Measurements of Prolibytherium magnieri, post-cranial material Vertebrae B.U.2z0170 B.U.zo17r B:U.201%72” “B:U-zor7z3 Length of centrum 28 mm 30 mm 23 mm 27 mm Total height of vertebra 34 mm — 102 mm — Depth of centrum — 19 mm 18 mm 16 mm Width of anterior end of centrum = I5 mm I9 mm 22 mm Width of posterior end of centrum — 27 mm 29 mm 30 mm Scapula B.U.20153 Width of neck of scapula (Minimum) 25 mm Depth of neck of scapula (Minimum) 14mm Width of glenoid (Anterior-Posterior) 30 mm Depth of glenoid (Transverse) 25 mm Humerus B.U.20154a Distal end Width across epicondyles 35 mm Antero-posterior depth of distal articular surface 18 mm Radius B.U.20154b

Width of proximal articular surface 35 mm

GEBEL ZELTEN, LIBYA 125

TABLE 9g (cont.)

Depth of proximal articular surface:

Minimum 13 mm Maximum 18 mm Total length of bone 210 mm Width of distal articular surface 29 mm Depth of distal articular surface 21 mm Metacarpal B.U.20155 Width of distal end 27 mm Depth of distal end 15 mm Phalanges B.U.20156 3B.U.20157 B.U.20158 Length 37 mm 37 mm 33 mm Width of proximal end 13 mm I4 mm 13 mm Depth of proximal end I4 mm 14 mm I4 mm Width of distal end II mm II mm Io mm Depth of distal end og mm og mm og mm Tibia B.U.20159 B.U.2z0160 B.U.20161 B.U.20162 B.U.20163 Width of proximal end 47 mm 46 mm a — — Depth of proximal end 46 mm 44 mm — — — Width of distal end — — 29 mm 27 mm 30 mm Depth of distal end — — 21 mm 21 mm 23 mm Calcaneum B.U.20164 Total length 70 mm Length of tuber calcis 48 mm Width across sustentaculum I9 mm Depth of tuber calcis 18 mm Maximum depth of bone 28 mm Astragalus B.U.20165 3B.U.20166 B.U.20167 B.U.20168 Maximum length 32 mm 36 mm 33 mm 33 mm Minimum length 26 mm 30 mm 26 mm 25 mm Width proximally 20 mm 20 mm 20 mm 18 mm Width distally 18 mm 20 mm 19 mm 18 mm Maximum depth 15 mm 16 mm 16 mm I5 mm

Family BOVIDAE Gray 1821. Subfamily BOVINAE Gill 1872. Diacnosis: This subfamily is defined by Pilgrim (1939 p. 249).

Tribe BOSELAPHINI Simpson 1945. DiacGnosis: This tribe is defined by Gentry (1970 p. 245).

126 LOWER MIOCENE RUMINANTS

Genus PROTRAGOCERUS Depéret 1887

TYPE SPECIES: Protragocerus chantret Depéret 1887.

MATERIAL: M.26687 An isolated right horn core fragment. B.U.20113 Anisolated left horn core fragment with part of the frontal attached. B.U.20114 An isolated right horn core fragment.

DEscriPTiIon: M.26687 (pl. 13, fig. 1) is the best preserved of the three specimens; part of the orbital region is preserved in the ventro-lateral region of the specimen and in the medial part of the orbital roof a small supraorbital foramen runs dorso- medially for a short distance in the bone, before turning anteriorly to emerge in the antero-medial region of the horn core, as in Protragocerus gluten. The lateral region of the orbit extends slightly more laterally than the lateral face of the pedicle, which results in a weak lateral swelling in this region; however the horn core is more laterally situated than in Protragocerus gluten which may indicate that the Gebel Zelten species is more primitive. The region of the frontals lying between the bases of the horn cores, is slightly higher than the edge of the orbit and in B.U.20113 the area immediately medial to the horn core base is concave as far as the median suture, which is strong and slightly raised.

The horn core is set diagonally on the skull roof and in B.U.20113 it slopes laterally at an angle of about 20° from the vertical; however in M.26687 this angle is smaller. The horn core also slopes posteriorly at an angle of 35° in all three specimens; which is very similar to the posterior slope in Protragocerus gluten (Pilgrim 1937). The surface of the horn core carries many pronounced vertical ridges and grooves which terminate a short distance above the skull roof, there are no cross striations.

The anterior and posterior keels are strong but do not show any signs of the torsion exhibited by more advanced forms. There is no medial keel but the medial face of the horn core is convex antero-posteriorly and a slight swelling may be detected near the middle of the face, this could be an incipient third keel. The medial face is concave proximo-distally, which indicates that the horn cores, though divergent at their bases, may have become more convergent distally. The lateral face is convex proximo-distally at the base but shows slight concavity distally. The cross-section of the horn core agrees closely with that of Protragocerus gluten (Pilgrim 1937) and the agreement with Protragocerus chantret (Thenius 1956) is more distant. On the basis of this evidence the horn cores from Gebel Zelten, although more primitive than other specimens of Protragocerus, do exhibit features which indicate their close relationship and also incipient features which could lead to the characteristics exhibited by the more advanced species of the genus.

In B.U.20113 sufficient of the roof is preserved for an estimate of the width across the orbits to be made, this was certainly not greater than 6-5 cm and not less than 6-0 cm which is much smaller than Protragocerus gluten and indicates an animal in the same size range as the small mandible which has been identified as Gazella sp. The features of the horns and mandible which separate them into different sub- families are very conclusive and the chances that these two specimens could be from the same species are very small indeed.

GEBEL ZELTEN, LIBYA 127

Genus EOTRAGUS Pilgrim 1939

Dracnosis: The genus Eotragus is defined by Pilgrim (1939 p. 137). TYPE SPECIES: Eotragus haplodon (von Meyer) 1846.

Eotragus sp.

MATERIAL: M.26688 A single isolated hor core. M.26689 A single isolated horn core.

DESCRIPTION: The supraorbital foramen penetrates the roof of the orbit and the canal curves inside the bone, to open on the anterior face of the horn core base near the centre of the base. The horns were almost vertical and perhaps slightly con- vergent ; they were situated supraorbitally and were very near the edge of the orbit. The surface of the horn core is excavated by a large number of proximo-distal grooves but transverse grooves are totally absent (pl. 13, fig. 1). A strong keel is present on the posterior face of the horn core; this keel is stronger in M.26689 than it is in M.26688 but as the latter is from an older individual the strength of the keel may have decreased with age. The cross-section of the horn core agrees closely with that of Eotragus sansamensis (Thenius 1952 fig. 5). The lateral face is convex proximo-distally agreeing with Eotragus haplodon (Thenius 1952) and the Gebel Zelten specimens also agree in size with this species ; however the horns were certainly longer in the Gebel Zelten species than in E. haplodon as the opposing faces converge more gradually in the former species.

TABLE IO Horn cores

Protvagocerus Eotragus B.M.26687. B.U.20113 B.M.26688 3B.M.26689 Maximum antero-posterior

thickness of horn core 27°38 mm — 20-6 mm 16-9 mm Maximum antero-posterior thickness of pedicle 25°5 mm 25°7 mm 19'4 mm 16-7 mm Maximum transverse width of horn core 18-0 mm — I77I mm 14-3 mm Maximum transverse width of pedicle 17-7 mm — 1770 mm 15°55 mm

Subfamily ANTILOPINAE Baird 1857

Diacnosis: The subfamily was defined under the name Gazellinae by Pilgrim (1939 P. 30).

Genus GAZELLA De Blainville 1816 Diacnosis: The genus was defined by Gentry (1970 p. 292).

128 LOWER MIOCENE RUMINANTS

Gazella sp.

REMARKS: Species of Gazella are identified as much on the basis of the horn cores and occipital region as by the dentition. The confused state of classification within the genus is being gradually resolved (Gentry 1964, 1966, 1970) but at the moment the material from Gebel Zelten is insufficient to identify the specimen with any existing species or to allow the establishment of a new species.

MATERIAL: M.26685 An isolated left mandible, dentition showing medium wear. M.26686 A left mandibular fragment, Ms partially erupted.

Mandible. The mandible is clearly bovid and the dentition confirms this. For purposes of comparison mandibles of Gazella capricornis and G. pilgrimi were used: with reference to the former Gentry (1970) states:

‘Gazella capricornis (Wagner) is the gazelle of the famous Pikermi lower Pliocene fauna from Attica. ... Its teeth frequently show primitive characters: basal pillars on other teeth than the upper and lower first molars, strong ribs between parastyle and mesostyle of upper molars, medial wall of lower molars not very flattened, fairly shallow mandibular horizontal rami, and probably a longer premolar row.’

The species G. pilgrimi; with which Gentry synonymized G. gaudryi; is a more advanced form from Samos in which the dentition lacks primitive characters.

In the Gebel Zelten species the mandible is deep with the deepest point lying below the anterior end of Mg. The lateral face of the mandible is convex with a strong swelling in the posterior region as in Syluicapra. A small posterior mental foramen lies below the anterior end of Pg. The anterior mental foramen is very large and double as in Sylvicapra. The diastema is short and though the anterior region is missing its length was probably similar to that of Gazella or Sylvicapra. The medial face of the mandible (pl. 13, fig. 8) is convex dorsally but a shallow concavity runs antero-posteriorly along the ventral part of this face. The mandibular foramen is large and below it is a shallow mandibular groove as in Gazella.

Lower dentition. The molars of the Gebel Zelten species are more brachyodont than in G. capricornis or G. pilgrimi and cingula are lacking in all three species. On Ms the mesostylid is as strong as that of G. capricornis but is weaker than in G. pilgrimi. The metaconid is high, transversely flattened and set diagonally on the tooth causing the metastylid to be produced lingually. In G. capricornis the lingual cuspids are parallel to the axis of the tooth. The metaconid has a weak lingual swelling. The metastylidis very weak andislost after medium wear. The entoconid is as high as the metaconid and has a more feeble lingual swelling, it is set more nearly parallel to the axis of the tooth than the metaconid. The protoconid is high and crescentic with no trace of a ‘Palacomeryx fold’, its anterior region joins the mesostylid very early in wear. The posterior end of the protoconid meets the metaconid and entoconid at the lingual end of the median valley. The hypoconid is lower than the protoconid but is less isolated than in the palaeomerycids. The posterior end of the hypoconid joins the entoconid, isolating the hypoconulid which forms the accessory column; this is lower than the hypoconid and is crescentic,

GEBEL ZELTEN, LIBYA 129

resembling closely the accessory column of G. capricornis or G. pilgrimu. Excluding the differences mentioned the Mg of the Gebel Zelten species is very similar to the two European species.

Mg resembles the anterior region of Mg. The mesostylid is stronger in Mz than in Mg (pl. 13, fig. 2), this may be a compensatory strengthening as the metaconid is more nearly parallel to the axis of the tooth in Mg. Mj is more worn but otherwise similar to Mg. Each molar has a strong ectostylid and M3 has a weak stylid in the posterior valley, these stylids are similarly developed in G. capricornis.

The Py, of the Gebel Zelten species is short with a high protoconid which gives off a strong postero-lingual metaconid. The anterior region consists of a paraconid which is stout and directed antero-lingually. A parastyle does not appear to have been developed as the lingual face of the paraconid does not bear a groove (pl. 13, fig. 2). The posterior region is wide and the hypoconid is stout with a strong labial swelling (pl. 13, fig. 2) which is separated from the protocone by a deep labial groove. The entoconid is strong and wide but it is heavily worn and the presence of an entostylid cannot be established; the absence of a lingual groove on the face of the entoconid(pl. 13, fig. 3) indicates that an entostylid was probably not present.

The Py, of G. capricornis is similar to that of the Gebel Zelten species. The hypoconid is strong in G. capricornis and has a strong labial swelling as in Gazella sp.. The metaconid, entoconid and entostylid are similar in the two species but the anterior regions differ as the paraconid and parastylid are widely separated in G. capricornis but are joined in Gazella sp.. The Py of G. pilgrimt is smaller than in Gazella sp. and the hypoconid is more feebly developed. The entoconid and ento- stylid are more widely separated in G. pilgrimi than in Gazella sp. and, as in G. capricornis, the paraconid and parastylid are also widely separated in G. pilgrimt.

The P3 is much shorter than the Py, (table 11); the lingual region is badly broken (pl. 13, fig. 3). The labial wall has a strong hypoconid swelling asin Py. The Ps of G. capricornis is highly developed and is as long as the Py (table 11); its paraconid and parastylid are strongly separated as on the Py, and in contrast to the Pg of Gazella sp. in which the anterior region is not divided. The Ps of G. pilgrimt is smaller than that of Gazella sp. but the paraconid and parastylid are separated as in G. capricornis. The Pe is missing in M.26685 but it appears to have been greatly reduced in length and was probably comparable to the Pe of G. pilgrim: or G. capricorms.

The molars of the Gebel Zelten gazelle are therefore similar to those of the European Pliocene species and they were similar in length and width to those of G. capricornis; however in that species the molars are more hypsodont than in Gazella sp. and even after medium wear the molars of G. capricornis are almost as high as the unworn dentition of M.26686. The molars of G. capricornis are more advanced than those of the Gebel Zelten gazelle. The premolar row is relatively long in G. capricornis and short in Gazella sp. and G. pilgrimi ; thus in this respect Gazella sp. resembles the more advanced G. pilgrimi. It is not possible to place the Gebel Zelten gazelle with either of these species as it shows features which are present in both species and also exhibits some features such as degree of brachyodonty and details of the premolars, which are more primitive than in either species.

LOWER MIOCENE RUMINANTS

130

UII 9-6 wu 1-6 WU 0-8 wut 2.2 wut ¢.Z

%9-oL wut o-.VE

wut 2.1

WIW 9-01

UWI 0-01 wut ¥z

WU 9-9 wut 4.9 wut 9-S

uu L-6 wut S.6

wu +. WU Z-S wut Z.€ UWIU Z-S UIPIM. qysue'] vIokI WE SWUAOIUAGDI *5)

%Wri-SS UIU 1-Z€

wu S-V1

WIW 0.9 UIUI O-O1

uur o.S Ulu 1-6 wut 2.21

wut L.9

wurut €.2 wut Z.¢ wut Z.z wut Z.¢ UIPIM qysueT LLib we wurasyrd “5

wu S.F uur L.€

UIUI 9-9 wut 0.S1

UIPIM. yysueT 93997 Wa “ds vIaz05)

Il alavy

Wu 0-6 wut £-g wut o.Z UIUL 8-9 uu 6.¢

%r-gS wu F-r€

wiul 9-F1

UIUI O-IT

uw S.6 UU OZ

WU 9-9 wut o.Z wut 2.9

wut $.¢ WIUI 9-g wut o-F uur z-Z — wut 0. UIPIM yysue'T Sgo9z' Wd “ds vazvy)

UMOIO jo doz 0} Yoou woz 7YsIoy UNUWIIxe[, (ooI—y}8U9] IeOJA)) sus] Iejpoy 0} IejoulsIg % MOI Ie[OUL JO Y}SUg] [eJOL,

MOL Teyouroid JO Y}Sug] [CIOL

GEBEL ZELTEN, LIBYA 131

The presence in the Zelten fauna of three species of bovids in the early Burdigalian is surprising as, with the exception of European sites and the Hsanda Gol of Asia, bovids are of later occurrence. The presence of three bovid genera representing the two most important subfamilies of the Bovidae indicates that the bovids were already well established in the early Miocene and therefore they must have diverged from their ancestral stock at some time during the Oligocene. The presence of bovids at Hsanda Gol is not yet well established though Trofimov (1968) figures very bovid like lower molars of Palaeohypsodontus and this may indicate that the ancestry of the bovids must be sought in the Oligocene of Asia.

Ill. THE EVOLUTION OF PRIMITIVE GIRAFFOIDS

The Giraffoidea was established by Simpson (1931) to include the Palaeomerycidae and Giraffidae ; Simpson (1945) demoted the Palaeomerycidae to subfamily level and transferred it to the Cervoidea, leaving the Giraffoidea containing the Giraffidae and the Lagomerycidae.

The Palaeomerycidae was established by Lydekker (1883) taking its name from Palaeomeryx von Meyer 1834. Lagomeryx Roger 1904, was established to include small species of Palaeomeryx, bearing branching ossicones. This situation persisted until Roman and Viret (1934) suggested that the two genera were synonymous; this suggestion was supported by Teilhard de Chardin (1939). De Chardin also suggested that a separate family should be created to accommodate Lagomeryx, Procervulus and ‘other so called cervids’. This was left to Pilgrim (1941) who proposed the name Lagomerycidae for a family including the two genera named by de Chardin and Climacoceras which had been placed in the Cervidae (MacInnes 1936).

Stirton (1944) discusses the relationships of the palaeomerycids and states that Lagomeryx and Palaeomeryx are synonymous; he also proposes the retention of the family name Palaeomerycidae expanded to include:

‘, . . the Old World and New World non-antlered but horned and some hornless cervoids. The genera with forked horns from the Old World are included in the subfamily Palaeomerycinae and those from North America with pointed, bulbous or flat tipped horns are assigned to the Dromomerycinae.’ (Stirton 1944).

Whitworth (1958) discusses the affinities of the palaeomerycids at some length, supporting the synonymy of Lagomeryx and Palaeomeryx and indicating the close affinity of palaeomerycids and cervids. This is contested by Ginsburg and Heintz (1966) who challenge the synonymy of Lagomeryx and Palaeomeryx and in their discussion of the genus Palaeomeryx they state:

‘En dehors d’Europe, le genre a été signalé par Teilhard de Chardin (1939) dans le Miocéne de Chine et par T. Whitworth (1958) dans le Miocéne d’Afrique orientale. Dans les deux cas, on peut affirmer qu'il ne s’agit pas de Palaeomeryx ne serait-ce que par la présence, chez deux formes, d’une P}.’

The presence of a P; in Palaeomeryx africanus is questioned by Gentry (1970) but a P; was certainly present in the species described by de Chardin (1939); which must therefore be removed from the genus Palaeomeryx and should revert to its

132 LOWER MIOCENE RUMINANTS

previously applied name of Lagomeryx simpsoni ; however the name Lagomeryx is not applicable as it was established (Roger 1904) to refer to European species which have been synonymized with Palaeomeryx: in this situation the species reverts to its original name—Heterocemas simpsoni Young 1937.

Pilgrim (1941) and de Chardin (1939) indicate close giraffoid affinities for the Palaeomerycidae (Lagomerycidae) and their evidence for this was summarized by Whitworth (1958):

‘(i) the unequal development of anterior and posterior external ribs on the upper molars; (ii) the characteristically corrugated enamel of the teeth; (111) the occurrence of a permanent velvet over the bony parts of the “antlers’’.’

Whitworth objected to the first on the grounds that variation occurs in the strength of the ribs in both giraffids and cervids and he states:

‘In fact, an unequal development of the anterior and posterior external nbs are found, to a varied degree in most cervids and giraffids alike; although generally speaking, the anterior rib is more angular and better defined in the deer than in the Giraffidae.” (Whitworth 1958).

This objection is valid as is his objection to the second piece of evidence on the basis of the great variability of the enamel in the molars of the giraffids and cervids. The third piece of evidence is the strongest and Whitworth’s grounds for rejecting it are correspondingly weak, he states:

‘Finally to regard the possession of non-deciduous, velvet covered ‘‘antlers’”’ as diagnostic of early Giraffoidea is contrary to Pilgrim’s own repeated opinion (1941 and 1947) that this was as likely to be the primitive cervoid condition as the giraffoid.’ (Whitworth 1958).

This distinction is probably basic to the classification of the giraffids, cervids and palaeomerycids and Pilgrim’s statement is given here:

‘.. . the skin covered “horn” was the most primitive and the nearest approximation to it is to be seen in the Lagomerycid and Giraffid “horn”. From this original type it is easy to surmise that the deciduous antlered horn of the Cervidae arose, though the intermediate stage is as yet unknown.’ (Pilgrim 1941).

Whitworth indicated that this could just as easily mean that the palaeomerycids were primitive cervids as giraffids but the ossicones are positive evidence in favour of the affinity of the palaeomerycids and giraffoids, whereas they can only be regarded as evidence for the affinity of palaeomerycids and cervids on purely hypothetical grounds. In this context Ginsburg and Heintz (1966) state:

‘Les ossicones de Palacomeryx offrent, avec ceux des Girafes, de si fortes simili- tudes que nous nous demandons s’il ne s’agit déja d’un Girafide vrai.’

The upper dentition of Zarafa agrees with that of Palaeomeryx in many features but it also agrees with Palaeotragus and the skull shows close affinities with the palaeotragines. Unfortunately the skull of Palaeomeryx is not known but the presence of a dentition with some palaeomerycid features on a skull with giraffid affinities indicates the closeness of the palaeomerycids and the giraffids.

Much of the confusion over the affinities of the palaeomerycids can be attributed to the poorly defined limits of the family. Ginsburg and Heintz (1966) have

GEBEL ZELTEN, LIBYA 133

removed Walangamia africanus and Heterocemas simpson from the genus Palaeo- meryx which limits the genus mainly to the Lower and Middle Miocene of Europe. They also suggested the removal of the Oligocene genera which lack ossicones, from the Palaeomerycidae, transferring them to a position as the common stock from which the giraffids and cervids evolved; this group was named the Dremotheriidae by Ginsburg and Heintz but it is of the same status as the other main groups and should therefore have superfamily status. This regrouping of the early ruminants results in an arrangement as shown (text fig. 12).

The North American members of the Palaeomerycidae were discussed in detail by Stirton (1944) who assigned them to the subfamily Dromomerycinae and indicated an origin for this group which was close to the origin of the cervids and palaeomery- cines. Simpson (1945) treated the Dromomerycinae as a subfamily of the cervidae; while Crusafont (1952) placed the tribe Blastomerycini in the Palaeomerycinae and transferred the Dromomerycinae to the Giraffoidea. Ginsburg and Heintz (1966) proposed the inclusion of the blastomerycines in the Dremotheriidae. The blastomerycines are small, hornless pecorans, at the same level of evolution as the

OLD WORLD NEW WORLD LOWER Cervoidea Dromomerycidae MIOCENE Blastomerycidae UPPER Giraffoidea OLIGOCENE MIDDLE Dremotheriidae OLIGOCENE LOWER OLIGOCENE UPPER Gelocidae EOCENE

Fic. 12. The interrelationships of the early, non-bovid ruminants.

134 LOWER MIOCENE RUMINANTS

dremotheriids and, with the exception of Eumeryx, they are an entirely New World group. At this level of pecoran evolution the forms are so similar that division into groups is difficult and it is proposed that the Dremotheriidae and Blastomerycidae should be treated as separate families of the Dremotherioidea. The Dromomerycidae probably originated from the Blastomerycidae.

Eumeryx culminis from the Hsanda Gol Formation was identified (Matthew and Granger 1924) as a blastomerycine and was so classified by Simpson (1945), it is the only blastomerycine from the Old World and is of Lower Oligocene age whereas the other blastomerycines are of Lower Miocene age. Stirton (1944) briefly discussed the position of Eumeryx and he states:

‘As to whether Ewumeryx is placed in the Cervoidea or Traguloidea is more or less arbitary but recognition of its pre-cervoid characters is quite important.’ (Stirton 1944).

In view of the great difference of distribution in space and time and the difficulty of establishing relationships purely on the dentition, at this level of ruminant evolution; the affinity of Eumeryx with the other blastomerycines is doubtful.

The Pecora probably originated from the Traguloidea during the Upper Eocene or Lower Oligocene and of the two traguloid families the Gelocidae are the most likely to have given rise to the Pecora. In the gelocids true selenodonty is developed from more bunodont forms; thus Lophiomeryx has very bunoid lower molars showing few signs of true selenodonty while Bachithertum and Prodremother1um have molars which are very similar to those of Dremotherium. A detailed study of this group is needed and it is here that the divergence of the Bovoidea and other higher ruminants probably occurred.

The system shown (text fig. 12) appears more natural than any yet proposed for the evolution of the early, non-bovid ruminants. The members of the Dremother- ioidea appear to form a natural unit and Palaeomeryx with its closely allied genera are included in the Giraffoidea. The whole pecoran classification at this level is in a state of flux as it relies upon very small differences and only the discovery of more localities and more complete material can resolve some of the outstanding problems.

IV. THE EFFECT OF THE OSSICONES ON GIRAFFOID EVOLUTION

With the exception of size, the most important differences between Zarafa and Prolibytherium are related to the frontal bones, ossicones and occipital region. The ossicones of Zarafa were developed supraorbitally and probably projected dorso- laterally asin Samotheriwm sinense (Bohlin 1926 pl. VI.). Each ossicone was probably a simple, conical projection about the same size as in Okapia. The ossicones of Prolibytherium are well known (pl. 7). If the origin of the giraffoids lies in the middle or late Oligocene then the divergent trends in the development of the ossicones must have been established very early to allow such great differences to evolve by the early Miocene; these trends may be revealed by the ossicones.

Frontal appendages are developed in the ruminants primarily for intraspecific combat and certain rules may be applied to their development; these were formu- lated by Geist (1965). Fighting in Givaffa (Innis 1958) and Okapia (Walther 1960

GEBEL ZELTEN, LIBYA 135

and 1962) consists of lateral display and the delivery of heavy blows to the sides and neck of the opponent with the head used as a club. Lateral display is the most primitive combat method used in the ruminants and is also found in the tylopods. With this combat method strong crushing or bending forces are not experienced along the length of the neck and therefore any tendency to increase the length of the neck is not inhibited by behavioural factors; indeed increase in length of the neck may confer a small selective advantage in combat as the length of the swing and therefore the force of the blow will be increased. This slight advantage would rein- force advantages resulting from other aspects of the animals behaviour such as feeding habits. With this combat method the force of blows delivered is on the lateral region of the frontals and therefore if a protuberance is developed it will be primitively in the supraorbital position; also head to head contact will not usually occur and a skin covered appendage will be effective and will retain its skin covering at least over most of its surface as in Givaffa or Okapia. Apart from the extant girafids many of the palaeotragines and giraffines were long necked relatively slender animals, with small ossicones and it is likely that lateral combat methods were practised by all of them.

In Prolibytherium the ossicones consist of a solid outer layer of bone and a cancellous interior; frontal sinuses are not present. The derivation of the ossicones in Proliby- therium cannot be established but certain features of the skull may indicate their mode of origin. The anterior palmation projects antero-laterally from the supraorbital region and it is possible that the anterior region was derived from a supraorbital tine similar to that of Zarvafa. At the posterior end of the ossicones the presence of paired pillars (pl. 10, fig. 2) suggests the derivation of this region from paired supraoccipital ossicones. The occurrence of supraorbital and parietal ossicones in other sivatheriids and palaeotragines indicates that giraffoids have the ability to develop ossicones in both regions.

The ossicones of Prolibytherium are the earliest advanced frontal appendages known in the ruminants. Geist (1965) relates the evolution of large frontal appendages to intraspecific combat involving frontal or head to head attack with wrestling and pushing between the heads. Geist states:

“.. . cervid antlers should be regarded primarily as structures binding opponents together during pushing and wrestling matches.’

Thus the ossicones of Prolibytherium present a stage of evolution which is more advanced than that of Zavafa. The ossicones would also function in threat postures with reference to which Coope (1968) states:

‘I believe that initially the “‘pedicels’’ evolved as threat display structures increasing the apparent surface area of the face and thus its deterrent value.’

Geist and Coope agree that the frontal appendages function in frontal display and attack when they are large and the effectiveness of the ossicones of Prolibytherium in increasing the apparent surface area of the face cannot be questioned.

Morphological and functional differences between the ossicones of Zarafa and Prolibytherium may suggest reasons for the divergence between the two main lines of giraffoid evolution. The selective advantage in intraspecific combat conferred by the large ossicones of the sivatheriids was a strong evolutionary force maintaining

136 LOWER MIOCENE RUMINANTS

the large ossicones and also the short neck and limbs necessary for their effective use. The more primitive combat methods of the palaeotragines and giraffines tended to increase the selective advantage of a long neck and, more important, made possible the elongation of the neck under other selective forces.

V. A CLASSIFICATION OF THE GIRAFFOIDEA

The discovery of Prolibytherium and Zarafa in the Burdigalian of North Africa indicates that the origin of the giraffoids must lie in the late Oligocene and that divergence within the group occurred at this time. This was also stated by Ginsburg and Heintz (1966):

‘On peut se demander si les soi-distant Cervides sans bois de la fin de l’Oligocene (Amphitragulus, Dremotherium, Blastomerycines) ne representant pas, non des Cervides primitifs, mais le stock commun d’ou sortiront a l’epoque suivant les Cervides, d’une part, et les Giraffoidea, d’autre part.’

Features of Zarafa and Prolibytherium indicate divergence very soon after their origin from the pregiraffoid stock and the establishment of a separate family to accommodate the sivatheres is proposed. The evolution of the giraffines and palaeo- tragines has followed similar trends and the divergence of these two groups probably did not occur until the middle Miocene, they are therefore retained as subfamilies within the family Giraffidae. This results in a classification of the Giraffoidea as summarized in text figure 13.

Superfamily GIRAFFOIDEA Simpson 1931 Family PALAEOMERYCIDAE Lydekker 1883

Canthumeryx New genus. L. Miocene; Africa. Climacoceras MacInnes 1936. M-—U. Miocene; Africa. Heterocemas Young 1937. U. Miocene; Asia. Palaeomeryx Von Meyer 1834. L. Miocene; Africa. M-U. Miocene; Europe. Procervulus Gaudry 1878. L—M. Miocene; Europe Propalaeoryx Stromer 1926. L. Miocene; Africa. Triceromeryx Villalta, Crusafont and Lavocat 1946. L. Miocene; Europe. Incertae sedis. Progivaffa Pilgrim 1908. L. Miocene; Asia.

The Palaeomerycidae. At present the family is best regarded as a level of ruminant evolution equivalent to the primitive representatives of the Giraffidae and Sivatherii- dae. This system was used quite drastically by Stirton (1944) who drew straight lines across his phylogenetic diagram to indicate levels of evolution and the limits of the family.

The genus Tviceromeryx is placed in the Palaeomerycidae as it shows close similari- ties to the members of the genus Palaeomeryx from Sansan. The presence of a

GEBEL ZELTEN, LIBYA 137

supraoccipital horn sets it apart from the other palaeomerycids and, owing to its strange nature, from all the sivatheriids and giraffids; however the material of Triceromeryx can be interpreted differently and it is possible that the supraoccipital ossicone does not belong to Tviceromeryx, this is also suggested by Churcher (1970). The lower dentition of Tviceromeryx is very palaeomerycine in form and each molar bears a strong ‘Palaeomeryx fold’. The P4 is almost indistinguishable from a Py, of Palaeomeryx sansaniensits (M.5409). The presence of a ‘Palaeomeryx fold’ is not evidence of palaeomerycid affinities as such a fold is also present in some specimens of Palaeotragus and Honanotherium but in these genera the occurrence of the fold is irregular. The genus Tviceromeryx is endemic to the Iberian peninsula and certainly does not effect the evolution of the advanced giraffoids.

The main problems relating to this family cannot be resolved until an exhaustive review of the European material is made. It is unlikely that the palaeomerycids gave rise to either of the other giraffoid families or to the cervids. They represent a Miocene expansion of the ruminants into an ecological niche which was later filled by the palaeotragines and cervids. Inter-relationships within the family are not known.

LOWER PLIOCENE Palaeotraginae Giraffinae Sivatheriidae UPPER MIOCENE MIDDLE Giraffidae MIOCENE LOWER MIOCENE Palaeomerycidae arafa Prolibytherium UPPER OLIGOCENE Dremotheriidae MIDDLE OLIGOCENE

Fic. 13. The early evolution of the Giraffoidea.

138 LOWER MIOCENE RUMINANTS

Family SIVATHERIIDAE New family

Birgerbohlinia Crusafont 1952. L. Pliocene; Europe Bramatherium Falconer 1845. M. Pliocene; Asia. Helladotherium Gaudry 1860. L. Pliocene; Europe, Asia, and N. Africa. Hydaspitherium Lydekker 1878. M. Pliocene; Asia. Libytherium Pomel 1893. U. Pliocene and Pleistocene; Africa. Prolibytherium Arambourg 1961. L. Miocene; Africa. Sivatherium Falconer and Cautley 1835. U. Pliocene and Pleistocene; Asia.

SIVATHERIIDAE: Prolibytherium is the earliest representative of this family but even in this genus the ossicones are highly developed and the cheek teeth show hypsodont tendencies. The retention of a lacrymal fossa indicates the primitive nature of the genus. The family is characterized by its short neck and limbs, and also by the large ossicones; these are probably related features but they also indicate that members of the family fed near the ground and grazing forms may have developed. Meladze (1964) suggests that the family reached its climax in the late Miocene but I would place this climax in the Pliocene when the group was represented by fairly abundant, massively built forms throughout the Old World. The African genus Libytherium is probably the only Upper Pliocene and Pleistocene genus from Africa and with it may be synonymized such forms as Griquatherium Cooke and Wells 1947 and Ovangiotherium van Hoepen 1932. The genus Helladotherium from Pikermi, is problematical as it lacks ossicones, indicating that it was probably the female form: with reference to this genus Matthew (1929) states:

‘It appears not at all improbable that Helladotherium may be the female of Bramatherium or Hydaspitherium. The teeth are indistinguishable, and the skulls are by no means as diverse in degree, but differ in the same manner, as Sivatherium and “Indratherium” of the Upper Siwaliks.’

No further work has been done on this problem and Helladotherium is here included in the family as a valid genus, with the qualification that it is probably a female form.

Meladze (1964) also suggests that the sivatheriids were adapted to life in the savannahs but I suggest that they were probably woodland or forest forms, feeding on low vegetation or grasses of the woodland floor. The family was very successful and it may have survived in Asia to sub-recent times as suggested by Colbert (1936).

Family GIRAFFIDAE Gray 1821 Subfamily PALAEOTRAGINAE Pilgrim 1911

Givaffokeryx Pilgrim 1910. L. Pliocene; Asia. Okapia Lankester Ig01. Pleistocene and Recent; Africa. Palaeotragus Gaudry 1861. U. Miocene; Europe, Asia, and Africa. L. Pliocene; Europe and Asia.

GEBEL ZELTEN, LIBYA 139

Samotherium Major 1888. U. Miocene; Africa. L. Pliocene Europe and Asia. Zarafa New genus. L. Miocene; Africa.

Subfamily. GIRAFFINAE Zittel 1893.

Bohlinia Matthew 1929. L. Pliocene; Europe.

Decennatherium Crusafont 1949. L. Pliocene; Europe.

Givaffa Brisson 1756. Pliocene; Europe and Asia. Pleistocene; Asia and Africa. Recent; Africa.

Honanotherium Bohlin 1926. Pliocene; Asia.

GIRAFFIDAE: A group of late Oligocene origin with Zarafa as the earliest known genus. The family is characterized by a tendency to increase the length of the neck and limbs. The giraffids reached their climax in the early Pliocene when they were well represented throughout the Old World. It was probably during the middle Miocene or slightly earlier that the group divided into two subfamilies; the palaeo- tragines and giraffines.

The palaeotragines were in many ways very progressive and the development of a hypso-brachyodont dentition in later members of the genera Palaeotragus and Samotherium indicates that they probably fed upon fairly low vegetation with a high proportion of ground vegetation in their diet. In any case they were utilizing an intermediate zone between the bovids and giraffines and in times of scarcity it is this zone which the other two groups would utilize at the expense of the palaeotragines. The subfamily has one extant member, the okapi which has survived in a tropical forest environment.

The giraffines have been relatively successful since the Pliocene. The series Honanotherium, Bohlinia, Givaffa suggested by Bohlin (1935) seems to be a natural sequence and is probably the true relationship.

VI. EAST AFRICAN RUMINANTS

During this work reference has been made to the Miocene ruminants of East Africa which were described by Whitworth (1958). A brief review of these ruminants has been made as I disagree with some of Whitworth’s identifications. These ruminants are treated as a separate section as they are nearly all previously described forms and all depend upon previously described material; also it was felt that their inclusion with the Libyan forms would confuse the description of a regionally defined ruminant group.

Superfamily TRAGULOIDEA Gill 1872 Family TRAGULIDAE Milne-Edwards 1864 Genus DORCATHERIUM Kaup 1833

The agreement between Dorcatherium chappuisi from Rusinga and Turkana (Arambourg 1933) is very close and there is no doubt that these specimens are from

140 LOWER MIOCENE RUMINANTS

the same species. The other three species described by Whitworth (1958) are D. pigotti, D. parvum and D. songhorensis ; these are distinguished on a size basis only and with reference to these species Whitworth states: ‘Further collecting may show that D. songhorensis is synonymous with D. parvum or D. pigotti, but for the present it seems preferable to treat the Songhor material separately.’ (Whitworth 1958).

I agree with Whitworth on this point, there are certainly two smaller species of Dorcatherium and as further collections have not been made it is not possible to synonymize D. songhorensis with either of these.

The genus Dorcatherium is the only traguloid genus previously identified from the Miocene of Africa but included in the material identified as Palaeomeryx africanus (Whitworth 1958) are a few lower molars which differ anatomically from the type specimen. These molars exhibit traguloid features and agree closely with the lower molars of Gelocus.

Superfamily TRAGULOIDEA Gill 1872 Family GELOCIDAE Schlosser 1886.

DiacGnosis: Traguloids in which true selenodonty is developed; strong cingula developed on the labial or lingual sides of the cheek teeth. Metapodials fused to form cannon bones and side toes reduced. Lower premolars very simple. Pj reduced and peg-like, separated from Pz: by a short diastema (After Schlosser 1886).

Genus GELOCUS Aymard 1855

Diacnosis: A medium sized gelocid in which the selenodonty is less advanced than in most members of the family. Metaconid rounded anteriorly but less so than in Lophiomeryx. Premolars very simple. (Mainly after Schlosser 1886).

Gelocus whitworthi sp. nov.

Diacnosis: A medium sized species of Gelocus possessing a rounded metaconid on the lower molars; the median valley of the lower molars is very open lingually. Length of lower molar row about 33 mm.

SYNONYMY: One specimen of this species—K.Sgr.368.49—was described by Whitworth (1958) with Palaeomeryx africanus.

DERIVATION OF NAME: The species is named after Dr. T. Whitworth who produced the first definitive account of African Miocene ruminants.

Ho.otyPe: K.Sgr.365.1949—a left mandibular fragment with lightly worn Me and Ms. From Songhor, Nyanza Province, Kenya. All specimens of this species are the property of the National Museum, Kenya.

GEBEL ZELTEN, LIBYA 141

LOCALITY AND HORIZON: Specimens are from Songhor and Rusinga Island Kenya. Both sites are referred to the Miocene.

MATERIAL: K.Sgr.265.1949 Holotype; a left mandibular fragment with lightly worn Me and M3. K.Sgr.368.1949 An isolated M, showing slightly heavier wear than the holotype.

K.Sgr.581.1949 An isolated right Mg showing light wear. K.Sgr.159.1949 An isolated right Mz showing medium wear. K.R.30 An isolated right M3, showing medium wear.

DeEscriPTIoNn: The lower molars are almost the same size as Walangania africanus. On the Mz the metaconid is conical with an anterior crest curving antero-lingually to meet the mesostylid at the antero-lingual corner of the tooth; this results in the concavity of the antero-lingual face of the metaconid similar to Gelocus communis but differing greatly from Walangania in which the metaconid is selenodont. The postero-lingual face of the metaconid is rounded with a feeble metastylid but lacking a fold such as is present in Dorcatherium. In Walangania the metastylid lies lingual to the anterior end of the entoconid whereas in Gelocus it is closely joined to the metaconid and lies lingual to the posterior end of the metaconid; as a result the median valley is very open lingually (pl. 13, fig. 5). The entoconid is conical with a strong anterior crest. The posterior end of the entoconid is forked and the labial branch meets the postero-lingual end of the hypoconid while the lingual branch is produced postero-lingually. This results in a strong vertical groove on the posterior face of the cuspid (pl. 13, fig. 4); a similar groove is present in Gelocus communis in which the lingual branch meets the postero-lingual extension of the hypoconid and the labial wing joins this part of the hypoconid. No such groove is present in Lophiomeryx, Bachytheriwm or Prodremotherium.

The protoconid is crescentic and extends antero-lingually around the anterior end ot the metaconid; causing the anterior fossette to open lingually (pl. 13, fig. 4) but not to the same extent asin Lophiomeryx. The posterior region of the protoconid extends lingually and joins the postero-labial face of the metaconid from which a wing is produced as in Walangania and Palaeomeryx; however in Gelocus this wing is much more anteriorly situated resulting in a shorter anterior fossette and a longer median valley region. The anterior wing of the entoconid also meets the protoconid in this region. The hypoconid is crescentic and lower than the protoconid. Me has strong anterior and posterior cingula.

The metaconid of M3 is more feeble than on Ms and the anterior crest is weaker. In Lophiomeryx the anterior end of the metaconid is extremely shortened and rounded while in Gelocus it bears a strong crest and is truly selenodont. The entoconid is more selenodont in M3 than Mg, this is mainly due to the presence of a strong crest in the posterior region. This crest extends posteriorly and lies lingual to the postero-lingual end of the hypoconid (pl. 13, fig. 4). The posterior end of the entoconid is not forked as in the Mg; this region is similar to Gelocus communis but differs from Lophiomeryx in which the entoconid has only a feeble posterior crest.

142 LOWER MIOCENE RUMINANTS

The protoconid is crescentic and its antero-lingual extension is stronger than in the Mg; resulting in the posterior shift of the lingual opening of the anterior fossette in the M3 relative to the Mg (pl. 13, fig. 4). The hypoconid is similar to that ot the Mz but the posterior region is shorter in the Mg as it joins the posterior extension of the hypoconid. Theaccessory column consists of a strong entostylid running postero- labially from the posterior end of the entoconid and fusing with the strong hypoconu- lid. The hypoconulid is crescentic and curves postero-labially from the face of the hypoconid (pl. 13, fig. 4). Atits posterior end the hypoconulid turns antero-lingually and almost encircles the entostylid. M3 has a strong anterior cingulum and both M2 and Mz have strong entostylids in the median valleys. Mg also has a stylid at the labial end of the posterior valley.

Features of their anatomy warrant the specific but not generic separatioa of these specimens from other species of Gelocus. Their presence in East Africa is important as representatives of the family Gelocidae are previously unrecorded from Africa.

TABLE 12

The Lower Dentition of Gelocus whitworthi

My Me M3 Length Width Length Width Length Width K.Sgr.365.1949 = —_ to-8 mm 67mm 14'7mm 6-9 mm K.Sgr.368.1949 975 mm 5:9 mm — — — — K.Sgr.581.1949 — — — — 14-5 mm 6-8 mm K.Sgr.159.1949 — oo 10-4 mm 6-7 mm — — K.R.30 — — — — 15-0 mm 6:38 mm

Family PALAEOMERYCIDAE Lydekker 1883 Genus PROPALAEORYX Stromer 1926

DiaeGnosis: ‘Pecora of medium size, with shallow mandible and rather brachyodont, selenodont lower cheek teeth, closed from Pz: to Mg; P isolated by a very short diastema. Enamel wrinkled. Lower molars with very strong metastylid and entostylid; pronounced median rib on lingual surface of metaconid, similar rib on entoconid ; accessory stylid in median, external valley developed to varying degree.’ (Whitworth 1958).

TYPE SPECIES. Pyvopalaeoryx austroafricanus Stromer 1926; from the Lower Miocene of South West Africa.

Propalaeoryx nyanzae Whitworth 1958

Dracnosis: ‘A species of Propalaeoryx with lower molar series (Mj-3) measuring about 45 mmin length. All lower molars have prominent accessory tubercle in the median, external valley. Teeth rather lower crowned than in type species, P. austroafricanus.’ (Whitworth 1958).

GEBEL ZELTEN, LIBYA 143

Ho.otyPpe: M.21368 (K.324.47) Figured Whitworth 1958, fig. 12. A fragment of left mandible with Mi and Mg showing slight wear.

Horizon: From the Lower Hiwegi Beds, Rusinga Island. Lower Miocene.

MATERIAL: M.21368 Holotype. K.1263.51 — Isolated left Me. K.614.49 Anterior region of right M3. K.774.52 Right mandibular fragment with Mz and M3. K.780.52 Newly erupted left Ms. K.193.51 Heavily worn left M2. K.Mt.21 Lightly worn right M1. eX. Left maxillary fragment with M1 and M2. K.Mt.67.51 Fragment of left maxilla with P4 and M!. (Listed as Palaeomeryx africanus by Whitworth 1958; table X.). K.246.59 Heavily worn left M1.

DeEscrIPTION: Whitworth (1958) described the lower dentition only and upper molars here identified as Propalaecoryx nyanzae agree with the lower molars in details such as size, brachyodonty, degree of rugosity, depth of fossettes and strength of labial ribs and styles. The lower dentition is redescribed as it agrees closely with Canthumeryx.

Lower dentition. The enamel of the molars is finely rugose. Moe has a strong anterior cingulum as in P. austroafricanus and Canthumeryx; this cingulum has a serrated occlusal edge and appears to consist of a series of small tubercles. The molars exhibit the same degree of brachyodonty as in Canthumeryx. On the Me the posterior crest of the metaconid is flexed labially in the unworn state and this flexion is indicated on the wear trace until medium wear has occurred. The lingual rib of the metaconid is as strong as that of Canthumeryx but does not equal the strength of this rib in Palaeomeryx furcatus or Palaeomeryx magnus in which the rib has a strong posterior fold. The metastylid is strong and projects lingually but it is weaker than the metastylid of some palaeomerycids. The entoconid is shortened in the posterior region and the posterior fossette opens lingually. The protoconid is crescentic and joins the anterior end of the metaconid; this end of the molar is very pointed as in Canthumeryx (pl. 14, fig. 1). The hypoconid is crescentic and its anterior end meets the protoconid in the median valley. The posterior end of the hypoconid is very long and is produced lingually.

The Mz is similar to the Mg in the anterior region. The posterior region of the hypoconid is produced lingually and meets the face of the strong entostylid. The hypoconulid is crescentic and joins the hypoconid labially and the entostylid lingually thus enclosing a large enamel island (pl. 14, fig. I).

Upper dentition. The molars are four rooted with the lingual and labial pairs of roots fused. The enamel is more rugose than in Walangania and M! has a strong anterior cingulum and a weaker posterior cingulum. The parastyle of M! is strong with a pronounced labial rib which forms the antero-labial corner of the tooth. The paracone is stout and less transversely flattened than that of Walangania; this

144 LOWER MIOCENE RUMINANTS

is a primitive feature resembling the traguloid condition. The labial rib of the para- cone is stronger than that of Walangamnia (pl. 14, fig. 2) and resembles closely that of Palaeomeryx in which the rib flexes slightly anteriorly. The mesostyle is similarly developed in Propalaeoryx, Walangania and Palaeomeryx. The metacone has a feeble labial rib and the metastyle is strongly developed (pl. 14, fig. 2). The protocone is similar to that of Palaeomeryx ; it is crescentic with a few small wings at the postero- labial end; these project into the medial valley and the anterior fossette (pl. 14, fig. 2). The anterior fossette is much shallower than in Walangania but resembles that of Palaeomeryx. The metaconule is crescentic; from its posterior region two strong accessory crests project into the fossette; these are strongly developed in all the molars and an enamel island is formed early in wear, this island is large and shallow (pl. 14, fig. 2). As wear continues the first enamel island is lost and a smaller one is developed at the posterior end of the fossette. Accessory crests are very feeble when present in Walangania.

M2? is similar to M! in all its main features. Each molar has a strong entostyle in the median valley and a very strong anterior cingulum which has a serrated edge as in the lower molars.

The P4 is similar to that of Palaeomeryx. The labial region of the tooth is not known but the lingual region of the metacone is preserved. The protocone is crescentic and from its posterior region an extremely strong accessory crest divides the fossette into anterior and posterior regions (pl. 14, fig. 2); in the posterior region a series of small tubercles and folds are present on the face of the protocone and metacone. A strong cingulum forms a swelling at the base of the lingual face of the protocone (pl. 14, fig. 3) a similar though weaker cingulum is occasionally present in Palaeomeryx and Walangania.

The upper dentition with its shallow fossettes, stout labial cusps, strong styles and cingula and complications of the cusps by accessory crests or tubercles is more primi- tive than the dentition of Palaeomeryx or Walangania and in many features agrees closely with the tragulid dentition. The description of the upper molars of Propalae- oryx shows that though it is more primitive than Palaeomeryx it shows closer affinities with this type of dentition than with the primitive bovids.

The affinities of Propalaeoryx. Stromer (1926) indicated that Propalaeoryx was a member of the Bovidae but Arambourg (1933) states:

‘Propalaeoryx austro-africanus Stomer appartient a une espéce d’assez grande taille et posséde d’incontestables traits de Cervidé primitif.’

Whitworth (1958) discusses the affinities of Propalaeoryx and states:

‘Indeed in everything except size, the lower molars of Propalaeoryx are precisely like those of Micromeryx from the Miocene of Europe.’

In spite of evidence presented by dentition, Whitworth continues his argument, citing isolated ruminant limb bones which he tentatively identifies as Propalacoryx ; especially a single metatarsal which:

‘“. . . exhibits a bovid condition in the housing of the extensor tendon while retaining a remarkably cervid-like development of the shaft.’ (Whitworth 1958).

The reasons for this identification rest upon the fact that the bone agrees in size

GEBEL ZELTEN, LIBYA 145

with Propalaeoryx nyanzae; however true bovid dentitions of a similar size have been found in these deposits. Whitworth considered this single bone to be sufficient to cast grave doubts on the supposed cervid affinities of the genus but I do not feel that this tentative evidence can outweigh positive evidence presented by the dentition in favour of palaemerycid affinities for Propalaeoryx. As stated the dentition of Propalaeoryx is similar to that of Palaeomeryx and the strong metastylid, which is much stronger than that of Walangania, is a palaeomerycid rather than a bovid character. The lower premolars of Propalaeoryx austroafricanus are virtually indistinguishable from those of some palaeomerycids or from Canthumeryx. For these reasons Propalaeoryx is classified as a member of the Palaeomerycidae.

It is possible that Canthumeryx and Propalaeoryx are representatives of a new group of ruminants which is of African origin but in the absence of more material I prefer to group them with an existing and certainly closely related family.

TABLE 13 The Dentition of Propalaeoryx nyanzae

Lower Dentition

Mi Me Ms Length Width Length Width Length Width B.M.21368 13°6 mm 72mm 13°7mm 7-9 mm — — K.614.49 — -- —- — -- 8-4 mm K.1263.51 12°5 mm 7-7 mm — — K.774.52 13-5 mm 78mm 19:0 mm 8-3 mm K.780.52 —_ o— 19-6 mm 8-2 mm Upper Dentition p4 M1 M2 Length Width Length Width Length Width K.193.51 —— —- — — 125mm 13:1 mm K.Mt.21 - — I4;2mm 14:4 mm — — K.X. — _- 13z0mm 14:°0mm 12:0mm — K.Mt.67.51 Ioomm 115mm 125mm 13:1 mm — ase K.246.59 — — 125mm 14:0mm — —

BOVIDAE Incertae Sedis Genus WALANGANIA Whitworth 1958

DracGnosis: Small, lightly constructed pecoran. Frontal appendages unknown. Mandible with prominent angle and large recurved coronoid process. Cheek teeth closed from Pg to Mg. First premolar lost. Lower molars brachyodont, selenodont and narrow. Protoconid and hypoconid angular and compressed antero-posteriorly. Upper molars square, brachyodont and selenodont. Paracone with strong labial rib, metacone lacking labial rib. Enamel of cheek teeth finely rugose. Pelvis and rear limbs of advanced pecoran type. (After Whitworth 1958).

146 LOWER MIOCENE RUMINANTS

Walangania africanus (Whitworth) 1958

Palaeomeryx africanus Whitworth 1958. Walangania gracilis Whitworth 1958. Kenyameryx africanus Ginsburg and Heintz 1966.

DiacGnosis: As for genus.

REMARKS: The species Palaeomeryx africanus was established on the basis of a large number of upper and lower dentitions as well as a considerable amount of post- cranial material. Ginsburg and Heintz (1966) removed the species from the genus Palaeomeryx on the basis of a supposed presence of a P; and the more primitive nature of the anterior premolars. The name Kenyameryx was suggested to accommodate the species. The presence of a Pj was disputed by Gentry (1970) who identifies the P; as a second deciduous premolar.

The genus Walangania is known from the holotype only ; this is an almost complete, associated skeleton in which the deciduous dentition and the first permanent molars are erupted. Except in the degree of wear the permanent molars are indistinguish- able from those which Whitworth described as Palaeomeryx africanus. Although Whitworth states that the ‘Palaeomeryx fold’ 1s absent in the holotype a fold is certainly present and is as strong as in many specimens of Palaeomeryx africanus. The dimensions of the teeth are also very similar in Walangania and Palaeomeryx africanus (table 14).

The main differences in the diagnoses of Walangania gracilis and Palaeomeryx africanus are the presence of a P; and a ‘Palaeomeryx fold’ in P. africanus but neither of these differences are acceptable on closer investigation. The other differences between the species result mainly from the juvenile condition of the Walangania holotype and are insufficient to warrant the continued separation of the two species.

Palaeomeryx africanus was described earlier in the publication than Walangamia gracilis but the generic name Palaeomeryx is not applicable to the material. The name Kenyameryx suggested by Ginsburg and Heintz (1966) cannot be used as the name Walangania has precedence. The trivial name ‘africanus’ is retained as it has page precedence over the trivial name ‘gracilis’.

The affinities of Walangania africanus. The affinities of Walangania gracilis were discussed by Whitworth (1958) who concludes:

‘For the present, Walangania must be regarded as a pecoran genus of doubtful systematic position, although the available evidence may slightly favour inclusion with the Cervidae.’

The affinities of Palaeomeryx africanus were implied by its generic name and there- fore with the synonymy of P. africanus and W. gracilis the affinities of the species Walangania africanus should lie with the Palaeomerycidae; however a study of the material indicates that even this affinity is very doubtful.

The lower molars of Walangania exhibit weak lingual ribs and styles these agree in strength with those of Eotvagus and are weaker than is usual in Palaeomeryx. The metaconid and entoconid are aligned almost parallel to the long axis of each lower molar in Walangania and Eotragus but in Palaeomeryx these cuspids usually have a diagonal orientation. The metaconids and entoconids of the lower molars

147

GEBEL ZELTEN, LIBYA

wu o.FT UIWI O-€T wut o.F1 WI 9-6 WU O-OT wut S.or WU O-II wu $.1T wiur S-11 UIPIM qysueT TIPIM. el 2IN

IN

wu O-7I

uur S.or urut $.11

yysueT

wut €-11

UIUI O-ZI UIUI O-ZI yysue'T

WU O-€T

UIUWI I-OL UWIUI 0-6 UIW O-OL wut €-6

UIPIM

TIN

uu o.Z

WIUI 0-9 wut +.9 WIUL 9-9 wut £-9

UFPIM

wu O-O1

uu £-01 mut £-6 wut ¢-o1 UIUL 9-6 yy8ueT

TA

wu F-o1

WIUI O-II wut S.1r ww S-Ir UIUI O-II

yy8ueT

UIUI O-OT

wut /.g WU 9-8 WIUI 9-8

UFPIM

(‘S2719DA8 Diunsuny YY Se poaystT st b Ioquiny)

rd

(zS6r sniueyy ur101,7) sisuatupsuns snsvuyoq (S) ggtie we (%)

6S€iz wa (€) (edAjojoH) gSE1z Wg (Zz) Lb-gh-185°y (1)

uoyyuag samoT wu $.9 (zS6r sntusyy, wo1q) sisuatupsuns snsvoyoq (S

)

= 6g€iz Wa (F) UI 9-9 Lv‘ LeSsy (€) wut 2.2 L¥-469'°x1 (Zz) UIUI 0-8 6b-Lz'18S yf (1)

yysueT

uoyyuag saddyQ

(‘snuvaiafo xhaamoanjog se ‘gS6x yy1ompY AA Aq pozst] ore € 0} I sroquInN)

SnUDILA{D DIUDSUDID AA JO WOIZI}UEp OUT,

vI ATAVE

148 LOWER MIOCENE RUMINANTS

tend to be more selenodont in Walangania and Eotragus than in Palaeomeryx. The labial ribs and styles of the upper molars are far weaker in Walangania and Eotragus than in Palaeomeryx. In all these features Walangamia and Eotragus agree and each differs from Palaeomeryx, the main feature in which Palaeomeryx and Walangania agree is the presence of a ‘Palaeomeryx fold’ on the lower molars of both genera; such a fold is unknown in the bovids but occurs in most other primitive ruminants, its presence in Walangania is insufficient to debar this genus from affinity with the bovids.

Gentry (1970) has also suggested bovid affinities for Walangania on the basis of its mandibular ramus and he states:

‘It also has a moderately deep ramus, so it is quite possible that it will one day be shown to be a bovid.’

The distinction between bovoids, giraffoids and cervoids is very difficult in species of Burdigalian age but the dentition of Walangamia exhibits features which indicate affinity with early bovids and for this reason it is here treated as a bovid; however all the features used exhibit variation and the tentative nature of these interpretations must be emphasized.

VII. REFERENCES

ARAMBOURG, G. 1933. Mammiféres miocénes du Turkana. Annis Paléont., Paris, 22 : 1-26, 5 figs, pls. 1-2.

1947. Mission Scientifique de l’?Omo 1932-1933. Tome 1. Géologie—Anthropologie. Muséum National d’ Histoive Naturelle, Paris, 231-562 : 91 figs, pls 1-11.

1961a. Prolibytherium magmeri, un Velléricorne nouveau du Burdigalien de Libye. (Note préliminaire). C. v7. Séanc. Soc. géol. Fy., Paris, 1961 (3) : 61, 1 fig.

1961b. Note préliminaire sur quelques Vertébrés nouveaux du Burdigalien de Libye. C. vy. Seanc. Soc. géol. Fr., Paris, 1961 4. 107, I fig.

1963a. Legenre Bunolistriodon Arambourg 1933. Bull. Soc. géol. Fr., Paris, (7), 5,903—911,

Brags i ple 1963b. Continental vertebrate faunas of the tertiary of North Africa. Jn Howell, F. C. and Bourliere. African ecology and human evolution, Chicago, 1963.: 56-63.

ARAMBOURG, C. & MAGNIER, P. 1961. Gisements de Vertebres dans le bassin tertiaire de Syrte (Libye). C. 4. Acad. Sci., Paris, 252 (8) : 1181-1183.

Brack, D. 1915. A study of the endocranial casts of Okapia, Givaffa, and Samotherium, with special reference to the convolutional pattern in the family Giraffidae. J. comp. Neurol., Philadelphia, 25, 4 : 329.

Boutin, B. 1926. Die Familie Giraffidae. Palaeont. Sin., Peking, Ser. C, 4, (1), 1-179, 252,

figs, pls 1-12.

1935. Some remarks on fossil Giraffidae. Bull. geol. Soc. China., Peking, 14, 83-89, pl. 1.

CHARDIN, T. de. 1939. The Miocene Cervids from Shantung. Bull. geol. Soc. China., Peking, 19 (3) : 269-278, 5 figs.

CHURCHER, C. S. 1970. Two new Upper Miocene Giraffids from Fort Ternan, Kenya, East Africa : Palaeotvagus primaevus n.sp. and Samotherium africanum n.sp. In Leakey & Savage 1970. Fossil Vertebrates of Africa. 2. 336p.

CLarK, Le Gros. 1939. The Brain of the Okapi. Proc. zool. Soc. Lond., 109 B., 153-159, 4 figs.

CoLBERT, E. 1933. A skull and mandible of Givaffokeryx punjabiensis, Pilgrim. Am. Mus.

Novit., New York, 632, 1-14, 6 figs.

1935a. Distributional and Phylogenetic Studies on Indian Fossil Mammals. V. The classification and the phylogeny of the Giraffidae. Am. Mus. Novit., New York, 800, 1-15.

GEBEL ZELTEN, LIBYA 149

CoLBERT, E. 1935b. Siwalik Mammals in the American Museum of Natural History. Tyan. Am. phil. Soc., Philadelphia. N.S., 26,x + 401 pp. 198 figs. 1936. Was the extinct giraffe Sivatherium known to the Ancient Sumerians? Am. Anthrop., 38. 1938. The Relationships of the Okapi. J. Mammal., Baltimore, 19, (1), : 47-64, 3 figs. CooprE, G. R. 1968. The Evolutionary Origin of Antlers. Deer, 1: 215-217. CrusaFont, M. 1952. Los Jirafidos Fésiles de Espana. Mems. Comun. Inst. geol., Barcelona, 8 : 9-239, 26 figs, pls 1-47. Desio, A. 1935. Studi geologici sulla Cyrenaica, sul Deserto Libico, sulla Tripolitania e sul Fezzan Orientali. Missione scient. R. Acc. d'Italia a Cufra (1931). I. Rome. FRASER, F.C. 1951. Vestigial metapodials in the Okapi and Giraffe. Pyvoc. Zool. Soc. Lond., 121 : 315-317, pls 1-2. Geist, V. 1965. The evolution of horn-like organs. Behaviour. 27, 175-214, 16 figs. Gentry, A. W. 1964. Skull Characters of African Gazelles. Ann. Mag. nat. Hist., London (13) 7: 353-382. 1966. Fossil Antilopini of East Africa. Bull. Br. Mus. nat. Hist. (Geol.) London, 12 : 45-106, 15 figs. 1970. The Bovidae (Mammalia) of the Fort Ternan Fossil Fauna. In Leakey & Savage 1970. Fossil Vertebrates of Africa. 2. 3306p. GinspurGc, L. & Herntz, E. 1966. Sur les affinités du genre Palaeomeryx (Ruminant du Miocéne européen). C. R. Acad. Sc., Paris, 262 : 979-982, 1 fig. Innis, A. C. 1958. The behaviour of the giraffe, Givaffa camelopardalis, in the Eastern Transvaal. Proc. zool. Soc. Lond., 131, (2) : 245-278, 6 figs, pl. 1. LANKESTER, E. R. 1907. The origin of the lateral horns of the giraffe. In foetal life on the area of the parietal bones. Proc. zool Soc. Lond., 1907 : 100-115, figs 24-36. 1908. On certain points in the structure of the cervical vertebrae of the Okapi and the Giraffe. Proc. zool. Soc. Lond., 1908 : 320-324, figs 60-70. tg10. Monograph of the Okapi. Bvit. Mus. Nat. Hist., London, 12 pp., pls 1-48. LYDEKKER, R. 1883. Siwalik Selenodont Suina. Pal. Indica. (10), II, Pt. 5, 142-177, 3 figs, pls 23-25. MacInnes, D. 1936. A new genus of fossil deer from the Miocene of Africa. J. Linn. Soc. London, 39 : 521-530, 5 figs. Macanier, P. 1962. Etude géologique du gisement de Vertébrés du Gebel Zelten (Libye). C. vy. somm. Séanc. Soc. géol. Fy. Paris, 1962, 2, Feb. 19 : 55-57, I fig. MattHew, W.D. 1929. Critical Observations upon Siwalik Mammals. Bull. Am. Mus. nat. Hist., New York, 56 : 437-560. MattHew, W. D. & GRANGER, W. 1924. New Insectivores and Ruminants from the Tertiary of Mongolia, with Remarks of the Correlation. Am. Mus. Novit., New York. 105: 7p figs I-3. Meap, C. S. 1906. Adaptive Modifications of the Occipital Condyles in Mammalia. Am. Nat., Cambridge, Mass., 40 : 475-483, 12 figs. Merapze, G. K. 1964. On the phylogeny of the Sivatheriinae. Int. Geol. Congress., 22. Reports of Soviet Palaeontologists. (8) Tertiary Mammals. (In Russian). 47-50, 1 fig. MILNE-EpDwarps, A. 1864. Recherches anatomiques, zoologiques et paléontologiques sur la famille des chevrotians. Ann. Sci. nat. Zool., Paris (5) 2 : 49-167, pls 2-12. Piterim, G. 1937. Siwalik Antelopes and Oxen in The American Museum of Natural History Bull. Am. Mus. nat. Hist., New York, 72 : 731-874, 81 figs. 1939. The fossil Bovidae of India. Mem. geol. Surv. India., Calcutta. 26 (1) : iii + 356 Pp., 35 figs, pls 1-7. 1941. The Relationship of certain Variant Fossil Types of ‘Horn’ to those of the Living Pecora. Ann. Mag. nat. Hist., London, (2), 7 : 172-184. Rieu, E. V. 1959. Apollonius of Rhodes. The voyage of Argo. Penguin. London. RoGeEr,O. 1904. Wirbeltierreste aus dem Obermiocan der bayerisch-schwabischen Hochebene. Ber. naturw. Ver. Schwaben, Ausburg, 36 : 1-22 pls. 1-4.

150 LOWER MIOCENE RUMINANTS

Roman, F. & VirET, J. 1934. La faune de Mammiferes du Burdigalien de la Romieu (Gers). Mem. Soc. géol. Fr., Paris, (N.S.) 9 (mém 21) : 1-67, pls I-12.

SavaGE, R. & WuitE, M. 1965. Two mammal faunas from the early Tertiary of central Libya. Proc. geol. Soc., London. 1623: 89-91.

ScHLossER, M. 1886. Beitrage zur Kenntnis der Stammesgeschichte der Hufthiere und Versuch einer Systematik der Paar- und Unpaarhufer. Morph. Jb., Leipzig, 12 : iv + 136, pls 1-6.

SELLEY, R. 1968. Near-shore marine and continental sediments of the Sirte basin, Libya. Proc. geol. Soc., London. 1648: 81-90, 1 fig.

1969. Near-shore marine and continental sediments of the Sirte basin, Libya. Q. Jl. geol. Soc. Lond., 124 : 149-460, 17 figs, pls 21-25.

SHUTTLEWORTH, A. 1943. The function of the femoro-patellar joint of the horse. Jl. R. Aymy vet. Cps., Aldershot, 15 : 2-6. 7 figs.

SIGOGNEAU, D. 1968. Le genre Dremotherium (Cervoidea) anatomie du crane, denture et moulage andocranien. Annis. Paléont., Paris, (Vertébrés) 54 (1) : 37-64, 33 figs, pls 1-6.

Simpson, G. G. 1931. A new classification of mammals. Bull. Am. Mus. nat. Hist., New

York, 59 : 259-293. 1945. The principles of classification and a classification of mammals. Bull. Am. Mus.

nat. Hist., New York, 85: xvi + 350.

Sisson, S. & GROSSMAN, J. 1953. The anatomy of the domestic animals. Fourth Edition, Saunders, London, : 972 pp., 736 figs.

SmitH, J. & SavaGE, R. 1956. Some locomotory adaptations in mammals. J. Linn. Soc. London, 42 : 603-622, 14 figs.

STiRTON, R. 1944. Comments on the relationships of the cervoid family Palaeomerycidae. Am. J. Sci., New Haven, 242 : 633-655, 2 figs.

STROMER, E. 1926. Reste Land- und Siisswasser-Bewohnender Wirbeltiere aus den Diaman- tenfeldern Deutsch-Siidwestafrikas. Jn Kaiser, E. Die Diamantenwuste Sudwest-africas, 2: 107-153. Berlin.

THENIUS, E. 1952. Die Boviden des steirischen Tertiars. Sbev. Ost. Akad. Wiss., Vienna, (1)

161, 409-439, 11 figs.

1956. Zur Entwicklung des Knochenzaptens von Protvagocerus Deperet (Bovidae,

Mammalia) aus dem Miozan. Geologie, Berlin, 5: 308-318, 5 figs.

TroFimov, B. 1958. New Bovidae from the Oligocene of Central Asia. Véerteby. palasaat., Peking, 2: 244-247, 1 fig.

VooruiEs, M. R. 1969. Taphonomy and Population Dynamics of an Early Pliocene Verte- brate Fauna, Knox County, Nebraska. Contrib. to Geol. University of Wyoming, Special Paper. Laramie, Wyoming; 1: 1-69: 29 figs.

WALTHER, F. 1960. ‘Antilopenhafte’ Verhalfensweisen im Paarungszeremoniell des Okapi

(Okapia johnstoni Sclater, 1901). Z. Tierpsychol., Berlin, 17 : 188-210.

1962. Uber ein Spiel bei Okapia johnstoni. Z. Saugetierk., 27 : 245-251.

WesB, D. 1965. The osteology of Camelops. Bull. Los. Ang. City. Mus. Sci., Los Angeles, 1: 1-54, 22 figs.

WEBER, 1928. Die Sdugetiere. Einfuhrung in die Anatomie und Systematik der Recenten und Fossilen Mammalia. Gustav Fischer, Jena, pt 1: xv—444, 315 figs pt 2: xxiv + 898, 573 figs.

WHITWoRTH, T. 1958. Miocene Ruminants of East Africa. Br. Mus. nat. Hist., London. Fossil Mammals of Africa, 15: iii + 50 pp., 18 figs.

Younea, C. 1937. Ona Miocene fauna from Shantung. Bull. geol. Soc. China, Peking, 17 : 209-238, 17 figs, pls 1-3.

W. R. HamiLton, Ph.D.,

Department of Palaeontology,

BriTisH Museum (NATURAL History), CROMWELL Roap, LonDoN SW7 5BD

IIL ANAS i

Fic. 1. Dorcatherium libiensis sp. nov. Right mandible, occlusal aspect. Holotype

(M.26684) 1-6 approx. Fic. 2. Canthumeryx sirtensis gen. et. sp. nov. Right M; and D4; occlusal aspect.

(M.26683) 1-6 approx. Fic. 3. Canthumeryx sirtensis gen. et sp. nov. Right Mj; lingual aspect. (M.26683)

x 1-4 approx. Fic. 4. Canthumeryx sirtensis gen. et sp. nov. Right mandible; occlusal aspect.

Holotype (M.26682). Fic. 5. Canthumeryx sirtensis gen. et sp. nov. Right P4 and Ps; occlusal aspect.

(M. 26683) 1-5 approx. Fic. 6. Palaeomerycidae indet. Ossicones; anterior aspect. (M.26690).

Bull. Br. Mus. nat. Hist. (Geol.) 21, 3 PLATE 1

PIgA iE 2

Zarafa zelteni gen. et sp. nov. Skull; lateral aspect. Holotype (M.26670).

Bull.

Bry. Mus. nat. Hist. (Geol.) 21, 3

PIL,

ATE

iS)

PLATE 3

Zarafa zelteni gen. et sp. noy. Skull; dorsal aspect. Holotype (M.26670).

~-

ol.) 21,

se!

(¢

nat. Hist.

Mus.

Bull. Br.

PLATE 4 Fic. 1. Zarafa zelteni gen. et sp. nov. Supraoccipital region; posterior aspect. Holotype (M.26670) 1-2 approx. Fic. 2. Zarafa zelteni gen. et sp. nov. Basicranial region; ventral aspect. Holotype (M.26670) -68 approx. Fic. 3. Zarafa zelteni gen. et sp. nov. Right mandible; lingual aspect. (M.26675) x +54 approx.

fe) o a — ss n +s S

Mus.

Bull. By.

Fic. 1. (M.26672) FIG. 2. approx. Fia. 3. approx. FIG. 4. approx. Fic. 5. approx.

Zarafa zelteni gen.

1-7 approx.

Zarafa zelteni gen.

Zarafa zelteni gen.

Zarafa zelteni gen.

Zarafa xelteni gen.

PLATE 5

et sp. nov. et sp. nov. et sp. nov. et sp. nov.

et sp. nov.

Left upper, juvenile dentition; occlusal aspect.

Right Me; occlusal aspect. Right Mg; lingual aspect. Right M3; occlusal aspect.

Right Ms; occlusal aspect.

(M.26677) (M.26677) (M.26675)

(M.26676)

Bull. Br. Mus. nat. Hist. (Geol.) 21, 3 PLATE 5

F*

PLATE 6

Fic. 1. Zarafa zelteni gen.et sp.nov. Right upper dentition; occlusal aspect. (M.26671) 1-7 approx.

Fic. 2. Zarafa zelteni gen. et sp. nov. Right upper dentition; labial aspect. (M.26671) 1-7 approx.

Bull. By. Mus, nat. Hist. (Geol.) 21, 3 PAGE 6

PLATE 7

Prolibytherium magnieri Arambourg 1961. Skull and ossicones, ventral aspect. (M.21901).

Bull. By. Mus. nat. Hist. (Geol.) 21, 3 PAC 7

PIL INAS,

Prolibytherium magnieri Arambourg 1961. Left maxilla and upper dentition; lateral aspect. (M.21901) 1-4 approx.

Bull. Br. Mus. nat. Hist. (Geol.) 21, 3 PLATE 8

PLATE 9

Prolibytherium magnieri Arambourg 1961. Left upper dentition and palate. (M.21901) 1-4 approx.

Bull. Br. Mus. nat. Hist. (Geol.) 21, 3 PLATE 9g

PLATE to

Fic. 1. Prolibytherium magnieri Arambourg 1961. Basicranial region; ventral aspect. (M.21901) 0-8 approx.

Fic. 2. Prolibytherium magnieri Arambourg 1961. Occipital region; posterior aspect. (M.21901) 0-6 approx.

Fic. 3. Prolibytherium magnieri Arambourg 1961. Left M3; occlusalaspect. (M.2668r) <2 approx.

Fic. 4. Prolibytherium magnieri Arambourg 1961. Left M3; lingual aspect. (M.2668r) x2 approx.

Bull. Br. Mus. nat. Hist. (Geol.) 21, 3 PLATE to

PLATE 11

Fic. 1. Prolibytherium magnieri Arambourg 1961. Right mandible; occlusal aspect. (M.21899) 1 approx.

Fic. 2. Prolibytherium magnieri Arambourg 1961. Right mandible; labial aspect. (M.21899) XI approx.

Fic. 3. Prolibytherium magnieri Arambourg 1961. Right petrosal; ventro-lateral aspect. (M.21g01) 1-7 approx.

Fic. 4. Prolibytherium magnieri Arambourg 1961. Right petrosal; dorso-medial aspect. (M.21901) 2 approx.

Bull. By. Mus. nat. Hist. (Geol.) 21, 3 PLATE t1

PAE TANASE mere,

hic. 1. Prolibytherium magnieri Arambourg 1961.

aspect. (M.26678) 1-3 approx. Fic. 2. Prolibytherium magnieri Arambourg 1961. (M.26678) 1-3 approx.

Endocranial cast; right lateral

Endocranial cast; dorsal aspect.

nat. Hist. (Geol.) 21, 3

PLATE 13

Fic. 1. Left. Protragocerus sp. Horncore; left lateral aspect. (M.26687). Right. Eotragus sp. Horncore; right lateral aspect. (M.26688).

Fic. 2. Gazella sp. Left mandible; occlusal aspect. (M.26685) 0-78 approx.

Fic. 3. Gazella sp. Left mandible; lingual aspect. (M.26685) 0-78 approx.

Fic. 4. Gelocus whitworthi sp. nov. Left Mz and M3; occlusal aspect. Holotype (K.Sgr. 305.1949) 3:3 approx.

Fic. 5. Gelocus whitworthi sp. nov. Left Mz and M3; lingual aspect. Holotype (K.Sgr. 305.1949) 3:3 approx.

Bull. By. Mus. nat. Hist. (Geol.) 21, 3 PIL NANI, 113}

PLATE 14

Fic. 1. Propalaeoryx nyanzae Whitworth 1958. Right Mz and M3; occlusal aspect. (K.774.52) 2:9 approx.

Fic. 2. Propalaeoryx nyzanzae Whitworth 1958. Left M! and P*4; occlusal aspect. (I<.Mt.67.51) 2-9 approx.

Fic. 3. Propalaeoryx nyzanzae Whitworth 1958. Left M! and P#4; lingual aspect. (K.Mt.67.51) 2-9 approx.

Bull. By. Mus. nat. Hist. (Geol.) 21, 3 PLATE 14

‘ : ol

A LIST OF SUPPLEMENTS TO THE GEOLOGICAL SERIES OF THE BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY)

Pp. 213; 30 Plates; 2 Text-figures. 1965. . Ex-Nacear, Z. R. Stratigraphy and Planktonic Foraminifera of the Upper — Cretaceous—Lower Tertiary Succession in the Esna-Idfu Region, Nile Valley Egypt, U.A.R. Pp. 291; 23 Plates; 18 Text-figures. 1966. f10. me. . Davey, R. J., Downtz, C., SARGEANT, W. A. S. & Pare G. L. Studies o or g x

figures. 1966. £7. . . APPENDIX. Davey, R. J., DowniE, C., SARGEANT, W. A. S. & WILLIAMS, c. : Appendix to Studies on Mesozoic and Cainozoic Dinoflagellate Cysts. ‘Pp. 1969. 8op. ° . Erziott, G. F. Permian to Palaeocene Calcareous Algae (Dasycladaceae) of th " Middle East. Pp. 111; 24 Plates; 17 Text-figures. 1968. {5.12}. “ . Ruopes, F. H. T., Austin, R. L. & Druce, E. C, British Avonian ( (Carboni ni- ferous) Conodont faunas. and their value in local and continental correlation. Pp. 315; 31 Plates; 92 Text-figures. 1969. {TII. * . Cuttps, A. Upper Jurassic Rhynchonellid Brachiopods from Northw [ Europe. Pp. 119; 12 Plates; 40 Text-figures. 1969. £4.75. ce . Goopy, P. C. The relationships of certain Upper Cretaceous Teleoae . special reference to the Myctophorids. Pp. 255; 102 Text-figures. 1969. {6. . OwEN, H. G. Middle Albian Stratigraphy in the Paris Basin. he 3 Plates; 52 Text-figures. 1971. {6.

from West Pakistan. Pp. 98; 42 Plates; 7 Text-figures. 1971. {£8.

. oe

BCG Printed in England by Staples Printers Limited at their Kettering, Northants establishment

te WALKER

roe te

: BL : NATURAL HISTORY) Doct hema Vol. 21 No. 4

ets

i fang iN

Ee Ar PINITIES"OF HALCYORNIS FROM THE LOWER EOCENE

BY COLIN JAMES OLIVER HARRISON AND CYRIL ALEXANDER WALKER

Pp. 151-169; 3 Plates, 9 Text-figures

BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) GEOLOGY Worl 2 New 4 LONDON : 1972

THE BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY), instituted im 1949, 1s issued in five series corresponding to the Departments of the Museum, and an Historical serves.

Parts will appear at irregular intervals as they become veady. Volumes will contain about three or four hundred pages, and will not necessarily be completed within one calendar year.

In 1965 a separate supplementary series of longer papers was instituted, numbered serially for each Depariment.

This paper is Vol. 21, No. 4 of the Geological (Palaeontological) series. The abbreviated titles of periodicals cited follow those of the World List of Scientific Periodicals.

World List abbreviation Bull. Br. Mus. nat. Hist. (Geol.)

© Trustees of the British Museum (Natural History), 1972

TRUSTEES OF THE BRITISH MUSEUM (NATURAL HISTORY)

Issued 31 October, 1972 Price £1.55

mie APFINITIES OF HALCYORNIS FROM THE POWER EOCENE

By C. J. O. HARRISON & C. A. WALKER

CONTENTS Page I. HusroricAL INTRODUCTION . . é : : 153 II. DESCRIPTION AND STATE OF PRESERVATION . . : : 154 IfI. COMPARISON WITH RECENT FORMS . 157 IV. DETAILED COMPARISON WITH CORACIIFORM AND PICIFORM SPECIES 163 V. POSSIBLE CHARACTERS OF HALCYORNIS ¢ : : ; 167 VI. GENERAL CONCLUSIONS . ; : : 2 5 F : 167 VII. SYSTEMATIC DESCRIPTION : : : : . . : 168 VIII. ACKNOWLEDGEMENTS. F : ‘ : 3 . 3 169 IX. REFERENCES. : : j 5 ; 3 ‘ : : 169 SYNOPSIS

Halcyornis toliapicus (Koenig) of the Lower Eocene, known from a single cranium, has been classified with the gulls (Koenig, 1825 and Lydekker, 1891) and the kingfishers (Owen, 1846). The specimen has now been fully prepared; it is redescribed here and its probable affinities re-appraised on the characters now apparent.

A comparison with Recent forms indicates a general similarity to skulls of Coraciiformes and Piciformes. A more detailed examination of a range of species representing families within those two orders indicates that the greatest similarity is to the Coraciiformes, particularly to the rollers of the Coraciidae and Leptosomatidae. The characters of the fossil specimen do not appear to indicate stronger affinities with either one of those families, however, and it is therefore proposed that Halcyornis should be regarded as the type- (and only known) genus of a new family within the Coraciiformes. A diagnosis is given of the Halcyornithidae.

I. HISTORICAL INTRODUCTION

In the earlier part of the 19th century the cranial portion of a small bird skull was found in the London Clay, Ypresian (Lower Eocene) of the Isle of Sheppey, Kent, England. It was figured by Koenig (1825), who noted the general resemblance of the rounded cranium with its even, tapering temporal fossae to that of some of the smaller gulls (e.g. Larus canus or L. ridibundus) and therefore reconstructed it as a small gull, calling it Larus toliapicus.

Owen examined the specimen, and noted (1846 : 554) the absence of the supra- orbital grooves which accommodate the lacrimal glands (grooves which are normally present in, and typical of, gulls and most other sea-birds). He compared the specimen with Recent material but may have used an incomplete collection. He noted the resemblance of the temporal fossae to those of the kingfisher, Alcedo atthis (although in the latter the fossae extend upwards until they almost meet in the mid- line) and, believing that it was probably an early kingfisher, called it, ‘Halcyornis tohapicus. Bird probably of the family Halcyonidae’. The kingfisher family is now known as the Alcedinidae.

154 AFFINITIES OF HALCYORNIS

Lydekker (1891 : 183) considered the specimen to be a gull. He rejected Owen’s suggestion because the temporal fossae on the posterior part of the cranium were further apart than in the kingfisher; and also because of the narrowness of the inter- orbital bar formed by the frontal bones. He commented that, ‘the imperfection of the fronto-parietal region renders it difficult to be sure as to the presence of super- orbital grooves, but the appearance suggests their presence’. He associated with the skull the distal end of a left humerus (BMNH No. A 10) which he considered to be of larid origin. There appears to be no reason for associating the humerus with the cranium; the former has therefore been considered as a separate specimen and has not been dealt with in the present paper.

We have compared Halcyornis with a range of Recent bird bones in the British Museum (Natural History). The most obvious character, and the one on which earlier identifications were based, is the presence of well-defined temporal fossae which occupy about two-thirds of the total width of the cranium. They are of fairly even shape, tapering posteriorly, with a slightly concave posterior margin. The search for similar Recent species showing this type of posteriorly-rounded cranium with relatively simple temporal fossae produced a number of alternative possibilities, allowance having been made for a reasonable degree of variation. Specimens with some similarity were found in the Procellariidae (Procellariiformes), Fregatidae (Pelecaniformes), Rhinochetidae (Gruiformes), Laridae (Charadriiformes), Cuculidae (Cuculiformes), Alcedinidae, Momotidae, Meropidae, Coraciidae, Upupidae, Leptosomatidae (Coraciiformes), Galbulidae, Bucconidae, Capitonidae, and Ramphastidae (Piciformes). So many families possess this type of skull that it is obviously an unsatisfactory character from the taxonomic point of view, but nevertheless provides an initial limited list of potentially related families

Il. DESCRIPTION AND STATE OF PRESERVATION

It was necessary to consider the other characters of the specimen, which was further prepared for this purpose, making some of the interorbital features more apparent.

The specimen consists of the posterior portion of a cranium, imperfect in all its aspects. The roof has been broken away to show part of the endocranial cast and the anterior region is broken off about half-way along the frontals. The interorbital bar is narrow; the orbital rim is apparently undamaged, with no evidence of any superorbital grooves. There isa fairly large, well-marked temporal fossa the margin of which arises anteriorly at the junction between the postorbital processes and the orbital rim, runs obliquely backwards towards the mid-line, and then curves round to pass outwards along the dorsal border of the occiput.

In lateral view the skull shows a cranium that is not particularly inflated or rounded, and the thick interorbital septum is perforated only by a single foramen for the optic nerve. This foramen is situated at the postero-ventral corner and is rather small. The orbital rim is raised slightly above the plane of the skull roof. The postorbital processes are broken on both sides, but would have been blade-like, with

ENGLISH LOWER EOCENE 155

narrow edges directed laterally. The temporal fossa forms, in this aspect, a well- marked groove between the postorbital process and the tympanic cavity.

The palatal surface is eroded and much of the detail on the basiparasphenoid region has been lost. The rostroparasphenoid is thick and the alaparasphenoids swing out to form a wide angle with the former. The eustachian tubes are visible. There is a well-marked depression just posterior to the ridge which runs across and joins the two alaparasphenoids. Much of the basioccipital is broken, but the occipital condyle, although eroded, is still present and was small. The foramen magnum is directed posteroventrally, but its exact shape is impossible to determine. The profile of the occiput in this view is relatively flat with a slight swelling around the foramen magnum.

The occiput is also damaged about the mid-line, but the shape is broadly crescentic, with a well-defined median ridge running down the parietals and supraoccipital to meet the dorsal rim of the foramen magnum in the mid-line. Either side of this ridge there is a well-marked groove, which probably indicates the border between the supraoccipital and opisthotic. There appear to have been no foramina in the supraoccipital. The temporal fossae are distinct from this angle and their posterior border forms a well-defined ridge along the upper edge of the occiput. The fossae do not, however approach the mid-line.

Fic. 1. Halcyornis toliapicus. BM(NH) No. A 130. Cranium. Dorsal View, ~ 4.

56

Fic. 2.

AFFINITIES OF HALCYORNIS

Halcyornis toliapicus. BM(NH) No. A 130. Cranium.

Lateral View,

ENGLISH LOWER EOCENE 157

MEASUREMENTS Maximum width across exoccipital region. ; 20°5 mm Minimum width between temporal fossae_.. : g mm Width between postorbitals : : 3 : 22 mm Width of interorbital bar. ' : é : 7 mm Minimum width of interorbital septum . : : 2.9 mia: Maximum depth of skull in orbital region. : 17 mm Maximum length of auditory meatus. ‘ : 9 mm

Tic. 3. Halcyornis tohapicus. BM(NH) No. A 130. Cranium. Frontal View, x 4.

Ill COMPARISON WITH RECENT FORMS

From the description it is apparent that for determining relationships, characters might be used other than those used in the past. The most obvious of these relate EO:

I. The upper edges of the orbit.

2. The interorbital septum and foramen.

3. The relative size and shape of the parasphenoid region.

IES He

APFINITIES OF BALCYORNIS

Halcyornis tohiapicus.

BM(NH) No. A 130.

Cranium.

Ventral View, X 4.

Halcyornis toliapicus.

BM(NH) No. A 130.

Cranium.

Posterior View, X 4

ENGLISH LOWER EOCENE 159

1. The edges of the orbit

Lydekker’s views on the affinities of Halcyornis appear to have been strongly influenced by his opinion that the appearance suggested the presence of supraorbital grooves, subsequently destroyed. On birds which normally live on or by salt water,

be - é

H |

Fic. 6. Diagrams of dorsal views of crania to show width of interorbital bridge. a, Alcedo; 8, Dacelo; c, Coccyzus; D, Halcyornis; ©, Larus: F, Puffinus; G, Rhynochetos; H, Fregata; 1, Coracias.

160 AFFINITIES OF HALCYORNIS

the supraorbital glands tend to become greatly enlarged and are housed in distinct grooves along the upper edges of the orbits (Text-fig. 6E, F). These glands, which assist salt-excretion, are much smaller in related fresh-water species. The grooves, where present, are on top of the skull and tend to occupy most of the upper edges of the orbits, reducing the thickness of the frontal bones in this region. They may be deepest towards the centre of their width, at times penetrating to the orbits below and forming a series of fenestrae, or they may give to the orbital edges an irregular or eroded appearance. A consistent feature is the distinct ridge of bone along the inner edge of each groove where the frontal bone is restored to its normal thickness.

Since these ridges of bone are not apparent on Halcyornis it would be necessary, were it held that supraorbital grooves had been present, to assume that the entire bony shelf supporting the supraorbital glands had worn away completely, back to this ridge. Although the presence of the groove reduces the thickness of the frontal bone by removing part of its upper surface, it does not affect the underlying surface within the orbit which curves downwards away from the orbit edge in the normal manner; and therefore an orbit worn back to the inner edge of a supraorbital groove would show a considerable depth of worn or broken surface at the edge between the top of the skull and the inside of the orbit. The skull of Halcyornis does not; the upper edge of the orbit does not taper to a very narrow edge but shows an abrupt surface about one millimetre deep, suggesting that a small amount of wear may have occurred on a normal orbit lacking a supraorbital groove (Text-fig. 2A, B). Another character of the supraorbital grooves is that it tends to terminate rather abruptly posteriorly, with the inner ridge curving towards the orbit edge; thus if a groove were to wear away, the orbit edge would show a sharp discontinuity at this point. There is no such discontinuity in Halcyorms.

The absence of a supraorbital groove suggests that the specimen is not related to the Procellariformes, Anseriformes or Charadriiformes (including Laridae). The Pelecaniformes (Text-fig. 6H) and Ciconiiformes, although in some cases associated with salt water, do not possess these supraorbital grooves. Since the groove is a relatively plastic and adaptable character in the families in which it now occurs, it might be argued that it could have been absent in some or all species in earlier epochs. It is known to be present in Miocene gulls but we have no definite proof that it was present earlier, and as an indication of relationship it can be assessed only in conjunction with other characters.

2. The interorbital septum

Halcyornis has a thick, unfenestrated interorbital septum (Text-fig. 3) with a single optic foramen in a low posterior position (Text-fig. 71). Most of the birds discussed in this paper have marked and often extensive interorbital fenestrae (Text-fig. 7); these include the Procellariiformes, Pelecaniformes, Ciconiiformes and Charadriiformes. In the skimmers (Rhynchops spp., Charadriiformes) a highly specialized method of feeding (Zusi, 1962) has necessitated extra strengthening of the skull, including the development of a thick, solid interorbital septum; the latter, however, is pierced by the optic foramen in a high position (Text-fig. 7C) and there are small paired fenestrae opening into the cranial cavity just above this. Inter-

ENGLISH LOWER EOCENE 161

orbital fenestrae are also present in the Rhinochetidae, Cuculidae, and Alcedinidae. On the other hand the coraciiform and piciform families listed earlier generally show a solid interorbital septum with a low posterior optic foramen like that of Halcyornis ; the only exceptions are the Upupidae with a pair of fenestrae opening into the cranial cavity from the upper orbits and the Momotidae with a tiny variable “‘relict’’ foramen in the mid-septum. On the basis of this character there is a strong case for linking Halcyornis with the Coraciiformes and Piciformes, although the example of Rhynchops indicates that exceptional circumstances could lead to some adaptive variation in the fenestration of the septum.

Fic. 7. Diagrams to show interorbital septa and position and size of interorbital foramina. A, Fregata; B, Puffinus; c, Rhynchops; v, Dacelo; ©, Merops; ¥, Upupa; G, Leptosomus; H, Eurystomus; 1, Halcyornis.

162 AFFINITIES OF HALCYORNIS

Another character in which Halcyornis resembles species of the Coraciiformes and Piciformes is the shape of the angle formed by the interorbital septum and the underside of the interorbital bridge section of the frontals (Text-fig. 8). On most of the Recent species examined the interorbital bridge is relatively thin and the septum extends well up towards the orbit where it meets the underside of the frontal at a sharp angle. In Halcyornis the interorbital bridge is much thicker towards the mid-line and tapers down more gradually, making an obvious junction with the septum at about a third of the distance between the top of the bridge and the base of the septum (Text-figs 2A, 2B, 3). This is less obvious anteriorly where the bone narrows, but can be clearly seen in lateral view as a discontinuity crossing the inner orbit. This heavier tapering interorbital bridge is apparent on skulls of Momotus (Coraciiformes) and Megalaima, Selenidera and Monasa (Piciformes). Eurystomus, Leptosomus and Merops (all Coraciiformes) show a similar low junction between septum and frontal inside the orbit; in these genera, however, there is a mid-line hollow on the top of the skull and the frontal is concave in cross-section, so that the apparently lower position of the junction is due to a general ventral displacement and cannot be regarded as an homologous development.

3. The parasphenord region

We have followed Jollie (1957) in regarding the externally visible ventral bones of the hind-skull as parasphenoid rather than sphenoid. The basiparasphenoid of Halcyornis is transversely elongated and roughly oblong, with a slight lateral taper (Text-fig. 4, Pls 1B, 3E). It is rather flat and this may have been emphasized by pressure and slight erosion. Anteriorly it terminates in a distinct edge where it abuts on the alaparasphenoids. The rostroparasphenoid is thick and has a small but distinct ventral prominence near the posterior end. The most conspicuous feature on each side is the alaparasphenoid, directed strongly laterally at the hind end of the rostroparasphenoid to form a prominent bulging ridge at the back of the base ot the orbit, curving upwards to meet the orbitosphenoid. The lateral ends of both alaparasphenoids and the basiparasphenoids are broken away.

In the Procellariformes (Pl. 3C) the basiparasphenoid tapers markedly towards the front and there is a narrow gap between it and the ventrally projecting edge of what appears to be the orbitosphenoid. The Pelicaniformes show a tapering basipara- sphenoid which becomes distinctly wedge-shaped in some forms, and in Fregata (Pl. 3B) there is again a gap between the basiparasphenoid and orbitosphenoid. In the Charadriiformes (Pl. 3D, F) this anterior taper is again apparent on the basi- parasphenoid, while the alaparasphenoid is more prominent than in the previous taxa but slants away weakly to the orbital and otic region, least so in some small Larus species such as L. ridibundus. The Cuculiformes (Pl. 3L) show a wedge- shaped basiparasphenoid, narrow anteriorly in some species but broad in others, and the alaparasphenoid is poorly defined.

In these characters, as in the previous ones, the greatest similarity to Halcyornis is shown by piciform and coraciiform birds. In most of these the basiparasphenoid is wide transversely, tending towards an oblong shape with little lateral taper; in the Capitonidae (Piciformes) and in the Coractidae (Pl. 3H) Leptosomatidae, and

ENGLISH LOWER EOCENE 163

Momotidae (Coraciiformes) it has a similar shape to the fossil, but in the Bucconidae (Pl. 31) and Galbulidae (Piciformes) and in the Alcedinidae (Pl. 3J) and Meropidae (Pl. 3G) (Coraciiformes) the distance between the orbits and the foramen magnum is shortened, making the basiparasphenoid narrower from front to back.

In both orders many species have a pair of small prominences on the ventral surface of the basiparasphenoids. These are very variable, being almost absent, for instance, on Alcedo (Alcedinidae). In Halcyornis this bone shows a pair of small areas with dull surfaces suggesting abrasion or heavy pressure, corresponding in position to these prominences although one is a little displaced by fracture.

The principal difference between the Piciformes and Coraciiformes lies in the development of the Alaparasphenoid. On the piciform skulls examined this element is barely apparent, although there is a small, poorly developed ridge on the skulls of the larger Megalaima species. In the Coraciiformes the alaparasphenoid shows some variation in development in the Meropidae, Leptosomatidae, and many of the Coraciidae it is well developed, as a distinct projecting ridge at the posterior basal edge of the orbit, the resemblance to Halcyornis being especially strong in the rollers Leptosomus, Eurystomus and Coracias.

In both orders the rostroparasphenoid is relatively stout ; Merops (Meropidae) and Dacelo (Alcedinidae) show a ventral prominence near the posterior end which is like that on the fossil.

Conclusions

The characters of Halcyornis listed above afford no good reason for linking it with either the Laridae or the Charadriiformes. The only orders with which it appears to show any consistent affinities are the Piciformes and the Coraciiformes; insofar as the parasphenoid region is concerned it seems more like the Coraciiformes, in particular the Coraciidae and the Meropidae.

IV. DETAILED COMPARISON WITH CORACIIFORM AND PICIFORM SPECIES

The species within these two orders do not show uniform skull characters, for they have evolved differences in feeding habits and bill-shape and therefore exhibit varying degrees of divergence from a hypothetical ancestral form. Specimens of the following species were used in this comparison:

CORACITFORMES

Alcedinidae Dacelo novaeguineae Alcedo atthis

Momotidae Momotus sp. Aspatha gularis Baryphthengus ruficapillus

Meropidae Merops apiaster

164 AFFINITIES OF HALCYORNIS

Leptosomatidae Leptosomus discolor Coraciidae Atelorms pittoides Coracias garrulus Eurystomus glaucurus Upupidae Upupa epops

PICIFORMES Galbulidae Galbula leucogastra Bucconidae Monasa morphoeus Capitonidae Megalaima virens M. haematocephala Ramphastidae Selenidera langsdor ffi

1. Cranial shape

The general roundness of the fossil cranium in lateral view (Text-fig. 2A, B) is similar to that of Monasa morphoeus, Megalaima haematocephala (Text-fig. 8B), Selentdera langsdorffi and the Momotidae species. Most of the others have a cranium extending further back and with a slight upward taper, apparently caused by a forward shift of the foramen magnum and a shortening of the basiparasphenoid area so that the occiput faces more ventrally. This shift of the foramen would seem to be linked with a more upright perching stance by the bird. In Leptosomus and Eurystomus this shift is accompanied by a shortening of the cranium but this appears to be compensated for by an increase in the height, the cranium bulging noticeably at the top. In Upupa there is no apparent shift but the cranium is generally enlarged and rounded.

2. Lemporal fossae

It has already been indicated that the size and shape of the temporal fossae (site of origin of the M. adductor mandibulae externus) varies considerably from species to species. The two examples of Megalaima demonstrate this. The fossae of the smaller species (M. haematocephala, Text-fig. 8B) are little larger than those of Halcyornis, but those of the larger bird (M. vivens, Text-fig. 8D) are considerably elongated and extend across to meet at the back of the skull. In the Coraciiformes small fossae similar in size to those of Halcyornis but differing a little in shape occur on the various Momotidae species (Text-fig. 8C). In Leptosomus they are also small, but on other species show a variable tendency to become elongated towards the mid-line, almost meeting in Atelornis pittordes and the Alcedinidae (Text-figs 6B, 8F). The temporal fossae of Halcyornis extend well forwards anterodorsally,

ENGLISH LOWER EOCENE 165

approaching the orbit closely and reducing the spur above the postorbital process to a narrow ridge. A slight discontinuity in the floor of the fossae, adjacent to the process, seems to indicate a secondary area of muscle attachment, probably of the M. dermotemporalis, an accessory muscle of the neck used in head movements. This area is relatively poorly developed in most of the forms considered here but is rather more apparent in Megalaima haematocephala, Monasa morphoeus and Upupa epops, although very shallow in Upupa. The narrower ridge of the postorbital process in Leptosomus discolor and Selenidera langsdorffi appears to be due to an

Fic. 8. Diagrams of left lateral posterior views of crania to show variation and extent of temporal fossae. a, Covacias; B, Megalaima haematocephala; c, Momotus; D, Megalaima vivens; E, Larus; F, Alcedo.

166 AFFINITIES OF HALCYORNIS

increase in the M. adductor mandibulae externus rather than M. dermotemporalis. The relatively small size of the M. dermotemporalis in the species which actively hunt live prey, and its development in species which tend to take insects from a solid substrate or to take fruit, suggests that the relatively plastic shape of the fossae is linked with the type of feeding behaviour.

3. Interorbital bridge and dorsal groove

Dorsally the cranium of Halcyornis tapers to a relatively narrow interorbital bridge formed by the frontals (Text-fig. 1), and there is a distinct groove along the mid-line (Text-fig. 3, Pl. 1A). The skull roof also bears a series of fine striae radiating from the edges of the orbits, mostly directed posteriorly (Text-fig. 1).

Over the two orders as a whole the trend appears to be for the interorbital bridge to become gradually broader (Text-fig. 6). The bridge is fairly narrow on Aspatha gularis and Atelornis pittoides, but only in Blcedo atthis (Text-fig. 6A) does it approach the narrowness of Halcyornis. Milne-Edwards and Grandidier’s (1876) plate of the skull of the Madagascan Roller, Bractypteracias leptosomus (Coraciidae), shows that this too had a narrow interorbital bridge. In the Alcedinidae the width of the bridge varies (Text-fig. 6A, B), the narrow bridge apparently correlated with a slender bill. The dorsal groove is apparent in species of Momotidae, Alcedinidae, Upupidae, Leptosomidae and most Coraciidae; poorly defined in Atelornis pittoides, Monasa morphoeus and Merops apiaster; virtually absent in the Ramphastidae;

E F e H

Fic. 9. Diagrams of the left external auditory meatus. a, Halcyornis; B, Eurystomus glaucurus; c, Leptosomus discoloy; D, Covacias garrulus; E, Dacelo novaeguineae; F, Megalaima vivens; G, Monasa morphoeus; a, Momotus martit; approx. } nat. size.

ENGLISH LOWER EOCENE 167

and replaced by a ridge in the Galbulidae and Capitonidae. The striae, which we suspect may vary with age and the degree of ossification, are more or less apparent in most of these species.

4. Occipital region

The fossil shows a very definite curved occiput of fairly even width (Pl. 1C). The lateral ends (formed by the exoccipitals) terminate in a narrow ridge which is also the posterior edge of the auditory meatus, and this ridge curves abruptly at the top to form a small arch almost at the point where the ridge of the upper edge of the occiput terminates.

Most of the skulls examined show an occiput similar to that of Halcyornis, but its degree of curvature varies from one species to another. On most of them the ridge of the upper edge of the occiput terminates laterally in a triangular-shaped surface which slopes down towards the meatus, the upper arch of which is less acute than in Halcyornis, lower and more anterior in relation to the exoccipitals. The rollers Eurystomus glaucurus and Coracias garrulus (Text-fig. 8A), however, are much more like Halcyornis. Firstly, the occiput is very similar in shape and relative size; secondly, in F. glaucurus and to some extent in C. garrulus the triangular surface is very small and the upper edge of the meatus terminates in a small acute arch, near the end of the occipital ridge, very similar to that of Halcyornis (Text-fig. 9). In the latter some allowance must be made for probable erosion of these surfaces.

V. POSSIBLE CHARACTERS OF HALCYORNIS

When a fossil bird is known only from an incomplete skull, any ideas concerning its appearance or habits must necessarily be highly conjectural. It is, however, possible to make a few intelligent guesses by extrapolation from Recent species. The size of the fossil skull, compared with those of other birds, suggests that Halcyornis was probably about the size of a medium-sized thrush (Turdus sp.). The position of the foramen magnum suggests that the bird did not have the very upright posture of the present-day rollers (Coraciidae) and kingfishers (Alecinindae) but probably adopted a more horizontal body posture on a perch or on the ground, like that of the smaller barbets (Capitonidae). The narrow bridge between the orbits suggests that the bill was probably slender and more like a typical kingfisher’s than a roller’s. Most birds of this type which catch living food by rapid and vigorous movements have temporal fossae extending well back around the skull; but the temporal fossae of Halcyornts are fairly small and rounded, suggesting that this species may have taken food, probably small creatures or fruit, from the sub- strate or from the surfaces of vegetation rather than by seizing it rapidly in the air.

VI. GENERAL CONCLUSIONS

From an examination of affinities within the higher taxa it is apparent that Halcyornis toliapicus has been wrongly assigned to the Laridae and that its true affinities seem to lie with the Coraciiformes and Piciformes, more specifically with the former. The subsequent comparison with families within those two orders

168 AFFINITIES OF HALCYORNIS

appears to confirm that conclusion and to reinforce the view that the closest affinities are with the Coraciidae. The similarities are not sufficient, however, to justify referring it to that family, and it appears to show the characters of several families, and also exhibits its own peculiarities and combination of characters. Hitherto Coraciiformes have not been described from the Lower Eocene, but the momotid Uzintornis! has been described from the middle Eocene of Wyoming, the bucerotid Gezsleroceros from the Middle Eocene of Germany, the coraciid Geran- opterus from the Upper Eocene of France, and the alcedinid Protornis from the Upper Eocene of Switzerland. Halcyornis, because of its earlier origin, might be a repre- sentative of a group ancestral to more than one of the Recent families. As a matter of taxonomic convenience it seems best for the present to treat Halcyornis as the unique representative of a new family of the Coraciiformes, systematically not far removed from the Coraciidae: the Halcyornithidae.

VI. SYSTEMATIC DESCRIPTION Order CORACIIFORMES Family HALCYORNITHIDAE noy.

FAMILY DIAGNOSIS: Cranium not particularly inflated or rounded. Interorbital bar narrow and deep. Postorbital process small and narrow. Temporal fossae tapering posteriorly, occupying about two-thirds of total width of cranium, slightly curved where they border the edge of exoccipital region. Occiput semi-lunar with well-defined median ridge on parietals. Interorbital septum thick, with single optic foramen in low position. Wide angle between alaparasphenoids and rostro- parasphenoid, latter thick.

Genus HALCYORNIS Owen, 1846 DracGnosis: The only genus of its family.

TYPE SPECIES: Halcyornis toliapicus (Koenig).

Halcyornis toliapicus (Koenig, 1825) (Pls 1-3; Text-figs 1-9)

1825 Larus toliapicus Koenig, pl. 16, fig. 193. 1846 Halcyornis toliapicus (Koenig): Owen: 554.

Dracnosis: The only species of the genus. Ho.otyreE: Imperfect cranium in Brit. Mus. (Nat. Hist.) No. A 130.

LocaLITy AND HORIZON: London Clay (Ypresian), Lower Eocene of Sheppey, Kent, England.

1According to Dr. J. Cracraft (pers. comm.), Uintornis is not a motmot.

ENGLISH LOWER EOCENE 169

Vill. ACKNOWLEDGEMENTS

We wish to thank Dr. A. J. Charig and Dr. J. Cracraft for criticizing the manu- script; Mrs. T. E. D. Brendell for making the detailed line drawings; Mr. F. Howie for preparing the specimen and Mr. T. W. Parmenter for taking the photographs.

IX. REFERENCES

Jotyig, N. T. 1957. The head skeleton of the chicken and remarks on the anatomy of this region in other birds. J. Morph., Philadelphia, 100 : 389-436. 26 figs.

Koenic, E. 1825. Icones fossilium sectiles. 44 pp. 19 pls. London.

LYDEKKER, R. 1891. Catalogue of the fossil birds in the British Museum (Natural History). XII-368 pp. 75 figs. London.

Mitne-Epwarps, A. & GRANDIDIER, A. 1876. Histoive physique, naturelle et politique de Madagascar. 13, Histoive naturelle des oiseaux. tome 2, Atlas 1. 87 pls, 236 figs. Paris.

OweEN, R. 1846. A history of British fossil mammals and bivds. XLVI-560 pp. London.

Zusi, R. L. 1962. Structural adaptations in the head and neck in the Black Skimmer. Publ. Nuttall Orn. Cl. 3, 101 pp. Cambridge, Mass.

C. J. O. Harrison, Ph.D. Department of Zoology, Sub-depariment of Ornithology, ZooLocicaL Museum,

TRING,

HERTs.

C. A. WALKER,

Department of Palaeontology

British Museum (Natural History) CROMWELL Roap,

Lonpon, SW7 5BD

TINA, a

Halcyornis toliapicus BM(NH) No. A 130. Stereo views of cranium, X 2. A, dorsal; B, ventral; c, posterior.

Bull. Bry. Mus. nat. Hist. (Geol.) 21, 4 PEATE 1

PA Ee 2

Halcyornis toliapicus BM(NH) No. A 130. Stereo views of cranium, 2. A, right lateral; B, left lateral; c, frontal.

Bull. Br. Mus. nat. Hist. (Geol.) 21, 4 IIL ANTES, 2

PLATE 3

Ventral views of bird crania to show variations in basisphenoid elements.

A, Rhynochetos jubatus B, Fregata ariel

c, Puffinus diomedia

D, Larus argentatus

E, Halcyornis

F, Rhynchops niger

G, Merops apiaster

H, Coracias garrulus

1, Monasa morphoeus J, Dacelo novaeguineae K, Upupa epops

L, Coccyzus erythropthalmus various magnifications

Bull Br. Mus. nat. Hist. (Geol.) 21, 4 PLATE 3

rY/ td Me cy

"

OF ii = , | we

A LIST OF SUPPLEMENTS TO THE GEOLOGICAL SERIES OF THE BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY)

Pp. 213; 30 Plates; 2 fae oe (6. . Er-Nacear, Z. R. Stratigraphy and Planktonic Foraminifera of the

Royse U.A-R.. Pp. 291; 23 Plates; 18 Text-figures. Sry £10.

. Davey, R. J., DowniE, C., SARGEANT, W. A. S. & Wittiams, G.L. St Mesozoic and Cainozoic a eect Cysts. be 248; 28 he eees é figures. 1966. £7.

Appendix to Studies on 1 Mesozoic and Cainozoic = Dinofagllate cysts. 1969. 80p. ee . Exuiort, G. F. Permian to Palaeocene Calcaruote Algae (Dacycdadaceaen a Middle East. Pp. rrr; 24 Plates; 17 Text-figures. 1968. £5.12}. nage . Ruoves, F. H. T,, Austin, R. L. & Druce, E. C. British Avonian ( ferous) Conodont faunas, and their value in local and continental cor Pp. 315; 31 Plates; 92 Text-figures. 196g. {11 : . Cuitps, A. Upper Jurassic Rhynchonellid Brachiopods from Nort Europe. Pp. 119; 12 Plates; 40 Text-figures. 1969. £4.75. . Goopy, P. C. The relationships of certain Upper Cretaceous Tele special reference to the Myctophorids. Pp. 255; 102 Text-figures. 1969 . Owen, H. G. Middle Albian Stratigraphy in the ed Basin. e 3 Plates; 52 Text-figures. 1971. £6. e . Sippigul, Q. A. Early Tertiary Ostracoda of the family Teach from West Pakistan. Pp. 98; 42 Plates; 7 Text-figures. 1971. {£

ablle

BEE #s: Printed in England by Staples Printers Limited at their Kettering, Northants, establishment

‘U. GITMEZ

DINOFLAGELLATE CYSTS AND ACRITARCHS FROM THE KIMMERIDGIAN (UPPER JURASSIC) OF ENGLAND, SCOTLAND AND FRANCE

~~ y ~»

anor te |

FON i)

A> 3 (ae

Lar URAL Wy

BY GULDEN USMAN GITMEZ Hacettepe University, Ankara, Turkey AND

WILLIAM ANTONY S. SARJEANT

University of Saskatchewan, Saskatoon, Canada

Pp. 171-257; 17 Plates; 27 Text-figures; 6 Tables

BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) GEOLOGY Wolhen Now's LONDON : 1972

THE BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY), instituted in 1949, 1s issued in five series corresponding to the Departments of the Museum, and an Historical series.

Parts will appear at irregular intervals as they become ready. Volumes will contain about three or four hundred pages, and will not necessarily be completed within one calendar year.

In 1965 a separate supplementary series of longer papers was instituted, numbered serially for each Department.

This paper is Vol. 21, No. 5 of the Geological (Palaeontological) series. The abbreviated titles of periodicals cited follow those of the World List of Scientific Periodicals,

World List abbreviation Bull. Br. Mus. nat. Hist. (Geol.)

© Trustees of the British Museum (Natural History), 1972

TRUSTEES OF THE BRITISH MUSEUM (NATURAL HISTORY)

Issued 1 November 1972 Price £4.65

DINOPLAGELLATE CYSTS AND ACRITARCHS FROM THE KIMMERIDGIAN (UPPER JURASSIC) OF ENGLAND, SCOTLAND AND FRANCE

By G. U. Gitmez & W. A. S. Sarjeant

CONTENTS Page I. INTRODUCTION ‘ : : 5 9 : 175 II. LocatTioN AND DESCRIPTION OF SAMPLES . b : . ; 176 1. Dorset (South coast) . é : : F : : 176 2. Oxfordshire c : : : : : c c 180 3. Cambridgeshire . ‘ : F : j : . 180 4. Warlingham, Surrey. ; A : : : : 180 5. Isle of Skye (Staffin Bay) . : . : 6 181 6. Eathie Haven (South of @ascee ; : ‘ : 181 7. The Boulonnais . : : : : : : : 181 8. Normandy . é ; i 2 : : : 182 9g. Le Havre, Seine Eeieceace : > : 4 : : 183 10, Lorraine. 4 : : : 3 184 11. Mont Crussol (Rhéne vale) : . é : 184 12. The Jura Mountains (Southern French ies) ‘ c ; 184 III. SystTEMATIC SECTION 0 : 5 5 185 Cyst-Family Fromeaceae Sige & eae : : : 185 Genus Chytroeisphaeridia Sarjeant . : C : 185 C. chytroeides Sarjeant . : : 6 : 185 C. mantelli sp. nov. : : : ; ‘ 186 C. pococki Sarjeant c : : : : 187 Genus Fvomea Cookson & Eisenack . : : . 188 F. warlinghamensis sp.nov. . ‘ : : 188 Genus Tenua Eisenack emend. Sarjeant . > 5 189 T. capitata (Cookson & ee) : : : 189 T. echinata sp. nov. ‘ : ‘ 6 190 T.sp. . : 2 190 Cyst-Family Gorn relreyaeeae Sotennt & Deaae ; : 191 Genus Cyryptarchaeodinium Deflandre emend. Gitmez . 191 C. calcavatum Deflandre 5 5 4 3 IOI C. cf. calcavatum Deflandre . : ‘ 192 Genus Gonyaulacysta Deflandre emend. Sarena : 193 G. cauda sp. nov. . é : : : 193 G. cf. giusepper Morteanot : : ¢ : 194 G. globata sp. nov. : 6 : : . 195 G. longicornis (Downie) : : F : 197 G. cf. mamillifera (Deflandre) ‘ : ; 199 G. nuciformis (Deflandre) : 5 ; ; 200 G. perforans (Cookson & Eisenack) . : ; 202 G. systvemmatos sp. nov. : : : : 204 G. sp. A. : : c : : ° : 205

174 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

Genus

Genus

Cyst-Family

Genus

Genus

Genus

Cyst-Family Genus

Genus

Cyst-Family

DADNDHDAD wn

OS OOK

OB Ob

Leptodinium Klement Suen’, Sarjeant

L. acevas (Eisenack)

L. amabilis (Deflandre) .

L. cf. cvassinervum (Deflandre)

ESD:

Occisucysta Glas

O. evitti (Dodekova)

O. monoheuriskos sp. nov.

Microdiniaceae Eisenack emend. Seeeant ad Downie

Dictyopyxis Coakean & Beemer

D. aveolata Cookson & Eisenack

D. cf. reticulata (Valensi)

Meiourogonyaulax Sarjeant

M. staffinensis Gitmez 1970

M. dicryptos sp. nov.

M. pila sp. nov.

M. sp. :

Egmontodinium en nov.

E. polyplacophorum sp. nov.

Pareodiniaceae Gocht emend. Great & Downie

A pteodinium Eisenack ¢

A. cf. maculatum Eisenack & Cankeon

Imbatodinium Vozzhennikova

I. antennatum sp. nov.

I. cf. villosum erate

Uncertain

Proximate cyst sp. indet.

Cyst-Family Genus

Cyst-Family

Genus

Genus

Genus Cyst-Family

Genus Cyst-Family

Genus

Adnatosphaeridiaceae Seu & oer

Adnatosphaeridium Williams & Downie

A. paucispinum (Klement) :

Hystrichosphaeridiaceae Evitt emend. Serene and Downie

Cleistosphaevridium heme Dae. Gaeicant aud Williams : .

Caspar

Olccspiaeridaan Dewey & wv allems

O. pulcherrimum (Deflandre & Cookson)

Systematophora Klement 4

S. ovata sp. nov.

Uncertain

Stephanelytron eae en

S. vedcliffense Sarjeant

S. cf. vedcliffense Sarjeant

Endoscriniaceae Vozzhennikova saenil cane & Downie

Endoscrinium (idement),

206 207 208 209 211 212 213 215 215 216 218 219 220 220 221

223 223 228 224 224 224 225 226 227 228 229 231 231 231 232 232 233 233 233 234 234 234

235

235 235 235 235 237 237 237 237 237 238

239 239

FROM ENGLAND, SCOTLAND AND FRANCE 175

Spa : : : 239 Cyst-Family Hexagoniferaceae Sanjeant & Dawe : : 240 Genus Hewxagonifera Cookson & Eisenack . 6 : 240 H. jurassica sp.nov. . : . : : 240

Cyst-Family Muderongiaceae Neale & Sarjeant emend. Sarjeant and Downie . d : : 241 Genus Muderongia Cookson & Beedle 5 5 9 241 M. simplex Alberti , 241

Cyst-Family Nelsoniellaceae Eisenack emend. Sarjeant and Downie . : : : : 242 Genus Scriniodinium ilemedee ; 6 6 2 242 S. bicuneatum (Deflandre) : : é 5 242 S. dictyotum Cookson & Eisenack . : 6 243 S.sp. . - : : : 3 o 244 Genus Sivmiodinium Alberti c : ; : é 245 S. grossi Alberti. : é : : é 245 Cavate cyst sp. indet. A. . : Q : : : 245 Cavate cyst sp. indet. B. . : : : c : 246

Incertae sedis

Group Acritarcha Evitt ; : 247 Subgroup Acanthomorphitae Desusie Evict & Sonieeint : 247 Genus Micrhystridium Deflandre emend. Sarjeant : 247 M. vecuvvatum Valensi . c é 5 ; 247 M.sp.. : 247

Genus Solisphaeridium Sele. [easeaine & Pocock emend. Sarjeant : : : : 248 S. claviculorum (Deflandre) i . 248 Subgroup Netromorphitae Downie, Evitt & Sagiena é 249 Organism A. . 249 Subgroup Pteromorphitae acme, Evite & Sane . 249 Genus Ptevospermopsis W. Wetzel . 4 c 3 249 P. harti Sarjeant . : : : ; : 249 Subgroup Uncertain . : : ; ; ‘ : 250 Acritarch sp. indet. . : : 3 : : : 250 IV. STRATIGRAPHICAL DISTRIBUTION CHARTS . : ; : F 250 V. CONCLUSIONS : : : F ; : : P 6 250 VI. ACKNOWLEDGEMENTS ‘ : 3 i ‘ j : : 252 VII. REFERENCES . é : : : ; j , ; ‘ 253

I. INTRODUCTION

THIS paper contains an account of assemblages of organic-walled microplankton (dinoflagellates and acritarchs) from the Kimmeridgian of England, Scotland and France. Since an account has already been given of assemblages from the lowest Kimmeridgian, the Baylei Zone (Gitmez, 1970), attention is concentrated on the higher zones: however, additional records from the Baylei Zone are included and data respecting this zone is incorporated into the stratigraphical discussion.

Many of the samples examined were collected by the second author (W.A.S.S.), with the help or under the guidance of a number of other geologists—in Dorset, from Dr. J. C. W. Cope (University of Wales, Swansea) ; in the French Jura and the Boulonnais, from Professor Derek V. Ager (then of Imperial College, London; now of the University of Wales, Swansea) ; in Normandy and Le Havre, from Dr, Michel

176 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

Rioult (Université de Caen) ; in Lorraine, from M. Pierre L. Maubeuge; and in Skye, from Dr. Dennis Field (University of Nottingham). In addition, a series of samples from the Warlingham borehole of H.M. Geological Survey were made available for study, through the courtesy of the Director, Dr. F. W. Anderson and Dr. H. Ivimey- Cook; and samples from Cromarty and from Oxfordshire and Cambridgeshire were furnished respectively by Dr. W. D. Ian Rolfe (Hunterian Museum, University of Glasgow) and Dr. Robin I. Whatley (University of Wales, Aberystwyth).

Preparation and preliminary study of the samples was done by the first author (G.U.G.), who also prepared most of the diagrams and photographs. The results were worked out jointly. It was found that many species were represented by insufficient individuals for satisfactory description; repreparation and further study of the samples were therefore to have been undertaken. This was precluded by the destruction by fire of the upper floor of the Geology building of the University of Nottingham in late March, 1970; all wet and dry samples and many microscope slides were lost and the research programme of the second author (W.A.S.S.) so seriously set back that further work on the Kimmeridgian cannot now be envisaged for some years to come. In consequence, it was considered that, since so little is known of the assemblages from these levels, the data currently available should be published forthwith.

In many instances, samples examined did not yield assemblages; relatively pure limestones in particular proved unproductive, the bulk of the assemblages being obtained from clays or argillaceous limestones. Details are given here of the negative as well as of the positive results.

II. LOCATION AND DESCRIPTION OF SAMPLES

Since full stratigraphic details and sample numbers for the specimens from the Baylei Zone have already been given (Gitmez, 1970), these are summarized only briefly here. Location, position and specimen numbers for higher horizons are given in full. Nottingham University sample numbers are quoted, since these were employed in all notes made by the authors. [The samples themselves were all destroyed in the conflagration. |

The history of the ammonite zonation proposed for the Kimmeridgian of England is outlined in Table 1. The zonation here adopted is based on that of Arkell (1956) and incorporates the modifications introduced by Cope (1967).

(1) Dorset (South coast):

The clay formation known as the Kimmeridge Clay represents long-continued deposition of muddy sediments. It is not a uniform deposit, for it includes several lithological types which may alternate in rapid succession; in addition to thick clays there are thin bands of mudstone and several prominent “‘stone-bands’’, formed by limestones of variable degrees of purity (see Arkell, 1933; Cope, 1967).

The Kimmeridge Clay is well exposed on the Dorset coast, through faulting in Ringstead Bay and again near Osmington Mills which causes repetition of the succession. Around the type locality of Kimmeridge, the Kimmeridge Clay reaches its maximum thickness (495 m), but towards the west, in the Weymouth district, the thickness is nearly halved and inland it reduces to go m.

177

FROM ENGLAND, SCOTLAND AND FRANCE

LOWER

MIDDLE

s}ug, soley

topAeq eIuoZOIg

NVIGUYOAXO

toyAeq eru0z01g te[Aeq eruozoIg

topAeq eIuojoIg

snyeqorinus aoopow Ad eIuesery as0pourAd erIueseyy a90pourAd erIueseyy ss0pourAd erueseyy syiqeznur soprourydeajsoorpny sljiqejnu eruese yy SI]IqeyNU eruaser ST]IGe}NU eIUISeYT snxopnoe snueydeysoorny SUIQeIN]A,, SISUOIOPOISSINe stjIqezynuropnesd sijiIqeznuropnesd sijiIqeynuopnesd snueydeysoorny snueydojsoorpny snueydojsoorpny snueydoajsoorny suesojo BULISOACIS PISOAPID SUPISOALIS VISOAPID) (saprozyourydsozesita ) SOUOZ PISOAPIN) stdsei0q4S sopzeurzoog SPSIS PISOAPID SNTIT VISOARID) snynz10s (soprojourydsozesit A ) so}izeuroeg ‘dds sozruejdqns stsuoA9q} eo seprozourydsozesit A sisuoAo]} POT SISU9A OM YOS}EIUI (soprozoutydsoze 811 A ) $o}1}eSIT A sopzeurz00q sipueis soytuefdqns sniojIpou TuU0}soTpNny sisuoA9T} ROM soprozyourydsozesit, (seqTTe4TV) sezzeUIQOeq soyyuedqns snzeurjoed saqizeurz00q snzeurjoed sozizeur}00g snzeurzoed soqrzeurjOI9g snzeurjoed soqz1zeuI}I0g epunyzoL PIAC[ACG epunyzor eIAo[Aeg epunyzoL eCIAOQTAeG (epunjor etacjAeg) (snuetseyjed eraojaeg) soprorseyjed viAojardg soprorseyed eraojAeg soprorserjed eraocjaeg (soprotseyyed eraojAeg) NVIGNVILUOd £961 adog 10932 9S61 S761 br61-€ 161 gSgi—9Ser peyipour ‘9S6r [[ayxIV TexrtVv WOSIOAPO NT PIPyIes jeddo

ULISPILOWUWITY 94} JO UOIJLOYISse[O [euOz 9Yy3 Jo AIO4STY OY,

I ATAV]

178 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

The samples studied were collected from two areas, the Isle of Purbeck and the Weymouth district (Text-fig. 1). Fourteen samples were examined from the Isle of Purbeck. (The colours are given according to the ““Rock Colour Chart”, produced by the Geological Society of America [1963] and based on the Munsell System.)

Autissiodorensis Zone:

I. KD 221—Shale containing shell fragments, medium grey in colour (N5), from c. 4 ft above Washing Ledge Stone Band, Kimmeridge (National Grid Reference: 909791).

2. KD 224—Clay containing shell fragments, medium dark grey in colour (N4), from 30 ft above Maple Ledge Stone Band, Kimmeridge. (National Grid Reference: 909788.) [Very few microfossils were obtained from this sample. |

3. KD 225—Clay containing shell fragments, medium dark grey in colour (N4), from immediately below the cementstone at the junction of the Autissiodorensis and Elegans Zones, Kimmeridge. (National Grid Reference: 909789.)

Elegans Zone:

4. KD 227—Clay containing shell fragments, medium dark grey in colour (N4), from 25 ft below the Yellow Ledge Stone Band, Kimmeridge. (National Grid Reference: 917780.) [This sample contains few microfossils. ]

Scitulus Zone:

5. CD 229—Clay, brownish grey in colour (5 YR 4/1), from 6 ft below the Cattle Ledge Stone Band, Cuddle. (National Grid Reference: 917780).

Wheatleyensis Zone:

6. CH 231—Clay, medium grey in colour (N5), from 22 ft below the Black Stone, Clavells Hard. (National Grid Reference: 920778.)

7. RD 234—Clay, medium grey in colour (N4), from 13 ft above the Rope Lake Head Stone Band, Rope Lake Head. (National grid reference: 934785.) [No assemblage was obtained from this sample. ]

Boundary of Pectinatus-Hudlestoni Zones:

8. FD 236—Marl, medium dark grey in colour (N4), from + mile west of Fresh- water Steps, (National grid reference: 946773.)

Pectinatus Zone:

9g. FD 237—Marl, containing shell fragments, medium dark grey in colour (N4), from 2 ft above the Freshwater Steps Stone Band, + mile west of Freshwater Steps. (National grid reference: 946773.)

10. ED 240—Marl containing shell fragments, medium grey in colour (N5), from 30 ft above the Freshwater Steps Stone Band, Egmont Bight. (National Grid reference: 948772).

ir. ED 242—Clay, medium dark grey in colour (N4), from 60 ft above Freshwater Steps Stone Band, Egmont Bight. (National grid reference: 948772).

12. HC 243—Clay, medium dark grey in colour (N4), from c. 100 ft below the Rotunda Nodules, in the base of Hounstout Cliff. (National grid reference: 951773).

Rotunda Zone:

13. CP 245—Clay containing shell fragments, medium light grey in colour (N6), from the Rotunda Nodule Bed, Chapmans Pool. (National Grid Reference: 956772).

14. HC 246—Clay, medium grey in colour (N5), from 140 ft below the Massive

FROM ENGLAND, SCOTLAND AND FRANCE 179

Bed, Hounstout cliff (National grid reference: 950774). [Few specimens were ob- tained from this sample. ]

ue Hounstout i | S Cliff at ares re) gmon fF Egmont Bay Point = w =: E s Pol = 2S 2 \ il Houns tout Cliff ao = 4 Freshwater o 2 = 55 8 (lili 2) a vii 4 @ Rope Lake Hole f wy Ih lavell’s Hard a s 2 f ae “i js o \I n x { Kimmeridge Bay e = | a =) lk: ic oO a o 7) Oo fo) 5 Clavell’s Hard & as teh = wo | = = Yellow a S || Ledge a Cuddle a ‘Stone Band m5 S x ] a Ee | oS Maple mo] a | S | | Ledge a % ® {| Stone Band : & a Kimmeridge Bay \ =f\Fossil Forest 2 © y | | Washing | E S \ Ledge Ay cS > N (pf Stone Band Lulworth Cove | 5) 2 ie 5 o- Se rat vo -——\"EH Durdle Door % —SSSaae we 9° 8 =) <x

Fic. 1. Sketch map of the Isle of Purbeck, showing the positions of the type sections of the Kimmeridge Clay. Section along the cliffs of Kimmeridge Clay (modified from Arkell, 1933).

180 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

In the Weymouth District, the Jurassic rocks have been folded and faulted along a N-Sline. The Kimmeridge Clay, after a gap, reappears in Ringstead Bay, rising from the sea and partly hidden by slips of Chalk and Greensand. In this part of the Dorset Coast the Kimmeridge Clay is overlying conformably the Ringstead Coral Bed, at the top of the Oxfordian.

Around Osmington Mills, the exposed clay in the cliffs mainly belongs to the Mutabilis and Pseudomutabilis Zones: it is quite deeply weathered and much slipped. The Cymodoce and Baylei Zones, with the Ringstead Coral Bed below, appear from time to time in the foreshore, as a result of periodic stripping of shingle by storms. An account of samples from the Baylei Zone at this locality was given earlier (Gitmez, 1970).

(2) Oxfordshire:

The Kimmeridgian, together with the Portland and Purbeck Beds, occupies a small area, being partially concealed by the overlap of the Cretaceous strata. Three samples from the Pallasioides Zone were examined; these were collected from a quarry at Littleworth, Wheatley (National grid reference: 595055).

t. LO 352—Clay, medium light grey in colour (N6), from 15 ft below the

Wheatley Sands. 2. LO 353—Clay, medium grey in colour (N5), from the quarry. 3. LO 360—Clay, light olive grey in colour (5 Y 6/1), from the top of the Kimmeridgian, Littleworth. [No assemblage was obtained from this sample. | (3) Cambridgeshire:

There are a few exposures of the Kimmeridge Clay in the north west of this area. Only one sample, from the Baylei Zone, was examined from this district: for details see Gitmez, 1970.

(4) Warlingham, Surrey:

The Warlingham borehole was drilled, under contract, in the Geological Survey programme of boring in 1956-1958 and was sited in a field beside the Woldingham Road, Warlingham (National Grid reference: TO 34765719). The boring commenced in Middle Chalk and passed through the Cretaceous and Jurassic rocks, terminating in the Lower Carboniferous at a depth of 5001 ft. At this locality the Kimmeridge Clay is 703 ft thick (between 2284 ft and 2987 ft in depth).

Twenty-nine samples, at 25 ft intervals, from the Rotunda Zone to the Mutabilis Zone, were studied for their organic-shelled microplankton content. These Kimmeridge Clay samples are between light bluish grey and medium bluish grey in colour (5 B 6/1), representing the zones as follows:

Rotunda Zone: 1. WB 29 from 2285’7 ” depth. 2. WiB28. 5.8 2310/6 tudepth. 3. WB27 ,, 23350” depth. Pectinatus Zone: 4. WB 26 from 2359’9 ” depth. 5. WB25 ,, 2384’9” depth. [Few specimens obtained. ] 6. WB24 ,, 2409’9” depth.

FROM ENGLAND, SCOTLAND AND FRANCE 181

Hudlestoni Zone: 7. WB 23 from 2434’6 ” depth. 8. WB22 ,, 24596” depth. [Few specimens obtained. ] 9. WBaxi ,, 2485’0” depth. [Few specimens obtained. | Wheatleyensis Zone: 10. WB 20 from 2510’0 ” depth. om, WBig-,, 25353 ° depth. mw. WBI8 ,, 25600” depth. Scitulus Zone: 13. WB17 from 2584’9 ” depth. 14. WB16 ,, 2610’0” depth. Elegans Zone: 15. WB15 from 2635’3 " depth. Autissiodorensis Zone: 16. WB 14 from 2660’1 " depth. ig WBI3 ,, 2068473” depth. 18. WBiz2 ,, 27093” depth. [Few specimens obtained.] Eudoxus Zone: mg. WB 11 from 2734‘11 ” depth. ga, WBi0 ,, 2760'5 "depth.

ae WE 9 ,, 2785'2" depth. 22, WE 8 ,, 28106” depth. Boe WE 7 ,,.. 28347 ” depth. ae WE 6 ,; 2860%0.” depth. ae WR 5 ,, 2885/1 " depth. 20, WB 4 ,, 29106.” depth.

Mutabilis Zone:

27. WB 3 from 2935’2 ” depth.

ze WB 2 ,, 2950'5 ” depth:

aoe WB i ., 2984’7°" depth. (5) Isle of Skye (Staffin Bay):

Three samples were examined from the Baylei Zone of Staffin Bay; these have been fully described by Gitmez, 1970.

(6) Eathie Haven (South of Cromarty):

The Kimmeridgian strata are visible in a narrow shore strip exposed only at low tide. The beds consist of carbonaceous shales, sandstones, bituminous shales and limestones. The total thickness has been calculated as approximately 38 m (Waterston, 1951, p. 33); the apparent thickness is less than the real thickness, because of the displacements consequent upon folding, which render it difficult to determine the thickness accurately (Text-fig. 2 a, b). Only one sample, from the Cymodoce Zone, was examined:

CS 421—Shale, olive black in colour (5 Y 2/1), from the first Meleagrinella Band of Eathie. (National grid reference: 778636.)

(7) The Boulonnais, Pas-de-Calais, France: The Kimmeridgian rocks are well exposed along the coast of the Boulonnais (see

182 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

Oo . . cS a Middle Old Red Sandstone

Kimmeridgjan

Fic. 2. A. Sketch map showing the geology around Eathie, Cromarty (after Waterston, 1952).

Ager and Wallace, 1966a, b). Eight samples were studied, seven of which (obtained from the basal formations, which are the equivalent of the Baylei Zone in Dorset) were described in a previous paper (Gitmez, 1970). In addition, one was obtained from the Subplanites Zone sensu Arkell, here considered probably equivalent to the Scitulus Zone of Cope, 1967.

CC 453—clay, light grey in colour (N7), from Argiles de la Créche, north of Cap de la Créche. No microfossils were recovered. (8) Normandy:

Only the lower Kimmeridgian is represented in Normandy: it appears beneath the unconformable Cretaceous and comprises clays and limestones with ammonites

FROM ENGLAND, SCOTLAND AND FRANCE 183

indicative of the Mutabilis, Cymodoce and Baylei Zones. Two samples were studied, one from the Baylei Zone (described by Gitmez, 1970), the other from the Cymodoce Zone:

BN 179—Marls, light olive grey in colour (5 Y 6/1), Benerville, Normandy. (9) Le Havre, Seine Inférieure:

On the shore at Cap de la Héve, clays and limestones of Kimmeridgian age are exposed, beneath the Cretaceous unconformity, at the foot of the cliffs. Three

Mutabilis Zone

_Jt.6ins. —~Al-= 1-7-7 - 7 - Second Limestone

lft é6ins. —> £

t 1Si fit: 4

lft. > Gop First Limestone

12 ft-

£ 2 ft. —> Meleagrinella Band S o Vv ° me} ° f 250 ft U 5 ; So ot -* |Astarte Muds ----F--- === || Olive Black Shale e ‘ g ‘\Green Grey Mudstone a ~ o a

Fic. 2. B. The section of the strata at Eathie, Cromarty (after Waterston, 1952).

184 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

samples were studied: two of these, from the Baylei Zone, have already been described (Gitmez 1970); the third sample was from the Mutabilis Zone.

HF 395—Marls, light olive grey in colour (5 Y 6/1), from the Marnes a ammonites, Cap de la Heéve.

(10) Lorraine:

In the east of the Paris basin, the Cretaceous usually rests on the Oxfordian; occasionally, however, some Lower Kimmeridgian rocks are still present below the unconformity. Only one sample was obtained from Lorraine:

LF 368—Marl, yellowish grey-light olive grey in colour (5 Y 7/1), collected from the Cymodoce Zone, about 2 ft above the Calcaires a Astartes. Roadside, about 1 km east of Gondrecourt. [This sample contains very few microfossils. | (1r) Mont Crussol (Rhone Valley):

The Jurassic rocks of Mont Crussol comprise a continuous and well-exposed, dominantly calcareous sequence from Upper Bathonian to Tithonian (see Karvé- Corvinus, 1966). The Middle and Lower Tithonian (which is equivalent to the Upper and Middle Kimmeridgian sensi anglico) is represented by limestones of various sorts. Thick limestones form the Kimmeridgian stage in its restricted, Continental acceptation (the Tenuilobatus, Pseudomutabilis and Beckeri Zones representing the Lower and Middle Kimmeridgian, sensu anglico).

Eight samples were examined, the first sample yielding very few microfossils, the others none at all;

1. MR547—Limestone, yellowish grey-light olive grey in colour (5 Y 7/1), from the lower boundary of the Platynota Zone (?Baylei Zone), small quarry above the west side of the Ravin d’Enfer.

2. MR548—Limestone, light grey in colour (N7), from the lower part of the Ataxioceras Zone (Cymodoce-Mutabilis Zones), same locality.

3. MR549—Limestone, very light grey in colour (N8), from the top of the Ataxioceras Zone, same locality.

4. MR550—Limestone, pinkish grey in colour (5 YR 8/1), from the Idoceras balderum Bed, same locality.

5. MR552—Limestone, yellowish grey in colour (5 Y 8/1), from the fossil band at the base of Pseudomutabilis Zone (i.e. Autissiodorensis Zone), ridge top above the Carriere Mallet.

6. MR553—Limestone, yellowish grey-light olive grey in colour (5 Y 7/1), from the Pseudomutabilis Zone, ridge top.

7. MR554—Limestone, light olive grey in colour (5 Y 6/1), from the lower boundary of the Beckeri Zone (i.e. Autissiodorensis-Elegans Zones), ridge top.

8. MR555—Limestone, pinkish grey in colour (7 YR 8/1), from the Beckeri Zone, near the summit of the ridge.

(12) The Jura Mountains (Southern French Jura):

As the original type locality for Alexander von Humboldt’s “‘Jurassic’’, this region is of particular interest. The most important recent work has (rather unexpectedly) been done by English geologists (see Ager and Evamy, 1963). The sequence is again predominantly calcareous and exposure is intermittent but adequate. Three

FROM ENGLAND, SCOTLAND AND FRANCE 185

samples from the Oignon Beds and two samples from the Virieu Limestone were examined:

1. OF 485—Limestone (a well-bedded calcilutite), yellowish grey in colour (5 Y 8/x), from the road side, Montard d’Oignon (type locality). Mutabilis Zone.

2. OF 486—Limestone, yellowish-light olive grey in colour (5 Y 7/1), from the top of the Oignon Beds, beneath a pisolite; roadside west of Lac du Chavoley. Mutabilis Zone.

3. OF 487—Pisolitic limestone, yellowish grey in colour (5 Y 8/1), from the junction of Oignon Beds and Bedded Virieu, 2 km north of St. Germain-de-Joux (Mutabilis Zone). [This sample contains few microfossils. |

4. BV 488—Limestone, greenish grey in colour (5 GY 6/1), from the base of Bedded Virieu (Lower Kimmeridgian), calcilutite above pisolite.

5. MV 489—Limestone, pinkish grey in colour (5 YR 8/1), from the Massive Virieu (?Upper Kimmeridgian), 200 yards west of Virieu-le-Grand (type locality). [The sample yielded very few microfossils. |

Ill. SYSTEMATIC SECTION Class DINOPHYCEAE Pascher Sub-class DINOFEROPHYCIDAE Bergh Order DINOPHYCIALES Lindemann Cyst-Family FROMEACEAE Sarjeant & Downie, 1966

Genus CHYTROEISPHAERIDIA Sarjeant, 1962

emend. Downie, Evitt & Sarjeant, 1963 Chytroeisphaeridia chytroeides Sarjeant, 1962b

Plate 1, figure 2

1962b Leiosphaeridia (Chytroeisphaeridia) chytroeides Sarjeant, 493-4, pl. 70, figs 13, 16, text- figs 11-12, tables 2-3.

1963 Chytroeisphaeridia chytroeides (Sarjeant); Downie, Evitt & Sarjeant, 9.

1964a Leiosphaeridia chytroeides Sarjeant; Sarjeant, table 3.

1964 Chytroeisphaeridia chytroeides (Sarjeant); Downie & Sarjeant, 103.

1967b C. chytroeides (Sarjeant); Sarjeant, table III.

1968 C. chytroeides (Sarjeant); Sarjeant, pl. III, fig. 10, table 2B.

1970 C. chytroeides (Sarjeant); Gitmez, pl. 14, fig. 5, table 4.

1970 C. chytroeides (Sarjeant); Gocht, 152, pl. 34, figs 20-24.

FIGURED SPECIMEN: I.G.S. slide PK10zA: Sample WB 2, Kimmeridge Clay. H.M. Geological Survey Borehole, Warlingham, Surrey, at 2959 feet 5 in. depth. Lower Kimmeridgian (Mutabilis Zone).

DIMENSIONS: Figured specimen: length (apex lacking) 45u, breadth 48u. Range of the English specimens: length (apex lacking) 12-72u, breadth 18-80u,, measured specimens 1029 innumber. Range ofthe Scottish specimens (40 specimens measured) : length (apex lacking) 20-50u, breadth 22-75u. 161 specimens from French assem-

186 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

blages were measured: length (apex lacking) 16-62u, breadth 22-65. There is no significant difference between the size of specimens from different zones in the Kimmeridgian or between English, Scottish and French specimens. The Kim- meridgian specimens exhibit a somewhat wider dimensional range than do the Oxfordian specimens; diameters of the latter, as quoted by Sarjeant (1962b), are 30-60u.

OBSERVED RANGE: Kimmeridgian (Baylei to Pallasioides).

TOTAL KNOWN RANGE: ?Lower Bathonian: certainly Callovian (Mariae) to Kimmeridgian (Pallasioides).

REMARKS: This species occurs in moderate abundance in all Kimmeridgian assemblages from England, Scotland and France, being most abundant in the Lower Kimmeridgian, numerically somewhat reduced in the Upper Kimmeridgian (Pallasioides Zone).

Chytroeisphaeridia mantelli sp. nov. Plate 1, figures 3-4; Plate 12, figure 3

DERIVATION OF THE NAME: Named in honour of Gideon Algernon Mantell, pioneer of the study of microplankton.

D1aAGnosis: Shell subspherical to elongate. The periphragm is coarsely granular and bears an irregular scatter of tubercles and of low knobs, giving it a somewhat warty appearance. An apical archaeopyle is present, with slits extending posteriorly along presumed reflected sutures, producing a ragged appearance. The operculum most often remains attached to the ventral side of the shell.

HoLtotyre: I.G.S. slide PK116, Sample WB13, Kimmeridge Clay, H.M. Geological Survey Borehole, Warlingham, Surrey, at 2684 ft 2 in. depth. Lower Kimmeridgian (Autissiodorensis Zone).

PARATYPE A: I.G.S. Slide PK114, Sample WB 13.

PARATYPE B: BM(NH) slide V.56338 (1) sample CH 231, from 22 ft below the Blackstone, Clavells Hard, Dorset. Middle Kimmeridgian (Wheatleyensis Zone).

DIMENSIONS: Holotype: overall length 65u, breadth 60u. Paratype A: overall length (apex lacking) 60u, breadth 65u. Paratype B: overall length (apex lacking) 65u, breadth 7ou. Range of Lower Kimmeridgian specimens: length (apex lacking) 25-75u., breadth 30-65u, measured specimens 12 in number. Range of Middle Kimmeridgian specimens: length (apex lacking) 40-80y, breadth 50-75u, measured specimens 16 in number. Range of Upper Kimmeridgian specimens: length (apex lacking) 23-70, breadth 28-75u, measured specimens 12 in number.

According to these measurements, the species attained its largest size in the Middle Kimmeridgian.

DescripTIon: The thick shell wall is apparently composed of two layers: the inner layer thin, the outer layer making up almost the entire wall thickness. The ornamentation of the periphragm is of three types: coarse granules, small irregularly

FROM ENGLAND, SCOTLAND AND FRANCE 187

formed lumps (verrucae) and rounded tubercles, irregularly scattered. A tabulation is indicated only by the slits; there is no suggestion of a cingulum. The sulcal notch was seen only in specimens in which the operculum was completely lost; it was not perceptible in specimens with the operculum still attached. This is considered to indicate that when the operculum is present, it is attached to the ventral side of the cyst.

REMARKS: C. mantelli differs from previously described species of the genus in its relatively thick wall, the ornamentation of the periphragm and the form of its apical archaeopyle. The most similar species is C. euteiches Davey (1969), from the Cenomanian; but the shell wall of this new species is not so thick as in C. euteiches (2-3u as quoted by Davey). Although C. euteiches has an apical archaeopyle, it is angular in outline and generally narrower, whereas in C. mantelli the archaeopyle is characteristically wide, with deep slits passing posteriorly from its margin. (Davey mentioned that the apical archaeopyle of C. euteiches also has small slits extending posteriorly from the margin.)

This new species, in its surface ornamentation and wall structure, also shows a broad accord with the diagnosis of Tenwa as emended by Sarjeant (1968b). Since cingulum and sulcus are not indicated and since, although the cyst wall shows a considerable ornament, spines are not present, it was allocated to the genus Chytroeisphaeridia. However, it should be noted that the form of the archaeopyle is closer to that of Tenua than to that of typical species of Chytroeisphaeridia. This species is thus intermediate in morphology between the genera Chytroeisphaeridia and Tenua.

C. mantelli was recorded from all zones of the Kimmeridge Clay; it was more abundant in the Middle Kimmeridgian (Eudoxus to Elegans Zones) than in the other subdivisions of the Kimmeridgian. Thirty-seven specimens from England and five specimens from France were examined; it was not observed in the Scottish assemblages.

Chytroeisphaeridia pococki Sarjeant, 1968 Plate 1, figure 5 1965 Chytroeisphaeridia sp. Sarjeant, pl. 1, fig. 13.

1968 Chytroeisphaeridia pococki Sarjeant, 230, pl. 3, fig. 9. 1970 C. pococki Sarjeant; Gitmez, pl. 9, fig. 7, pl. Io, fig. 3, table 4.

FIGURED SPECIMEN: BM(NH) Slide V.53961(3). Sample SC 444, from Great Ouse River Board Pit, Stretham, Cambridgeshire. Lower Kimmeridgian (Baylei Zone).

Dimensions: Figured specimen: length (apex lacking) 18u, breadth 22u. Range of Lower Kimmeridgian specimens: length (apex lacking) 18-85, breadth 22-78y, measured specimens 178 in number.

Range of Middle Kimmeridgian specimens: length (apex lacking) 35-75, breadth 35-80, measured specimens I13 in number.

Range of Upper Kimmeridgian specimens: length (apex lacking) 28-80y, breadth 35-80u, measured specimens 61 in number.

B

188 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

There is no difference in dimensions between the English, Scottish and French specimens. Lower and Middle Kimmeridgian specimens are of comparable dimensions to those of the Oxfordian holotype (dimensions, as quoted by Sarjeant; length [apex lacking] 45u, breadth 55,).

OBSERVED RANGE: Kimmeridgian (Baylei to Pallasioides).

TOTAL KNOWN RANGE: Callovian (Lamberti to Mariae) and Kimmeridgian (Baylei to Pallasioides) of Europe; Upper Jurassic of Canada.

REMARKS: C. fococki is present in the Kimmeridgian assemblages from England, Scotland and France: although present at all levels, it was found to be most abundant in the Lower and Middle Kimmeridgian, being especially common in the Warlingham borehole samples.

Genus FROMEA Cookson & Eisenack, 1958 Fromea warlinghamensis sp. noy. Plate 1, figures 6, 8; Plate 9, figures 5-6

DERIVATION OF THE NAME: Named after the type occurrence in the Warlingham borehole, Surrey.

D1aGnosis: Cyst broadly ovoidal to almost spherical, with a relatively thick wall. Archeopyle apical, subhexagonal to almost circular; a rather rounded projection, not always observable, appears to be a sulcal tongue. The shell surface is densely and coarsely granular. A cingulum is indicated by indentations at the margins; less frequently, it is traceable across the surface.

Ho.otyPeE: I.G.S. slide PK115, Sample WB 13, Kimmeridge Clay, H.M. Geological Survey borehole, Warlingham, Surrey, at 2684 ft 3in. depth. Lower Kimmeridgian (Autissiodorensis Zone). PARATYPES (a) BM(NH) slide V.56339(2), sample CH 231, slide V.56339(2) Kimmeridge Clay 22 ft below the Black Stone (Wheatley- ensis Zone), Clavell’s Hard, Dorset. (b) BM(NH) slide V.56340(1), sample LO 353, top of Kimmeridge Clay (Pallasioides Zone), Littleworth, Oxfordshire.

Dimensions: Holotype: length (apex lacking) 47-3, breadth 4o-5u. Paratype (a): length (apex lacking) 81-5u, breadth 79:5u. Paratype (b): length (apex lacking) 71y, breadth 65u. Range of specimens observed (25 in number); length (apex lacking) 42-95y, breadth 40-88y.

DESCRIPTION: The pronounced original sphericity characteristic of this species results in a variety of structures produced by compression ; paratype (a) in particular, shows an irregular series of bulges. The granules are of variable size and are characteristically circular; the distinctly polygonal granules on the outbulges on paratype (a) probably result from pressure by mineral grains. The wall appears to be composed of a single layer: it may be as much as 2—3y in thickness.

OBSERVED RANGE: Kimmeridgian (Autissiodorensis to Pallasioides Zones).

REMARKS: This new species is distinguished from Fromea amphora, the only other species to date placed in this genus, by its much more spherical shape and coarsely

FROM ENGLAND, SCOTLAND AND FRANCE 189

granular surface. The known range of the latter species is Barremian to Albian (Cookson and Eisenack, 1958): the stratigraphic hiatus between the type species and this Upper Jurassic species may well be removed by future studies.

Genus TENUA Eisenack, 1958c emend. Sarjeant, 1968 Tenua capitata (Cookson & Eisenack, 1960b) comb. nov. Plate 1, figures 11-12

1960b Hystrichosphaeridium capitatum Cookson & Eisenack, 252. pl. 39 fig. 9. 1964 4H. capitatum Cookson & Eisenack; Sarjeant, table 3.

1964 H. capitatum Cookson & Eisenack; Downie & Sarjeant, 120.

1970 Tenua cf. capitata (Cookson & Eisenack); Gitmez. pl. 10, fig 4. table 4.

DESCRIPTION: Cyst spherical to elongate, with an apical archaeopyle and rounded antapex, bearing processes whose length sometimes reaches to one-third of the shell breadth and which number around fifty. The processes are hollow, capitate or briefly bifurcate, their distribution appearing random.

FIGURED SPECIMEN: BM(NH) slide V.56341(1) sample HF 185, from the Exogyra Marls, c. 1 m above the Upper Hard Band, Cap dela Héve, Le Havre. Lower Kimmeridgian (Baylei Zone).

Dimensions: Range of the English specimens: length (apex lacking) 30-65y, breadth 22-50u (7 specimens measured). Range of the French specimens: length (apex lacking) 30—40u, breadth 22—33u (2 specimens measured). 3 specimens from the Scottish assemblages were recorded and measured: length (apex lacking) 30-58y, breadth 30-58. Overall range of process length (all localities) 3-10. There is not much difference between these dimensions and the dimensions of the Australian specimens quoted by Cookson and Eisenack (length 64-66y, breadth 28—44u,, process length 8).

OBSERVED RANGE: Kimmeridgian (Baylei to Mutabilis). TOTAL KNOWN RANGE: Jurassic (Oxfordian to Kimmeridgian).

REMARKS: This species, under the name of Hystrichosphaeridium capitatum, has previously been recorded from the Oxfordian to Kimmeridgian of Australia by Cookson and Eisenack. It is transferred to the genus Tenwa on the basis of shell outline, the presence of an apical archaeopyle, and the form and number of the processes. The processes do not clearly reflect any tabulation.

Well-preserved specimens were observed in moderate numbers in samples from the Lower Kimmeridgian only; nine specimens from the Baylei Zone of England and three specimens from that zone in France; three specimens from the Cymodoce Zone of Scotland; and two specimens from the Mutabilis Zone of England were recorded.

These specimens are similar to that figured by Cookson and Eisenack. Although the number of the processes appears greater than in the Australian specimens, it was not possible to make precise comparisons since the number of processes was not mentioned by Cookson and Eisenack.

190 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

Tenua echinata sp. nov. Plate 1, figures 1, 9 1969 Tenua sp. Gitmez, 245-6 pl. 8, fig. 3, text-fig. 3, table 4.

DERIVATION OF THE NAME: Latin, echinatus, spiny, prickly; referring to the spiny surface of the shell.

Di1aGnosis: Cyst spherical, subspherical or broadly ovoidal, covered with spines and looking like a prickly ball. Spines very short, broad-based and conical, uniformly distributed over the whole surface, over 200 in number. There is no indication of tabulation, cingulum or sulcus. Archaeopyle usually present, apical in position; the operculum usually remains attached on one side but is sometimes completely lost.

HototypPEe: BM(NH) slide V.52796(1). Sample OM 131, from the base of the Kimmeridge Clay, Liostrea delta Bed, Osmington Mills, Dorset. Lower Kimmeridgian (Baylei Zone).

PARATYPE: I.G.S. slide PK.119, sample WB 16, from H.M. Geological Survey borehole, Warlingham, Surrey, at 2610 ft depth. Middle Kimmeridgian (Scitulus Zone).

DIMENSIONS: Holotype: overall length 50u, breadth 50u; length without apex 43u.; length of the spines 2u.

Paratype: overall length 72u, breadth 60; length of the spines 2-5u.

Overall range of the English specimens (15 specimens measured): length 45—90u, breadth 40-80u, length without apex (3 specimens were observed without apex) 43-60u; length of the spines 1-5-2-5u.

Dimensions of the single French specimen encountered: length (apex lacking) 38u, breadth 4ou; length of spines 2y.

DescripTIon: The shell surface is smooth. The apex is typically detached in archaeopyle formation. The margin of the archaeopyle is roughly polygonal, with tears extending from the angles along the presumed lines of a reflected tabulation. When the operculum remains attached on one side, this side is probably the ventral side. The spines appear to be solid.

OBSERVED RANGE: Kimmeridgian (Baylei to Pallasioides Zones). Not recorded to date from the Mutabilis, Autissiodorensis and Elegans Zones.

REMARKS: This new species of Tenwa differs from previously described species of the genus in its characteristic shape and in the nature of its spines. It is an in- frequent species: 15 specimens from English assemblages and only a single specimen from France were recorded.

Tenua sp. Plate 1, figures 7, 10 1970 Chytroeisphaeridia pococki Gitmez: pl. 9, fig. 8, table 4. DESCRIPTION: Shell spheroidal to broadly ovoidal, densely granular. The shell

FROM ENGLAND, SCOTLAND AND FRANCE IQI

wall may or may not be composed of two layers: if the wall is indeed bipartite, then both layers are thin. The outer wall (periphragm?) bears numerous spines whose character is seen clearly only at the margins, especially around the antapex. The spines are slender and short: the relative length of particular spines is, however, highly variable, median spines being consistently shorter than those of the antapical region. At their tips, the spines are most often knobbed, capitate or briefly bifurcate. The cingulum is faintly indicated by two parallel lines in the equatorial region. An apical archaeopyle is developed, with a scalloped margin suggesting partial reflection of a tabulation.

FIGURED SPECIMEN: BM(NH) slide V.53619(1). Sample SS 627, from too ft above the second dolerite sill, Staffin Bay, Skye. Lower Kimmeridgian (Baylei Zone).

Dimensions: Figured specimen: length (apex lacking): 50u, breadth 55u, spine length 2-3y.

REMARKS: This specimen was mentioned earlier as Chytroeisphaeridia pococki by one of the authors (Gitmez, 1970), but later examination by high power phase contrast objective showed the presence of spines and other details which made it clear that this specimen is different from C. pocockit. The observation was based on a single, fairly well preserved specimen. It is generally similar to Tenua verrucosa Sarjeant and Tenua villersense Sarjeant; the shape of the spines compares closely with those of T. villersense, but they are very short, as in T. verrucosa. This may be a representative of an undescribed species intermediate between 7. verrucosa and T. villersense.

Cyst-Family GONYAULACYSTACEAE Sarjeant & Downie, 1966 Genus CRYPTARCHAEODINIUM Deflandre, 1939b emend. Gitmez, 1970

Cryptarchaeodinium calcaratum Deflandre, 1939b emend. Gitmez, 1970

1939b Cryptarchaeodinium calcavatum Deflandre, 145, pl. 6, fig. 6. 1941a C. calcavatum Deflandre; Deflandre, 19, pl. 5. figs. 7-9; text-figs 9-10.

1962 C. calcavatum Deflandre; G. & M. Deflandre, fiche 1908. 1964 C. calcavatum Deflandre; Downie & Sarjeant, 104.

1964 C. calcavatum Deflandre; Eisenack, 153-4.

1964 C. calcavatum Deflandre; Sarjeant, table 2.

1965 C. calcavatum Deflandre; Gorka, 303, pl. 2, figs 3-5, table 1. 1967b C. calcavatum Deflandre; Sarjeant, table IV. 1970 C. calcavatum Deflandre; Gitmez, 246-8 pl. 1, figs 1-2, text-fig. 4, table 4.

OBSERVED RANGE: Kimmeridgian (Baylei and Rotunda Zones).

TOTAL KNOWN RANGE: ?Oxfordian to Kimmeridgian (Rotunda).

Remarks: Deflandre first observed this species in the Kimmeridgian assemblages from Orbagnoux (1939); a fuller diagnosis was given later (1941). In 1965, Gorka recorded this species for the first time from Poland, in sediments considered to be of

Oxfordian age. She observed the archaeopyle formation (by loss of plate 3’’) and determined a tabulation similar to that of the Kimmeridgian specimens, except that

192 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

she mentioned the presence of a second antapical plate: this plate (2’’’’) was never observed in the Kimmeridgian specimens. According to the measurements quoted by Gorka, the Polish specimens from the Oxfordian are larger than the English and French Kimmeridgian specimens, with longer sutural spines. The attribution of these Polish specimens to the species C. calcavatum must, all in all, be considered very doubtful.

Cryptarchaeodinium cf. calcaratum Plate 2, figure 3; text-figure 3

DESCRIPTION: Shell spherical, with a small, blunt apical horn and rounded, dome- shaped antapex. Tabulation: 4’, 6’, 6c, 7’, 2p, ?Ipv and 1”. Apical plates small; plates 1’ and 3’ form the apical horn. The precingular plates are more or less constant in shape and size, the boundary between plates 5” and 6’ not being clear. Cingulum equatorial, a laevorotatory spiral; cingular plates 3c and 4c are the largest and occupy the dorsal side. The postcingular plates are of variable shape and size: plates 1’”, 2’”, and 7’” are relatively small and triangular, plate 4’ (the largest) is more or less square. Two posterior intercalary plates, rp and 2p, are present and placed on either side of plate rpv: the boundary between Ip and Ipv was not clear. Plate 2p is very small. The single antapical plate (1’’’’) is convex and its greatest portion is positioned on the dorsal side of the cyst: this does not seem to be the result of distortion in compression. The sulcus is short, being very wide on the epitract but narrowing on the hypotract. On the plate boundaries, crests of irregularly spinous character rise up: some crest spines are briefly bifurcate. The shell surface is minutely granular. An archaeopyle was not observed.

Fic. 3. Cvryptarchaeodinium cf. calcavatum Deflandre. General appearance, showing the tabulation: left, in ventral view; right, in dorsal view. BM(NH) slide V.56342 (1). X Cc. 1400,

FROM ENGLAND, SCOTLAND AND FRANCE 193

FIGURED SPECIMEN: BM(NH) slide V.56342(1). Sample CS 421, from the first Meleagrinella Band of Eathie, Cromarty. Lower Kimmeridgian (Cymodoce Zone).

DIMENSIONS: Overall length 40-50y, breadth 35-5ou, horn length 5—6y, length of the spines on the sutures 4-6. Two specimens were measured.

REMARKS: Two specimens observed, one from the Lower Kimmeridgian (Cymodoce Zone) of Scotland and one from the Upper Kimmeridgian (Pectinatus Zone) of Dorset, are similar to C. calcaratum except in their possession of an apical horn and in slight differences in the shape of the plates on the hypotract.

Genus GONYAULACYSTA Deflandre, 1964 emend. Sarjeant, 1969 Gonyaulacysta cauda sp. nov. Plate 2, figures I-2, 4-5 1969 Gonyaulacysta sp. B Gitmez, pl. 6, fig. 3, text-fig. 14, table 4.

DERIVATION OF THE NAME: Latin, cauda, tail, appendage; in reference to the antapical spines.

Diacnosis: The broadly ovoidal cyst possesses a poorly developed apical horn with long spines arising from its tip. Tabulation: 4’, 1a, 6”, 6c, 6’, Ip, Ipv and 1’’”. Spiny crests separate the plates. The single antapical plate is characteristically surrounded by long (nearly three times longer than the other sutural spines), thin, simple spines. Cingulum helicoid, laevorotatory ; sulcus moderately broad, extending on both epitract and hypotract to the same length. Surface densely granular. Precingular archaeopyle, if present, formed by loss of plate 3”’.

Horotyre: BM(NH) slide V.53965(2) from the sample CC 447, Argiles de Moulin Wibert of Cap de la Créche, Boulonnais, France. Lower Kimmeridgian (Baylei Zone).

PARATYPE: BM(NH) slide V.56343(1). Sample HC 243, from c. 100 ft below Rotunda Nodules, base of Hounstout Cliff, Dorset. Upper Kimmeridgian (Pectinatus Zone).

DIMENSIONS: Holotype: overall length 78u, breadth 50y, apical horn length 8y; length of the sutural processes 3-5u, antapical processes 8; breadth of the cingulum 3-Su-

Paratype: length 80u, breadth 65u, apical horn length 12u, antapical processes length row.

A third specimen could not be measured because of its poor preservation.

DEscriPTION: The slightly helicoid, laevorotatory cingulum of moderate breadth, divides the cyst into two unequal parts. The conical epitract terminates in a poorly developed apical horn, the hypotract is dome-shaped. The epitract is larger than the hypotract, almost two-thirds of the shell length. Apical plate I’ is elongate, its anterior and posterior ends being narrow; together with plate 3’, it forms the apical horn. Plate 2’ is quite large; 4’ is the smallest of the apical plates. The single anterior intercalary plate 1a is large; as a result plate

194 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

6” is reduced. The precingular plate 1” is long and narrow; plates 2”, 3’, 4’ and 5’ are large. One of the specimens observed has a precingular archaeopyle, formed by loss of plate 3’. Six postcingular plates occupy the hypotract, together with the single antapical plate and the posterior plates. Plate 1’’’ is quadrate and as small as the adjacent sulcal plate; all the other postcingular plates are relatively large. A crescent-shaped plate, rpv, separates the sulcus from antapical plate 1’’’’.. The boundary surrounding the antapical plate 1’’’’ bears longer spines than the other sutural spines. All the sutural spines are simple, solid, thin, threadlike.

REMARKS: This new species is rare, only three specimens being recorded from the following samples: CC 447, RB 219, HC 243 (The first two are from the Baylei Zone, the last from the Pectinatus Zone). It has not been observed in the Middle Kimmeridgian. Two specimens were well preserved, with tabulation and mode of archaeopyle formation easily determinable, but the third was badly preserved. With their long spines distributed like tassels around the antapex, these specimens are different from all previously described proximate dinoflagellate cysts.

Gonyaulacysta cf. giuseppei (Morgenroth, 1966) Sarjeant, 1969 Plate 3, figures 3-4, text-figure 4

DESCRIPTION: Cyst subspherical to globular, with the tabulation 4’, 6’’, 6c, 6’”’, ip and 1’’’’.. The cingulum is strongly helicoid, laevorotatory, dividing the cyst into two equal parts: the epitract ends in a short apical horn, the hypotract has a conical aspect with very convex antapex. The apical plates combine to form the

ee Bt

Fic. 4. Gonyaulacysta cf. giuseppei (Morgenroth). Showing tabulation and archaeopyle: left, in ventral view; right, in dorsal view. BM(NH) slide V.56344. ™ c. Ioro.

FROM ENGLAND, SCOTLAND AND FRANCE 195

apical horn. The precingular plates, except plate 6’’, are quite large. Plate 3” is subtriangular in shape and lost in archaeopyle formation, together with some parts of the cingulum (plate 3, figure 4). The postcingular plates are of variable size and shape: plates 1’ and 2’”’ are both reduced to accommodate the long posterior intercalary plate 1p. Plates 3’ and 4’” are the largest of all the plates. A single convex plate occupies the antapex. The sulcus is broad, extending between the apex and the antapex. The surface of the shell is granular. Crests on the plate boundaries are low and membraneous.

FIGURED SPECIMEN: BM(NH) slide V.56344(1). Sample CC 448 from Calcaires de Moulin Wibert, south side of Cap de la Créche, Boulonnais. Lower Kimmeridgian (Baylei Zone).

DIMENSIONS: Figured specimen: overall length 78u, breadth 62u, length of apical horn 64. Range of the observed specimens: overall length 65—78y, breadth 58-62u, horn length 5-6u. (Measured specimens 4 in number.) Morgenroth gave the following dimensions for G. giusepper (Eocene): length 67-87y, breadth 67—78u, horn length 6-8u. The specimens from the Kimmeridgian are thus slightly smaller than the true G. giusepper.

REMARKS: Four specimens from the Lower Kimmeridgian assemblages of the Baylei and Mutabilis Zones (one from France, three from England) are closely similar to G. giusepper, recorded from the Lower Eocene of Germany by Morgenroth (1966). The only major difference is in the shape of the precingular archaeopyle; G. giwsepper has a very large, markedly polygonal archaeopyle, but in the Kimmeridgian specimens observed, the archaeopyle is somewhat smaller and tapers so markedly towards the apex that it appears almost triangular. Since the big stratigraphical gap makes it improbable that these specimens are conspecific with G. giuseppet, they are compared with, rather than attributed to that species.

Gonyaulacysta globata sp. nov. Plate 3, figures 1-2; text-figures 5 A-B

DERIVATION OF THE NAME: Latin, globus, ball, sphere, in reference to the overall shape of the cyst.

Diacnosis: A proximate cyst, subspherical to broadly ovoidal with a strong apical horn. The sutural crests are low, generally well defined, and reflect the tabulation 4’, 1a, 6”, 6c, 6’, ?Ipand 1’’”’”. Cingulum moderately narrow, helicoid, laevorotatory, dividing the theca unequally, the epitract being longer than the hypotract. The sulcus is broad. The surface of the shell is densely granular. A precingular archaeopyle, formed by loss of plate 3’’, is developed in some instances.

HototyPe: I.G.S. slide PK.122, Sample WB 20, H.M. Geological Survey Borehole, Warlingham, Surrey at 2510 ft depth. Middle Kimmeridgian (Wheatleyensis Zone).

PARATYPE: BM(NH) slide V.56345. Sample FN 236, from the White Stone

Band, + m west of Freshwater Steps, Dorset. Middle Kimmeridgian (boundary of Pectinatus-Hudlestoni Zones).

196 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

Fic. 5. Gonyaulacysta globtaa sp. nov. A. Holotype: left, in ventral view; right, in dorsal view. I.G.S. slide PK 122. x c.800. B. Paratype: showing the epitractal tabulation: left, in ventral view; right, in dorsal view. BM(NH) slide V.56345. x c.655.

FROM ENGLAND, SCOTLAND AND FRANCE 197

Dimensions: Holotype: overall length gou, breadth 68y, apical horn length Iry. Paratype: overall length g2u, breadth 75u, horn length 12u. Range of the observed specimens: overall length 85-92u, breadth 62~-75u, horn length 11-12u. (Four specimens were measured.)

Description: The cyst is globular, relatively thin-walled. Four apical plates combine to form the slender horn. Plate 1’ is narrow and elongate, the other apical plates are approximately polygonal but with an apical prolongation. A single anterior intercalary plate is present and quite large, apical plate 4’ and precingular plates 5” and 6” being correspondingly reduced. The four other precingular plates are large. The cingular plates are poorly defined, but appear to number six.

The hypotract is dome-shaped, composed of large reflected plates; plate 4’’ is the largest of all. Plate 1’” is greatly reduced; plates 5’” and 6’” are relatively small. The boundary between the plates Ip and 2’” was not confirmed. The single antapical plate, 1’’’”’, is also large.

The sulcus is narrow in its anterior portion, broadening to contact with the cingulum and thenceforward remaining of constant breadth in its posterior portion. It is relatively short and extends to the antapex.

In one specimen only, a precingular archaeopyle was seen, formed by loss of plate gee OBSERVED RANGE: Kimmeridgian (Wheatleyensis-Pectinatus Zones).

REMARKS: This species is extremely infrequent: of four specimens encountered, the holotype and paratype only are moderately well preserved, the other two being folded, crushed and severely damaged.

In its combination of overall morphology and tabulation G. globata differs from previously described species of Gonyaulacysta. The most closely similar species is G. nuciformis, but G. globata differs in having a relatively thin cyst wall and dis- similar ventral antapical tabulation.

Gonyaulacysta longicornis (Downie, 1957) Sarjeant, 1969, emend. Plate 2, figure 6; Plate 4, figure 1; text-figure 6

1957 Gonyaulax longicovnis Downie, 420, pl. 20, fig. 8; text-figs 2a—b; table 1. 1962 G. longicornis Downie; G. & M. Deflandre, fiche 1830.

1964 G. longicoynis Downie; Downie & Sarjeant, 115.

1964 G. longicorynis Downie; Sarjeant, table 2.

1964 G. longicoynis Downie; Eisenack, 371-2.

1966 ©Gonyaulacysta longicornis (Downie); Sarjeant, nomen nudum, 131.

1967b G. longicornis (Downie); Sarjeant, nomen nudum, table 1.

1967b Gonyaulax longicorvnis Downie; Vozzhennikova, table 12.

1969 Gonyaulacysta longicornis (Downie); Sarjeant, ro.

1970 G. longicornis (Downie); Gitmez, table 4.

EMENDED DIAGNOSIS: This species of Gonyaulacysta is characterized by a very long apical horn (not less than one-third of the whole length). Tabulation: 4’, 6”, 6c, 6’”, Ip and 1”. Cingulum slightly helicoid, dividing the cyst unequally: the epitract being longer than the hypotract. On the sutures, short, roughly denticulate

198 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

crests rise up. Precingular archaeopyle sometimes present and formed by loss of plate 3’. Surface of the shell coarsely granular.

Ho.otyPe: C. Downie’s collection, Micropalaeontological Laboratory, University of Sheffield, KL 11.15; from the Upper Kimmeridge Clay of Norfolk, England.

FIGURED SPECIMENS: I.G.S. slide PK.120, Sample WB 18, Kimmeridge Clay, H.M. Geological Survey Borehole, Warlingham, Surrey, at 2560 ft depth. Middle Kimmeridgian (Wheatleyensis Zone).

BM(NH) slide V.56346(2). Sample LO 353, from the Littleworth Quarry, Oxfordshire. Upper Kimmeridgian (Pallasioides Zone).

Dimensions: Holotype, as quoted by Downie: overall length 92u, breadth 58u, horn length 36y.

Figured specimen from Warlingham Borehole: overall length 115, breadth 75y, horn length 46u. Figured specimen from Littleworth: overall length 125, breadth 85u, horn length 4ou.

Range of the specimens from England and France: overall length 80-155y, breadth 55-100p, horn length 25-60u. Measured specimens 34 in number.

Fic. 6. Gonyaulacysta longicorvnis (Downie). Specimen: left, in ventral view; right, in dorsal view. BM(NH) slide V.56346 (2). x c.704.

FROM ENGLAND, SCOTLAND AND FRANCE 199

DescriPTion: The thin-walled shell is roughly polygonal in shape, with a long apical horn and conical hypotract. The apical horn, with its solid tip, is formed by four long apical plates; plate 1’ is elongate, the others are almost triangular. The apical plates, together with six precingular plates, make up the epitract; this is separated from the hypotract by a narrow cingulum and is always longer than the hypotract. Six cingular plates of variable size occupy the cingulum. The sulcus extends onto the epitract and hypotract, between the apex and the antapex. Six postcingular plates of variable shape and size are present: plate 1’ is reduced to accommodate the posterior intercalary plate, Ip; plates 2’, 3’ and 5’’’ are more or less uniform in size and plate 4’” is the largest of all the plates.

OBSERVED RANGE: Kimmeridgian (Baylei to Pallasioides Zones) [See discussion below].

TOTAL KNOWN RANGE: Kimmeridgian (Baylei to Pallasioides Zones).

REMARKS: G. longicornis has been known hitherto only from the Upper Kim- meridgian of England. Though the specimens in the French and English assem- blages were generally badly preserved, it was possible to determine the tabulation and the mode of archaeopyle formation. This species was doubtfully included in the genus Gonyaulacysta by Sarjeant (1969), in the absence of knowledge of the type of archaeopyle; the new observations confirm this reallocation.

The observed specimens are closely similar to Downie’s figured specimen, but show slight differences in tabulation. The apical horn is not developed from plate 1’ only, as figured by Downie; instead, it is made up of four apical plates. The posterior intercalary plate was not shown on the figure of the holotype, but was observed in all specimens encountered.

G. longicornis is similar to Pareodinia nuda (Downie) in the shape of the apical horn and general appearance; but no tabulation has been yet determined for P. nuda and an intercalary archaeopyle was considered by Sarjeant (1967a pp. 254) to be developed in the latter species.

In England, G. Jongicornis was found in most horizons of the Kimmeridgian from Aulacostephanus to Pallasioides, but it was absent from the lowest zones and the Rotunda Zone. In France, in contrast, it was recorded only from the Baylei Zone. Thirteen specimens from France and forty-six specimens from England were recorded.

Gonyaulacysta cf. mamillifera (Deflandre, 1939b) Sarjeant, 1969 Plate 4, figure 7; text-figure 7

DESCRIPTION: Relatively large, globular shell, broadly ovoidal to subspherical in shape. The more or less equatorial cingulum is slightly helicoid and divides the cyst into two equal parts; the epitract terminates in a mammelon form (in two of the observed specimens, this was well-developed, but the third one has a feebly- developed apical prominence); the hypotract is rounded. Tabulation: 4’, 6’, 6c, 6’, Ip, 1’’’.. Plate boundaries are marked by low membraneous crests. The sulcus is long, extending further on the epitract than on the hypotract; it narrows

200 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

towards the apex. The shell wall is thin, its surface densely granular and punctate. A precingular archaeopyle is present, formed by the loss of plate 3”.

FIGURED SPECIMEN: I.G.S. slide PK.130, Sample WB 29, from H.M. Geological Survey Borehole, Warlingham, Surrey, at 2285 ft 7in. depth. Upper Kimmeridgian (Rotunda Zone).

DIMENSIONS: Figured specimen: length g2u, breath 80u. Range: length 76-92, breadth 60-80, measured specimens 3 in number. (Holotype dimensions: length g2u, breadth 84u, as given by Deflandre).

REMARKS: G. mamullifera has only been previously recorded from the Kimmeridgian of France. In this investigation, three specimens probably attributable to this species were observed: one from the Baylei Zone of France and two from the Rotunda Zone of England. In their general aspect they are closely similar to G. mamuillifera ; however, there are differences in the reconstructed tabulation and the ornamentation ofthesutures. In the observed specimens the sutures are in the form of membraneous crests, not spinose, as described by Deflandre. The tabulation is generally similar, but the shapes of postcingular plates 1’’’ and 2’” are different. Comparison between the apical plates of this form and G. mamiutllifera was not possible, because Deflandre was unable to determine the apical tabulation. Allocation to this species must, therefore, be provisional only.

Gonyaulacysta nuciformis (Deflandre) Sarjeant, 1969 Plate 3, figure 5; text-figure 8

1938 Palaeoperidinium nuciforme Deflandre, 180, pl. 8, figs 4-6.

1962a P. nuciforme Deflandre; Sarjeant, pl. 1, fig. 8; tables 3-4.

1962b Gonyaulax nuciformis (Deflandre); Sarjeant, 482-3, pl. 69, fig. 6; text-fig. 4; tables 2-3. 1964 G. nuciforymis (Deflandre); Downie & Sarjeant, 115.

1964 G. nuciformis (Deflandre); Sarjeant, table 2.

1964 Palaeoperidinium nucifornis Deflandre; Eisenack, 609.

1965 Palaeoperidinium nuciformoides Gorka, 300-1, pl. 2, figs 1-2; table 1.

1966 P. nuciforymoides (Deflandre) G. & M. Deflandre, fiche 3030.

1966 ?Gonyaulacysta nuciformis (Deflandre); Sarjeant, nomen nudum 132.

1967b Gonyaulax nuciformis (Deflandre) ; Vozzhennikova, table 11.

1967b Gonyaulacysta nuciformis (Deflandre); Sarjeant, nomen nudwm, table tr.

1968b G. nuciformis (Deflandre); Sarjeant, nomen nudum, 227, pl. 3, fig. 4; table 2A. 1969 G. nuciformis (Deflandre); Beju, nomen nudum, to, pl. 3, fig. 1; table 1.

1969 G. nuciformis (Deflandre) ; Sarjeant, Io.

1970 G. nuciformis (Deflandre) ; Gitmez, 3, pl. 6, fig. 1; table 4.

DESCRIPTION: The shell is ovoidal to spherical, with the tabulation 4’, Ia, 6”, 6c, 6’’, Ip, Ipv and 1’’’. The epitract and hypotract are more or less equal in size; the epitract ends with an apical horn of variable length, the hypotract is dome- shaped with rounded antapex. The cingulum is helicoid, laevorotatory. The sulcus is broad and extends onto both the epitract and hypotract. The shell is densely granular and relatively thick; because of this, determination of the tabula- tion is difficult. A precingular archaeopyle was developed by some specimens, formed by loss of plate 3”.

FROM ENGLAND, SCOTLAND AND FRANCE 201

Fic. 7. Gonyaulacysta cf. mamillifera (Deflandre). Showing the tabulation: left, in ventral view; right, in dorsal view. I.G.S. slide PK 130. X c.772.

Fic. 8. Gonyaulacysta nuciformis (Deflandre). Tabulation: left, in ventral view; right, in dorsal view. Specimen I.G.S. slide PK tog. x c.1112.

202 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

FIGURED SPECIMEN: I.G.S. slide PK.109, Sample WB 7, from H.M. Geological Survey Borehole, Warlingham, Surrey, at 2834 ft 7in. depth. Lower Kimmeridgian (Eudoxus Zone).

DIMENSIONS: Figured specimen: overall length 7ou, breadth 60u, horn length Ou.

Range of the Lower Kimmeridgian specimens (61 specimens were measured): overall length 43-118p, breadth 40-85y, horn length 4-12y.

Range of the Middle Kimmeridgian specimens (28 specimens were measured): overall length 68—-105y, breadth 60-85y, horn length 5-12y.

Range of the Upper Kimmeridgian specimens (18 specimens were measured): overall length 65—102p, breadth 55-85, horn length 6—-12u.

The following dimensions were quoted by Deflandre for the Oxfordian specimens from France: overall length 60-65u,, breadth 47-53 (approximately). Dimensions of the Callovian specimens from England, as given by Sarjeant: overall length 56-58y, breadth 50-64u. Gorka gave the following dimensions for Polish Upper Jurassic specimens: overall length 38-54y, breadth 40-44u. Dimensions of the Roumanian specimens (Oxfordian to Kimmeridgian) are given by Beju as overall length 62—78y, breadth 60-78u. The Polish specimens are thus smaller than the others and the Roumanian specimens are more spherical. The Kimmeridgian specimens are larger than the specimens from lower stages.

OBSERVED RANGE: Kimmeridgian (Baylei to Pallasioides).

TOTAL KNOWN RANGE: Upper Callovian (Lamberti) to Upper Kimmeridgian (Pallasioides).

REMARKS: G. nuciformus was first recorded from the Upper Jurassic of France by Deflandre (1938). Subsequently, this Upper Jurassic species has been observed in several assemblages from Western Europe, and the geographic range has been extended by its observation from the Callovian to Kimmeridgian of Roumania by Beju (1969). It is generally present in moderate abundance in Kimmeridgian assemblages; however, it was not observed in the Middle and Upper Kimmeridgian of France. Also there is a progressive reduction in the number of specimens in the English assemblages through the Upper Kimmeridgian.

The combination Gonyaulacysta nuciformis in Beju (1969), although proposed in correct form, was not validly published since a pre-print distributed at a meeting does not constitute effective publication (cf. ‘International Code of Botanical Nomenclature’, Art. 29). The combination is, therefore, correctly attributed to Sarjeant (1969).

Gonyaulacysta perforans (Cookson & Eisenack) Sarjeant, 1969 Plate 4, figure 6; text-figure 9

1958 Gonyaulax perforans Cookson & Eisenack, 30, pl. 2, figs 1-4, 7-8; text-figs 8-9. 1961 G. perfovans Cookson & Eisenack; Alberti, 6, pl. 11, figs 4-6; tables a—c.

1962 G. perfovans Cookson & Eisenack; G. & M. Deflandre, fiches 1849-1852.

1963 G. perforans Cookson & Eisenack; Baltes, 584, pl. 4, figs 1-6, table 1.

1964 G. perfovans Cookson & Eisenack; Downie & Sarjeant, 115.

FROM ENGLAND, SCOTLAND AND FRANCE 203

1964 G. perfovans Cookson & Eisenack; Sarjeant, table 2.

1964 G. perforans Cookson & Eisenack; Eisenack, 397-8.

1965 G. perforans Cookson & Eisenack; Baltes, 12, pl. 3, figs 93-4.

1966 Gonyaulacysta perforans (Cookson & Eisenack); Sarjeant, nomen nudum, 131. 1967b Gonyaulax perforans Cookson & Eisenack; Vozzhennikova, table 12.

1967 G. perforans Cookson & Eisenack; Millioud, pl. 2, fig. 15; text-fig. I.

1969 G. perforans Cookson & Eisenack; Baltes, fig. 3.

1969 Gonyaulacysta perforans (Cookson & Eisenack) ; Sarjeant, Io.

DescRIPTION: The cyst is elongate, with a long apical horn. The epitract and hypotract are separated by the helicoid, laevorotatory cingulum and are more or less equal in size. Rabulation: 44, 64,0, ip,and«1 "4 Plate boundaries are demarcated by membraneous, delicate porate crests, which are well developed around the apex and the antapex. The sulcus is long and narrow, extending from apex to antapex. The shell wall is thin, the surface granular and occasionally perforate. An archaeopyle was rarely observed; when developed, it forms by the loss of plate 3”.

Fic. 9. Gonyaulacysta perforans (Cookson & Eisenack). Tabulation: left, in ventral view; right, in dorsal view. I.G.S. slide PK 131. X c.1491.

204 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

FIGURED SPECIMEN: I.G.S. slide PK.131, Sample WB 29, from H.M. Geological Survey Borehole, Warlingham, Surrey, at 2285 ft 7 in. depth. Upper Kim- meridgian (Rotunda Zone).

DIMENSIONS: Figured specimen: overall length 108, breadth 7ou, horn length 33u. Range of the Middle Kimmeridgian specimens (7 specimens measured): overall length 73-100n, breadth 54-73u, horn length 16-18y. 15 specimens measured from the Upper Kimmeridgian: overall length 80—110y, breadth 52-75, horn length 12-40u. Cookson and Eisenack gave the following dimensions for the specimens from New Guinea: length 136-168y, breadth 93-109gy. Dimensions of the specimens from Germany are quoted by Alberti as length 130-145y and breath 81-103u. The British and French Middle Kimmeridgian specimens are thus smaller than the Upper Kimmeridgian specimens, but both are smaller than the New Guinea and German specimens.

OBSERVED RANGE: Kimmeridgian (Wheatleyensis to Pallasioides). TOTAL KNOWN RANGE: Upper Jurassic to Lower Cretaceous (Albian).

REMARKS: G. pervforans was originally recorded from the Upper Jurassic of New Guinea by Cookson and Eisenack; later Alberti observed it in the Barremian assem- blages from Germany and Baltes recorded it from the Albian of Roumania. It is recorded in English assemblages for the first time. The observed specimens are similar to those figured by Cookson and Eisenack, except for small differences in tabulation and in the length of the apical horn.

Although the New Guinea and German specimens are larger than the Kimmeridgian specimens, from the figures they seem to have a proportionately smaller horn. (Since the horn length was not specified, it is possible to deduce this only from the figures.)

Cookson and Eisenack did not mention the presence of apical plate 4’, but in their figure a boundary is shown between the plates 3’ on the ventral side and 3’ on the dorsal side; therefore, the plate on the ventral side should be the fourth apical plate, as observed in the Kimmeridgian specimens. Similarly the elongate plate which they figure below the postcingular plate 1’’’ should be the posterior inter- calary plate, Ip.

Gonyaulacysta systremmatos sp. nov. Plate 5, figures 7-8

1970 Gonyaulacysta sp. C. Gitmez, 265-7, pl. 4, figures 10-11, text-fig. 15, table 4.

DERIVATION OF NAME: Greek, systremmatos, anything consolidated, generally a ball or round object; in reference to the ball-like shape of the shell.

DiaGnosis: Thick-walled shell, almost spherical, with a moderately long apical horn. Tabulation: 4’, 1a, 6”, 6-7c, 7’”, Ip, Ipv, 1’. The plate boundaries are demarcated by delicate crests of variable height. The cingulum is helicoid, laevoro-

tatory; the sulcus is short and broadens posteriorly. The surface is densely granular. A precingular archeopyle, formed by loss of plate 3’’, is generally developed.

FROM ENGLAND, SCOTLAND AND FRANCE 205

HoLotyPre: BM(NH) slide V.53966(z), from the sample CC 447, Argiles de Moulin Wibert, Cap de la Créche, Boulonnais, France. Lower Kimmeridgian (Baylei Zone).

Dimensions: Holotype: overall length 68u, breath 62, horn length 8y, height of the crests 4-5u. Range of the observed specimens: overall length 66-78, breadth 60-65, horn length 7-16; measured specimens 5 in number.

Description: The helicoid, laevorotatory cingulum divides the cyst unequally. The epitract is somewhat longer than the hypotract; both are more or less dome- shaped. The number of the cingular plates is hesitantly mentioned, because the character of the small plate beside 6c is doubtful; it may be either a small cingular plate or a short sulcal plate.

Four apical plates make up the apex, plate 1’ occupying the anterior prolongation of the sulcus. Plates 2’ and 3’ are small; plate 4’ is almost as large as plate 1’. The single, small anterior intercalary plate is placed between the plates 4’ and 6’. The precingular plates are generally large, plate 6’’ being of reduced size because of the presence of intercalary plate Ia.

The postcingular plates are of variable size and shape: plate 1’ is very small; plate 2’” is also reduced and does not have a boundary with the antapical plate. Plates 3’”, 4”, 5’” and 6’” are relatively large; plate 7’” is in contrast reduced, having nearly the same size as plate 2’”. The quite broad intercalary plate, Ipv, separates the sulcus from the single antapical plate 1’’””’.

4st

REMARKS: This is an infrequent species, six specimens being encountered, all from the Baylei Zone of France. They were badly preserved, being somewhat crushed, folded or covered by debris; the holotype was the best oriented for study. In its general form, this new species of Gonyaulacysta differs from all described species. The most closely comparable species is G. palla Sarjeant, which has a similarly spherical shape and comparable tabulation; but G. systemmatos differs in its apical horn, sutural crests, absence of plate 1a, and presence of plate Ipv. In the possession of a seventh postcingular plate, it is comparable with G. fetchamensis and G. ehrenbergi1, but it is markedly different in overall morphology from both these species.

Gonyaulacysta sp. A Plate 9, figures 1-2 1970 Gonyaulacysta sp. A. Gitmez, 263-4, pl. 3, fig. 3, text-fig. 13.

FIGURED SPECIMEN: BM(NH) slide V.56347(2), Kimmeridge Clay (Pectinatus Zone) 60 ft above Freshwater Steps Stone Band, Egmont Bight, Dorset.

DIMENSIONS: Figured specimen: overall length 103u, length of apical horn 32u, overall breadth 66-5u, length of crest spines c. 1:5—2u.

RemaARKS: This form was originally described on the basis of two specimens from the lowest Kimmeridge Clay (Baylei Zone) of Normandy. The discovery of a third specimen at a higher horizon is thus of interest: its dimensions are markedly larger

206 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

than those of the specimens described earlier (overall length 65u, breadth 42y) but its proportions are similar. Yet further specimens of this type must be located before nomenclatural proposals can justly be made.

Gonyaulacysta sp. B Plate 4, figures 2-3, text-figure 10

DESCRIPTION: Only one specimen of this species has so far been observed. It possesses a subspherical cyst, bearing a moderately well developed apical horn. The strongly helicoid, laevorotatory cingulum divides the cyst into two more or less equal parts: the hypotract is somewhat flattened at the antapex. The sulcus is sigmoidal and narrow. Tabulation: 4’, 6’, 6c, 5’, Ip, ?Ipv and 1’’”.

The apical plates 1’ and 4’ are small and elongate; together with plates 2’ and 3’, they form the apical horn. The precingular plates are quite large. Plate 3” is lost in archaeopyle formation. On the hypotract, crests demarcate five postciagular plates. Plate 1’ and 5’” are reduced because of the presence of posterior plates, but the other postcingular plates are large. A single antapical plate occupies the antapex.

The wall is moderately thin, its surface finely granular and in part tuberculate. Crests on the plate boundaries are delicate; the denticulation is very deep, virtually giving the crests the appearance of a row of bifid spines.

FIGURED SPECIMEN: I.G.S. slide PK.117, Slide WB15, from H.M. Geological Survey Borehole, Warlingham, Surrey, at 2635 ft 3ia. depth. Middle Kimmeridgian (Elegans Zone).

Fic. 10. Gonyaulacysta sp. B. Tabulation: left, in ventral view; right in dorsal view. Specimen I.G.S. slide PK 117. X ¢.1351.

FROM ENGLAND, SCOTLAND AND FRANCE 207

DIMENSIONS: Overall length 48y, breadth 45u, horn length 5u, sutures 3-5. high.

RemARKS: In its general morphological features and sutural characteristics, this species may be distinguished from all previously described species of Gonyaulacysta. The shape of the shell and crests is most comparable to G. serrata Cookson & Eisenack (1958) suggesting a relationship between the two species, but the form of the apical horn is different ; since a tabulation could not be determined for G. serrata, a detailed comparison of these two species is impossible.

Gonyaulacysta sp. C Plate 6, figures 1-2, text-figure I1

Description: The cyst is subspherical, bearing a short, tapering apical horn. The thickness of the wall is uneven because of irregularly distributed granules. The sutural crests are in the form of low ridges giving rise occasionally to delicate mem- branes: they indicate a reflected tabulation of 4’, Ia, 6’’, 6c, 6’’’, ?>Ipv, 1'’’’.. Four apical plates together form the horn; plate 1’ is elongate in shape. The precingular plates, except plate 6’’, are of almost uniform shape and size; plate 6’’ is reduced to accommodate the intercalary plate fa.

Six cingular plates make up the slightly helicoid, laevorotatory cingulum: plate 6c is very small, the others are relatively large. The ends of the cingulum are widely separated by a very broad sulcus, which further widens in its posterior portion.

Fic. 11. Gonyaulacysta sp. C. Tabulation: left, in ventral view; right, in dorsal view. Specimen I.G.S. side PK 118. c.984.

208 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

(The antapical end of the sulcus and the plate boundaries near to the antapex were not very clear because of bad preservation of this part of the cyst.) The first post- cingular plate, 1’, is narrow, elongate and small. Plate 4’’’ appears the largest of the post-cingular plates. No archaeopyle was observed in this specimen.

FIGURED SPECIMEN: I.G.S. slide PK.118, Sample WB 16, from H.M. Geological Survey Borehole, Warlingham, Surrey, at 2510 ft depth. Middle Kimmeridgian (Scitulus Zone).

DIMENSIONS: Overall length 77u, breadth 60, apical horn length 8y.

REMARKS: Only one specimen has so far been observed. It is distinguished from all previously described species in its overall morphological features; it is possibly a new species, but further specimens are needed to decide this.

Gonyaulacysta sp. D Plate 6, figures 4-5; text-figure 12

DeEscripTION: A specimen with an almost spherical cyst and short, blunt, tapering apical horn. The shell wall is composed of two layers, periphragm and endophragm, which are of the same thickness; the periphragm is coarsely granular. The apical horn is formed of both shell layers. Tabulation: 4’, 1a, 6”, 6c, 6’, Ip and 1’’’”. Plate 1’ is elongate and occupies the anterior prolongation of the sulcus. Plates 2’ and 3’ are large; plate 4’ is slightly reduced to accommodate the anterior intercalary plate, Ia. Four apical plates together make up the apical horn. Plate Ia is quite large; because of this, plate 6’ is reduced. The other precingular plates are of more or less similar shape and size. The cingulum is of moderate breath,

Fic. 12. Gonyaulacystasp.D. Tabulation and archaeopyle formation, Plate 3” is partially lost in the archaeopyle formation. Specimen BM(NH) slide V.56346 (4). c.1035.

FROM ENGLAND, SCOTLAND AND FRANCE 209

possessing six cingular plates: plate 1c is smaller and broader than the other cingular plates. Postcingular plate 1’” is very small and gives the appearance of hanging onto one corner of the cingulum. Plate 2’” and 6’” are relatively small; plates 3’”, 4’” and 5’” are large. An elongate posterior intercalary plate (Ip) is placed below plate r’’’. The single antapical plate, 1’’’’, is large and convex. The broad sulcus extends between the apex and antapex.

Plate boundaries are demarcated by low crests and generally well defined. A precingular archaeopyle was seen in some specimens; it forms by loss of plate 3” (in the figured specimen, plate 3” is partially detached).

FIGURED SPECIMEN: BM(NH) slide V.56346(4). Sample LO 353, from the Littleworth Quarry, Oxfordshire. Upper Kimmeridgian (Pallasioides Zone).

DIMENSIONS: Figured specimen: overall length 65u, breadth 60, horn length 5y. Range: overall length 65—80u, breadth 60-75y, horn length 4-7. Measured speci- mens 5 in number; 3 other observed specimens could not be measured because of bad preservation.

REMARKS: This unnamed species of Gonyaulacysta differs from all described species in its general aspect and peculiar apical horn. The most comparable species is G. palla Sarjeant, both species having similarly spherical cysts and tapering apical horns. In Gonyaulacysta sp. C, the apical horn is broad based and short and its tip appears conical: although G. palla has a tapering horn, it is slender and relatively long. The number of the plates on the epitract is the same for both species, but plate 4’ is not placed at the top of the horn as in G. palla. The hypotractal plates (especially plates 1’’’, 2’”’ and 1p) appear similar in shape to those of G. palla, but their number is different: G. palla has seven postcingular plates, whereas this species has six. The crests of G. palla are spiny. The similarity between these two species is thus only in the overall shape.

Eight specimens, all from the same quarry in Littleworth, were recorded; un- fortunately, all the specimens encountered are somewhat crushed, folded or covered by debris, which makes them difficult to examine in detail. The figured specimen was the best oriented for study. It may be a new species, but needs further, better preserved specimens for typification.

Gonyaulacysta sp. E Plate 6, figure 9; text-figure 13

DESCRIPTION: Only one moderately well preserved specimen of this type was observed, in the Lower Kimmeridgian assemblages from the Warlingham Borehole. It possesses an elongate cyst, with conical epitract and dome-shaped hypotract, thus looking rather like a pear. The apical horn is slender and tapering. The wall is densely granular. Plate boundaries are well defined by moderately high membraneous crests. Tabulation: 4’, 1a, 6”, 6c, 6’’, Ip, Ipv and 1’. Plate 1’ is characteristi- cally long and broad, extending down almost two-thirds of the epitract. Plates 2’ and 3’ are similar to each other; plate 4’ is greatly reduced because of the anterior intercalary plate 1a. For the same reason, plates 5’ and 6” are also reduced.

210 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

The other precingular plates are quite large. The six postcingular plates are of variable shape and size. Plate 1’’’ is small and triangular; plates 2’ and 6’” are moderately large; plates 3’, 4’ and 5’” very large, together almost occupying the whole dorsal side of the hypotract. A single plate occupies the antapex. The posterior intercalary plate (Ip) is elongate. A crescent-shaped plate, Ipv, separates the sulcus from the antapical plate 1’’’. The cingulum is deep, formed by 6 cingular plates, of which 5c and 6c are quite small; it is almost circular and divides the cyst into two unequal parts, with the epitract twice as long as the hypotract. The sulcus is very short and broad. An archaeopyle was not observed.

FIGURED SPECIMEN: I.G.S. slide PK.1113. Sample WB 13, from H.M. Geological Survey Borehole, Warlingham, Surrey, at 2684 ft 3in. depth. Lower Kimmeridgian (Autissiodorensis Zone).

DIMENSIONS: Figured specimen: overall length 65u, breadth 5ou, apical horn length 7-5u.

REMARKS: The long epitract, densely granular shell wall, elongate apical plate 1’ and large postcingular plates 4’ and 5’” distinguish this form from all described species. It almost certainly represents an undescribed species of Gonyaulacysta, but before this can be decided, more specimens must be awaited.

Fic. 13. Gonyaulacysta sp. E. Tabulation: left, in ventral view; right in dorsal view. Specimen I1.G.S. slide PK 113. X c.669.

FROM ENGLAND, SCOTLAND AND FRANCE Ziit

Gonyaulacysta sp. F. Plate 6, figures 3, 6; text-figure 14

DescripTION: The cyst is elongate, with a long and tapering apical horn. The cyst wall is composed of two layers; a fairly thin endophragm and a thicker peri- phragm. The endophragm bulges into the lower half of the apical horn; the anterior portion of the horn is formed by the periphragm only, so that there is a cavity between the wall layers at the anterior end of the horn. Tabulation: 4’, 1a, 6”, 6c, 6’’, Ip, Ipv and 1’. The four apical plates together make up the apical horn. Plate 1’ is elongate, extending down two-thirds of the epitract. The anterior intercalary plate (1a) is quite large: because of this, precingular plate 6’’ is reduced. The other precingular plates are relatively large. The postcingular plates 1’’’ and 6'’’ are small, the others are moderately large. The posterior intercalary plate (1p) is elongate: the boundary between the sulcus and the posterior ventral plate (1pv) was not confirmed. A single narrow plate occupies the antapex.

The cingulum is strongly helicoid, laevorotatory, dividing the cyst unequally, the epitract being longer than the hypotract. The sulcus is broad and largely con- fined to the hypotract.

Fic. 14. Gonyaulacysta sp. F. Tabulation and archaeopyle formation (the operculum has partially slipped inside the cyst). Left, in ventral view; right, in dorsal view. Specimen BM(NH) slide V.56348 (1). c.983.

212 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

The plate boundaries are marked by crests in the form of very low, thick mem- branes. The surface of the shell is densely granular. A precingular archaeopyle forms by loss of plate 3’’; on the figured specimen, the operculum is partially slipped inside.

FIGURED SPECIMEN: BM(NH) slide V.56348(1). Sample ED 240, from 30 ft above Freshwater Steps Stone Band, Egmont Bight, Dorset. Upper Kimmeridgian (Pectinatus Zone).

DIMENSIONS: Figured specimen: overall length 92y, breadth 52u, apical horn length 26.

REMARKS: The description of this species was based on a single specimen observed in the assemblages examined. In its overall morphology it is typically a species of Gonyaulacysta, but it is markedly different from all previously described species. It may be compared with G. perforans, since both have a similarly elongate theca, with long apical horn, and a comparable tabulation, but the crests on the sutures are completely different: G. perforans has well developed, porate membraneous crests, whereas this form has low, thick membraneous crests. G. perforans apparently has a single-layered wall, but in this species the shell wall is distinctly two layered. This specimen certainly represents a new species of Gonyaulacysta but, since the only specimen observed is not well preserved, no new name is given.

Gonyaulacysta sp. G Plate 6, figures 7-8; text-figure 15

DEscRIPTION: The cyst is broadly ovoidal, with a strong apical horn. Tabulation: 4’, 6”, 6c, 6’”, Ip, ?Ipv and 1’’’. The apical plates 1’ and 4’ are broad and long, plates 2’ and 3’ are rounded, smaller. The precingular plates are moderately large, except for plates 1’ and 6” which are somewhat smaller than the others. Plate 3’ is typically lost in archaeopyle formation. The postcingular plates 1’”, 5’ and 6’” are small; in contrast, plates 3’” and 4’” are very large and occupy almost the whole dorsal side of the hypotract. The single antapical plate, 1’’’”’, is quite large and convex; plate Ip is small. On the posterior portion of the sulcus some small plates were suggested, but their presence could not be confirmed: they may constitute a subdivided posterior ventral plate Ipv.

The cingulum is narrow, helicoid, laevorotatory, comprised of six cingular plates. It divides the cyst unequally into two parts, the epitract being smaller than the hypotract. The sulcus is short, mainly confined to the hypotract and broadening posteriorly.

The plate boundaries are marked by high, delicate crests, which have smooth edges and are irregularly perforate. The surface of the shell is densely granular. The dense granulation and high crests render the tabulation difficult to determine.

FIGURED SPECIMEN: BM(NH) slide V.56349(1). Sample HC 243, from c. roo ft below the Rotunda Nodules, in the base of Hounstout Cliff, Dorset. Upper Kimmeridgian (Pectinatus Zone).

FROM ENGLAND, SCOTLAND AND FRANCE 213

Fic. 15. Gonyaulacystasp.G. Tabulation and archaeopyle formation (plate 3” is missing). Left, in ventral view; right, in dorsal view. Specimen BM(NH) slide V.56349 (1). X €.1024.

Dimensions: Figured specimen: overall length 75u, breadth 45u, apical horn length tou, breadth of the cingulum 5y.

REMARKS: This single specimen differs from the described species of Gonyaulacysta in its general shape and distinctive sutural crests. The tabulation and perforate crests are similar to those in G. perforans, but in the other morphological characters those species are dissimilar.

Gonyaulacysta sp. H Plate 13, figure 1, text-figure 16

DEscrIPTION: Cyst subpolygonal to ovoidal, with a strong apical horn of moderate length. The cyst wall is rather thick and composed of two layers, the periphragm alone forming the horn; the surface of the periphragm is densely and finely granular Tabulation 4-?5’, 1a, 6’, 6c, 5’, op, ?Ipv, 1’. The sutures are indicated by low ridges, from which arise small prominences, too blunt and short to be called spines. The number of apical plates is doubtful only because a small circular plate appears to cap the horn: four other plates can clearly be seen, plate 1’ being unusually large and broad. Plate 6” of the precingular series is reduced to accommodate an obliquely positioned anterior intercalary plate: plate 3” is lost in archaeopyle formation.

214 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

The cingulum is narrow and pronouncedly laevorotatory, its two ends differing in antero-posterior position by three times its breadth. The sulcus is broad: it is widely separated from the apex by the enlarged plate 1’, from the antapex (apparent- ly) by a narrow posterior ventral plate, whose boundary was only doubtfully determined.

Only five postcingular plates appear to be present, the first being reduced: no posterior intercalary plate was determinable. The single antapical plate is relatively small.

FIGURED SPECIMEN: BM(NH) slide V.56339(1). Sample CH 231, Kimmeridge Clay (Wheatleyensis Zone) 22 ft below the Blackstone, Clavells Hard, Dorset.

DIMENSIONS: Figured specimen; overall length 120u, length of apical horn 30, breadth 72-5u, crests c. I-5u high.

REMARKS: This single specimen certainly represents an undescribed species of Gonyaulacysta, characterized by its shape and tabulation. In general proportions, it is akin to a number of other Upper Jurassic and Lower Cretaceous species, all of which, however, have more elaborate crests and a more complex tabulation.

Fic. 16. Gonyaulacystasp.H. Tabulation and archaeopyle formation (plate 3” is missing). Left, in oblique ventral view; right, in oblique dorsal view. Specimen BM(NH) slide V.56339 (1). X c.640.

FROM ENGLAND, SCOTLAND AND FRANCE 215

Genus LEPTODINIUM Klement, 1960b emend. Sarjeant, 1969 Leptodinium aceras (Eisenack) Sarjeant, 1969, emend. Plate 5, figures 1-3, text-figure 17

1958 Gonyaulax aceras Eisenack, 391, pl. 2, figs 1-2.

1962 G. acevas Eisenack; G. & M. Deflandre, fiche 1752.

1964 G. acevas Eisenack; Downie and Sarjeant, 113.

1964 G. acevas Eisenack; Eisenack, 311.

1966 ?Gonyaulacysta aceyas (Eisenack); Sarjeant, nomen nudum, 131. 1967b G. acevas (Eisenack); Sarjeant, nomen nudum, table t.

1967b Gonyaulax acevas Eisenack; Vozzhennikova, table II.

1969 Leptodinium aceras (Eisenack); Sarjeant, 12.

EMENDED DIAGNOSIS: Cyst broadly ovoidal to subspherical, without appendages, with tabulation 4’, 1a, 6’, 6c, 6’”, Ip, ipvand 1’. Cingulum helicoid, laevorotatory ; sulcus narrow, extending on both epitract and hypotract. Sutures in the form of low membraneous crests. Surface coarsely granular. Archaeopyle rarely developed, formed by loss of plate 3’’.

HototyPe: Tubingen, Geol.-Palaont. Institut, Pr. 1125, Ob. Apt. Nr.g. Aptian, North Germany.

FIGURED SPECIMENS: I.G.S. slide PK.108. Sample WB 7, from H.M. Geological Survey Borehole, Warlingham, Surrey, at 2834 ft 7in. depth. Lower Kimmeridgian (Eudoxus Zone).

Fic. 17. Leptodinium acevas (Eisenack). Tabulation: left, in ventral view; right, in dorsal view. I.G.S. slide PK 108. X c.1041.

216 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

I.G.S. slide PK.112. Sample WB 13, from the Warlingham Borehole, at 2684 ft 3 in. depth. Lower Kimmeridgian (Autissiodorensis Zone).

DIMENSIONS: Holotype: overall length 85u, breadth 71u. Range of the observed specimens, which are 16 in number; overall length 60-95y, breadth 52—90u.

DEscRIPTION: The moderately thick-walled cyst is divided into two unequal parts by the relatively narrow cingulum. The epitract is slightly longer than the hypotract: both are dome-shaped.

Four apical plates, of variable shape and size, make up the apex: plate 1’ is elongate and sigmoidal, almost as long as the sulcus, and extending down two-thirds of the epitract. Plates 2’ and 3’ are quite large; plate 4’ is reduced because of the larger anterior intercalary plate (Ia), which is also the reason for the reduction of precingular plate 6’. The other precingular plates are relatively large. The cingular plates are narrow and long, plate 1c being smallest. Six postcingular plates are present, with plate 1’ reduced and triangular. Plate 4’’’ is the largest plate of the cyst. The intercalary plate (1p) is placed below the plate 1’”, its outbulge causing the sulcus to become narrower. A quite large plate, Ipv, separates the single antapical plate 1’’” from the sulcus and plate Ip.

OBSERVED RANGE: Kimmeridgian (Autissiodorensis to Pectinatus).

TOTAL KNOWN RANGE: Kimmeridgian (Autissiodorensis to Pectinatus) and Aptian.

REMARKS: The diagnosis is emended to include reference to the tabulation and the mode of archaeopyle formation. (In his original diagnosis, Eisenack was unable to give the tabulation pattern.) Sixteen specimens were observed in the Kimmeridgian assemblages from England: although they were not perfectly preserved, it was possible to determine the tabulation, one of them (the figured specimen) showing it particularly well. A precingular archaeopyle was observed in only two of the specimens; an archaeopyle of this type is figured by Eisenack, who recorded this species from the Aptian of Germany and placed it in Gonyaulax. Since it has no apical horn it was transferred to the genus Leptodinium by Sarjeant (1969).

Leptodinium amabilis (Deflandre) Sarjeant, 1969 Plate ro, figures 5-6, text-figure 18

1939b Gonyaulax amabilis Deflandre, 143, pl. 6, fig. 8.

1941b G. amabilis Deflandre; Deflandre, 11, pl. 3, figs 8-9, text-figs 1-2. 1962 G. amabilis Deflandre; G. & M. Deflandre, fiche 1755.

1964 G. amabilis Deflandre; Downie and Sarjeant, 113.

1964 G. amabilis Deflandre; Eisenack, 315-316.

1964 G. amabilis Deflandre; Sarjeant, table 2.

1966 Gonyaulacysta amabilis (Deflandre); Sarjeant, nomen nudum, 130. 1967b G. amabilis (Deflandre) ; Sarjeant, nomen nudum, table 1.

1967b Gonyaulax amabilis Deflandre; Vozzhennikova, 91, table 11. 1969 Leptodinium amabilis (Deflandre) ; Sarjeant, 12.

1970 L. amabilis (Deflandre) ; Gitmez, 269-70, pl. 12, figs 1-2.

FROM ENGLAND, SCOTLAND AND FRANCE 217

DeEscriPTIoNn: The cyst is broadly ovoidal, with the tabulation 4’, 6”, 6c, 6’”, Ip, Ipv, 1’. The strongly spiral cingulum divides the cyst more or less equally. The sulcus is long and extends onto both epitract and hypotract, being narrow on the epitract. Moderately high crests arise from the sutures, distally feebly denticulate or smooth. The surface of the shell is smooth or finely granular. A precingular archaeopyle is sometimes developed, by loss of plate 3”.

FIGURED SPECIMEN: BM(NH) slide V.56350(1)._ Sample OF 485, from the road side, Montard d’Oignon, France. Lower Kimmeridgian (Mutabilis Zone).

Dimensions: Figured specimen: overall length 38u, breadth 34u. Range of the observed specimens (8 in number): length 38-5o0u, breadth 34—45u, sutures length 3-4u. Deflandre gave the dimensions of the holotype as length 38y, breadth 32u. The observed specimens are larger than the holotype.

RemaRKs: L. amabilis has been previously recorded only from the Kimmeridgian of France. It was infrequent in the samples from England and France, five speci- mens from the Lower Kimmeridgian (Baylei to Mutabilis Zones) and three specimens from the Upper Kimmeridgian (Pectinatus) being observed. It is recorded from English assemblages for the first time: the mode of archaeopyle formation for this species is also recorded for the first time. In general structure and tabulation, the observed specimens correspond closely to the holotype.

Fic. 18. Leptodiniwm amabilis (Deflandre). Tabulation and archaeopyle formation (plate 3” is missing): left, in ventral view; right, in dorsal view. Specimen BM(NH) slide V.56350 (I). X c.1690.

218 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

Leptodinium cf. crassinervum (Deflandre) Sarjeant, 1969 Plate 3, figure 8, plate 5, figures 4-6; text-figure 19

DeEscrIPTION: The shell is broadly ovoidal to polygonal in shape. The cyst wall is thick (c. 2:54) and densely granular. The crests are membraneous in character and arise from slight thickenings of the periphragm; although they are low, they are quite obvious. They give rise to occasional short spines, up to c. 4—4-5u in height. Tabulation: 4’, Ia, 6”, 6c, 6’”, Ip, Ipv, 1’”’’ and 2s. Plate 1’ is elongate, extending down almost two-thirds of the epitract. Plates 2’, 3’ and 4’ are more or less equal in size. The precingular plates are relatively large, except plate 6’’ which is reduced because of the large anterior intercalary plate (1a). The six postcingular plates are of variable shape and size: plate 2’ is reduced to accommodate the intercalary plate (1p) which is rather large. Plate 1’’’ is quite small; the other postcingular plates are relatively large, plate 4’’’ being the largest of all the plates. A crescent- shaped posterior ventral plate, Ipv, separates the sulcal plates from the single antapical plate.

The cingulum is moderately narrow, formed by six plates (plate 6c is very small) ; it is slightly spiral, laevorotatory and divides the theca unequally: the epitract is twice as large as the hypotract. The sulcus is short and broad; its posterior portion is formed by two sulcal plates, one small, the second quite large.

Fic. 19. Leptodinium cf. crassinervum (Deflandre). Tabulation and archaeopyle forma- tion: left, in ventral view; right, in dorsal view. Specimen BM(NH) slide V.56346 (1). x c.960.

FROM ENGLAND, SCOTLAND AND FRANCE 219

On the dorsal side of the shell, a large precingular archaeopyle is formed by loss of plate 3”.

FIGURED SPECIMENS: BM(NH) slide V.56346(1), (3) and V.56351(1), Sample LO 353 from the Littleworth Quarry, Oxfordshire. Upper Kimmeridgian (Pallasioides Zone).

DIMENsIons: Range of the observed specimens: overall length 68—80u, breadth 50-68. Measured specimens 4 in number.

Deflandre gave the following dimensions for L. crassinervum: length 82u, breadth 69u. The English Kimmeridgian specimens found which are similar to Deflandre’s specimen, are slightly smaller.

REMARKS: L. crassinervum has been recorded only from the Kimmeridgian of France (by Deflandre) ; the species was based on a single specimen and, because of its poor preservation, Deflandre was unable to determine the tabulation. Later Sarjeant re-studied the holotype and, on the basis of its similarity to other Jurassic species, re-attributed it to Gonyaulacysta. Recently, it was transferred to Leptodinium, on the basis of its lack of an apical horn.

Four specimens observed, all from the Pallasioides Zone of England (sample LO 353), exhibit a strong resemblance, in their long epitract and thick shell wall, to L. crassinervum, but could not be attributed to that species with confidence, in absence of knowledge of the tabulation of the holotype.

Leptodinium sp. Plate 3, figure 9; text-figure 20

DESCRIPTION: Cyst spherical to subspherical, an appearance of polygonality being imparted by the crests. Tabulation: 4’, Ia, 6”, 6c, 6’’, Ip, 1’; plate boundaries bearing relatively high, delicate, distally denticulate crests.

The first apical plate, 1’, is long and narrow, occupying the anterior extension of the sulcus. Plates 2’ and 3’ are comparable in shape and size, but plate 4’ is markedly larger. Between the plates 1’ and 4’, a small, elongate intercalary plate (Ia) is accommodated. The precingular plates are generally large, except plate 6”, which is narrow. The postcingular plates are also all quite large. Plate 6’” is slightly reduced. Because of the bad orientation of the specimen, the exact shape and size of plates 1’” and 2’” is not very clear, but plates 3’” and 4’” appear the largest of all the plates. The single large antapical plate (1’’’’) is pronouncedly convex.

The cingulum is strongly helicoid, laevorotatory, occupied by six relatively large

cingular plates. It divides the cyst unequally: the sulcus is sigmoidal in shape and _ extends between the apex and antapex, narrowing to the two ends.

The surface is smooth and the wall is transparent. A precingular archaeopyle

_ formed by loss of plate 3’’ was observed in some of the specimens.

FIGURED SPECIMEN: BM(NH) slide V.56352, sample HC 246, from 140 ft below _the Massive Bed, Hounstout Cliff, Dorset. Upper Kimmeridgian (Rotunda Zone).

D

Fic. 20. Leptodinium sp. Tabulation and archaeopyle formation (plate 3” is missing): left, in oblique ventral view; right, in oblique dorsal view. BM(NH) slide V.56352. X c.1620.

DIMENSIONS: Figured specimen: overall length 4ou, breadth 4ou. RANGE: Length 40-60u, breadth 30-52u. Measured specimens 11 in number. OBSERVED RANGE: Kimmeridgian (Wheatleyensis to Rotunda).

REMARKS: Twelve specimens were observed in the Middle-Upper Kimmeridgian assemblages of England, which are comparable in their morphological features to Leptodinium, but differ from the other species of this genus. They possibly represent a new species. Unfortunately, all the specimens observed are somewhat crushed, folded and in a bad orientation; the figured specimen was the best. The small size and the transparent shell wall of the specimens increase the difficulty of determining the tabulation. Accordingly, no new specific name for these forms can yet be proposed.

The most closely comparable species is L. amabilis, which is similarly small. However, these Kimmeridgian forms are clearly distinguished by the character of their crests, the presence of an anterior intercalary plate, the absence of the posterior ventral plate and the details of the rest of the tabulation.

Genus OCCISUCYSTA Gitmez, 1969

Occisucysta evitti (Dodekova) Gitmez, 1970

1969 Gonyaulacysta evitti Dodekova, 14-15, pl. 1, figs 1-8, table r. 1970 Occisucysta evitti (Dodekova) ; Gitmez, 269.

DESCRIPTION: Cyst spherical, with the tabulation 4’, ?1a, 6’, 6-7c, 7’, Ip, Ipv, 1’’’’, No true apical horn appears to be present, the apical prominence being

FROM ENGLAND, SCOTLAND AND FRANCE 221

formed by the confluence of crests. The cingulum is strongly helicoid, laevorotatory, dividing the cyst unequally: the epitract is slightly longer than the hypotract. The sulcus is short. The sutures bear delicate, perforate, membraneous crests, with denticulate edges. Around the apex, the crests are apparently higher than else- where. The surface is granular and tuberculate, also possessing lines of small spines which form “‘double sutures”, parallel to the true sutures. A precingular archae- opyle is formed by loss of plates 2’’ and 3”.

Ho.otyPe: Dodekova’s collection, Jmp/DO-16. Tithonian, Bulgaria.

Dimensions: Holotype: overall length 82u, breadth 82u, crests 4u, high on the sutures, apical crests 8y high.

RemARKS: The description and dimensions mentioned here are as given by Dodekova. This species is characterized by a two-plate precingular archaeopyle. Although Dodekova did not mention the anterior intercalary plate (1a), the photo- graphs of the holotype of G. evztti show that the plate above 6” (which was indicated as plate 4’) is, in fact, plate 1a, a boundary being present at its anterior end, separating off a small plate 4’. The position of the plates 7’” and 1pv is exactly the same asin O. balios. In tabulation and the character of archaeopyle, therefore, this species corresponds to the genus Occisucysta and is accordingly reallocated to it, despite the lack of a true apical horn. Erection of a second genus, to accommodate hornless forms, may prove desirable in the future.

Occisucysta monoheuriskos sp. nov. Plate 7, figures 10-11, text-figure 21

DERIVATION OF THE NAME: Greek, monos, one, single, heurisko, find, discover; referring to the discovery of a single specimen.

Diacnosis: A species of Occisucysta with a globular cyst. Tabulation: 4’, 6”, 7c, 7'", Ip, Ipv, 1’, 2s. The epitract and hypotract are almost equal in size, separated by the only slightly spiral, laevorotatory cingulum. The sulcus is broad and short, stretching from about mid-point on the epitract to about mid-point on the hypotract. The surface is finely granular and sparsely tuberculate. Sutures are in the form of spine rows; the spines are distally closed, oblate or bifid, generally simple, but the spines near to the apical horn are connected distally and thus appear more complicated. Short spines surround the distal end of the apical horn like a corona. A two-plate precingular archaeopyle is typically present, forming by loss of plates 2’’ and 3”’.

HoLotyre: BM(NH) slide V.56353(z). Sample CS 421, from the Sand Stone dyke, first lower Meleagrinella Band of Eathie Haven, South Cromarty, Scotland, Lower Kimmeridgian (Cymodoce Zone).

Dimensions: Holotype: overall length 7ou, breadth 65y, horn length roy; length of the sutural spines 5y, length of the spines surrounding the apical horn 3y.

Description: The cyst wall is relatively thick, c. 1:5u. The cylindrical horn rises from the top of the dome-shaped epitract and is formed by two of the four apical

222 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

plates, 1’ and 2’. Plate 1’ is elongate and broad, corresponding in position to the anterior prolongation of the sulcus. The other three apical plates are relatively large and of more or less equal size. The six precingular plates, together with the apical plates, make up the epitract; no intercalary plates are present. Except for plate 6’, which is narrow and elongate, the precingular plates are quite large; plates 2’ and 3” are lost in archaeopyle formation. The cingulum is occupied by 7 plates of variable size, the first and last (1c and 7c) being small. The ends of the cingulum are widely separated by a very broad sulcus, which widens further in its posterior portion. Two small plates are developed in the mid-portion of the sulcus.

The hypotract is hemi-spherical. The first postcingular plate (1’’’) is small, quadrate and placed in the flank of the sulcus. Plate 2’” is roughly triangular, and, because of the presence of the intercalary plate (1p), is reduced in size. Plates 3'", 4’ and 5’” are very large. Plate 6’” is narrow and elongate; plate 7’” is small. The single, convex antapical plate (1’’’’) is quite large. Two intercalary plates, Ip and Ipv, together form a crescent shape; Ipv separates the sulcus from the antapex.

REMARKS: O. monoheuriskos has been found in only one sample (CS 421) and only one well preserved specimen has so far been observed. Its general shape, tabulation and the sutural features, combined with the two-plate precingular archaeopyle, agree with the diagnosis of the genus Occisucysta. Since all the morphologic features are discernible on the one well-preserved specimen, it is proposed without hesitation as a new species of Occisucysta.

O. monoheuriskos is distinguished from the two other described species of the genus by its apical horn of different form and aspect, its lack of an anterior inter-

Fic. 21. Occisucysta monoheuriskos sp. nov. Tabulation and archaeopyle formation (plates 2” and 3” are missing): left, in ventral view; right, in dorsal view. Holotype: BM(NH) slide V.56353 (1). xX c.1041.

"maaan

FROM ENGLAND, SCOTLAND AND FRANCE 223

calary plate, and a number of details of the tabulation. In its possession of a cylindrical apical horn distally surrounded by spines and its spinous sutures, it shows some similarity to Occisucysta sp. of Gitmez (1969), but since bad preservation precluded any detailed knowledge of Occisucysta sp., an extended comparison is not possible.

Cyst-Family MICRODINIACEAE Eisenack, 1964, emend. Sarjeant and Downie, 1966

Genus DICTYOPYXIS Cookson and Eisenack, 1960b Dictyopyxis areolata Cookson and Eisenack, 1960b Plate 7, figure 9

1955 Membranilarvnax ovulum Valensi, 590, pl. 2, fig. 4, pl. 5, fig. 6.

1960b Dictyopyxis aveolata Cookson and Eisenack, 255-6, pl. 39, figs 12-14.

1961 Dictyopyxidia aveolata (Cookson and Eisenack); Eisenack, nomen nudum, 316.

1962b Dictyopyxis aveolata Cookson and Eisenack; Sarjeant, 494, pl. 70, fig. 19; text-fig. 13, tables 2-3.

1964 Dictyopyxidia aveolata (Cookson and Eisenack); Downie and Sarjeant, nomen nudum 110.

1964 Dictyopyxis aveolata Cookson and Eisenack; Sarjeant, table 3.

1966b Ellipsoidictyum areolata (Cookson and Eisenack); G. & M. Deflandre, fiches 3318-9.

1967b Dictyopyxidia aveolata (Cookson and Eisenack); Sarjeant, nomen nudum table 2.

1968 Dictyopyxis aveolata Cookson and Eisenack; Sarjeant, 229-30, pl. 1, fig. 1; text-fig. 5.

1970 Dictyopyxis sp. Gitmez, 275-6, pl. 1, fig. 1, table 4.

DEscriPTION: This form is characterized by its highly reticulate surface. Each small field is demarcated by high sutures. The cyst is ovoidal to subspherical in shape. The arrangement of reticulae as a pattern of tabulation suggested by Sarjeant (1968) was not determined; however, some small fields on the equatorial plane together form a median line, which may well be equivalent to the cingulum. The apex is lost in archaeopyle formation.

FIGURED SPECIMEN: BM(NH) slide V.53956(1). Sample OM 420, from 2o ft above the Ringstead Coral Bed, west of Osmington Mills, Dorset. Lower Kimmeridgian (Baylei Zone).

Dimensions: Figured specimen: overall length (apex lacking) 55u, breadth 5ou. Range of the English specimens (3 specimens were measured): overall length 50-55n, length without apex 30u, breadth 40-50u. Scottish specimens: length (apex lacking) 45-60u, breadth 40-554 (3 specimens measured). French specimens: overall length 50-60y, length without apex 35-50p, breadth 35-52. (6 specimens measured).

Cookson and Eisenack gave the range for the Australian specimens as 86-124 length and 54—66u breadth, which makes them larger than the European Kimmer- idgian specimens.

OBSERVED RANGE: Lower Kimmeridgian (Baylei to Mutabilis).

TOTAL KNOWN RANGE: Upper Callovian (Lamberti) to Lower Kimmeridgian (Mutabilis).

224 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

REMARKS: D. areolata was described originally from the Oxfordian to Lower Kimmeridgian of Australia; subsequently, Sarjeant recorded it from the Oxfordian of England (1962) and the Callovian of France (1968). The specimens illustrated by Valensi as Membranilarnax ovulum, from Magdalenian flints of presumed Upper Jurassic age, appear attributable to this species.

Three specimens from Scotland, eight specimens from France and four specimens from England were observed; they occur only in the Lower Kimmeridgian assem- blages. These specimens agree in their morphologic features with the holotype of D. areolata.

Dictyopyxis cf. reticulata (Valensi) Sarjeant, 1968 Plate 7, figures 4-5; Plate 12, figures 1-2

DESCRIPTION: The shell is ovoidal with a reticulate surface; the reticulation is irregular. The crests surrounding the small areas are not so high as in D. areolata. Spines rise up from the crest nodes; they are solid, simple, oblate or bifid. There is no obvious tabulation, but more regularly formed reticulae make up a median band equivalent to a cingulum. A large apical archaeopyle, with polygonal outline, is developed.

FIGURED SPECIMENS: BM(NH) slide V.56354(1). Sample OF 485, from the road side, Montard d’Oignon, France. Lower Kimmeridgian (Mutabilis Zone).

BM(NH) slide V.56355(z). Sample 486, from the top of the Oignon Beds, west of Lac du Chavoley, France. Lower Kimmeridgian (Mutabilis Zone).

DIMENSIONS: Range of the French specimens (7 specimens measured) overall length 50-61y, length without apex 30-55u, breadth 23-65y, spine length 6-8y. Scottish specimen: length (apex lacking) 35yu, breadth 48y. Valensi gave the dimensions for this species as 45y length (apex lacking) and 52y breadth, spine length, 3. The Kimmeridgian specimens are similar in size, but their spines are longer than those of the holotype.

Remarks: D. reticulata was recorded from the Bajocian of Calvados and Bathonian of Vienne, as a species of Palaeoperidinium, by Valensi. It was trans- ferred to the genus Dictyopyxis by Sarjeant (1968). Single specimens from the Cymodoce Zone of Scotland and France and eight specimens from the Mutabilis Zone of France were recorded, which are closely similar to the specimen figured by Valensi, except for a greater length of the spines arising from the crest nodes. In view of this minor difference in morphology and the stratigraphical hiatus, these nine specimens were compared with, but not placed in, D. reticulata.

Genus MEIOUROGONYAULAX Sarjeant, 1966a Meiourogonyaulax staffinensis Gitmez, 1970 Plate 9, fig. 4 1970 Meiourogonyaulax staffinensis Gitmez, 276-8, pl. 3, fig. 1, text-fig 20 a, b. FIGURED SPECIMEN: BM(NH) slide V.56356, specimen ED 242, Kimmeridge Clay

FROM ENGLAND, SCOTLAND AND FRANCE 225

(Pectinatus Zone) 60 ft above Freshwater Steps Stone Band, Egmont Bight, Dorset.

DIMENSIONS: Overall length (apex lacking) 84, overall breadth g1-5y, length of cyst alone 72u, breadth 81-5u, height of crests 5-gu.

RANGE OF DIMENSIONS: Overall length (apex lacking) 45-98u, overall breadth 42-915.

OBSERVED RANGE: Kimmeridgian (Baylei to Pallasioides).

REMARKS: This species, hitherto recorded only from the Baylei Zone, ranges throughout the Kimmeridgian (though it has not been encountered in some zones). The size range here quoted is significantly greater than that originally quoted (overall length 48—80y, overall breadth 45-78): the specimen figured is one of the largest encountered.

Meiourogonyaulax dicryptos sp. nov. Plate 7, figure 6; text-figure 22

DERIVATION OF THE NAME: Greek, di-, two, double; kvypto, cover; referring to the two-layered shell wall.

DiaGnosis: Cyst subspherical to spherical, with the tabulation 4’, 6”, 6c, 7’”, Ip, 1’, poorly marked by low ridges. Cingulum more or less equatorial and circular. Sulcus deep, short, confined to the hypotract, broadening towards the antapex. Cyst wall thick and composed of two layers: thick endophragm and thin periphragm. The periphragm bulges out on the apex to form a blunt apical horn, a cavity being developed between the two layers. Surface of the cyst is densely granular. Apical archaeopyle always present: sometimes the operculum remains attached to the shell, ventrally.

Fic. 22. Meiourogonyaulax dicryptos sp. nov. Tabulation and archaeopyle formation: left, in ventral view; right, in dorsal view. Holotype, specimen BM(NH) slide V.56357 (1). xX €,1063.

226 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

HoLotyPE: BM(NH) slide V.56357(1). Sample BN 179, from the Marnes a Harpagodes, Benerville, Normandy. Lower Kimmeridgian (Cymodoce Zone).

Dimensions: Holotype: length (apex lacking) 4ou, breadth 52.

RANGE: overall length 70-gou, breadth 52-88y, length (apex lacking) 40-64y. Measured specimens 8 in number.

DESCRIPTION: The spherical to subspherical cyst is divided into two equal parts by the moderately narrow cingulum. Both epitract and hypotract are dome-shaped; the epitract bears a small, blunt, hollow apical horn. Four apical and six precingular plates make up the epitract: plate 1’ is elongate, the other apical plates are broader. Plates 1’ and 6” are somewhat reduced and plates 2” and 5” are the largest of the precingular plates. Seven postcingular plates are present on the hypotract. Plates 1’” and 2’” are reduced because of the presence of a large intercalary plate (rp). Plates 3’, 4”, 5’” and 6’” are relatively large; plate 7’” is reduced. The single antapical plate (1’’’’) is quite large and convex. The cingulum is poorly indicated, formed by six plates of variable size. The sulcus is deep and, in its posterior portion, very wide. An apical archaeopyle forms by loss of the apical plates.

OBSERVED RANGE: Kimmeridgian (Cymodoce, Pectinatus and Rotunda Zones).

REMARKS: Eight specimens (one from France, seven from England) were recorded. These specimens differ from the other species of the genus in the character of their tabulation and possession of a hollow apical horn. In the presence of a seventh postcingular plate, this new species differs from the typical Meourogonyaulax tabulation: however, it corresponds in all other respects.

Meiourogonyaulax pila sp. nov. Plate 4, figure 5, plate 7, figure 3; text-figure 23

DERIVATION OF THE NAME: Latin, pila, ball; referring to the shape of the cyst.

Diacnosis: Cyst almost circular in outline, without an apical horn and rounded at the antapex. Wall moderately thin, surface finely granular. Tabulation: 4’, 6’’, 6c, 6’’’, Ip, Ipv and 1’’’".. Plate boundaries faintly indicated. The cingulum forms a feebly laevorotatory spiral. Apical archaeopyle typically present, with operculum attached ventrally.

HorotyrPe: BM(NH) Slide V.56358; sample FD 236, from + mile west of Fresh- water Steps, Dorset. Middle Kimmeridgian (on the boundary of the Pectinatus and Hudlestoni Zones).

PARATYPE: I.G.S. Slide PK.121, sample WB 19, from H.M. Geological Survey Borehole, Warlingham, Surrey, at 2535 ft 3in. depth. Middle Kimmeridgian (Wheatleyensis Zone).

DIMENSIONS: Holotype: overall length 60, breadth 55u. Range of the observed specimens (Ig in number): length 55—75u, breadth 55—7ouw.

DESCRIPTION: The spherical cyst 1s divided by the moderately wide and more or less equatorial cingulum into two parts, the epitract and the hypotract ; these may

} | '

|

FROM ENGLAND, SCOTLAND AND FRANCE 227

Fic. 23. Meiourogonyaulax pila sp. nov. The holotype, with attached operculum. Left, in ventral view; right, in dorsal view. BM(NH) slide V.56358. x c.1018.

be equal in size or the hypotract may be slightly longer than the epitract. The sulcus is relatively short, mainly placed on the hypotract.

Plate 1’ is elongate, the other three apical plates are more or less similar in shape and size. The precingular plates are generally almost equal in size, except that plate 6” is smaller than the others. The postcingular plates are of variable size: plate 1’” is relatively small, plate 4’” is the largest of all the plates. A quite large intercalary plate (Ip) is situated between the plates 2’” and 1’’”’; plate 2’” is corres- pondingly reduced. A large posterior ventral plate (Ipv) separates the sulcus from the antapex. A single antapical plate, broad and slightly convex, occupies the antapex. Plate boundaries are marked by low ridges.

RemaRKs: In general appearance, M. pila is similar to the members of the genus Canningia, for example C. minor and C. ringnesit; however, since the species of Canningia exhibit no tabulation, except for the weakly indicated cingulum, there is no possibility of confusion with M. pila. Because of the tabulation and apical archeopyle, the species is clearly attributable to Metourogonyaulax. M. pila was recorded from the Middle and Upper Kimmeridgian (Scitulus to Rotunda Zones) from England only: it was not observed in Lower Kimmeridgian assemblages.

Meiourogonyaulax sp.

Plate 4, figure 4, plate 7, figure 12

DESCRIPTION: Cyst broadly ovoidal, almost spherical in shape, without apical horn and with rounded antapex. There is often no indication of tabulation, but in

28 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

2

some of the specimens two apical and three precingular plates were recognised with difficulty on the dorsal side. The cingulum is frequently weakly indicated. The cyst wall is thick and is densely granular. An apical archaeopyle is typically present, formed by the rupture of the apex along a more or less straight line with small V- shaped notches along the edge, corresponding to the positions of sutures. Generally, the operculum remains attached ventrally.

FIGURED SPECIMENS: BM(NH) slide V.56359(1). Sample LO 352 from the Littleworth quarry, Oxfordshire. Upper Kimmeridgian (Pallasioides Zone). 1.G.S. slide PK.100, sample WB 1, from H.M. Geological Survey Borehole, Warlingham, Surrey, at 7984 ft 7 in. depth. Lower Kimmeridgian (Mutabilis Zone).

DIMENSIONS: Figured specimen from Warlingham Borehole: overall length 75y, breadth 68. Figured specimen from Littleworth: overall length 63y, breadth 6o0u. Range: overall length 48-95y, breadth 46-85. Measured specimens 15 in number.

ReMARKS: A group of specimens (22 in number) observed in the Kimmeridgian of England are, in their general appearance, with granular surface and attached operculum, similar to Canningia ringnesi (recorded from the Upper Cretaceous of the Arctic by Manum and Cookson, 1964). However, they differ in that they have a relatively thick cyst wall and tabulation (albeit poorly indicated). Because of the mode of archaeopyle formation and the slight indication of a tabulation, these specimens are allocated to the genus Meiourogonyaulax. It is possible that they may be intermediate forms between the two genera, in which the tabulation is becoming progressively less apparent.

Genus EGMONTODINIUM gen. nov.

DERIVATION OF NAME: Named after the type locality—Egmont Bight, Dorset.

DiaAGnosiIs: Proximate cyst, spherical to ovoidal. Tabulation typically 4’, 5 or 6 ac., 6’, 6c, 6’, op, 2pv, Ope, 1’’”’: additional, very small platelets may be developed at crest nodes and the posterior tabulation is subject to some variation. No apical or other horns are developed. Crests or spinelets may arise from the sutures and spines may also be present, singly or in rows, on some plates. Archaeopyle apical, formed by schism along the anterior circle; the operculum frequently remains attached.

TyPE SPECIES: Egmontodinium polyplacophorum sp. nov. Kimmeridge Clay (Kimmeridgian: Pectinatus Zone), Egmont Bight, Dorset.

REMARKS: This genus is distinguished from all others yet described inits tabulation. The plates surrounding the apex might be termed anterior intercalaries: the authors, however, feel that this would be inappropriate, since they are not merely interposed between existing reflected plate series but constitute an additional series. The new name “‘anterior circle plate” is thus coined for them. The plates surrounding the antapex are similarly designated “‘posterior circle plates”, following the precedent of another Jurassic genus, Pluriarvalium.

The most comparable genus is Ellipsoidictywm Klement 1960, whose complex tabulation was described in detail by Gocht (1970, pp. 150-2): however, the tabula-

FROM ENGLAND, SCOTLAND AND FRANCE 229

tion of the epitract of this genus is markedly dissimilar and a close affinity cannot be considered probable.

The familial allocation of this genus is based on its proximate character and apical archaeopyle: the tabulation does not accord with that specified by its authors for this cyst family and a reallocation may prove necessary in the futute.

Egmontodinium polyplacophorum sp. nov. Plate 8, figures 1-4; Plate 9, figure 3; Plate 11, figures 5-6, 8; text-figure 24

DERIVATION OF NAME: In reference to the large number of plates developed.

Diacnosis: A species of Egmontodinium having an ovoidal cyst, thin walled and without prominent granulation or punctation. Sutures variably ornamented with delicate crests or with rows of spinelets, simple or bifurcate and sometimes distally connected; the sutural ornamentation is most prominent around the antapex. The tabulation is as for the genus, but shows some variation in detail through the presence or absence of additional small plates at sutural nodes.

Ho.LotyPe: BM(NH) slide V.56360(2b), Kimmeridge Clay (Pectinatus Zone) 60 ft above Freshwater Steps Stone Band, Egmont Bight, Dorset. Paratypes: a. BM(NH) slide V.56360(1)._ b. BM(NH) slide V.56360(2a), showing apical archaeopyle. c. BM(NH) slide V.56347(1), also showing archaeopyle. [All paratypes are from the same locality and horizon as the holotype. |

DimEnSIONS: Holotype: length of cyst 76u, breadth 58, maximum height of crests 3°5u. Paratype a: length 78, breadth 61-5u, maximum height of crests 3u. Paratype b: length (apex lacking) 68, breadth 59:5u, maximum height of crests c. 5u. Paratype c: length (apex lacking) 6Iu, breadth 661, maximum height of crests c. 5:5u. Range: overall length 60-80y, length without apex 50-68y, breadth 45-65u., maximum height of crests c. 3-5u. Measured specimens: 15.

DescriPTIOn: The cyst surface is typically smooth, but may exhibit minute granulation or punctation. Ornament is normally confined to the sutures: in some instances, however, one or a few isolated spines, or a short row of spines not traversing the plate, may be present within a plate boundary.

Four apical plates are developed, the first being the largest and situated in the anterior prolongation of the sulcus. They are surrounded by six (possibly sometimes by only five) elongate plates constituting the anterior circle, plate 3ac being broader than the others in the holotype. The opercular suture opens along this circle, so that the plate boundaries with the apical plates are present in the operculum and those with the precingular plates on the cyst proper. The initial opening of the opercular suture appears to occur on the dorsal side: the operculum frequently retains a ventral attachment with the abandoned cyst. The holotype shows no opening: paratype (a) is partly open, though the operculum remains attached and almost in place; paratypes (b) and (c) lack the operculum.

The six precingular plates are almost of equal size. In the holotype, a small round plate is present at the node of the crest separating plates 3’ and 4” and that bounding the operculum: no such plate was distinguished on the paratypes. The

230 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

4’

6a (paca Sac 6" 2

Fic. 24. Egmontodinium polyplacophorum gen. et. sp. nov. Reconstruction of the tabula- tion, A~-B Holotype, BM(NH) slide V.56360 (2b) in ventral view and dorsal view. C-D Paratype (a), BM(NH) slide V.56360 (1) in ventral and dorsal view. x c.958.

FROM ENGLAND, SCOTLAND AND FRANCE 231

cingulum is made up of six elongate plates with clear boundaries. The sulcus is broad and occupies only the central part of the ventral surface, being separated from the apex by the large plate 1’ and from the antapex by two posterior ventral plates and by two plates of the posterior circle. In the holotype, a small plate is present at the junction of the sulcus and plate 6’”.

Of the six postcingular plates, the first is extremely small (as is the case in many species of Gonyaulacysta) and may be masked by its bounding crests: plate 2’’’ is also reduced, to accommodate the second posterior ventral plate, an equivalent of the posterior intercalary plate but displaced to the ventral side. The other four post- cingular plates are quite large. The plates of the posterior circle are quite variable in form: in particular, plates 3pc and 5pc sometimes exhibit a remarkable “‘tail’’ extending along the sulcus separating two dorsal postcingular plates. The boundaries between the ventral posterior circle plates are in some instances distinguishable only with difficulty, if at all: the holotype is unusually clear in structure and exceptionally favourably orientated. The antapical plate is polygonal and quite large: on two specimens (paratypes a and c) it is partially subdivided by a row of proximally connected spines, but this row only traverses half the plate.

The ornamention of the sutures is highly variable, from rows of isolated, simple or bifurcate spines, with or without distal or (more commonly) proximal connections, to simple delicate crests of moderate height. The highest crests are generally those bounding the cingulum and antapex. (The character of the crests may be modified by accidents of preservation.) The crests or spines are usually little more than one- tenth of the cyst breadth in height: difficulty is often experienced in distinguishing particular crests.

OBSERVED RANGE: Middle to Upper Kimmeridgian (Wheatleyensis to Pectinatus Zones).

Cyst-Family PAREODINIACEAE Gocht, emend. Sarjeant & Downie 1966 Genus APTEODINIUM Eisenack, 1958 Apteodinium cf. maculatum Eisenack & Cookson Plate 12, figure 6

FIGURED SPECIMEN: I.G.S. slide PK.105, sample WB 4, from H.M. Geological Survey Borehole, Warlingham, at 2g10 ft 6in. depth. Lower Kimmeridgian (Eudoxus Zone).

DIMENSIONS: Figured specimen: overall length 83, breadth 80u.

RANGE: (7 specimens measured): length 50-83u, breadth 45-80y, apical horn length 3-4. Range of the dimensions of Australian specimens, as given by Eisenack and Cookson: length 74-105, breadth 70-105y. The Kimmeridgian specimens are thus of comparable size.

REMARKS: Nine specimens recorded, seven from English and two from French assemblages, are similar to A. maculatum, as recorded and described from the Albian by Eisenack and Cookson. They differ from the Australian specimens in

232 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

having a precingular archaeopyle and in the absence of the small thickened areas with circular outlines that give A. maculatwm its characteristic appearance; in only one specimen, from the Warlingham assemblage, were similar small circular areas observed. The Kimmeridgian specimens must be thus considered only comparable to, and not definitely conspecific with, the Australian specimens.

OBSERVED RANGE: Lower to Middle Kimmeridgian (Baylei to Rotunda Zones). Not yet observed from the Cymodoce and Elegans Zones.

Genus IMBATODINIUM Vozzhennikova, 1967b Imbatodinium antennatum sp. nov. Plate 11, figures 2-3 1970 Imbatodinium sp. Gitmez, 282, pl. 7, fig. 5, table 4.

DERIVATION OF THE NAME: Latin, antenna, feeler, a sensory appendage on the head of an insect—with reference to the similarity of the process rising up from the apical horn of this species to an insect’s antenna.

Diacnosis: A proximate cyst, elongate to ovoidal in shape, with a strong apical horn. On the distal end of the horn, there is a thread-like projection of variable length ending in a small knob. In some cases, a cingulum is weakly developed; but there is no indication of tabulation or sulcus. The surface of the cyst is granular. An intercalary archaeopyle is often developed.

Ho.otyPe: I.G.S. slide PK.124, sample WB 23, from H.M. Geological Survey Borehole, Warlingham, Surrey, at 2434 {t 6in. depth. Middle Kimmeridgian (Hudle- stoni Zone).

PARATYPE: BM(NH) slide V.56361(1), sample ED 242, from 60 ft above the Fresh- water Steps Stone Band, Egmont Bight, Dorset. Upper Kimmeridgian (Rotunda Zone).

Dimensions: Holotype: overall length 73u, breadth 35y, apical horn length without projection gu, with projection I6u. Range of the observed specimens (16 in number): overall length 62-100p, breadth 28—5ou, overall length of horn 12— 30p, horn length without projection 7-16y, length of the projection 4-14u.

DESCRIPTION: The cyst is elongate, broadening in the posterior median region. The apical horn is well developed, bearing a thread-like process of variable length, generally between half and one-third of the overall horn length. This thread-like process ends in a small bulge, which appears as a knob. The cingulum is only weakly developed, but may be suggested by faint surface marking. The epitract is longer than the hypotract, comprising almost two-thirds of the overall length.

OBSERVED RANGE: Lower to Upper Kimmerdigian (Baylei to Rotunda Zones). Not yet observed from the Cymodoce and Elegans Zones.

REMARKS: This new species of Imbatodinium is distinguished from the previously described species in its general shape, presence of an intercalary archaeopyle and

FROM ENGLAND, SCOTLAND AND FRANCE 233

characteristic shape of the apical horn. In horn shape, it is similar to J. villoswm, which was recorded from the Upper Jurassic of Russia by Vozzhennikova, but it differs from J. villosum in the absence of the sutural spines distributed all over the surface.

I. antennatum is present in the Kimmeridgian assemblages of England, Scotland and France; however, it is infrequent ; one specimen from France, one specimen from Scotland and fifteen specimens from England being recorded. It is rare in the Lower Kimmeridgian, in which only four specimens were observed. The number increases in the upper horizons: five specimens were recorded from the Middle Kimmeridgian and seven specimens from the Upper Kimmeridgian.

Imbatodinium cf. villosum Vozzhennikova, 1967b Plate 11, figure 1

Description: The cyst is broadly ovoidal, elongate, with a broad based apical horn, distally bearing a process. There is neither tabulation nor sulcus; the cingulum is only faintly indicated. The epitract is longer than the hypotract, comprising almost three-quarters of the whole length of the cyst. The surface of the cyst is coarsely granular and covered by short, thick spines. An archaeopyle, intercalary in position, is occasionally present.

FIGURED SPECIMEN: BM(NH) slide V.56362(1), sample HC 243, from c. 100 ft below the Rotunda Nodules Bed, Chapmans Pool, Dorset. Upper Kimmeridgian (Pectinatus Zone).

DIMENSIONS: Figured specimen: overall length 80u, breadth 4op, horn length 17u. Range: overall length 70-87y, breadth 40-52y, overall length of horn 8-17, horn length without process 4-13, apical process length 4-6; length of the spines over the surface 2:5-4u. Dimensions of J. villosum as given by Vozzhennikova: overall length 70:5—100p, breadth 27—40°5y, apical horn length 10-5-13°5y.

REMARKS: Seven specimens observed in the Upper Kimmeridgian (Pectinatus to Pallasioides Zones) assemblages of England are similar to J. villosum, but smaller and the spines distributed on the surface are relatively shorter; for these reasons, they are compared with, rather than attributed to J. villosum. Vozzhennikova recorded this species from the Upper Jurassic of the Moscow Province, U.S.S.R.

Cyst-family UNCERTAIN Proximate cyst sp. indet. Plate 11, figures 4, 7, 9

DeEscripTION: The shell is broadly ovoidal to subspherical, with two blunt apical horns: the apex is rounded. Tabulation is very faintly indicated; four apical plates, six precingular and one antapical plate were determined with difficulty. The cingulum is indicated by inbulges on the sides of the cyst; the sulcus was not observed. The epitract is longer than the hypotract, so far as is determinable from

234 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

the feeble indications of the cingulum. The archaeopyle is well developed, but in a very different way from the observed archaeopyle types in recorded species: the apex as a whole, together with one of the precingular plates, is thrown off in the archaeopyle formation, though the apex remains attached to the cyst. The surface of the cyst is densely granular; irregularly formed lumps (verrucae) were present on three of the observed specimens.

FIGURED SPECIMENS: BM(NH) slide V.56363(1) and V.56364, sample FD 236, from +} mile west of Freshwater Steps, Dorset. Middle Kimmeridgian (from the boundary of the Pectinatus and Hudlestoni Zones).

DIMENSIONS: Range of the observed specimens (4 in number): overall length 66-83, breadth 60-75u.

REMARKS: Four specimens were observed, in the English Middle and Upper Kimmeridgian assemblages (Wheatleyensis to Rotunda Zones). In its mode of archaeopyle formation, this form is different from all previously described proximate cysts. Because of the bad preservation and dense surface ornamentation, full details of the morphology could not be obtained; in consequence, no new taxon is proposed.

Cyst-Family ADNATOSPHAERIDIACEAE Sarjeant and Downie, 1966 Genus ADNATOSPHAERIDIUM Williams & Downie, 1966 Adnatosphaeridium paucispinum Klement, 1960b, comb. nov. Plate 10, figures 1-4

1960 Cannosphaeropsis paucispina Klement, 72, pl. 10, figs 9-10. 1964 C. paucispina Klement; Downie & Sarjeant, 1or. 1964 C. paucispina Klement; Sarjeant, table 3.

DESCRIPTION: Cyst subspherical to ovoidal in shape, with a thin, smooth shell wall bearing slender, hollow processes, open distally and branched or broad, funnel- shaped, fenestrate. The processes are connected distally by trabeculae. An apical archaeopyle with a zig-zag margin was seen in almost all observed specimens.

FIGURED SPECIMENS: BM(NH) slide V.56365(1) and V.56366(2), sample BN 179, from the Cymodoce Zone of Benerville, Normandy.

DIMENSIONS: Range (8 specimens were measured): length 40-55, breadth 35-60n, length without apex (6 of the measured specimens have apical archaeopyles) 30—50p, process length 15-30u.

OBSERVED RANGE: Lower to Middle Kimmeridgian (Cymodoce to Hudlestoni/ Pectinatus boundary).

TOTAL KNOWN RANGE: Upper Oxfordian (Malm Alph.) to Middle Kimmeridgian (Hudlestoni/Pectinatus Boundary).

REMARKS: This species has previously been recorded from the Upper Oxfordian of Germany by Klement (1960). It is now placed in the genus Adnatosphaeridium on the basis of the presence of an apical archaeopyle. The observed specimens

FROM ENGLAND, SCOTLAND AND FRANCE 235

were recorded from the Lower Kimmeridgian (Cymodoce Zone) of Normandy and Middle Kimmeridgian (Boundary of Hudlestoni and Pectinatus Zones) of Dorset.

Cyst-Family HYSTRICHOSPHAERIDIACEAE Evitt, emend. Sarjeant and Downie, 1966

Genus CLEISTOSPHAERIDIUM Davey, Downie, Sarjeant and Williams, 1969 Cleistosphaeridium sp. Plate 15, figure 3

DEscRIPTION: The cyst is spherical to subspherical, the wall thin and granular. Transparent processes, approximately 60 in number, are present: they are simple, conical, hollow and closed distally; their length is generally less than one-third of the cyst length. An apical archaeopyle is sometimes developed.

FIGURED SPECIMEN: BM(NH) slide V.56367(1), sample ED 242, from 60 ft above the Freshwater Steps Stone Band, Dorset. Upper Kimmeridgian (Pectinatus Zone).

DIMENSIonsS: Figured specimen: Cyst length (apex lacking) 50u, breadth 5oyn, process length 12u. Range (16 specimens measured): cyst length 38-60, length without apex 40-50u, breadth 38-66, length of the processes 10-22u.

OBSERVED RANGE: Lower to Upper Kimmeridgian (Autissiodorensis to Rotunda Zones). Not yet observed from the Scitulus-Hudlestoni Zones.

RemARKS: Although quite a number of specimens (31 in number) were recorded, the preservation was consistently very bad: they were always found covered by debris, so that there was no chance to examine them in detail sufficient to justify giving a specific name. They accord in general appearance, number and type of processes and apical archaeopyle, with the genus Cleistosbhaeridium. The most similar species is perhaps C. machaerophorum Deflandre and Cookson (1955), which was recorded from the Miocene of Australia, but the bad preservation precluded any detailed study.

Genus OLIGOSPHAERIDIUM Davey and Williams, 1966 Oligosphaeridium pulcherrimum (Deflandre and Cookson) Davey and Williams, 1966 Plate 13, figure 3, text-figure 25

1954 Hystrichosphaeridium pulcherrimum Deflandre & Cookson, text-fig. 6, nomen nudum.

1955 H. pulcherrimum Deflandre & Cookson, 270, pl. 1, fig. 8, text-fig. 21.

1955 H. pulcherrimum Deflandre & Cookson; Valensi, 592, pl. 4, fig. 1.

1957 4H. pulcherrimum Deflandre & Cookson; Delcourt & Sprumont, 59, pl. 1, fig. 4, pl. 2, figeei2.

1964 H. pulcherrimum Deflandre & Cookson; Downie & Sarjeant, 121.

1966 Oligosphaeridium pulcherrimum (Deflandre & Cookson); Davey & Williams, 75-6, pl. 10 fig. 9, pl. 11, fig. 5; table r.

1967b O. pulcherrimum (Deflandre & Cookson); Sarjeant, table 6.

E

236 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

1967 Hystrvichosphaeridium pulcherrimum Deflandre & Cookson; Clarke and Verdier, 54-5 1. ro, figs 4-5. 1970 Oh ae ie pulcherrimum (Deflandre & Cookson); Gitmez, 290 pl. 7, fig. 7, table 4. DESCRIPTION: The subspherical cyst possesses processes of two types. Some processes are tubular, distally open, widening distally and assuming a funnel shape with fenestrate walls (characteristic for this species) ; the other processes are simple, bifid or foliate. Because of the complication of the processes and bad preservation, accurate determination of the reflected tabulation was not possible, but the tabu- lation 6’, 5’, Ip, 2’’”’, 3s may be suggested, with five additional simple processes that could not be named and appear random in distribution. The surface of the shell is smooth. Apical archaeopyle developed.

FIGURED SPECIMEN: BM(NH) slide V.56368(1), sample CC 449, from c. ro ft from the top of the Calcaires du Moulin Wibert, Cap de la Creche, Boulonnais. Lower Kimmeridgian (Baylei Zone).

Dimensions: Figured specimen: length (apex lacking) 4ou, breadth 50u, process length 25-28u. Range (4 specimens observed): length (apex lacking) 35-48p, breadth 45-55y, process length 18-35y. Dimensions of the holotype: overall diameter c. 118u, process length 26-38u (as quoted by Deflandre and Cookson).

REMARKS: This species first recorded from the Jurassic by Gitmez (1970). Un- fortunately, all specimens observed are in too poor condition for examination in detail. Four specimens were recorded, all from the Lower Kimmeridgian (Baylei Zone) assemblages, two of them from Dorset and two from France.

Fic. 25. Olhgosphaeridium pulcherrimum (Deflandre & Cookson). Specimen showing the suggested tabulation: left, in ventral view; right, in dorsal view. BM(NH) slide V.56368 (@a e648"

FROM ENGLAND, SCOTLAND AND FRANCE 237

Genus SYSTEMATOPHORA Klement, 1960b Systematophora ovata sp. nov. Plate 14, figures I-3 1970 Systematophora sp. Gitmez, 296, pl. 8, fig. 5, table 4.

DERIVATION OF THE NAME: Latin, ovatus, egg-shaped, with reference to the shape of the cyst.

Diacnosis: A species of Systematophora having an elongate, ovoidal cyst bearing short processes (not more than one-fourth of the cyst breadth). The processes are located in groups: there are ten such groups, one occupying the apex, an opposite one the antapex, whilst eight groups are distributed between the apex and the antapex, four of them on the epitract, the other four on the hypotract. There is no connection between the groups of processes or between the processes in each group. The processes are simple, bifid distally or broad based, foliate and deeply forked at their distal end. The surface of the shell is finely granular. When an archaeopyle is developed, it is apical in position.

Hototyre: BM(NH) slide V.53962(1), sample SC 444, from the Great Ouse River Board Pit, Stretham, Cambridgeshire. Lower Kimmeridgian (Baylei Zone).

PARATYPE: BM(NH) slide V.56343(2), sample HC 243, from c. 100 ft below the Rotunda Nodules, Hounstout Cliff, Dorset. Upper Kimmeridgian (Pectinatus Zone).

Dimensions: Holotype: shell length (apex lacking) 35u, breadth 28y, process length 6-8. Paratype: shell length 58y, breadth 4oy, process length 8y. Range: shell length 58—60u, apex lacking 35-48y, breadth 28—45u, process length 8-1Iy. Measured specimens were 6 in number.

DESCRIPTION: The cyst wall is composed of two layers, the periphragm forming the processes. Both of the layers are thin and transparent. There is no connection between the processes at their proximal and distal ends; they arise separately from each other, positioned around the margins of ovoidal or polygonal fields whose shape is clearly shown by the broad bases of the processes.

REMARKS: This new species was observed very infrequently in the Lower and Upper Kimmeridgian: one specimen was recorded from the Baylei Zone, three specimens from the Pectinatus Zone and one specimen from the Pallasioides Zone of England. Only one specimen was observed in the Scottish assemblages and none from France. The preservation of the specimens was moderately good. This new species differs from all previously described species of the genus on the basis of shape and character of its processes, in combination with the shape of the cyst.

Cyst-Family UNCERTAIN Genus STEPHANELYTRON Sarjeant, I1961a

Stephanelytron redcliffense Sarjeant, 1961a Plate 14, figure 6

238 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

t960c Organism A. Sarjeant, 404, pl. 13, fig. 13, table 2.

1961a Stephanelytron vedcliffense Sarjeant, 109-110, pl. 15, fig. 11, text-figs Io, 15. 1962a S. vedcliffense Sarjeant; Sarjeant, table 4.

1962b S. vedcliffense Sarjeant; Sarjeant, 495, pl. 70, fig. 7, tables 2-3.

1964 S. vedcliffense Sarjeant; Downie and Sarjeant, 146.

1964 S. vedcliffense Sarjeant; Sarjeant, table 4.

1967 _ S. vedcliffense Sarjeant; Brito, pl. 2, fig. 3.

1968b S. vedcliffense Sarjeant; Sarjeant, 225, pl. 3, fig. 5, table 2A.

Description: The cyst is broadly ovoidal, rounded at both ends. The tubular processes extend down the flanks in six rows and surround the apex and antapex in transverse rows; there is also a median transverse row of processes. A corona, broad-based and bearing striations, is present on the antapex. The cyst wall is composed of two layers; both are thin and transparent, without ornamentation. The periphragm forms the processes. There is no connection between the central cavity and the cavity of processes. An apical archaeopyle is typically developed.

FIGURED SPECIMEN: BM(NH) slide V.56366(1), sample BN 179, from the Cymodoce Zone of Benerville, Normandy.

DIMENSIONS: Figured specimen: cyst length 4ou, breadth 32y, process length 5y, corona length 8. Range: length 40-60u, length (apex lacking) 45-5o0y, breadth 32-45p, process length 4-5u, corona length 5—8y. Measured specimens 4 in number. Holotype dimensions, as given by Sarjeant: length 36, breadth 30y, process length 5u, corona length Iou. The Kimmeridgian specimens exhibit similar dimensions, except in the length of the corona, which is greater in the holotype.

REMARKS: This species was originally recorded from the Oxford Clay of England, and later from the Lower Oxfordian of Normandy, by Sarjeant (1961, 1968). Five specimens, all from the same horizon in the Cymodoce Zone of Normandy, were observed, which are closely similar to those figured by Sarjeant.

Stephanelytron cf. redcliffense Sarjeant, 1961a Plate 14, figure 7

DESCRIPTION: Two specimens, also from the Cymodoce Zone, show a general resemblance to S. vedcliffense but differ in that the processes are markedly thinner. Forms of intermediate character were not encountered.

FIGURED SPECIMEN: BM(NH) Slide V.56365(2), sample BN 179 from the Cymodoce Zone of Benerville, Normandy.

Dimensions: Figured specimen: cyst length (apex lacking) 35y, breadth 30u, process length 5, corona length 72. The second specimen could not be measured because of its bad preservation and orientation.

REMARKS: These two specimens from the Lower Kimmeridgian (Cymodoce Zone of France) may represent a new species or might be extremes in the range of morphological variation of S. redcliffense. Fuller information must be awaited.

FROM ENGLAND, SCOTLAND AND FRANCE 239

Cyst-Family ENDOSCRINIACEAE Vozzhennikova, emend. Sarjeant and Downie, 1966

Genus ENDOSCRINIUM Klement, 1960b emend. Vozzhennikova, 1967a Endoscrinium sp. Plate 14, figures 9-11, text-figure 26

DEscrIPTION: A species of Endoscrinium possessing a subspherical to broadly ovoidal periblast, without apical or antapical horns. The periphragm is irregularly studded with pores of varying shapes and sizes; the endophragm is finely granular. The sutural crests are well defined, in the form of low ridges. Reflected tabulation: 4’, 6”, 6c, 5’”, Ip, pv, 1’’’’ and 8s. Plate 1’is elongate, plates 2’ and 3’ are relatively small, the boundary between them was not confirmed; plate 4’ is quite large. The precingular plates and postcingular plates are large, plate 3’’’ being the largest of all the plates. The cingulum is strongly laevorotatory, occupied by six plates; plates 1c and 6c are greatly reduced, the other cingular plates are of constant size. The cingulum divides the cyst more or less equally. The sulcus is broad, occupied by sulcal plates of varying shape and size.

A precingular archaeopyle is developed, and formed by loss of plate 3”.

FIGURED SPECIMEN: BM(NH) slide V.56369(1), sample OF 485, from well-bedded calcilutite, Montard d’Oignon, France. Lower Kimmeridgian (Mutabilis Zone).

Dimensions: Figured specimen: overall length 78u, breadth 75u, endoblast length 60u, breadth 55w.

REMARKS: Only one well preserved specimen has so far been observed. The

Fic. 26. Endoscrinium sp. Tabulation and the archaeopyle formation: left, in ventral view; right, in dorsal view. Specimen BM(NH) slide V.56369 (1). x c.836.

240 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

perforation of the periblast and the presence of sulcal plates distinguish this species from all described forms. The general appearance of the cyst and the tabulation are most comparable to E. luridum, suggesting a relationship between two species.

Cyst-Family HEXAGONIFERACEAE Sarjeant & Downie, 1966 Genus HEXAGONIFERA Cookson and Eisenack, 1961a emend. Cookson and Eisenack, 1962 Hexagonifera jurassica sp. nov. Plate 14, figures 5, 8 1970 Hexagonifera sp. Gitmez, 2, pl. 1, fig. 12, table 4.

DERIVATION OF THE NAME: So named because this is the first species of the genus observed in Jurassic assemblages.

Dracnosis: A species of Hexagonifera with broadly ovoidal to elongate cyst. The periblast has a blunt, hollow apical projection formed by the periphragm only. The surface of the periblast is delicate, smooth or minutely granular: the endoblast, in contrast, is thick-walled and densely granular, with occasional tubercles. A circular cingulum divides the periblast almost equally; sometimes the epitract is slightly smaller than the hypotract. There is no definite indication of tabulation or of a sulcus. An apical archaeopyle is typically developed.

Horotyre: I.G.S. slide PK.123, sample WB 20, from H.M. Geological Survey Borehole, Warlingham, at 2510 ft depth. Middle Kimmeridgian (Wheatleyensis Zone).

PARATYPE: BM(NH) slide V.53621(1), sample SS 627, from 100 ft above the second dolerite sill, Staffin Bay, Skye. Lower Kimmeridgian (Baylei Zone).

Dimensions: Holotype: overall length 85u, breadth 72p, endoblast length 73y, breadth 66u, Paratype: overall length (apex lacking) 46, breadth 50u, endoblast length (apex lacking) 40u, breadth 42u. The size range of specimens from different horizons is shown in Table 1; according to these measurements, the Lower Kim- meridgian specimens are smaller than those from the Middle and Upper Kimmeridgian. (There is only a slight size difference between the Middle and Upper Kimmeridgian specimens). No dimensional difference was observable between the English and French specimens.

OBSERVED RANGE: Lower to Upper Kimmeridgian (Baylei to Pallasioides). Not yet observed from the Cymodoce, Mutabilis, Elegans and Scitulus Zones.

Description: The subspherical to ovoidal endoblast is completely enclosed by the delicate periblast. The epitract of the periblast is conical, with no apical projection superimposed on the cone shape. The epitract of the endoblast is rounded and dome-shaped. The antapex of both periblast and endoblast is rounded. A poly- gonal apical archaeopyle is usually present ; generally the operculum remains attached to the shell. The tabulation is generally indeterminable, but the dorsal tabulation

FROM ENGLAND, SCOTLAND AND FRANCE 241

could be distinguished with difficulty on some specimens: two apical, three pre- cingular, three postcingular and one antapical plates were recognised.

REMARKS: This new species is distinguished from previously described species of the genus by the presence of a cingulum and poorly developed tabulation.

These specimens are similar to H. chlamydata Cookson & Eisenack, (1952) in having a granular endoblast, but it is impossible to compare the periblast since it is usually badly preserved or not preserved at all. The presence of a slight apical prominence, a cingulum and poorly developed tabulation distinguish this species from H. chlamydata.

TABLE 2 Range of the dimensions of Hexagonifera jurassica from the different levels of the Kimmeridgian Lower Middle Upper Kimmeridgian Kimmeridgian Kimmeridgian Periblast: Length 45-50. 70-102. 80-105u Length without apex 42-55 60-75. 55-70 Breadth 35-50u 58-90 50-90 Endoblast: Length 40-48u. 58-88. 65-90u. Length without apex 38-59. 55-75u. 45-05 Breadth 35-52 58-75u. 45-75v Measured specimens I2 29 19

Cyst-Family MUDERONGIACEAE Neale & Sarjeant, emend. Sarjeant & Downie, 1966 Genus MUDERONGIA Cookson & Eisenack, 1958 Muderongia simplex Alberti, 1961

Plate 15, figures 1-2

1961 Muderongia simplex Alberti, 12, pl. 2, figs 1-6, pl. 12, figs 1-2, table c. 1964 MM. simplex Alberti; Downie & Sarjeant, 134.

1964 WM. simplex Alberti; Eisenack, 525-6.

1966b M. simplex Alberti; G. & M. Deflandre, fiches 3249-50.

1967b M. simplex Alberti; Sarjeant, table 12.

DEscRIPTION: The cyst is flattened, bearing an apical, two lateral and two antapical horns. Its outline is almost rhombic, as a result of differential development of the horns. The lateral horns are short and rounded at their free end. The anta- pical horns are unequal in length; one of them is very short and blunt, the other is well developed. The endoblast lies close to the outer margin of the periblast, generally stretching out into the horns but not reaching their tips, so that lateral,

242 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

apical and antapical pericoels are present. The tabulation is not well indicated but lines on the epitract, observable only with difficulty, simulate plate boundaries. A narrow cingulum is present but poorly marked. Both periphragm and endo- phragm are transparent; the surface of the periblast is granular. A well-developed apical archaeopyle was observed in one of the specimens; the other specimens ob- served often have a split on the flank of the apical region, indicating that the archaeo- pyle has not developed fully.

FIGURED SPECIMENS: I.G.S. slides PK.128 and PK.129, sample WB 29, from H.M. Geological Survey Borehole, Warlingham, Surrey, at 2285 ft 7in. depth. Upper Kimmeridgian (Rotunda Zone).

Dimensions: Range of the observed specimens (II in number): overall length 78-110u, breadth 68—goy, endoblast length 65-88y, breadth 55-75u, overall length (without apex) 70, endoblast length (without apex) 48u. Alberti gave the following ranges from Cretaceous specimens of the species: overall length 68-175u, breadth 63-133u-

REMARKS: M. simplex has been recorded from Valanginian to Barremian of Poland, Bulgaria and Germany. Eleven specimens were recorded, for the first time from England and the Jurassic, all from one horizon, the Rotunda Zone of the Warlingham Borehole. These specimens are similar to the specimens figured by Alberti, the only difference being that the apical horn is not so long as the apical horn of previously recorded specimens and the notches at the ends of the lateral horns were not seen on the Kimmeridgian specimens. With its second, blunt antapical horn, M. simplex is similiar to the type species of the genus, M. mcwhaet, but it differs in its short and rounded-ended lateral horns, whereas those of the type species are long and curved, downwardly directed.

Cyst-Family NELSONIELLACEAE Eisenack, emend. Sarjeant & Downie, 1966 Genus SCRINIODINIUM Klement, 1957 Scriniodinium bicuneatum (Deflandre) Sarjeant, 1967a Plate 15, figure 4

1938 Palaeoperidinium bicuneatum Deflandre, 180, pl. 8, fig. 7. 1957 PP. bicuneatum Deflandre; Downie, 422, pl. 20, fig. 2, table r. 1964 P. bicuneatum Deflandre; Downie & Sarjeant, 137.

1964 PP. bicuneatum Deflandre; Eisenack, 591-2.

1964 FP. bicuneatum Deflandre; Sarjeant, table 2.

1967a Scriniodinium bicuneatum (Deflandre) ; Sarjeant, 248.

1967b S. bicuneatum (Deflandre) ; Sarjeant, table 11.

1970 S. bicuneatum (Deflandre) ; Gitmez, 308, pl. 5, fig. 5, table 4.

OBSERVED RANGE: Lower to Upper Kimmeridgian (Baylei to Pallasioides). Not yet recorded from the Elegans and Scitulus Zones.

TOTAL KNOWN RANGE: Oxfordian (prob. Cordatum)—Kimmeridgian (Rotunda).

FIGURED SPECIMENS: BM(NH) slide V.56370, sample MR 547, from the lower

FROM ENGLAND, SCOTLAND AND FRANCE 243

boundary of the Platynota Zone (Baylei Zone), west side of the Ravin d’Enfer, Crussol, France.

DIMENSIONS: Range: overall length 80-115u, breadth 65-100, endoblast length 75-90, breadth 62-83u. (Measured specimens 36 in number). Holotype dimen- sions, as given by Deflendre, are 1oou length, 65 breadth, well within the quoted range.

REMARKS: S. bicuneatum was originally recorded from the Oxfordian of Normandy; the species was based on a single specimen which was not well preserved. In 1957, it was, for the first time, observed in the English assemblages, when Downie recorded it from the Pectinatus Zone of Dorset. His specimens also were poorly preserved.

It is abundantly present in the Kimmeridgian samples examined from England (55 Specimens being recorded); but rare in Scottish and French assemblages (5 specimens from the Baylei Zone of France and only one from the Cymodoce Zone of Scotland). The preservation was generally not good. The presence or absence of an endoblast was not confirmed by Deflandre, nor later by Sarjeant, who re-examined the holotype, noted the general similarity to the members of Scrimiodinium and redesignated this species as S. bicuneatum: its presence can now be confirmed.

The authors consider that Deflandre’s figures are in an inverse orientation, with antapex uppermost. The figured specimen is shown in the correct orientation: the ventral tabulation is not discernible, but apical, pre- and postcingular plate series can be distinguished without difficulty.

Scriniodinium dictyotum Cookson & Eisenack, 1960a Plate 15, figures 5-7, Plate 16, figure 6; text-figure 27

1960a Scriniodinium dictyotum Cookson & Eisenack; 248-9, pl. 37, figures 8, 9. 1962a S. dictyotum Cookson & Eisenack: Sarjeant, 262, pl. 1, fig. 9, tabs 3-4. 1962b S. dictyotum Cookson & Ejisenack: Sarjeant, pl. 69, fig. 11.

1964 S. dictyotum Cookson & Eisenack: Downie & Sarjeant, 145.

1964 S. dictyotum Cookson & Eisenack: Eisenack, 755.

1964a S. dictyotum Cookson & Eisenack: Sarjeant, table 2.

1967 S. dictyotum Cookson & Eisenack: Sarjeant, table 11.

1968 S. dictyotum Cookson & Eisenack: Sarjeant, 236, pl. 1, fig. 7, table 2b.

1970 S. dictyotum Cookson & Eisenack: Gitmez, 310.

1970 S. dictyotum subsp. dictyotum Cookson & Eisenack: Gitmez, 310.

1970 S. dictyotum subsp. osmingtonensis Gitmez, 310-11, pl. 1, fig. 3, pl. 8, fig. 12.

1970 S. dictyotum subsp. papillatum Gitmez, 311, pl. 9, fig. 11.

1970 S. dictyotum subsp. pyvum Gitmez, 311-13, pl. 13, figs 1-2, pl. Io, figs 1: Text-fig. 33a—b.

Remarks: A text-figure clarifying the differences (in form of the apex) between the four subspecies distinguished by Gitmez (1970) is here presented. The typical subspecies, S. dictyotum dictyotum, was not encountered in the basal Kimmeridgian, although recorded by Sarjeant (1962a, 1962b, 1964a, 1967b) from the Oxfordian of England and France: four specimens were, however, obtained from the Pectinatus Zone (specimen ED 240) of Dorset. The other subspecies were found only in the Lower Kimmeridgian: observed ranges:

244 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

S. dictyotum osmingtonensis Baylei only: Dorset. S. dictyotum papillatum Baylei to Mutabilis: Dorset and France. S. dictyotum pyrum Baylei only: Dorset and France.

d / S.dictyotum subsp. papillatum

S. dictyotum subsp. osmingtonensis alin : S.dictyotum subsp. dictyotum S. dictyotum subsp. pyrum

Fic. 27. Scriniodinium dictyotum (Cookson & Eisenack). Diagram showing the variation in the character of the apex in the four subspecies of the above.

Scriniodinium sp. Plate 13, figure 4

Description: The periblast is broadly ovoidal, with a hollow apical horn and rounded antapex. The surface of the periblast is smooth. The endoblast is sub- spherical to elongate, with rounded ends; its surface is smooth or minutely granular. Boundaries of reflected plates were indicated very feebly or not at all; in consequence, the tabulation could not be determined. The cingulum is relatively narrow and slightly helicoid, laevorotatory. A broad sulcus could be distinguished on some of the specimens observed. No archaeopyle was seen.

FIGURED SPECIMEN: I.G.S. slide PK.107, sample WB 7 from H.M. Geological Survey Borehole, Warlingham, Surrey, at 2834 ft 7in. depth. Lower Kimmeridgian (Eudoxus Zone).

DIMENSIONS: Figured specimen: periblast length 105y, breadth 76y, endoblast length 80u, breadth 7ou. Range: overall length 53-105u, breadth 42-76y, endo- blast length 42-80u, breadth 35~7ou. Measured specimens 5 in number.

OBSERVED RANGE: Lower to Upper Kimmeridgian (Mutabilis to Pectinatus).

REMARKS: Six specimens [two from the Lower Mutabilis, two from the Middle (Scitulus) and two from the Upper (Pectinatus) Kimmeridgian], all from English assemblages, were recorded. In general appearance, they are similar to Dingodinium curopaeum, which was recorded from the Aptian of Germany by Eisenack (1958c). However, these specimens are larger; Eisenack did not record specimens over 65-67u length. Since there is a very considerable stratigraphic gap and also a difference in dimensions, and in the absence of data regarding the archaeopyle, they were not placed in D. ewropaeum but were assigned to the genus Scriniodium.

FROM ENGLAND, SCOTLAND AND FRANCE 245

Genus SIRMIODINIUM Alberti, 1961 Sirmiodinium grossi Alberti, 1961 Plate 16, figures 7-8

1961 Sivmiodinium grossi Alberti, 22, pl. 7, figs 5-7, pl. 12, fig. 5; table c. 1964 S. gvossit Alberti; Downie & Sarjeant, 145.

1965 S. gvossi Alberti; G. & M. Deflandre, fiches 2787-2788.

1966 S. gvossi Alberti; Sarjeant, p. 212, pl. 22, fig. 7, table 5.

1967b S. gvossi Alberti; Sarjeant, table 11.

DeEscripTION: The cyst is dorso-ventrally flattened. The periblast is roughly pentagonal in shape, with a blunt apical horn and flattened antapex. The endo- blast is subspherical to ovoidal, both ends rounded. The circular cingulum is very deep and divides the cyst unequally ; the epitract is smaller than the hypotract. The tabulation is poorly indicated; on the dorsal side of the cyst, two apical, three pre- cingular and three postcingular plates were recognised, but the ventral tabulation was not established. The surface of the endoblast is apparently smooth, the surface of the periblast minutely granular. In one of the five specimens observed, an apical archaeopyle was observed, formed by loss of the whole apex; in the other specimens, the apex is still attached to the shell in the position of the first apical plate and a median dorsal, precingular plate (possibly 3’’) is also surrounded by splits, suggesting © that a combination archaeopyle of an undescribed type is developed. In all the observed specimens there is a posterior dorsal aperture of circular shape.

FIGURED SPECIMENS: BM(NH) slide V.56373(1) and V.56374(1), sample CP 245, from the Rotunda Nodule Bed, Chapmans Pool, Dorset. Upper Kimmeridgian (Rotunda Zone).

DIMENSIONS: Range: overall length 65-80u, length without apex 48y, breadth 48-70u, endoblast length 59-70u, without apex 43u, breadth 40-60n. Measured specimens 5 in number. Holotype: overall length gtu, breadth 86u, endoblast length 72y, breadth 61, as given by Alberti.

OBSERVED RANGE: Upper Kimmeridgian (Rotunda Zone). TOTAL KNOWN RANGE: Upper Kimmeridgian (Rotunda Zone) to Upper Barremian.

REMARKS: Five specimens were observed, all from the same horizon in Dorset. These specimens are characterized by their archaeopyle formation and their posterior dorsal aperture. Alberti (1961) recorded the holotype from the Upper Hauterivian to Upper Barremian of Germany, Sarjeant (1966) encountered it also in the Lower Hauterivian. Its presence also in the Upper Jurassic suggests a direct relationship with Scriniodinium, from which it may well have evolved by enlargement of the archaeopyle.

Cavate cyst sp. indet A Plate 16, figure 3

DEscRIPTION: The periblast is broadly ovoidal, with a well developed anterior

246 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

=

pericoel. The endoblast is spherical. There are no apical or antapical projections; tabulations is not indicated. The relatively narrow cingulum divides the cyst unequally; the epitract is longer than the hypotract. The surface of the periblast is perforated; the endoblast has a granular surface. An archaeopyle was not observed.

FIGURED SPECIMEN: I.G.S. slide PK.102C, sample WB 2, from H.M. Geological Survey Borehole, Warlingham, Surrey, at 2959 ft 5in. depth. Lower Kimmeridgian (Mutabilis Zone).

Dimensions: Figured specimen: periblast length 63, breadth 52u, endoblast length 48y, breadth 4Qu.

RemMARKS: This observation was based on a single specimen. In its general shape, it looks similar to members of Deflandrea and Hexagonifera, but an archaeopyle was not observed. It may represent a new genus, but before this can be decided, better preserved specimens must be awaited.

Cavate cyst sp. indet. B Plate 16, figures 2, 4

DESCRIPTION: The cyst is broadly ovoidal to elongate in shape, bearing a short, blunt, hollow apical horn, and one or two antapical horns. The endoblast is sub- spherical to spherical, with rounded ends. Tabulation and sulcus are not indicated. The broad cingulum divides the cyst unequally: the epitract is longer than the hypotract. The periphragm is densely perforated. The endophragm is smooth or minutely granular. An irregular breakage on the epitract, suggesting the beginning of opening of an apical or intercalary archaeopyle, was observed in some of the specimens.

FIGURED SPECIMENS: I.G.S. slide PK.104, sample WB 4, from Warlingham Borehole at 2910 ft 6 in. depth; and I.G.S. slide PK 106, sample WB 5, from the Borehole at 2885 ft 1 in. depth. Lower Kimmeridgian (Eudoxus Zone).

DIMENSIONS: Range (22 specimens were measured) : overall length 45-70u, breadth 33-60u, endoblast length 33-53y, breadth 31-5ou.

REMARKS: A group of specimens (44 in number), all from the Lower Kimmeridgian (Mutabilis to Pectinatus) of the Warlingham Borehole, were recorded which resemble, in their general appearance (with apical and antapical projections) the species of the genus Deflandrea. However, since no regular archaeopyle formation was observed, they are not attributed to that genus: they may indeed well be representatives of a new genus. The cavate cyst sp. indet. A (previously mentioned) shows similarities to these specimens, in their perforated periphragm and similar overall appearance, but has no projections at the apex and the antapex.

FROM ENGLAND, SCOTLAND AND FRANCE 247

INCERTAE SEDIS Group ACRITARCHA Evitt, 1963 Subgroup ACANTHOMORPHITAE Downie, Evitt & Sarjeant, 1963 Genus MICRHYSTRIDIUM Deflandre, emend. Sarjeant, 1967c

Micrhystridium recurvatum Valensi, 1953 Plate 17, figures 1-2

1953 Micrhystridium vecurvatum Valensi, 43, pl. 6, figs 1-4, pl. Io, fig. ro. 1955 WM. vecurvatum Valensi; Valensi, 589, pl. 1, fig. Io.

1960c M. vecurvatum Valensi; Sarjeant, 392, pl. 14, fig. 19, text-fig. 1a, table 2. 1962b M. rvecurvatum Valensi; Sarjeant, 489, text-figs 8b, f, tables 2-3. 1963 MM. vecurvatum Valensi; Wall & Downie, 778.

1964 M. vecurvvatum Valensi; Downie & Sarjeant, 133.

1964 MM. vecurvatum Valensi; Sarjeant, table 4.

1964 IM. vecuvvatum Valensi; Gocht, 123, pl. 16, fig. 13, text-fig. 43. 1965b M. vecurvatum Valensi; G. & M. Deflandre, fiches 2346-2351.

1965 MM. vecurvatum Valensi; Sarjeant, 177-178, pl. 1, figs 11-18, table 1. 1967 M. vecuvvatum Valensi; Dodekova, 27, pl. 3, fig. 10, table 1.

1967c M. vecuvvatum Valensi; Sarjeant, pl. 1, figs 1, 3-5, 9, text-fig. 1H. 1968 MM. vecurvatum Valensi; Sarjeant, table 2A.

DEscRIPTION: The cyst is spherical to subspherical, bearing simple, hollow, distally closed, curved processes, about 32-38 in number. The surface of the cyst is smooth or very finely granular.

FIGURED SPECIMEN: I.G.S. slide PK.127, sample WB 26, from H.M. Geological Survey Borehole, Warlingham, Surrey, at 2359 ft gin. depth. Upper Kimmeridgian (Pectinatus Zone).

Dimensions: Figured specimen: diameter Iou, process length 4u. Range: dia- meter 13-20u, process length 3-124, measured specimens 29 in number. Valensi gave the holotype diameter as Ion, Sarjeant gave the mean diameters of the speci- mens from Normandy as 14y. In contrast, Dodekova gave the average diameters of her Kimmeridgian specimens from Bulgaria as 22u; these specimens appear well outside the normal size range and may well represent a distinct species.

OBSERVED RANGE: Lower to Upper Kimmeridgian (Cymodoce to Rotunda). Not yet observed from the Elegans and Scitulus Zones.

TOTAL KNOWN RANGE: Bajocian to Upper Kimmeridgian (Rotunda Zone).

Micrhystridium sp. Plate 17, figures 7-8 1970 Micrhystridium inconspicuum Gitmez, pl. 1, fig. 8, table 4.

DescripTion: A form of Micrhystridium having a spherical shell, with thick shell wall (c. In). Processes are simple, conical, about 30 in number, and slightly curved.

248 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

The length of the processes is not more than one quarter of the shell diameter; they are hollow, distally closed, their cavity connected to the shell interior. The surface of the shellis smooth. An opening in the form of a split was observed in the single specimen seen.

FIGURED SPECIMEN: BM(NH) slide V.53953(1), sample OM 418, from 10 ft above the Ringstead Coral Bed, Osmington Mills, Dorset. Lower Kimmeridgian (Baylei Zone).

DIMENSIONS: Figured specimen: shell diameter Ig x 20u, process length 4—5y.

REMARKS: A single specimen recorded from the Baylei Zone of Dorset, was earlier placed into M. inconspicuum (Gitmez, 1970); subsequently, examination by high power objective has shown that it is different from M. inconspicuum as recently redefined by Deflandre and Sarjeant (1970).

With its short processes, this specimen is similar to the Oxfordian species M. vavrispinum Sarjeant (1960c). However, its cyst diameter is greater than that of M. varispinum and its processes are slightly longer and more numerous (Sarjeant gave the cyst diameter of the latter species as IIp, process length 2-2-5 and the process number as 14-20). It may represent a new species, but more specimens must be awaited.

Genus SOLISPHAERIDIUM Staplin, Jansonius & Pocock, 1965 emend. Sarjeant, 1968b Solisphaeridium claviculorum (Deflandre) Sarjeant, 1968b Plate 17, figures 9-10

1938e Hystrichosphaeridium claviculorum Deflandre, 191-2, pl. 10, fig. 4.

1963 Baltisphaeridium claviculorum (Deflandre); Downie & Sarjeant, 9r.

1964 B. claviculorum (Deflandre); Downie & Sarjeant, 88

1964 8B. claviculoyrum (Deflandre) ; Sarjeant, table 3.

1966 B. claviculovrum (Deflandie); Davey, Downie, Sarjeant & Williams, 174.

1968 ?Solisphaeridium claviculorum (Deflandre) ; Sarjeant, 233, pl. 2, figs 13, 15, table 2A. 1970 S. claviculovrum (Deflandre); Deflandre & Sarjeant, 6, pl. 1, fig. 5.

DESCRIPTION: Cyst spheroidal, moderately thin-walled, bearing about 22 processes. These processes are simple and straight, their length about four-fifths of the cyst diameter; they are closed at both the distal and the proximal ends but contain an elongate cavity which does not connect to the central cavity of the shell. The surface of the cyst is finely granular. An opening, in the form of a split, was observed.

FIGURED SPECIMEN: I.G.S. slide PK.101, sample WB1, from H.M. Geological Survey Borehole, Warlingham, Surrey, at 2984 ft 7in. depth. Lower Kimmeridgian (Mutabilis Zone).

DIMENSIONS: Shell 18 x 20, process length 17u. Holotype: overall diameter 58u, process length 14-16y (as given by Deflandre).

Remarks: A single specimen of this species was originally recorded from the Upper Jurassic of France and attributed to the genus Hystrichosphaenidium. In 1963,

FROM ENGLAND, SCOTLAND AND FRANCE 249

Downie and Sarjeant transferred it to the genus Baltisphaeridium on the basis of its spine characters. Recently, one of the authors (W.A.S.S., 1968b), after re-examina- tion of the holotype and recording further specimens from the Lower Oxfordian, placed this species in the genus Solisphaeridium hesitantly, because of the absence of clear knowledge of the mode of archaeopyle formation. Subsequent study of the holotype has since confirmed his judgement (Deflandre and Sarjeant, 1970).

A single specimen was observed in the Warlingham Borehole sample from the Mutabilis Zone which conforms in morphology to this species; since only one speci- men was found, the presence of the species at this level, after such a stratigraphic hiatus, cannot be regarded as definite since pollution cannot be altogether ruled out.

Subgroup NETROMORPHITAE Downie, Evitt & Sarjeant, 1963 Organism A Plate 16, figure 1; plate 17, figure 3 1970 Organism A Gitmez, 321, pl. 11, figure 9, table 4.

DEscriPTION: The cyst is ellipsoidal to elongate. One pole is rounded, the other is flattened: the lateral walls are slightly outbowed. The cyst wall is thick (about Iu), without ornamentation, processes or division into fields. The surface is smooth but porate; distribution of the pores is irregular, being generally densest around the flattened pole and on the sides of the cyst. An opening was observed in the flattened pole: its outline appears to be roughly circular.

FIGURED SPECIMEN: BM(NH) slide V.53948(3), sample RB 219, from the Rhactor- hynchia inconstans Bed, Ringstead Bay, Dorset, Lower Kimmeridgian (Baylei Zone).

DIMENSIONS: Figured specimen: cyst length 77y, breadth 28u. Range: length 42-77, breadth 18-48. Measured specimens 12 in number.

REMARKS: This new form was observed in the assemblages from the Baylei and Mutabilis Zones of Dorset and Le Havre only. It resembles in general outline mem- bers of the genus Palaeostomocystis, especially the species P. laevigata Drugg, 1967 (Upper Cretaceous of California): but none of the specimens observed contains an internal cyst or sac-like body.

Subgroup PTEROMORPHITAE Downie, Evitt & Sarjeant, 1963 Genus PTEROSPERMOPSIS W. Wetzel, 1952 Pterospermopsis harti Sarjeant, 1960c Plate 17, figure 6

1960c Ptervospermopsis harti Sarjeant, 402-3, pl. 14, fig. 16, text-fig. 3, table 2. 1962b P. harti Sarjeant; Sarjeant, table 3. 1964 PP. harti Sarjeant; Downie & Sarjeant, 143.

FIGURED SPECIMEN: I.G.S. slide PK.111, sample WB 8, from H.M. Geological Survey Borehole, at 2810 ft 6 in. depth. Lower Kimmeridgian (Eudoxus Zone).

250 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

DIMENSIONS: Figured specimen: overall 30 x 30u, cyst 12 X I2u. Overall dimensions of the other English specimen 32 x 32u, cyst 17 x 17u. French specimen: overall 30 x 30y, cyst I2 X I2u.

OBSERVED RANGE: Lower to Upper Kimmeridgian (Eudoxus to Pectinatus). (See below.) Total Known Range: Upper Oxfordian (Pseudocordata) to Upper Kimmeridgian (Pectinatus).

REMARKS: This species was originally recorded from the Corallian of Yorkshire. It is veryrare in the Kimmeridgian, only threespecimens being recorded, one eachfrom France (Mutabilis), Warlingham Borehole (Eudoxus) and Dorset (Pectinatus Zone).

Subgroup UNCERTAIN Acritarch sp. indet. Plate 17, figures 4-5

DEscRIPTION: The cyst is elongate, ovoidal with rounded ends, bearing about I4-16 processes. These processes are simple, straight or slightly curved, conical and closed distally. The shell wall is composed of two layers, the outer layer forming the processes; the processes are hollow but their cavities are not in contact with the central cavity. The surface of the cyst is smooth or minutely granular. An irregular opening was observed.

FIGURED SPECIMENS: I.G.S. slides PK.125 and PK.126, sample WM 26, from H.M. Geological Survey Borehole, Warlingham, Surrey, at 2359 ft 9 in. depth. Upper Kimmeridgian (Pectinatus Zone).

DIMENSIONS: Range: Shell length 15-35y, breadth 8-15, process length 6—1oy. Measured specimens 5 in number.

REMARKS: All the specimens were recorded from the same horizon in the Warling- ham Borehole (Pectinatus Zone). They differ from described species in their elongate shape and irregular opening. They may represent a new species and perhaps even a new genus.

IV. STRATIGRAPHICAL DISTRIBUTION CHARTS See Tables 3 to 6.

V. CONCLUSIONS

In the Kimmeridgian samples from England, Scotland and France described by Gitmez (1970) and herein, 36 genera, comprising 110 species, of dinoflagellate cysts and 6 genera, comprising 16 species, of acritarchs were recognised. From these, two new genera, 23 new species and 4 new varieties have been proposed: more than a dozen other forms certainly represent new species but were not proposed as such, because of poor preservation or (more frequently) inadequate numerical repre- sentation. In addition, there was a further list of species, probably new, whose preservation, presentation or orientation entirely precluded their description: the microplankton of the Kimmeridgian Stage is thus much more rich and more varied than has hitherto been recognised.

TABLE 3, \

‘The numerical distribution of microfossils inthe Kimmeridgi: i Cambridgeshire and Scotland (Staffin Bay na Canney alan casemsblates fromm Dorase Op ordste)

DORSET

CAM-| OXFORD- SCOTLAND BS. SHIRE

Baylei Zone Autissio-

2 Peetii agence 4 2 ‘ectinatus as

4 *4 °

Zone 28 g

ensis Zone

+ Pallasioides Baylei EI he bylei 2 Ea: Zone

28 || Zone 5S

Zone

Wheatley- Hudlestoni =Pectin.

OM 131 OM 418 OM 419 OM 420 HD 191 RB 218 KD 227 CD 229 CH 231 RD 234 FD 236 ED 237 ED 240

LO 352 LO 353 LO 360 SS 625 SS 626 SS 627

Chytrocisphaeridia chytrocides: ar 17 % 16 - ‘ Teagieee «1g NCD 2” Pa agra € mantelli ; 3 ee: $ e

C. pococki F

Chee Maecamonitd Siena EAGT gees dler es - ewer aie to en tice |e ak Tenua capitate “4 * 2 . q : 5 ci 7 2 : . . - - S 1. echinata Owe ee a 3 +p ills 6

T. hystnx it, a Bs EE wo Ke ig} eS Wat wees ere oilltallreecn oiling oc T. pilosa an ca te . : 5. Se 4

tsp. 3 é ‘Acanthaulax venusta . Or ec ) oe ae . . A op 6 8 ao 5 a tS Cry tarchaeodinium calcaratum { 5 y : 4 * , . . . : . - : (A t |} 5 C cf. calcaratum, a ee Pee! co p 8 Se gk OG 5 : Gonyaulacysta aculeata. ‘ . . . . . . 9 fs e ; 5 : 2 K 2 t

G. angulasa : . . . : : i “

G, cauda

ype +

cf, belicoides i cies a : : Is oC wo oe 2 ian So Ta ag hyaloderma ee ee a ee ee tee AAI colli Nl See et * jorassica 3) 13) (6) 20, (kao 163)» 1 is Oe ee, “ee allio eee es ctr ey jurassica var. longicornis ee es oo 1 5 ae 5 3 as, jae

Jongicornis é 5 .

5. ef, mamillifera Be hee rite oop 6 o : oop fied H Sh biat ie 5. nuciformis ye Ey id ae Uh oes aU OG «SG Sows m SS A MD 4 ee (ik eg 5. perforans Re od er ee as oes ee afr acre enn te wee II aie

¥. serrata, 1 . - . . 2 a a ° FE 6 ' 5. sp. A

5. sp. D

5. sp. E

sp.G a ee i oO 2 : 5. Sp. H . . . . . - . 8 . . 2 Leptodinium aceras 4 A F c - : fi “ L. amabilis ee cohen “oe op eg >, gag aeons a eee) L. arcuatum 2 . . 1 . . z ‘ - . * -

L. ef. crassinervum . . . : a . . = - . - . - . . = : 4 L. egemenil im © % : 5 é 6 Wee we 1. ef, subtile fc Paice Te ee G8 figs. AS L-sp. een he se 8G 5 5) Ge 2 - A a mw 6 Occisucysta balios ©. monoheuriskos fon 6

©. sp. (Gitmez, 1970) Seg oF er te 4 ee | gon OG Pee oan BNI Wo oS Oe Belodinium dysculum + ee poo 23 3) ea oS ee ee eae cee Se HES NHI we see = A a Dictyopyxis areolata 2 : Flee foe cay ge 0 ee et Hie gee eee com Gu \

D. cf. reticulata . . : . - . . . . . t . , . 5 . : a o © Histiophora cf. ornata 6 “ SR OR eR ME RD ps Meiourogonyaulax dicryptos aac ec: f io) bo hs F M. pila 6 5 E 5 : 5 A o ‘ fd 5 3 M. stafinensis Salou : =e : 3 M. sp. A (Gitmez, 1970) M. sp. B, herein. a te Os oe tf Apteodinium granulatum ie ie ee sy eek a i A. cf, maculatum, ea eee mc ea. rc? | t r 93 aly pk Ree Serres, ||(hisy Saar Pareodinia ceratophora Aa | a Eo a a owe 2 a) | 20), ae eee ava, ‘Trichodinium sp. Re ee! Mc oes ee aie wor oo eos a ee Al) | Se rete Imbatodinium antennatum 1 eS ae 2 1. cf. villosum Be OE ai ge oe Cee ene eect res o>) |(| Ke memo | We on 8 1

w wo ——

Nannoceratopsis pellucida, yo ty ah tee at 1 Egmontodinium polyplacophorum, a eo Ek Ss cat Proximate cyst sp. indet. Hela ieee Ro ee oe Heslertonia pellucida es in Wy ose aes 0 oe ae Epiplosphaera reticulospinosa tt ae eee eS ee ae oe i 8 Allis ao = yea OS Adnatosphaeridium paucispinum ee na ee cee ee ee eo : Sp gece 0 tee Cleistosphaeridium ehrenbergt 6 & Jan ot 2 : Se : 4 oe

C. polyacanthum 1 ;

C. polytrichum a Sanna €. tribuliferum 4

C. sp. (Gitmez, 1970) C. sp., herein. . = > : q : : Hystrichosphaeridium petilum 5 5 Dass x. a > eee Oligosphaeridium pulcherrimum A 2 - Polystophanephorus sarjean|

Prolixosphacridium ef, deirense

P. granulosum

P. parvispinum

Systematophora arcolata

S. orbifera

5. ovata

‘Taeniophora iunctispina é ‘ Stephanelytron cf. scarburghense 4 2 : ae Ry AS Ue Xa pay es : Endoserinium cf. campanula Fo Rr ad ; Be PSA TEE °

E. galeritum F. luridum A : ~ 5 Co ee

E. oxfordianum + + “ : 0 cS 5 5 5 - = 5 * : - Psaligonyaulax apaleta a Ae co é ‘ : : : Ss S P. sp, (Gitmez, 1970) i 3 4 . 5 5 . - . . * a ire : ; ; Hexagonifera jurassica : e ci 5 5 D 8 eo ea Faryocavatus tuberous o * : . . . : 4 5 Scriniodinium bicuneatum . 5 . : = . . ot as

S. crystallinum S 4 n 4 - : |

5. dictyotum subsp, dictyotum i " c a . x 5 . . .

5, dictyotum subsp, osmingtonensis S, dictyotum subsp, papillatum

Geet) ae w

7

fee || 2 ||

funn © Bn

w

wu

microplankton

raminiferal shell linings

TABLE 4

The ni i i je numerical distribution of microfossils in the Kimmeridgian assemblages from the Warlingham Borehole, Su rey

Mutabilis Zone

Scitulus Wheatleyensis Hudlestoni Zone Zone Zone

Pectinatus Rotunda Zone Zone

rensis

ar —o Zone 2 Ie

WB WB WB 3 WB 4 WB 5 WB 6 WB WB § WB 9 WB 10 WB 1 WB i WB 13 WB 14 WB 15 WB 16 WB 17 WB 18 WB 19 WB 21 WB 22 WB 23 WB 24 WB 25 WB 26 WB 27 WB 28

Chytrocisphaeridia chytrocides sh) om Ze <pl ey aan ay C. mantelli Sf ie a pococki

Cramea warlinghamensis ee lee ; Nae B58

Tenua capitata 4 : Dp eo :

‘7. echinata ‘ . : if Be mS ry c

T. hystrix Jb SO ae : j 1 50 chet i

7. pilosa : 5 3 us

Gonyaulacysta angulosa : F 7 Ps 5 2 é a 3

G, cladophora ‘ . : , ‘ x H : , 6 4 % . 5

ehrenbergii ¢ ; am oe 5

ef, givseppei . 3 a

_ globata “os

granulata 1 i

, granuligera eG : F ;

jurassica ! t 2

jurnssica var. longicornis : F I

longicornis Tia ar

"cf, mamillifera ary

nuciformis 6 : mas

perforans :

sp. B , ead r

sp. C : 2 ip ar ce er

G. sp. E he 8 5 ;

Leptodinium aceras 5 Py

L. arcuatum .

L. sp. : A : S 1

Occisucysta balios 2 TS :

Histiophora cf. ornata

Meiourogonyaulax pila . .

M. staffinensis : ; I

M.sp., herein Te

‘Apteodinium granulatum ‘ f . t

A-cf. maculatum : é 1 :

Pareodinia ceratophora 5 uO me A ey 1 ri 7

Imbatodinium antennatum 5 : 5 1 i

Proximate cyst sp. indet, :

Cleistosphaeridium ehrenbergi a gu

C, polyacanthum 5 . 1

© tibuliferam i my s M _ 5 : -

C. sp,, herein Z 5 rae 1 we 1 ‘ ie | Oey

Hystrichosphaeridium petilum : St siete hina ae: ae :

Prolixosphaeridium granulosum : 1 a 5 ee rome 1

P. parvispinum : : : 1

Systematophora areolata 23 4 5 G 4 6 5

S orbifera - Ze ‘ ; u 7 a c ) G4 a i Yin 2 é a

Endoscrinium Turidum Sa Pe ef a5 > Pees :

E. oxfordianum > 5 rs . = 2 2 2 : . 6 A :

Psaligonyaulax apaleta 1 3 Beas I “ ‘ E eS Tes

Hexagonifera jurassica fs 5 Tes 3 5 pas 1 o tb a oe oe a

Parvocavatus tuberosus c 1 : ry . . . . 1

Muderongia simplex E - Pe ;

Scriniodinium bicuneatum , tee : ankS Pee cy ra Oe ahs 7

crystallinum, 1

dictyotum subsp. papillatum ; 1 ; Pre a5 2 ee eee ee eee ee

playfordi z 5 sii 1 any é [ie

S. sp. 1 5 6 1 I Mer 8 1

Netrelytron parum

Cayate cyst sp. indet. A

Cavate cyst sp. indet. B F

Micthystridium fragile 66 2 7

2 e 7B & 3

° ES Ges nanid ry

G. G. G. G. G G. G

o9900 ee

w

w rear

wR : I ya M. inconspicuum

M. recuryatum 5 x ; M. sydus A i 1 é ee Oe as eae as j 5 I Solisphaeridium claviculoram og = : : : S. stimuliferum a We Oh Tk - ee : : Veryhachium hyalodermum 3 1 : : ap stig et 5 5 ok oy eo Pterospermopais australiensis 2 : 2 pees ; 3 é tee ae P, harti , : ww inad 2: 1 Acritarch sp, indet. ears aes thaw " : oar am © oc Total number of determined microplankton 158 93 84 101 119 95 6x 40 43 81 47 32 tor 72 88 112 52 49 87 65 10 24 56 47 42 97 80 39

Ria muctie © oe 2 r

re)

Indeterminate microfossils 40 15 25 16 17 7 48 6Y 57 18 22 17 108 33 117 45 106 50 118 153 20 32 53 34 7 201 27 27 Foraminiferal shell linings Gy Rate eae) Ber XO: Ci BRP eye ie ee eae eee GREE S eae Ate st ok) Naw ea BS One Pollen & Spores 88 60 94 94 of 83 75 43 58 83 81 84 72 70 67 80 58 83 75 62 82 82 77 77 93 92 92 94 Wood fragments ea 3 Toeegvenxdeeyes 2 31 os) i he a Ta Chas 7 he ke ¢ I

185 290 286 113 145 188 172 143 209 200 161

Torat 291 280 204 211 227 189 190 172 176 193 151 132 281 176 274

-

wine \ Sy

TABLE 5

‘The numerical distribution of microfossils in the Kimmeridgian assemblages from France

Baylei Zone Cymodoce Autissiodorensis

Zone Mutabilis Zone

Zone

Boulonnais J ura Mont Mountains, Crussol

Normandy

M. Crussol

Normandy

Lorraine oe Gz es 2

Le Havre

CC 447 CE 448 CC 449 CC 450 CC 451 CC 452 CC 455

184

VN HE 185 HE 186 MR 547 BN 179 MR 548 MR 550 HF 395 OF 485 OF 486 OF 487 MR 552 MR 553

MR 554

MR 555

Zone Palla Zone

Scitulus Jura M.

Boulon,

MV 489

Chytroeisphacridia chytroeides C mantelli

C. pococki D 1 Tenua capitata f : T. echinata Coa at 5 A 1 3

T. hystnx v ES 5 r 2 1 qq G aes 3 eB 7 T. pilosa : Taree, WE ; Acanthavlax venusta . : ‘ 1 a és

Cryptarchacodinium sp : 1 :

Gonyaulacyata aculeata A 1 t

angulosa . 1 1 1 1

cauda 1

cladophora 1 ‘ , 5

ehrenbergit 2 1 ” - 2 5. eisenacki

cf. eisenacki 3

cf. giuseppei 1 A 5 granulata bo Bie fs : ce 1 5 granuligera Tf ha ce) 6) Ge Ge ee 1

cf, helicoidea : t

hyaloderma f : 1 jurassica 2 4

jurassica var. longicornis

longicornis 6 3 cf, mamillifera 1 cae ae : nuciformis 1 + ck oe ip mo 2 4 >. serrata 2 1

>. systremmatos (ob

5. sp. A uk

Leptodinium amabilis 1

L. arcuatum * 1 . . 1 1

L. clathratum 4 1 : :

L. egemenii 5 9 3 3 L. ef. subtile ‘ ; 5 ae 1 Occisucysta balios 1 : 1 (ee “ Dictyopyxis arcolata ch 1 2 Z 5 1 D. cf. reticulata 5 1 ea 4 Histiophora ef. ornata Lite 1

Meiourogonyaulax dicryptos : = é ,

M. staffinensis 1 Oe Go 4 Lon a © 3 Apteodinium granulatum A. cf. maculatum. Pareodinia ceratophora Imbatodinium antennatum - 1 . 5

Heslertonia pellucida ses : ror 2 1 Epiplosphaera reticulospinosa Adnatosphaeridium paucispinum ‘ E 2 9 : Cleistesphaeridium ebrenbergi Cie i | Taner cr zo o ae a C. polyacanthum: A

C. tribuliferum. C Bhi hig 2 C. sp. (Gitmez, 1970) : ried % ey x A 3 Hystrichosphaeridium petilum hee 6 1 1 Oligosphaeridium pulcherrimum : - Polystephanephorus sarjeantii 1 Prolixosphaeridium granulosum. ce 5 : ee 5 moO 1 2 Systematophora areolata 75 5 (9 bey FER SH c 51 6 6 1 3 1 34 2 6 S. orbifera 33073) <5) 4 ay Ss 3

Taeniophora iunctispina hati eS yo | Keke & oo)

Stephanelytron redeliffense 5

S. cf. redcliffense

S, scarburghense anes ; 1 T Endoscrinium cf. campanula die eve fide 9)

galeritum 2 >

Juridum Hg E n t 2 : t oxfordianum : ‘ - oo an

E. sp. ‘ cn) : Tae Psaligonyaulax apaleta a ume oe a. Pirie E Se Hexagonifera jurassica 6 5 1 (es eres a e 4 eg Parvocavatus tuberosus I 22 1 5 rae 3 s 4 Scriniodinium bicuncatum 4 ain eee 1

S. crystallinum i 1 : clas a 5 17 5. dictyotum subsp. papillatum ef PG io tte 1

5. dictyotum subsp, pyrum c ol ms \ oo gee

S. cf, galeatum 3 Tee a 6 2) 6 : 5. playfordi ieee week oe oo yea es 6 ‘etrelytron parum < mee bY ia = 6 1 a N. atogastum cae ae c 3

Baltisphacridium inusitatum ee ty Micrhystridium fragile Ae 159 (ithe old . M. inconspicuum. Suen? OS eae orp Em 12 M. recurvatum Pry eter oo aces 2 17 = «8 5 re M. sydus ~ & 5 «© f © 2» » oe 1 eae Be SoGmes Solisphaeridium brevispinosum. Hote ig Ss, a, eo ; mc en dr on MS Side er ce Ae rs S. stimuliferum ip 4 OY 5) 2 | Es Organism A ste Soaeee 2 a ie ce Se Veryhachium hyalodermum Re a te welts 1 bp atuhe Mec E

Staplinium cistum ao Var wis, eet : ids hie Va 103 re Pterospermopais australionsix Poh ae ed eee? Mos moo dO Om Ce P, harti tie teat p : ee he) Ps bo © P. helios eee Ces ae é j wor oo 1S ce OU Gee He ‘Total number of determined microplankton 207 87 56 110 82 k Pe REY ge ek I gg Undetermined microfossils Ren ee ar 188 48 van a Hy s a Fe 87 42 ss 164) BOF GS, 7) 4a Foraminiferal shell linings Sonts® 3| a0) ar Re aa Ac eee ec oe oe Pollen & Spores daexG) es eateiaMMeeaMeoM 37) 20 54) 65) 224) ms 77) 38 Go 8 st? 9 17 13 Wood fragments . 38 u 2 5

Peat a w

10

pow

5 Perey

wo

ra:

74

w on Se

a Ss oe

B | CC 453 au

ToraL 322 235 141 329 146 204 114 31x g28 163 79 308 196 — = — 378 334 168 I0q 222 10 17 13

nes TABLE 6

‘Tho distribution of Kimmoridgian microplankton comparing with their Previously known stratigraphic range

Kimmoridgion Previously known stratigraphic range

Zone

Scitulas Zone Hudlestoni Zone Pectinatus Zone Rotunda Zone

Pallasioides Zone

Eudoxus Zone Mutabilis Zone Aut

Baylei Zone

[s

Lower Cretaceous (Berriasian-Valanginian)

Portlandian (Tithonian, Volgian)

Scotland France England

| England France England France England Fran England England England

England Kimmeridgian

* | France

% | Oxfordian

* * * x * x * * * x *

Chytroelsphaeridia chytroeides Sain

mantelli bs = a c ‘ ‘ . x Ki Sue fn ne Tes) Ved mcr BRry

pocoe jromes warlinghamensis

Tenvia capitata x T, echinat x x x

» %) Pre-Callovian

+ > #| Callovian

KKH

eK

eR KH *

Birt

KKM # x

* * x xe ne » * ne x* * x Pre * “

T. hystrix

T. pilosa

T. »p. ‘ : SuaeAL ze

Acanthaulax venusta ied ek: eet x Seas tps : oie ems

Cryptarchaeodinium calcaratom x ae - ee ed aS

C ef, calearatum ‘ x ‘ ‘ . . ’ . . ’ x

. sp. (Gitmez, 1970) : i - ets

Gonyaulacysta aculeata x

angulosa x x x

tal

cauda

eladophora ehrenbergii eisenacki x Aes “A. cane x x ef. eisenacki x x : 3 cf. giuseppel x x | a eee fie f x

wR KK Re “

*

x * Pe *

x x

G. ef helicoidea x hyaloderma x x

x

x

*

*

* wu KE HHH: nA

jurneaica x x Jorassica var. longicornis x x longicornis

ef, mamillifera

nuciformis x x

. 2 perforans ; ; «| foes 2p & : serrata Meter oS a . ee Pee ; : me SS Bs

* Ae

systremmatos

Leptodiniam aceras Pos ac eee 1 amatillis ee

* * 4 x a

* Ee oo

L cf, subtile x Sy x i : 5 6 x : x Lap. Se Soe aera x C c Occixueysta balios op eS 5 oS : ae aise ©. monoheuriakos ‘i Re 0 os bac 3 a 2 ©. sp. (Gitmez, 1970) x Belodinium dysculam yore. ea Dictyopyxis areolata 5 ots ft D. cf. reticulata Er Mistiophora ¢f. ornata x x) 55 ‘Melourogonyaulax dicryptos 5° oe (a) rs M. pila a og arte = M. staffinensis Fats ty ot? SEY cy OE x x Peareete eo) 0o M. sp. (Gitmez, 1970) Mops Jel ye a or: mes) et | M. sp. herein Se Geol we 3 - 8 = ob o Apteodinium granulatum me A. cf maculatum eee: ae Pareodinia ceratophora ‘Trichodinium sp. x Sate dl An Met 5 st Wd | Imbatodinium antennatum 3 5 ae sae A : 3 5 tee <0 ay 1. cf. villosum Se ee Te ORNS. Ses ‘ ae ry por et nt lds oe, Se Xe cree jannoceratopais pellucida x Sy 4a ae : i omc -

PI

x * +h Pio Kanne * * x “”

x

*

* x x ”

*

*

nH

nom

*

*

eet

*

»

”

xo x*

*

¥

KHnK x nn

* * * » x * » Py P) *

* x

gmontodinium polyplacophorum Sa mee ~ 8 oo Proximate cyst sp. indet - ; 2 2 - g . 5 ‘ : 5 Fe SS Heslertonia pellucida tA x 5 2. ¢ Epiplosphaera reticulospinosa Adnatosphaeridium pauelspinum Cleistosphaeridium ehrenbergi ©. polyacanthum © polytrichum © tribuliferam © 5p: (Gime; 1079)

* *

+ Wie x - Ke KW

x x

KKn nH *

herein

Tipurichosphaeriafom petilamn

Oligospbacridium poleherrimom

Polystephanephorus sarjeantil

Prolixosphacridium cf, deirense

P. granulosum

¥, parvispinum

‘Systematophora areolata

5S. orbifera

S. ovata

Taeniophora junctispina x

Stophanelytron redcliffense Aus ear x

x x x bie x Awe:

a“ Cale Wadia ite ee

KR RHR KKK: -»

* RK

ne KR RK

wo RK Ko KK * * * a Me * * » card wenn *

5. cf, redeliffense

S. cf, searburghense Endoscrinium ef, campanula E, galeritum

E. luridum

K. pear

* ca

"K RAKAAKA RA yw we

1 OLA eases *

PAI cay res ‘apalota iW ap, (Gitmex, 1970) nifera juramica

RATES eben Muderongia simplex Sesoel faa nium blewneatun

er

i. reas subsp. dictyotum: S, dictyotum subsp, ete a dictyotum subsp.

5, dictyotum

AM: K-

Br sc ah oe

PEWISESUSeCs (Wisse 5 2 o>

HHH KH HH He RHR HK Pegaso. 2

= A ee a

.- > Ee. SS

0 ec Ee S68 iae Serene reef pu «=

RR RH RK RR ooo gee Oe

+ ae crn

Biebloa

cont 1 00 EO 2505 a

PERIYEEM =) le a = RRS i

Ke MRA KH

ooh i ee

haltlophine rin

FROM ENGLAND, SCOTLAND AND FRANCE 251

The assemblages are dominated by proximate cysts, of which 68 species have been described. Gonyaulacysta is the dominant genus, represented by 29 species and some 800 specimens. The most abundant species is G. juvassica (301 specimens) ; this is, however, confined to the Lower Kimmeridgian, predominating in the Baylei and Cymodoce Zones and occurring only sparsely in the higher zones of the Lower Kimmeridgian. It has already been shown to be present in the Lower and Middle Callovian and abundant in the Lower Callovian and Oxfordian: its frequency makes it a useful stratigraphic index for these two stages and part-stage. G. nuciformis, represented by 184 specimens, ranges throughout the Kimmeridgian and has been previously recorded from Upper Callovian and Oxfordian: its presence or absence in the Portlandian remains to be elucidated. Three other species, G. longicornis, G. granulata and G. granuligera, are each represented by 50-55 specimens; G. granulata occurs in the uppermost Oxfordian also, the remaining two species being only recorded from the Kimmeridgian. The two latter species are most abundant in the Lower Kimmeridgian, though ranging throughout the stage. G. longicornis has a more uneven distribution: it is infrequent in the Lower Kimmeridgian, attains maximum frequency in the Middle, and was encountered only in the topmost (Pallasioides) zone of the Upper Kimmeridgian. G. perforans occurs in the Middle and Upper Kimmeridgian in sufficient numbers to serve as a useful index for those levels: other species with restricted ranges are numerically less frequent. Four other tabulate genera with precingular archaeopyles, Leptodinium, Acanthaulax, Cryptarcheodinium and Occisucysta, are also well represented.

Proximate cysts with apical archaeopyles (genera Tenua, Chytroeisphaeridia, Fromea, Egmontodinium, Dictyopyxis, Meiourogonyaulax) are much less important, Meiourogonyaulax ranges throughout the Kimmeridgian: the high-crested species M. staffinensis is especially conspicuous and a good index fossil. Dictyopyxis was not encountered higher than the Lower Kimmeridgian: the other genera appear less useful as stratigraphic indices. Two genera of proximate cysts with intercalary archaeopyles, Pareodinia and Imbatodinium, are present: the former is long-ranging, but the latter (known to date only from the Kimmeridgian) is potentially important. Proximochorate cysts were poorly represented by only two genera: the absence of Spiniferites is noteworthy and suggests that prior records from the Oxford and Kimmeridge Clays were a product of contamination.

Chorate and cavate cysts were present in comparable numbers (19 and 21 species Tespectively). Chorate cysts with complex process groupings (Systematophora, Polystephanephorus, Epiplosphaera and Taeniophora) are characteristic, the two latter genera being encountered only in the Lower Kimmeridgian. Stephanelytron, hitherto known only from the Oxfordian, ranges only as high as the Cymodoce Zone and appears an important index. The long-ranging genus Prolixosphaeridium may Tepay fuller study, since the circumscription of the species erected to date appears most unsatisfactory. Cleistosphaeridium is abundant throughout: in contrast, the two species with tubulai spines both had only a sparse representation.

Of the cavate cysts present, Endoscrinium and Scriniodinium (both abundant in the Oxfordian) dwindle in importance markedly after the Lower Kimmeridgian. Netrelytron, Psaligonyaulax and Hexagonifera range throughout the stage, the former

F

252 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

consistently in low numbers: Muderongia and Sirvmiodinium, characteristically Lower Cretaceous, enter only in the uppermost zones.

Acritarchs are only a minor constituent of these assemblages and consistently occur only in low numbers. Acanthomorphs are the most numerous acritarchs, especially the various species of Micrhystridium and Solispheridium stimuliferum: pteromorphs, prismatomorphs, netromorphs and polygonomorphs are also infre- quently present.

The stress of this study was placed on the lower zones of the Kimmeridgian and the horizons of the Middle and Upper Kimmeridgian require further study before a full stratigraphical picture can be painted. However, it is clear that there are three distinct phases: in the first phase (Baylei, Cymodoce and, in some measure, Mutabilis) the assemblages retain a characteristically Oxfordian allure: in the second, they have a distinct identity which may be termed characteristically Kimmeridgian: and in the third (Rotunda and Pallasioides) the assemblages begin to acquire Lower Cretaceous characters though still retaining a dominantly Upper Jurassic allure.

Thirty-eight species were observed to range throughout the Kimmeridgian: 26 further species were recorded both from the Lower and Upper and must be assumed to range through the Middle Kimmeridgian. Thirty seven species were recorded only from the Lower Kimmeridgian, seven others in the Lower and Middle Kim- meridgian. Five species occurred only in the Middle and Upper, four only in the Middle, nine only in the Upper Kimmeridgian. The imbalance in the number of characteristic species certainly reflects the imbalance in study. It is already possible to distinguish from the study of an assemblage, whether it is from the Lower, Middle or Upper Kimmeridgian: it should be possible, in the future, to determine that the sample came from a particular ammonite zone or from one of two adjacent zones.

Whether it will be possible in the future to distinguish, as well as to correlate between the assemblages from France, England and Scotland is less clear: the present authors consider that the similarities between the English and Scottish assemblages and those from northern France are so great as to make it virtually certain that these areas were all part of one plankton province in one water body (which agrees with what is known of Jurassic palaeogeography). Fuller information is needed on the assemblages from central France (Crussol and the Jura) before any meaningful commentary can be made, but it should be noted that surprisingly few of the species described from the latter locality by Deflandre (1939, 1941) were encountered in this study. Of 19 species described by Klement (1960) from the Kimmeridgian of South Germany, only seven were identified in the assemblages studied. It thus appears probable that these assemblages were drawn from a different plankton province, and hence a different water body. Future studies of these microfossils promise to aid in elucidating the pattern of water circulation in the Jurassic.

VI. ACKNOWLEDGEMENTS

The bulk of the work by the first author was carried out, under supervision by the second author, during her tenure of a research studentship awarded by the Scientific and Technical Research Council of Turkey. Samples were collected under

FROM ENGLAND, SCOTLAND AND FRANCE 253

guidance from, or provided by, a number of geologists listed in the “‘Introduction’”’, to all of whom the authors would like to express their thanks. Both authors, during their work at the University of Nottingham, have received considerable help from Mr. R. D. Hendry and his staff, especially Mr. John Eyett and Mr. K. J. Cass in photography ; they would also like to thank Professor the Lord Energlyn of Caerphilly for his support and encouragement. Mr. L. A. Riley is thanked for making some corrections to the stratigraphical conclusions.

VII. REFERENCES

Acer, D. V. & Evamy, B.D. 1963. The Geology of the Southern French Jura. Proc. Geol. Ass., London, 74, 3 : 325-356, figs 1-6.

Acer, D. V. & Ramsay, J. G. 1963. Summer Field Meeting in the French Jura and Alps, July 1963. Proc. Geol. Ass., London, 74, 4 : 483-515, figs 1-4.

Acer, D. V. & SmirH, W. E. 1965. The Coast of South Devon and Dorset between Brans- combe and Burton Bradstock. Geol. Ass. Guide, London, no. 23 : 1-21, figs 1-5.

Acer, D. V. & WaALLAcE, P. 1966a. The Environmental History of the Boulonnais. Pyroc.

Geol. Ass., London, 77, 4 : 385-418, figs 1-9.

1966b. Easter Field Meeting in the Boulonnais, France. Proc. Geol. Ass., London,

77, 4 : 419-436, pls. 19-21, figs 1-4.

ALBERTI, G. 1961. Zur Kenntnis mesozoischer und alttertidrer Dinoflagellaten und Hystrichosphaerideen von Nord- und Mitteldeutschland sowie einigen anderen europdischen Gebieten. Palaeontographica, Stuttgart Abt.A, 116 : 1-58, pls. 1-12, table 1.

AnpeErson, F. W. & DuNnHam, K. C. 1966. The Geology of Northern Skye. Mem. Geol. Surv. Scotland, Edinburgh, 216 : 1-216, pls. 1-4, figs 1-23.

ARKELL, W. J. 1933. The Jurassic System in Great Britain. University Press, Oxford.

681 pp., 47 pls., 97 figs.

1930. Geology of Oxfordshire, in L. F. Salzman (ed.). Victoria History of the County of

Oxford, vol. 1, University Press, Oxford. (1-26, 1 map.)

1941. Dorset Geology, 1930-1940. Pyoc. Dorset Nat. Hist. Archaeol. Soc., Dorchester,

61, 117.

1946. Standard of the European Jurassic. Bull. Geol. Soc. Am., New York, 57 : 1-34, tables 1-4.

1951. Dorset Geology, 1940-1950. Proc. Dorset Nat. Hist. Avchaeol. Soc., Dorchester, 72 : 176-194.

1956. Jurassic Geology of the World. Oliver & Boyd, Edinburgh. 1—806, 102 figs, 28 tabs.

Battes, N. 1963. Dinoflagellate si Hystrichosphaeride cretacice din Platforma moesica. Petrol si Gaze, Bucuresti, 12 : 581-597, pls. 1-8.

1965. Observatii asupra microflorei cretacice inferioare din zona R. Bicaz. Petrol si Gaze, Bucuresti, 16, 3-17, pls. 1-4, fig. 1.

1967a. The microflora of the Albian “‘Green Sands” in the Moesic Platform (Rumania). Rev. Palaeobot. Palynol., Amsterdam, 5, 183-197, pls. 1-4, figs 1-4.

1967b. Albian microplankton from the Moesic Platform, Rumania. Mucvopaleontology, New York, 131 : 1, 327-336.

1969. Discontinuitati in repartita microflorei cretacice din Platforma moesica. Petrol st Gaze, Bucuresti, 20, 332-338, figs 1-5.

254 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

Beju, D. 1969. Jurassic Microplankton from the Carpathian Foreland of Romania. A pre- print for the “Colloquium on the Mediterranean Jurassic, 1969”. I-12, pls. 1-8, figs 1-7.

Brito, I.M. 1967. Osacritarcha. Notes prelim. Estud. Serv. geol. min. Brasil No. 138, 1-21, pls I-II.

Brown, R.W. 1956. Composition of Scientific Words. Reese Press for the Author, Baltimore, Md., 1-882.

CLARKE, R. F. A. & VERDIER, J. P. 1967. An investigation of Microplankton Assemblages from the Chalk of the Isle of Wight, England. Verh. K. ned. Akad. Wet., Amsterdam, 24, 3 : 1-96, pls. 1-17, figs 1-32.

Cookson, I. C. & Etsenack, A. 1958. Microplankton from Australian and New Guinea

Upper Mesozoic sediments. Pvoc. R. Soc. Vict. Melbourne, 70, 1 : 19-79, pls. 1-12, t-figs

I-20.

1960a. Upper Mesozoic microplankton from Australia and New Guinea. Palaeontology, London, 2, 2 : 243-261, pls. 37-39, t-figs 1-4. 1960b. Microplankton from Australian Cretaceous sediments. Mucropaleontology, New

York, 6, 1 : 1-18, pls. 1-3, t-figs 1-6.

1961. Upper Cretaceous microplankton from the Belfast no. 4 bore, South-Western

Australia. Proc. R. Soc. Vict., Melbourne, 74, 1 : 69-76, pls. 11-12, t-figs I.

1962. Additional microplankton from Australian Cretaceous sediments. Muicropaleon-

tology, New York, 8, 4 : 485-507, pls. 1-5, t-figs 1-2.

Cooxson, I. C. & HucHEs, N. F. 1964. Microplankton from the Cambridge Greensand (mid- Cretaceous). Palaeontology, London 7, 1 : 37-59, pls. 5-11.

Cope, J. C. W. 1967. The Palaeontology and stratigraphy of the lower part of the Upper Kimmeridge Clay of Dorset. Bull. Br. Mus. nat. Hist. (Geol.), London, 15, 1 : 1-79, pls. 1-33, figs 1-22.

Craic, G. Y. 1965. The Geology of Scotland. Oliver & Boyd, Edinburgh, 1-556, fig 16.

Davey, R. J. 1969. Non-Calcareous Microplankton from the Cenomanian of England, Northern France and North America, pt. 1. Bull, By. Mus. nat. Hist. (Geol.), London, 17, 3 : 103-180, pls. 1-11, figs 1-16.

Davey, R. J., Downie, C., SARJEANT, W. A.S. & Wittiams,G.L. 1966. Studies on Mesozoic and Cainozoic Dinoflagellate cysts. Bull. By. Mus. nat. Hist. (Geol.), London. Suppl. 3 : 1-248, pls. 1-26, figs 1-64.

; ; ; 1969. Appendix to “Studies on Mesozoic and Cainozoic Dinoflagellate Cysts’. Bull. Br. Mus. nat. Hist. (Geol.), London. Appendix to suppl. 3, 1-24.

Davey, R. J. & Wittiams, G. L. 1966. See Davey, R. J. and others.

DEFLANDRE, G. 1938. Microplancton des mers jurassiques conservé dans les marnes de Villers-sur-Mer (Calvados). Tvav. Stn. zool. Wimereux., Paris, 13 : 147-200, pls. 5-11, t-figs I-10. 1939a. L’étude microscopique des silex. Sciences Naturelles, 1 : 259-264, t-figs 1-24. 1939b. Sur les dinoflagellés des schistes bitumineux d’Orbagnoux (Jura). Bull. Soc. Fy. Microsc., Paris, 8, 4 : 141-145, pl. 6. 1941a. La vie créatice des voches. Le véle bdtisseur des étves microscopique et la genése des houilles et des pétyéles. Presses Universitaires de France, Paris, Collection ““Que Sais-Je”’. 1-128, t-figs 1-25. 1941b. Le microplancton Kiméridgien d’Orbagnoux et l’origine des huiles sulfurées naturelles. Mém. Acad. Sci., Paris, no. 65, 1-32, pls. 1-7, t-figs 1-7. 1966a. Addendum 4 mon Mémoire: Microfossiles des Silex Crétacés. Arch. orig. Centre document. C.N.R.S., Paris no. 419, I-9, pl. 1. 1966b. Etude micropaléontologique des Silex du site de Pincevent. Gallia préhistoive, Gy, Tao), A 8 Shei, Jol, ©. DEFLANDRE, G. & Cookson, I. C. 1954. Sur le microplancton fossile conservé dans diverses roches sédimentaires australasiennes s’étageant de Crétacé inférieur au Miocene supérieur. C. rv. hebd. Séanc. Acad. Sci., Paris, 239, 19 : 1235-8, t-figs I-17.

FROM ENGLAND, SCOTLAND AND FRANCE 255

DEFLANDRE, G. & Cookson, I. C. 1955. Fossil microplankton from Australia late Mesozoic and Tertiary sediments. Aust. J. mar. Freshwat. Res., Melbourne, 6 : 242-313, pls. 1-9, t-figs I-59.

DEFLANDRE, G. & DEFLANDRE, M. 1962. Fichier micropaléontologique générale. Dino- flagellés III. Arch. orig. Centre document C.N.R.S. Paris, 383, fiches 1751-1947.

—, 1966a. Fichier micropaléontologique générale. Dinoflagellés V. Arch. orig. Centre docum. C.N.R.S., Paris, fiches 2876-3175.

—, 1966b. Fichier micropaléontologique générale. Dinoflagellés VI. Avch. orig. Centre docum. C.N.R.S., Paris, fiches 3237-3307.

DEFLANDRE, G. & SaRJEANT, W. A. S. 1969. Nouvel examen de quelques holotypes de Dinoflagellés fossiles et d’Acritarches. Arch. orig. Centre docum. C.N.R.S. No. 466, I-10, plea:

DeEtcourt, A. & SPRUMONT, G. 1957. Quelques microfossiles du Wéaldien de Féron-Glageon. Bull. Soc. géol. belg., 6, 1 : 57-68, pls. 1-3.

DopveEKova, L. 1967. Les dinoflagellés et acritarches de |’Oxfordien-Kimméridgien de la

Bulgarie du Nord-est. Ann. Univ. Sofia, 60 : 1-29, pls. 1-3, figs 1-2.

1969. Dinoflagellés et Acritarches du Tithonique aux environs de Pleven, Bulgarie Centrale du Nord. Bull. geol. Inst. Sofia, 18 : 13-24, pl. 1-5, text-figs A—H, table 1. DownleE, C. 1957. Microplankton from the Kimeridge Clay. Q.J.G.S. London, 112 : 413-

434, pl. 20, t-figs 1-6.

DownlE, C. & SARJEANT, W. A.S. 1964. Bibliography and Index of fossil dinoflagellates and acritarchs. Mem. Geol. Soc. Am., Washington, no. 94 : 1-180.

Etsenack, A. 1958. Mikroplankton aus dem norddeutschen Apt nebst einigen Bemerkungen iiber fossile Dinoflagellaten. Neues Jb. Geol. Paléont. Abh., Stuttgart, 106 : 383-422, pls.

21-27.

1961. Einige Erérterungen iiber fossile Dinoflagellaten nebst Ubersicht iiber die zur Zeit bekannten Gattungen. Neues Jb. Geol. Paléont. Abh., 112, no. 3 : 281-324, pls. 33-37, t-figs I-7.

1964. Katalog der fossilen Dinoflagellaten, Hystrichosphaven und verwandten Mikrofossilien. Schweizerbart, Stuttgart. 1-895, 9 pls.

EIsENAcK, A. & Cookson, I. C. 1960. Microplankton from Australian Lower Cretaceous sediments. Pyvoc. R. Soc. Vict., Melbourne, 72, 1 : 1-11, pls. 1-3.

GitMEz, G. U. 1970. Dimnoflagellate Cysts and Acritarchs from the Basal Kimmeridgian (Upper Jurassic) of England, France and Scotland. Bull. By. Mus. nat. Hist. (Geol.). 18, 7 : 231-331, pl. 1-14, text-figs 1-34.

Gocut, H. 1964. Planktonische Kleinformen aus dem Lias/Dogger-Grenzbereich Nord-und

Sitiddeutschlands. Neu. Jb. Geol. Paldont. Abh., Stuttgart, 119; 113-133, pls. 15-17, figs

I-57.

1970. Dinoflagellaten-Zysten aus den Bathonium des Erdélfeldes Aldorf (N W— Deutschland). Palaeontographica, Stuttgart, Abt. B, 129 : 125-165, pls. 26-35.

Gé6rKa, H. 1965. Les microfossiles du Jurassique Supérieure de Magnuszew (Pologne). Acta. palaeont. pol., Warszawa, 10 : 291-334, pls. I-5.

KarveE-Corvinus, G. 1966. Buiostratigraphie des Oxfordium und untersten Kimmeridgium am Mont Crussol, Ardéche, im Vergleich mit Siiddeutschland N. Jb. Geol. Paliéont Abh. Stuttgart 2 : 101-141, pls. 22-26.

KLEMENT, K. W. 1960. Dinoflagellaten und Hystrichosphaerideen aus dem Unteren und Mittleren Malm Siidwestdeutschlands. Palaeontographica, Stuttgart, Abt.A, 114 : 1-104, pls. 1-7, t-figs 1-37.

Manvum, S. & Cooxson, I. C. 1964. Cretaceous microplankton in a sample from Graham Island, Arctic Canada, collected during the second ‘“‘Fram-” Expedition (1898-1902), with notes on microplankton from the Hassel Formation, Ellef Ringnes Island. Shr. Norske Vidensk-Akad. mat. Nat. Kl. (N.S.)., Oslo, 17 : 1-36, pls. 1-7.

Mititioup, M.E. 1967. Palynological study of the type localities at Valangin and Hauterive. Rev. Palaeobot. Palynol., Amsterdam, 5 : 155-167, pls. 1-2, fig. I.

256 KIMMERIDGIAN DINOFLAGELLATES & ACRITARCHS

MorcEenrotH, P. 1966. Mikrofossilien und Konkretionen des Nordwesteuropdischen Untereozans. Palaeontogvaphica, Stuttgart, Abt. B, 119 : 1-53, pls. 1-11, table 1.

Norris, G. & SARJEANT, W. A.S. 1965. A descriptive index of genera of fossil Dinophyceae and Acritarcha. Bull. Geol. Surv. N.Z., Wellington, no. 40 : 1-72.

Roman, F. 1936. Geology of the Rhone Valley (Lyons-Crussol). Proc. Geol. Ass., London, 47 : 153-182, figs 40-48.

SARJEANT, W. A. S. 1960. New hystrichospheres from the Upper Jurassic of Dorset. Geol. Mag., London, 97, 2 : 137-144, pl. 6, t-figs 1-4.

— 1961a. Microplankton from the Kellaways Rock and Oxford Clay of Yorkshire. Palaeontology, London, 4, i : 90-118, pls. 13-15, t-figs I-15.

1961b. Systematophora Klement and Polystephanosphaera Sarjeant. J. Paleont., Tulsa, Oklahoma, 35, 5 : 1095-1096.

1962a. Upper Jurassic microplankton from Dorset, England. Micropaleontology, New York, 8, 2 : 255-268, pls. 1-2, t-figs 1-8, tables 1-4.

196zb. Microplankton from the Ampthill Clay of Melton, South Yorkshire. Palaeon- tology, London 5, 3 : 478-497, pls. 69-70, t-figs 1-13, tables 1-3.

1963. Fossil dinoflagellates from Upper Triassic sediments. Nature, Lond. 199, no. 4891 : 353-354, t-figs 1-3.

1964. The stratigraphic application of fossil microplankton in the Jurassic. In P. L. Maubeuge (ed.) Colloque du Jurassique, Luxembourg 1962, Volume des comptes-vendus et mémotres, Inst. Grand-Ducal, Sect. Sci. Natur., Luxembourg, 441-448, tables 1-4.

1965. Microplankton from the Callovian of Normandy. Revue Micropaléont. Paris, 8, 3 : 175-184, pl. I, fig. 1.

1966. (See Davey, R. J., and others, 1966.)

1967a. The genus Palaeoperidinium Deflandre 1934 (Dinophyceae). Gvrana palynol., Uppsala, 7, I : 241-258, figs 1-6.

1967b. The stratigraphical distribution of fossil Dinoflagellates. Rev. Palaeobot.

Palynol., Amsterdam, 1 : 323-343, tables, 1-1. 2

— 1967c. Observations on the Acritarch genus Micrhystridium (Defl.). Revue Micro- paléont., Paris 9, 4 : 201-208, pl. 1, fig. 1.

1968. Microplankton from the Upper Callovian and Lower Oxfordian of Normandy. Revue Micropaléont., Paris, 10, 4 : 221-242, pls. 1-3, figs 1-5, tables 1-2.

1969. (See Davey, R. J. and others, 1969.)

SARJEANT, W. A. S. & DownlE, C. 1966. The classification of Dinoflagellate Cysts above generic level. Gvana Palynol, Uppsala, 6, 3 : 503-527.

Scuurz, E. & Mar, D.H. 1966. Erlauterungen zur Tabelle der stratigraphischen Verbreitung des Phytoplanktons im Lias und Dogger. In H. Doring et. al. (eds.): Erlauterungen zu den sporenstratigraphischen Tabellen vom Zechstein bis zum Oligozan, Zentr. Geol. Inst. Abh., Berlin, H8 : 35-45, table 1.

STAPLIN, F. L., JANsonius, J. & Pocock, S. A. J. 1965. Evaluation of some acritarchous hystrichosphere genera. Neues Jb. Geol. Paldont Abh., Stuttgart, 123 : 167-201, pls. 18-20.

VALENSI, L. 1953. Microfossiles des silex du Jurassique moyen. Remarques pétrographiques.

Mem. Soc. geol. Fr., Paris, 68 : 1-100, pls. 1-16, t-fig. 1-7.

1955. Etude micropaléontologique des silex du Magdalénien de Saint-Amand (Cher).

Bull. Soc. préhist. Fr. Paris, 52, 9-10 : 584-596, pls. 1-5.

VOZZHENNIKOVA, T. F. 1967a. Intvoduction to the study of fossilized peridinid algae. National Lending Library for Science Technology, Boston Spa, Yorks, England. (English trans- lation by K. Syers, edited by W. A. S. Sarjeant). 233+ xiii, figs 1-51, tables 1—10.

—— 1967b. Iskopayemiye peridinot yurskikh, myeolovikh, palaeogenovikh otlozhenty S.S.S.R., Moscow, Nauka, 1-347, pls. 1-120, 14 figs. 12 tables.

Walt, D. & Downtz, C. 1963. Permian hystrichospheres from Britain. Palaeontology, London 5, 4 : 770-784, pls. 112-114, t-figs 1-4.

Waterston, C. D. 1951. The stratigraphy and palaeontology of the Jurassic Rocks of Eathie (Cromarty). Tvans. R. Soc. Edin., Edinburgh, 62 : 33-52, pls. 1-2, figs 1-4.

FROM ENGLAND, SCOTLAND AND FRANCE 257

WETZEL, W. 1966. Characteristik des marinen Planktons und Pseudoplanktons der Amaltheen- Schichten Deutschlands und Lothringens. Neues Jb. Geol. Paldéont. Abh., Stuttgart, 124, 3 : 313-326, pls. 30-31.

Wittiiams, D. B. & SaryjEANT, W. A. S. 1967. Organic-walled Microfossils as depth and shoreline indicators. Mar. Geol., Amsterdam, 5 : 389-412, figs 1-16.

Wittiams, G. L. & DownlE, C. 1966. See Davey, R. J. et al.

Mrs G. U. GitmeEz, Ph.D., Dept. of Geology, HACETTEPE UNIVERSITY, ANKARA,

TURKEY

W. A. S. SaRJEANT, B.Sc., Ph.D., F.G.S., F.L.S., Dept. of Geology,

UNIVERSITY OF SASKATCHEWAN,

SASKATOON,

CANADA

Fic. 1. FIG. 9. Fie. 2 Fic. 3 Fic. 4 Fic. 5 Fig. 6 Fie. 8 1ges / FIG. 10 Fic. 11 FIG. 12

PLATES:

Tenua echinata sp. nov. General appearance and archaeopyle formation of the holotype, V.52790(I) x716. Paratype, I.G.S. slide PK 119 450. Chytroeisphaeridia chytroeides Sarjeant IEG Sa slid pe ke Tro2 Aw Gr53o: Chytroeisphaeridia mantelli sp. nov. Holotype, showing archaeopyle with attached operculum. I.G.S. shde PK 116 560. Paratype A, I.G.S. slide PK 114 500. Chytroeisphaeridia pococki Sarjeant BM(NH) slide V.53961(3) 1600.

Fromea warlinghamensis sp. nov. Holotype (b) I.G.S. slide PK 115 720. Paratype BM(NH) slide V.56340 (1) x5IOo.

Tenua sp. BM(NH) slide V.53619 (1) 690. Same specimen taken by phase contrast. 690. Tenua capitata (Cookson & Eisenack) Specimen with attached operculum. BM(NH) slide V.56341 x8 40. Same specimen by phase contrast. 8 40.

Bull. By. Mus. nat. Hist. (Geol.) 21, 5

IPIEAN IIS, at

ey

aor

eon F 7

ne

coo™I

IPIL IN IID, 2

Gonyaulacysta cauda sp. nov. Ventral view of the holotype. BM(NH) slide V.53965 (2) 614. Dorsal view of the holotype. x 6r4.

Ventral view of the paratype. BM(NH) slide V.56343 (1) 716. Dorsal view of the paratype, by transparency. 716. Cryptarchaeodinium cf. calcaratum

Dorsal view. BM(NH) slide V.56342 xc. 1

,000.

Gonyaulacysta longicornis (Downie)

A specimen from the Middle Kimmeridgian.

1.G.S. slide PK 120.

Gonyaulacysta cladophora (Deflandre) I.G.S. slide PK r10. 464. Sample WB7 (Eudoxus Zone).

Same specimen by phase contrast. x 464.

x 496.

IPIL JANIS, 2

Bull. By. Mus. nat. Hist. (Geol.) 21, 5

Fic. Fic.

Fila. Fig.

Fic.

Fic. Fie.

Fic.

FiG.

iS)

Ww

PLATE 3

Gonyaulacysta globata sp. nov. Holotype, I.G.S. slide PK 122. 625. Paratype BM(NH) slide V.56345. x 6r4. Gonyaulacysta cf. guiseppei Morgenroth BM(NH) slide V.56344. Ventral view. 600. Dorsal view of the same specimen, showing the archaeopyle. 600. Gonyaulacysta nuciformis (Deflandre) Ventral surface of specimen. I.G.S. slide PK 109. 576. Gonyaulacysta eisenacki (Deflandre) BM(NH) slide V.56375. Showing the ventral tabulation. x 8go. Dorsal view of the same specimen. x 8go. Leptodinium cf. crassinervum (Deflandre) Specimen with detached operculum. BM(NH) slide V.56351 (1) x 384. Leptodinium sp. BM(NH) slide V.56352. 572.

Bull. By. Mus. nat. Hist. (Geol.) 21, 5 IP ILE, 3}

Fic.

Fic.

Fic.

Fia.

Fig.

Fic.

Fic.

PLATE 4

Gonyaulacysta longicornis (Downie) A specimen from the Upper Kimmeridgian. Gonyaulacysta sp. B Ventral view of a specimen from the Middle Kimmeridgian.

I.G.S. slide PK 117.

x 840.

Dorsal view of the same specimen. Meiourogonyaulax sp.

BM(NH) slide V.56359 (1).

x 840.

BM(NH) slide V.56346 (2).

The operculum is still attached. 614.

Meiourogonyaulax pila sp. nov. BM(NH) shde V.56358. Holotype. Gonyaulacysta perforans (Cookson & Eisenack)

1.G.S. slide PK 131.

x 1430.

x 6901.

Gonyaulacysta cf. mamillifera

1.G.S. shde PK 130.

x 759.

x 480.

Bull. Br. Mus. nat. Hist. (Geol.) 21, 5 IPILININIS, A

Fic. Fic. Fic.

TG: Fic. Eines

Fic. Fic.

WO NH

Ore

on

DIGAINE ¢

Leptodinium aceras (Eisenack) Ventral surface. I.G.S. slide PK 108. 640. Dorsal surface of the same specimen. 640. A specimen from the Lower Kimmeridgian. I.G.S. slide PK 112. 626. Leptodinium cf. crassinervum (Deflandre) Tabulation on the ventral side. BM(NH) slide V.56346 (1 & 3). 768. Dorsal surface of the same specimen; plate 3” is lost in archaeopyle formation Dorsal view of another specimen, with operculum still attached. BM(NH) slide V.56351 (1). 480. Gonyaulacysta systremmatos sp. nov. Holotype BM(NH) slide V.53966 (1), showing the ventral surface. 848. Dorsal view of the holotype. 848.

x 768.

Bull. Br. Mus. nat. Hist. (Geol.) 21, 5 PLATE 5

Fic. Fic.

Fic. Fiac.

Fic. Fia.

EK Fic.

Fia.

H

one

Oi

IPL NAD,

Gonyaulacysta sp. C I.G.S. shde PK 118. taken by phase contrast. 656.

Same specimen, by Nomarski interference contrast.

Gonyaulacysta sp. F Left lateral view. BM(NH) slide V.56348 (1). x 512.

Right lateral view, as seen by Nomarski interference contrast.

Gonyaulacysta sp. D Ventral view. BM(NH) slide V.56346 (4). 580. Dorsal view of the same specimen. 580. Gonyaulacysta sp. G Ventral view. BM(NH) slide V.56349 (1). 768.

Dorsal view. 768.

Gonyaulacysta sp. E

Ventral view of the specimen.

I.G.S. slide PK 113.

x 656.

x 680.

x 512.

6

PLATE

Mus. nat. Hist. (Geol.) 21, 5

Bull. Br.

On

. 10. . It.

PLATE 7

Leptodinium egemenii Gitmez

Ventral surface of the paratype, BM(NH) slide V.52798 (3), by phase contrast. Dorsal view of the paratype: plate 3” lost in archaeopyle formation. 574.

Meiourogonyaulax pila sp. nov. General appearance of the paratype, showing the attached operculum. I.GiS. slide PK 121. 544. Dictyopyxis cf. reticulata (Valensi) Ventral surface of specimen. BM(NH) slide V.56354 (1). 768. Dorsal view, showing the median band equivalent to a cingulum. 768. Meiourogonyaulax dicryptos sp. nov.

Holotype, with apical archaeopyle; taken using Nomarski interference contrast.

BM(NH) slide V.56357 (1). x 704.

Histiophora cf. ornata Specimen with apical archaeopyle. BM(NH) slide V.52964 (1). 768. Same specimen, taken by phase contrast. 768.

Dictyopyxis areolata Cookson & Eisenack BM(NH) slide V.53956 (1). Specimen with apical archaeopyle. 614. Occisucysta monoheuriskos sp. nov.

Ventral view of the holotype. BM(NH) slide V.56353 (1). 615.

Dorsal view of the holotype, showing the two-plate precingular archaeopyle.

Meiourogonyaulax sp. Specimen with attached operculum. I.G.S. slide PK 100. 1200.

« O15.

7

PLATE

Mus. nat. Hist. (Geol.) 21, 5

Bull. Br.

FIGS. I-2.

Fics. 3-4.

PPA Ess

Egmontodinum polyplacophorum gen. et. sp. nov. Holotype, BM(NH) slide V.56360 (2b). 1. In ventral view. 2. In dorsal view,

by transparency. 1215. Paratype (a), BM(NH) slide V.56360. 3. In oblique ventral view, by transparency |

details of posterior not visible. 4. In oblique dorsal view. 1215.

PIL JAI, 8

Bull. By. Mus. nat. Hist. (Geol.) 21, 5

nN

PLATE 9

Gonyaulacysta sp. A Gitmez and herein BM(NH) slide V.56347 (2). 720. Detail of apex by phase contrast, showing the short spines on crests. 1215. Egmontodinium polyplacophorum gen. et. sp. nov. Paratype (c), showing apical archaeopyle. BM(NH) slide V.56347 (1). 720. Meiourogonyaulax staffinensis Gitmez BM(NH) slide V.56356. xc. 720. Fromea warlinghamensis sp. nov. Paratype (a), BM(NH) slide V.56339 (2). 720.

Detail of surface, showing bulges and polygonal patterning considered to be produced

by pressure of mineral grains. 215.

9

PLATE

Mus. nat. Hist. (Geol.)

Bull. Br.

Fic. Fic. Fic. Fic.

Fic. Fic.

Fic.

PLATE to

Adnatosphaeridium paucispinum (Klement)

BM(NBH) slide V.56365 (1). 480. Same specimen taken by phase contrast. x 480.

Specimen with apical archaeopyle. BM(NH) slide V.56366 (2).

phase contrast. 544.

Dorsal view of the same specimen by phase contrast. 544.

Leptodinium amabilis (Deflandre) Ventral view. BM(NH) slide V.56350 (1). 1200.

Cleistosphaeridium ehrenbergi (Deflandre)

BM(NH) slide V.56376 Sample LO353, by phase contrast.

x 651.

Ventral view, by

. Dorsal view of the same specimen, showing the precingular archaeopyle. 1200.

Bull. Br. Mus. nat. Hist. (Geol.) 21, 5 PLATE ro

Fic.

Fic. Fic.

Fic. Fic. Fic. Fia.

Fic. Fic.

0

IPILVNANIS, ann

Imbatodinium cf. villosum Vozzhennikova BM(NH) slide V.56362 (1); specimen from the Upper Kimmeridgian of Dorset. 608. Imbatodinium antennatum sp. nov. Paratype, showing the intercalary archaeopyle. BM(NH) slide V.56361 (1). 768. Holotype, showing the intercalary archaeopyle. I.G.S. sude PK 124. 845. Proximate cyst sp. indet. Specimen with attached operculum. BM(NH) slide V.56363 (1). 513. Same specimen, taken by Nomarski interference contrast. 513. General view of another specimen; BM(NH) slide V.56364. 451. Egmontodinium polyplacophorum gen. et. sp. nov. Holotype in ventral view, showing the tabulation. BM(NH) slide V.56360 (2b). x 640. Dorsal view of the holotype, by transparency. 640. Paratype (c), showing apical archaeopyle. BM(NH) slide V.56347 (1). x 1024.

PIL ANTES, wi

Mus. nat. Hist. (Geol.) 21, 5

Bull. Br.

/

Nor *

a

.

om

ines

FIG.

nee

1BKE, AL, Fic.

Fic.

Fic.

Fie. Fic.

oe)

9.

JIC PNAD IS, 12

Dictyopysxis cf. reticulata (Valensi) In presumed oblique ventral aspect. BM(NH) slide V.56355 (1). 768. Same specimen in presumed oblique dorsal aspect showing the apical archaeopyle. < 768. Chytroeisphaeridia mantelli sp. nov. Paratype B, showing the archaeopyle and slits. BM(NH) slide V.56338 (1). 532. Apteodinium granulatum Eisenack Ventral view of specimen. I.G.S. slide PKi1o2z. 592. Dorsal view of the same specimen, showing the precingular archaeopyle. 592. Apteodinium cf. maculatum Eisenack & Cookson I.G.S. slide PK 105. 464. Pterospermopsis australiensis Deflandre & Cookson BM(NH) slide V.56353 (2) Sample CS421. x 1000. Polystephanephorus sarjeantii Gitmez Holotype by phase contrast. BM(NH) slide V.52792 (2), sample OM 131. 580. Holotype showing the apical archaeopyle. (Phase contrast). 580.

y 2

IPL NA

Ib 5)

2

Mus. nat. Hist. (Geol.)

Bull. Br.

Fic.

Fic.

Fic.

Fic.

Fia.

Fic. Ges

NOW

IPL AN IIS 03)

Gonyaulacysta sp. H Oblique ventral view BM(NH) slide V.56339 (1), sample CH231. 720. Systematophora orbifera Klement Showing apical archaeopyle. BM(NH) slide V.56377, sample OM 418. Oligosphaeridium pulcherrimum (Deflandre & Cookson) BM(NH) slide V.56368 (1), sample CC 449. 800. Scriniodinium sp.

Ventral view of the specimen I.G.S. slide PK 107. 448.

Prolixosphaeridium granulosum (Deflandre) BM(NH) slide V.52799 (4); from the sample OM 131. 768. Specimen with attached operculum. I.G.S. slide PK 103. 573.

x 720.

Specimen lacking the operculum. BM(NH) slide V.53960 (3). Sample SC 444.

x 1024.

Ih 5)

2

Mus. nat. Hist. (Geol.)

Bull. By.

Fic. Fic. Fic.

Fic.

Fic. Fia.

Fic. Fic. Fic.

Fic. Ge

nN

10. JE

PLATE 14

Systematophora ovata sp. nov. Paratype. BM(NH) slide V. 56343 (2) 717. Holotype, with apical archaeopyle. BM(NH) slide V.53962 (1). 800. Same specimen, under phase contrast. x 800. Parvocavatus tuberosus Gitmez BM(NH) slide V.56353 (3). Sample ES 421. x1075. Hexagonifera jurassica sp. nov. Holotype, with attached operculum. I.G.S. slide PK 123. 350. Paratype with apical archaeopyle. BM(NH) slide V.53621 (1). 8109. Stephanelytron redcliffense Sarjeant BM(NH) slide V.56366 (1). 8 o. Stephanelytron cf. redcliffense Sarjeant Specimen with apical archaeopyle. BM(NH) slide V.56365 (2). 768. Endoscrinium sp.

Ventral view of the specimen by phase contrast. BM(NH) slide V.56369 (1).

Ventral view of same specimen, by ordinary light. x 48o.

Dorsal view of same specimen showing the precingular archaeopyle. _ 480.

x 640.

14

PEATE

Bull. Br. Mus. nat. Hist. (Geol.) 21, 5

FIG.

PLATE 15

Muderongia simplex Alberti I.G.S. slide PK 128. 720. Specimen showing the apical archaeopyle. I.G.S. slide PK 129. 922. Cleistosphaeridium sp.

BM(NH) slide V.56367 (1). 720.

Scriniodinium bicuneatum (Deflandre) BM(NH) slide V.56370. 752.

Scriniodinium dictyotum subsp. osmingtonensis Gitmez BM(NH) slide V.52799 (1) sample OM 131. Holotype, without prominence at the apex. X47I. Scriniodinium dictyotum subsp. papillatum Gitmez Holotype, with precingular archaeopyle, showing the blunt, mammelon-shaped apical prominence. BM(NH) slide V.53940 (1). 480. Scriniodinium dictyotum subsp. pyrum Gitmez

Specimen with strong apical horn. BM(NH) slide V.56371.

Bull. Br. Mus. nat. Hist. (Geol.) 21, 5 PLATE 15

Fic.

Fia. Fic.

Fic.

Fic.

Fic.

Fic. Fic.

con]

PADI NANI; 1(6) Organism A BM(NH) slide V.53948 (3). General appearance. 691. Cavate cyst sp. indet. B I.G.S. shde PK 104. x 640. I.G.S. slide PK 106. x 640. Cavate cyst sp. indet. A 1.G.S. slide PK 102C (3). 640. Pterospermopsis helios Sarjeant BM(NH) slide V.53963 (1). Sample SC 144 (Baylei Zone). 2100. Scriniodinium dictyotum subsp. dictyotum Cookson & Eisenack Specimen with a slight apical prominence. BM(NH) slide V.56372 (1). 496. Sirmiodinium grossi Alberti Dorsal view of specimen. BM(NH) slide V.56373 (1). 896. Specimen with apical archaeopyle. BM(NH) slide V.56374 (1). 941.

Bull. By. Mus. nat. Hist. (Geol.) 21, 5 PLATE 16

Fic.

Eig. 2 Fig. 3 Fic. 4 Fic. 5 Fic. 6 Fic. 7 TGs 3 FIG. 9 Fic. 10

PLATE 17

Micrhystridium recurvatum Valensi

EGS. slide PK 127, x1276;

The other side of the same specimen. 1276.

Organism A

BM(NH) slide V.53948 (3). Phase contrast objective used. 768. Acritarch sp. indet. Specimen with opening. I.G.S. slide PK 125. x 1200.

Another specimen with opening.

I.G.S. slide PK 126.

Pterospermopsis harti Sarjeant

I.G.S. side PK 111. 1085.

Micrhystridium sp. BM(NH) shde V.53953 (1). 1280. Phase contrast view of same specimen. 1280. Solisphaeridium claviculorum (Deflandre)

I.G.S. slide PK tor. 1320. Phase contrast view of specimen.

x 1320.

x 1400.

V7,

PLATE

Mus. nat. Hist. (Geol.) 21, 5

Bull. Br.

A LIST OF SUPPLEMENTS ~ TO THE GEOLOGICAL ‘SERIES OF THE BULLETIN OF | : THE BRITISH MUSEUM SS HISTORY) :

Pps 213: 30 ‘Plates: 2 sa eri 706s ae . Et-Nacear, Z. R. Stratigraphy and Planlctovae Ronneaieat ° Cretaceous—Lower Tertiary Succession in the Esna-Idfu Regio Egypt, U.A.R. Pp. 291; 23 Plates; 18 Text-figures. 1966. . Davey, R. J., Downte, C., SARGEANT, W. A. S. & WILLIAM: Mesozoic and Cainozoic eee ieee ae ae figures. meee Sy fate

Appendix to Studies on n Mesozoic ee Cainozoic Dinoflag 1969. 8op. eos ; ELLIOTT, Lee Permian to Pace me Ae @ us Middle East. Pp. 111; 24 Plates; 17 Text-figures. 1968, — . Ruopes, F. A Ls SosEy R.. L, & Drvce, E. Pes 1 Ee

Pp. 315; 31 Pinter: 92 Tete heae 1969. fe ee . Cuitps, A. Upper Jurassic Rhynchonellid | "Brachiopetis om Europe. Pp. 119; 12 Plates; 40 Text-figures. 1969. i

- Goopy, P. C. The relationships of -certain Upper special reference to the Myctopheid = 255; ween

3 Plates: 52 Texthnie ‘Ig7T. fe: oe Ee . Sippigur, Q. A. Early Tertiary Oiees ee the cay Trachyleberic from West Pakistan. ae 98; 42 bene xe Bie go of EGYL eS ‘

abPLe

JE8 Printed in England by Staples Printers Limited at their Kettering, Northants, establishment,

-TERTIARY CYTHERETTINAE OF. _ NORTH-WEST EUROPE See”

é KE » | 2 9 MAY i973

aac .°

£ ey & S Shay Wr

_ BULLETIN OF MUSEUM (NATURAL HISTORY) Ree Vol. 21 No. 6 _ LONDON : 1972

x

MID-TERTIARY CYTHERETTINAE OF NORTH-WEST EUROPE

BY

MICHAEL CHARLES KEEN_,

Glasgow University

Pp. 259-349; 23 Plates, 30 Text-figures

BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) GEOLOGY Vor 21 No, 6 LONDON : 1972

THE BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY), instituted in 1949, is issued in five series corresponding to the Departments of the Museum, and an Historical series.

Parts will appear at irregular intervals as they become ready. Volumes will contain about three or four hundred pages, and will not necessarily be completed within one calendar year.

In 1965 a separate supplementary series of longer papers was instituted, numbered serially for each Department.

Ihlis paper is Vol. 21, No. 6 of the Geological (Palaeontological) series. The abbreviated titles of periodicals cited follow those of the World List of Scientific Periodicals.

World List abbreviation Bull. Br. Mus. nat. Hist. (Geol.)

© Trustees of the British Museum (Natural History), 1972

TRUSTEES OF THE BRITISH MUSEUM (NATURAL HISTORY)

Issued 29 December, 1972 Price £4.60

MID-TERTIARY CYTHERETTINAE OF

NORTH-WEST EUROPE

By M. C. KEEN

CONTENTS

INTRODUCTION AND ACKNOWLEDGEMENTS LOCALITIES .

STRATIGRAPHICAL DISTRIBUTION 5 EVOLUTION OF MID-TERTIARY Caen ECOLOGICAL DISTRIBUTION OF RECENT SPECIES . PALAEOECOLOGY z

GEOGRAPHICAL DISTRIBUTION .

CLASSIFICATION OF THE SUBFAMILY CYTHERETTINAE .

THE SHELL STRUCTURE OF CYTHERETTA SOFT PARTS . ‘ : LARVAL STAGES THE SHELL STRUCTURE OF FLE x US THE TYPE SPECIES OF CYTHERETTA SYSTEMATIC DESCRIPTIONS

Order Podocopida Miiller

Sub order Podocopina Sars Family Tvachyleberididae Sylvester Bradley

Genus Cytheretta Miller . . A judaea (Brady) 2 = : subvadiosa (Roemer)

costellata (Roemer) : costellata costellata (Roemer)

costellata grandipora subsp. nov. costellata cvatis subsp. nov.

costellata antecalva subsp. nov. bambruggensis Keij.

crassivenia Apostolescu

decipiens Keij

aff. decipiens Keij

haimeana (Bosquet)

ruelensis Sp. NOV.

eocaenica Keij

oligocaenica sp. nov.

geoursensis sp. Nov.

cavita sp. nov.

cellulosa sp. nov.

aff. cellulosa

Superspecies C. laticosta (Reuss)

C. laticosta (Reuss)

C. forticosta sp. nov.

C. porosacosta sp. nov.

Superspecies C. tenuipunctata (Bosquet) C. tenuipunctata (Bosquet)

C. tenuipunctata tenuipunctata (Bosquet) C. tenuipunctata absoluta subsp. nov..

NANHHOHHAGHHAHAAHHAAADS

Page 263 265 207 270 272 273 274 274 277 280 280 281 281 282 282 282 282

282 283 284 285 289 289 290 291 293 293 294 294 295 295 296 297 298 299 300 301 302 304 304 304 3097 309 309 310

262 MID-TERTIARY CYTHERETTINAE C. tenuipunctata livata subsp. nov. . ¢ : - < 311 C. tenutstriata (Reuss) . : : j é c 312 C. tenuistriata tenuistriata (Reuss) ° : é : . 312 C. tenuistriata ornata subsp. nov. : : A . : 313 C. minipunctata sp. nov. . : : 5 5; 5 : 314 C. buttensis sp. nov. é é : ; : : 315 C. buttensis buttensis subsp. nov. é : : : ‘ 316 C. buttensis reticulata subsp. nov. : 5 : : 316 Discussion of Superspecies C. pane : : : 318 C, minoy (Lienenklaus) . : é : : ; ; 319 C. posticalis Triebel : c c : : é ¢ 320 C. posticalis posticalis Triebel : : . : : 320 C. posticalis parisiensis subsp. nov. . : é : : 320 C. headonensis Haskins . g : . : ¢ c 321 C. vesca sp.nov. . c : 2 : : 5 : 322 C. stigmosa Triebel . c 5 : 5 : : , 323 C. stigmosa stigmosa Triebel_ . : : : ; : 323 C. aff. stigmosa Triebel . > . : : : : 323 C. stigmosa gallica subsp. nov. . : c é : : 323 C. vegularis sp. nov. : : : : é : : 325 C. bullans sp. nov. ; . é é ; : é 327 C. sagri Deltel : : : : c : : : 327 C. sagrvi sagvi Deltel 5 . 5 5 : : e 329 C. sagyi inconstans subsp. nov. : : c : : 330 C. sagyi martini subsp. nov. . ; < - . : 331 C. samothracia Deltel 5 c : : : : 5 332 C. minipustulosa sp. nov.. ; : ; 3 5 : 333 C. gibberis sp. nov. . : : : 5 . . : 334 C. postornata sp. nov. a c 5 3 c 5 : 336 C. perita Deltel : 0 ; : 5 : c 336 C. sculpta Ducasse c ° 6 ° : é : 337 Cytherettasp. A. . . - : < ; : : 338 Cytherettasp.B. . = 5 : F ; : 338 Cytheretta sp. C. . : 5 0 c . 5 ‘ 338 Genus Flexus Neviani . - - - : : : 338 F. plicatus (von Munster) D - 5 2 . 2 339 F. concinnus (Triebel) . 5 : . : : c 339 F. gutzwillevi (Oertli) . 4 : : : : F 340 F. solentensis sp.nov. . 5 5 é : 340 F. solentensis solentensis subsp. nov. . 3 . c c 340 F. solentensis congestus subsp. nov. . “ 6 - : 341 F. ludensis sp. nov. : : c : 5 : : 342 F. lenijugum sp. nov. a a , c ; 9 : 343 F. schoelleri (Keij) . c : : : : 5 : 344 Flexus sp. A . : : 6 5 : : : : 345 XV. CONCLUSIONS . ; : 0 5 < 5 : : 2 345 SUMMARY

Fifty eight species and subspecies of Cytheretta and ten of Flexus are described from the Middle and Upper Eocene and the Oligocene of western Europe. New species and subspecies are: from the Bartonian of the Paris Basin, Cytheretta costellata gyandipora, C. costellata cratis, C. carita, C. cellulosa, C. vuelensis, and Flexus ludensis; from the Upper Eocene of the Hampshire

OF NORTH-WEST EUROPE 263

Basin, C. costellata antecalva, C. forticosta, C. porosacosta, F. solentensis solentensis and F. solen- tensis congestus; from the Oligocene of the Paris Basin, C. tenuipunctata absoluta, C. tenui- punctata livata, C. tenuistriata ornata, C. minipunctata, C. buttensis buttensis, C. buttensis reticulata, C. posticalis parisiensis, C. vesca, and C. stigmosa gallica; and from the Oligocene of the Aquitaine Basin, C. oligocaenica, C. vegularis, C. bullans, C. gibberis, C. sagrt inconstans, C. sagyi martini, C. minipustulosa, C. postornata, and F. lenijugum. The stratigraphical distribution, ecology, and classification are also discussed.

I. INTRODUCTION AND ACKNOWLEDGMENTS

THE following study of the subfamily Cytherettinae is mainly concerned with Upper Eocene and Oligocene species, but also includes some from the Lutetian, from possible Miocene, and from the Pliocene and Recent. The area covered includes the Hampshire Basin, the Paris Basin, Belgium and the Aquitaine Basin.

The study is taken from a Ph.D. thesis completed at the University of Leicester in 1967. Since then the Geology Department has had a scanning electron microscope installed, and through the permission of Prof. P. C. Sylvester-Bradley it has been possible to re-photograph the ostracods. This has sometimes brought out characters which are not very clear under an optical microscope, and very high magnifications are possible which show features not visible at all with an ordinary microscope. The photographs were taken by Mr G. Mc. Turk, to whom thanks are extended.

The stratigraphy of the Upper Eocene and Oligocene in Western Europe is complicated but the nomenclature of the stage names is even more so. There are some thirteen of the latter in common usage between the Lutetian and the Chattian, so to avoid confusion formation names are used where possible. When stage names are used however, they refer to the current usage in the particular area under dis- cussion. When two or more areas are mentioned, the classification adopted is that of Wrigley & Davis (1937). Recent reviews of the stratigraphy can be seen in Batjes (1959), Cavelier (1964, 1965), Curry (1965, 1966), and Vigneaux (1964).

There are three main problems for the correlation of the mid-Tertiary within western Europe. The first is the relationship between the Middle and Upper Eocene; the second is the difficulty of recognizing the Bartonian in Belgium and the Paris Basin; and finally the placing of the Eocene-Oligocene boundary. Recent symposia at Bordeaux (1962), Paris (1968), and Marburg (1969) have failed to adopt any firm conclusions on any of these points. Most of the traditional concepts are under debate, and with so many ideas in the air correlation charts become redundant very quickly. Fig. 1 lists the horizons sampled for Cytheretta and also gives a tentative correlation.

The species concept adhered to in this work is narrower than has perhaps been usual with Tertiary ostracods. The reasons for this are, first, that by studying one subfamily it is possible to follow its geographical and stratigraphical ramifications. Secondly, comparisons have been made with type material whenever possible. Finally, it is my belief that it is only through the detailed studies of small groups of ostracods that they will take their rightful place for use in Tertiary stratigraphical correlation,

MID-TERTIARY CYTHERETTINAE

264

OLIGOCENE

UPPER EOCENS

MIDDLE EOCENE

HAMPSHIRE BASIN

PARIS BASIN; BELGIUM FALUNS D’ORMOY

MORIGNYJARGILE DE BOOM JEURRE

MARNES A HUITRES BERG _N.comta

C. DE SANNOIS ARGILE VERTE TONGEREN HAMSTEAD BEDS

SMARNES | OYSTER MARLS SUR RAey ie BEMBRIDGE LST.

HEADON BEDS

GY PSE

BARTON BEDS

\

MARNES A R ludensis

SABLES MOYENS:

UPPER

MARINES

CRESNES

BEAUCHAMP BRACKLESHAM AUVERS BEDS

SABLES DE WEMMEL Sy NeiUes) [pls Ls |pS SABLES DE BRUXELLES

CALCAIRE GROSSIER

MIDDLE

RHINE oa am

ASTRUP KASSEL

RUPELTON ~ UNT. MEERESSAND

MELANI ENTON~ LATDORF

EOZAN V

IASTERIES

AQUITAINE

ATAL AYE |

MARNES BLEUES

Correlation of Mid-Tertiary Beds in western Europe.

Fie. 1.

OF NORTH-WEST EUROPE 265

I should like to record my grateful thanks to Prof. P. C. Sylvester-Bradley for his supervision throughout the work and for the use of the facilities of the Department of Geology at the University of Leicester. A study such as this also needs the cooperation of researchers in other countries, and I should particularly like to thank Dr H. J. Oertli, Mlle B. Deltel, the late Prof. J. Cuvillier, and Mme R. Damotte in France; Dr P. Marks in Holland; and Dr E. Triebel in Germany. The work was made possible by a N.E.R.C. NATO Research Studentship. The text-figures have been drawn by Mrs N. Farquharson.

The ostracods described in this paper are in the collections of the British Museum (Nat. History) Palaeontology Department.

il. LOCALIZIES

Most of the samples used came from classical localities, either collected personally or donated by other workers. The sections are often poorly exposed, so few detailed measurements were taken. The relevant parts of some of the sections are given below. Other localities can be seen in Fig. 2.

I. Cormeilles-en-Parisis (P.C.M.).

This famous locality situated in the western suburbs of Paris reveals strata ranging from the marnes a L. inornata, through the gypsum beds, to the Couches de Sannois and Marnes a Huitres at the top.

P.C.M.18—23; Couches de Sannois, sandy clays with shell bands. The samples come from beds 40, 42, 44, 45, 46, and 47 respectively of Albissin (1955).

EC.M.24:; basal Marnes a Huitres.

P.C.M.25; sandy clay 115 cm above base.

P.C.M.26; oyster bed 230 cm above base.

FC. M.27; Brown clay with Polymesoda immediately above oyster bed.

2. Moiselles (PMS)

A sand pit beside the RN I some twenty miles north of Paris. Sables de Beau- champ, Sables d’Ezanville, and Calcaire de St. Ouen are exposed. Only one sample, PMS.8, has yielded well preserved Cytheretta species; this is from the Sables de Beauchamp, 480 cm below the base of the Sables d’Ezanville.

3. Marnes a P. ludensis of the Paris Basin.

Two localities yielded ostracods: Chavengon and Verzy. At both localities the Marnes a P. ludensis is thin; about 150 cm at Chavengon, PCC.1 near the base, PCC.2 100 cm above; about go cm at Verzy, PVY.z2 at base, PVY.3 40 cm higher, PVY.4 30 cm higher still.

266 MID-TERTIARY CYTHERETTINAE

LONDON Bag! Re.

; -~ oe Cness

<<

RHINE

CHANNEL 7%

CAS, AQUITAINE -BORDEAUX, BASIN

Wie SCALE? (0 50 100 150 200 250 Km. BEE a)

ke C EPS oe ee Palaeogene areas St Geours Zo eis Gaas ~ pe

Fic. 2. Localities sampled for Cytherettinae.

OF NORTH-WEST EUROPE 267

4. Biarritz (RO). The cliff section at Biarritz exposes strata ranging from the Lutetian to the upper

part of the Oligocene. Cytheretta species are described from the following samples, collected at the foot of the cliffs. No detailed section was measured.

RO 264; base of the Couches de 1’Atalaye.

RO 265; entrance to the tunnel at the Musée de la Mer.

RO 266; southern end of the Grande Plage.

RO 267; Rochers de la Villa Eugenie.

RO 268; southern end of the Falaise lou Cout.

RO 269, 270, 271; in ascending order below the Phare St Martin, collected over about 10 m of strata.

5. Other localities in the Aquitaine Basin.

(a) Bartonian clays of Lespontes.

(b) Stampian of Gaas. Two localities were sampled, Espibos (AGE) and Les- barritz (AGL). The former of these is a completely overgrown quarry, where two samples were collected by digging, AGE.1 being slightly lower on the slopes than AGE.2z. At Lesbarritz, AGL.1 was from the cream marls exposed at the base of the section below the coral horizon.

(c) St Geours de Maremne. The controversial locality at Escornbéou, which is of topmost Oligocene or lowermost Miocene age.

6. Headon Beds of the Hampshire Basin.

About 100 samples were examined; all cytherettinid species come from the Middle Headon Beds. Samples mentioned in the text are: Headon Hill, EHH.42, coming 165 cm above the base of the Venus Bed; Whitecliff Bay, EWB(A) from the base of the Brockenhurst Beds, EWB(B) 10 cm higher, EWB.19 from 100 cm above the top of the Barren Sands, and EWB.22 100 cm above EWB.19. Other localities sampled are Colwell Bay, Milford, and Brockenhurst.

7. Other localities.

Bracklesham Beds of Selsey Bill, Whitecliff Bay, and Bramshaw.

Auversian of Auvers-sur-Oise, and Le Ruel.

Stampian of Auvers-St-George, St Cloud, Ormoy, and Morigny in the Paris Basin, and of Weinheim Trift in the Mainz Basin.

Upper Oligocene of Astrup, near Osnabruck.

Ledian of Bambrugge, Belgium.

Ill. STRATIGRAPHICAL DISTRIBUTION

The genus Cytheretta is probably the commonest ostracod genus found in the mid- Tertiary of north-west Europe. It often forms up to 30% of the ostracod fauna, with an average of about 15°%. The earliest species so far known is C. nerva Aposto-

268 MID-TERTIARY CYTHERETTINAE

lescu. This was originally described from the Sables de Bracheux (Thanetian) of the Paris Basin, but has since been recorded as the subspecies C. nerva montensis Marliére from the Montian of Belgium and Dutch Limburg (Marliére, 1958; Deroo, 1966). C. multicostata Apostolescu is also found in the Sables de Bracheux. Other species are known from the London Clay (Ypresian) of the London and Hampshire

C._bernensis C. buttensis . klahni ip AQUITAINE

minipunctata <a C. minor Gaienese C. bullans C.posticalis C. minipustulosa C.regularis C. tenuipunctata C. postornata C. tenuistriata C.sagri : Csti C. triebeli C. samothracia C oligocaenica Somos C. variabilis C.vesca

—

C.rhenana

RUPELIAN

1

S C.headonensis i|. 2) C. porosacostds a Wee N2 C.stigmosa

1

N \ a C.perita

C laticosta (AQUITAINE)

| | | | | | | : | ! z —S L ee | | C costellata | { |

C. bambruggensis | C costellata I C.crassivenia C decipiens C. forticosta ? C. haimeana C. ruelensis

C costellata C. crassivenia Gicceipiene C. grignonensis “C.laticosta” Cc eocaenica

C. haimeana i = — = = pe a ee, a eS eS ee

C.carita | C.cellulosa

C. eocaenica =

| LUTETIAN | AUVERSIAN | BARTONIAN |

C. crassivenia 4 re ; C.laticosta C. scrobiculoplicata

CUISIAN

C. scrobiculoplicata “C.laticosta” ?

YPRESIAN

SPARNACIAN |

C. multicostata | C. nerva |

THANETIAN

._C.montensis 2. \

iG nerva

Paracytheretta

=

L

bs MONTIAN = =TACEOUS|| DANIAN a | | lf | |)

h

Fic. 3. Suggested lineages of Cythevetta in northwest Europe.

OF NORTH-WEST EUROPE 269

basins (Jones 1956; Haskins, 1968; Eagar, 1965) and from the Sables de Cuise of the Paris Basin (Keij, 1957; Apostolescu, 1964). It is clear that Cytheretta is found near the beginning of the Tertiary, and it is necessary to look into the Cretaceous for its ancestry.

The related genus Paracytheretta Triebel ranges from the Senonian to the Palaeocene (Morkhoven, 1963); the type species P. veticosa Triebel was described from the Middle Palaeocene of Denmark. Deroo (1966) has described four genera belonging to the subfamily from the Maastrichtian of Dutch Limburg. These include two new genera which are discussed in more detail below. Puri’s statement (1958) that Cytheretta stemmed trom Paracytheretta late in Palaeocene times therefore needs some modification. It would appear that the genus as now understood probably arose polyphyletically from late Cretaceous forms. Unfortunately Palaeocene and Lower Eocene species are not very well known, nor are Danian relatives, so it is impossible to discuss the origins of Cytheretta in any detail. It is not present in the well studied Maastrichtian faunas, but is in the Montian. The genus probably originated in western Europe, where it is most abundant, but it was already present in the Caribbean region during the Palaeocene (van den Bold, 1957), and also in Greenland (Szcezechura, 1971).

Two distinct lineages can be seen amongst the early forms of Cytheretta (Fig. 3). The first is the superspecies C. laticosta (Reuss) which is found throughout the Eocene. The second is the more complicated group which includes C. nerva, C. scrobiculoplicata (Jones), C. crassivenia Apostolescu, C. costellata (Roemer), C. bambruggensis Keij, C. decipiens Keij, C. grignonensis Apostoiescu and C. haimeana (Bosquet). The last six of these are Middle Eocene species, although C. crassivema is also known from the Sables de Cuise (Lower Eocene). It was during Lutetian times that the group underwent a great explosion, both in numbers and in geo- graphical range. Several are found in the Auversian (Sables d’Auvers, Sables de Beauchamp), but apart from C. costellata did not survive into the Bartonian. The Upper Eocene was essentially a period during which the descendants of the Middle Eocene radiation continued to evolve. Compared with the Middle Eocene, the number of species is poor but individuals are abundant. A third group emerged during the Middle Eocene, exemplified by C. eocaenica Keij. The history of this group is fragmentary. Its ancestors are unknown; it is not found in the Upper Eocene of the region, yet in the Lower Oligocene of Aquitaine C. oligocaenica sp. nov. is found which is so close in morphology as to be separable only with difficulty from C. eocaenica. It is possible that this group was ancestral to the Oligocene groups and to the Miocene and Recent species of the Mediterranean.

There were two distinct provinces in the Anglo-Paris region during the Eocene. The London Clay (Ypresian) of the western part of the London Basin has so far yielded only one common member of the genus, C. scrobiculoplicata, which is often extremely abundant (Eagar, personal communication). Bowen (1953) records one specimen of the C. laticosta group from Enborne, in the south-west London Basin. No younger species are known, however, because of the unfossiliferous nature of the succeeding strata. The London Clay of the Hampshire Basin also contains C. scrobiculoplicata, together with early members of the superspecies C. laticosta

270 MID-TERTIARY CYTHERETTINAE

(Haskins, 1968). For the remainder of the Eocene the superspecies C. laticosta is by far the most abundant Cytheretta. It is joined by C. haimeana, C. costellata, and C. eocaenica during the Middle Eocene, but in the Upper Eocene in most samples examined it is almost the sole representative of the genus. The Paris Basin formed the other province. During the Middle Eocene the C. haimeana group was common, no particular species being predominant; in the Upper Eocene however, C. costellata became completely dominant. In Belgium the picture is not clear, but it is ap- parently similar to the Paris Basin, although C. eocaenica is by far the commonest species in the Sable de Léde (Ledian).

There was a certain amount of communication between these regions, because C. costellata is occasionally found in the Barton Clay (Bartonian) of the Hampshire Basin, while the superspecies C. laticosta is present in the Paris Basin and Belgium. The communication was much clearer during the Middle Eocene than during the Upper Eocene.

The Oligocene saw an almost complete replacement of the Eocene species. New groups such as the superspecies C. tenuipunctata and C. sagri are dominant. The only groups with known Eocene ancestors are C. rhenana Triebel, C. stigmosa Triebel, and C. oligocaenica. The first two are related to species trom the Headon Beds of the Hampshire Basin, C. headonensis Haskins and C. aff. stigmosa Triebel, which in turn are probably related to C. eocaenica. C. oligocaenica is also related to the Middle Eocene C. eocaenica. This reflects the general situation amongst the ostracods, i.e. at the base of the Oligocene a completely new fauna is found in western Europe. Provinces existed, as in the Eocene, but with a different constitu- tion owing to palaeogeographical changes. The Paris Basin, Belgium, Mainz Basin, Rhine Graben and Swiss Basins formed one unit, and Aquitaine another. Within these are found sub-provinces due to geographical separation with the attendant evolution of distinct sub-species.

The genus Flexus Neviani is first reported from the Sables de Cuise (Cuisian) of the Paris Basin (Apostolescu, 1964). Cytheretta decipiens was placed in the genus Flexus by Puri (1958), but tor reasons given below it is here retained in Cytheretta. However, this was probably the ancestor of F. concinnus (Triebel) (Keij, 1957; and see below). An undescribed species of Flexus has been observed in the Lutetian (Fisher Bed VII) of Whitecliff Bay, which is thought to have been derived from an early member of the superspecies C. laticosta. The genus is fairly common, though never abundant, in the Upper Eocene; it is represented by F. solentensis sp. nov. and F. ludensis sp. nov., the origins of which are unknown. In the Oligocene the genus is represented by F. concinnus; the type species F. plicatus (von Munster) comes from the Chattian. As used here, the genus is undoubtedly polyphyletic.

IV. EVOLUTION OF MID-TERTIARY CYTHERETTINAE

Various types of evolutionary pattern were exhibited by the Cytherettinae during Tertiary times. The C. haimeana species group is a good example of cladogenesis as defined by Sylvester-Bradley (1962). Using his terminology (Fig. 4) it is possible to recognize a first period of stabilization in the Palaeocene and Ypresian. During

OF NORTH-WEST EUROPE 271

5 SECOND PERIOD OF STABILITY 4 DIVERSIFICATION

3 ISOLATION AND RETICULATION 2 ERUPTION

] FIRST PERIOD OF STABILITY

MIOCENE

plicatus

gutzwilleri

OE GOCE NE

| Flexus concinnus

‘i

38

Le costellata costellata aff. decipiens antecalva cratis

————

ruelensis

= a in w > 5 =

costellata = =; = grandipora

| | |

Za et Z [e) Re ac =z © 143 oz Lt | mel | od uw ke =) —

| bambruggensis __ aff. coStellata scroviculoplicata costellata

Y PRESIAN

scrobiculoplicata

t

multicostata nerva

montensis

PALAEOCENE

Fic. 4. The evolution of the C. haimeana species group.

272 MID-TERTIARY CYTHERETTINAE

this lengthy period, gradual evolution (stasigenesis) was taking place whereby C. montensis gave rise to C. nerva, which in turn led to C. scrobiculoplicata. In the Upper Ypresian (=Cuisian) a period of eruption began which continued into the Lutetian. During this period seven new species appeared. Perhaps significantly this coincided with a transgression otf the sea, giving large areas of shallow marine waters ideal for Cytheretta, and presumably leading to a reduction in the selection pressure. The Middle Eocene (Lutetian and Auversian) was a time of extreme variation, and by the end of the Auversian most ot the species were extinct. The Bartonian saw a second period of stabilization, with two gecgraphical subspecies of C. costellata, and the descendants ot C. decipiens which were now almost generically distinct. Flexus s.s. appeared in the Oligocene, representing the last members of the group. Final extinction occurred in the Quaternary. Within the species group the evolution of the subspecies of C. costellata shows a pattern very similar to the classic Zaphrentis delanouei as interpreted by Sylvester-Bradley (1951). In the Lutetian, C. costellata costellata consisted almost entirely of one morphotype; great variation occurred in the Auversian, with seven morphotypes, followed by stabilisa- tion in the Bartonian where the subspecies consist mainly of one morphotype in each geographical region.

As opposed to this pattern, the superspecies C. laticosta shows gradual evolutionary change, without any period of explosive evolution.

A second period of eruption ot the Cytherettinae occurred near the base of the Oligocene, once again coinciding with the spread of shallow seas over much of Europe. The superspecies C. tenuipunctata shows great eruption in the Lower Rupelian, with the presence of many geographical subspecies. This was followed by stabilization in the Upper Rupelian and Chattian.

Specific examples of evolution are dealt with in the taxonomic descriptions. These are C. costellata, C. eocaenica, C. laticosta, C. buttensis, C. sagri, and F. solen- tensis.

V. ECOLOGICAL DISTRIBUTION OF RECENT SPECIES

Three species are known from the Mediterranean:

C. judaea (Brady) the type species, is described by Kruit (1955) from the marine, shallow water (5-15 m), sandy deposits of the Rhone delta; Puri, Bonaduce & Malloy (1965) report it being found in association with Posidonia meadows in shallow banks around the islands and peninsulars of the Gulf of Naples, where the salinity is about 38%, and the bottom water temperature 14-15°C; Rome records it from the shallow waters at Monaco (0-3 m) associated with Posidonia; Ascoli (1965) found it in waters up to 74 m deep in the Adriatic Sea. Brady originally recorded it from the coast of Syria.

C. adriatica Ruggieri is apparently more restricted, but no details are available; it is found in beach sands at Rimini associated with C. judaea.

C. belgica (Brady) is reported by Kruit (1955) trom the marine, sandy sediments of the Rhone delta in water of 10-15 m depth. These specimens may however be identical to C. adriatica.

OF NORTH-WEST EUROPE 273

Three species are present around the coasts of southern North America from the Bahamas, Florida, the Gulf coast and New York. These are C. edwardsi (Cushman), C. sahni Puri and C. tracy: Blake. C. sahni is reported by Hulings and Puri (1965) from a sand mud mixture in water less than 65 ft deep and with a salinity greater then 30%,. No details are known for the other two.

In the same region two species of Protocytheretta are present:

P. daniana Brady is reported by Puri and Hulings (1957) from clastic sediments at depths of 36-65 ft, more commonly in the deeper part; salinity, 28-35:6%% 9, clear water, temperature between 10° and 30°C; it is absent in the carbonate province of Florida. Curtis (1960) found it in sands, silts, and clays, particularly in water with an estuarine influence but predominantly marine; temperature 24-25°C. Kornicker (1965) mentions that it is found around the Bahamas.

P. multicarinata Swain was recorded by Swain (1955) from the nearshore Gulf of Mexico.

Cytheretta knysnaensis Benson is recorded by Benson & Maddocks (1964) from the Knysna estuary, South Africa, where it is found in muddy sands free from wave action but with a fair current; salinity is 28-4°%,, pH 8; it is associated with a partly marine and partly estuarine fauna. However, following the description of this species there are reasons for believing it may not be a true Cytheretta; the inner margin, radial pore canals and central muscle scars are different, there is a vestibule and an internal eye sinus, which Cytheretta does not have.

To sum up, Cytheretta is normally found in shallow (10-30 m) nearshore clear waters of normal salinity; the temperature is warm (‘Mediterranean’) with a bottom water temperature of about 15°C; the substratum is a sand or sandy clay, often covered with seaweed. The genus Protocytheretta seems to be able to tolerate slightly brackish conditions with a salinity as low as 28%.

VI. PALAEOECOLOGY

The genus undoubtedly inhabited a similar environment during the Eocene as the recent species. This accounts for its abundance in the Tertiary of the Paris Basin, London Basin, Hampshire Basin, Belgium, Mainz Basin and the Swiss basins. These were all shallow water areas, except for the eastern part of the London Basin where the genus has not been recorded so far. The clear distinction between the London and Hampshire basins on the one hand and the Paris Basin on the other is probably in part geographical and in part related to the contrast between the sedi- ments of the two regions. In the former the ostracods are mainly found in argil- laceous deposits, while they are foundin arenaceous and calcareous depositsin the latter.

The Upper Eocene of Biarritz has yielded only one or two specimens of Cytheretta, while Cytherella and Pontocyprella are extremely abundant. On the other hand, some inland exposures contain fairly abundant Cytheretta. The Oligocene of Biar- ritz also contains abundant Cytheretta. This is undoubtedly ecologically controlled. The Upper Eocene saw deep water at Biarritz which shallowed eastwards, eventually giving way to continental deposits; at the end of the Eocene the water shallowed with a change from argillaceous to arenaceous deposition.

274 MID-TERTIARY CYTHERETTINAE

The Headon Beds of the Hampshire Basin contain sediments deposited in a variety of environments ranging from freshwater to shallow marine. Ostracods are found in most of these environments, but Cythereita is restricted to the marine phases; as soon as brackish conditions prevailed Cytheretta disappeared. Thus it is found in association with such genera as Pterygocythereis, Tvachyleberidea, Bradleya, Leguminocythereis, Brachycythere and Haplocytheridea. ‘Whenever such genera as Neocyprideis or Cytheromorpha become abundant, Cytheretta is no longer present.

VII. GEOGRAPHICAL DISTRIBUTION

Fossil species have so far been reported from the Palaeocene, Eocene, Oligocene, Miocene, Pliocene and Quaternary of Europe; from the Palaeocene, Eocene, Oligo- cene, and Miocene of North America and from the Palaeocene of Greenland (Szcze- chura, 1971); Latham (1938) recorded Cytheretta costellata (Roemer) from the salt range, Punjab, but this is a misidentification and is probably a species of Buntonia. In space Cytheretta would appear to be restricted to the coasts of the North Atlantic and adjacent seas; the case of the South African C. kysnaensis has already been mentioned. Its greatest development is in Europe, where some hundred species, both fossil and recent, have been described.

Flexus has only been reported from western Europe where it ranges from the Lower Eocene to the Quaternary, the latest recorded species being F. tviebeli Ruggieri from the Upper Pliocene and Lower Quaternary of Italy (Ruggieri, 1952).

Protocytheretta is restricted to North America where it ranges from the Oligocene to Recent. Itis found in the Gulf of Mexico, and also off the west coast of California and Mexico (Swain, 1969).

Recently described genera are mainly known from their type areas only. Thus Acuticytheretta and Semicytheretta are only known from western Europe, while Bensonia, Grekoffiana, and Argenticytheretta are only recorded from south America. For a discussion of these see the next section.

VIII. CLASSIFICATION OF THE SUBBAMILY CYTHERETLINAE

Following Hazel (1967) the Cytherettinae are regarded as a subfamily of the Trachyleberididae ; this based on the muscle scars, hinge, and soft parts.

The following genera have been included by various authors within the subfamily: Cytheretta Muller, 1894; Flexus Neviani, 1928 (syn. Eucytheretta Puri, 1958) ; Pseudo- cythereis_ Skogsberg, 1928; Buntonia Howe, 1935; Paracytheretta Triebel, 1941; Loculicytheretta Ruggieri, 1954; Ambocythere Van den Bold, 1957; Protocytheretta Puri, 1958; Netrocytheridea Howe and Laurencich, 1958; Neocytheretta Morkhoven 1963; Acuticytheretta Deroo, 1966; Semicytheretta Deroo, 1966; Bensonia Garcia, 1969; Grekkofiana Garcia, 1969; and Argenticytheretta Garcia, 1969.

Loculicytheretta was placed in the subfamily by both Ruggieri (1954) and Howe (Treatise, 1961), but its distinctive characters would seem to exclude it. Deroo (1966) placed Netrocytheridea into the subfamily ; however its shape and entomodont hinge would appear to exclude it. The only real similarity is in the irregular shape

OF NORTH-WEST EUROPE 275

of the inner margin. Ambocythere was included by Morkhoven (1962, 3) because of its supposed ressemblance to Buntonia; its false radial pore canals, branching radial pore canals, and lateral shape do not support its inclusion. Neocytheretia from Indonesia is not considered to belong to the subfamily. It has eye spots (Cytheretta is blind), a completely different type of ornamentation, and a different hinge; only the irregular inner margin resembles Cytheretta.

Pseudocythereis was placed in the subfamily by Puri; no material has been available for study, only the original description of Skogsberg could be examined. The type species is Cythereis (Pseudocythereis) spinifera Skogsberg, and the author was un- doubtedly comparing it with C. rubra Miller as far as the soft parts were concerned, but the description of the shell is inadequate. Therefore, no conclusion could be reached concerning it, except to agree with Puri that there are strong resemblances to Cytheretta.

Paracytheretta has a very well developed anterior hinge ear in the left valve and no other Cytheretta species approach it in this respect. The hinge is apparently similar to Cytheretta (Morkhoven, 1963), the surface of the valve is reticulate with three longitudinal ridges.

The development of three longitudinal ridges is a common feature of the subfamily, but not all such forms are necessarily closely genetically related. Puri recognized three such genera, Paracytheretta, Eucytheretta, and Protocytheretta. The latter was said to be Cytheretta-shaped, while the others were Cythereis-shaped; the first two were then differentiated on the hinge, which is Cytheretta-like in Eucytheretta and Cythereis-like in Paracytheretta. However, Triebel’s original description of the hinge of Paracytheretta, as shown by Morkhoven (1963) is the same as for Cytheretta. There is, however, no doubt that Paracytheretta is a valid genus because of its distinctive shape. Flexus has a totally different shape, much more like that of Cythereita. Flexus and Protocytheretta can be separated, both on shape and ornamentation. The only species studied that Puri placed into Protocytheretta is P. schoellert Keij from the Oligo-Miocene of Aquitaine; apart from this species the genus would be restricted to North America. A study of descriptions and illustrations of P. daniana, the type species, suggests that P. schoelleri does not belong to the same genus, and therefore, that the genus is probably restricted to North America (see description and discussion in P. schoelleri below).

Flexus has already been shown to be polyphyletic. It is used here for all species shaped like the type species, F. plicatus (von Munster) and having three prominent longitudinal ridges. This is not a very satisfactory arrangement, but present knowledge makes it very difficult to divide it into new genera.

Similarly, no attempt has been made to further subdivide the genus Cytheretia. As mentioned above, there are six distinct groups in the Eocene and Lower Oligocene of western Europe. However, the recognition of such monophyletic groups is difficult to correlate with easily diagnosed morphological characteristics, so it has been thought better for the present to limit their taxonomic recognition to the rank of superspecies. Further investigation, particularly of Lower Tertiary forms, should lead to a phylogenetic division of the genus; this must be combined with a study of the American species, and so is beyond the scope of the present work.

B

276 MID-TERTIARY CYTHERETTINAE

The six groups mentioned above are:

. haimeana group

. laticosta group

. cocaenica group

. tenuipunctata group . sagryt group

. rhenana group

Acuticytheretta certainly has a Cytheretta-like appearance in general outline, has unequal valves and similar muscle scars to Cytheretta. The inner margin is very wide posteriorly, but not irregular; and the hinge is simpler, with a large anterior tooth in the right valve, and a small anterior tooth at the end of the smooth (?) bar of the left valve, and a posterior tooth in the right valve. It could represent a primitive member of the subfamily.

Semicytheretta has a similar hinge and muscle scars to Cytheretta, but has a narrow and regular inner margin and a different shape.

Deroo also describes Cythere euglypha Bosquet and Cythereis euglyphoidea Van Veen, which from the illustrations could also be included in the subfamily. Their shape, unequal valves, muscle scars, and ornamentation suggest this; the hinge has faintly crenulate anterior and posterior teeth in the right valve, and the inner margin is narrow and regular. They are placed in the genus Anticythereis Van den Bold. These could be early primitive members of the subfamily.

Morkhoven (1962, 3) included Buntonia in the subfamily because the first thoracic legs of the male are similar to those of Cytheretta. In general appearance certain Buntonia species can be confused with cytherettinids, although there are noticeable differences in the hinge and the inner margin. It is provisionally included within the subfamily.

Grekoffiana is described as being similar to Protocytheretta, but with a regular inner margin. The type species is G. australis Rossi de Garcia. In the discussion of the new genus (1969; 218), de Garcia placed Protocytheretta daniana (Brady) Benson and Coleman in Grekoffiana. This might be taken to imply the species as interpreted by Benson and Coleman, and not Brady’s species; on page 220 we have mention of ‘Grekofiiana daniana (Brady) dans Benson et Coleman (1963) (p. 26, Tab. 5, Figs 5, 7, 9 & 10)’; again, perhaps implying the species as interpreted by Benson and Coleman, but in this case there should have been some statement to this effect. The idea is thus given that Brady’s species is being included in the new genus, even though it is in fact the type species of Protocytheretta. P.danianais generally thought of as having a typical cytherettinid sinuous inner margin, although following Benson and Coleman it may in fact show considerable variation. Thus two problems arise: firstly, if P. daniana can have a regular or a sinuous inner margin, 1.e., it is an infra- specific character, it can hardly be used to diagnose a new genus; and secondly, the type species of a genus cannot be included in a new genus. There are grounds then for wondering whether Grekoffiana is in fact a valid genus, and not a synonym of Protocytheretta.

Bensonia is described as being similar to Cytheretta, but with a regular inner margin. The hinge and overall appearance certainly look like a typical cytheret-

AANA AO

OF NORTH-WEST EUROPE 277

tinid. Cytheretta knysnaensis is placed in this genus, and as stated previously, there are certain features, particularly the implied presence of an eye, which would exclude it from Cytheretta. The presence or absence of any occular structures cannot be determined from Rossi de Garcia’s description, so it is difficult to know whether Bensomia is a true cytherettinid genus. It is provisionally placed in the subfamily. Argenticytheretta is illustrated with what appears to be an eye tubercle, and in overall shape does not resemble Cytheretta. It also has a regular inner margin. From this it would seem that there exists a group of Cytheretta-like ostracods, principally occurring on the two sides of the south Atlantic, but which differ from true cytherettinids in several respects. Perhaps it is a case of homeomorphy? Until more species are described in detail, this question must remain unanswered. The genera included within the subfamily are:

Cytheretta Flexus (syn. Eucytheretta) Paracytheretta Protocytheretta (syn. Grekoffiana?) Acuticytheretta Semicytheretta

Provisionally: Buntonia Pseudocythereis

Doubtfully: Bensonia

Argenticytheretta

The characters taken to be diagnostic of the Cytherettinae are: inequivalve, the left valve being markedly larger and differently shaped than the right; a modified holamphidont hinge; presence of a fulcral point; numerous and simple radial pore canals; generally sinuous inner margin; unornamented, or with a predominantly longitudinal ornamentation; development of few posterior spines and many anterior denticles ; lack ot occular structures, lateral spines, and caudal process.

IP EPS rn Li SrieUC RUB Om CY ie RE na A

Cytheretta typically has an elongate-ovate shape. The dorsal margin is straight to convex, often with a posterior hinge ear and sometimes with an anterior hinge ear in the left valve. The posterior margin is obliquely curved with a sharp, high postero-dorsal angle and a gentle slope round to the ventral margin. The latter is often concave, always so in the right valve. The valves are usually very unequal in size and shape, the left valve being much larger and with a greater height; in lateral view the outline of the left valve over-reaches that of the right valve in most places. This over-reach is particularly prominent in the postero-dorsal angle where the hinge ear is developed and in the anter-dorsal angle where the antero-dorsal lobe (see below) of the hinge of the left valve rests on the antero-dorsal platform of the right. The shape in dorsal view varies. Sexual dimorphism is distinct, the males being more elongate.

The carapace may be completely unornamented, as in the type species. More usual, however, is an elongate ornamentation which can take the form of either

278 MID-TERTIARY CYTHERETTINAE

longitudinal rows of pits, usually better developed towards the posterior, or longi- tudinal ridges. The latter often have reticulation developed between them, and some ridges converge to form a sub-central plexus (Fig. 5). For descriptive purposes the ridges are numbered from the dorsal margin in a position just to the posterior of the sub-central plexus (Fig. 5). The anterior and antero-dorsal regions of both valves are often smooth, while the rest of the carapace is ornamented. The orna- mentation is generally similar for both valves.

The hinge is modified holamphidont with several accessory elements (Fig. 6, 7; Plates 5, 7,8). In the left valve there is an anterodorsal lobe, which is an outgrowth of the selvage and rests on to an antero-dorsal platform in the right valve. The anterior socket is partially opened ventrally; it is bordered by an antero-ventral lobe which fits into the modified anterior part of the anterior tooth of the right valve, the antero-ventral sinus, which is a hollow in the tooth. The antero-median tooth

719

SS Ste

Fic. 5. C. costellata (Roemer), showing the system of numbering the ridges. X75.

is generally small, the postero-median bar is crenulate and swollen at the posterior to give a postero-median swelling. This is sometimes almost as prominent as the antero-median tooth, as in the type species; it forms one edge of the posterior socket, which is almost open ventrally. The posterior socket often forms a posterior hinge ear, sometimes with a lobe. In the right valve the anterior tooth has an almost vertical face towards the anterior; it is sometimes stirpate (=stepped), or with a gentle posterior slope. In lateral view the tooth has a concavity in the antero- ventral position, the antero-ventral sulcus. The antero-median socket is shallow, the postero-median groove is crenulate, and there is usually a shallow postero- median socket. The posterior tooth is generally pessular (i.e. with more or less parallel sides) and is sometimes almost as prominent as the anterior tooth, as in the type species.

There is considerable variation in the details of the duplicature. The selvage is generally more prominent in the right valve. At the anterior it usually has a position close to the outer margin, but in some species it is further away so that a prominent flange groove is formed; there is a wide flange groove developed along the ventral margin of the right valve, with a narrow one along the posterior. The selvage is

OF NORTH-WEST EUROPE 279

Py packs: I vl. HINGE IN DORSAL VIEW ; Postero-median Y> Antero- median Postero -median bar tooth

swelling

Posterior socket eee Anterior radia Radial pore canals pore canals

dm. Dorsal muscle scars Selvage am. Adductor muscle scars fm, Frontal muscle scar f. Fulcral point Left Valve as. Anterior socket avi. Antero-ventral lobe np. Normal pore canals

Fic. 6. Internal structure of the left valve of Cytheretia.

HINGE IN DORSAL VIEW

Antero- Postero- = Postero- ~Posterior Antero-dorsal Anterior median median median tooth platform tooth socket groove socket Fl Bae eoeove Antero-ventral sinus I Anterior re : indentation Ventral indentation | Posterior Posterior een Bedian \ ape ment indentation \segment | segmen of inner Flange g Crlvage g ewe i

Flange

Flange groove

Right Valve

Fic. 7. Internal structure of the right valve of Cytheretta.

280 MID-TERTIARY CYTHERETTINAE

sometimes very strongly developed at the posterior. The flange usually forms the outer margin. A weak list is often present.

The outline of the inner margin is a characteristic feature; it is broad with three principal indentations (Fig. 7); its outline can be divided into three segments for descriptive purposes. It ends against the anterior and posterior hinge elements. The anterior and ventral radial pore canals are long, simple, often crossing one another, usually bulbous near the outer margin (Pl. 1). The anterior pore canals tend to be concentrated around the antero-ventral angle. The posterior radial pore canals are rather different (Pl. 1, Fig. 9); the inner lamella in which they are situated is not fused to the outer lamella, so that a vestibule is in fact formed. The posterior radial pore canals are numerous, straight and very closely packed; often they form two groups, a ventral one of short pore canals similar to the anterior and ventral radial pore canals, and a dorsal one of closely packed pore canals.

The central group of muscle scars consists of a row of four adductors; the lower two often touch and the third one often elongate; and a large irregularly “U’-shaped frontal scar with a smaller scar in front. A fulcral point is prominent, forming a raised boss; this feature is also a well known characteristic of the Cytherideinae (von Morkhoven, 1963). A dorsal group of about three muscle scars can be detected. There are two mandibular scars which almost touch, or one elongate scar near, and just anterior to, the ventral indentation (Pl. 1, Fig. 7).

There is no eye spot or occular sinus, as Cytheretta is blind. The feature often described as an eye spot is the antero-dorsal lobe of the hinge of the left valve.

x, SORT PARTS

The soft parts have not been examined during this study, and the only published descriptions are for C. rubra, C. edwardsi (Cushman) 1906 and C. tracyi Blake, 1929 (Hazel, 1967). The principal characters ‘are the three-jointed exopodite of the second antennae, absence of a seta on the posterior border of the protopodite of the first thoracic legs’ (Hazel, 1967; 40, after Miller, 1894) and an asymmetry of the first pair of thoracic legs in the male which is developed to varying degrees.

XI. LARVAL STAGES

Only the last three larval stages have so far been recognized. They tend to be more triangular in shape, with a pointed posterior. Ornamentation in the last moult is similar to that of the adult, but in earlier moult stages is only partially developed. The hinge is very simple; in the left valve the antero-dorsal lobe is developed, followed by an anterior socket open ventrally, smooth bar and posterior socket, also opened ventrally; the corresponding features are present in the right valve.

About one in fifty of the adult specimens appears to be weakly calcified. The hinge is an adult hinge, but weak; the inner margin is straight, without the normal indentations, and very narrow. Pokorny (1965) suggested that such a feature might be due to a parasitic infection and Morkhoven (1963) that the animal died soon after moulting and before calcification was complete. Oertli (1965) mentions the case of

OF NORTH-WEST EUROPE 281

Falumia where the true adult form is very rare, and the thinly calcified moult stage was probably mature, i.e. a case of neotony. Sandberg (1965) supported this by citing cases where such immature forms contained eggs. However, in the case of Cytheretta Morkhoven’s explanation is probably the correct one.

In two species, C. eocaenica and C. oligocaenica, a few very large specimens are found. They are about double the size of the normal adult, but otherwise identical. This is a fairly frequent phenomenon in several families of the Ostracoda. Benson (1965) suggests that such forms might be a gerontic stage; and Kesling in the Treatise (1961) considers that they represent post-maturation instars.

SITs ARIEhd; SIRT IVI Salix Oed, (ON 1a IT>,<10/s)

Internal details are the same as for Cytheretta. As defined here, all Cytheretta-like forms with three prominent longitudinal ridges are referred to this genus.

SO eh sh Vere Shel s Ol (OY irr Reo WA

Miller described Cytheretta in 1894 with C. rubra Miller as type species. In 1950 Ruggieri put C. rubra into the synonymy of C. subradiosa (Roemer), originally described from the Pliocene of Italy. This was accepted by Triebel (1952), Puri (x958) and Hazel (1967), but queried by Morkhoven (1963). Thanks to Prof. Ruggieri, who kindly supplied the material, it has been possible to study both forms. Morkhoven also mentioned that C. rubra might be a synonym of I/yobates? judaea Brady, 1868; Hazel accepted this view, but places both C. rubra and C. judaea within the synonymy of C. subradiosa. Fortunately Brady’s material is preserved in the collections of the Hancock Museum, No. B67, presently under the care of Dr. K. McKenzie. There is no doubt that C. judaeais the senior synonym. The problem now is whether this is synonymous with C. subradiosa.

C. subradiosa was described from the Pliocene of northern Italy at Castellarquato ; none of Roemer’s material is available for study, but Ruggieri has obtained topotype material. The specimens described below come from the Pliocene of San Arcangelo, near Rimini. Miller’s specimens were from the Gulf of Naples; the specimens des- cribed below come from the beach sand at Rimini.

A comparison of the Pliocene specimens with the recent has shown a number of differences which are believed to distinguish C. judaea from C. subradiosa. The position of the Quaternary forms is unknown.

The major difference is in the line of the inner margin, particularly the posterior segment. In C. judaea this cannot be separated from the median segment, but in C. subradiosa it is steep, giving a narrow and deep posterior indentation ; the posterior indentation in C. judaea is very broad. There are more radial pore canals in C. subradiosa, 32 compared with 24 in C.judaea. The central muscle scars are situated more to the dorsal in C. subradiosa; the four adductors are larger and the frontal is more dorsally situated in respect to the adductors. There is a slight difference in shape in that C. judaea is more obliquely rounded in the anterior margin. The hinge is almost identical although the anterior tooth of the right valve is slightly more prominent in C. judaea.

282 MID-TERTIARY CYTHERETTINAE

Finally, Muller included two distinct species in his description of C. rubra (see below). One is probably synonymous with C. judaea and the other with Cytheridea striatopunctata Terquem 1878.

XIV. SYSTEMATIC DESCRIPTIONS Subclass OSTRACODA Latreille, 1806 Order PODOCOPIDA Miller, 1894 Suborder PODOCOPINA Sars, 1866 Family TRACHYLEBERIDIDAE Sylvester-Bradley, 1948 Subfamily CYTHERETTINAE Triebel, 1952

Genus CYTHERETTA Miller, 1894

1894 Cytheretta Miller, p. 382

1906 Pseudocythevetta Cushman, p. 382 1928 Cylindrus Neviani, p. 106

1941 Prionocytheretta Mehes, p. 60

Dracnosis: The carapace is elongate-ovate, with a gently curving postero-ventral angle so that the posterior margin is markedly asymmetrical. Generally very inequivalve, with a large left valve. Ornamentation consists of longitudinal ridges or rows of pits, or the carapace may besmooth. The hinge is modified holamphidont. The line of inner margin is sinuous; the duplicature is wide; selvage, flange, and list are developed to varying extents; the radial pore canals are simple, long, bulbous, and curved; the normal pore canals are simple. The central muscle scars consist of a row of four adductors and a large irregularly ‘U’-shaped frontal; a fulcral point is well developed.

TYPE SPECIES: Ilyobates? judaea Brady.

Discussion: The type species of Cylindrus is C. jurinei (von Munster) which is a true Cytheretta; the name Cylindrus is also preoccupied (Fitzinger, 1833, Mollusca). Neither of the other two have been studied; Puri (1958) states that the shell structure of Pseudocytheretta is identical with that of Cytheretta, and Hazel (1967) discusses the genus in some detail and concludes that it is synonymous with Cytheretta. Triebel (1952) places Prionocytheretta in the synonymy of Cytheretta.

OF NORTH-WEST EUROPE 283

Cytheretta judaea (Brady) (Pl. 1, fig. 1, 4-7, 9; Text Fig. 8)

1868 Ilyobates ? judaea Brady: 112, pl. 13, figs 17, 18.

1894 Cytheretta vubra Miller: 382, pl. 8, figs. 9, 10, 16; pl. 39, figs 8-22, 24. 1912 Cytheretta vubya Miller; Miiller: 366.

1950 Cytheretta subyadiosa (non. Roemer; pars); Ruggieri: 9.

1953 Cytheretta subvadiosa (non. Roemer; pars); Ruggieri: 102.

1955 Cytheretta vubya Miller; Kruit: 482, pl. 5, figs 9a—c.

1958 Cytheretta subyadiosa (non. Roemer; pars); Puri: 186, pl. 1, figs 3-7. 1959 Cytheretta subvadiosa (non. Roemer; pars); Ruggieri: 190.

Diacnosis: A species of Cytheretta with sub-parallel dorsal and ventral margins and unornamented surface. The inner margin is almost straight.

MATERIAL: Thirteen valves and carapaces from the beach sand at Rimini; Io 3792-4; 3810. TYPE LOCALITY AND HORIZON: Haifa, Recent.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Recent—Rhone delta, Monaco, Naples, Adriatic Sea, Syria, Aegean Sea; Quaternary of Italy?

DESCRIPTION: Sexual dimorphism is not strong; the males are more elongate. Sex ratio 1:5. The left valve has a weak posterior hinge ear; the dorsal margin is slightly convex; the anterior margin is strongly obliquely rounded. The ventral margin is almost straight and roughly parallel to the dorsal margin; the posterior margin is obliquely rounded. The greatest height is just to the posterior of centre. The ventral margin of the right valve is concave. In dorsal view the carapace is ovate, tapered towards the anterior.

The shell is smooth, but with conspicuous opaque areas; there is a large one in the central region and a smaller one in the anterior. These are approximately constant in position in all specimens, producing the black areas seen in PI. 1, fig. 4.

The antero-dorsal lobe of the hinge of the left valve is flat; the postero-median swelling is pronounced. In the right valve the posterior tooth is almost equal in prominence to the anterior tooth.

Fic. 8. Cytheretta judaea (Brady); left valve; x75.

284 MID-TERTIARY CYTHERETTINAE

The anterior indentation of the inner margin is wide and downward pointing; the anterior segment is gently curved; the median and posterior segments cannot be separated; the posterior segment is large and broadly rounded. There are some 24 anterior, 27 ventral, and 32 posterior radial pore canals. The posterior set can be divided into two groups; a ventral group of ten, widely spaced and bulbous, and a dorsal group of 22, very close together and straight. There are some 75 scattered normal pore canals. The selvage is close to the outer margin, with a flange groove along the ventral margin.

DIMENSIONS: Left valve Right valve Carapace ie H L/H i H L/H Ww Female 0-75 0°43 174) 0-75 0:40 1°88 0°42 Male 0-79 0°43 1:84 0-78 0-39 2°00 0-40

Discussion: Miller (pl. 8, fig. 13) and Puri (pl. 1, figs 9-13) illustrate large speci- mens with longitudinal rows of puncta. These must represent a different species; as well as the differences already mentioned, the anterior margin is much squarer than the obliquely rounded C. judaea. Triebel (1952; 17) suggests that these belong to Cytheridea striatopunctata Terquem, described by Terquem (1878) from the Pliocene of Rhodes. Several specimens of this type were found in the sample from Rimini, and were first thought to belong to C. adviatica Ruggieri; however, none have such strong longitudinal ridges as those illustrated by Puri (1958, pl. 2, figs 1-5). They are perhaps members of a gradational series in which C. adrviatica represents the mor- photype with the strongest developed ornamentation.

See also the generic discussion (p. 281).

Cytheretta subradiosa (Roemer) (Pir, fig? 8)

1838 Cytherina subyadiosa Roemer: 517, pl. vi, fig. 20.

1880 Cytherella calabra Seguenza: 326, 366, pl. 17, fig. 56.

1900 Cytheridea subyadiosa (Roemer); Namias: 105, pl. 15, fig. 17. 1905 Batrdia subyadiosa (Roemer); Cappelli: 306, pl. 9, fig. 4. 1928a Cythevidea subyadiosa (Roemer) ; Neviani : 66.

1928b Cytheridea subradiosa (Roemer); Neviani: 131.

1950 Cytheretta subvadiosa (Roemer); Ruggieri; 9-11 (pars).

1953 Cytheretta subvadiosa (Roemer); Ruggieri: 102 (pars)

1959 Cytheretta subradiosa (Roemer); Ruggieri: 190 (pars).

DiaGnosis: An unornamented species of Cytheretta with sub-parallel dorsal and ventral margins. The inner margin has a steep posterior segment.

MATERIAL: Five carapaces and valves, together with several fragments, from the Lower Pliocene of San Arcangelo, near Rimini. Io 3795.

TYPE LOCALITY AND HORIZON: Castellarquato; Pliocene.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Pliocene of Sicily, Calabria, Castel- larquato, and Rimini; Quaternary of Italy?

OF NORTH-WEST EUROPE 285

DEscrRIPTION: The left valve has a very weak posterior hinge ear; the dorsal margin is almost straight, although in some specimens there is a distinct convexity in the central region which interrupts the smoothness of the margin. The anterior margin is obliquely rounded. The ventral margin is straight in the presumed females, but concave in the one specimen thought to be a male. The posterior margin is broadly rounded. The right valve has a greater height towards the posterior; the ventral margin is concave. The carapace is ovate in dorsal view.

The shell is smooth, with no ornamentation.

The hinge of the left valve has a flat antero-dorsal lobe; the antero-ventral lobe is weak; the postero-median swelling is pronounced and equal in size to the antero- median tooth. In the right valve the posterior and anterior teeth are equal in size.

The anterior indentation of the inner margin is wide and downward pointing; the ventral indentation has a flat base instead of the usual ‘V’ shape; the posterior indentation is narrow, deep and curved. The anterior segment is almost straight; the median segment is slightly irregular, with a gentle upward slope towards the posterior; the posterior segment is very steep. There are some 32 anterior, 34 ventral, and 32 posterior radial pore canals; and 25 scattered normal pore canals. The selvage runs very close to the margin; there is a small flange groove along the ventral margin.

DIMENSIONS: Carapace

Bb H Ww L/H Female 0°75 0°45 0:37 1-67 Male 0-81 0°44 0:36 1°84

Discussion: From the few specimens studied, there would appear to be consider- able variation in shape. See also the generic discussion (p. 281).

Cytheretta costellata (Roemer)

DiacGnosis: A species of Cytheretta with a characteristic ornamentation of ten longitudinal ridges which unite and bifurcate in a regular pattern; posterior margin with four strong spines, anterior margin with some nine denticles. Strongly inequivalve. Duplicature with an anterior flange groove and a ventral list.

DESCRIPTION: Seven morphotypes have been recognized, chiefly on ornamentation. These are described below.

The ornamentation consists of ten longitudinal ridges with varying ornamentation between (see Text-fig. 5). The second ridge bifurcates towards the posterior and the lower part eventually joins ridge no. 3; from the junction a thin ridge runs into no. 4. Ridge no. 4 is prominent, and forms the dorsal limit to the subcentral plexus towards the anterior. No. 5 forms the ventral limit and again is prominent; in the posterior of the valve it joins no. 7. No. 6 appears to be enclosed between 5 and 7, and joins one of these two. No. 7 is very prominent and extends further towards the posterior than any other; 8 and g join together in a position level with the sub-central plexus, 9 bifurcating just to the posterior. An area of reticulation of varying size is present at the anterior.

286 MID-TERTIARY CYTHERETIINAE

The antero-dorsal lobe of the hinge of the left valve is swollen; the antero-ventral lobe is poorly developed; the antero-median tooth is prominent and downward pointing; the postero-median swelling is pronounced but small in comparison with the antero-median tooth; the posterior socket is circular rather than ovate.

There are some 30 anterior pore canals, about 30 very closely packed posterior radial pore canals, and 15 ventral radial pore canals. The selvage is prominent; a flange groove is present, particularly prominent in the right valve, both anterior and posterior. Four spines come from the flange at the posterior and some g denticles are developed along an anterior fringe. A faint list is developed, particularly along the ventral side.

The central muscle scars are in a slight pit; they are small, the two ventral scars close together, the third elongate, and the dorsal-most scar circular.

Sexual dimorphism is pronounced; sex ratio 1-75.

Discussion (I): Roemer (1838) originally described the species from the Tertiary of the Paris Basin; the specimen illustrated has six ridges and four or five posterior spines. Its shape is similar to the Lutetian forms of the species to be described below. Bosquet (1852) recorded it from the Sables inférieurs, Calcaire grossier, and the sables moyens; it was commonest in the Calcaire grossier. His diagram shows eight ridges and four spines. Keij (1957) redetermined Bosquet’s material and records several species and genera amongst it. Roemer’s material is thought to be lost and Keij’s revision of Bosquet’s material has been followed in the interpretation of the species. There is, therefore, still some doubt about the identity of Roemer’s

MORPHOTYPES | suBsrecies SUB SPECIES A leilptly Ci | Dial Mee alee Co cumeneoutos [|= [|= [= [wo] = cratis A CS

LOCALITY

enue [=| | = [me eae faa woweites se) [se [v9 [77 [64/86] — | | erensoe covers ao) |esa|wo [s0 [isa] m0] - | seacncesnanca [wo | = [= [= [= [=f costellata pewwenven feo l= [= T= [=P [J

Fic. 9. Morphotypes and subspecies of C. costellata (Roemer). The numbers in brackets refer to the number of specimens examined.

OF NORTH-WEST EUROPE 287

species. Bosquet (1852) mentions Roemer in his acknowledgments for supplying material however. Jones and Sherborn’s new variety C. costellata var. triangulata from the type Bracklesham Beds is a female dimorph.

Amongst the material studied it is possible to recognize seven morphotypes which fall into four groupings, here recognized as subspecies. Three of these are chrono- logical subspecies and the fourth is geographical. (Text-fig. 9). For a discussion of the subspecies as used here see Sylvester-Bradley 1951.

Morphotype A:

The left valve, particularly of the female, is triangular with its apex at the ant- erior; the dorsal margin is gently convex with a prominent posterior hinge ear; the ventral margin is slightly convex, sweeping round at the posterior in a continuous curve to form the ventral part of the posterior margin; the dorsal part of the posterior margin turns sharply round to meet the hinge ear. The anterior margin is obliquely rounded. The right valve is much lower than the left with a strongly convex dorsal margin and a concave ventral margin. In dorsal view the carapace is ovate.

The ornamentation between the ridges consists of a fine punctation. In the anterior part of the valve the ridges tend to become submerged within a large area of reticulation. A strong marginal rim runs around the anterior and dorsal margins. The sub-central plexus is simple. Ridge no. 6 is joined at both ends to no. 7 by a fine thread-like continuation.

The outline of the inner margin differs slightly between the two valves. The anterior indentation is small and pointed; the anterior segment slopes down quite steeply to the small and ill-defined ventralindentation. In the left valve the median segment is at first level and then slopes upwards to the posterior segment, where the slope steepens markedly and sweeps up close to the dorsal margin; then it curves sharply downwards to form a deep and narrow posterior indentation. In the night valve the median segment is entirely level. In the males the posterior segment reaches much closer to the ventral margin.

Morphotype B:

Similar in most respects to Morphotype A. However, a very strong and character- istic sub-central plexus is developed, especially well seen in a specimen from Bam- brugge (Pl. 2, fig. 5). It consists of a raised platform between ridges 4 and 5 with three pits on its surface. This is a character developed in all the remaining mor- photypes and is generally very prominent. The shape is slightly different to Morpho- type A as the ventral margin is more convex and so the carapace appears to be much less triangular.

Morphotype C:

The shape of this and the remaining morphotypes is similar to that of Morphotype B but with the development of an anterior hinge ear in the left valve.

The ornamentation between the ridges consists of a punctation intermediate in size between that of Morphotype A and that of Morphotype F. This and the

288 MID-TERTIARY CYTHERETTINAE

remaining morphotypes show a slight difference in ornamentation between the two valves. In the left valve ridge no. 6, joins No. 7 in the anterior but ends without joining either 5 or 7 at the posterior; in the right valve of some specimens ridge no. 7 bifurcates near the posterior margin, the dorsal branch is joined by no. 5 and the ventral branch joins no. 8. There is a prominent sub-central plexus. The anterior area of reticulation is much reduced compared with Morphotypes A and B.

Morphotype D:

Large pits are developed between the longitudinal ridges, but can only be seen by straining. The pits are not developed in the posterior regions and cover a larger area of the valve in some specimens than in others. Other features are similar to Morphotype E.

Morphotype E:

This has a strong development of large pits between the ridges. The large pits are two abreast between the median portions of ridges no. 3 and 4, and in this respect Morphotype E differs from Morphotype F, which has a network of small pits. Ridges 5 and 7 join in the posterior, leaving an unattached no. 6 between them. The anterior area of reticulation is smaller than in Morphotype A, but larger than in Morphotype F.

Morphotype F:

The anterior margin is evenly rounded and in this respect differs from Morphotypes A-E. Pitting is developed between the ridges; the pits are large between ridges 3-7, but form a network of small pits between 1, 2, and 3. The anterior area of reticulation is narrow, only two reticules wide; this results in the longitudinal ridges being more prominent in the anterior region than in Morphotypes A-E. The sub- central plexus is prominent. In the left valve ridge no. 5 is the strongest at the posterior and 6 and 7 join it; a few examples can be seen where no. 5 joins no. 7, and a few in which the ridges do not join at all. In the right valve no. 7 bifurcates near the posterior, the dorsal portion usually joining 5 and the ventral portion joining 8. This pattern is also present in a few left valves.

The inner margin differs from that of Morphotypes A-E; the posterior segment is not so steep and does not reach so far towards the dorsal margin; the posterior indentation is therefore not so deep.

Morphotype G:

This is similar to Morphotype F except for the development of a smooth area at the anterior. This varies in extent ; in some specimens it is small, in others it extends almost to the sub-central plexus. Both males and females are affected. In the left valve ridge no. 5 is the most prominent at the posterior; no. 6 joins 5 or 7 or neither. The right valve is similar to that of Morphotype F.

OF NORTH-WEST EUROPE 289

Cytheretta costellata costellata (Roemer) (Pl. 2, fig. 1-10)

1838 Cytherina costellata Roemer: 517, pl. 6, fig. 24.

1852 Cythere costellata (Roemer); Bosquet: 58 (pars), pl. 2, fig. 11.

1857 Cythere costellata (Roemer); Jones: 32, pl. 5, fig. 11.

1874 Cythere costellata (Roemer); Brady, Crosskey and Robertson: 152, pl. 16, figs 13-15. 1889 Cythere costellata (Roemer) var. tviangulata Jones and Sherborn: 30, pl. 1, fig. 21. 1955 Cytheretta costellata (Roemer); Apostolescu: 26, pl. 5, figs 75, 76.

1957 Cytheretta costellata (Roemer); Keij: 132, pl. 22, fig. 7.

DiaGnosis: A subspecies of C. costellata with fine puncta between the ridges and usually with a simple sub-central plexus.

MATERIAL: Material was examined from the following localities; Lutetian IV of Damery, Fisher Beds 21-24 of Selsey, the Keij Collection at Utrecht (Belgian Lutetian and Ledian), Ledian of Bambrugge. For numbers see Text-fig. 9. Io 3796-3803.

TYPE LOCALITY AND HORIZON: Lutetian of the Paris Basin.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: See Text-fig. Io.

DEscripPTion: This is composed of Morphotypes A, B, and C. It is small in size (see Text-fig. 1r and Discussion II).

REMARKS: The Ledian specimens at Utrecht have not been included in Text-fig. 9 because these were only measured and not divided into morphotypes. Most of them are Morphotype A; a few show very faint pitting in places and are therefore Mor- photype D. The specimens mentioned by Keij (1957; 133) as being reticulate are not C. costellata.

Cytheretta costellata grandipora subsp. nov. (Pl. 3, figs 1-8)

1852 Cythere costellata; Bosquet: 58 (pars). 1957 Cytheretta costellata (Roemer); Keij: 132 (pars), pl. 6, fig. 9.

DERIVATION OF NAME: Latin—grandis, large, and pora, pit; refers to the large pits developed between the ridges.

D1acnosis: A subspecies of C. costellata characterized by the presence of large pits between the longitudinal ridges with a well developed sub-central plexus.

HoLotyPeE: Io 3804, a female left valve.

PARATYPES: Io 3805, 3806, 3808-9, 3811-13.

MATERIAL: See Text-fig. 9.

TYPE LOCALITY AND HORIZON: Carriere de Moiselles, Sables de Beauchamp.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Sables d’Auvers, Auvers-sur-Oise ; Sables de Beauchamp, Moiselles; Sables de Cresnes, Le Ruel. All are in the Paris Basin.

290 MID-TERTIARY CYTHERETTINAE

DESCRIPTION: This consists of Morphotypes A-G, but principally A-E; thus there is a great range in ornamentation and size. For size, see Text-figs 9 and Io.

Cytheretta costellata cratis subsp. nov. (Pl. 4, figs 1-8; Pl. 5, figs 1-3. Text-fig. 5) DERIVATION OF NAME: Latin—cratis, wickerwork; refers to the ornamentation.

Diacnosis: A subspecies of C. costellata characterized by the development of pits between the ridges and a narrow area of reticulation at the anterior.

HoLotyPe: Io 3814, a female left valve.

PARATYPES: Io 3815-3822.

MATERIAL: See Text-fig. 9.

TYPE LOCALITY AND HORIZON: Verzy; Marnes a P. ludensis.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Marnes a P. ludensis of Verzy and Chavencon in the Paris Basin.

DEscRIPTION: This consists entirely of Morphotype F. For dimensions see Text- fig. Il.

ENGLAND PARIS BASIN BELGIUM

costellata antecalva costellata cratis MIDDLE & UPPER BARTON BEDS MARNES A P.ludensis

costellata_grandipora SABLES MOYENS

costellata costellata CALCAIRE GROSSIER

UPPER BRACKLESHAM BEDS SABLES DE LEDE

Fic. 10. Stratigraphical and geographical distribution of the subspecies of C. costellata (Roemer).

OF NORTH-WEST EUROPE 291

Cytheretta costellata antecalva subsp. nov. (Pl. 4, figs 9-11) 1968 Cytheretta costellata (Roemer); Haskins: 165; Pl. 2, fig. 1-8.

DERIVATION OF NAME: Latin—ante, anterior, and calvus, smooth; refers to the smooth anterior region.

Diacnosis: A subspecies of C. costellata characterized by a smooth area at the anterior; pits are present between the longitudinal ridges.

HototyPeE: Io 3823, a female left valve. PARATYPES: Io 3824-6.

045

Height

e a (e@ \

©

a 7 \l iK b

Aoi i

040

KEY o Bracklesham--—— @ Damery............ © Belgium —.—.—. ee

0:35 0-60 0-70 0-80 0-90mm Length ———

Fic. 11. Size distribution of C. costellata (Roemer). PVY=Verzy; PCC=Chavencon; PMS = Moiselles.

292 MID-TERTIARY CYTHERETTINAE

TYPE LOCALITY AND HORIZON: Barton; Upper Barton Beds, Bed H (Chama Bed).

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Middle and Upper Barton Beds, Barton; Middle Barton Beds, Alum Bay.

DESCRIPTION: This consists of Morphotypes F and G, but mainly the latter. DIMENSIONS: Holotype, L=o-77 mm.

REMARKS: Although the number of specimens is small, a further set of specimens was described by Haskins (1968) from Alum Bay which also had a smooth anterior region, so that this is apparently a well marked character.

Discussion (II): Each assemblage studied shows a range of variation, indicated rather crudely by the proportion of morphotypes it contains. In fact gradations between morphotypes occur and the division into discrete groups is therefore arbitrary. The differences between subspecies are marked by changes in the proportion of the constituent morphotypes. Most of these changes can be cor- related with stratigraphical position, but in the case of C. costellata cratis subsp. nov. and C. costellata antecalva subsp. nov. it seems probable that we are dealing with contemporary subspecies that are geographically separated. The suggested relation- ships between the subspecies of C. costellata are shown in Text-fig. Io.

In general there is an increase in size with time. This is clearly seen in Text-figs rr & 12. A comparison between the morphotypes found at Auvers and Moiselles (Text-fig. 12) shows that Morphotype A tends to be the smallest and E the largest.

a ae eer a

045+

0-40 S28

Morphotypes

0-65 070 0-75 0-80 0-85 0-90

Fic. 12. Size distribution of morphotypes of C. costellata (Roemer) from Auvers-sur-Oise.

OF NORTH-WEST EUROPE 293

This is much clearer in the case of the females than the males. The species from Verzy are smaller than those from Chavengon; as these are of the same age, and because most other ostracods are similarly affected, this is thought to be ecologically controlled. The specimens from the Belgian Sable de Léde are much larger than those from Damery and Bracklesham.

The ridge pattern remains remarkably constant, apart from the posterior involve- ment between ridges nos. 5,6and 7. With time there is an increase in complication of the ornamentation between the ridges and a decrease in the width of the anterior area of reticulation. At the posterior ridge no. 7 is at first the most prominent and no. 5 and 6 join it; this changes until no. 5 is the strongest and no. 6 and 7 join it (Text-fig. 13).

The anterior margin changes from obliquely rounded to evenly rounded. The Ludian forms are proportionally higher than the Lutetian ones.

6 See ei 7 shila

Fic. 13. Changes in ridges 5, 6 and 7 of C. costellata (Roemer).

Cytheretta bambruggensis Keij

(Pl. 3, fig. 9) 1957 Cytheretta bambruggensis Keij: 131, pl. 6, fig. 10, pl. ro, figs 9-11.

MATERIAL: One right valve from the Sables de Léde of Bambrugge. Io 3827. TYPE LOCALITY AND HORIZON: Bambrugge (Belgium) ; Sables de Lede.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Lutetian, Ledian and basal Sables de Wemmel, Belgium; ‘Bartonian’ of the borehole at Delden, the Netherlands.

Cytheretta crassivenia Apostolescu (Pl. 6, figs 2, 5)

1852 Cythere costellata (non Roemer; pars); Bosquet: 58. 1955 Cytheretta crassivenia Apostolescu: 261, pl. 5, figs 77-79. 1957 Cytheretta cvassivenia Apostolescu; Keij: pl. 6, fig. 4; pl. 10, figs 12-14.

MATERIAL: Two valves from the Lutetian of Grignon; nine valves and carapaces from the Lutetian of Damery; two carapaces from the type Sables d’Auvers; eleven valves and carapaces from the Sables de Beauchamp of Moiselles. Io 3828, 9.

TYPE LOCALITY AND HORIZON: Montmirail; Lutetian.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Lutetian, Sables d’Auvers of the type locality, Sables de Beauchamp of Moiselles in the Paris Basin; Sables de Léde of Belgium.

294 MID-TERTIARY CYTHERETTINAE

Discussion: Ridge no. 4 is strong and runs from the anterior to the posterior; beneath it are two short, but conspicuous ridges. In the specimens from Grignon and Damery (Lutetian) these two ridges do not quite join, the anterior one finishing just above the anterior end of the posterior ridge; in the specimens from Moiselles, however, they actually join. The reticulation between the ridges tends to be slightly stronger in the Moiselles specimens than in the Lutetian ones.

Cytheretta decipiens Keij (Pl. 6, figs 8—r0)

1955 Cytheretta concinna (non. Triebel) Apostolescu: 261, pl. 4, figs 72-74. 1957 Cytheretta decipiens Keij: 133, pl. 6, fig. 8, pl. ro, figs 15-16.

MATERIAL: One carapace from the type Sables d’Auvers, five valves and carapaces from the Sables de Beauchamp of Moiselles. Io 3830-32.

TYPE LOCALITY AND HORIZON: Forest (Belgium), Cuisian.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Cuisian, Lutetian and Ledian of Belgium; Lutetian, Sables d’Auvers, and Sables de Beauchamp of the Paris Basin.

Discussion: It has been decided to retain this in the genus Cytheretta rather than Flexus because of the presence of several minor longitudinal ridges, apart from the three main ones. The ridge pattern is, in fact, very similar to that of C. crassivema and also to C. haimeana, C. scrobiculoplicata and C. bambruggensis. If Keij’s sup- position is correct, i.e. C. decipiens was the ancestor of Flexus concinnus (Triebel), at least some part of the genus Flexus has clearly been derived from Middle Eocene Cytheretta. The transition can be closely placed to the Bartonian.

Cytheretta aff. decipiens Keij (Pl. 6, fig. 3)

MATERIAL: One carapace from the Marnes a P. ludensis at Chavencon. Io 3833.

Dimensions: L, 0-66; H, 0:34; L/H, 1-94.

Discussion: This is clearly related to C. decipiens, but has lost most of the minor longitudinal ridges; there are still several ventral ridges however. This is probably the form called C. concinnus by Keij, but it differs from the latter in several respects. However, it could be placed in an evolutionary sequence, which, starting with

C. decipiens, eventually gave rise to F. concinnus. As only one specimen was avail- able for study, this question has been left open.

OF NORTH-WEST EUROPE 295

Cytheretta haimeana (Bosquet) (Pl. 6, figs 1, 6)

1852 Cythere haimeana Bosquet: 61, pl. 2, fig. 14.

1852 Cytheve costellata (non. Roemer) Bosquet (pars): 58.

1955 Cytheretta haimeana (Bosquet); Apostolescu: 262, pl. 5, figs 84-85. 1957 Cytheretta haimeana (Bosquet); Keij: 136, pl. 6, fig. 7, pl. ro, figs 7, 8.

MATERIAL: One valve from the Lutetian of Damery; ten valves and carapaces from the type Upper Bracklesham Beds; five valves and carapaces from the type Sables d’Auvers; eight valves and carapaces from the Sables de Beauchamp of Moiselles. Io 3834-5.

TYPE LOCALITY AND HORIZON: Grignon; Lutetian IV.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Cuisian and Lutetian of the Paris Basin (numerous localities) and the type Sables d’Auvers and Sables de Beauchamp of Moiselles; Upper Bracklesham Beds of England.

Discussion: The specimens from Moiselles and Auvers are larger than those from Damery. The longitudinal ridges are more prominent because the cross ridges are not so strongly developed as in the Lutetian forms.

Cytheretta ruelensis sp. nov. (Pl. 6, figs 4, 7; Pl. 5, figs 4-7)

DERIVATION OF NAME: From the hamlet of Le Ruel.

Diacnosis: A species of Cytheretta with 12 longitudinal ridges, of which no. 4 and 5 are prominent, and with coarse puncta between the ridges.

Hototype: Io 3837, a female left valve.

PARATYPE: Io 3838.

MATERIAL: Four carapaces and one left valve.

TYPE LOCALITY AND HORIZON: Le Ruel; Sables de Cresnes.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: So far only known from the type locality.

DESCRIPTION: Sexual dimorphism is very distinct, the males being more elongate. The left valve has an almost straight dorsal margin with a prominent posterior hinge ear. The anterior margin is obliquely rounded; the ventral margin is almost straight; the posterior margin is obliquely rounded. The right valve has a convex dorsal margin and a concave ventral margin. The carapace is narrowly ovate in dorsal view.

The ornamentation consists of 12 longitudinal ridges with coarse puncta between them. Ridge no. 1 forms most of the dorsal margin, leaving it towards the anterior where it swings downwards to join a narrow anterior area of reticulation. Ridges 2 and 3 are not strong ; towards the posterior no. 2 bifurcates and no. 3 joins the lower and stronger part. No. 4 and 5 are very strong running almost the whole length

296 MID-TERTIARY CYTHERETTINAE

of the carapace; a prominent gap is developed between them to the posterior of the sub-central plexus. The sub-central plexus is not very well developed and lies between ridges 4and5. There are four rows of coarse puncta between ridges 4 and 5 forming two double rows; there are four more closely packed rows between 3 and 4; two between 2 and 3, and 5 and 6; and one between the remaining ridges.

The internal features could not be clearly seen. DiIMENsIons (Carapace):

i H L/H Ww Female 0:75 0°43 74 0°35 Male 0°85 0°44 1-93 0°37

Discussion: C. ruelensis shows a similarity to the C. haimeana group, but the ornamentation is quite distinct and unlike any of these in detail. The puncta between the ridges are unlike any other inter-costal ornamentation developed in this group.

One specimen of a related form has been found in a sample from Moiselles (Pl. 7, fig. 11). The ridges form the same pattern as in C. ruelensis but they are all of about equal prominence. This is Cytheretta sp.A.

Cytheretta eocaenica Keij (Plt, fig. 2} Pl..7, digs aro)

1852 Cythere jurinei (non. von Munster); Bosquet: 56 (pars), pl. 2, fig. 9. 1955 Cytheretta juvinei (non. von Munster); Apostolescu: 263, pl. 5, figs 86-89; pl. 6, figs go—-91. 1957 Cytheretia eocaenica Keij: 134, pl. 6, fig. 6; pl. ro, figs 2-4.

Diacnosis: (After Keij, 1957): ‘A species of the genus Cytheretta with the following characteristics; anterior margin obliquely rounded, obtusely angular ventrally; with horizontal rows of rounded depressions posteriorly; inner margin with three inward protrustions in the ventral half of the valve.’

MATERIAL: 109 valves and carapaces from the Sables de Lede of Bambrugge were available for study, donated by Dom. R. Rome, together with nine valves and carapaces from the Lutetian IV of Damery. Io 3839-44.

TYPE LOCALITY AND HORIZON: Grignon; Lutetian IV.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Cuisian and Lutetian of the Paris Basin; Lutetian and Ledian of Belgium; Upper Bracklesham Beds of England.

Discussion: Keij describes the ornamentation as consisting of 3—5 horizontal rows of rounded depressions in the postero-ventral part of the valve with additional widely scattered depressions in the female. The specimens from Bambrugge have some nine rows of pits in the postero-ventral region and some five in the antero-ventral region. There is a smooth area near the centre of the valve representing the sub- central plexus from which run two prominent rows of pits towards the direction of the anteroventral angle. The Damery specimens fit Keij’s description with five postero-ventral rows of pits; the more ventrally placed pits, as developed at Bam-

OF NORTH-WEST EUROPE 297

brugge, are absent. However, the two prominent anterior rows can also be distinguished.

Another feature of the Bambrugge sample is the presence of a few very large indi- viduals, about rin 8. These are similar in shape and ornamentation to the smaller specimens. The smaller individuals are definitely adults: the hinge, sexual dimor- phism, thick shell and inner margin are all well developed. This is probably an example of post-maturation moulting.

DIMENSIONS: v H WwW L/H Female (normal) 0-78 0°47 0°40 1:66 Female (large) 0:93 0°59 0-49 1°58

Sex ratio: I : 3.

Cytheretta oligocaenica sp. nov. (Pl. 8, figs 1-13) DERIVATION OF NAME: From Oligocene; refers to the strata in which it is found.

Diacnosis: A species of Cytheretta with an obliquely rounded anterior margin and slightly convex ventral margin in the left valve; ornamentation consists of up to 8 rows of pits in the postero-ventral region; there is often an ill-defined sulcus in the postero-lateral position.

Ho.otyPe: Io 3845, a female left valve. PARATYPES: Io 3846-9.

MATERIAL: Couches du Phare (Biarritz), 55 valves and carapaces; Couches de VAtalaye (Biarritz), 11 valves and carapaces; St. Geours de Maremne, 9 valves and carapaces; Blaignan, 5 carapaces.

TYPE LOCALITY AND HORIZON: Biarritz; Couches du Phare (RO 271); Stampian.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Couches de l’Atalaye and Couches du Phare, Biarritz; Faluns Bleues of St. Geours de Maremne; Argile a algues, Blaignan.

DESCRIPTION: Sexual dimorphism is distinct. Sex ratio, 1: 3. The left valve has a prominent posterior hinge-ear and a convex dorsal margin; the anterior margin is obliquely rounded; the ventral margin is slightly convex; the posterior margin is obliquely rounded. The greatest height is just to the anterior of centre. The right valve has its greatest height about one quarter of the way from the posterior margin and this results in a slight triangular shape as the almost straight dorsal margin slopes towards the anterior margin; the ventral margin is concave.

The ornamentation consists of up to 8 rows of pits in the postero-ventral region. An ill-defined sulcus is often present in the postero-lateral position.

The antero-dorsal lobe of the hinge of the left valve is swollen; the antero-ventral lobe is large and prominent; the antero-median tooth is large; the postero-median

298 MID-TERTIARY CYTHERETTINAE

swelling is prominent. The anterior tooth of the right valve is crescentic-shaped in lateral view; in dorsal view both teeth of the right valve are large and squat (Plate 8, Fig. 8-13).

The inner margin has a narrow anterior indentation and small ventral and posterior indentations; the anterior segment is semi-circular; the median segment is long with a gentle upward slope; the posterior segment is short and semi-circular. There are some 35 anterior and 17 ventral radial pore canals. The posterior radial pore canals are very closely packed and there are about 80 of them; they are not divisible into two groups. The selvage is very close to the anterior margin; there is a small posterior flange groove and a larger ventral one; a weak list is present.

As with C. eocaenica, there are a few very large individuals, about 10%.

DIMENSIONS:

L H Ww L/H Female 0°85 0:53 0:51 I-60 Male 0-86 0°49 0-41 I-71 Large form, male 0-98 0:58 _- 1:70

Discussion: This is very similar to C. eocaenica Keij. It differs in shape; the antero-ventral angle is rounded and not angular as in C. eocaenica, the ventral margin is slightly convex, not straight, and the greatest height is just to the anterior of centre, whereas it is about one third of the way from the anterior in C. eocaenica. The right valve is much higher in the posterior in C. oligocaenica. There are more anterior radial pore canals in the latter, 35 compared with 27-30. The inner margin is similar in both species, and in this respect they differ from C. juvinez (von Munster), which has a very large and circular posterior indentation.

See also C. geoursensis sp. nov.

Cytheretta geoursensis sp. nov. (Pl. ro, figs I, 3, 5)

DERIVATION OF NAME: From St. Geours de Maremne.

Diacnosis: A large species of Cytheretta with a prominent posterior hinge ear, convex ventral margin of left valve, and a finely punctate outer surface.

Ho.otyPe: Io 3850, a female left valve. PARATYPES: lo 3851-2. MATERIAL: Eight valves and carapaces.

TYPE LOCALITY AND HORIZON: Marniére d’Escornbéou, near St. Geours de Maremne (Aquitaine Basin); Faluns bleues, Oligocene(?) (Chattian?).

STRATIGRAPHICAL RANGE AND DISTRIBUTION: So far only known from the type locality.

DESCRIPTION: Sexual dimorphism is distinct; sex ratio, 1:3. The left valve has a very prominent posterior hinge ear; the dorsal margin is strongly convex; the anterior margin is obliquely rounded; the ventral margin is strongly convex, par-

OF NORTH-WEST EUROPE 299

ticularly towards the posterior; the posterior margin is obliquely rounded. The right valve has a slightly convex dorsal margin and a slightly concave ventral margin.

The whole outer surface of the carapace is covered with a very superficial appearing punctation. Two or three rows of pits in the postero-lateral position can be seen on some specimens, and one right valve has these quite prominently developed.

The hinge, inner margin, selvage and list are as for C. oligocaenica. There are 27 anterior radial pore canals, but the exact number of posterior radial pore canals could not be determined.

DIMENSIONS: 1B H Ww L/H Female 0:98 0:58 0-60 1:69 Male 1:07 0:58 0:50 1:84

Discussion: This could perhaps be regarded as a post-maturation moult stage of C. oligocaenica sp. nov. similar to those already described. However, unlike these, this is not merely a larger version of the small forms. It differs from C. oligocaenica in shape, particularly the strongly convex ventral margin of the left valve; in ornamentation; and in the number of anterior radial pore canals (27 compared with 35).

Cytheretta carita sp. nov. (Rito; tigs|5, S)o—11: Plivno, fig. 7: Lext-fig> 14)

DERIVATION OF NAME: Latin—carita, to be without or to be deprived; refers to the lack of ornamentation.

Diacnosis: A smooth species of Cytheretta with an obliquely rounded anterior margin and two ‘dimples’ in the posterior.

HototyPe: Io 3853, a female left valve. PARATYPES: Io 3854-8.

MATERIAL: 39 valves and carapaces from the type Auversian; 57 from Moiselles; 5 from Le Ruel, Sables de Cresnes; 1 from the Marnes a P. ludensis at Verzy.

TYPE LOCALITY AND HORIZON: Carriere de Moiselles; Sables de Beauchamp. STRATIGRAPHICAL RANGE AND DISTRIBUTION: ‘Bartonian’ of the Paris Basin.

DESCRIPTION: Sexual dimorphism distinct; sex ratio, 1 : 2°5._ The dorsal margin of the left valve has a prominent posterior hinge ear, is straight until about the mid- point when it slopes down to the anterior margin, which is very obliquely rounded. The ventral margin is slightly convex and curves into the posterior margin, which is rounded. The dorsal margin of the right valve is more irregular; a slight hinge ear is present at the posterior and the anterior tooth projects beyond the margin. There is a marked concavity between the hinge ear and the posterior margin. The ventral margin is slightly concave. In dorsal view it is ovate.

The valve issmooth. At the posterior are two dimples, one in the postero-ventral angle and the other higher along the posterior margin.

300 MID-TERTIARY CYTHERETTINAE

In the hinge of the left valve, the antero-ventral lobe is weakly developed, leaving the anterior socket almost open ventrally; the antero-dorsal lobe is flat to lobate and small; the antero-median tooth is large; the postero-median swelling small. In the right valve the anterior tooth is large and the posterior tooth is slightly reniform.

The anterior indentation of the inner margin is narrow; the ventral deep and prominent; and the posterior broad, but deep. The median segment is almost flat; the posterior segment is steeper, but not very pronounced. The muscle scars are normal, the two ventral scars being close together. They are situated in a slight pit.

The selvage runs along the anterior margin of both valves, so there is no anterior flange groove. A ventral flange groove is well developed, as well as a small posterior one. A very weak list is present along the ventral and posterior duplicature.

Fic. 14. Cytheretta cavita sp. nov.; female right valve; x75.

DIMENSIONS: 1 H WwW L/H Female 0:87 0°55 0°43 1°58 Male 0:93 0°53 0-41 I-75

Discussion: This differs from other smooth Cytheretta species such as C. rhenana Triebel by its shape and the presence of the posterior dimples. It shows a very close relationship to C. eocaenica, but lacks the longitudinal rows of pits of the latter.

Cytheretta cellulosa sp. nov. (Pl. 9, figs 1-4, 6, 7) DERIVATION OF NAME: Latin—cellulosus, full of little cells, referring to the orna- mentation.

DraGnosis: A species of Cytheretta with 13 rows ot pits which converge at the posterior; dorsal area is smooth.

HoLotyPeE: Io 3859, a female left valve. PARATYPES: Io 3860-63.

MATERIAL: Eight valves and carapaces from the type Auversian; fourteen valves and carapaces from Moiselles.

OF NORTH-WEST EUROPE 301

TYPE LOCALITY AND HORIZON: Auvers-sur-Oise; Sables d’Auvers. STRATIGRAPHICAL RANGE AND DISTRIBUTION: Auversian of the Paris Basin.

DESCRIPTION: Sexual dimorphism is pronounced; sex ratio, 1: 3. The dorsal margin of the left valve has a prominent posterior hinge ear and is convex. The anterior margin is slightly obliquely rounded; the ventral margin is straight in the central portion; the posterior margin is very obliquely rounded. In the male, the ventral margin has a very straight appearance. The right valve has a concave ventral margin. Carapace is ovate in dorsal view.

The ornamentation of the left valve consists of 13 rows of small pits. In the posterior region these rows converge and merge into one another; the pitting is very fine. In the central region the rows are distinct and consist of a single line of pits larger in the centro-dorsal region. The central rows converge upon an ill-defined sub-central plexus, which is almost smooth. To the anterior of this the rows consist of double rows of puncta; there are five of these, together with three ventral rows of single puncta which are continuous along the ventral region. The dorsal area is smooth. The right valve is similar except that there are single rows of pits in the anterior region.

The hinge of the left valve has a swollen antero-dorsal lobe; a small antero-ventral lobe so that the anterior socket is virtually open ventrally, as is the posterior socket ; a large antero-median tooth; and a small inconspicuous postero-median swelling. The anterior tooth of the right valve is large, and the posterior tooth is prominent.

The inner margin could not be clearly seen. There are some 20 anterior radial pore canals. The central muscle scars are in a slight pit, and the two ventral ones are close together. One very prominent dorsal muscle scar can be seen above the fulcral point. The selvage runs close to the anterior margin; there is a ventral flange groove and a small posterior one. A weak list is present along the anterior, ventral and posterior parts of the duplicature.

DIMENSIONS: Ly H Ww L/H Female 0:80 0:49 0:38 1-65 Male 0°93 0-51 0°43 1:82

Discussion: C. cellulosa bears a ressemblance to C. tenuipunctata (Bosquet), but the pattern of pits is different and the outline of the valve is completely different. C. eocaenica Keij from the Ledian of Bambrugge often develops pits over a large area, although never to the same extent ashere. The shape, however, is similar, and it is thought likely that C. eocaenica is the ancestor of C. cellulosa, with such forms as those from Bambrugge as intermediates.

Cytheretta aff. cellulosa 1968 Cytheretta minor (non Lienenklaus) Haskins: 167, pl. 1, figs 30-35.

Discussion: Haskins records this from the Barton Beds in Alum Bay. In many respects it appears to be intermediary between C. cellulosa and C. minor.

302 MID-TERTIARY CYTHERETTINAE

Superspecies Cytheretta laticosta (Reuss)

DiaGnosis: A group of species of the genus Cytheretta characterized by three longi- tudinal ridges. The dorsal ridge is convolute; the median ridge contains four depressions along its ventral side; the ventral ridge is the most prominent of a group of ridges developed in the ventral part of the valve.

DESCRIPTION: The carapace generally has a massive appearance with a thick shell; it is strongly inequivalve, and sexual dimorphism is pronounced. In the left valve there is a prominent posterior hinge ear. In the right valve the anterior tooth projects beyond the margin and the ventral margin is concave. In dorsal view the carapace is swollen posteriorly in both males and females, but this is more pronounced in the latter.

The ornamentation consists predominantly of three thick longitudinal ridges. The dorsal ridge runs along the margin, but does not reach the anterior margin; the median ridge swells in the central region and contains four depressions on its ventral side, each of which has a normal pore canal opening into it. This runs from the anterior margin to the posterior, where it curves upwards to form a margin to the hinge ear. The ventral ridge also contains four depressions with normal pore canals, much better seen in the right valve. In the left valve this ridge is really the most pronounced of a series of ventral ridges, of which there are seven in all. In the right valve the lower ventral ridges are indistinct or absent, and this gives the whole valve a strongly tri-costate appearance. A prominent marginal rim runs around the anterior margin and the anterior portion of the ventral margin. In the right valve there is a posterior marginal rim. There are some eight marginal antero-ventral denticles, each of which bears a radial pore canal, and three posterior ones. The latter are larger in the right valve. The whole surface is finely punctate, although this can only be seen on well preserved specimens.

The hinge of the left valve has a swollen, knob-like, antero-dorsal lobe; the antero- ventral lobe is small; the antero-median tooth is prominent ; postero-median swelling is small. In the right valve the anterior tooth is large and the posterior tooth circular in plan.

The selvage is prominent, particularly in the right valve; there is a well developed flange groove in the anterior, ventral, and posterior; and a list is present in the postero-veatral region. There are some 32 anterior radial pore canals, 20 ventral, and 18 posterior; and 29 normal pore canals which are arranged in sympathy with the ornamentation (Text Fig. 15). The central and dorsal muscle scars are as for the genus.

Cytheretta laticosta (Reuss) (Pl. 2, fig3; Plo xo, figs?2) 4, 6).8, 93. Pl. a2) figs 1—2 55 Lext-tig. 15):

1850 Cypridina laticosta Reuss: 87, pl. 11, fig. 13.

1857 Cythere plicata var. laticosta (Reuss); Jones: 32, pl. 5, fig. 8 (pars). 1889 Cytheve plicta var. laticosta (Reuss); Jones and Sherborn: 29 (pars). 1957 Cytheretta laticosta (Reuss); Keij: 137, pl. 18, figs 15-18; pl. 21, fig. 16. 1968 Cytheretta laticosta (Reuss) ; Haskins: 166 (pars), pl. 2, figs 23, 27, 28.

OF NORTH-WEST EUROPE 303

Dracnosis: A member of the superspecies C. laticosta with an evenly rounded anterior margin, convolute dorsal ridge and small areas of coarse puncta around the median and ventral ridges.

MATERIAL: Three carapaces from the Lower Barton Beds of Alum Bay; numerous specimens from the Middle and Upper Barton Beds of Barton, and Alum Bay; fifteen valves and carapaces from the Marnes a P. ludensis of Verzy, and five from Chavengon. Io 3864-70.

TYPE LOCALITY AND HORIZON: Barton Clay of Barton (see discussion).

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Barton Clay of Barton, Alum Bay and Whitecliff Bay; Marnes a P. ludensis of Verzy and Chavencon; Argile d’Asche of Oedelem (Belgium).

DESCRIPTION: Sex ratio, I : 1:6. The carapace is massive, the shell is thick. The dorsal ridge is thick with convolutions forming some six depressions, which bear normal pore canals. The median ridge has a prominent swelling in the central part and has an area of coarse puncta around the four depressions. The ventral ridge likewise has an area of coarse puncta.

The inner margin has a narrow and deep anterior indentation, particularly in the right valve, and a prominent ventral one. The anterior segment is sharply rounded. The median and posterior segments are inseparable; they slope up to the posterior indentation, which is quite close to the ventral margin.

DIMENSIONS:

Left valve Right valve IL H L/H WwW 1 H L/H

Female o-80 + 0:02 0:50+0-01 1I:60+0-04 0:46 0-79 0:42 1:88 Male 0:86 + 0:02 0-49 + 0:01 1:76 +0:04 0:45 0:89 0:46 1:94 Discussion: Reuss described this as coming from the London Clay of Barton, Hampshire. This relates to the idea prevalent in the early 19th century that the clay at Barton was of the same age as that at London; in fact, the London Clay is Lower Eocene, while the Barton Clay is Upper Eocene.

Fic. 15. Cytheretta laticosta (Reuss) showing distribution of normal pore canals in relationship to ornamentation; 75.

304 MID-TERTIARY CYTHERETTINAE

Cytheretta forticosta sp. nov. (Pl. 11, figs 1-4, 8, 9; Pl. 12, figs 6-12)

1857 Cythere plicata (non. von Munster); Jones: 32 (pars). 1968 Cytheretta laticosta (Reuss); Haskins: 166 (pars), pl. 2, figs 19-22, 24-26, 29.

DERIVATION OF NAME: Latin—fortis, strong; costa, ridge. Refers to the three strong longitudinal ridges.

DraGnosis: A member of the superspecies C. Jaticosta with an obliquely rounded anterior margin and simple median ridge.

HoLotyPeE: Io 3871, a female left valve.

PARATYPES: Io 3872-8.

MATERIAL: Numerous specimens from the localities mentioned below.

TYPE LOCALITY AND HORIZON: Fisher Beds 17-18, Upper Bracklesham Beds, Whitecliff Bay.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Upper Bracklesham Beds of White- cliff Bay, Selsey Bill, and Bramshaw (New Forest) ; Sables de Cresnes, Le Ruel.

DESCRIPTION: Sex ratio, I : 2:5. The carapace is massive. The left valve of the female has a circular outline; the dorsal margin is strongly convex; the anterior margin is obliquely rounded; the ventral margin is short and straight; the posterior margin is very obliquely rounded. In the right valve the dorsal outline is irregular, and the ventral margin is concave.

The dorsal ridge has five depressions, each of which bears a normal pore canal. The four pits on the median ridge are simple; those on the ventral ridge are prominent in both valves. The surface of the valveis punctate. In the right valve the ventral group of ridges are distinct in the anterior and posterior regions.

The inner margin is similar to that of C. laticosta except that the joint median and posterior segments have two undulations.

DIMENSIONS:

Left valve Right valve IL, H L/H W L H L/H

Female 0:77 40:03 0°52 0:02 1:'46+0:04 0°52 0:83 0-45 1°84 Male 0:87 + 0:03 0°54 40:02 1:62 +0:05 0:52 0-90 0-48 1:88

Cytheretta porosacosta sp. nov. (Pl. 11, figs 5-7; Pl. 12, figs 3, 4)

1857 Cythere plicata (non. von. Munster); Jones: 32 (pars), pl. 4, fig. 16; pl. 5, fig. 8. 1889 Cytheve plicata (non. von Munster); Jones and Sherborn: 29 (pars), pl. 1, fig. 18. 1968 Cytheretta laticosta (Reuss); Haskins: p. 166 (pars).

DERIVATION OF NAME: Latin—porosus, full of holes; costa, ridge. Refers to the heavily punctate ornamentation.

Diacnosis: A member of the superspecies C. lJaticosta with an evenly rounded anterior margin, simple ridge, and wide areas of large puncta.

OF NORTH-WEST EUROPE 305

Ho.otyPe: Io 3879, a female left valve. PARATYPES: Io 3880-83. MATERIAL: Numerous valves from the localities mentioned below.

TYPE LOCALITY AND HORIZON: Colwell Bay, Isle of Wight; Middle Headon Venus Bed.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Brockenhurst Beds of Brockenhurst and Whitecliff Bay; Middle Headon Beds of Colwell Bay, Headon Hill, Whitecliff Bay, Milford, and Brockenhurst.

Description: The dorsal margin of the left valve is slightly irregular due to the over-reach of the dorsal ridge; otherwise it is straight to slightly convex. The anterior margin is evenly rounded; the ventral margin is slightly concave; and the posterior margin obliquely rounded.

In the left valve the dorsal ridge has several faint depressions, but is fairly simple. The median ridge is narrow, with four very faint depressions on the ventral side. There is no strong ventral ridge. In the right valve however, the ventral ridge is distinct due to the absence of the lower ventral ridges. Large areas of coarse puncta are developed amongst the ventral group of ridges and around the median ridge.

DIMENSIONS: Left valve Right valve i H L/H I H L/H Female 0-74+0:02 0-44+0-01 1:68-+0-:025 0:74 0:38 1:92 Male 0-78-+0-:02 0-42+0-:01 1:84+0:04 0:82 0:40 2:05

Discussion: The three species described are only part of the superspecies C. laticosta. Other forms not studied are found in the London Clay of the London and Hampshire basins and the Lower and Middle Bracklesham Beds.

The most noticeable difference between these species is the shape of the carapace. This is most marked in the female dimorph. The species represent three quite distinct groups (Text-fig. 16). The L/H ratio is:

Female Male C. forticosta 1:46-0:04 1:62-0:05 C. laticosta I-60-L0:03 1:76-0:04 C. porosacosta 1-68-+L0:025 1:84-+0-04

The anterior margin changes from obliquely rounded in C. forticosta to evenly rounded in C. laticosta and C. porosacosta.

The three longitudinal ridges are more complex in C. laticosta. The dorsal ridge is convolute and complicated. The four depressions in the median ridge are prominent and punctate; in C. forticosta there are only four large puncta and no depressions; in C. porosacosta the depressions are almost absent. The ventral ridge has a prominent central area with coarse puncta; these are not present in C. forticosta, while in C. porosacosta the puncta cover a large area of the valve. In the right valve of C. forticosta the ventral group of ridges are stronger than in C. laticosta and C. porosacosta.

306 MID-TERTIARY CYTHERETTINAE

The trends seen in these three species are, therefore:

(i) Change from a rounded to a more elongate shape.

(ii) Anterior margin from obliquely rounded to evenly rounded.

(11) Development of coarse puncta, especially around the median and ventral ridges.

The group of specimens from Le Ruel have a mean L/H ratio of 1-43 and have a more rounded appearance than the English C. forticosta. The specimens from Selsey Bill and Bramshaw are more elongate, with a mean ratio of I-50. The Whitecliff

0:60

o * Bramshaw

e@ Selsey Bill

© Whitecliff Bay

x Barton

© Ludian, Paris Basin +

Standard deviation

Headon Beds

J a a)

Height - mm——

0-65 0-70 0-75 0:80 0-85 Length -mm—_ >

Fic. 16. Size distribution of species of the superspecies C. laticosta (Ruess).

OF NORTH-WEST EUROPE 307

Bay specimens have a mean of 1-46 and have an even distribution. There is con- siderable overlap between these specimens, however, which does not occur between these specimens of C. laticosta and C. porosacosta.

The variation in size is most noticeable between C. porosacosta on the one hand and C. laticosta and C. forticosta on the other. This is thought to be environmental because C. porosacosta is found in beds with a mixed marine and brackish water fauna. The thinner shell of C. porvosacosta and the general weakness of the three ridges is probably also connected with this. The specimens of C. forticosta from Bramshaw and Selsey are seen to be smaller than those from Whitecliff Bay; and C. laticosta from the Ludian of the Paris Basin are smaller than those from the Barton Clay. This may be environmental. The specimens from Le Ruel show a large range in size, due perhaps to the nature of the Sables de Cresnes, which are coarse-grained current-bedded sands, often with rolled macro-fossils.

Eventually, specimens of an intermediate nature may be found between these three species, in which case they will become subspecies. It is thought highly likely that such intermediates exist, but at the moment there are present three distinct groups with no overlap.

Superspecies Cytheretta tenuipunctata (Bosquet)

The superspecies C. tenuipunctata comprises a group of ostracods with similarities in shape of lateral view, ornamentation, and internal structures. The following species and subspecies are included:

. tenuipunctata tenuipunctata (Bosquet) . tenuipunctata absoluta subsp. nov. . tenuipunctata livata subsp. nov. . tenuistriata tenuistriata (Reuss) . tenuistriata ornata subsp. nov. . bernensis Oertli . buttensis sp. nov. buttensis subsp. nov. . buttensis reticulata subsp. nov. . mintpunctata sp. nov. The following are tentatively included: C. triebeli Oertli C. variabilis Oertli C. ramosa ramosa (Lienenklaus) C. vamosa sublaevis Triebel

MAA A AAA AO

There are further groups of ostracods from the Oligocene of western Europe which are also closely related to the superspecies. These are discussed below.

Dracnosis: A superspecies of the genus Cytheretta with up to 13 longitudinal ridges, often stronger in the ventral half of the valve, which form a regular pattern, although differing slightly in detail between species. There are three ridges in the anterior part which run from the region of the sub-central plexus towards the antero-ventral angle. In lateral view the carapace is elongate with pronounced sexual dimorphism ; D

308 MID-TERTIARY CYTHERETTINAE

in dorsal view it is ovate or tapered towards the anterior. The valves are not strongly inequivalve.

DESCRIPTION: The shape varies, but in general it is elongate in lateral view. In dorsal view it is usually ovate, or slightly triangular with its apex at the anterior. Sexual dimorphism is pronounced; sex ratio, 1 : 2. The valves are inequivalve, but not strongly so.

Up to 13 longitudinal ridges are developed, which may be strong or weak. In some species no ridges are present in the dorsal half of the valve. The ornamentation between the ridges varies from species to species. A complete development is seen in Text-fig. 17.

Ridge no. I runs from near the posterior hinge ear, disappearing to the anterior of the sub-central plexus. No. 2 is often formed of a series of short curved ridges, usually broken and bifurcating. No. 3 joins no. 2 in the posterior, runs above the sub-central plexus, and just to the anterior of it joins no. a. In some species, and particularly in the right valve, ridge a is very prominent, sloping sharply towards the antero-ventral angle. The sub-central plexus is an irregular, smooth area, varying from species to species, but usually prominent. Three parallel ridges, a, b, and c, run from the sub-central plexus towards the antero-ventral angle; these are diag- nostic of the superspecies. There are another three parallel ridges, 4, 5, and 6 to the posterior of the sub-central plexus. Ridge no. 7 joins no. 4 at the posterior and no. 8 at the anterior. Ridges 8-13 are approximately parallel to the ventral margin. The ridges form a complicated pattern at the posterior, seen in Text-fig. 17. Ridges 9-13 disappear amongst fine puncta at the posterior.

The antero-dorsal lobe of the hinge of the left valve is prominent and slightly swollen; the antero-ventral lobe is prominent; the antero-median tooth is well developed, but the postero-median swelling is virtually absent. In the nght valve the anterior tooth is slightly reniform in shape; in dorsal view the two teeth appear to be almost equal in size.

The selvage is prominent along the ventral and posterior margins. A wide flange groove with a well marked flange is present in the ventral region; in the posterior the flange groove is narrow and a small fringe is developed. Along the anterior

enki br

Fic. 17. Ridges of the superspecies Cytheretta tenuipunctata (Bosquet). The specimen is C. tenuistrviata oynata subsp. nov; X75.

OF NORTH-WEST EUROPE 309

margin a few very small denticles are present; a radial pore canal opens from each of them. There are some 35 anterior radial pore canals and some 40 closely spaced posterior pore canals. The inner margin varies slightly from species to species.

Cytheretta tenuipunctata (Bosquet)

Diaenosis: A species of the superspecies C. tenuipunctata with a prominent sub- central plexus and a medium to coarse pitting between the ridges, the pits often being in double rows in the median part of the valve. The ridges are strongly developed.

Discussion: This was described by Bosquet from the Argile a N. comta of Belgium. It is very similar to C. tenwistriata (Reuss), and Keij (1957) believed that the two were synonymous. This was because of the similarities in ornamentation. The two species are here regarded as distinct, but are placed within the same superspecies. C. tenumstriata is a much larger ostracod than C. tenuipunctata, but in itself this need not be important. In the Paris Basin however, the two co-exist without inter- mediaries, so that the difference in size is very obvious; thus they formed two separate populations which must be assumed to be specifically distinct. The ornamentation between the ridges consists of a coarser pitting in C. tenuipunctata than in C. tenuistriata.

Cytheretta tenuipunctata tenuipunctata (Bosquet)

1852 Cythere jurvinei (non. von Munster); Bosquet: 56 (pars). 1852 Cythevre jurinet var. tenuipunctata Bosquet: 56, pl. 2, fig. ro. 1957 Cytheretta tenuipunctata (Bosquet); Keij: 138, pl. 5, fig. 21; pl. 6, fig. 5. TYPE LOCALITY AND HORIZON: Berg, near-Kleine Spouwen, Belgium; Argile a N. comta (Rupelian).

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Upper Tongrian (?) and Lower Rupelian of Belgium.

Dimensions: (After Keij):

Lectotype (Male R) 0:96 x 0:47 Female L 0°88 x 0-51

Discussion: Keij (1957) has selected a lectotype and redescribed species as having 6-8 ridges which gradually vanish towards the anterior and posterior margins; between the ridges are fine pits, mostly in two rows; towards the dorsal margin the surface is pitted, but without ridges; the areas along the anterior, posterior, and dorsal margins are smooth; and a sub-central plexus is well developed.

C. tenuipunctata tenuipunctata differs from the two subspecies described below by the absence of the dorsal ridges.

310 MID-TERTIARY CYTHERETTINAE

Cytheretta tenuipunctata absoluta subsp. nov. Pl. 14, figs 1-4, 6, 7; Text-fig. 18)

DERIVATION OF NAME: Latin—absolutus, complete; refers to the ornamentation which completely covers the valve, unlike C. tenwipunctata tenuipunctata.

DiacGnosis: A subspecies of C. tenuipunctata in which the full 13 ridges of the superspecies C. tenuipunctata are developed; the ornamentation between the ridges consists of a fine pitting, in double rows in the median and dorsal parts of the valve, and single rows between the ventral ridges.

Ho.otyPeE: Io 3884, a female left valve.

PARATYPES: Io 3885-9.

MATERIAL: 55 adult valves and carapaces and 68 larval stages from Cormeilles. TYPE LOCALITY AND HORIZON: Cormeilles-en-Parisis; Marnes a Huitres. STRATIGRAPHICAL RANGE AND DISTRIBUTION: Marnes a Huitres of the Paris Basin.

DescripTION: The dorsal and ventral margins of the valves are sub-parallel, particularly in the left valve. In the latter there is a prominent posterior hinge ear; the dorsal margin is slightly convex; the anterior margin is evenly rounded; the ventral margin is straight or slightly concave ; the posterior margin obliquely rounded In the right valve the ventral margin is concave and there is a concavity in the postero-dorsal position. In dorsal view the carapace is ovate.

The ridges are well defined, particularly in the right valve. The ornamentation between the ridges consists of a fine pitting. Between ridges I, 2, 3 and 4, 6and7, a, b, c, there is a double row of pits; between the ventral ridges there is a single row. The anterior region is reticulate.

The inner margin (Text-fig. 18) has a broad anterior and posterior indentation and a well marked ventral indentation. The anterior segment is rounded, and the joint median and posterior segments are undulating.

Two larval stages have been recognised (Text-fig. 19). In these the complete ridge pattern of the adult is developed with small puncta between them; these are in rows of three between ridges no. 6 and 7, and in double rows between the others. Ridges 8 and 9 are very strong in the posterior region and in larval stage 7 there is a distinct swelling in this region.

DIMENSIONS:

Left valve Right valve IU, H L/H W L H L/H Female 0°87 + 0:03 0-49 + 0-01 1:76+ 0:05 0:43 0:87 0:46 1-89 Male I-00 + 0:02 O51 + 0°01 1:96 + 0:02 0:43 0:98 0:50 1:96

Discussion: This subspecies differs from C. tenuipunctata tenuipunctata in the larger number of ridges, caused by their presence in the dorsal regions. It differs from C. tenuipunctata lirata subsp. nov. in having a slightly different shaped dorsal margin, the latter having a weak anterior hinge ear in the left valve, and in the ornamentation between the ridges which show a great deal of variation in C. tenuipunctata livata.

OF NORTH-WEST EUROPE 311

Fic. 18. Cytheretta tenuipunctata (Bosquet) absoluta subsp. nov.; female right valve; x75.

0:60

@ Expected size, after the method of Anderson

(1964)

0:50 ° © 00f Oo [orole} ° ° x 828 te Oe) E E £ © xox cy x x t x x z= >» Bee ko Xo Ox x 040 XM EKIN, x x me 8(33) ° x @ o x * 766) x 030 x n 1 1 1 0:50 060 0.70 0.80 0:90 10

Length-mm—

Size distribution of the adults and larval stages of Cytheretta tenuipunctata (Bosquet)

Fic. 19. absoluta subsp. nov.

Cytheretta tenuipunctata lirata subsp. nov. (Pl. 14, figs 5, 8-10; Pl. 15, fig. 10; Pl. 16, fig. 4) 1895 Cytheve juvinei (non. von Munster); Lienenklaus: 8 (pars). DERIVATION OF NAME: Latin—lirata, earth or ridge formed by ploughing; refers

to the ornamentation.

Dracnosts: A subspecies of C. tenuipunctata in which the full 13 ridges of the superspecies C. tenuipunctata are developed; the ornamentation between the ridges varies from pitting to reticulation; a weak hinge ear is developed in the left valve.

HototyPe: Io 3890, a female left valve.

PARATYPES: Io 3891-95. MATERIAL: 37 valves from Auvers-St-George.

312 MID-TERTIARY CYTHERETTINAE

TYPE LOCALITY AND HORIZON: Auvers-St-George; Stampian. STRATIGRAPHICAL RANGE AND DISTRIBUTION: Stampian of the Paris Basin.

DESCRIPTION: The shape of the dorsal margin of the left valve is convex with two slight concavities to the anterior and posterior of the posterior and anterior hinge ears respectively. Otherwise the shape is similar to C. tenuipunctata absoluta. The 13 ridges are developed, although a few specimens show a smooth area in the antero- dorsal region, but this is small. Ornamentation between the ridges shows a large amount of variation; pits are usually developed, sometimes in a double row, but more often in a single row; in other specimens a fine reticulation is present, and in others the longitudinal ridges are very strong with a relatively inconspicuous orna- mentation between them. The sub-central plexus is well developed. There is an anterior area of reticulation. The inner margin is similar to that of C. tenuipunctata absoluta.

DIMENSIONS: Left valve Right valve 1 H L/H #w ih, H L/H Female o0-89+0:03 0:49+0-01 1°82+0-:03 0:29 0:90 0°45 2:00 Male 0:96 0:49 1-07 0:29 0:98 0:48 2-04

Discussion: See C. tenuipunctata absoluta subsp. nov.

Cytheretta tenuistriata (Reuss)

DracGnosis: A species of the superspecies C. tenuipunctata of large size, with well developed sub-central plexus and distinct ornamentation between the ridges.

Cytheretta tenuistriata tenuistriata (Reuss) (Pl. 16, figs 5, 7)

1853 Cytherella tenuistriata Reuss: 676, pl. 9, fig. ro.

1905 Cythereis jurvinet (non. von Munster); Lienenklaus: 31 (pars). 1952 Cythevetta tenuistriata (Reuss); Triebel: 22, pl. 3, fig. 12-15. 1956 Cytheretta tenuistyviata (Reuss); Oertli: 61, pl. 6, fig. 163-165.

MATERIAL: Specimens from several localities around Weinheim; Io 3896-7. TYPE LOCALITY AND HORIZON: Weinheim, Mainz Basin; Unterer Meeressand.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Unterer Meeressand of the Mainz Basin, Blaue Ton of Switzerland; Rupelian.

DIAGNOSIS AND DESCRIPTION: The male has an anterior hinge ear in the left valve which, together with the posterior hinge ear, gives the dorsal margin an undulating appearance with two concavities. The anterior hinge ear is not present in the female, so the dorsal margin of the left valve is regular, and the valve has its greatest height close to the anterior end. The ornamentation is not strong; the ridges are weakly defined and in between them are double rows of small pits. The anterior and posterior areas are smooth.

OF NORTH-WEST EUROPE 313

Dimensions (After Triebel):

L Female I-08-1:'I13mm Male I:22-1:28mm

Discussion: See C. tenuistriata ornata below.

Cytheretta tenuistriata ornata subsp. nov. (Pl. 13, figs 1-12; Text-figs 17, 20)

1852 Cythere jurinei var. tenuipunctata Bosquet: 56 (pars). 1895 Cytherve jurinei (non. von Munster); Lienenklaus: 8 (pars).

DERIVATION OF NAME: Latin—-orno, ornamented; refers to the strong ornamenta- tion.

Draenosis: A subspecies of C. tenwistriata with strong ornamentation.

Ho otyPeE: Io 3898, a female left valve.

PARATYPES: Io 4020-7.

MATERIAL: 32 valves from Auvers-St-George, 3 valves from Morigny; Io 3899. TYPE LOCALITY AND HORIZON: Auvers-St.-George ; Stampian.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Stampian of Auvers-St.-George, Jeurre, and Morigny in the Paris Basin.

DeEscRIPTION: The dorsal margin of the left valve of both the male and the female is almost straight with two slight concavities, one to the anterior of the posterior hinge ear and the other to the posterior of the weakly developed anterior hinge ear. The anterior margin is evenly rounded; the ventral margin is concave; and the posterior margin is obliquely rounded. In the right valve two concavities are present, one at the postero-dorsal angle and the other at the antero-dorsal angle.

The ridges are well developed and broad; there is a well marked anterior zone of reticulation. The ornamentation between the ridges consists of puncta arranged in three rows, or two rows between the ventral ridges. The sub-central plexus is large and prominent.

The inner margin has three broad indentations; the posterior segment has a greater slope than the median segment.

Three larval stages have been recognized, no. 6, 7, and 8. Ridges are weakly developed, being stronger in the right valve than in the left. In no. 8 all the ridges are present with a similar arrangement to those of the adult; there is a fine punctation between them. In no. 7 only the more ventral ridges are seen (ridges 4-13), and in no. 6 a few very faint lines can be seen in the postero-ventral region, where there is a slight swelling.

DIMENSIONS:

Length of combined left and right valves: Female I-05 + 0-01 Male I'I5 + 0:03

314 MID-TERTIARY CYTHERETTINAE

Left valve Right valve L H L/H 4W iL, H L/H Female 1:03 0:58 1-78 0:31 I-02 0-51 2°00 Male I'I5 0:58 1-98 0:31 I-14 0:56 2°04

Mean length of larval stages: No. 8, 0:88; no. 7, 0°74; no. 6, 0-61. Discussion: This differs from C. tenwistriata tenuistriata (Reuss) in the stronger ornamentation. The ridges are much more pronounced, there are no smooth area

in the anterior and posterior regions, and the pitting between the ridges is finer than in C. tenuistriata tenuistriata. See also C. tenuipunctata and C. minipunctata sp. nov.

Fic. 20. Cytheretta tenuistriata (Reuss) ovnata subsp. nov.; female right valve; x75.

Cytheretta minipunctata sp. nov. (Pl. 16, figs 1-3)

DERIVATION OF NAME: Latin—minus, small; punctata, small pits; refers to the ornamentation.

Dracnosis: A species of the superspecies C. tenwipunctata with six faint ridges and a finely punctate ornamentation.

HototyPe: Io 4028, a female left valve. PARATYPE: Io 4029. MATERIAL: 9 valves and carapaces from Cormeilles.

TYPE LOCALITY AND HORIZON: Cormeilles-en-Parisis; Bed no. 44 of Albissin, Couches de Sannois superieur.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Lower marine Sannoisian of the Paris Basin.

DEscRIPTION: The dorsal and ventral margins of the left valve are sub-parallel and almost straight; a posterior hinge ear is developed; the anterior margin is obliquely rounded towards ventral; the posterior margin is slightly pointed. In the

OF NORTH-WEST EUROPE 315

right valve, the dorsal margin is convex and the ventral margin concave. In both valves there is an elongate dimple in the postero-ventral region.

Ornamentation consists of some 6 faint ridges which are developed in the ventral half of the valve, particularly in the median portion. In between the ridges are 2 or 3 rows of puncta; these also cover a large part of the dorsal area of the valve, where ridges are not developed.

Details of the interior are not clearly seen.

DIMENSIONS: ie H L/H Ww (Average) Carapaces—Female 0:94-1:00 0-54-0'58 I-74 0-48 Male I-IO-I°15 0:58 1:93 0°50

Discussion: This shows resemblances to C. tenuistriata tenuistriata in size and shape, but the ornamentation is completely different. The ornamentation is similar to that of C. buttensis buttensis subsp. nov, but has fewer ridges and is a different shape.

Cytheretta buttensis sp. nov.

DERIVATION OF NAME: French—butte, a hill; refers to the Butte de Cormeille and other buttes of the Paris region where the Sannoisian is found.

Dracnosis: A species of the superspecies C. tenuipunctata showing great variation in the development of ridges and ornamentation. There are usually rows of small puncta between weak ridges; sometimes reticulate.

DEscripTION: The left valve has a posterior hinge ear, almost straight dorsal margin, evenly rounded anterior margin, straight ventral margin, and tapered posterior. The dorsal margin of the right valve is strongly convex, with its greatest height in about the centre. There is a dimple in the postero-ventral region of both valves. Ovate in dorsal view.

The inner margin (Text-fig. 21) is fairly regular; the posterior and anterior indenta- tions are large, the anterior segment is flat, and the joint median and posterior segments have a gentle slope.

The larval stages are very triangular in shape with a pointed posterior end. The ornamentation is similar to that of the adults.

Three morphotypes have been recognized, which constitute two subspecies (Text- fig. 22).

Morphotype A:

This is characterized by the development of longitudinal ridges, usually stronger in the right valve. These follow the basic pattern of the superspecies; no. 7, with a sinuous course, joins no. 9 just before the anterior area of the reticulation; no. 8 joins no. 7 as a faint ridge. Between the ridges are parallel rows of small puncta; there are three rows between the ridges in the median and dorsal part of the valve and two in the ventral part. The sub-central plexus is only weakly developed.

316 MID-TERTIARY CYTHERETTINAE

Morphotype B:

Similar to Morphotype A except for the development of faint cross ridges between the longitudinal ones. The surface between the ridges is punctate, as in Morpho-

type A.

Morphotype C:

Strong cross ridges are developed between the longitudinal ones, which gives the valve a reticulate appearance. The surface between the ridges is smooth. The sub-central plexus 1s very weak. An additional ridge 1s present between no. 2 and 3.

Cytheretta buttensis buttensis subsp. nov. (Pl. 15, fig. 10) 1960 Cytheretta tenuistviata Mehrotra (non Reuss) p. 80, pl. 1, figs 11-12.

DIAGNOSIS AND DESCRIPTION: This consists entirely of Morphotype A.

Ho.ortyPe: Io 4030.

PARATYPE: Io 4031.

MATERIAL: See fig. 22.

TYPE LOCALITY AND HORIZON: Cormeilles-en-Parisis; Bed no. 46 of Mlle. Albissin, Couches de Sannois Supérieur.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Couches de Sannois supérieur.

DIMENSIONS (Carapace) H L/H Ww

Female 0-92 +0:04 o51 +002 1:80+0-04 0-48 Male I'02 +0:04 0°52 +002 1:95 + 0:03 0°45

Cytheretta buttensis reticulata subsp. nov. (Pl. 15, figs 1-8; Text-fig. 21) DERIVATION OF NAME: Latin—reticulatus, net-like; refers to the ornamentation.

DIAGNOSIS and DESCRIPTION: Consists predominantly of Morphotype C together with Morphotypes A and B.

Ho.ortyPe: Io 4032.

PARATYPES: Io 4033-8.

MATERIAL: See Fig. 22.

TYPE LOCALITY AND HORIZON: Cormeilles-en-Parisis; Bed no. 47 of Mlle. Albissin, Couches de Sannois supérieur.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Couches de Sannois supérieur- Marnes a Huitres infeérieurs.

OF NORTH-WEST EUROPE 317

DIMENSIONS: (Carapace) ; iD H L/H WwW Female 0-85 + 0:025 0-48 +001 1:79 40:03 0:42 Male 0:97 + 0°02 0O50+00I 1:94 +0:03 0°43

Fic. 21. Cythevetta buttensis reticulata sp. subsp. nov.; female left valve; x75.

Discussion: Morphotype A resembles C. bernesis Oertli from the Rupelian of Switzerland; they differ in dorsal outline, as C. bernesis is tapered towards the anterior. The longitudinal ridges are also unequally developed in the latter; ridges no. 3 is strong, while nos. 4, 5 and 6 are weak, and they also form a slightly different pattern. C.vamosa sublaevis Triebel from the Chattian Cyrenenmergel of the Mainz Basin is similar, but the longitudinal ridges are very weak and can hardly be seen in the right valve; the surface of the valve is also more uneven, with marked longi- tudinal swellings in the position of ridges nos. 4 and 9; these swellings are much more apparent in C. ramosa ramosa (Lienenklaus).

MORPHOTYPE ‘°/o Subspecies

SAMPLE

23 (79) 33 19 48 reticulata 24 (72) 23 7 70

Fic. 22. Distribution of morphotypes of Cytheretta buttensis sp. nov. from Cormeilles. Note that the stratigraphical sequence has youngest at the bottom.

COUCHES DE SANNOIS

MARNES A HUITRES

318 MID-TERTIARY CYTHERETTINAE

DISCUSSION OF THE SUPERSPECIES ¢C. LPENUIPUNCTATLA

The stratigraphically arranged series of specimens from the Paris Basin shows certain trends. The earliest species, C. minipunctata sp. nov. and C. buttensis sp. nov. buttensis subsp. nov., have very weakly developed ridges with several rows of puncta between them. In the later species the ridges become stronger and the puncta in between them become larger with fewer rows, as in C. tenuipunctata absoluta and C. tenuipunctata livata. In the latter the ridges are very strong and the pitting is reduced to a single row of large pits.

It is considered that the relationships suggested in Fig. 23 represent a possible phylogeny.

The barrier may be ecological rather than geographical sensu stricto. C. mintpunc- tata, or something like it, is a possible ancestor of C. tenuistriata with its two recorded geographical sub-species.

It is interesting to note that in Switzerland the finely punctate species with weak ridges, C. variabilis and C. bernesis, occur at a stratigraphically lower horizon than C. tenuistriata tenuistriata, as in the Paris Basin. In the Mainz Basin these postu- lated early forms are not present, possibly because the Sannoisian is poorly exposed, but more likely because they were not present in the area; they have not been found in samples collected nor recorded in the works of Triebel, Stchepinsky and Gramann. In the Chattian of the Mainz Basin there are, however, two finely punctate forms, C. ramosa ramosa (Lienenklaus) and C. vamosa sublaevis (Triebel). The valve of the former has an uneven surface, similar to C. variabilis, and both of these are only tentatively included in the superspecies. C.vamosa sublaevis has a much smoother valve with fine punctae between weak ridges.

C. stigmosa Triebel has a similar ridge pattern, but varies in shape and has much larger pitting. The two must be closely related, however.

The L/H ratio of all the female left valves from the Paris Basin were averaged, and gave an answer of I-79 + 0-04.

FALUN C.tenuipunctata lirata DE D JEURRE ¢, tenuipunctatastenuipunctatal| 5 9) Pa 9) pe 0 (BELGIUM) 2 C. tenuipunctata absoluta MARNES s A ic < HUITRES £ pas. SS 5 C. buttensis reticulata Hypothetical form Ee COUCHES o oO DE C. buttensis buttensis SANNOIS

Fic. 23. Suggested relationship between Cytheretta buttensis sp. nov. and Cythevetta tenuipunctata (Bosquet).

OF NORTH-WEST EUROPE 319

This standard deviation compares favourably with that of individual species, so it is concluded that this is a further character to be considered in diagnosing the superspecies.

C. minor (Lienenklaus), C. stigmosa Triebel and C. regularis sp. nov. show similari- ties to the superspecies C. tenutpunctata (Bosquet) in ornamentation, but differ in shape. C. posticalis Triebel has a similar shape to C. tenuipunctata but has almost no ornamentation.

Cytheretta minor (Lienenklaus) (Pl. 16, fig. 6)

1905 Cythereis jurinei (von Munster) var. minor Lienenklaus, p. 32. 1952 Cythevetta minor (Lienenklaus) Triebel, p. 24, pl. 4, figs. 22-3.

Diaenosis: A small species of the genus Cytheretta with smooth dorsal and antero- dorsal areas and double rows of puncta between weak ridges.

MATERIAL: 25 valves and carapaces from Weinheim (Trift). Io 3704. TYPE LOCALITY AND HORIZON: Weinheim; Unterer Meeressand.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Mainz Basin, Unterer Meeressand; Paris Basin, Couches de Sannois supérieur. Falun de Morigny.

DescripTION: In the left valve there is a posterior hinge ear; the dorsal margin is very slightly convex and the greatest height is about one quarter of the way from the anterior. The anterior margin is obliquely rounded; the ventral margin is almost straight and the posterior margin is slightly tapered. In the right valve the dorsal margin is slightly convex and the ventral margin is concave. A few denticles are present along the anterior margin of the right valve. Ovate in dorsal view.

The dorsal and antero-dorsal regions are smooth, so that ridges no. 1 and 2 are absent. To the anterior of the sub-central plexus ridges b and c are prominent, but where ridge a would be is a diffuse area of puncta. There is a double row of puncta between the ridges. In most specimens the ridges are merely areas without puncta, but this gap between the double rows of pits is greater than that between the contained single rows. In a few specimens, however, actual ridges are present.

DIMENSIONS: (Carapaces).

iG H L/H Ww Female 0-88 0°53 1:66 0°45 Male 0°85 0:50 1-70 0°43

Discussion: This could be included in the superspecies due to the similarity of the ridge pattern, even though ridge no. a is absent. The size seems to vary; the dimensions given by Triebel are less than those of the material examined from Weinheim, where it is the commonest Cytheretta species; those from the Paris Basin are in better agreement with Triebel. It shows a great resemblance to C. tenwistriata tenuistriata (Reuss), with which it is associated in the Mainz Basin. It is much smaller, however, and there are no specimens of intermediate size; it has a slightly different shape, and lacks ridge a.

320 MID-TERTIARY CYTHERETTINAE

Cytheretta posticalis Triebel

Diacnosis: A large species of Cytheretta with an elongate shape, sub-parallel dorsal and ventral margins, and prominent posterior hinge ear. It is almost smooth, with a faint ornamentation of ridges and puncta in the posterior and ventral parts of the valve.

Cytheretta posticalis posticalis Triebel

1905 Cytherers juvinu Lienenklaus (non von Miinster), p. 31. 1952 Cytheretta posticalis Triebel, p. 23, pl. 3, figs 18-21. 1956 Cytheretta posticalis Triebel, Oertli, p. 59, pl. 6, figs 160-162.

TYPE LOCALITY AND HORIZON: Welschberg (Mainz Basin), Unterer Meeressand.

STRATIGRAPHICAL RANGE AND HORIZON: Mainz Basin: Unterer Meeressand, Schleichsand, Cyrenenmergel ; Switzerland: Meeressand, Blaue Tone (both Rupelian).

Cytheretta posticalis parisiensis subsp. nov. (Pl. 18, figs 1-4, 6) DERIVATION OF NAME: Named after Paris.

Diacnosis: A subspecies of C. posticalis showing a large amount of variation in areas of ornamentation.

HototyPeE: Io 4039. PARATYPES: Io 4040-42. MATERIAL: 22 valves from Auvers-St.-George; 5 from Morigny.

TYPE LOCALITY AND HORIZON: Auvers-St.-George, Stampian.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Auvers-St.-George, Morigny; Stampian of the Paris Basin.

DESCRIPTION: Sexual dimorphism is pronounced, the males being more elongate. The left valve has a strong posterior hinge ear and a weak anterior one; this causes the dorsal margin to have an undulating appearance. This is particularly true of the male. The anterior margin is evenly rounded; the ventral margin, almost parallel to the dorsal, is slightly concave in the male and convex in the female. The posterior margin is tapered. In the right valve the dorsal margin is almost straight and the ventralis concave. It is ovate in dorsal view.

The ornamentation varies. Some specimens are completely smooth; some have a few faint double rows of puncta in the ventral portion of the postero-median region; others have a few faint ridges with puncta between in the postero-ventral angle.

The internal features are as for the superspecies C. tenuipunctata with inner margin similar to C. tenuistriata ornata.

Discussion: C. posticalis posticalis differs only in the ornamentation, which is restricted to a few ridges in the postero-ventral angle. Some specimens of

OF NORTH-WEST EUROPE 321

C. posticalis parisiensis are exactly the same, but the great variation within the sample is taken to indicate subspecific differentiation.

C. klahni Stchepinsky from the Stampian Marnes a Cyrénes of Alsace has orna- mentation confined to the median and posterior parts of the valve and consists of double rows of small puncta. In this respect it is similar to some of the specimens of C. posticalis parisiensis. However, the lateral outline has a rhomboidal appearance due to the shape of the posterior margin, which differs from C. posticalis; and in dorsal outline it is more tapered. It is much smaller, but is probably a related

species. DIMENSIONS: Left valve Right valve ie H L/H ie H L/H Female 0:95 0°53 1-79 0-91 0°47 1:94 Male I'05 0°53 1-98 1:03 0-49 2:10

Cytheretta headonensis Haskins (Pl. 18, figs 11-14; Text-fig. 24)

1857 Cytherideis colwellensis Jones (pars) p. 49, pl. 14, figs 2oa—c. 1870 Cytheve ? Jones p. 157 and 159. 1887 Xestolebevis auvantia non Baird, var. Jones and Sherborn, vol. 4, p. 456. 1889 Cytherideis colwellensis Jones, Jones and Sherborn p. 45. 1968 Cytheretta vrhenana headonensis, Haskins, p. 167, pl. 3, figs 11-18.

Diacnosis: An unornamented species of Cytheretta with a strongly obliquely rounded anterior margin.

MATERIAL: 3 valves from Headon Hill; 16 from Colwell Bay; 17 from Milford; 7 from Whitecliff Bay. Io 4043-7.

TYPE LOCALITY AND HoRIzON: Although there are grounds for believing that Headon Hill is the type locality due to etymology, in the type description Whitecliff Bay is quoted; Middle Headon Beds.

Fic. 24. Cytheretta headonensis Haskins; female right valve; 75

322 MID-TERTIARY CYTHERETTINAE

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Middle Headon Beds of Headon Hill, Colwell Bay, Whitecliff Bay, and Milford.

Discussion: This is very similar to C. rhenana Triebel; the left valve of C. headon- ensis has a much more obliquely rounded anterior margin, as well as more anterior radial pore canals (34 compared with 27).

Jones figured two different ostracods as his new species Cytheridets colwellensis; all of the material is preserved in the British Museum (Nat. Hist.), No. 1 6431 (13). Pl. 4, fig. 13 of Jones is a species of Neocyprideis, which has now been selected as the lecto- type to avoid taxonomic complications, even though the type is a moult stage and thus not very satisfactory; fortunately it is a very common species so that its diag- nosis is possible with topotype material of the adult.

Cytheretta vesca sp. nov. (PL 27; figs:S) Oyar2)

DERIVATION OF NAME: Latin—vescus, weak, little; refers to the fragile appearance of the carapace.

DiaGnosis: A species of Cytheretta with a thin shell and a weak ornamentation consisting of rows of small puncta in the posterior and latero-ventral areas of the carapace.

HoLotyPeE: Io 4048, a female left valve.

PARATYPES: Io 4049-51.

MATERIAL: 9g valves.

TYPE LOCALITY AND HORIZON: Auvers-St.-George; Stampian.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: So far known only from the type locality.

DESCRIPTION: Sexual dimorphism is distinct, the males being more elongate; sex ratio 1:1. The left valve has a weak posterior hinge-ear and a convex dorsal margin; the anterior margin is obliquely rounded; the ventral margin is straight; the posterior margin is obliquely rounded, particularly in the female. The ventral margin of the right valve is concave. The carapace is ovate in dorsal view.

The ornamentation consists of a few indistinct rows of very small puncta in the posterior and latero-ventral areas of the carapace.

The hinge of the left valve has a small antero-dorsal lobe, although the corres- ponding antero-dorsal platform of the right valve is very prominent. The antero- ventral lobe is weak; the antero-median tooth is small. The postero-median swelling is prominent in lateral view, but cannot be seen in dorsal view. The anterior tooth of the right valve is large, projecting beyond the dorsal margin; the posterior tooth is equally prominent in dorsal view, but smaller in lateral view.

The selvage forms the anterior margin; along the ventral margin there is a promin- ent flange and a wide flange groove; there is a narrow flange groove along the pos- terior. A list is present. There are some 45 normal pore canals, but the number of radial pore canals could not be determined, nor could the shape of the inner margin.

OF NORTH-WEST EUROPE 323

DIMENSIONS: Left valve Right valve ib H L/H 6, H L/H Female 0°73 0°42 1°74 0-72 0°37 1°95 Male 0-70 0:38 1-84 0°70 0°33 212

Discussion: This differs from C. rhenana in having a weak ornamentation. The ornamentation is much weaker than that of C. stigmosa. It differs from both of these in shape; the dorsal margin of the left valve is more convex and the posterior margin is more obliquely rounded. The carapace ot C. vesca is much less massive than these.

Cytheretta stigmosa Triebel

Diacnosis: A species of the genus Cytheretta with an obliquely rounded anterior margin. The ornamentation consists of longitudinal rows of large pits with ridges in between which become stronger ventrally.

Cytheretta stigmosa stigmosa Triebel 1952 Cytheretta rhenana stigmosa Triebel, p. 26, pl. 5, figs 28, 29.

TYPE LOCALITY AND HORIZON: Welschberg; Unterer Meeressand.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Rupelian Meeressand and Schleich- sand of the Mainz Basin. Cavelier (1965, determinations by Apostolescu) records it from the Sannoisian of Sannois in the Paris Basin.

Discussion: See C. stigmosa gallica sub sp. nov.

Cytheretta aff. stigmosa stigmosa (Pl. 16, fig. 8)

LOCALITY AND STRATIGRAPHICAL POSITION: Whitecliff Bay, Isle of Wight; Middle Headon, Beds (WB18). Io 4052.

Discussion: Six poorly preserved specimens, together with one well preserved one, were available for study. No internal characters could be seen. The shape is similar to C. stigmosa. The ornamentation is very similar except that the pits are rather smaller and there are more of them per row than in C. stigmosa stigmosa. The dorsal and antero-dorsal regions are smooth. It is unknown whether this is a distinct subspecies or the same as that from the Rhine Valley.

Cytheretta stigmosa gallica subsp. nov. (Pl. 17, Figs 1, 2, 5, 10; Text-fig. 25) DERIVATION OF NAME: Latin—Gallica, country of the Gauls.

Diacnosis: A subspecies of C. stigmosa in which the ornamentation covers the whole valve and the longitudinal ridges are very narrow.

324 MID-TERTIARY CYTHERETTINAE

HototyPeE: Io 4053, a female left valve.

PARATYPES: Io 4054-56.

MATERIAL: 44 valves, 4 carapaces.

TYPE LOCALITY AND HORIZON: Auvers-St-George ; Stampian.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: So far only known from the type locality.

DESCRIPTION : Sexual dimorphism can be observed, the males being more elongate; sex ratio, 1: 1°5. The left valve has a posterior hinge ear; the dorsal margin is convex with the greatest height of the valve in line with the sub-central plexus. The anterior margin is obliquely rounded; the ventral margin is almost straight; the posterior margin is obliquely rounded, but only slightly so. In the right valve the ventral margin is concave.

Ridges with pitting between them are present in the ventral part of the valve and would correspond to ridges nos. 7-13 in the terminology adopted for C. tenuipunctata; ridges b and c can also be seen. In the median and dorsal areas are some seven longitudinal rows of large pits with weak and irregular ridges between them. There is an area of small pits, not arranged in rows, in the antero-dorsal region; to the an- terior of the poorly defined sub-central plexus is an area of larger pits, while along the anterior margin there is reticulation.

In the hinge of the left valve the antero-dorsal lobe is slightly swollen; the antero- ventral lobe is small; the antero-median tooth is small but prominent; and the postero-median swelling is of equal size and prominence as the antero-median tooth.

The selvage runs close to the anterior and posterior margins so that there is only a very small flange groove present in these regions which is better seen in the right valve. Along the ventral margin of the right valve there is a wide flange groove.

The inner margin has a beak-shaped anterior indentation, prominent ventral, and a high and fairly narrow posterior indentation. The anterior segment is gently rounded, the median is short and curves into the steep posterior indentation. The distribution of pore canals could not be seen.

Fic. 25. Cytheretta stigmosa Triebel gallica subsp. nov.; male left valve; x75.

OF NORTH-WEST EUROPE 325

DIMENSIONS: L H L/H Ww Female 0-75 0°47 I-57 0:38 Male 0-74 0°42 I-76 0:36

Discussion: C. stigmosa stigmosa has unornamented dorsal and antero-dorsal areas and smaller pits with wider areas between the rows. Thus it differs from the almost reticulate appearance of C. stigmosa gallica. C. minor is rather similar but has double rows of puncta instead of single rows of large pits and unornamented areas as in C. stigmosa stigmosa; its lateral outline is also different. See also C. regularis sp. nov. and C. bullans sp. nov.

C. stigmosa is here regarded as a separate species rather than a subspecies of C. rhenana because the two are found together and should therefore be regarded as distinct species or as varieties or morphotypes of a single species. There is also a lack of intermediaries and the difference cannot be sexual because sexual dimor- phism can be recognized with each group. It should be pointed out, however, that small unornamented forms and pitted forms similar to C. rhenana and C. stigmosa are found together not only in the Rhine Valley, but in the Paris Basin, Aquitaine Basin and the Hampshire Basin. They do not always occur in the same sample, but do occur at the same locality and in adjacent horizons. They perhaps inhabited differ- ent ecological zones and could represent ecologically separated subspecies. As this is not proven, the evidence still favours them as distinct species.

Cytheretta regularis sp. nov. (Pl. 17, figs 3, 4, 6, 7; Text-fig. 26)

DERIVATION OF NAME: Latin—regularis, regular; refers to the smooth and regular lateral outline of the left valve.

Diaenosis: A species of Cytheretta with a straight dorsal margin in the left valve, sub-parallel dorsal and ventral margins, and an evenly rounded posterior margin. Ornamentation consists of longitudinal rows of pits with a smooth antero-dorsal region. Sexual dimorphism is not pronounced.

HototyPe: Io 4057.

PARATYPES: Io 4058-60.

MATERIAL: 12 valves and carapaces from Espibos, 6 from Lesbarritz.

TYPE LOCALITY AND HORIZON: Lesbarritz, Gaas (AGLI); Stampian.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Lesbarritz and Espibos, Gaas; Stampian.

DESCRIPTION: Sexual dimorphism is not very pronounced, the males being slightly more elongate; sex ratio 1:2. The dorsal margin of the left valve is straight with no posterior hinge ear; the anterior margin is slightly obliquely rounded; the ventral margin is almost straight ; and the posterior margin is evenly rounded. In the right valve the dorsal margin is convex and the ventral is concave. The dorsal and ventral margins are almost parallel. In dorsal view the carapace is ovate.

326 MID-TERTIARY CYTHERETTINAE

The ornamentation consists of 13 longitudinal rows of pits between narrow ridges. The sub-central plexus is fairly prominent although not appearing so in the electron scanning photographs, and to its anterior are two prominent ridges which slope towards the antero-ventral angle. The anterior and posterior regions are covered by numerous small pits. The extreme antero-dorsal area is smooth.

In the hinge of the left valve the antero-dorsal lobe is small and slightly swollen; the antero-ventral lobe is poorly developed; the antero-median tooth is small and the postero-median swelling is hardly noticeable. In the right valve the anterior tooth is much smaller and globose in shape.

The selvage is very close to the margins of the valve with a flange groove developed along the ventral margin and a small one along the posterior margin. The inner margin has three prominent indentations; the anterior one is rather ill defined in the specimens available, but is narrow; the ventral one is very long and narrow; the posterior one is small. The anterior segment is unevenly rounded; the median segment is short with a gentle curve; the posterior segment is long with a fairly steep slope. There are some 27 anterior radial pore canals, tending to be grouped into five sets; 35 closely spaced posterior radial pore canals; 10 ventral radial pore canals; and some 38 normal pore canals, which are not related to the ornamentation except that they mainly open into the pits of the outer surface.

The central muscle scars are in a slight pit with four equal small and circular adductor muscle scars along the posterior edge of the pit and a large frontal muscle scar on the anterior edge. The fulcral point is not very prominent.

DIMENSIONS: Left valve Right valve L H L/H 8 H L/H Female 0-70 0:38 1-84 0:68 0°35 1-94 Male 0:70 0°37 1°88 0-68 0°33 2:06

Discussion: This differs from C. stigmosa in shape, particularly of the dorsal margin. Theornamentation is very similarto C. stigmosa gallica and also to C. bullans. It differs from the latter in the shape of the anterior margin, in size, and in the constancy of development of the ornamentation. It differs from C. minor in having single rows of large pits instead of double rows of puncta, as well as in shape.

Fic. 26. Cytheretta vegularis sp. nov.; male right valve; x75.

OF NORTH-WEST EUROPE 327

Cytheretta bullans sp. nov. * (PION nes 5.77 bl. 22, fg. 1)

DERIVATION OF NAME: Latin—bulla, bubble; refers to the ornamentation which resembles strings of bubbles.

Dracnosis: A species of the genus Cytheretta with parallel dorsal and ventral margins and evenly rounded anterior and posterior margins. Ornamentation consists of longitudinal rows of pits often with a large unornamented anterior region.

HoLotyPeE: Io 4061.

PARATYPE: lo 4062.

MATERIAL: 18 carapaces.

TYPE LOCALITY AND HORIZON: Chateau Romefort, Blaignan; Argile a Algues, Sannoisian.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Known only from the type locality.

DESCRIPTION : Sexual dimorphism is distinct, the males being more elongate; sex ratio, 1: 2. The dorsal and ventral margins of the left valve are parallel. The dorsal margin of the left valve is straight without a posterior hinge ear; the anterior margin is evenly rounded; the ventral margin is very slightly concave; the posterior margin is evenly rounded. In dorsal view it is ovate with a tapered anterior end.

The degree of development of the ornamentation varies. There are some thirteen rows of pits with ridges between to the posterior of the ill-defined sub-central plexus. To the anterior are two prominent ridges which slope towards the antero-ventral angle. There is an anterior area of reticulation with small pits in between. The antero-dorsal area is smooth. This unornamented anterior area varies in size; in some specimens the whole anterior area is smooth and in others the ornamentation is restricted to six or seven rows of pits in the postero-median position. It must be emphasized that this is not a form of sexual dimorphism, as might be inferred from Pi. x8, figs 5, 7.

Internal characters could not be seen.

DIMENSIONS: Carapace

Te H L/H Ww Female 0°85 0:48 177 0°43 Male 0-90 0°45 2:00 0-40

Discussion: See C. vegularis. It differs from C. stigmosa in lateral shape, and from C. minor in having single rows of large pits instead of double rows of puncta, as well as in shape.

Cytheretta sagri Deltel

Diacnosis: A species of Cytheretta with thirteen longitudinal ridges, often only present in the posterior. The inner margin has a characteristic shape with a depressed median segment markedly separated from the anterior and posterior segments.

328 MID-TERTIARY CYTHERETTINAE

DESCRIPTION: Six merphotypes divisible into three subspecies have been recog- nized; these are described below. The shape and ornamentation vary to a great extent, but the internal structures appear to be constant. Another feature that is constant is the five posterior spines.

The hinge of the left valve has a prominent swollen antero-dorsal lobe; a strong antero-ventral lobe; a large antero-median tooth; a very weak postero-median swell- ing, and a large posterior socket. In the right valve the anterior tooth is much larger than the posterior one.

The selvage runs very close to the anterior margin; a wide flange groove is present along the ventral margin with a narrow one along the posterior. The inner margin is very characteristic of the species. The anterior and posterior indentations are narrow; the ventral indentation is narrow and ‘V’-shaped. The anterior and posterior indentations are semi-circular; the median segment is sharply differentiated from these, lying close to the ventral margin and with a postero-ventral indentation.

There are 25 anterior radial pore canals, 33 posterior, and 14 ventral. The central muscle scars are in a pit; the two lowest adductors are almost joined, and the frontal is inside the pit. The fulcral point is large and prominent.

Morphotype A:

Sexual dimorphism is pronounced, the males being more elongate. The left valve has a strong posterior hinge ear and a weak anterior one; the dorsal margin between these is symmetrically convex in the female and asymetrically convex in the male with the steep slope towards the posterior. The anterior margin is slightly obliquely rounded with some nine denticles in the ventral half. The ventral margin of the female is straight, while that of the male is concave. The posterior margin is evenly rounded. The dorsal margin of the right valve has a marked protuberance in the anterior half caused by the high position of the antero-median socket of the hinge; the ventral margin is concave. In dorsal view the female is ovate and tapered towards the anterior; the male is more bullet-shaped.

The ornamentation consists of thirteen longitudinal ridges. In the right valve ridges nos. 4 and 8 are sometimes very strong with a slight depression developed between them. Ridge no. 6 is thin, bifurcating at its anterior end just to the posterior of the sub-central plexus, one part joining no. 5 and the other no. 7. This ridge is always weak in the right valve, but in some left valves it is strong, stronger in fact than no. 7, so that it appears that no. 7 joins it instead of the other way round. To the anterior of the weak sub-central plexus are four prominent ridges sloping towards the antero-ventral angle. Faint cross-ridges and meandriform punctation are developed between the longitudinal ridges (see Pl. 21, fig. 5 for meandriform punctation).

Morphotype B:

This differs slightly from Morphotype A in shape; it has a more rectangular outline due to the evenly rounded anterior margin, and in dorsal view it is ovate, not tapered. The whole of the anterior margin is denticulate with some twelve denticles. The

OF NORTH-WEST EUROPE 329

ornamentation is similar to that of Morphotype A, except that there is a smooth area of varying extent in the antero-dorsal region. The internal details could not be seen.

Morphotype C:

The female left valve has no anterior hinge ear, so the antero-dorsal angle is a smooth curve, unlike Morphotypes A and B. In dorsal view it is ovate. The orna- mentation is restricted to the posterior part of the valve where 9-13 ridges can be seen. Ridge no. 6 is clearly recognizable and of equal strength to the other ridges. No internal details could be seen.

Morphotype D:

This is similar in shape to Morphotype C and the ornamentation is also restricted to the posterior. It differs in the inequality of the ridges; no. 6 in particular is weaker.

Morphotype E:

The female left valve has no anterior hinge ear, but the lateral outline of the carapace differs from Morphotypes C and D in being almost triangular with a very prominent posterior hinge ear. The carapace is unornamented over a large anterior and antero-dorsal area; longitudinal ridges are present over the remaining surface with a very fine meandriform punctation between them. Ridge no. 6 is very weakly developed. This differs from Morphotype B in shape and in having a much weaker ornamentation without the cross-ridges present between the longitudinal ridges.

Morphotype F :

This is very similar to Morphotype E, except that the ornamentation is restricted to the posterior half of the valve. Ridge no. 6 is very faint and thread-like, leaving a prominent gap between nos. 5 and 7; in this respect it differs from Morphotype C.

Cytheretta sagri sagri Deltel (Pl. 19, figs. 1-4; text-fig. 28) 1964 Cytheretta sagvi Deltel, p. 156, pl. 3, figs 56-57.

DIAGNOSIS AND DESCRIPTION: A subspecies of C. sagvi consisting predominantly of Morphotype A with Morphotype D.

The last two larval stages have been recognized. The ornamentation in no. 8 consists of two prominent ridges, nos. 4 and 8 of the adult, with the other ridges of the adult stage weakly developed. Cross-ridges are sometimes present; puncta are present between the ridges. There are four posterior spines and eleven anterior denticles, each bearing one of the eleven anterior pore canals. In the seventh larval stage the two ridges nos. 4 and 8 are present. There are seven anterior radial pore canals and denticles and two posterior spines.

330 MID-TERTIARY CYTHERETTINAE

MATERIAL: See fig. 27. Io 4063-6. TYPE LOCALITY AND HORIZON: Lesbarritz, Gaas; Stampian. STRATIGRAPHICAL RANGE AND DISTRIBUTION: Stampian of Gaas (Lesbarritz Espibos) and Bastennes-Gaujacq, Aquitaine Basin. DIMENSIONS: Morphotype A Left valve Right valve ik, H L/H 4W 1G H L/H Female 084 0:45 1:87 40:22 0:84 0:42 2:00 Male 0-00" “O47 “"E-9r ~~ 0-247 —'0-g0) Fo-45" 9 2-0G

Morphotype D (Carapace) :

Female 0:92 0°53 1-74 0-46

MORPHOTYPE SAMPLE

RO 270 RO 271

Fic. 27. Distribution of Morphotypes of Cytheretta sagvi Deltel. Cytheretta sagri inconstans subsp. nov.

(Pl. 19, figs 5-7, 9)

DERIVATION OF NAME: Latin—inconstans, the opposite of standing firm, or inconstant ; refers to the great variation of shape and ornamentation.

OF NORTH-WEST EUROPE 331

DIAGNOSIS AND DESCRIPTION: A subspecies of C. sagri showing variation in shape and ornamentation. The latter consists of thirteen longitudinal ridges which in some specimens cover the whole valve and in others only the posterior region. It consists predominantly of Morphotype C with A and B.

HototyPe: Io 4067, a female left valve.

PaRATYPES: Io 4068-70.

MATERIAL: See fig. 27.

TYPE LOCALITY AND HORIZON: Chateau Romefort, Blaignan; Argiles a Algues, Sannoisian.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: So far only known from the type locality.

DIMENSIONS: Carapaces Morphotype B Morphotype C L H L/H Ww L H L/H Ww

Hemale 078) 0:43. 1:81 0-38 0°92 «0°53 ) 1-74) 0-47 Male ©788 40°43) “1-93-0309 0-95) 0:47" 2:02" 0742

Cytheretta sagri martini subsp. nov. (Pl. 20, figs 1-4)

DERIVATION OF NAME: From the Phare St. Martin, Biarritz.

DIAGNOSIS AND DESCRIPTION: A subspecies of C. sagri of a triangular shape in lateral view and with a weak ornamentation. It consists of Morphotypes D, E and F, particularly the last two.

HorotyPe: Io 4071, a female left valve. PARATYPES: Io 4072-6. MATERIAL: See fig. 27.

Fic. 28. Cythevetta sagyi Deltel; female left valve. 75

332 MID-TERTIARY CYTHERETTINAE

TYPE LOCALITY AND HORIZON: Couches du Phare, Biarritz (RO 270) ; Stampian. STRATIGRAPHICAL RANGE AND DISTRIBUTION: So far only known from the type

locality. DIMENSIONS: Morphotype E Left valve Right valve 1, H L/H 1G H L/H Female 0:83 0°45 1-84 0-81 0-40 2°03 Male 0-86 0°45 I‘QI 0-90 0°43 2°09 Morphotype F (Carapace): is H L/H w Female 0°85 0-47 1-81 0°39 Male 0-87 0°45 I'93 0:40

Discussion: C. sagri Deltel may be the form described by Reuss (1869) as Cythera multinervis sp. nov. (p. 482, pl. 6, fig. 2).

The Oligocene of Aquitaine contains a group of closely related species: C. sagrz Deltel, C. gibberis sp. nov., C. minipustulosa sp. nov., C. postornata sp. nov., and C. samothracia Deltel; C. perita Deltel from the Upper Eocene is perhaps related to this group.

C. sagri (Morphotype A), C. minipustulosa and C. samothracia have an unusual meandriform punctation between the longitudinal ridges, but the pattern of the ridges differs amongst the three species. C. sagyi (Morphotypes C, D and G), C. gibberis, C. perita and C. postornata are similar in that the ornamentation is restricted to the posterior. C. gibberis differs in the unusual shape of the right valve with its dorsal ““‘hump”’; C. perita differs in shape in having only four posterior spines and in the shape of the inner margin; C. postornata also differs in shape and in ornamenta- tion, which consists of six sulca, one of which reaches to the centre of the carapace.

C. tenuipuncta (Bosquet), C. tenuistriata (Reuss) and C. buttensis sp. nov. reticulata subsp. nov. are similar to C. sagri (Morphotype A), but have a different ridge pattern and lack the characteristic meandriform punctation. C. buttensis reticulata has cross-ridges similar to C. sagvi (Morphotypes A and B), but differs by the features already mentioned. C. posticalis Triebel has the ornamentation restricted to the posterior, but this is much weaker than the omamentation of C. sagvi (Morphotypes C, D and G) and is developed in a more ventral position; it also differs in shape.

Cytheretta samothracia Deltel (Pl. 21, figs 5, 6, 8) 1964 Cytheretta samothvacia Deltel, p. 158, pl. 3, figs 58-60.

DIAGNOSIS AND DESCRIPTION: Sexual dimorphism is pronounced; the female carapace is quadrate in lateral view, the male is rectangular. There are five promi- nent posterior spines. Ornamentationconsists ofeleven longitudinal ridges, including

OF NORTH-WEST EUROPE 333

two prominent parallel ridges in the median part of the valve and two in the dorsal part. Between the ridges is a meandriform punctation and a faint reticulation.

MATERIAL: 36 valves and carapaces from Biarritz. Io 4077-79. TYPE LOCALITY: Bastennes-Gaujacq ; Stampian.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Bastennes-Gaujacq, Lourquen; Stampian. Couches du Phare, Biarritz; Stampian.

DIMENSIONS: Left valve Right valve i H L/H 1 H L/H Female 0:95 0:56 1-70 0:95 0°49 I-94 Male I-04 0°55 1-89 I-00 0°50 2:00

Discussion: In shape and ornamentation this is easily distinguished from other species of Cytheretta. See also C. sagri and C. minipustulosa.

Cytheretta minipustulosa sp. nov. (Pl. 21, figs 1-4; Text-fig. 29) DERIVATION OF NAME: Latin—mini, small; pustulosa, full of pimples; refers to the ornamentation between the longitudinal ridges.

Draenosis: A species of Cytheretta with eleven longitudinal ridges, four of which join in the anterior to form two concentric ovals open towards the posterior. Between the ridges are faint cross-ridges and a meandriform punctation.

Ho.otyPe: Io 4080, a male right valve.

PARATYPES: Io 4081-83.

MATERIAL: I5 valves and carapaces from Biarritz; 1 valve from Gaas.

TYPE LOCALITY AND HORIZON: Biarritz; Couches du Phare superieur.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Couches du Phare, Biarritz (Stampian); Espibos (Gaas), Stampian.

Description: In lateral view the carapace is rectangular. Sexual dimorphism is not very pronounced, the males being more elongate. The left valve has a posterior

Fic. 29. Cytheretta minipustulosa; male right valve; x75.

334 MID-TERTIARY CYTHERETTINAE

hinge ear and a weak anterior one; the dorsal margin between is slightly convex. The anterior margin is almost evenly rounded. The ventral margin is slightly con- cave, particularly in the male. The posterior margin is evenly rounded with five spines. In the right valve the dorsal margin has a protuberance due to the high position of the antero-median socket. The ventral margin is concave; the posterior has five spines, and there is a marked concavity in the postero-dorsal angle. In dorsal view the female is tapered, while the male has more or less parallel sides.

Ornamentation consists of eleven longitudinal ridges. Ridges nos. 2 and 6, and nos. 3 and 5 join in the anterior and form two concentric ovals, open towards the posterior. Ridge no. 4, which is weak, runs down the centre. Nos. 5 and 6 join towards the posterior. Between the longitudinal ridges are faint cross-ridges and a meandriform punctation. There is no sub-central plexus.

The hinge of the left valve has a swollen antero-dorsal lobe; a strong antero-ventral lobe ; a deep anterior socket ; a small antero-median tooth; and a weak postero-median swelling. In the right valve the anterior tooth is large and pointed; the posterior tooth is small.

The selvage is strong. It runs close to the anterior margin but a small flange groove is present; the flange groove along the ventral margin is not large. The selvage is very strong in the posterior forming a projecting ridge, to the posterior of which is the flange groove. The latter has more the appearance of a platform; the flange is weak.

The inner margin has a characteristic shape. The anterior and posterior segments are narrow and deep; the ventral indentation is small, but because of the shape of the posterior segment it 1s very prominent. The anterior segment is semi-circular; the median segment is small and overshadowed by the steeply curved posterior segment, which sweeps up close to the dorsal margin.

The two lower adductor muscle scars touch; the third is elongate; the topmost one is triangular. The fulcrum is not very prominent. No pore canals could be seen.

DIMENSIONS: Left valve Right valve L H L/H Ww 1g H Ue Female 0:88 0-48 1°83 — 0-98 0-48 2:04 Male I-00 0-48 2:08 0:40 I-09 0-50 2-18

Discussion: This differs from C. sagvi Deltel by the ridges which form two con- centric ovals open towards the posterior, in its elongate shape, and in the shape of the inner margin. It differs from C. samothracia Deltel in lacking the two sets of paired ridges, as well as in its more elongate shape.

Cytheretta gibberis sp. nov. (Pl. 16, figs 9, 10; Pl. 19, figs 10)

DERIVATION OF NAME: Latin—gibberis, hump on the back; refers to the shape of the female right valve.

OF NORTH-WEST EUROPE 335

DiaGnosis: A species of Cytheretta in which the right valve of the female is very high in the posterior. Ornamentation is restricted to the posterior and consists of twelve radial sulca, the central one being longer than the others.

HorotyPe: Io 4084, a female right valve.

PARATYPES: lo 4085-86.

MATERIAL: g valves and carapaces.

TYPE LOCALITY AND HORIZON: Couches du Phare superieur (RO 271), Biarritz.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: So far only known from the type locality.

DESCRIPTION: Sexual dimorphism is distinct, particularly in the right valve. The left valve has a weak posterior hinge ear, in front of which is a slight concavity; the remainder of the dorsal margin is convex. The anterior margin is obliquely rounded ; the ventral margin is straight; the posterior margin is evenly rounded with a few small spines. The right valve of the temale has a very high posterior dorsal margin, which over-reaches the left valve in the position of the concavity adjacent to the hinge ear. The dorsal margin slopes steeply to the position of the anterior tooth, in front of which is a concavity with the antero-dorsal platform of the hinge. The anterior margin is evenly rounded; the ventral margin is straight. The ventral part of the posterior margin bears some four spines, although the exact number could not be determined; in the dorsal part there is a large concavity. The right valve of the male is not so high posteriorly and has a concave ventral margin. In dorsal view the carapace is ovate.

The ornamentation is restricted to the posterior and consists of some eleven short, radiating sulca with a long central one which reaches to the central region of the carapace. There is a slight postero-ventral depression in the right valve.

Owing to the poor preservation of the material, the internal features could not be completely observed. The hinge of the left valve has a strong antero-dorsal lobe and a prominent antero-median tooth. The right valve has a large antero-dorsal plat- form and a large anterior tooth; the posterior tooth is fairly small and lies along the postero-dorsal concavity, almost at right angles to the dorsal margin.

The selvage is prominent with a wide flange groove along the anterior, ventral and posterior of the right valve. A weak list is present in the anterior and postero-ventral regions. The inner margin could not be clearly seen; the anterior and posterior indentations are deep and narrow; and the anterior segment is short and semi- circular.

DIMENSIONS: Carapace Right valve L H L/H W IE, H L/H Female 0:86 0°52 1-65 0°43 0:88 0:48 1°83 Male a — — ~ 0:92 0°47 I-96

Discussion: The shape of the female right valve distinguishes this form from all other Cytheretta spp. in which the ornamentation is restricted to the posterior.

336 MID-TERTIARY CYTHERETTINAE

Cytheretta postornata sp. nov. (Pl. 20, figs 5-8; Pl. 22, fig. 12)

DERIVATION OF NAME: Latin—post, posterior; ornata, ornament; refers to the ornamentation which is restricted to the posterior.

Diacnosis: A species of Cytheretta in which the ornamentation is restricted to the posterior. This consists of 6-8 prominent sulca, one of which reaches to the centre of the carapace.

HototyPe: Io 4087, a female left valve.

PARATYPE Io 4088.

MATERIAL: IO carapaces.

TYPE LOCALITY AND HORIZON: Biarritz; Couches de l’Atalaye (RO 264).

STRATIGRAPHICAL RANGE AND DISTRIBUTION: So far only known from the type locality.

DESCRIPTION: Sexual dimorphism is not very strong. The dorsal margin of the left valve is strongly convex, less soin the right valve. The anterior margin is evenly rounded; the ventral margin is straight in the left valve, slightly concave in the right. The posterior margin is obliquely rounded and has five spines.

Ornamentation is restricted to the posterior half of the carapace and consists of some 6-8 sulca. These are mainly short, except for the central one which reaches to the centre of the carapace. Within this long sulcus is a fine threadlike ridge. No internal details could be seen.

DIMENSIONS: Carapaces

ie H L/H Ww Female 0:83 0-49 I-69 0°43 Male 0-84 0°47 I-79 0°42

Discussion: This is similar to C. sagri Deltel (Morphotypes C, D and F), but differs from these in L/H ratio as well as ornamentation. C. posticalis Triebel has a com- pletely different shape; C. perita Deltel differs in shape, L/H ratio, ornamentation, and has only four posterior spines.

Cytheretta perita Deltel (Pl. 19, fig. 8) 1964 Cytheretta perita Deltel, p. 155, pl. 3, figs 53-55. DIAGNOSIS AND DESCRIPTION: The female is triangular in lateral view. There are

some eight anterior denticles and four posterior spines. Ornamentation is restricted to the posterior, where there are six short ridges.

MATERIAL: I0 valves and carapace from Lespontes. Io 4089. TYPE LOCALITY: Coupe de Lespontes, Peyrehorade; Bartonian.

OF NORTH-WEST EUROPE 337

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Coupe de Lespontes (Moulin de Basat; Ferme Le Vigneau). DIMENSIONS: Carapace 16 H L/H Ww Female 0-81 0°43 1-88 0°35

Discussion: This differs from C. posticalis Triebel in shape, size, inner margin, and in having stronger posterior ridges. See also C. postornata sp. nov.

Cytheretta sculpta Ducasse (Pl. 20, figs 9, 10) 1964 Cytheretta sculpta Ducasse, p. 225, pl. 1, figs 2-4.

Diacnosis: A species of Cytheretta with a prominent anterior hinge ear in the left valve. The ornamentation consists of eleven longitudinal ridges with faint cross- ridges between them.

MATERIAL: 9 carapaces. Io 4090-01, Argiles a Algues, Blaignan. TYPE LOCALITY AND HORIZON: Villeneuve-de-Blaye, Eocéne supérieur.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Upper Eocene and Sannoisian of the Bordeaux region.

DESCRIPTION: Sexual dimorphism is distinct, the males being more elongate. The dorsal margin of the left valve is slightly convex with a prominent anterior hinge ear. The anterior margin is obliquely rounded; the ventral margin is slightly concave; the posterior margin is evenly rounded. The right valve has a more typical Cyther- etta shape due to the lack of the anterior hinge ear. There are some twelve denticles along the whole of the anterior margin and four spines along the posterior margin; these are present in both valves. The carapace is tapered towards the anterior end in dorsal view.

The ornamentation consists of eleven longitudinal ridges; nos. 2, 4, 7 and 10 run the whole length of the carapace; no. 1 forms the dorsal margin and in the left valve curves sharply downwards by the hinge ear to join no. 2. No. 3 is faint; no. 5 joins no. 4 just before the anterior margin; no. 6 is faint; no. g joins no. 8 in the centre. Another ridge is present just below the hinge ear of the left valve. There is a strong anterior marginal rim. Between the ridges are faint cross-ridges. There is no sub- central plexus.

No internal features could be seen. The antero-dorsal lobe of the hinge is very strong.

DIMENSIONS: Carapaces

Y, R L/H Ww Female 0-70 0:40 1°75 0°33 Male 0°73 0:38 I-92 0°33

Discussion: The ornamentation is unlike that of any other described species of Cytheretta.

338 MID-TERTIARY CYTHERETTINAE

Cytheretta sp. A (PIS GO fies 11) MATERIAL: I carapace. lo 4092.

LOCALITY AND HORIZON: Moiselles; Sables de Beauchamp.

DIMENSIONS: Left valve, male: L, 0:74; H, 0-39; L/H, x-90.

Discussion: This is very similar to C. ruelensis sp. nov.; the ridge pattern is the same, but the ridges are all of about equal strength. This is probably an individual of a species ancestral to C. ruelensis.

Cytheretta sp. B (Pl. 9; fig22)

MATERIAL: 2 broken right valves, 2 distorted carapaces. Io 4093.

LOCALITY AND HORIZON: Biarritz: Couches a Pentacrinus de la Cote des Basques (RO 254; RO 255); Couches des Bains (RO 258).

DESCRIPTION: This has a posterior hinge ear in the left valve and four posterior spines. The ornamentation consists of eleven longitudinal ridges, one of which forms the dorsal margin; ridge no. 6 is short, not reaching to the anterior half of the valve. There is a strong anterior marginal rim and a wide anterior area of reticulation. A weak reticulation is present between the longitudinal ridges.

DIMENSIONS:

Right valve: L, 0-75; H, 0:39; L/H, 1-92.

Discussion: This is of interest as the only Cytheretta sp. found in the Couches a Pentacrinus. There are no other species with which it can be compared.

Cytheretta sp. C (Pl. 3, fig. 10) MATERIAL: I right valve, L, 0-70.

LOCALITY AND HORIZON: Sables d’Auvers, Auvers-sur-Oise.

Discussion: The ornamentation of this valve is very similar to that of C. bambrug- gensis Keij, but it differs from the latter in having its greatest height situated more to the posterior. The specimen was unfortunately destroyed while being photo- graphed with the electron scanning microscope, but is left here for the record.

Genus FLEXUS Neviani 1928

1928 Flexus, Neviani, p. 26. 1958 Eucytheretta Puri, p. 188.

TYPE SPECIES: Cythere plicata von Munster.

OF NORTH-WEST EUROPE 339

DraGnosis: Similar to Cytheretta but with the development of three prominent longitudinal ridges. Ornamentation between the ridges varies. The carapace tends to be more elongate than Cytheretta.

Discussion: See Introduction.

Flexus plicatus (von Munster) (Pli22. tiger)

1830 Cytheve plicata von Munster, p. 63.

1838 Cythere plicata von Munster, Roemer, p. 518, pl. 6, fig. 26.

1850 Cypridina plicata (von Munster), Ruess, p. 83, pl. 10, fig. 21. 1896 Cythere plicata von Munster, Lienenklaus, p. 141.

1952 Cytheretta plicata (von Munster), Triebel, p. 28, pl. 5, figs 34-35. 1956 Cytheretta plicata (von Munster), Oertli, p. 65, pl. 8, fig. 194. 1958 Eucytheretta plicata (von Munster), Puri, p. 188, pl. 3, figs 1-6.

TYPE LOCALITY AND HORIZON: Astrup, near Osnabruck; Upper Oligocene. MATERIAL: 2 carapaces from Astrup. Io 4094.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: F’. plicatus has been recorded from a great variety of localities and horizons. Authenticated occurrences however seem to be restricted to the Upper Oligocene of Astrup and Doberg.

Discussion: This occurs together with a form resembling F. concinnus (Triebel) ; samples from the Upper Oligocene of Bithl near Weimer (Kassel) contain only the latter. This is the form figured and described by Speyer (1863, pl. 4, fig. 2) and men- tioned by Lienenklaus (1894, p. 198).

Flexus concinnus (Triebel) (BiP22Mfigs2) 355)

1852 Cythere plicata Bosquet (pars) (non von Miinster), p. 60, pl. 2, fig. 13. 1895 Cythere plicata Lienenklaus (non von Minster), p. 17.

1905 Cythereis tlicata Lienenklaus (non von Miinster), p. 37, 64.

1952 Cytheretta concinna Triebel, p. 27, pl. 5, figs 31-33.

1957 Cytheretta concinna Keij, p. 132, pl. 10, fig. 6.

MATERIAL: Alzey: 10 valves; Auvers-St-George: 3. lo 4095-97. TYPE LOCALITY AND HORIZON: Welschberg; Unterer Meeressand.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Unterer Meeressand of the Mainz Basin; Stampian of Jeurre, Auvers-St.-George, and Morigny in the Paris Basin; Sables de Berg and Argiles a N. comta, Belgium (Sables de Wemmel and Argiles d’Asche, Belgium ?).

F

340 MID-TERTIARY CYTHERETTINAE

Flexus gutzwilleri (Oertli) (Pl. 22, fig: 4) 1956 Cytheretta gutzwilleri Oertli, p. 64, pl. 8, figs 189-192. MATERIAL: 31 valves and carapaces from the topmost Couches du Phare (RO 270, 271). Io 4098-99.

TYPE LOCALITY AND HORIZON: Therwil (near Basel); Cyrenenmergel (Lower Chattian).

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Lower Chattian of Therwil; Couches du Phare, Biarritz.

Flexus solentensis sp. nov.

DERIVATION OF NAME: After the Solent.

Diacnosis: A small species of the genus Flexus with a thick anterior marginal rim and thick longitudinal ridges. Two subspecies have been recognized.

Flexus solentensis solentensis subsp. nov. (Pl. 23, figs 7-10; Text-fig. 30)

1957 Cytheretta gracilicosta Keij (non Reuss), p. 135, pl. 10, fig. 5. 1968 Cytheretta gracilicosta Haskins (non Reuss), p. 166, pl. 3, figs 1—10.

HototyPeE: Io 4100, a female left valve. PARATYPES: Io 4101-2.

MATERIAL: Barton: EBA 1 (Bed F), 4 valves; EHC 2 (Bed D), 2 carapaces. Alum Bay: Middle Barton Beds, 5 valves and carapaces.

TYPE LOCALITY AND HORIZON: Barton; Middle Barton Beds, Bed F.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Middle Barton Beds of Barton and Alum Bay.

DIAGNOSIS AND DESCRIPTION: Sexual dimorphism is distinct; sex ratio, I : 3 The left valve has a strong posterior hinge ear and a very weak anterior one; the

Fic. 30. Flexus solentensis solentensis; female left valve; X75.

OF NORTH-WEST EUROPE 341

dorsal margin is convex. The anterior margin is almost evenly rounded; the ventral margin is concave, particularly in the right valve. The posterior margin is obliquely rounded. In dorsal view all three ridges can be seen, giving the carapace a tapered appearance with the apex at the anterior.

Ornamentation consists of three longitudinal ridges which end against a thick anterior marginal rim. The dorsal ridge forms the dorsal margin between the two hinge ears, curving downwards just to the posterior of the anterior hinge ear. The median ridge is roughly parallel to the dorsal ridge, but with a less accentuated course. The ventral ridge is almost straight. At the posterior the ventral ridge joins the median one and this remaining thin ridge then joins the thin posterior part of the dorsal ridge. Between the ridges is a coarse reticulation of irregular cross-ridges.

The hinge of the left valve has a small swollen, but very prominent, antero-dorsal lobe; the antero-ventral lobe is small; the antero-median tooth is large and the postero-median swelling is almost as big. In the right valve the anterior margin is pointed and the posterior tooth is almost equal in size. The inner margin does not appear to have a very well developed anterior indentation; the ventral and posterior indentations are narrow and deep. The median and posterior segments form a continuous steep curve, going a long way towards the dorsal margin.

There are 24 anterior, 22 posterior and 20 ventral radial pore canals. The selvage is prominent; there is a flange groove along the anterior, posterior and ventral margins; a list is developed in the antero-ventral and postero-ventral areas.

DIMENSIONS: Left valve Right valve 1B H L/H WwW iD H L/H Female 0°52 0-31 1-68 0°25 0-51 0:27 1-89 Male 0°54 0:28 1:93 — — _- —

Flexus solentensis congestus subsp. nov. (Pl. 23, figs 11-15)

DERIVATION OF NAME: Latin—congestus, dense, thick; refers to the longitudinal ridges.

Ho.otyPe: Io 4103, a female left valve.

PARATYPES: lo 4104-5.

MATERIAL: EBA 4, 7 valves and carapaces (5 females, 2 males).

TYPE LOCALITY AND HORIZON: Barton; Upper Barton Beds (Chama Bed, H).

STRATIGRAPHICAL RANGE AND DISTRIBUTION: So far only known from the type locality and horizon.

DIAGNOSIS AND DESCRIPTION: Shape and internal features are as for the nominate subspecies. The ornamentation consists of three very thick ridges which merge into

a thick anterior marginal rim. The ridges are thicker than the intervening areas; the latter have a fine, uneven reticulation.

Fe

342 MID-TERTIARY CYTHERETTINAE

DIMENSIONS: Left valve Right valve if, H L/H WwW L H L/H Female 0-51 0-30 I-70 0:24 0-51 0:25 2°04 Male 0°54 0:29 1-86 0°25 — a —

Discussion: The oldest specimens of F. solentensis show similarities to F. ludensis sp. nov. in shape, size and ornamentation. In detail, however, the ridges are thicker and the reticulation between them consists of cross-ridges rather than the uneven reticulation of F. Judensis. They are quite probably related species however. In younger beds the ridges thicken until in the Upper Barton Beds the end member of the series is met with and is here distinguished as a separate subspecies, F. solentensis congestus. The remainder of the Barton Clay has not yielded any ostracods. The overlying Brockenhurst Beds contain a form of F. Judensis which must have migrated into the area with the Headon Beds transgression.

F’. gracilicostus (Reuss) shows similarities to F. solentensis and F. ludensis, but has much finer ridges and a smaller and more even reticulation between them. In dorsal view it is more ovate and the three ridges do not stand out as in F. solentensis and F. ludensis. F. gracilicostus is also much larger.

Flexus ludensis sp. nov. (Pl. 23, figs 1-6, 16) DERIVATION OF NAME: After the Marnes a P. ludensis in which it is found.

DiaGnosis: A small species of the genus Flexus with thick longitudinal ridges; at the posterior the dorsal and ventral ridges join the median one; at the anterior the ridges join a strong marginal rim. Between the ridges is an uneven reticulation. At the posterior are three small spines.

HototyPeE: Io 4106, a female left valve.

PARATYPES: Io 4107-12.

MATERIAL: Verzy: PVY 2, 8 valves and carapaces; PVY 4, 22. Chavencon: PCC 2, 9 valves and carapaces. Whitecliff Bay: EWB(A), 2 valves and carapaces; EWB(B), 3; EWB 19, 2; EWB 22,1. Headon Hill: EHH 42, 4 valves and carapaces.

TYPE LOCALITY AND HORIZON: Verzy; Marnes a P. ludensis.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Marnes a P. ludensis of Verzy and

Chavengon. Brockenhurst Beds, Whitecliff Bay; Middle Headon Beds, Headon Hill, Whitecliff Bay.

DESCRIPTION: Sexual dimorphism is pronounced; sex ratio, I : 2:5. The left valve has a strong posterior hinge ear and a weak anterior one; the dorsal margin is convex. The anterior margin is almost evenly rounded with a few marginal denticles in the ventral portion. The ventral margin is straight in the anterior half, curving round into the posterior margin in the posterior half. The posterior margin is obliquely rounded and has three small spines in the median portion. The right valve

OF NORTH-WEST EUROPE 343

has a concave ventral margin. In dorsal view all three ridges can be clearly seen, giving the carapace a tapered appearance with the apex at the anterior.

Ornamentation consists of three strong, prominent longitudinal ridges which end against a strong anterior marginal rim. The dorsal ridge forms the dorsal margin between the two hinge ears; it curves sharply downwards just to the posterior of the anterior hinge ear. The median ridge is slightly sinuous, running roughly parallel to the dorsal ridge but with a less accentuated course; it is very faint in the extreme posterior, but can be traced right to the margin, just before which it bifurcates. The ventral ridge is almost straight, curving upwards just before reaching the anterior marginalrim. All the ridges are faint at the posterior and tend to disappear amongst the reticulation, but the dorsal and ventral ridges appear to join the median ridge. The anterior marginal rim is particularly strong in the right valve. Between the ridges is an uneven reticulation; there is a particularly prominent “ridge’’ running between the median and dorsal ridges just to the posterior of centre. The specimens from the Headon Beds lack this “‘ridge’’. The area between the dorsal ridge and the antero-dorsal angle is almost smooth.

The hinge of the left valve has a swollen antero-dorsal lobe, prominent antero- ventral lobe, large antero-median tooth and a small postero-median swelling. In the right valve the posterior and anterior teeth are about equal in size and rather small. The selvage is prominent, with a small anterior and posterior flange groove and wide ventral one; the flange is particularly prominent along the anterior margin. A list is strongly developed in the antero-ventral and postero-ventral regions. No other internal details could be clearly seen.

DIMENSIONS: Left valve Right valve ip H L/H Ww 1 H L/H Female 0:50 0:30 1:67 0:23 0°51 0:26 1:96 Male 0°52 0:28 1-86 0:24 0°51 0:25 2°04

Discussion: The specimens from the Headon Beds are slightly different from the Ludian ones, particularly with the reticulation between the ridges. The similarities are so strong, however, that it was thought unjustifiable to separate them. See also F. solentensis sp. nov.

Flexus lenijugum sp. nov. (Pl. 21, figs 7, 9; Pl. 22, figs 9, 10) DERIVATION OF NAME: Latin—lenis, smooth; jugum, ridge. Refers to the orna- mentation.

D1aenosis: A species of Cytheretta with an almost straight posterior margin bearing four spines ; apart from the longitudinal ridges the carapace is smooth.

Hototype: Io 4113, a female left valve. PARATYPE: Io 4114. MATERIAL: 8 carapaces.

344 MID-TERTIARY CYTHERETTINAE

TYPE LOCALITY AND HORIZON: Chateau Romefort, Blaignan; Argiles a algues.

STRATIGRAPHICAL RANGE AND DISTRIBUTION: So far only known from the type locality.

DESCRIPTION: Sexual dimorphism can be discerned, but it is not very prominent. There are anterior and posterior hinge ears 1n the left valve; the dorsal margin is evenly rounded; the ventral margin is slightly concave in the left valve and strongly so in the right; the posterior margin is almost straight and has four spines. It is tapered towards the anterior in dorsal view.

The ornamentation consists principally of three longitudinal ridges. The dorsal ridge forms the dorsal margin; in the right valve it is continuous with the anterior margin rim; in the left, it ends beneath the anterior hinge ear. The median ridge is short and has another weak ridge above it in the posterior. The ventral ridge is strong, joining the anterior marginal rim and, at the posterior, the weak ridge above the median ridge. The anterior marginal rim is strong, running from the anterior hinge ear to the ventral margin.

No internal features could be seen.

DIMENSIONS: Carapaces

18 H L/H Ww Female 0-80 0°43 1-86 0:38 Male 0-80 0-41 1:95 0:37

Discussion: F. lenijugum resembles F. plicatus (von Minster) with the lack of ornamentation between the longitudinal ridges. It differs in shape; FP. plicatus has a more tapered posterior margin in lateral view. It also differs in the configuration of the ridges; the dorsal ridge does not form the dorsal margin in F. plicatus, nor does it join the anterior marginal rim; the ventral ridge is continuous with the anterior marginal rim, not merely joining it. F. lenijugum differs from all other described species by the absence of ornamentation between the ridges.

Flexus schoelleri (Keij) (Pl. 22, figs 6-8)

1955 Pavracytheretta schoellert Keij, p. 119, pl. 16, fig. 4; pl. 19, figs 11-12. 1956 Cytheretta schoelleri (Keij) Oertli, p. 65, pl. 8, figs 196-197.

1965 Protocytheretta schoellert (Keij) Moyes, p. 56, pl. 6, fig. 13.

1969 Protocytheretta schoelleri (Keij) Carbonnel, p. 111, pl. 8, figs 1-3.

MATERIAL: Couches du Phare: RO 269, 10 valves and carapaces; RO 270, 5; RO 271, 3. Io 4115-7. St. Geours-de-Maremne: ASG I, 3 valves and carapaces; ASG 252 eNSGr3y 2:

TYPE LOCALITY AND HORIZON: Moulin de Gamachot, Upper Aquitanian (?).

STRATIGRAPHICAL RANGE AND DISTRIBUTION: Couches du Phare, Biarritz; Faluns

Bleues, St. Geours-de-Maremne; Aquitanian and Burdigalian of the Bordelais and Rhone.

OF NORTH-WEST EUROPE 345

Discussion: This was placed by Puri (1958) into his new genus Protocytheretta, defined as ‘Cytheretta’-shaped, but with three longitudinal ridges; Flexus was regarded as being ‘Cythereis’ shaped and with three longitudinal ridges. F. schoel- lert has a truncated posterior margin which is not at all ‘Cytheretta’-like and the rod- like ridges are completely different from those of P. daniana (Brady). (See Hulings and Puri, 1964, p. 327 for an illustration of P. daniana.) F. schoelleri is probably not related to any other Plexus species here described, but is included in the genus on the purely morphological grounds that it has three longitudinal ridges.

The specimens from the Couches du Phare are smaller than the typical F. schoelleri (length of female carapace = 0-60 compared with 0:78).

Flexus sp. A (PlF 22h fie: 5)

MATERIAL: I carapace. lo 4118.

LOCALITY AND HORIZON: Bambrugge; Sables de Lede.

Dimensions: L, 0-64; H, 0:36; W, 0:31; L/H, 1-78.

Discussion: This is almost certainly a new species, but lack of material prevents a description. The configuration of the ridges is similar to FP. concinnus (Triebel), but its shape is different from the latter both in dorsal and in lateral view, and it has a much stronger anterior marginal rim.

XV. CONCLUSIONS

The Cytherettinae have proven useful for helping to establish a correlation between the various localities in the Anglo-Paris-Belgian area in the Eocene and between this region and Germany in the Oligocene. In particular they support the idea of cor- relating the Sables de Lede with the Upper Lutetian of the Paris Basin, placing the Sables moyens in the Middle Eocene and correlating them with the Upper Brackle- sham Beds of Hampshire, and correlating the Barton Beds with the Marnes a P. ludensis. Unfortunately the Cytherettinae provide little information concerning the relationship of the type Lattorfian with other areas of western Europe. The four species recorded from the Headon Beds suggest a relationship with the Bartonian on the one hand (C. porosacosta, F. ludensis) and with the Oligocene on the other (C. headonensis, C. aff. stigmosa), although the latter are related to the Eocene C. carita and C. cellulosa. Other ostracods however support a Bartonian age for the Headon Beds (Keen, 1968). Detailed correlation between England, France, and Belgium is possible using the evolution of C. costellata and C. laticosta.

The Aquitaine Basin formed a very distinct province, the only group in common with the northern areas being the C. eocaenica group. This suggests that the English Channel as now known could hardly have existed during the Eocene and Oligocene.

The presence of the Cytherettinae in Tertiary sediments is a good indication of shallow marine conditions, close to shore. Of the main species groups present, only the superspecies C. laticosta seems to have preferred muddy waters. The C. haimeana group were most abundant in clear waters in which calcareous or sandy sediments

346 MID-TERTIARY CYTHERETTINAE

were accumulating, the C. eocaenica, C. tenuipunctata, C. rhenana, and C. sagri groups inhabited clear waters where sands were being deposited.

REFERENCES

AsissIn, M. 1955. See Girard d’Albissin, 1955.

ANDERSON, F. W. 1964. ‘The law of ostracod growth. Palaeontology, London 7 : 85-104.

APOSTOLESCU, V. 1955. Description de quelques ostracodes du Lutetien du Bassin de Paris.

Cah. géol. Thoiry, 28/29 : 241-279, pls. 1-8.

1956. Contribution a l'étude des ostracodes de |’Eocene infeérieur (s.l1.) du Bassin de Paris.

Rev. Inst. franc., Petvole, Paris 11 : 1327-1352, pls. 1-4.

1964. Repartition stratigraphique generale des Ostracodes du Paleogene des bassins de

Paris et de Bruxelles. In colloque sur le Paleogene, Bordeaux, 1962. Mém. Bur. Rech. géol. min., Paris, 28 : t.II, 1035-1040.

Ascout, P., 1965. Ecological study on Ostracoda from bottom cores of the Adriatic Sea. Pubbl. Staz. zool. Napoli, 33 (suppl) : 213-246, 4 pls.

Bates, D. A. J. 1958. Foraminifera of the Oligocene of Belgium. Imst. Roy. Sci. Nat. Belgique, Bruxelles 143 : 188p., 13 pls.

Benson, R. H. 10965. In discussion of Sandberg, 1965, p. 510.

Benson, R. H. & CoLEMAN, G. L. 1963. MJRecent marine ostracodes from the eastern Gulf of Mexico. Paleont. Contr. Univ. Kansas., Arthropoda, Art. 2 : 1-52, 8 pls.

Benson, R. H. & Mappocks, R. F. 1964. Recent ostracodes of Knysna Estuary, Cape Province, Union of South Africa. Paleont. Conty. Univ. Kansas., Anthropoda, Art. 5 : 1-39, 6 pls.

Bop, VAN Den, W. A. 1957. Ostracoda from the Paleocene of Trinidad. Muicropaleontology, New York 3 : 1-14, pl. 1-4.

BosguEt, J., 1952. Description des entomostraces fossiles des terrains Tertiaires de la France et de la Belgique. Mém. cour. Acad. vr. Sci. Belg., Bruxelles, 24 : 142 p., 6 pls.

Bowen, R.W.C. 1953. Ostracoda from the London Clay. Proc. Geol. Ass. Lond., 64 : 276— 292, 8 text figs.

Brapy, G.S. 1868. La cote de Syrie; In: De Folin, L., and Perier, L., Les fonds de la mer,

legbars) i 3 fore, 1, 12 jo),

Brapy, G. S., CrosskEy, H. W. & Ropertson, D. 1874. A monograph of the post-Tertiary entomostraca of Scotland, including species from England and Ireland. Palaeontogr. Soc. (Monogr.) London: 232 p., 16 pls.

CAPELLI, G. B. 1905. Contribuzione allo studio degli Ostracodi fossili dello strato a sabbie grigie della Farnesina presso Roma. Boll. Soc. geol. ital., Roma 24 : 303-343, pl. 9-10. CARBONNEL, G. 1969. Les ostracodes du Miocene Rhodanien. Docum. Lab. Géol. Fac. Sci.

Lyon, 32 : 469 p., 16 pls. CAVELIER, C. 1964. L’Oligocene inférieur du bassin de Paris. In Colloque sur le Paleogene, Bordeaux, 1962. Mém. Bur. Rech. géol. min., Paris 28 : t.I, 65-73. 1965. Le Sannoisien de Sannois (Seine-et-Oise) dans le cadre du bassin de Paris et sa

signification stratigraphique. Bull. Soc. géol. Fr., Paris (7) 7 : 228-238.

Curry, D. 1965. The Palaeogene Beds of South-east England. Proc. Geol. Ass. Lond. 76:

151-174.

1966. Problems of correlation in the Anglo-Paris-Belgian Basin. Proc. Geol. Ass. Lond.

77 : 437-467.

Curtis, D. M. 1960. Relation of environmental energy levels and ostracod biofacies in east Mississippi Delta area. Bull. Am. Ass. Petvol. Geol., Chicago 44 : 471-494.

CusuMan, J. A. 1906. Marine Ostracoda of Vineyard Sound and adjacent waters. Proc. Boston Soc. nat. Hist., Boston 32 : 359-385, pls. 27-38.

DELTEL, B. 1961. Les ostracodes du Paleogene moyen et supérieur d’Aquitaine meridionale. These 3em Cycle, Bordeaux., No. 95 : 215 p., 19 pls.

OF NORTH-WEST EUROPE 347

DELTEL, B. 1964a. Ostracodes du Paleogene d’Aquitaine meridionale. In Colloque sur le Paleogene, Bordeaux, 1962. Mém. Bur. Rech. géol. min., Paris, 28: t.II,

2041-2048.

1964b. Nouveaux ostracodes de l’Eocene et de l’Oligocene de l’Aquitaine meridionale. Act. Soc. Linn. Bordeaux, 100 (for 1962-63) 127—221,: 6pls.

DeErRoo, G. 1966. Cytheracea (Ostracodes) du Maastrichtien de Maastricht (Pays-Bas) et des regions voisines; resultats stratigraphique et paleontologiques. Meded. geol. Sticht., s’Gravenhage, Ser. C., 2, no. 2: 197 p., 27 pls.

DrooGER, C. W., KAAsscHIETER, J. P. H. & Kety, A. J. 1955. The microfauna of the Aqui- tainian-Burdigalian of south-western France. Ver. Ned. Ak. Wet. Nat., Deel, 21, no. 2: 135 p., 20 pls.

DucassE, O. 1964. Quelques espéces nouvelles d’ostracodes de l’Eocene terminal Girondin. Act. Soc. Linn. Bordeaux, 100 (for 1962-63) : 223-248, 3 pls.

EaGcar, S. H. 1965. Ostracoda of the London Clay (Ypresian) in the London Basin : 1. Reading district. Revue Micropaléont., Paris, 8 : 15-32, 2 pls.

Garcia, E. R., DE. 1969. Amendment a la sub-famille Cytherettinae Triebel 1952 (famille Cytherettidae (Triebel) Howe 1961. Proc. 3rd African Micvopal. Coll. (Cairo 1967), Cairo: 217-223, 2 pls.

GIRARD D’ALBISSIN, N. 1955. Etude du Sannoisien de Ile-de-France. Awnns. Cent. Etud. Docum. paléont., Paris, no. 11:

Haskins, C. W. 1968. Tertiary Ostracoda from the Isle of Wight and Barton, Hampshire, England. Revue Micropaléont., Paris, 11 : 161-175, 3 pls.

Hazet, J. E., 1967. Classification and distribution of Recent Hemicytheridae and Trachyle- beridae (Ostracoda) off northeastern North America. Geol. Surv. U.S. Prof. Paper, Washing- ton, 564 : 49 p., 2 figs., 11 pls.

Hutines, N.C. & Puri, H.S. 1965. The ecology of shallow water ostracods off the west coast of Florida. Pubbl. Staz. zool. Napoli, 33 (suppl.) : 308-344.

Jones, T. R. 1856a. Notes on the Entomostraca of the Headon and Osborn Series. Mem.

Geol. Surv. G.B., London, (Forbes 1856) : p. 157.

1856b. A monograph of the Tertiary Entomostraca. Palaeontogy. Soc. (Monogr.),

London : 68 p., 6 pls.

1870. Notesonthe Tertiary EntomostracaofEngland. Geol. Mag., London, 7 : 155-159.

Jones, T. R. & SHERBORN, C. D. 1889. A supplemental monograph of the Tertiary Ento- mostraca of England. Palaeontogr. Soc. (Monogr.), London : 55 p., 3 pls.

Keen, M.C. 1968. Ostracodes de l’Eocene supérieur et l’Oligocene inferieur dans les bassins de Paris, du Hampshire et de la Belgique et leur contribution a |’échelle stratigraphique. In Colloque sur l’Eocene, Paris, 1968. Mém. Bur. Rech. géol. min., Paris, 58 : t.1, 137-145.

Kez, A. J. 1955. See Drooger ef al., 1955.

1957. Eocene and Oligocene Ostracoda of Belgium. Mem. Inst. Roy. Sci. Nat. Belgique,

Bruxelles, 136 : 210 p., 23 pls.

Kornicker, L.S. 1965. Ecology of Ostracoda in the north-western part of the Great Bahama Bank. Pubbl. Staz. zool. Napoli, 33 (suppl) : 345-360.

Kruit, C. 1955. Sediments of the Rhone Delta. I. Grain size and microfauna. Ver. Ned. Geol. Gen., Deel, 15 : 359-499, 6 pls.

Kuiper, W. N. 1918. Oligocine und Miocaine Ostrakoden aus den Niederlanden. Thesis, Groningen : 90 p., 3 pls.

LatHam, M. H. 1938. Some Eocene Ostracoda from North-West India. Proc. R. Soc. Edinb. 59 : 38-48.

LIENENKLAUS E. 1894. Monographie der Ostrakoden des nordwestdeutschen Tertiars.

Z. dt. geol. Ges., Berlin, 46 : 158-268, pl. 13-18.

1895. Die Ostrakoden des Mitteloligociins von Jeurre bei Estampes. Verh. naturh. Ver. preuss. Rheintl., Bonn, 10 : 125-156, 3 pls. 1896. Ostracoda; in Kisstinc, E. : Die fauna des Mitteloligocans im Berner Jura.

Abh. schwz. pal, Ges. Basel 22 : 22-33, pl. 2.

348 MID-TERTIARY CYTHERETTINAE

LIENENKLAUS, E. 1905. Die Ostrakoden des Mainzer Tertiar-Beckens. Ber. senck. natf. Ges., Berlin, 1904/5 : 3-74, 4 pls. MARLIERE, R. 1958. Ostracodes du Montien de Mons et resultats de leur étude. Mém. Soc. Belg. géol. Bruxelles 5 : 1-53, pl. 1-6. MeuEs, G. 1941. Die Ostracoden der Oberoligozans der Umgebung von Budapest. Geol. hung. pal., Budapest 16 : 1-96, pl. 1-5. MorkKHOoVEN, F. P. C. M. van. 1962. Post-Palaeozoic Ostracoda. Their morphology, taxonomy, and economic use. Elsevier, Amsterdam 1 : 204 p. 1963. Post-Palaeozoic Ostracoda. Elsevier, Amsterdam 2 : 478 p. Moyes, J, 1965. Les ostracodes du Miocene Aquitain. Thesis, Bordeaux: 338 p., 13 pls. Mutier, G. W. 1894. Ostracoden des Golfes von Neapel und der angrensenden Meere- sabschnitte. Fauna u Flova Neapel, Berlin : 21 : 403 p., 40 pls. 1912. ‘Ostracoda’, Das Tierreich, Berlin, 31 : 1-434, 92 pls. Munster, G. von. 1830. Uber einige fossile Arten Cypris und Cythere. Neues Jb. Miner. Geol. Palaont., Stuttgart,1 : 60-67. Namias, I. 1900. Ostracodi fossili della Farnesina e Monte Mario presso Roma. Palaeontogr. italica, Pisa 6 : 79-114, pl. 14-15. Neviani, A. 1928a. Ostracodi fossili d’Italia. I” Vallebiaja Calabriano. Memorie Accad. pont. Nuouri Lincei, Roma, 11 : 1-120, 2 pls. 1928b. Oialcuni Ostracodi raccolti sulle spiaggie di Barie di Rimini. <Adti. Accad. pont. Nuoumi Lincet, Roma 81 : 123-132, 3 figs. OeErRTLI,H. 1956. Ostrakoden aus der Oligozénen und Miozanen Molass eder Schweiz. Schweiz Pal. Abh., Basel, 74 : 119 p., 16 pls. 1965. In discussion of General Problems of ecology and Palaeoecology. Pubbl. Staz. zool. Napoli, 33 (suppl) : 538-9. PoMEROL,C. 1964. Le Bartonien du bassin de Paris. In Colloque sur le Paleogene, Bordeaux, 1962. Mém. Bur. Rech. géol. min., Paris 28 : t.I, 153-168. Puri, H. S. 1952. Ostracode genera Cytheretta and Paracytheretta in America. J. Paleont., Tulsa, Okla. 26 : 199-212, pl. 39, 40. 1958. Ostracode subfamily Cytherettinae. Florida Geol. Surv. Geol. Buil., Tallahasse 40 : 183-194, pls. I-3. Puri, H. S. & Hutines, N. C. 1957. Recent ostracod facies from Panama City to Florida Bay. Tvans. Gulf-Cst. Ass. geol. Soc., New Orleans, 7 : 167-190. Puri, H.S., Bonapuce, G. & MaLLoy, J. 1965. Ecology of the Gulf of Naples. Pubbl. Staz. zool. Napoli. 33 (suppl) 33 : 87-199, 67 figs. Reuss, A. E. 1850. Due fossilen Entomostracaen des osterreichischen Tertiarbeckens. Naturw. Abh, Berlin, 1 : 41-92, pl. 8-11. 1851. Uber die fossilen Foraminiferen und Entomostraceen der Septarienthone der Umgebung von Berlin. Z. dt. geol. Ges., Berlin, 3 : 49-92, pls. 3-7. 1853. Einige Foraminiferen Bryozoen und Entomostraceen des Mainzer Beckens. Neues Jb. Miner. Abh. geol. u. Pet., Stuttgart, 21 : 670-679, pl. 9. 1855. Ein Beitrag zur genaueren Kenntniss der Kreidegebilde Mecklenburgs. Z. dt. geol. Ges., Berlin 7 : 261-292, pls. 8-11. 1869. Zur fossilen kauna der Oligozénschichten von Gaas. Sitz.-Ber. k. Akad. Wiss. math.-naturw. Cl., Wien, 57; 480-486, pls. 5-6. RoEMER, F. 1838. Die Cytherinen des Molasse Gebirges. Neues Jb. Miner. geol. u. Pall., Stuttgart 6 : 514-510, pls. 5-6. Rome, Dom. R. 1964. Ostracodes des environs de Monaco, leur distribution en profondeur, nature des fonds marins explores. Pubbl. Staz. zool. Napoli 33 (suppl) : 200-212. Ruaaier1, G. 1950. Gli Ostracodi della Sabbie grige quaternarie (Milazziano) di Imola. G. Geol. Bologna, ser. 2, 21 : 1-57, pl. 1, 34 figs. 1953. Eta e faune di un terrazzo marino sulla costa ionica della Calabria. G. Geol. Bologna, ser. 2, 23 : 20-144, 6 pls., 3 figs.

OF NORTH-WEST EUROPE 349

RuGGIERI, G. 1959. Enumerazione degli Ostracodi marini del Neogene, Quaternario e Recente italianid escritti o elencati nell’ultimo decennio. Atti Soc. ital. Sci. nat., Milan 98 :183-208. 1962. Gli Ostracodi marini del Tortoniano (Miocene modio-superiore) di Enna, nella

Sicilia Centrale. Palaeontogr. Italica, Roma 56 : 1-68, pls. 11-17, 15 figs.

SANDBERG, P. A. 1965. Notes on some Tertiary and Recent brackish-water Ostracoda. Pubbl. Staz. zool. Napoli 33 (suppl) : 498-514.

SEQUENZA, G. 1880. Le formazioni tertiarie nella provincia di Reggio (Calabria) Atti Accad. naz. Lincet Memorie, Roma 6 : 1-416.

SPEYER, O. 1863. Die Ostracoden der Casseler Tertiarbildungen. Ber. Ver. Naturk., Cassel 24 : 1-62, pls. 1-4.

SKOGSBERG, T. 1928. Studies on marine ostracodes. Part II. External Morphology of the genus Cythereis with description of twenty-one new species. Occ. Pap. Calif. Acad. Sci., San Francisco, 15 : 1—155, 6 pls.

SonmMEz, N. 1963. Deux nouveau genres d’Ostracodes du Paleogene de Thrace (Turquie):. Revue Micropaléont., Paris,6 : 76-84, 1 pl.

SONMEZ-GoKCEN, N. 1964. Notice sur le nouvel age determine par les Ostracodes de la serie a Congeria du Neogene des environs de Catalca (Thrace). Bull. Miner. Res. Explor. Inst Ankara 63 : 47-59.

STCHEPINSKY, A. 1963. Etude des Ostracodes du Stampien d’Alsace et complement a l’étude des ostracodes du Sannoisien d’Alsace. Bull. Serv. Carte géol. Als. Lovr., Strasbourg, 16 : 151-173.

STRAUB, E. W. 1952. Mikropalaontologische Untersuchungen Tertiar zwischen Ehingen und Ulma.d.donau. Geol. Jb., Stuttgart 66 : 433-524, 3 pls.

Swain, F. M. 1955. Ostracoda of San Antonio Bay, Texas. J. Palaeontol., Tulsa, Okla. 29 : 561-646, pl. 59-64.

1969. Taxonomy and ecology of near-shore Ostracoda from the Pacific coast of North and Central America. In The Taxononomy, Morphology, and Ecology of Recent Ostracoda, ed. Neale, J. W., Oliver and Boyd, Edinburgh, 423-474, 11 pls.

SYLVESTER-BRADLEY, P.C. 1951. The subspecies in palaeontology. Geol. Mag., London 88 :

88-102.

1962. The taxonomic treatment of phylogenetic patterns in time and space, with examples from the Ostracoda. Syst. Assoc. Publ., London 4 : 119-133.

SZCEZECHURA, J. 1971. Paleocene Ostracoda from Nugssuaq, West Greenland. Meddr. Gronland, Kobenhavn, 193 : 42 p., 8 pls.

TERQquEM, M. O. 1878. Les Foraminiferes et les Entomostraces-Ostracodes du Pliocene de V’Ile de Rhodes. Mém. Soc. géol. France, Paris, 3e ser., I : 1-135, 14 pls.

TRIEBEL, E. 1952. Ostracodem der gattung Cytheretta aus dem Tertiar des Mainzer Beckens. Notissbl. Hess. L. Amt. Bodenforsch., Wiesbaden 6 : 15-30, pl. 2-5.

ViIGNEAUX, M. 1964. Le bassin d’Aquitaine. In Colloque sur le Paleogene, Bordeaux, 1962. Mém. Bur. Rech. géol. min., Paris 28 : t.2, 177-220.

WRIGLEY, A. & Davis, A.G. 1937. The occurrence of Nummulites planulatus in England, with

a revised correlation of the strata containing it. Proc. Geol. Ass. Lond. 48 : 203-227.

M. G. KEEN, B.Sc., Ph.D., Department of Geology, THE UNIVERSITY, Griascow, G12 800

PIC IN ICIS, Cytheretta judaea (Brady)

FIGs I, 4-7, 9. Specimens from Recent beach sand, Rimini.

Fic. Fic. Fic. Fic. Fic. Fic.

Fic.

a,

OW DAB

Left valve, female, lo 3807 x 70, L = 0-75 mm. Right valve, male, Io 3792, « 100, L = 0-79 mm. Right valve, male, lo 3810, x 70, L = 0-79 mm. Left valve, female, lo 3793, x 70, L = 0:75 mm. Enlargement of Io 3810 x 140.

Posterior radial pore canals of Io 3793, x Ioo.

Cytheretta subradiosa (Roemer)

Right valve, male. Io 3795 < 100, L = 0-81 mm, Lower Pliocene, Rimini.

BOE MIN 1 THE BRITISH MUSEUM (NATURAL HISTORY) GEOLOGY Vol 2n Ne: 6

CORRIGENDA

Plate 2, caption to Fig. 8 For “‘punctuation” read “‘punctation’’.

Plate 15, caption to Fig. 10 For “Io 4031’ read “Io 4030”.

Plate 19, caption to Fig. 5-7, 9 For “‘Calcaire a Algues” read “‘Calcaire 4 Algues’’.

Bull. By. Mus. nat. Hist. (Geol.) 21, 6 PLATE 1

FIGs.

Fic. Fic. Fic. Fic. Fic. Fic. Fic. Fic. Fic. EKG, 1

IPIL JAIN, 2

All, except figs. 7, 8, x 70

I-10 Cytheretta costellata costellata (Roemer)

OO ON AUNHW DN H

Fig. 5 from Sables de Lede, Bambrugge; Fig. 6 from Upper Bracklesham Beds, Selsey; remainder from Lutetian IV, Damery. All except Fig. 5 are MORPHOTYPE A.

Left valve, female, lo 3796, L = 0-69 mm.

Left valve, male, lo 3798, L = 0-72 mm.

Right valve, female, lo 3797, L = 0-69 mm.

Right valve, male, lo 3799, L = 0:72 mm.

Left valve, female, Io 3802, L=o-72mm. MORPHOTYPE B. Left valve, female, lo 3803, L = 0:64 mm.

Detail of lo 3797, x 140

Detail of Io 3797 showing “ punctuation ’”’ between the ridges, x 300 Female carapace, dorsal view, lo 3800, L = 0-70 mm.

Male carapace, ventral view, Io 3801, L = 0:75 mm.

Bull. By. Mus. nat. Hist. (Geol.) 21, 6 PLATE 2

PLATE 3

All, except Fig. 8, x 75

Fic. 1-8 Cytheretta costellata grandipora subsp. nov.

IE, i Ine, 2 Fig. 3 Fic. 4 Fig. 5 Fic. 6 ING, 97 EiGe 1S) lane, ©) aK, WO)

MORPHOTY PE C, Left valve male, Io 3812, L = 0-83 Auvers-en-Oise. MORPHOTYPE, E. Right valve, female, lo 3811, L = 0-74 Auvers-en-Oise. MORPHOTYPE, D, Left valve, male, Io 3809, L = 0-83 Auvers-en-Oise. MORPHOTYPE, E, Right valve, female, lo 3805, L = 0-77 Moiselles. MORPHOTYPE, D, Left valve, female, lo 3808, L = 0-74 Auvers-en-Oise. MORPHOTYPE, E, Right valve, male, Io 3807, L = 0-85 Moiselles. MORPHOTYPE E, Left valve, female, lo 3804, L = 0-76 Moiselles. HOLOTYPE

Enlargement of Io 3804 showing “ pores ’”’. x 150.

Cytheretta bambruggensis Keij. Right valve, female, lo 3827, L = 0-74. Sables de Lede, Bambrugge.

Cytheretta sp. C. Right valve, female, L=o-70 Auvers-en-Oise. Specimen destroyed.

Bull. By. Mus. nat. Hist. (Geol.) 21, 6 PLATE 3

PLATE 4

2S 7/8)

Fic. 1-8 Cytheretta costellata cratis subsp. noy. All specimens are MORPHO- TYPE F from the Marnes a P. ludensis, Verzy.

Fig. 1 Left valve, female, Io 3814, L = 0:73 HOLOTYPE

Fig. 2 Left valve, male, Io 3817, L = 0-76

Fig. 3 Right valve, female, lo 3815, L = 0-70

Fig. 4 Right valve, male, Io 3818, L = 0-83

Fie. 5 Left valve, larval no. 8, Io 3820, L = 0-60

IMG, © Right valve, larval no. 8, Io 3819, L = 0-60

Fig. 7 Female carapace, ventral view, Io 3816, L = 0-73 8

Male carapace, dorsal view, Io 3821, L = 0-76

Fic. 9-11 Cytheretta costellata antecalva subsp. nov. All specimens are MORPHOTYPE G from the Middle Barton Beds, Barton.

Fic. 9 Left valve, male, Io 3825, L = 0:84 Fic. 10 Left valve, female, Io 3823, L = 0-77. HOLOTYPE. Fic. 11 Right valve, female, lo 3824, L = 0-77

Bull. Br. Mus. nat. Hist. (Geol.) 21, 6 PLATE 4

PILI Tic. 1-3 Cytheretta costellata cratis subsp. nov.

ING, it Right valve, female, Ilo 3822, L = 0-67, x 80, showing the inner margin. From the Marnes a P. ludensis, Chavencon.

Rie 2 Hinge of lo 3822. x 400

Bie. 3 Central muscle scars and fulcral point of Io 3822. x 800.

Tlic. 4-7 Cytheretta ruelensis subsp. nov. Io 3837, Sables de Cresnes, Le Ruel. x 80. Female carapace. HOLOTYPE.

ITE, AL Posterior view Fig. 5 Dorsal view Fic. 6 Left valve ErGan7, Right valve

PLATE 5

Bull. By. Mus. nat. Hist. (Geol.) 21, 6

IPILNIDID © All x 70 Fic. 1, 6 Cytheretta haimeana (Bosquet) le. it Left valve, female, lo 3834, L = 0-60. Lutetian IV, Damery. Fic. 6 Left valve, female, lo 3835, L = 0-70. Sables de Beauchamp, Moiselles. Fic. 2,5 Cytheretta crassivenia Apostolescu. lie, 2 Left valve, female, Io 3828, L = 0-66. Lutetian IV, Damery. Fic. 5 Left valve, female, lo 3829, L = 0-70. Sables de Beauchamp, Moiselles.

Big. 3 Cytheretta aff. decipiens Keij Left valve, female, of a carapace Io 3833, L = 0-66. Marnes a P. ludensis, Chavencon.

Fic. 4,7 Cytheretta ruelensis sp. nov.

Fig. 4 Ventral view of male carapace, Io 3838, L = 0-85. Sables de Cresnes, eskuell Fig. 7 Left valve of Io 3838.

Fic. 8-10 Cytheretta decipiens Keij Specimens from Sables de Beauchamp. Moiselles.

Fie. 8 Left valve, female, Io 3830, L = 0:69 FIG. 9 Left valve, male, Io 3832, L = 0-81 FIG. 10 Right valve, female, lo 3831, L = 0-73

Fic. 11 Cytheretta sp. A Left valve, Io 4092, L = 0-74. Sables de Beauchamp, Moiselles.

PLATE 6

Bull. By. Mus. nat. Hist. (Geol.) 21. 6

ee

oe

beg A PEV Ps Mase, cpa eet er

PLATE 7

Cytheretta eocaenica Keij EE, © from Lutetian IV, Damery; remainder from Sables de Lede, Bambrugge.

IEW, Left valve, female, Io 3841, x 65, L = 0-79

Jue, 2 Left valve, male, lo 3844, x 65, L = 0-88

Fic. 3 Left valve, female, lo 3840, x 65, L = 0:88

Fie. 4 Hinge of lo 3844, x 125

Fie. 5 Central muscle scars of Io 3844, x 250

Fic. 6 Left valve, female, lo 3839, x 65, L = 0-80

LEE 7 Right valve, male, Ilo 3842, x 65, L = 0°83

Fic. 3 Anterior tooth and hinge bar of Io 3844, x 650

BIG. 9 Left valve, female, lo 3843, L = 0-93. Post-maturation moult stage. o)

Fic. 1 Detail of pitting of Io 3839, x 750.

Bull. Br. Mus. nat. Hist. (Geol.) 21, 6 PLATE 7

Specimens of Fic. 1, 2, 4 from the Couches du Phare, Biarritz; specimen of remaining

Pi AEs

Cytheretta oligocaenica sp. nov.

Fic. from Faluns Bleues. St. Geours-de-Maremne.

Fic. Fic. Fic. Fic. Fic. Fic. Fic. Fic. Fic. Fic. Fic. Fie. Fic. Jae,

MONI ANN WN H

No}

Left valve, female, Io 3845, x 70, L = 0-85, HOLOTYPE

Carapace, dorsal view, male, Io 3847, x 70, L = 0-86 Central muscle scars of lo 3849

Right valve, female, lo 3846, x 70, L = 0-84

Right valve, female, Io 3849, x 100, L = 0:84 Right valve, female, Ilo 3849, x 100, L = 0-84 Hinge of Io 3849, x 125

Io 3849, x 70

Anterior tooth of Io 3849, from dorsal, x 350 Anterior tooth of Io 3849, from anterior, x 350 Anterior tooth of Io 3849, from lateral view, x 350 Posterior tooth of Ilo 3849, from ventral, x 350 Posterior tooth of Io 3849, from posterior, x 350 Posterior tooth of Io 3849, from lateral view, x 350.

PLATE 8

Bull. By. Mus. nat. Hist. (Geol.) 21, 6

PLATE 9 Alyx 70

Fic. 1-4,6,7 Cytheretta cellulosa sp. nov.

Fic. Fic. Fic. Fic. Fic. Jae.

Fic

Fic. Fig. Fic. Fic. Fic.

18ne,

Fic. 2 from Sables de Beauchamp, Moiselles; remainder from Sables d’Auvers- en-Oise.

N OF WN H

Left valve, female, lo 3859, L = 0:78; HOLOTYPE Right valve, female, Io 3860, L = 0-74

Left valve, male, Io 3861, L = 0:96

Male carapace, dorsal view, Io 3863, L = 0:93 Right valve, male, lo 3862, L = 0:93

Posterior view of Io 3863.

. 5, 8, 9-11 Cytheretta carita sp. nov.

Specimens from the Sables de Beauchamp, Moiselles.

H HOW) CON

lal

12

Female carapace, dorsal view, L = 0-87; specimen destroyed Female carapace, anterior view; specimen destroyed

Left valve, female, Io 3853, L = 0-89; HOLOTYPE

Right valve, female, lo 3854, L = 0-85

Left valve, male, Io 3855, L = 0:94

Cytheretta sp. B Right valve, Io 4093; Marnes a Pentacrinus, Biarritz.

Bull. By. Mus. nat. Hist. (Geol.) 21, 6 PLATE 9

Fic. Fie. Fic.

Fic

Fic. Fic. Fia. Fic.

Fic.

Fic.

I

3 5

PEATE ro * 7) Cytheretta geoursenis sp. nov. Specimens from the Faluns bleues, St. Geours-de-Maremne.

Left valve, male, Io 3852, x 70, L = 1:07 Left valve, female, Ilo 3850, x 60, L = 0:98; HOLOTYPE Right valve, female, lo 3851, x 70, L = 0-93

. 2, 4, 6, 8, 9, Cytheretta laticosta (Reuss)

aoanr N

Specimens from the Middle Barton Beds, Barton.

Left valve, female, lo 3865, x 70, L = 0-77

Left valve, male, Io 3864, x 70, L = 0:90

Right valve, male, Io 3866, x 70, L = 0-go

Enlargement of Io 3864, central area between ventral and medium ridges, showing punctation. x 250.

Further enlargement of Io 3864, showing a normal pore canal and sur- rounding puncta. x 850.

Cytheretta carita sp. nov. Male carapace, ventral view, Io 3858, x 70, L = 0-93; Sables de Beau- champ, Moiselles.

Bull. By. Mus. nat. Hist. (Geol.) 21, 6 PLATE to

PATE ime

x 70, except Fic. 7

Fic. 1-4, 8,9 Cytheretta forticosta sp. nov.

Fic. 1-4 from Upper Bracklesham Beds, Whitecliff Bay

Fic. Fig. Fic. Fic. Fic.

Fic.

Fic.

Fic. Fic.

Fic.

Con W NH

Left valve, female, Io 3871, L = 0-79; HOLOTYPE

Left valve, male, Ilo 3872, L = 0:87

Right valve, female, lo 3874, L = 0:84

Right valve, male, Io 3873, L = o-g1

Female carapace, ventral view, Io 3875, L = 0.80; Upper Bracklesham Beds,

Selsey. Male carapace, dorsal view, Io 3876, L = 0-92; Upper Bracklesham Beds,

Selsey.

Cytheretta porosacosta sp. nov.

Left valve, male, Io 3880, L = 0-79 Middle Headon Beds, Colwell Bay. e Left valve, female, Ilo 3879, L = 0-75. HOLOTYPE; Middle Headon Beds,

Colwell Bay. Enlargement of Io 3880 showing punctation x 140.

Bull. Br. Mus. nat. Hist. (Geol.) 21, 6 PLATE 11

PAD E, 12

Fic. 1-2, 5 Cytheretta laticosta (Reuss)

Fic. Fia. Fic.

Fic.

Specimens from Middle Barton Beds, Barton; x 7o.

Male carapace, ventral view; Io 3869, L = 0-88 Female carapace, dorsal view, Io 3868, L = 0:77 Left valve, female, Io 3869, L = 0-81

Cytheretta porosacosta sp. nov. x 70

Right valve, male, Io 3882, L = 0-81; Middle Headon Beds, Milford. Right valve, female, Io 3881, L = 0-76; Middle Headon Beds, Milford.

Cytheretta forticosta sp. nov.

Right valve, female, lo 3878, x 70, L = 0:86; Upper Bracklesham Beds, Selsey.

Dorsal muscle scars, x 350

Posterior duplicature, x 350

Anterior tooth, x 350

Posterior tooth, dorsal view, x 350

Anterior tooth, dorsal view, x 350

Central muscle scars, x 350.

Bull. By. Mus. nat. Hist. (Geol.) 21, 6 PLATE 12

IPI INADIS, 03}

Cytheretta tenuistriata ornata subsp. nov.

Fic. Fic. Fic. Fie. Fia. Fic. Eee Fic. Fia.

I 2 3 4 5 6 i 8

Ne}

Fic. 10 awe, iit 1G, U2

Left valve, female, Io 3898, L = 1:03, x 50; HOLOTYPE Right valve, male, lo 4021, L = 1:14, x 50

Right valve, female, Io 3899, L = 1-02

Left valve, female; specimen destroyed

Left valve, male, Io 4020, L = 1°15

Right valve, male; specimen destroyed

Left valve, 8th moult stage, L = 0-87; specimen destroyed Right valve, 8th moult stage, Io 4025, L = 0-86

Left valve, 7th moult stage, L = 75:; specimen destroyed Right valve, 7th moult stage, Io 4026, L = 0-80

Left valve, 6th moult stage, Io 4024, L = 0°58

Right valve, 6th moult stage, lo 4027, L = 0°59

Specimens from the Falun d’Auvers-St.-Georges.

Bull. By. Mus. nat. Hist. (Geol.) 21, 6 PLATE 13

~ SENS ay

OR ee

PL AME: oa Cytheretta tenuipunctata absoluta subsp. nov.

Fic. 1-4, 6, 7; x 60; specimens from the Marnes a Huitres, Cormeilles. KE, Left valve, female, Ilo 3884, L = 0:84; HOLOTYPE

IRC, 2 Left valve, male, Io 3886, L = 0:98

FIG. 3 Right valve, female, lo 3885, L = 0-87

Fic. 4 Right valve, male, lo 3887, L = 0-98

BiG, © Female carapace, dorsal view, Ilo 3888, L = 0-90 Fic. 7 Male carapace, ventral view, Io 3889, L = 0-98.

Cytheretta tenuipunctata lirata subsp. nov.

Fic. 5, 8-10, x 50; specimens from the Falun d’Auvers-St.-Georges. Fig. 5 Left valve, female, lo 3890, L = 0-90; HOLOTYPE ETGaaO Right valve, female, L = 0-90; specimen destroyed

Fic. 9 Right valve, male, lo 3893, L = 1-o1

Fic. I0 Left valve, male, Io 3892, L = 1:04.

Bull. Br. Mus. nat. Hist. (Geol.) 21, 6 PLATE 14

dedomtn oe = MRELOTE, We bo

Te,

PLATE 15 x 60, except Fic. 9, x 50 Cytheretta buttensis reticulata sp. nov., subsp. nov. Fic. 1-8, Specimens from Cormeilles.

MORPHOTYPE A.

Ie, 1 Left valve, female, Io 4032, L = 0-85, Marnes a Huitres; HOLOTYPE

FIG. 2 Left valve, male, Io 4033, L = 0.97, Marnes a Huitres.

Fic. 3 Right valve, female, Io 4038, L = 0-82, Couches de Sannois

FIG. 4 Female carapace, dorsal view, Io 4034, L = 0-91, Marnes a Huitres

Fig. 5 Male carapace, dorsal view, L = 0-95, Marnes a Huitres. Specimen destroyed.

MORPHOTYPE B

Fic. 6 Right valve, female, L = 0-83, Couches de Sannois, Io 4036

FIG. 7 Left valve, female, Io 4037, L = 0-88, Couches de Sannois.

MORPHOTYPE C lee, Right valve, male, lo 4035, L = 0-96, Couches de Sannois

Cytheretta buttensis buttensis sp. nov. 18K, 1© Left valve, male, Io 4031, L = 0-91, MORPHOTYPE C, Couches de Sannois; HOLOTYPE

Cytheretta tenuipunctata lirata subsp. nov. Fig. 9 Right valve, female, L = 0-90; specimen destroyed.

Bull. Br. Mus. nat. Hist. (Geol.) 21, 6 PLATE 15

PAPAS 16

Cytheretta minipunctata sp. nov.

FIG. I-3, x 50. Specimens from the Couches de Sannois, Cormeilles Fic. I Left valve, male, Io 4028, L = 1-10. HOLOTYPE rem Male carapace, ventral view, Io 4028

Fic. 3 Female carapace, ventral view, lo 4029, L = 0:08.

Cytheretta tenuipunctata lirata subsp. nov. Fic. 4 Left valve, male, lo 3894, L = 1:04, x 55; Falun d’Auvers-St.-Georges.

Cytheretta tenuistriata tenuistriata (Reuss).

Fic. 5,7 Specimens from the Unterer Meeresand, Alzey Trift FIG. 5 Left valve, male, lo 3896, L = 1:25, x 40

FIG. 7 Right valve, female, Io 3897, L = I-10, x 50.

Cytheretta minor (Lienenklaus). Fic. 6 Left valve, female, Io 3704, L = 0-88; Unt. Meeresand, Alzey Trift, x 60.

Cytheretta aff. stigmosa Triebel. Fic. 8 Left valve, female, Io 4052, L = 0:76, x 65; Mid. Headon Beds, White- cliff Bay.

Cytheretta gibberis sp. nov.

Fic. 9, 10, x 70 Specimens from the Couches du Phare, Biarritz Fig, “9 Right valve, male, Io 4086, L = 0-92

Fic. 10 Right valve, female, Io 4084, L = 0-88; HOLOTYPE.

Bull. By. Mus. nat. Hist. (Geol.) 21, 6 PLATE 16

’ ae ae = z. yee tome ‘ ‘x

Sh A

ae ee We face Ome ek

PLATE 17

Cytheretta stigmosa gallica subsp. nov.

Fic. 1, 2, 10 Specimens from the Falun d’Auvers-St.-Georges, x 60 Fic. I Left valve, female, lo 4053, L = 0-75; HOLOTYPE

Fie. 2 Left valve, male, Io 4055, L = 0°73

Bie. 5 Female carapace, dorsal view, L = 0-73; specimen destroyed EIG. 10 Male carapace, dorsal view, Io 4056, L = 0-74.

Cytheretta regularis sp. nov.

Fic. 3, 4, 5, 6, 7 Specimens from the Stampian of Gaas (Lesbarritz), x 60 BiG Left valve, female, lo 4057, L = 0-70; HOLOTYPE

Big. 4 Right valve, female, Io 4058, L = 0:68

Ime, © Left valve, male, Io 4059, L = 0-70

Fig. 7 Right valve, male, lo 4060, L = 0-68.

Cytheretta vesca sp. nov.

Fic. 8, 9, 12 Specimens from the Falun d’Auvers-St.-Georges, x 60 Fic. 8 Left valve, female, lo 4048, L = 0:73; HOLOTYPE Fic. 9 Right valve, male, Io 4051, L = 0-71

Fic. 12 Right valve, female, lo 4049, L = 0-72.

Cytheretta headonensis Haskins.

Fic. 11, 13, 14 Specimens from the Middle Headon Beds, x 50 Fig. 11 Right valve, male, Ilo 4044, L = 0-80; Milford

Fic. 13 Left valve, male, Io 4043, L = 0-78; Headon Hill Fig. 14 Right valve, female, lo 4046, L = 0:80; Headon Hill.

Bull. By. Mus. nat. Hist. (Geol.) 21, 6 PLATE 17

me 3 Dd I, oo gOee TF

PREAH 2s

Cytheretta posticalis parisiensis subsp. nov. Fic. 1-4, 6 Specimens from the Falun d’Auvers-St.-Georges Fie. I Left valve, female, Io 4039, L = 0:95, x 50; HOLOTYPE

BiG. 2 Left valve, male, Io 4041, L = 1-05, x 50

FiG. 3 Right valve, female, lo 4040, L = 0-91, x 50

Fic. 4 Right valve, male, lo 4042, L = 1-03, x 50

Fia. 6 Anterior radial pore canals of left valve, lo 3702, x IIo.

Cytheretta bullans sp. nov.

Fic. 5, 7, x 70 Specimens from the Calcaire a Algues, Blaignan Fic. 5 Left valve, female, Ilo 4061, L = 0:85; HOLOTYPE FIG. 7 Right valve, male, Io 4062, L = o-go.

Cytheretta rhenana Triebel. Fic. 8 Left valve, female, lo 4047, L = 0°87, x 60; Unt. Meeresand, Alzey Trift.

Cytheretta headonensis Haskins. Fic. 9 Left valve, female, lo 4045, L = 0:83, x 70; Mid. Headon Beds, Headon Hill.

Bull. Br. Mus. nat. Hist. (Geol.) 21, 6 PLATE 18

PLATE 19

x 715

Cytheretta sagri sagri Deltel. Fic. 1-4 Specimens from the Stampian of Gaas

MORPHOTYPE A, Gaas (Espibos)

le, 1 Left valve, female, Io 4063, L = 0:84 Gee Right valve, male, Io 4066, L = 0-90 Fig. 3 Right valve, female, Io 4064, L = 0°84.

MORPHOTYPE D, Gaas (Lesbarritz) Fic. 4 Left valve, female, Io 4076, L = 0-92.

Cytheretta sagri inconstans subsp. nov. Fic. 5-7, 9 Specimens from the Calcaire a Algues, Blaignan.

MORPHOTYPE C Fic. 5 Left valve, male, lo 4069, L = 0-95 le, © Left valve, female, lo 4070, L = 0:95.

MORPHOTYPE B BG 7 Left valve, female, Io 4067, L = 0-78; HOLOTYPE Fie. 9 Female carapace, dorsal view, Io 4067.

Cytheretta perita Deltel. Fic. 8 Left valve, female, lo 4089, L = 0-81; Bartonian of Le Vigneau.

Cytheretta gibberis sp. nov. Fic. 10 Female carapace, dorsal view, Io 4085, L = 0:86; Couches du Phare, Biarritz.

Bull. Br. Mus. nat. Hist. (Geol.) 21, 6 PLATE 19

PLATE 20

x 75

Cytheretta sagri martini subsp. nov. Fic. 1-4 Specimens from the Couches du Phare, Biarritz

MORPHOTYPE E

Iie, it Left valve, female, Io 4071, L = 0-83; HOLOTYPE Fic. 2 Right valve, female, lo 4072, L = 0-81

Fie. 3 Left valve, male, Io 4073, L = 0°86.

MORPHOTYPE F Fic. 4 Left valve, male, lo 4074, L = 0-87.

Cytheretta postornata sp. nov.

Fic. 5-8 Specimens from the Couches de l’Atalaye, Biarritz Fie. 5 Left valve, male, Io 4088, L = 0-84

IG, © Right valve, female, lo 4087, L = 0:83; HOLOTYPE Fic. 7 Right valve, male, lo 4088

le, Left valve, female, lo 4087.

Cytheretta sculpta Ducasse.

Fic. 9, 10 Specimens from the Argiles a Algues, Blaignan Fic. 9 Right valve, male, Io 4o91, L = 0-73.

Fic. 10 Left valve, female, lo 4090, L = 0-70.

Bull. Br. Mus. nat. Hist. (Geol.) 21, 6 PLATE 20

PLATE 22 x 70, except Fig. 5 which is x 160

Cytheretta minipustulosa sp. nov.

Fic. 1-4 Specimens from the Couches du Phare, Biarritz

lie, i Right valve, male, Io 4080, L = 1:09; HOLOTYPE IMG, 2 Right valve, female, Io 4082, L = 0-98

EG 3 Male carapace, dorsal view, Io 4081, L = 1-00

Fic. 4 Left valve, female, Io 4083, L = 0:88.

Cytheretta samothracia Deltel.

Fic. 5, 6, 8 Specimens from the Couches du Phare, Biarritz Fic. 5 Enlargement of the antero-dorsal area of Io 4077 Ine, © Left valve, female, Io 4077, L = 0:95

Fic. 8 Left valve, male, Io 4078, L = 1-04.

Flexus lenijugum sp. nov.

Fic. 7,9 Specimens from the Argiles a Algues, Blaignan

HG. 7 Left valve, male, Io 4114, L = 0-80

ne, & Right valve, female, lo 4113, L = 0:80; HOLOTYPE.

Bull. By. Mus. nat. Hist. (Geol.) 21, 6 PLATE 21

PLATE 22 x 60

Flexus plicatus (von Munster). BiG. 1 Left valve, Io 4094, L = 0-80; Chattian, Astrup.

Flexus concinnus (Triebel).

Ie, 2, 3}

Fig. 2 Left valve, male, Io 4095, L = 0:94; Unt. Meeresand, Alzey Trift Fic. 3 Left valve, male, Io 4096, L = 0-81; Falun d’Auvers-St.-Georges

Flexus sp. A Fic. 5 Left valve, female, Io 4118, L = 0-71; Sables de Lede, Bambrugge.

Flexus gutzwilleri (Oertli). Fic. 4 Left valve, female, Io 4098, L = 0-79; Couches du Phare, Biarritz.

Flexus schoelleri (Keij).

Fic. 6-8 Specimens from the Faluns bleues, St. Geours-de-Maremne Fic. 6 Left valve, female, Io 4115, L = 0-71

Fic. 7 Left valve, male, lo 4117, L = 0-74

Fic. 8 Right valve, female, Io 4116, L = 0-71.

Flexus lenijugum sp. nov.

Fic. 9, 10 Specimens from the Argiles a Algues, Blaignan

Fic. 9 Male carapace, dorsal view, lo 4114, L = 0-80

Fic. 10 Female carapace, dorsal view, Io 4113, L = 0:80; HOLOTYPE

Cytheretta bullans sp. nov. Fic. 11 Female carapace, dorsal view, Io 4061, L = 0-85; HOLOTYPE. Calcaire a Algues, Blaignan.

Cytheretta postornata sp. nov. ETG. £2 Female carapace, dorsal view, Io 4087, L = 0-83; HOLOTYPE. Couches de l’Atalaye, Biarritz.

Bull. By. Mus. nat. Hist. (Geol.) 21, 6 PLATE 22

JPIGPNIPIS, 2S

x 70

Flexus ludensis sp. nov. Fic. 1-6, 16 Specimens from the Marnes a P. ludensis, Verzy, except Fig. 6

Fic. Tes Fic. Fic. Fic. Fic. Fic.

Left valve, female, Io 4106, L = 0:50; HOLOTYPE

Left valve, male, Io 4107, L = 0-52

Right valve, male, Io 4111, L = 0°51

Male carapace, ventral view, Io 4109, L = 0:52

Right valve, female, Io 4110, L = 0:51

Left valve, male, Io 4112, L = 0-55; Mid. Headon Beds, Whitecliff Bay Female carapace, dorsal view, Io 4108, L = 0-50.

Flexus solentensis solentensis sp. nov. Fic. 7-10 Specimens from the Middle Barton Beds

Fic. Fic. Fic. Fic.

7 8

9 10

Female carapace, dorsal view, Ilo 4101, L = 0-54; Alum Bay Left valve, Io 4101

Right valve, female, L = 0:51; Barton; specimen destroyed Left valve, male, Io 4100, L = 0:54; Barton; HOLOTYPE.

Flexus solentensis congestus subsp. nov. II-I5 Specimens from the Upper Barton Beds, Barton

Rie: Fic. Fic. Fie. Fia. Fic.

It 1i7} 13 14 15

Left valve, female, Io 4103, L = 0-51; HOLOTYPE Male carapace, ventral view, Io 4104, L = 0°54

Left value, Io 4104

Right valve, female, Io 4105, L = 0-51

Male carapace, anterior view, Io 4104.

Bull, Br. Mus. nat. Hist. (Geol.) 21, 6 PLATE 23

ee = - © 7 | vy : : am wy. is ia 7 _ =. ae oe 5

age «2, £

&

\i

Sle

| _

4

A LIST OF SUPPLEMENTS TO THE GEOLOGICAL SERIES OF THE BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY)

. Cox, L. R. Jurassic Bivalvia and Gastropoda from Tanganyika and Kenya. Pp. 213; 30 Plates; 2 Text-figures. 1965. 6.

. Et-NaceGar, Z. R. Stratigraphy and Planktonic Foraminifera of the Upper Cretaceous—Lower Tertiary Succession in the Esna-Idfu Region, Nile Valley, Egypt, U.A.R. Pp. 291; 23 Plates; 18 Text-figures. 1966. {10.

. Davey, R. J., Downtz, C., SARGEANT, W. A. S. & Witiiams, G. L. Studies on Mesozoic and Cainozoic Dinoflagellate Cysts. Pp. 248; 28 Plates; 64 Text- figures. 1966. £7.

. APPENDIX. DAVEY, R. J., Downie, C., SARGEANT, W. A. S. & WILLIAMS, G. L. Appendix to Studies on Mesozoic and Cainozoic Dinoflagellate Cysts. Pp. 24. 1969. 8op.

. Ertiott, G. F. Permian to Palaeocene Calcareous Algae (Dasycladaceae) of the Middle East. Pp. 111; 24 Plates; 17 Text-figures. 1968. £5.12}.

. Ruoves, F. H. T., Austin, R. L. & Druce, E. C. British Avonian (Carboni- ferous) Conodont faunas, and their value in local and continental correlation. Pp. 315; 31 Plates; 92 Text-figures. 1969. {TII.

. Cuitps, A. Upper Jurassic Rhynchonellid Brachiopods from Noveinatatens :

Europe. Pp. 119; 12 Plates; 40 Text-figures. 1969. £4.75.

. Goopy, P. C. The relationships of certain Upper Cretaceous Teleosts with

special reference to the Myctophorids. Pp. 255; 102 Text-figures. 1969. £6.50.

ay . OWEN, H. G. Middle Albian Stratigraphy in the Paris Basin. Pp. 164;

3 Plates; 52 Text-figures. 1971. 6.

. Stippigul, Q. A. Early Tertiary Ostracoda of the family Trachyleberididae

from West Pakistan. Pp. 98; 42 Plates; 7 Text-figures. 1971. £8.

ae Bie Printed in England by Staples Printers Limited at their Kettering, Northants, establishment

ee.

INDEX TO VOLUME a1

New taxonomic names and the page numbers of the principal references are printed in bold type.

An asterisk (*) denotes a figure.

References to ‘tab.’ without indication of a page number refer to no. 5 (Gitmez & Sarjeant): the tables follow p. 250.

Acanthaulax 251 Aulacostephanoides mutabilis zone 177, 181 venusta tab. 3, 5, 6 Aulacostephanus antissiodorensis zone 177-8, Acanthomorphitae 247-9 181, 184 accessory septa 10 eudoxus zone 177, 181 Acritarchs from Kimmeridgian 171-257 pseudomutabilis zone 177, 184 incertae sedis 247-50 Auversian 267 sp. indet. 250, tab. 4, 6; pl. 17, figs. 4 5 Acuticytheretta 274, 276-7 Bachitherium 134, 141 Adnatosphaeridiaceae 234-5 Bactrynium 5 Adnatosphaeridium 234-5 Bairdia subvadiosa 284 paucispinum 234-5, tab. 3, 5, 6; pl. ro, Baltisphaeridium 249 figs. 1-4 claviculorum 248 Aegivia 14 tnusitatum tab. 3, 5, 6 grvayi 14, 25; pl. 4, fig. 3 barbets, see Capitonidae Aegivomena 3, 14, 20 Baryphthengus ruficapillus 163 aquila 14, 25; pl. 1, fig. 3; pl. 3, fig. 5; pl. 4, baylet zone 175-7; see Pictonia baylei figs. 1-2 Belgium 264*, 266* Aegiromeninae 3, 12, 14-15, 17, 18*, 20; pl. 4 Belodinium dysculum tab. 3, 6 Ahtiella 15 bema 10-11 Airtonia 9 Bensonia 274, 276-7 Alcedinidae 153-4, 161, 163-4, 166-8 knysnaensis 277 Alcedo 159*, 163, 165*, 166 Biarritz 267 atthis 153, 163 Billingsellacea 17, 19*, 23 Aleodon brachyrhamphus 40 Billingsellidae 23 alveolus Io Bilobia 15 Ambocythere 274-5 Bimuria 12 Amphitragulus 111, 113, 136 siphonata 12* anderidia 7, 9-10, 12, 20 Birgerbohlinia 138 Andescynodon 39, 44 [footnote] Blastomerycidae 79, 133* Anopliidae 21 blastomerycines 134, 136 Anoptambonites 12 Bohlinia 139 Anseriformes 160 Bos 95 Anticythereis euglypha 276 Boselaphini 125-7 euglyphoidea 276 Boulonnais, Pas-de-Calais, Kimmeridgian 181-2 Antilopini 127-31 Bovidae 125-31, 145-8 Apteodinium 231-2 Bovini 125-7 gvanulatum [no text] tab. 3-6; pl. 12, brachial ridges 1o figs. 4, 5 brachiopods, shell microstructure 4 maculatum 231-2 morphology of shell 4-9 cf. maculatum 231-2, tab. 3-6; pl. 12, fig. 6 brachiopore, ptycholophous 6 Aquitaine Basin 264*, 266*, 267 Brachycythere 274 Argenticythevetta 274, 277 Bracklesham beds 267 Arhkellites hudlestoni zone 177 Bractypteracias leptosomus 166 Aspatha gularis 163, 166 Bradleya 274

Atelornis pittoides 164, 166 Bramatherium 138

352 INDEX

Brunton, C. H. C. 1-26 Bucconidae 154, 163-4 Bucerotidae 168 Buntonia 274-7

Cadomella 4, 19* Cadomellacea 4 Cambridgeshire, Kimmeridge Clay 180 distr. of microfossils in tab. 3 Camelus 95 Canningia minor 227 vingnesit 227-8 Cannosphaeropsis paucispina 234 Canthumeryx 81, 82-4, 94, 136, 143, 145 sirtensis 76, 81-4; pl. 1, figs. 2-5 lower dentition 81-4 Capitonidae 154, 162, 164, 167 Capreolus 98, 101, 103, 105-6, 106*, 108, 110-3, I16—7, 119-22, 121* cavate cysts, indet.sp. A 245-6, tab. 4, 6; pl. 16, fig. 3 sp. B 246, tab. 4, 6; pl. 16, figs. 2, 4 Cervidae 131 Cervoidea 78, 133* Charadriiformes 154, 160, 162-3 Chiniquodontidae 30, 31*, 65 Chonetacea 4, 9, 19*, 21, 25; pls. 1, 5-8 ancestral stocks 9-13 community on sea floor 22* lophophore 4 morphology 4-9, 5* phylogenetic relationships with Plectam- bonitacea 18* shell structure 16-20 spirolophe 4 Chonetacean brachiopods, shell structure 1-26 Chonetes (Eochonetes) celtica 16 primigenius 15, 21 Chonetidina 4 chonetids 3-4, 18* Chonetoidea 3, 10, 15, 21 Chonostrophia 9 Chytroeisphaeridia 185-8, 251 chytroeides 185-6, tab. 3-6; pl. 1, fig. 2 euteiches 187 mantelli 186-7, tab. 3-6; pl. 1, figs. 3-4; pl. 12, fig. 3 pococki 187-8, 190-1, tab. 3-6; pl. 1, fig. 5 sp. 187 Ciconiiformes 160 clavicular plates 11 Cleistosphaeridium 235, 251 ehvenbergi [no text] tab. 3-6; pl. ro, fig. 7 machaerophorum 235 polyacanthum tab. 3-6 polytrichium tab. 3, 6 tribuliferum tab. 3-6 spp. 235, tab. 3-6; pl. 15, fig. 3 Climacoceras 85, 131, 136 Clitambonitacea 23

Coccyzus 159* erythopthalamus pl. 3, fig. L Cooperina 5-6, 19* Coracias 159*, 163, 165* gavrulus 164, 166*, 167; pl. 3, fig. H Coraciidae 153-4, 162—4, 166-8; see Rollers Coraciiformes 153-4, 161-2, 163-4, 167-8 Cormeilles-en-Parisis 265 Cricodon 36, 37*, 39-40 metabolus 40; pl. 3 Crompton, A. W. 27-71 Cryptarchaeodinium 191-3, 251 calcavatum 191-2, 193, tab. 3, 6 [no fig.] cf. calcavatum 192*, 192-3, tab. 3, 6; pl. 2, fig. 3 sp. tab. 5, 6 Cuculidae 154, 161 Cuculiformes 154, 162 cusp nomenclature, ruminant molars and pre- molars 77*, 77-8 Cylindrus jurinei 282 Cynidiognathus 33*, 34 Cynodontia [inford.] 30 Cynodonts, postcanine occlusion in 27—71 Cynognathidae 30, 31*, 65 Cynognathus 34-5; zone 34 Cynosauridae 30 Cynosuchoides 34 Cypridina laticosta 302 plicata 339 [cyst] organism A of Sarjeant, 1960 238 Cythera multinervis 332 Cythere 321 costellata 289, 293, 295 euglypha 276 haimeana 295 jurinet 296, 309, 311, 313 tenuipunctata 309, 313 plicata 304, 338-9 laticosta 302 Cythereis 268* euglyphoidea 276 jurvinel 312, 320 minor 319 plicata 339 vubya 275 (Pseudocythereis) spinifera 275 Cytherella 273 calabra 284 tenuistriata 312 Cytheretta 262, 265, 267, 269-70, 272-7, 279*, 280-1, 282, 283-338 larval stages 280-1 lineages, suggested, in NW Europe 268* soft parts 280 type species 281-2 valves, internal structure of 279* absoluta, see tenuipunctata adviatica 272, 284 antecalva, see costellata

INDEX

bambruggensis 268*, 269, 271*, 293, 294, 338; pl. 3, fig. 9 belgica 272 bernensis 268*, 307, 317-8 bullans 263, 268*, 325-6, 327; pl. 18, figs. 5, 7; pl. 22, fig. 11 buttensis 268*, 272, 315-6 morphotype A 315, 317*; pl. 15, figs. I-5 morphotype B 316, 317*; pl. 15, figs. Saal/ morphotype C 316, 317*; pl. 15, fig. 8 buttensis 263, 307, 315, 316, 317*, 318, 318*; pl. 15, fig. 10 reticulata 263, 307, 316-7, 317*, 318*, 332; pl. 15, figs. 1-8 carita 262, 268*, 299-300, 300%, 345; pl. 9, figs. 5, 8-11; pl. 10, fig. 7 cellulosa 262, 268*, 300-1, 345; pl. 9, figs. I-4, 6-7 aff. cellulosa 301 concinna 294, 339 costellata 268*, 269-70, 272, 274, 278*, 285-93, 286*, 291*, 293*, 345 morphotype A 286*, 287, 292*; pl. 2,

figs. 1-4, 6-10

morphotype B 286*, 287, 292*; pl. 2, fig. 5

morphotype C 286*, 287-8, 292*; pl. 3, fig. I

morphotype D 286*, 288, 292*; pl. 3, figs. 3, 5

morphotype E 286*, 288, 292*; pl. 3, figs. 2, 4, 6-8

morphotype F 286*, 288, 292*; pl. 4, figs. 1-8

morphotype G 286*, 288, 292*; pl. 4, figs. 9-11 :

antecalva 263, 271*, 286*, 290%, 291-3; pl. 4, figs. 9-11 costellata 271*, 272, 286*, 289, 290*; pl. 2 cratis 262, 271*, 278*, 286*, 290, 290*, 292; pl. 4, figs. 1-8; pl. 5, figs. 1-3 grandipora 262, 271*, 286*, 289-90, 290*; pl. 3, figs. 1-8 triangulata 287 crassivenia 268*, 269, 271*, 293-4; pl. 6, figs. 2, 5 crvatis, see costellata decipiens 268*, 269-72, 271*, 294; pl. 6, figs. 8-10 aff. decipiens 271*, 294; pl. 6, fig. 3 edwardsi 273, 280 eocaenica 268*, 269-70, 272, 281, 296-7, 208, 300-1; pl. 1, fig. 2 [no pl. caption]; pl. 7 group 276, 345-6 forticosta 263, 268*, 304, 305-7; pl. 11, figs. 1-4, 8-9; pl. 12, figs. 6-12 gallica, see stigmosa

353

geoursensis 298-9; pl. ro, figs. 1, 3, 5 gibberis 263, 268*, 332, 334-5; pl. 16, figs. 9-10; pl. 19, fig. 10 gracilicosta 340 gvandipora, see costellata grignonensis 268*, 269, 271* gutzwilleri 340; see Flexus haimeana 268*, 269-70, 271*, 294, 295, 296, 345; pl. 6, figs. 1, 6 group 276 species-group, evolution of 270, 271* headonensis 268*, 270, 321*, 321-2, 345; pl. 17, figs. 11, 13-14; pl. 18, fig. 9 inconstans, see sagri judaea 272, 281-2, 283*, 283-4; pl. 1, _ figs. 1, 4-7, 9 jurinet 282, 296, 298 Rlahni 268*, 321 knysnaensis 273-4, 277; see Bensonia laticosta [superspecies] 269-70, 272, 302, 303-7, 306*, 345 [species] 268*, 302-3, 303*, 304-6; pl. 1, fig. 3 [me pl. caption]; pl. ro, figs. 2, 4, 6, 8-9; pl. 12, figs. 1-2, 5 group 276 livata, see tenuipunctata martini, see sagri minipunctata 263, 268*, 307, 314-5, 318; pl. 16, figs. 1-3 minipustulosa 263, 268*, 332, 333-4, 333*; p. 21, figs. 1-4 minor 268*, 301, 319, 325-7; pl. 16, fig. 6 montensis 271*, 272; see nevua multicostata 268, 268*, 271* nevva 267, 268*, 269, 271*, 272 montensis 268, 268*, 271*, 272 oligocaenica 263, 268*, 269-70, 281, 297-8, 299; pl. 8 ornata, see tenuistriata parisiensis, see posticalis perita 268*, 332, 336-7; pl. 19, fig. 8 plicata 339; see Flexus porosacosta 263, 268*, 304-7, 345; pl. 11, figs. 5-7; pl. 12, figs. 3, 4 posticalis 268*, 319, 320, 332, 336-7 parisiensis 263, 320-1; pl. 18, figs. I-4, 6 posticalis 320, 321 postornata 263, 268*, 332, 336, 337; pl. 20, figs. 5-8; pl. 22, fig. 12 vamosa vamosa 307, 317-8 sublaevis 307, 317-8 regularis 263, 268*, 319, 325-6, 326*; pl. 17, figs. 3, 4, 6, 7 reticulata, see buttensis vhenana 268*, 270, 300, 322-3, 325, 346; pl. 18, fig. 8 headonensis 321 stigmosa 323 group 276

354 INDEX

vubya 275, 280-2, 283 yuelensis 262, 268*, 271*, 295-6, 338; pl. 5, figs. 4-7; pl. 6, figs. 4, 7 sagri 268*, 270, 272, 327-9, 332-4, 346 morphotype A 328, 332; pl. 19, figs. 1-3 morphotype B 328-9, 332; pl. 19, figs. 729 morphotype C 329, 332, 336; pl. 19, figs. 5, 6 morphotype D 329, 332, 336; pl. 19, fig. 4 morphotype E 329; pl. 20, figs. 1-3 morphotype F 329, 336; pl. 20, fig. 4 morphotype G 332 inconstans 263, 330*, 330-1; pl. 19, figs. 5-7, 9 martini 263, 330*, 331-2; pl. 20, figs. 14 sagyt 329-30, 330*, 331*; pl. 19, figs. 1-4 group 276 sahnt 273 samothracia 268*, 332-3, 334; pl. 21, figs. 5, 6, 8 schoellert 344; see Flexus scvobiculoplicata 268*, 269, 271*, 272, 294 aff. scvobiculoplicata 271* sculpta 337; pl. 20, figs. 9, Io stigmosa 268*, 270, 318-9, 323, 325-7, 345 gallica 263, 323-5, 324*, 326; pl. 17, figs. I, 2, 5, 10 stigmosa 323 aff. stigmosa stigmosa 270, 323; pl. 16, fig. 8 sublaevis, see vamosa subvadiosa 281, 283, 284-5; pl. 1, fig. 8 tenuipunctata [superspecies] 307-9, 309-17, 318-9 [species] 268*, 270, 272, 301, 309, 314, 324, 332, 346 absoluta 263, 307, 310-1, 311*, 312, 318, 318*; pl. 14, figs. 1-4, 6-7 lirata 263, 307, 310, 311-2, 318, 318%; pl. 14, figs. 5, 8-10; pl. 15, fig. 9;

pl. 16, fig. 4 tenuipunctata 307, 309, 310, 318* group 276

tenuistviata 268*, 309, 312, 316, 318, 332 ornata 307, 308*, 313-4, 314*, 320; pl. 13 tenuistviata 307, 312-3, 318-9; pl. 16, figs. 5, 7 tvacy1t 273, 280 triebeli 268*, 307 varviabilis 268*, 307, 318 vesca 263, 268*, 322-3; pl. 17, figs. 8, 9, 12 sp. A 296, 338; pl. 6, fig. 11 sp. B 338; pl. 9, fig. 12 sp. C 338; pl. 3, fig. 10 Cytherettinae 263, 282-345 characters of subfamily 277

classification 274-7 ecological distribution of Recent species 272-3

evolution of mid-Tertiary 270-2 geographical distribution 274 localities sampled for 266* mid-Tertiary of NW Europe 259-349 palaeoecology 273-4 shell structure 277—80

Cytheridea stviatopunctata 282, 284 subvadiosa 284

Cytheridets colwellensis 321-2

Cytherina costellata 289 subradiosa 284

Cytheromorpha 274

Dacelo 159*, 161*, 163 novaeguineae 163, 166*; pl. 3, fig. J Dama dama 81 Davidsoniacea 4, 6, 19*, 23 Daviesiella 9 Daviesiellidae 9 Dawsonelloides canadensis 17, 25; pl. 7, figs. 1-2 Decennatherium 139 Defiandrea 246 Devonalosia wrightorum 26; pl. 8, fig. 4 Diademodon 35-40, 37*, 39 [footnote], 48, 65*, 67-8; pl. 2, fig. B Diademodontidae 30, 31*, 35-40, 44 [footnote], 66 postcanine teeth 37* Diarthrognathus 30 Dicrocerus 85 ‘Dictyoclostus’ sp. 26; pl. 9, fig. 3 Dictyopyxidia areolata 223 Dictyopyxis 223-4, 251 aveolata 223-4, tab. 3, 5, 6; pl. 7, fig. 9 veticulata 224 cf. veticulata 224, tab. 3, 5, 6; pl. 7, figs. 4, 5; Dire hes m2 Sp. 223 Dingodinium europaeum 244 Dinoferophycidae 185-246 Dinoflagellate cysts from Kimmeridgian 171-257 Dinophyceae 185-246 Dinophyciales 185-246 cyst-family uncertain 233-4, 237-8 Dorcatherium 79, 80, 139-40, 141 chappuist 80, 139 libiensis 76, 80; pl. 1, fig. 1 naui 79-80 parvum 140 pigotti 80, 140 songhorensis 140 Dorset, distr. of microfossils in tab. 3 Dremotheriidae 79, 133*, 137* Dremotherioidea 79 Dremothevium 87, 89-91, 106, 111-3, 134, 136 Dromomerycidae 79, 133* Dromomerycinae 131

INDEX 355

Dvinia prima 32, 33* Dviniidae 30, 32 Dyoros 6-7

sp. 6*, 8*

East African ruminants 139-48 Eathie Haven, Kimmeridge Clay 181, 182*, 183* Egmontodinium 228, 229-31, 251 polyplacophorum 228, 229-31, 230*, tab. 3, 6; plate 8; pl. 9, fig. 3; pl. 11, figs. 5, 6, 8 Ellipsoidictyum 228 aveolata 223 endopuncta 7 Endoscriniacea 239-40 Endoscrinium 239-40, 251 cf. campanula tab. 3, 5, © galeritum tab. 3, 5, © luridum 240, tab. 3-6 oxfordianum tab. 3-6 sp. 239*, 239-40, tab. 5-6; pl. 14, figs. 9-11 Eochonetes 3, 10, 13, 21 celtica 16 primigenius 15 Eomarginifera lobata 26; pl. 9, fig. 2 Eoplectodonta 10-13, 13* tvansversalis 14, 25; pl. 1, fig. 4; pl. 2, figs. 4-6; pl. 3, figs. 1-4 Eoplicanoplia 17 Eotragus 127, 146, 148 haplodon 127 sansaniensis 127 sp. 76, 127; pl. 13, fig. 1 (right) Eozostrodon 38, 68 Eozostrodontidae 31* Epiplosphaera 251 veticulospinosa tab. 3, 5, © Equus 95 Etampes 266* Eucytheretta 274-5, 277, 338; see Flexus plicata 339 Eumeryx culminis 134 Europe, NW, mid-Tertiary Cytherettinae 259- 349 correlation of mid-Tertiary beds 264* Eurystomus 161*, 162-4 glaucurus 164, 166*, 167 Exaeretodon 44, 56, 58-60, 64, 65*, 67 frengnelli 58, 59*, 60

Falunia 281 Flexus 262, 268*, 270, 272, 274-5, 277, 294, 338-9, 339-46 shell structure 281 concinnus 270, 271*, 294, 339; pl. 22, figs. 2,3 congestus, see solentensis decipiens 270; see Cytheretta gvacilicostus 342 gutzwilleri 271*, 340; pl. 22, fig. 4 lenijugum 263, 343-4; pl. 21, figs. 7, 9; pl. 22, figs. 9, 10

ludensis 262, 270, 342-3, 345; pl. 23, figs. I-6, 16 plicatus 270, 271*, 275, 339, 344; pl. 22, fig. I schoellevi 344-5; pl. 22, figs. 6-8 solentensis 270, 272, 340, 342-3 congestus 263, 341-2; pl. 23, figs. 11-15 solentensis 263, 340*, 340-1; pl. 23, figs. 7-10 tviebeli 274 sp. A 345; pl. 22, fig. 5 foraminiferal shell linings tab. 3-5 France, numerical distr. of microfossils in Kimmeridgian tab. 5 Fregata 159*, 161*, 162 ariel pl. 3, fig. B Fregatidae 154 Fromea 188-9, 251 amphora 188 warlinghamensis 188-9, tab. 3, 4, 6; pl. 1, figs. 6, 8; pl. 9, figs. 5, 6 Fromeacea 185-91

Galbula leucogastra 164 Galbulidae 154, 163-4, 167 Galecranium 32 Galeophrys 32 Galesauridae 30, 31*, 32-5, 65

postcanine teeth 33* Galesaurus 34 Gazella 127-31

capricornis 128-30

gaudryi 128

pilgvimi 128-30

sp. 76, 126, 128-31; pl. 13, figs. 2, 3

dentition, lower 128-9 mandible 128

Gebel Zelten, Libya 76-7

Lower Miocene ruminants 73-150 Geisleroceros 168 Gelocidae 133*, 140-2 Gelocus 140, 141

communis 141

whitworthi 76, 140-2; pl. 13, figs. 4, 5 Geranopterus 168 Givaffa 78, 86-8, 91, 95-101, 103, 105-6, 109-10,

I13, 118, 118*, 122-4, 134-5, 139 Giraffidae 75, 79, 85-103, 131, 137*, 138-9 Giraffinae 75, 137*, 139 Giraffoidea 75-150; 80-1, 131, 133*, 136 giraffoids, effects of ossicones on evolution of 134-6

evolution of primitive 131-4 Givaffokeryx 88, 91, 138 GitMe£z, Mrs G. V. & SARJEANT, W. A. S. 171-257 Glochinodon 34 Glochinodontoides 34.

gracilis 33*; pl. 2, fig. A gomphodont cynodonts 29-68 Gomphodontosuchus 40, 44, 58-60, 64

350 INDEX

braziliensis 44, 54, 59-60 Gonambonitacea 23 Gonyaulacysta 193-214, 219, 231, 251 ? acevas 215 aculeata tab. 3, 5, © amabilis 216 angulosa tab. 3-6 cauda 193-4, tab. 3, 5, 6; pl. 2, figs. I, 2, 4, 5 cladophora [no text] tab. 3-6; pl. 2, figs. 7, 8 ehrenbergit 205, tab. 4—6 eisenacki [no text] tab. 3, 5, 6; pl. 3, figs. 6, 7 evitti 220-1 fetchamensis 205 cf. giuseppei 194*, 194-5, tab. 4-6; pl. 3, figs. 3, 4 globata 195-7, 196* tab. 3 4, 6; pl. 3, figs. Ui, 2 granulata 251, tab. 3-6 gvanuligeva 251, tab. 3-6 helicoidea tab. 3, 5, © hyaloderma tab. 3, 5, © jurassica 251, tab. 3-6 longicornis tab. 3-6 longicornis 197-9, 198*, 251, tab. 3-6; pl. 2. fig. 6; pl. 4, fig. 1 cf. mamiullifera 199-200, 201*, tab. 3-6; pl. 4, fig. 7 nuciformis 197, 200-2, 201*, 251, tab. 3-6; pl. 3, fig. 5 palla 205, 209 perfovans 202-4, 203*, 212-3, 251, tab. 3, 4, 6; pl. 4, fig. 6 sevvata 207, tab. 3, 5, © systremmatos 204-5, tab. 5, 6; pl. 5, figs. 7, 8 sp. A 205-6, tab. 3, 5, 6; pl. 9, figs. 1, 2 sp. B 193, 206*, 206-7, tab. 4, 6; pl. 4, figs. 2, 3 sp. C 204, 207-8, 207*, 209, tab. 4, 6; pl. 6, ED i, D sp. D 208-9, 208*, tab. 3, 6; pl. 6, figs. 4, 5 sp. E 209-10, 210%, tab. 4, 6; pl. 6, fig. 9 sp. F 211-2, 211*, tab. 3, 6; pl. 6, figs. 3, 6 sp. G 212-3, 213*, tab. 3, 6; pl. 6, figs. 7, 8 sp. H 213-4, 214*, tab. 3, 6; pl. 13, fig. 1 Gonyaulacystacea 191-223 Gonyaulax 216 acevas 215 amabilis 216 longicoynis 197 nuciformis 200 perfovans 202-3 Gravesia (gigas, gravesiana, ivius) zones 177 Grekoffiana [‘Gvrekkofiana’] 274, 276-7; see Protocytheretta australis 276 daniana 276 Griquatherium 138 Gruiformes 154 gulls 153-4, 160

Halcyonidae 153 Halcyornis from Lower Eocene, affinities 151-69 toliapicus 151-69, 155*, 156*, 157*, 158*, 159*, 161*, 166*, 168; plates1, 2; pl. 3, fig. E cranium, compared with Recent forms 157-63 detailed comparison with coracii- formes and piciformes 163-7 measurements 157 state of preservation 154-7 interorbital septum and foramen 160-2 parasphenoid region 162-3 orbit, upper edges of 159-60 Halcyornithidae 153, 168 Hamitton, W. R. 73-150 Hampshire Basin 264*, 266*, 267 Haplocytheridea 274 Harrison, C. J. O. & WALKER, C. A. 151-169 Headon beds 267 Helaspis luma 26; pl. 9, fig. 1 Helladotherium 138 Hestertonia pellucida tab. 3, 5, 6 Heterocemas 136 simpsont 132-3 Hexagonifera 240-1, 246, 251 chlamydata 241 jurassica 240-1, tab. 3-6; pl. 14, figs. 5, 8 Sp. 240 Hexagoniferacea 240-1 Histiophora cf. ornata [no text] tab. 3-6; pl. 7, figs. 7, 8 Honanotherium 137, 139 horns, definition 78 Horridonia horrida 26; pl. 9, fig. 4 Hydaspitherium 138 Hystrichosphaeridiacea 235-7 Hystrichosphaeridium capitatum 189 claviculorum 248 petilum tab. 3-6 pulchevrimum 235-6

Ictidosauria 30, 31* Idoceras balderum bed 184. Ilyobates ? judaea 281-3 Imbatodinium 232-3, 251 antennatum 232-3, tab. 3-6; pl. 11, figs. 2,3 villosum 233 cf. villosum 233, tab. 3, 6; pl. 11, fig. 1 Sp. 232 ‘Indratherium’ 138 inner side septum 10 Inversella 15 Ischignathus 60, 67

Jura Mountains Kimmeridgian 184-5 Juresania 17

KEEN, M. C. 259-349 Kenyameryx africanus 146

INDEX 357

Kimmeridge Clay microplankton 171-257 zonal classifications 177 Kimmeridgian microplankton, distribution com- pared with previously known stratigraphical range tab. 6 Kingfishers, see Alcedinidae Kistecephalus zone 39 [footnote] Kuehneotheriidae 31*

Lagomerycidae 131-2 Lagomeryx 131-2 simpsoni 132 Lariidae 154, 160, 163, 167 Lavus 159*, 165* argentatus pl. 3, fig. C canus 153 vidibundus 153, 162 toliapicus 153, 168 Leavachia 33* durenhaget 32 Ledian 267 Leguminocythereis 274 Le Havre, Kimmeridgian 183-4 Leiosphaeridia (Chytroeisphaeridia) chytroeides 185 Leiostrea delta bed 190 Leptaenisca 20 Leptellinidae 13 Leptelloidea 15 leptelloides 25; pl. 5, fig. 3 Leptestia 15 musculosa 25; pl. 5, fig. 4 Leptodinium 215-20, 251 aceras 215-6, 215*, tab. 3, 4, 6; pl. 5, figs. 1-3 amabilis 216-7, 217*, 220, tab. 3, 5, 6; pl. 10, figs. 5, 6 arcuatum tab. 3-6 clathvatum tab. 5, 6 crassineyuum 219 cf. crassinervum 218-9, 218*, tab. 3, 6; pl. 3, fig. 8; pl. 5, figs. 4-6 egemenii [no text] tab. 3, 5, 6; pl. 7, figs. 1, 2 subtile tab. 3, 5, © sp. 219-20, 220*, tab. 3, 4, 6; pl. 3, fig. 9 Leptosomatidae 153-4, 162-4, 166 Leptosomus 161*, 162-4 discoloy 164-5, 166* Libya, Lower Miocene ruminants 73-150 Libytherium 138 Loculicytheretta 274 London Basin 266* Lophiomeryx 134, 140-1 lophophore 4-5, 7, 10-3 lophophore platform 11 Lorraine, Kimmeridgian 184 Luangwa drysdalli 40 Lucina inoynata, Marnes a 265 Lystrosaurus zone 39 [footnote] Lyttoniacea 6, 19*

Mainz Basin 266* Manda formation (M. Trias, Tanzania) 40, 43-4 Massetognathus pascauli 56, 58-60 Nn. sp. 44, 56-8, 57*, 60, 62*, 63-8, 65*; plate 7 Megalaima 162-4 haematocephala 164-5, 165* vivens 164, 165*, 166* Megathivis 7, 11 Meiourogonyaulax 224-8, 251 dicryptos 225-6, 225*, tab. 3, 5, 6; pl. 7, fig. 6 pila 226-7, 227*, tab. 3, 4, 6; pl. 4, fig. 5; pl. 7, fig. 3 staffinensis 224-5, 251, tab. 3-6; pl. 9, fig. 4 sp. A tab. 3, 6 sp. B 227-8, tab. 3, 4, 6; pl. 4, fig. 4; pl. 7, fig. 12 Membranilarnax ovulum 223-4 Meropidae 154, 163 Merops 161*, 162-3 apiaster 163, 166; pl. 3, fig. G Micrhystridium 247-8, 252 fragile tab. 3-6 inconspicuum 247-8, tab. 3-6 vavispinum 248 vecurvatum 247, tab. 3-6; pl. 17, figs. I, 2 sydus tab. 3-6 sp. 247-8, tab. 3, 6; pl. 17, figs. 7, 8 Microdiniacea 223-31 Micromeryx 144 microplankton, organic-walled 171-257 microstructure of brachiopod shell 4 Mid-Tertiary beds of W. Europe, correlation 264* Miocene, Lower, ruminants of Libya 73-150 Moiselles 265 Momotidae 154, 161, 163-4, 166, 168 Momotus 162-3, 165* martit 166* Monasa 162 morphoeus 164-6, 166*; pl. 3, fig. I Mont Crussol (Rhone) Kimmeridgian 184 Moorellina 7 Morganucodon 38, 68 motmots, see Momotidae Mudevorgia 241-2, 252 mewhaet 242 simplex 241-2, tab. 4, 6; pl. 15, figs. I, 2 Mudevorgiacea 241-2 muscles, mechanical advantage of, in Giraffidae 101 ‘mutabilis’ zone 177

Nanictosaurus 34 Nanictosuchus 34 Nannoceratopsis pellucida tab. 3, 6 Nelsoniellacea 242-5 Neochonetes 17 Neocyprideis 274, 322 colwellensis 322

358 INDEX

Neocytheretta 274-5

Netrelytvon 251 parum tab. 3-6 stegastum tab. 3, 5, ©

Netvocytheridea 274

Netromorphitae, organism A 249, tab. 3, 5, 6;

plSn6; fies ipl aes

nisusiids 17, 23

Normandy, Kimmeridgian 182-3

Ntawere Formation (Trias, Tanzania) 40 (Zambia), undescribed cynodont reptile

from 36 Nucula comta, Argiles 4 (Belgium) 309, 339

Occisucysta 220-3, 251 balios 221; tab. 3-6 evittt 220-1 [no fig. ] monoheuriskos 221-3, 222*, tab. 3, 6; pl. 7, figs. Io, 11 RJD: AH), ison 3, ©) Okapia 78, 86-91, 95-101, 103, 105-7, 109-13, 118-24, 118*, 119*, 120*, 121*, 134-5, 138 oldhaminids (brachiopods) 4 Oligocene, Upper 267 Oligokyphus 67 Oligosphaeridium 235-6 pulcherrimum 235-6, 236*, tab. 3, 5, 6; plier igs Ovangiotherium 138 Orthacea 23 ossicones, definition 78 effect on giraffoid evolution 134-6 of Prolibytherium 107, 116 pair of, indet. 85 Ovis 118, 124 outer side septum 10, 12 Oxfordshire, Kimmeridge Clay 180 distribution of microfossils in tab. 3

Pachygenelus 30 Palaeohypsodontus 131 Palaeomerycidae 75-6, 78-9, 81, 82-5, 131-2, 136-7, 137*, 142-5 palaeomerycid, indet. 85; pl. 1, fig. 6 Palaeomeryx 81-3, 92-5, 98, 105, 114, 121-2, 131-4, 136, 141, 144-6, 148 africanus 76, 131, 140, 143, 146 furcatus 143 MAgNUS 143 sansaniensis 137 sp. 84 ‘Palaeomeryx’ fold 77, 80-1, 84-5, 94, 128, 137, 146, 148 Palaeoperidinium bicuneatum 242 nuciforyme 200 nuciformoides 200 veticulatum 224 Palaeostomocystis laevigata 249 Palaeotraginae 75, 85-103, 137*, 138-9 Palaeotvagus 92-5, 98, 132, 137-9

microdon 87, go-2 vyouentt 94, 116 Pavacytheretta 268*, 269, 274-5, 277 veticosa 269 schoelleri 344 Paveodinia 251 cevatophova tab. 3-6 nuda 199 Pareodiniacea 231-3 Paris Basin 264*, 266* Parvocavatus tuberosus [no text] tab. 3-6; pl. 14, fig. 4 Pascualgnathus 39, 44 [footnote] Pavlovia pallasianus zone 177 pallasoides zone 177 votunda zone 177-8, 180 Pecora 134 Pectinatites pectinatus zone 177-8, 180 (Arkellites) hudlestont zone 177-8, 181 (Virgatosphinctoides) elegans zone 177-8, 181 scitulus zone 177-8, 181 wheatleyensis zone 177-8, 181 Pelecaniformes 154, 160, 162 Pentacrinus, Couches a 338 Pholadomya ludensis, Marnes a 264*, 265, 290, 294, 299, 303, 342, 345 Pholidostrophia 9 Piciformes 153-4, 161-3, 164*, 167 Pictonia baylei zone 177 Platycraniellus 34 Plectambonitacea 3, 9-13, 17, 19*, 20, 23, 25; plates 1-5 phylogenetic relationships with Chonetacea 18* shell structure 13-6 Plectambonites seviceus 21 Plectodonta 12-3 Plurviarvalium 228 Podocopida 282-345 Podocopina 282-345 pollen tab. 3-5 Polymesoda 265 Polystephanephorus 251 savjeantit [no text] tab. 3, 5, 6; pl. 12, figs. 8, 9 Pontocyprella 273 Posidonia 272 postcanine occlusion in cynodonts and tri- tylodontids 27—71 Prionocytheretta 282 Procellariidae 154 Procellariiformes 154, 160, 162 Procervulus 131, 136 Procynosuchidae 30, 31*, 32-5, 39 [footnote], 65 postcanine teeth 33* Procynosuchus 32 Prodvemothevium 134, 141 Productacea 4, 18*, 19, 19*, 23, 26; plate 9 Productidina 4—6 productids 4, 18*

INDEX 359

Proexaeretodon 60, 67 Progivaffa 136 Prolibytherium 84-5, 88-9, 104 magniert 75-6, 104-25, 134-8, 137*; plates 7-12 astragalus 121-2, 125 atlas 122, 123* brain 110-3, 112* calcaneum 121, 125 cervical vertebrae 122, 123* dentition, upper 113-4, 116-7 lower 114-7 frontal 107 functional interpretations 124 humerus 118-9, 119*, 124 jugal 106 lacrymal 105-6 mandible 113 maxilla 105, 106* metacarpal 120, 125 metatarsal 122 occipital 108-9 ossicones 107, 108* palatine 106 parietal 107 petrosal 110, 110* phalanges 122, 125 post-cranial skeleton 117-25 radius 119-20, 120*, 124-5 scapula 118, 118*, 124 skull 105-17 sphenoid 109 squamosal 109-10 thoracic vertebrae 122-4, 123* tibia 120-1, 121*, 125 ulna 120 vertebral column 122-4 Prolixosphaeridium 251 cf. deivense tab. 3, 6 granulosum [no text] tab. 3-6; pl. 13, figs. Om parvispinum tab. 3, 4, 6 Propalaeoryx 81-4, 136, 142, 143-5 austroafricanus 142 nyanzae 76, 142-5; plate 14 Protochonetes 15-6, 21 ludloviensis 15, 17 styiatellus 15-6 Protocynodon 34 Protocytheretta 273-7, 345 damiana 273, 275-6, 345; see Grekoffiana multicarinata 273 schoellevi 275, 344 Protornis 168 Protragocerus 126 chantrei 126 gluten 126 sp. 76, 126, 127; pl. 13, fig. 1 (left) proximate cyst sp. indet. 233-4, tab. 3, 4, 6; pl. 11, figs. 4, 7,9

Psaligonyaulax 251% apaleta tab. 3-6 sp. tab. 3, 6 Pseudocythereis 274-5, 277 spinifera 275 Pseudocytheretta 282 pseudopuncta 7, 13-4, 16-7; plate 1 Pteromorphitae 249-50 Plevospermopsis 249-50 australiensis [no text] tab. 3-6; pl. 12, fig. 7 harti 249-50, tab. 3-6; pl. 17, fig. 6 helios [no text] tab. 3, 5, 6; pl. 16, fig. 5 Plterygocythereis 274 Ptychoglyptus 15 ptycholophe 5, 11 ptycholophous brachiopore 6 Puffinus 159*, 161* diomedia pl. 3, fig. C Purbeck, Isle of, Kimmeridgian 178, 179*

Ramphastidae 154, 164, 166 Rasenia (cymodoce, mutabilis) zones 177 Retichonetes vicinus 17, 25; pl. t, fig. 2; pl. 7, fig. 3 Rhactorhynchia inconstans bed 249 Rhine, Germany 264*, 266* Rhinochetidae 154, 161 rhynchonellids 13 Rhynchops 160-1, 161* niger pl. 3, fig. F Rhynochetos 159* jubatus pl. 3, fig. A Richthofeniacea 6, 10, 19* rollers (birds) 153, 166-7; see Coraciidae Rugosochonetes silleest 17, 25; pl. I, fig. 1; pl. 7, fig. 4; pl. 8, figs. 1-3 Spp. 17 rugosochonetids 6 Ruminants, Lower Miocene of Gebel Zelten, Libya 73-150 East African 139-48

Samotherium 98, 139 sinense 87, 134 SARJEANT, W. A. S. 171-257 Scalenodon 36, 40, 41-56, 60 angustifrons 40, 41, 42, 44 [footnote], 44-9, 45*, 46*, 47*, 50-1, 53, 56, 58-60, 62*, 63, 65*, 66-7; plate 4 attridgei 29, 42, 43, 53-4, 55* (textfigs. 10A, B) 58, 60, 63-4, 67; plate 6 charigi 29, 42, 44, 54-6, 55* (textfig. 10C), 60, 63-4 hirschsoni 29, 42, 43, 49-53, 50*, 52*, 53* (textfigs. 7-9), 54, 58, 62* (textfig. 13), 63-4, 65*, 67; plate 5 Scalenodontoides 60 macrodontes 59-60 schizolophe 5, 11 Scotland, distribution of microfossils in tab. 3

360 INDEX

Scriniodinium 242-4, 251 bicuneatum 242-3, tab. 3-6; pl. 15, fig. 4 cvystallinum tab. 3—6 dictyotum 243-4 dictyotum 243, 244*, tab. 3, 6; pl. 16, fig. 6 osmingtonensis 243-4, 244*, tab. 3, 6; pl. 15, fig. 5 papillatum 243-4, 244*, tab. 3-6; pl. 15, fig. 6 pyrum 243-4, 244*, tab. 3, 5, 6; pl. 15, fig. 7 cf. galeatum tab. 5, 6 playfordi tab. 3-6 Selenideva 162 langsdorffi 164-5 Semicytheretta 274, 276-7 Sentolunia 10, 21 Sericoidea 14-5, 21 yestvicta 14, 25; pl. 4, fig. 4; pl. 5, figs. 1, 2 shell structure, Chonetacean brachiopods 1-26 side septa, inner and outer 10 Silphedestes 34 Silphedestidae 30 Sivmiodinium 245, 252 grossi 245, tab. 3, 6; pl. 16, figs. 7, 8 Sivatheriidae 75, 79, 85, 103-25, 103-4, 137*, 138 Sivatherium 105, 109, 138 Skye (Staffin Bay), Kimmeridge Clay 181 socket ridges 11 Solisphaeridium 248-9 brevispinosum tab. 3, 5, 6 claviculovrum 248-9, tab. 4, 6; pl. 17, figs. 9, 10 stimuliferum 252, tab. 3-6 Sowerbyella 3, 13, 21 (Viruella) liliifera 25; pl. 2 figs. 1-3 Sowerbyellidae 9-10 12, 14, 18* Sowerbyellinae 12-4, 18*, 20; plate 2 Spiniferites 251 spores tab. 3-5 Staffin Bay, Skye, Kimmeridge Clay 181 Stampian 267 Stapilinum cistum tab. 3, 5, © Stephanelytron 237-8, 251 vedcliffense 237-8, tab. 5, 6; pl. 14, fig. 6 cf. vedcliffense 238, tab. 5, 6; pl. 14, fig. 7 cf. scarburghense tab. 3, 5, © Stevaspis zone 177 Strophalosiacea 4, 6, 18*, 19, 19*, 26; plate 8 stropheodontids 9 ~ Strvophochonetes 10, 15-6, 21 celtica 16 cingulatus 15 primigenius 15-7, 25; pl. 5, figs. 5-8; pl. 6, figs. I-4 Strophomenacea 4, 9-10, 18*, 19*, 23 Strophomenida 4-5, 14 phylogeny of superfamilies 19*

strophomenids 3 Subplanites (gvandis, wheatleyensis, spp.) zones 177 Sylvicapra 128 Systematophora 237, 251 aveolata tab. 3-6 orbifeva [no text] tab. 3-6; pl. 13, fig. 2 ovata 237, tab. 3, 6; pl. 14, figs. 1-3 SP. 237

Taeniophova 251 tunctispina tab. 3, 5, 6 taleolae 7, 13 Tenua 187, 189-91, 251 capitata 189, tab. 3-6; pl. 1, figs. 11, 12 echinata 190, tab. 3-6; pl. 1, figs. 1, 9 hystrix tab. 3-6 pilosa tab. 3-6 veyvucosa IQ villersense 191 sp. 190-1, tab. 3, 6; pl. 1, figs. 7, 10 tenuilobatus zone 177 terebratulids 13 Thaerodonta 13 Thecideacea 4—6, 19*, 23 Thecidellina 5, 11 Thecospiva 4-5, 19* therapsid reptiles 30 Thrinaxodon 33*, 34-5, 39 [footnote], 48, 65* liorhinus 34; plate 1 Tinodon 38 Toquimia 15 Trachyleberidea 274 Trachyleberididae 274, 282-345 Tragulidae 79-81, 139-40 Traguloidea 139-45 Tvaversodon stahleckeri 54, 60 Traversodontidae 30, 31*, 40-61, 65, 68 distribution 60-1 Tribolodon 34 Triceromeryx 130-7 Trichodinium sp. tab. 3, 6 Triplesiacea 4, 23 triplesiidines (brachiopods) 4 Trivachodon 36, 39, 65* angustifrons 4% Trirachodontidae 30, 31*, 35-40, 39 [footnote], 66 postcanine teeth 37* Trithelodon 30 Tritylodon 61-4, 62*, 65*, 67 Tritylodontidae 31*, 61-4 Tritylodontids, postcanine occlusion in 27-71 Turdus sp. 167

Uintornis 168 Upupa 161*, 164

epops 164-5; pl. 3, fig. K Upupidae 154, 161, 164, 166

INDEX 361

Veryhachium hyalodermum tab. 3-6 dentition, upper 90, 92-3

Virgatites miatschkovensis zone 177 lower 90, 93-5

Virgatosphinctoides (elegans, nodiferus, scitulus, ear region 89

wheatleyensis) zones 177 femur 97, 102 Viruella 13 frontal 87 lihifera 25; pl. 2, figs. 1-3 functional interpretation ror, 103

humerus 96, 102 jugal 87

Walangania 83, 141, 143-4, 145, 146-8 africanus 76, 133, 141, 146-8 dentition 147 gracilis 76, 146 WALKER, C. A. 151-69 Warlingham (Surrey) borehole, Kimmeridge Clay 180-1 numerical distribution of microfossils in Kimmeridgian tab. 4 Weymouth district, Kimmeridgian 179*, 180 wood fragments tab. 3-5

lacrymal 87

lumbar vertebrae 99*, 100-1 mandible 89

maxilla 86

metacarpal 96—7, 102 metatarsal 98, 103

nasal 86-7

occipital 87-8

palatine 87

parietal 87

pectoral girdle 95

pelvic limb 97-8

phalanges 97, 102 post-cranial material 95-103 radius 96, 102

Xenambonites 15 Xestoleberis auvantia 321

Zaphrentis delanouei 272 scapula 95-6, 102 Zarafa 85, 105, 109-10, 114-6, 118, 118%, 120-2, skull 86-95 I21*, 132, 134-6, 137*, 139 sphenoid 88 zelteni 75—6, 86-103; plates 2-6 squamosal 88-9 astragalus 98, 103 thoracic vertebrae 99*, 100 axis 98-100, 99* tibia 97-8, 103 calcaneum 98, 103 ulna 96

cervical vertebrae 99*, 100 vertebrae 98-102, 99*

il * x Oe by rote F y te < ’ ti; he » " mt i , ‘ ' i © et ‘ ‘ ; ( s P b a { y,»' « +

Peay

gy

a * 4 ny or aa Ati

=f ta :

Senet nL Weeicegires trey ener

383

witty et Hints gesersatestezersesceass etzsete rere rereerers

ceeregts Sorener satires