'O ^
'V v

-4* 'P

'V ^

t

VOLUME 3 • PART 3

Palaeontology

DECEMBER 1960

PUBLISHED BY THE
PALAEONTOLOGICAL ASSOCIATION
LONDON

THE PALAEONTOLOGICAL ASSOCIATION

The Association was founded in 1957 to further the study of palaeontology. It holds
meetings and demonstrations, and publishes the quarterly journal Palaeontology.
Membership is open to individuals, institutions, libraries, &c«*on payment of the
appropriate annual subscription:

1960 1961 onwards

Institutional membership .. . . . £2. 2s. ($6.50) £5. 5s. ($15.50)

Ordinary membership . £2. 2s. ($6.50) £3. 3s. ($ 9.50)

Student membership . . . . . . . . £2. 2.y. ($ 6.50)

There is no admission fee. Student members will be regarded as persons receiving
full-time instruction at educational institutions recognized by the Council. Subscrip-
tions are due each January, and should be sent to the Treasurer, Professor P. C.
Sylvester-Bradley, Department of Geology, The University, Leicester, England.

Palaeontology is devoted to the publication of papers (preferably illustrated) on all
aspects of palaeontology and stratigraphical palaeontology. Four parts are published
each year and are sent free to all members of the Association. Members who join for
1960 will receive Volume 3, Parts 1 to 4.

Back parts may be purchased separately at a cost of £2 ($6) post free for each part;
orders should be sent to the approved agents, Messrs. B. H. Blackwell, Broad Street,
Oxford, England.

Manuscripts on all aspects of palaeontology and stratigraphical palaeontology are
invited. They should conform in style to those already published in this journal, and
should be sent to Mr. N. F. Hughes, Sedgwick Museum, Cambridge, England.
A sheet of detailed instructions for authors will be supplied on request.

PALAEONTOLOGY

VOLUME 3 • PART 3

CONTENTS

The median abdominal appendage of the Silurian Eurypterid Slimonia acuminata
(Salter). By Charles d. waterston 245

A new Echinoid from the Lower Cretaceous (Albian) of Kent. By Raymond
casey 260

Calathospermum fimbriatum sp. nov., a Lower Carboniferous Pteridosperm
cupule from Scotland. By p. D. w. Barnard 265

The external anatomy of some Carboniferous ‘Scorpions’, Part 2. By Leonard
j. wills 276

The Peltaspermaceae, a Pteridosperm family of Permian and Triassic age. By
JOHN A. TOWNROW 333

Photonegative young in the Triassic Lamellibranch Lima lineata (Schlotheim).

By R. P. S. JEFFERIES 362

The Cretaceous species of Pyripora d’Orbigny and Rhammatopora Lang. By
H. DIGHTON THOMAS and G. P. LARWOOD 370

The Permian Leiopteriid Merismopteria and the origin of the Pteriidae. By
j. m. dickins 387

Characters and relationships of the Mesozoic Pelecypod Pseudavicula. By J. m.
dickins 392

THE MEDIAN ABDOMINAL APPENDAGE OF THE
SILURIAN EURYPTERID SLIMONIA ACUMINATA

(SALTER)

by CHARLES D. WATERSTON

Abstract. The results of a study of a series of specimens exhibiting the median abdominal appendage of
Slimonia acuminata (Salter) from the Upper Silurian of Logan Water, Lesmahagow, Lanarkshire, are given.
The anatomy of the opercular appendage of both sexes is redescribed in the light of knowledge obtained from
recent work on related structures in other eurypterids, and the segmented nature of the appendages demon-
strated. Stages in the development of the appendage in both sexes in ontogeny are described for the first time.

INTRODUCTION

Sexual dimorphism was first noted in eurypterids by Woodward (1866-78, p. 61) who
described two forms of median abdominal appendage in Pterygotus bilobus Salter. Later
in the same work, Woodward (pp. 114-19) gave a description of both types of sexual
appendage in Slimonia acuminata (Salter) and compared them with the opercular struc-
tures of Limulus. Since the appearance of Woodward’s monograph, much detailed re-
search has been done on eurypterid anatomy in Europe and America and important
works dealing with the form of the median appendage and its function have been
published. An historical summary of earlier work has been given by Stormer (1934a,
pp. 43-50) who later published two masterly descriptions of the opercular appendages
of German Lower Devonian eurypterids: Grossopteris overathi (Gross) (19346, pp.
289-90) and Pterygotus rhenaniae Jaekel (1936, pp. 14-23). Recent detailed descriptions
of American eurypterids, in which particular attention is given to the sexual appendage,
include Kjellesvig-Waering (1948, pp. 22-23) on the Carboniferous Adelophthalmus
mazonensis (Meek and Worthen), and Caster and Kjellesvig-Waering (1956, pp. 22-27)
on the Upper Silurian form Dolichopterus jewetti Caster and Kjellesvig-Waering.

The purpose of the present work is to re-examine the opercular structures of Slimonia
acuminata in the light of our greatly increased knowledge of related structures in other
eurypterids, and to describe the growth stages of the median appendage of both sexes,
an aspect which appears to have received little attention from previous workers. The
study has been made possible by the large size of the appendages in Slimonia which are
often excellently preserved in the Upper Silurian mudstones of Logan Water, Lesmaha-
gow, Lanarkshire, from which locality all the specimens examined have come. The
relatively large number of available specimens having the opercular appendage preserved
has also facilitated the work.

History of previous work. Neither in Salter’s original description of the species (1856,
p. 29) nor in Page’s original figure of the genus (1856, p. 135, fig. 3) is the form of the
median appendage of Slimonia acuminata recorded. Huxley and Salter (1859, pp. 59, 60),
however, figured both types of genital appendage. The type A appendage, using Stormer’s

[Palaeontology, Vol. 3, Part 3, 1960, pp. 245-59, pis. 42-43.]

B 6612 R

246

PALAEONTOLOGY, VOLUME 3

classification (1934 a, p. 46), was regarded as a portion of the epistoma of a species
referred with hesitation to '‘Pterygotus’ acuminatus, and the type B appendage was
regarded as the epistoma of that species.

The sexual significance of the central lobe of the operculum was first realized by
H. Woodward (1866-79, pp. 114-20) who figured both types of appendage and com-
pared them with the opercular structures of Limulus. Unfortunately the supposed sexual
dimorphism which he recognized in Limulus was based on two distinct species (Pocock
1902, p. 259, footnote). Woodward regarded the tricuspid type A appendage as being
female and the type B appendage as male, and in both he figured what he regarded as
genital openings on the deltoidal plates of the operculum. (For definitions of parts see
text-figs. 1, 3.) Sutures were not noted in the type A appendage, but in the type B form,
Woodward described the transverse furrows which occur at the distal end of the append-
age and which he thought were variable in number and were due to there being a number
of similar appendages lying one on top of another, the end of each projecting beyond
that of the one above it. Laurie (1893, pp. 512-15) adds little to Woodward’s description
of the type A appendage, except to refute the idea that Parka decipiens , sometimes found
in association with it, represents egg-capsules, a supposition which had been used by
Slimon and Woodward to support their claim that the type A form was female. Laurie
gave a careful account of the type B appendage which he regarded as being a single
structure, the distal end of which was eversible. The variation in the number of furrows,
a fact which Laurie did not question, he regarded as being due to the different extent to
which it was protruded in different cases. In neither appendage did Laurie mention the
genital openings described by Woodward, and their existence was queried by Clarke
and Ruedemann (1912, p. 64), who quoted the findings of Woodward and Laurie and
adopted the sex determination given by Woodward. In describing the type B appendage,
they stated that there are two transverse furrows at the distal end.

Gaskell (1908) had compared the opercular appendages of eurypterids with those of
living members of the Arachnida, and concluded that the type A appendage of Eurypterus
belonged to the male, and not the female, as had been previously supposed. Stormer
(1934#) found the occurrence of secondary sexual characters in eurypterids, such as
clasping organs, to support Gaskell’s sex determination, and figured the two types of
genital appendage of Slimonia accordingly. Through a knowledge of the position of
transverse sutures in the type A appendage of other eurypterids, Stormer indicated
sutures in his drawing of the tricuspid appendage of Slimonia, although these are not
repeated in his later figure (1955, fig. 21, 2c).

In a paper concerned primarily with the gill-like structures of eurypterids, P. F.
Moore (1941, p. 66) figured a type A appendage of Slimonia acuminata in which two
pairs of ridged areas, interpreted as muscle scars, are indicated near the base of the
appendage. Moore did not indicate the presence of transverse sutures on the specimen
which he figured.

THE TYPE A APPENDAGE

Anatomy of the adult appendage. The appendage is situated behind the paired deltoidal
plates (delt.p., text-fig. 1) of the operculum. The anterior part of the appendage is
hastate, the hastation being bounded anteriorly by the inner post-lateral sutures (i.p.l.s.)
of the deltoidal plates. The sutures meet at the anterior tip of the median appendage and

C. D. WATERSTON: SLIMONIA ACUMINATA (SALTER)

247

continue anteriorly as a single median suture (med.s.), separating the paired deltoidal
plates, to the anterior margin of the operculum. The deltoidal plates meet the opercular
plates (op.p.) in the outer post-lateral sutures (o.p.l.s.) which may be straight or angled
causing the deltoidal plates to be four- or five-sided.

The shape of the type A appendage is broadly fusiform, the point of greatest breadth
being approximately level with the posterior margins of the opercular plates. The length

text-fig. 1. a, Reconstruction of the dorsal surface of the type A median abdominal appendage of
Slimonia acuminata (Salter) x f. b, Reconstruction of the dorsal surface of the type A median abdomi-
nal appendage of Pterygotus rhenaniae Jaekel, after Stormer, X 2 approx, ap. = male gonopore,
d.f. = dorsal furrow, d.s. = distal segment, delt.p. = deltoidal plate, i.p.l.s. = inner post-lateral
suture, int. = thin integument, m.s. = middle segment, med.s. = median suture, musc.sc. = muscle
scar, o.p.l.s. = outer post-lateral suture, op.p. = opercular plate, p.s. = proximal segment, v.d. =
median canals or vasa deferential, lst.t.s. = first transverse suture, 2nd.t.s. = second transverse suture.

is between three and four times the breadth. The anterior part of the appendage is
hastate, the posterior portion is trispinate. The posterior spine is normally slightly
longer and more acute than the paired spines which extend farther laterally than the
greatest breadth of the appendage. A marked feature is the asymmetry of the paired
spines, one being more posteriorly directed than the other, which is set slightly angled
towards the posterior from the normal of the long axis of the appendage.

A longitudinal median ridge extends posteriorly from the rear of the hastate portion
of the appendage. The margins closely follow the outline of the appendage and taper to
a narrow keel in the posterior median spine. The median ridge occupies about one-third
of the width of the appendage.

The type A appendage is divided into three segments by two transverse sutures. The

248

PALAEONTOLOGY, VOLUME 3

first separates the proximal segment from the middle segment, the former making up at
least two-thirds of the total length of the adult appendage. The form of the suture varies
throughout ontogeny (see text-fig. 2), but in adult specimens it is divided into three
parts. The median section is curved as the suture passes over the median ridge, which

text-fig. 2. Growth stages of the type A median abdominal appendage of Slimonia acuminata (Salter).
a, Kelvingrove Museum .09. 123. aa, X2. b, Kelvingrove Museum .09. 123. aca, Xl. c. Geological
Survey Museum 87280, xl.D, British Museum (Natural History) 45157, X -J. e, Royal Scottish Museum

1859. 35. 5, X|.

is wide at this point since the edges of the ridge flare outwards as the posterior margin
of the proximal segment is approached. The two lateral sections of the suture also curve
outwards as they are traced posteriorly to the margins of the appendage. The form of
the suture appears to vary in different specimens owing to the accident of flattening of
the median ridge during preservation. The second suture separates the middle segment
from the distal segment. It also has a median curved portion passing over the median
ridge, which again broadens as it approaches the posterior margin of the middle seg-
ment. The lateral parts of the suture pass outwards to the tips of the lateral spines in a

C. D. WATERSTON: SLIMONIA ACUMINATA (SALTER)

249

straight line bisecting them into unequal portions. The part of the spine formed by the
middle segment is narrower than the posterior portion formed by the distal segment.

Stormer (1934a, text-fig. 19) indicated possible positions for the transverse sutures of
the type A appendage, basing his figure on an illustration given by Woodward (1866-78,
pi. 17, fig. 2). Examination of this specimen (B.M.N.H. 45157) shows a crack running
across the appendage almost midway down the proximal segment which was shown in
Woodward’s figure and interpreted by Stormer as a suture. The position of the first
suture is, however, correctly shown in Woodward’s illustration in the more distal posi-
tion, while the central part of the second suture is drawn between the lateral spines. The
sutures are particularly well seen in specimens exhibiting the dorsal side of the appendage.

With regard to the finer structures of the type A appendage, Woodward (1866-78,
p. 116, text-fig. 35) figured two circular openings, one on each of the deltoidal plates,
which he interpreted as ovipores. These are not mentioned by Laurie (1893) and a re-
examination of the figured specimen (B.M.N.LI. 44445) does not support Woodward’s
claim. What were thought of as gonopores appear as two very small wrinkles in the
integumen which are neither rounded nor pore-like. It is true that other workers, for
example Holm (1898, p. 43) and Clarke and Ruedemann (1912, p. 66 and text-fig. 122),
have described what they regard as gonopores as occurring in a rather similar position
in other genera. These, however, occur at the lateral points of the hastate basal portion
of the proximal segment of the median appendage and not on the deltoidal plates. They
are thus not homologous to the pores described by Woodward. In the course of the
present study many perfectly preserved specimens have been examined and no similar
structures have been seen. They are therefore regarded here as merely an accident of
preservation.

In the description of the median ridge it has been stated that it extends posteriorly
from the rear of the hastate portion of the appendage. Preparation of one specimen
(R.S.M. 1859. 35. 5) suggests that the median ridge may be accompanied by a comple-
mentary median furrow in the dorsal surface of the appendage, the opening of which
does not extend as far as the ridge itself. The specimen displays a mould of the interior
of the dorsal surface of the appendage, and the median ridge was broken away at the
midline of the proximal segment rather less than half-way from the anterior tip of the
segment. Where the ridge was broken away, two natural margins were revealed con-
verging anteriorly, where they meet to form the boundaries of a lanceolate groove.
Hair-like thickenings of the integumen occur along these margins, the hairs being directed
posteriorly and diagonally inwards towards the mid-line. The hairs appear to extend
beyond the margins and protrude into the groove.

The ridged areas regarded by Moore (1941) as muscle scars flank the median ridge on
the proximal segment posterior to the hastation. On Moore’s specimen (Sedgwick
Museum A 16236) two pairs of ridged areas are seen, the larger being anterior to the
smaller, which has a more central position on the appendage (see PI. 42, fig. 2). Similar
areas may be seen in the Geological Survey Nos. 87282, 87283, and 87285; in the Royal
Scottish Museum 1859. 35. 5; and in part in the Kelvingrove Museum .09. 123. pi. The
areas are marked by narrow branching ridges radiating in a post-lateral direction with
intervening smaller, parallel ridges which may be very numerous (see PI. 42, fig. 3). An
examination of the muscular attachments on the dorsal surface of the operculum of
Limulus has confirmed, in the writer’s opinion, Moore’s view that the areas represent

250

PALAEONTOLOGY, VOLUME 3

muscular attachments. The pattern of the markings compares closely with that of the
attachment of the ‘branchial muscles’ in Limulus (Benham 1885), although the muscles
of the type A appendage of Slimonia cannot be regarded as homologous with these.

Development of the type A appendage. Examination of a series of type A appendages
has shown how different growth stages, exhibiting very different shapes, are related to
one another. It has thus been possible to work back from the known adult appendage
to the previously unrecognized nepeonic and juvenile forms.

The smallest type A appendage seen has a total length of 24-0 mm. and a maximum
breadth of 6-0 mm. and is preserved in the Slimon collection of the Kelvingrove museum
No. .09. 123. aa. (see text-fig. 2a, and PI. 42, figs. 4-6). The maximum width of the
appendage is at the posterior angles of the hastation, from which it tapers posteriorly.
The proximal segment comprises over three-quarters of the total length of the appendage,
as compared with two-thirds in adult examples, and its posterior margin, marked by the
first suture, is anterior to the rear margins of the opercular plates. The latter feature has
been noted in a nepeonic example of the type A appendage of Eurypterus fischeri Eich.
by Holm (1898, p. 43, pi. 1, fig. 11). The suture is curved, the convexity towards the
anterior, and the post-lateral angles are acute. Lateral spines are not developed, the
middle segment tapering towards the posterior, where the post-lateral angles are more
acute than in the proximal segment and are posteriorly directed. The second suture, like
the first, is curved with the convexity towards the anterior. The distal segment is ellipti-
cal, being longer than it is broad, and divides into two minute points at the posterior
extremity of the appendage. The median ridge extends from the rear of the hastation to
the distal segment, where it terminates before reaching the posterior margin. The oper-
cular plates of this specimen are interesting in that they indicate the transverse fused
suture extending laterally from the posterior points of the deltoidal plates, a feature
which is rarely seen in Slimonia (PL 42, fig. 4).

The next growth stage is shown in the Kelvingrove Museum specimen No. .09. 123. aca.,
in which the length of the appendage is 39-0 mm. and the maximum breadth 14-0 mm.
The broadest part lies in the proximal segment behind the hastation. The proximal seg-
ment still constitutes about three-quarters of the total length of the appendage. The first
suture, however, lies posterior to the rear margins of the opercular plates (see text-fig.
2b). In the post-lateral angles of neither the proximal nor the middle segment is the angle
so acute as in the smallest form, although they are still posteriorly directed. The middle
segment at this stage is proportionately longer in relation to the other segments than in
any other stage in ontogeny, comprising about one-sixth of the total length of the
appendage as compared with one-ninth in the smallest specimen. The distal segment is
triangular, tapering posteriorly to a point, showing no indication of the small paired

EXPLANATION OF PLATE 42

Figs. 1-6. Slimonia acuminata (Salter). Median abdominal appendage Type A. 1, Entire appendage.
X R.S.M. 1859.35.5. 2, Entire nepeonic appendage. X 3, Kelvingrove .09. 123. aa. 3, Distal part
of the nepeonic appendage showing the distal, middle and part of the proximal segments. Alcohol
immersed X 6, Kelvingrove .09.123. aa. 4, Diagrammatic representation of fig. 3. 5, Part of the
proximal segment to show the two pairs of muscle scars. X 1|, Sedgw. Mus. A 16236. 6, Part of the
proximal segment to show muscle scars. X 1 J, R.S.M. 1859.35.5.

Palaeontology, Vol. 3,

PLATE 42

WATERSTON, Slimonia acuminata (Salter)

C. D. WATERSTON: SLIMONIA ACUMINATA (SALTER) 251

horns noted at the posterior extremity of the smallest specimen. The length of the distal
segment is smaller than the maximum breadth.

An appendage exhibiting a slightly later stage in development is Geological Survey
87280, counterpart 87281, which is probably the specimen which was figured by Huxley
and Salter (1859, pi. 15, fig. 3) as a problematical fragment. The length is 48-0 mm. and
the maximum breadth 13-0 mm., but the point of maximum breadth lies midway down
the length of the proximal segment instead of at its anterior as in smaller examples
(text-fig. 2c). In other respects this specimen resembles No. .09.123 aca., except that the
median ridge is more pronounced, and as a consequence the sutures have become
divided into a central and lateral parts such as is seen in the more adult specimens. The
relationship of the first suture with the posterior margins of the opercular plates is not
seen, since the appendage has become separated from the rest of the operculum.

The British Museum specimen B.M. (N.H.) 45157 is small but exhibits adult characters
(text-fig. 2d). The posterior margins of the opercular plates are now level with the mid-
point of the length of the proximal segment. The trispinate distal feature is present
although not fully developed. The post-lateral angles of the middle segment have been
produced to form the anterior portion of the lateral spines. The antero-lateral parts of
the distal segment have also been expanded to form the posterior portions of the lateral
spines, angles being developed in the lateral margins of the distal segment between the
lateral spines and the posterior median spine. All three spines are less acute than in
adult specimens, in which they are still further produced (text-fig. 2e).

In summary, the development of the type A appendage appears to exhibit the following features:

1 . A greater growth rate in the length of the proximal segment in relation to the length of the opercular
plates from the condition seen in the nepeonic form, in which the proximal segment does not extend
beyond the posterior margins of the opercular plates, to the adult condition in which the rear margins
of the opercular plates are approximately on a level with the mid-line of the proximal segment.

2. The gradual posterior migration of the point of maximum breadth in the proximal segment from
the anterior part of the segment to half-way down the length.

3. The reorientation of the post-lateral angles of the middle segment from a posterior to a postero-
lateral or lateral direction, with their elongation to form the anterior parts of the lateral spines.

4. The changing shape of the distal segment from elliptical to triangular, and finally to the expansion
of the antero-lateral parts to form the posterior portions of the lateral spines, and the production of the
posterior tip to form the median spine.

5. The early loss of the paired horns at the posterior of the nepeonic appendage.

6. The growth of the median ridge with the consequent change in shape of the sutures from a simple
curve to central and lateral parts.

THE TYPE B APPENDAGE

Anatomy of the adult appendage. Relationships between the deltoidal plates, the oper-
cular plates, and the hastate anterior part of the median appendage are the same as those
in the type A appendage.

The shape of the type B appendage is fairly narrow and conical, the point of greatest
breadth approaching the posterior. It is made up of three segments of which the proximal
is by far the largest, comprising approximately eight-ninths of the whole. A median
ridge extends across the proximal segment from the lateral points of the hastation to the
first suture. The ridge is waisted towards the centre of the segment, but flares out
laterally in an ogee curve towards the rear of the segment, and occupies over two-thirds
of the width of the segment at the posterior margin. In the Royal Scottish Museum

252

PALAEONTOLOGY, VOLUME 3

specimen 1859. 35. 7 longitudinal folds of the integumen commence within the median
ridge at the posterior third of the proximal segment, and after running parallel for a
short distance in the mid-line, curve outwards towards the post-lateral extremities of the
median ridge in a more accentuated ogee curve than that followed by the margin of the
ridge. This gives the appearance of two ducts or channels passing downwards and out-

text-fig. 3. a, Reconstruction of the dorsal surface of the type B median abdominal appendage of
Slimonia acuminata (Salter), X J. b, Reconstruction of the dorsal surface of the Type B median ab-
dominal appendage of Pterygotus rhenaniae Jaekel, x2 approx, d.s. = distal segment, delt.p. =
deltoidal plate, m.s. = middle segment, med.s. = median suture, o.v.d. = female gonopores,
op.p. = opercular plates, ? p.m.p.s. = ? posterior margin of the proximal segment, ? p.m.m.s. =
? posterior margin of the middle segment, p.s. = proximal segment, lst.t.s. = first transverse suture,

2nd.t.s. = second transverse suture.

EXPLANATION OF PLATE 43

Figs. 1-6. Slimonia acuminata (Salter). Median abdominal appendage Type B. 1, Invagination of the
right side of the distal segment of the nepeonic form showing the striate ornament, x 144, Kelvin-
grove .09.123.aav. 2, Distal part of the adult appendage showing part of the proximal, the middle,
and the distal segments with post-lateral complexes, note radial fold pattern in integumen of distal
segment. X 11, R.S.M. 1859.35.7. 3, Distal part of a flattened adult appendage showing part of the
proximal, the middle, and distal segments with post-lateral complexes, note radial fold pattern in
integumen of distal segment. X 1 -1, Geol. Surv. Scot. 6643. 4, The entire adult appendage, xf,
R.S.M. 1859.35.7. 5, Distal part of the juvenile form showing part of the proximal segment, the
middle segment, and the distal segment with lateral spurs. x3, Kelvingrove .09.123.vq. 6, Entire
appendage of the nepeonic form. X 3, Kelvingrove .09.123.aav.

Palaeontology, Vol. 3.

PLATE 43

WATERSTON, Slimonia acuminata (Salter)

C. D. WATERSTON: SLIMONIA ACUMINATA (SALTER)

253

wards to the post-lateral points of the median ridge (see text-fig. 4e, and PI. 43, fig. 5).
The first suture is semicircular in the area of the median ridge, with the convexity
towards the anterior. The lateral parts of the suture curve outwards to the post-lateral
angles of the proximal segment in shallower curves the convexities of which are also
directed anteriorly. The centre part of the second suture is also semicircular, and con-
centric to the first, thus confining the median portion of the middle segment to a narrow
semicircular band. The central part of the distal segment has the form of a half-circle,
being bounded anteriorly by the curved second suture, whereas the posterior margin is
almost straight. Paired sigmoidal ridges extend laterally from the centre of the segment,
near the posterior margin, to join the post-lateral complexes of the appendage.

Previous authors have described the type B appendage as terminating distally in a
more or less truncated cone, which is marked by two or three deep furrows (Laurie
1893, p. 513). Laurie advances a convincing argument for refuting Woodward’s idea that
the type B appendage consists of three similarly shaped median plates, lying one on top
of another, the end of each projecting a little beyond the one above, and correctly inter-
prets the transverse furrows as structures belonging to a single median appendage. In
addition to claiming that there may be two or three such furrows, Laurie says that
the preservation of markings on the remains of these animals seemed to him to
depend so much on the details of fossilization, and perhaps also on the condition of the
animal at death, and that their presence on some specimens, and absence on others, is
not of much weight as an argument. The writer sympathizes with this view since there is
a great deal of variation in the appearance of the distal structures of the type B appen-
dage due to the accidents of fossilization, and particularly to the flattening of a three
dimensional structure. Careful examination of many adult specimens, however, makes
it clear that throughout the variation there can be traced constant features which are due
to the original form of the appendage and not to accident. Of these, the presence of two
transverse sutures is one constant feature, and the occurrence of what are here termed
the post-lateral complexes is another.

The form of the median portions of the two transverse sutures, or furrows, which
divide the appendage into proximal, middle, and distal segments has been described.
Posterior to the appendage, however, there occurs in some specimens (e.g. R.S.M.
1859. 35. 7 and Geol. Surv. Scot. 6643) a semicircular piece of integumen which prob-
ably served to attach the distal part of the appendage to the abdomen of the animal
(see text-fig. 4e, f, g). The furrows which exist between the integumen and the posterior
margin of the distal segment is probably what Laurie regarded as the third furrow, but
is not a true suture.

The post-lateral complexes occur in all specimens which have been examined and are
made up of structures developed from the lateral parts of the middle and distal segments.
The complexes occur in the post-lateral angles of the appendage. Since their anatomy is
more easily understood when their development through ontogeny is known, the
description of these features is included in the following account of the development of
the type B appendage.

Development of the type B appendage. A small appendage exhibiting the dorsal surface,
which is here identified as a type B appendage in the nepeonic condition, is in the Slimon
Collection, Kelvingrove Museum No. .09. 123 aav. (text-fig. 4a; PI. 43, figs. 1, 2). Its

254

PALAEONTOLOGY, VOLUME 3

text-fig. 4. Growth stages of the Type B median abdominal appendage of Slimonia acuminata (Salter).
a, Kelvingrove Museum .09. 123. aav, x2. b, Distal portion of Kelvingrove Museum .09. 123. aav,
X4. c, Kelvingrove Museum .09. 123. vx, X 1. d, Distal portion of Kelvingrove Museum .09. 123. vx,
x4. E, Royal Scottish Museum 1859. 35. 7, x|. f, Distal end of Royal Scottish Museum 1859. 35. 7,
X 1. g, Geological Survey Scotland 6643, distal end X 1.

total length is 22-0 mm. and the maximum breadth, which occurs just posterior to the
hastation, is 8-0 mm. Behind the hastation the form of the appendage is obconical. The
proximal segment does not extend beyond the posterior margins of the opercular plates,
and the bounding first suture is arcuate in form. The middle segment is broadest at the

C. D. WATERSTON: SLIMON1A ACUMINATA (SALTER)

255

anterior and tapers posteriorly, but flares slightly at the post-lateral angles. The second
suture is in the form of a simple arch. The distal segment is broadly triangular in form.
Embayments or invaginations occur in the dorsal surface of the distal segment, one on
each side near the margins. That on the right side is more clearly seen than the one on
the left owing to a slight anticlockwise twist in the specimen. The invaginations are
situated towards the anterior of the segment and are bounded anteriorly by spurs which
are seen immediately behind the second suture. From the lateral tips of the spurs, which
form the antero-lateral angles of the segment, the margins are continued in a straight
line towards the posterior tip of the segment. A remarkable feature, which has been seen
in this specimen only, is the ornament of extremely fine striae which it exhibits. This is
best seen on the right-hand side of the distal segment where the striae radiate across the
surface of the segment from the inner margin of the invagination. Those striae, which
extend anteriorly, cross the second suture without interruption and continue on the
right postero-lateral portion of the middle segment. Striae which originate within the
invagination at the posterior margin of the spur extend parallel to the edge of the seg-
ment in a posterior direction and then sweep diagonally across the rear part of the seg-
ment (text-fig. 4b; PI. 43, fig. 1 ). Ornamentation can be traced also on the left posterior
angle of the hastation, where striae radiate across the appendage from a point near the
re-entrant angle at the rear of the hastation, and continue across the suture and over the
left deltoidal plate. It would appear most probable that in life the entire surface of
the appendage was sheathed in a fine integumen bearing striated ornament.

The 'median ridge’ divides anteriorly into three lobes on the proximal segment of the
appendage. There is a larger central lobe which does not extend forwards as far as the
hastation, whereas the two lateral lobes almost reach the margins of the segment antero-
laterally. The lobes unite at a point about one-quarter of the length of the segment from
its posterior margin, and continue rearwards as a single median ridge which extends
over the middle and distal segments. The margins of the ridge flare outwards at the
sutures and it terminates posteriorly on the distal segment in two lobes.

The next growth stage is exhibited by the Kelvingrove specimens .09. 123. vq and
.09. 123. vx. The latter (text-fig. 4c) has a length of 30-0 mm. and a maximum breadth
of 8-0 mm. A similar growth stage is shown in R.S.M. 1865. 11. 11, Geol. Surv. 87287,
Geol. Surv. Scot. 11825 and 11826. At this stage the shape of the appendage has already
assumed the conical form of the adult, the point of greatest breadth being near the
posterior of the proximal segment. The proportions of the segments and the form of the
sutures are also similar to those of the adult. Differences occur, however, in the form of
the post-lateral complexes, which are much simpler than in the mature appendage. The
centre part of the distal segment is approximately semicircular in form, being bounded
anteriorly by the median-curved portion of the second suture, and posteriorly by an
almost straight margin. From this central region two lateral curved spurs or horns are
developed from the distal segment, one on each side of the segment. They extend antero-
laterally from the sides of the semicircular median part of the segment and then curve
so that their tips are directed laterally or slightly post-laterally. Their anterior margins
form a simple curve, convex towards the anterior, whereas the posterior margins are
sigmoidal, the inner ends terminating near the median line of the appendage. The
median-curved portion of the second suture terminates against the anterior margins of
the spurs, and is presumably coincident with these margins as it is traced laterally.

256

PALAEONTOLOGY, VOLUME 3

Intermediate growth stages such as are shown in specimens Geol. Surv. 87288, 87289,
and B.M. (N.H.) 59653, occur between that just described and the adult form. These,
however, demonstrate only the constancy of the conical form of the appendage after the
nepeonic stage has been passed, and a gradual increase in the complexity of the post-
lateral regions.

The finest adult type B appendages seen are R.S.M. 1959. 35. 7 and Geol. Surv. Scot.
6643 (text-fig. 4e, f, g). Here the lateral horns of the distal segment, after growing
antero-laterally, appear to curve fairly sharply until they are growing in a post-lateral
direction. They extend post-laterally until their tips are posterior to the rear margin of
the central part of the distal segment. In Geol. Surv. Scot. 6643 the appendage has been
greatly flattened, and where the lateral spurs curve round from an antero-lateral to a
postero-lateral direction they have been flattened upon themselves at the points of
curvature, to form two round bodies (PI. 43, fig. 6; text-fig. 4g). This is not seen in
R.S.M. 1859. 35. 7, which has not suffered such severe flattening.

The lateral margins of the middle segment appear to extend posteriorly from the post-
lateral angles of the proximal segment to the posterior part of the antero-lateral margin
of the spurs of the distal segment. The supposition finds support in the simpler conditions
seen in the juvenile forms (text-fig. 4c, d). A margin traverses the spurs of the distal
segment and terminates laterally at the point where the lateral margins of the middle
segment meet the distal spurs. This margin is rather indistinct and is interpreted by the
writer as the posterior margin of the middle segment on the ventral side of the appen-
dage. On the dorsal side, the posterior margin of the middle segment would appear in its
lateral parts to coincide with the antero-lateral margin of the spurs of the distal segment.
Another indistinct margin may be traced extending from the post-lateral angles of the
proximal segment to a point just anterior to the position of curvature of the spurs of
the distal segment. It is possible that this parallels the margin seen traversing the distal
spurs and represents the posterior margin of the proximal segment on the ventral side
of the appendage.

In summary, four features appear to be common to each of the post-lateral complexes of the adult
type B appendage :

1 . A distal member in a post-lateral orientation, which is regarded as a post-lateral extension of the
spur of the distal segment clearly seen in juvenile forms.

2. Another member in a post-lateral orientation, lying anterior to the distal spur and posterior to
the post-lateral angle of the proximal segment, which is thought to represent the lateral portion of the
middle segment.

3. A margin traversing the posterior part of the distal spur, which is regarded as the posterior margin
of the middle segment.

4. A margin traversing the middle segment between the post-lateral angle of the proximal segment
and a point anterior to the position of curvature of the distal spur, which is regarded with hesitation
as the posterior margin of the proximal segment on the ventral side of the appendage.

Although these features appear to be constant despite variation in the condition of
preservation of the specimens, the interpretation of the structures presents many difficul-
ties owing to the degree of specialization attained in Slimonia.

The form of the nepeonic type B appendage may be related to that of the adult
appendage by the differential growth-rate diagram of text-fig. 5. The lines superimposed
upon the outlines of the appendages have been constructed in the following ways. I,
crosses the proximal segment three-quarters of its length from the anterior margin and

C. D. WATERSTON: SLIMONIA ACUMINATA (SALTER) 257

is of purely diagrammatic significance. II, touches the post-lateral angles of the proximal
segment, between which it crosses the middle segment. These points are not in doubt in
either form. Ill, touches the post-lateral angles of the middle segment, between which
it crosses the distal segment at the anterior margin of the invaginations. There is no
difficulty in constructing this line in the nepeonic form, but to do so in the adult form
requires the assumption that the spurs, which lie anterior to the invaginations and
posterior to the second suture in the nepeonic appendage, become extended in the adult
to form the posterior member of the post-lateral complex. If this is allowed, then the

text-fig. 5. Diagram to illustrate the possible relationship through differential growth-rates of the
form of the nepeonic (A) and adult (B) median abdominal appendage, type B, of Slimonia acuminata

(Salter).

points about which the distal spurs are curved in the adult must represent the anatomical
equivalent of the invaginations of the nepeonic form. IV, is a line crossing the posterior
lobes of the distal segment. Support is given for the validity of the assumptions made in
construction by the regular pattern which emerges in the resulting diagram. While the
lines are parallel with and at right angles to the long axis of the appendage in the nepeo-
nic form, they form a symmetrical series of curves in the adult which become more
intense as the posterior of the appendage is approached. The adult form would therefore
appear to be derived from the nepeonic form by greater differential growth at the
margins of the distal part of the appendage as compared with a smaller growth-rate in
the median region.

SEX AND FUNCTION OF THE APPENDAGES
While dimorphism of the median abdominal appendages of eurypterids is obvious,
the determination of sex has proved difficult since no direct comparison can be made
with closely related living forms. Following the work of Pocock (1902) on the secondary
sexual characters among Xiphosura, and Gaskell’s (1908) comparison of the opercular
structures of eurypterids with those of Thelyphonus , Stormer (1934 a) argued that the
type A appendage was male and the type B female. He found support for this view
principally in the secondary sexual characters seen in eurypterids such as clasping organs
of Mixopterus and Eurypterus. More direct evidence became available in the wonder-
fully preserved genital appendages of the lower Devonian Pterygotus rhenaniae Jaekel,

258

PALAEONTOLOGY, VOLUME 3

which is of particular interest for the present study, since the genital appendages of
Slimonia acuminata may be compared closely with this form.

The type A appendage of P. rhenaniae is narrowly lanceolate and is made up of three
segments (text-fig. 1). The form of the appendage is less specialized than in Slimonia but
is very similar to the nepeonic condition of the Scottish species, particularly in the shape
of the distal segment with its terminal incision dividing the posterior margin into two
lobes. On the dorsal side of the German species, Stormer (1936) was able to detect a
central aperture towards the posterior of the proximal segment which he regarded as
probably the combined aperture of the vasa deferentia. This hypothesis is supported by
the occurrence in P. rhenaniae of two parallel canals anterior to the opening. While no
trace of vasa deferentia or genital aperture has been seen on the type A appendage of
Slimonia, it may well be that the male aperture opened into the dorsal furrow which
occurs on the proximal segment of the appendage in an equivalent anatomical position
to the male gonopore of the German species.

The type B appendage of P. rhenaniae is broad and pear-shaped, and is also made up
of three segments whose proportions are very similar to those of the adult type B appen-
dage in Slimonia (text-fig. 3). Towards the posterior of the proximal segment, two fairly
large ovate openings occur which may be ovipores. Similar openings have not been seen
in Slimonia. On the other hand, post-lateral complexes, which are a specialized feature
of the Slimonia type B appendage, have not been described from P. rhenaniae.

The comparison which can be made between the appendages of Pterygotus rhenaniae
and Slimonia acuminata is sufficiently close to leave no doubt that the trispinate appen-
dage of Slimonia is equivalent to the lanceolate appendage of Pterygotus rhenaniae, and
Stormer’s determination of a male sex for these appendages is here adopted. In the same
way the similarities between the pear-shaped or conical appendages in both species are
sufficient to prove their equivalence, and a female sex is presumed for this type. The
differences in the more detailed features of the anatomy, however, suggest that the
functioning of the appendages was distinct in the two species.

It is reasonable to suppose that the post-lateral complexes of the type B appendage of
Slimonia may represent ovipositor mechanisms. This appears not unlikely when the
longitudinal median structures of the proximal segment are considered. As has been
stated above these structures have the appearance of two ducts or channels which pass
posteriorly and laterally in ogee curves on either side of the hinder portion of the proxi-
mal appendage, and terminate in the region of the post-lateral complexes. These chan-
nels could be regarded as genital ducts leading to the post-lateral complexes. In P.
rhenaniae, however, the ovipores are found on the proximal segment. If the post-lateral
complexes of Slimonia are regarded as fulfilling an ovipository function, it is necessary
to postulate the participation of the middle and distal segments in the female genital
system, in addition to the proximal segment of the appendage. Laurie (1893) regarded
the distal portion of the type B appendage of Slimonia as being eversible. The writer
favours this suggestion because of the regular radial fold pattern that is found on the
median part of the distal segment in nearly all specimens of the type B appendage.
Were the distal part of the appendage eversible it would appear to favour the supposi-
tion that the function of the post-lateral complexes was ovipository.

Acknowledgements. For the loan of material which has made the present work possible, grateful
acknowledgement is made to: Dr. E. I. White and Dr. H. W. Ball, British Museum (Natural History);

C. D. WATERSTON: SLIMONIA ACUMINATA (SALTER)

259

Dr. W. H. C. Ramsbottom and Mr. J. Smith, Geological Survey, London; Mr. R. B. Wilson, Geological
Survey, Edinburgh; Mr. A. G. Brighton, Sedgwick Museum, Cambridge; Dr. S. M. K. Henderson and
Mr. C. E. Palmar, Glasgow Museum, Kelvingrove. For the provision of photographic facilities thanks
are due to Professor F. H. Stewart, Grant Institute of Geology, Edinburgh. Publication has been
assisted by a grant from the Carnegie Trust which is gratefully acknowledged.

REFERENCES

benham, w. b. s. in lankester, e. r., benham, w. b. s., and beck, e. J. 1885. On the muscular and
endoskeletal systems of Limulus and Scorpio: with some notes on the anatomy and generic charac-
ters of scorpions. Trans. ZooL Soc. London , 11, 311-84.
caster, k. e. and kjellesvtg-waering, E. n. 1956. Some notes on the genus Dolichopterus Hall.
/. Paleont. 30, 19-28.

clarke, J. m. and ruedemann, r. 1912. The Eurypterida of New York. Mem. N.Y. St. Mas. 14.
gaskell, w. h. 1908. The origin of Vertebrates. London.

holm, G. 1898. Uber die Organisation von Eurypterus Fischeri d'Eichwald. Mem. Acad. Sci. St.
Petersb. 8th ser. 8, 2.

huxley, t. h. and salter, j. w. 1859. On the anatomy and affinities of the genus Pterygotus. Mem.
Geol. Surv. Monogr. 1.

kjellesvig-waering, e. n. 1948. The Mazon Creek Eurypterid. A revision of the genus Lcpidoderma.
Illinois Sci. Pap. 3, no. 4.

laurie, M. 1893. The anatomy and relations of the Eurypterida. Trans. Roy. Soc. Edinb. 37, 509-28.
moore, p. f. 1941. On gill-like structures in the Eurypterida. Geol. Mag. 78, 62-70.
page, d. 1856. Advanced text book on Geology, 1st ed. Edinburgh.

pocock, r. J. 1902. The taxonomy of Recent Limulus. Ann. Mag. Nat. Hist. (7), 5, 256-66.
salter, J. w. 1856. On some new Crustacea from the uppermost Silurian rocks. Quart. J. Geol. Soc.
London, 12, 26-34.

stormer, l. 1934a. Merostomata from the Downtonian Sandstone of Ringerike, Norway. Skr.
Norske Vidensk. Akad., I.M.-N-K1. 1933. 10.

19346. Uber den neuen, von W. Gross beschriebenen Eurypteriden aus dem Unterdevon von

Overath im Rheinland. Jb. Preass. Geol. L.-A., 55, 284-91.

1936. Eurypteriden aus dem Rheinischen Unterdevon. Abb. Preuss. Geol. L.-A, n.f. h. 175.

— - — • 1955. Merostomata in Treatise on invertebrate paleontology, Part P, pp. P4-41. Lawrence,
Kansas.

woodward, H. 1866-78. British fossil Crustacea belonging to the order Merostomata. Palaeontogr.
Soc.

CHARLES D. WATERSTON

Royal Scottish Museum,

Manuscript received 8 July 1959 Edinburgh

A NEW ECHINOID FROM THE LOWER
CRETACEOUS (ALBIAN) OF KENT

by RAYMOND CASEY

Abstract. Holaster cantianus sp. nov. is a large holasterid of Lower Cretaceous (Lower Albian) age and is
found in the Lower Greensand (Folkestone Beds) of the Folkestone neighbourhood of Kent. It is designated
type species of a new subgenus, Labrotaxis, to which is also referred the Upper Albian-Cenomanian Holaster
latissimus J. L. R. Agassiz. The diagnostic feature of Labrotaxis is the primitive condition of the meridosternous
plastron.

The common occurrence of a large echinoid belonging either to Holaster or to Car diaster
in the Folkestone Beds (Lower Albian) of the Folkestone neighbourhood of Kent has
long been known (Price 1874; Topley 1875; Casey 1949), but owing to the imperfect
state of preservation of the specimens no attempt at specific determination or description
has been hitherto made. The following account is based on material accumulated by the
author over a period of many years and now incorporated in the collections of the Geo-
logical Survey Museum, London. The paper is published by permission of the Director
of the Geological Survey and Museum.

Order holasteroida Wyatt Durham and Melville 1957
Family holasteridae Pictet 1857
Genus holaster J. L. R. Agassiz 1836
Subgenus labrotaxis nov.

Type species. Holaster ( Labrotaxis ) cantianus sp. nov.

Diagnosis. Holaster with well-developed frontal sulcus bounded by carinae. Plastron of
primitive meridosternous type, the labrum spanning the oral margin of sternal 2' to make
contact with sternal 2.

Remarks. Among the Spatangoida and the Holasteroida the structure of the plastron
has high taxonomic value. In the former order it is of amphisternous type, characterized
by the labrum being joined at its posterior end to two large, equally developed sternal
plates (2', 2). The Holasteroida include those forms in which the plastron is merido-
sternous, with the labrum abutting against a single sternal plate (2'). Lambert (1893)
showed that in Holaster intermedins (Munster), of very early Cretaceous (Neocomian)

explanation of plate 44

Figs. 1-4, Holaster ( Labrotaxis ) cantianus sp. nov., Lower Greensand (Folkestone Beds), Folkestone,
Kent. Author’s collection, now in the Geological Survey Museum, London. 1, Adoral surface, the
sutures inked-in. Zm 32, reduced XO 8. 2, Part of adapical surface showing pores and ornament.
The apical system lies on the right-hand border of the photograph; the frontal sulcus runs obliquely
upwards. G.S.M. 97304, enlarged x5. 3, Two of a group of three specimens (the third is on the
other side), the larger being the holotype. G.S.M. 74118-19, natural size. 4, Smaller specimen illus-
trated in fig. 3 enlarged X 2.

[Palaeontology, Vol. 3, Part 3, 1960, pp. 260-4, pi. 44.]

Palaeontology, Vol. 3.

PLATE 44

C AS EY, Holaster ( Labrotaxis ) cantianus sp.nov.

RAYMOND CASEY: A NEW ECHINOID

261

age, the meridosternous type of plastron is already typically developed, the labrum being
well separated from the second sternal plate (2). His illustrations of the plastron of H.
nodulosus (Goldfuss), the type species of Holaster, from the base of the Upper Creta-
ceous, show this second sternal plate even farther removed from the labrum, its position
behind sternal plate 2' foreshadowing the single row of plastronal plates peculiar to the
subgenus Sternotaxis. On the other hand, the plastron of Labrotaxis (PI. 44, fig. 1 ; text-fig.
Id), having the labrum just touching sternal plate 2, represents a more primitive stage in
the evolution of the meridosternous condition than is seen in H. intermedins. In all
other features, especially the intercalary type of apical system, Labrotaxis conforms to
the Holasteridae. It lacks the ambital fasciole of Cardiaster and the sharp frontal carinae
and unequal pores of Pseudholaster. An Upper Albian-Cenomanian form, H. Iatissimus
J. L. R. Agassiz, is also referable to Labrotaxis and the scope of the subgenus may be
expected to widen with increasing knowledge of the plastronal structure of other members
of the genus.

Holaster ( Labrotaxis ) cantianus sp. nov.

Plate 44, figs. 1-4; text-fig. 1

Holaster sp., Price 1874, p. 140.

? Holaster, Topley 1875, p. 138.

Cardiaster ? sp., ibid., p. 413.

Holaster, large undescribed species, Casey 1949, p. 225.

Holotype. G.S.M. 74118, Lower Greensand (Folkestone Beds) (Lower Albian, tardefurcata Zone,
regularis Subzone), East Cliff, Folkestone, Kent.

Description. Test large (up to 70 mm. long), fairly thin, of heart-shaped outline, widest
at the anterior three-fifths of the length and contracting posteriorly, the length slightly
exceeding the width. Adapical surface elevated, the highest point of the test placed well
forward, at the apical system or just in front of it. From the anterior border the profile
makes a sweeping curve to the summit and thence declines in a gentle curve to the
posterior border, which is truncated either vertically or obliquely downwards and in-
wards. A cross-section of the test through the summit is subtrigonal in outline, with the
sides and base nearly flat, but rounded at the ambitus and the apex. Posterior to the
summit the sides of the test are gently vaulted, the line of their convergence being
marked by a faint carina extending from the centre of the disk to the posterior extremity.
Anterior border indented by the frontal sulcus, which is defined by two blunt carinae
diverging at an angle of 25° from their point of origin at the apical system. Periproct
oval, placed high in the concave posterior face.

Adoral surface more or less flattened, depressed around the peristome, which is very
eccentric anteriorly, transversely oval and without a labral process. Plastron tumid,
increasing in elevation posteriorly, with about seven ill-defined nodules connected into
a zigzag. Nodules also on either side of the periproct and at the base of the posterior
face and a line of them, albeit faint, is traceable on each of the frontal carinae.

Ambulacral areas fairly wide, the paired ambulacra subpetaloid. Near the apex the
ambulacral plates are raised a little at the median vertical sutures, giving a sunken
appearance to the poriferous avenues on that part of the test. The pairs of poriferous
avenues have a simple adoral divergence, and the pores are uniserial, subequal, and
feebly conjugate. Pore-pairs in ambulacra I, II, IV, and V close to the adradial suture

B 6612

S

262

PALAEONTOLOGY, VOLUME 3

d c

text-fig. 1 . Holaster ( Labro taxis ) cantianus sp. nov. a, b, c, diagrammatic representation of specimen
viewed from top, side, and posterior end; based on G.S.M. 74118-21. d, under surface showing plating
of the plastron, with the labrum (1) spanning the oral margin of sternal plate 2' and touching sternal
plate 2; G.S.M. Zm 32. e, apical system; G.S.M. 97304. a-d approximately natural size, ex4.

RAYMOND CASEY: A NEW ECHINOID

263

near the apex, remote at the ambitus, the pores slit-like and nearly horizontal, becoming
rapidly smaller, closer, and circumflex towards the ambitus. Poriferous avenues in
ambulacra II and IV unequal, the anterior series having smaller and more approximated
pores. Poriferous avenues of ambulacrum III lodged in the sides of the sulcus and
adjacent to the adradial suture, the pores minute and oblique.

Apical system elongate, composed of angular plates flush with the surrounding
corona. Oculars II and IV meet in the mid-line and separate the anterior from the
posterior genitals. Oculars II and IV and genitals 1 and 4 each support a tubercle.
Madreporite covers all of genital 2.

Interambulacral areas built up of long curved plates, averaging eleven plates to a
column from apex to ambitus. Plastron composed of a series of alternating cuneiform
plates surmounted by an asymmetric, subtrigonal labrum, conforming to the characters
of the subgenus.

All plates finely granulate, those of the upper surface studded irregularly with very
small tubercles which become larger and more thickly clustered towards the anterior
carinae and around the apical system, and larger still along the sides of the sulcus. A
line of miliaries, sometimes irregular or failing at the ambitus, runs vertically up the
sulcus, parallel and close to each of the inner lines of pores in ambulacrum III. Most of
the lower surface is covered with primary tubercles. On the plastron the tubercles are
crowded and are graduated in size, the smallest at the centre of the plastron, the largest
at the margins. Periplastronal areas devoid of tubercles. The tubercles have crenulated
bosses, perforated summits, and wide scrobicular circles limited by the granules.

Remarks. This echinoid is ubiquitous in the Folkestone Beds of the Folkestone district,
ranging from th e Jacobi Subzone of the nodosocostatum Zone to the mammillatum Zone.
The holotype is one of three specimens found in association in the regularis Subzone,
on which horizon the species reaches its maximum. Specimens used in this account are
listed below under the registration numbers of the Geological Survey Museum:

97304 mammillatum Zone (top stone band), Copt Point, Folkestone, Kent.

74118-20; 74121 tardefurcata Zone ( regularis Subzone, 15-25 ft. below 97304), East Cliff, Folkestone.
Zm 32, 35 tardefurcata Zone ( regularis Subzone, bottom stone band), as before.

97305 tardefurcata Zone (regularis Subzone), Mill Point, Folkestone.

97306 tardefurcata Zone ( trivialis Subzone), Sandling Junction sandpit, Hythe, Kent.

Zm. 522, 653 nodosocostatum Zone ( jacobi Subzone, Red Bed), as before.

Holaster ( Labrotaxis ) latissimus Agassiz is a more orbicular species with rather more
pronounced anterior carinae; the profile lacks the posterior descent of H. ( L .) eantianus ,
tuberculation is not so strong on the apex, and the ambulacral plates are more numerous,
the poriferous avenues generally narrower. Unlike the Cenomanian form figured by
d’Orbigny (1853, pi. 387-8), the English Upper Albian specimens of this species are less
elevated than H. eantianus (cf. Wright 1878, pi. 67, fig. 2b). Holaster ampins d’Orbigny,
described from the Albian of Grandpre (Ardennes), France, and united with H. latissi-
mus by Desor (1858, p. 337), is an orbicular, depressed form with long, sloping posterior
end. The earlier Lower Greensand forms, H. benstedi Forbes (Lower Aptian) and H.
wrighti Lambert (Upper Aptian), like the foreign Aptian H. cordatus Dubois and H.
prestensis Desor, are much smaller and more rotund.

264

PALAEONTOLOGY, VOLUME 3

REFERENCES

casey, r. 1949. Field meeting at Folkestone and Sandling. Proc. Geol. Assoc. Lond. 60,223-5.
desor, e. 1858. Synopsis des echinides fossiles. Paris.

lambert, j. 1893. Etudes morphologiques sur le plastron des Spatangides. Bull. Soc. Sci. hist, et not.
de VYonne , 2e sem. 1892, 55-98.

- — — - 1917. Note sur quelques Holasteridae. Ibid., 2e sem. 1916, 1-33.
orbigny, a. 1853-5. Paleontologie frangaise. Terrains ere faces. 6. Echinodermes. Paris.
price, f. G. h. 1874. On the Lower Greensand and Gault of Folkestone. Proc. Geol. Assoc. Lond. 4,
135-50.

topley, w. 1875. The geology of the Weald. Mem. Geol. Surv.

wright, t. 1 864-82. A monograph on the British fossil Echinodermata from the Cretaceous forma-
tions. Palaeont. Soc. London.

R. CASEY

Geological Survey and Museum,
London, S.W. 7

Manuscript received 20 June 1959

CALATHOSPERMUM FIMBRIATUM SP. NOV.,
A LOWER CARBONIFEROUS PTERIDOSPERM
CUPULE FROM SCOTLAND

by P. D. W. BARNARD

Abstract. A new cupule generically comparable to Calathospermum scoticum Walton is described containing
ovules identical in form to the seeds known as Salpingostoma dasu Gordon. The specimens of this new species
illustrate some interesting morphological features new to the genus. The cupule is believed to be equivalent to a
large part of a frond and its branching may be interpreted as a pinnate system.

The material on which these observations are based consisted originally of a block of
volcanic ash containing petrified plant remains. It was collected by the late Professor
W. T. Gordon. The plants so far described from this source are Tetrastichia bupatides
Gordon (1938), Salpingostoma dasu Gordon (1941), and Eosperma oxroadense Barnard
(1959). The present communication concerns a large cupule belonging to Calathosper-
mum Walton (1940) which is believed to have borne Salpingostoma dasu seeds.

The various fragments of the original block of ash, now broken and cut up, have
yielded more than ten specimens of this new cupule. Three of these were sectioned by
Gordon who prepared thirty-two petrological si ides, some of which were lent to Professor
J. Walton who at the time was describing Calathospermum scoticum. The series, how-
ever, did not permit a detailed description, but Walton noted that Gordon's specimens
differed from C. scoticum in having more divided cupule segments and in having a
different vesture. The remaining specimens have been examined by the peel technique
of Lacey and others (1956). Rock slices of specimens of Salpingostoma dasu not used
by Gordon in preparing petrological sections were also examined by the peel technique.
Some new specimens of S. dasu have also been discovered in the course of these studies.

GENERAL MORPHOLOGY

The cupule is believed to have been borne terminally on the main (primary) rachis
of what was probably a special fertile frond (text-fig. 1). Below the cupule the primary
rachis, which probably exceeded 7 cm. in length, bore two rows of opposite pinnae at
intervals of 10-12 mm. In transverse section the primary rachis is nearly flat adaxially
and strongly convex abaxially and measures 6 mm. in width by 3 mm. in height (text-
figs. 1, 2a-d). The pinnae on the primary rachis are divided near their base into three
portions, possibly the bases of pinnules.

The cupule itself is divided into two halves which may correspond to the two main
arms (secondary rachises) typical of many pteridosperm fronds. The base of the cupule
would thus be equivalent to the region of bifurcation (text-fig. If). Within 3 mm. or
so of its origin the secondary rachis gives rise to a pair of lateral pinnae and ends in a
terminal pinna (text-fig. 1g). These (primary) pinnae form the basic segments of the
cupule.

[Palaeontology, Vol. 3, Part 3, 1960, pp. 265-75, pi. 45.]

P. D. W. BARNARD: CALATHOSPERMUM FIMBRIATUM SP. NOV. 267

Each primary pinna or segment bears in turn about five secondary pinnae alternately;
of these, the two proximal appear to have been fertile, and the three more distal,
vegetative. The rachises of the fertile pinnae may branch dichotomously and may bear
from one to four ovules. In Calathospermum scoticum the seed stalks (fertile rachises)
which arise from a single axis at the base of the cupule were called collectively the
central (stalk) system by Walton. This term is used here for the rachises of the four
fertile pinnae which arise from the bases of the lateral primary pinnae. The marginal
system in C. scoticum is here represented by the remaining fertile pinnae. The sterile
secondary pinnae branch in an alternate manner and end in from three to five or some-
times more ultimate, uninerved, cylindrical processes (pinnules) which I have called
strands (s in text-figs. 1, 3). As shown in the reconstruction (text-fig. 1), some of these
strands exceed 5 cm. in length and have an average diameter of 1 mm. The tip of the
cupule was not preserved in any of the specimens examined so that the exact total
length is uncertain. In all but one of the specimens examined the rachises of both the
central and marginal systems as traced upwards become poorly preserved and shrunken
in appearance and apparently end some 2 to 3 cm. above their origin.

Recognizable attached ovules, two of which are shown in (text-fig. 3c, d), have only
been seen in one specimen (no. 2); this contained at least three and probably four
small, poorly preserved ovules. Two of them belonging to the marginal system are
borne terminally on the undivided rachises of fertile pinnae. One of these (text-fig. 3c)
arises from the abaxial margin of the terminal pinna and the other from a lateral pinna.
The stalk of the third ovule (text-fig. 3d) had its origin on the left abaxial lateral pinna,
and separated from it at a level in between that of the lowest fertile pinna rachis and
that of the first strand (a in text-fig. 3a). It does not appear to have affected either of
the two ‘branches’ between which it is interpolated.

ANATOMY

The epidermis. As seen in surface sections of the rachis the epidermal cells are mainly
rectangular, though some have an inclined end wall, and others are five-sided with a
gable end. When seen in transverse section they appear nearly square, and in radial
longitudinal section, rectangular. The average dimensions are: length, 102 yu.; radial
height, 34/x; tangential width, 38 p. Stomata appear to be present on the primary
rachis (text-fig. 4a, b). They may also have been present on the outside of the cupule
lobes.

In the lowest 3 cm. of the cupule, large epidermal hairs with an average diameter of
45 /j. are found on the inside of the segments and their subdivisions. They are very long,
straight, and without apparent cross walls. I estimate that they were at least 5 mm.
long and may even have exceeded 10 mm.

Fine epidermal hairs, with an average diameter of 13 p, are present around a few of
the seed stalks and also around the ovules in specimen 2 (text-fig. 3b; PI. 45, fig. 5).

text-fig. 1. Calathospermum fimbriatum sp. nov. Reconstruction of the frond with a quarter of the
cupule cut away to reveal the contents ; the fertile rachises, one bearing an ovule shown in section and
another a Salpingostoma seed; together with transverse sections a to J through the regions indicated.
In section h, ms, the fertile rachises of the marginal system (on the left only); the fertile rachises of the
central system are inclosed — cs — . In section I, s, strand.

• J

o

o &
^7 0C> °f

°Xq

lmiiM

P. D. W. BARNARD: CALATHOSPERMUM FIMBRIA TUM SP. NOV. 269

Similar hairs occur in Salpingostoma dasu, where they are found on the body of the
seed and on its stalk, but only for some 5 to 10 mm. below the base of the seed.

The cortex. This is differentiated into two distinct zones, an inner parenchymatous
ground tissue, and an outer fibrous zone or sterome.

The sterome is composed of a number of rows of longitudinally elongate cells of
small diameter and fairly thick walls. The average dimensions are: length, 500 p+;
diameter, 33 p; wall thickness, 4 p. The sterome is broken radially by numerous gaps
with such an irregular disposition that it cannot be called a ‘sparganum’. Some of these
gaps appear to underlie stomata in the epidermis.

The distribution of the fibres is probably related to the mechanics of the cupule: thus
in the primary rachis the sterome is generally better developed abaxially than adaxially;
it is greatly reduced in the conical base of the cupule, but increases again all round after
the primary division into segments, only to decrease again in the pinnules. The sterome
thickens in the base of the seeds as indicated by Gordon.

The ground tissue consists of large empty isodiametric parenchyma cells about 100 p
in diameter, amongst which are scattered slightly larger ‘secretory’ cells with black, often
crystalline contents, and nests of somewhat similar cells which are surrounded by clear
parenchyma cells which are somewhat radially elongate with respect to the centre of the
nest. These nests of dark cells occur very regularly in the adaxial groove in the xylem
of the primary rachis, and give a scalariform appearance to longitudinal sections, but
elsewhere their distribution is more random.

The vascular system. The xylem of the vascular bundles is often extremely well
preserved whilst the phloem cannot be discerned, but it may be represented by part or
all of the cavity surrounding the xylem. The xylem is mesarch and consists of annular
protoxylem tracheids, reticulate to pitted centrifugal tracheids, and pitted centripetal
tracheids (text-fig. 4c and d), with pits about 7 p in diameter, distant and the pit aper-
tures horizontal and opposite. The tracheids range in diameter from 14 p (protoxylem)
to 40 p (metaxylem).

In the primary rachis the bundle was probably a continuous U-shaped strand, though
as seen now it is always irregularly broken especially at the bottom. The xylem contains
from 6 to 8 protoxylem groups, one situated in each of the abaxial ridges of the strand.
The most abaxial pair of protoxylem groups divide just above the ‘pinna node’ and the
new groups so produced depart from the central bundle to form separate traces which
run parallel to the parent bundle until they pass into the pinnae at the next ‘node’.
As they pass out they divide to form two xylems of unequal size, the larger of which
soon divides again.

text-fig. 2. Calathospermum fimbriatum sp. nov. a-k, Series of transverse sections (specimen 1)
showing the origin of the lateral pinnae on the primary rachis, and the basal segmentation of the
cupule. Owing to the slightly oblique nature of the sections the fertile rachises of the central system
depart from the adaxial primary pinnae between I/J and the abaxial primary pinnae between J/K.
A, G.C. 2190; b, G.C. 2196; c, G.C. 2198; d, G.C. 2203 (peel 23); e, G.C. 2207; f, G.C. 2209; g, G.C.
2210; h, G.C. 221 1 ; i, G.C. 2212; J, G.C. 2213; k, G.C. 2214; l, restored outline of the same specimen
to show the position of the petrological sections and the small portion peeled. The letters a to k
indicate the sections illustrated, m to o, basal sections of a series of oblique transverse sections (speci-
men 2) continued in text-fig. 3. In n, cs, fertile rachises of the central system. Owing to the oblique
nature of the section the adaxial rachises are shown above their first division whilst abaxially they are
still attached, m, G.C. 2218; n, G.C. 2221; o, 2223. All x 3, except Lx 1.

P. D. W. BARNARD: CALATHOSPERMUM FIMBRIATUM SP. NOV. 271

As the primary rachis enlarges into the base of the cupule the strand divides to give
two C-shaped portions, each of which then breaks up into three; its further subdivisions
follow those of the cupule. Each division of the xylem occurs at a distance of from 5 to
15 mm. below that at which the individual parts of the cupule they supply become free
from one another. The strands and the fertile rachises contain a single oval xylem bundle
with an excentric mesarch protoxylem. Re-examination of the stalks of Salpingostoma
dasu revealed a single oval xylem bundle in three specimens. This is shown in Gordon’s
pi. 1, fig. 6, but it is rather obscured in that specimen by the pyrite. The single oval
bundle when traced upwards into the seed becomes horseshoe-shaped (Gordon’s term)
before dividing to produce the six oval integumentary bundles.

CORRELATION WITH SALPINGOSTOMA

There are three reasons for relating the seed Salpingostoma dasu to this cupule: the
discovery of small ovules in some of the cupules, the presence of fine hairs in some of the
other cupules, and the close histological similarity.

The best ovule is shown in text-fig. 3d, where one can clearly see the apical processes.
Unfortunately the body of this ovule was lost in the preparation of petrological sections.
Its stalk, surrounded by a belt of fine hairs, is shown in PI. 45, fig. 5.

When traced upwards the seed stalks in the cupules become collapsed and poorly
preserved just before they disappear. The stalks of Salpingostoma seeds fade out similarly
downwards. I believe that rupture of the stalks was not confined to a specialized
abscission zone, but rather that it occurred at any weak point. Most of them appear
to have parted in the hairless region, but an occasional one broke slightly higher so as to
leave part of the hairy region in the cupule.

The approximate dimensions of the ovules is shown in the table alongside those of
C. seoticum and S. dasu after Walton (1949).

Salpingostoma

Calathospermum

dasu

C. fimbriatum

C. seoticum

Number of tentacles

6

6

9+

Length of ovular body

14 mm.

3-5 mm.

3 mm.

„ „ tentacles

25 mm.

4 mm.+

12 mm.

Diameter of ovular body

6 mm.

T2 mm.

T9 mm.

„ „ tentacles

T2 mm.

0-25 mm.

0-21-0-25 mm.

„ „ lagenostome

T6 mm.

0-6 mm.

1-4 mm.

„ „ salpinx

0-4 mm.

0-3 mm.

0-47 mm.

„ „ hairs

1 2-20 /x

10-15 p

12-14 p

That the size differences may be due to differences in maturity was mentioned by
Walton. In this instance it would appear probable that the ovules were abortive and

text-fig. 3. Calathospermum fimbriatum sp. nov. a-g. Continuation of the series of oblique transverse
sections (specimen 2) commencing in text-fig. 2m. a, Showing a, the anomalous departure of the stalk
of an ovule from the left abaxial primary pinna (segment); ms, fertile rachises of the marginal system.
G.C. 2225. b, The first vegetative secondary pinna has now separated from the left abaxial segment.
The ovule stalk surrounded by the interrupted line is shown in PI. 45, fig. 5. G.C. 2229. c. Showing
the body of an ovule of the marginal system sectioned medinally; it is illustrated again in PI. 45, fig. 6.
The left abaxial segment is now represented by three secondary pinnae, from one of these a strand s
has separated. G.C. 2235. d, Petrological section showing the six tentacles and lagenostome of the
ovule with the anomalous stalk attachment. G.C. 2236. e, G.C. 2245. g, G.C. 2248. All X 3.

272

PALAEONTOLOGY, VOLUME 3

text-fig. 4. Calathospermum fimbriatum sp. nov. a. Transverse section of primary rachis showing
epidermis with probable stoma and a portion of the sterome. G.C. 2282 (peel 6), X 250. b, Surface
section of primary rachis showing epidermis with probable stomatal pit. G.C. 2283 (peel 1), X250.
c. Radial longitudinal section of pitted centripetal tracheid. G.C. 2286 (peel 57), X 300. d, Radial
longitudinal section of reticulate to pitted centrifugal tracheids. G.C. 2285 (peel 52), X 300. e. Restored
outline of holotype to show extent of specimen and regions peeled, unpeeled portion stippled; trans-
verse series a, 10 peels; c, 60 peels and d , 88 peels; b , longitudinal series of 100 peels. 1 and 3 indicate
positions of sections illustrated in PI. 45, figs. 1,3, x 1. F, Restored outline of specimen 2 to show
the positions of the petrological sections, regions peeled, and the positions of the sections in text-figs.

2, 3, xl.

that the mature seeds had been shed. A few small detached seeds the same size as the
ovules in the cupule have been discovered in the course of this investigation. It is
possible that these could be the seeds of this cupule and S. dasu the seeds of a different
and possibly larger cupule. One of these new seeds contained microspores (? pollen) in
its lagenostome (PI. 45, fig. 7). Large quantities of similar microspores, many of them

P. D. W. BARNARD: CALATHOSPERMUM FIMBRIATUM SP. NOV. 273

still adhering in tetrads and in larger masses, are present in cupule specimens 2 and 3.
The single seed found containing microspores might, therefore, represent a post-pollina-
tion stage ovule which became detached preventing further development.

The histology of S. dasu differs from that of the cupule in only two minor details.
The xylem in Salpingostoma was only observed to consist of annular and reticulate
(scalariform) tracheids, whereas in the cupule rachis pitted forms were also present.
The megaspore membrane of S. dasu as revealed by maceration of a small portion of
Gordon’s seed no. 8 (PI. 45, fig. 8) is similar to that found in one of the new small seeds,
except that the fibrils of the membrane are more widely separated (expanded) and the
meshes of the cellular reticulum on the outer surface are approximately 350 ju, as com-
pared to 250 /x in the smaller seed.

Genus calathospermum Walton 1940

Emended diagnosis. Cupule borne terminally on a dorsiventral rachis and containing
numerous stalked ovules. Cupule consisting of six main segments which may be simple
or divided. Ovules borne terminally on rachises which arise marginally on the segments
or from the base of the cupule or from both. Ovules similar to the seeds of the type
known as Salpingostoma Gordon 1941.

Calathospermum fimbriatum sp. nov.

Plate 45, figs. 1-7

Cupule about 90 mm. long and 30 mm. broad at its widest part, borne on a primary
rachis 6 by 3 mm. in diameter. Cupule bilaterally symmetrical and divided into equal
halves from just above the base, each half again divided into three segments. Each
segment subdivided into five pinnae, the two basal of which are seed-bearing, and the
three higher sterile. Fertile pinna rachises with from one to four ovules. Sterile pinnae
ending in three or more strands, 1 mm. in diameter.

Material. Holotype specimen no. 5, slide nos. 2279 to 2330; Gordon collection, Geology
Department, King’s College, London: and paratypes slides nos. 2190 to 2385; together
with peel collection and rock specimens from the Green Ash within the Cementstone
Group (Upper Tournaisian) of the Calciferous Sandstone Series (Lower Carboniferous)
at Oxroad Bay, in East Lothian, Scotland.

Discussion. The view adopted here that C. fimbriatum is a modified frond is the one
advanced by Walton in 1 949. This appears fully justified by the morphology and anatomy
of the cupule stalk in C. fimbriatum which is unquestionably a rachis. If only an isolated
portion of such a stalk had been found (e.g. PI. 45, fig. 2) it would have been described
as a rachis and placed in the form genus Lyginorachis.

Some light may be thrown on the question as to whether the stalk is a primary or
secondary rachis by comparison with Lyginopteris oldhamia (Binney) as described by
C. Louvel (1959), and Tetrastichia bupatides Gordon (1938). In L. oldhamia the primary
rachis (petiole) bears opposite pinnae and above the bifurcation the secondary rachises
bear alternate pinnae. The primary rachis contains a W-shaped xylem strand formed
by the partial fusion of two V-shaped traces which supply the secondary rachises. In
T. bupatides there are no pinnae on the petiole (primary rachis); on the secondary

274

PALAEONTOLOGY, VOLUME 3

rachises the pinnae are alternate (original observation). In the petiole the xylem strand
is butterfly shaped (a modified W) and separates into two modified V-shaped traces at
the bifurcation.

In comparing the two species of Calathospermum a number of differences become
apparent. The cupule segments in C. scoticum are not pinnately divided yet they contain
six xylem strands. The seed-bearing branches of the central system in C. scoticum arise
from a central stalk which appears as a direct continuation of the primary rachis (cupule
stalk), and the marginal system is more extensively developed in that twelve pinnae con-
tribute to it as opposed to the eight in C. fimbriatum. The four extra pinnae in C.
scoticum arise from the lateral pinnae (cupule segments). In C. fimbriatum (specimen 2)
the ovule which arises in an anomalous position appears to be a homologue of one of
the extra marginal pinnae in C. scoticum.

The two cupules also differ markedly in their relative proportions. The cupule of
C. fimbriatum is twice as long as that of C. scoticum and the diameters of their stalks
are in the same proportion. However, the diameters of the two cupules are about equal.
The fertile rachises in C. fimbriatum arise over a distance of about 9 mm., whereas in
C. scoticum they all arise within 3 mm. of one another. These proportional differences
suggest the following explanation for some of the morphological differences: that the
number of fertile pinnae per cupule is determined by some factor correlated with growth
in diameter, and that the basal origin of the central system in C. scoticum is due to a
telescoping of the region over which the fertile pinnae arise.

Walton (1949) has compared C. scoticum and Dip/opteridium tielianum (Kidston)
Walton. It is interesting to note that the frond of C. fimbriatum appears to be inter-
mediate between D. tielianum and C. scoticum in having pinnate cupule segments. The
fact that C. fimbriatum is basically pinnately branched contrasts strongly withD. tielianum
where the fertile rachises arise as a direct continuation of the primary rachis.

Walton (1949), in a very full discussion, compared Calathospermum with Gnetopsis
elliptica Renault, Megatheca tliomasii Andrews, Calathiops renieri Walton, and Cala-
thiops bernhardti Benson. The first three of these species are only known as detached

EXPLANATION OF PLATE 45

Figs. 1-7. Calathospermum fimbriatum sp. nov. 1, Transverse section of primary rachis containing two
dividing C-shaped traces from the extreme base of the cupule; holotype, slide no. G.C. 2296 (peel 3),
X 13-5. 2, Transverse section (oblique 30°) of a small detached rachis ( Lyginorachis sp.?) identical
in structure to a primary rachis of C. fimbriatum and showing the characteristic broken dentate
U-shaped xylem and the bases of two pinnae; G.C. 2378, X 13-5. 3, Transverse section through the
holotype showing eight rachises of the central system, together with six rachises of the marginal
system surrounded by the cupule segments in various stages of division (compare with text-fig. 1h
and i); G.C. 2324, x3-5. 4, Oblique transverse section showing six segments and centrally two
fertile pinna rachises; specimen 6, G.C. 2342, x3-5. 5, Transverse section of the stalk of an ovule
surrounded by hairs (text-fig. 3b); G.C. 2229, x36. 6, Transverse section through the body of an
ovule (text-fig. 3c), showing the crumpled megaspore membrane, superficial layers of integument
with five grooves and surrounding hairs; G.C. 2234, x 38. 7, Microspore (?pollen) from lagenostome
of small detached ovule; G.C. 2384 (peel 10), X 666.

Fig. 8. Salpingostoma ciasu, portion of megaspore membrane showing the cellular reticulum on the
outer surface; G.C. 2383, X21.

Figs. 9, 10. Calathiops sp. W. Hemingway’s photographs of a specimen from Coseley. 9, Specimen in
nodule, xl. 10, Counterpart, x 3.

Palaeontology, Vol. 3.

PLATE 45

BARNARD, Calathospermum

P. D. W. BARNARD: CALATHOSPERMUM FIMBRIATUM SP. NOV.

275

cupules; in the fourth, Calathiops bernhardti Benson (1935), the cupules are secund
on pinnae which are borne alternately on a major rachis. This is a very distinct type,
and although the cupules appear similar to C. fimbriatum the frond bearing them was
very different and probably much larger.

In the M. Benson manuscripts in the British Museum (Natural History) there are
some photographs from W. Hemingway, two of which illustrate a compression (speci-
men WH/2356) of what appears to have been a cupule bearing two (? four) ovules.
Though it has not been possible to locate the specimen, these photographs are repro-
duced here as PI. 45, figs. 9, 10. The specimen was from Coseley, near Dudley, south
Staffordshire (Westphalian B) and was referred by Hemingway to the genus Gnetopsis.
The cupule is highly divided and ended in at least ten strands though its basal segmenta-
tion is not clear. It closely resembles the cupules of Calathiops bernhardti.

More recently Walton (1953) has indicated some features of anatomy and symmetry
common to Calathospermum and some genera of the Trigonocarpales. In the Neuro-
pterids which supposedly bear Trigonocarpalean seeds; the seeds are attached to pinna
rachises and appear to be morphologically equivalent to pinnules as they are in Cala-
thospermum. The cupule of Calathospermum appears thus morphologically comparable
to the vegetative part of a Neuropterid frond and not to the integument of the Trigono-
carpalean seed. In Calathiops bernhardti noted above the cupules seem to be comparable
to pinnae. A possible explanation is that a transference of cupule function (see Corner
1958) may have occurred so that morphologically smaller regions of a frond became
involved.

Acknowledgements. I am indebted to Professor J. H. Taylor for the loan of slides and cut blocks from
the Gordon collection, to Professor J. Walton for help and advice, to Dr. K. L. Alvin for helpful
criticism of the manuscript, to Mr. McGregor for photographic assistance, and to the University of
London together with the Clothworkers’ Company for the William Gilles Research Fellowship
during my tenure of which this work was accomplished.

REFERENCES

Barnard, p. d. w. 1959. On Eosperma oxroadense gen. et sp. nov., a new Lower Carboniferous seed
from East Lothian. Ann. Bot. n.s., 23, 284-96, pi. 1.
benson, M. 1935. The fructification, Calathiops Bernhardti n. sp. Ann. Bot. 49, 155-60, pi. 5.
corner, E. J. h. 1958. Transference of function. J. Linn. Soc. {Bot.), 56, 33-40.

Gordon, w. t. 1938. On Tetrastichia bnpatides: a Carboniferous pteridosperm from East Lothian.
Trans. Roy. Soc. Edinb. 59, 351-70, pi. 1-6.

1941. On Salpingostoma dasu: a new Carboniferous seed from East Lothian. Ibid. 60, 427-64,

pi. 1-6.

lacey, w. s. and others. 1956. A rapid cellulose peel technique in palaeobotany. Ann. Bot. n.s., 20,
635-7.

louvel, c. 1958. Evolution du systeme libero-ligneux dans les petioles de Lyginopteris oldhamia
(Binney). C.R. Acad. Sci. Paris, 247, 2411-13.
walton, J. 1940. Introduction to the Study of Fossil Plants. London.

1949. Calathospermum scoticum — an ovuliferous fructification of Lower Carboniferous age from

Dumbartonshire. Trans. Roy. Soc. Edinb., 61, 719-28, pi. 1-3.

1953. Evolution of the ovule in the pteridosperms. Advanc. Sci. Lond. 10, 223-30.

P. D. W. BARNARD

Department of Botany,
Birkbeck College,
London, W.C. 1

Manuscript received 8 October 1959

THE EXTERNAL ANATOMY OF SOME
CARBONIFEROUS ‘SCORPIONS’
PART 2

by LEONARD J. WILLS

Abstract. Part 2 is concerned with the anatomy of eight Orthostern ‘scorpions’, developed by the technique
described in Part 1 ( Palaeontology , 1, 261-82). Virtually complete skins of two are described. The first is a
paratype of Buthiscorpius buthiformis (Pocock), the description of which is supplemented from a second, less
complete, example. The second is Mazoniscorpio mazonensis gen. et sp. nov. Each of the remaining five is in-
complete as only half-nodules were available. They comprise a new species of Buthiscorpius — B. major, a new
genus and species — Wattisonia coseleyensis; and three unidentifiable forms. Each provides valuable data about
one or more organs : yet there is still no absolutely convincing evidence as to how any of them breathed. There
follows a revised diagnosis and description of the Lobostern Eoscorpius tuberculatus Peach, here made the type
species of Benniescorpio gen. nov. The paper ends with a brief discussion of the anatomical and ecological
conclusions to be inferred from the descriptions in both Parts, followed by four supplementary notes correcting
statements in Part 1 and describing new development-techniques.

INTRODUCTION

The majority of Carboniferous ‘scorpions’ have been placed in Pocock’s group Ortho-
sterni (see Part 1, p. 267), because a few among them are known to resemble Recent
forms in having unlobed parallel-sided sternites. I propose describing here details of the
anatomy of eight Orthostern ‘scorpions’ etched out of ironstone nodules. After the
descriptive part, a few general conclusions are stated, but I have made no attempt to
revise our knowledge of Carboniferous ‘scorpions’ in general or to make a critical
review of their taxonomy. In fact this last problem can never be successfully achieved
until the type specimens have been developed to show details analogous to those
achieved by the present etching technique.

I have throughout numbered the segments according to their position on the adult
animal: Prosoma, carapace i to vi; Abdomen, mesosomatic vn to xn plus metasomatic
xm ; Tail, metasomatic xiv to xvm plus the sting. According to this scheme the genital
operculum lies on the first adult mesosomatic segment (vn); whereas it is generally
stated by zoologists to lie on the second mesosomatic (viiith) somite of the whole body,
since the first (vnth) or pregenital disappears in embryonic growth ‘retaining in the adult
only its neuromere, which becomes incorporated in the thoracic ganglionic mass as the
6th pair of its ganglia’. . . . ‘The original number (12) of abdominal segments is restored
in the course of embryological development by sub-segmentation of the 8th embryonic
segment’ (Petrunkevitch 1955, p. P68). Presumably Carboniferous ‘scorpions’ behaved
in the same way, but there is, of course, no evidence. See also Stormer 1955, pp. PI, P6,
re segmentation in Chelicerata in general and Merostomata in particular.

Repositories. B.M., British Museum (Natural History); B.U., Birmingham University, Geology
Department; G.S.M., Geological Survey Museum, London; G.S.E., Geological Survey, Edinburgh;
M.M., Manchester Museum.

[Palaeontology, Vol. 3, Part 3, 1960, pp. 276-332, pis. 46-57.]

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’

277

List of abbreviations used in the illustrations, act, anterior claw; acs, aculeus of sting; app, anterior plate
of sternum of pecten; ats, anterior tarsal spur; ap, anterior process of carapace; avn, anterior V-notch
of sternite; b, boss on rachis of pecten; belt, basal joint of chelicera; bo, border; C xiv-xvm, caudal
rings of adult segments xiv-xviii; c 1-4, coxae of legs 1-4; ca, carapace; ch, chelicera; cl, claw; ell,
claw-lobe; cp, coxa of pedipalp; cr, cephalic region of carapace; d, dagger; ebc, end of broken claw;
et, eye tubercle; /, fulcra; f 1— 4, femur of legs 1-4; fp, femur of pedipalp; frf free fingers; go, genital
operculum; gs, pad or Gehstachel ; gr, granule; h, hair; hch, hand of chelicera; hpd, hand of pedipalp;
L, left; L.L. 1-4, left legs 1-4; hn, lamella of pecten; lo, lobe on metatarsus; md, mandibular process
of coxa 1 or 2; me, median eye; mg, median groove of carapace; mt, metatarsus; mts, metatarsal spur
(arising from base of metatarsus); pa, patella; pa 1-4, patella of legs 1-4; pap, patella of pedipalp; peg,
posterior cephalic groove; pel, posterior claw; pd, pedipalp; pe, pecten; pgs, poison-gland of sting;
ppp, posterior plate of sternum of pecten; ps, platform spine; pts, posterior tarsal spur; pvn, posterior
V-notch of sternite; R, right; Rch, right chelicera; R.L. 1-4, right legs 1-4; ra, rachis of pecten; S ix-
xii, sternites of adult segments ix-xii; Sxm, sternal plate of adult segment xm; se, spinule on doublure
of sternite; set, seta, bristle, movable hair; sf, sensory field; sg, sting; sk, stop-knob; spe, sternum of
pecten; spi, spine; st, sternum of prosoma; Tvii-xii, tergites of adult segments vh-xii; Txm, tergal
plate of adult segment xm; ta, tarsus; tas, tarsal spur; th, teeth of pecten; ti, tibia; tr, thoracic region
of carapace; tr 1-4, trochanter of legs 1-4; tri, sensory hair-bases (trichobothria); trp, trochanter of
pedipalp.

SECTION A — ALMOST COMPLETE EXOSKELETONS

ORTHOSTERNI PoCOCk 1911
buthiscorpius Petrunkevitch 1953
Buthiscorpius buthiformis (Pocock)

Plates 46-48; text-figs. 1-9

Anthracoscorpius buthiformis Pocock 1911, pp. 24-28, pi. 1, fig. 2, pi. 2. fig. 1, text-figs. 6-8.

Eoscorpius buthiformis Petrunkevitch 1913, p. 35.

Eoscorpius buthiformis Petrunkevitch 1949, pp. 152-3.

Buthiscorpius buthiformis Petrunkevitch 1953, pp. 26, 32, figs. 31-34, 122.

Material, (i) Holotype, B.M. In. 18596, Middle Coal Measures, Sparth Bottoms, Rochdale, Lancs.
Pocock 1911, pi. 11, fig. 1, text-fig. 6; Petrunkevitch 1953, figs. 31-33, 122. (ii) Paratypes in British
Museum, all from Coal Measures, Coseley, S. Staffs.: (1) In. 31262, Pocock 191 1, pi. 1, fig. 2 a (here
redescribed). (2) In. 22832, Pocock 1911, text-fig. 8. (3) In.1555, Pocock 1911, pi. 1, fig. 2. (4, 5) Two
specimens in Mr. Egginton’s collection cited by Pocock 1911, p. 27 (present whereabouts unknown).
One specimen, B.M. In. 7883, figured as a paratype by Pocock (1911, text-fig. 7) has been made into
the holotype of Compsoscorpius e/egans Petr, (iii) Other material: B.U.720 (here described).

Remarks. In 1913, and again in 1949, Petrunkevitch assigned Anthracoscorpio buthiformis Pocock to
the genus Eoscorpius Meek and Worthen, but in 1953 he erected Buthiscorpius for Pocock’s holotype
and all but one of his paratypes. Dr. E. I. White, with the sanction of the Trustees of the British
Museum, has allowed me to develop both halves of one of the paratypes (B.M. In.3 1 262) with the
primary aim of discovering how a typical Orthostern ‘scorpion’ breathed. Unfortunately only negative
evidence on this point has emerged. Each half is now encased in a transparent block of Marco, save
for the two pedipalp hands, fragments of legs and comb, and part of the sting which are mounted as
micro-slides.

The results of the development confirm Pocock’s view that this specimen is conspecific with the holo-
type of his ‘ Anthracoscorpio ’ buthiformis. As regards the generic identification it is best, in my opinion,
to use Petrunkevitch’s genus Buthiscorpius at present, pending a revision of the generic diagnoses of
Anthracoscorpio Kusta (1884) and Eoscorpius Meek and Worthen (1868), on which also awaits the
diagnosis of the Family Eoscorpioniidae to which undoubtedly all three genera belong. It may well
prove impossible to distinguish the three genera, in which case Eoscorpius has priority over the other
names.

B 6612

T

278

PALAEONTOLOGY, VOLUME 3

Revised diagnosis, based on B.M. In.31262 and B.U.720. To the original diagnosis
(Pocock 1911, p. 24) with additions by Petrunkevitch (1953, p. 32) can be added: Last
caudal ring very long (one-third of whole metasoma), poison capsule rather small and
probably somewhat shorter than 6th caudal ring. Sternum and coxae of the legs as
described by Pocock in B.M. 1.1555, with coxae 3 and 4 fused throughout the length of
coxa 3. Genital operculum small. All legs with two claws and two tarsal spurs, and 3rd
and 4th legs each with one large metatarsal spur. Pecten probably small. Sternites ix to
xii unlobed, overlapping one another briefly and devoid of pulmonary stigmata.

Description of paratype, B.M. In.31262 (PL 46, 47; text-figs. 1-5)

Remarks. This specimen was collected by Egginton from the Coal Measures at Coseley, South
Staffordshire. It was figured by Pocock (1911, pi. 1, fig. 2a) as a paratype of A. buthiformis, the figure
incorporating data from both halves of the nodule and showing the chelicera more clearly than does
the specimen.

As received the specimen lay in two halves of a nodule. One, now numbered In. 31262a (and here-
after referred to as A), showed dorsal features seen from the ventral side with Right side elements
appearing on the left (PI. 46, fig. 1) — Pocock’s pi. 1, fig. 2a, is reversed. On A a thin sliver of stone
defined by a crack passing almost parallel to the surface and visible in the photo had been stuck on
with shellac. I dissolved this and repaired with ‘Seccotine’. The crack had severed the Right appendages
which consequently broke away during the etching. The other half nodule, In.31262B, showed ventral
features seen in dorsal aspect (PI. 47, fig. I). In both halves a good deal of very tenuous chitin was
preserved, but in some places the surfaces were casts of the outside of the skin. This arrangement
implies that when the nodule was split open the fracture passed between the dorsal and ventral skins,
and that any intervening film of matrix was lost, for the dorsal features of A do not appear on B, and
vice versa; e.g. the carapace is seen on A and the sternum and coxae on B.

There was no kaolinite reinforcement and only a very little iron pyrites, and the chitin proved to be
very thin, transparent, and fragile. Consequently the mounted preparations are disappointingly frag-
mentary except in one or two cases.

After etching, the exterior of the dorsal surface (minus the sting) was displayed, the skin being pre-
served in places where it had escaped destruction on the original surface of fracture of the nodule
(PI. 46, fig. 2).

Approximate measurements in mm. Length of carapace, 3-5; abdomen, 8-5; tail, 11; sting, c. 4;
total, 27.

The carapace is damaged at the forward end, but it can be seen that the sides converge
slightly towards the front where the antero-lateral corners are rounded and united by a

EXPLANATION OF PLATE 46

Figs. 1-5. Buthiscorpius buthiformis (Pocock), B.M. In.31262A and B. 1, Ventral view of the dorsal
surface of A, as received. X 6. 2, Dorsal view of the dorsal surface of A, after development. X 6.

3, Dorsal view of the sting, Slide B.M. In.31262A/2, at about the same magnification as fig. 2. X7-5.

4, Ditto, x 22-5 to compare with text-fig. 2. 5, Tip of the rachis of the comb with three teeth, each
with a narrow sensory field. Slide B.M. In.31262B/ll. X 60.

EXPLANATION OF PLATE 47

Figs. 1-6. Buthiscorpius buthiformis (Pocock), B.M. In.31262A, B. 1, Dorsal view of the ventral surface
of specimen B, as received. X 6. 2, Ventral view of the ventral surface of specimen B after develop-
ment. x 6. 3, Femur, patella, and hand of R. pedipalp. Slide A/1. Xl5. 4, Tibia, metatarsus,
tarsus, and claws of 1st or 2nd L. leg. Slide B/4. X 15. 5, Part of metatarsus, tarsus, and claws of
1st or 2nd R. leg. Slide B/l. X 15. 6, Metatarsus, tarsus, and one surviving claw of ? 3rd L. leg.
Slide B/3. Xl5.

For key to abbreviations see p. 277.

Palaeontology, Vol. 3.

PLATE 46

WILLS, Buthiscorpius

Palaeontology , Vol. 3.

PLATE 47

WILLS , Buthiscorpius

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’

279

curved anterior margin. The shield is somewhat arched, with a shallow median furrow
that is blocked in front by the ocular tubercle which is situated about one-quarter of the
length of the carapace from the front. The eyes are conspicuous (text-fig. 1) : each appears
to be a hemisphere with its polar axis pointing almost vertically. As Petrunkevitch
(1953, p. 32) states, ‘the space between the eyes is elevated’. This elevation is defined by
a pair of small ridges separated by a median groove. On the right side of the carapace

text-fig. 1. Buthiscorpius buthiformis (Pocock). B.M. In.31262A, after development. Front half of
carapace to show the eye-tubercle with large median eyes separated by two diverging ridges and a
narrow median groove. Behind the tubercle is the front end of the median furrow, x c. 10.

text-fig. 2. Buthiscorpius buthiformis (Pocock). B.M. In.31262A. Dorsal view of the metasomatic
segments and sting. For key to abbreviations see p. 277.

the antero-lateral corner is fairly well preserved and shows no lateral eye ridges or lateral
ocelli. The posterior margin is mucronate and passes into linear lateral margins. No
ornamentation can be seen.

The mesosomatic tergites are quite normal, being very short in the front of the body
and increasing in length so that tergite xn is at least twice as long as tergite vn. The
width increases from tergite vn to tergite x, and then remains almost constant back to
the articulation of tergite xn with the tergal plate xm. The posterior margins are defined
by lines of granular and mucronate tubercles, the anterior margins are linear. Wide
strips of intersegmental skin are exposed, showing that the animal was entombed in a
distended state.

The metasomatic segments. The dorsal surfaces of all the metasomatic segments
etched out perfectly (PI. 46, fig. 2), the end of the tail floating free in the acid (it now lies
free in the Marco block, text-fig. 2); but only certain parts of the ventral surface of the
tail can be examined, for it proved impossible to remove all the matrix from below the
free-floating end in A, and only the skin of the 1st and of parts of the 2nd and 3rd
metasomatic segments is preserved in B (text-fig. 2). The shapes and relative proportions

280

PALAEONTOLOGY, VOLUME 3

text-fig. 3. Buthiscorpius buthiformis (Pocock). B.M. In.31262B/2. The poison-capsule with the
missing aculeus restored (broken line), a, Dorsal view of dorsal skin (stippled) and of inverted proximal
process of the ventral skin (unstippled) with semi-annular thickening, tk, and ? periproctal skin, ps;
transverse fold, tr; external and internal fold-crests, efc, ifc. b, Ventral view of ventral skin (stippled)
with its proximal process (unstippled) restored to its original position, c. Sagittal section of specimen
with its proximal part (broken line) restored as in b. d, Tentative restoration of the poison-capsule in
relation to caudal ring 5 and the anus, in sagittal section.

of the metasomatic segments are fairly normal, the first four all having much the same
length, the fifth being somewhat longer and the sixth (perhaps abnormally long) being
about three times the length of the first. The first (Txin) has the usual sharp taper
backwards from the full width of the body to the narrow width of the caudal rings (the
latter have been flattened and appear wider than they originally were). The whole seg-
ment consists of a dorsal and a ventral plate united by a wedge of pleural skin. The two
plates taper sharply backwards to articulate with the first caudal ring. With the wedge

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’ 281

of pleural skin they resemble a pair of old-style bellows. The sides of each plate are
defined by marginal ridges which unite in blunt spines at the apex of each pleural
wedge. On the dorsal plate there are also two sharp keels ending in formidable thorns,
and a strong median spine. Both plates have linear anterior margins, that on the ventral
plate lying much farther back than that on the dorsal, and with a correspondingly
broader anterior border of intersegmental skin (possibly an adaption to the habit of
inverting the tail over the animal’s back).

Caudal rings xiv-xvi each have a pair of dorsal keels armed with spinelets and ending
in a large thorn like those on Txm. In Cxvn and Cxvm the corresponding keels are
wider, flatter, and free from spinelets and thorns.

Owing to flattening and to the obscuring by matrix, details of the sides and ventral
surfaces of the caudal rings cannot be completely elucidated, but probably ventro-
lateral keels marked the limits of the lower surface of each ring. The keels end in rather
flat angular processes which are progressively larger as the end of the tail is approached,
the pair on the last caudal ring being conspicuous and separated by a sinus, behind
which the anus would lie.

The sting. The poison capsule lay symmetrically inverted over Cxviii. Unfortunately
the aculeus itself was missing, probably as a result of my having ground away the back
of the nodule a tiny fraction too much, the cut removing the dorsally projecting point
of the sting. The rest of it, about 3 mm. long, etched out complete (PI. 46, figs. 3, 4;
text-fig. 3) as an almost cylindrical object. It is not absolutely certain which of its sur-
faces is the ventral one that, being inverted over Cxviii, was first laid bare by the etch,
but the supposed dorsal side lies upwards on Slide A/2. Being quite transparent it can
be examined from both sides and by reflected or transmitted light. The originally
flask-shaped capsule was not flattened in fossilization, but was partially collapsed in a
remarkably symmetrical way by the flexing of its curved sides inwards in two large and
two small longitudinal folds, and transversely by one deep fold on the dorsal, and
by two lesser ones on the ventral side; and the median proximal process on the ventral
side was inverted so that its end pointed distally, text-fig. 3a, b, c. This process ends in
a triangular piece that continues into thin plates which may be remains of the skin and
muscles lying near the anus and connecting the sting to the last caudal ring.

Taking account of the inversion of the process, and the positions and sizes of the
infolds (which can be traced out within the transparent sclerite), it seems likely that the
capsule bulged dorsally and laterally near its proximal end, and that it had shallow
lateral grooves (as occur in some Recent forms) which determined the position of the
two conspicuous deep longitudinal infolds. On text-fig. 3d I have attempted a reconstruc-
tion of the actual sting as it would have appeared in sagittal section.

Organs of the ventral surface (text-fig. 4)

The original fracture that split open the nodule passed through the chelicerae and the
basal parts of the pedipalp, and then between the carapace and the inside of the coxae of
the legs and the sternum. In the mesosoma it passed between the tergites and the ventral
organs (genital operculum, the sternum of the pecten and the four sternites). As a result
the chelicerae and the coxae of the pedipalp were much damaged, and most of the two
combs were lost. The etching of specimen B, however, revealed the outside or ventral
surface of the coxo-sternal area and of the mesosomatic ventral organs in fair detail.

282

PALAEONTOLOGY, VOLUME 3

Chelicera. Only part of the hand of the chelicera shows in specimen A (PI. 46, fig. 2,
ch). It appears to be of normal proportions.

PedipaJp. Apart from the coxae, both pedipalps are fairly well displayed, and the
hands were both extracted (Slide A/1, PI. 47, fig. 3, and Slide B/2). The R. coxa is prob-
ably visible where it projects beyond the side of the carapace in A; the left one is seen in
B to lie dorsally to the coxa of the 1st leg, but its shape is difficult to make out. The
trochanter, femur, and patella are quite normal, the L. femur showing the row of granules
noted by Pocock — ‘the femur of the left chela with an anterior granular crest, such as is
present in most recent scorpions’. The whole of the R. patella (pap) is preserved in
Slide A/1 (PI. 47, fig. 3). It shows a ‘stop knob’ (see p. 304) on its anterior side and

text-fig. 4. Buthiscorpius buthiformis (Pocock). B.M. In.31262B. Ventral surface of body after de-
velopment. Cf. PI. 47, fig. 2. For key to abbreviations see p. 277.

text-fig. 5. Buthiscorpius buthiformis (Pocock). B.M. In.31262B/4. Claws and end of tarsus (in pos-
terior view) of ? 2nd L. leg. Cf. PI. 46, fig. 11. X c. 60. For key to abbreviations see p. 211 .

traces of small granules. Owing to crushing the original shape of the hand (hpd) is hard
to determine, but the palm must have been stouter than the present outline would
suggest. The fingers are somewhat longer than the palm and they taper gradually, to
end in points slightly hooked towards each other. The biting edges are a little thickened,
but devoid of granules. Under a high power of the microscope minute hair-bases, some
perhaps trichobothria, can be seen near the tips and elsewhere.

In general the whole hand resembles that extracted from B. major sp. nov. (G.S.M.
Za. 2926, PI. 52, figs. 1, 2); but is relatively more massive and with shorter fingers. The
whole pedipalp was large and massive in proportion to the legs.

Sternum of the prosoma (st) is six-sided, but is best described as pentagonal with the
posterior side deeply excised, as a result of which the sclerite resembles a bluntly tanged
arrowhead. Near the apex of the excavation there is a deep pit with its bottom pointing
backwards. The sternum therefore agrees closely with Pocock’s drawing (his pi. 1, fig. 2)
of the ventral aspect of the coxo-sternal area of one of his paratypes (B.M. 1.1555).
Along the left side coxae 3 and 4 (c 3, 4) can be seen to abut against it, coxa 4 being
much longer than coxa 3, and stretching back almost to the 2nd sternite. The two seem
to be fused throughout the part in which they are adjacent to one another, as are the

FIG. 4

fig. 5

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’ 283

corresponding articles in Recent scorpions. This is an important observation, confirmed
by what is seen in B.U. 720 (p. 288), because it negatives one of Petrunkevitch’s tentative
diagnostic characteristics of Eoscorpiidae — coxae of 3rd and 4th pairs of legs ‘probably
not yet grown together along their line of contact, retaining independent motion’ (1955,
p. P73), and shows that in the present genus at any rate the coxal arrangement was in
this respect the same as it is today. Coxae 1 and 2 also conform exactly to the pattern
found in Recent scorpions. The mandibular processes are well displayed, those on coxa 2
being the larger and lying side by side on the middle line. The other articles of the four
legs are minute and their skin is so tenuous and devoid of any reinforcement that claws,
spurs, and even individual joints were inevitably detached during the etching. Though
several parts were recovered and mounted as Slides A/3-A/8 and B/l, B/3-B/8, it was
impossible in some cases to be certain about their individual provenance, nor can one be
certain that the surviving claws and spurs represent the original full complement. For
these reasons only approximate estimates of the lengths of the appendages can be made.
See Table, p. 289. The parts recovered (a few of which are illustrated on PI. 47, figs.
4-6) do, however, compare so closely in shape or structure with the better preserved
examples in B.U. 720 (PI. 48, fig. 6; text-fig. 6) that it may be safely assumed that these
limbs were organized on the same pattern. The detailed structure of the claws with their
pad ( Gehstachel ) can be seen in Slide B/4 (text-fig. 5). No denticles can be seen on the
claws, a feature possibly related to the small size of the individual.

Mesosoma. The genital operculum (go) is small, its two halves filling the posterior sinus
of the sternum. The left half has been pushed under the right half far enough to damage
and displace the median edges of both.

The sternum of the pecten ( spe ) appears to be short and possibly bilobed, but no details
are visible. Remains of both combs were seen, but only a few teeth were recovered
(Slides 10-12, PI. 46, fig. 5). Each is a flattened sack, pointed at its proximal end and
rounded distally, with a narrow sensory field thickly covered by peg-organs (Part 1,
p. 274). The field occupies a lanceolate strip running from end to end. This arrangement
is probably the normal one in Carboniferous ‘scorpions’, but the example here photo-
graphed reveals it more clearly than any of the other specimens so far found. What I
saw of the combs suggested that each was short, with teeth few in number and attached
to a narrow slight rachis — the whole resembling the comb of a Recent scorpion more
than the fan-like ones found in Lichnophthalmus, Pareobuthus (Part 1), and Bennie-
scorpio tuber culatus (Peach) (below).

The sternites of adult segments ix-xii, exposed on the original ventral half nodule (B)
but not noticed by Pocock, revealed disappointingly meagre details after etching, prob-
ably because the ventral skin was extremely thin, crushed against the dorsal, and
damaged when the nodule was split open. Sternite ix was better seen on the surface
originally exposed (PI. 47, fig. 1) than it is now after etching. It is short and ill-defined in
front where it may have merged into the sternum of the pecten with but little marginal
thickening of either sclerite. The others are much of a size, each about one-third longer
than sternite ix. Each agrees roughly in length with its corresponding tergite. Traces of
linear anterior margins can be seen here and there. The posterior margins are un-
ornamented, but appear dark, mainly as the result of the infolding of the intersegmental
skin in a deep doublure. It is this belt of overlap that showed up as a broad stripe

284

PALAEONTOLOGY, VOLUME 3

between adjacent sternites on the internal surface as originally exposed (PI. 47, fig. 1).
The postero-lateral angles of sternites x-xii are rounded and clearly overhang (in ventral
aspect) the pleural skin which here merges with the infolded intersegmental skin. There
is no sign of any stigmata on any of the sternites or on the pleural or infolded inter-
segmental skin; nor can I detect any minute hairs or spinules on the infolded skin such
as occur in Pareobuthus and Lichnophthalmus (Part 1). The depth of the intersegmental
infold where one sternite overlaps the next one behind, although about one-third of the
length of the sternite, would appear to be too small for an adequate cover to gills : how
this animal breathed remains a mystery.

Description of B.U.720 (PI. 48; text-figs. 6-9)

Remarks. This specimen appears to resemble the paratype of B. buthiformis just described (B.M.
In.31262) so closely that it may confidently be referred to the same species. In particular it agrees in
regard to dimensions and proportions, the shape of the carapace, the position of the median eyes, the
apparent absence of lateral eyes, and in the organization of the coxo-sternal area. However, the front
end of the carapace is not well preserved, and there are features towards its antero-lateral corners that
in some lightings might be interpreted as lateral eyes. Were these certainly present the specimen would
have to be referred to Compsoscorpius Petr., one species of which, C. elegans Petr., is based on another
of Pocock’s paratypes of Anthracoscorpio buthiformis (B.M. In. 7883). The Birmingham specimen is
particularly instructive because its legs were extracted almost intact with spurs, bristles, thorns, and
spines still attached — features used extensively in the classification of Recent scorpions.

This was the first specimen to which, with Dr. Isles Strachan's help, I applied the embedding and
hot-etching technique described in Part 1. The fossil, preserved in one half of a small ironstone nodule
of the ‘pennystone’ type, was found by me many years ago amongst duplicate material of unknown
origin in the Geology Department's possession. It had been almost certainly obtained from the Coal
Measures of South Staffordshire, probably from above the Thick Coal at Coseley. Impressions of the
dorsal surface as originally exposed were made in collodion, plastone, and dental wax, and it was
photographed some years ago (PI. 48, figs. 1, 2). As it was seen that bits of the original chitinous skin
were preserved, I attempted to embed it in balsam with a view to etching with HF, but I was not
satisfied and I dissolved off the balsam. Nothing further was done until 1956 when we took advantage
of the development of the new transparent plastics to embed it in Marco and etched it with hot HC1.
As a result, the following organs (some in a broken state) were isolated, and mounted: the sternites,
the sternum, and attached coxae of the pedipalp and legs, the trochanter and hand (minus the movable
finger) of the pedipalp, the remaining joints of three left legs, and parts of the right legs. In this instance,
unlike the preservation in B.M. In.31262, the chitinous skin in some parts had been covered with, and
cavities had been filled by, kaolin which functioned to keep the original shape and relief of the various
parts even after they had been extracted. Embedded in the kaolin, however, were many bristles which
have the appearance of having been torn away from the skin, as described and figured in Part 1 (p. 263 ;
pi. 50, fig. 16), the figure being a photo of the metatarsus of the 3rd leg of the present specimen. At the
end of the etching a very good transparent cast was obtained of the originally exposed surface, with a
fair amount of chitin fragments embedded in it (PI. 48, fig. 3).

EXPLANATION OF PLATE 48

Figs. 1-7. Buthiscorpius buthiformis (Pocock), B.U.720. 1, Dorsal view of the dorsal skin of the body
and of the ventral skin of four caudal rings before development (cf. text-fig. 6a). x 3. 2, Ditto, to
show carapace and eye-tubercle. X 5. 3, Dorsal view of Marco-cast after development. x5. 4,
Left pedipalp hand in dorsal view. Slide k. X Ik. 5, Ditto, in ventral view, x 7-i. 6, Ventral view
of ventral organs reassembled in approximately their original relative positions; based on Slides a,
b, ci, h (cf. text-fig. 6b). X 1\. 7, Hairs at the entrance to the mouth, probably on ventral side of
the base of the chelicerae. Slide g. X 225.

For key to abbreviations see p. 277 .

Palaeontology, Vol. 3.

PLATE 48

WILLS, Buthiscorpius

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’

285

Dimensions. The body was preserved in full relief, and consequently appears narrower in proportion
to its length than it would had it been flattened. Approximate measurements and estimates given
below agree closely with those of the holotype, of B.M. In. 31262, and other paratypes. As scorpions
go, these specimens represent either immature individuals or adults of a really small species, probably
the latter. Approximate measurements in mm. Length of carapace, 3 ; of abdomen, 9} ; of tail (4 segments
only), 7|; whole body (estimated), 24-25; greatest width of carapace (apparent), 3-8; of mesosoma
(apparent), 4-4; of mesosoma (estimate assuming it were flattened), 6 0. For dimensions of the legs
see Table, p. 289.

text-fig. 6. Buthiscorpius buthiformis ( Pocock). B.U. 720. a, Dorsal view of dorsal sclerites. b, Ventral
aspect of ventral parts with appendages restored. Cf. PI. 48, fig. 6. For key to abbreviations see p. 277.

Dorsal features of the body

The carapace (PL 48, figs. 2, 3; text-fig. 6a) was exposed as a poor internal mould, and
practically all the chitin had been lost. It was strongly arched, and before development
it appeared to be longer than wide, but afterwards it was seen to be almost square, or
wider than long if allowance be made for the arching (compare PI. 48, figs. 2 and 3). The
shape of the damaged anterior margin appears to have been slightly emarginate (as in
Eoscorpius typicus Petr.). The posterior margin is practically straight. In front of, and
parallel to the latter at about one-sixth of the length of the carapace, is a slight transverse
post-cephalic groove — a feature frequently found in fossil and Recent scorpions that
may perhaps be connected with the attachment of the antero-posterior muscle (Werner
1934, p. 63, figs. 39, 40). (A similar groove in Proscorpius and Palaeophonus is interpreted
by Petrunkevitch as proof that the first tergite was concealed under the carapace and

286

PALAEONTOLOGY, VOLUME 3

that therefore these Silurian forms had seven mesosomatic tergites, not the usual six.
This assumption appears to me to be entirely unwarranted, and to introduce a false basis
to his classification in Petrunkevitch 1955, p. P69.)

A fairly large eye-tubercle carrying the two median eyes lies about a third of the
carapace-length from the anterior margin. It is surrounded by an almost circular de-
pression. The eyes are damaged but appear to lie closer to one another than in B.M.
In.31262 and to be without intervening ridges.

The six mesosomatic tergites are best seen in PI. 48, figs. 1, 3, which show them to
increase greatly in length from front backwards; but their margins are poorly defined
because all the chitin has gone, and we are looking at an internal mould (see text-fig. 6a,
Tvii-xii). The body was preserved fully distended with the intersegmental skin stretched
out between the tergites, giving the impression of great length in relation to breadth.
This power of extending the length of the tergum was probably possessed by many
Carboniferous ‘scorpions’. It, and the misleading appearance that may arise from it, is
fully discussed on p. 320.

The dorsal plate of the 1st metasomatic segment (Txm), poorly preserved, appears to
taper unusually slowly. There are no strong keels on it. Possibly part of the sternal plate
xiii is to be seen at the right postero-lateral corner.

External casts of the ventral sides of four caudal rings of the tail were originally
present (PI. 48, figs. 1, 2; text-fig. 6a). Partially flattened, they appeared stout and
abnormally wide with traces of costal ridges, but the chitin having been lost no details
could be seen. During preparation most of the tail was ground away, and now only the
first caudal ring can be seen in the Marco-cast (PL 48, fig. 3). The fifth ring and the sting
were not present in the half nodule.

Appendages and ventral features of the body

The chelicerae, originally poorly exposed, now show as casts in the Marco, but no
details can be made out and no part was recovered.

The L. pedipalp was originally shown up by patches of white kaolin that indicated the
presence of the trochanter, femur, patella, and hand, extended in all over a distance of
about 1\ mm. Of these joints the L. trochanter, together with L. and R. coxae, were
recovered on slide g (PI. 48, fig. 6; text-fig. 6b), and the L. hand, minus the free finger, on
Slide k (PI. 48, figs. 4, 5). The L. coxa is well seen in slide g. It has the normal shape,
including the maxillary lobe extending towards the middle line and provided with fine
hairs (see below). The hand (PI. 48, figs. 4, 5) is very slender and produced into a rod-like
finger. The whole hand measures 5-5 mm. (as compared with 7 mm. in In.31262), made
up of about equal lengths of palm and fixed finger. Its maximum breadth is 1-5 mm. On
the dorsal inner surface at about one-third of the length from the proximal end is a
small but prominent knob, but little else can be seen on this side. On the other, one or
two minute hair-bases are visible near the base and tip of the finger (as in Buthiscorpius
major and Mazoniscorpio mazonensis, below), but there is nothing that can be regarded
as a row of granules along the biting edge. The free finger was not recovered.

Any differences that may appear to exist between this not completely flattened hand
and the crushed examples in B.M. In.31262 (PI. 47, fig. 3), above, are due to differences
in original size and in the mode of preservation. The long slender finger compares well
in shape with those figured for several other supposed Eoscorpionids — Eoscorpius typi-

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’

287

cits, Compsoscorpius, Alloscorpius and Trigonoscorpius, and with the Archaeoctonid
genus Eoctonus ; but in all cases the fingers appear more robust than does the one which
has survived in this specimen.

The organs around the mouth (text-fig. 7). In Recent scorpions the entrance to the mouth
is virtually a hollow tube-like filter formed by the chelicerae and labrum above, by the
maxillary lobes of the pedipalpal coxae at the sides, and below by the mandibular pro-
cesses of the 1st and 2nd legs. All of these may be thickly covered by brushes of forward-

FIG. 8

FIG. 9

text-fig. 7. Buthiscorpius buthiformis (Pocock). B.U.720. a, Mouth parts showing hairs attached or
(dh) detached from the dorsal side of the mouth, now in the kaolin, originally on the chelicerae.
Highly enlarged, b, Diagram-section through the mouth (mo). For key to other abbreviations see p. 277 .

text-fig. 8. Buthiscorpius buthiformis (Pocock). B.U. 720/M. ?Part of the lung-books — chitinous skin
at the pleural ends of four sternites, drawn in dorsal aspect with lighting from right to emphasize the
ends of overlapping films of chitin and the blotchy chitin on right (heavy shading); and sections along

aa and bb. Cf. text-fig. 9.

text-fig. 9. Buthiscorpius buthiformis (Pocock.) B.U. 720/M. Hypothetical sections through the right
half of a body segment to illustrate the position of Slide jM (cf. text-fig. 8), which is outlined by a small
rectangle, in relation to a possible pulmonary pouch with a stigma (st), a sternite (5), and a tergite (t);
laminate respiratory organ, diagrammatic (lb); pleural skin (ps).

directed bristles and hairs, forming a perfect filter to exclude all solids. In the fossiliza-
tion of the present remarkable fossil, this entrance became filled with ironstone, but
after solution of the latter it now appears (on Slide g) as an open cavity into which pro-
ject short fine hairs attached to the chitin of the pedipalpal coxae and to the mandibular
processes of the 1st and 2nd legs, and some similar hairs lying in the kaolin and not
attached to any visible chitin, but in a position corresponding to the chelicerae and/or
labrum (PI. 48, fig. 7; text-figs. 6b, 7a, b).

Taken in conjunction with the hairs actually still attached to the chelicera of B. major

288

PALAEONTOLOGY, VOLUME 3

(PL 52, fig. 4), the present specimen suggests that some at least of the Carboniferous
‘scorpions’ employed the same method of feeding as do their present-day descendants,
and one that would appear more appropriate in terrestrial than in aquatic animals (the
aquatic Eurypterus fischeri Eichw., however, also had hairs around its mouth. See Holm
1898.)

The coxae of the 1st and 2nd legs (text-fig. 6) closely resemble the corresponding parts
in Recent scorpions in possessing the strong mandibular processes referred to above, but
differ in that the process on the 2nd leg is not markedly larger than that on the 1st. This
difference from Recent scorpions was noted by Pocock ( 191 1, p. 28) in his description of
a paratype of his Anthracoscorpio buthiformis (B.M. 1.1555). The same characteristic
appears also in B.M . In. 31262 (above and text-fig. 4). The posterior edges of the coxae of
the second leg meet in an angle of 120°. As these edges coincide with the front of the
sternum, they reveal its anterior shape.

The sternum and the coxae of the 3rd and 4th legs. The sternum was badly broken
during development, and is now represented only by some small fragments in slide g
(PI. 48, fig. 6; text-fig. 6, st)\ but at one stage it was well exposed, and I noted it as being
roughly rectangular, wider than long, with a forwardly directed anterior side and a nearly
straight or very slightly emarginate posterior side. In view of the 120° angle between the
posterior edges of the two second coxae, noted above, its true shape must have been
roughly pentagonal, and somewhat different from that found in B.M. In. 31262 (above,
p. 282; text-fig. 4). It agrees, however, with the shape shown by Petrunkevitch as charac-
teristic of Eoscorpiidae (1955, fig. 40/1) which is based on one of Pocock’s paratypes of
Anthracoscorpio buthiformis (1911, pi. 1, fig. 2, B.M. 1.1555). See also Petrunkevitch
1953, fig. 34.

Coxa 3 is stout and short, barely half the length of the slender coxa 4 to which it is
completely fused. Both abut against the sides of the sternum.

This specimen confirms the findings in B.M. In. 31262, namely that the shape of the
sternum, the relation of the coxae to it, and the fusion of coxae 3 and 4 are all essen-
tially the same as in Recent scorpions.

Other parts of the legs. In death the legs had been flexed under the body and were
fossilized in this position. Supported by sheaths of kaolin, those that were still buried
in the ironstone etched out almost intact, but, with the exception of the 1st L. leg which
survives on Slide g, they broke off at various distances from their bases. The fragments
belong to L. legs 2-4 and R. legs 3 and 4. They were extracted and mounted as Slides a ,
b, d , e, f. On PI. 48, fig. 6, photographs of them have been placed in approximately
correct relationship to one showing the coxae and 1st L. leg on Slide g. It will be noted
that the distal parts appear in the figure as if they lie in the same plane as the coxae,
whereas they were preserved disposed in planes more or less at right angles to it, and in
postures exactly analogous to those assumed by the legs of any dead scorpion of today.
The legs increase in length from in front backwards (Table, p. 289), the 4th being
twice as long as the 1st. When fully extended it would have been capable of reaching
to the end of the 2nd caudal ring. The general shape and proportions of the different
articles of the legs are shown in text-fig. 6 and on PI. 48, fig. 6, to be extraordinarily like
homologous features in Recent forms. Clearly this particular Carboniferous ‘scorpion’

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’

289

was already adapted to moving about easily on land, a feat that Lichnophthalmus (Part
1 ) with its stiletto heels could never have achieved.

As in Recent forms, each leg terminated in a pair of non-denticulate curved claws
diverging from an angular basal pad. This was well seen in the 3rd L. leg, but the slide (a)
was subsequently damaged and the part lost in remounting. It agreed exactly with the
claw-parts on the 3rd leg of In. 3 1262 (text-fig. 5).

The absence from B. buthiformis of the denticles that occur on the claws of all the
larger ‘scorpions’ here described is perhaps to be correlated with the small size of the
species.

TABLE

Lengths in mm. of the body, the pedipalp, and legs (exclusive of the claws) of B. buthiformis

Body
( less tail)

Pedipalp

arm+hand

1st leg

2nd leg

3rd leg

4th leg

Holotype B.M. In. 18596

121

? 11-5
? 7-0+4-5

Paratype B.M. In.31262

12-5

14-5

80+6-5

? c. 7-8

00

?c. 9

c. 15

B.U.720 ....

12-5

c. 12
76-5 + 5-5

8-2

11-3

13-3

16

The three distal articles (tarsus, metatarsus, and tibia) of each leg carried a variety of
immovable spines and spinelets on keels, crests, and round their distal ends, together
with movable hairs, thorns, and bristles (some ? trichobothria); but the three proximal
ones (patella, femur, and trochanter) are, like the coxae, virtually devoid of any such.
Of considerable interest is the occurrence of two tarsal spurs sited on the skin between
the tarsus and metatarsus of each leg and of one metatarsal spur on the skin between
the metatarsus and tibia of the 3rd and 4th legs (these names for the spurs have been
adopted here as indicating the positions on the leg, whereas the terms ‘tarsal spur’ of
English authors and ‘Grunddorn’ and ‘Tarsalsporn’ of Werner 1934, p. 35, do not).
Now, in the classification of Recent scorpions the number and distribution of such spurs
are characteristics of a whole genus or even a family. When I later developed Mazoni-
scorpio I found the number and distribution of spurs to be just as in the present case,
and I suspect that the same is true for all the Orthostern Carboniferous ‘scorpions’.
Further, it is most remarkable that this arrangement can be matched exactly in Recent
scorpions, but only in members of one family — the Buthidae.

The distribution of fixed spines, movable hairs, and bristles on each of the three
distal articles of the legs can be epitomized as follows:

Tarsus with some small rather blunt bristles (? rudimentary thorns or Dornen of Werner), especially
a large group on the distal half of leg 3.

Metatarsus with two rows of short bristles on legs 1-3, scattered bristles on 4, a distal group of
bristles on 3 and 4. Two rows of short spinelets on 2 and 3, but none on 1 and 4. One end-spine on
all legs.

Tibia with no rows of bristles, but a few scattered ones at the distal end on legs 2, 4, and ? 3.

A few small but conspicuous hair-bases can be seen on the tarsus of legs 3 and 4, on the metatarsus
of all four, and on the tibia of legs 1-3. These may perhaps be regarded as the thecae of slender sensory

290

PALAEONTOLOGY, VOLUME 3

hairs (? trichobothria), because in no case has a bristle survived in attachment to them, whereas there
are many short bristles to be seen (above) which do not appear to spring from any definite hair-bases.
It is interesting to note that among Recent scorpions trichobothria are not known to occur on the legs
except in immature individuals.

The genital operculum and pecten were missing.

The sternites. Scraps of sternites i and n are mounted on Slides 1 and n, part of sternite
ii and most of hi and iv on Slide h ; and on slide m bits of them which were extracted
from the pleural region between the sternites and tergites. The chitin of the sternites is
very thin and devoid of ornament, hairs, and hair-bases, except for the minutest of prickles
and hair-bases on the posterior and postero-lateral margins (Slides h and /). Sternites
hi and iv are roughly rectangular, slightly wider than long, with the postero-lateral
corners rounded and posterior margin slightly emarginate. In Slide n this edge can be
seen to be double for a very short distance forward, suggesting that one sternite over-
lapped the next behind, but the preservation is not good enough to allow us to determine
the extent of overlap or to see whether or not there are pulmonary stigmata on the
postero-lateral border, where they occur in the Triassic scorpion Mesophonus (Wills
1947); but it can be stated definitely that there are no stigmata on the external surface
of any of the four sternites, so far as they are preserved (and this covers the greater part
of sternites hi and iv).

On Slide m was mounted a part of the pleural intersegmental area lying between the
tergites and sternites. It is sketched in supposed dorsal aspect on text-fig. 8. On it there
are four narrow strips of chitin (1-4) lying lengthwise along one side. They are some-
what blotchy, and darker and less transparent than the rest, and may represent the skin
originally dorsal to the ends of the sternites. Nos. 2-4 stand up at right angles to the
general plane of the specimen which seems to consist of two or three sheets of excessively
thin chitin, the top sheet being continuous with the upturned part of No. 3 (see section
A-A on the text-figure). 1 find it impossible to interpret the specimen with any degree of
certainty, but I am inclined to regard what I have sketched (it lies upward on Slide m)
as possibly respiratory structures originally lying dorsal to the R. ends of the four
sternites, the shortest (top of figure) being the most anterior. If this hypothesis is correct,
it would suggest that a pulmonary pouch and stigma existed between the external end of
each sternite and the corresponding dark strip which was part of the pleural skin, and
that in this pouch lay the thin chitinous laminae, forming some kind of lung-book.
This speculative explanation is shown in text-fig. 9.

mazoniscorpio gen. nov.

Type species M. mazonensis sp. nov.

Mazoniscorpio mazonensis sp. nov.

Plates 49, 50, and 51, figs. 4-6; text-figs. 10-13
Holotype. B.U.721A, B, and C. Pennsylvanian, Mazon Creek, Illinois.

Remarks. The holotype is on permanent loan to the Geology Department, Birmingham University,
from the Botany Department, University of Illinois, Urbana. The large nodule had been split in two
roughly on the plane of the dorsal skin. The half labelled A originally displayed dorsal features seen in
ventral view (PI. 49, fig. 1) with grooves on the carapace appearing as ridges. After development it
showed the external aspect of the dorsal skin (PI. 49, fig. 2) except where that had been lost. Here it

LEONARD J. WILLS: SOME CARBONIFEROUS 'SCORPIONS’

291

presented a mould of the inside of that skin on the Marco-cast. This has now been filled in and become
a transparent Marco-block, 721 A. The last segment of the tail and the sting were sawn off the nodule
and developed separately. They were mounted in Marco on glass as 721 C.

The half labelled B originally exhibited bits of dorsal skin, the natural cast of the inside of the dorsal
skin, and indications in places of ventral organs, all seen in dorsal view (PI. 50, fig. 1). After develop-
ment it displayed many ventral organs in ventral view together with traces of the dorsal features as
impressions in the Marco-cast (PI. 50, fig. 2; PI. 51, fig. 4). The Marco-cast has now been filled in as a
Marco-block, B.U.721B.

The specimen was remarkable for its large size — about 7 cm. in length — and is so now for the com-
plete preservation of almost every part, like an insect in amber, within the three Marco-blocks. The
missing organs, except for parts of the ventral skin which is probably present but crushed on to the
dorsal tergites, can be accounted for as follows. Part of one sternite fell away, but was recovered; and
I extracted and mounted the left halves of two sternites in order to be able to examine them in trans-
mitted light. I also recovered the end of one leg, fragments of the finger of one pedipalp, and a few
tiny bits of other parts.

Diagnosis of genus and species. Large Orthostern ‘scorpion’, e. 70 mm. long and 1 1 mm.
wide at widest tergite xi; dorsal side and prosomatic appendages conforming to the
pattern of a Recent buthid scorpion with a long sting, large chelicerae, powerful pedi-
palps, and slender legs which have the buthid arrangement of spurs.

Carapace almost square with a median groove and two cephalic and two postcephalic
arched lobes ; median eyes small on front of an eye tubercle of two kidney-shaped bosses
separated by a narrow groove, close to front margin of carapace; front part of carapace
has deep doublure and is coated with fine hairs, rear part covered irregularly with small
granules which are tiny on the postcephalic lobes. Tergites coated with fine hairs,
otherwise unornamented. Tergal plate xm and all caudal rings with strong dorsal keels,
caudal ring xvm short, sting long, flask-shaped with strongly curved aculeus.

Chelicerae large, projecting in front of the carapace for a distance equal to not less
than half its length. Pedipalps powerful. Coxo-sternal area with a bluntly pentagonal
sternum; mandibular process of 1st leg larger than that of 2nd, coxa 4 about twice the
length of coxa 3; legs relatively slender, coated with fine hairs, two denticulate claws
and two tarsal spurs on each leg; one metatarsal spur on 3rd and 4th legs.

Genital operculum with two pairs of arched lobes and a narrow median ridge carrying
a flatfish leaf-shaped plate. Sternum of pecten ill-defined, combs large, each with at least
sixteen teeth. Sternites roughly rectangular, overlapping backwards. Sternite ix ill-
defined, possibly triangular behind. Sternite x lamellate, in two halves, each half with
deep posterior doublure, rounded at postero-lateral corner and here covered externally
and on doublure by minute hairs ; no stigmata. Sternites xi, xn similar to Sx, but possibly
not divided into two halves. Sternal plate xiii large with pronounced rounded postero-
lateral corners, possibly half the plate was covered by Sxn.

Description. The carapace measures 8 mm. in length and 9 mm. in breadth behind. The
abdomen is 20 mm. long and 10 mm. wide at its widest part (tergite xi). The tail, as
preserved in A, is c. 20 mm. in length, but a few mm. were lost in the sawcut that severed
the last caudal ring and tail. The last caudal is 5 mm., and the sting can be estimated at
not less than 15 mm. These figures give the total length of the ‘scorpion’ as 68-70 mm.

Dorsal surface

Prosoma. Carapace (PI. 49, fig. 3; text-fig. 10). This is nearly rectangular, with a slight
taper forwards. The front corners are rounded and devoid of lateral eyes. The anterior

292

PALAEONTOLOGY, VOLUME 3

margin has a slight median projection. The whole front end has a very deep doublure
extending back at least as far as the hind end of the eye-tubercle. The carapace was
originally rather strongly arched with a deep median and two slight postcephalic
grooves, which together divide the shield into two lobes of the cephalic region in front,
and a pair of postero-lateral postcephalic lobes behind. The median groove narrows in
front where it is continued as a narrow furrow between the two halves of a small circular
eye-tubercle, situated very near to the front margin of the carapace which it overhangs.
The small eyes, originally hemispherical and looking upwards, occupy the front half of
the tubercle. The posterior margin of the carapace is a broad band of thick skin rounded
at the corners, with a few small hair-bases. Many parts of the skin of the shield are
ornamented by numerous small granules which show up most prominently on the folds
at the sides of the grooves. The postcephalic lobes, however, are covered with similar,
but quite tiny granules. Numerous hair-bases, some with short setae still attached,
occur on the external surface of the R. anterior corner (some also are probably on the
doublure). In addition, the original existence of a coating of minute fine hairs over
much of the front end is evidenced, where the chitin of the shield is missing, by the
actual hairs (now attached to the Marco-cast) which were torn from their bases but left
in the rock when the skin was broken away. A similar coating of fine hairs has been
observed in other parts of the animal (below).

EXPLANATION OF PLATE 49

Figs. 1-7. Mazoniscorpio mazonensis gen. et sp. nov., B.U.721A. 1, Ventral view of dorsal skin, as
received. Photographed under alcohol. X2-7. 2, Dorsal view of dorsal skin and R. appendages
after development. X2-7. 3, Chelicerae, carapace, and tergite vii, after development (cf. text-
fig. 10). X 5. 4-7, The ends of the four L. legs, fig. 4 being the 1st and fig. 7 the 4th, photographed
during development. x7.

For key to abbreviations see p. 277.

EXPLANATION OF PLATE 50

Figs. 1-5. Mazoniscorpio mazonensis gen. et sp. nov., B.U.721B except figs. 3, 4 (fragments derived
from 721 /A). 1, Dorsal view of dorsal skin and in places traces of ventral features. As received,
x 2-2. 2, Ventral view of the ventral organs of anterior part of the body before detachment of
sternites, and of the Marco-cast containing bits of chitin and impressions of dorsal and ? ventral
skin of the posterior end (cf. text-fig. 12). X 5. 3, Tip of the fixed finger of the hand of the R. pedi-
palp. Slide B.U.721 A/4/6. x40. 4, Metatarsus with spines and tarsal spurs, tarsus and claws of
2nd or 3rd leg, showing hairs and hair-bases. Slide B.U.721 A/3. X 14. 5, Fragments ? left side
of sternite ? xn, figured on PI. 6, fig. 6; photographed by transmitted light to show hairs on both
surfaces. Slide B.U.721B/1. X40.

For key to abbreviations see p. 277.

EXPLANATION OF PLATE 51

Figs. 1-3. Buthiscorpius major sp. nov. G.S.M. Za 2926. 1, As received. Ventral view of dorsal surface,
photographed under alcohol. X 3. 2, 2nd Marco-cast often etching. Dorsal view of dorsal sur-
face of body, with tergites vii-xii and two fragments of the sting. X 3. 3, 1st Marco-cast repaired,
dorsal view of dorsal surface of body and tail, after etching, x 3.

Figs. 4-6. Mazoniscorpio mazonensis gen. et sp. nov. 4, B.U.721 B. Progress photo of ventral organs
to show the L. pedipalp and L. legs with claws and spurs. Some parts outlined in ink. x21. 5,
Ventral view of the skin of the L. ends of sternites ? x, xi, with the posterior margin and median
continuation of ? Sx. Slide B.U.721/2. x 6-8. 6, Part of the L. side of ? Sxn having hairs on both
sides (see PI. 50, fig. 5). Slide B.U.721B/1. X6-8.

Palaeontology, Vol. 3,

PLATE 49

W I L LS, Mazoniscorpio

Palaeontology , Vo I. 3,

PLATE 50

mi»Wf

9ff jxmSSnKi

i. ;&f <

P>t 3

m ffil-g&X

/** «** &ay

mf

WILLS, Mazoniscorpio

Palaeontology, Vol. 3.

PLATE 51

4 x 21

WILLS, Buthiscorpius and Mazoniscovpio

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’

293

Mesosoma. Tergites vii-xii (PI. 49, figs. 1, 2). These are quite normal in shape and
proportions, increasing in length backwards from 1 mm. to 3 mm., and in breadth from
the carapace to tergite xi and then decreasing slightly. Each is bounded by a linear
anterior margin and an unornamented posterior margin formed by a narrow infold or
doublure which passes into the wide anterior border of the next segment. These margins

no. 10

text-fig. 10. Mazoniscorpio mazonensis, gen. et sp. nov. B.U.721A. Chelicerae, carapace, and 1st
tergite. X 5. cc, coxa of L. chelicera outlined by broken line; cp, coxa of L. pedipalp; et, eye-tubercle;
fb, finger of L. chelicera bent backwards ; fd, fixed finger of R. chelicera;#, free finger, ditto;#, femur
of L. pedipalp; me, median eye; mg, median groove; pi, posterior lobe covered with minute granules;
ps, pleural skin; R.ch, right chelicera; sa, setae and hair-bases of ant. outer skin; sm, short setae of sa

in Marco; Tvii, tergite of 7th segment.

text-fig. 11. Mazoniscorpio mazonensis gen. et sp. nov. B.U.721C. Pre-anal caudal ring (Cxvm) and
poison-capsule (pc) and aculeus (ac) of the sting, a, In ventral view showing the shield-like surface of
Cxvm and the folded ventral surface of the capsule. B, Dorsal aspect with aculeus (broken line)
restored to its original position, dk, dorsal keel, c, Aculeus in plan and section, d, Sagittal diagram-
section of present postures with spaces between the crushed skin opened up. Arrows show direction of
crushing pressure, e. Tentative restoration, anus (an).

originally appeared as rather wide lines (PL 49, fig. 1). They carry, possibly on the
doublure, a few small hair-bases. The surface of the tergites is devoid of granules, but
have a uniform coating of minute hairs without visible hair-bases, which is similar to
that on the front of the carapace. There are folds across the ends of some of the tergites
which suggest that in life they were strongly arched from side to side.

Metasoma. Tergal plate xm. The outer surface of this was completely exposed by the

u

294

PALAEONTOLOGY, VOLUME 3

etch. It has the normal outline, rapidly narrowing backwards to about one-third its
anterior breadth to accommodate the 1st caudal ring, ft carries a pair of strong, knobbly
dorsal keels which end well in front of the true posterior margin at points where they
meet a transverse ridge that mimics the appearance of the real margin. As a result of
having these keels, the tergal plate resembles the dorsal sides of the caudal rings. Assum-
ing that the keels were for muscle attachments, it would appear that the arching of the
tail and sting over the body of the ‘scorpion’ involved the last segment of the abdomen
as well as the caudal segments.

Metasoma. Sternal plate of segment xm is poorly exposed, most of its chitin having
been lost or else pressed against the inside of the tergal plate. From what is left in
specimen B it appears to have been wider than the latter and to have ended in two
postero-lateral processes (PI. 51, fig. 4).

Caudal rings and sting (PI. 49, fig. 2; text-fig. 11). Caudal rings xiv-xvii etched out
complete as a sort of bridge from the body to the wall of Marco to which Cxvn was
attached (i.e. the sawcut referred to above). It was not feasible to clear all the matrix
from below the bridge, and for that reason they can only here and there be viewed by
transmitted light, and details of their ventral sides are unknown. Owing to compression
the rings appear wider than they originally were. Each carries a pair of strong dorsal
keels (it cannot be seen that these are denticulate) and a pair of lateral or dorso-lateral
keels. They all resemble closely the corresponding rings in Bathiscorpias buthiformis
(text-fig. 2).

The caudal ring of the pre-anal segment xvm and the sting are preserved in specimen
C (text-fig. 1 1 ). The caudal ring is short and this may be due to the sawcut that severed
it from specimen A, but it appears to be complete. It is not much longer than the pre-
ceding ring, and in this respect differs from the corresponding segment in Buthiscorpius.
As the ring and sting are completely transparent every detail can be examined. The
dorsal surface of the ring has two short, posteriorly elevated dorsal keels, but is other-
wise devoid of conspicuous features, except for a dark blotching of the skin and two
minute longitudinal folds at the distal end (text-fig. 11b). These may have led to a
median sinus on the posterior margin, but this, if present, is concealed by the folded
sting. The blotching and folds recall features seen in the pre-anal segments in Meso-
phomis (Wills 1947, p. 69, text-fig. 34). The ventral surface in its flattened state is clearly
seen to have an almost rectangular shield-like shape that is defined by folds at the front
and sides, and by a strong posterior margin and doublure behind. Near this margin
were long setae, a few of which are still in place. In addition there are on various parts
of the ring several small sensory hairs (one still in place) attached to hair-bases, and
many short hairs like those that coat the front of the carapace and the tergites. These
latter appear to be on the ventral surface only.

The sting (text-fig. 11) consists of a crushed and crumpled, originally flask-shaped,
poison-capsule and a long, strongly curved aculeus. The latter is now broken away
from the neck of the capsule, but was seen during development to rise vertically from it
(the neck is still visible in this position, but the broken aculeus had to be mounted on its
side). As found, the aculeus pointed distally with its concave side ventral. To attain this
posture (text-fig. 11d) it must have been inverted during the consolidation of the rock
by pressure applied to the whole sting as it stood up almost vertically over the pre-anal
ring in the posture usually to be seen in dead scorpions. In the capsule various folds of

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’ 295

the skin can be seen, but its exact original shape is hard to determine. Text-fig. 11e
shows a possible reconstruction. Some parts of it are closely covered by numerous small
hair-bases, often with short setae attached. The aculeus itself (text-fig. 11c) tapers
gradually to a fine point. Its smooth, thick chitin is folded longitudinally in a way that
strongly suggests that it contained a pair of poison-ducts (as is the case in Recent scor-
pions and in the Triassic Mesophonus. See Wills 1947, p. 75.) Near the base of the
broken part of the aculeus is a single tiny hair-base.

Ventral organs

As appears to have been the case in several of the ‘scorpions’ developed, the original
fracture of the nodule followed the inside of the dorsal skin, and where the ventral
skin had been closely apposed it suffered some damage. In the present case most of the
ventral organs can be seen in specimen B as far back as the last sternite (Sxn), but im-
portant parts of them are damaged or missing, so that some uncertainty as to their
shapes is inevitable.

The ventral half as originally exposed in dorsal aspect is shown on PI. 50, fig. 1. The
outlines of parts of the chelicera, of the two pedipalpal trochanters, of the mandibular
processes of coxae 1 and 2, the end of the 1st R. leg, and bits of other appendages could
be made out, but after development many other features appeared (PI. 50, fig. 2; PI. 51,
fig. 4; text-fig. 12), most of which are still visible in the Marco-block B.

Prosoma. Chelicerae. These show best in specimen A (PI. 49, fig. 3; text-fig. 10). The
basal joint of the L. chelicera can be seen to lie below the left front corner of the cara-
pace, followed by the hand which appears to have been bent backwards on itself (text-
fig. 10). The hand of the R. chelicera shows the distal parts of the fingers, while their
tips can be seen in specimen B (it is possible that the tips referred to belong to the L.
chelicera). The hands are at least half as long as the carapace (cf. Liehnophthalmus in
Part 1, p. 288). In specimen B the Marco-block carries many short hairs scattered over
the impression of the R. hand, and one or two larger ones on the fingertips.

Pedipalp. The original fracture damaged the coxae and trochanters of the pedipalps,
but their general shape can be seen to be quite normal. The distal ends of the coxae
carried rather large hairs (? trichobothria) and so did the trochanters which also have a
coating of smaller bristles like those on the carapace, tergites, &c. The rest of the
appendage consists of massive articles of the normal shape in Recent scorpions. The
whole R. hand except the very tips of the fingers is preserved in specimen A, and in B the
L. one is complete save for the end of the fixed finger, but bits of the missing part were
recovered (Slides A/4, A/6; PI. 50, fig. 3). The mounts show that there was a row of large
granules along the biting edge with a second row of widely spaced still larger ones
farther from the edge, a few long, slender, certainly trichobothrial bristles and numerous
smaller ones without conspicuous bases, and that the skin was thick and markedly
cellular in texture. The femur, patella, and palm of the hand have few or no hairs that
can be seen. The patella and hand are crumpled by several large lengthwise folds which
imply that in life both were strongly keeled. No ‘stop-knob’ can be seen on the patella.

These features can be matched in large Recent forms, in Buthiscorpius major (PI. 52,
fig. 3), and in Liehnophthalmus (Part 1, p. 278, pi. 49, fig. 7). The general shape is also
much the same as in B. buthiformis (PI. 47, fig. 3) which, however, is relatively shorter
and, perhaps because of its small size, has no granules on the biting edges.

296

PALAEONTOLOGY, VOLUME 3

The coxo-sternal region and legs. These only differ in minor points from those of
certain Recent Buthids and from the other Carboniferous Orthosternid ‘scorpions’ here
described which show these organs, in particular B. buthiformis, B.U.720. The relation
of coxae to the sternum is that stated by Petrunkevitch (1955, p. P73) to characterize the
superfamily Scorpionoidea Leach 1815, namely — ‘First and 2nd pairs of coxae with well-
developed maxillary lobes (fig. 40, 1), those of the 2nd pair meeting in median line and
wedged in between maxillary lobes of 1st pair; 3rd and 4th pairs of coxae abutting
against sternum’. His fig. 40, 1 is based on one of Pocock’s paratypes of Buthiseorpius
buthiformis (B.M. In. 1555).

text-fig. 12. Mazoniscorpio mazonensis gen. et sp. nov. BU.721B. A, Ventral view of the ventral
organs with the positions of some dorsal elements (broken line), cf. PI. 50, fig. 2, PI. 51, fig. 4. b, Tenta-
tive restoration of the same, the existence of the plate marked ? Sxii being in doubt. c/Sxi, internal
(dorsal) skin of the pouch above Sx (this skin belonging to Sxt); fch, finger of ? R. chelicera; mpc ,
maxillary lobe of coxa of pedipalp; vSx, internal (ventral) skin of the pouch above Sx pressed against
the external skin of the same. For key to other abbreviations see p. 277.

The coxo-sternal region and the left legs are well seen in specimen B (PI. 50, fig. 2;
PI. 51, fig. 4; text-fig. 12).

The sternum is pentagonal with its two anterior edges, defined by the back edges of
the two coxae 4, making an angle of 130°, and with its posterior edge slightly emarginate
where it adjoins the genital operculum. Against its sides abut coxae 3 and 4.

The big coxae of the pedipalp appear, in ventral view, to underlie the mandibular pro-
cesses of coxae 1 and 2. The maxillary process of the L. pedipalp with a few largish hair-
bases is well exposed (text-fig. 12, mpc). The mandibular processes of the 1st leg are
much broader than those of the 2nd leg (in Recent scorpions the reverse is the case).
Both processes have a felt of minute hairs on the inner sides of their tips. The rest of

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’ 297

coxa 1 is small, little more than an articulation for the small trochanter. In contrast the
small mandibular processes of coxae 2 pass into large horn-shaped articular ends, the
posterior sides of which define the front of the sternum. The coxae of the 3rd and 4th
legs abut against the sides of the sternum, that of the 4th leg being about twice as long
as that of the 3rd. The two on the L. side seem to be still united, whereas there appears
to be a gap between those on the Right. This appearance is probably the result of the
loss or fracture of the thin connecting chitin.

The four L. legs are preserved virtually complete in specimen B (PI. 51, fig. 4), but in
specimen A all the R. legs except the 3rd are broken (PI. 49, fig. 2). In general shape
and proportions all are closely similar to the much smaller ones described in B. buthi-
formis, but nearly every joint can be seen to be covered with a felt of small setae com-
parable with those on the carapace, tergites, &c., and the claws are denticulate, as in all
the larger specimens here described. The pads ( Gehstachel ) at the base of the claws are
quite small, as in other orthosterni. Spines, sensory hairs, and spurs on the interseg-
mental skin at the base of the tarsus and metatarsus are conspicuous and arranged as in
B. buthiformis (p. 289), in particular two tarsal spurs on all four legs, and a single meta-
tarsal spur on the 3rd and 4th. The ends of the four legs were displayed simultaneously
at one stage of the etching and were photographed (PI. 49, figs. 4-7). The figures show
that the tarsus gets progressively longer with the increase of length of leg from the 1st
to 4th. That on the second appears to be very slender, but this may be due to an accident
of preservation. Details of the claws, spurs, spines, and sensory hairs (? small tricho-
bothria) are well shown on the end of the 2nd or 3rd R. leg which broke loose during the
etching of specimen A. Mounted as Slide A/3 (PI. 50, fig. 4) it displays very clearly the
spiny ends of the sides and lower surface of the metatarsus, a feature not noticeable in
the metatarsus of other specimens, but one that I note below as being conspicuously
developed on the tibia of the 3rd leg of G.S.M., Za 2924 (text-fig. 19e). Slide A/3 also
makes clear how the pad is connected by prominent ridges to the bases of the claws.

Mesosoma. There is very little ventral skin preserved behind the genital operculum and
it is not easy to trace any margins to the sternum of the pecten and the first sternite
(S ix). The next two were better preserved, their L. postero-lateral parts having etched out
well. I photographed them under water (PI. 50, fig. 2, PI. 51, fig. 4) to obtain a record.
Then in order to be able to examine them from both sides, and because they appeared
to be almost detached by the solution of an underlying (in ventral view) film of matrix,
I prised them away without breaking them and mounted them as Slide B/2 (PI. 51 , fig. 5).
At the point on the Marco-cast from which the supposed Sx was detached, there is an
area covered with thin chitin. Somewhat similar, but less distinct patches follow where
the next sternite (Sxi) was lying and where the supposed last sternite (Sxii) is indicated
in text-fig. 12. 1 think that the fragment of sternite mounted as Slide B/l (PI. 51, fig. 6)
may be the L. end of Sxii, derived from above this last patch, but its exact provenance is
unknown. If I am right in this interpretation, Sxii would in ventral view lie partly over
Txii and partly over Txm, as shown in text-fig. 12. If I am wrong, those segments
marked ?S ix-Sxi must be in reality Sx-Sxn, and the part marked ?Sxn must be merely
the front half of Sxm.

Accepting my preferred interpretation (text-fig. 12) the following features may be
noted.

298

PALAEONTOLOGY, VOLUME 3

The genital operculum (text-fig. 1 3) is a complex organ, the skin of which is still pre-
served in slight relief. It is almost oval in outline with the anterior edge making an
obtuse, forward-pointing angle. This edge at one point can be seen to be parallel to the
posterior margin of the sternum. Both its sides are obscured by the flattened 4th coxae.
The posterior edge curves gently backwards near the middle line, is overhung by the
external lobes (see below), and is inturned as a narrow doublure. The whole operculum
is clearly a bilaterally symmetrical organ consisting of two pairs of lobes, the two external

text-fig. 13. Mazoniscorpio mazonensis gen. et sp. nov. B.U.721B.

Parts of the sternum, genital operculum, and the two coxae 4. X 10.
c4, coxa 4; el, external lobe; il, internal lobe; si, slit in median ridge;

st, sternum.

ones (el) being arched and falling away towards the posterior margin and towards the
two internal lobes (il) which are also slightly arched. Down the middle line is a con-
tinuous narrow ridge or keel which, towards the front, carries a leaf-shaped flattish
structure. Behind this the ridge at one point shows a narrow slit which may represent
the opening of the genital duct, but is more probably an accident of fracture.

With this specimen before us it is now clear that the structure in Lichnophthalmus
pulcher Petr., tentatively interpreted as an anterior plate of the sternum of the pecten
(Part 1, p. 272, text-fig. 4), is really the genital operculum, since it also consists of two
pairs of lobes with a median ridge. I pointed out the possibility that this might be the
case, and noted that, should it turn out to be so, then coxa 4 lay alongside the genital
operculum (as is the case here), and did not abut against it as it does in Pareobuthus,
Eobutlius, and other Tsobuthidae’. See also Addenda, p. 331.

Behind the genital operculum, on the 8th adult segment, the sternum of the pecten is
poorly defined, but appears to have been short. It carried a pair of large combs on which
the raches are very broad at their bases but taper to a point. Possibly both combs have
been broken and the teeth displaced backwards near the middle line. The rachis carried,
at any rate distally, bosses with sensory hairs (a feature common to all the combs
examined) ; fulcra cannot be recognized but may well exist ; sixteen teeth canbe counted
on the L. comb.

Behind the sternum of the pecten a long stretch of the ventral skin, partly destroyed
and partly covered by the combs, could easily account for two sternites, but I think it
was more probably occupied by a single large one (the supposed sternite ix), the pos-
terior margin of which was a broad V evidenced by a groove in the Marco. Since the
median suture of Sx can be traced across it, the groove may indicate the limit of the
area on Sx covered by the V-shaped end of S ix.

Much of the supposed sternite x has survived. It was a roughly rectangular flat
lamellate organ with a median division (perhaps in the form of a suture) and a small

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’ 299

median posterior notch between the two halves. (It is interesting to note that in Pareo-
buthus sternite x also shows a line of very thin skin between its two lobes, which
appears to be absent from the other sternites. Wills 1925, pi. 3, fig. 2.) Much of the L.
half was originally seen inplace (PI. 50, fig. 2; PI. 51, fig. 4), but is now mounted as Slide
B/2 (PI. 51, fig. 5). The R. half had been displaced and crushed sideways, but many of
its details can still be seen in the Marco-block (text-fig. 12a), including indications of
the median notch.

Only the R. end of the supposed sternite xi was seen during the etching, and this came
away attached to the adjacent piece of sternite x (Slide B/2, PI. 51, fig. 5). As noted
above, another end of a sternite was detached by the etching before its position had been
noticed. It may be the L. end of sternite xii, but it could equally well be the R. end of
sternite xi. Here it is regarded as the L. end of sternite xn (Slide B/l, PI. 50, fig. 5; PI. 51,
fig. 6). Sternites xi, xii are very badly preserved, and no sign of a median suture can be
seen. The absence of chitin is due to the ventral skin having been pressed against the
dorsal, and the two having broken away from specimen B. The ventral skin can be seen
in places in specimen A, but no details can be made out.

In text-fig. 12a I have shown by broken lines the position of the intersegmental skin
between tergites xi, xii, and the tergal plate of segment xm, and on text-fig. 12b the ends
of all the tergites, as in specimen A. From these it will be seen that on the proposed
interpretation each sternite covers not only its corresponding tergite, but about half of
the next one behind, Sxn eventually concealing the front half of Sxm, which last is only
represented by a few scraps of chitin, whereas the whole of tergal plate xm is preserved
in specimen A (PI. 49, fig. 2). Each overlap formed a pouch opening to the sides and
behind. Some parts of the overlapping portions of each of the supposed sternites x-xii
can be examined — the L. ends as Slides B/l and B/2 and the R. end of Sx in the Marco-
block B. In every case they consist of an external layer and a large thin-skinned doublure,
and in B/l and B/2 both surfaces can be seen to be covered closely by a felt of very
minute hairs (PI. 51, fig. 6). At the L. end of Sx, a patch of thin chitin can be seen in
the Marco-block at the point from which Slide B/2 was detached, and a similar patch
shows at the right end where sternite x has been displaced (text-fig. 12a, <7Sxi). These
are regarded as the dorsal or inner linings of the pouches, the skin itself being the over-
lapped portion of the next segment behind. On the same figure the letters vSx point to
the crumpled and displaced outer skin and doublure of the overlapping Sx.

These observations appear to prove that at either end of each sternite the postero-
lateral corner concealed a pouch, the ventral lining of which was the posterior doublure
of that sternite, and the dorsal lining of which was the anterior part of the next sternite
behind (or the sternal plate, Sxm, in the case of the last one). In this respect the sternites
compare exactly with the leaf-appendages ( B/attfiisse ) of Eurypterus, but the overlap
being relatively narrower they had less freedom of movement, and were correspondingly
more like true sternites than leaf-appendages. The general arrangement also matches
closely the structure of the sternites in Pareobuthus (Wills 1925) and Liclmophthalnms
(Part 1, p. 274), though the sternites in the present case are not bilobate, and have hairs
instead of spinelets on the doublure.

Imagine a gill within each pouch and we have a structure comparable with that of a
Blattfuss of Eurypterus fischeri Eichw. as described by Holm 1898. The available space
for the gill, however, would seem to be relatively much smaller than in Eurypterus, and

300

PALAEONTOLOGY, VOLUME 3

the sternites ill-adapted to promote a circulation of water through the gills. Such an
interpretation would nevertheless seem to imply a truly aquatic life for this particular
Orthostern ‘scorpion’, an environment similar to that inferred for the Lobosterni
described in Part 1.

Alternatively, imagine some air-breathing organ, perhaps a lamellate structure akin to
a gill-book, occupying the pouches and protected from drying up by the close-fitting,
hair-covered corners of the ‘sternite’, and we have an arrangement that can be pictured
as a first stage in the evolution of a scorpion’s lung-books. By the fusion of the outer
edge of the first sternite to the overlying ‘sternite’, except for a short strip at either end,
the arrangement found in the Triassic Mesophomis could follow. Here the lung-book
opening (stigma) is either on the postero-lateral margin or on the adjacent doublure
that connected that sternite to the next one behind (Wills 1947). Starting again from that
arrangement, it is easy to postulate a simple migration of the stigma from the edge to
the outer surface of the sternite to account for the siting of the pulmonary opening in
present-day scorpions.

On this second hypothesis this particular Orthostern ‘scorpion’ would rank as an
air-breather, though probably only adapted to life in a moist environment.

There is no satisfactory evidence as to which hypothesis is correct, but the amount of
overlap in the present case is greater than in B. buthiformis, and large enough to make
me favour the aquatic one.

SECTION B — ONLY HALF-NODULES AVAILABLE

Buthiscorpius major sp. nov.

Plate 51, figs. 1-3, Plate 52; text-figs. 14-16

Holotype. G.S.M. Za 2926. Coal Measures (Ammanian), base of Communis Zone, Kilburn Coal,
Trowell Colliery, Nottinghamshire.

Remarks. As originally exposed, there were visible most of the carapace, the mesosomatic tergites, the
tergal plate of the 1 3th segment, all the caudal segments of the metasoma except the sting, and several
bits of the legs (PI. 5 1 , fig. 1 ). After photography the specimen was embedded in Marco, but during the
grinding away of some unwanted matrix, the Marco-mount cracked right across, and a fresh start had
to be made. In trying to extract the nodule from the cracked mount the specimen was broken in two,
but not at the place where the mount had split (PI. 51, fig. 3). The rock containing the tail adhered to
the Marco, and was later developed by solution. This first mount with its cracks repaired provides a
record, in the form of a cast, of what was originally visible and also of what was etched out in the
caudal region. It is referred to in the sequel as the 1st Marco-cast.

The part of the nodule which broke away from this first mount contained the body and appendages.
It was remounted and developed by solution, and after several parts of the scorpion had been extracted,
the second mount remained as the 2nd Marco-cast. Parts of the sting and of one leg, however, were

EXPLANATION OF PLATE 52

Figs. 1-5. Buthiscorpius major sp. nov., G.S.M. Za 2926. 1, L. chelicera and pedipalp minus its coxa,
in dorsal view. Slide Za 2926/1. x 10. 2, R. pedipalp minus its coxa in dorsal view, except the
broken end of free finger which was inverted in mounting. Slide Za 2926/2. x 10. 3, Tip of the free
linger of L. pedipalp in ventral view. Slide Za 2926/1. X 66. 4, Hand of the L. chelicera in ventral
view. Slide Za 2926/1. x66. 5, Coxa (below) and trochanter of ?3rd or 4th R. leg. Slide Za 2926/3.
x 10.

For key to abbreviations see p. 277.

Palaeontology, Vol. 3,

PLATE 52

WILLS, Buthiscorpius

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’

301

left attached to it in their original positions and can still be seen in place (PI. 5 1 , fig. 2). It was surprising
to find the sting on this Marco-cast as the rest of the tail was etched out on the 1st Marco-cast. It must
have been bent back over the tail since the aculeus points towards the head of the scorpion.

At first it was assumed that the visible dorsal organs were exposed in dorsal aspect on the half-
nodule, but on development it was found that in fact they were exposed in ventral view, for the 2nd
Marco-cast has the eye tubercle projecting upwards and the tergites overlapping one another back-
wards as ridges (see text-fig. 14 and PI. 51, fig. 2). This explained the disappointing fact that the ventral
parts — the sternum, genital operculum, pecten, and stemites — were not discovered during develop-
ment. They lay in the other half-nodule which had not been collected. However, one chelicera, both
pedipalps, and the coxa and trochanter of the 3rd or 4th right leg were isolated with the original
brown chitin virtually free from matrix. Their most intimate details of structure can be examined in
transmitted light (PI. 52), and it can be demonstrated that these parts were organized in Carboniferous
times on almost exactly the same lines as they are in a Recent buthid scorpion.

B. major is undoubtedly an Eoscorpioniid comparable with Eoscorpius Meek and Worthen, Buthi-
scorpius Petr., and Compsoscorpius Petr. The carapace of the holotype of E. carbonarius lacks its front
half, and so comparison with it in respect of the shape and proportions of the carapace, and the posi-
tion and nature of the eyes is precluded. I thought at one time the present specimen showed one or
more lateral eyes, but am now convinced that I was mistaken. Had lateral eyes been present the speci-
men could be ascribed with assurance to Compsoscorpius elongatus Petr. ; but as they are not, compari-
son is closest with Buthiscorpius buthiformis Pocock, as described above, though the median eyes
appear to be almost touching one another instead of being separated by ridges as they are in that
species (text-fig. 1). The preservation, however, is too poor for certainty on this point. The fossil is,
however, almost twice the size of the holotype of B. buthiformis and considerably larger than any of
the specimens attributed to that species by Pocock or by me. For convenience in description and
reference I name G.S.M. Za 2926 Buthiscorpius major sp. nov.

Diagnosis. Large Eoscorpionid ‘scorpion’, about twice the size of Buthiscorpius buthi-
formis Pocock ; carapace ornamented with granules, some being large, mimicking lateral
eyes which are absent; median eyes small and near to one another on an eye-tubercle
without visible ridges between the eyes ; eye tubercle about two-fifths of carapace length
from the front; tergites with mucronate posterior margins; tail short and relatively
shorter than in B. buthiformis ; caudal ring xvm not much longer than the previous one,
and shorter than the flask-shaped sting. Pedipalp hand with fingers longer in proportion
to the palm than in B. buthiformis.

Dimensions. The holotype lay squashed almost flat on the ironstone. For purposes of comparison of its
dimensions with those of B. buthiformis Pocock as described above, it must be recognized that the
tergites of the present specimen are telescoped from back to front and flattened, so that the length-
dimensions are relatively less and the breadth-figures relatively greater than corresponding measure-
ments in the Birmingham specimen which was fully distended lengthwise and strongly arched from
side to side; and that both the above-described specimens are larger than the holotype of B. buthiformis
which (without the sting) is only 22 mm. long. Making allowances for differences in preservation and
for possible differences in age of the individuals, I consider it probable that the present specimen
represents an adult of a species that was about twice the size of B. buthiformis. Hence the specific name
major is proposed.

Approximate dimensions in mm. Carapace, length, 7-8; width, ? 7. Abdomen, length, 12; width
(maximum at Tx), 8-5; width of Txm at front, c. 7, at back, c. 3-5. Tail, length Cxiv to Cxvm, 13;
width (crushed), c. 3. Sting, c. 6. Total length, 38-40. Chelicera (hand), 2. Pedipalp, trochanter, 2; femur,
4-5; patella, 4; palm of hand, 3-5; fingers, 4-5 ; total length (excluding coxa), 18-5.

Description. The body

Carapace. PL 51, figs. 2, 3, text-fig. 14. The exact outline of the carapace is difficult to
make out, the sides being in places distorted or broken away. It is best seen in the 1st

302

PALAEONTOLOGY, VOLUME 3

Marco-cast. It was probably almost as wide as long if allowance be made for arching,
with rounded antero-lateral corners and the anterior side slightly emarginate, but it is
possible that there was a median anterior triangular process, but if so it is now broken
and distorted. Probably the carapace sloped sharply downwards at the front and sides.
The median eyes are represented by one well preserved on the right and the other
crushed. They are situated, at about one-third of the carapace-length from the front,

FIG. 14 fig. 15

text-fig. 14. Buthiscorpius major sp. nov. G.S.M. Za 2926. Dorsal view of dorsal surface as now
visible in the two Marco-casts, with a diagram section along line ab to show the cast of the inside of the
overlapping tergites. For key to abbreviations see p. 277.

text-fig. 15. Buthiscorpius major sp. nov. G.S.M. Za 2926/1. Details of L. chelicera; a, in dorsal;
b, in ventral view, art, articulation, ? trichobothria shown as rings.

on a slight eye-tubercle, on the hinder side of which are several large granules. The
tubercle drops away behind into a forwardly bifurcating median groove flanked by two
ridges which also bear granules. On the left, half-way between the eye-tubercle and the
lateral margin, is a group of four or five larger granules in the form of knobs in the
Marco which might be mistaken for lateral ocelli were they in the usual position. Other
rather smaller knobs also occur on the antero-lateral margin. None of the knobs, how-
ever, are large enough or round enough to warrant the assumption that they are lateral
eyes; for we have casts of true ones for comparison in Compsoscorpius elongatus Petr.
(B.M. 1.15862, figured by Petrunkevitch 1949, figs. 148, 150).

Mesosomatie tergites. PI. 51, figs. 1-3, text-fig. 14. The general shape of the tergites
is that normally found in Carboniferous ‘scorpions’, as fully described on p. 320.
Hardly any of the chitin has survived the etching, but PI. 51, fig. 1, shows by dark
patches that it had been broken into a mosaic, as in Lichnophthabnus (Part 1, p. 270).
There was little ornamentation except for some flat mucronate tubercles of dark-brown

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’ 303

colour along the posterior margins which are best seen in PI. 51, fig. 3. Each tergite is
defined by a slender linear margin at the front and sides, and by a sharp infold (doublure)
behind, which can best be seen on Txi in the 2nd Marco-cast (fig. 2). The arrangement
is illustrated by the section AB on text-fig. 14. In Tvii-Tix there is a median depression
just behind the anterior margin (as there is in many Recent forms). The tergites are
connected to the head in front, to the tergal plate xm behind, and to one another by
strips of intersegmental skin that appear rather narrow because of the partial telescoping
of the segments. It will be recalled that the corresponding strips are fully exposed in the
Birmingham specimen of B. buthiformis which accounts for quite a considerable increase
in its length (p. 286; see also p. 320). Bits of the pleural skin connecting tergites to
sternites and showing the linear margins are mounted on Slides Za 2926/8-/11.

Metasomatic segments. PI. 51, figs. 1-3; text-fig. 14. The tergal plate of the 1st meta-
somatic segment (Txiii) is imperfectly preserved, but appears to have been of the normal
tapering shape. Two slight ridges may represent dorsal keels on either side of the mid-
line, but they are much less prominent than those on B. buthiformis (PI. 46, fig. 2).

The caudal rings of the tail (Cxiv-xviii) appear unusually broad in relation to their
length as a result of flattening and perhaps because the tail seems to have been more or
less turned on to its side. Some show longitudinal ridges, but little detail can be made out,
even where the chitin has been exposed by the etching (in 1st Marco-cast, PI. 51, fig. 3).
The last ring may be imperfect, but appears to have been not much longer than the
preceding one and in this differs considerably from its opposite number in B. buthiformis.

The sting was long and curved, but only the actual aculeus and a posterior portion of
the poison-capsule lay in the piece of rock that had been collected. These bits have been
left attached to the 2nd Marco-cast just as they emerged from the etching (PI. 51, fig. 2).
In text-fig. 14 the sting is indicated in the position it occupied relative to the last caudal
ring as was determined from a comparison of the two Marco-casts.

The appendages

Cheiicerae. PL 52, figs. 1,4; text-figs. 14, 15. The hand of the left chelicera is preserved
intact on Slide Za 2926/1 and the broken coxal joint and bits of the right hand are
probably present in Slide (2. The left hand is c. 2 mm. long and 1 mm. wide in the
crushed state. The fixed finger bears two prominent teeth on one of the two biting edges
which in life converged from the broad base of the finger to its apex. The free finger is
bifid, the two branches closing one on either side of the fixed finger, and each carrying
two flatfish teeth. The whole structure must have functioned as a perfect crusher, and as
such has been handed down unaltered to many genera of Recent scorpions. (I have
found that the free finger is bifid in all the Recent scorpions which I have examined, but
I have never seen this character mentioned in descriptions or diagnoses or shown in
illustrations.) In Carboniferous times, as today, the cheiicerae not only crushed, but
helped to strain off any solid particles as the juices were sucked in, as is evidenced by a
conspicuous group of small setae on the sides and bases of the fingers (PI. 52, fig. 4).
There were also a few sensory bristles attached to prominent, but small, hair-bases, two
of the latter being visible in the photograph (tri). See also text-fig. 15. Recent scorpions
retain the same general equipment of hairs and bristles.

Pedipalps. PI. 52, figs. 1-3, text-fig. 16. The two pedipalps were extracted whole and
free of matrix (Slides Za 2926/1 and / 2). They lie neatly flattened, but curved as in life,

304

PALAEONTOLOGY, VOLUME 3

with the movable finger on the outside. Neither show the coxal joint, though possibly
this forms part of the opaque debris that covers the trochanter and base of femur in
Slide /I (PI. 52, fig. 1), but this is very obscure. The trochanter is short and wide; the
femur has two small knobs on its inner side and displays numerous small hair-bases ;
the patella (tibia) is longer, with a large ‘stop-knob’ on its inner side (this is a con-
spicuous feature in many Recent scorpions), and has a few small hair-bases at its proxi-
mal end. Both femur and patella have tuberculated ribs from which the knobs project.

text-fig. 16. Buthiscorpius major sp. nov. G.S.M. Za 2926/1. L. chelicera and pedi-
palp drawn to same scale (cf. PI. 52, fig. 1 ) showing distribution of granules (solid
dots) and hair-bases (rings), some being ? trichobothria. The sketch does not differen-
tiate between features on either side of the flattened skin. For key to abbreviations

see p. 277.

The hand is long and slender, and nearly half the length of the whole limb (excluding
the coxa). The free finger is rather more than half the total length of the hand, is not
hooked at the end, and is a little shorter than the hooked fixed finger, if we measure
from the apex of the angle between them. There is a conspicuous thickened process at
the articulation. The biting edges are each marked by a continuous single row of
granules with a number of isolated larger granules at intervals to the side of the main
row. These are easily seen on the fixed finger (PI. 52, figs. 1, 3; text-fig. 16), but I think
they also occur on the free finger which has been crushed somewhat. For Recent scor-
pions the arrangement of granules on the pedipalp fingers is used in classification. The
arrangement in Za 2926 can be closely matched with that characteristic of the Buthid
genus AnomaJobuthus, the Vejovid genera Vejovis and Hadrurus, and less closely with
the Bothriurid genus Jophorus (see Werner 1934, Abb. 341, 360, 361, 382).

On the great pincers there are also a few hair-bases which are almost certainly tricho-

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’

305

bothria scattered on the palm of the hand and at the base and tip of the fingers. Their
apparent distribution is shown on text-fig. 16, but it must be recalled that it is not easy
to discriminate between organs on the two surfaces of the transparent tubular hand in
its flattened state. At the tip of the fingers there are also a number of smaller hair-bases
which may have carried ordinary small setae. The distribution and type of hair were
evidently much the same as in Mazoniscorpio in which the actual hairs can still be seen
in place on the fragments of pedipalp mounted as B.U. 721 A/4, /6 (PI. 50, fig. 3).

The general shape and proportions of the articles composing the pedipalps are similar
to those figured in outline by various authors for the following genera Eoscorpius, Buthi-
scorpius, Compsoscorpius, AUoscorpius, Europhthalmus , Eoctomis, and Buthiscorpius and
Mazoniscorpio as figured here; but the fingers are longer in proportion to the palm than
in B. buthiformis, and much longer than in M. mazonensis.

Legs. Some fragments of legs isolated by solution were mounted as Slides Za 2926/3-
/7 and one bit was left in place on the 2nd Marco-cast. Most of the leg joints, however,
lie in the other half of the nodule which was not collected. The fragments show distinct
keels or ribs, and some have a number of granules and hair-bases; but all are of little
interest since they cannot be related to particular legs, with the possible exception of
the two joints on Slide 13 which appears to be the coxa and trochanter of the 3rd right
leg (PI. 52, fig. 5).

Unidentifiable ‘scorpion’, G.S.M. Za 2924

Plates 53, 54; text-figs. 17-19

Remarks. This is labelled ‘? Scorpio, ? Shipley Clay Pit’. The horizon in the Ammanian Coal Measures
from which it is said to come lies below the Top Hard Coal in the Shipley Clay Pit, near Ilkeston,
Derbyshire.

Owing to the absence or imperfect preservation of all the diagnostic parts, it is not possible to make
even a generic identification of this fossil, but the stemites are not markedly lobed, and for this reason
it falls into Pocock’s group Orthosterni.

G.S.M. Za 2924 is the much distorted, crumpled, and partly dismembered remains of a large
scorpion (PI. 53, fig. 1). The length of the body, without the tail, and in its broken and crushed condi-
tion, is about 20 mm., and the maximum width about 8 mm.; but I consider that originally the body
length was probably 27-30 mm. If this is correct — and my view is borne out by the very large size of
the appendages — the animal was half as big again as Buthiscorpius major. It has been preserved with an
infilling or reinforcement of kaolin plus a good deal of crystalline iron pyrites on and inside many of
the organs. As a result, some of the leg segments retain their original uncrushed shape with claws and
hairs standing out in life-like menacing attitudes (PI. 54, figs. 4, 5), others have one joint bent sharply
back on the next (PI. 54, fig. 8), while the pecten still displays its individual teeth flexed from the plane
of the rachis (like the barbs of an ostrich feather) and arranged en echelon along it (PI. 54, figs. 1-3).
Lying across the scorpion were a number of twigs of Asterophyllites, also preserved uncrushed and
retaining their original shapes by reason of a kaolin infilling of the pith-cavities on the outside of
which is a mere film of coaly material (PI. 54, fig. 1). All these factors and the large size contributed
greatly to the difficulty of extraction and mounting. In all, thirty-two separate mounts were made; but
the majority consist of fragments, the positions of which on the body are unknown or only known in a
general way.

Owing to the flexed posture of the appendages and to distortion of the body almost every part of the
skeleton had been broken through when the nodule had been split open. Most of the left pedipalp,
except the hand, lay exposed (PI. 53, fig. 1), but during etching the only appendage extracted in any-
thing like its entirety was the comb of the pecten which was found completely detached from the body,
and with the end of the ? 2nd L. leg lying across it (PI. 54, figs. 2, 3). As solution of the matrix pro-
ceeded, various pieces of the legs appeared and were recovered, but it was almost impossible to relate

306

PALAEONTOLOGY, VOLUME 3

them to pieces lying at other levels. PL 53, fig. 2, and text-fig. 18a-c show the general shape and relative
positions of the fragments at three stages and as finally interpreted. I would have been well advised to
cease the development at the stage shown in the progress photograph, PI. 53, fig. 2, but I decided to try
to mount all parts in order to be able to examine them more easily and, where possible, by transmitted
light or from both sides. Unfortunately, while trying to perfect the preparation containing the three
sternites that are shown on the progress photograph, I dropped it. The bits that I rescued and mounted
are useless. ‘Striving to better, oft we mar what’s well.’

There are two Marco-casts of Za 2924. The first was the outcome of the failure of the liquid plastic
to adhere to the exposed surface of the fossil. The nodule was freed from this mount (first Marco-cast),
and a fresh start was made. The second Marco-cast shows what was revealed after etching away the
nodule, but a good deal of the chitin that had originally been exposed on the specimen had adhered to
the first Marco-cast which therefore gives, on the whole, a better record of what was originally observable.

Description. Dorsal surface of Prosoma and Mesosoma

Carapace. Only about one-third of this was preserved, and since the fracture ran
diagonally from the right antero-lateral almost to the left postero-lateral corner, no
eyes or eye-tubercle can be seen. The margin and bordering intersegmental skin are well
displayed.

Tergites. As originally exposed (PI. 53, fig. 1) it was difficult to define the number and
dimensions of the tergites, some of which, particularly Tvm and Tix, were badly twisted
into a sort of hump (text-fig. 17, h). After development, however, the second Marco-cast
shows six tergites increasing slightly in length from Tvn to Tix, and then becoming con-
siderably longer (as is normally the case). Part of the difficulty of interpretation appears
to arise from a distension of the body, which has drawn out the usually infolded inter-
segmental skin (isk on text-fig. 17). This effect can be seen particularly well in front of

EXPLANATION OF PLATE 53

Figs. 1-3. ‘Scorpion’, indet., G.S.M. Za 2924. Dorsal aspect, as received; photographed under alcohol.
X 3. 2, The same after etching, in ventral view, photographed under water. Impression of pedipalp
on Marco-cast outlined. x4-3. 3, Fragment of a tergite of the same. Slide Za 2924/32. x86.

Fig. 4. ‘Scorpion’, indet. ? Metatarsus with two large spines, tibia and part of patella of ? 1st R. leg
(cf. text-fig. 25). M.M. Slide L. 8194/2. X46.

Fig. 5. ‘ Glyptoscorpius ’, Calciferous Sandstone, Cementstone Group, Newton Farm, Foulden, Ber-
wicks. B.M. In. 25982. Structureless skin of a barbed tooth or filament on a comb which exactly
matches Peach’s pi. 29, fig. 17 in Trans. Roy. Soc. Edinburgh, vol. 30, p. 188. The irregular pattern
is due to bits of adherent matrix. X 165.

Fig. 6. ‘ Glyptoscorpius ’, Calciferous Sandstone, Glencartholm, Dumfriesshire. B.M. In. 42706. Peg-
organs on an unbarbed tooth or filament of the comb. X 165.

For key to abbreviations see p. 277.

EXPLANATION OF PLATE 54

Figs. 1-9. ‘Scorpion’, indet., G.S.M. Za 2924. 1, The ends of the two combs in ventral view, partially
etched, with twigs of Asterophyllites lying across them. X c. 4. 2, End of ? 2nd L. leg with two
tarsal spurs and setae, lying across and below the crumpled L. comb; and the R. comb with a dis-
placed tooth. Dorsal view. Slides Za 2924/1, /2. x8. 3, Ditto, in ventral view. x8. 4, End of 1st
L. leg with the sharply flexed patella concealing part of the tibia. Slide Za 2924/6. x 8. 5, The distal
part of the same to show spines on metatarsus and denticulation of the tarsal spur and claws. X 24.
6, 7, The two sides of tarsus and metatarsus of 4th L. leg with tarsal spurs. Slide Za 2924/3. x8. 8,
The tarsus, metatarsus, and part of tibia with large metatarsal spur, ? L. leg 3. The tarsal spurs not
shown in this view. Slide Za 2924/4. x 8. 9, Structureless part of the chitin of a tergite with very
minute hair-bases. Slide Za 2924/11. x85.

For key to abbreviations see p. 277.

Palaeontology, Vo I. 3.

PLATE 53

WILLS, indet. ‘scorpions’ and ‘ Glyptoscorpius

Palaeontology, Vol. 3.

PLATE 54

WILLS, indet. ‘scorpion

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’

307

Tvn and behind Txii. The latter gives the semblance of an additional tergite (see p. 321).
In Slide 31 a fragment of the posterior part of a tergite with its border (doublure) of
intersegmental skin is preserved in full relief which shows the outer surface and doublure
joining at an angle of nearly 90°.

text-fig. 17. G.S.M. Za 2924. Outline of parts originally exposed or revealed on
Marco-cast after etching. Grid of f-inch squares as on text-fig. 18. H, apex of the
‘hump’ of crushed tergites; isk, inter-segmental skin. For key to other abbreviations

see p. 211 .

The chitin of the tergites is well preserved in places, and a fragment mounted as
Slide 32, PI. 53, fig. 3, shows a few hair-bases on a dark-brown sheet having a distinctly
cellular texture which is closely comparable with the ‘reticulate structure’ of a tergite of
Pareobuthus salopiensis (Wills 1925, pi. 3, fig. 17), and of Lichnophthalmus pulcher (Part

I, p. 271). Elsewhere, however, bits of a very thin yellow-brown material (perhaps inter-
segmental skin or possibly an outer surface-film of the chitin) has adhered to the Marco-
casts, which, apart from a few minute hair-bases, appears to be quite structureless (Slide

II, PL 54, fig. 9).

308

PALAEONTOLOGY, VOLUME 3

Ventral surface of Mesosoma and segment xm

The only parts of the ventral body-skin recognized were the two sternites and ? ventral
plate of segment xm to whose unlucky fate reference has already been made, and some
scraps of one or two caudal segments (Slide 18), of which nothing can be deciphered
with certainty. The three ventral body-plates were fairly well exposed at one stage in
the development (PI. 53, fig. 2; text-fig. 18b). At the time I formed the opinion that they

text-fig. 18. G.S.M. Za 2924. Stages in etching from the ventral side. Same grid as in text-fig. 17-
A, Early stage with the combs and distal articles of the L. legs exposed, b. Final stage with ? sternites
xi, xii, and sternal plate xm, proximal parts of three legs, and patella and tibia of two. Dorsal features
dotted in. c, Reconstruction. L. appendages with distal parts restored to the proximal ones; alternate
appendages with fine and course stipple. Broken lines, ventral sclerites and R. leg fragments; dotted
lines, dorsal sclerites. For key to abbreviations see p. 277.

were sternites x, xi, xii, but judging from their position, as now known, in relation to
the tergites (indicated on text-fig. 18b), they would appear (unless they had been dis-
placed backwards during fossilization) to be sternites xi and xii and the ventral plate of
segment xm. Towards the middle of each plate the posterior margin appeared to be
slightly notched, and the skin was folded in such a way that the development photo
(PI. 53, fig. 2) gives the impression of a median suture comparable with that on sternite x
in Pareobuthus salopiensis (Wills 1 925, fig. 2 and pi. 3, fig. 2) and Mazoniscorpio mazonensis
(PI. 50, fig. 2). I made a note at the time that the L. half (right in photo) of the fore-
most sternite was a double structure or folded on itself. Owing to the loss of the prepara-
tion all that can be stated with certainty is that none of the posterior margins were
strongly lobed, but gently emarginate, that each overlapped the next one behind by an
infold of the intersegmental skin, and that no stigmata occurred on the outer surface of
the sternites themselves. Nothing that was seen would rule out the possibility that there
were stigmata on the marginal bend-over or on the intersegmental bordering skin or
doublure at the postero-lateral corners as in the Triassic scorpion Mesophonus (Wills

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’ 309

1947). On the other hand, comparison with Mazoniscorpio appears to be so close that the
two animals may well have had the same type of breathing organ.

The appendages

As the coxae (except that of the pedipalp), the sternum, and the genital operculum
are missing, the relations of the appendages (including the pecten) to the body cannot
be stated. Those of the right side are too poorly preserved to be identified: those of the
left side appear to have been detached during burial, but without much displacement.
It proved difficult to trace the connexions between the distal joints of the legs which
were revealed and isolated during the early stages of etching and the proximal joints
lying at lower levels. It is now clear that in most cases the intermediate joints lay in the
counterpart, which was not collected. In text-fig. 18a, b, the joints of the appendages
exposed at two stages of etching are shown in approximately the correct positions in
relation to a common grid, and in B in relation to the dorsal organs and the left pedi-
palp. Text-fig. 18c is an attempted reconstruction of the L. side, with alternate appen-
dages lightly and heavily stippled.

The chelicerae were very obscurely exposed and nothing was discovered about them
during development.

The left pedipalp with the exception of most of the hand was exposed on the original
surface in poor preservation, and etching revealed nothing further except strong granules
on the femur. The appendage was a powerful one with the relative sizes of the joints
much the same as in Mazoniscorpio and Buthiscorpius major.

The legs. Parts of all four left legs were isolated and mounted. They are almost un-
crushed and in several cases have the joints sharply folded, one on the next, in the
natural flexed postures of death. In many cases keels, spines, spurs, claws, and bristles
project in their original shapes and positions (PI. 54, figs. 4-8; text-figs. 18, 19b-e). One
cannot fail to notice the resemblance of the fossils to the legs of dried specimens of large
Recent scorpions, which is so close that it is almost unbelievable that this particular
Carboniferous one was not a terrestrial animal adapted to the same general habits as
its present-day descendants.

The general layout of the legs when the various parts are restored to their original
positions (text-fig. 18c) appears to have been the same as in Buthiscorpius and Mazoni-
scorpio, namely, the anterior legs were much shorter than the posterior ones; each ended
in a double claw with a pad, and each leg probably carried two tarsal spurs: the 3rd
(and probably the 4th) leg had a large metatarsal spur. In addition the trochanters can
be seen to have been almost pyramidal in shape (Slide 19), the tibias in some cases were
ribbed and spiny, and those of the 3rd and 4th legs were very spiny at their distal ends
(Slide 4, PI. 54, fig. 8; text-fig. 19e). Several other articles also show that they too were
strongly ribbed, often with short spines. The distribution of hair-bases indicates that
many parts carried bristles, a number of which are preserved.

All the above features can be matched in large present-day scorpions, but there are
certain minor differences in the structure of the claws and spurs.

(a) There are denticles along the inner edges of the claws (well seen in Slide 6 where the chitin is
undamaged and unobscured by pyrites, PI. 54, fig. 5; text-fig. 19d).

(b) Also there are small denticles or spines on some of the tarsal spurs and on the one metatarsal
spur found (text-fig. 19e).

B0612

X

310

PALAEONTOLOGY, VOLUME 3

Both types of denticles are unknown in Recent scorpions, but occur in all the Carboniferous ones
here described, except B. buthiformis.

(c) The pad (Gehstachel) of the claws of the ? 2nd leg (text-fig. 19c) appears to have been a conical
structure not unlike the shortest type of ‘dagger’ in Lichnophthalmus (Part 1, text-fig. 8). It has a spine
on one side that may correspond to one of the ‘platform spines’ of that genus. On the other hand, the

fas

text-fig. 19. G.S.M. Za 2924. a, Slide 2924/2. The tip of R. comb. Plain areas, kaolin; stippled,
chitin. The sensory field shown where visible, b, Slide 2924/1. End of 2nd L. leg with the metatarsus
restored to its original position. Note uncrushed tarsus and setae still projecting from the skin at
right angles, c. Slide 2924/1. Claws and pad ( Gehstachel ) of 2nd L. leg drawn looking down on the
conical pad, with a possible platford spine, d, Slide 2924/6. End of 1st L. leg, to show denticles on the
two claws and on one of the tarsal spurs, and details of setae and spines, e, Slide 2924/4. Parts of tibia
and metatarsus of ? 3rd L. leg, to show the spiny end of tibia and slightly denticulated metatarsal spur
of triangular cross-section. For key to abbreviations see p. 277.

pad on the ? 1st leg (text-fig. 19d) has no spine and in general resembles a large Gehstachel of a Recent
scorpion.

{d) There are minute sensory hairs on one of the tarsal spurs on Slide 6 while the other smaller spur
seems to be quite bare. Hairs are, I believe, unknown on the spurs of Recent forms, but have been
seen on spurs and/or claws in Pareobuthus (Wills 1925, text-fig. 3a), Mazoniscorpio (PI. 50, fig. 4),
G.S.M. Za 2925 (text-fig. 22), and in Lichnophthalmus (they are not referred to in my description in
Part 1).

(e) This animal may have had sensory hairs on its claws, since all the forms mentioned above have
them, but the tips of those extracted are either broken or heavily encrusted with pyrites, and no hairs
or hair-bases can be seen.

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’

311

The pecten (PI. 54, figs. 1-3). Both combs of the pecten were preserved, detached from
the rest of the body, and in a somewhat crumpled state, yet with the teeth curved and
overlapping as in life. The chitin has been reinforced by kaolin which fills the inside of
the teeth and preserves their original shape, but it is in many places encrusted by pyrites.
One comb was partly obscured and distorted by the end of the 2nd L. leg which lay
across it (PI. 54, fig. 2). Both were slightly damaged in mounting (Slides 1, 2), but the
structure of the dorsal and ventral sides are excellently shown by the several fragments,
and the whole can be rearranged, as has been done on PI. 54, figs. 2, 3.

The combs were large, each measuring at least 5 mm. in length. The rachis of each is
somewhat folded and broken. It is impossible to see its shape, but the chitin of which it
is composed consists of irregular polygonal areoles of thicker or darker skin, sometimes
carrying a hair-base, surrounded by thinner or lighter skin. This is exactly the arrange-
ment on parts of the rachis of Pareobuthus (Wills 1925), Lichnophthahnus (Part 1), and
Buthiscorpius (above), and of the other Carboniferous ‘scorpions’ mentioned in dis-
cussing the comb of Lichnophthalmus in Part 1, p. 274; and it would appear to be the
structure depicted by Petrunkevitch’s sketches of Eoscorpius typicus (Petrunkevitch
1913, fig. 7) and of Isobuthus rakovnicensis (Petrunkevitch 1953, fig. 19). The arrange-
ment is analogous to the papillae carrying hairs on the combs of some Recent scorpions.

The teeth are long and, being preserved in the round, appear unusually narrow when
compared with others described in this paper that are crushed. At least fifteen are
attached obliquely to each rachis with a row of lappet-like fulcra covering the attach-
ment. The fulcra and teeth alternate, each of the former overlapping the two half-bases
of adjacent teeth (text-fig. 19a). A sensory field of peg-organs, as described in Lichno-
phthahnus (Part 1, p. 275), can be recognized on the teeth where the chitin is unob-
scured by pyrites. The kaolin infilling of the teeth bears minute polygonal markings,
presumably recording a cellular inner layer of chitin patterned by the nerve-endings
that supplied the peg-organs. (It may, however, only reflect the microcrystalline structure
of the kaolin.)

Unidentifiable ‘scorpion’, G.S.M. Za 2925

Plate 55; text-figs. 20-22

Remarks. This specimen, as received, bore the label ‘? Anthracoscorpio, Trowell Colliery’, and the
Survey catalogue states that it came from above the Kilbum Coal which is taken as the base of the
Ammanian in the Nottinghamshire Coalfield.

The specimen originally showed a dorsal view of ? a fragment of the carapace, the six tergites of the
mesosoma, the dorsal plate of segment xm, and some bits of appendages of a largish scorpion, but it
had been broken off obliquely so that most of the carapace and parts of the first three tergites and the
whole tail had been lost. On both sides of the animal lay long fragments of some plant (PL 55, figs. 1, 2).

I etched it and extracted a number of pieces now mounted as Slides 1-13 (Za 2925/1-/13), but the
excellent Marco-cast that I had obtained was accidently burnt. As a result there is nothing except two
photographs against which the following observations can be checked. Plate 55, fig. 2, taken in air
shows more clearly the segmentation, whereas fig. 1 records better the distribution of the remnants of
chitin (dark patches).

Although parts of the sternites, of the pecten, and of three or four legs were etched out and mounted,
and now show interesting details described below, the absence of the prosoma with the carapace,
sternum, and coxae makes exact identification impossible.

312

PALAEONTOLOGY, VOLUME 3

Description. Dorsal surface of the abdomen

Tergites. Behind a fragment of the carapace the six tergites of the mesosoma with
tergal plate xm (PL 55, figs. 1, 2) measured c. 20 mm. in length. Each tergite was bounded
by a linear anterior margin and an ill-defined posterior margin with infolded doublure
(text-fig. 20). Their greatest width at Tx is about 13 mm. The last tergite and the front
of the tergal plate xm are each c. 10 mm. wide and the hind end of the latter is c. 6 mm.
across. The animal had been squashed flat, and for that reason the fossil appeared to be
broader than it was in life, and wider than normal in relation to its length.

Ventral surface of the abdomen

Sternites. The ventral skin had been pressed against the dorsal during fossilization,
but only the impression of the linear anterior margins of the four sternites ix-xii and of
the sternal plate of segment xm are traceable on the photos (PI. 55, fig. 2), on which text-
fig. 20 has been based. Two fragments believed to be parts of sternites were recovered
(Slides Za 2925/6, /10). They are made of excessively thin skin which in parts is in two
superimposed sheets. One, presumably the exterior, is uniformly transparent and carries
along its edge a few diminutive granules and one or two tiny hair-bases; the other is
blotchy. The supposed exterior sheet has a narrow doublure (? posterior margin), where
it is seen by itself in Slide Za 2925/6 (text-fig. 21a), but in the double part the doublure
seems to merge into the internal sheet. To the right (in the figure) the doublure and the
internal sheet appear to separate and form a possible stigma leading into the space
between the sheets — an arrangement reminiscent of the stigmata on the doublure of the
sternites of Mesophonus (Wills 1947, pp. 26-35). However, the specimen is very frag-
mentary and nothing reliable can be deduced from it. The same applies to the second
mount (Slide Za 2925/6; text-fig. 21b) where a somewhat similar structure can be made
out at both ends. It must be recalled that at the postero-lateral corners of a sternite there
is a complex concertina folding of the intersegmental and interpleural skin, and it may
well be that this is what we see in these two specimens.

Prosomatic appendages

Legs. Parts of two legs lay exposed on the left side of the specimen as received (PI. 55,
figs. 1,2; text-fig. 20), but it is impossible to say with certainty to which legs they belong.

EXPLANATION OF PLATE 55

Figs. 1-9. ‘Scorpion’, indet., G.S.M. Za 2925. 1, As received, photographed under alcohol. The dark
patches are chitin. x6-6. 2, As received, photographed in air, features accentuated in ink (cf.
text-fig. 20) — continuous lines, anterior margins of tergal plates Tvn-Txm; dotted areas, posterior
marginal infolds of carapace and tergal plates; broken lines Six-Sxm, supposed anterior margins
of sternal plates. X 6-6. 3, Postero-ventral view of L. comb with rachis and teeth outlined in ink.
x 20. 4, Tip of the last tooth of the comb, showing sensory field of peg-organs on one side only.
Slide Za 2925/5. X 290. 5, Sensory hair (? trichobothrium) attached to hair-base, and a second hair-
base on the other surface of the rachis. Slide Za 2925/5. X 290. 6, Metatarsus, tarsus and claws of
2nd or 3rd leg. Slide Za 2923/3. X 20. 7, End of tibia with metatarsal spur, metatarsus, tarsus,
and claws (one detached) of ? 3rd leg. Slide Za 2925/4. X 20. 8, End of tibia with metatarsal spur,
metatarsus, and tarsus with claw-lobe but minus claws; ? 4th leg. Slide Za 2925/2. X 20.

For key to abbreviations see p. 277.

Palaeontology, Vol. 3.

PLATE 55

TNUl

siM

WILLS, indet. ‘scorpion'

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’

313

text-fig. 20. G.S.M. Za 2925, as received. The tergites defined
by linear anterior margins and by posterior infolds (stippled).
Impressions of the anterior margins of four sternites and sternal
plate xm indicated by dots. For key to abbreviations see p. 277.

text-fig. 21. G.S.M. Za 2925. Fragments of supposed sternites in plan with tentative sections in
which the upper layer, as drawn in plan, is shown to the right. Unstippled, one layer; light stipple, two
layers; heavy stipple, three or four layers, a, Slide 2925/10. b. Slide 2925/6. (The long strip on the left

has been restored to its original position.)

314

PALAEONTOLOGY, VOLUME 3

The hinder one shows a robust trochanter and elongated femur and patella, each with
two longitudinal keels. It may be the 2nd or 3rd leg.

In the process of etching the distal ends of three legs were extracted, all with the
chitinous skin transparent and virtually free from matrix — features that permit a very
accurate and detailed description.

(i) The specimen on Slide Za 2925/3 probably belongs to the 2nd leg. It has been
completely flattened and consequently all the parts appear wider than they were in life.
The structure of the tarsus and claws, and the end of the metatarsus with its spurs are

text-fig. 22. G.S.M. Za 2925. A, Slide 2925/3. Metatarsus with two tarsal spurs (one denticulate),
tarsus and denticulate claws carrying ? trichobothria, and a massive pad; ? 2nd L. leg. b, Slide 2925/2.
End of tibia with metatarsal spur, metatarsus with two tarsal spurs (one denticulate), tarsus with claw-
lobe, and a sinus accommodating the pad (the claws were lost), c, Slide 2925/4. Spiny process on a
keel of the metatarsus of 3rd or 4th L. leg; cf. PI. 55, fig. 7, spi. Highly magnified. For key to other

abbreviations see p. 277.

shown in perfect detail (PI. 55, fig. 6; text-fig. 22a). The metatarsus ends in a rather
spiny margin which has a distinct lobe on the outer side: on the inner side, on inter-
segmental skin are two tarsal spurs of which one at least has a small spine-like denticle
on its side. The tarsus has two spiny ribs and two hair-bases (? trichobothria). Its outer
side is produced into a prominent claw-lobe, as it is in many Recent scorpions. The leg
ends in two claws, each of which is inwardly curved and sharp-pointed, with two denticu-
late keels that converged from the base towards the point. Near the tip each claw shows
a single hair-base defined by a strong annular thickening which suggests that it carried
a sensory hair (? trichobothrium). At the base of the claws is a large rather blunt pad
articulated in a sinus on the inner edge of the tarsus-end, i.e. on the opposite side to the
claw-lobe.

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’ 315

The general shape of the claws and tarsus can be matched in Recent scorpions, but
the denticulation and ? trichobothria on the claws and on the spurs are traits that can,
apparently, be expected on any Carboniferous ‘scorpion’ except those of small size.

(ii) The specimen on Slide Za 2925/2 lay farther back and nearer the body than did
(i), and since it has a metatarsal spur, it is almost certainly the end of either the 3rd or
4th leg (p. 289), and in view of the length of the joints it may well be the 4th. It shows
the end of the tibia and the next two articles (PI. 55, fig. 8; text-fig. 22b). The tibia had
two strong keels and ended distally in a crenulate margin. It carried a rather blunt
metatarsal spur. The metatarsus had two keels marked by rows of small spines. Its
distal end is less strongly crenulate than the tibia, but has two rather blunt lobes. On the
adjacent connecting skin are the two tarsal spurs, one of which has distinct spine-like
teeth on its edge. The tarsus also had two slight keels with spinelets and one or two hair-
bases, and it shows very clearly on the outer side a fat rounded claw lobe with three
hair-bases (? trichobothria). The pad of the claws is in position, but the two claws
which were seen at one stage in the etching were lost.

(iii) The third specimen, on Slide Za 2925/4, PI. 55, fig. 7, also shows a metatarsal
spur, and is therefore to be regarded as the end of either the 3rd or 4th leg. It is less
crushed than (i) and (ii), but the metatarsus is twisted on itself, and all three joints meet
in sharp angles. Originally it lay across the body and well behind the others. Only the
distal part of the tibia is preserved and this is partly obscured by matrix. At its termina-
tion, which is toothed rather than crenulate, are two sharp-pointed structures that were
probably the ends of two keels. Attached to the intersegmental skin there is a strong
metatarsal spur with a few spine-like teeth on its edges. The metatarsus shows two keels
(? three), one on an edge and one standing out boldly in relief. Both keels carry sharp
spinelets which culminate at the distal ends in sturdy spines on the margin which is also
spiny elsewhere. One of the keels ends in a remarkable large spiny process sketched in
text-fig. 22c. This has almost parallel sides, with spiny edges, and its end is obliquely
truncated and has two terminal spinelets. It can hardly be seen in PI. 55, fig. 7, because
it is masked by one of the two tarsal spurs. So far as my experience goes it is a unique
structure. The tarsus, as preserved with two or possibly three keels marked by spinelets,
appears to broaden distally to end in a sharp claw-lobe carrying two or three hair-bases
(? trichobothria). The apparent shape of the segment and of the lobe is the result of the
crushing, and originally the latter must have resembled the corresponding organ in (ii).
The claws were broken in the mounting: one is still attached and the other floated away
as a tiny fragment (inset on fig. 7) which, however, shows the two rows of teeth, exactly
as in (i). Each claw shows a single hair-base (? trichobothrium).

Mesosomatic appendages

The pecten. One branch of the pecten, about 4 mm. long, was etched out, but details
of the rachis can only be seen in a few places owing to adherent matrix. It appears to
have been small and narrow, with an oblique proximal articular end, and to have been
covered with minute bristles (PI. 55, fig. 5) attached to conspicuous hair-bases that were
in some cases sited on areoles of darker skin. No subdivision of the rachis can be seen.
Along one side is a row of at least seventeen overlapping teeth attached obliquely to the
rachis, the last of these being the terminal one. The teeth stand out with intervening
spaces in a remarkable way which is easily seen with a binocular microscope, but are

316

PALAEONTOLOGY, VOLUME 3

difficult to photograph (PI. 55, fig. 3). Each tooth has on one side a sensory field covered
with minute thin-skinned circles (PI. 55, fig. 4). Even with a high-power objective I can-
not be certain that a papilla or peg occurs at the centre of any of these, but it is probable
that there was originally one in each, as in the other examples described.

EJmdentifiable ‘scorpion’, M.M. L.8194

Plate 53, fig. 4; Plate 56, fig. 1 ; text-figs. 23-25

Occurrence. Coal Measures, probably about 100 feet above the Arley Mine Coal, Communis Zone,
Sparth Bottoms, near Rochdale, Lancashire.

Remarks. This specimen, as received, showed the abdomen and one — probably the last — caudal seg-
ment of a diminutive ‘scorpion’. It consisted of two parts that had been glued together. The crack is
seen in PI. 56, fig. 1, and in text-fig. 23. In the hope of discovering the structure of the prosoma the
two pieces were separated and the smaller anterior portion with the remains of tergites vn to ix and
some leg segments, and (as proved later) one chelicera was treated with shellac to hold the exposed
parts together. This treatment proved a mistake, because the shellac was affected by the hot HC1 and
also was not a strong enough support. The other portion was embedded in plastic, etched in the usual
way, and can now be studied from both sides of a Marco-block, but it is by no means an easy specimen
to interpret because there is very little chitin preserved, and because the dorsal and ventral skins have
been somewhat displaced laterally and then have been pressed tightly together with virtually no inter-
vening matrix. Consequently impressions of sternites appear in the dorsal aspect and of tergites in the
ventral. On the R. side the ends of two or three sternites are folded over on to the dorsal surface with
the result that the tergites are here sandwiched between two layers of sternite (+ pleural skin). See
text-fig. 23b. It is hardly surprising that little can be learnt from the specimen about the nature of the
creature’s respiratory organs.

In the absence of the carapace, the coxo-sternal region, and pincers it is impossible to identify the
genus or species of this tiny ‘scorpion’, but nothing has been found that would preclude it from being
assigned to the Eoscorpioniidae.

Dimensions. M.M. L.8194 is part of the smallest of all the Carboniferous ‘scorpions’ that I have
developed. It is preserved with the body extended, but even so the six tergites (Tvii-Txii) and the
tergal plate xm only measure 8 mm. in length, to which must be added another 8 mm. for the tail (to
the supposed last caudal ring). Allowing 2-5 to 3 mm. for the carapace and 1-5 mm. for the sting,
I estimate the whole animal was 23-24 mm. long, but the tergites are all separated from one another
by strips of intersegmental skin that make up about half the 8 mm. length, stated above. (The specimen
provides a fine example of the difficulty of counting the number of tergites in a poorly preserved ‘scor-
pion’, see also p. 321.) The width of tergite xn, which appears to be the widest, is in its flattened state
about 3-5 mm.

The chitin of the tergites and appendages has a curious pustulose appearance due to a pronounced
cellular texture (PI. 53, fig. 4), and is penetrated by frequent pore canals. It can be compared with that
to be seen in Spongiophonus (Wills 1947, p. 2, pi. 9, figs. 8, 13), though the cell-walls are less regular.

Description. Dorsal surface of the abdomen

Tergites vii-xii and tergal plate xm (PI. 56, fig. 1 ; text-fig. 23b). Before the anterior
section of the specimen was removed it showed parts of three short anterior tergites
(vii-ix); but these, with the exception of tiny fragments (Slides L8 194/6, /? 3), were lost
during preparation. The tergites show as dark stripes on PI. 56, fig. 1 , which was photo-
graphed before etching. In the Marco-block tergites x-xn and tergal plate xm together
with the intersegmental skin can be examined in detail. The tergites themselves are seen
to be defined by linear anterior margins of thickened skin. The posterior margins are
poorly preserved; but in each case there is a distinct ridge visible in oblique lighting,

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’ 317

which represents the position of the posterior marginal fold along which the interseg-
mental skin could be turned in under the tergite when the animal was not distended (see
p. 320). The tergites are progressively longer from Tvii to Txii and, as mentioned above,
the strips of intersegmental skin are almost as long as the tergites which they separate.

The R. ends of Txii and Txm appear to have been partially concealed by the bending
over of the ends of sternites xi and xii. Txm is poorly preserved, but its anterior margin
seems to be a thick line of dark chitin. The posterior margin would appear to be ab-
normally wide, but I may have been misled in interpreting this crushed and defective
part of the fossil.

text-fig. 23. Manchester Museum L8194. a, Ventral aspect of abdomen, b. Dorsal ditto. The heavy
broken line defines the original fracture. ? bpe , possible bases of combs. For key to other abbreviations

see p. 277 .

Ventral surface of the abdomen

Sternites, text-fig. 23a, b. The sternites lie firmly pressed against the inside of the
tergites, but they can be examined by reflected light on the ventral surface of the Marco-
cast. They are unlobed, roughly rectangular plates, longer than the corresponding
tergites, with their lateral ends distinctly rounded. Here and there can be seen fragments
of the linear anterior margins. There is probably a median suture in Sx (as there is in
Mazoniscorpio , p. 298), but the appearance of one on Sxi is, I think, due to a fold or to
the impression of the right end of the sternite where it has been turned over on to the
dorsal surface (compare text-fig. 23a and b). Sxi has on the median section of its pos-
terior margin several large mucrones. The true L. ends of Six and Sx can be seen to
overhang at the postero-lateral corners, suggesting that there was a deep doublure and
pouch as in Mazoniscorpio (p. 299). Since there is no sign of any stigmata on the exposed
surface of the sternites, the above-mentioned pouches may have been connected with
respiration, as has been argued above (pp. 299, 300).

It is difficult to say whether these ventral plates are sternites ix-xii or sternites x-xn
and sternal plate xm. I favour the former explanation and consider that the three hinder
plates are somewhat displaced backwards on the L. side, so that Sxn lies partly below

318

PALAEONTOLOGY, VOLUME 3

the tergal plate, Txm. The displacement is effected by the widening out of the inter-
segmental skin behind the L. postero-lateral angles of the sternites.

Caudal segments. Only the supposed last caudal segment is preserved attached to the
Marco-cast. It is badly preserved, but seems to have been the usual cylinder reinforced
by two or more keels.

Appendages of the Prosoma

Only isolated fragments were recovered in preparing the anterior part of the specimen.
Chelicera, Slide L8 194/1, text-fig. 24. This tiny specimen shows one of the fingers of a

text-fig. 24. Manchester Museum L8194/1. Two aspects of the chelicera with an enlargement of a
spine with three spinules. Stippled areas are partly concealed by matrix.
text-fig. 25. Manchester Museum L8194/2. Probably part of the 1st R. leg, in ventral aspect. For

key to abbreviations see p. 277.

chelicera, and because it is not bifid I incline to the view that it is the fixed one. How-
ever, the part of the specimen where the palm of the hand should be found is obscured
by matrix. The finger itself is about 0-4 mm. long and shaped like the beak of a parrot.
From its sharp point descend two biting edges which diverge towards the base. Near
here are two blunt denticles on one edge: the other carries nearer the point a single
sharp spine of which only the tip is visible. Under a J-in. objective this latter can be
seen to have three hair-like spines projecting from its sides (text-fig. 24, inset).

A walking leg , Slide L8 194/2, text-fig. 25. As this small specimen is almost free from
any matrix, the two apposed layers of chitin are transparent and show the cellular
structure noted above. Parts of three joints (the four-jointed appearance is due to a fold
across the metatarsus) are preserved, which I interpret as patella (fragment only), tibia,
and metatarsus, the last ending in one very large and two smaller spines, each being an
outgrowth of the metatarsus and not a movable spur. The specimen measures less than
2 mm. and, by virtue of its small size and distinctive shape, is probably part of the first leg.

Other unidentifiable fragments from the prosoma are mounted on Slides L8 194/3, /6.

Diagnosis. A ‘scorpion’ with large tergal plate xm covered with numerous granules, the
larger ones ranged along two low dorsal keels. Sternal plate xm much larger than the

FIG. 24

FIG. 25

wattisonia gen. nov.

Type species W. coseleyensis sp. nov.

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’

319

tergal, and apparently possessing a median sinus in its posterior margin where a low
median keel ended; anterior margin linear, and crossed by a short median line from
which the median keel runs to the supposed posterior sinus.

Wattisonia coseleyensis sp. nov.

Plate 56, figs. 2-4; text-figs. 26, 27

Holorype. B.U.722A, B, and Slides W2A/1, /3; W2/2. Almost certainly from the Ten-foot Measures
above the Thick Coal, Coseley, South Staffordshire.

Remarks. This small specimen lies in two half-nodules collected by Mr. J. T. Wattison of Shrewsbury
many years ago, and generously presented to me for development. The two halves lie embedded in two
Marco-blocks which bear the labels I originally gave them, namely W2 and W2/A on 722B and 722A
respectively. Only five tergites and the tergal plate of segment xra were exposed in dorsal view in 722B
and in ventral view in 722A. As etching proceeded, 722a revealed nothing fresh except the L. end of
sternal plate xrn which had been bent over on to the dorsal side (PI. 56, fig. 2). This was detached and
mounted as Slide W2A/1 (PI. 56, fig. 4). Part of Tvm also floated free and was mounted as Slide
W2A/3 (PI. 56, fig. 3). The embedded sclerites are now completely transparent and display in B.U.722A
better than in any other specimen that I have developed how the tergites articulated with one another
and how the ‘scorpion’ could distend itself.

B.U.722B, on etching, revealed the downtumed ends of the tergites, especially the R. ends, part of
which broke away and was mounted as Slide W2/2. Also the remaining two-thirds of Sxm which had
not been bent over dorsally was exposed. This part is not fully transparent, but it is beautifully pre-
served. The shape of Sxm appears to be unique, and for this reason I propose to name this ‘scorpion’-
fragment Wattisonia coseleyensis in honour to its discoverer.

I was misled during the etching of both specimens by a laminate structure carrying minute hairs on
one side into thinking that I had found a detached sternite, but close examination later showed it to be
part of a pinnule of ? Neuropteris. A part was mounted as W2A/4 and /5.

Diagnosis. As for the genus with the following addition; small ‘scorpion’ with normal,
rather hairy tergites. Sternal plate xm thickly coated by posteriorly directed setae
attached to small hair-bases. All other organs unknown.

Description. Dorsal surface

Tergites. Parts of the chitinous skin of tergites vni-xii occur in both specimens as a
result of their complex structure which is best revealed in the external aspect etched out
in B.U.722A (PI. 56, fig. 2) and in a fragment of Tvm mounted as W2A/3 (PI. 56, fig. 3).
As these specimens throw light on the normal structure of the dorsum of Carboniferous
‘scorpions’ I describe them here in detail. From them it can be proved that each tergal
sclerite consists of the external skin of the tergite proper covered with scattered hair-
bases, and defined in front by a linear anterior margin which curves round at each end
into linear lateral margins. These disappear behind, where the posterior margin, defined
by a line of minute granules and hair-bases, begins. This margin is really part of an
extensible infold by which the external surface passes into the inturncd posterior border
or doublure. The chitin of the latter was doubtless very thin, and only in one or two
places can it be proved to exist in this fossil. In several specimens that I have developed,
this posterior infold, filled with the Marco, has been revealed by the etching. It is prob-
ably responsible for the rather indefinite, often wide, distinctive strips that indicate the

320

PALAEONTOLOGY, VOLUME 3

posterior sides of the tergites when they are exposed in dorsal aspect on the unetched
surface of a nodule (PI. 51, figs. 1-3; PI. 55, figs. 1, 2). Reverting to Wattisonia , in front
of the anterior margin is the anterior border, which in a forward tergite may be half as
long as the external surface of the tergite itself (PI. 56, fig. 3), but which is probably
relatively shorter on the hinder ones compared with the greater length of the sclerites.
The anterior border carries a line of hair-bases just in front of the anterior margin (a
similar line of actual hairs can be seen in Benniescorpio tuberculatus (Peach), text-fig.
28). Outside the linear lateral margin is the lateral border of pleural skin which also
carries hair-bases.

The specimens show that when the ‘scorpion’ was in a contracted posture, the anterior
border, the anterior margin, and the front half of the external skin of one tergite were
covered by the external skin of the preceding tergite. It follows that the doublure of the

TVIII TIX TX TXI TXH TXIII

text-fig. 26. Wattisonia coseleyensis gen. et sp. nov. B.U.722A. The structure and method of expan-
sion of the dorsum, a, Section through two complete tergites (reconstruction), ab, anterior border;
am, anterior linear margin; d posterior border or doublure; pm, posterior margin with granules.
b, Diagram section to scale through L. side of the contracted body (Tvm shown slightly displaced as
in the specimen, c, Ditto, with the intersegmental skin and tergites fully extended.

latter must have stretched forward internally to join the anterior border of the under-
lying tergite (see text-fig. 26a); but this union is nowhere visible, probably by reason of
the extreme tenuity of the chitin. Flattened as all these parts arc, except the posterior
marginal fold, the area of overlap consists of three layers — the external surface plus
doublure of one tergite overlying the front end of the external surface and the anterior
border of the next behind. Parts of several such areas show as dark strips on Tix-xii
(PL 56, fig. 2). It will be noticed that they are much broader than the lighter intervening
strips where only the external skin is present, except in the case of Txn in which the
overlap on to Txih was quite small. In a poorly preserved specimen the three-layer
strips might easily be mistaken for whole tergites, each apparently showing its own

EXPLANATION OF PLATE 56
Fig. 1. ‘Scorpion’, indet., M.M. L. 8194. As received. X6.

Figs. 2-4. Wattisonia coseleyensis gen. et sp. nov. 2, Dorsal view of dorsal surface with part of Sxm,
after etching (some features accentuated in ink). B.U.722A. x6. 3, End of ? Tvm or Tix. Slide
722A/3. x 24. 4, Setae on Sxm. Slide 722A/1. x70.

Fig. 5. Benniescorpio tuberculatus (Peach). Supposed Sxi (cf. text-fig. 28), G.S.M.E. 9675. X 12.

For key to abbreviations see p. 277.

Palaeontology , Vol. 3.

PLATE 56

WILLS, indet. ‘scorpion’, Wattisonia , and Beimiescorpio

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’ 321

anterior margin; but, as pointed out above, the margin that can be seen really belongs
to the next segment behind.

Tergite vn is missing. The others increase in length from Tviii to Txn (see text-fig.
26c), and in the amount of overlap from Tviii to Txi. Nearly half of each tergite is over-
lapped by the next one in front. The overlapping of the tergites implies a potential
increase in the length of the mesosoma (on distension) that has long been postulated to
explain the overlapping of the ventral sternites in Carboniferous ‘scorpions’. The dorsal
overlap has been noted above in several cases, but this specimen shows very clearly the
enormous amount of increase that was possible. On text-fig. 26b I have plotted to scale
a diagrammatic section of the L. side of 722A, and text-fig. 26c shows the same with
the skin fully extended (a condition doubtless never attained in life). The lengths of the
anterior borders and adjacent posterior doublures are minimal. It turns out that at the
maximum extension the length from the front of Tviii to the back of Txn was doubled.
In the past a want of appreciation of this feature may have led to misinterpretation of
poorly preserved specimens ; and in this connexion it would be interesting to etch out
the type specimen of Mazonia woodiana Meek and Worthen to determine whether the
supposed seventh mesosomatic tergite exists or whether the intersegmental skin between
the carapace and the first tergite has been
misinterpreted as an extra tergite. I have
examined a good cast of the holotype, and I am
inclined to think that this second alternative is
the explanation of the supposed abnormality
which was questioned long ago by Scudder in
Zittel (Eastman’s edition), as pointed out by
Woodward ( Geol . Mag. 44, p. 544). If the extra
tergite should prove to be fictional, another
of Petrunkevitch’s (1955, p. P70) superfamilies
and families would disappear. (Also his whole
suborder Protoscorpionina is based on the
Carboniferous Mazonia woodiana and on the
Silurian genera Palaeophonus , Dolichophonus ,
and Proscorpius, which three, he states, have
the first tergite (of the supposed seven characterizing his suborder) concealed under the
carapace. None, however, of the published figures of these three forms shows more than
six mesosomatic tergites, and the transverse mark across the hinder part of the carapace,
which is the sole evidence for his statement about the concealed tergite, can be matched
in many Recent scorpions and probably in Carboniferous ones also (p. 285). I therefore
regard as ill-founded his 1955 diagnosis of the family Palaeophonidae ‘First abdominal
tergite concealed under carapace, its anterior edge indicated by a transverse furrow’.)

Tergal plate Txm of Wattisonia is imperfectly preserved. What remains of its anterior
border suggests that it was only to a small extent overlapped by Txn. This would imply
that this articulation did not play a large part in the distension. Txiii appears to taper
backwards more slowly than is usual, but it has the normal semi-annular posterior
margin with a narrow doublure. Its surface is devoid of hairs and hair-bases, but is
ornamented with numerous, circular granules, the larger of which lie along the two
dorsal keels (cf. Buthiscorpius and Mazoniscorpio).

text- fig. 27. Wattisonia coseleyensis gen. et
sp. nov. B.U.722B. Sternal plate xra. Stippled
part is preserved in Marco-block W2. abo,
anterior border; am, anterior margin; Ik,
lateral keel; mk, median keel.

322

PALAEONTOLOGY, VOLUME 3

Ventral surface

Sternal plate XIII. No ventral organs were discovered except the sternal plate of seg-
ment xm, the structure of which appears to be unique; but it must be pointed out that
the corresponding plate in all the forms here described is in no case satisfactorily dis-
played for comparison. About two-thirds of it lies semi-transparent in the Marco-
block B.U.722B, the rest, shown in the progress photo of 722A (PI. 56, fig. 2), as it
etched out, is mounted as Slide W2A/1 (PI. 56, fig. 4).

In its flattened state Sxm is much wider than Txm, implying that in life it was strongly
arched downwards. It cannot be satisfactorily photographed, but is drawn in text-fig. 27.
It appears to have been defined at the sides by two lateral keels, the R. one of which can
be seen on the left side of 722B. In front it is defined by a linear anterior margin with
a narrow anterior border carrying a line of hairs. Across the middle of this margin is a
short thin line — probably a sharp fold of the skin — which is continued backwards as a
low median ridge that widens towards what appears to be a median sinus in the posterior
margin. The ridge was a median keel ending perhaps in a blunt spike, and the sinus may
not be original, but the result of the fracture and flattening of the keel and posterior
margin. The whole surface of the sternal plate is thickly coated with rather long setae,
many of which are still attached to their hair-bases (PI. 56, fig. 4).

SUPPLEMENT TO PART 1

LOBOSTERNI PoCOCk 1911
benniescorpio gen. nov.

Type species Eoscorpius tubevculatus Peach

Benniescorpio tuberculatus (Peach)

Plate 56, fig. 5; Plate 57; text-figs. 28, 29

Eoscorpius tuberculatus Peach 1881, p. 398, pi. 23, figs. 8, 8 a, 8 d, 8e.

Centromachus tuberculatus Thorell and Lindstrom 1885, p. 25.

Archaeoctonus tuberculatus Pocock 1911, p. 19.

Archaeoctonus tuberculatus Petrunkevitch 1913, p. 34, and 1949, pp. 138-9, figs. 138, 176.

Alloscorpius tuberculatus Petrunkevitch 1953, p. 29, figs. 27, 28.

Material. Holotype, G.S.E. 9675 (or 456), 9676 (or 457). This specimen consists of part and counter-
part preserved in grey shale with many plant remains. It was collected by James Bennie from the Coal
Measures at Blair Point, near Dysart, Fife. Petrunkevitch (1953) states that it is now the only surviving
specimen of several on which Peach (1881) founded his species Eoscorpius tuberculatus, making the
assumption of specific identity between them, though some came from the Calciferous Sandstone of
Lower Carboniferous age while this ‘the first and best specimen’ was from the Coal Measures. The
two halves are not complete mirror-images of one another owing to thin slivers of shale having been
lost. With the consent of Mr. R. B. Wilson I embedded both halves in Marco. This has made easily
visible details that could not be seen before.

EXPLANATION OF PLATE 57

Figs. 1-3. Benniescorpio tuberculatus (Peach). 1, Carapace, G.S.E. 9676. X6-75. 2, Ventral view of
external impression of carapace, and dorsal view of the rest (cf. text-fig. 28). G.S.E. 9675. x6-75.
3, Ventral view of supposed sternites xi, xn and bits of Txm. G.S.E. 9676. X 6-75.

Palaeontology, Vol. 3.

PLATE 57

WILLS, Benniescorpio

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’ 323

Remarks. When Peach described this specimen it was the first fossil scorpion to be
found in Britain. Only four others were known — two from Europe (Corda 1835; 1839)
and two from America (Meek and Worthen 1868). His reference of it to one of the
American genera — Eoscorpius — must be read in this context. Since it is now clear that
Peach's description is faulty in almost every detail (in part because it combines charac-
teristics of ‘scorpions’ of vastly different ages), I name it Benniescorpio after its dis-
coverer, James Bennie, who was a pioneer collector from the Scottish Carboniferous
rocks (Bennie 1885). Since it possesses large lobed sternites I regard it as belonging to
Pocock’s group, Lobosterni.

As pointed out by Peach, the carapace is upside down in relation to the body and lies
near the tergal plate of segment xhi. This means that the dorsal aspect of the actual
chitin of the carapace is seen on 9676, which elsewhere shows a ventral view of the ven-
tral skin or of the internal cast of that skin where the chitin has been lost. On the
other hand, 9675 displays the external cast of the carapace in ventral view, and at the
same time the dorsal aspect of the chitin of the tergites and of the sternites (i.e. their
inner surfaces), or of external casts of their outer surfaces (often with detached hairs
adherent to the shale) where the chitin has been lost. There are also other complications
resulting from folding and crumpling of the thin ventral skin and the displacement of
the sclerites: in particular it seems necessary to postulate that one half of the supposed
sternite xi has been folded dorsally on top of the other.

Diagnosis of genus and species. A rather large and relatively broad ‘scorpion’ with an
almost square (8 mm.2) carapace, having the general shape, position of median eyes,
and ornament of Lichnophthalmus pulcher Petr. ; no lateral eyes ; the rest of the body
about 20 mm. long, tergites only partly exposed but apparently normal in shape, non-
tuberculate, but with lines of setae along the anterior borders and posterior margins of
some of them; tergal plate xm tuberculate with posterior margin abnormally broad; tail
and sting unknown. Ventral organs somewhat displaced and obscure, but apparently an
oval sternum and a large bilobed genital operculum; combs of pecten large, fin-like,
with numerous teeth, and resembling those of Lichnophthalmus pulcher , Petr. Supposed
sternite xi very long and broad, bilobed, coated with hairs. Supposed sternite xii also
large and hairy, but its shape is unknown. Neither sternite has any stigmata.

Description. Dorsal organs

Carapace. This has been somewhat crushed, but the actual skin of its front part is
wonderfully preserved in 9676, and part of the right side missing from 9676 can be
restored from 9675 (PI. 57, figs. 1,2; text-fig. 29), but the hinder part has been lost.
When complete it must have been nearly square in its crushed state. The present width
in front is 8 mm. and its length is 7 mm., plus perhaps 2 mm. not preserved. The front
margin is almost straight with a slight median projection (cf. Lichnophthalmus , Part 1,
pi. 49, figs. 2, 3, 11, ap) and curves round in two semicircles at the sides. The carapace
retains sufficient relief to show that it had a strongly elevated heart-shaped cephalic part
with two anterior-lateral domes separated by a median furrow that divides in front into
two grooves which almost encircle a prominent median eye-tubercle situated very near
to the front of the shield. The tubercle carries two large, originally circular, eyes pointing
upward and somewhat forward. The eye-tubercle and the two domes probably overhung

324

PALAEONTOLOGY, VOLUME 3

the anterior margin, as suggested by Peach, but the front corners of the carapace are
perfectly exposed and exhibit nothing that can be taken for lateral eyes, which he states
to be present. The external cast of the carapace on 9675 shows impressions of several
parallel narrow ridges separated by fine linear markings. Though these lay on the
external surface they may have been related to muscle attachments on the inside. They
lie approximately on the posterior-lateral edge of the cephalic area. The postcephalic
posterior-lateral corners of the shield are crumpled or missing everywhere, except for
one doubtful fragment marked ? on text-fig. 29 which can be seen on PI. 57, fig. 1, to
show a bit of a supposed lateral margin.

text-fig. 28. Benniescorpio tuberculatus (Peach). G.S.E. 9675, 9676. Holotype in dorsal aspect
(except for the carapace), drawn in the same position as Peach’s pi. 23, fig. 8, but with features that are
only visible in 9676 restored to their true positions in relation to 9675 : light stipple, ventral body plates
with one layer of chitin preserved or as casts; dark stipple, ditto with two layers of chitin (one in-
verted); unstippled, dorsal plates, combs and external cast of carapace. Hairs (thick ends are proxi-
mal) represented here and there. Inset, diagram-section through Txi and ? Sxi as preserved in 9675
and 9676, i.e. the sternite in normal and inverted posture respectively (the outside being indicated by
the hairs), mr, median ridge. For key to other abbreviations see p. 277.

Irregularly scattered small and large granules or tubercles are seen in several parts,
the largest being concentrated on the sides of the median furrow. Their occurrence led
Peach to bestow the specific name tuberculatus on this specimen. (However, he also
described the tergites as covered with tubercles, which they are not; but he supported
his view by pi. 23, fig. 8 f, which refers to a specimen from the Calciferous Sandstone of
Redhall.) The tuberculation recalls that on the carapace of Lichnophthalmus pulcher
(Part 1, pp. 269, 271). Two long setae are seen attached, one to the eye-tubercle and one
on the R. cephalic dome.

From the above description it will be seen that Petrunkevitch’s ascription of the
specimen to Alloscorpius and his 1953, fig. 27, are erroneous, the large postcephalic area
which he shows being in reality part of tergal plate xm.

Were it possible to identify a Carboniferous ‘scorpion’ by the carapace alone, this

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’ 325

creature could with assurance be designated Lichnophthalmus pulcher Petr., though in its
crushed state the carapace gives the appearance of having been relatively narrower
behind than in that species.

Tergites vn-xn. Parts of the six tergites are preserved in chitin and are exposed in
dorsal view in 9675; and there are impressions on the 1st sternite of the linear margins
of the other ends of Tvii and Tviii, a fact which implies that they had been bent back
dorsally and then squashed on to the sternite. Peach’s description implies that the ends
of the tergites are present, but none of the lateral margins that should define them is
visible. What is exposed is the central sections of whole tergites which must have been
much wider than the parts preserved suggest. Tvii is very short — about 0-5 mm.; the
next is c. 0-75 mm., and the rest increase in length gradually to Txii which is c. 3-25 mm.
long. Though not always visible, there was certainly a linear anterior margin on each,
and the posterior margin was an infold just in front of which in two cases (probably in
all originally) is a row of setae, and on Tvii and Txii a few isolated granules or mucrones.
The rest of the tergite in each case is devoid of ornament, but there is evidence that it
was hairy in places, long setae being preserved here and there. I have already mentioned
that Peach's description of the tergites was not based upon this specimen and is quite
inapplicable to it.

Tergal plate xm. Parts of this are exposed on each of the pieces, but its outline cannot
be seen. It appears to have been at least 5 mm. long, plus 0-75 mm., the length of the
anterior border which carried a line of hairs. At right angles to this border near the
point where the latter passes into folded skin is a short keel which may mark the mid-
line. Small groups of largish tubercles seem to lie fairly symmetrically in relation to the
mid-line suggested by the keel, but there are other granules whose distribution appears
to be random. The posterior margin is indicated by a strong line. There is little to
suggest that this plate had the marked tapering that normally characterizes Txm, and
it may be inferred from this that the tail was broader than usual.

Ventral organs

These were made of excessively thin skin, that of the sternites being coated with long
slender hairs (PL 56, fig. 5) which pointed backwards, as in Wattisonia , above, p. 319.
The skin has been crushed on itself, and crumpled or lost in places, while in others shale
intervenes between the folds. This has resulted in the counterpart on 9676 not being
everywhere the mirror-image of the fossil seen on 9675. 1 have attempted to incorporate
in one text-fig. 28 details taken from both halves as they would appear in dorsal aspect on
9675 (cf. Peach’s pi. 23, fig. 8). Some of the lines on the ventral skin seem to represent the
impressions of dorsal features. For all these reasons the following description is tentative.

Supposed sternum of the prosoma (st), genital operculum {go), and sternum of the peeten
(spe.). These lie at the left front of the specimen, but are very poorly preserved. The skin
is devoid of the long hairs so conspicuous on the sternites (below). The sternum would
seem to be an oval plate passing backwards into the two halves of the genital operculum,
but the posterior part of the latter is not visible. Between these and the comb is a very
obscure area that may represent the sternum of the peeten.

Peach may be referring to the above when he states on p. 399: ‘Nothing can be said
of the first two of the ventral plates, except that one of them bears the combs.’

The peeten. Both combs of the peeten lie folded on themselves and displaced to one

B 6612

Y

326 PALAEONTOLOGY, VOLUME 3

side of the body; but, as stated above, its sternum cannot be recognized with certainty.
The specimen was the first Carboniferous ‘scorpion" to display the detailed structure of
the comb. Meek and Worthen (1868) say of the bit seen in Eoscorpius carbonarius : ‘The
single detached comb-like organ, seen lying in the matrix on the left side of the abdomen,
shows some eleven or twelve of the little laminae or divisions, but apparently had more,

as it is incomplete, at least at one end." Peach (1881, p.
399) nat urally described it fully and accurately as follows :
‘They (the combs) seen to be made up of a broad tri-
angular rachis ornamented with an irregular embossed
scale-like pattern, which reminds one of that on Eury-
pterus and Pterygotus , and edged at the lower side with
a row of comparatively large leaf-like teeth. These are
constricted at their bases, they then suddenly expand,
the sides then become parallel, and as suddenly become
truncated to a blunt point. ... In their present crushed
state the individual leaflets overlap each other like the
splints of a Venetian blind." The combs are therefore
closely comparable with those of Lichnophthalmus (Part
1, p. 274). I would add that the bosses appear to be
limited to that part of the rachis that adjoins the teeth,
those nearest the latter constituting a row of fulcra
which perhaps give the appearance responsible for the
item ‘constricted at their bases’ in Peach’s description.
Some of the chitin, including several long hairs, is preserved in 9675. The piece of comb
conspicuous in 9676 is unrecognizable in 9675.

The supposed steniite xi lies to the right of the pecten (PI. 56, fig. 5, PI. 57, figs. 2, 3;
text-fig. 28). Of this Peach says: ‘The third is an apron-like flap, narrow in front and
widening posteriorly, and rounded at the angles. It is as deep as three of the dorsal
plates, opposite which it is placed. Within the postero-lateral angles of this plate two
fine slits occur which are the openings into the air sacs.’ This last statement is not true;
perhaps he was misled into regarding hairs or a narrow seam of chitin as openings
(stigmata). I think Peach was also mistaken in other respects. Examination of both
halves and of the now transparent chitin forces me to regard his ‘apron-like flap’ as a
bilobed sternite with its two lobes symmetrically folded on to one another so that the
two lateral margins lie almost exactly superimposed where they adjoin the tergites, and
the two posterior margins likewise lie one almost above the other behind, while the
common mid-line forms the edge now adjacent to the combs. Further, about one-fifth
of the length of the sternite in front of its hind margin is another line, perhaps marking
the extent of overlap of Sxi on to Sxii. Where the skin is doubled, both this line and the
linear posterior margin are duplicated. The parts where the double structure is preserved
in chitin has a dark stipple on text-fig. 28. As I interpret it, this area represents the
remnants of the L. lobe lying turned over dorsally and seen in dorsal view in 9675.
Most, if not all, of the R. lobe is exposed in ventral view in 9676, and it is this skin
which is seen in 9675 to pass under one of the tergites (see section in text-fig. 28). I can
detect no pleural skin connecting sternites to tergites, as described by Peach, p. 399:
‘longitudinally folded thinner skin which is constricted opposite the articulations’.

text-fig. 29. Benniescorpio tubev-
cidatus (Peach). G.S.E. 9675, 9676.
Carapace, anterior part preserved
in chitin on 9676; R. posterior-
lateral part showing ? muscle scars
(ms) restored from the external cast
on 9675. ec, posterior-lateral edge
of the inflated cephalic area; ?,
doubtfully part of the carapace.

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’ 327

At the front end of Sxi are a number of lines, one of which may be its anterior linear
margin. The others are in my opinion the impressions of the margins of two or possibly
three of the anterior tergites turned over and twisted diagonally backwards.

If the above interpretation be correct, the sternite was not only abnormally long (as
pointed out by Peach), but also abnormally wide. The large size of the sternites is a
noticeable feature in other Lobostern ‘scorpions’.

Remains of a second large sternal plate, clearly displayed on both specimens (PI. 57,
figs. 2, 3), and here regarded as Sxn, lies behind the one I have just described, but the
directions in which the hairs on it point suggest that it has been displaced and twisted
round through about 120°. On one side the skin has also been bent on itself so that the
hairs on it point in the opposite direction to those on the main sheet. The original outline
of the sternite can nowhere be seen, but it must have been comparable in size with Sxi.

One area of chitin which lies ventrally below Txm, and another scrap which is below
the inverted carapace, both have hairs pointing posteriorly. It may be part of the same
sclerite.

It is a pity that so little can be deduced with certainty about the original shape and
organization of this animal, individual parts of which show so much. The carapace
agrees closely with that of Lichnophthalmus pulcher Petr., but there is no proof, only a
strong presumption that it belongs to the associated body. The pecten also matches well
that of Lichnophthalmus, and one sternite at least is bilobed; the sternites, however, have
the long hairs of Wattisonia, whereas Lichnophthalmus showed only minute spinules,
and those sited on the doublure.

The above description shows that the sternites of this creature in no way resemble
those of any Orthostern ‘scorpion’ ; and therefore it must be excluded from the Eoscor-
pionidae ( Auct .). On the other hand, it is almost certainly an aquatic Lobostern with
unique characteristics meriting its own generic name.

CONCLUSIONS

A. Anatomical. 1. The carapace, including its eyes, the dorsal plates of the body, the
caudal rings, and the poison-capsule and sting of the Carboniferous ‘scorpions’ here
studied differ in no essential feature from the homologous parts of Recent scorpions
which lack lateral eyes (I can, however, confirm Petrunkevitch’s statement that there are
Carboniferous forms with lateral eyes ; e.g. Compsoscorpius elongatus Petr, and C. elegans
Petr.).

2. Our knowledge of the ventral anatomy is much less complete; but in all the forms
here examined there is also an essential similarity between the macro- and microscopic
structure of the following organs and the homologous parts of present-day scorpions.
In the prosoma — the chelicerae, pedipalps, mouth parts, and legs (though the latter may
differ in detail); in the mesosoma — the sense-organs on the pecten.

3. The ‘scorpions’ here described fall into two groups when the shape, structure, and
probable functions of the sternites and pectines are considered.

Group A. Sternites strongly overlapping, each larger than its corresponding tergite, and resembling a
pair of the laminate, gill-bearing appendages of an Eurypterid. The combs large, fin-shaped, and with
many teeth.

(i) Sternites strongly bilobed and with spinules on the doublures :

(a) Coxa 3 with epicoxite, abuts against an hexagonal sternum; coxa 4 abuts against the genital

328

PALAEONTOLOGY, VOLUME 3

operculum which is bilobed as in Recent scorpions. At least one leg ends in denticulate claws with
‘stiletto heels’. Sternites with hairy coating on external surface. Pareobuthus salopiensis

( b ) Chelicerae abnormally large, sternum unknown, coxa 4 does not abut against the genital
operculum which has a median ridge flanked by paired, two-bossed lateral wings; legs with denticulate
claws with ‘stiletto heels’; some tarsal spurs replaced by leaf-like structures; no hairs observed on
external surface of the sternites. Lichnophthalmus pulcher

(ii) Sternites unlobed, some with a median suture; hairs on their external surfaces and on the
doublures. Chelicerae abnormally large; coxae 1, 2 with mandibular processes and the whole coxo-
sternal area as in Recent scorpions; genital operculum as in (i) (b) \ distribution and shape of spurs and
claws as in Recent Buthids, but all claws and some spurs are denticulate. Mazoniscorpio mazonensis

Group B. Sternites slightly overlapping and not much larger than corresponding tergites, resembling
those of Recent scorpions but with no stigmata on their external surface. Stigmata not yet observed
with certainty, but possibly present on the posterior-lateral doublures. The combs not fully known,
but probably with narrower rachis and fewer teeth than in Group A, and therefore more similar to
those of Recent scorpions.

This group cannot yet be subdivided. Its members have the following additional characteristics:
Coxo-sternal area and the genital operculum essentially the same as in Recent scorpions. The shape
and distribution of spurs and claws on the legs and the shapes and proportions of the chelicerae and
pedipalps, and the distribution of granules, setae, and trichobotheria on them can be closely matched
in present-day Buthid scorpions; but the claws and some of the spurs are denticulate in all the larger
specimens. Buthiscorpius buthiformis, B. major

and probably G.S.M. Za 2924, Za 2925 and M.M. L8194.

There is not enough of Benniescorpio tuber culatus or Wattisonia coseleyensis preserved
to allow them to be placed with certainty, but I consider that the former may belong to
Group A(i)(n) and the latter to some part of Group A.

I have also examined the holotype of Archaeoctonus glaber Peach (G.S.E. 9675,
9676). Judged by its sternites and the little that can be seen of its pecten, it would fall in
Group B; but the abnormal structure of its coxo-sternal region and the remarkable
shortness of its legs, both features being regarded by Pocock as diagnostic characteristics
of his genus Archaeoctonus, cannot be matched in any of the specimens here described.
On the other hand, the claws, so far as they can be examined (the specimen is adherent
to shale), appear to resemble all the larger claws described above in being denticulate
(though denticles are only obscurely visible on one leg) and in possessing (at any rate
on the 4th leg) a large dagger-type pad (figured by Peach on pi. 22, fig. Hi) recalling that
on the supposed 3rd leg of Lichnophthalmus (Part 1, text-fig. 7).

B. Ecological. In drawing inferences from the anatomy as to the mode of life and
habitat of Carboniferous ‘scorpions’ it must be remembered that the homologous
organs in Recent forms, however similar in structure they may be — and this dearth of
change since Palaeozoic times is one of the most awkward facts facing an evolutionist —
have all been inherited from their Carboniferous ancestors, and not vice versa. Can it
be assumed that these unchanged organs have today the same functions as they had in
Carboniferous times, or is it possible that there has been an adaptation of the functions
to new environments without any modification of the shape and structure of the organs?

An example of this problem is found in the uniformity of structure (though not
always in shape) of the pecten with its sensory hairs on the rachis and its sensory fields
of peg-organs on the teeth, whether the comb belongs to a Recent terrestrial scorpion,
to a supposedly terrestrial Carboniferous or Triassic ‘scorpion’, or to an almost cer-

LEONARD J. WILLS: SOME CARBONIFEROUS ‘SCORPIONS’ 329

tainly aquatic one like Lichnophthalmus (or for that matter the certainly aquatic ? Mero-
stome ‘ Glyptoscorpius ’).

Another example is the sting which is an organ that at first sight would appear to be
adapted exclusively to a terrestrial way of life, but which is known to have been de-
veloped in at least two purely aquatic Merostomes, Carcinosoma scorpionis Grote and
Pitt (late Silurian) and Mixopterus kiaeri Stormer (? early Devonian).

With such reservations in mind, the following inferences suggest themselves : (i) From
anatomical conclusions 1 and 2, above, it is arguable that the mode of locomotion,
perception of light by the eyes, and of other vibrations by the peg-organs of the pecten,
by setae and by trichobothria and the methods of hunting, capture, and stinging of their
exclusively animal food, practised by those Carboniferous ‘scorpions’ here described,
were the same as those employed by Recent scorpions, despite the fact that most of the
latter live in tropical dry regions, under stones or buried in sand (to retain their internal
moisture), whereas their Coal Measures ancestors, whether aquatic or terrestrial, were
denizens of swamps and rain-forests of an equatorial belt.

(ii) Ecological analysis of anatomical conclusion 3 and of the grouping there set out
is more complex.

(a) The structure of the pecten, the sternites, and the tarsalia of the legs (particu-
larly the claws and spurs) of animals in Group A(i) {a) ( b ), and of the pecten and sternite
in Group A(ii) appear to be better adapted for an aquatic than for a terrestrial life; yet
the tarsalia of Mazoniscorpio (Group A(ii)) so closely resemble their Recent homologues
(in the Buthidae) that this creature must have been capable of walking on land. How-
ever, there would seem to be room for gills above the laminate sternites of all the forms
in Group A and in Benniescorpio ; and there is visual proof that there were no stigmata
on the sternites of any of the specimens examined. I infer that all the animals in Group A
were gill-breathers, and that they were, nevertheless, capable, like Limulus, of spending
part of their life on land. It is noteworthy that in two specimens in Group A in which
the genital operculum is known, it is different in shape from and more complex than
that organ in ‘scorpions’ of Group B where it closely resembles the operculum of
Recent forms. There is nothing to indicate whether this difference can be correlated
with a difference in habitat or in reproductive processes (in a third specimen in Group
A, Pareobuthus, the genital operculum is simply bilobed as in Recent forms).

(b) I am inclined to infer that all the ‘scorpions’ in Group B were terrestrial, since
every part of the exoskeleton, barring the sternites, has been handed down to the scor-
pions of today virtually unmodified, there having been no adaptations needed to fit the
animal for its present exclusively terrestrial habitat, except perhaps the development of
true air-breathing book-lungs above the sternites. It may ultimately turn out that even
this type of respiration had already been acquired by the ‘scorpions’ in Group B ; yet
undoubted stigmatic openings for such book-lungs have not so far been observed either
on the external surface of the sternites, as in Recent forms, or on their doublures, as in
Triassic ones, or on any other part of the body. Nor does there seem enough overlap of
the sternites to have housed gills. The only other method of respiration that suggests
itself to me is by direct oxidation of the blood through the skin, which in several cases
has been seen to be penetrated by numerous pore-canals, and which in every case is
extremely thin on the ventral surface. It must be recalled that in Triassic and Recent
scorpions respiration takes place through a film of chitin that covers the laminae of the

330

PALAEONTOLOGY, VOLUME 3

lung-books. In the Carboniferous ‘scorpions’ of Group B such oxidation may perhaps
have been particularly effective in the small infolds of the skin created by the overlap of
the sternites (i.e. in the same positions as the gills of Group A), which were kept per-
petually moist by the water-laden atmosphere of the coal-swamps. This could have been
a first step towards the development of true book-lungs which by Triassic times had
come into existence above the sternites, but which had stigmata on the posterior-lateral
doublures, that is, in the same positions as the gill-pouches of Lichnophthalmus and
Mazoniscorpio, and not, as today, on the external surface of the sternites.

In appraising the validity of the hypothesis of respiration through the thin ventral
skin it should be borne in mind that experiments on Recent scorpions have shown that
‘they have considerable respiratory reserves’ (Cloudley-Thompson 1958, p. 74) — in
other words a scorpion can survive for long periods under conditions in which there is
a deficiency of air; e.g. at the bottom of deep burrows that have collapsed or with seven
out of its eight lungs blocked.

To summarize the foregoing conclusions in a general statement I advance the follow-
ing tentative opinions:

1 . There were two races of ‘scorpions’ living side by side in the very humid surround-
ings of the Carboniferous coal-swamps. One race, mostly but not entirely, with lobate
sternites were aquatic animals breathing by gills in gill-pouches lying above deeply
overlapping sternites, and possessing large fin-like combs. Some of these were adapted
to crawl easily over weeds, others had normal scorpion legs. Probably all were capable
of living out of water for fairly long periods. The second race were terrestrial creatures,
with short orthostern sternites, breathing air in some way at present unknown, but
possibly by book-lungs housed above the sternites with openings into modifications of
the gill-pouches seen in the other race. Mazoniscorpio seems to possess some charac-
teristics of each race, having the large combs and deeply overlapping sternites of the
first, yet with all its other organs fashioned exactly as their homologues in the second
race.

2. Both races had many organs fashioned like their homologous parts in Recent
Buthid scorpions, but it was from the second that the Buthidae have descended. (It is
interesting in this connexion to find that the Buthidae have a more primitive type of
embryonic development than the Scorpionidae, see below; and that those genera that
have metatarsal spurs also have pentagonal sternums as in the Carboniferous ‘scorpions’
here described.)

3. Judging from published information, there were Carboniferous ‘scorpions’ whose
ventral structures (particularly the coxo-sternal parts) differed greatly from those dis-
covered by my research. Development of such ‘scorpions’ by the present technique
might show how far these differences really exist and how far each specimen is unique
or referable to either of the two races, postulated above.

4. What has been learnt from this research does nothing towards solving the impasse
in which palaeontologists find themselves, in the matter of classification of the Carboni-
ferous ‘scorpions’, as a result of imperfection of preservation and rarity of specimens
(see also Part 1, p. 267). This impasse is hardly to be wondered at when we consider
that zoologists with complete specimens of hundreds of species from half a dozen conti-
nents at their disposal have been forced to employ an admittedly artificial classification

LEONARD J. WILLS: SOME CARBONIFEROUS 'SCORPIONS’

331

of Recent scorpions. This classification takes into account such things as the shape of
the pincers and the sternum, the number of trichobothria on the pincers and of spurs on
the legs, while placing within a single Order creatures whose embryological develop-
ments may be radically different. ‘We see that although so much alike in outward
appearance, there exist great internal differences between various scorpions. In some as
in the Buthidae the embryo is left to feed itself on the yolk of the egg, while in others,
as in the Scorpionidae, embryonic nourishment is a complex affair resulting from a
process of mutual adaptation between the organs of mother and young’ (M. Vachon,
Endeavour, April 1953, p. 89). Existing palaeontological classifications have made an
analogous mistake in assuming that all Carboniferous ‘scorpions’ were terrestrial air-
breathers, though Pocock, when instituting his Group Lobosterni, expressed more than
a suspicion that ‘respiratory lamellae’ lay above the lobes of the sternites.

ADDENDA

1. An error in my description o/ Lichnophthalmus pulcher Petr. In the light of discoveries in Mazoni-
scorpio, my preferred interpretation of the sternum of the pecten of Lichnophthalmus pulcher Petr, in
Part 1 (p. 272 and text-fig. 4) is clearly erroneous; for what I tentatively interpreted as anterior plates
of a complex sternum of the pecten (text-fig. 4 a, b, p, p') can be matched closely with the genital oper-
culum of Mazoniscorpio (present text-fig. 12), and the posterior plates with the whole sternum of the
pecten in the latter.

The discovery of the shape and structure of the genital operculum and sternum of the pecten in
Mazoniscorpio also throws doubt on Pocock’s interpretation of these parts in Eobuthus holti Pocock
(1911, pi. 11, fig. 2 and text-fig. 1).

2. The pecten of scorpions and ‘Glyptoscorpius’. In Part 1, p. 263, 1 stated that the pecten is an organ
only found in scorpions, whether fossil or living, provided we except the ‘doubtfully Eurypterid
Glyptoscorpius' , and that I had made preparations of the teeth of a Glyptoscorpius ‘comb’ and had
found ‘them to be devoid of the peg-organs so characteristic of the scorpion comb, whether fossil
or Recent’. This statement remains true for the specimen B.M. In.25982 that I examined which is
undoubtedly conspecific with G. caledonicus Salter, as figured by Peach (1882, pi. 29, figs. 17, \la-d),
agreeing, as it does, in size and in having barb-like spines along the two sides of the teeth or filaments.
In all my preparations the skin is quite structureless (PI. 53, fig. 5). However, I have now found that a
very similar comb (B.M. In.42706) attributed in the British Museum catalogue to Glyptoscorpius, which
differs from G. caledonicus Salter in that the teeth or filaments have no spines along their sides, differs
also in possessing typical peg-organs rather widely scattered over its surface (PI. 53, fig. 6).

Both these combs are parts of gigantic Arthropods, of a size met with in some Merostomata, but
monstrous when compared with the largest-known scorpion. The teeth (filaments of Peach) of the
first specimen (In.25982) range up to about 12 mm. in length, and those on the second comb (In.42706)
up to 6 or 7 mm. : yet the peg-organs on the latter are about the same size as those on the tiny Buthi-
scorpius buthiformis (PL 46, fig. 5) and on that of the medium-sized Lichnophthalmus pulcher (Part 1,
pi. 50, figs. 2, 3) and the unnamed G.S.M. Za 2925 (PI. 55, fig. 4), in all of which they are closely
packed together. I wish to thank Dr. Waterston and Dr. E. I. White for allowing me to examine these
important specimens of what is certainly a geological enigma.

3. Improvement in technique. Since writing Part 1, 1 have made a great improvement in the technique
described on p. 264. Instead of inverting the specimen after etching in order to get rid of the debris
and isolated pieces of the skin, I now suck up the acid with a pipette until the specimen is nearly dry,
lift it out of the evaporating basin and, inclining it slightly over a white basin full of water, I wash it
with a very gentle stream of water from a pipette. This carries the remaining acid, the mud, and any
loosened bits of chitin into the basin, from which the chitin can be extracted with a brush or pipette
as described in Part 1 ; but as a rule most parts of the skin remain attached to the body. When the
washing is finished, the specimen is put on the microscope stage and flooded gently for inspection,
&c., before the next etch. As soon as the ‘well’ in the Marco has been established round the fossil, the
washing should aim at moving the mud into one comer of the ‘well’, but even this operation should be

332

PALAEONTOLOGY, VOLUME 3

carried on over a white basin in case any chitin fragments get washed over the sides. The mud and
fragments are then sucked up from the corner by pipette and ejected into water in a white basin.

In dealing with the larger specimens, I have found that it is possible to etch the appendages free
from the matrix, and yet retain them attached to the body or to the Marco-cast right up to the end.
As is shown by PI. 50, fig. 2 and PL 51, fig. 4, such isolated parts of the body can eventually be dried
and then flooded with Marco, there to remain floating in their natural postures within the final Marco-
block.

When the fossil has been embedded in the Marco-block, details cannot be so easily seen as when the
individual organs have been extracted and mounted. For this reason in one or two cases I intentionally
broke away particular organs with a needle during the etching, and mounted them on slips, after
having previously recorded their position on the animal by a photo (PL 50, fig. 2; PL 51, figs. 4, 6).
The chief value of this modification of the original technique lies in the opportunity it gives to see the
organs in their original positions before they may become detached, accidentally or intentionally.

4. Extraction of chitin from shale. In dealing with the chitinous skin of ‘scorpions’ preserved in shale ,
I have found that a covering of Marco, ground level and sealed with glass, increases greatly the con-
trast between chitin and shale and at the same time gives a transparency to the chitin that it does not
possess when ‘dry’ (see the photos of Benniescorpio, Pl. 56, fig. 5 ; PL 57). In one case after the specimen
had been covered with Marco the underlying shale on the reverse was wetted and dried several times
and finally all of it was picked off with a needle, leaving the chitin beautifully transparent and attached
to the Marco. The great objection to this treatment is that the Marco cannot be removed after it has
polymerized.

REFERENCES

bennie, j. 1886. On the prevalence of Eurypterid remains in the Carboniferous shales of Scotland.
Proc. Phys. Soc. Edinburgh , 9, 499.

cloudsley-thompson, J. l. 1955. On the function of the pectines of scorpions. Ann. Mag. Nat. Hist.
(12), 8, 556-60.

1958. Spiders, scorpions, centipedes and mites. London.

corda, a. j. c. 1835. Ueben den in der Steinkohnlenformation Chomle gefundenen fossilen Scorpion.
Verhandl. Gesell. Vaterl. Museums Bohmen. April 1835, 36.

1839. Ueben eine fossile Gattung der Afterscorpione. Ibid. April 1839, 14-18, pl. 1.

meek, F. b. and worthen, a. h. 1868a. Preliminary notice of a scorpion, a Eurypterusl and other
fossils from the Coal Measures of Illinois and Iowa. Am. Jour. Sci. Arts. (2), 45, 25.

1 8 6 8 A . Geological survey of Illinois, 3, Arachnida, 560-3.

peach, b. n. 1881. On some new species of fossil scorpions from the Lower Carboniferous rocks of
Scotland and the English borders, &c. Trans. Roy. Soc. Edinb. 30, 399-412, pl. 22, 23.
petrunkevitch, a. 1913. A monograph of the terrestrial Palaeozoic Arachnida of North America.

Trans. Connecticut Acad. Arts. Sci. 18, 1-137.

1949. A study of Palaeozoic Arachnida. Ibid. 37, 69-315.

1953. Palaeozoic and Mesozoic Arachnida of Europe. Mem. Geol. Soc. Amer. 53.

1955. Arachnida in Treatise on Invertebrate Paleontology, Part P, pp. P42-162. Lawrence,

Kansas.

pocock, r. i. 1911. The Carboniferous Arachnida. Palaeontogr. Soc.

stormer, l. 1955. Chelicerata and Merostomata, in Treatise on Invertebrate Paleontology, Part P,
pp. Pl-41. Lawrence, Kansas.

thorell, t. and lindstrom, g. 1885. On a Silurian scorpion from Gotland. Kongl. Svenska Vet.-
Akad. Handlingar, 21, (9), 1-33.

whitehead, t. h. and eastwood, t. 1927. The geology of the southern part of the South Staffordshire
coalfield. Mem. Geol. Surv. Gt. Brit.

wills, L. J. 1925. The morphology of the Carboniferous Scorpion Eobuthus Fritsch. J. Linn. Soc.
London , 36, 87-98, pl. 3.

— — 1947. A monograph of British Triassic Scorpions. Palaeontogr. Soc.

PROFESSOR L. J. WILLS

Farley Cottage,
Romsley,

Manuscript received 21 June 1959 near Birmingham

THE PELTASPERMACEAE, A PTERIDOSPERM
FAMILY OF PERMIAN AND TRIASSIC AGE

by JOHN A. TOWNROW

Abstract. The three genera comprising the Peltaspermaceae, Lepidopteris (leaf), Antevsia (pollen organ),
Peltaspermum (seed organ) are discussed. The following are redescribed: L. stormbergensis (Seward) Townrow,
L. martinsii (Kurtze) comb, nov., A. zeilleri (Nathorst) Harris, A. extans (Frenguelli) comb, nov., and P.
thomasi Harris. The evidence for referring isolated organs to their parent plant, and the morphology and
affinities of the family are considered. The Peltaspermaceae range from the Thuringian to the Rhaetic (inclusive),
the earliest species coming from Europe.

The Peltaspermaceae were established by Dr. H. Hamshaw Thomas as a small family
of Triassic (including Rhaetic) age. They have now been found in the Upper Permian,
but only once from rocks later than the Rhaetic. The leaves, Lepidopteris, have been
known for many years. The pollen organ, Antevsia, has hitherto been known from one
species; herein a second is described. The seed organ, Peltaspermum, is known from
two species, but only the type species, P. rotula, in detail.

The Rhaetic P. rotula is peltate and radially symmetrical, and is thus unique among
seed organs. This has led to the belief that the Peltaspermaceae are very isolated
morphologically. Examination of the earlier species suggests that the family is not so
isolated, and that comparison is possible with Carboniferous fossils.

The ages and distribution of the species of the Peltaspermaceae are shown in Table 1.

lepidopteris Schimper

Type species L. stuttgardiensis (Jaeger) Schimper 1869

The genus Lepidopteris has been redefined and discussed by Frenguelli (1943) and
Townrow (1956). Of the five leaves referred to Lepidopteris, the youngest, L. ottonis
(see Antevs 1914; Harris 1932; Lundblad 1950), and the two oldest, L. stormbergensis
(see Thomas 1933; Townrow 1956) and L. martinsii (see below), are known in detail;
but the other two, L. stuttgardiensis and L. madagascariensis, are very incompletely
known (Townrow 1956). The finding of thirty-five new specimens of L. stormbergensis
has added new information on the range of variation of that species, and on the
morphology of Lepidopteris. This latter with characters of the rachis is taken in the
general discussion of Lepidopteris', they both are of wider interest, and support
the reference of reproductive organs to the leaves.

The gross form of the leaf. In L. stormbergensis the pinnae are offset towards the upper
(adaxial) side of the rachis, for on this side they are traceable almost to the mid-line of
the rachis ; but upon the lower (abaxial) side are overlapped by the rachis. Which side
is the upper or lower is determined by reference to the leaf base. L. martinsii is probably
the same as L. stormbergensis', the main rachis forms a trench whose bottom lies at a
deeper level than the bottom of the trenches formed by the pinna rachises (text-figs.
1 A— d, 4a). The matter has not been discussed in the other species.

[Palaeontology, Vol. 3, Part 3, 1960, pp. 333-61, pi. 58.]

334

PALAEONTOLOGY, VOLUME 3

The pinnules of L. stormbergensis, and probably of the other species, are set obliquely
on their rachises, so that if the pinna is considered to be inclined the pinnules lie
horizontally (text-fig. 5a, b, e, h; and see Schimper 1869, pi. 34, fig. 1 ; Harris 1932, pi.
6, fig. 2 and pi. 8, figs. 15, 16). Both asymmetry of pinna insertion and an oblique pinnule
insertion seems to be common, if not normal, in living and fossil pinnate leaves.

TABLE 1

The Peltaspermaceae

Organ

Species

Geological horizon and age

Distribution

Leaf

Lepidopteris

martinsii

Kupferschiefer, Obere Zechstein,
Marl Slate, Hilton Plant Bed,
Lower Marls and Limestone; Upper
Permian

Germany, England

Leaf

Lepidopteris

stormbergensis

Uppermost Beaufort, Molteno, Esk
Series, Hawkesbury Sandstone, Es-
tratos de Potrerillos, ? Sakanema;
uppermost Lower and Middle Trias

South Africa, Queensland,
New South Wales, the
Argentine (? Madagascar)

Pollen

organ

Antevsia extans

Uppermost Beaufort, Molteno, Es-
tratos de Potrerillos; uppermost
Lower and Middle Trias

South Africa, the Argentine

Seed

organ

Peltaspermum

thomasi

Molteno; Middle Trias

South Africa

Leaf

Lepidopteris

stuttgardiensis

Schillfsandstein; Lower Keuper

S.-W. Germany (? Urals)

Leaf

Lepidopteris

madagascariensis

Sakanema; Middle and ? Upper
Trias

Madagascar

Leaf

Lepidopteris

ottonis

Rhaetic zone fossil

Sweden, Germany,E. Green-
land, China, Kazaghistan
(? Tonkin)

Pollen

Antevsia zeilleri

organ

Seed

organ

Peltaspermum rotula

1 Scoresby Sound Beds, Scanian Coal
j Measures; Rhaetic

Sweden, E. Greenland

In all species of Lepidopteris zwischerfiedern occur. These are pinnules set directly
on the rachis between adjacent pinnae. In L. martinsii some of these pinnules lie on the
abaxial rachis surface, appearing to be in series with the pinnules on the pinnae (text-
fig. 5j). In four specimens of L. stormbergensis the lowest pinnule of a pinna is orientated
differently from its neighbours, and is set partly on the abaxial rachis surface (text-fig.
4g). In two of these specimens there are also pinnule tufts, or their remains, inserted

text-fig. 1. a-j, Lepidopteris stormbergensis. a-d, Series of rachises showing increasing degree of
lumpiness; A, c, d, adaxial surface exposed, b, abaxial surface exposed; x7-2. e-h, Rachis cuticles
from specimens figs, a-d, showing increasing proliferation at trichome bases; X200. J, Trichome set
on a single cell, top somewhat torn; x 876. k, L. ottonis, trichome set on a single cell; X 392. L, L.
martinsii, trichome set on a single cell, and showing vertical extensions of cell outlines; the pale area
in the centre of the trichome is probably a feebly cutinized part of trichome base; X879 (v/5793).

JOHN A. TOWNROW: THE PELTASPER MACEAE

335

336

PALAEONTOLOGY, VOLUME 3

purely on the abaxial rachis surface, but lying in series with the pinnae (text-fig. 4h).
In the other specimens the zwischerfiedern are lateral, joined to one another by a wing
of tissue with lamina-like cuticle, but the lowest zwischerfiedern are inserted nearer the
abaxial rachis surface than the pinnae. In L. ottonis and L. stuttgardiensis no difference
in level between the pinnae and zwischerfiedern is apparent in published figures.

I suggest that this series may be interpreted as a progressive simplification of the leaf.
In L. mcirtinsii the series of pinnules normally continues, as zwischerfiedern, on to the
abaxial rachis surface; in L. ottonis and L. stuttgardiensis it is wholly lateral and at one
level, while L. stonnbergensis is intermediate.

The leaf-base and vascular trace. Five leaves of L. stonnbergensis showed the bulbous
leaf-base (text-figs. 4e, f; 5c, d, h). It is cutinized on one side only (the abaxial) and
here there are strong wrinklings suggesting the compression of a terete organ, and blisters
are missing. The cell rows run laterally (text-fig. 4c). On the adaxial surface short
cells, without any pattern, are visible, and also two bulges which recall the vascular
trace scars on an abscission surface, and from each bulge an indistinct rib can be traced
upwards. These ribs join about 1 cm. from the leaf-base; the one rib can then be traced
the whole length of the leaf (text-fig. 4f). The ribs are wholly internal, giving dark
matter on incomplete maceration, but leaving no mark on the cuticle. I suggest they
represent the remains of vascular tissue. There is no information about the vascular
trace from the other species.

The structure of the rachis. The swellings on the rachis of Lepidopteris were first regarded
as scales, but were later shown to be blisters of the cuticle (Antevs 1914). In L. stonn-
bergensis there is a series of leaves from some with a smooth rachis to others with a
markedly blistered one. From this series one can suggest the nature and origin of the
blisters.

On leaves with a smooth rachis, the rachis cuticle shows trichomes, or trichome bases
(often numerous), either set on a single cell, or with slight proliferation of the epidermal
cells around the trichome base (text- figs. 1 a, b, e, f ; 2e), the other cells lying in longitudinal
rows. On leaves with a slightly blistered rachis the proliferation around the trichome
bases is greater (text-figs, lc, d, g, h; 2f), the area of proliferation corresponding to a
small blister. In leaves with a markedly blistered rachis the proliferation is greater still,
either to give an area up to twenty cells across in which the cells are set in a concentric
pattern (text-fig. 6f), or, rarely, the compressed blisters overlap and the cells appear to
be set irregularly. Sometimes, over large blisters, a trichome base is not visible, but
usually some trace of it remains. More than one trichome may contribute to a single
blister (cf. text-fig. 6f).

text-fig. 2. a— d, Runiex hydrolapathum. a, Trichome on a smooth petiole (cuticle prep.), b, Section
of outer tissues of a smooth petiole passing through a trichome. c, Trichome over a swelling, showing
slight proliferation of cells (cuticle prep.), d. Section of a swelling beneath a trichome; a-d x326.
e-k, Lepidopteris stonnbergensis. e, A naturally macerated stoma, showing arcs of dark matter
(probably lignine lamellae) flanking stomatal pit; X 730. f. Cuticle of a smooth rachis, showing one
trichome and trichome bases, g, Rachis cuticle showing one trichome base, and cell outlines over
wing (to left), h. Cuticle of a slightly lumpy rachis, showing trichome bases, proliferated cells, and
two stomata, k, Stem cuticle, showing stoma, trichome base over a proliferation, j, L. martinsii.
Rachis cuticle, trichome base over a proliferation and four stomata (left plain) (v/5963 a), l, L.
ottonis. Rachis cuticle showing a trichome base and slight proliferation, e-l all x 167.

JOHN A. TOWNROW: THE PELTASPERMACEAE

337

text-fig. 3. A-c, Lepidopteris stormbergensis. A, Cuticle from lamina showing cell outlines, papillae,
and stomata; X 167. b. Stoma without cutin papillae; X 876. c. Stoma of normal form; x876.
d-e, Antevsia extans. Trichome from the rachis; X 392. e. Stoma from the rachis; X392. F, Pelta-
spermum thomasi. Stoma from the rachis; X 392 (v/23400). g-k, L. martinsii. g, Cuticle from lamina
showing cell outlines, papillae, and stomata; X 167 (v/5963 b). h, Stoma lacking cutin lappetts; X 876
(the Hellstedt leaf). J, Stoma with large lappets of solid cutin; x876 (v/5963 b). k, Section, approxi-
mately transverse, reconstructed through stoma shown at j; x876 approx.

text-fig. 4. a-b, Lepidopteris martinsii. a, Part of a leaf showing decurrent pinnules and more or
less paired imprints of swellings on rachises; X 16 (v/5963). b, A pinna apex, showing venation and
imprints of swellings; x2 (v/2595). c-h, L. stormbergensis. c, d, Cells from abaxial (c) and adaxial
(d) surfaces of leaf base ; X 106. E, f. Surfaces of the leaf base (e abaxial, F adaxial), showing suggested
vascular traces; X7-2. g. Part of a leaf, showing insertion of lowest pinnule of pinnae partly on
abaxial rachis surface; X 7-2. H, Leaf, showing pinnule tufts set on the abaxial rachis surface; x 1-7.
J, k, Paripteris gigantea. J, Leaf, abaxial side; xO-4. k, Central part of same leaf showing blisters on
rachises, two pinnule tufts on abaxial rachis surface (to left) and venation in one pinnule; xl-6

(v/ 1296).

340

PALAEONTOLOGY, VOLUME 3

Rumex hydrolapathum [Polygonaceae] provides an analogy. In this plant the petiole
bears trichomes, and in some leaves the trichomes are set on small swellings, which
show, in a cuticle preparation, proliferated epidermal cells around the trichome base,
and in section proliferation of the sub-epidermal layer, which raises the swelling (text-
fig. 2a-d). Presumably, therefore, the swellings of L. stormbergensis are also formed
by sub-epidermal proliferation at the site of a trichome; and the proliferation of the
epidermis takes place to accommodate this local increase in girth. In Lepidopteris,
however, the blisters contain a dense coaly residue, soluble on maceration, suggesting
that the blisters were occupied by thick-walled lignified cells rather than cellulose
parenchyma.

L. ottonis and L. martinsii show the same features as L. stormbergensis (text-figs.
Ik, l; 2j, l), except that in L. ottonis solitary trichomes and small blisters are only seen
at the leaf or pinna apices, and blisters lacking any sign of a trichome are commoner
(Antevs 1914, pi. 2, figs. 6-8). Further, the reproductive structures show, as far as is
known, blisters on their rachises just like those of L. stormbergensis, though usually
smaller.

The trichomes appear to be unicellular, more or less spherical (text-fig. 1j, k, l), and
are very thinly cutinized, the cuticle being minutely functate.

The stomata. One or two naturally macerated leaf fragments with stomata were found.
These stomata showed two arcs of dark matter flanking the stomatal pit, and lying on
the guard cell surface (text-fig. 2e). These are interpreted as lignine lamellae upon the
outward surface of the guard cell, seen generally in living gymnosperms.

The stem of the Lepidopteris plant. Text-fig. 5c shows what is possibly the stem of L.
stormbergensis. Both ends are broken, and at present it measures 2-3 cm. long and
0-5 cm. wide. The surface is of low relief, but shows a number of more or less contiguous
raised areas. The cuticle (text-fig. 2k) is as on the rachises. The stem is wider than any
rachis seen, and differs further in lacking a wing (Townrow 1956; and text-fig. 2g).

This stem closely resembles the stem referred to L. ottonis (see Harris 1932), though it
is smaller and with a more delicate cuticle. Since both stems are cutinized they are
presumably primary organs. There is no information as to how the leaves were borne.

Lepidopteris stormbergensis (Seward) Townrow
Plate 58, fig. 1; text-figs. 1a-h, j; 2e-h, k; 3a-c; 4c-h; 5a-h; 6a-c; 8h
For synonomy and diagnosis see Townrow 1956

Description and discussion. Nearly all the material of L. stormbergensis comes from the

EXPLANATION OF PLATE 58

Fig. 1. Lepidopteris stormbergensis (Seward) Townrow. Leaf with abnormally lumpy rachis, xl.

Figs. 2, 3. Peltaspermum thomasi Harris, v/23400, Brit. Mus. (Nat. Hist.) Part and counterpart of
type, x2.

Figs. 4-9. Antevsia extans (Frenguelli) comb. nov. 4, 5, Pollen grains, sulcus facing observer (4) and
to left (5), x 500. 6, The type (from du Toit 1927, pi. 29, fig. 3), X T7. Specimen showing branch-
ing, X 1. 8, 9, Upper and lower surfaces of disk and pollen sacs (different specimens), x3. (st — -
branchlet, or position of same.)

Fig. 10. Antevsia zeilleri (Nathorst) Harris. Pollen grains in three views, X 500.

Palaeontology, Vol. 3

PLATE 58

T O W N R O W, Peltaspermaceae

JOHN A. TOWNROW: THE PELTASPERMACEAE

341

Burnera Waterfall locality, Upper Umkomaas, Natal; but two leaves are from the
Australian Trias (v/32106 and v/32472, Brit. Mus. (Nat. Hist.)). The Waterfall locality is
regarded as falling within the Middle Triassic Molteno (see Haughton 1954; Townrow
1957). At the Waterfall the leaves of L. stormbergensis are found scattered more or less
evenly through the thickness of plant-bearing rock; and this is different from the other
common leaves, which are markedly patchy.

L. stormbergensis proves to be a variable leaf, and the variation is set out in Table 2
and in text-figs. 3a-c; 5a-h; 6a-c.

table 2

Range of variation in L. stormbergensis

Variation

Characters showing variation

Leaf length

Leaf width

Pinna length

Pinnule length

Pinnule width

Largest

20 cm.

13 cm.

7 cm.

14 mm.

7 mm.

Normal

12 cm.

8 cm.

4 cm.

6 mm.

4 mm.

Smallest

7 cm.

2-5 cm.

1 -3 cm.

3 mm.

2 mm.

Size of

Cell size

sinuosities

Height and

Cuticle thickness

{away from

of cell

width of

veins)

outlines

papillae

Upper

Lower

Largest

60 p

5 V

4 p, 10 p

3-5 p

3-0 p

Normal

45 p

2-3 fj.

Ip, 5 p

2-0p

L0 p

Smallest

24 p

0-5 p

0 0

1-5 p

0-5 p

Number of

Stomatal

subsidiary

frequency

Stomatal

cells

mm}

index

Most

8

108

9-3

Normal

5 or 6

34

7-0

Least

3

20

4-7

The stomata are all of essentially the same pattern, but they differ in detail. The
commonest sort (about 75 per cent.) shows subsidiary cells bearing a cutinized papilla,
now hollow (text-fig. 3c). A second sort, making up nearly all the remainder, shows a
lappett of solid cutin (cf. text-fig. 3j). One or two leaves showed some stomata of an abnor-
mal sort (text-fig. 3b). These stomata (about 1 per cent.) are important, for they match
a sort of stoma seen on the pollen sacs of the pollen organ referred to L. stormbergensis.

Though variable, I do not believe that species other than L. stormbergensis are present.
The groups given in Table 3 intergrade, and no other character gives ground for splitting
up the material. The characters vary about one mean, and if more than one species was
involved this would scarcely be so. Such characters are the number of pinnae per leaf
(about 32), the interval between veins (about 0-25 mm.), the stomatal density and index,
and the stomatal form.

A list of records believed to be synonyms of L. stormbergensis is given by Townrow
(1956); in only one of them was the cuticle examined (Thomas 1933). The variation in

B 6612

z

342

PALAEONTOLOGY, VOLUME 3

text-fig. 5. a-h, Lepidopteris stonnbergensis. Leaves to show variation in gross form, stem shown
in c at a\ xO-75 (g is v/32472). J, L. martinsii. Large tripinnate leaf, showing pinnules on abaxial
surface, and zwischerfiedern between pinnae; xO-75 (redrawn from Gothan and Nagalhardt 1921,

pi. 3, fig. 1).

JOHN A. TOWNROW: THE PELTASPERMACEAE

343

gross form of the present material easily includes the variation seen in these records, and
the cuticle of Dr. Thomas’s specimen is typical.

I have examined two Australian specimens of L. stormbergensis (v/ 32106 and v/32472,
Brit. Mus. (Nat. Hist.)) and Mr. J. M. Pulley has very kindly sent me some excellent
photographs of other leaves from Queensland (described but not figured by Walkom
1924 and 1929), which, though only one has a cuticle, with nearly straight cell outlines,

TABLE 3

Variation in a number of characters in Lepidopteris stormbergensis

Leaf
length 1

Stomata! density

Lumpiness

Cuticle thickness (/x)

( per H.P. field)

Stomatal index

Specimen

width

of rachis

Upper

Lower

Upper

Lower

Upper

Lower

Averages

i

3-6

A

Stomatal density

8

18

3-6

A

B

3-5

2-0

30

1-0

30

2-8

1-4

1-2

10-4

9-6

5-6

5-0

upper 2-6
lower 1-5

11

A

30

1-5

2-2

10

7-3

4-3

16

B

20

10

1-6

1-2

5-3

4-1

both 2-0

D

O

A

2-7

C

3-5

3-0

3-0

1-4

10 5

5-6

cC

O

10

1-5

B

20

0-5

2-8

10

8-4

5-2

Stomatal index

9

A

20

10

2-2

1-4

7-1

4-9

upper 8-7

12+

1-7

D

3-0

10

2-6

1-4

6-6

3-9

lower 5-1

23+

D

2-0

0-5

4-2

3-0

100

7-6

—

15+

2-3

B

2-0

20

1-4

1-2

5-5

4-6

both 6-9

7

1-7

D

30

2-5

2-4

2-6

7-6

9-3

Stomatal density

6°

E

20

1-5

1-8

2-8

7-3

9-9

upper 1 -48

34

C

20

1-5

1-4

1-6

5-0

5-2

33

1-5

D

20

1-5

10

1-2

50

6-6

lower 2-4

eu

22

C

20

1-5

1-4

2-0

5-6

7-2

both 1 -96

D

o

27+

E

1-5

10

2-0

3-0

6-6

9-6

cC

0

4

1-3

E

3-5

0-5

1-8

2-6

6-5

8-4

Stomatal index

5°

B

D

B

3-5

2-5

0-5

0-5

2-2

0-2

3-0

2-2

7-5

0-8

9-4

9-3

upper 6-25
lower 8-45

30 1

E

30

20

2-4

3-4

7-6

110

31 +

D

2-5

1-5

2-8

1-8

90

6-6

both 7-35

Notes: (i) Rachis quantities A-E refer to 5-figd. rachises (text-fig. 1a-d; pi. 58, fig. 1)
taken as standards, (ii) Leaves marked + exceptional in one value, (iii) Leaves
marked the length/width probably less than 1-5, but specimen too small for
certainty, (iv) Counts averages of 10J in. fields.

I identify as L. stormbergensis. These photographs show the same variation as the Natal
leaves. I think that there is no doubt that the same leaf is involved in both Australia
and South Africa.

The possible habit of L. stormbergensis. The leaves of L. stormbergensis can be placed
in two groups. In the first the leaves are narrow, with smooth or nearly smooth rachises,
and with more stomata on the upper (adaxial) leaf surface than on the under (abaxial)
surface. In the second group the leaves are wide, the rachises blistered, and the distribu-
tion of stomata normal (Table 3).

In many, but by no means all, waterside herbs of the British flora the leaves arising
from the upper and lower nodes are of different form. The difference parallels the
variation between the two groups found in L. stormbergensis (see Table 4 for some
examples). Also, to refer again to Rwnex hydrolapathum, the swellings on the petiole

344 PALAEONTOLOGY, VOLUME 3

TABLE 4

Stomatal densities (per H.P. field) and indices of some waterside herbs

Species

Leaf

position

(mode)

Upper surface

Lower surface

Upper surface

Lower surface

Stomatal

density

Stomatal

index

Stomatal

density

Stomatal

index

Stomata

counted

Cells

counted

Stomata

counted

Cells

counted

Alisnia

Upper

2-8

9-1

2-0

7-5

21

211

21

257

plantago

Lower

3-5

4-6

1-8

8-0

35

266

22

254

Menianthes

Upper

1-8

4-75

3-66

8-7

22

441

44

461

trifoliata

Lower

4-0

6-8

4-7

7-0

39

533

47

630

Mentha

Upper

0

0

12-2

17-2

0

317

72

346

aquatica 1

Lower

2-8

11-0

4-7

13-0

28

230

47

317

Myosotis

Upper

6-5

12-0

13-6

27-0

65

All

136

365

palustris 2

Lower

6-8

15-0

61

14-0

68

392

61

Ml

Nasturtium

Upper

8-1

20-0

11-6

27-0

81

398

116

308

officinale

Lower

9-3

25-0

5-7

16-5

93

271

57

292

Polygonum

Upper

0-9

2-7

5-0

17-3

9

325

50

245

amphibium

Lower

13-8

22-0

7-0

180

138

489

69

317

P. hydropiper 1

Upper

0-9

2-5

18-0

22-0

9

355

85

331

Lower

4-0

13-2

5-0

13-2

35

231

41

269

Ranunculus

Upper

6-0

18-0

3-7

16-5

71

323

44

226

flamula

Lower

7-8

20-0

1-25

5-5

78

306

15

264

Rumex

Upper

2-3

9-2

3-0

12-0

28

270

36

230

hydrolapathum

Lower

4-8

14-5

4-4

10-5

48

282

44

376

Sium

Upper

2-9

7-1

15-7

26-0

29

382

157

543

angustifolium

Lower

4-7

11-7

4-2

9-6

47

352

42

397

Typha

Upper

21-5

20-0

24-0

22-0

129

501

145

528

latifolia 3

Lower

22-0

24-0

16-0

27-0

65

All

136

365

Veronica

Upper

8-0

16-0

14-5

25-0

80

422

145

437

beckabunga

Lower

7-6

22-0

5-5

17-0

89

317

66

321

1 Hair bases not counted. 2 On side shoots. 3 In stomatiferous areas only.

are characteristic of leaves from the upper nodes ; the leaves from the lower nodes are
smooth.

In Dicotyledonous trees stomata are generally few or absent on the leaf upper surface.
In terrestrial Dicotyledenous herbs stomata may be frequent on the upper leaf surface,
but they are regularly still more abundant on the under leaf surface, whatever the
position of the leaf on the plant (see Salisbury 1927 ; Walter 1951). There are exceptions
to this statement, but among plants which have more protected upper leaf surfaces as
the Graminae, Trifolium (see Erban 1916), or cushion-forming alpines (see Salisbury
1927).

In these respects L. stormbergensis is more like a waterside herb than a tree or terres-
trial herb. It might be argued that its rather thick cuticle is against this, but in fully adult
leaves of Rumex hydrolapathum or Alismci plantago the cuticle, as prepared by oxidative
maceration, may be up to 2-5 p thick.

JOHN A. TOWNROW: THE PELT ASPERMACEAE

345

Lepidopteris martinsii (Kurtze) comb. nov.

Text-ligs. 1l; 2j; 3g-k; 4a, b; 5j; 6d

1839 Alet/iopterin martinsii Germar MS; Kurtze, p. 34, pi. 3, fig. 2. German material, type
specimen.

1907 Callipteris martinsii (Germar); Gothan, pp. 1-4, figs. 1, 2. German material figured.

1921 Callipteris martinsii (Germar); Gothan and Nagalhardt, pp. 451-3, pi. 6, figs. 5, 6;

pi. 7, figs. 1-3. Good German material with cuticle.

1958 Callipteris martinsii (Germar); Stonely, pp. 313-15, pi. 37, figs. 2, 5; text-figs. 5, 6.

English material, cuticle figures.

(For full synonomy see Gothan 1907 and Stoneley 1958.)

Diagnosis (emended). Leaf bipinnate, sometimes tripinnate, pinnules parallel-sided with
very obtuse apex. Zwischerfiedern simple or pinnate, in smallest leaves absent, forming

text-fig. 6. A-c, Lepidopteris stormbergensis. Pinnules to show venation; xl-33 (b is v/32472).
d, L. martinsii. The Hellstedt leaf, showing form and swellings; X 1. e, Paripteris gigantea. Portion
of naturally macerated cuticle, showing cell outlines, veins (v) and supposed stomata (st) ; x 53
(v/1293). f, L. ottonis. Cuticle over a large swelling, showing three trichome bases each over a
proliferated area of cells, somewhat concentric cell arrangement at swelling margins, and two

stomata; x66.

346

PALAEONTOLOGY, VOLUME 3

a series from pinna base towards abaxial leaf surface. Blisters prominent. Cell outlines
usually straight, sometimes showing minute processes running horizontally, but regu-
larly with peg or strap-shaped processes at the cell corners running vertically. Stomatal
pit usually overhung by small lappetts of solid cutin borne partly on the anticlinal
surface of the subsidiary cells. Encircling cells often absent.

Description. The material described comes mostly from the Marl Slate of Kimberley
(Notts.), with some from the Hilton Plant Bed (Westmorland). These horizons are
approximately equivalent to the Kupferschiefer (Stoneley 1958). One additional leaf
from the Zechstein of Hellstedt (referred to as the Hellstedt leaf) was most kindly lent
me by the Director, Paleobotaniska Avdelningen, Naturhistoriska Riksmuseet, Stock-
holm. The English material is all in the collections of the British Museum (Nat. Hist.).
It is rather fragmentary, few specimens have a cuticle, and only the Hellstedt leaf is
cutinized over the main rachis.

The leaf is shown in text-figs. 4a; 5j. The larger specimens figured in the literature
are tripinnate, and the smaller bipinnate. In some German specimens the series of
zwischerfiedern can be seen continuing on to the abaxial rachis surface, but in my
specimens the lowest pinnule of a pinna is simply somewhat decurrent (see, for
example, Gothan and Nagalhardt 1921 and text-figs. 4a; 5j). The rachises of all the
specimens show rather prominent swellings, or imprints which are interpreted as their
remains; frequently rather regularly in pairs (text-figs. 4a; 6e). On the cuticle trichomes
and trichome bases surrounded by proliferated cells can be seen (text-figs. 1 l ; 2j). The
venation is shown in text-fig. 4b; the distance between the veins at the margin is about
0-75 mm.

The features of the cuticle are shown in text-figs. 2j ; 3g. On both lamina and rachis
one cuticle is thicker than the other. 1 suppose that the thinner cuticle is from the lower
surface of the leaf. The cell outlines in two of the English specimens (v/ 5963b and
v/5964) show in places minute processes, less than 1 p high (text-fig. 3g). In all the
specimens, especially over the rachis, the cutin on the periclinal cell walls penetrates
rather deeply, sometimes lying crushed on one side, sometimes still upright so that the
various levels can be determined by focusing. All also show projections of cutin inwards
at least at the cell corners, up to 3 p long (text-figs. 1l; 3h). In v/5963 the cells also
show each a single (often obscure) papilla set more or less centrally; all the other leaves
have a smooth cuticle.

The stomatal density is about 40 mm.2 (upper surface) and 80 mm.2 (lower surface);
and the stomatal index about 7 (upper surface) and 14 (lower surface) (count of 20
high-power fields). About three-quarters of the stomata examined showed lappetts of
solid cutin overhanging the stomatal pit, and in about a tenth of these lappetts were so
large as almost to close the stomatal pit, but normally were much smaller (text-fig.
3g, j). The stomata without lappetts showed a rim of thick cutin (text-fig. 3h). The
lappetts were set partly on the anticlinal surface of the subsidiary cells, so that they
pointed obliquely upwards, and did not lie in the same plane as the cuticle surface (text-
fig. 3j, k).

Various thin places and irregular holes, presumably pathological, and smaller holes
and irregularities that were probably caused by sand damage (the holes were about the
same size as the grains of the matrix) were present.

JOHN A. TOWNROW: THE PELTASPERMACEAE 347

Discussion. Many specimens of L. martinsii have been figured, and these form an inter-
grading series; most fall between the extremes, and the series is not easily divisible into
more than one entity. The cuticle has been much less studied (see Gothan and Nagal-
hardt 1921 ; Florin 1931 ; Stoneley 1958), but here too there seems to be a gradation of
forms that cannot conveniently be divided. Not enough cutinized specimens have been
examined, however, to say whether there is any correlation between leaves of a certain
form and certain cuticle characters. Stoneley (1958) points out that in the German
specimens whose cuticle has hitherto been figured the cuticle is papillate, whereas in the
English specimens the cuticle appeared to be smooth. However, staining in safranin
shows that in at least one English leaf there are papillae, and in at least one German
leaf there are not.

The similarity between L. martinsii and species of Lepidopteris has been noted before
(e.g. by Gothan 1907), and L. martinsii is now removed from Cailipteris , and placed in
Lepidopteris, because it agrees in the construction of the leaf, particularly in its series of
zwischerfiedern; in the detailed construction of the blisters on the rachis; and in overall
stomatal construction, which is of a rather unusual sort. This extends the range of the
genus Lepidopteris, and therefore of the Peltaspermaceae, from the Thuringian to the
Rhaetic.

L. martinsii may be distinguished from L. ottonis and L. stormbergensis by its parallel-
sided and obtuse pinnules, by the details of its stomata (text-figs. 5j; 3j) and by the
presence of cutin projections inwards at the cell corners. A tripinnate specimen may be
told from the bipinnate L. stuttgardiensis, but a smaller specimen may look like L.
stuttgardiensis. They may then be distinguished because L. stuttgardiensis has, it appears,
shorter pinnules.

antevsia Harris 1937

Type species Antevsia zeilleri (Nathorst) Harris

Diagnosis (emended). Microsporophyll bipinnate, primary branching alternate, in one
plane, second branching irregular. Main and branch rachises with blister-like swellings.
Ultimate branches bearing four to twelve pollen sacs. Pollen sacs sessile, unilocular,
2-5 mm. long and 1-2 mm. wide, dehiscing by a longitudinal slit on ventral face.
Pollen-sac wall massive, showing midrib and a few stomata on dorsal surface. Interior
cuticle lining pollen sac absent, all other parts cutinized, and showing stomata overhung
by hollow or solid cutin papillae. Pollen grains oval with slight projection at each end,
monosulcate, sulcus same length as grain, lips of sulcus touching or almost touching
along its whole length. Cuticle without other markings. Length 23-40 p, width 13-28 p,
depth 12-25 p, ratio length/width about 1-7.

Antevsia zeilleri (Nathorst) Harris
Plate 58, fig. 10; text-figs. 7a-d, k, l; 8e, f, j, k; 9e

1910 Antholithus zeilleri Nathorst, pp. 20-24; pi. 2, figs. 59, 60; pi. 4. Type specimen pi. 4,
fig. 1 ; Swedish material.

1914 Antholithus zeilleri (Nathorst); Antevs, pp. 10-15, pi. 3 (except figs. 17, 18). Nathorst’s
material redescribed.

1922 Antholithus zeilleri (Nathorst); Johansson, p. 28, pi. 1, figs. 11-13. Additional Swedish
material.

text-fig. 7. a-d, Antevsia zeilleri. a, b. Specimen figured by Nathorst, 1908, pi. 4, fig. 1, showing the
two surfaces of the organ (a upper, b under); x7-2. c, d, two pollen grains, area of darker cuticle
interpreted as the internal cuticle of furrow; X875. k, Macerated specimen showing pairs of pollen
sacs and fragment of branchlet (dark); X 14. l, A specimen showing four fertile branchlets, one
showing lateral bulges clearly, another retaining its pollen sacs; X7-2. e-j, A. extans. E, f, Fertile
branchlet ending in disk and four dehisced pollen sacs, e upper side, f lower side (in transfer), f, Lit
from bottom right; x7-2. g-j, Three pollen grains, area of darker cuticle as in c, d; x875.

JOHN A. TOWNROW: THE PELTASPERMACEAE

349

1932 Lepidopteris ottonis ( Antholithus zeilleri ) (Nathorst); Harris, pp. 62-65, pi. 7, figs. 1, 2,
10, 11; text-figs. 27e, f. Abundant Greenland material.

1937 Antevsia zeilleri (Nathorst) Harris, p. 35, diagnosis.

Description. The second species of Antevsia raised questions that made a re-examination
of the type species desirable, and the Director of the Paleobotaniska Avdelningen,
Naturhistoriska Riksmuseet, Stockholm, kindly lent me Nathorst’s material.

The whole organ is bipinnate, about 9 cm. long, the rachis bearing alternate branches
in one plane about 1 -5 cm. apart; these then divide, in a plane more or less at right angles
to the rest of the organ, into two to five ultimate branchlets. The rachis and branches
show blisters, but the branchlets do not.

The sterile region of the branchlets is 1 -5-2-0 mm. wide, and was probably thick in
life, for the cuticle shows irregular wrinklings, presumably arising during compression
(text-fig. 7l). The cuticle of the two surfaces differs slightly in thickness, and in cell
pattern (text-fig. 8e). The fertile region is the same width, or very slightly wider than the
sterile part, and the cuticle is only slightly different (text-fig. 8k) ; stomata number about
20 mm.2, and in four out of six specimens the cells were papillate. The margin of the
fertile region is usually produced into a number of bulges, which lie more or less opposite
to one another, and are 0-25-1-0 mm. wide, and the same high (text-fig. 7a, b, l). These
lobes now appear to be of the same thickness as the branchlet, and the cuticle on both
their sides is as in text-fig. 8f.

The pollen sacs lie in two rows up the under side of the branchlet end, as preserved
from 0-25-0-5 mm. in from the margin. They tend to lie in opposite pairs, one pollen
sac to each lobe, but this is not entirely regular, and there is sometimes an unpaired
terminal pollen sac. They commonly overlap laterally, and the inner faces of the pollen
sacs approximate, so that there is very little, or no, discernible tissue of the branchlet
between them (text-fig. 7a, b, k, l). Near the pollen sac’s outer edge the cuticle of the
lateral bulge shows a varying number of cell rows whose orientation is the same as on
the pollen sac itself. There is usually a strong fold, interpreted as caused during fossiliza-
tion, at the apparent junction of the lateral bulge and pollen sac. These features are
much the same whether the pollen sacs are dehisced or not. I suggest that they show
that the pollen sacs were borne wholly on the under surface of the branchlet, and that the
sides of the thick branchlet overlapped the line of insertion of the pollen sacs.

The bluntly pointed pollen sacs measure about 2 mm. long and 1 mm. wide. The
dehiscence slit runs all the way up the inner face to just below the tip on the outer face
(text-fig. 9e) and at this point the cuticle commonly shows a little group of equidimen-
sional cells on which the cell rows on the rest of the pollen sac converge. The pollen-
sac cuticle is strongly dorsi-ventral. Near the dehiscence line it is thick, showing markedly
elongated cells (text-fig. 9e), and opposite to the dehiscence line there is another area of
thick cuticle showing roughly rectangular cells (54 x 25 p), but at the base one or two
strips of markedly elongated cells, about four cells wide and thirty high (text-fig. 8j), and
a few stomata (cf. text-fig. 8g). Between these areas the cuticle is thin, showing elongated
cells (61x19 p). The outer face of the unmacerated pollen sac is sometimes raised in the
area opposite the dehiscence line, and in one or two cases (text-fig. 7a) there is a sugges-
tion of a small mid-rib. The cell rows sometimes converge slightly towards the pollen-
sac base, suggesting that the base was slightly contracted.

350

PALAEONTOLOGY, VOLUME 3

Four pollen sacs were still filled with pollen, and several others contained some
grains. All these grains were alike and no others were seen (PI. 58, fig. 10; text-fig.
7c, d). When the grains lie sulcus uppermost a curving line can be seen arising at each
end of the sulcus (text-fig. 7c) which presumably marks the interior margin of the furrow ;
while when the grains lie on one side, a further curved line can be seen running under
the sulcus (text-fig. 7d), and this I suppose represents the bottom of the furrow. The
dimensions of a sample of twenty grains are as follows (extremes in parentheses):
length 36-2 p (46-25 p, 29 p); width 22-6 p (29-8 p ., 20 p); depth 21-35 p (26-4 p, 19-8 pi);
interior width of furrow 8-73 p (13-2 p 6-6 p); depth of furrow 6-3 p (9-9 p, 3-3 p), this
value was measured directly from grains lying on their side; the proportions of the
grains are: average ratio of length/width 1-6 (1-55 longest grain, 2-3 shortest grain,
TO broadest grain, 1-45 narrowest grain); and average ratio of length/depth 1-7 (cor-
responding figures of extreme specimens, T8, 2-32, 1-12, 1-46).

Antevsia extans (Frenguelli) comb. nov.

Plate 58, figs. 4-9; text-figs. 3d, e; 7e-j; 8g; 9a-d; 10a

1927 Sagenopteris sp. (non Presl gen.) du Toit, pp. 399-400, pi. 29, fig. 2. Single pollen-sac
group; Aliwal North, South Africa.

1944 Fanerotheca extans Frenguelli (part), pp. 393-402, pi. 1; pi. 2, fig. 1; pi. 4, fig. 1. (Ex-
cluded are pi. 2, figs. 2-4; pi. 3; pi. 4, fig. 2. ) Three specimens; near Cacheuta, the Argentine.

71915 Equisetaceous tubers (?) Walkom, p. 31, pi. 3, figs. 3, 4. Ill-preserved fragments; Den-
mark Hill, Ipswich, Queensland.

Diagnosis (emended). (Main rachis unknown) presumed branches twice pinnate, ultimate
branchlets 0-5-1 -0 mm. wide expanding to 2-25 mm. at their ends and bearing four
radially arranged pollen sacs. Pollen sacs attached partly marginally and partly on
(presumed) under surface, when dehisced about 5 mm. long and 2 mm. wide. Cuticle
1 p thick or less, on upper surface of branchlet end showing equidimensional cells,
stomata, and trichome bases, between pollen sacs showing elongated cells only.

Description and discussion. The material consists of five specimens from the Waterfall
locality (see p. 340), and one, v/ 20796 (Brit. Mus. (Nat. Hist.)), from an unknown locality
in the Argentine. The most complete specimens (PI. 58, fig. 7) show alternate branching.
The larger branches show small blister-like swellings, but the branchlets cellular striae
only. The cuticle shows trichomes, trichome bases sometimes proliferated, and stomata
like those of L. stonnbergensis (see text-figs. 3d, e; 9b). On the larger branches the

text-fig. 8. a-d, Peltaspermum thomasi. A, Cuticle of outer surface of marginal lobe, showing
elongated cells possibly marking the course of a vein. ( Note : the specimen was torn, and slightly
rotated, in preparation); X 167. b, c, Cuticle of upper and lower surfaces of seed head (b lower, c
upper), showing cell outlines and stomata. In b darker area of rectangular cells marks course of a
vein: in c stippled area marks a fold in which cells are not visible; x 167. d, Cuticle of main rachis
showing stoma and two trichome bases, one with proliferated cells; x 167 (a-d all v/ 23400). E, f, j, k,
Antevsia zeilleri. e, Cuticle from ultimate branchlet, a part without pollen sacs, f. Cuticle from under
surface of a lateral bulge, note cells at top right are similar in shape to the cells at bottom right in
j; e, f x 167. J, Cuticle at pollen-sac base, showing elongated cells over mid-rib (centre), and cells
probably of branchlet (bottom right); x88. k, Cuticle from upper surface of ultimate branchlet
from a place with pollen sacs on lower surface; X 167. g, Antevsia extans, a stoma from a pollen sac;
X 327. h, Lepidopteris stonnbergensis, cell outlines showing form of the sinuosities and papillae;

X 327.

JOHN A. TOWNROW: THE PE LTAS PE R M ACE A E

351

352

PALAEONTOLOGY, VOLUME 3

cuticle is of a different thickness on the two sides, and in the largest specimen the pollen
sacs were all the same way up — with the dehiscence line hidden by the rock. This side
corresponds to the branch surface with thicker cuticle, so I suppose the pollen sacs were
borne with the dehiscence line facing downwards.

The ultimate branchlet is expanded at its apex, and this part is termed the disk.
Viewed from the upper side the branchlet lies at a lower level than the disk surface and
cellular striae cannot be traced from one on to the other: but viewed from the under
surface, the disk surface and branchlet lie at the same level, and the cellular striae con-
tinue from one on to the other. The branchlet therefore was attached partly to the under
surface of the disk as well as to its margin (PI. 58, fig. 9; text-fig. 7e, f).

Viewed from the under side the edges of the open pollen sacs meet laterally and at
the same level (PL 58, fig. 8 ; text-fig. 7f) and radiate out from the disk. Viewed from
the upper side the pollen sacs and disk are in complete continuity, and though the centre
of the disk is collapsed (presumably it was thick) it is not possible to see where the
pollen sac ends and the disk begins (PI. 58, fig. 8 ; text-fig. 7e). Supposing that the empty
pollen sacs were supported by the matrix during compression (Walton 1936) the union
of the pollen sacs and disk lies at the edge of the collapsed region. These appearances
mean that the pollen sacs were borne partly marginally and with part of their area of
attachment on the under surface of the disk. It seems clear that they were not borne
wholly on the under surface of the disk.

The cuticle of both surfaces of the disk are shown in text-fig. 9b, c. Both surfaces
show irregular wrinklings, and solitary trichomes were present on the upper surface.
One pollen sac shows the cells of its interior surface clearly. Other pollen sacs are similar,
though less clear. Along a thicker rib running for about half the length of the pollen
sac (here termed the mid-rib) the cells are elongated, and the other cell rows diverge
from it (text-fig. 9a). The cuticle of the pollen sacs (text-fig. 9b) is very delicate, and the
cells obscure, partly owing to wrinkling of the cuticle, but the cells over the mid-rib
are plainer and more elongated than the others. A few stomata are scattered over the
outer face of the pollen sacs, mostly near the mid-rib (text-figs. 9b; 3e). Most of these
stomata show neither radially disposed subsidiary cells, nor cutinized papillae. There
is no sign of an interior cuticle. The whole pollen-sac wall is of dense substance, and
only translucent in one or two places next to the dehiscence slit; in one place the
pollen-sac wall is now 0-2 mm. thick.

No unopened pollen sacs were available as a source of uncontaminated pollen, but
there were pollen grains adhering to the macerated pollen sacs. Of these forty-six were
from 23 to 40 n. long and of one sort, eight were well over 40 ^ long, but of the same sort,
while three were clearly different. I accept only those grains from 23 to 40 ^ long as
belonging to A. extans, though it is possible that the larger grains belong also. The
form of the grains is shown in PI. 58 and text-fig. 7g-j. They show the same features as
A. zeilleri and are interpreted in the same way. The dimensions of a sample of twenty
grains are as follows: length 31 /x (35 n and 23 ^ ) ; width 17-5 /i (23 n and 13 n); depth
16 /x (24 jtx and 12 jix); interior width of furrow 8 n (5 ^ and 12 /x); depth of furrow 3 /j.
(5 /x and 1 -6 /x), this last value I regard as rather unreliable as it was taken from ten
grains only. The proportions of the grains are: length/width T7 (T8, T7, 2-2, T4) and
length/depth 1-6 (1-8, 1-5, 2-3, 1-4).

Du Toit’s pollen-sac group (see PI. 58, fig. 6) is identified with the present material,

JOHN A. TOWNROW: THE PELTASPERM ACEAE

353

for it shows the same form of pollen sac, disk, and branchlet, and the dimensions agree.
No microscopic detail was available to du Toit. Frenguelli’s material also agrees with
the present in dimensions, in showing blisters on the larger branches, and, rather
obscurely, a disk to which the pollen sacs are attached (text-fig. 10a) and possibly a
median rib. No microscopic detail is given. The specimens quoted above as distinct
differ; I believe they belong to another plant, but here discuss them no further.

Discussion of Antevsia. The main rachis of A. ex tans is unknown, so that the two species
can only be compared as regards their lateral branches, branchlets, pollen sacs, and
pollen. The pollen sacs of both species are strikingly alike. The walls are massive, and
the presence of stomata (a most unusual feature) suggests that photosynthetic tissue
may have been present. Both species show a mid-rib. The nature of this structure is
unknown, but it shows those characters which in compressions indicate vascular tissue ;
a strand of thicker substance, and epidermal cells modified in shape over it. The pollen
of the two species is indistinguishable. I can find no difference in form and an effort
to separate the two statistically (sample of 46 grains of A. extans and 50 of A. zeilleri )
failed since the Standard Error of each sample was greater than the difference between
the Means.

These resemblances are held to justify the inclusion of both A. zeilleri and A. extans
in one genus.

The differences lie in the number and the manner in which the pollen sacs are borne
(p. 349 and see text-fig. 7). These differences I regard as variations on a common plan,
and they can be explained away in at least three ways. Either we may regard A. extans
as arising from A. zeilleri by specialization of the fertile part and reduction of the pollen
sacs to four — perhaps two pairs; or A. zeilleri can be regarded as derived from A. extans
either by amplification in the number of pollen sacs, or by the condensation of several
fertile apices of A. extans to form one fertile apex of A. zeilleri. All three alternatives
have to face considerable difficulties, and I do not think it is possible at present to make
a choice between them.

Antevsia does not closely resemble other known Mesozoic pollen organs. Two show
some approach, Sphenobaiera fur eat a (see Krausel 1943, 1955) and Antholithus wettsteini
(see Krausel 1955), but both differ in the shape and number of pollen sacs at each fertile
apex, and apparently in the construction of the pollen-sac wall and stomatal construction.
The isolated pollen grain Monosulcites minimus Cookson (see Couper 1953, 1958) is, in
figures, indistinguishable from the pollen grains of Antevsia. M. minimus is an artificial
assemblage ranging the Mesozoic which cannot in Couper’s opinion be subdivided as
yet.

peltaspermum Harris 1937

Type species P. rotula Harris

Diagnosis (emended). Whole organ with alternate branches in one plane. Seed-bearing
heads borne at branch ends, of thick substance, showing five to fifteen (probably
vasculated) marginal lobes over arching surface of head bearing the seeds. Rachis and
branches showing blister-like swellings, upper surface of seed head similar. Seeds borne
on lower surface of head, either two, lateral to insertion of branch ( P . thomasi), or ten
to twelve in a ring around insertion of branch ( P . rotula). Cuticle showing stomata of
Lepidopteris sort. Seeds ( P . rotula only) with prominent micropyle and free nucellus.

text-fig. 9. a-d, Antevsici extans, a. Interior of pollen sac in transfer, showing cells and mid-rib;
x 156. b, Cuticle of upper surface of disk showing cells, two stomata, and two trichomes; x 167.
c, Cuticle of lower surface of disk; X 167. d, Cuticle of pollen sac, a fold to right, two stomata;
<r/= dehiscence line, w=midrib; xl56. e, A. zeilleri. Cuticle at apex of pollen sac showing cells
along the dehiscence line and equidimensional cells near apex; x 100.

text-fig. 10. a, Antevsia ex tans. Specimen from the Argentine, showing branching (redrawn from
Frenguelli 1944, pi. 1, fig. 1); x 1. b-f, Peltaspermum thomasi. b, The part of type specimen, showing
pinnate arrangement of branches; X 6. c, d. First seed head on left showing seed-scar (c on left),
seed, veins, and two marginal lobes (part and counterpart; the detached object is of unknown nature);
x 9. E, Third seed head on left of part, showing four marginal lobes and veins (the other lobes are
broken off); x9. f, Second seed head on left of part, showing two seeds and three marginal lobes;

X 9. All from v/23400.

356

PALAEONTOLOGY, VOLUME 3

Peltaspermum thomasi Harris
Plate 58, figs. 2, 3; text-figs. 3f; 8a-d; 10b-f; 11

1933 Lepidopteris natalensis Thomas (in part only), pp. 251, 254, pi. 24, fig. 75; text-fig. 55.

Type specimen described (under same name as leaf).

1937 Peltaspermum thomasi Harris, pp. 34—35. Diagnosis.

Diagnosis (emended). Main rachis showing small blisters and lateral wing; seed-bearing
heads attached marginally, showing five or six marginal lobes, and pinnately branched
veins on lower surface; interval between veins about 0-5 mm. Seeds two per head,
attached laterally to branch insertion, about 3 mm. long and 1-5 mm. wide.

Description. The material consists of the type and two other specimens (v/23400,
v/35798, and v/35799, Brit. Mus. (Nat. Hist.)), all from the Burnera Waterfall locality.
The type was described by Thomas (1933), but my interpretation differs from his.

In v/23400 the specimen is compressed on one bedding plane, but the others traverse
several, and probably preserve the original shape of the organ more nearly. Upon this
view the seeds pointed down towards the base of the organ. The cuticle of one side of
the rachis is thicker than the other, and this suggests that the organ may have been
upright or inclined rather than pendulous.

The rachis and branches have a cuticle like that of L. stormbergensis, showing trichomes,
stomata, and a wing (text-figs. 3f; 8d). Though not all are clearly shown the evidence
is fully consistent with the view that the branching was pinnate, for the wing on the
branches is decurrent on to the wing on the rachis without sign of twisting in the cell
rows (PI. 58, fig. 2; text-fig. 10b). Like the pinnae of the leaf the branches are attached
towards one (presumably the upper) surface of the rachis (text-fig. 11a).

The round or cordate seed-bearing heads are attached at one edge and not in the
middle. The lower epidermis of the branch passes gradually into the lower epidermis of
the head, and the upper into the upper epidermis of the head, but most of the attachment
lies on the lower surface of the head, so that head and branch do not lie at the same level
(text-fig. 10d, e). The whole head is of thick substance, and its upper surface is raised
in irregular wrinkles, especially towards the margin, but at the extreme margin there is
a flat rim about 0-5 mm. wide (text-fig. 10d, e). Presumably, therefore, the whole head
was thick, or dome-shaped, and the flat rim is a compression border (Walton 1936).
At the margin of the heads, lying just inside the compression border (text-fig. 10c, e, f),
there are semicircular lobes, about 1 mm. high at a maximum. In one head (text-fig.
10e) there were six occupying the whole circumference. The cuticle of the lobes is
similar to the cuticle of the head, and there are no internal cuticles, or signs of a micro-
pyle, so that they cannot be seeds. The arrangement of epidermal cells (text-fig. 8a)
suggests the lobes may have been vasculated, and there appear to be as many lobes as
veins in the head.

The lower cuticle of the heads is only broken by the seeds, and by scars believed to
mark the insertion of seeds. Three heads show clearly (text-fig. 10c-f) a system of raised
lines, over which the cells are elongated, and which on incomplete maceration show
a strand of dark substance (text-fig. 8b). These lines are therefore veins. The cuticle of
the upper surface of the head is shown in text-fig. 8c.

In two heads there are two seeds lying laterally to the branch insertion, and more or

JOHN A. TOWNROW: THE PELTASPERM ACEAE

357

less over the first dichotomy of the vein (text-fig. 10f). In another head only one seed
was present, but there is a scar in the corresponding position (text-fig. 10c). In the
other heads there is one seed shown. I suggest that two seeds per head is probably the
normal.

The seeds (which may not be mature) are about 3 mm. long and 1 -5 mm. wide, with a
micropyle which is slightly curved away from the branch. I have not macerated any
seeds, since there are so few.

Discussion of Peltaspermum. The two species of Peltaspermum are superficially rather
different. P. rotula has peltate radially symmetrical seed heads, 10-12 seeds, and 10-15
marginal lobes : P. thomasi has bilaterally symmetrical seed heads, attached marginally,
2 seeds and 5 or 6 marginal lobes. Nevertheless, the two species appear to be organized
on the same plan, so they are left in one genus. Thus the seed heads of P. rotula were
unattached, but there is evidence that they were borne in two rows on a central rachis
(Harris 1932, p. 66). The seed heads of P. rotula were certainly thick (Harris 1932,
pp. 65-66), and I suggest that the rim shown in Harris’s reconstruction is a compression
border. This is supported by the appearance of some Swedish material of P. rotula
(Lundblad 1950) which is preserved ‘in the round’. Lundblad found that in her Swedish
material there were as many ridges between the seeds as marginal lobes (cf. also Harris
1932, pi. 8, figs. 1, 2, 9), while the cuticle of the lower surface of the heads is stated to
consist of elongated (rectangular or polygonal) cells. Lundblad’s discovery suggests
that, as in P. thomasi, the lobes may have contained vascular tissue which continued
within the head, between the seeds.

GENERAL DISCUSSION

(a) Reference of isolated organs to the parent plant. Lepidopteris ottonis, Antevsia zeilleri,
and Peltaspermum rotula are all referred to the same plant because they all show blisters
on their rachis cuticles and the same sort of stomata, two features not found together
in other plants of the holoarctic Rhaetic floras. In addition, leaf, pollen, and seed organ
are found together in several localities in Greenland and Sweden, and in some of these
L. ottonis is the local dominant (Harris 1932; Lundblad 1950).

The same arguments are applied to Lepidopteris stormbergensis, Antevsia extans, and
Peltaspermum thomasi, and as far as I know no other plant in their flora shows the
combination of blisters and stomata of the Lepidopteris sort. Evidence from association
is weaker than in the case of the L. ottonis plant. If the identifications suggested hold,
L. stormbergensis and A. extans are found together in three localities, the Waterfall,
near Cacheuta, and Aliwal North; but other plants are also common to these three
localities. P. thomasi is known only from the Waterfall. However, the evidence for
relationship between the L. ottonis and L. stormbergensis plants automatically supports
also the reference of each isolated organ to its parent plant.

( b ) The Morphology of Antevsia and Peltaspermum. Previous authors (e.g. Harris 1932;
Thomas 1932; Hirmer 1939; Frenguelli 1944) have considered Antevsia to be a micro-
sporophyll. I agree, taking ‘sporophyll’ to mean an organ that is like a leaf, but not
pursuing the question of the ultimate meaning of the term. Antevsia shows pinnate
branching like a compound leaf, and not spiral or whorled branching like a stem. The
organ also is flattened in one plane, at least as regards the first branching.

B 6612

Aa

358

PALAEONTOLOGY, VOLUME 3

Likewise, I think that Peltaspermum is a sporophyll(but see Thomas 1933). P. thomasi
is flattened in one plane, and shows pinnate branching like the leaf; and probably P.
rotula is similar. The seed-bearing heads of P. thomasi are bilaterally symmetrical, and
show a pinnate plan of venation. P. rotula, however, is unique (see Harris 1932; Hirmer
1939; Lundblad 1950), but is referred to the youngest member of the family, and I
suggest that the form of its seed heads can be regarded as secondary and derived, just
as peltate Dicotyledonous leaves are regarded as derived.

text-fig. 11. Peltaspermum thomasi. Part (a) and counterpart (b); v/35798; x3.

Peltaspermum thomasi can be compared with the multi-ovulate cupule Gnetopsis
elliptica Renault 1885 (and see Walton 1949), for they are about the same size, and the
basic construction of the seed-bearing portions appears to be similar. There are im-
portant differences; for example, G. elliptica has seeds which somewhat resemble
Lagenstoma (see Oliver and Salisbury 1911). This similarity between Gnetopsis and
P. thomasi strengthens Harris’s (1932) suggestion that Peltaspermum is best regarded as
a multi-ovulate cupule (for a discussion of the cupule see, for example, Halle 1937;
Walton 1949); and makes it improbable that the seed heads of Peltaspermum are more
or less unmodified pinnules such as bear seeds in some pteridosperms (see Halle 1929,
1931).

( c ) The relationships of the Peltaspermaceae. As they are interpreted, the Peltaspermaceae
resemble the Pteridospermae (= Cycadofilicales) and no other group of gymnosperms.
They show no particular approach to the other groups of (presumed) pteridosperms in
the Mesozoic, the Corystospermaceae (Thomas 1933) and Harrisia marsilioides (see
Lundblad 1950) with its leaf Ptilozamites. Gothan and Weyland (1954) have classified

JOHN A. TOWNROW: THE PELT ASPERMACEAE

359

the Peltaspermaceae with the Caytoniales (probably having affinities with pteridosperms,
see Harris 1951), but the only similarity is that both groups have reproductive organs
built on a pinnate plan.

Among Palaeozoic pteridosperms, the Peltaspermaceae show one distinct point of
resemblance to the Lyginopterideae, namely that two vascular traces enter the leaf base
which then unite. This is quite different from the structure of the medulloseaen petiole.

The leaf Lepidopteris shows some interesting similarities with the leaf Poripteris
Gothan (1941, 1953). In both the rachises are blistered (text-fig. 4j, k, and Gothan
1941), though the structure of the blisters in Paripteris is not known; in both there are
zwischerfiedern, which continue on to the abaxial leaf surface, where they may form
tufts (text-fig. 4j, k, see Bertrand 1930); and both have stomata typically surrounded
by radially disposed subsidiary cells (text-fig. 6e). None of these features appears to
be very common among fossil leaves, and, so far as I can discover, no other leaves show
the combination. There are differences ; Paripteris is at least tripinnate, paripinnate, and
of neuropteroid pinnule form and venation.

The reproductive organs of Lepidopteris cannot be closely compared with any known
Palaeozoic pollen or seed organ. Antevsia has been tentatively compared with Urna-
topteris (see Harris 1932) and Crossotheca (see Harris 1932; Thomas 1936; Frenguelli
1944), but Crossotheca has bilocular pollen sacs (at least in the better-known species)
partly adnate to the ‘fertile pinnule’, and round pollen grains without a sulcus but with
a triradiate mark (see Kidston 1906, 1923; Florin 1937; Remy 1955). A. extans can be
visualized in terms of a Potoniea in which the pollen sacs have been reduced to four
and the ‘fertile pinnule’ to the disk (see Halle 1933; Florin 1937; Remy 1955), and this
comparison gains force slightly since P. adiantiformis has been found in organic con-
nexion with Paripteris gigantea (Gothan 1913). But the differences in form between
Potoniea and Antevsia are great, and the pollen is different. The isolated Rotliegend
fossil Pteridospermostrobus gimmianus Remy (1954) appears to have a structure a little
like Antevsia , but its pollen is quite different from Antevsia.

Gnetopsis elliptica is the only fossil that appears to be at all like Peltaspermum ; it has
only been found isolated, and its affinities are uncertain, though it is not medulloseaen.

Among the Pteridospermae, therefore, the Peltaspermaceae show one distinct point
of resemblance to the Lyginopterideae, and a rather tenuous resemblance to the
Potonineae. Nowhere do they show any resemblance to the Medulloseae. They must
still be considered a group on their own, but perhaps lying not far from the Lygino-
pterideae. I can find no support for Thomas's (1932) suggestion that they might show
some approach to the Angiosperms.

The earliest record of Lepidopteris is from the Thuringian of Europe; the southern
hemisphere records are all Triassic. It thus seems more probable that the family
originated in the northern hemisphere and spread over the rest of the world from there,
than that it is southern in origin.

Acknowledgements. I am indebted to the Director, Paleobotaniska Avdelningen, Naturhistoriska
Riksmuseet, Stockholm, and to Mr. F. M. Wonnacott, British Museum (Nat. Hist.), for lending me
material with great readiness. I am grateful to Dr. W. G. Chaloner, University College, London, for
help concerning the description of pollen grains, and I am extremely grateful to Professor T. M. Harris,
F.R.S., for a great deal of help at all stages of this work.

360

PALAEONTOLOGY, VOLUME 3

REFERENCES

antevs, e. 1914. Lepidopteris ottonis (Gopp.) Schimp. and Antholithus zeilleri Nath. K. svensk.
Vet.-Akad. Hand!. 51 (7), 1-18, pi. 1-3.

Bertrand, p. 1930. Bassin houiller de la Sarre et de la Lorraine: I. Flore fossile; Ier fascicule,
Neuropteridees. Et. Git. Min. France, 52, pi. 1—30.
brik, M. i. 1952. Iskopaemaya flora i stratigrafiya nizhnemezozoiskich otlozhenii basseina v Ilek.
VSEGEI, Moscow, 1-70, pi. 1-22.

couper, r. A. 1953. Upper Mesozoic and Cainozoic spores and pollen grains from New Zealand.
N.Z. geol. Surv. Pal. Bull. 22, 1-77, pi. 1-9.

1958. British Mesozoic microspores and pollen grains; a systematic and stratigraphic study.

Palaeontographica, 103b, 75-179, pi. 15-31.

du toit, a. L. 1927. The fossil flora of the upper Karroo Beds. Ann. S. Afr. Mus. 22, 289-420, pi.
16-32.

erban, m. 1916. Uber die Verbreitung der Spaltoffnungen in Beziehung zur Schlaftstellung der
Blatter. Ber. Deutsch. Bot. Ges. 34, 880-90.

florin, r. 1931. Untersuchungen zur Stammesgeschichte der Coniferales und Cordaiatales. Erster
Teil. K. svensk. Vet.-Akad. Hand!. (3), 19, 1-107, pi. 1-58.

1937. On the morphology of the pollen grains in some Palaeozoic Pteridosperms. Svensk. bot.

Tidskr. 31, 305-38, pi. 1-3.

frenguelli, J. 1943. Resena critica de los generos atribudos a la Serie de Thinnfeldia. Rev. Mus. La
Plata, n.s., 2, pal. 12, 225-342.

— — 1944. Contribuciones al conocimiento de la Flora del Gondwana Superior en la Argentina.

xvm. Fanerotheca extans n.g., n.sp. Not. Mus. La Plata, 9, 385N-02, pi. 1 -4.
gothan, w. 1907. in potonve, h. Abbildungen und Beschreibungen fossiler Pflanzenreste. Callipteris
martinsii. Kgl. Preuss. Geol. Landesanst. 5 (96), 4 pp.

1913. Die oberschlesische Steinkohlflora. I. Teil. Fame und farnahnliche Gewachse. Ibid.,

n.s., 75, 1-278, pi. 1-53.

1941. Palaobotanische Mitteilungen. 5. Die Unterteilung der karbonischen Neuropterideen.

Pal. Zeit. 22, 421-8.

1953. Die Steinkohlenflora der westlischen paralischen Steinkohlreviere Deutschlands. Bei.

Geol. Jb. 10, 1-83, pi. 1-44.

gothan, w. and nagalhardt, k. 1921. Kupferschieferpflanzen aus dem niederrheinischen Zechstein.

Jb. Preuss. Geol. Landesanst. 42, 440-460, pi. 5-7.
gothan, w. and weyland, h. 1 954. Lehrbuch der Palaobotanik. Berlin.

halle, t. g. 1929. Some seed-bearing Pteridosperms from the Permian of China. K. svensk. Vet.-Akad.
Hand/. (3) 8, 3-24, pi. 1-5.

■ 1931. On the seeds of Emplecopteris triangularis. Bull. geol. Soc. China, 11, 301-3, pi. 1.

1933. The structure of certain fossil spore-bearing organs believed to belong to the Pteridosperms.

K. svensk. Vet.-Akad. Hand!. (3), 12, 1-103, pi. 1-15.

— — - 1937. The position and arrangement of the spore-producing members of the Palaeozoic pterido-
sperms. Com. Rend. 2nd Cong. Strut. Carb., Heerlen, 227-35, pi. 6, 7.

Harris, t. m. 1932. The fossil flora of Scoresby Sound, East Greenland. Part 2: Description of seed
plants incertae sedis together with a discussion of certain Cycadophyte cuticles. Medd. Gronland,
Kobenhavn, 85, 1-112, pi. 1-9, figs. 1-39.

1937. The fossil flora of Scoresby Sound, East Greenland. Part 5: Stratigraphic relations of the

plant beds. Ibid. 112, 1-112, pi. 1.

1951. The relationships of the Caytoniales. Phytomorph. 1, 29-39.

haughton, h. s. in du toit, a. l. 1954. The Geology of South Africa. 3rd ed. Edinburgh.
hirmer, m. in wettstein, f. 1939. Fortschritte der Botanik. 5, Berlin.

Johansson, n. 1922. Die ratische Flora der Kohlengruben bei Stabbard und Skromberga in Schouen.
K. svensk. Vet.-Akad. Handl. 63 (5).

kidston, r. 1 906. On the microsynangia of the Pteridospermae, with remarks on their relationship to
existing groups. Phil. Trans, roy. Soc. Lond. 198b, 41 3—45, pi. 25-28.

JOHN A. TOWNROW: THE PELTASPERMACEAE

361

kidston, r. 1923. Fossil plants of the Carboniferous rocks of Great Britain. Mem. geol. Surv. G.B.
(Pal.), 2 (4), 275-376, pi. 69-91.

krausel, r. 1943. Die Ginkgophyten der Trias von Lunz in Niederosterreich und von Neue Welt
bei Basel. Palaeontographica, 87b, 59-93, pi. 1-15.

in krausel, r. and leschik, g. 1955. Die Keuper Flora von Neue Welt bei Basel. I. Koniferen

und andere Gymnospermen. Schweiz. Pal. Abhandl. 71, 1-28, pi. 1-9.
kurtze, A. 1839. Commentatio de Petrefactis quae in schisto bituminoso Mansfeldensi reperiuntur. Jena.
lundblad, A. b. 1950. Studies in the Rhaeto-Liassic floras of Sweden. I. Pteridophyta, Pteridospermae
and Cycadophyta from the mining district of N.W. Scania. K. svensk. Vet.-Akad. Handl. (4), 1,
1-82, pi. 1-13.

nathorst, a. g. 1910. Palaobotanische Mitteilungen. 6. Antholithas zeilleri n.sp. mit noch erhaltenen
Pollenkornen aus den rhatischen Ablagerungen Schonens. Ibid. 43, 20-24, pi. 2, 4.

Oliver, f. w. and Salisbury, e. h. 1911. On the structure and affinities of the Palaeozoic seeds of the
Conostoma group. Ann. Bot. 25, 1-48, pi. 1-3.

remy, w. 1954. Beitrage zur Kenntniss der Rotliegendflora Thiiringens, Teil II; Fruktifikationen.
Sitz. Deutsch. Akad. Wiss., Kl. Mat. Nat., 1-20, pi. 1-4.

1955. Die Systematik der Pteridospermen unter Beriicksichtigung ihrer Pollen. Zeit. Geol. 2V

312-25, pi. 1-3.

Renault, b. 1885. Course de botanique fossile. Paris.

Salisbury, E. h. 1927. On the causes and ecological significance of stomatal frequency, with special
reference to the woodland flora. Phil. Trans, roy. Soc. Lond. 216b, 1-65.
schimper, w. p. 1869. Traite de Paleontologie Vegetale. Paris.

stoneley, H. M. m. 1958. The Upper Permian flora of England. Bull. Brit. Mus. (Nat. Hist.) Geol.
3, 293-337, pi. 36-40.

thomas, h. h. 1932. The old morphology and the new. Proc. Linn. Soc. Lond., session 145, 17-32.

1933. On some Pteridospermous plants from the Mesozoic rocks of South Africa. Phil. Trans.

roy. Soc. Lond. 222b, 193-265, pi. 23, 24.

1936. Palaeobotany and the origin of the Angiosperms. Bot. Rev. 2, 397-418.

townrow, J. A. 1956. The genus Lepidopteris and its southern hemisphere species. Norsk. Videns.-
Akad. Oslo, Mat. Nat. KL 1-28.

1957. On Dicroidium, probably a pteridospermous leaf, and other leaves now removed from this

genus. Trans, geol. Soc. S. Afr. 60, 21-56, pi. 2, 3.
turutanova-ketova, A. i. 1958. Florichiskaya characteristika nekotorych nizhne-mesozoiskich pro-
duktivnych tolshch vostochnogo sklona srednego Urala. Bot. Zhurn., Moscow and Leningrad, 43,
664-78, pi. 1-3.

walkom, a. b. 1915. Mesozoic floras of Queensland. Part I. The flora of the Ipswich and Walloon
Series, (a) Introduction, (b) Equisetales. Publ. QTand. geol. Surv. 252, 1-66, pi. 1-10.

1924. On fossil plants from Bellevue near Esk. Mem. Q'land. Mus. 8, 77-92, pi. 15-21.

1928. Fossil plants from the Esk District, Queensland. Proc. Linn. Soc. N.S.W. 53, 458-68,

pi. 26-28.

Walter, H. 1951. Grundlagen der Pflanzenverbreitung, I. Teil, Standortslehre. Stuttgart.
walton, J. 1936. On the factors which influence the external form of fossil plants with descriptions of
the foliage of some species of the Palaeozoic Equisetalean genus Annularia Sternberg. Phil. Trans,
roy. Soc. Lond. 226b, 219-37, pi. 31-32.

1949. Calathospermum scoticum, an ovuliferous fructification of Lower Carboniferous age from

Dunbartonshire. Trans, roy. Soc. Edinb. 61, 719-28, pi. 1-3.

JOHN A. TOWNROW,
Botany Department,
University of Reading

Manuscript received 20 May 1959

PHOTONEGATIVE YOUNG IN THE TRIASSIC
LAMELLIB RANCH LIMA LIN EAT A (SCHLOTHEIM)

by R. P. S. JEFFERIES

Abstract. The occurrence of young specimens of Lima lineata (Schlotheim) within the lunules of adults of this
German Triassic species is held to be due to photonegative habits and points to a possible origin of nest-building
in the Limidae. The non-occurrence of this feature in some parts of Germany and its abundance in other parts is
believed to depend on whether or not the adult was byssally attached.

In the deeply excavated lunules of double- valved specimens of Lima lineata (Schlotheim)
from the Wellenkalk (Middle Triassic), near Heidelberg and Wurzburg, it is common to
find one or two double-valved specimens of young of the same species. Two things show
that this association is not accidental ; firstly the young are nearly always at the extreme
anterior end of the lunule — of thirty-nine young observed thirty-two were anterior, six
were median (including four which were members of a pair and which were as anterior
as the other member of the pair would allow), and only one was posterior; secondly
the young nearly always have their lunules pressed close to the floor of the adult lunule
(twenty-nine out of thirty-seven determinable cases). The purpose of this note is to dis-
cuss this phenomenon.

Seilacher (1954) has shown, from mode of crushing and orientation of attached epizoa,
that double-valved specimens of Lima lineata were usually embedded with the commis-
sure vertical and resting on the antero-dorsal margin (PI. 59, fig. 7). This orientation,
which I have seen in two specimens in situ at Lengfurt, near Wurzburg, was the life-
orientation of the species. This is shown by the fact that modern Limidae normally rest
in the same position (Studnitz 1931), except when lying at random in the nest, and by the
fact that the byssal gape, which is fairly large in Lima lineata (PI. 59, fig. 4), is on the
antero-dorsal side.

The elongate antero-dorsal margin of the Limidae is an adaptation to this mode of
life and has led to the markedly posterior position of Limid umbones. The antero-
dorsal side of Lima lineata, in particular, makes a very firm base to rest upon; it is a
broad, heart-shaped area (PI. 59, figs. 1-6) consisting of the excavate, cordate lunule
bordered to right and left by rather flat anterior umbonal ridges which have specially
coarse radial ribs. When resting on the sea-floor in the position described these anterior
umbonal ridges would be the only parts of the shell actually touching the floor and their
coarse ornament would help to prevent slipping at the contact. In 1910 Douville (p. 645)
had already noticed the adaptive significance of the flattened antero-dorsal surface in
P/agiostoma, in which subgenus L. lineata should probably be placed (Philippi 1900,

p. 621).

Hence Lima lineata rested in life with the lunule downwards. Communication be-
tween the sea and the lunule was, nevertheless, possible through the gap between the
prominent umbones. It was doubtless through this gap that the young specimens entered.
They could scarcely have found their way so commonly into such a small gap, however,

[Palaeontology, Vol. 3, Part 3, 1960, pp. 362-69, pi. 59.]

363

R. P. S. JEFFERIES: LIMA LINEATA (SCHLOTHEIM)

unless they were photonegative with a definite tendency to enter small crannies. And in
this connexion it is interesting to find the same photonegative tendency in Recent Lima
(Crozier 1921). The actual position which the young adopted at the extreme anterior
end of the lunule (PI. 59, figs. 1-5) may also, with light coming from above, have been
the darkest place inside the lunule. At least it was clearly the farthest point from the
source of illumination. The orientation which the young adopted, with their lunules
pressed to the floor (or ‘ceiling’) of the adult lunule, meant that their relationship with
their substratum was the same as that of the adults to the sea-floor and suggests that
they were byssally attached.

Before proceeding farther a digression on the habits of Recent Limids is necessary.
These fall morphologically into three groups which Pelseneer (1911, p. 32) called Man-
tellum, Limatula, and Radula. In nomenclature these groups are unsound ( Man tel lum
Bolten 1798 = Lima Bruguiere 1792 (same type species) and Radula Klein 1753 is pre-
Linnaean (Arkell 1931, p. 128)). From the present standpoint, however, nomenclature
is unimportant since the three groups seem to have real existence. "Mantellum' (e.g.
Lima hians) includes forms with extensive gapes round the whole shell margin (except
the hinge line); they have the interesting habit of using the byssus to form a sort of
nest, usually under stones or inside empty shells. MacGinitie and MacGinitie (1949,
p. 346) say of the American species Lima dehiscens that it ‘builds nests in cavities by
spinning byssus threads and attaching them to the surrounding rocks in a scattered
formation. The completed nest is suggestive of the nest that certain spiders build and in
which they wait for their prey.’ (My italics.) Studnitz (1931, p. 301) gave a very similar
account of nest-building in the European Lima hians which usually makes its nest inside
empty My tikis or Cardium shells. There seems to be no fixational byssus in ‘ Mantellum',
for Pelseneer contrasts these ‘Limes nidificatrices’ with the ‘Limes a byssus fixateur’
represented by "Radula'. Pelseneer’s second group, Limatula, consists of small equila-
teral forms with a fairly wide byssal gape but no gape round the rest of the shell margin.
Little is known of the habits of this group, but there is no evidence of nest-building in
the British species L. subauriculata (C. M. Yonge, personal communication). Pelseneer’s
third group, "Radula', includes very large species (e.g. Lima excavata) with a flattened
or excavate antero-dorsal margin and no gape but the byssal gape, with a fixational
byssus and without the nest-building habit. "Mantellum' can swim by clapping its valves
(Studnitz 1931 ; MacGinitie and MacGinitie 1949) and it is possible that Limatula and
‘ Radula ' can also do so though the habit does not seem to have been recorded.

Now "Radula' is very like Plagiostoma and, indeed, there is no clear distinction
between them apart from age (Philippi 1900, p. 621). By the same token it is "Radula',
of all Pelseneer’s three Recent groups, which Lima lineata most nearly resembles. It is
similar in its large size, excavate lunule, and absence of any gape except the byssal
gape ; and this comparison suggests that L. lineata was not a nest-builder.

There is more direct evidence to the same effect. Seilacher (1954, p. 167) stressed the
fact that double-valved Lima lineata, which had evidently been buried with commissure
vertical, nevertheless had their valves closed. He concluded from this that they must
have been buried shortly after death, for on death the ligament would have opened the
valves automatically if the shell had been lying free on the sea-floor. This deduction
does not seem to go far enough, however, for on death the valves would open at once.
To produce fossils with their valves closed burial would therefore have to be simul-

364

PALAEONTOLOGY, VOLUME 3

taneous with death or slightly precede it. In fact it seems likely that double-valved Lima
lineata found in the life-position with their valves closed were nearly always killed by
burial. The same conclusion would not, of course, apply to valves buried with com-
missure horizontal for these would be closed by weight of sediment.

From this argument it follows that the attached animals on Lima lineata (or at least
those found on the lower half or two-thirds of an animal killed by only partial burial)
must have been attached during the life of the Lima. This is confirmed by the presence
on one specimen of L. lineata (PI. 59, fig. la) of a Lopha sp. which is truncated against
a growth-line of the ‘ host ’ and so must have lived when this growth-line represented the
edge of the valve. Now attachment of epizoa in such abundance as is usual could
scarcely have happened if the Lima had been surrounded in life by a nest of byssal
threads. The occurrence of Lopha sp. in the Muschelkalk has been previously noticed
by Seilacher (1954) and Cox (1932).

Furthermore, there is evidence of a fixational byssus in Lima lineata , which, as pointed
out above, probably does not occur in Recent nest-builders. Seilacher (1954, p. 176)
noticed that small, single-valved shells of various species were often present in the lunules
of his specimens (cf. PI. 59, fig. 6). Near Freudenstadt (text-fig. 1), from which most of
Seilacher’s material came, this phenomenon was quite regular (about 30% of speci-
mens). It is clearly different from the attachment of double-valved young in the lunule,
for the fact that the shells are single-valved and of many different species proves that
they became associated with the L. lineata only after their death. Seilacher's explana-
tion, that they represent debris attached to the byssus of the Lima, seems the only
reasonable one.

EXPLANATION OF PLATE 59

Figs. 1-7. Lima lineata (Schlotheim). 1, Antero-dorsal aspect. X 1. Diedesheim. Heidelberg Univer-
sity collection. Note young specimen at extreme anterior end of the adult lunule, oriented with its
lunule to the floor of that of the adult. 2, Antero-dorsal aspect. xO-75. Leimen. Heidelberg Univer-
sity collection. Remarks as for fig. 1 . 3, Oblique antero-dorsal aspect. xO-75. Rohrbach. Heidel-
berg University collection. Remarks as for fig. 1 . 4, Antero-dorsal aspect. xO-75. Lengfurt. British
Museum no. LL8481. This specimen was found in situ in the quarry with the antero-dorsal margin
downwards. The particularly coarse costation of the anterior umbonal ridges and the byssal gape
between the two halves of the lunule are clearly visible. Of the two young individuals in the lunule,
one is at the extreme anterior end and the other is as anterior as the first member of the pair would
allow. Both have the lunule pressed close to the floor of the adult lunule. 5, Antero-dorsal aspect.
xO-75. Leimen. Heidelberg University collection. Remarks as for fig. 1. 6, Oblique antero-dorsal
aspect. xO-75. Giingrich near Pirmasens. Heidelberg University collection. Note the presence of
abundant byssally attached debris at the posterior end of the lunule, and the shallowness of the
lunule, possibly due to the persistence of the byssus in the adult. The debris includes two valves of a
young Lima lineata (at a) and a small gastropod b as well as other material. The specimen of
Anodontophora fassaensis (Wissmann) at c was probably part of the byssally attached debris but has
been displaced by compaction of the sediment. 7, Oblique lateral aspect. X 0-75. Lengfurt. British
Museum no. LL8482. This figure shows the specimen resting on the antero-dorsal margin in the
orientation in which it was found in situ. The epizoa are Placunopsis plana (Giebel), Lopha sp. and
small bryozoa.

Fig. la. Lopha sp. X 2-25. Enlarged view of specimen also visible in fig. 7. Note the truncation against
a growth-line of the ‘ host ’ which proves that the Lopha must have lived when the ‘ host ’ was of the
size represented by the growth line.

Figs. 1-3 and 5 are by Dr. A. Seilacher and figs. 4 and 6-la are by J. Pope.

Palaeontology, Vol. 3.

PLATE 59

JEFFERIES, Lima lineata ( Schlotheim) and Lopha sp.

R. P. S. JEFFERIES: LIMA LINEATA (SCHLOTHEIM)

365

However, though L. lineata was probably not a nest-builder, its young had a habit of
entering crannies and fixing themselves inside and it was probably from this that the
Limid habit of building nests inside cavities arose. This is not to suggest, of course, that
Lima lineata was the direct ancestor of nest-building Limidae; but if L. lineata was
photonegative the ancestor of ‘ Mantellum' may also have been so.

text-fig. 1. Locality map. Young inside the lunule are abundant near Heidelberg and Wurzburg but
apparently absent near Freudenstadt. Byssally attached debris, on the other hand, is abundant near
Freudenstadt but rare near Heidelberg and Wurzburg.

Text-fig. 2 shows the relation between habit and ‘length’ in Lima lineata from the
Wellenkalk, near Heidelberg and Wurzburg, ‘length’ being the largest dimension
parallel to the antero-dorsal margin. One hundred and one specimens have been used
in the construction of the figure (not counting young specimens inside the lunules of
other specimens) and their localities of origin are as follows:

(i) Lengfurt near Wurzburg, 3 specimens (collected by the author, British Museum
numbers LL8481-3).

(ii) Leimen, 54 specimens; Diedesheim, 35; Auerbach, 4; and 1 specimen each from
Tauberbischofsheim, Neckarelz, Rohrbach, Eschelbronn, and Hoffenheim (in the
Heidelberg University collection).

The sample presented in text-fig. 2, as pointed out above, is part of a population
which died suddenly. For this reason the age and size distribution of the sample, apart

366

PALAEONTOLOGY, VOLUME 3

Inside Sunule. Not in lunule. One young

‘Length8 No young. in lunule.

135.
I30«
125*
120.
IIS .

no.

105"

IOO.

95*

90.

85-

80*

75-

70.

66.

60.

55*

50.

45-

40*

35.

30-

25.

20

IS-

O-

S'

cm J

Each dot represents one specimen.

TWo young
in Sunule.

text-fig. 2. Distribution of habit with ‘length’ in Lima lineata from the Wellenkalk near Heidelberg
and Wurzburg. ‘Length’ is the greatest dimension parallel to the antero-dorsal margin and is plotted

to the nearest 5 mm.

from the considerable vagaries of preservation and collection, should be the same as in a
living population of the species. It therefore seems legitimate to conclude that in the
Heidelberg-Wurzbtirg area a young Lima lineata did not try to enter an adult lunule
until it was about 7 mm. long. Perhaps below this size it was not photonegative. Although
in the figure almost 100% of the specimens between 5 and 20 mm. long are found inside
lunules, the percentage was probably very much less in life. Unprotected individuals of
this size would probably have been disarticulated after death and so would not have
been included in the sample which was purposely restricted to double-valved speci-
mens. Also it is clear that small Limas inside adult lunules would be much more easily

367

R. P. S. JEFFERIES: LIMA LINEATA (SCHLOTHEIM)

trapped by sudden deposition of sediment than would similar individuals living outside.
At about 20 mm. the young left the lunules, which had presumably become uncomfort-
ably small. Then, at a length of about 55 mm. or greater, the animals started in their
turn to serve as ‘host’ to young, and this continued till death.

text-fig. 3. Relationship between ‘length’ of young in lunule and ‘length’ of ‘host’ adult, based on
all observed specimens with young in the lunule from the Wellenkalk. ‘Length’ is the greatest dimen-
sion parallel to the antero-dorsal margin.

Text-fig. 3, which is based on all observed specimens from the Wellenkalk with young
in the lunule, includes two specimens from Giingrich near Pirmasens besides those
used in the construction of text-fig. 2. It shows that the size of a young specimen was
related to the size of the ‘host’, i.e. roughly speaking the adult was six times as large as
the associated young. At first this might suggest that a young Lima in the lunule of a
large ‘host’ had attached itself when the ‘host’ was smaller and grown pari passu with it.
This is probably not the correct explanation, however, for the percentage linear growth-
rates of the adults would probably be negligible by comparison with those of the young.
It seems more likely that young Limas were capable of leaving a lunule they had out-
grown, and that then they either lived free, or found more spacious accommodation.

The fact that Seilacher did not find double-valved young inside the lunules of his
specimens from near Freudenstadt proves that they must be very rare there. On the
other hand, text-fig. 2 shows that near Heidelberg and Wurzburg thirty-three specimens
out of a total of seventy-three longer than 50 mm. had one or two young inside the
lunule. This is probably linked with the fact that, as mentioned above, 30% of Freuden-
stadt specimens had byssal debris in the lunule while stray fragments which might

368

PALAEONTOLOGY, VOLUME 3

represent such debris were present in only four of the seventy-three specimens from near
Heidelberg and Wurzburg. Even though such debris will often have been removed during
preparation the difference is probably significant. It suggests that perhaps the typical
adult near Heidelberg lay free on the sea-floor, whereas the typical adult near Freuden-
stadt was byssally fixed. This is not improbable, for the Wellenkalk near Freudenstadt
is of more marginal facies and only half as thick as near Heidelberg and Wurzburg
(55 m. as opposed to 98 m. (Schmidt 1928, opp. p. 24)) and so the sea near Freudenstadt
was probably shallower and more turbulent.

Persistent byssal attachment in the adult might affect the young in two ways. Firstly
there would be a mass of byssus and byssally attached debris blocking the gap between
the umbones of the adult and making it difficult for the young to enter the lunule.
Secondly there was probably a tendency for byssal fixation to be correlated with a very
shallow lunule. Thus the specimen from Giingrich shown in PI. 59, fig. 6, has byssal
debris and a much shallower lunule than is usual near Heidelberg (cf. PI. 59, figs. 1-5)
and the same also seems true of the specimens with byssal debris from near Freudenstadt
of which Seilacher has published drawings (1954, fig. 7, p. 177). These shallow lunules,
which doubtless resulted from the pull of the byssus on the lunular mantle edge, would
probably not have provided adequate space for a young Lima even if it had succeeded
in entering.

Byssal attachment persisted in the Heidelberg-Wurzburg area at least up to a length of
30 mm. This is shown by two specimens of this size in the Heidelberg collection from
Diedesheim which are pressed lunule downwards against the inner concave side of a
large valve of Lima lineata. It is at least possible that byssal attachment persisted up to
a length of about 55 mm., at which length the first young are found in the lunule (text-
fig. 2).

Before leaving this subject, however, it is as well to note that one specimen in the
Heidelberg collection from Diedesheim has both a young in the normal position in the
lunule and a shell fragment which might represent byssal debris. On the other hand, it is
quite possible that the position of this particular fragment is quite accidental.

The southern boundary of the area in which young occur in the lunule is not at pre-
sent clear. However, single specimens in the Heidelberg collection from Grotzingen and
Pforzheim (text-fig. 1) were without young.

Young specimens do not occur in the lunules of Lima striata (Schlotheim) in the
Heidelberg collection. The lunule of this Upper Muschelkalk species was shallower than
is usual in L. lineata and this perhaps suggests permanent byssal attachment. The Wellen-
kalk species L. radiata Goldfuss sometimes has young in the same position and orienta-
tion as L. lineata.

The conclusions of this paper can be summarized as follows :

1 . The occurrence of young in the lunules of Lima lineata is not accidental since their
positions and orientations are almost constant.

2. In life L. lineata rested with the lunule on the sea-floor but the lunule was open to
the sea by the gap between the umbones. The young could only have entered the lunule
with such frequency if they had been photonegative with a tendency to enter crannies.

3. L. lineata was not a nest builder since it resembles Recent non-nest-building
Fimidae, had a fixational byssus, and was usually covered with epizoa which attached
during the Lima's, lifetime and could not have penetrated a nest of byssal threads.

R. P. S. JEFFERIES: LIMA LIKEATA (SCHLOTHE1M)

369

4. Nevertheless, the nest-building habit could well have evolved from the habit of
byssal attachment inside crannies shown by young L. Iineata, for Limid nests are pre-
ferentially made inside cavities.

5. Tn the Heidelberg area young Limas inside lunules are 10-20 mm. long. A sub-
stantial proportion of individuals more than 55 mm. long acted as ‘host’. And the
young was usually about one-sixth the length of the associated adult.

6. Adult L. Iineata in south Germany were probably byssally attached. In consequence
the lunule was shallow and the gap between the umbones was blocked by byssus and
byssal debris so that young could not enter the lunule. Farther north the adults were not
byssally attached and so the young could enter.

Acknowledgements. It is a pleasure to thank Dr. and Frau A. Seilacher of Frankfurt University for
hospitality and extensive discussion. I should also like to thank Dr. Rosier of Heidelberg for allowing
me to examine the collections at that University, Mr. G. F. Elliott for help with the literature, and
Dr. R. G. S. Hudson for advice on presentation. The photographs in PI. 59 are partly the work of
Dr. Seilacher and partly of Mr. J. Pope of the Iraq Petroleum Company. Mr. G. Rosser provided
advice on statistics.

REFERENCES

arkell, w. j. 1929-37. A monograph of British Corallian lamellibranchia. Palaeontogr. Soc.

cox, l. R. 1932. Further notes on the Transjordan Trias. Ann. Mag. Nat. Hist. (10), 10, 93-113, pi. 7.

crozier, w. j. 1921. Notes on some problems of adaptation. 5. Phototropism in Lima. Biol. Bull. 41,
102-5.

douville, h. 1910. Observations sur les ostreides. Origine et classification. Bull. Soc. Geol. France
(4), 10, 634-45.

macginitie, g. e. and macginitie, n. 1949. Natural history of marine animals. New York.

pelseneer, p. 1911. Les Lamellibranches de l’expedition du Siboga. Partie anatomique. Siboga
Exped. Monogr. 53a, 1-125, pi. 1-6.

philippi, h. 1900. Beitrage zur Morphologie und Phylogenie der Lamellibranchia. 3. Lima und ihre
Untergattungen. Zeit. Deutsch. Geol. Ges. 52, 619-23, pi. 24.

schmidt, m. 1928. Die Lebewelt unserer Trias. 461 pp., Ohringen.

seilacher, a. 1954. Okologie der triassischen Muschel Lima Iineata (Schlot.) und ihrer Epdken.
Neues Jahrb. Geol. Pal. ( Monatsh .), 163-83.

studnitz, g. v. 1931. Die Morphologie und Anatomie von Lima inflata, der Feilenmuschel, nebst
biologischen Untersuchungen an Lima hians Gmel. Zook Jahrb. Abt. Anat. Out. Tiere, 53,
199-316.

R. P. S. JEFFERIES

British Museum (Natural History),
London S.W. 7

Manuscript received 3 November 1959

THE CRETACEOUS SPECIES OF PYRIPORA
D’ORBIGNY AND RHAMMATOPORA LANG

by H. DIGHTON THOMAS and G. P. LARWOOD

Abstract. The Cretaceous species of Pyripora d’Orbigny and Rhammatopora Lang are revised, and Rhamma-
topora gasteri sp.nov. is described.

In a recent publication (Thomas and Larwood 1956) we have discussed the status of the
genera Pyripora d'Orbigny and Rhammatopora Lang. We have now investigated in
greater detail the Cretaceous species of these genera, in the course of which we have
examined several hundred zoaria of species of Pyripora. This examination has shown,
inter alia, that P. [Herpetopora] anglica Lang is distinct from P. [H.] clavigera Lang,
a junior synonym of P. laxata (d’Orbigny), and that the two species may be differen-
tiated on the ratio of the length and width of the aperture.

Most of the zoaria of Pyripora which we have examined encrust comparatively smooth
substrates. Commonly these are fragments of Inoeeramus shell (40%), pieces of echinoid
test (40%), or fragments of Ostrea valve (10%). Less frequently zoaria are also found
encrusting belemnite guards (5%) and occasionally the remains of other organisms (5%),
e.g. brachiopod shell fragments, corals, crinoid plates, and other polyzoa (percentages
approximate).

We wish to thank Mr. A. G. Brighton and Mr. R. R. Clarke for the loan of material from the Sedg-
wick Museum, Cambridge, and the Norwich Castle Museum, respectively; Prof. E. Voigt, of the
Geologisches Staatsinstitut, Hamburg, for the loan of German specimens, for information readily given,
and for permission to reproduce his photograph of the type specimen of Hippothoa laxata d’Orbigny;
and Dr. Anna B. Hastings, of the Zoological Department, British Museum (Nat. Hist.), for discussions
on the Recent species of Pyripora. Except where otherwise stated, the photographs are by Messrs. M. G.
Sawyers and J. V. Brown of the British Museum (Nat. Hist.).

The specimens are referred to by their registered numbers in the various museums : those prefixed by
the letter D (e.g. D. 2062) are in the British Museum (Nat. Hist.).

Pyripora d’Orbigny

Diagnosis. Zoarium encrusting, uni- or pluri-serial ; zooecia pyriform, budded distally
and uni- or bi-laterally or both, proximally tapered, often with a very slender tubular
proximal cauda; aperture longitudinally oval; cryptocyst usually present forming
a proximal shelf or a narrow shelf round most of the opesia; gymnocyst well developed
proximally; no calcified basal wall; no ovicells, spines, or avicularia.

Remarks. In all the specimens of the Recent species Pyripora catenularia (Fleming), and
in all the hundreds of zoaria, and therefore thousands of zooecia of P. laxata which we
have examined, no ovicells were seen. Hence it is certain that ovicells are not developed
in the genus. We therefore exclude from Pyripora those species which resemble that
genus in other characters but possess ovicells.

We have found that, although the size of zooecia may vary considerably in a given

[Palaeontology, Vol. 3, Part 3, 1960, pp. 370-86, pis. 60-62.]

H. D. THOMAS AND G. P. LARWOOD: PYRIPORA D'ORBIGN Y

371

zoarium of a species of Pyripora, the ratio of the length and width of the aperture varies
essentially within narrow limits, and this ratio, in conjunction with the lengths of the
zooecia (excluding the caudae), is of great value in recognizing the species.

We reject certain species from Pyripora (see p. 381) and include only those given in the
following key.

Key to the Cretaceous species of Pyripora

1 . Majority of caudae narrow and longer than the rest of the zooecia.

(A) Lc commonly well over 1 00 mm.

( а ) ha = 0-25 to 0-43 mm., la = 0T4 to 0 25 mm., apertural

ratio 1-40 to 2-35 ....... P. laxata (d’Orbigny)

(б) ha = 0-44 to 0-69 mm., la = 0T7 to 0 27 mm., apertural

ratio 2-22 to 3-29 . . . . . . P. cinglica (Lang)

(B) Lc = 0T6 to 0-52 mm., ha = 0T0 to 0T5 mm., la = 0-06 to

0 07 mm. ......... P. filum (Voigt)

2. Majority of caudae broad and shorter than the rest of the zooecia.

(A) Zoarial budding uni- and bi-lateral occasionally becoming

pluriserial; caudae nearly as long as the rest of the zooecia P. texana Thomas and Larwood

(B) Zoarial budding uni- or bi-lateral never becoming plui iserial ;
caudae distinctly shorter than the rest of the zooecia.

(a) Zoarial budding unilateral; zooecia markedly rounded
distally; Lz = 1-20 to L50 mm., Iz = 0-50 mm.; ha =

0-40 mm.; caudae short, very narrow proximally . . P. faxensis (Voigt)

(b) Zoarial budding uni- and probably bi-lateral; zooecia
more pointed distally; Lz = 050 to 0-75 mm., lz = 0-35
to 0-50 mm.; ha = 0-40 to 0-70 mm.; caudae very short
and broad . . . . . . . . P. parvicauda (Voigt)

3. Caudae absent. Zooecia very small, Lz = 0-30 to 0-35 mm.,

lz = 0-25 to 0-30 mm.; gymnocyst very narrow . . P.filimargo (Voigt)

Pyripora laxata (d’Orbigny)

Plate 60, figs. 1,2; Plate 61, fig. 2

Hippothoa laxata d’Orbigny 1852, pi. 71 1, figs. 12-15; 1853, p. 386.

Hippothoa cruciata Reuss 1854, p. 134, pi. 28, fig. 1.
non Hippothoa desiderata Novak 1877, p. 86, pi. 2, figs. 1, 2.

Hippothoa labiata Novak 1877, p. 86, pi. 3, figs. 1-5.

Herpetopora clavigera Lang 1914«, p. 6, pi. 2, figs. 4, 5 : Voigt 1930, pp. 409, 553, pi. 1, figs. 2. 3;

1949, p. 10, pi. 1, figs. 4-6.

Herpetopora cruciata (Reuss); Lang 1914a, pp. 6, 7.

Herpetopora dispersa (von Hagenow); Voigt 1930, pp. 409, 553, pi. 1, fig. 1 ; [non Cellepora dis-
persa von Hagenow 1839, p. 280; 1846, p. 629, pi. 23 b, fig. 55].

Herpetopora comptoniensis Brydone 1936, p. 88, pi. 42, fig. 17.

Pyripora cruciata (Reuss); Thomas and Larwood 1956, p. 375 [part ini],

Lectotype (here chosen). Specimen encrusting an Echinocorys test. Senonian, zone of Belemnitella
mucronata, Meudon, south-west of Paris. D’Orbigny Collection No. 7911, Museum National d’His-
toire naturelle, Paris.

Diagnosis. Pyripora in which the lateral budding is usually bilateral but sometimes
unilateral; zooecia inflated, 0-24 to 0-45 mm. wide, with steep lateral walls, distally
rounded or slightly tapering, proximally tapered; calcification of frontal membrane

372

PALAEONTOLOGY, VOLUME 3

and of operculum, and regeneration, relatively common; aperture longitudinally oval,
ha = 0-25 to 043 mm., la = 0-14 to 0-25 mm., apertural ratio = 140 to 2-35 but
mostly 1-7 to 1-9; cryptocyst a narrow shelf extending about three-quarters of the way
round the opesia; caudae very narrow, tubular, varying greatly in length but usually
much longer than the rest of the zooecia, expanding rapidly into the triangular proximal
gymnocyst of the zooecium; triangular kenozooecia, with a rounded pore, occasionally
occur.

Description. Zoarium encrusting, with distal-median, and usually bilateral, but some-
times unilateral, budding. The distal-median budding is usually single, but very occa-
sionally it may be double. Bilateral budding is more common than unilateral, though
both occur at times in the same zoarium. This budding is usually from the distal-lateral
walls of zooecia and may be at right angles to them or directed distally and obliquely at
a lesser angle. In any one series of zooecia unilateral budding may occur on either side.
Sometimes the budding, whether uni- or bi-lateral, may develop from the caudae.

Caudae slender, tubular and of very variable length — in rare instances absent or very
short. The caudae may, however, be several times the length of the rest of the zooecia.
In any series of zooecia successive caudae are commonly increasingly longer. They are
generally straight but may be variously curved. They rapidly expand into the triangular,
proximal, smooth gymnocyst of the zooecia: the length of the gymnocyst varies, but it
is generally well developed.

Zooecia pyriform, longitudinally oval, inflated; lateral and distal walls steep, the latter
rounded or slightly pointed distally; calcified basal walls absent in the fossil state. Aper-
ture longitudinally oval, of smaller size but relatively wider than in Pyripora anglica
(Lang) (cf. text-figs. 1 b-d). The apertural ratio = 140 to 2-35 but is mostly T7 to

EXPLANATION OF PLATE 60

Figs. 1, 2. Pyripora laxata (d’Orbigny). 1, Type specimen, Mus. nat. Hist. nat. Paris, d’Orbigny
Collection No. 791 1, encrusting test of Echinocorys, normal zooecia with, at left, one zooecium with
calcified frontal membrane with central pore, x20. Photograph by Prof. E. Voigt. 2, Part of the
holotype of Herpetopora clavigera Lang, Brit. Mus. (Nat. Hist.) 40246, several series of uni- and
bi-laterally budded zooecia which have overgrown each other; one zooecium has a calcified frontal
membrane and operculum, X 20.

Figs. 3, 4. Pyripora anglica ( Lang). 3, Paratype, D. 10074, uni- and bi-laterally budded zooecia, one
with a calcified frontal membrane and operculum, and one heterozooecium, x20. 4, Holotype,
D. 11359, uni- and bi-laterally budded zooecia, some with the frontal membrane and operculum
calcified and some regenerated. The linear series of partly developed zooecia is typical of some
zoaria, X 20.

EXPLANATION OF PLATE 61

Fig. 1. Pyripora anglica (Lang), paratype, D. 21957. Part of a uni- and bi-laterally budded zoarium
showing the tendency for successive zooecia to develop increasingly longer caudae, the occurrence of
calcified zooecia in linear series and a triangular heterozooecium, X 20.

Fig. 2. Pyripora laxata (d’Orbigny). Holotype of Herpetopora comptoniensis Brydone, Sedgwick
Museum, B. 36997. Normal zooecia, one with a calcified frontal membrane, X 20.

Figs. 3, 4. Pyripora filum (Voigt), Voigt Collection No. 1532. 3, Normal zooecia with thread-like
caudae and strongly inflated distal portions encrusting branches of a cyclostomatous polyzoan, X 35.
4, Two smaller normal zooecia of the same specimen with shorter caudae, x 35. Photographs by
Department of Photography, University of Durham, King’s College, Newcastle upon Tyne.

Fig. 5. Pyripora texana Thomas and Larwood, holotype, D. 40541 . Uniserially budded normal zooecia
in part of the zoarium, x 20.

Palaeontology, Vol. 3.

PLATE 60

THOMAS and LARWOOD, Pyripora

Palaeontology, Vol. 3.

PLATE 61

T H O M AS and L A R W O O D, Pyripora

H. D. THOMAS AND G. P. LARWOOD: PYROPORA D'ORBIGNY 373

1-9. Cryptocyst narrow, slightly descending, extending most of the way round the aper-
ture and producing an oval opesia only slightly smaller than the aperture. Calcification
of the frontal membrane and of the operculum is common. It is often complete, but at
times incomplete, leaving a small, circular, uncalcified portion near the middle, showing
that the calcification has spread from the margin inwards. The calcified opercula are
small, straight sided, and evenly rounded distally. Lateral budding frequently, but not
always, takes place from zooecia which are thus calcified. Regeneration of zooecia is
fairly common and may occur more than once in a given zooecium (text-fig. \e). Some-
times regenerated zooecia have the frontal membrane and operculum completely cal-
cified (text-fig. 1 f-g).

Kenozooecia are fairly common, and are inflated, triangular, and apparently vicarious :
they usually have a rounded, median pore and often bud to give normal zooecia.

Avicularia, spines, and ovicells absent.

Measurements (see text-fig. 1 a)

Zooecia Lz = 0-66 to 5-18 mm. (or more). ho — 0-23 to 0-41 mm.

lz = 0-24 to 0-45 mm. lo = 0T3 to 01 4 mm.

ha = 0-25 to 0-43 mm. Lc = up to 4 00 mm. (or more).

la = 0T4 to 0-25 mm. lc = about 0 07 mm. (unless absent).

Apertural ratio = 1-40 to 2-35, but mostly 1-7 to 1-9.

Measurements of a large number of zooecia have been made on specimens ranging
from the Cenomanian to the Danian. Lz is widely variable, for it includes measurement
of the cauda as well as the rest of the zooecium; lz, ha, la, ho, and lo are less variable
than Lz or the caudal measurements (Lc, lc), and are consequently of greater diagnostic
value. Lc is extremely variable; caudae may be lacking, or they may be up to three or
four or more times as long as the rest of the zooecia; the caudae are, however, always
very narrow.

Remarks. This revision shows that wide variation may occur not only between zoaria but
also within the zoaria themselves.

In 1839 von Hagenow briefly described (p. 280) a polyzoan from Rugen as Cel/epora
dispersa : unfortunately he gave no figure. Subsequently (1846, p. 629, pi. 23 b, fig. 55),
he gave a more detailed description of the species, accompanied by a figure, on which the
interpretation of Cel/epora dispersa must be based. In the upper right-hand part of the
figure he showed three circular heterozooecia which can be matched in a species from
Rugen which Lang (19146, p. 439, pi. 34, fig. 3) described as Marssonopora dispersa (von
Hagenow). We agree with Lang’s interpretation. Thus, in contrast to Voigt’s opinion
(1930, p. 409) von Hagenow’s specific name dispersa is not available for the Chalk
species lacking avicularia.

Lang (1914a, p. 7) considered Hippothoa gracilis d'Orbigny (1852, pi. 711, figs. 12-15;
1853, p. 386) to be a species of his genus Herpetopora, but the nature of its orifice and
the occurrence of ovicells show clearly that it does not belong to that genus, and, there-
fore, not to Pyripora.

Prof. E. Voigt has recently sent us photographs of the type specimen of Hippothoa
laxata d’Orbigny (1852, pi. 711, figs. 12-15; 1853, p. 386), d’Orbigny Collection No. 7911.
From these it is clear that the species is the same as Herpetopora clavigera Lang of which

b b

B 6C12

374

PALAEONTOLOGY, VOLUME 3

text-fig. 1. a. Measurements: Lz = length of zooecium; lz = width of zooecium; ha = length of
aperture ; la = width of aperture ; ho = length of opesia ; lo = width of opesia ; Lc = length of cauda ;
lc = width of cauda. b, c, Distal ends of two normal zooecia of Pyripora anglica (Lang) : ( b and c from
D. 40489). d-g. Distal ends of zooecia of P. laxata (d’Orbigny): ( d — D. 40135, normal zooecium;
e — D. 40158, repeatedly regenerated; f— D. 40364, repeatedly regenerated with completely calcified
frontal membrane and operculum; g — D. 40140, regenerated with completely calcified frontal mem-
brane and operculum).

we have much well-preserved material. Hippothoa laxata d’Orbigny therefore becomes
the first available name for the species dealt with here.

Hippothoa desiderata Novak (1877, p. 10, pi. 2, figs. 1, 2) was placed by Lang (1914n,
p. 7) in Herpetopora, but Hippothoa desiderata was described by Novak as possessing
small, hemispherical ovicells. The species is thus excluded from Pyripora.

Among the large number of specimens from the English, and some foreign, Chalk
which we have examined there are variations in the branching and in the caudal and

H. D. THOMAS AND G. P. LARWOOD: PYRIPORA D’ORBIGNY

375

other zooecial characters which show that Hippothoa laxata d’Orbigny (1852; 1853),
H. cruciata Reuss (1854), H. labiata Novak (1877), H. dispersa Marsson (1877) non von
Hagenow, Herpetopora clavigera Lang (1914 a) and H. comptoniensis Brydone (1936)
are all synonymous.

The species is widely distributed in Europe — see the localities to which the references
in the synonymy apply. Prof. E. Voigt has recently sent us a photograph of a specimen
encrusting a Gryphaea from the Maastrichtian of Burunduk Kaja, Crimea (Coll. Inst.
Hist. Geol. Fac. Geol. Moscow, No. 8084-5/6), which he had received from Prof. Najdin
of Moscow. He has also told us of a specimen he had found on an Echinocorys from the
Upper Maastrichtian of Hallembaye, near Haccourt, north-east of Liege, Belgium.

Stratigraphical distribution. Cenomanian, zone of Schloenbachia various, to Danian
(see remarks on Pyripora anglica).

Specimens. A. British. From numerous localities in south-east and east England, ranging from Dorset
and the Isle of Wight to Lincolnshire. (1) British Museum (Natural History) : ( a ) Zone of Belemnitella
mucronata — 43 specimens including 40246, D. 3243, D. 20607-8, D. 28898, D. 40135, and D. 40140.
(b) Zone of Gonioteuthis [Actinocamax ] quadrata — 169 specimens including D. 2647, D. 3244,
D. 20610, D. 28081-100, D. 28795, D. 29054, D. 29063, D. 40158, and D. 40364. (c) ZomofOffaster
pilula — D. 28917, D. 29842-5, D. 40532-3. (d) Horizon and locality unknown— D. 35352. (2) Nor-
wich Castle Museum: Zone of B. mucronata — N.C.M. 76.937(34 — 39). (3) Sedgwick Museum:
(a) Zone of Ostrea lunata — B. 60756-60, B. 61001-2, B. 80750-60. ( b ) Zone of B. mucronata — B.
80676, B. 80691-749. (c) Zone of G. [A.] quadrata — B. 80675, B. 80677-88. (d) Zone of B. mucronata or
G. [A.] quadrata — B. 80689-90. (<?) Zone of O. pilula — B. 80674. (/) Zone of Schloenbachia varians—
B. 36997, type specimen of Herpetopora comptoniensis Brydone.

B. Danish. Danian. Faxe, Seeland— ?D. 9113, ?D. 29014, ?D. 40537-8.

C. German specimens lent by Prof. E. Voigt, Geologisches Staatsinstitut, Hamburg, (a) Upper
Maastrichtian, Hemmoor, Coll. Voigt 3171, 3173. (b) Lower Maastrichtian, Hemmoor, Coll. Voigt
3166; Luneburg (Zeltburg), Coll. Voigt 3167, 3168; and Riigen, Coll. Voigt 3163-5. (c) Maastrichtian,
division unspecified, Hemmoor, Coll. Voigt 3172. (d) Zone of B. mucronata or G. [A.] quadrata,
Grube Alsen, near Lagerdorf, Coll. Voigt 3169. (e) Zone of G. [A.] quadrata, Lagerdorf, Coll. Voigt
3170.

Pyripora anglica (Lang)

Plate 60, figs. 3, 4; Plate 61, fig. 1

Herpetopora anglica Lang 1914u, pp. 6, 7, pi. 2, figs. 1-3; 19146, p. 436; 1921, p. 83; 1925, p. 229,
pi. 20, fig. 7, pi. 21, fig. 18: Canu andBassler 1920, p. 81, text-fig. 22; 1923, p. 18, text-fig. lc;
Bassler 1935, p. 124; 1953, p. G157, text-fig. 119, 8.

Pyripora cruciata (Reuss) Thomas and Larwood 1956, p. 370 [ partim ].

Holotype. D. 11359. Specimen encrusting large fragment of Volvinoceramus involutus. Senonian, base
of zone of Micraster coranguinum or top of zone of M. cortestudinarium, Chatham, Kent; W. Gamble
Collection.

Diagnosis. Pyripora like P. laxata (d'Orbigny) but ha = 0-44 to 0-69 mm., la = 0T7 to
0-27 mm., apertural ratio = 2-22 to 3-29, but mostly about 2-5.

Measurements

Zooecia Lz = 0-69 to 4 89 mm. (or more). ho = 0-42 to 0-67 mm.

lz = 0-25 to 0-47 mm. lo = 01 6 to 0-26 mm.

ha = 0-44 to 0-69 mm. lc = 0 to 3-87 mm. (or more).

la = 0T7 to 0-27 mm. lc = about 0 08 mm., unless absent.

Apertural ratio = 2-22 to 3-29 (mostly about 2-5).

376 PALAEONTOLOGY, VOLUME 3

Remarks. P. anglica (Lang) very closely resembles P. la.xata (d’Orbigny) in its zoarial
and zooecial characters. The mode of budding of the zoarium is the same; caudae are of
the same narrow type though apparently very slightly wider, frequently of increasing
length in a given single series ; but the zooecia, excluding the caudae, are usually longer.
The aperture is larger and distinctly longer and narrower than in P. laxata (cf. text-
figs. 1 b-d), and the gymnocyst proportionately smaller. Other characters, such as the
presence of a narrow, proximal cryptocyst, the occurrence of series of sealed zooecia,
and the presence of regenerated zooecia and triangular kenozooecia are the same.

Lang (1914a) diagnosed Herpetopora anglica as having ‘zooecia that taper somewhat
distally as well as proximally; with regularly elliptical termens [apertural margins]’. As
distinct from this, he defined H. clavigera as a ‘ Herpetopora with zooecia hardly tapering
distally and consequently more or less pear-shaped, having their outlines bounded by
steeply curved sides . . .’. These distinctions in the general shape of the zooecia of H.
anglica and H. clavigera are not clear either in the new material which we have examined
or in Lang’s original specimens. In his key to the genus Herpetopora (1914a, p. 7) Lang has
described the termen [apertural margin] as ‘elliptical rather than oval’ in H. anglica ;
in other words the aperture is longer and narrower in this species, the apertural ratio
being about 2-5. This and the somewhat larger zooecia are the only significantly constant
differences from H. clavigera Lang, and are clearly confirmed here by many measure-
ments on abundant material of both species.

P. anglica (Lang) is less common than P. laxata (d’Orbigny) and has a more restricted
stratigraphical range in the English Chalk. Lang (1913, p. 172) has referred to Herpeto-
pora clavigera [ Pyripora laxata (d'Orbigny)] and to Herpetopora anglica [= Pyripora
anglica (Lang)] as species which ‘divide the English Chalk between them at the junction
of the quadratus and Marsupites zones’. Our examination of these species supports Lang’s
statement but has shown that Pyripora laxata (d’Orbigny) first occurs in the Cenomanian
where it is represented by one well-preserved zoarium, Sedgwick Museum, B. 36997
(the type specimen of Herpetopora comptoniensis Brydone). The species next occurs in
the zone of Offaster pilula and becomes very abundant in the succeeding zones, possibly
surviving into the Danian from which we have seen four rather poorly preserved speci-
mens. In contrast, Pyripora anglica (Lang) first occurs in the zone of Terebratulina lata
and becomes more abundant in the zones of Holaster planus, Micr aster cortestudinarium,
and M. coranguinum, and survives, much reduced in numbers, into the upper part of the
zone of Offaster pilula (see text-fig. 2).

Stratigraphical distribution. Turonian, zone of Terebratulina lata, to Senonian, zone of
O ffaster pilula.

Specimens. From numerous localities in south-east and east England. (1) British Museum (Natural
History): ( a ) Zone of Offaster pilula — D. 28239, D. 29846, D. 40529-31. ( b ) Zone of Marsupites
testudinarius — D. 29847, D. 40526-8. (c) Zone of Micraster coranguinum — D. 4189, D. 8213, D. 10074,
D. 11126, D. 21218, D. 21955-8, D. 24361-2, D. 28240-1, D. 28899-900, D. 29098, D. 31825, and 22
other specimens, (d) Zone of M. cortestudinarium — D. 8214, D. 11359, D. 21949-51, D. 28242-3,
D. 29847, D. 40504-12. (e) Zone of M. coranguinum or M. cortestudinarium — D. 11372. (/) Zone of
Holaster planus — D. 8387, D. 21219, D. 29848, D. 40489, and 30 other specimens, (g) Zone of
Terebratulina lata — D. 20611, D. 40332-4. (//) Horizon and locality unknown — D. 38849. (2) Sedg-
wick Museum: (a) Zone of Marsupites testudinarius — B. 80662-7, B. 80670-3. (b) Zone of Micraster
cortestudinarium — B. 80655-60.

Aperfural Ratio -ha/la

text-fig. 2. Stratigraphical distribution of Pyripora laxata (d’Orbigny) and P. anglica (Lang). The
range in value of apertural ratios is shown for 415 specimens from the Chalk. The figures to the left
of each line indicate the number of measurable specimens available to us from each zone. (Maasn. — -
Maastrichtian; 0.1. — zone of Ostrea lunatcr, B.m. — zone of Belemnitella mucronata ; G.[A.]q. — zone of
Gonioteuthis [Actinocamax] quadrat a; O.p. — zone of Off aster pilula ; E.c. — subzone of Echinocorys
scutata var. cincta; E.d. — subzone of E. scutata var. depressula; M. — zone of Marsupites testudinarius;
M.a. — subzone of M. testudinarius ; U. — subzone of Vintacrinus westfalicus ; M.c.a. — zone of Micraster
coranguinum ; M.c.t. — zone of M. cortestudinarium ; H.p. — zone of Ho/aster planus', T.l. — zone of
Terebratulina lata ; R.c. — zone of Orbirhynchia [ Rhynchonella ] cuvieri', H.s.g. — zone of Holaster sub-
globosus; S.v. — zone and subzone of Schloenbachia varians; S.c. a. — subzone of Stauronema carteri).

378

PALAEONTOLOGY, VOLUME 3

Pyripora filum (Voigt)

Plate 61, figs. 3, 4

Herpetopora filum Voigt 1930, pp. 410, 553, pi. 1, figs. 8-10.

‘Inon Herpetopora filum Voigt; Balavoine 1956, p. 157, pi. 6, fig. 2.

Type specimen. That figured by Voigt 1930, pi. 1, fig. 10. Maastrichtian. Maastricht.

Diagnosis. Pyripora with small, strongly inflated zooecia; apertural rim thick and
prominent; ha = 0-10 to 0T5 mm., la = 0-06 to 0-07 mm.; small proximal cryptocyst
present; Lc = 0-16 to 0-52 mm., lc = 0-02 to 0-155 mm.

text-fig. 3. Three normal zooecia of Pyripora filum (Voigt).

E. Voigt Collection 1532. Upper Maastrichtian, Maastricht.

Description. Zoarium encrusting, with single distal-median and uni- and bi-lateral bud-
ding. Bilateral budding is apparently more common than unilateral, but both may occur
in the same zoarium (e.g. Voigt 1930, pi. 1, fig. 9 and Voigt Collection 1532). Lateral
budding is commonly from the distal-lateral walls of zooecia and is directed distally and
obliquely.

Caudae straight, or only slightly irregular, tubular, very narrow, of variable length
with a maximum of 1-00 mm. (see Voigt, 1930, p. 410) but usually much shorter. A
cauda may be twice as long as the rest of the zooecium. The caudae expand abruptly
into a proximal, triangular, smooth gymnocyst which is nearly as long as the aperture
and which rises steeply to the apertural rim and extends, much narrowed, round the
aperture.

Zooecia longitudinally oval, strongly inflated; lateral and distal walls fairly steep.

H. D. THOMAS AND G. P. LARWOOD: PYRIPORA D’ORBIGNY

379

Aperture longitudinally oval, small, with a prominent wide rim. Cryptocyst a slight,
crescentic, sloping shelf within the proximal margin of the aperture. Opesium only
slightly smaller than the aperture. Calcification of the frontal membrane may be com-
plete (e.g. Voigt, 1930, pi. 1, fig. 9); calcification of the operculum and regeneration of
the zooecia have not been seen.

Kenozooecia not seen.

Spines, avicularia, and ovicells absent.

Measurements

Zooecia Lz = 0-28 to 0-65 mm.

Iz = 0T2 to 0T5 mm.
ha = 0T0 to 015 mm.
la = 0 06 to 0 07 mm.

ho = 0 09 mm.
lo = 0 06 mm.

Lc = 016 to 0-52 mm. (occasionally up to 100 mm.),
lc = 0 02 to 0155 mm.

Apertural ratio = 1-66.

Remarks. P. filum (Voigt) is distinguished from other caudate Cretaceous Pyripora
species by its unusually small size, its prominent, wide apertural rim, and its straight,
narrow caudae (see text-fig. 3).

The type specimen from Maastricht is, according to Voigt’s description, more robust
than the material which he has seen from the Gonioteuthis [ Actinocamax ] mammillata
Chalk of Ifo, Sweden. Balavoine (1956, p. 157, pi. 6, fig. 2) has described, as Herpetopora
filum Voigt, a specimen from limestones containing Hercoglossa danica at Vigny, Seine-
et-Oise, France. Its very small size (Lz = 0T2 to 0T5 mm., lz = 0-05 to 0-06 mm.)
distinguishes it from P. filum (Voigt) as described above.

Stratigraphical distribution. Senonian, G. [A.] mammillata Chalk (upper part of zone
of G. [A.] quadrata) and Upper Maastrichtian.

Specimen. Part of a zoarium encrusting a ramose cyclostomatous polyzoan; Upper Maastrichtian,
‘Nd Geul-Tal b. Berg b. Maastricht’; E. Voigt Collection 1532.

Pyripora texana Thomas and Larwood
Plate 61, fig. 5

Pyripora texana Thomas and Larwood 1956, pp. 371, 372, text-figs. 2, 3.

Diagnosis. Pyripora with somewhat irregular uni- or bi-lateral budding, sometimes be-
coming pluriserial; Lz = 0-72 to 0-93 mm., lz = 0-21 to 0-33 mm.; ha = 0-33 to
0-36 mm., la = 0T5 to 0-21 mm.; caudae relatively broad, about as long as the rest
of the zooecia, Lc = 0-30 to 0-50 mm., lc = 0-07 to 0T0 mm.

Remarks. This earliest known species of Pyripora has been described recently by the
present authors.

Pyripora faxensis (Voigt)

Herpetopora faxensis Voigt 1930, pp. 410, 553, pi. 1, fig. 4.

Type specimen. That figured by Voigt 1930, pi. 1, fig. 4. Danian. Faxe, Seeland, Denmark.

Diagnosis (based on Voigt’s 1930 description and figure). Pyripora with zooecia inflated,
well rounded distally, tapered proximally, Lz = 1-20 to 1-50 mm., lz = 0-50 mm.;

380

PALAEONTOLOGY, VOLUME 3

aperture oval, not much longer than wide, ha = 040 mm. ; proximal cryptocyst appar-
ently present ; caudae short, tubular, rapidly widening distally into a long, wide, triangu-
lar, proximal gymnocyst.

Measurements (according to Voigt 1930, p. 410)

Zooecia Lz = 1-20 to 1-50 mm. ha = 0-40 mm.

lz = 0-50 mm. (no other measurements recorded)

Remarks. Voigt (1930, p. 410) has suggested, ‘Apparently this [P. faxensis] is the largest
species of the dispersa group — and is thereby distinguished from the other species’
[translation]. Voigt’s term ‘ dispersa group ’ implies all the Cretaceous species of Pyripora,
other than P. texana, which possess caudae. In addition to the generally greater size of
zooecia in P. faxensis, their marked distal rounding, the oval aperture, which is almost
as wide as long, the short, straight caudae, and the long, wide, triangular, proximal
gymnocyst also appear to be of diagnostic value. There are no specimens in the collec-
tions examined.

Stratigraphical distribution. Danian.

Pyripora parvicauda (Voigt)

Herpetopora parvicauda Voigt 1930, pp. 411, 553, pi. 1, fig. 11.

Type specimen. That figured by Voigt 1930, pi. 1, fig. 11. ‘Oberer Emscher.’ Lengede-Broistedt, Gr.
Ilsede.

Diagnosis (based on Voigt’s 1930 description and figure). Pyripora with zooecia distally
pointed; Lz = 0-50 to 0-75 mm., lz = 0-35 to 0-50 mm.; caudae very short and broad;
la = 040 to 0-70 mm.; cryptocyst apparently present; gymnocyst narrow distally and
laterally, wider and somewhat triangular proximally.

Measurements (according to Voigt 1930, p. 411)

Zooecia Lz = 0-50 to 0-75 mm. ha = 0-40 to 0-70 mm.

lz = 0-35 to 0-50 mm. (no other measurements recorded).

Remarks. Pyripora parvicauda is distinguished particularly by its ‘very short caudal
part’ which is ‘hardly developed’ (Voigt, 1930, p. 411, translation). The distally
pointed zooecial outline is also distinctive. There are no specimens in the collections
examined.

Stratigraphical distribution. Senonian, upper part of the Emscherian (? zone of Micraster
coranguinum or zone of Marsupites testudinarius).

Pyripora fiiimargo (Voigt)

Herpetopora fiiimargo Voigt 1930, pp. 411, 553, pi. 1, fig. 12.

Type specimen. That figured by Voigt 1930, pi. 1, fig. 12. ‘Mammillatensenon’ (upper part of zone of
Gonioteuthis [ Actinocamax ] quadrata). Ifo, Sweden.

Diagnosis (based on Voigt’s 1930 description and figure). Pyripora with zooecia in which
Lz = 0-30 to 0-35 mm., lz = 0-25 to 0-30 nun.; caudae absent; cryptocyst apparently
absent ; gymnocyst very narrow.

H. D. THOMAS AND G. P. LARWOOD: PYROPORA D’ORBIGN Y

381

Measurements (according to Voigt 1930, p. 411)

Zooecia Lz = 0-30 to 0-35 mm. lz = 0-25 to 0-30 mm.

(no other measurements recorded)

Remarks. The absence of caudae is particularly distinctive in this species, and, as Voigt
(1930, p. 411) has stated, the zooecia are very small with an ‘exceptionally narrow
border [gymnocyst]’ (translation). There are no specimens in the collections examined.

Stratigraphical distribution. ‘ Mammillatensenon’ (upper part of zone of Gonioteuthis
|/4]. quadratd).

Species previously assigned to Herpetopora (= Pyripora ) but here excluded

from Pyripora

1. Herpetopora danica Lang (1914u, p. 7, pi. 2, figs. 6, 7). Prof. E. Voigt has very
kindly lent us a well-preserved specimen, E. Voigt Coll. 3174, from the Upper Maastrich-
tian of Hemmoor, encrusting a piece of echinoid test. This specimen is like the holotype
in its fundamental features, but at least three zooecia have prominent hyperstomial
ovicells. It is clear that although ovicells occur they are not frequent and that in less
extensive zoaria they may not therefore be present — as is the case in the holotype of
Herpetopora danica Lang (D. 19429) and in the paratype (D. 44407). However, as all
other zoarial and zooecial characters are so similar, the species as a whole may be
regarded as ovicelled and therefore excluded from Pyripora. We would refer it provi-
sionally to Membranipora s.l.

2. Herpetopora danica Lang subsp. titania Voigt (1949, p. 9, pi. 1, figs. 1-3). We agree
that this subspecies, the type of which has been lent to us by Prof. Voigt (E. Voigt
Coll., 135), belongs to Herpetopora danica. As it belongs to an ovicelled species it also is
excluded from Pyripora.

RHAMMATOPORA Lang

ICrisia Mantell 1844, p. 285; 1854, p. 269.

IHippothoa Morris 1854, p. 125 [partim\.

Membranipora Vine 1890, p. 484 [partial]; 1891«, p. 385 [partial]; ? 1891(5, p. 395; 1892, p. 154;
1893, p. 335 [partial].

Rhammatopora Lang 1915, p. 496 [partial — Crisia johnstoniana Mantell excl. ] ; Canu, 1916,
p. 58; Thomas and Larwood 1956, p. 372.

Pyripora; Bassler, 1935, pp. 23, 181 [partial]; 1953, p. G158 [partial].

Type species (by original designation). Meaibranipora gauttina Vine 1890, p. 484, pi. 19, figs. 13 a-d,
Gault. Cambridge.

Diagnosis. Zoarium encrusting, uniserial; zooecia pyriform, budded distally and uni-
or bi-laterally or both, proximally tapered, with slender, tubular, proximal caudae;
aperture longitudinally oval; cryptocyst forming a narrow descending shelf, particularly
on the proximal side of the opesia; gymnocyst well developed proximally; differs from
Pyripora in having an apertural rim slightly raised with numerous small apertural spine-
bases; no calcified basal wall; no avicularia or ovicells.

Remarks. We have already indicated elsewhere (Thomas and Larwood, 1956, p. 372)
that the ‘rhamma’, described by Lang (1915, p. 497) as a feature of generic importance,
is of no systematic significance: it is an irregular crack developed along the caudae

382

PALAEONTOLOGY, VOLUME 3

during fossilization. However, Rhammatopora Lang may be distinguished from Pyripora
d’Orbigny by its numerous small apertural spine-bases. We thus differ from Canu (1916,
p. 59) and from Bassler (1935, pp. 23, 181 ; 1953, p. G158) who regarded the two genera
as synonymous.

Charixa vennensis Lang, the type species of Charixa Lang (1915, p. 500), is a multi-
serial species which becomes uni-serial in places (op. cit., pi. 17, fig. 5). In other respects,
e.g. the presence of apertural spine-bases, it resembles Rhammatopora, and it seems to
bear a similar relationship to R. gaultina as Pyripora pyriformis (Michelin) does to P.
laxata (d’Orbigny) — see Thomas and Larwood (1956, p. 370). But more material is
needed to elucidate further the structure of Charixa.

Distelopora Lang (1915, p. 502) differs from Rhammatopora in possessing a very wide,
distinctive cryptocyst, which contrasts with the slight, narrow, proximal cryptocyst of
Rhammatopora.

Lang (191 5, p. 498) regarded Crisia johnstoniana Mantell as a species of Rhammatopora.
As he pointed out, only Mantell's figures are available for the interpretation of the
species. They are much stylized, the dimensions of the specimens are not recorded, and
there is confusion as to the true locality of the material. Crisia johnstoniana Mantell
was recorded by Mantell (1844, p. 285; 1854, p. 269) from the Lower Greensand of
Maidstone, Kent, and possibly from the Shanklin Sands of the Isle of Wight. The
systematic position of Crisia johnstoniana is quite uncertain, and we doubt whether it can
be referred to Rhammatopora.

Stratigraphical distribution. Albian to Cenomanian, upper part of zone of Schloenbachia
varians.

Key to the species of Rhammatopora

1 . Aperture up to 045 mm. long and up to 0-25 mm. wide ; number of apertural

spine-bases 15-20 (commonly 18) . . . . . . .1. R. gaultina (Vine)

2. Aperture up to 0-27 mm. long and up to 017 mm. wide; number of apertural

spine-bases apparently 10-13 (commonly 12) . . . . .2. R. gasteri sp. nov.

Rhammatopora gaultina (Vine)

Plate 62, figs. 1, 2

Membranipora gaultina Vine 1890, pp. 461, 484, pi. 19, fig. 13 a-d\ 1 891 <7, pp. 385, 396; 1892,
pp. 154, 155, pi. 6, fig. 15; 1893, p. 335.

Membranipora gaultina var. Vine 1 891 A, p. 395.

Rhammatopora gaultina (Vine); Lang 1915, pp. 498, 499, pi. 17, fig. 2.

Rhammatopora pembrokiae Lang 1915, pp. 498, 499, pi. 17, figs. 3, 4.

Rhammatopora vinei Lang 1915, p. 498, pi. 17, fig. 1.

Pyripora gaultina (Vine); Bassler, 1935, p. 189.

EXPLANATION OF PLATE 62

Figs. 1, 2. Rhammatopora gaultina (Vine). 1, Holotype of Rhammatopora pembrokiae Lang, D. 22987,
uni- and bi-laterally budded normal zooecia with minute apertural spine-bases, X 20. 2, Holotype of
Rhammatopora gaultina (Vine), D. 2062, normal zooecia with minute apertural spine-bases, X 20.

Figs. 3, 4. Rhammatopora gasteri sp. nov. Holotype, D. 40472. 3, Part of zoarium with normal zooecia
with minute apertural spine-bases, X 35. 4, Detail of two zooecia and part of a damaged third
zooecium, the most distal zooecium with two very distinct spine-bases on the distal part of the
apertural rim, X 75. Photographs by Dept, of Photography, University of Durham, King’s College,
Newcastle upon Tyne.

Palaeontology, Vol. 3,

PLATE 62

THOMAS and LARWOOD, Rhammatopora

H. D. THOMAS AND G. P. LARWOOD: PYRIPORA D’ORBIGNY 383

Lectotype (chosen Lang 1915, p. 499). D.2062. Incomplete zoarium encrusting a fragment of Inocera-
mus. Albian, Gault. Cambridge. T. Jesson Collection.

Diagnosis. Rhammatopora with aperture 0-27 to 0-45 mm. long, 0-16 to 0-25 mm. wide;
opesia 0-22 to 0-38 mm. long, 0-15 to 0-17 mm. wide; 15 to 20, commonly 18, apertural
spine-bases.

Description. Zoarium encrusting; budding distal-median and, in places, uni- and bi-
lateral almost at right angles.

0 0-5 I’Omm.

1 i i i i 1 i i i i I

text-fig. 4. a, b, Rhammatopora gaultina (Vine), a— D.2062, lectotype, a normal zooecium; b —
D. 22905, paratype of R. pembrokiae Lang, distal end of a zooecium showing complete calcification of
the frontal membrane and operculum, c, d, Rhammatopora gasteri sp. nov. D. 40472, holotype, two
normal zooecia showing apertural spine-bases.

Caudae generally long, narrow, tubular, widening distally into a wide, smooth
gymnocyst which descends steeply, and is narrower laterally and distally to the aperture.
The caudae of successive zooecia in a given series are increasingly long.

Zooecia narrowly rounded distally and tapering proximally into the caudae. Aperture
longitudinally oval. Cryptocyst a narrow, descending, proximal shelf which may extend
some way round the sides of the aperture. Opesia longitudinally oval, only slightly
smaller than the aperture. Apertural rim slightly raised, with from 1 5 to 20, commonly
18, small apertural spine-bases. Regeneration is fairly common. Complete calcification
of the frontal membrane and operculum occurs occasionally, as in D. 22905.

Heterozooecia, other than avicularia, are possibly occasionally present. In D. 2062,
the lectotype, there is possibly one heterozooecium which is reminiscent, in its triangular
form, of the kenozooecia of Pyripora laxata (d’Orbigny).

Avicularia and ovicells absent.

384

Measurements

PALAEONTOLOGY, VOLUME 3

Zooecia Lz = 070 to 2-33 mm.

Iz = 0-24 to 0-39 mm.
ha = 0-27 to 0-45 mm.
la = 0T6 to 0-25 mm.

ho = 0-22 to 0-38 mm.
lo = 0 15 to 0T7 mm.
Lc = 015 to L95 mm.
lc = 0 03 to 0 08 mm.

Apertural ratio = 1 -42 to 1-94.

Remarks. Lang (1915) separated Vine’s specimens of Membranipora gaultina into two
species. He placed those from the Red Chalk of Hunstanton, Norfolk (D. 2051-2,
D. 20488-90), in Rhammatopora vinei, and retained, with a revised diagnosis, the speci-
mens from the Gault of Cambridge as R. gaultina (Vine). He also introduced another
new species, R. pembrokiae. According to Lang, the lateral gymnocyst of R. gaultina is
‘small’ and its aperture ‘oval to ovate’, whereas the lateral gymnocyst of R. pembrokiae
is ‘comparatively wide’ and its aperture ‘ovate to elliptical’. We have found, however,
that the apertural ratios of Lang’s specimens of R. pembrokiae are the same as those of
R. gaultina and R. vinei. There is considerable variation in the width of the lateral
gymnocyst in specimens which Lang assigned to R. pembrokiae, and in some zooecia
the lateral gymnocyst is as narrow as in the lectotype of R. gaultina (Vine).

The distinction which Lang (1914#, p. 7, footnote) made between ‘oval’, ‘ovate’, and
‘elliptical’ apertures is not clear. Apertures which are more or less pointed, or which
have almost parallel sides, or which are more nearly circular may be found in single
zoaria assigned by Lang to R. pembrokiae and, as already mentioned, the apertural
ratios are similar in the three species he recognized. Thus the characters used by Lang to
distinguish the species are really not distinctive. Furthermore, the dimensions of R.
pembrokiae and R. gaultina are similar and we consider the former to be a junior synonym
of the latter.

The specimens from the Red Chalk, which Lang named R. vinei, are badly preserved,
but, although it is not possible to identify all the diagnostic characters in any one zooe-
cium, they can be distinguished in the zoaria as a whole. The dimensions of the species
are similar to those of R. gaultina. Apertural spine-bases are occasionally preserved
(e.g. in D. 2051 and D. 43483), but other pittings on the worn apertural margins may
well be due to wear. Thus R. vinei is also synonymous with R. gaultina.

Stratigraphical distribution. Albian to Cenomanian, Chalk Marl.

Specimens. British Museum (Natural History) : D. 2062 — lectotype, see above. D. 22987, holotype,
and D. 21729-34, D. 22880-907, D. 22986, D. 22988-3008, all paratypes, of R. pembrokiae Lang;
zoarial fragments encrusting pieces of shell; Cenomanian, Chalk Marl, north-east of Cambridge; F.
Mockler Collection. D. 2051, holotype, and D. 2052, D. 20488-90, all paratypes, of R. vinei, Lang;
zoarial fragments encrusting pieces of Inoceramus; Red Chalk, Hunstanton, Norfolk; T. Jesson Collec-
tion. D. 43483 — zoarial fragments encrusting Inoceramus sp.; Red Chalk, Hunstanton, Norfolk; J. E.
Lee Collection. D. 43482 — zoarial fragments encrusting Inoceramus sp.; Red Chalk, Speeton, York-
shire; W. Bean Collection.

Rhammatopora gasteri sp.nov.

Plate 62, figs. 3, 4

Holotype. D. 40472. Zoarial fragment encrusting part of a valve of Ostrea sp. Cenomanian, upper
part of zone of Schlaenbachia various. Pit south-west of Fox Inn, Southerham, Lewes, Sussex. C. T. A.
Gaster Collection.

H. D. THOMAS AND G. P. LARWOOD: PYRIPORA D’ORBIGN Y

385

Diagnosis. Rhammatopora with aperture 0-16 to 0-27 mm. long, 0-08 to 0-17 mm. wide;
apparently 10 to 13, commonly 12, apertural spine-bases.

Description. Zoarium encrusting; budding distal-median and uni- and bi-lateral com-
monly at right angles.

Caudae generally as long as, or longer than, the rest of the zooecia, narrow, tubular,
and straight. In the holotype, the only available specimen, there are numerous short
caudae, but these are near the beginning of individual series of zooecia in which succes-
sive caudae become increasingly longer: the longest measurable cauda is 1-17 mm. The
caudae widen distally into a smooth, triangular gymnocyst which extends, much
narrowed, round the aperture.

Zooecia distally rounded and tapering proximally into the caudae. Aperture longi-
tudinally oval. Cryptocyst a narrow descending proximal shelf which, apparently, does
not extend along the sides of the aperture. Opesia longitudinally oval, only slightly
shorter than the aperture and of the same maximum width. Apertural rim slightly raised,
apparently with from 10 to 13, commonly 12, very small apertural spine-bases. Complete
calcification of the frontal membrane and operculum occurs occasionally. Regeneration
of zooecia has not definitely been seen.

Heterozooecia, other than avicularia, have not been seen.

Avicularia and ovicells absent.

Measurements

Zooecia Lz = 0-45 to 1-53 mm. ho = 0T4 to 0 25 mm.

lz = 075 to 0-27 mm. lo = 0 08 to 077 mm.

ha = 076 to 0-27 mm. Lc = 0-23 to 177 mm.

la = 0 08 to 077 mm. lc = 0 03 to 0 05 mm.

Apertural ratio = 1-33 to 2 00.

Remarks. The holotype is the only specimen known to us. Although it is somewhat worn,
the important details of most of the zooecia can be clearly determined without difficulty.
Spine-bases occur on the apertural rims of some of the zooecia, are regularly spaced,
and are exactly like those found in zooecia of similar size in other genera. Owing to
wear, it is not possible to trace them all round an aperture, but in many zooecia they
are so clear that they can be interpreted only as spine-bases.

R. gasteri sp. nov. differs from R. gaultina (Vine) in its smaller size and particularly
in its fewer apertural spine-bases (cf. text-figs. 4 a-cf). The species is named after Mr.
C. T. A. Gaster.

Stratigraphieal distribution. Cenomanian, zone of Schloenbachia various.

Specimen. D. 40472 — holotype, see above.

REFERENCES

balavoine, p. 1956. Bryozoaires recueillis dans le Calcaire de Vigny (Seine-et-Oise) par M. A.

Huilleret. Bull. Soc. Geol. Fr. (6), 6, 157-61, pi. 6.
bassler, r. s. 1935. Bryozoa. Foss. Catal., 1, Anim. 67, 1-229.

• 1953. Treatise on invertebrate paleontology, Part G, Bryozoa, G1-G253. Lawrence, Kansas.

brydone, r. m. 1936. Further notes on new or imperfectly known Chalk Polyzoa, 3, 61-90, pi. 30-42.
London.

386

PALAEONTOLOGY, VOLUME 3
canu, F. 1916. Bryozoaires. Rev. crit. Palezool., 20 Ann. ( 2), 58-63.

canu, f. and bassler, r. s. 1920. North American early Tertiary Bryozoa. Bull. U.S. nat. Mus. 106,
1-879, pi. 1-162.

1923. North American later Tertiary and Quaternary Bryozoa. Ibid. 125, 1-302, pi. 1-M7.

hagenow, k. F. von. 1839. Monographic der Riigen’schen Kreide-Versteinerungen, 1, Phytolithen
und Polyparien. Neues Jb. Min. Geol. Paldont. 7, 253-96. Polyzoa — 263-96, pi. 4, 5.

— — in GEiNiTZ, H. B. 1846. Bryozoa. Grundriss der Versteinerungenskunde, 586-635.
lang, w. d. 1913. Report of a visit to the exhibits of Polyzoa and Corals in the Geological Depart-
ment of the British Museum (Natural History). Proc. Geol. Assoc. 24, 169-73.

1914a. On Herpetopora, a new genus containing three new species of Cretaceous Cheilostome

Polyzoa. Geol. Mag. 51, 5-8, pi. 2.

19146. Some new genera and species of Cretaceous Cheilostome Polyzoa. Ibid. 51, 436-44,

pi. 34.

■ 1915. On some new uniserial Cretaceous Cheilostome Polyzoa. Ibid., 52, 496-504, pi. 17.

1925. Persistence in fossils. Proc. Geol. Assoc. 36, 227-39, pi. 20-22.

mantell, G. A. 1844. Medals of creation. 1st ed. 1, 1-456; 2, 457-1016. London.

1854. Ibid., 2nd ed. 1, 1M46; 2, 447-930. London.

morris, j. 1854. A catalogue of British fossils. 2nd ed., 1-372. London.

novak, o. 1877. Beitrag zur Kenntniss der Bryozoen der bohmischen Kreideformation. Denkschr.
Akad. 1 Viss. Wien, 37 (2), 79-162, pi. 1-10.

orbigny, A. d’. 1851-4. Bryozoaires. Paleontologie frangaise. Terrain Cretace, 5, 1-1 192, pi. 600-800.
reuss, a. e. von. 1854. Beitrage zur Charakteristik der Kreideschichten in den Ostalpen besonders im
Gosauthale und am Wolfgangsee. Denkschr. Akad. Wiss. Wien, 7, 1-156, pi. 1-31.

Thomas, h. d. and larwood, g. p. 1956. Some ‘uniserial’ membraniporine Polyzoan genera and a
new American Albian species. Geol. Mag. 93, 369-76.
vine, G. r. 1890. A Monograph of the Polyzoa (Bryozoa) of the Red Chalk of Hunstanton. Quart. J.
geol. Soc. Lond. 46, 454-86, pi. 19.

• 1891a. Notes on the Polyzoa and Microzoa of the Red Chalk of Yorkshire and Norfolk. Proc.

Yorks, geol. polyt. Soc., N.s. 11 (3), 363-96, pi. 17.

18916. Report of the committee ... on the Cretaceous Polyzoa. Rep. Brit. Aw. Adv. Sci. ( for

1890), Leeds, 378-96.

— — • 1892. Fossil Polyzoa: further additions to the Cretaceous lists. Proc. Yorks, geol. polyt. Soc.,
n.s. 12, 149-61, pi. 6.

• 1893. Report of the committee ... on the Cretaceous Polyzoa. Rep. Brit. Ass. Adv. Sci. (for 1892),

Edinburgh, 301-37.

Voigt, e. 1930. Morphologische und stratigraphische Untersuchungen iiber die Bryozoenfauna der
oberen Kreide. 1 Teil. Die cheilostomen Bryozoen der jtingeren Oberkreide in Nordwestdeutsch-
land, im Baltikum und in Holland. Leopoldina, 6, 379-579.

- — — 1949. Cheilostome Bryozoen aus der Quadratenkreide Nordwestdeutschlands. Mitt. geol. (St.)
Inst. Hamb. 19, 1-49, pi. 1-11.

H. DIGHTON THOMAS

British Museum (Natural History),
London, S.W. 7

G. P. LARWOOD
Sedgwick Museum,

Manuscript received 1 August 1959 Cambridge

THE PERMIAN LEIOPTERII D MER1SMO PTERIA
AND THE ORIGIN OF THE PTERII DAE

by J. M. DICKINS

Abstract. The musculature and dentition of Merismopteria Etheridge jun. 1892 are described and illustrated.
Muscles are present both in front and behind the ‘ clavicle ’ or buttress ridge. Lateral and cardinal teeth may be
present. Merismopteria is closely related to and may be a synonym of Leiopteria Hall 1883. Modern pteriids
are considered to be derived from the Palaeozoic leiopteriids which lack a chondrophore and have a flattish
ligament area with chevron-shaped or parallel-longitudinal ligament grooves and it is concluded that the ‘pteriid’
type of ligament with a chondrophore is independently developed in the pterioids and the pectinoids.

In describing a specimen from the Bowen River, Queensland, Etheridge jun. (1892,
p. 271) proposed a new generic name Merismopteria, designating as type Pterinea
macroptera Morris (1845, p. 276, pi. 13, fig. 2; 3). The characters of this genus, including
the nature of the clavicle and the anterior musculature, and the relationship to other
genera, especially Leiopteria Hall, have caused considerable speculation.

Under the heading ‘Generic Characters’ Etheridge gave the following diagnosis:

‘ Pteronitiform in appearance, the anterior end lobe-like and well developed; posterior end alate.
Area excavated along the cardinal edge of both valves, and deeply ridged for the reception of a liga-
ment ; cardinal teeth wanting but a strong clavicle descends in each valve before the anterior adductor
muscles; one or more lateral teeth in each valve. Anterior muscle scars double and strong, the
superior scar situated towards the umbones. External ornament of concentric ridges. ’

Etheridge states under the heading ‘Observations’, however, that there is ‘one lateral
tooth only, either oblique or horizontal’.

In order to clarify the characters of Merismopteria an attempt has been made to find
the type specimen or specimens described by Morris and according to Morris collected by
Strzelecki from Spring Hill, Van Diemen’s Land (Tasmania). According to a letter from
the British Museum (Natural History) the types are not in this institution where most of
the Strzelecki Collection is housed, so that it appears that they are lost. Two specimens,
however, from the Strzelecki Collection (PL2865-6) are housed in the British Museum.
These have the locality label ‘ Carboniferous Australia’. These and other specimens have
been kindly lent by the British Museum, and, together with specimens in the Australian
Museum, Sydney, and in the Bureau of Mineral Resources, Canberra, afford consider-
able additional information on Merismopteria.

Type Specimens of Pterinea macroptera Morris. Although Morris’s figured specimen
(or specimens) from Spring Hill, Tasmania, appear to be lost it is not proposed to
choose a neotype at present.

Mr. M. R. Banks, of the Geology Department of the University of Tasmania, in a
letter of 19 November 1958, says : ‘ I have so far been unable to find the Permian rocks at
Spring Hill. . . . Permian rocks occur a few miles to the east at Eastern Marshes and
about ten or fifteen miles to the west near Waddamana.’ He also says the University of
Tasmania has no specimens from this area. Although it is possible that PL2865 and

[Palaeontology, Vol. 3, Part 3, 1960, pp. 387-91.]

388 PALAEONTOLOGY, VOLUME 3

PL2866 are from the type locality it appears better to defer choice of a neotype until the
locality of the specimens can be confirmed or until suitable topotype material is avail-
able.

PL2865 and PL2866 are contained in a light buff to white (? leached) dense silt to fine
sandstone with many quartz grains, some mica, and, apparently, a clay matrix.

Dimensions of British Museum Specimens

Locality

Carboniferous Australia

McCallum’s Block, Bowen Coal
Field, about 4 miles south-west
of Collinsville, probably above/
Big Strophalosia Bed and below |
Derby ia Bed.

Number

Length

Height

Width

PL2865 (internal

Left valve

38

19+

2-5

impression)

Right valve

36

21 +

20

PL2866 (internal

Left valve

36

23

30

impression)

Right valve

32

21

2-5

PL541 (internal

Left valve

103 +

63

180

impression)

Right valve

103 +

60+

160

PL540 (external
impression)

Right valve

46

31

60

text-fig. 1. Composite internal impression of a left valve of Merismopteria. 2, Cardinal teeth or
Merismopteria. 3, Musculature of a right valve of Pinctada vulgaris (based on Newell 1938). a, b, c,
impressions of pedal muscles; d, impression of ‘posterior’ adductor; e, impression of ‘anterior’
adductor ; /, pallial line ; g, cardinal teeth or their impressions ; h, ‘ clavicle ’ (anterior buttress ridge) of
its impression; /, posterior lateral tooth; L, left valve; R, right valve.

Musculature. Etheridge considered that the anterior muscle scar was situated behind the
clavicle, an observation which I have been able to confirm (Dickins 1957, p. 30). On the
other hand, in apparently closely related leiopteriids, for example Dozierella Newell
(1940, p. 284), the anterior muscle is found in front of the ‘clavicle’ which thus repre-
sents a buttress ridge. The two specimens from the Strzelecki Collection allow a resolu-
tion of this apparent contradiction. Muscles are present both in front of and behind the
‘clavicle’. In front of the ‘clavicle’ in PL2865 the presence of a relatively large anterior
adductor muscle is indicated by muscle growth lines (the muscle track). The anterior
adductor muscle is bounded behind by the ‘clavicle’ which is thus a buttress ridge con-
firming the interpretation made by Newell (1940, p. 283).

J. M. DICKINS: THE PERMIAN LEIOPTER IID MERISMO PTERIA 389

In PL2866, however, two distinct muscles occur behind the buttress. The larger,
which is not much smaller than the anterior adductor, is situated immediately behind the
buttress and lies below the smaller. The smaller is placed on the anterior side of the
internal impression of the beak. Although they are unusually large, it seems likely that
these muscles are connected with the action of the foot (pedal muscles).

Further information is given on the musculature by PL541, from McCallum’s Block,
Bowen Coal Field. The posterior adductor scar is large and kidney-shaped, situated over
the posterior umbonal ridge. The intermittent pits of the pallial line run from the base
of the ventral pedal muscle along the anterior side of the umbonal ridge towards the
posterior adductor. A number of small round pits occur on the surface of the shell
within the pallial line, marking the mantle attachment to the shell.

The musculature is thus more primitive than that characteristically found in adult
living pteriids, the anterior adductor and the pedal muscles being considerably more
highly developed.

Hinge structure. Examination of specimens in the Australian Museum, Sydney, and in
the Bureau of Mineral Resources, Canberra, confirms Etheridge’s description of the
ligament (1892, p. 271) and shows this structure is of the same type as that described
for Leiopteria? carrandibbiensis Dickins (1957, p. 30, pi. 4, figs. 13-17, text-fig. 4). A
specimen from the collections at Canberra is figured. The ligament is borne on an area
which is elongated on either side of the umbo, and on which are a number of parallel-
longitudinal ligament grooves. Recognition of a posterior lateral by Morris, de Koninck,
and Etheridge is confirmed by the presence of a lateral tooth in PL2866 and in speci-
mens in the Australian Museum. In PL2866 a lateral tooth is visible in the left valve
towards the rear of the cardinal margin. By analogy with living pteriids at least a single
tooth could be expected in the right valve, although such a tooth has not been visible in
most of the specimens which I have examined. However, even though the lateral den-
tition is difficult to see in internal impressions, because it diverges only slightly from the
cardinal plate, it does appear possible that in some specimens or species lateral teeth are
absent. Living pteriids vary in this way.

In PL2865 the impressions of small cardinal teeth are visible immediately in front
of the umbos. In the right valve these comprise a projection with a groove on either
side and in the left valve a socket for the projection of the right valve with a clasping
lamella on either side fitting in the sockets of the right valve. These are similar to those
of living Pteriidae and by analogy, like the lateral teeth, may be variable. The details
of the dentition would appear to be of doubtful value for generic separation.

Shell structure. The main shell layer is composed of prisms at right angles to the surface
(specimens in the Australian Museum). The prisms are of a similar size to those recorded
for L.? carrandibbiensis. Under the hand lens the structure appears the same as that
found in Atomodesma. It may be assumed that during life a thin inner nacreous layer
was present.

Relationship to Leiopteria Hall 1883. Amongst described genera Merismopteria appears
to be closest to Leiopteria. On the basis of Hall’s diagnosis (1884, p. xiii) there is little
to choose between the two, but the detailed characters of Hall’s species still remain
obscure and until more is known about these species it is not possible to decide whether

cc

B 6612

390

PALAEONTOLOGY, VOLUME 3

Merismopteria is a synonym of Leiopteria or not. Meantime perhaps L.? carrandibbiensis
Dickins should be assigned to Merismopteria rather than Leiopteria. Such species as
Liopteria? dutoiti Harrington (1955, p. 120, pi. 24, figs. 5, 11) and Liopteria bonaerensis
Harrington (1955, p. 120, pi. 24, fig. 10) may be more closely related to Merismopteria
than is suggested by their generic assignment to Leiopteria. (According to Neave (1939,
p. 960) Liopteria is an emendment proposed by Fischer 1886 for Leiopteria.)

ORIGIN OF THE PTERIID AE

In this paper the Pteriidae are restricted to those forms, characteristically with a well-
developed prismatic shell layer, which possess the ‘pteriid’ type of hinge, that is with a
single distinct ligament pit on the area. This does not include the Mesozoic forms such
as Meleagrmella, Oxytoma, and MaccoyeUa which properly belong to the Pectinacea
(see Ichikawa 1958). Pteriidae in this restricted sense are at least very rare in Palaeozoic
rocks, and I know of no definite occurrence. However, Bakevellia King from the Upper
Permian has multiple ligament pits on the area like Perna and is thus apparently derived
from a form with a single ligament pit. Most of the Palaeozoic forms which have been
referred to the Pteriidae have areas with a number of parallel-longitudinal or chevron-
shaped ligament grooves and are more satisfactorily referred to the Leiopteriidae or
other families.

Although Merismopteria differs in some important characters from modern Pteriidae,
it shows sufficient similarity to suggest that the Pteriidae were derived from leiopteriids
or related shells. The general shape and shell structure of Merismopteria are similar to
modern pteriids, and its dentition and posterior musculature to species of Pinctada. In
all these respects it is significantly different from the pectinoid forms. It differs markedly
from the pteriids only in the anterior musculature, which is apparently more primitive
and in the ligament structure. The poor development of the subauricular notch in right
valves in Palaeozoic forms appears to be of lesser importance. From these data it seems
that the ‘pteriid’ type of hinge was developed independently in the Pteriacea and the
Pectinacea, a suggestion which has already been made by Newell (1942, p. 26). Alterna-
tive explanations such as derivation of the pteriids from the aviculopectinids seem
unlikely. Such a derivation would require modification of the shell in a great number of
respects, including shell shape, shell structure, and musculature to produce a type of
shell already in existence but requiring essentially only modification of the hinge.

Bernard (1895-7) observed that the ligament was internal when it first appeared in the
young pelecypod and concluded this was the primitive condition of nuculoids. Newell
(1942, p. 28), after referring to Bernard’s observation, states that ‘in some genera such
as Pecten, Nueula, Pteria, Lima, and Ostrea, this condition apparently continues to
maturity’. Newell thus suggests the early internal ligament corresponds to the adult
resilifer.

However, the phylogeny of at least some of these forms suggests that the early
ligament recorded by Bernard is unlikely to be analogous with the adult resilifer, and
indeed Bernard’s ontogenetic studies also seem to show this. If Bernard’s and Newell’s
suggestion were correct this would seem to contradict the conclusion that the ‘pteriid’
ligament is derived independently in the Pectinacea and the Pteriacea.

There seems little doubt, however, that the ancestors of the pteriids in the restricted
sense are represented in the Upper Palaeozoic mainly by forms which have parallel-

J. M. DICKINS: THE PERMIAN LEIOPTERIID MERISMO PTERIA

391

longitudinal or chevron-shaped grooves, and all these forms can be satisfactorily
placed in the superfamily Pteriacea. Earlier in the Palaeozoic the superfamilies such as
the Pteriacea, the Pectinacea, and the Mytilacea become less easy to distinguish as these
groups converge but none of these earlier forms can be expected to possess the ‘pteriid’
type of ligament.

Acknowledgements. This work was partly carried out in the Department of Geology of the University
of Queensland in fulfilment of requirements for the Ph.D. degree, and I am grateful to Professor W. H.
Bryan and Dr. D. Hill for their co-operation and advice. The loan of specimens by the British Museum
(Natural History) is gratefully acknowledged, and the Director, Dr. J. W. Evans, and Mr. H. O.
Fletcher, of the Australian Museum, Sydney, are thanked for making collections available for examina-
tion. Publication is by permission of the Director of the Bureau of Mineral Resources, Geology and
Geophysics.

REFERENCES

Bernard, f. 1895-7. Notes sur le developpement et la morphologie de la coquille chez Lamelli-
branches. Soc. geol. France, Bull .; Premiere note, 23, 104-59, 1895; Deuxieme note (Taxodontes),
24, 54-82, 1896; Troisieme note (Anisomyaires), 24, 412-49, 1896; Quatrieme et derniere note, 25,
559-66, 1897.

dickins, j. m. 1957. Lower Permian Pelecypods and Gastropods from the Carnarvon Basin, Western
Australia. Bur. Min. Resour. Aust. Bull. 41.

etheridge, R. jun. 1892. In jack, R. L. and etheridge, R. jun., Geology and Palaeontology of Queens-
land and New Guinea. Brisbane and London.

hall, J. 1883-4. Lamellibranchia I. Monomyaria. Natural history of New York, Palaeontology, 5 (1);
Plates and Explanations, 1883; Text and Additional Notes and Explanations, 1884.

Harrington, h. j. 1955. The Permian Eurydesma fauna of eastern Argentina. J.Paleont. 29, 112-28.

ichikawa, K. 1958. Zur Taxionomia und Phylogenie der Triadischen ‘Pteriidae’ (Lamellibranch).
Palaeontographica, 3 (A), 131-212.

morris, J. 1845. Description of fossils; in strzelecki, p. de, Physical Description of New South
Wales and Van Diemen's Land, 270-91.

neave, s. a. 1939. Nomenclator Zoologicus , 2, London.

Newell, n. d. 1938. Late Palaeozoic Pelecypods: Pectinacea. Kansas Geol. Surv. 10.

— — 1940. Invertebrate fauna of the late Permian Whitehouse Sandstone. Bull. geol. Soc. Amer. 51,
261-323.

1942. Late Palaeozoic Pelecypods: Mytilacea. Kansas Geol. Surv. 10 (2), 1-115.

J. M. DICKINS

Bureau of Mineral Resources, Geology and Geophysics,
Canberra, A.C.T.

Manuscript received 28 August 1959

CHARACTERS AND RELATIONSHIPS OF THE
MESOZOIC PELECYPOD PSEUDAVICULA

by J. M. DICKINS

Abstract. The hinge structure and the musculature are described and illustrated. An anterior ear and a sub-
auricular notch are present in right valves, the ligament is of the ‘pteriid’ type with a chrondophore, and the
musculature is similar to that of Chlamys. The main shell layer is composed of concentric interlocking lamellae
and the valves although inequal are biconvex. It is concluded that Pseudavicula is a pectinoid and appears to be
most closely related to forms placed in the subfamily Streblochondriinae of the family Aviculopectinidae.

When Etheridge jun. proposed the generic name Pseudavicula (1892, p. 449), with
Lucina anomala Moore (1870, p. 251, pi. 14, fig. 4) as type, he gave the following diag-
nosis :

‘Shell in general outline meleagriniform, but devoid of an anterior wing or ear. Valves compressed,
closed, nacreous within, bi-convex, inequilateral. Cardinal margin thin, no area, or hinge teeth, but
probably with a small ligament. Position wing moderately developed, but with little or no emargina-
tion. Adductor muscular scars subcentral, of medium size.’

The name Pseudavicula was apparently first published by Hudleston (1890, p. 244). It
is clear from Hudleston’s paper, however, that the name was derived from Etheridge’s
work and he states ‘the diagnosis of Pseudavicula is not known to me’. He referred his
specimens to Pseudavicula anomala (Moore) 1870 and placed Avicula orbicularis Hudle-
ston 1884 in synonymy with P. anomala. Hudleston’s use of the name thus does not
effect the choice of type species.

In Pseudavicula, Etheridge placed two closely related species Lucina anomala Moore
and Lucina? australis Moore (1870, p. 251, pi. 14, fig. 5). He also doubtfully referred
Avicula alata Etheridge sen. (1872, p. 342, pi. 20, fig. 8) to this genus (p. 563). In 1907 he
added, from Queensland, another species, P. papyracea Etheridge jun. (1907, p. 319).

In describing specimens from South Australia, Whitehouse (1925, pp. 27, 28) sug-
gested that P. australis was a synonym of P. anomala and that the right valve of his
specimens had no byssal sinus. He also showed that a ‘pteriid’ type of hinge was
present (p. 28, pi. 1, figs. 2, 3) although this does not seem to have become widely
known.

The three species, referred without doubt by Etheridge to Pseudavicula, form a dis-
tinctive closely related group in the Roma ‘Formation’ of Aptian age, and are wide-
spread in Queensland, New South Wales, and South Australia, occurring sometimes
in large numbers. P. papyracea has also been recorded by Glaessner (1945, p. 159, pi. 6,
fig. 11; 1958, p. 201) from New Guinea, and I have found P. papyracea or a closely
related species in the Canning Basin of Western Australia in deposits which are thus
no doubt of Cretaceous age (Frezier Sandstone of Lindner and Drew: see McWhae,
Playford, Lindner, Glenister, and Balme 1958, p. 108). The anterior ear and subauricu-
lar notch are also shown in right valves of these specimens.

The following features are revealed by a study of specimens in the collections of the

[Palaeontology, Vol. 3, Part 3, I960, pp. 392-6, pi. 63.]

J. M. DICKINS: MESOZOIC PELECYPOD PSEUDAVICULA

393

Geological Survey of Queensland and afford additional information on the character
and relationship of the genus.

Structure of hinge and cardinal region. The characters of Pseudavicula are shown excel-
lently in internal and external impressions from the south-west corner of Portion 53V,
Parish of Bendemere, left bank of Yuelba (or Yuleba) Creek, Queensland. (Yuelba
Creek is about 36 miles east of the township of Roma.) All the right valves appear to
be smooth but the left valves apparently may be smooth or have delicate radial ribbing.
A left valve showing radial ribbing is figured in the accompanying plate. According to
Etheridge (1907, p. 326), P. papyracea is separable from P. anomala by the absence of
radial ribbing so that, in this case, it is difficult to decide what species are present — apart
from the radial ribbing the specimens do not appear to differ significantly.

text-fig. 1, Internal impression of a right valve of Pseudavicula. 2, Impression of ligament of Pseuda-
vicula. 3, Musculature of a right valve of Chlamys islandicus (based on Newell 1938). a, b, adductor
impressions; c, pallial line; d, line of posterior edge of chondrophore; e, line of anterior edge of

chondrophore.

The umbo is almost but not quite terminal, and a small anterior ear is visible in
several right valves, one of which is figured. The anterior ear is relatively very small
and would easily be lost by abrasion before burial or by erosion or breakage during
weathering or during breaking out from the rock. This no doubt explains why the
ear has not previously been recorded. The ear is separated below from the rest of the
shell by a small but distinct subauricular notch (the term ’subauricular notch’, used by
Cox 1943, p. 153, is preferred to the perhaps misleading term ’byssal notch’). Except in
distorted specimens the anterior ear is in the same general line as the rest of the cardinal
margin.

The structure of the ligament is of the ‘pteriid’ type, characteristic of the aviculopecti-
nids and the living pteriids. The ligament is lodged on a flattish ligament area ; behind
the umbo the area thickens and deepens to form a chondrophore or resilium pit. The
chondrophore is marked off at the back by a distinct but rounded oblique boundary,
the upper part of which points to the front. The anterior boundary is not clear in any
of the specimens but appears to be more upright than the posterior boundary and
situated level with the umbo. The area has parallel-longitudinal growth-lines and
lacks any longitudinal ligament grooves. The structure is different from that found in

x2Vs

xYa

5

394 PALAEONTOLOGY, VOLUME 3

Deltopecten, where no chondrophore boundary is present, and is distinct from the leio-
pteriids where distinct longitudinal or chevron-shaped ligament grooves are present.

Musculature. The musculature of the right valve is shown in the accompanying text-
figure, where a right valve of Chlamys is also illustrated. The right valve of Pinctada has
been figured in the preceding paper on Merismopteria. An important feature of Pseuda-
vicula is the closer resemblance to Chlamys than to Pinctada: that is, it is more closely
related to the Pectinacea than to the Pteriacea.

Convexity of valves and shell structure. Etheridge (1907, p. 319) discussed the question of
relative convexity of the right and left valve and concluded that in all three species, the
left valve is more convex and the umbo higher than in the right. An examination of the
collections in the Geological Survey of Queensland confirms Etheridge’s conclusion.
Although both valves are convex, in all specimens examined the left valve was more
convex than the right. The inequality, however, varies considerably.

The specimen figured by Etheridge (1892, pi. 21, fig. 14) which was later the only
specimen referred by him to Pseudavicula papyracea (1907, p. 319), and which must thus
become the holotype, is preserved in the Geological Survey of Queensland. The speci-
men is exfoliated but shows a layer of concentric interlocking lamellae, such as Newell
(1938) has shown to occur in Aviculopecten and Ichikawa (1958) has recently shown to
occur in Meleagrinel/a and Oxytoma. The specimen figured by Etheridge (1892, pi. 21,
fig. 16), which in my opinion is also referable to P. papyracea, also shows concentric
interlocking lamellae, which are just visible to the naked eye.

Relationships of Pseudavicula. The small anterior ear and especially the subauricular
notch in Pseudavicula, together with other characters, suggest derivation from members
either of the Pectinacea or the Pteriacea in which the anterior ear and subauricular
notch were more strongly developed. In considering whether Pseudavicula belongs to the
Pectinacea or to the Pteriacea, the musculature and the shell structure are of significance,
as the type of hinge present occurs in both superfamilies. Each definitely indicates
relationship to the Pectinacea. The muscle pattern is of the ‘pectinoid’ type; the shell

EXPLANATION OF PLATE 63

Figs. 1-5. Pseudavicula cf. papyracea Etheridge jun. 1907. Specimens from south-west corner of
Portion 53V Parish of Bendemere, left bank of Yuelba Creek, about 36 miles east of Roma. The
numbers are those of the Geological Survey of Queensland. 1 , Latex impression of a left valve, X 2,
No. F. 23 1 7. 2, Latex impression of right valve showing anterior ear, X 2, No. F. 23 1 6, dorso-lateral
view. 3, Same, lateral view, X 2. 4, Internal impression of dorsal part of a right valve, X 4, No.
F. 2315. 5, Complete view of same shell, X 2.

Figs. 6-12. Merismopteria sp. All specimens are from rocks of Permian age. 6-8, British Museum
(Natural History) No. PL2865, ‘Carboniferous’, Australia. 6, Plasticene impression to show
cardinal teeth, x2. 7, Normal view, X 1. 8, Specimen tilted up at front to show muscle growth-
lines, x2. 9-10, British Museum (Natural History) No. PL2866, ‘Carboniferous’, Australia. 9,
Normal view, x 1. 10, Dorsal view of right valve, X 1. 11, British Museum (Natural History) No.
PL541, McCallum's Block, Bowen Coal Field, lateral view of left valve, xj. 12, Commonwealth
Palaeontological Collection Type No. 107, Field No. JB35, east end of Steamer’s Beach, Jervis Bay,
New South Wales, from beds probably equivalent to the Conjola Beds of David and Stonier. In-
ternal impression of right valve showing longitudinal ligament grooves and a posterior lateral
tooth, x 2; the front part of the shell is missing.

Palaeontology, Vol. 3.

PLATE 63

‘‘if /X?

DICK I NS, Australian Pelecypods

J. M. DICKINS: MESOZOIC PELECYPOD PSEUDAVICULA

395

structure is characteristic of that found in the aviculopectinids and distinctly different
from that found in the pterioids where, characteristically, a prominent prismatic layer is
developed in each valve and the lamellar structure is absent. In the Upper Palaeozoic the
biconvex Streblochondriinae of Newell (1938) appear likely to be ancestral to Pseuda-
vicula.

In the Jurassic the biconvex Pseudomonotis laevis (Blake and Hudleston) Arkell (1931,
p. 200, pi. 24, figs. 11, 11 a, 12) appears to be closely related to Pseudavicula and may
even be referable to this genus. In the Cretaceous Pseudomonotis super stes Spitz (1914,
p. 201, pi. 18, figs. 6, 7) from the Gieumal Sandstone of the Himalayas appears to be
closer to Meleagrinella Whitfield 1 885 than to Pseudavicula unless two species are present
among Spitz’s specimens.

Conclusion. The present evidence supports a previous suggestion, based on a study of
Permian forms, that a number of Mesozoic genera which have usually been placed by
previous authors in the Pteriidae are descended from the aviculopectinids (Dickins 1957,
p. 44). Ichikawa (1958) has independently arrived at a similar conclusion, which he sup-
ports with ample and convincing evidence. Ichikawa (1958, p. 193) also agrees with
Newell (1938) in considering the Aviculopectinidae are better placed with the pectinoids
than with the pterioids. The author has discussed evidence in the immediately preceding
paper which suggests that by the end of the Carboniferous period the Pectinacea and
the Pteriacea were already distinct groups, evolving separately, and thus would place
the Aviculopectinidae and derived forms in the superfamily Pectinacea and separate
them from forms placed in the Pteriacea.

It can be concluded that Pseudavicula is derived from aviculopectinid forms and is
possibly close to forms placed by Newell in the subfamily Streblochondriinae.

Acknowledgements. This work was partly carried out in the Department of Geology of the University
of Queensland in partial fulfilment of requirements for the Ph.D. degree. Mr. A. K. Denmead, Chief
Government Geologist, Geological Survey of Queensland, made the collections available for study, and
his co-operation and the help of Mr. R. J. Paten of the Geological Survey of Queensland are gratefully
acknowledged. The paper is published by permission of the Director of the Bureau of Mineral
Resources, Geology and Geophysics.

REFERENCES

arkell, w. J. 1931. British Corallian Lamellibranchia. Part 5. Palaeontogr. Soc.

cox, l. r. 1943. The English Upper Lias and Inferior Oolite species of Lima. Proc. malac. Soc. 25,
151-87.

dickins, j. m. 1957. Lower Permian Pelecypods and Gastropods from the Carnarvon Basin, Western
Australia. Bur. Min. Resour. Aust. Bull. 41.

etheridge, r. sen. 1872. Description of the Palaeozoic and Mesozoic fossils of Queensland. Quart. J.
geol. Soc. Lond. 28, 317-60.

etheridge, R. jun. 1892. In jack, r. L. and etheridge, r. jun., Geology and Palaeontology of Queensland
and New Guinea. Brisbane and London.

1907. Lower Cretaceous fossils from the sources of the Barcoo, Ward and Nine Rivers, South

Central Queensland, Part I — Annelida, Pelecypoda and Gastropoda. Rec. Aust. Mus. 6, 317-29.

glaessner, m. f. 1945. Mesozoic fossils from the Central Highlands of New Guinea. Proc. Roy. Soc.
Vic. 56 (n.s.) (2), 151-68.

1958. New Cretaceous fossils from New Guinea. Rec. S. Aust. Mus. 13 (2), 199-226.

hudleston, w. H. 1890. Further notes on some Mollusca from South Australia. Geol. Mag. 27,
241-6.

396

PALAEONTOLOGY, VOLUME 3

Ichikawa, k. 1958. Zur Taxionomia und Phylogenie der Triadischen ‘Pteriidae’ (Lamellibranch).
Pale out ographica, 3 (A), 131-212.

MCWHAE, J. R. H., PLAYFORD, P. E., LINDNER, A. W., GLENISTER, B. F., and BALME, B. E. 1958. The Strati-
graphy of Western Australia. J. geol. Soc. Aust. 4 (2).
moore, c. 1870. Australian Mesozoic geology and palaeontology. Quart. J. geol. Soc. Loud. 26,
226-60.

Newell, n. d. 1938. Late Palaeozoic Pelecypods : Pectinacea. Kansas Geol. Surv. 10.
spitz, a. 1914. A Lower Cretaceous fauna from the Himalayan Gieumal Sandstone together with a
description of a few fossils from the Chikkim Series. Rec. geol. Surv. India , 44 (3), 197-224.
whitehouse, f. w. 1925. On Rolling Downs fossils collected by Prof. J. W. Gregory. Trans. Roy. Soc.
S. Aust. 49, 27-36.

J. M. DICKINS

Geological Section,

Bureau of Mineral Resources, Geology and Geophysics,
Canberra, A.C.T.

Manuscript received 28 August 1959

THE PALAEONTOLOGICAL ASSOCIATION

COUNCIL 1960

President

Professor O. M. B. Bulman, Sedgwick Museum, Cambridge
Vice-Presidents

Dr. R. G. S. Hudson, Trinity College, Dublin
Professor T. N. George, The University, Glasgow
Treasurer : Professor P. C. Sylvester-Bradley, Department of Geology,

The University, Leicester

Assistant Treasurer: Dr. T. D. Ford, Department of Geology,

The University, Leicester

Secretary: Dr. Gwyn Thomas, Department of Geology,

Imperial College of Science, London, S.W. 7
Assistant Secretary: Dr. C. H. Holland, Department of Geology,

Bedford College, London, N.W. 1

Editors

Dr. W. H. C. Ramsbottom, Geological Survey Office, Ring Road Halton, Leeds, 15
Mr. N. F. Hughes, Sedgwick Museum, Cambridge
Dr. W. S. McKerrow, University Museum, Oxford

Other members of Council

Dr. D. V. Ager, Imperial College of Science, London
Dr. F. T. Banner, British Petroleum Company, Sunbury on Thames
Mr. M. A. Calver, Geological Survey Office, Leeds
Dr. W. G. Chaloner, University College, London
Dr. A. J. Charig, British Museum (Natural History), London
Dr. L. R. Cox, British Museum (Natural History), London
Dr. R. H. Cummings, The University, Glasgow
Dr. J. C. Harper, The University, Liverpool
Professor F. Hodson, The University, Southampton
Professor L. R. Moore, The University, Sheffield
Dr. Dorothy H. Rayner, The University, Leeds
Mr. J. D. D. Smith, Geological Survey and Museum, London
Dr. I. Strachan, The University, Birmingham
Dr. C. J. Stubblefield, Geological Survey and Museum, London

Overseas Representatives

Australia: Professor Dorothy Hill, Department of Geology, University of Queensland, Brisbane,
Queensland

Canada: Dr. D. J. McLaren, Geological Survey of Canada, Department of Mines and Technical
Services, Ottawa

New Zealand: Dr. C. A. Fleming, New Zealand Geological Survey, P.O. Box 368, Lower Hutt
Pakistan: Dr. M. Haque, Geological Survey of Pakistan, P.O. Box 15, Quetta, West Pakistan
West Indies and Central America: Dr. L. J. Chubb, Geological Survey Department, Kingston, Jamaica
Eastern U.S.A.: Professor H. B. Whittington, Museum of Comparative Zoology, Harvard Uni-
versity, Cambridge 38, Mass.

Western U.S.A.: Dr. J. Wyatt Durham, Department of Paleontology, University of California,
Berkeley 4, Calif.

PRINTED IN GREAT BRITAIN
AT THE UNIVERSITY PRESS, OXFORD
BY VIVIAN RIDLER
PRINTER TO THE UNIVERSITY

VOLUME 3 • PART 4

c?

Piv

Palaeontology

MARCH 1961

PUBLISHED BY THE
PALAEONTOLOGICAL ASSOCIATION
LONDON

Price £2

THE PALAEONTOLOGICAL ASSOCIATION

The Association was founded in 1957 to further the study of palaeontology. It holds
meetings and demonstrations, and publishes the quarterly journal Palaeontology.
Membership is open to individuals, institutions, libraries, &c., on payment of the
appropriate annual subscription:

Institutional membership . .... £5. 5 s. ($15.50)

Ordinary membership £3. 3s. ($ 9.50)

Student membership £2. 2s. ($ 6.50)

There is no admission fee. Student members will be regarded as persons receiving
full-time instruction at educational institutions recognized by the Council. Subscrip-
tions are due each January, and should be sent to the Treasurer, Professor P. C.
Sylvester-Bradley, Department of Geology, The University, Leicester, England.

Palaeontology is devoted to the publication of papers (preferably illustrated) on all
aspects of palaeontology and stratigraphical palaeontology. Four parts are published
each year and are sent free to all members of the Association. Members who join for
1961 will receive Volume 4, Parts 1 to 4.

Back parts may be purchased separately at a cost of £2 ($6) post free for each part ;
orders should be sent to the approved agents, Messrs. B. H. Blackwell, Broad Street,
Oxford, England.

Manuscripts on all aspects of palaeontology and stratigraphical palaeontology are
invited. They should conform in style to those already published in this journal, and
should be sent to Mr. N. F. Hughes, Sedgwick Museum, Cambridge, England.
A sheet of detailed instructions for authors will be supplied on request.

PALAEONTOLOGY

VOLUME 3 * PART 4

CONTENTS

Upper Devonian Foraminifera from Western Australia. By irene crespin 397

Alimentary Caeca of Agnostids and other Trilobites. By A. A. opik 410

New species of Mandelstamia (Ostracoda) from the English Mesozoic. By

j. w. neale and T. i. kilenyi 439

The feeding mechanism of the Permian Brachiopod Prorichthofenia. By m. j. s.
rudwick 450

Acanthoclymenia, the supposed earliest Devonian Clymenid, is a Manticoceras.

By m. r. house 472

The Carboniferous Rhynchonellid Pugnoides triplex (M‘Coy). By D. parkinson 477

Hoegisporis, a new Australian Cretaceous form genus. By Isabel c. cookson 485

The stratigraphical palaeontology of the Lower Greensand. By Raymond casey 487

The Palaeontological Association: Report and accounts for 1959 622

Index to Volume 3 625

UPPER DEVONIAN FORAMINIFERA FROM
WESTERN AUSTRALIA

by IRENE CRESPIN

Abstract. Eight new species of arenaceous foraminifera are described from the Upper Devonian of the Pillara
Range and Bugle Gap areas, Fitzroy Basin, Western Australia, giving the first record of authentic Devonian
foraminifera in Australia. The species are Rhabdammina virgata, Saccammina glenisteri, Sorosphaera adhaerens,
Lagenammina ampidlacea, Colonammina imparilis, Hyperammina devoniana, Tolypammina helina, and T.
nexuosa.

INTRODUCTION

While investigating the conodont fauna of the Devonian of Western Australia, Dr.
B. F. Glenister of the University of Western Australia treated many limestones from the
Upper Devonian of the Pillara Range and Bugle Gap areas, Fitzroy Basin, with weak
acetic acid. The insoluble residues from these limestones contained not only abundant
assemblages of conodonts but also numerous well-preserved radiolaria and arenaceous
foraminifera. This discovery of authentic Devonian foraminifera in Australia was
briefly recorded by Glenister and Crespin (1959).

Little literature is available on Devonian foraminifera. Good illustrations are scarce
and at times it is difficult to study some of the described forms even generically. How-
ever, several Ordovician and Silurian foraminifera have been described and some of the
figures are helpful at least generically. Workers on Devonian foraminifera include Miller
and Carmer (1933), Bartenstein (1937), Rauzer-Chernoussova (1938), Ireland (1939),
Cushman and Stainbrook (1943), Stewart and Lampe (1947), Pokorny (1951), and Sum-
merson (1958). Several papers on Devonian foraminifera of the U.S.S.R. are known to
have been published and recently the writer received three of them from Dr. Bukova
(1952, 1955, 1958).

In 1918 Chapman described and figured three genera and five species of what he con-
sidered to be foraminifera from thin sections of a limestone of Devonian age from the
Nemingha area, Tamworth, northern New South Wales. However, Wood (1957) stated
that these forms were ‘oolite grains more or less affected by dolomitization and by
mechanical distortion’. Parr (in Teichert and Talent 1958) reported poorly preserved
arenaceous foraminifera from Devonian rocks in the Buchan area, east Gippsland,
Victoria.

Eight new species and two unidentified species of foraminifera from the Upper
Devonian of Western Australia are described in this paper. The tests of all forms are
characterized by delicate wall structure and many are incomplete because of this. It is
interesting to note the similarity of this assemblage from the Devonian with species
described from the Silurian overseas.

All species described herein are housed in the Commonwealth Palaeontological Col-
lection at Canberra, Australia.

[Palaeontology, Vol. 3, Part 4, 1961, pp. 397-409, pi. 64-67.]
B 6612 D d

{INSTITUTION ®

398

PALAEONTOLOGY, VOLUME 3

Stratigraphical notes. The stratigraphy of the Fitzroy Basin has been discussed at length
by McWhae, Playford, Lindner, Glenister, and Balme (1958) and by Guppy, Lindner,
Rattigan, and Casey (1958). The Devonian succession in the Pillara Range and Bugle

text-fig. 1. Locality map.

Gap areas, in which the present microfauna was found, was briefly mentioned by
Glenister and Crespin (1959) and is included here.

Epoch

Age

PHlara Range area

Bugle Gap area

Lower Permian

Sakmarian

Grant Formation

Grant Formation

\

Fairfield Beds

Fairfield Beds

Upper

Famennian

Bugle Gap Limestone

Devonian

Virgin Flills Formation

Virgin Hills Formation

Frasnian

Gogo Formation

Gogo Formation

Sadler Formation

Sadler Formation

Middle Devonian

Givetian

Pillara Formation

Pillara Formation

Precambrian

Ordovician

IRENE CRESPIN: UPPER DEVONIAN FORAMINIFERA 399

The localities from which the samples containing foraminifera were collected (see
text-fig. 1) are:

1. Section at south-east end of Bugle Gap, 18° 44f' S., 126° 05' E. This section probably starts in
the Gogo Formation (0-22 feet) and continues in the Virgin Hills Formation.

2. Two hundred yards south of section 1, and an approximate continuation of it (Virgin Hills
Formation).

3. Between Margaret River and Needle Eye Rocks. Base of section at 18° 16' S., 125° 51' E. and
top at 18° 16' S., 125° 52' E. (Virgin Hills Formation).

Notes on some of the Palaeozoic arenaceous genera. The foraminifera so far found in the
Ordovician, Silurian, and Devonian are all of the arenaceous type. All the forms from
the Upper Devonian of Western Australia have delicate walls composed of medium to
fine, angular quartz grains. Many previously described species have been referred to
genera described from Recent material, such as Bathy siphon. Hyper ammina , Lagenam-
mina, Psammosphaera, Rhabdammina, Saccammina, Sorosphaera, Thurammina, and
Webbinella. Genera described from the Ordovician, Silurian, and Devonian include
Colonammina Moreman (1930), Arenosiphon Grubbs (1939), Ceratammina Ireland
(1939), Fairiella Summerson (1958), Kerionammina Moreman (1933), Kettneranunina
Pokorny (1951), Bifurcammina Ireland (1939), Moravammina Pokorny (1951), Soro-
sphaeroidea Stewart and Larnpe (1947), Steganammina Moreman (1930), Webbinelloidea
Stewart and Larnpe (1947), Wiekkoella Summerson (1958), and Vasicekia Pokorny
(1951), and several others by Bukova (1952, 1955).

It is unfortunate that many genera and species of foraminifera described and figured
from the Ordovician, Silurian, and Devonian are difficult to study. This is especially the
case with regard to the Recent genus Sorosphaera Brady 1879, of which several species
are described from older rocks, and the Devonian genus Sorosphaeroidea Stewart and
Larnpe 1947. Species of both these polythalamous genera are figured by Stewart and
Larnpe (1947, pi. 78) and the difference between them is difficult to detect. They in-
dicate that in Sorosphaeroidea the chambers are more intimately joined and are arranged
more nearly in a plane than in Sorosphaera. Neither of these features seems consistent
enough in the many specimens available for study in the Western Australian material to
justify their reference to Sorosphaeroidea rather than Sorosphaera. Summerson (1958)
uses the genus Sorosphaeroidea for his Middle Devonian specimens and described two
species of the genus. However, the variation within a species as shown by the Western
Australian specimens suggests that these two species may be synonymous.

I am in agreement with Grubbs (1939, p. 544) who, when discussing his species Soro-
sphaera irregularis, stated that ‘several species have been erected on which the deter-
mination apparently was based on the number of cells, but from the material at hand, it
is evident that a rupture of some of the composite individuals would obviously produce
several new ones having variable numbers of chambers. Thus a specific description would
possibly apply to only a single individual. Forms of such variable attributes are here
grouped into one species.’ Gutschick and Treckman (1959, p. 232), discussing their
Pennsylvanian species Sorosphaera papilla , express agreement with Grubbs and add that
‘it seems superfluous to attempt to split this form into several species based on material
in the collection’. The variation within Sorosphaera adhaerens sp. nov. is illustrated on
PI. 66.

Some of the specimens from the Ordovician, Silurian, and Devonian referred to the

400

PALAEONTOLOGY, VOLUME 3

genus Bathysiphon Sars 1878 require generic revision. Cushman (1950), Galloway ( 1933),
and Glaessner (1945) refer to Sars’s definition of Bathysiphon as having a thick wall
and being typically composed of broken sponge spicules overlain by fine-grained
arenaceous material. Grubbs (1939) introduced a new genus Arenosiphon for a cylindri-
cal, commonly tapering test composed of quartz grains of nearly uniform size and with
apertures at either end of the test. He considered that several small Silurian forms
previously referred to Bathysiphon should be included in this genus. Ireland (1939), in
describing his species Bathysiphon rugosus from the Silurian of Oklahoma, states that
the wall is composed of well-cemented medium-sized sand grains and continues ‘the
rough exterior of this form and the large size of the sand grains show it ( B . rugosus )
to be different from all other described species of Bathysiphon .

It is most probable that these specimens with walls composed of angular quartz
grains and included under Bathysiphon and Arenosiphon should be included in the genus
Rhabdammina Sars 1869. Species of this genus have been described from the Ordovician
and Silurian. The chitinous inner layer of the tests and the outer wall of sand grains
suggest that the specimens from the Virgin Hills Formation belong to the family Astro-
rhizidae and are most probably referable to the genus Rhabdammina rather than to Bathy-
siphon as suggested in Glenister and Crespin (1959). Ireland’s Silurian species Bathysiphon
rugosus referred to in the previous paragraph closely resembles Glaessner’s Paleocene
species Rhabdammina cylindrica.

Specimens referable to the genus Colonammina Moreman 1930, a small attached
Lagenammina-Yike form from the Silurian of Oklahoma, have been recognized in t_he
assemblage from the Virgin Hills Formation.

Future investigations on the wall structure of numerous specimens from the Ordo-
vician, Silurian, and Devonian may prove that these early primitive forms now referred
to Recent genera are generically distinct from the Recent ones, and that some of the
names now erected for them are valid. However, extensive collections are necessary for
such work. The present work is based on a small collection only.

SYSTEMATIC DESCRIPTIONS
Family astrorhizidae
Genus rhabdammina M. Sars 1869
Rhabdammina virgata sp. nov.

Plate 64, figs. 7-10; Plate 65, figs. 1, 2

Occurrence. Holotype (C.P.C. No. 394) and paratypes B and C (C.P.C. Nos. 396, 397) from locality
between Margaret River and Needle Eye Rocks, Virgin Hills Formation. Paratypes A, D, and E
(C.P.C. Nos. 395, 398, 399) from section south-east end of Bugle Gap, 44-66 feet above base, Virgin
Hills Formation.

Dimensions. Length of tube — holotype, 1-51 mm.; paratype A, 103 mm.; paratype B, 2-60 mm.;
paratype C, 1-63 mm.; paratype D, 0-97 mm.; paratype E, L16 mm. Width of inner chitinous tube of
paratype A, 01 1 mm.

Diagnosis. This species has a free, elongate tube which may be curved, sinuous, straight,
or tapering, and with apertures at both ends of the tube. The outer wall is thin and com-

IRENE CRESPIN: UPPER DEVONIAN FORAMINIFERA 401

posed of angular quartz grains of varying sizes, giving a rough surface; the inner wall is
a thin chitinous layer.

Holotype. Test free, an elongate tube, slightly curved, chiefly uniform in width but with
occasional constrictions. Wall thin, composed of medium and fine angular quartz
grains, finely cemented. Surface rough. Inner wall smooth, composed of thin chitinous
material. Apertures at both ends of test.

Paratype A. Specimen with portion of outer wall of quartz grains in siliceous cement
removed and exposing the thin, smooth, chitinous inner layer.

Paratype B. Specimen long, showing sinuous character of tube, outer surface rough with
wall composed of moderately large quartz grains; chitinous inner wall showing at one
end of tube.

Paratype C. Similar to holotype.

Paratype D. Tube coiled in early portion then becoming almost erect. Surface rough with
several large quartz grains in cement.

Paratype E. Tube tapering rather rapidly, with smooth chitinous layer of inner wall
showing at one end.

Observations. R. virgata is one of the commonest species in the Virgin Hills Formation
assemblage, being especially numerous in the residue from the locality between Mar-
garet River and Needle Eye Rocks. All specimens from this locality are white in colour;
at the south-east end of Bugle Gap they are brown with considerable limonitic replace-
ment. Comments on the genus Rhabdammina are given in the previous section. The
species name is taken from the Latin virgatus — rodlike.

Family rhizamminidae
Genus marsipella Norman 1878
Marsipella sp.

Plate 65, fig. 9

Occurrence. Figured specimen (C.P.C. No. 400) from section between Margaret River and Needle Eye
Rocks, Virgin Hills Formation.

Dimensions. Length of figured specimen, 0-49 mm.

Observations. A fragment of an elongate tubular test with an outer wall composed of
coarse, angular, quartz grains and an inner one with a roughened surface suggests the
genus Marsipella. The present specimen closely resembles M. aggregata Moreman
(1933) from the Silurian of Oklahoma.

Family saccamminidae
Genus saccammina M. Sars 1869
Saccammina glenisteri sp. nov.

Plate 65, figs. 3-7

Occurrence. Holotype (C.P.C. No. 401) and paratypes A and B (C.P.C. Nos. 402, 403) from south-east
end of Bugle Gap, 66-88 feet above base of section. Virgin Hills Formation. Paratype D (C.P.C.

402

PALAEONTOLOGY, VOLUME 3

No. 405) from same locality as above but 0-22 feet above base of section and probably Gogo Forma-
tion. Paratype C (C.P.C. No. 404) section between Margaret River and Needle Eye Rocks, Virgin
Hills Formation.

Dimensions. Maximum width of holotype, 0-37 mm.; paratype A, 0-39 mm.; paratype B, 0-42 mm.;
paratype C, 0-34 mm. Thickness of wall of paratype C, 0 04 mm.

Diagnosis. This species has a small, more or less sphaerical test, usually free but occa-
sionally lightly attached, and composed of small angular quartz grains cemented
together. An apertural opening is usually present.

Holotype. Test free, almost sphaerical, small. Wall thin, composed of medium to small
angular quartz grains in fine siliceous cement. Surface rough. Aperture a small elongated
slit-like opening in centre of test.

Paratype A. Test less regularly sphaeroidal in shape than holotype. Surface rough but
grains more even in size. Aperture elongate and slightly more open than in holotype.

Paratype B. Test irregularly sphaeroidal with rough surface as angular edges of quartz
grains protrude from cement. Wall thin. Aperture circular.

Paratype C. Test broken showing hollow central cavity, thin wall of test and roughened
inner wall.

Paratype D. Two sphaeroidal tests gently attached to fragment of Rhabdammina.

Observations. Some species of Saccammina described from the Ordovician, Silurian,
and Devonian have apertural openings at the end of a slightly protruding neck but this
feature is not essential to the determination of the genus. The present species, S. glenis-
teri , has a small aperture which may be slit-like or rounded and it does not protrude
beyond the surface of the test. S. ingloria Bukova (1952) from the Devonian of the Ural
region resembles S. glenisteri in the composition of the wall which is made up of angular
quartz grains closely cemented, the angular grains also forming a rough surface to the
inner wall. It differs, however, in its short neck for the apertural opening.

The species is named after Dr. B. F. Glenister who discovered the present microfauna
in the Upper Devonian rocks of Western Australia.

EXPLANATION OF PLATE 64

Figs. 1-6. Hy perammina devoniana sp. nov. 1, Holotype. Showing globular proloculus and long, nar-
row, tubular second chamber. 2, Paratype A. Large globular proloculus followed by delicate tubular
second chamber. 3, Paratype E. Showing thin wall of sinuous tubular second chamber. 4, Paratype
B. Globular proloculus followed by tubular second chamber gradually decreasing in size. 5, Para-
type D. Showing globular proloculus and sinuous tubular second chamber. 6, Paratype C. Showing
proloculus followed by straight tubular second chamber.

Figs. 7-10. Rhabdammina virgata sp. nov. 7, Holotype. Showing openings at either end of tube,
rough outer surface and inner chitinous wall. 8, Paratype B. Showing sinuous character of tube.
9,ParatypeE. Showing rapidly tapering test with smooth chitinous inner wall. 10, Paratype A. Por-
tion of outer wall removed to show smooth chitinous inner layer.

All figures x 92

Palaeontology, Vol. 3.

PLATE 64

CRESPIN, Upper Devonian Foraminifera, W. Australia

IRENE CRESPIN: UPPER DEVONIAN FORAMINIFERA 403

Genus sorosphaera Brady 1879
Sorosphaera adhaerens sp. nov.

Plate 66, figs. 1-6

Occurrence. Holotype (C.P.C. No. 406) and paratypes A, B, C, and D (C.P.C. Nos. 407, 408, 409, 410)
from section south-east end of Bugle Gap, Virgin Hills Formation, between 66 and 88 feet above
base of section.

Dimensions. Average width of chamber — holotype, 0-37 mm.; paratype A, 0-28 mm.; paratype B
0-34 mm. ; paratype C, 0-30 mm. ; paratype D, 0-47 mm.

Diagnosis. Test arenaceous consisting of an aggregate of two or more chambers intimately
attached to one another around a central chamber or in linear arrangement. The cham-
bers are rounded on dorsal surface, flattened on ventral side indicating attachment. The
wall is thin and the aperture, when observed, is central.

Holotype. Test arenaceous, medium sized, consisting of an aggregate of eight chambers
attached in irregular linear arrangement along a fragment of Tolypammina. Each
chamber rounded on dorsal surface and flattened on ventral one. Chambers almost
uniform in size and firmly attached to one another, line of attachment being slightly
polygonal in outline. Wall thin, composed of quartz grains which are large for size of
chamber and set in fine siliceous cement. Surface rough. Aperture a minute rounded
opening in centre of rounded portion of chamber, the area around this opening slightly
constricted in one of the chambers.

Paratype A. Test consisting of six chambers arranged in branching fashion, all strongly
attached to one another. Two attached chambers in stem and two in each branch. Wall
composed of coarse quartz grains with little visible cement. Surface very rough. Aper-
ture observed in only one chamber.

Paratype B. Test consisting of three rounded chambers attached to one another in a
linear direction, with suggested polygonal outline at point of attachment and flattened on
lower surface. Dorsal surface very rough, wall on both surfaces composed of coarse,
angular, quartz grains. Minute aperture present in middle chamber, close to line of
attachment between chambers.

Paratype C. Test consisting of two chambers strongly attached to one another and also
attached to fragment of Tolypammina. Aperture in one chamber near line of attachment,
and in second chamber it appears as an opening in short neck composed of quartz
grains.

Paratype D. Single chamber strongly attached to fragment of Tolypammina. Wall thin,
composed of coarse quartz grains. Aperture almost rounded opening.

Observations. Many moderately well-preserved specimens of Sorosphaera adhaerens
were available for study in one sample and it seems certain that the variation in the
number of chambers present in the different specimens is comparable with the habit of
growth of other polythalamous foraminifera. The persistent uniformity of composition
of the chambers’ walls, which includes the size of the component quartz grains, and the
persistent delicate wall structure indicate that the variation in the number of chambers
is all within the one species.

404

PALAEONTOLOGY, VOLUME 3

S. adhaerens shows some similarity with S. pentachord and S. trichora of Summerson
(1958) from the Middle Devonian of Ohio, in the shape of the chambers and the penta-
gonal lines of attachment of chambers. However, in the Western Australian form the
chambers appear to be more intimately attached.

Genus proteonina Williamson 1858
Proteonina sp.

Plate 65, fig. 8

Occurrence. Figured specimen (C.P.C. No. 411) from south-east end of Bugle Gap, between 0-22 feet
above base of section, probably Gogo Formation.

Dimensions. Maximum width of figured specimen, 0-43 mm.; length, 0-58 mm.

Observations. The test of this form is very compressed, with a short apertural neck
situated to one side of the test. The wall is coarsely arenaceous and the aperture is at the
end of a compressed tubular neck. Only one specimen has been found. Its shape re-
sembles that of P. pseudospiralis Cushman and Stainbrook (1943) from the Devonian of
Iowa.

Genus lagenammina Rhumbler 1911
Lagenammina ampullacea sp. nov.

Plate 66, figs. 6-8

Occurrence. Holotype (C.P.C. No. 412) and paratypes A and B (C.P.C. Nos. 413, 414) from section
south-east of Bugle Gap, 66-88 feet above base of section. Virgin Hills Formation.

Dimensions. Length of test, including neck — holotype, 0-44 mm.; paratype A, 0-41 mm.; paratype B,
0-49 mm. Maximum diameter of holotype, 0-41 mm.; paratype A, 0-35 mm.; paratype B, 0-37 mm.

Diagnosis. The species is almost sphaerical in shape with a short protruding neck. The

EXPLANATION OF PLATE 65

Figs. 1, 2. Rhabdammina virgata sp. nov. 1, Paratype C. Showing openings at both ends of slightly
curved tubular chamber. 2, Paratype D. Early portion of tubular chamber coiled, later being almost
straight.

Figs. 3-7. Saccammina glenisteri sp. nov. 3, Holotype. Almost sphaerical test showing slit-like aper-
ture. 4, Paratype C. Test broken to show thin wall and central cavity. 5, Paratype A. Showing
rough surface of test. 6, Paratype B. Irregular-shaped test attached to fragment of Rhabdammina.
7, Paratype D. Two sphaerical tests attached to fragment of Rhabdammina.

Fig. 8. Proteonina sp. Showing compressed test with short apertural neck.

Fig. 9. Marsipella sp. Broken specimen showing roughened surface of inner wall.

Figs. 10-13. Colonammina imparilis sp. nov. 10, Holotype. Showing roughened surface of unattached
surface and groove with smooth surface indicating area of attachment. 11, Paratype B. Showing
smooth flat area of attachment. 12, Paratype A. Area of attachment flattened and smooth, and
protruding neck of homy material. 13, Paratype C. Area of attachment near lower portion of
test.

All figures X 92

Palaeontology, Vol. 3.

PLATE 65

S&i l
- sm-'h

CRESPIN, Upper Devonian Foraminifera, W. Australia

IRENE CRESPIN: UPPER DEVONIAN FORAMINIFERA

405

wall is moderately thin and composed of angular quartz grains. The surface is rough.
The aperture is circular at the end of the neck.

Holotvpe. Test free, almost sphaerical in shape with a short protruding neck. Wall
moderately thin, composed of quartz grains of varying sizes in a siliceous cement. Sur-
face rough. Aperture circular at the end of short neck.

Paratype A. Test irregularly sphaerical, with surface rougher than holotype.

Paratype B. Test with neck incomplete, but measuring at least a third of the length of the
entire test; coarse quartz grains in wall of neck.

Observations. L. ampullacea appears to be the first record of the genus Lagenammina in
the Devonian. Several species have been described from the Silurian. L. ampullacea
differs from the two species L. sphaerica and L. cucurbit a described by Moreman (1930;
1933) from the Silurian of Oklahoma, in its shorter neck; this character, however,
shows some resemblance to L. bulbosa Dunn (1942) and L. cornuta Grubbs (1939),
both from the Silurian.

The species name is derived from the Latin ampullaceus, flask-like.

Genus colonammina Moreman 1930
Colonammina imparilis sp. nov.

Plate 65, figs. 10-13

Occurrence. Holotype (C.P.C. No. 415) and paratypes A, B, and C (C.P.C. Nos. 416, 417, 418) from
section south-east end of Bugle Gap, 66-88 feet above base of section, Virgin Hills Formation.

Dimensions. Length of test including neck — holotype, 0-57 mm. ; paratype A, 0-43 mm. ; paratype B,
0-50 mm. ; paratype C, 0-45 mm. Maximum width of holotype, 0-50 mm.

Diagnosis. The test is attached and is irregularly sphaerical in shape on the unattached
side. It has a protruding neck on the convex surface and the wall is composed of well-
cemented angular quartz grains.

Holotype. Test attached, medium sized, subsphaerical with a protruding erect neck.
Wall moderately thin, composed of medium to fine angular quartz grains well cemented.
Surface rough. On one side of test immediately below neck where it protrudes from the
rounded test is a groove indicating attachment. This groove reveals a wall of fine siliceous
material. The neck is composed of chitinous material. Aperture is a small rounded
opening at end of the neck.

Paratype A. Test smaller than holotype and more irregular in shape and with a long,
protruding, erect neck of horny material. Area of attachment flattened and smooth.
Wall of convex surface thinner than holotype and composed of angular quartz grains.

Paratype B. Test irregular in shape, somewhat flattened on one side. Wall thin composed
of quartz grains of varying sizes which also cover the protruding neck. Neck slightly
curved.

Paratype C. Test almost sphaerical, but with smooth area of attachment on one side of
lower portion of test. Neck protruding, erect, composed of chitinous material.

406

PALAEONTOLOGY, VOLUME 3

Observations. Several specimens of Colonammina imparilis are present at the type
locality. All show variation in shape and position of area of attachment and the protrud-
ing neck composed of horny-like material. This genus was originally described by More-
man (1930) from the Silurian of Oklahoma. Although Moreman has described two
species, C. verruca and C. conea, the variation in shape found in the specimens from
Western Australia suggests that probably these should be referred to one species only.
The specific name is derived from the Latin imparilis , unequal.

Family hyperamminidae
Genus hyper ammina H. B. Brady 1878
Hyperammina devoniana sp. nov.

Plate 64, figs. 1-6

Occurrence. Holotype (C.P.C. No. 419) and paratypes A, B, C, D, E (C.P.C. Nos. 420, 421, 422, 423,
424) from section south-east end of Bugle Gap, 66-88 feet above base of section, Virgin Hills Forma-
tion.

Dimensions. Length of incomplete test of holotype, 0-82 mm. ; diameter of globular proloculus, 0- 1 6 mm.

Diagnosis. The species has an elongate test composed of a globular proloculus followed
by a tubular second chamber which may be straight or sinuous. The wall is thin and
composed of very fine quartz grains.

Holotype. Test free, delicate, elongate, incomplete, consisting of a small but prominent
globular proloculus followed by a long, rather narrow, tubular second chamber. Wall
thin, composed of small quartz grains in siliceous cement. External surface rough, inter-
nal wall smooth. Aperture at open end of tube.

Paratypes A, B , C, D, E. These specimens illustrate the variation within the species
of the shape of the tubular second chamber, paratype E especially exhibiting the sinuous
character of that chamber. This specimen also shows the delicate character of the wall,
with its smooth inner surface.

Observations. Although the genus Hyperammina is recorded from the Devonian by
Cushman and Stainbrook (1943) and Bartenstein (1937), no species have been described
as far as the author can discover. However, species are described from the Ordovician
and Silurian. H. devoniana shows some resemblance to H. hastula Moreman (1933)

EXPLANATION OF PLATE 66

Figs. 1-5. Sorosphaera adhaerens sp. nov. 1, Holotype. Aggregate of eight chambers firmly attached
to one another in an irregular linear direction along fragment of Tolypammina. 2, Paratype A. Six
chambers arranged in branching fashion. 3, Paratype B. Three chambers attached to one another in
linear direction. Surface rough. 4, Paratype D. Single chamber attached to Tolypammina. 5, Para-
type C. Two chambers strongly attached to one another and to fragment of Tolypammina.

Figs. 6-8. Lagenammina ampullacea sp. nov. 6, Holotype. Showing almost sphaerical test with short
protruding neck. 7, Paratype A. Showing rough surface and short neck. 8, Paratype B. Showing
long, incomplete, neck with coarse quartz grains in wall of neck.

All figures X 92

Palaeontology, Vol. 3.

PLATE 66

C RES PIN, Upper Devonian Foraminifera, W. Australia

IRENE CRESPIN: UPPER DEVONIAN FORAMINIFERA

407

from the Silurian of Oklahoma, but differs from that species in its more delicate tubular
second chamber which at times becomes sinuous. This latter character is shown in H.
baltica Eisenack (1954) from the Silurian. The prominent globular proloculus is charac-
teristic of all specimens figured from the Ordovician, Silurian, and Devonian. H. graci-
lenta Gutschick and Treckman (1959), from the Pennsylvanian of Indiana, closely
resembles H. devoniana in its prominent globular proloculus followed by a delicate
tubular chamber.

Family ammodiscidae
Genus tolypammina Rhumbler 1895
Tolypammina helina sp. nov.

Plate 67, figs. 1-5

Occurrence. Holotype (C.P.C. No. 425) and paratypes A and B (C.P.C. Nos. 484, 485) from section
200 yards south of Bugle Gap, south-east end, 88-100 feet above base of section. Virgin Hills Formation.

Dimensions. Width of tubular second chamber near spiral portion of proloculus — holotype, 0-13 mm. ;
paratype A, Oil mm. Maximum width of spiral portion — holotype, 0-39 mm.; paratype A, 0-41 mm.

Diagnosis. The proloculus of this species is followed by a tubular second chamber which
is coiled for about two whorls, then becoming erect or somewhat irregular in direction.
The wall is very thin.

Holotype. Test attached throughout length, consisting of proloculus followed by a tubu-
lar second chamber which is coiled for at least two whorls, then becoming erect. This
chamber is narrow near the proloculus then gradually increases in width. Wall very thin,
as shown by fracture of attached surface, and composed of fine quartz grains in siliceous
cement. Surface rough. Aperture not observed but probably at open end of tube.

Paratype A. Tubular second chamber more loosely coiled than in holotype and very
gradually increasing in width. Wall as seen on attached surface, very thin, composed of
fine quartz grains of varying sizes. Surface rather rough.

Paratype B. Tubular chamber irregularly and loosely coiled, and then winding in
different directions.

Observations. Tolypammina helina cannot be compared with any described species from
the Ordovician or Silurian and it appears that no species has been described from the
Devonian. It differs from T. nexuosa sp. nov. in its coiled portion of the tubular cham-
ber and in the tendency in some tests for the chamber to be erect rather than sinuous
as in that species.

The species name is derived from the Greek helinos , tendril.

Tolypammina nexuosa sp. nov.

Plate 67, figs. 6-8

Occurrence. Holotype (C.P.C. No. 486) and paratype A (C.P.C. No. 487) from south-east of Bugle
Gap, 66-88 feet above base of section, Virgin Hills Formation. Paratype B (C.P.C. No. 488) 200 yards
south of south-east end of Bugle Gap, 44-66 feet above base of section.

Dimensions. Maximum length of test — holotype, 1-15 mm.; paratype A, 0-73 mm.; paratype B, 1-33
mm. Maximum width of tube — holotype, 0- 1 5 mm. ; paratype A, 0- 1 3 mm.

408

PALAEONTOLOGY, VOLUME 3

Diagnosis. This attached species consists of a proloculus followed by a long second tubu-
lar chamber which winds loosely in irregular directions with the proloculus and the end
of the tube at times becoming attached. The wall is thin and the surface smooth.

Holotype. Test attached, consisting of a sphaerical proloculus followed by a long tubular
second chamber which is loosely coiled in the early portion, then becoming sinuous
and at times curling back on itself and becoming attached to the proloculus. Wall thin,
composed of fine quartz grains well cemented, and with occasional coarser grains.
Unattached surface smooth and at times polished. Attached surface where broken shows
roughened inner wall. Aperture not visible and is probably attached to outside wall of
sinuous chamber.

Paratype A. Proloculus broken, tubular second chamber gradually increasing in size
as it winds around to form an irregular circle. Wall thin. Surface smooth, at times
polished.

Paratype B. Sinuous tubular second chamber gradually increasing in width.

Observations. T. nexuosa is common in some residues. It winds in many irregular direc-
tions and for the most part is attached. The species closely resembles T. tortuosa Dunn
(1942) from the Silurian of the Mississippi Basin. It differs from that species in its smooth
and at times polished surface of the unattached side. As with T. tortuosa, although many
specimens of T. nexuosa are available, no two are exactly alike. It also resembles T. poly-
verta Ireland (1958) from the Pennsylvanian of Kansas.

The species name is derived from the Latin nexuosus, complicated.

Acknowledgements. Thanks are expressed to the Director of the Bureau of Mineral Resources, Geology
and Geophysics, Canberra, Australia, for permission to publish this paper; to Dr. B. F. Glenister for
making available the material for study; and to F. Hadzel of the Bureau of Mineral Resources for the
excellent illustrations.

REFERENCES

bartenstein, h. 1937. Neue Foraminiferen-Funde in Mittel-Devon der Eifel. Senckenbergiana, 19,
334-8, 516.

brady, h. b. 1879. Notes on some of the reticularian Rhizopoda of the Challenger Expedition.
Quart. J. Micr. Soc., N.s., 19, 1-46, pi. 3-5.

bykova, e. v. 1952. Foraminifera of the Devonian of the Russian Platform and Ural region.
V.N.I.G.R.I. 60 (5), 5-64, pi. 1-14.

explanation of plate 67

Figs. 1-5. Tolypammina lielicina sp. nov. 1, Holotype. Unattached surface of test with smooth wall,
minute proloculus, followed by coiled tubular chamber which later becomes erect. 2, Attached sur-
face of holotype. 3, Paratype A. Unattached surface with loosely and irregularly coiled tubular
chamber. 4, Attached surface of Paratype A. 5, Paratype B. Loosely and irregularly coiled tubular
chamber later becoming sinuous. Unattached surface.

Figs. 6-8. Tolypammina nexuosa sp. nov. 6, Holotype. Unattached surface with smooth wall, long
sinuous tubular second chamber curling back on itself. 7, Paratype A. Tubular chamber gradually
increasing in size and winding around to form an irregular circle. Unattached surface. 8, Paratype
B. Smooth sinuous tubular chamber gradually increasing in size.

All figures x 92

Palaeontology, Vol. 3.

PLATE 67

C RES PIN, Upper Devonian Foraminifera, W. Australia

IRENE CRESPIN: UPPER DEVONIAN FORAMINIFERA 409

bykova, E. v. 1955. Foraminifera and radiolaria of the Devonian of the Volga-Ural province.
V.N.I.G.R.I. 87, 5-141, pi. 1-24.

1958. On the discovery of chitinous foraminifera in Devonian sediments of northern Kazakhstan.

Acad. Sci. U.S.S.R., Doklady, 120 (4), 879-82.

chapman, f. 1918. Devonian foraminifera: Tamworth District, New South Wales. Proc. Linn. Soc.
N.S.W. 43 (2), 385-94, pi. 39-41.

cushman, j. a. and stainbrook, m. a. 1943. Some foraminifera from the Devonian of Iowa. Cush.
Lab. For am. Res. 19 (4), 73-79, pi. 13.

dunn, p. h. 1942. Silurian foraminifera of the Mississippi Basin. /. Paleont. 16, 317-42, pi. 42-44.
eisenack, a. 1932. Neue Microfossilen der baltischen Silur. II. Pal. Zeit. 14, 257-77, pi. 11, 12.

1954. Foraminiferen aus der baltischen Silur. Senckenbergiana, 35, 51-72, pi. 1-5.

glenister, b. f. 1958. Upper Devonian ammonoids from the Manticoceras Zone, Fitzroy Basin,
Western Australia. J. Paleont. 32, 58-96, pi. 5-15.

and crespin. i. 1959. Upper Devonian microfauna from the Fitzroy Basin, Western Australia.

Aust. J. Sci. 21 (7), 222-3.

grubbs, d. m. 1939. Fauna of the Niagaran nodules of the Chicago area. J. Paleont. 13, 543-60,
pi. 61-62.

guppy, d. j., Lindner, A. w., rattigan, j. h., and casey, j. n. 1958. The geology of the Fitzroy Basin,
Western Australia. Aust. Bur. Min. Resour. Bull. 36.
gutschick, r. c. and treckman, j. f. 1959. Arenaceous foraminifera from the Rockford Limestone of
northern Indiana. /. Paleont. 33, 229-50.

Ireland, h. a. 1939. Devonian and Silurian foraminifera from Oklahoma. J. Paleont. 13, 190-202.
— — - 1956. Upper Pennsylvanian arenaceous foraminifera. Ibid. 30, 831-67.

krestovinkov, v. n. and rauzer-chernoussova, d. p. 1938. On the foraminifera from the transitional
beds between the Devonian and the Carboniferous (Etroeungt Zone) of Kazakhstan, South Urals
and Samaraskaya Luka. Acad. Sci. U.S.S.R., Doklady, 20, 587-9.
mcwhae, j. r. h., playford, p. e., lindner, a. w., glenister, b. f., and balme, b. e. 1958. The strati-
graphy of Western Australia. J. geol. Soc. Aust. 4 (2), (1956).
miller, A. k. and carmer, a. n. 1933. Devonian foraminifera from Iowa. J. Paleont. 7, 423-31,
pi. 50.

moreman, w. l. 1930. Arenaceous foraminifera from Ordovician and Silurian limestones of Okla-
homa. Ibid. 4, 42-59, pi. 5-7.

1933. Arenaceous foraminifera from the Lower Paleozoic of Oklahoma. Ibid. 7, 393-7, pi. 47.

pokorny, v. 1951. The Middle Devonian foraminifera of Celechovice, Czechoslovakia. Kral.Ceske.

Spol. Nauk. Trida Mat-Prirod. Vestnik, Pragus (1953), 9.
stewart, g. a. and lampe, l. 1947. Foraminifera from the Middle Devonian Bone Beds of Ohio.
/. Paleont. 21, 529-36, pi. 78-79.

- — and priddy, r. r. 1941. Arenaceous foraminifera from the Niagaran rocks of Ohio and Indiana.
Ibid. 15, 366-75, pi. 54.

summerson, c. h. 1958. Arenaceous foraminifera from the Middle Devonian Limestone of Ohio.
Ibid. 32, 544-88, pi. 81-82.

teichert, c. and talent, j. s. 1958. Geology of the Buchan area, east Gippsland, Victoria. Geol.
Surv. Viet. Mem. 21.

wood, A. 1957. The Devonian ‘foraminifera’ from Tamworth, New South Wales. Mem. Nat. Mus.
Viet. 22, (5), 1-4, pi. 1.

IRENE CRESPIN

Bureau of Mineral Resources,
Geology and Geophysics,
Canberra, Australia

Manuscript received 28 August 1959

ALIMENTARY CAECA OF AGNOSTIDS AND OTHER

TRILOBITES

by A. A. opik

Abstract. Genal and other caeca of trilobites have left distinctive imprints of veins and lines on the test; all such
‘ornamental lines and veins’ are referred to as ‘alimentary prosopon’. In agnostids the alimentary prosopon
consists of scrobiculae and rugae; the genal caeca of agnostids form a left and a right reniform ramified gland,
connected with the glabella (oesophagus) by one or two pairs of short diverticula. The caeca are few and re-
latively thick, and reticulation is scant. The scrobicules probably correspond to mesenteries, and the preglabellar
frontal furrow probably corresponds to a median mesentery. The frontal sulcus probably indicates that the front
of the oesophagus was bilobed. The pygidial caecal apparatus of agnostids is similar to that of the cephalon.
The alimentary apparatus of agnostids is similar to that of Burgessia.

In Redlichici and olenellids the genal caeca consist of radiating and anastomosing capillary vessels, whereas
in the fixed cheek, including the palpebral lobes, ingluvial sacs replace the caeca. In Redlichia each segment of
the thorax bears two pairs of veins, which are regarded as intestinal appendages (caeca or diverticula).

In ptychopariids ( Papyriaspis ) the genal caeca are of capillary size; the ocular ridges consist of two pairs of
lines, interpreted as diverticula, with the palpebral lobes and the eye at their tips. The caecal veins of the postero-
lateral border terminate in a swelling (opistopleural terminal glandular swelling). Each segment of the thorax
bears two pairs of intestinal appendages, and the anterior segments possess as well the 'glandular swelling’ at the
tips of the pleurae. The intestinal appendages are probably homologous with the ocular ridges, and the opisto-
pleural swellings with the eyes. The anterior part of the pygidium repeats the structure of the thorax; in the
posterior part of the pygidium the segments fuse and the caeca radiate, anastomose, and reticulate, and so repeat
the pattern of the genal caeca. This is also demonstrated on the pygidium of an undescribed ptychopariid, and
a pleura of Centropleura shows structures similar to those described in ptychopariids.

Trilobites have segmentally arranged intestinal appendages; their eyes are segmental and located at the tips of
a pair of pleurae; caeca are developed only on the pleural lobe; and the cephalic, frontal, border, and doublure
including the rostral shield are pleural elements and not part of the axial lobe.

As examples of agnostids with well-developed cephalic and pygidial caeca Glyptagnostus and its two species —
reticulatus (Angelin) and stolidotus sp.nov. — are described. The caecal prosopon of G. stolidotus consists of
conspicuous rugae, and the cephalic and pygidial glands are almost mirror images of one another.

INTRODUCTION

The aim of the present paper is a description of the alimentary glands known in the
literature as ‘genal caeca’ of trilobites. The material is arranged in two parts: first, the
principal morphological observations and interpretations are presented, and, secondly,
the results are tested in the taxonomy and morphology of Glyptagnostus.

The segmental arrangement of caecal glands, their presence in all tagmata, and their
duplication in each segment, are the main results of the study. A new trilobite model can
be visualized when the present results are combined with ideas concerning the origin of
the pygidium and the hypostoma, and the pleural nature of the rostral shield (Opik 1958).
A preliminary note has already been published (Opik 1959).

Observations on the anatomy and organization of trilobites are steadily accumulating,
and, hand in hand with this progress, theories arise on the origin, phylogeny, and pro-
geny of trilobites. It becomes increasingly evident that trilobites are neither Crustacea
nor Merostomata, although they are commonly interpreted by comparison with the
latter. In the present study a somewhat different approach has been attempted: Cam-

[ Palaeontology, Vol. 3, Part 4, 1961, pp. 410-38, pi. 68-70.]

A. A. OPIK: ALIMENTARY CAECA OF AGNOSTIDS 411

brian trilobites are studied here solely on their own merits, and post-Cambrian trilobites
(and possible relatives and descendants) are not taken into consideration because they
had no influence on Cambrian life and its evolution. Of course, trilobites are here
regarded as arthropods, and the general organization of arthropods is considered in the
interpretation of the observations.

The morphology of trilobites is not yet known completely, for the obvious reason of
incomplete preservation. Still, recent finds of well-preserved material have provided
more information; in particular, the cuticular sculpture, which is termed the ‘alimentary
prosopon’, is exceptionally well preserved in the Australian material, which has been
collected in undisturbed rocks without any trace of metamorphism. The term ‘alimen-
tary prosopon’ refers to traces of the cephalic caeca and their homologues on the thorax
and the pygidium. The alimentary prosopon is in effect a map of the pleural alimentary
apparatus of trilobites.

G. Lindstrom, P. E. Raymond, C. D. Walcott, A. Born, R. Richter, and L. Stormer
are the main contributors to the knowledge of the alimentary apparatus of trilobites.
The most recent summary of current knowledge is given by Hupe (1953a, pp. 82-84) as
follows: the glabella contains the oesophagus, from which caeca extend into the cheeks
and the frontal limb; the intestine, which has no lateral ramifications, extends from the
rear of the oesophagus along the axis of the thorax and the pygidium. He also refers to
the observation by Yolborth of a metamerized intestine in Illaenus.

Hupe also suggests that the genal caeca emerge from the oesophagus and are connected
to it by diverticula. He (1953a, p. 81 , fig. 35) assumes that at the maximum development
of metamerization the head of a trilobite is equipped with four pairs of diverticula with
ramified caeca. The foremost pair are the ocular ridges. The caeca of the frontal limb
emerge direct from the front of the glabella, i.e. from the anterior end of the oesophagus.

The observations here presented confirm Hupe’s theory of the connexion of the genal
caeca with the oesophagus, and the function of the ocular ridges as alimentary diverti-
cula.

The Australian Cambrian trilobites, however, permit some amplification of Hupe’s
description of the cephalic alimentary apparatus of trilobites in general. Thus, the
frontal caeca extend on to the marginal border; the ocular ridge is a double ridge and its
distal extension is the double palpebral lobe; the ocular ridge represents not a single
diverticulum, but a double pair of diverticula and, consequently, not four but five pairs
of diverticula probably existed.

Furthermore, the posterior marginal border of the cephalon carries two pairs of
diverticula, and the veins of the posterior pair terminate in the form of a ‘glandular
swelling’ or opistopleural terminal swelling. The frontal caeca probably emerge not
from the oesophagus but from the anterior ocular ridge, the extension of which is the
‘parafrontal band’ in front of the glabella.

Cephalic caeca can be regarded as established in trilobites. Post-cephalic caeca in the
thorax and pygidium, which have so far escaped the scrutiny of palaeontologists, can
now be added to the model of the trilobite.

Diverticula (intestinal appendages) are regularly arranged on the pleurae of the thorax
and pygidium; each segment bears two pairs of them. In trilobites with a developed
pygidial border and progressive fusion of the pleural segments towards the rear, the
posterior portion of the pygidial pleural platform bears caeca that are similar to the

412

PALAEONTOLOGY, VOLUME 3

cephalic caeca. These new observations on the occurrence of post-cephalic alimentary
prosopon have been made on the ptychopariid Papyriaspis (thorax and pygidium);
Olenus, Red/ichia, and Centropleura (thorax only) ; and on the dolichometopid Amphoton,
and in Proceratopyge (ramified caeca on pygidium).

All that is preserved of the alimentary diverticula and caeca are their moulds on the
inner surface of the test that appear externally as low ridges and lines. As each mould
is usually more pronounced than the corresponding outer line, the observed pattern
simply reflects the external structure of the soft parts of the trilobite body. To use a term
proposed by Gill (1949), the pleural alimentary apparatus of trilobites manifests itself as
an alimentary ‘prosopon’ of the test.

The ornament of trilobites consists of terraced lines, lines, netting, granules, pits, and
puncta. This external ornament, if well enough developed or if the test is thin, may also
be reflected on the internal surface of the test. But the external ornament is, of course,
more pronounced on the external than on the internal surface and thereby differs funda-
mentally from the alimentary prosopon.

The two pleurae of Centropleura (PI. 68, figs. 3, 4) illustrate the difference. In fig. 3 the
test is preserved with its external ‘ornamental’ lines, and the opistopleural vein only is
indicated. The pleura in fig. 4 — a rubber cast of the parietal (internal) surface — shows
the opistopleural and propleural veins, whereas the external lines are only partly im-
pressed, perhaps by compaction pressure of the sediment.

The preservation of the details of the alimentary prosopon is not always clear on un-
treated fossils and may escape detection. For better results the calcareous test is washed
with ordinary soap and a nylon toothbrush.

With some rare exceptions noted by Hupe, Ordovician and younger trilobites lack the
alimentary prosopon. But, presumably, more examples will be found if diligently
searched for. The pleural alimentary apparatus of most Ordovician and younger trilo-
bites was probably ‘deep-seated’, not in touch with the parietal surface, and therefore
did not produce any cuticular sculpture.

Heavy surface ornament or a thick test usually masks the alimentary prosopon. Trilo-
bites with a thin smooth test preserved in limestone yield the best results. However, speci-
mens of a single species from the same bed differ because within the soft body of different
individuals the position of the caeca may vary in relation to the parietal surface: caeca
that are deep-seated in one individual may lie near the surface in another.

The alimentary prosopon may be preserved in shale as well. For example, steinkerns
of Redlichia, preserved in siliceous shale and chert (Opik 1958, pi. 4; pi. 6, figs. 4, 6),
show the pleural lines (intestinal appendages).

Modern methods of whitening fossils for study and photography, and appropriate
magnification of printed illustrations, facilitate the study and presentation of the details
of the alimentary apparatus. A good example is the illustrations of the olenid Acerocare
tullbergi in Henningsmoen (1957, pi. 30, fig. 9), which shows pleural lines similar to those
of Redlichia.

AGNOSTIDS

The scrobiculate test of agnostids. In many agnostids the cheeks have grooves that
radiate from near the dorsal furrows. Authors have termed such cheeks scrobiculate,
wrinkled, furrowed, grooved, rugose, or covered with ramified furrows or with ribs and

A. A. OPIK: ALIMENTARY CAECA OF AGNOSTIDS

413

furrows. Well-illustrated examples can be examined in Westergaard (1946). In some
agnostids with an apparently smooth test the grooves can be detected when whitening
is applied. Grooves are lacking in a few agnostids only.

Scrobiculation and caeca. The diagrams (text-figs. 2-4) illustrate the common scrobicu-
late pattern of the head of agnostids. This pattern is regular and is bilaterally symmetri-
cal; it has been observed in many species of several genera. It signifies the existence of a
pair of dendritic ramified glands under the test of the cheeks of the agnostid head con-
nected with the glabella (containing the stomach) by one or two pairs of short ducts
(diverticula). The glands usually surround the front of the glabella but do not meet; in
Hvpagnostus (text-fig. 3), however, the frontal glabellar lobe (stomach) is obsolete, but
appears to be replaced by a duct connecting the glands. In one form ( Corrugatagnostus ,
text-figs. 5, 6) a radiating arrangement occurs as a variation of the fundamental dendritic
pattern. As shown below, these glands are homologous with the genal caeca of other
trilobites. The caeca of agnostids are characterized by their relatively small number of
ramifications, the dendritic arrangement, and the relatively large diameter of a single
caecum (exceeding capillary size).

In a few agnostids the pleural lobes are scrobiculate in both shields. These include
several species of Pseudagnostus Jaekel — especially P. securiger (Lake) — Oidalagnostus
Westergaard, Lotagnostus Whitehouse, and Glyptagnostus Whitehouse (PI. 70). More
familiar is the combination of a scrobiculate cephalon with a smooth pygidium.

To explain the absence of scrobiculation in the pygidia of some agnostids, we note
that all agnostids should have caeca even in the absence of an external prosopon;
accepting this, it should follow that in the cephalon of some forms the caeca have a
relatively shallow position near the parietal surface and produce a cuticular sculpture,
whereas in pygidia of the same forms the caeca are mostly deep seated and produce no
external caecal prosopon.

The pygidia of such forms are convex and provide ample space to accommodate the
glands below the surface. Corrugatagnostus Kobayashi (text-figs. 5, 6), however, has a
relatively flat pygidium and therefore shows a well-developed caecal prosopon. This,
however, is not the only possible solution. For example, agnostids that have scrobiculate
pygidia ( Oidalagnostus and Pseudagnostus ; Lotagnostus and Glyptagnostus ; Corrugatag-
nostus) constitute three separate groups, each of a relatively late geological age (upper-
most Middle Cambrian, Upper Cambrian, and Ordovician). Each of these three groups
could have acquired, independently from the others, their ramified pygidial caeca as
a modification of an earlier, simpler form of glands. The earlier, Middle Cambrian,
agnostids have smooth pygidia; and therefore in the early agnostids the pygidial caeca
may have had a form that was incapable of producing an external prosopon, and need
not have been simply ‘deep seated’. In this case, the pygidia would have had lateral in-
gluvial sacs only, and no ramified glands.

Comparison with Burgessia. The idea that the scrobiculate pattern of the agnostid test is
an alimentary prosopon is illustrated by comparison of the carapace of Burgessia
Walcott (text-fig. 1) with the cephalon of Diplagnostus (text-fig. 2). As Burgessia has no
pygidium, the comparison refers only to the cephalon of agnostids and the carapace of
Burgessia.

Burgessia is an arthropod with a carapace and is definitely not an agnostid trilobite.

e e

B 0612

text-figs. 1-6. Diagrams of the alimentary apparatus of Burgessia and agnostids. 1 , Burgessici Walcott
1912. Anterior part, a simplified interpretation. After Walcott (1931). 2, Diplagnostus. Cephalon,
based on Scandinavian (Westergaard 1946, pi. 8) and Australian specimens. 3, Hypagnostus. Cephalon,
based on H. exsculptus (Angelin). Adapted from Westergaard (1946, PI. 6, fig. 1). 4, Ptychagnostus.
Cephalon, combined from several illustrations in Westergaard, and some Australian specimens. The
left half (4 a) shows two pairs of diverticula; the right half (4 b) shows one pair of diverticula and arcuate
scrobicules. This serves also as an explanation of Tomagnostus, Glyptagnostus, &c. 5, 6, Corrugatag-
uostus. Adapted from Whittard (1955). 5, Cephalon; 6, Pygidium. A — axial lobe (pygidium); B—
border with doublure; C — caeca (rugae); Cn — caecal node; D — diverticulum; Da — anterior diverti-
culum; Dp — posterior diverticulum; F — dorsal (axial) furrow; Gd— posterior lobe of glabella; Gf —
glabellar furrow (lateral indentations); Gu — anterior lobe of glabella; H — transverse glabellar furrow;
K — outline of carapace (in Burgessia) ; Mf — preglabellar median furrow (mesentery); mm — muscle
spots (in H); Ms — frontal sulcus; R — reticulation; T — stomach; Ti — intestine; V — scrobiculae

(mesenteries); Va — arcuate scrobicule.

A. A. OPIK: ALIMENTARY CAECA OF AGNOSTIDS 415

111 Burgessia, however, the alimentary apparatus itself is preserved and is not known only
from imprints on the test. The apparatus consists of a bifurcated stomach, an intestine,
and a pair of dendritic reniform glands — the ramified caeca. A bifurcate stomach is
shown (text-fig. 1) in accordance with the text in Walcott (1931, p. 18), although Wal-
cott’s diagram shows an ‘anterior central lobe of the stomach’, i.e. a trilobate stomach.
Walcott’s interpretation in the text is preferred to the alternative indicated in his illustra-
tion.

The alimentary apparatus of Diplagnostus (text-fig. 2) differs from that of Burgessia
(text-fig. 1) in the following characters:

1 . The diverticula are long in Burgessia, short in the agnostid — a minor difference.

2. In Burgessia the stomach lies in front of the diverticula and the intestine behind it, whereas, in the
agnostid, behind the diverticula lies the posterior lobe of the glabella that, presumably, also con-
tains a part of the stomach — a major difference.

3. Similar bilateral caecal glands are accommodated under the carapace of Burgessia, and under the
cephalic shield of the agnostid; but the carapace of Burgessia covers ‘five cephalic, eight thoracic,
and one abdominal segment ’, whereas in the agnostid the cephalic shield alone covers a similar
apparatus — a fundamental difference.

In my opinion the caecal glands of Burgessia and of the agnostid are homologous in
form and position, and in relation to the stomach, and are functionally identical. But
still a difference remains, because, as indicated above, in the agnostid the glands are
confined to the cephalon, and in Burgessia they have expanded beyond the cephalon.

Furthermore, the carapace of Burgessia is externally non-lobate; but internally it is
divided into an axis provided with the alimentary canal and ‘pleural’ lateral lobes pro-
vided with caecal glands; consequently, Burgessia is in this sense internally trilobate and
is comparable with the agnostid trilobite.

The similarity of the alimentary apparatus of Burgessia to the cephalic alimentary
apparatus of the agnostid suggests that both kinds of animals have retained the same
ancestral design of the alimentary apparatus, though they remain far apart in other
aspects of their organization. On the other hand, the apparent simplicity of the alimen-
tary apparatus of Burgessia and of the agnostid, and not common descent, may be the
cause of the similarity. Metabolism and digestion are primary functions and presumably
all ancestral animals had a similar simple alimentary apparatus. For such reasons the
anatomy of the digestive organs may have little meaning in the problem of a possible
phylogenetic relationship between Burgessia and agnostids.

Within the class of the trilobites, however, which includes the uncomplicated agnos-
tids and the more complicated ptychopariids, the diversity of the organization of the
alimentary apparatus is probably of some phylogenetic significance. Thus, agnostids
may have retained a primitive and, therefore, ancestral alimentary system, from which
the complicated system of other trilobites was derived. This, of course, is no indication
that agnostids are ancestors of the others.

Morphology of the caecal glands of agnostids. The alimentary prosopon of agnostids is
manifested as a wrinkling of the pleural lobes. The wrinkles consist of radiating swells
or rugae, separated by furrows or scrobiculae. The scrobiculae are narrow and sharp,
and have been regarded by taxonomists as the substantial element of the wrinkled
ornament of agnostids. But the wrinkling and its regular and recurrent radiating

416 PALAEONTOLOGY, VOLUME 3

arrangement is already a hint that some organs existed that have produced the ‘ orna-
ment’.

The rugae are channels and the scrobiculae narrow crests on the internal, parietal,
surface of the test. In turn, these parietal channels are moulds of the caeca. The crests
are partitions (parietal septa) between the caeca and probably served as carriers of mem-
branous mesenteries.

The scrobicules (parietal septa) are often discontinuous, especially at a distance away
from the margin, and produce externally a pitted pattern, whereas near the margin of the
test they are continuous and radial. The rugae also more or less radiate along the margin
and may reticulate near the centre of the cheek: thus, the caeca in the cheeks of agnos-
tids are roughly reticulate, and the cheeks may be pitted. Still another form of reticula-
tion occurs in Glyptagnostus and is described below.

The preglabellar median furrow separates the left and right caecal glands from one
another. This furrow, a crest on the inner surface, indicates a median parietal septum
and a corresponding mesentery that separates and keeps in place the caeca. In some
agnostids this furrow is externally effaced, or almost so, but it is indicated by a crest on
the internal surface. In most species of Peronopsis, and in some other forms, the pregla-
bellar median furrow is absent altogether, but this does not necessarily imply the absence
of a median mesenterial partition. The external absence of some organs may mean that
they are absent altogether, as, for example, the absence of eyes certainly means blind-
ness, but the absence of an external manifestation of a mesentery that has a mechanical
function is no indication of its total absence. In smooth agnostids (e.g. Grandagnostus)
even the external trilobation is absent, but internally such forms still remain ‘trilobed’
trilobites. In taxonomy, incidentally, the absence or presence of the median preglabellar
furrow has been regarded as a character of generic and even family significance.

The traces of one or two pairs of diverticula can be observed in the cephalon of
agnostids (text-fig. 4), and are referred to as anterior and posterior diverticula.

The best published examples of the anterior and posterior cephalic diverticula in one
specimen can be selected from Westergaard (1946) as follows: Leiopvge laevigata perru-
gata (pi. 14, fig. 2), Diplagnostus planicauda vestgothicus (pi. 8, fig. 26), Ptychagnostus
atavus (pi. 11, fig. 13), Cotalagnostus lens claudicans (pi. 6, fig. 25), and Hypagnostus
exsculptus (pi. 6, fig. 1).

The traces of diverticula are not visible in all specimens; their definition depends not
only on the preservation of the fossil, but also on their erratic external manifestation.
They occur as ‘bars’ crossing the dorsal furrows; the outer ends of these bars are the
nodes from which the caeca radiate, and the nodes themselves have a fixed position.
The bars may be strong, weak, or missing altogether, presumably because one diverticu-
lum or another may dive under the parietal crest of the dorsal furrow instead of crossing
it straight and forming a bar. The anterior diverticula always arise from the postero-
lateral flanks of the anterior glabellar lobe, and the posterior ones from immediately
in front of the anterior glabellar furrows (lateral indentations) of the posterior lobe.

The fixed position of the diverticula and their connexion with the caecal nodes, which
also maintain a fixed position, is proof that organization and not chance or accident
determines the distribution of the alimentary prosopon of the agnostids.

The glabella, from which the diverticula branch off, is assumed to have contained the
oesophagus (proventriculum, stomach), because no other space is available for its

A. A. OPIK: ALIMENTARY CAECA OF AGNOSTIDS

417

accommodation. The glabella consists of two main lobes (anterior and posterior)
separated by the transverse glabellar furrow. This furrow presumably constricted the
stomach from above, and muscles attached to the outer ends of the furrow constricted
the stomach laterally. In other words, the stomach of the agnostid probably consists of
an anterior and a posterior sac or cavity connected beneath the transverse furrow by a
constricted passage.

The posterior sac is evidently a relatively late acquisition, and is an ingluvies formed
by an enlargement of the original anterior portion of the intestine. No direct evidence
can be produced in support of this assumption, but the following reasoning is pertinent:

1. Agnostids are trilobites.

2. In the trilobite Redlichia (Opik 1958), for example, the four posterior lobes of the glabella dis-
play occipital similarity, and, therefore, in the ancestors of Redlichia , the posterior four segments
of the cephalon were probably added gradually, and existed at some stage as freely articulating
segments.

3. Features preserved in some agnostids (furrows, muscle spots) indicate that the posterior lobe of
the glabella also consists of four segments; and by analogy with other trilobites the transverse
glabellar furrow apparently belongs to the foremost of these four segments.

4. The transverse glabellar furrow is probably a vestigial segmental division, and a relic of an articu-
lation joint between two segments or two tagmata — the original head and thorax of an arthropod
ancestral to the agnostids. Hence, the anterior, frontal lobe of the glabella, to which the four
posterior lobes were added, may represent the original stomach. In some agnostids ( Dip/ag -
nostus, Tomagnostus, Glyptagnostus, and, sporadically, Ptychagnostus atavus) the frontal lobe
bears a frontal median sulcus, probably indicating a waning bilobation of the front of the oeso-
phagus that is well developed in Burgessia. The same conclusion remains in force when this
sulcus is regarded as the extension of the median frontal furrow with its mesentery.

Discussion of examples. Diplagnostus (text-fig. 2) has already been discussed and com-
pared with Burgessia, and therefore needs no further comment.

Hypagnostus (text-fig. 3) embraces species of agnostids that have a waning or obsolete
frontal lobe of the glabella. The obsolete frontal lobe probably indicates a reduction of
the original stomach, and a transfer of its functions to the posterior sac of the stomach.
This, however, cannot be assumed to be true in every case of effacement of the frontal
lobe, because a single effacement of the external relief of the test is no indication of
loss of organs. The reduction of the anterior sac of the stomach is apparent only if the
place of the frontal lobe is occupied by the scrobiculae and rugae of the caecal prosopon.
In our example, Hypagnostus exsculptus (Angelin), the frontal caeca are as strong
(prominent) as those on the flanks, and the place of the frontal lobe is completely
covered by them. But no evidence is available of the loss of the passage (duct) below
the transverse furrow, connecting the posterior sac (glabellar cavity) with the frontal
caeca.

Condydopyge, with its expanded anterior lobe, is an example of a tendency opposite
to that of Hypagnostus (see Westergaard 1946, pi. 2, fig. 1). The same trilobite, by the
way, clearly exemplifies occipital similarity.

The caecal pattern of Ptychagnostus and Tomagnostus is shown in text-fig. 4. In these
forms the distal caeca are finer and more numerous than in Hypagnostus ; diverticula
(one pair, or two pairs) are present; and a frontal sulcus occurs in Tomagnostus, and
sporadically in Ptychagnostus. A pair of arcuate scrobicules accentuates the bilateral
symmetry of the distribution of the glands.

In Corrugatagnostus (text-figs. 5, 6) the frontal lobe of the glabella is probably

418

PALAEONTOLOGY, VOLUME 3

obsolete. The frontal preglabellar furrow (the median mesentery), as illustrated, for
example, by Whittard (1955, pi. 1, fig. 12), is present. The pygidial caeca of Corrugatag-
nostus are almost a replica of the cephalic pattern, and the anterolateral pygidial caeca
are reticulate. The scrobiculae that radiate inwards from the margin are almost straight.

Glyptagnostus stolidotus sp. nov. provides an example of ramified pygidial caeca and is
described separately below.

Efficiency of the agnostid alimentary apparatus. The kind of food of agnostids is as yet
unknown, as are their cephalic and other appendages. The small size of agnostids,
their universal distribution in various sediments, and their abundance in bituminous
rocks indicate that they fed on microbios. The convex carapace and relatively large and
spacious cheeks and pleural lobes of the pygidium contained voluminous caecal glands.
These glands accommodated amounts of food, the weight of which probably exceeded
the weight of the agnostid body itself, and which should have lasted for some time. On
the other hand, it is also possible, but less probable, that the caecal glands were so large
because the nutritive value of the food was low and had to be compensated by volume.
Nevertheless, and in spite of their blindness, it is improbable that agnostids were ben-
thonic mud-eaters.

REDLICHIA AND OLENELLIDS

In Redlichia and in several olenellids, authors have noted a ‘confluence’ of the
palpebral lobes with the frontal lobe of the glabella, and a ‘confluence’ of the two
posterior glabellar lobes with the swollen fixed cheek. In Redlichia forresti ‘the cephalic
dorsal furrows are deep at their junction with the glabellar furrows and shallow in con-
tact with the glabellar lobes’ (Opik 1958, p. 26). These ‘ shallow contacts’ are bars across
the dorsal furrows and are interpreted here as sites of lateral diverticula. These bars
(diverticula) may either interrupt the dorsal furrows almost completely, or remain in-
conspicuous, or even be invisible, and this variability may occur in a single specimen.
The diverticula apparently did not remain rigidly at a definite level, but varied in depth
below the cuticle within certain limits. When the diverticula were shallow the dorsal
furrow was interrupted, whereas deeper-seated diverticula passed under the furrow with-
out forming a bar across it. The ‘confluence of lobes’ and the swollen fixed cheek, and
the two pairs of swellings in the cheeks seen in some olenellids, can be explained by
assuming that the diverticula terminated in the form of ingluvial sacs located between
the dorsal furrows and the eye. Of course, such ingluvies could have been encased in
tissue that was itself penetrated by delicate caeca; however, in Olenellus and Redlichia
no radiate or reticulate caecal prosopon is seen in the space between the eye and the
glabella.

The outer cheeks and the brim of olenellids, like those of ptychopariids, bear radiating
capillary caecal veins which have their source at the base of the eye and the front of the
glabella. The same is also shown in the diagram of the cephalon of Redlichia (text-fig. 8),
although in some species of the genus the caecal veins are not only ramified but also
reticulate.

A special variation of caecal pattern of Redlichia is the presence of a pair of facial lines
on the brim. Opik (1958, p. 31) suggests that the facial lines and the lines on the pleurae
of Redlichia may represent vestiges of similar organs. If so, they are alimentary caeca,

A. A. OPIK: ALIMENTARY CAECA OF AGNOSTIDS

419

and ‘the facial lines, consequently, may indicate the position of pleural elements within
the otherwise fused cephalon’. Because of some instability of the facial lines they were
regarded as ornament by Opik. Actually they belong to the alimentary prosopon of
Redlichia, as suggested by Hupe (1952, p. 268), who regards them as genal caeca.

Nothing as yet is known about the intestinal appendages of the thorax of olenellids,
but observations are available for the related Redlichia. According to Opik (1958, p. 31),
veins occur on each side of the pleural furrow. Each segment of the thorax has two pairs

text-fig. 7. Cephalon of an olenellid ( Paedeumias ). Adapted from Opik (1937, p. 130). DD — diverti-
cular ingluvies; Es — ocular striga (palpebral furrow); Ld — inner palpebral lobe, an ingluvies; Lu — -

outer palpebral lobe.

text-fig. 8. Redlichia. Cephalon and anterior part of thorax. Outlines adapted from Opik (1958).
BM — border (rim) of cranidium, with rostral shield (ventral); C — capillary genal caeca; Dc — facial
line (diverticular vein homologous with Dd or Du); DD — ingluvies (swelling of fixed cheek) connected
with oesophagus (glabella) by two pairs of diverticula (arrow on left side indicates one of them) ; Dd — -
opistopleural vein (intestinal appendage); Du — propleural vein (intestinal appendage); Eb — base of

eye; L — palpebral lobe; Ti — intestine.

of veins, in accordance with similar structures seen in Papyriaspis (see below). In the
diagram (text-fig. 8) these lines are interpreted as slightly ramified intestinal appendages
(caeca) issuing from the intestine.

The intestine is assumed to be annulated by analogy with the lobate oesophagus
(glabella), and because such lobation conforms with the external shape and internal
structure of the tergites ; the constrictions of the intestine also correspond to the position
of the apodemes at the junction of the somites.

PTYCHOPARIIDS ( PAPYRIASPIS )

Plate 68, figs. 1, 2; Plate 69, fig. 1 ; text-figs. 9-13

Present state of knowledge. Ptychopariids are a large group of trilobites that were spread
over the seas of the globe in the Cambrian Period. Unlike the agnostids they are multi-
segmented, and most of the species have eyes. The caecal prosopon is preserved in many

420

PALAEONTOLOGY, VOLUME 3

of these fossils. Indeed, the present knowledge of the caeca of trilobites is based on
ptychopariids. The most complete diagram of cephalic caeca is given by Hupe (1953a,
p. 83), for Ptychoparia striata Emmrich from the Bohemian Middle Cambrian. Hupe’s
schematic reconstruction of the alimentary apparatus of trilobites (1953a, p. 81) is also
based on ptychopariids and represents the view that caeca were restricted to thecephalon
and that they were present only on the cheeks (genal caeca) and the frontal brim.

Hupe correctly shows that the caeca were connected with the oesophagus by at least
(or by the maximum number of) four pairs of lateral diverticula. The ocular ridges con-
stitute the anterior pair of these diverticula.

The observations of Papyriaspis, as presented below, allow for amplifications and
modifications of this picture; an observation of special interest is that caeca are present
on the pleurae of all tagmata of the trilobite.

Geological origin of the material. The present study is illustrated by examples of Papy-
riaspis lanceola Whitehouse 1939. P. lanceola occurs in the upper levels of the Middle
Cambrian of Queensland and is associated with Ptychagnostus punctuosus (Angelin)
and P. nathorsti (Brogger). The specimens here illustrated were collected from the V-
Creek Limestone at locality M418, about 3 miles east of Morstone, in the Undilla Basin.
Particulars of the geology of the area are given in Opik (1956, 1960).

The rock (V-Creek Limestone) is a fine-grained impure (argillaceous) limestone, well
laminated and almost as fissile as shale. Many of the fossils are, therefore, partly flattened,
but minute details of the test are nevertheless preserved.

Two photographs (PI. 68, fig. 1 ; PI. 69, fig. 1) illustrate the mode of preservation of
Papyriaspis in the V-Creek Limestone. PI. 68, fig. 1, shows the cranidium and the
anterior segments of the thorax with all the details shown in text-figs. 9-12. The test of
this specimen is slightly corrugated, apparently by differential compaction of the sedi-
ment, and the duplication of the ocular ridges is somewhat obscured by a wrinkle.
The present of two pairs of these ridges, however, has been established in numerous
other specimens.

The pygidium (PI. 69, fig. 1) was selected because it was found on the same piece of
limestone as the specimen shown in PI. 68, fig. 1 ; but it is preserved without distortion,
except for flattening of the pleurae.

The cranidium illustrated in text-fig. 12 (a tracing from a photograph of a rubber cast)
was collected at V-Creek Crossing. In this figure the missing parts of the right postero-
lateral limb are reconstructed from the left.

Text-figs. 9 and 10 (pleura and pygidium) are tracings from a photograph of a com-
plete specimen from locality M418.

Pleural caeca o/Papyriaspis. The description and discussion of the alimentary prosopon

EXPLANATION OF PLATE 68

Figs. 1, 2. Papyriaspis lanceola Whitehouse, Middle Cambrian V-Creek Limestone, Loc. M418, near
Morstone, north-western Queensland. 1, Cranidium and anterior part of thorax, x4. (CPC 573.)

2, Left free cheek, X 6. (CPC 574.)

Figs. 3, 4. Centropleura sp., Middle Cambrian Devoncourt Limestone, north-western Queensland.

3, Right anterior pleura, Loc. D16, X 7. Outer surface, test preserved. (CPC 575.) 4, Left anterior
pleura, Loc. D13a, x4. Rubber cast of internal (parietal) surface, test preserved. (CPC 576.)

Palaeontology, Vol. 3.

PLATE 68

OP IK, Middle Cambrian trilobites

A. A. OPIK: ALIMENTARY CAECA OF AGNOSTIDS

421

of Papyriaspis should start with the thorax, and especially with its segments, because (1)
the segments of the thorax are the simplest units of the tergite, and (2) they display
features that have not yet been described. Moreover, the nomenclature applied to the
segments of the thorax can be used with advantage in describing the pygidium and the
cephalon.

Text-fig. 9 illustrates the structure of a
pleura of a thoracic segment of Papyriaspis,
and of ptychopariid trilobites in general.

The pleural furrow divides the pleura into
two separate lobes: an anterior propleuron
and a posterior opistopleuron. The lobes are
unequal and each has a vein (caecal vein)
with a distinctive shape. These veins are re-
garded here as the prosopon of the intestinal
appendages. The veins extend from the dor-
sal furrow almost to the inner edge of the
doublure ( Papyriaspis has a very narrow
doublure). The propleural vein arises at the
anterior end of the dorsal furrow; it is wavy

and has a ramified termination. Delicate veinlets branch off the main vein along its
posterior slope, but are not shown in the diagram.

Dd Ds Lt

text-fig. 9. Papyriaspis lanceola Whitehouse.
Pleura of ninth segment of a complete speci-
men. A — axial lobe; B — border with doublure;
C — terminal caeca; Dd — opistopleural diver-
ticular vein (intestinal appendage); Ds — sub-
sidiary caecal vein; Du — propleural diverti-
cular vein (intestinal appendage); Lt — opisto-
pleural (terminal glandular) swelling; Nf —
articulating furrow (homologous with occipital
furrow); Wf — pleural furrow.

FIG. 10 FIG. 11

text-fig. 10. Papyriaspis lanceola Whitehouse. Pygidium anchylosed with last segment of thorax.
Adapted from a complete specimen (the same as text-fig. 9). A — axial lobe; Aa — articulating half-ring;
B — border with doublure; C — ramified caeca; Cn — caecal node (ramification starts here); Dd — opisto-
pleural vein (intestinal appendage); Du — propleural vein (intestinal appendage); fi — interpleural
groove (intersegmental joint); m — margin of anchylosed segment; Nf — articulating furrow; R — •

reticulate caeca; Wf — pleural furrow.

text-fig. 11. Papyriaspis lanceola Whitehouse. Free cheek (PI. 69, fig. 1). B — border with doublure
and, probably, with a caecal trunk; C — capillary-size and ramified genal caeca; Eb — base of eye; p —
pits (interspaces between ends of caeca); Q — lateral field; q — anterior field; q/Q — dividing line (caecal
‘lineament’) between Q and q; Sd — posterior branch of facial suture (posterior edge of free cheek);

Su — anterior branch of facial suture (anterior edge of free cheek).

The opistopleural vein has two branches. The main (posterior) branch arises near the
posterior end of the dorsal furrow; it is also wavy and has fine veinlets. It terminates in

422

PALAEONTOLOGY, VOLUME 3

an elongate swelling, the terminal opistopleural glandular swelling, which has veinlets
on its outer end. Between the main branch and the pleural furrow runs the subsidiary
vein; it dies out before reaching the dorsal furrow but terminates in the opistopleural
swelling.

The opistopleural swellings are strongest on the first segment of the thorax and gradu-
ally decrease in size towards the pygidium. From about the middle of the thorax (ninth
or tenth segment) the swellings and the subsidiary vein are not discernible.

Pygidial caeca o/" Papyriaspis. In the pygidium (PI. 69, fig. 1 ; text-fig. 10) the first three
pleurae repeat the structure of the pleurae of the thorax; over the posterior half of the
pygidium, the ends of the veins branch increasingly towards the rear, and over the same
area the pleurae fuse completely and the pleural furrows and interpleural grooves dis-
appear. Thus, the veins persist in a ramified and irregularly reticulate pattern similar to
that of genal caeca as seen in the free cheek (PI. 68, fig. 9).

Around the terminus of the axis the reticulate pattern prevails, with narrow pit-like
meshes.

The veins of the pleurae of the thorax and of the pygidium are referred to as intestinal
appendages or caeca because (1) the veins of the thorax and the pygidium are homolo-
gous, and (2) the ramification of the veins on the pygidium increases towards the rear,
and the final pattern resembles that of the genal caeca. Each vein with its ramified caeca
has a comparable example in the head of the same trilobite (text-fig. 12). As the pygi-
dium has no eyes and no ocular ridges its caecal pattern should be compared with that
of blind ptychopariids, as for example E/yx latifrons Angelin (see Hupe 1953a, p. 83)
or Dasometopus breviceps (Angelin) (in Westergaard 1950).

Thus, the thoracic and pygidial veins are morphologically and topologically similar
to the genal caeca and therefore probably functioned in the same way.

Assuming, however, that the pygidial caeca are unrelated to the alimentary apparatus,
one has to assume the same for the genal caeca, because the complete similarity of the
veins and their arrangement on the head and the pygidium suggest a uniform explana-
tion for both.

In summary, alimentary diverticula, caeca, or intestinal appendages in trilobites are
present in the cephalon, in the thorax, and in the pygidium. In the thorax and in the
pygidium they are arranged segmentally, and two pairs of intestinal appendages belong
to any one segment.

This arrangement ought to be expected, because of the metamerism of the trilobite
body. An organization in which alimentary caeca issue from the oseophagus only is,
however, feasible in arthropods and Crustacea that have a carapace as, for example,
Bwgessia (text-fig. 1).

Cephalon of Papyriaspis. The complete caecal prosopon is preserved in the cephalon
of the ptychopariid Papyriaspis (PI. 68, fig. 1). The tracing of the prosopon (text-fig. 12)

EXPLANATION OF PLATE 69

Fig. 1. Papyriaspis lanceola Whitehouse, a pygidium, outer surface of the test, X6. Middle Cambrian
V-Creek Limestone; Loc. M418, near Morstone, north-western Queensland. (CPC 577.)

Fig. 2. Pygidium of a ptychopariid (new genus), outer surface of the test, X 12-5. Upper Cambrian
Georgina Limestone; Loc. W42, south of Glenormiston, western Queensland. (CPC 578.)

Palaeontology, Vol. 3.

PLATE 69

OP IK, Cambrian trilobites

A. A. OPIK: ALIMENTARY CAECA OF AGNOSTIDS

423

was made from another specimen (not illustrated). This tracing is simplified because the
numerous minute veins of the reticulate portions are too fine and dense to be presented
in a diagram of convenient size. The free cheek is seen in PI. 68, fig. 2, and its tracing in
text-fig. 11.

text-fig. 12. Papyriaspis lanceolci Whitehouse. Cranidium. Tracing from a photograph of an isolated
cranidium. Bb — brim with ramified and reticulate capillary caeca; Bf — marginal frontal furrow; BM —
rim (border) with ramified caeca and with doublure (no rostral shield in Papyriaspis\)\ Cno — postocu-
lar caecal node; Cw — ramified caeca of posterolateral limb (not crossing the suture); Dd — occipital
opistopleural vein; Ddn — occipital opistopleural diverticulum; Dg — glabellar diverticula; Ds —
subsidiary vein ; Du — occipital propleural vein ; Dun — occipital propleural diverticulum ; e — palpebral
furrow; Ed — posterior ocular ridge; Eg — parafrontal band; Es — ocular striga; Eu — anterior ocular
ridge; Gf — glabellar furrows (and muscle spots); Ld — inner palpebral lobe; Lt — opistopleural (ter-
minal glandular) swelling; Lu — outer palpebral lobe; N — occipital ring (axial lobe); Nf — occipital
furrow; Nm — occipital muscle spots; R — reticulate capillary caeca on fixed cheek; Sd — posterior
branch of facial suture; Su — anterior branch of facial suture; Sx — normal position of anterior branch

of facial suture in ptychopariids; W — posterolateral limb; Wf — posterolateral (pleural) furrow.

The following peculiarities in the distribution of caeca of the free cheek (genal caeca)
are noted;

1 . A division into frontal and lateral fields. The caeca of the frontal field cross the
suture from the brim ; these caeca arise at the anterior ocular ridge. The caeca of the
lateral field arise at the base of the eye (at the base of the visual surface); these
caeca do not cross the posterior branch of the suture.

2. All genal caeca merge into the thickened border of the cheek as if the border con-
tained a peripheral trunk. The pits along the inner margin of the border are the
interspaces between the extreme ends of the caecal veins (PI. 68, fig. 2). Similar pits
are present also in the marginal frontal furrow of the cranidium.

The cranidium (PI. 68, fig. 1 ; text-fig. 12) of Papyriaspis has a peculiar character not
seen in other known ptychopariids: the anterior branches of the facial sutures are
markedly concave, and not subparallel, convex outward, or divergent, as in other ptycho-

BM Bb Bf
1 A

W

Li

Dd Ds Du NmNfN Dun Ddn Dd Wf

424

PALAEONTOLOGY, VOLUME 3

pariids. At the same time the caecal veins of the brim are not parallel to, or limited by,
the suture, but extend across it and continue, as observed above, into the anterior field
of the free cheek. The dividing line between the anterior and posterior fields of the free
cheek indicates a ‘most probable’ position of the anterior suture in ptychopariids in
general. In Papyriaspis, however, the sutures deviate from this position.

Of course, we cannot expect the caecal veins to be strictly parallel to the sutures,
because the sutures and the caeca are different features without any common function.
We should, however, expect to find a roughly common direction of caeca and sutures
among ptychoparids, for this, mechanical, reason: the sutures are lines of weakness,
and the direction parallel to the caecal veins is mechanically weaker than that across
them. An example of almost exactly parallel suture and caecal veins is seen in Logano-
peltoides zenkeri (Billings) (Rasetti 1945, pi. 1, fig. 2).

The cranidial caeca behind the eye, along the posterior branch of the facial suture, do
not cross the suture. In other words, the posterior suture follows the caecal lineament.
So, the position of the anterior branch of the suture in ptychopariids is presumably
variable within genera and species, whereas the position of the posterior branch (and*
therefore, the shape of the posterolateral limbs) appears to be a conservative character.
This conclusion, however, needs the support of a larger number of observations.

Two different kinds of diverticula can be seen in the cranidium of the ptychopariid
Papyriaspis'. (1) the axial diverticula that are located at the dorsal furrows, and (2) those
of the ocular ridges. The axial diverticula can be discussed in two groups: (a) glabellar
diverticula located at the flanks of the glabella, and ( b ) occipital diverticula that arise at
the outer ends of the occipital lobe.

Three pairs of glabellar diverticula are present. They arise at the glabellar furrows as
ramified bundles that rapidly multiply their branches and merge into a unified reticulate
network on the fixed cheeks.

Two pairs of occipital diverticula (veins) are discernible: an anterior (‘ propleural’)
pair in front of the marginal furrows and a posterior (‘ opistopleural’) pair, forming the
crests of the posterior borders. Each of the posterior diverticula passes into a pair of
veins that have the appearance of pleural intestinal appendages.

The anterior veins extend almost to the tips of the posterolateral limbs; the offshoots
of these caeca merge into the reticulation of the fixed cheeks. They are referred to as pro-
pleural in text-fig. 12 because they occupy the same position relative to the marginal
furrow as the propleural veins of the thoracic segments relative to the pleural furrow.
All propleural veins probably represent homologous intestinal appendages.

The posterior, opistopleural, veins of the cranidium are exact replicas of the opisto-
pleural veins (including the subsidiary vein) of the anterior pleural segments of the
thorax (cf. text-figs. 9, 12). The opistopleural swellings of the cranidium are prominent
and are situated at the tips of the posterolateral limbs, which are, therefore, identified
as the tips of the occipital pleurae. Hence, occipital similarity is well represented mor-
phologically in Papyriaspis and is expressed as a complete segmental homology between
the thorax and the occipital segment.

The palpebral lobes, the ocular ridges, and the parafrontal band (a term introduced
by Hupe 1953, p. 263) constitute a single diverticular system. Hupe interpreted the
parafrontal band as the vestige of a segment (segment ‘x’), the pleurae of which are seen
in the ocular ridges. It should be noted that the parafrontal band has a median position.

A. A. OPIK: ALIMENTARY CAECA OF AGNOSTIDS 425

because of which the ‘band’ could be interpreted itself as being axial and, therefore, part
of the glabella. It is, however, an element of the brim and of the pleural lobe in general.

In Papyriaspis the parafrontal band is variable in appearance. In some specimens it is
missing, in others it is only a short bar not visibly connected with the ocular ridges; a
few specimens only show the connexion of the band with the ocular ridges. Similar
observations have been made on other Middle Cambrian ptychopariids from Australia.
The parafrontal band is as narrow as the ocular ridge with which it is connected, and
presumably functioned as a duct connecting the left and the right genal caecal glands.
It had, perhaps, the special function of maintaining, for example, the equality of pressure
in the cheeks and the eyes. The variable external appearance of the parafrontal band
indicates that the depth of the duct in the tissue varied from one individual to another.

The ocular ridges are double. Each ridge consists of an anterior and a posterior band
or ridge; both hands extend into the palpebral lobe, which is also double, and consists
of an outer and an inner lobe. Running lengthwise between the anterior and posterior
ocular ridges and the outer and inner palpebral lobes is a furrow — the ocular striga. The
striga is a furrow within the palpebral lobe, and should not be confused with the actual
palpebral furrow that separates the whole palpebral lobe from the fixed cheek.

The ocular ridges and the inner palpebral lobes are major diverticula (diverticular
trunks). From the anterior ocular ridges and the parafrontal band arise the caeca of the
brim, including the marginal frontal border, and the caeca of the anterior fields of the
cheeks. From the inner edge of the posterior ocular ridge (and its extension — the inner
palpebral lobe) caeca arise and merge with the ramification of the propleural caeca.

The edge of the palpebral lobe is defined by the facial suture; but it is assumed that
this lobe contains an ingluvies from which (from the base of the eye outside the suture)
the caeca of the free cheeks arise.

The alimentary prosopon of the cephalon of Papyriaspis as described above requires
further interpretation. Thus, the paired ocular ridges are probably homologous with the
paired caecal veins of the occipital segment and with the intestinal appendages of the
thoracic segments; if so, the ocular strigae are homologous with the pleural furrows, and
the palpebral lobes with the opistopleural terminal swelling at the tips of the pleurae.

Since the ocular strigae divide the palpebral lobe into two parts, a second explanation
is possible. It is, however, much more complicated, and therefore less probable than the
first. Each part of the palpebral lobe, perhaps, is a separate opistopleural swelling, and
the ocular ridges should therefore each be an opistopleural diverticulum; if so, the inter-
vening striga should be a segmental joint replacing the obsolete propleuron of the
posterior segment. It should, however, be noted that, so far as is known, even three pairs
of ocular ridges may represent a single segment, because the opistopleuron contains two
veins (including the subsidiary vein).

The visual faces of the eyes have, of course, no inherent connexion with the alimentary
apparatus except for the support given by the ingluvial buttresses, which are located
in the cavities of the palpebral lobes. The palpebral lobes are surrounded by radiating
caeca, and each lobe being a source of caeca certainly contained a diverticular ingluvies.

Weighing the probable explanations, preference should be given to the simpler alterna-
tive. It is, therefore, fair to conclude that each eye is situated at a pleural tip, and its
position corresponds to that of the opistopleural terminal swelling.

If the classification of eyes by Raw (1956; 1957) is accepted, the eyes of Papyriaspis

426

PALAEONTOLOGY, VOLUME 3

should be classed as segmental. The opistopleural terminal swellings may indicate
vestiges of completely obsolete eyes that ‘were apparently in the pro-Arthropod borne
on the pleura’ (1956, p. 679).

The swellings on the posterolateral limbs of Papyriaspis and on the pleural tips of its
thorax are blind, but the blindness by itself justifies no conclusion that eyes are Tost’ or
that such eyes may not have existed at all in the ancestors. It is, nevertheless, safe to
assume that the opistopleural swellings are ingluvial expansions of intestinal append-
ages similar to the palpebral lobes, and that the swellings and the lobes have a homo-
logous position; but it is just as speculative to assume the loss of corresponding eyes as
to say that these swellings were about to develop visual surfaces of their own.

THE PYGIDIUM OF A PTYCHOPARIID
Plate 69, fig. 2; text-fig. 14

The illustrated pygidium belongs to an undescribed genus from the early Upper
Cambrian of Queensland. The ramified pattern of the anterior pleural veins (intestinal
appendages) changes into a reticulate pattern at the rear of the pygidium. Behind the
axial terminus strong median ramified caeca can be seen. The doublure is narrow and
the axis short, as if its actual terminus is concealed under the test. Behind the terminus
the test is swollen and forms a shallow boss.

It is improbable that the median terminal caeca arise from the anal segment, and it is,
therefore, reasonable to assume that the axis actually extended below the dorsal test into
the ‘boss’ as a testless terminus which may have contained the end segment.

PLEURA OF CENTROPLEURA
Plate 68, figs. 3, 4; text-fig. 13

One of the illustrated pleurae (fig. 4) belongs to the foremost segment of a Centro-
pleura from the Middle Cambrian Devoncourt Limestone of Queensland and is ex-
plained in text-fig. 13. The pleura is narrow and relatively long and shows strong

text-fig. 13. Centropleura. Pleura (PI. 68, fig. 3). B — border with doublure; Dd — opistopleural vein
(intestinal appendage); Du — propleural vein (intestinal appendage); Pd — opistopleuron ; Pu — pro-

pleuron ; Wf — pleural furrow.

propleural and opistopleural veins. The specimen is a rubber cast of the internal face
of the test on which the outer ornamental lines are only partly reflected. Externally
(as seen in the other pleura — PI. 68, fig. 3) the ornamental lines are strong and the opisto-
pleural vein only is slightly indicated. The pleura of Centropleura has a remote and
superficial similarity to a truncated insect wing.

A. A. OPIK: ALIMENTARY CAECA OF AGNOSTIDS

427

CONCLUSIONS

1. The alimentary prosopon of agnostids, OJenellus, and Redlichia , ptychopariids,
and Centropleura reflects the pleural alimentary apparatus. The surveyed data belong
to various and representative families, and the results should be valid for agnostids and
for Cambrian opisthoparian trilobites. Post-Cambrian Opisthoparia probably differ
little from the Cambrian Opisthoparia in principle. This inference is based on the general
morphological similarity of Cambrian and post-Cambrian trilobites, and on phylogenetic
considerations. No observations are available for Cambrian eodiscids and post-Cam-
brian proparian trilobites.

2. Agnostids, Redlichia, and OJenellus,
and ptychopariids represent three different
groups with respect to the form of their
caecal glands.

Agnostids have dendritic caecal glands
with relatively few ramifications; the dia-
meters of single caeca exceed capillary size ;
their diverticula are short ducts. Caeca of
capillary size have not been observed in
agnostids.

Redlichia and OJenellus have capillary
ramified radiating and reticulate caeca on
the cheeks (including the brim), and
probably two to four pairs of sac-like in-
gluvies in the fixed cheeks and palpebral
lobes.

Ptychopariids have capillary caeca that
branch to form a reticulate pattern. The
cephalic diverticula of ptychopariids con-
sist of ramified bundles of caecal veins.

3. Alimentary appendages (diverticula and caeca) are not confined to the cephalon.
They are present in each segment of the thorax and pygidium in the form of intestinal
appendages. Their distribution and arrangement are segmental.

4. In Redlichia, Centropleura, ptychopariids, and olenids, each segment of the thorax
bears two pairs of lateral intestinal appendages, one pair in front of and another behind
the pleural furrows.

5. In ptychopariids the paired intestinal appendages of the posterior, fused, portion
of the pygidium develop a type of reticulation similar to that of the genal caeca.

6. In the ptychopariid Papyriaspis, swellings at the pleural tips of the anterior thoracic
segments and the tips of the posterolateral limbs connect with the opistopleural caecal
vein. These swellings probably functioned as glands.

7. The ocular ridges of Papyriaspis consist of two parallel ridges each of which passes
into the palpebral lobes; the ocular ridges are probably homologous with the pleural
intestinal appendages and constitute a single segment ; the palpebral lobes and the eyes
are situated at the tips of the segment represented by the ocular ridges. The cephalic
eyes are, consequently, segmental eyes.

text- fig. 14. Pygidium of a new genus of ptycho-
pariids. Semi-diagrammatic tracing of the speci-
men figured on PI. 69, fig. 2. Aa — articulating
half-ring; B — border with doublure; fi — inter-
pleural groove; Nf — articulating furrow; R — reti-
culate caeca (‘genal caeca’); Wf — pleural furrow.

428

PALAEONTOLOGY, VOLUME 3

8. The palpebral lobes (with the cephalic eyes) and the opistopleural glandular swell-
ings have homologous positions within their segments; the opistopleural swellings,
therefore, may represent vestiges of pleural eyes ; or the palpebral lobes (excluding the
eyes) and the opistopleural swellings may be homologous, or may contain homologous
organs. Any of these alternatives is reasonable because the cephalic eyes and the palpe-
bral lobes to which the eyes adhere are different and unrelated organs and may have
followed independent trends in their history.

9. The caecal prosopon occurs only on the pleural lobes of the trilobite and is not
observed on the axial lobe; ramification and reticulation of caeca develop concurrently
with the fusion of pleurae. Consequently, the frontal border, consisting of brim, rim,
and doublure, comprises fused pleurae. This conclusion is supported by the observation
that in all trilobites the sutures are confined to the pleural lobe (Opik 1958, p. 29). More-
over, ‘all axial parts of the tergum are surrounded by pleurae’ (loc. cit.) ; the pleural
doublure is continuous and the frontal doublure together with the rostral shield is a
part of it.

10. The pleurae that are contained in the frontal border and the free cheeks probably
belong to the segments that are indicated in the glabella; there is no evidence to indicate
that the frontal border contains any other segments. These pleurae intervene between the
axial lobe (glabella) and the ventral hypostoma (see Opik 1958, p. 30).

DESCRIPTION OF G LYPTAG NO S TUS
Glyptagnostus Whitehouse 1936

Plate 70; text-figs. 15, 16

Species of Glyptagnostus. The type species is G. toreuma Whitehouse 1936. According
to Westergaard (1947, p. 5) G. toreuma is synonymous with G. reticuJatus (Angelin 1851),
and so is G. angelini Resser 1938. Kobayashi (1949) supports this interpretation. The
synonymy is subjective and reflects the present state of knowledge of these forms.
G. reticulatus nodulosus Westergaard 1947 completes the specific list of the genus. A new
species— Glyptagnostus stolidotus— is described below.

Concept of the genus. The reticulate or blistered surface of the test has been regarded as
the main diagnostic character of the genus. The distribution, size, and number of the
blisters are highly variable and provide no specific criteria. An exception is G. reticulatus
nodulosus Westergaard, a subspecies with an ‘over-blistered’ cephalon and an ‘under-
blistered’ pygidium, in which the radiating and furcating rugae have preserved their
continuity.

The inclusion of the non-reticulate new species stolidotus in Glyptagnostus necessitates
a modification of the generic concept and a re-evaluation of the generic significance of
the ‘blisters’.

The reticulation (blistering) is created by scrobicules crossing the rugae. Most of these
cross-scrobicules are shallow and many are incomplete. In such cases (see, for example,
Westergaard 1947, pi. 1, fig. 8) the rugae are laterally notched. In the new G. stolidotus
laterally notched rugae also occur, but no complete blisters are present. It appears,
therefore, that blistering or reticulation is the specific character of G. reticulatus and its

A. A. OPIK: ALIMENTARY CAECA OF AGNOSTIDS

429

subspecies, but it cannot be regarded as a generic character. Nevertheless, the genus
Glyptagnostus is not invalidated by such an interpretation.

The main diagnostic character of Glyptagnostus is the presence of rugae (and scrobi-
cules) on both cephalon and pygidium. In the pygidium six or seven pairs of first-order
rugae are present; they fork to form about twice as many second-order rugae, which in
larger specimens themselves fork and produce short rugae (terminal caeca) of the third
order. In the cephalon about ten pairs (nine to eleven) of first-order caeca (rugae) are
present, and their number multiplies by forking into rugae of the second and third orders.
A similar number of caeca (rugae) is seen in Diplagnostus (text-fig. 2), but the pattern
in Glyptagnostus is that of a Ptychagnostus with a pair of arcuate scrobicules (text-fig. 4b).

The structure of the glabella and the pygidial axis (text-fig. 15) supplements the
generic diagnosis of Glyptagnostus. The glabella is distinct in having a sub-pentagonal
flattened anterior lobe with a bifurcate sulcus, and two pairs of lateral lobes including
the basal lobe. Among agnostids with scrobiculate cephala Lotagnostus Whitehouse
(1936) alone has a similar glabella.

The structure of the median sulcus in Glyptagnostus is unique, but is rarely well pre-
served. It consists of a deep notch in the middle of the frontal lobe and an oval depression
which may even reach the transverse furrow. It is illustrated, for example, by Wester-
gaard (1947, pi. 1, fig. 5).

The pygidial axis is trifid, with a median train of three unequal ‘ bulbs ’, the anterior
of which carries the blunt axial spine with a superimposed small node. The median
train of bulbs is flanked by a pair of depressed areas bounded by shallow, incised dorsal
furrows. In these depressions, in the best-preserved specimens, seven pairs of pits (muscle
spots, apodemes) can be observed (PL 70, fig. 9). They are also illustrated by Wester-
gaard (1947, pi. 1), and so is the small median terminal node on the second axial ‘bulb’.

The term ‘bulb’ (or ‘axial bulb’) indicates that these features do not correspond
exactly to, and are therefore distinct from, segmental axial lobes.

The pygidial axis is also quadrilobate, with three discontinuous transverse furrows.
The three anterior axial lobes are of about equal size and occupy the anterior half of the
axis. The fourth lobe forms the posterior half of the axis, carries the two posterior ‘ bulbs
and is pointed at the rear. The articulating furrow is wide and deep, the articulating
half-ring narrow and elevated.

As regards a similarity with Lotagnostus more comment is needed. The species referred
to are: Lotagnostus trisectus (Salter) (Westergaard 1922, pi. 1, fig. 11), L. trisectus mut.
ponepunctus (Matthew) (1903, pi. 17, fig. 8), Agnostus americanus Billings (Rasetti 1944,
pi. 36, figs. 1, 2), Lotagnostus asiaticus Troedsson 1937, and Lotagnostus obscurus Palmer
1955.

In all these species the pygidial articulating furrow is narrow (normal); the axis is
tripartite longitudinally and trilobate transversally and laterally not depressed, but
swollen ; and the number of first-order rugae (where they are discernible) is visibly greater
than in Glyptagnostus. Moreover, only one pair of lateral lobes (the basal lobes) is pre-
sent in the glabella of Lotagnostus.

Diagnosis. Cephalon and pygidium rugose (scrobiculate); about ten pairs of cephalic
and six or seven pairs of pygidial first-order caecal rugae ; arcuate scrobicules in cephalon ;
two pairs of glabellar lateral lobes including basal lobes; pygidial axis trifid with a

F f

B 6612

430

PALAEONTOLOGY, VOLUME 3

median train of three unequal ‘bulbs’, the anterior ‘bulb’ with axial spine (node);
lateral axial depressions with about seven pairs of pits (muscle spots); three anterior
axial lobes, three discontinuous transverse furrows, and a large pointed terminal lobe;
articulating furrow deep and wide. The type species has reticulate rugae.

Stratigraphic distribution in Queensland. G. stolidotus is the oldest known species. It
occurs in beds which can be correlated with the upper half of the Scandinavian Upper
Cambrian zone with Agnostus pisiformis. G. stolidotus has been found in numerous out-
crops of the Georgina Limestone (lower levels), the Pomegranate Limestone (lower
horizon on Wills Creek), and the O’Hara Shale (Tower O’Hara fauna’).

G. reticulatus follows stolidotus immediately in time. Apparently the life-spans of the
two species overlap for a short interval (represented by a few feet of sediment only).
G. reticulatus occurs in Queensland in two separate successive horizons.

In the lower horizon, immediately above stolidotus, G. reticulatus is associated with
Olenus at Pomegranate Creek, and with some of the latest genera of the Tower O’Hara
fauna’ at Wills Creek. The complete specimens (PI. 70, figs. 10, 11) came from this
locality. In the upper horizon, which is known from many localities of the Georgina
Limestone, G. reticulatus is associated with species of Pseudagnostus, Proceratopyge,
and Eugonocare, as described by Whitehouse (1936; 1939). This horizon corresponds
to the ‘Glyptagnostus stage’ of Whitehouse, from which G. toreuma was described. These
two horizons with Glyptagnostus can be correlated with the three lowermost subzones
of the Scandinavian Olenus zone, which also contain G. reticulatus (Westergaard 1947).
Explanations of the geographical and geological names used above are found in Opik
(1956; 1960).

Phylogeny. The early specimens of G. reticulatus and its late representatives described
as G. toreuma are very similar, and the derivation of toreuma is in no need of discussion.
G. ‘ toreuma ’ is distinguished by the abundance of very large specimens in its population.

G. stolidotus antedates reticulatus in every locality and may be interpreted as the
ancestor of reticulatus. If so, Queensland should be the centre from which Glyptagnostus
spread over the Cambrian seas of the world — a conclusion that is not supported by
other evidence. It is evident, also, that G. reticulatus and stolidotus are different in the
structure of the pygidial axis and are quite distinct species without intermediate links in
Queensland. It should be assumed, therefore, that stolidotus is not itself the immediate
ancestor of reticulatus, but is still very close to the unknown actual progenitor of
reticulatus.

The origin of G. stolidotus itself is not evident. The structure of its glabella, pygidial
axis, and cephalic caeca indicates relationship with the Middle Cambrian Ptychagnostus ,
but none of the known species of the latter is close enough to be considered a direct
ancestor of Glyptagnostus.

Glyptagnostus reticulatus (Angelin)

Plate 70, figs. 9-11; text-fig. 15

The illustrated specimens come from two different localities and horizons. The speci-
men illustrated on Plate 70, fig. 9 (CPC 587) is from the Georgina Limestone (Locality
W 1 57), from the highest level of the occurrence of Glyptagnostus reticulatus, described as

A. A. OPIK: ALIMENTARY CAECA OF AGNOST1 DS

431

G. toreuma Whitehouse. The pygidium illustrated here is identical with the holotype
of toreuma, which is also a pygidium. It is 6-3 mm. long (without the articulating half-
ring) and 6-5 mm. wide. It shows the small node on the axial spine and the terminal node
on the middle axial bulb. Notable are the elevated rugae of the main caecal trunks (for
explanations of terms, see text-fig. 16) flanking the axis, similar to the trunks of stolidotus
(PI. 70, fig. 3).

The two complete specimens (PI. 70, figs. 10, 11) are from the same limestone bedding
plane from Wills Creek (Loc. D126) and represent the lower horizon of the beds with

nl Nf Aa bl

b3 VM V

fig. 16

text-fig. 15. Glyptagnostus reticulatus (Angelin). Diagrams snowing the structure of glabella and
pygidial axis. The reticulation is omitted (see PI. 70, figs. 9-11) to facilitate comparison of axial struc-
tures of G. reticulatus and G. stolidotus (text-fig. 16). The pygidium represents a specimen from Pome-
granate Creek with exceptionally well-preserved muscle spots (pits). Cephalon : ba 1 — basal lobe ; ba2 —
lateral glabellar lobe; Gd — posterior lobe of glabella; Gu — anterior lobe of glabella; Ms — frontal
sulcus; Msu — depressed extension of frontal sulcus. Pygidium: Aa — articulating half-ring; bl —
anterior bulb; b2 — middle bulb; b3 — posterior bulb; F — dorsal (axial) furrow; mm — muscle spots
(pits); nl — node on axial spine; n2 — terminal axial node; Nf — articulating furrow.

text-fig. 16. Glyptagnostus stolidotus sp. nov. Diagram of pygidium. Aa — articulating half-ring;
bl — anterior bulb ; b2 — middle bulb; b3 — posterior bulb; Cl — first-order caecum ; C2 — second-order
caecum ; C3 — third-order caecum ; Ct — main caecal trunk ; D — diverticulum ; F — dorsal (axial) furrow ;
mm — muscle spots (five posterior pairs — ‘pits’; two anterior pairs — ‘spots’); nl — tip of axial spine
(node); n2 — terminal axial node; Nf — articulating furrow (with a pair of muscle pits or appendiferi) ;

Y — lateral scrobicula; VM — median scrobicula (postaxial median furrow).

Glyptagnostus reticulatus. The larger specimen (CPC 588), 1T4 mm. long and 54 mm.
wide, has about the same number of blisters on the cephalon and pygidium, whereas in
the smaller specimen (CPC 589), 84 mm. long and 3-9 mm. wide, the pygidium has
visibly fewer blisters than the cephalon. This is an example of the variability of the
reticulation already indicated in the discussion of the genus.

Diagnosis. Test rugose, rugae reticulate (blistered); outline of cephalon subelliptical,
node on posterior glabellar lobe, on its anterior half, elongate and relatively prominent;
anterior half of pygidial axis divided by discontinuous transverse furrows into three
lobes; anterior lobe of pygidial axis tripartite, low node superimposed on axial spine;

432

PALAEONTOLOGY, VOLUME 3

posterior bulb of axis relatively long (about one-quarter of axial length), middle bulb
narrow (about one-fifth of axial length); seven pairs of muscle spots (pits of even depth)
in axis.

Glyptagnostus stolidotus sp. nov.

Plate 70, figs. 1-8; text-fig. 16

Diagnosis. Rugose, without blisters (except for occasional blisters on cheeks near the
basal lobes); outline of cephalon subcircular, small rounded central node on posterior
glabellar lobe ; two anterior axial lobes on pygidium defined by discontinuous transverse
furrows, a third lobe occasionally indicated; anterior lobe of pygidial axis undivided, or
tripartition indicated occasionally by vestigial furrows; posterior bulb of pygidial axis
small (one-eighth of axial length), middle bulb wide (one-third of length of axis); five
posterior pairs of muscle spots on pygidium as deep pits, two anterior pairs occa-
sionally discernible as ‘spots’ only.

Holotype. The cephalon, CPC 584 (PI. 70, fig. 6), from Loc. B525 at Wills Creek,
Boulia area. A cephalon is selected as the holotype to facilitate comparison with the
lectotype of Glyptagnostus reticulatus (Angelin) in Westergaard (1947, pi. 1, fig. 2),
which is also a cephalon.

Comment on illustrated material. All the illustrated specimens have been collected from
a single bed, which contains only one kind of cephala and of pygidia; it is therefore
certain that the shields are conspecific, and that the material in hand represents a popu-
lation within narrow limits. The collecting points are recorded as B525 and B537, which
are about a quarter of a mile apart. Twenty-five cephala and fifteen pygidia were selected
for examination.

The small pygidium (PI. 70, fig. 1 ; CPC 579) is 4-8 mm. long and 5-0 mm. wide. Third-
order caeca are present; they are uncommon in such small specimens.

The fragmentary pygidium (PI. 70, fig. 2; CPC 580) is 5-9 mm. long. The diverticula
are separated from the axial lobe by a shallow furrow. Third-order caeca are numerous.
The posterior bulb is small and is connected with the left gland.

Another fragmentary pygidium (PI. 70, fig. 3; CPC 581), the largest found, is 6T mm.
long. It is photographed with a tilt to the right to show the gland in a horizontal plane.
Third-order caeca are numerous; the posterior bulb is visibly part of the left gland.
The edges of the anterior caeca are serrate. The main caecal trunks flank the axis as
prominent ridges, as is also indicated in the pygidium of G. reticulatus (PI. 70, fig. 9).
On the end of the middle bulb the tiny terminal node is preserved, just as in fig. 9.

EXPLANATION OF PLATE 70
All specimens x 6

Figs. 1-8. Glyptagnostus stolidotus sp. nov., Upper Cambrian Pomegranate Limestone, lower horizon
at Wills Creek, Boulia area, western Queensland. Fig. 6 (holotype cranidium) from Loc. B525
(CPC 584); all other specimens Loc. B537. 1, CPC 579; 2, CPC 580; 3, CPC 581 ; 4, CPC 582; 5,
CPC 583; 7, CPC 585; 8, CPC 586.

Fig. 9. Glyptagnostus reticulatus (Angelin), pygidium, Upper Cambrian Georgina Limestone; Loc.
W157, south of Glenormiston, western Queensland. (CPC 587.)

Figs. 10, 11. Glyptagnostus reticulatus (Angelin), Upper Cambrian Pomegranate Limestone, upper
horizon at Wills Creek, western Queensland. Loc. D126. 10, CPC 586; 11, CPC 589.

Palaeontology, Vol. 3.

PLATE 70

OPIK, Upper Cambrian trilobites

A. A. OPIK: ALIMENTARY CAECA OF AGNOSTIDS 433

The small cephalon (PI. 70, fig. 4; CPC 582) is 4-0 mm. long and 4T mm. wide; its
glabella appears slender; incipient reticulation of the cheeks is seen near the left basal
lobe, and the basal lobes are themselves transversely divided by a shallow furrow.

The small cephalon (PI. 70, fig. 5; CPC 583) is 3-9 mm. long and 4-1 mm. wide. No
reticulation is present.

The holotype cephalon (fig. 6) is 5-7 mm. long and 6-0 mm. wide. Incipient reticulation
is present at the dorsal furrows; the basal lobes are divided; and third-order caeca are
developed. The central glabellar node is preserved and the sulcus is indicated.

Another cephalon (PI. 70, fig. 7; CPC P 585), 6-5 mm. long, shows the divided basal
lobe mentioned above. It appears that the cheek bears a major caecal duct from which
first-order caeca branch off to left and right.

The large cephalon (PI. 70, fig. 8; CPC 586), 6-3 mm. long and 6-5 mm. wide, is
flattened. It seems that each cheek has two main caecal ducts, one directed towards the
posterior lateral corners, the other surrounding the anterior glabellar lobe. A strong
and long scrobicule at the level of the transverse glabellar furrow seems to divide each
gland into an anterior and posterior portion.

Derivation of the name. The specific name stolidotus means ‘wearing a garment with
many folds’, as distinct from reticidatus, ‘covered with a network’.

Depository. Commonwealth Palaeontological Collection (CPC), Bureau of Mineral
Resources, Geology and Geophysics, Canberra.

Occurrence and age. All the illustrated specimens have been collected from a bed in the
Pomegranate Limestone, at Wills Creek, Boulia area, north-western Queensland. The
age is Upper Cambrian, upper half of the zone of Agnostus pisifonnis (by correlation).

Alimentary apparatus. The caecal pattern of the cephalon corresponds to text-fig. 4b,
a Ptychagnostus with arcuate scrobicules, but without discontinuous scrobicules and,
therefore, without pitted cheeks.

Notable is the connexion of the posterior (pygidial) axial bulb with the caeca and
always with the left gland; consequently the left gland is the larger. Moreover, the
posterior bulb is not an original part of the axial lobe but a caecum in the median posi-
tion. In very small pygidia (1-0-1 -3 mm. long) no posterior bulb is present, which in-
dicates that it is a late acquisition. These pygidia have a Ptychagnostus- like axis and only
five pairs of lateral caeca, which can be mistaken for segmental ribs.

The terminus of the axial lobe is the middle bulb; it bears the terminal node, perhaps
a rudiment of a telsonic spine. This middle bulb (b2, text-fig. 16) most probably con-
tained the anus. The diverticula (D, text-fig. 16) arise from the third axial lobe just in
front of the fifth pair of muscles, in front of, and at a distance from, the anus.

Little difference exists in the arrangement of caeca in the cephalic and pygidial glands.
The similarity is apparent when, for example, the pygidium (PI. 70, fig. 2) is turned rear
to front and compared with the cephalon (PI. 70, fig. 8). In this position the shields
appear almost as mirror images, and even the glabella and the pygidial axis reveal a
striking similarity.

434

PALAEONTOLOGY, VOLUME 3
COMMENTS AND EXPLANATORY NOTES
The comments and notes are arranged alphabetically. Terms in current usage, some
of which are elementary, are applied in the present paper in a restricted sense that may
be different from their general or vernacular usage, and new terms are introduced to
designate as yet unnamed morphological features of trilobites. Hence, the collection of
definitions here presented may facilitate the understanding of the text, because the
meaning of the terms is not always immediately apparent from the context.

Some of the definitions are accompanied by comments that refer incidentally to the
topic of the paper or are excursions into related problems.

Agnostids. The systematic position of agnostids within the trilobites is a subject of contention. Agnos-
tids are regarded as a subclass or a separate order and suborder of the class of trilobites. It seems, how-
ever, that the gap between the conventional trilobites and the agnostids is bridged by Pagetia and the
eodiscids. The distinctive character of the caecal glands of agnostids may indicate a separate super-
family or order; the alimentary apparatus of the intermediate eodiscids is, however, unknown, and
the distinction between agnostids and ptychopariids may therefore be only apparent.

Alimentary apparatus {or system). All organs contributing to the food supply, including all lateral
appendages of the intestine and oesophagus. In trilobites it was perhaps a ‘pressurized plumbing’,
because the caeca had no outlets. The pressure in the caeca may have contributed to a mechanical
rigidity of the pleural lobe of the butter-crab.

Alimentary canal. The continuous median alimentary trunk from mouth to anus without diverticula or
other appendages; in trilobites it was located in the axial lobe.

Alimentary prosopon. A morphological term for the cuticular sculpture of veins and swellings seen in
the test of a trilobite, and assumed to be the external expression of alimentary organs (see also Orna-
ment, Prosopon).

Articulating furrow. Each segment of the thorax has in the front of its axial lobe a transverse furrow
(the ‘articulating furrow’) that separates the lobe from its ‘articulating half-ring’. At the outer ends of
the articulating furrow are the appendiferi supporting the legs. Consequently, the articulating furrow
and its homologues in the cephalon and pygidium mark the frontal portions of each segment.

Base of the eye. See Eye.

Border with doublure. A phrase to emphasize the fact that in all trilobites the outer margin of the test
has a doublure, i.e. a marginal duplication of the test with a hairpin section. The test has no free edge.

Brim. The meaning is evident from text-fig. 12.

Bulb. ‘Bulbs’ are divisions of the pygidial axis of Glyptagnostus that do not coincide with the original
segmentation. The anterior bulb (bl in text-fig. 15) contains two anterior axial lobes and extends partly
into the third lobe; the middle bulb (b2) is the composite axial terminus proper; the posterior bulb
(b3) in adult specimens appears morphologically as a part of the pygidial axis, but morphogenetically it
is a secondary expansion of the caeca of the left pleural lobe. The posterior bulb is flanked by a pair or
even two pairs of segmentally arranged muscle-spots indicating the original extent of the axial terminus
before its reduction and transformation into the middle bulb (b2).

Caecum {-a). Ramified lateral extensions of the alimentary canal. Caeca have no outlets (they are blind),
and their function was glandular (hepatic caeca), providing for digestion of food delivered from the
alimentary canal by way of the diverticula (see also Genal caeca, Pygidial caeca).

Caecal node. A point at which a diverticulum or a major caecum branches.

Diverticulum (-a). Bilaterally arranged outlets (ducts) leading from the alimentary canal into the caeca.

Dorsal (axial) furrow’. The furrow (and any part of that furrow) that completely surrounds the axial
lobe of a trilobite and separates it from the pleural lobe or lobes.

A. A. OPIK: ALIMENTARY CAECA OF AGNOSTIDS

435

Duct. See Diverticulum.

Eye. The eye, or the ‘visual surface’, belongs to the free cheek and is located on the flank of the palpe-
bral lobe; below the visual surface the test of the free cheek has a vertical band — the base of the eye.

Facial line. Facial lines occur not only in Redlichia (Opik 1958), but also in several specimens of Para-
doxides illustrated by Westergaard (1936; pi. 2, fig. 6; pi. 3, fig. 2; pi. 5, fig. 13; pi. 6, fig. 6; pi. 7,
figs. 3, 5, 9).

Facial sutures. Divisional lines in the cephalic test of trilobites that open for moulting. The position of
the sutures is constant for a species and may vary within certain limits in a genus. No evidence is avail-
able to prove that facial sutures correspond to, or may have developed from, intersegmental joints.

Fusion of pleurae. This expression refers to a morphological fact that has been successfully exploited in
taxonomy. Fusion of pleurae is part of the process of fusion of pygidial segments. In the first stage of
fusion the interpleural grooves disappear and each opistopleuron is fused with the next following pro-
pleuron to form a pleural rib; the ribs are divided by pleural furrows, which are oblique to the original
segmentation defined by the interpleural grooves. According to the hypothesis of Stormer (1941),
the segmentation of the trilobite is already a secondary one and the pleural furrows are the vestiges of
the primary segmental boundaries. The fact of fusion of the pygidial pleurae may mean, therefore, the
restitution of a status quo, and this status quo refers to a condition already completely lost by the
ancestors of trilobites.

Hypostoma. In Papyriaspis the hypostoma apparently did not touch the doublure. In Pagetia (Opik
1952) a wide space remains open between the anterior edge of the hypostoma and the doublure, and no
evidence exists as yet for a postulated connective ‘wide rostral plate’. Pagetia, however, has a pregla-
bellar median furrow (a parietal mesenterial septum in front of the hypostoma). According to Palmer
(1957) the hypostoma in olenellids comes into contact with the rostral shield gradually, beginning with
a median connective stalk that has the position of a preglabellar median furrow. All these observations
are in support of the idea of an original discontinuity between the hypostoma and the glabella (Opik
1958).

Genal caeca. Caeca located in the cheeks and brim of trilobites and assumed to be the only ones that have
existed.

Glabella. The axial lobe of the cephalon without the occipital ring. Many authors include the occipital
ring.

Glandular swelling. See opistopleural swelling.

Ingluvies. Expansion of the alimentary canal. In trilobites they may be bilaterally arranged sacs or an
axial boss.

Interpleural groove. A term in current usage to denote the dividing lines or furrows in the pygidium that
represent the fused but still visible intersegmental joints.

Intestinal appendages. Diverticula with their ramified caeca. It is, however, difficult to apply the terms
diverticula and caeca to the pleural veins because they are linear and not differentiated clearly into duct
and caeca. For this reason the general term ‘intestinal appendages’ is applied.

Intestine. The gut behind the oesophagus, but not including the lateral appendages.

Mesentery. A membrane attached to the inside (parietal) surface of the test and serving as a partition
between, and a support for, organs.

Occipital furrow. The occipital furrow is part of the occipital segment; it is a homologue of the articu-
lating furrows of the segments of the thorax.

Occipital ring. The axial lobe of the occipital (posterior) segment of the cephalon. ‘Neck ring’ is also
in use. It is morphologically similar to, and homologous with, all other axial lobes of the trilobite.

Occipital similarity. Refers to the similarity of the occipital segment of the cephalon to the anterior (or

436

PALAEONTOLOGY, VOLUME 3

any other) thoracic segment. This similarity is also recognizable between the glabellar lobes and the
axial lobes of the thorax (Opik 1958).

Ocular ridge. Also called the ‘eye line’; it is a low narrow elevated band that extends in many trilobites
from the anterior end of each palpebral lobe towards the frontal lobe of the glabella. In some trilobites
a double ocular ridge has been observed. The absence of the ridge does not necessarily indicate that
the corresponding organs are obsolete.

Ocular striga. In Papyriaspis the faint furrow that divides the ocular ridge into posterior and anterior
ridges and extends into the palpebral lobe.

Oesophagus. The anterior expanded part of the alimentary canal, located in the cephalon in the space
between the glabella and the hypostoma: proventriculum and stomach.

Olenellids. Resser (1928) has illustrated several specimens of olenellid trilobites with facial sutures.
They are, of course, vestigial and remained fused during moulting. One of Resser’s specimens (1928,
pi. 1, fig. 4) exhibits the vestige of the posterior branch of the suture, the course of which is similar to
that of Papyriaspis. It may not be a suture at all, but a caecal vein similar to the caeca of the postero-
lateral limb of Papyriaspis (Cw in text-fig. 12). These veins (and caeca) may have preserved their posi-
tion even after the complete disappearance of the suture in Olenellus.

Opistopleural terminal glandular swelling. This phrase can be abbreviated to ‘ opistopleural swelling'.
It has been observed as yet in Papyriaspis only. It is interpreted as part of the alimentary apparatus
because of its connexion with the intestine via the opistopleural vein, and because of the absence of an
external opening. The resolving power of the alimentary prosopon has, however, its limits, and there-
fore other interpretations are possible, as for example segmental excreting glands and ‘poison-glands’.
Such glands should have external openings and should have no ducts connecting with the intestine.

Opistopleuron. That part of the pleura posterior to the pleural furrow. It is not a whole pleura but a part
of it.

Ornament. Surface features, markings, in the test of trilobite that cannot be interpreted as representing
or containing definite organs. Some of them may have a mechanical function (see also Terraced lines,
Prosopon).

Palpebral furrow. The palpebral furrow (‘e’ in text-fig. 12) is a depression crossed by caecal lines, and
is therefore part of the fixed cheek. Its probable function was to provide for additional strength at a
critical part of the test.

Palpebral lobe. The ‘top’ or the ‘lid’ of the eye separated from the fixed cheek by the palpebral furrow.
The extension of the ocular striga divides the lobe into two parts.

Parafrontal band. Hupe’s interpretations of the significance of the parafrontal band are complicated
and involve a number of other preglabellar features, such as the transverse crest on the brim in some
trilobites and the frontal boss. Hupe (1953, p. 265) suggests that in Hindermeyeria an elongate boss
represents the parafrontal band and is a segment (an axial lobe) between the glabellar front and the
brim. He considers that the brim contains reduced segments (including their axial lobes) and assumes an
axial continuity between the hypostoma and the glabella. This interpretation is fundamentally different
from that of the present author. The preglabellar boss need not be interpreted as a now-and-then
recurrent axial lobe of an otherwise obsolete segment. The preglabellar boss probably develops in
some trilobites as a secondary ingluvial enlargement of the frontal lobe or in other trilobites as an
ingluvial protuberance projecting from the front of the oesophagus into the pleural brim.

Pleural furrow. Its origin and significance are unknown and it is an attribute of trilobites. The only
obvious function of the furrow is a mechanical strengthening of the pleurae. According to Stormer, it is
the vestige of the primary segmentation (see Fusion of pleurae) ; some muscles may have been attached
to it.

Posterolateral furrow. The occipital homologue of the pleural furrow.

A. A. OPIK: ALIMENTARY CAECA OF AGNOSTIDS

437

Preglabellar median furrow. Most common in agnostids and eodiscids, it is interpreted as a parietal
median septum carrying a mesentery. In the pygidium of many agnostids a ‘postaxial median furrow’
probably has similar function. In conventional trilobites the preglabellar median furrow is rare. Ex-
amples are the Upper Cambrian Liostracina Monke and the Ordovician Tornquistia Reed and Meso-
taphraspis Whittington & Evitt.

Propleuron. That part of a pleura situated in front of the pleural furrow (see Opistopleuron).

Prosopon. External markings and features in the test that have been commonly classed as ‘ornament’,
but signify the presence of organs. This term, introduced by Gill (1949), has been little used, but no
better term is available. Alimentary prosopon in this paper is a ‘subsidiary term’, as envisaged by Gill.

Walcott (1912, pi. 24, fig. 2; p. 192) describes a butter-crab of Ptychoparia cordillerae with a dense
longitudinal striation on its pleurae interpreted as branchiae. Similar structures have been observed in
the test of pleurae of several Australian trilobites and may also be regarded as prosopon.

Whitehouse (1939, pp. 185-7) describes a unique specimen of Redliehia showing imprints of nodular
veins on the thorax, in the position of the opistopleural veins, as well as a pair of such veins on the
posterior border of the cephalon. These veins are prominent and if they were organs common to all
specimens of Redliehia their detection would be simple. They coincide in position with the opisto-
pleural intestinal appendages and, if the coincidence is not accidental, they probably represent an
individual and unique hypertrophic and, perhaps, pathological enlargement of alimentary appendages.

Pygidial caeca. The term genal caeca cannot be applied to the pygidium, which has no cheeks.

Reticulation. A network of caecal ducts that may be anastomotic or only apparently anastomotic.
Caecal dusts may cross one another at various levels in the tissue without forming a reticulate ‘plumb-
ing’ system. Both are seen in the genal caeca of Papyriaspis, but the prosopon is too delicate to keep
them apart.

Rim. The frontal border with doublure and rostral shield (if present).

Rugae. Rugae are wrinkles of the test of agnostids. They correspond to the subcuticular caeca.

Scrobicula (-ae). Scrobiculae in agnostids are external radiating furrows between rugae; internally
the scrobiculae are parietal septa carrying mesenteries.

Sutures. See Facial sutures.

Terraced lines. Occur on the surface of the test and on the external side of the doublure. They cannot
be interpreted, therefore, as muscle attachments: muscles attached to the outer surface of the test
would be shed together with the test during moulting.

Vein. A single linear element of the alimentary prosopon.

Whitening of fossils. The modern method of coating fossils with chlorammonia, magnesia, or bismuth
oxide has increased many times the definition of delicate features of the test and the resolving power
in palaeontological photography. Without whitening the present study would be almost impossible.

Acknowledgements. The author thanks Miss J. Gilbert-Tomlinson and Dr. J. J. Veevers for the reading
of the manuscript. This paper is published by permission of the Director, Bureau of Mineral Resources,
Geology and Geophysics, Canberra, A.C.T.

REFERENCES

gill, e. d. 1949. Prosopon, a term proposed to replace the biologically erroneous term ornament.
J. Paleont. 23, 572.

henningsmoen, g. 1957. The trilobite family Olenidae. Skrift. norsk. Vidensk.-Akad., mat.-nat. Kl.,
1957 (1).

hupe, p. 1953. Contribution a l'etude du Cambrien inferieur et du Precambrien III de l’Anti-Atlas
marocain. Mem. Serv. geol. Maroc, 103. [1952 on title-page.]

1953a. Classification des trilobites (1). Ann. Paleont. 39, 59.

438

PALAEONTOLOGY, VOLUME 3

kobayashi, t. 1949. The Glyptagnostus hemera, the oldest world-instant. Jap. J. Geol. Geogr. 21, 1.
matthew, G. F. 1903. Report on the Cambrian Rocks of Cape Breton. Geol. Surv. Canada.
opik, a. a. 1937. Trilobiten aus Estland. Publ. geol. Instn Tartu Univ. 52.

1952. The hypostoma of Pagetia. J. Paleont. 26, 272.

1956. Cambrian geology of Queensland; in El sistema Cambrico, su paleogeografia y el prob-

lema de su base. XX lnt. geol. Congr., Mexico, 2, 1.

1958. The Cambrian trilobite Redlichia: organization and generic concept. Bur. Min. Resour.

Aust., Bull. 42.

1959. Genal caeca of agnostids. Nature, 183, 1750.

1960. Cambrian and Ordovician geology of Queensland. /. geol. Soc. Aust. 6, 2.

palmer, A. R. 1955. Upper Cambrian agnostids of the Eureka district, Nevada. J. Paleont. 29, 86.

1957. Ontogenetic development of two olenellid trilobites. Ibid. 31,105.

rasetti, f. 1944. Upper Cambrian trilobites from the Levis conglomerate. Ibid. 18, 229.

1945. New upper Cambrian trilobites from the Levis conglomerate. Ibid. 19, 462.

raw, F. 1956. Origin of winged insects. Ann. Mag. nat. Hist. (12) 9, 673.

1957. Origin of chelicerates. J. Paleont. 31,139.

resser, c. E. 1928. Cambrian fossils from the Mohave Desert. Smithson misc. Coll. 81 (2).
stormer, l. 1941. Studies on trilobite morphology. Part II. Norsk geol. Tidsskr. 21, 49.
troedsson, g. t. 1937. On the Cambro-Ordovician faunas of western Quruq tagh, eastern T’ien-
shan. Palaeont. sinica, N.s. B, 2.

walcott, c. d. 1912. Middle Cambrian Branchiopoda, Malacostraca, Trilobita, and Merostomata.
Smithson misc. Colt. 57, 145.

■ 1931. Addenda to descriptions of Burgess shale fossils. Ibid. 85 (3).

westergaard, a. h. 1922. Sveriges olenidskiffer. Sverig. geol. Unders. Avh., ser. Ca, 18.

■ 1936. Paradoxides oelandicus beds of Oland, with the account of a diamond boring through

the Cambrian at Mossberga. Ibid., ser. C, 394.

1946. Agnostidea of the Middle Cambrian of Sweden. Ibid. 477.

1947. Supplementary notes on the Upper Cambrian trilobites of Sweden. Ibid. 489.

1950. Non-agnostidean trilobites of the Middle Cambrian of Sweden. II. Ibid. 511.

whitehouse, F. w. 1936. The Cambrian faunas of north-eastern Australia. Parts 1 and 2. Mem. Qld.
Mus. 11, 59.

1939. Idem. Part 3. Ibid. 11, 179.

whittard, w. F. 1955. The Ordovician trilobites of the Shelve Inlier, west Shropshire. Palaeontogr.
Soc.

A. A. OPIK

Bureau of Mineral Resources, Geology and
Manuscript received 9 November 1959 Geophysics, Canberra

NEW SPECIES OF MANDELSTAMIA (OSTRACODA)
FROM THE ENGLISH MESOZOIC

by John w. neale and t. i. kilenyi

Abstract. The diagnosis of Mandelstamia is translated into English and the distribution of the genus is examined.
Four new species from the English Kimmeridgian are described and one new species from the Lower Cretaceous.

The genus Mandelstamia has been described in two publications. In 1955 Lubimova des-
cribed Mandelstamia Lubimova 1955 from the Russian Upper Jurassic, together with five
species belonging to the genus. In 1956 an authoritative ‘Diagnosis’ of Mandelstamia
gen. nov. was given. No problem is involved regarding nomenclature or priorities but the
Diagnosis given below is translated from the 1956 publication which differs little from
the 1955 description.

Family cytheridae.

‘Genus Mandelstamia Lubimova gen. nov.

Type species — Mandelstamia facilis Lubimova sp. nov.

Upper Jurassic. Lower Volgian Beds. Zone of Virgatites virgatus Buch.

Kiubyshev District, Basin of the River Irgiz, Village of Ukrainka.
vnigri Coll. No. 226-17.

‘Diagnosis. Carapace irregularly oval in shape, weakly convex, sometimes concave in outline in the
anterior third of the dorsal part of the valve. Left valve insignificantly larger than the right and envelop-
ing the latter. Anterior end bow-shaped and rounded off in
more sloping fashion than the posterior. Dorsal margin
straight. Ventral margin concave in approximately the middle
part of the valve. Shell ornament cellular or cellular-tubercu-
late. Inner margin coincides with the outer margin of the
shell [y/c c Hapy>KHLiM KpaeM. This statement appears in
both the 1955 and 1956 texts and is difficult to understand.

Possibly it should read ‘the inner margin coincides with the
line of concrescence’ (ahhhh cpaujeHua).] Pore canal zone
broad at the anterior end and moderately developed at the
posterior end of the shell. Pore canals straight and sparse.

‘Hinge with the bar in the left valve and consisting of three
different elements. In the right valve (fig. 53 b, see fig. lb) a
lamellar, half-moon shaped tooth appears in the anterior part
of the hinge, merging with the margin in front, and behind — -
with the inner margin of the valve. Middle part of the hinge
appears as a narrow, smooth groove arranged above the an-
terior tooth and passing behind it. Posterior end of the hinge
appears as a small, compressed, half-moon shaped tooth, also
merging with the edge of the valve. The hinge in the left valve
(fig. 53a, see fig. la) has a shallow, parallel-sided socket in the
anterior part joined with a straight, smooth ridge forming the
middle part of the hinge. Posterior part of the hinge also con-
sists of a socket but of considerably lesser dimensions than the
anterior. ’

[Palaeontology, Vol. 3, Part 4, 1961, pp. 439-49, pi. 71.]

1 B

I A

text-fig. 1. Structure of Mandelsta-
mia facilis Lubimova. A, Inside of
left valve. B, Inside of right valve.
X94. From Lubimova 1956, p. 141.

440

PALAEONTOLOGY, VOLUME 3

The following remarks on the genus are drawn from both papers quoted above and
represent a free translation:

‘The genus Mandelstamia is distinguished by the characteristic structure of the hinge and five species
occur in the Upper Jurassic sediments of the Middle Volga Region and the Obshchii Syrt. These species
are characterized by the elongate shape of the shell, the characteristic appearance of a vertical trans-
verse furrow and well-marked cellular ornament. In appearance the genus is close to the genus Cyclo-
cytheridea Mandelstam 1955, particularly in general shape and valve ornament, but may be distinguished
by the weaker development of the hinge with the lack of large sockets open at the anterior and posterior
ends in the left valve, the rather greater elongation of the shell, and the strongly marked concavity in
the anterior third of the dorsal part of the shell. The genus Mandelstamia also shows some resemblance
in shape and shell ornament to the genus Palaeocytheridea Mandelst. (Mandelstam 1947, p. 243) from
the Middle Jurassic sediments of Mangyschlak, but is distinguished from it mainly on the structure of
the hinge which has not got notched sockets in the left valve or notched teeth in the right valve. ’

Hitherto, apart from one new species from the Kimmeridgian of Dorset described by
Malz (1958), this genus has not been recorded outside the U.S.S.R. In the course of work
on the English Kimmeridge and marine Lower Cretaceous ostracod faunas five new
species of the genus have been discovered and these are described below. The distribu-
tion of all species of the genus known to date is shown in Table 1.

Acknowledgements. J. W. N. would like to acknowledge the great kindness and help of Professor V. V.
Drushchitz of the University of Moscow, and Drs. P. S. Lubimova and M. I. Mandelstam of the All-
Union Petroleum Geological Exploration Institute, Leningrad, who have made available the necessary
Russian literature; also the help of Mr. E. N. Blackmore, A.R.C.A., who very kindly made the draw-
ings for text-fig. 3.

Abbreviations. In giving dimensions the following abbreviations are used throughout the text : L, length ;
H, height; W, width; Hi, hinge length; M/a, width of anterior marginal area; M/p, width of posterior
marginal area. In all cases the dimensions are in millimetres. Numbers preceded by the index letters
‘HU’ indicate the catalogue numbers in the collection of the Geology Department, University of Hull,
where all the specimens here described and figured are stored.

SYSTEMATIC DESCRIPTIONS
Family cytheridae Baird 1850

Subfamily loxoconchinae Sars 1925
Genus mandelstamia Lubimova 1955

Mandelstamia rectilinea Malz 1958

Plate 71, figs. 1-4, 6

Mandelstamia rectilinea Malz 1958, p. 38, pi. 11, figs. 58-63
Material. Fifty-two valves and carapaces. HU.2.J.1.21; HU. 3. J. 20. 2-52.

Distribution. Rasenia mutabilis Zone, L. Kimmeridgian, Black Head, 3./ miles north-east of Wey-
mouth, Dorset; ‘Lower Kimmeridge Clay of Ely’; ‘Lower Kimmeridge Clay of Ringstead’.

Measurements

L H W Hi M/a M/p
Left valve HU. 2. J. 1.21. 0-69 0-39 0 16 0-45 0 06 0 03

Description. Carapace oblong shaped, both ends equally rounded, the posterior being
higher than the anterior. The two valves are about equal in size. In dorsal view the

J. W. NEALE AND T. I. KILENYI: MANDELSTAMIA (OSTRACODA) 441

carapace is oval shaped with a ‘waist’ just above the middle. The greatest width is in
the posterior half, about one-third of the way from the posterior end. The two valves
DISTRIBUTION OF KNOWN SPECIES OF MANDELSTAMIA.

<

>

O

2

CD

O

<

u!

y

a

a.

>

2

:

i

:

:

2. <

0

1

o

->

' -

£

2

M. ABDITA LUBIMOVA

M IGNOBILIS LUBIMOVA

M. ANGULATA KILENYI

M TRIEBEL 1 KILENYI

M RECT ILINE A MALZ

M. MACULATA KILENYI

M SP 1 KILENYI

M SEXTI NEALE

B ERRlASl AN

UPPER VOLGIAN

PUPBECK IAN

LOWER
V O L G IAN

NIKITI NELLA
NIKITINI

VIRGATITES

VIRGATUS

1

PORTL ANDIAN

PAVLOVIA

PANDERI

PER ISP HI NOTES
BLEICHERI

I

K IMMER IDGIAN

■

■

UPPER

KIMMERIDGIAN

■

1

—

LOWER

UPPER OXFORDIAN

s

UPPER OXFORDIAN

LOWER OXFORDIAN

LOWER OXFORDIAN

U. s. S. R.

U. K.

TABLE 1.

are symmetrical. The dorsal margin is straight, both cardinal angles being more or less
marked, whilst the ventral margin is concave for the first third of its length. The surface
of the valve is reticulate, the shell being built up of a system of pits and ribs. The size of
these pits is larger on the anterior half of the valve, the two size groups being separated

442

PALAEONTOLOGY, VOLUME 3

by a distinct vertical line. In polarized light a faint uniaxial interference figure occurs,
while in addition every pit shows a small Brewster-cross and one or two interference rings.
There is a vertical depression on each valve, about one-third of the way from the anterior
end.

The marginal areas are narrow, inner margin and line of concrescence coinciding. The
selvage is well developed, the selvage lip being narrow, and in most cases the inner
lamella does not disappear under it ventrally. Radial porecanals are simple and straight,
their basal part being wide, and getting gradually narrower towards the exterior end.
They number between seven and twelve anteriorly and between two and four posteriorly.
The hinge is lophodont and in the right valve consists of three elements. The anterior one
is a smooth, ellipsoidal ridge, the median one a smooth, narrow groove, while the
posterior element resembles the anterior, but is smaller. In the left valve the terminal ele-
ments are smooth half-moon-shaped sockets, the anterior one being the wider. These
are joined by a smooth ridge which runs in the middle of the contact margin. The muscle
scar pattern consists of an oblique row of four equally sized scars and one anterior scar
in line with the top of the row. Sexual dimorphism is not apparent.

Remarks. Mandelstamia rectilinea corresponds in many respects with the type species,
M.faeilis Lubimova, but its hinge seems to be simpler, and in M. facilis the size of the
surface pits appears to be uniform.

Mandelstamia triebeli Kilenyi sp. nov.

Plate 71, figs. 5, 9, 10, 14, 15
Derivation of name. In honour of Dr. E. Triebel.

Holotype. A left valve. HU.2.J.1.22. Paratypes. Five hundred and thirteen valves and carapaces.
HU. 3.J.21. 1-513.

Occurrence of holotype and paratypes. Sixty feet above the base of the Rasenia mutabilis Zone, L.
Kimmeridgian, Black Head, 3J miles north-east of Weymouth, Dorset.

Distribution. Rasenia mutabilis and Aulacostephanus pseudomutabilis Zones, Lower Kimmeridgian.
Black Head, 3| miles north-east of Weymouth, Dorset.

Measurements

L

H

W

Hi

M/a

M/p

Holotype
Left valve :

0-60

0-34

0 15

0-40

005

004

Paratypes
Left valve :

0-55-0-61

0-30-0-35

0 15-0-17

0-40

0-06

0-04

Right valve :

0-57-0-64

0-32-0-36

0-15-0-17

0-40

0-06

0-04

Diagnosis. A species of Mandelstamia in which the valves are equal in size, the carapace
tapering towards the posterior end. Anterior end rounded, posterior cardinal angle
more or less prominent; dorsal margin straight, ventral convex. Marginal areas broad.

Description. Carapace elongated, tapering slightly posteriorly. The two valves are equal
in size. In dorsal view the valves are trapezoid shaped, with a slight depression just
above the middle, the greatest width being in the posterior half of the carapace. The two
valves are nearly the same shape, the only significant difference being in the posterior end,
which is more angular in the right valve. Dorsal margin straight and cardinal angles
marked, especially on young moults. The anterior end is equally rounded on both valves,
the ventral margin being slightly convex, curving gently upwards towards the posterior

J.W. NEALE AND T. I. KILENYI: MANDELSTAMIA (OSTRACODA) 443

end and having a short, straight section in the middle. The posterior end is rounded,
but is slightly angular on the right valve. No trace of sexual dimorphism was found. Line
of concrescence and inner margin coincide; the selvage is wide, the selvage lip being
wider on the right valve than on the left. Radial porecanals are few (about six to eight
anteriorly) and simple, with their interior part wider than their exterior.

The hinge is lophodont, consisting of two terminal ellipsoidal ridges (right valve)
and corresponding sockets (left valve). The median element on the left valve is a smooth
straight ridge, fitting into a smooth groove on the opposite valve. The surface of the
valve is strongly reticulate, the shell being built up of a network of strong ribs and pits,
the size of these latter being fairly uniform all over the valve. In polarized light a small
Brewster-cross can be seen in every pit, the rest of the shell remaining dark. Muscle-scar
pattern consists of an oblique row of four scars in vertical superposition, and two
anterior scars. The upper anterior scar is larger than the others and half-moon shaped.

Remarks. Mandelstamia triebeli differs from M. rectilinea Malz in the uniform reticula-
tion and pronounced posterior tapering of the shell seen in side view.

Mandelstamia angulata Kilenyi sp. nov.

Plate 71, figs. 11, 12, 16-18
Derivation of name. Angulatus (Lat.) — angled.

Holotype. A female (?) left valve HU. 2. J. 1.23. Paratypes. Fifty-six valves and carapaces. HU.3.J.22.
1-56. Occurrence of holotype and paratypes. Eleven feet above base of the Rasenia cymodoce Zone,
Black Head, 3 J miles north-east of Weymouth, Dorset.

Distribution. Pictonia baylei and Rasenia cymodoce Zones, Lower Kimmeridgian, Black Head, Dorset.
Measurements

Holotype

L

H

W

Hi

M/a

0-45

0-26

0-35

003

Paratypes

L

H

Hi

M/a

M/p

Left valve :

0-42-0-45

0-25-0-27

0-33-0-35

0-03

0-01

Right valve:

0-42-0-44

0-22-0-23

0-35

0-03

0-01

Diagnosis. A small species of Mandelstamia with pointed posterior end. Dorsal margin of
the right valve is convex, that of the left valve straight. Left valve slightly rounded
posteriorly, right valve pointed. Sexual dimorphism doubtful.

Description. Carapace small, ovoid-triangular. Left valve slightly larger than the right,
with only a slight overlap on the dorsal and anterior part. In dorsal view the carapace
is pear shaped, with a marked ‘waist’ at about the middle, the greatest width being in the
posterior half of the carapace. The two valves differ in shape. In the right valve the dorsal
margin is convex, both cardinal angles are rounded, and the posterior end is pointed,
although the extreme end is blunt. The ventral part of the margin is straight with a short
concave section in the middle, while the anterior end is rounded. The left valve, on the
other hand, differs from the right both dorsally and posteriorly. The left valve dorsal
margin is nearly straight, the anterior cardinal angle is marked, while the posterior end is
more rounded than that of the right valve. Both valves are highest at the anterior
cardinal angle. Some of the left valves have a more pointed posterior end and these are
tentatively regarded as males. The surface of the valve is reticulate, the shell structure

444

PALAEONTOLOGY, VOLUME 3

being characterized by a network of ribs, the pits between the ribs being roughly hexa-
gonal shaped. The inner lamella is relatively broad and is well developed posteriorly also.
Along the entire free margin the selvage is wide and forms a narrow selvage lip ventrally.
The inner margin and line of concrescence coincide. Radial porecanals are straight and
simple, about eight on the anterior part and four or five on the posterior part.

The hinge is lophodont. In the left valve the hinge consists of two oval-shaped ter-
minal sockets with a straight, smooth bar between them. The right valve bears the com-
plementary ridge-groove-ridge arrangement. The hinge is narrow and rather delicate.
The muscle-scar pattern was not clearly seen. In one case an oblique row of four scars
was observed, the lower three being larger than the dorsal one and elongated longitudin-
ally.

Remarks. This species is similar in certain respects to Mandelstamia triebeli but has a
more angular contour and a more pointed posterior end.

Mandelstamia maculata Kilenyi sp. nov.

Plate 71, figs. 19-25

Derivation of name. Maculatus ( Lat.) — spotted, speckled.

Holotype. A complete female carapace. HU. 2. J. 1.24. Paratypes. Eighty-nine valves and carapaces.
HU.3.J.23.1-89. Occurrence of holotype. Forty-four feet above the base of the Subplanites (V.)
grandis Subzone, Upper Kimmeridgian, Rope Lake Head, Kimmeridge, Dorset. Occurrence of para-
types. Kimmeridgian, New Closes Cliff, Speeton, Yorkshire. Horizon uncertain.

EXPLANATION OF PLATE 71
Magnification x 50 except where stated

Figs. 1-4, 6. Mandelstamia rectilinea Malz 1958. Sixty feet above base of Rasenia mutabilis Zone, L.
Kimmeridgian, Black Head, Dorset. 1, Right valve. External view. HU.3.J.20.2. 2, Left valve.
External view. HU.2.J.1.21. 3, Left valve, in transmitted light. HU.2.J.1.21. x60. 4, Left valve.
Dorsal view. HU.2.J.1.21. 6, Left valve. External view. HU.3.J.20.3.

Figs. 7, 8, 13. Mandelstamia sp. 1. 7, 8. Ten feet above base of Pavlovia pallasioides Zone; 13, 17 feet
below top of Pectinatites pectinatus Zone, U. Kimmeridgian, about one-third of a mile east of Fresh-
water Steps, Dorset. 7, Right valve. External view. HU.3.J.30.1. 8, Left valve. External view.
HU.3.J.30.2. 13, Left valve. External view (juvenile). HU.3.J.30.7.

Figs. 5, 9, 10, 14, 15. Mandelstamia triebeli Kilenyi sp. nov. 60 feet above base of Rasenia mutabilis
Zone, L. Kimmeridgian, Black Head, Dorset. 5, Right valve in transmitted light. HU.3.J.21.3.
x60. 9, Right valve. External view. HU.3.J.21.2. 10, Holotype. Left valve. External view.
HU.2.J.1.22. 14, Right valve. External view. HU.3.J.21.5. 15, Left valve. External view.

HU.3.J.21.7.

Figs. 11, 12, 16-18. Mandelstamia angulata Kilenyi sp. nov. 11 feet above base of Rasenia cymodoce
Zone, L. Kimmeridgian, Black Head, Dorset. 1 1 , Male right valve. External view. HU.3.J.22.7.
1 2, Male left valve. External view. HU.3.J.22.8. 16, Female right valve. External view. HU.3.J.22.2.
17, Female carapace. Dorsal view. HU. 3.J.22. 10-11. 18, Female (?) left valve. External view.
HU.3.J.22.6.

Figs. 19-25. Mandelstamia maculata Kilenyi sp. nov. 19-22, 24, 25, Zone uncertain, U. Kimmeridgian,
New Closes Cliff, Speeton, Yorkshire; 23, 44 feet above base of Subplanites ( V .) grandis Subzone,
U. Kimmeridgian, Rope Lake Head, Dorset. 19, Female right valve. External view. HU.3.J.23.7.
20, Female right valve. External view. HU.3.J.23.8. 21, Female left valve. Dorsal view. HU. 3.
J.23.5. 22, Female right valve. Dorsal view. HU.3.J.23.4. 23, Holotype. Female carapace. Dorsal
view. HU. 2. J. 1.24. 24, Female left valve. External view. HU.3. J.23.5. 25, Male right valve.
External view. HU.3.J.23.6.

Palaeontology, Vol. 3.

PLATE 71

N EA L E and KILE NY I, Mandelstamia

J. W. NEALE AND T. I. KILENYI: MAN DEL ST AMI A (OSTRACODA) 445

This and all the species described below differ from the type species in hinge structure. The difference
is not regarded as being of subgeneric importance.

Distribution. Subplanites (V.) grandis Subzone. Upper Kimmeridgian, Rope Lake Head, Kimmeridge,
Dorset, and Kimmeridgian, New Closes Cliff, Speeton, Yorkshire, zone uncertain.

Diagnosis. Carapace elongated, tapering posteriorly. Anterior end rounded, posterior
rather more angular. Ventral margin of the right valve strongly concave, that of the left
valve straight, the ventral side of the valves overhanging the ventral margins to form a
sort of ala. Surface covered with equally dispersed pits. The hinge of the right valve has
the following structure: The anterior element is a minute oval-shaped ridge with four
small rounded projections, the middle two of which are larger than the others. The
median element is a wide and smooth groove, which ends on both sides in a rounded pit,
which is slightly deeper than the groove itself. The posterior element is like the anterior
but smaller. The left valve terminal elements are oval-shaped sockets, with three or four
faint loculi in them. The median element lies in the middle of the contact margin and
consists of a smooth bar, both ends of which are thickened and slightly projecting.
Marginal areas relatively broad with a few straight, simple porecanals. Sexual dimor-
phism very pronounced, males being much longer than females.

Measurements

L

H

W

Hi

M/a

M/p

Holotype:

0-72

0-41

0 31

Left valve :

0-72

0-40

017

0-50

0 08

004

$Right valve :

0-72

0-42

017

0-50

007

0 04

dRight valve:
Juvenile forms :

0-86

0-45

0-59

007

006

Instar 8:

0-56

0-34

„ 7:

0-46

0-26

„ 6:

0-41

0-25

Description. Carapace trapezoidal (?) or elongate oblong shaped (d1). The left valve
larger than the right. Overlap is doubtful, but there may be slight overlap dorsally. In
dorsal view carapace is pear shaped with a marked depression just above the middle.
The greatest width is in the posterior part of the carapace, about one-third of the length
from the posterior end. The dorsal margin is straight on both valves, and the anterior
end rounded; both cardinal angles are marked. The posterior end of the left valve is
rounded, but is more angular in the right valve, and the posterodorsal margin is straight.
The ventral margin is straight (left valve) or slightly concave (right valve), and the side of
the valve overhangs the ventral margin, especially on the right valve, forming a sort of
ala. Only one male specimen was found (a right valve), and this is oblong shaped and
much longer than the female. There are also a few denticles on the anterior margin. In
side view the greatest height of the female valve is at the anterior cardinal angle, the male
valve is highest at the middle. The surface of the valve is strongly reticulate, the pits being
arranged in vertical rows. The inner lamella is broad ; the inner margin and line of con-
crescence coincide. The selvage is wide and projecting, the selvage lip being very pro-
minent on the right valve. Radial porecanals are simple and straight and only about six
occur on the anterior margin. The hinge is strongly developed, and is of a new type
which has been fully described in the diagnosis above.

The structure of the shell is a network of ribs similar to that in other species of Mandel-

cg

B 6612

446

PALAEONTOLOGY, VOLUME 3

stamia. In polarized light only faint interference figures can be seen in the pits and the
overall cross is absent. The ventral parts of the valve, where the ‘ala’ is developed, appear
dark in every position of the microscope stage.

Mandelstamia sp. 1 . Kilenyi
Plate 71, figs. 7, 8, 13

Material. Twenty-three valves. HU.3.J.30.1-23.

Occurrence. Upper Pectinitites and Lower Pavlovia Zones, U. Kimmeridgian, Freshwater Steps to
Hounstout Cliff, Dorset.

L

H

W

Hi

M/a

Left valve:

0-50

0-32

016

0-30

004

Right valve:

0-48

0-28

015

0-30

004

Diagnosis. Shape like Mandelstamia maculata but the hinge is more advanced towards
the merodont hinge type, the terminal teeth being more markedly dentate. The median
ridge on the left valve is straight and has no terminal widening. Radial porecanals few.
Description. Only a few, badly preserved specimens were found. The contour of the
valves is similar to that of Mandelstamia maculata but this form is more tumid. In dorsal
view the ‘waist’ is less prominent, the carapace being more oval-shaped. The hinge
structure is more advanced towards the merodont type. The terminal ridges are more
markedly dentate, and the corresponding sockets loculate; the median element is
straight with no projections or widening at the ends. Inner margin and line of concres-
cence coincide; the inner lamella is narrow. Radial porecanals few and simple. Selvage
wide, well developed along the entire free margin.

Mandelstamia sexti Neale sp. nov.

Text-figs. 2, 3, 4

Derivation of Name. Sextus (Lat.) — sixth. An allusion to the fact that this species is confined to the
sixth stratum of the D. Beds at Speeton and is not found above or below.

Holotype. Female carapace, HU.l.C.4.49. Paratypes. One hundred and four valves and carapaces
belonging to various growth stages. HU.l.C.4:47,48,50-54; HU.l.C.5.1-97. Other material: Several
hundred valves and carapaces from the Speeton Clay D6 Beds (D6B-D6I). Occurrence of holotype and
paratypes. Speeton Clay D6A, Berriasian, Lower Cretaceous. Middle Cliff, Speeton, Yorkshire.

Distribution. This species is unknown outside the Speeton Clay Blue Bed (D6) which is about 6 feet
thick, in the coastal exposures.

Diagnosis. A species of Mandelstamia with well-differentiated hinge, and differing in
shape from previously described species in that the height is somewhat greater in pro-
portion to the length. The terminal hinge cusps in the right valve are divided to form four
or five small teeth; the median bar in the left valve is finely denticulate. Sexual dimor-
phism not very marked, the presumed males having shells which in side view taper
slightly more posteriorly than those of the presumed females.

Measurements

L

H

W

Hi

M/a

M/p

Dimensions of Holotype:

0-71

0-40

0-35

. .

L.V. female HU. l.C.4.48.

0-69

0-40

016

0-47

006

004

J. W. NEALE AND T. I. KILENYI: M AN DELSTA MIA (OSTRACODA)

447

Description. Shape elongate-oval, tapering somewhat posteriorly, rather more so in the
case of the male than in the case of the female. Anterior margin forms an asymmetrical
curve which bulges ventrally, while the dorsal and ventral margins are relatively straight,
although the latter may, on occasion, be slightly concave. Posterior margin rounded-sub-
angular with a straight postero-dorsal section. Cardinal angles marked. In dorsal view
the shell is widest posteriorly, the greatest width lying about three-quarters of the length
from the anterior end. The inflation in the female is a little greater than in the male but
sexual dimorphism is not pronounced.

Ornament is typical of Mandelstamia
and consists of deep concentrically
arranged pits. Radial pore canals (text-
fig. 2) are straight and sparse. The flange
and selvage are very well developed and
in well-preserved specimens the inner
margin and the line of concrescence are
narrowly separated anteriorly. Elsewhere,
the inner margin and line of concrescence
coincide. Hinge structure as described
above differs from that of Mandelstamia
facilis Lubimova. The tiny denticles on
the hinge bar in the left valve consist of
minute elliptical granules of calcite with
the long axes of the ellipses set at right
angles to the length of the bar. These are
similar to the granules which beset the
margin of the shell and it is perhaps
better to term the bar ‘granulate’ rather
than ‘denticulate’. This median bar fits
into a shelf-like groove under the margin
of the right valve. The latter overlaps the
left valve dorsally except at the cardinal
angles.

The pattern of growth is the same as that generally found in the Ostracoda and the
various growth stages are indicated on the graph shown in text-fig. 4.

Remarks. In this species the height is slightly greater in proportion to the length than in
other species of Mandelstamia and it differs from M. maculata in the more pronounced
taper posteriorly and in the straighter ventral margin.

B

text-fig. 2. Structure of Mandelstamia sexti Neale
sp. nov. Adult male right valve. HU.l.C.4.47.
Specimen destroyed. A, Inside. B, Dorsal view.
x90.

REFERENCES

Aioehmoba, El. C. 1955. OcxpaKOAbi Me.3030iicKiix otaojkchhh cpeAHero noBOAJKbn n oGujero cbipTa in
AioGiiMOBa, Id. C. h XaGapoBa, T. H. 1955. OcTpanoAbi Me3030HCKiix otaokchhh BoAro-YpaAbCKofi
oGAacTH. TpyAbi Bcecoio3Horo neijmuioro May'Jiio-iiccAeAOBaTeAbCKoro reoAoro-pa3eAOHHoro iiHCTiiTy ra

(BHIirPH) AemiHrpaA- 1-189, Tab. 1-6, figs. 1-19, pi. 1-13.

Aioehmoba, El. C. 1956. in MaTepnaAbi no naAeoHTOAoriift. Bcecoio3HbiH HayHHO-iiccAeAOBaTeAbCKini
reoAoniHecKHii HHCTHTyT (BCETEH). MocKBa. 1-356. pi. 1-43.

TEXT-FIG. 3

J. W. NEALE AND T. I. K1LENYI: M A N DELSTA MIA (OSTRACODA)

449

malz. h. 1958. Die Gattung Macrodentina und einige andere Ostracoden-Arten aus dem Oberen Jura
von NW-Deutschland, England und Frankreich. Abh. senckenb. mturf. Ges. 497, 1-67, pi. 1-11.
Mah4eai>iutam, M. H. 1947. OcrpaKO/iw 11.3 otao/kciiuh cpe^Heii iopi>r noAyocrpoBa MaurwinAaKa. C6op-
hhk CTaTeii no MHKpo(J)ayne Hecjvi HHbix MecTopoK^eumi KaBKa3a, 9m6w ii Cpe.meii A3hh. (BHHrPH)

JOHN W. NEALE
University of Hull

T. I. KILENYI
Sir John Cass College, London

Manuscript received 25 September 1959

text-fig. 3. Mandelstamia sexti Neale sp. nov. 1, 2, Hinge structure. 1, Left valve. HU.l.C.4.53 (a)
from inside, (b) dorsal view. xl50. 2, Right valve. HU.l.C.4.54 ( a ) from inside, (b) dorsal view.
Xl57. 3, Female left valve. Adult. HU. l.C.4.48 from outside, x 84. 4, Holotype. Female carapace.
Adult. HU.l.C.4.49 (a) from left, (b) dorsal view. X 82. 5, Left valve. Instar 7. HU.l.C.4.51 from
outside. x89. 6, Left valve. ?Male. Instar 8. HU. I.C.4. 50 (u) from outside, (b) dorsal view. xllO.

0-5

i

m

<T>

I

2

2

(A

0-4--

0-3-

0-2

0-1

MANDELSTAMIA SEXTI

+ RIGHT VALVES
. LEFT VALVES
° CARAPACES

O EXTRAPOLATED INSTARS

LENGTH MMS

0-2 03 CM 05 06

text-fig. 4. Growth stages in Mandelstamia sexti Neale sp. nov.

0-1

0-7

0-8

THE FEEDING MECHANISM OF THE PERMIAN
B RACHIOPOD PRORICHTHOFENIA

by M. J. S. RUDWICK

Abstract. The morphology of Prorichthofenia uddeni (Bose) and P. permiana (Shumard), from the Permian of
Texas, U.S.A., is analysed and interpreted in functional terms. The methodology of such functional interpreta-
tions is discussed briefly. It is inferred that most of the anomalous characters of the genus (especially the un-
usually thin and recessed dorsal valve, the peculiar form of hinge and the internal spines) relate to an unusual
feeding mechanism. On this interpretation rapid movements of the dorsal valve created powerful currents and
eddies, from which food-particles were collected by the mantle surfaces. It is inferred that the spines projecting
across the aperture of the shell, and those on the dorsal valve, served to supplement the particle-collecting capacity
of the mantle surfaces, and to exclude harmfully large particles. Experiments with working models of the shells
confirm that the morphology of both species would have been highly adapted to the efficient operation of this
mechanism. The specific differences in morphology are taken to represent a minor functional differentiation.

INTRODUCTION

The detection of adaptation in fossils. Though adaptation stands at the centre of the
modern debate on the mechanisms of evolutionary change, the problem of the recogni-
tion of adaptation in fossils, or the inference of function from structure has received
surprisingly little attention. The problem is especially acute in the study of aberrant
fossils such as the richthofeniids; for then neither homological nor analogical compari-
sons with living animals can yield reliable inferences about the functional significance of
the morphology. Therefore a method must be used that does not depend upon specific
similarities with living animals.

Many functions of the body demand, for their efficient operation, predictable modi-
fications of the anatomy. For these, it is often possible to specify the nature of the ‘ideal’
structure that would be able to fulfil this function with perfect efficiency. But in actuality
the materials (anatomical and environmental) involved in the function are never ‘per-
fect’ in their properties. For any given set of materials, the ‘ideal’ structure must, there-
fore, be modified into the paradigm. This is the structure that can fulfil the function with
maximal efficiency under the limitations imposed by the nature of the materials. The
degree of approximation between any paradigm and an observed fossil structure is a
measure of the degree of efficiency with which the structure would have been physically
capable of fulfilling the function ; but it cannot establish the probability that the structure
did fulfil it. But by analogy with adaptation in living animals, there are strong grounds
for inferring that a fossil structure capable of fulfilling a certain function with great
efficiency did fulfil that function, especially if it can also be shown that the structure
would have been inefficient or inoperable as the agent of any other conceivable function.
Thus, by transforming rival possible functions into their respective paradigms, rival
structural predictions can be made; and these can be tested by direct comparison with
the observed structure of the fossil.

The ease and confidence with which a function can be inferred by this method is

[Palaeontology, Vol. 3, Part 4, 1961, pp. 450-71, pi. 72-74.]

M. J. S. RUDWICK: THE PERMIAN BRACHIOPOD P RO RICHTHOFENIA 451

directly proportional to the efficiency of the adaptation. A structure that was very
efficient will approximate closely to the paradigm of its function, and thereby can be
recognized as an adaptation with relative ease. A less efficient structure will be more
ambiguous, because it will not be very similar to its paradigm, and is likely to show
some points of fortuitous resemblance to the paradigms of other functions. A non-
adaptive structure can never be recognized as such; for its apparent lack of correspon-
dence to any paradigm might always be due to failure to consider the correct function
and the correct paradigm. Thus there can be positive and cumulative evidence that a
structure was an efficient adaptation; but it is methodologically impossible ever to
demonstrate that a structure was non-adaptive.

This method involves an analysis of adaptation only as a static phenomenon. Theories
of its causal origin (e.g. by natural selection) or of its temporal origin in a particular
instance (by a particular evolutionary lineage) are irrelevant to the detection of an
adaptation.

Material studied. The earlier descriptions of richthofeniids were based on rather poorly
preserved material; and the morphology remained imperfectly understood until well-
preserved silicified specimens were discovered in the Permian of Texas, U.S.A. More
recently, such specimens have been recovered in great numbers from these rocks, by
dissolving large masses of limestone in weak acid (Cooper 1950). The present work is
based on a small collection of this material, from the Permian of the Glass Mountains,
near Marathon, Brewster County, Texas. The specimens are deposited in the Sedgwick
Museum, Cambridge (registration numbers E. 15,577-95; E. 15,706-35; E. 15,740-91;
E. 17, 124-39); some were donated by Professor A. Williams and some were received in
exchange from the U.S. National Museum. The collection includes two of the species
described by King (1930): Prorichthofenia uddeni (Bose) from the Leonard formation at
Old Word Ranch, and P. permiana (Shumard) from the overlying Word formation in
Hess Canyon.

The basic morphology o/Prorichthofenia. The normal form of a brachiopod shell is modi-
fied to an extreme degree in the richthofeniids. Nevertheless, the basic structure, of two
shelly valves hinged together on the posterior side, is still recognizable. But the ventral
valve is modified into an irregular cone; and the dorsal valve is reduced to a thin flat
operculum, which closes the interior of the ventral valve some way below its rim (text-
fig. lc). The ventral valve is cemented to the substratum by its apex and by tubular
external spines. Solid internal spines are usually developed on the internal surface of
the ventral valve and on the lower surface of the dorsal valve.

For the purposes of this paper, the two species studied differ most significantly in the
form of the internal spines on the upper part of the ventral valve. In P. uddeni they occur
in a single row, they often branch and anastomose, and in some specimens they form
a continuous mesh over the opening of the ventral valve (PI. 73, fig. 2). In P. permiana
they are relatively stouter, they occur not in a row but in an irregular ‘thicket’, and
they rarely branch or anastomose (PI. 73, figs. 7-9). The smaller specimens of both
species have no internal spines.

The plane of the dorsal valve (when closed) will be taken as ‘horizontal’; and the
axis at right angles to this as ‘dorso-ventral’. The space enclosed by the dorsal valve
will be termed the inner shell cavity (text-fig. lc, I.S.C. ); the space above the dorsal

452 PALAEONTOLOGY, VOLUME 3

valve, but below the rim of the ventral valve, will be termed the outer shell cavity
( O.S.C. ).

RECONSTRUCTION OF THE ‘SOFT PARTS’

The want/e tissue. In living articulate brachiopods there is a uniform relationship be-
tween the valves and the mantle tissue that secretes them. There are normally two dis-
tinct shell layers: a thin outer ‘lamellar’ or primary layer and a thicker inner ‘prismatic’
or secondary layer (text-fig. 1 b). The primary layer is deposited only by the extreme edge

text-fig. I. Homological relations between a ‘normal’ brachiopod (a, b) and a richthofeniid (c, d),
shown in diagrammatic median longitudinal sections, a. Anatomy of a living ‘normal’ brachiopod;
b, the same, after destruction of ‘soft parts'; c, ‘hard parts' of a richthofeniid; d, reconstruction of
anatomy of richthofeniid. a.m., adductor muscle; att., attachment area; c.p., cardinal process; d.i.sp.,
dorsal internal spines; d.m., divaricator muscles; D.V., dorsal valve; e.sp., external spines; h.a., hinge
axis; inner mantle cavity; I.S.C., inner shell cavity; lop/i., lophophore; M.C., mantle cavity;

m.s., muscle scars; nit., mantle tissue; O.M.C., outer mantle cavity; O.S.C., outer shell cavity; s., shelf;
S.C., shell cavity; v.i.sp., ventral internal spines; V.V., ventral valve; visceral cavity lightly stippled;
mantle tissue and lophophore densely stippled; primary layer of shell, solid black; secondary layer of

shell shaded.

of the mantle lobe (text-fig. la, mt.), while the secondary layer is deposited, on the inner
side of the primary layer, by the rest of the outer epithelium of the mantle lobe (Wil-
liams 1956). A precisely analogous mode of formation occurs in lamellibranchs (Beed-
ham 1958). For any shell formed in this way the shell surfaces on which the secondary
layer is exposed correspond to the areas originally covered by the mantle (text-fig. lb,
cf. la).

In Prorichthofenia the two shell layers are readily recognizable even in silicified speci-
mens, by their distinctive surface characters.

The shell structure is complicated by the development of vesicular texture in parts of the secondary
layer, and by the existence of an apparently distinct internal ‘third layer’. Waagen (1884-7, pp. 730-1),
King (1930, p. 97) and others have identified this internal layer as the homologue of the external
primary layer of other brachiopods, and the outer and vesicular layers as an external ‘investment’
without parallel in other brachiopods. This leaves unexplained the very close resemblance (e.g. in
strong growth-lines and the bases of external spines) between the outer layer and the primary layer of
other brachiopods ; and the resemblance between the internal layer and the primary layer can be ex-
plained in terms of analogy. The internal layer bears vague growth-lines; but these merely represent

M. J. S. RUDWICK: THE PERMIAN BRACHIOPOD PRO RICHTHOFENIA 453

the former position of the internal ‘shelf’ on which the dorsal valve rests (p. 455, PI. 72, figs. 1M-). It
also bears posteriorly a pair of flat ‘sectors’ separated by a rounded groove, resembling the cardinal
area and arched deltidium (seen from within) of a ‘strophic’ brachiopod (Rudwick 1959). These
represent the former positions of the ‘fulcral ridges’ (i.e. the functional hinge-line) and the median gap
occupied by the cardinal process (p. 457), and are therefore analogous to a true cardinal area and delti-
dium; but there is no independent evidence that they are homologous. Thus the internal layer appears
to be merely a modified part of the inner surface of the secondary layer; and the outer layer then repre-
sents the primary layer.

The primary layer is confined to the outer surface of the ventral valve, and is not
present on any part of the dorsal valve (text-fig. lc). Therefore, if the shell was formed by
the ‘normal’ mode, the whole inner surface of the ventral valve, and both surfaces of the
dorsal valve, must have been covered with mantle tissue (text-fig. id). If this tissue was
permanently in this position, a large area of the mantle would have been permanently
exposed to the external environment (cf. Williams’s (1958) reconstruction of the mantle
in oldhaminoids). The alternatives are to postulate (a) that the ventral mantle was fre-
quently retracted off the walls of the outer shell cavity (in fact this would have been
difficult, if not impossible, in specimens with spines in this position; cf. text-fig. 1 d)\ or
(b) that the process of shell formation was abnormal on the dorsal valve, and that its
upper surface was not covered by mantle tissue at all (cf. Stehli’s (1956) interpretation
of oldhaminoids). The mantle must have been abnormal either in being permanently
exposed, or in its mode of shell secretion. In other words, Prorichthofenia (and the
oldhaminoids) differed from present-day brachiopods either functionally (analogically)
or homologically. The manifest similarity in shell structure clearly favours the former.
Prorichthofenia is therefore reconstructed with an outer shell cavity permanently lined
with mantle tissue, forming an outer mantle cavity (text-fig id, O.M.C.).

The mantle tissue of Prorichthofenia, like that of living articulate brachiopods, was
probably very thin, at least on the side of the outer shell cavity ; for the edge of the dorsal
valve often lies very close to the inner surfaces of the ventral valve. Its proximity to some
of the spines which project across the outer shell cavity shows that the tissue sheathing
the spines must have been equally thin.

The ‘ body ’. In living articulate brachiopods the ‘body’ (i.e. the organs enclosed in the
visceral cavity) is confined to a small median posterior portion of the shell cavity, and
the body-wall lies closely against the muscles both anteriorly and laterally. In Prorich-
thofenia the muscle attachments are similar to those found in living articulate brachio-
pods. On the ventral valve both adductors and divaricators were attached either to the
floor of the inner shell cavity or, more commonly, to the posterior face of the median
ridge that runs up the anterior wall of the cavity (PI. 72, fig. 10). On the dorsal valve
there is a pair of adductor scars in front of the hinge axis (text-fig. 4c; PI. 72, figs. 4, 6,
11: a.m .); and the cardinal process is situated in the median plane, on the posterior
border of the valve (text-fig. 4c; PI. 72, figs. 5, 6, 11, 12: c.p.). Thus the muscles would
have been confined to the median posterior portion of the inner shell cavity (text-fig. id).
Therefore, by homology, this was probably also the position of all the other organs of
the ‘body’; and the remainder of the inner shell cavity would have formed an inner
mantle cavity (text-fig. id, I.M.C.).

The lophophore. The ‘hard parts’ of Prorichthofenia give no indications of the form,
position, or even existence, of the lophophore. If it was present, and occupied a position

454

PALAEONTOLOGY, VOLUME 3

homologous with its position in living brachiopods, it would have been attached to
the anterior body wall below the hinge, and either suspended freely in the inner mantle
cavity or perhaps attached to the lower surface of the dorsal valve (text-fig. \d; here its
size and form are purely diagrammatic).

FUNCTIONAL MORPHOLOGY OF THE DORSAL VALVE
The valve as a protective structure. It is generally accepted that one important function
of the valves of a brachiopod is that of protecting the soft tissues and organs of the animal
from the effects of harmful agents in the external environment. This function can only be
fulfilled efficiently if the valves (a) are wholly external to the soft parts, and ( b ) have
edges that are identical in form. Then, when the valves are closed, the edges will fit
tightly together, and will seal all the soft parts from direct contact with the external
environment. This ‘paradigmatic’ specification is fulfilled accurately in the shells of all
living articulate brachiopods. But in Prorichthofenia the dorsal valve is more or less
deeply recessed within the ventral valve. Even when it was closed, a considerable area of
mantle tissue would still have been unprotected and therefore exposed to the action of
predators and of harmful solutes and suspensions. It is true that the tight seal around
the edge of the closed dorsal valve might have prevented them from penetrating to the
‘body’ of the animal; but the valve is so thin that even when closed it would have given
little protection against larger predators. Thus it could not have provided efficient pro-
tection against any type of harmful external agent.

Richthofeniids have sometimes been compared with other operculate coralloid
organisms, such as goniophyllid rugose corals (e.g. Calceola), other aberrant brachio-
pods (e.g. Scaccinella , Gemmellaroia ), and the more highly modified rudist lamelli-
branchs (e.g. Radiolites, Hippurites). As Cloud (1948, p. 327) has suggested, the com-
mon coralloid form may be due to ecological convergence. But in the form of the
‘operculum’ the richthofeniids differ markedly from the other organisms. In the latter,
the ‘operculum’ lies across the aperture of the ‘coralloid cone’ and is relatively thick
and robust; and it could therefore have functioned efficiently as a protective structure.

EXPLANATION OF PLATE 72

Figs. 1-14. Prorichthofenia permiana. 1, Antero-lateral view of right lateral wall; dorsal valve pre-
served half-open, against arcuate zone (E. 17128, x2). 2, Lateral view of left lateral wall; dorsal
valve broken away except near hinge (E. 17129, x2). 3, Antero-lateral view of right lateral wall;
dorsal valve broken away except near hinge; note pustular surface of outer shell cavity (E. 15581,
x 2). 4, Antero-ventral view of hinge ; dorsal valve closed (E. 15589, X 3). 5, Anterior view of hinge ;
dorsal valve half-open (with small adherent brachiopod) (E. 17128, X 3). 6, Anterior view of hinge;
dorsal valve fully open (E. 15787, X 3). 7, Dorsal view of hinge; dorsal valve closed (E. 15783, x4).
8, 9, Oblique views of hinge structure of ventral valve (dorsal valve missing) (E. 15710, E. 15778, x4).
10, Posterior view of ventral muscle scars (E. 15755, X 3). 11, Ventral view of dorsal valve (E. 15744,
x3). 12, Oblique postero-dorsal view of the same dorsal valve, showing spinules (E. 15744, x4).
13, View of anterior edge of another dorsal valve, showing fluted spines projecting from ventral
surface and spinules on dorsal surface (E. 15786, x4). 14, Dorsal view (posterior side to left) of the
same dorsal valve (hinge region broken off), to show spinules (E. 15786, x 3).

Dorsal and ventral views orientated with posterior side uppermost; anterior and posterior views
with dorsal side uppermost, unless otherwise stated, a.g., accommodation groove; a.k., articulation
knob; a.m., adductor muscle scar; a.z., arcuate zone; c.p., cardinal process; D.V., dorsal valve;
fulcral ridge; //., hinge; h.a., hinge-axis; l.p., lateral plate; s., shelf.

Palaeontology, Vol. 3.

PLATE 72

RUDWICK, Prorichthofenia

M. J. S. RUDWICK: THE PERMIAN BRACHIOPOD P RO RIC HTHO FEN I A 455

This comparison places the richthofeniids in an isolated and anomalous position. Though
it might be inferred that they were merely less well adapted than the other organisms
(invoking for a causal explanation such factors as ‘low selection pressure' or ‘phylo-
gerontism’), this interpretation is methodologically inconclusive; for the dorsal valve,
though relatively inefficient for protection, might have been highly efficient for some
other function.

The range of movement of the valve. The total range of movement of the dorsal valve can
be reconstructed from specimens in which it is preserved in different positions.

When fully closed it rests on a shelf (cf. King 1930, p. 97), which encircles the inner
surface of the ventral valve (PI. 72, figs. 1-3; PI. 73, fig. 16: s). It makes an accurate con-
tact with the shelf, so that during life the inner mantle cavity would have been tightly
sealed from the outer mantle cavity and the external environment.

It is clear that the dorsal valve was able to move through a wide angle: parts of the
walls of the cavity, and their derivatives (viz. the internal spines), lie extremely close to
the inferable course of the edge of the dorsal valve, yet could not have obstructed its
movement, (a) Parts of the lateral walls of the outer shell cavity are close to the edge of
the valve. These parts are never pustular or spiny, but smooth or marked with faint
arcuate striae. These arcuate zones ( PI. 72, figs. 1,2: a.z.) represent parts of the surface of
revolution described by the edge of the dorsal valve in rotating around the hinge axis.
(b) In P. uddeni the mesh itself lay only just above the course of the edge of the dorsal
valve (text-fig. 3 a, d). Externally the mesh appears gently domed; internally it can be
seen that this form is geometrically continuous with the arcuate zones, (c) Similarly in
P. permiana the lowest of the anterior and lateral spines are inclined at such an angle
that their lower sides were very close to the edge of the dorsal valve (text-fig. 2b, d). (d)
In some specimens of P. permiana, however, the lowest spines lack the sharp point that
normally terminates the smaller spines, and instead have a distinctive blunt termination
(cf. PL 73, figs. 12, 13). This is not due to posthumous damage; almost certainly it is the
result of resorption. The positions of these resorbed points are always very close to the
course on which the edge of the dorsal valve would have moved; so that if these spines
had not been shortened to that particular degree by resorption they would have ob-
structed the dorsal valve (text-fig. 2a, c). (At an earlier growth stage they were presum-
ably sharp-pointed and uppermost in position; but with the upward migration of the
dorsal valve during growth, they would have come to occupy a lower relative position,
in which only terminal resorption could prevent them from obstructing the dorsal valve.)

The posterior wall of the outer shell cavity sets an upper limit to the possible range of
movement of the dorsal valve. In most of the larger specimens the form of this wall
suggests very strongly that the valve habitually moved into this extreme position, and in
some specimens it is preserved there (text-fig. 3a, c ). The wall is conspicuously flattened,
and the outline of this flattened area (PI. 73, fig. 1) corresponds exactly to the outline of
the dorsal valve (if preserved) and to the outline of the shelf. In one specimen of P. per-
miana there is no flattened area, and the posterior wall bears solid spines; but all except
the uppermost spines have resorbed points (PI. 73, figs. 12, 13) which lie in a single plane
(PI. 73, fig. 7); clearly this plane was the limiting position into which the dorsal valve
could move (text-fig. 2c).

Thus no ‘hard part’ could have prevented the dorsal valve from moving through the

456 PALAEONTOLOGY, VOLUME 3

wide angle (between 60° and 90°: see text-figs. 2, 3) between the shelf and the flattened
area; and the valve is preserved in all positions between and including these limits. The

FIG. 3

text-figs. 2, 3. Median longitudinal sections of representative specimens of P. permiana (text-tig. 2)
and P. uddeni (text-fig. 3), based on camera-lucida drawings. All x T3. Reconstructed parts stippled
or shown with dashed outline. Note inferred arcuate course of anterior edge of dorsal valve. Mesh of
P. uddeni shown as row of solid dots; reconstructed parts, as open circles. All other spines shown as
projections from shell, cut off at base by white line (the spines shown are those near, but not necessarily
in, the median plane). Inferred 'resorbed points’ of spines of P. permiana shown by short arrows, sp.,
median septum of P. uddeni. (Text-fig. 2: a, E. 15787; A, E. 15785; c, E. 15786; d,E. 17136; e,E. 17135;
f E. 15583; g. E. 15778; //, E. 15756; /, E. 15746; j, E. 15591. Text-fig. 3: a, E. 17123; b, E. 15723;
c, E. 15717; d, E. 17122; e, E. 15716;/, E. 17127.)

structural relations between the dorsal valve and the outer shell cavity remain inexplic-
able except on the assumption that the dorsal valve habitually moved through this wide
angle when the animal was alive. (There is no direct evidence of the angle of movement

M. J. S. RUDWICK: THE PERMIAN BRACHIOPOD P RO RIC HTHO FEN I A 457

in specimens in which the ventral valve does not extend far above the shelf (text-figs. 2 e-j ;
3e,f); but a wide angle of opening would have been physically possible.)

The relation of the valve to the lophophore. Among living brachiopods, only in Lacazella
( Lacaze-Duthiers 1861) is the angle of opening comparable to that postulated for Pro-
richthofenia ; but unfortunately the feeding mechanism of Lacazella is still undescribed.

In the diverse lophophores of all living brachiopods that have been investigated,
steady one-directional water currents are set up by the lateral cilia on the filaments (e.g.
Atkins 1956), and the brachia are virtually immobile and non-extensible. If the lopho-
phore of Prorichthofenia was normal in this respect, it follows that, whatever the form
of the lophophore, both incurrent and excurrent streams of water would have had to

•9-

P-

JW

text-fig. 4. Hinge structure of Prorichthofenia, shown by block diagrams (cut edges of valves shown
solid black), a. Dorsal valve removed; b, dorsal valve closed; c, dorsal valve open, a.m., adductor
muscle scar; a.g., accommodation groove; a.p., articulation pit; a.z., arcuate zone (extreme posterior
part); c.a., false cardinal area, c.p., cardinal process;/./-., fulcral ridge; h.a., hinge-axis; l.p., lateral
plate; M.P., median plane; psd., false ‘pseudodeltidium’; 5., shelf.

traverse the outer mantle cavity simultaneously. In whatever parts of the cavity currents
flowed, contamination between them would have been inevitable, unless some ‘soft part’
diaphragm (analogous to the partition between the siphons of a lamellibranch) extended
vertically across the cavity to separate them from one another; but it is difficult to con-
ceive how a vertical diaphragm could have been related structurally to a dorsal valve
that moved through such a wide angle. Therefore, if the lophophore functioned nor-
mally, the recessed position of the dorsal valve would have made its filtering system
inefficient.

The nature of the hinge. In most brachiopods a wide opening of the valves is prevented
either by the proximity of the umbones behind the hinge axis, or by the tight enclosure
of the teeth between the walls of the sockets.

The structure of the hinge of Prorichthofenia (text-fig. 4; PI. 72, figs. 4-12) differs
significantly from that of other brachiopods. When the dorsal valve is preserved resting
on the shelf, the rectangular projection on its posterior side fits into the top of a vertical
recess in the posterior wall of the ventral valve. Near the posterior corners of the pro-
jection, a pair of small articulation knobs ( a.k .) fit into a pair of articulation pits (a.p.) on
the inner sides of a pair of vertical lateral plates (l.p.), which project forwards from the
posterior wall of the outer shell cavity. This clearly constitutes a spindle-like bearing.
The hinge-axis thus lay slightly in front of the posterior border of the rectangular pro-
jection. Along the same line, the dorsal valve is supported ventrally by a pair of sharp-

458

PALAEONTOLOGY, VOLUME 3

edged horizontal fulcral ridges ( fr .), one on either side of the median plane, running
parallel to the posterior wall but separated from it by a deep accommodation groove ( a.g .),
into which the posterior border of the projection sank when the dorsal valve opened.
This clearly constitutes a ‘knife-edge’ type of bearing.

This remarkable hinge structure would have had four significant mechanical properties.
(a) Both elements of the dual articulation would have permitted the dorsal valve to
rotate through a very wide angle. ( b ) The ‘knob and pit’ articulation would have pre-
vented any lateral slewing or longitudinal shearing during rotation, (c) The fulcral ridge
articulation would have enabled the hinge to withstand considerable stresses during
rotation, (d) The frictional resistance to the rotation would have been extremely slight,
because the minute articulation knobs and the sharp-edged fulcral ridges are the only
bearing surfaces between the valves.

The musculature of the valve. The positions of the muscle attachments show that the
system of muscular leverage was normal (text-fig. 1). It is clear that in every position of
the dorsal valve the line of action of the adductors would have been anterior, and that of
the divaricators posterior, to the hinge-axis. Therefore contractions of the adductors
and divaricators, respectively, would have been capable of lowering and raising the
valve through the wide angle already postulated. It is necessary, however, to consider the
nature of the resistance that the muscles would have had to overcome. The possible
sources of resistance would be the same for any brachiopod. (a) Resistance due to
friction at the bearing surfaces of the hinge. ( b ) Resistance due to gravity, (c) Resistance
due to the inertia of the dorsal valve, (d) Frictional and inertial resistance of the water
displaced by the moving valve (this would depend, above all, on the angular velocity of
the valve). In Prorichthofenia {a) would have been minimized by the structure of the

EXPLANATION OF PLATE 73

Figs. 1-6. Prorichthofenia uddeni. 1. Anterior view of flattened area (outlined) and hinge region, with
posterior attachment of mesh; cf. text-fig. 3c (E. 15717, X 2). 2, Dorsal view of mesh (broken in two
places anteriorly) (E. 17123, x 2). 3, Enlargement of left lateral part of same mesh; anomalous zone
visible at top of figure ( E. 17123, X 4). 4, Dorsal view of aperture (mesh broken away except margin-
ally), showing spines on lower surface of dorsal valve (E. 17122, X 2). 5, Enlargement of left postero-
lateral part of mesh shown in fig. 2, to show anomalous zone; normal part of mesh visible at bottom
of figure; cf. fig. 3 (E. 17123, x4). 6. Enlargement of part of ‘imperfect’ mesh (broken posteriorly)
on anterior side of aperture; (E. 15724, x4).

Figs. 7-16. P. permiana. 7, Dorsal view of aperture (cf. text-fig. 2c), to show distribution of spines
(broken on left antero-lateral sector), spinules and pustules; note ‘resorbed points’ of posterior
spines lying in a single plane (E. 15786, x 1-5). 8, The same, of another specimen (cf. text-fig. 2b)
(E. 15785, X 1-5). 9, The same, of another specimen (cf. text-fig. 2 a); note ‘resorbed points’ of right
posterior spines, in same plane as flattened area (E. 17130, xl-5). 10, 11, Enlargements of anterior
spines shown in fig. 7 ; note sharp points on unbroken spines (E. 15786, x4). 12, 1 3, Enlargements of
posterior spines of same specimen; note characteristic blunt (‘resorbed’) points (E. 15786, x4).
14, Enlargement of antero-lateral spines of specimen shown in fig. 8 (E. 15785, x4). 15, Postero-
ventral view of dorsal valve (posterior part broken away), to show spines along anterior edge (above)
and ‘resorbed’ stumps behind; cf. PI. 72, fig. II (E. 15743, X 3). 16, Oblique postero-ventral view
of inner shell cavity of specimen cut in median plane; dorsal valve nearly closed; cf. text-fig. 2d
(E. 17136, x 2).

Dorsal views orientated with posterior side uppermost; anterior and posterior views, with dorsal
side uppermost, an., anterior limit of anomalous zone; h., hinge; s., shelf; sp., median septum.

Palaeontology , Vol. 3.

PLATE 73

RU D W IC K , Pro rich thofenia

M. J. S. RUDWICK: THE PERMIAN BRACHIOPOD P RO RIC H T H O FEN I A 459

hinge; and ( b ) and (c) by the unusually thin and light dorsal valve. Therefore the
muscles would have encountered much less resistance on contraction than their counter-
parts in ‘normal’ brachiopods — excepting only the unknown factor of the resistance
of the water. The sizes of the muscle scars (PI. 72, figs. 4-6, 10, 11) show that the
muscles were certainly not reduced in size relative to those of comparable ‘normal’
brachiopods. If their strength corresponded to the demands made upon them, the
dorsal valve must habitually have moved with great rapidity.

This does not imply that the intrinsic speed of the muscles was necessarily high. The
lines of action of both sets of muscles, and especially that of the divaricators, pass close
to the hinge-axis (text-fig. 4c; PI. 72, figs. 4-6, 1 1 ). Hence even a relatively slow contrac-
tion of either set of muscles would have served to rotate the valve rapidly. It is possible
that the valve would have opened and closed with equal rapidity: observations on living
brachiopods suggest that adductors have a higher intrinsic speed than divaricators (the
valves are usually ‘snapped’ shut but opened slowly, yet the adductors are farther from
the hinge-axis than the divaricators). More significantly, the cardinal process is so
unusually close to the hinge-axis that the contraction of the divaricators would have
occurred under almost isometric conditions. (For instance, in P. penniana , specimen
E. 17128, the length of the muscle would have shortened by only about 5% of its total
length during a full opening of the dorsal valve.) This would have given maximal effi-
ciency in a rapid contraction, by reducing to a minimum the ‘viscous’ effects in the
muscle.

To postulate that the dorsal valve moved rapidly up and down thus gives a consistent
functional explanation of the thin opercular form of the valve, of the peculiar structure
of the hinge, and of the apparently powerful musculature. Powerful muscles were re-
quired to overcome the resistance of the water displaced by the valve; other sources of
resistance were minimized by the lightness of the valve and by the friction-free hinge;
yet the stress at the hinge could be borne by the robust fulcral ridges.

THE BASIC FEEDING MECHANISM

An experimental study of a moving dorsal valve. A working model of Prorichthofenia has
been constructed, and the physical effects of a rapidly moving valve have been studied
experimentally. The model (PI. 74, figs. 5, 6), constructed at natural size to avoid the need
for dimensional corrections, represents a large specimen of Prorichthofenia. The interior
is exposed to view by the removal of the left lateral wall; the cut surface (a plane parallel
to, but well distant from, the median plane) is cemented against the front wall of a per-
spex tank. Fine nylon threads are attached to the dorsal valve at the positions cor-
responding to the adductor scars and the cardinal process; they run across the inner
shell cavity, and pass through fine holes in the wall of the ventral valve at the points
corresponding to the ventral muscle scars. Thus the degrees of leverage and the lines of
action of both sets of muscles are accurately reproduced. Outside the model the threads
pass over very small pulleys to a device that simulates the contractions of the muscles.
The contracting forces are provided by a pair of elastic rubber threads, which, during
contractions at varying velocities and under varying loads, simulate closely the physical
behaviour of actual muscles. Either rubber thread can be made to contract, and so to
move the dorsal valve of the model, by extending it and then releasing a trigger. The

460

PALAEONTOLOGY, VOLUME 3

power of the contraction can be controlled by varying the degree of extension of the
thread or by use of stouter or finer threads.

The tank is filled with water. The movements of the water, when the dorsal valve is
opened or closed, are made visible by the use of a suspension of very small oil droplets
with the same density as the water (a mixture of olive oil and nitrobenzene, adjusted to
this density, is used). These droplets become visible when brightly illuminated. Light
from a photoflood lamp above the tank is passed through a condenser and through a
narrow slit, and so illuminates only those droplets that lie in a thin vertical sheet of
water. For most of the experiments the zone of illumination was arranged to coincide
with the median plane of the model. Since the dorsal valve is carved from Perspex, the
flow within the inner shell cavity is visible even when the dorsal valve is closed. The
movements of the water, as revealed by the oil droplets, were studied visually and with
cine photography. The length of exposure of the films reproduced here (PI. 74) was such
that the droplets moved far enough to appear as streaks on the film; in this way the
velocity and the direction of the water movements can be determined with ease.

Two varieties of the model were used in the experiments. One simulates the more
typical specimens of P. uddeni (cf. text-fig. 3 a-d). The other simulates specimens of P.
permiana such as those figured in text-fig. 2 a-d.

The use of models representing the ‘hard parts’ is justified by the evidence that much
of the mantle tissue was very thin and that the ‘body’ occupied only a small part of the
inner shell cavity; thus the results of the experiments would scarcely be modified by the
reconstruction of the ‘soft parts’. The impossibility of inferring the actual strength and
rates of contraction of the muscles may seem to be more serious. But in fact the basic
features of the flow patterns are constant over a wide range of angular velocities of the
dorsal valve (the experiments have covered velocities from about 100° to about 1,000°
per second).

Flow patterns in models of Prorichthofenia. Owing to the development of arcuate zone
(p. 455) on the lateral walls of the outer shell cavity, the currents caused by the dorsal
valve are mainly confined to the median region; it is therefore sufficient to describe and
figure the currents as they are seen in the median plane in the models.

When the dorsal valve moves upwards (PI. 74, figs. 1, 3, 5) the space below it expands
in volume and the space above and behind it contracts. Therefore there is a downflow
into the former and an upflow out of the latter (text-fig. 5a: d., u.). Some of the downflow

EXPLANATION OF PLATE 74

Sequences of cine film to show water currents induced in models of Prorichthofenia by rapid move-
ments of the dorsal valve. The first frame of each sequence is at the top, and is the first frame
exposed after the dorsal valve had begun to move. Photographed at 24 frames per second (exposure
Fg second per frame); except fig. 2, which was taken at 16 frames per second (exposure F second per
frame), s. marks position of shelf in each model.

Figs. 1 , 2, Model of P. uddeni ( X 1 ) ; enlargement of region near dorsal valve ; angular velocities 400° and
360° per second, respectively.

Figs. 3, 4, Model ofF. permiana (x 1); enlargement as figs. 1, 2; angular velocities 340° and 530° per
second, respectively.

Figs. 5, 6. Model of P. permiana (x0-5), showing almost whole model; angular velocities 360° and
540° per second, respectively. (The threads representing the muscles can be seen crossing the inner
shell cavity, and the divaricator thread is also visible outside the model.)

Palaeontology, Vol. 3.

PLATE 74

RUDWICK, Water currents induced in models of Prorichthofenia

M. J. S. RUDWICK: THE PERMIAN BRACHIOPOD P RO RICH THO FENIA 461

is derived from the upflow by an overflow eurrent (or.) which passes over the anterior
edge of the dorsal valve. The remainder of the downflow is derived from an inflow (i.)
over the anterior lip of the ventral valve; and the remainder of the upflow passes into an
outflow (o.) over the posterior lip. Even at low angular velocities, a trailing eddy ( t .) de-
velops against the lower surface of the dorsal valve. After the valve has come to rest the
downflow continues, though with diminishing velocity, to flush out the shell cavity,
emerging posteriorly as an upflow (text-fig. 5 b, PI. 74, figs. 1 e,f; 3 /). At the same time
(unless the dorsal valve has moved very slowly) the trailing eddy moves upwards with
the upflow, away from the shell.

When the dorsal valve moves downwards (PI. 74, figs. 2, 4, 6) the currents are simply
reversed (text-fig. 5c). After the valve has come to rest on the shelf, the downflow con-

text-fig. 5. Water currents in model of Prorichthofenia, shown by diagrammatic longitudinal sections
of shell, with dorsal valve opening (a), fully open ( b ), closing (c), and fully closed (d); cf. PI. 74. This
diagram also shows (by dotted lines) the inferred courses of suspended particles heavier than water.
d., downflow; i., inflow; o., outflow; ov., overflow; t., trailing eddy; «., upflow; minor eddies omitted.

tinues to flush out the outer shell cavity (text-fig. 5 d; PI. 74, figs. 2 d-f; 4c,/), though with
diminishing velocity ; and the trailing eddy moves away from the shell (PI. 74, figs. 6 d-f).
It is important to note that during this movement the water below the level of the shelf
is never disturbed.

When the valve moves with relatively high angular velocity (more than about 500° per
sec.) turbulence develops in the currents that flush out the shell cavity after the valve has
come to rest.

The larger specimens of the two species differ in the total angular range of the dorsal
valve and in the form of the outer shell cavity. This causes some differences in the flow
patterns. In P. permiana the shelf is well below the anterior lip of the outer shell cavity,
whereas in P. uddeni it is very near the lip (text-figs. 2, 3). Therefore in the lower part of
its total range of movement the gap between the edge of the dorsal valve and the anterior
wall of the outer shell cavity is much narrower in P. permiana than in P. uddeni, and is
differently orientated. Hence, as the dorsal valve of the model rises, the downflow is more
prolonged in P. permiana than in P. uddeni ; and it sweeps down to the floor of the inner
shell-cavity in P. permiana, but along the plane of the shelf in P. uddeni (PI. 74).

The functional significance of the flow patterns. The experiments show that a rapidly
moving dorsal valve would have created in the water some distinctive patterns of power-
ful currents. If food particles of any kind were suspended in the water, these flow pat-
terns could have been the basis of a mechanism of food collection.

A. In its simplest form, such a mechanism would depend on the fact that when the

B 6612 h h

462

PALAEONTOLOGY, VOLUME 3

dorsal valve of the model of P. permiana opens, the inner shell cavity is flushed out by
the downflow, whereas the closure of the valve does not disturb this region at all. There-
fore, if the dorsal valve of the living P. permiana was opened rapidly and then closed
again, some particles would have been swept down into the inner mantle cavity as the
dorsal valve rose, and then trapped there when it returned to the shelf (this could not
have occurred in adult P. uddeni). The downflow, as the dorsal valve rose, would have
been much more powerful than the inhalant currents of living brachiopods, and would
have been capable of sucking relatively large particles into the inner mantle cavity.

This mechanism would be closely analogous to that of the living Septibranchs (Yonge
1928), a group of lamellibranchs which have lost the filter-feeding mechanism of food
collection. A muscular septum, which can be raised or lowered, divides the septibranch
mantle cavity into a dorsal and a ventral chamber. The feeding cycle begins when the
septum is lowered slowly, the water in the ventral chamber being transferred to the
dorsal through the pores in the septum. The pores are then closed and the septum is
raised suddenly. This causes a powerful inflow of water through the inhalant siphon into
the ventral chamber, and simultaneously a powerful ejection of water from the dorsal
chamber through the exhalant siphon. The force and rapidity of the flow are such that
comparatively large fragments of detritus and small animals are sucked into the ventral
chamber. In this process the muscular septum and ventral and dorsal chambers are,
respectively, the analogues of the dorsal valve and the inner and outer mantle cavities of
Prorichthofenia.

Once the particles were trapped in the inner mantle cavity there are two possible
mechanisms of collection, (a) If a special food-collecting organ (e.g. the lophophore)
existed in the cavity, it might have been able to strain the particles out of the water and
transport them to the mouth (as, in the Septibranchs, they are transported by the palps).
If the filaments on the lophophore, like those of living brachiopods, bore cilia and mucus
cells on their frontal surfaces, and were swept through the inner mantle cavity, they
might have captured and transported small particles in the normal way. Or if they were
flexible and muscular, like those of living brachiopods, they might have been adapted to
seize and hold larger particles. But despite these possibilities, the circumstantial evidence
presented in the remainder of this paper strongly suggests that the lophophore was either
lost completely (cf. text-figs. 7, 8) or at least reduced to a subordinate function. ( b ) If
there was no collecting organ (i.e. if the lophophore was lost) the particles could only
have been transported by the mantle tissue lining the cavity.

But first they would have had to settle on the surface of the tissue. The microscopic
particles utilized by normal filter-feeding animals are so small or so light that they re-
main in suspension almost indefinitely. But larger fragments of detritus, having been
kept in suspension externally by current or wave action, would have settled on to the
floor of the inner mantle cavity soon after the dorsal valve had closed. Small free-swim-
ming animals, which might also have been trapped in the cavity, would have settled out
only if they were first narcotized or poisoned, for which process the tightly sealed inner
mantle cavity would have been highly effective. In living brachiopods the mantle sur-
faces are ciliated and secrete mucus. But the cilia invariably function as a rejection
mechanism, and transport particles towards the edge of the mantle (Orton 1914;
Richards 1952; Chuang 1956, &c.). Nevertheless, it is possible that on parts, at least, of
the mantle surfaces of Prorichthofenia the cilia might have been orientated in the oppo-

M. J. S. RUDWICK: THE PERMIAN BRACHIOPOD P RO R1C H THO FEN I A 463

site direction and have been able to transport particles to the mouth. Such ‘acceptance
tracts’ of cilia occur, for example, on the mantle of Lucina, which has adopted a some-
what comparable means of particle capture (Allen 1958). Alternatively, it is possible
that the mantle cilia were capable of reversal, and thereby were able to transport particles
either towards the mouth or towards the mantle edges according to circumstances:
though reversal is not known in the mantle cilia of living brachiopods, it occurs in the
frontal cilia of many species (Atkins 1958, p. 576).

B. If the mantle surfaces functioned as areas for the collection and transport of particles,
the simplest form (A) of the mechanism could have been modified significantly. For
many particles would have been thrown centrifugally on to the mantle from the strongly
curved parts of the currents, as shown by the dotted lines in text-fig. 5. The rate at which
this would have occurred would have depended (a) on the rotational velocity of the cur-
rents, and (b) on the size, form, and density of the particles present. The greater the
angular velocity of the dorsal valve, the more efficient the method would have been ;
and, as in (A), it would have been most effective with relatively large and heavy particles
of detritus or free-swimming animals, rather than with planktonic micro-organisms.
But particles striking the mantle surface would have had to be retained there. This
would have been possible if the mantle tissue secreted mucus, but would have been aided
(especially if the flow was turbulent) by a rugose or papillose mantle surface. It is prob-
able that the surfaces of the outer mantle cavity, at least, had this character; for except
on the arcuate zones the surfaces of the outer shell cavity are conspicuously covered
with small pustules (PI. 72, fig. 3; PI. 73, figs. 7-13) or even spinules (PI. 72, figs. 12-14).

This reconstruction emphasizes the importance of the outer mantle cavity as a food-
collecting area; and provides a functional explanation of the recessed dorsal valve, the
permanently exposed mantle tissue, and the pustular surfaces of the shell.

Both (A) and (B) imply that Prorichthofenia had evolved a feeding mechanism which
was fundamentally different from that of all living brachiopods, both in the mode of
creation of water currents and in the type of food particle utilized.

THE FUNCTIONS OF THE INTERNAL SPINES
Spines on the ventral valve. A. P. uddeni. On most of the larger specimens of P. itddeni
(text-fig. 3 a-d) the ventral spines are so fully branched and anastomosed that they form
a continuous mesh across the outer shell cavity. The simplest interpretation of this mesh
is that its function was protective (cf. Stehli 1954, p. 286). Whatever the feeding mech-
anism within the shell, a mesh could have prevented the entry of large and possibly
harmful ‘particles’ (e.g. either inert particles of debris or actively moving animals). This
interpretation can be tested by comparing the actual form of the mesh with the paradigm.

A paradigmatic mesh would have the following characters, (a) The bars must be stout
enough to withstand the stresses to which the mesh is subjected, yet slender enough to
minimize the reduction in the effective area of the aperture and the frictional resistance
to the flow. These conflicting demands are most effectively reconciled if the bars are
elongated in cross-section in the direction of flow, (b) The spaces between the bars must
be uniform in effective size (i.e. in the size of the largest particles that could pass through
them), (c) The mesh must cover the whole aperture, leaving no part unprotected.

This specification does not set a standard far beyond attainment in ‘natural’ meshes.

464

PALAEONTOLOGY, VOLUME 3

For example, the osculum of the Recent sponge EuplecteJla is covered by a ‘sieve-plate'
(Ijima 1901), in which the bars are fairly uniformly slender, and usually elongated in
cross-section; the spaces between them though irregularly polygonal in form, are sub-
equal in effective size; and the mesh gives a uniform degree of protection to all parts of
the osculum (text-fig. 6a).

The same specification can be compared with the mesh of P. uddeni (PI. 73, figs. 2, 3).
(a) The bars are fairly uniformly slender, and usually elongated in cross-section. ( b ) The
spaces between the bars appear to be rather unequal in size. But this is a direct corollary
of the mode of growth of the mesh. It was evidently not formed once and for all at a
final stage of growth, but developed gradually during ontogeny. It must have migrated

text-fig. 6. Oscula of euplectellid sponges, to show approximations to paradigmatic mesh (a) and
grille ( b ). a, Euplectella sp., X 1 -5 (from specimen in Sedgwick Museum) ; b, Regadrella komeyamai, X 1

(after Ijima 1901, pi. 9, fig. 3).

upwards, as the ventral valve grew in height, by continuous accretion on the upper
surface of each bar and simultaneous resorption on its lower surface. Unless the size of
the largest particles tolerated increased at precisely the same rate as the size of the whole
aperture, the mesh would have had to increase in relative complexity during ontogeny.
Clearly this was in fact occurring : for many of the larger spaces in the mesh are partially
subdivided by projections (PI. 73, fig. 3). Thus exact uniformity in the sizes of the spaces
was unattainable. Nevertheless, they were moderately uniform: in the mesh figured in
PI. 73, figs. 2, 3, they are quite closely grouped about a norm corresponding to a spherical
particle of diameter 0-7 mm.; the minimum is 04 mm. and the maximumis 0-9 mm.
(c) The mesh normally covers the whole of the aperture of the outer shell cavity. But near
the posterior border there is an ‘anomalous zone’: here most of the bars are much
stouter than in the rest of the mesh; they are often angular in cross-section; and the
spaces between them are much more irregular in size (PI. 73, figs. 2, 5). (This part of the
mesh was not included in the measurements summarized above ; the spaces there range up
to 1 -3 mm. in effective width.) However, on comparing the position of the anomalous
zone with the water-currents in the model, it is clear that no large particle could have
penetrated through it to the inner mantle cavity. As the dorsal valve rose, no inward-
flowing current would have traversed it at any phase (in the final phase it is traversed by
the recoil of the trailing eddy (text-fig. 7c, d)). As the dorsal valve was lowered, it was
traversed by an inward-flowing current (text-fig. 7f-j; PI. 74, fig. 2). But any large
particles that entered the outer mantle cavity during this phase (text-fig. li, j ) could

M. J. S. RUDWICK: THE PERMIAN BRACHIOPOD PRO R/CHTHOFENIA 465

never have penetrated to the inner mantle cavity. For, as the valve rose again, the close
proximity of the mesh to the edge of the valve would have prevented them from passing
over the edge in the overflow; and at the end of the movement they would have been
expelled through the anomalous zone (cf. text-fig. la-d). Thus in conjunction with the
postulated feeding mechanism the deficient degree of protection in the anomalous zone
would have been immaterial. Its chief function may have been to strengthen the mesh
by anchoring it posteriorly to the wall of the outer shell cavity. In some specimens of
P. uddeni the mesh apparently did not extend into the area corresponding to the anoma-
lous zone, for the posterior wall of the outer shell cavity extends only a little above the
hinge, and the mesh was not attached to it ; but in terms of protection such specimens
would have been no less efficient than the more typical specimens with a complete mesh.

The mesh of P. uddeni thus approximates closely to the paradigm of a protective
mesh; and may therefore be interpreted with confidence as a protective device that
served to exclude from the inner mantle cavity all particles larger than a certain critical
size. It would have been "self-cleansing’; for a particle caught by it out of the downflow
as the dorsal valve rose would have been propelled away from it by the upflow as the
dorsal valve closed again (text-fig. 7).

B. P. permiana. Ideally, any aperture can be protected with equal efficiency against
the entry of large particles (assuming they are spherical) either by a mesh or by a
grille. The ventral spines of P. permiana cover the aperture in rather the same manner as
those of P. uddeni , and this suggests that they might have fulfilled the same function
by actualizing the alternative basic "design’.

A paradigmatic protective grille would have the following characters, {a) The bars
must be strong yet slender (as in a mesh), (b) They must all lie in a single row (i.e. a
"thicket’ of bars would increase the resistance but not the degree of protection), (c) The
spaces between them must be uniform in effective width, (d) They must cover the whole
aperture. For a circular aperture with radially arranged bars, the most efficient arrange-
ment would be one of intercalated "cycles’ of bars of differentiated lengths (rather like
the septa of a Fiexacoral); the aperture would then be given complete protection with a
minimum total length of bars and therefore minimal resistance to flow.

As before, this specification is not far beyond attainment in "natural’ grilles. The
osculum of Regadrella komeyamai , a Recent sponge closely related to Euplecte/la, is
covered by a grille (the "corona’: Ijima 1901, pp. 259—60) that approximates closely to
the paradigm: the bars are slender, and are arranged in a single row, and collectively
cover the whole aperture; and the spaces between them are subequal in width (text-fig.
6b). (Since they project upwards, differentiated ‘cycles’ are not strongly developed.)

The same specification can be compared with the spines of P. permiana (PI. 73, figs. 7-
14). (a) Many of the spines are very stout, relative to the size of the whole aperture and to
the width of the spaces between the spines. Moreover, they are generally circular in
cross-section, and they are ‘roughened’ with sharp longitudinal ridges, (b) They are
placed in an irregular ‘thicket’, (c) The spaces between them are fairly uniform in
width. They are arranged roughly in intercalated ‘cycles’: the longest spines converge
towards the centre of the aperture, and the relatively broad spaces between their bases
are filled by intercalated shorter spines, (d) They cover the whole aperture, except for
a posterior space (not due to imperfect preservation) corresponding in position to the

466

PALAEONTOLOGY, VOLUME 3

anomalous zone in the mesh of P. uddeni and explicable in the same manner (PI. 73,
figs. 8, 9 ; text-fig. 2 a-d).

FIG. 7

text-figs. 7, 8. Reconstructions of feeding mechanisms of P. uddeni (text-fig. 7) and P. permiana
(text-fig. 8); shown by five successive phases (a-e) of upward movement of dorsal valve and five
phases (f-j) of downward movement. Suspended particles shown by irregular stippling; direction
and relative velocity of particles shown by ‘tails' behind dots. Visceral cavity shown by close regular
stippling. A few ‘large’ particles (above critical size of mesh) are shown in text-fig. 7.

M. J. S. RUDWICK: THE PERMIAN BRACHIOPOD PRO RICHTHOFENIA 467

Points (c) and (d) indicate that the spines would have been fairly efficient in preventing
the entry of large particles. But points (a) and ( b ) indicate that they would have given
great resistance to the flow of the currents (especially if the currents were rapid) and
would have greatly reduced the effective area of the aperture. Any grille must represent
a compromise between protection and obstruction, but these spines are not an effective
compromise ; yet a comparison with those of P. uddeni shows that the inefficiency is not
due to any limitation inherent in the material utilized. Therefore the grille should not be
judged merely ‘less efficient’ than the mesh, for its obstructive character might have
been related to another function, coexisting with the function of protection.

*

a.

b.

c

text-fig. 9. Morphology of grooved spines of P. permiana. a. Cross-section of a spine, x 10 (from
a photograph of the end of a broken spine on E. 15786). b. Cross-section of grooved cylinder used in
hydrodynamical experiments, c. Spine with anastomosing grooves, X 10. d. Branched spine, X 10;
arrows in grooves of these spines indicate inferred course of captured particles towards base of spine
(these spines are figured in PI. 73, figs. 11,13 respectively).

If some of the suspended particles collided with the spines while traversing the grille,
the grille might have acted as a supplementary food-collecting device. This would be an
extension of the mechanism of collection by the mantle-surfaces, as postulated above : for
the spines must have been sheathed with outgrowths of the mantle tissue (text-fig. Id).
When a non-turbulent fluid flows past a cylindrical obstacle, the rate at which suspended
particles collide with the cylinder is always less than if they were undeflected by the
laminar flow of the fluid around it. The ratio between these rates of collision is the
‘collision efficiency’ (see East and Marshall 1954, pp. 31-32). For a given fluid, this is
a complex function of the diameter of the cylinder, the size and density of the particles,
and the velocity of flow. The rate of collision is highest when (a) the particles are re-
latively large and dense; ( b ) the velocity of flow is high; and (c) the diameter of the
cylinder is relatively great. Hence the paradigm for a food-collecting grille would have
the following characters: {a) the spines would be relatively stout; ( b ) they would be
closely spaced and abundant ; and (c) they would be sited where the currents would be
most rapid and most prolonged. The grille of P. permiana approximates closely to this
paradigm. The stoutness, close spacing, and thicket-like arrangement of the spines have
been noted already. In addition, they are most abundant along the anterior side of the

468

PALAEONTOLOGY, VOLUME 3

aperture, which in the model is traversed by long-continued rapid currents ; they are less
abundant towards the centre, where in the model the currents are weaker; and they are
moderately abundant along the posterior side, where in the model there are fairly
powerful currents (PI. 73, figs. 7-9 ; PI. 74, figs. 3-4). If the dorsal valve moved sufficiently
rapidly for the currents to become turbulent, the efficiency of this particle-capturing
mechanism would have been enhanced still further.

The spines are marked with conspicuous, sharp, longitudinal ridges separated by
rounded grooves (PI. 73, figs. 10-14; text-fig. 9). If they were sheathed with uniformly
thin mantle tissue, the surface of the tissue would have been marked by similar ridges and
grooves. Experiments with a fluted cylinder (text-fig. 9b) show that the effect of such
ridges is to increase the eddying in the wake of the cylinder at moderate Reynolds
Numbers, and to lower the critical velocity at which the wake becomes turbulent.
Possibly the collecting efficiency of the spines might have been increased thereby; but
this is unlikely to represent the true function of the ridges, because (a) the scale involved
is such that even with rapid currents the Reynolds Numbers would have been low, and
( b ) the ridges are no coarser on the stout than on the slender spines (PI. 73, figs. 10-13).
This uniformity in the width of the grooves suggests a different function. Even after
entanglement in mucus, particles that had been captured by a spine would have been
liable to become detached again by the force of the strong currents flowing past. But if
the ciliary tracts which transported them to the base of the spine were sheltered within
the longitudinal grooves, this risk would have been lessened. This reconstruction is sup-
ported by the fact that the grooves, but not the ridges, are invariably continuous to the
base of the spine (text-fig. 9c, d). It implies that the spines served to collect relatively
small particles, for the grooves average only 0- 1 5 mm. in breadth.

Thus every character that would have made the spines inefficient merely as a protective
device would have made them highly efficient as a device combining the function of
protection with that of collecting particles.

Spines on the dorsal valve. The dorsal spines project from the lower surface of the dorsal
valve, and lie near its anterior edge (text-figs. 2 a-d; 3a, d). Those of P. uddeni are very
slender, fairly regularly spaced, and arranged somewhat in a ‘thicket’ (PI. 73, fig. 4);
those of P. permiana are stouter, grooved, and arranged less regularly (PI. 72, fig. 13,
PL 73, fig. 15). They cannot have had a protective function, for they are enclosed within
the mesh or grille and could have encountered particles smaller than the critical size
transmitted by the protective structure (on no specimens do they strictly ‘interlock’ with
the ventral spines; cf. King 1930, p. 97 ; Stehli 1954, p. 286). At every phase of the move-
ment of the dorsal valve, they would have stood in the path of the strong outward-
flowing currents (text-figs. 7, 8; PI. 74). If suspended particles collided with them, they
could have acted as another supplementary food-collecting device. They are in a para-
digmatic position for this function. If they had been nearer the edge of the dorsal valve
they would have interfered with the shelf (text-fig. 2 a-d; 3a, d); if they had been any
farther from the edge they would have failed to intercept the main force of the currents.
But since the dorsal valve grew by accretion, the optimal position of one growth stage
would have become an inefficient position (i.e. too far from the edge) at a later stage.
The distribution of the spines was apparently kept optimal during ontogeny by the re-
sorption of the more posterior spines and by the formation of new spines nearer the

M. J. S. RUDWICK: THE PERMIAN BRACHIOPOD PRO RIC H THO FEN1A 469

edge; for there is a series of smooth knobs, which appear to be the stumps of resorbed
spines, behind the actual spines (PI. 72, fig. 11 ; PI. 73, fig. 15).

The specific differences in the form and arrangement of the spines possibly relate to a
difference in the size of particles collected. Those of P. permiana, being similar to the
ventral spines, would probably have collected relatively small particles in their grooves.
The more slender and uniformly spaced spines of P. uddeni suggest the collection of
larger particles, in the size-range comparable to the width of the spaces between the
spines (on the specimen figured in PI. 73, fig. 4, the spines are about 0-4 mm. apart,
whereas the mesh would have admitted particles up to a spherical size of 1 -0 mm.).

THE DEVELOPMENT OF THE FEEDING MECHANISM
Development during ontogeny. The smallest specimens in the collection have a thin
recessed dorsal valve hinged in the same manner as larger specimens. Therefore rapid
movements of the valve were probably habitual from an early stage in ontogeny, and
food particles could have been captured either by being trapped below the dorsal valve
or by impinging directly on the mantle surfaces. Then the gradual development of
spines later in ontogeny must represent the acquisition of (a) a means of protection that
was altogether lacking in earlier stages, and ( b ) a supplementary means of collecting
particles.

(a) When the dorsal valve opened, ‘ harmfully large particles ’ would have been sucked
downwards only if the power of the downflow overcame their inertia or their ability to
escape (if they were actively swimming animals). The water currents induced by the
movement of a small dorsal valve would have been much less powerful than those in-
duced by a larger, even if the angular velocity was the same. Therefore during the growth
of an individual increasingly large particles would have become liable to be sucked into
the inner mantle cavity; and the risk from harmfully large particles would have in-
creased progressively. Hence a protective device across the aperture might have been
unnecessary in the early stages of growth but increasingly advantageous in later stages.
Moreover, since the downflow in the models is most rapid near the anterior side, the
need for protection would have occurred in that region first, and would have spread
over the rest of the aperture only in later stages. This corresponds to the mode of forma-
tion of the mesh and the grille, for in both species the spines developed first from the
anterior wall of the outer shell cavity, and only gradually spread across the remainder
of the aperture. At the same time the quality of the protection improved ( PI. 73, cf. fig. 6
with fig. 3).

( b ) The rapidity of the currents near the anterior side would also have made that area
the most suitable for spines that served to capture food particles; and, as noted above,
this is the area in which the first spines appear. In P. permiana, particle capture on the
ventral spines (supplemented by the dorsal spines) probably became far more important
in adult specimens than the original means of capture. In P. uddeni these original means
probably became altogether inoperable, owing to the gradual changes in the orienta-
tion of the shelf and in the form of the outer shell cavity; and therefore, since the
ventral spines became modified into a primarily protective structure, the dorsal spines
must have become the major food-collecting device in adult specimens of this species.

The morphogenesis of the two species thus seems to reflect a gradual divergence into

470

PALAEONTOLOGY, VOLUME 3

two varieties of the feeding mechanism. The first-formed spines functioned, perhaps,
both for protection and for food collection (it may be noted that the paradigms have
some features in common). But in P. uddeni the ventral spines became modified into
a primarily (if not exclusively) protective device, and the dorsal spines developed into
the major food-collecting device; whereas in P. permiana both sets of spines developed
into food-collecting devices, though the ventral spines continued to confer some degree
of protection. It is also possible that there was a divergence in the size-range of food-
particles utilized. This suggested functional differentiation between the two species may
perhaps be confirmed when preliminary accounts of the palaeoecology of Prorichtho fenia
(Newell et al. 1953) have been supplemented by more detailed studies.

Development during phylogeny. The origin of the richthofeniids is obscure, even in purely
morphological terms. Until the morphology of other, less aberrant, productoids has
been fully analysed in functional terms, it will be impossible to determine whether the
feeding mechanism of Prorichthofenia represents an evolutionary novelty or merely a
development of a mechanism already utilized by some of its forerunners.

The mesh of an unnamed species of Prorichthofenia (Newell et al. 1953, pi. 21, fig. 33;
Stehli 1954, p. 286) approximates even more closely to the protective paradigm than that
of P. uddeni. This species occurs at a higher stratigraphical level than P. uddeni, and may
possibly indicate some evolutionary progression in the development of the feeding
mechanism.

The inferred efficiency of this feeding mechanism cannot properly be used as evidence
for any theory of the origin of adaptation. But any causal explanation that interprets the
bizarre morphology of richthofeniids as non-functional or detrimental, as "phylo-
gerontic’ explanations have commonly done, is definitely unacceptable. Similarly there
is no warrant for ‘phylogerontic’ explanations of the rather sudden extinction of the
family towards the end of the Permian.

Acknowledgements. I would like to thank Professor O. M. B. Bulman, F.R.S., Professor C. F. A.
Pantin, F.R.S., Mr. N. F. Hughes, and Dr. K. A. Joysey, who read the typescript of this paper and
suggested many improvements; Sir G. I. Taylor, F.R.S., and Dr. P. G. Saffmann, for discussion of
hydrodynamical aspects of the work; Mr. A. F. Huxley, F.R.S., for discussion of physiological aspects;
Dr. G. A. Cooper, for advice on taxonomy; Professor Sir James Gray, F.R.S., for giving me facilities
for cine photography; Dr. D. Atkins, for the gift of specimens of living brachiopods; and Mr. A. G.
Brighton, for access to specimens in the Sedgwick Museum.

REFERENCES

allen, j. A. 1958. On the basic form and adaptations to habitat in the Lucinacea (Eulamellibranchia).

Phil. Trans. Roy. Soc. (B), 241, 421-84, pi. 18.

Atkins, d. 1956. Ciliary feeding mechanisms of brachiopods. Nature, 177,706-7.

1958. A new species and genus of Kraussinidae (Brachiopoda) with a note on feeding. Proc.

Zoo!. Soc. Load. 131, 559-81.

beedham, g. e. 1958. Observations on the mantle of the Lamellibranchia. Quart. J. micr. Sci. 99,
181-97.

chuang, s. h. 1956. The ciliary feeding mechanisms of Lingula unguis (L.) (Brachiopoda). Proc. Zool.
Soc. Lond. 127, 167-89.

cloud, p. e. 1948. Some problems and patterns of evolution exemplified by fossil invertebrates.
Evolution, 2, 322-50.

M. J. S. RUDWICK: THE PERMIAN BRACHIOPOD P RO RIC H T HO FEN I A 471

cooper, G. A. 1950. Permian faunas of the Glass Mountains, Texas, and their environment. Trans.
N.Y. Acad. Sci. 12,80-81.

east, t. w. R. and marshall, j. s. 1954. Turbulence in clouds as a factor in precipitation. Quart. J. R.
met. Soc. 80, 26-47.

ijima, i. 1901. Studies on the Hexactinellida. Contribution I. (Euplectellidae). J. Coll. Sci. Tokyo, 15,
1-300, pi. 1-14.

king, r. e. 1930. The geology of the Glass Mountains, Texas. Part II. Univ. Texas Bull., No. 3042.
lacaze-duthiers, h. de. 1861. Histoire naturelle des Brachiopodes vivants de la Mediterranee. Ann.
Sci. nat. 15, (4), 259-330, pi. 1-5.

NEWELL, N. D., RIGBY, J. K., FISCHER, A. G., WHITEMAN, A. J., HICKOX, J. E., and BRADLEY, J. S. 1953. The
Permian Reef Complex of the Guadalupe Mountains Region, Texas and New Mexico. San Francisco.
orton, j. h. 1914. On ciliary mechanisms in Brachiopods and some Polychaetes, with a comparison
of the ciliary mechanisms on the gills of Molluscs, Protochordata, Brachiopods, and Cryptocepha-
lous Polychaetes, and an Account of the Endostyle of Crepidula and its Allies. J. Mar. biol. 4s.s.
U.K., n.s. 10,283-311.

richards, j. r. 1952. The ciliary feeding mechanism of Neothvris lenticularis (Deshayes). J. Morph.
90, 65-91.

rudwick, m. j. s. 1959. The growth and form of brachiopod shells. Geol. Mag. 96, 1-24.
stehli, f. G. 1954. Lower Leonardian Brachiopoda of the Sierra Diablo. Bull. Amer. Mus. nat. Hist.
105, 257-358, pi. 17-27.

1956. Notes on Oldhaminid Brachiopods. J. Paleont. 30, 305-13, pi. 41-42.

waagen, w. 1884-7. Salt Range Fossils: Productus Limestone Fossils. Palaeont. Indica (13), 1.
williams, a. 1953. The morphology and classification of the Oldhaminid Brachiopods. J. Wash.
Acad. Sci. 43, 279-87.

1956. The calcareous shell of the Brachiopoda and its importance to their classification. Biol.

Rev. 31,243-87.

yonge, c. m. 1928. The structure and function of the organs of feeding and digestion in the Septi-
branchs Cuspidaria and Poromya. Phil. Trans. Roy. Soc. (B), 216, 221-63, pi. 12-14.

M. J. S. RUDWICK
Sedgwick Museum,

Revised manuscript received 4 February 1960 Cambridge

ACANTHOCLYMENIA , THE SUPPOSED EARLIEST
DEVONIAN CLYMENID, IS A MANTICOCERAS

by MICHAEL R. HOUSE

Abstract. Evidence is presented to show that the supposed earliest clymenid, the Frasnian Acanthoclymenia
neapolitana of New York State, is in fact a Manticoceras. A lectotype is chosen which shows the suture and
indicates the position of the siphuncle. The family Acanthoclymenidae Schindewolf thus becomes a synonym
of the Gephuroceratidae. Revisionary comments are made concerning the origin of the clymenids.

INTRODUCTION

The records of clymenids from the Frasnian, or basal Upper Devonian, in North
America have posed particular problems for two reasons. Firstly, these records have
been taken to show that, since clymenids do not occur until the Platyclymenia Stufe of
the Famennian in Europe, the American forms must belong to a different faunal pro-
vince. Secondly, the elucidation of phylogeny in the Clymeniina has been rendered un-
certain by the supposed occurrence of complex-sutured forms at the earliest appearance
of the group. The purpose of this note is to show that the supposed Frasnian clymenids
in fact belong to the genus Manticoceras , and are in no way anomalous : thus the problem
of the origin of the Clymeniina can now be stated more clearly.

Historical survey. In 1892 John M. Clarke described as Clymenia ( Cyrtoclymenia ) Nea-
politana some ammonoids from the Lower Portage Shale (Cashaqua Shale) of Shurtleff’s
Gully, New York State. This locality is probably the one mentioned by Luther (1894,
p. 228) as in the eastern part of the town of Livonia; Foord, in the same year, reported
this record in the Geological Magazine and stressed the anomalously low horizon.

Clarke added some further details in a subsequent description of the species (1898,
p. 131), and gave several new localities for it, but these were all within the Cashaqua
Shale. In 1900 Hyatt (in Eastman-Zittel, p. 548) proposed the genus Acanthoclymenia
with C. neapolitana as the type and sole representative apparently without examining
the original material. Later Schindewolf (1934, p. 347), after examining the types, and
sectioning one, stated that he was unable to locate the siphuncle. But A. K. Miller, who
gives the fullest synonomy for the species and who reprinted Clarke’s description and
figures, stated that he ‘definitely located the siphuncle in two of the hypotypes’ and that
it was ‘dorsal and marginal in position and is in contact with the dorsal wall of the
conch or essentially so’ (Miller 1938, p. 192). Miller did not quote the museum numbers
of these specimens and gave no new illustrations of them.

Accepting Miller’s evidence, Schindewolf (1955, p. 422) erected the family Acantho-
clymenidae with Acanthoclymenia neapolitana as the only species. The stratigraphical
anomaly of the genus as a clymenid is well illustrated in the evolutionary diagram of the
Clymeniina which Schindewolf has published (1955, p. 422; 1957, p. L38).

[Palaeontology, Vol. 3, Part 4, 1961, pp. 472-6, pi. 75.]

MICHAEL R. HOUSE: AC AN THO C LY M EN I A IS A MANTICOCERAS 473

DISCUSSION

In his earliest description Clarke illustrated two specimens showing a dorsal struc-
ture which he took to be the siphuncle (1892, p. 63, figs. 3, 4). At least the second of
these (Clarke’s fig. 4) must be one of the hypotypes which satisfied Miller, for he reillus-
trated it (1938, p. 180, fig. F), and presumably would not have done so were he uncon-
vinced of its authenticity. This specimen is NYSM. 11264, and the inner, and only
complete, whorl is figured here (PI. 75, figs. 8, 9; text-fig. Id). The other specimen which
Clarke figured as showing the siphuncle is NYSM. 3625, and this specimen is also figured
here (PI. 75, figs. 1-4; text-fig. 1a, b): since this specimen shows the dorsal structure
better than any of the other types it is to be presumed that it is the other specimen men-

text-fig. 1 . Mcinticoceras neapolitanum (Clarke), a, b, Whorl section and suture based on the lecto-
type, NYSM. 3625, from the Frasnian Cashaqua Shale in Shurtleff’s Gully, New York State, c, d.
Suture and outline based on NYSM. 1 1264 from the same horizon and locality. All x 12-J.

tioned by Miller. It appears to have formed the basis for Clarke’s diagram of the mature
suture. This last specimen, NYSM. 3625, which was initially figured by Clarke (1892,
p. 63, fig. 3), is here designated the lectotype.

Evidence that the siphuncle is ventral in position. Clarke’s first account shows that he did
not see the siphuncle, for he wrote: ‘The siphonal funnel is long, conspicuously de-
veloped, and open along its inner surface. It does not appear to have extended across the
air chamber . . . and I have seen no evidence of a true siphonal tube connecting these
funnels’ (Clarke 1892, p. 59). It is clear from this that Clarke saw a deep adapical flexure
in the dorsal part of the septum and presumed this to be a siphonal funnel and hence
supposed this to show that the siphuncle was dorsal. In 1898, without presenting further
evidence, the presumption was categorically stated.

The lectotype is illustrated here with enlarged photographs which show the dorsal
region (PI. 75, figs. 2-4). This is a specimen which Clarke figured as showing a dorsal
siphuncle and is probably a specimen for which Miller made the same assertion. The
specimen has been developed slightly. The adapical flexure in the mid-dorsal part of the
septum is clearly seen (marked ‘dl’ upon the plate). Both figs. 2 and 3 show this septal
fold passing steeply down to the preceding whorl. Despite the presence of a small adher-
ing fragment, it will be seen that the fold almost reaches the wall of the preceding whorl

474

PALAEONTOLOGY, VOLUME 3

before it is truncated by the break which bounds the specimen. But both figs. 2 and 4
show the low crescent formed by the truncated extremity of the septal fold and the wall
of the preceding whorl. A small transverse slit alone marks a connexion with the sub-
sequent chamber. Clearly there is no tubular siphuncle, and none could pass through the
small slit. The siphuncle therefore cannot be dorsal in position and the adapical
flexure of the septum represents a deep mid-dorsal lobe. The siphuncle must be ventral
in position and pass through the mid-ventral lobe (marked ‘vl’ on PL 75, fig. 1).

The smaller specimen which Clarke figured as illustrating a dorsal siphuncle is re-
figured here (PI. 75, figs. 8, 9) and, although the specimen is indifferently preserved, a
ventral view shows a mid-ventral lobe continued adapically into an elongate structure
which can only represent a siphuncle, ventral in position. In the earliest stages Clarke
himself noted that the siphuncle was ventral in position.

Evidence that the suture is o/ Manticoeeras type. In the adult suture of Acanthoclymenia
as drawn by Clarke (1898, p. 133, text-fig. 105) and Schindewolf (1957, p. L39, text-fig.
4c) no saddle is shown in the wide ventral lobe. If one were added the suture would be
typical of Manticoeeras. At first Clarke stated that no saddle was present on the mid-
ventral line (1892, p. 59), but he later changed his opinion and wrote that ‘the ventral
lobe also appears to be minutely divided at its apex forming a ventral saddle’. Clarke
did not show this saddle on his suture diagram. The lectotype has been developed
slightly in the ventral region and Clarke’s later opinion has been confirmed. The saddle
lies between a mid-ventral and ventro-lateral lobe (marked ‘vl’ and ‘vll’ on PL 75,
fig. 1). Miller's statement (1938, p. 192) that he was unable to verify the existence of this
saddle may therefore be dismissed.

The adult suture now presented shows all the sutural elements typical of Manticoeeras
(text-fig. 1b). Further, the earlier stages (text-fig. lc) show a suture of Archoceras type,
so that it can be demonstrated that the ontogeny is also typical of Manticoeeras.

All the remaining specimens in the New York State Museum have been examined but
none show evidence bearing on the position of the siphuncle or contradictory evidence
on the form of the suture. In order to dispel any thought that the lectotype is atypical,
the better preserved of the syntypes which do not show the suture or siphuncle are also
illustrated here (PL 75, figs. 5-7, 10, 11). It will be seen that the proportions of coiling,
the whorl form, and the details of the ornament are similar to those of the lectotype.

THE ORIGIN OF THE CLYMENIDS

The uncertainties concerning the ancestry of the clymenids are legion. Nor does the
elimination of Acanthoclymenia solve the problem, but it emphasizes the sudden entry of
the group in the Platyclymenia Stufe of the Famennian. The earliest genera, Platy-

EXPLANATION OF PLATE 75

Figs. 1-1 1. Manticoeeras neapolitanum (Clarke). All specimens are from the Cashaqua Shale and from
Shurtleff ’s Gully, Livingstone County, N.Y., except perhaps fig. 1 1 which may have come from a
different locality. 1-4, The lectotype, NYSM. 3625. Views illustrating the dorsal structure, ‘vl’ =
ventral lobe, ‘vll’ = ventro-lateral lobe, Tl’ = lateral lobe, ‘ul’ = umbilical lobe, ‘dF = dorsal
lobe. Magnifications: l,xl0. 2,x8-4. 3,x6-2. 4, x6-l. 5-7, Syntype, NYSM. 3632. Allx5-6.
8, 9, Syntype, NYSM. 11264. Both x7. 10, Syntype, NYSM. 3629. x4. 11, Specimen figured by
Clarke, NYSM. 3631. x4.

Palaeontology, Vol. 3.

PLATE 75

HOUSE, Manticoceras neapolitanum (Clarke)

MICHAEL R. HOUSE: ACANTHOCLYMENIA IS A MANTICOCERAS

475

clymenia , Rectoclymenia, Cyrtoclymenia, and Hexaclymenia, all have very simple sutures:
in the case of the first three the suture consists merely of a wide lateral lobe and a
mid-dorsal lobe; in the case of Hexaclymenia there is a broad ventral lobe, a sub-
umbilical lobe, and a mid-dorsal lobe (Schindewolf 1923, p. 62, 1957, pp. L37 et seq.).
Forms with more complex sutures appear at higher stratigraphical levels and it is reason-
able to infer that they evolved from the early, simple-sutured stocks.

Several hypotheses have been proposed to explain the origin of the clymenids, and
in conclusion these may be briefly reviewed.

1. Monophyletic origin from goniatites. Schindewolf (1949) has argued strongly for
Arehoeeras, or a related anarcestid, as the single ancestor of the clymenids. This would
involve a quite sudden migration of the siphuncle to a dorsal position. Arehoeeras , a
genus omitted in the systematic section of the recent Treatise on invertebrate Paleontology ,
Part L, was erected by Schindewolf (1937, p. 243; also discussed by Gallwitz in 1938).
It ranges from the middle Frasnian to the lower Cheiloceras Stufe of the Famennian. It
has a simple suture consisting of a deep ventral lobe, a lateral lobe, and a mid-dorsal
lobe (essentially that of the early stages of Manticoeeras neapolitanum here figured, text-
fig. lc).

2. Polyphyletic origin from goniatites. Sobolew contended that many clymenid genera
evolved independently from goniatite genera with somewhat similar shell form and suture
pattern (Sobolew 1914, see also Schindewolf 1949, p. 198). There are serious objections
to this hypothesis. It would be highly curious for unrelated stocks to evolve a dorsal
siphuncle independently. Further, there can be little doubt Sobolew linked homoeo-
morphs. Nevertheless the evidence Sobolew presented of a slight dorsal migration of the
siphuncle in some goniatites deserves further attention.

3. Origin from nautiloids. Most early authorities up to Branco (1880) included the
clymenids among the nautiloids. But Branco’s demonstration that clymenids possess
the typical sub-globular ammonoid protoconch convinced most subsequent authors that
their affinities lie with the Ammonoidea. However, the very early stages of those
Devonian centroceratids and other nautiloids somewhat homoeomorphic with the early
clymenids are virtually unknown. So this possibility cannot be completely eliminated.

Acknowledgements. I am particularly indebted to the Commonwealth Fund of New York for the
Fellowship which enabled me to study Devonian faunas in the United States. Mr. Clinton F. Kilfoyle
has been of great help in locating specimens in the New York State Museum at Albany, where all
Clarke’s types are housed. Professor John W. Wells kindly allowed me the use of photographic and
laboratory equipment at Cornell University, N.Y.

REFERENCES

branco, w. 1880. Beitrage zur Entwicklungsgeschichte der Fossilen Cephalopoden. 2, Goniatiten,
Clymenien, Nautiliden, Belemniten und Spiruliden. Palaeontographica, 27, 12-81.
clarke, j. m. 1892. The discovery of Clymenia in the fauna of the Intumescens zone (Naples Beds)
of western New York, and its geological significance. Amer. J. Sci. (3), 43, 57-63.

1898. The Naples fauna (fauna with Manticoeeras intumescens) in western New York. N.Y.S.

Geol. Ann. Rept. 16, 29-161.

Eastman, c. r. and zittel, k. a. von. 1900. Text-book of palaeontology. 1. London.

foord, a. H. 1892. Note on the discovery of Clymenia in North America. Geol. Mag. 29, 173, 4.

476 PALAEONTOLOGY, VOLUME 3

gallwitz, h. 1938. Archoceraten aus dem unteren Oberdevon Deutschlands. Zbl. Miner. Geol.
Paldont., B, 367-83.

luther, d. d. 1894. Report on the geology of the Livonia Salt Shaft. N.Y. State Mas., Rep. 47, 217—
352.

miller, a. k. 1938. The Devonian Ammonoids of America. Geol. Soc. Amer., Spec. Pap. 14, 262 pp.
schindewolf, o. h. 1923. Entwurf einer naturlichen Systematik der Clymenoidea. Zbl. Miner. Geol.
Paldont. 23-50, 59-64.

1934. Uber eine oberdevonische Ammoneen-Fauna aus den Rocky Mountains. Neues Jb. Min.

Geol. Paldont. 72, 331-50.

1937. Zwei neue, bemerkenswerte Goniatiten-Gattungen des rheinischen Oberdevons. Jb.

Preuss. Geol. Landes, 58, 242-55.

1949. Zur Phylogenie der Clymenien (Cephalop., Ammon.). Neues Jb. Min. Geol. Paldont., B,

197-209.

1955. Zur Taxionomie und Nomenklatur der Clymenien. Ein Epilog. Ibid. 417-29.

1957. Clymeniina in R. C. Moore (Ed.). Treatise on Invertebrate Paleontology. Part L. Mollusca 4,

L37-47. Geol. Soc. Amer. & Univ. Kansas Press.
sobolew, d. 1914. Uber Clymenien und Goniatiten. Palaeont. Zeit. 1, 348-78.

M. R. HOUSE
University of Durham,
Durham City

Manuscript received 9 November 1959

THE CARBONIFEROUS RHYNCHONELLID
PUGNOIDES TRIPLEX (M‘COY)

by D. PARKINSON

Abstract. This small shell is described, its variation characteristics studied, and its relative growth features
analysed. Although the species displays a fairly wide range of variation, collections from different localities in
the Eh zone have not been found to differ significantly from each other.

Weller (1914, p. 192) describes Pugnoides as follows: ‘Shells rhynchonelliform, below
medium size, subovate in outline, with the fold and sinus well developed. Both valves
marked by rounded or subangular plications which become obsolete in the posterior
portion of the shell. Internal characters of both valves essentially as in Camarotoechia.'

The genus resembles Pugnax in its short costae ; it differs from it in internal characters,
the most obvious of which is the presence of a median septum in the brachial valve. This
feature is absent or obsolescent in Pugnax.

Pugnoides triplex (M‘Coy)

Plate 76, figs. 1-3.

Diagnosis. Small rhynchonellid, subovate in outline; width a little more than length;
median fold on brachial valve with corresponding broad sinus in pedicle valve; costae
short, angular to sub-rounded, three (occasionally two, four, or five) on fold of brachial
valve and usually three on flanks; two costae normally in sinus of pedicle valve; umbo
of pedicle valve small, pointed, and sub-erect; median septum in brachial valve; dental
lamellae in pedicle valve.

Type specimens. M‘Coy’s original figured type has not been traced. Topotypes are
preserved in the British Museum (Natural History) (PL 76, fig. 1). These are sandstone
casts which according to Davidson (1858-63, p. 104) were found in ‘yellow or reddish
sandstone forming the base of the Carboniferous system at Kildress in Tyrone’.

M‘Coy’s figure of " Atrypa triplex’ (1844, pi. 22, fig. 17) is copied by Davidson (pi. 23,
fig. 17). M‘Coy describes his species (1844, p. 157) as ‘Transversely oval, gibbous, beaks
very small, pointed, surface with nine short angular ribs which reach but half way to the
beak, front elevated with three of the ridges; the three ridges on each side slightly lower
than the mesial ones’. He states that the shell is remarkable for its three equal lobes of
three ridges each and comments on its very small size.

Costation. The costae are normally angular to sub-angular. A few specimens have been
noted with sub-rounded costae in the ventral sinus which have developed a short
median incision near the anterior margin.

Out of 218 shells examined 185 have a tri-costate dorsal fold; there are 10 specimens
with only two central costae; 17 specimens have four and 6 specimens have five.

The number of costae on the flanks is less constant, although most shells have three

[Palaeontology, Vol. 3, Part 4, 1961, pp. 477-84.]

B 6012 I i

478 PALAEONTOLOGY, VOLUME 3

on each side. The lateral costae diminish in size posteriorly and where more than three
are developed, the fourth is usually not prominent, and if there is a fifth it is short and
shows itself as little more than a wrinkling of the lateral margin of the shell.

The tricostate feature of the anterior margin of the dorsal fold develops early in
growth, though occasionally the third costa does not appear until later and sometimes
not at all. Where there is a fourth costa (PI. 76) it normally forms in adult individuals as
a bifurcation of one of the existing costae. This feature is sometimes seen in the lateral
costae, although the usual mode of increase is by addition towards the posterior lateral
margin of the shell.

Internal characters. The interior structure, so far as it has been possible to determine it,
resembles that of Camarotoechia as the genus is interpreted by Weller (1914, p. 175). The
hinge plates appear to be simple in outline. Except immediately below the dorsal umbo

text-fig. 1. Serial sections of a specimen of Pugnoides triplex of dimensions L = 9-7 mm., W = 10-8
mm., D = 8 0 mm., taken at 0-3 mm. intervals. x5.

where they appear to fuse, the hinge plates do not extend beyond the inner surfaces of the
septalial plates which connect them to the median septum of the brachial valve. These
plates enclose a V-shaped septalium. Anteriorly the median septum breaks away from
the septalial plates. The crura are apparently formed by the anterior extensions of the
inner margins of the hinge plates. The dental lamellae in the pedicle valve at first diverge
both ventrally and anteriorly; they soon become parallel ventrally, but continue to
diverge anteriorly. Faint muscle scar impressions have been seen in some individuals.

Text-fig. 1 illustrates serial sections of a specimen collected by Mr. A. Ludford from
Dx reef limestone near the Red Lion Inn, Cauldon, Staffordshire.

Occurrence. The Kildress sandstone of Co. Tyrone, Northern Ireland, which yielded
M‘Coy’s types is probably of Seminula (S2) age. (George 1958, p. 282). All the English
specimens known to the writer were collected from reef limestone of the Lower Dibuno-
phyllum Zone (Dx) in north Staffordshire and Derbyshire. Davidson records and figures
‘ Rhynchonella pleurodon var. triplex' (p. 105, pi. 23, fig. 16) from the Carboniferous
shale near Carluke, Lanarkshire. He also instances its occurrence in ‘the shales of the
upper portion of the Carboniferous limestone at Settle’, in Yorkshire.

Remarks. Davidson confused this species with young specimens of 'Rhynchonella'
pleurodon (Phillips). However, the features distinguishing these two variable forms
appear to be constant. The costae of pleurodon always extend to the posterior margin,
whilst those of triplex are confined to the anterior part of the shell. In specimens similar
in size the costae of pleurodon are narrower, closer spaced, and more numerous than
those of triplex.

Immature specimens of Pugnax pseudopugnus D. Parkinson (1954, p. 570) are some-

479

D. PARKINSON: PUGNOIDES TRIPLEX (M'COY)

times difficult to distinguish from Pugnoides triplex. Both species have a tricostate dorsal
fold, but the internal characters differ markedly. Externally the main distinguishing
feature is in the shape of the costae which in Pugnax pseitdopugnus are more angular,
deeper, and more acute than in Pugnoides triplex.

VARIATION IN BIOMETRICAL CHARACTERS
Assemblages have been analysed of measurable individuals from four localities: (1)
Weaver Hills, north Staffordshire (095462), 72 specimens; (2) Dielasma Bed, Treak

text-fig. 2. Scatter diagrams of a community of P. triplex from Weaver Hills, north Staffordshire.

Cliff, north Derbyshire (134832), 28 specimens; (3) Dowel Dale, west Derbyshire
(075676), 22 specimens; (4) Narrowdale Hill, north Staffordshire (123574), 28 speci-
mens. Except for Narrowdale the lateral extent of the exposures is only a few feet and the
vertical thickness a foot or less. Some of the Narrowdale shells were collected by Mr. A.
Ludford and the writer from one thin pocket in the reef limestone; the others, in the
British Museum (Natural History), if not from the same bed, were probably obtained
from near-by at approximately the same stratigraphical level.

The scatter diagrams (text-figs. 2, 3, 5) indicate the range of variation in the size and
shape of the shell. The measurements were made in the conventional directions. The

480

PALAEONTOLOGY, VOLUME 3

length (L) is the maximum distance in a straight line from the posterior margin at the
ventral umbo to the anterior margin ; the width ( W) is the maximum distance in a straight
line perpendicular to the length between the two lateral margins; the depth ( D ) is the
maximum distance through the shell perpendicular to the plane containing the length
and width. The thickness ( T ) is measured centrally in the same plane as the depth.
These features should be clear from the figures of PI. 76 (see also Parkinson 1954,
fig. 1, p. 563). In this particular species D is more often than not no greater than T. The
sinus is seldom deeper than half a millimetre and there is little to be gained in the statis-
tics by distinguishing between depth and thickness. D is preferred to T because it can be
more accurately measured with a dial gauge.

The range of variation, though moderately wide, is about normal for the Carboniferous
rhynchonellids. As usual, the L-W scatter is much narrower than those for D-L (not
reproduced) and D-W.

RELATIVE GROWTH

2

X

E-

O

z

UJ

The analyses were made by the reduced major axis procedure described by Kermack

and Haldane (1950). In text-fig. 2 L is
plotted against W and D on an arith-
metical grid of the seventy-two specimens
from Weaver Hills. The data for L- W were
analysed on the assumption of isometry
and the relationship was found to be L =
0-80IP+0-885. The line representing this
equation fits the data fairly well and shows
a uniform increase in length relative to
width as the shell grows. It has, of course,
to be assumed that the size of the shell is
a measure of its age. In specimens with a
width greater than 6 mm. the shell is nearly
always broader than long. If the prolonga-
tion of the growth line to the L axis is
justified it further indicates that in the early
growth stages L is on average greater than
W. At zero width the mean length — if
such were physically possible — would be
0-885 mm. But if the fifteen specimens
less than 6 mm. wide are considered sep-
arately it is found that on the average they
are equal in length and width. The calcu-
lated mean values of L and W, for W less

WIDTH

MM.

text-fig. 3. Scatter diagrams of the Weaver Hills
collection. The equations to the fitted allometric than 6 mm., curiously enough, happen to
lines (reduced major axes) are L = 1-180 IE0 876 give the same figure of 5-253 mm.

and D - 0-222 IT1'466. The inference is, on the assumption that

growth throughout is isometric, that at first L and W are equal and increase at the
same rate, but that later the width begins to increase at a greater rate than the length.

D. PARKINSON: PUGNOIDES TRIPLEX (M‘COY)

481

This suggests the probability that differential growth is allometric, i.e. that it accords
with the equation L = fiWa (see Zuckerman et al. 1950, for a full discussion), rather
than isometric (L = aW+b). The data were therefore replotted on a double logarith-
mic grid. The computed reduced major axis L = 1-180 Iff0'876 (or log L = 0-876 log Iff
+0-072) is a satisfactory fit to the scatter in text-fig. 3.

In text-fig. 4 the curves for both isometric and allometric growth are drawn on arith-
metical co-ordinates for the early growth stages, together with the line L = Iff, which is
never far from the allometric line and meets it a little below 4 mm. (Calculation by put-
ting L = Iff in the allometric equation shows the two values to coincide at 3-8 mm.)

Although an isometric relationship holds approximately for the L-W scatter, this is
not the case for depth-width as can be seen from text-fig. 2, in which a curved line
drawn centrally through the points passes through the origin of the graph. This line in
fact is derived from the allometric equation D = 0-222 Iff1'466 (or log D = 1-466 log
Iff— 0-653). The reduced major axis, which is plotted on logarithmic co-ordinates in
text-fig. 3, fits the data reasonably well.

The equation for the reduced major axis of the D-L relationship is D = 0-169 L1'673
(log D = 1-673 log L — 0-771). The scatter diagram is not reproduced in the paper.

Similar graphs have been drawn for the samples from the other three localities and
analyses carried out. Tests of significance made by the shortened method of Leigh-
Dugmore (1953) indicate that the different collections could all belong to the same
population. The individual graphs are not reproduced here, but composite scatter dia-
grams for all the 150 specimens measured are drawn in text-fig. 5. The calculated equa-
tions are log L = 0-830 log Iff+0-107 and log D = 1-248 log Iff— 0-478.

Previous work on differential growth in brachiopods (Parkinson 1952, 1954, 1959;
Prentice 1956) has shown that in many instances the values of the allometric growth
constants a and have changed at some stage during growth. In the case of the small
shell described here there are no such changes, but it is interesting to consider what
might have happened if the shell had grown to a larger size.

Inspection of text-figs. 3 and 5 shows that in each case if the two growth-lines are
produced to higher values they will meet at some point which is representative of the
width of the shell where length and depth have become equal. The most reliable estimate
is obtained by considering the data as a whole, since the differences between the four
samples are not statistically significant. The appropriate values can be obtained from the
allometric equations representative of the 150 specimens measured:

Thus

log D = log L = 0-83 log tff+0-107
= 1-248 log Iff- 0-478
log Iff (1-248-0-83) = 0-107+0-478
0-585

Hence

*°s ^-o-4i8- 140

and

logL = either (0-83 x l-40)+0-107 = 1-269

= or (1-248x1 -40) -0-478 = 1-269

D = L = antilog 1-269 = 18-6 mm.

and

Iff = antilog 1-40 = 21-5 mm.

482 PALAEONTOLOGY, VOLUME 3

The same result is obtained graphically (text-fig. 6) by finding the point of intersection
of the reduced major axes.

text-fig. 4. Comparison of the isometric and allometric reduced major axes for the early growth
stages of P. triplex. Although the departure from isometry for the L-W relation is small, allometry
gives a truer representation because for small specimens the mean values of L and W are equal.

This means that if the shell had grown to a size exceeding some 25 mm. in width the
depth (and soon the thickness) would have become greater than the length. Since no
specimens have been found where the length exceeds 11 mm. the value of 18-6 at which
D = L is hypothetical. It is also very approximate because if the Weaver Hills collection
is considered separately D = L at a value as low as 13-5 mm. The inference is that the
shell would be unlikely to grow much larger without changes in the allometric growth

D. PARKINSON: PUGNOIDES TRIPLEX (M'COY) 483

constants a and /?, otherwise the shape of old age specimens would be uncharacteristic
of the species.

These data serve to illustrate the principle discussed fully elsewhere (Parkinson 1960)

text-fig. 5. Scatter diagrams of the composite sample of P. triplex. The equations to the fitted lines
are L = 1-279 W °'830 and D = 0 333 W 1-248.

text-fig. 6. Illustrates the changing relative dimensions of the shell with continued growth. Beyond
the point of intersection of the reduced major axes the depth becomes greater than the length if the
allometric parameters remain constant in value.

that changes in the allometric growth parameters are normally to be expected in the
brachiopoda.

Acknowledgement . I am greatly indebted to Dr. H. M. Muir-Wood for helpful discussions, for
the loan of specimens from the British Museum (Natural History) and for the photographs re-
produced in PI. 76. The figured topotype, a very small sandstone cast, is, in particular, a very
difficult subject for photography. Thanks are also due to Mr. A. Ludford for specimens from the
Weaver Hills area and to Mr. H. R. Willcocks for drawing the graphs.

REFERENCES

davidson, t. 1858-63. A monograph of the British fossil Brachiopoda. Palaeontogr. Soc.
george, t. n. 1958. Lower Carboniferous palaeography of the British Isles. Proc. Yorks. Geol. Soc.
31, 227-318.

kermack, k. a. and haldane, j. b. s. 1950. Organic correlation and allometry. Biometrika, 37,
30-41.

leigh-dugmore, c. h. 1953. A rapid method of estimating the correlation coefficient from the range
of the deviations about the reduced major axis. Biometrika, 40, 218-19.
m‘ coy, f. 1 844. Synopsis of the characters of the Carboniferous Limestone fossils of Ireland. Dublin,
Parkinson, d. 1952. Allometric growth in Dielasma hastata from Treak Clift', Derbyshire. Geol. Mag.
89, 201-16.

1954. Quantitative studies of brachiopods from the Lower Carboniferous reef limestones of

England. I, Schizophoria resupinata (Martin), II, Pugnax pugnus (Martin) and P. pseudopugnus
n.sp., Ill, Pugnax acuminatus (J. Sowerby) and P. mesogonus (Phillips). J. Paleont. 28, 367-81,
563-74, 668-76.

484 PALAEONTOLOGY, VOLUME 3

Parkinson, d. 1960. Differential growth in Carboniferous brachiopoda. Proc. Geol. Assoc. 71,403-
28.

prentice, J. e. 1956. Gigant opr o ductus edelburgensis (Phillips) and related species. Proc. Yorks. Geol.
Soc. 30,229-58.

weller, s. 1914. The Mississippian brachiopoda of the Mississippi valley basin. State Geol. Surv.
Illinois Mon. 1,1-508.

zuckerman, s. and others. 1951. A discussion on the measurement of growth and form. Proc. Roy.
Soc. 137 b, 433-523.

D. PARKINSON

6 Clowes Avenue,
Southboume,

Manuscript received 20 July 1959 Bournemouth

H OEG ISPORIS, A NEW AUSTRALIAN CRETACEOUS

FORM GENUS

by ISABEL C. COOKSON

Abstract. A distinctive microspore occurring in certain Australian Cretaceous deposits is described as Hoegi-
sporis lenticulifera gen. et sp. nov. and briefly discussed.

The microspore for which the new genus Hoegisporis is herein proposed, although well
characterized and readily recognizable, does not possess those clear features by which
dispersed spores can usually be distinguished from pollen grains. It shows no sign of
either a tetrad scar or any kind of germinal aperture — Hoegisporis may well represent an
inaperturate pollen grain, but whilst its origin remains obscure the term nricrospore, in
the broad sense, seems the better application.

hoegisporis gen. nov.

Inaperturate microspore with a thin exine that is strengthened by a variable number
of prominent, lenticular thickenings around the equator. Type species. H. lenticulifera
sp. nov.

Hoegisporis lenticulifera sp. nov.

Plate 76, figs. 4-9. Holotype figs. 6, 7, Nat. Mus. Vic. P 20510.

Age and occurrence. Probably Aptian: 'Santos’ Ltd. Oodnadatta Bore, S.A., between 1,032 and
1,052 ft. Albian: Oodnadatta Bore, at 327 ft.; Moora Bore, W.A., between 86 and 170 ft.; Regan's
Ford on Moore River, W.A., Wapet’s seismic shot hole L 8 at 240 ft.; Lower Gearle Siltstone, W.A.,
Wapet's Rough Range no. 1 Bore at 2,750 ft. and Wapet’s Rough Range South no. 1 Bore, W.A.,
between 2,758 and 2,867 ft.? Upper Albian to Cenomanian: Osborne Formation, W.A., Subiaco Bore,
at 358 ft.; near Gingin, W.A., Wapet's seismic shot hole B 1 between 190 and 220 ft.; Galbraith,
N.Q. Frome Broken Hill Co.’s Wyaaba no. 1 Bore between 1,155 and 1,156 ft.; Haddon Downs, S.A.,
Bore no. 1 at 431 ft. and Bore no. 5 at 801 ft. Probably Cenomanian: Brickhouse Bore, W.A., at 1,210 ft.
Further details regarding these localities may be found in the papers by Cookson and Eisenack, and
Cookson and Dettmann cited below.

Description. Microspore always much flattened and showing the equatorial outline,
approximately circular in polar view with a ± wavy outline. Exine less than 1 /x thick,
intectate, finely (less than 1 p) and closely pilate, frequently dotted with larger club-
shaped outgrowths or clavae of variable size; equatorial exinous thickenings 6-11 in
number with a circular outline in surface view. Dimensions. Type-diameter 59 /*,
equatorial thickenings c. 7x5^ in optical section. Range-diameter 33-60ft, equatorial
thickenings 7x4/x to 12x7/x, clavae c. 1— 4/x wide, up to c. 3-5 p long.

Comments. It is probable that those examples in which the ornament consists only of
pila are specifically distinct from those in which clavate prominences are also present.
However, a considerably larger number of examples than, at present, is available will be
necessary before this question can be fully resolved.

The only described spore or pollen type with which Hoegisporis lenticulifera appears
to be at all comparable is the angiospermous species Pollenites oculis metis Thiergart

(Palaeontology, Vol. 3, Part 4, 1961, pp. 485-6, pi. 76.]

486

PALAEONTOLOGY, VOLUME 3

(1940, pi. 7, fig. 1) from the Oligocene of Germany and the apparently similar uni-
dentified pollen, from the Russian Oligocene, figured by Pokrowskoi (1956, pi. 5, fig. 26)
and reproduced herein (PI. 76, fig. 9). However, in P. oculis noctis the conspicuous
exinous thickenings which characterize this species are associated with pores whereas
those of H. lenticulifera have no apertural connexion whatsoever. Furthermore, H.
lenticulifera , unlike P. oculis noctis, has been found, with one exception (Haddon Downs
no. 1 Bore at 431 ft.), in beds in which no recognizable angiospermous pollen grains
occur, so that an affinity with the Angiospermae seems unlikely.

Dr. W. G. Chaloner has drawn my attention to Leschik’s genus Camerosporites\
Camerosporites, however, has equatorial swellings which are described as hollow
chambers, whereas those in Hoegisporis are solid.

Although never frequent, H. lenticulifera has been isolated, to date, from several
widely separated Cretaceous deposits in Western Australia, two in South Australia, and
one in North Queensland. The exact age of some of these sediments is still in doubt but
present indications are that H. lenticulifera ranged from high in the Aptian to Ceno-
manian.

I wish to thank Professor R. Potonie, Krefeld, Professor O. Arbo Hoeg, University of Oslo, and
Mr. J. M. Wonnacott, British Museum (Natural History), for advice and information. Mr. Svein
Manum, University of Oslo, kindly made the photographs for PI. 76, figs. 4-8.

REFERENCES

cookson, i. c. and eisenack, a. 1960. Microplankton from Australian Cretaceous sediments. Micro-
paleontology, 6, 1-18.

and dettmann, m. e. 1959. On Schizosporis, a new form genus from Australian Cretaceous

deposits. Ibid. 5, 213-16, pi. 1.

eisenack, a. and cookson, i. c. 1959. Microplankton from Australian Lower Cretaceous sediments.
Proc. Roy. Soc. Vic. 72, in press.

leschik, G. 1955. Die Keuperflora von Neuewelt bei Basel. II, Die Iso- und Mikrosporen. Schweitz.
Palaeont. Abh. 72, 1-70.

pokrowskoi, i. 1956. Atlas oligocenovych sprovo-pyl'cevych kompleksov razlicnych rajonov SSSR.
Materia/y vsesojuznogo naucno-issledovateV skogo geologice skogo instituta ( VS EG El) Novaja serija.
vyp. 16.

thiergart, F. 1940. Die Micropalaontologie als Pollenanalyse im Dienst Braunkohlenforschung.
Schr. Brennstoff-Geol. 13, 1-89.

ISABEL C. COOKSON,
Department of Botany,
University of Melbourne,

Manuscript received 12 November 1959 Australia

EXPLANATION OF PLATE 76

Figs. 1-3. Pugnoides triplex (M'Coy). 1 a-c. Ventral, dorsal, and anterior views of topotype, x5,
Kildress, Co. Tyrone, Ireland, British Museum (Natural History) B 12652. 2 a-c. Dorsal, lateral,
and anterior views of specimen from Weaver Hills, x4, B.M. (N.H.) BB 39182. 3 a-c. Ventral,
dorsal, and anterior views of a specimen from the Weaver Hills which has four costae on fold of
brachial valve, x3, B.M. (N.H.) BB 39180.

Figs. 4-8. Hoegisporis lenticulifera gen. et sp. nov. 4, Paratype, Regan’s Ford, W.A., seismic shot hole
L8 at 240 ft., x 1,000, Nat. Mus. Vic. P 2051 1. 5, Subiaco Bore, W.A., at 358 ft., x 1,000. 6, 7, Two
views of holotype, Subiaco Bore, W.A., at 358 ft., x 1,000. 8, Rough Range South no. 1 Bore, W.A.,
between 2,758 and 2,867 ft., x 800.

Fig. 9. Reproduction of Pokrowskoi’s figure (1956, pi. 5, fig. 26, p. 241) of an ‘indeterminate pollen.
Angiospermae’ from the Oligocene of Russia, X600.

Palaeontology, Vol. 3

PLATE 76

COOKSON, Hoegisporis gen. nov.
PARKINSON, Pugnoides triplex (M‘Coy)

THE STRATIGRAPHICAL PALAEONTOLOGY OF
THE LOWER GREENSAND

by RAYMOND CASEY
CONTENTS

page

Introduction ........... 487

Stratigraphical relations of the Lower Greensand ..... 489

Zonation of the Lower Greensand ........ 492

Depositional history of the Lower Greensand ...... 499

Stratigraphical account . . . . . . . . .501

Southern Basin

Vectian Province ......... 502

Wealden Province . . . . . . . . .517

Northern Basin

Cambridge-Bedford Province ....... 566

Lincolnshire-Norfolk Province ....... 570

Palaeontology ........... 572

Faunal and floral lists . . . . . . . . . .601

References . . . . . . . . . . . .611

Abstract. The Lower Greensand in Britain comprises up to 800 feet of sediments laid down in a great variety
of near-shore environments stretching from the Isle of Wight northwards to the border of Yorkshire. The fauna
is dominantly molluscan, of neritic, littoral, and estuarine facies, with local abundance of brachiopoda, polyzoa,
or sponges. Despite losses from subsequent leaching, it is unexpectedly rich; the ammonite sequence is known
in unrivalled detail and affords a basis for division of the Aptian and Lower Albian Stages of the Lower Creta-
ceous into nine zones and twenty-four subzones. The whole fauna and flora of the Lower Greensand (microzoa
excepted) is listed with up-to-date names in the new zones and the application of the zonal scheme in the field
is demonstrated in a detailed description of the stratigraphy and life-succession region by region. Revised
correlations of the various local subdivisions of the Lower Greensand are set out in diagrams and tables. The
depositional history of the formation is reviewed, bringing out new information which has resulted from use
of a refined ammonite chronology. New taxa described in the systematic section are: Gastropods, 1 genus,
2 species; Lamellibranchs, 1 family, 8 genera, 20 species; Ammonites, 3 genera, 14 species; Brachiopods, 1 genus,
I species; Polyzoa, 1 species; Problematica, 2 genera, 1 species. Many species are recorded from the Lower
Greensand for the first time. Genera new to the British Cretaceous are the lamellibranchs Disparilia, Senis ,
Cimeocorbida, Eomiodon, and Protodonax , the ammonite Megatyloceras, and the boring polyzoan Graysonia.
Eomiodon and Protodonax occur in the Aptian of the Middle East but have not been recorded previously from
the Cretaceous of Europe.

INTRODUCTION

The Lower Greensand is a series of sandy deposits underlying the Gault and occupies
extensive tracts of country in southern England, reaching a thickness of 800 feet in the
Isle of Wight. It provides the most complete record of the Aptian and Lower Albian
stages of the Lower Cretaceous in Britain and marks the beginning of a great cycle of
marine sedimentation that continued until the end of the Mesozoic.

Historical accounts of the Lower Greensand will be found in Mantell (1822), Cony-
beare and Phillips (1822), Fitton (1824, 1836, 1847o, b ), Mantell (1851), Topley (1875),
Bristow(1889), Jukes-Browne(1900; 191 1), Stopes ( 191 5), Boswell ( 1929), and Kirkaldy

[Palaeontology, Vol. 3, Part 4, 1961, pp. 487-621, pi. 77-84.]

488 PALAEONTOLOGY, VOLUME 3

(1939), and fuller references to the literature are given by Stopes (1915), Kirkaldy
(1939), and Casey (1960a).

Interest in the stratigraphical palaeontology of the Lower Greensand was at its height
in the early part of the last century: Fitton, Bensted, and others were then active in the
field and the Sowerbys, Forbes, Mantell, and Morris were busy naming and describing
the fossils. Such team work among the field men and the palaeontologists was never
repeated and succeeding generations of palaeontologists found themselves more and
more out of touch with the Lower Greensand. Already by 1854 Sharpe found the fossils
of the Faringdon Sponge Gravels, a local facies of the formation, so foreign to his idea
of a Lower Greensand fauna that he declared them to be of Danian age. Later Kitchin
and Pringle (1921) refused to believe that fossils collected by Lamplugh and Walker
(1903) from the top of the Lower Greensand at Leighton Buzzard belonged even to the
Lower Cretaceous and went to extraordinary lengths trying to prove that the whole of
the Gault and its superstratum had been turned upside-down. Scepticism has also been
voiced about the provenance of the flowering plants described from the Lower Green-
sand (Harris 1956) and it is only a few years ago that the ammonites of this foimation —
now known to give an unrivalled sequence through Aptian and Lower Albian times —
had been written off as an ‘impoverished’ set by the experts (Spath 1930a; Arkell 1947h).

Not only in the ammonites, but in many other groups of Lower Greensand fossils,
poverty turns to riches with patience. These riches are the natural legacy of a formation
laid down in changing coastal waters. The Faringdon Sponge Gravels, the Shenley
Limestone, the Iron Sands of Seend, the Crackers, the Punfield Marine Band, the
regular is-mammillatum nodule-beds — these and many more reflect marine environments
of a diversity and individuality difficult to match in any other formation the world over.

The present memoir is really a corollary to my Monograph of the Ammonoidea of the
Lower Greensand (Part 1, Casey 1960a) and was first written as a stratigraphical review
of the Lower Greensand with special regard to ammonite occurrences. The importance
of ammonites for zoning and correlation makes such emphasis inevitable, but to make
the paper more useful I have tried to give an up-to-date account of the whole fauna.
The need for systematic work on all its animal-groups is patently obvious to anyone
who has to name Lower Greensand fossils. Woods’s great monograph on the Cretaceous
Lamellibranchia was written fifty years ago and it is not surprising that many more
species have been found since and that the names of others need revising. Except for a
recent paper by Cox (1960) on the family Pleurotomariidae, work on Lower Greensand
gastropods seems to have ceased from the time of Starkie-Gardner (1875-7). Elliott
(1947; 1959), Middlemiss (1959), and Owen (1956; 1960) have made a start on revising
the brachiopods, so well described by Davidson (1851-86) and Walker (1867-70) in
the last century, but for the corals, echinoids, and sponges we still rely largely on the
works of Duncan (1866-91), T. Wright (1864-82), and Hinde (1883; 1885) respectively.
Papers by Chapman (1894) and J. Wright (1905) are still the last words on the microzoa.
The only group of fossils in the Lower Greensand that has received adequate attention
is the plants, mostly drift-wood, which was the subject of illuminating work by the late
Dr. Marie Stopes (1911-15).

Aside from the restudy of museum specimens, there is a greater need for fresh material
gathered first-hand from the field. Much of the Lower Greensand consists of sands more
or less leached of organic matter and real advances in the study of the faunal sequence

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 489

depend on the finding at new levels of hard nodules or lenses in which fossils have
escaped dissolution. All that can be accomplished today is an interim stocktaking of the
fauna to show what groups are most badly in need of specialist attention and where
search in the field should be redoubled.

The following pages contain the results of some twenty-five years’ work on the Lower
Greensand as a leisure- time pursuit. During that time I have received the help of a
large number of collectors and enthusiasts. In addition to those friends named on a
previous occasion (Casey 1960u, pp. ii-iii), I am indebted to Miss Eileen Andrews for
assistance with some of the diagrams. Dr. W. G. Chaloner and Mr. C. W. Wright have
been especially helpful in providing information and I have had the benefit of advice
from many of my colleagues at the Geological Survey and the British Museum (Natural
History). Some of the work was done in the Geology Department of the University of
Reading during a period of leave granted me by the Department of Scientific and
Industrial Research. Geological Survey photographs are reproduced by permission of
the Controller, H.M. Stationery Office, and the paper is published by permission of the
Director, Geological Survey and Museum. Funds for publication were provided by Shell
International Research.

STRATIGRAPHICAL RELATIONS OF THE LOWER GREENSAND

It is traditional for those who write about the Lower Greensand to repeat the gibe
that the formation is seldom green and frequently not sand. How this misnomer came
to be accepted for a primary division of the British Cretaceous System is a matter of
historical interest and is fully explained by Jukes- Browne (1900, pp. 15-26). When the
term ‘Greensand’ was first introduced is a little uncertain, though it is thought to have
originated with William Smith between the years 1800 and 1812. What is certain is that
both he and Thomas Webster always used it for the green sands between the Chalk and
the Gault and not for the sands that underlie the Gault. Misapprehension of the position
of the ‘Greensand’ by William Phillips and Mantell led to much confusion and con-
troversy until it was realized that there were two sandy formations, one above and one
below the Gault. ‘Reigate Sands’, ‘Shanklin Sands’, ‘Ferruginous Sands’, ‘Carstone’,
and other terms had come into use for the lower member, but the wide circulation
enjoyed by Mantell’s books had helped to implant the word ‘Greensand’ too deeply
for it to be uprooted. ‘Lower Greensand’ and ‘Upper Greensand’ was the obvious
nomenclatorial compromise, for which Webster (1825) accepted responsibility.

Attempts to fit the Lower Greensand into d’Orbigny’s scheme of stages provoked
another lively controversy about the term ‘Neoconiian’, in which Fitton, Leymerie,
Cornuel, Judd, and others joined. It is now known that the true Lower Greensand, as
found in the Weald and the Isle of Wight, is of Aptian and Lower Albian age and is
younger than the Sandringham Sands, Claxby Beds, and other Neocomian strata to
which the name Lower Greensand was loosely applied in the past. The terms ‘Vectine’
(Fitton 1845) and ‘Vectian’ (Jukes Browne 1886), proposed for use in an adjectival
sense or as a stage name for the Lower Greensand Series, have not been adopted.
Defunct names of continental origin which have been applied in the past to parts of
the Lower Greensand or with which parts of the Lower Greensand have been correlated
are ‘Rhodanian’ and ‘Urgonian’. The first was proposed by Renevier (1854) for a

490

PALAEONTOLOGY, VOLUME 3

neritic, Orbitolina facies of the Aptian developed at Perte-du-Rhone (Ain), France;
the second name was given by d’Orbigny (1847a) to a calcareous facies of the Barremian
(topmost Neocomian) rich in rudists and Orbitolina and seen typically at Orgon, southern
France.

Relations with the Wealden Beds. In south-east England the Lower Greensand rests on a
thick series of fresh- and brackish-water sediments, the Wealden Beds, of which the
topmost member is the Weald Clay (Wealden Shales in the Isle of Wight). With rare
exceptions, wherever the two formations are seen in contact the junction is absolutely
sharp and there are signs in places that deposition of the Lower Greensand was pre-
ceded by gentle folding and erosion of the Wealden Beds. At the western end of the
Weald, south-west of Haslemere and west of Fernhurst, the base of the Lower Green-
sand oversteps various marker bands in the Weald Clay (Holmes 1959) and the same
type of contact on a much smaller scale was seen by Kirkaldy (1937, p. 106) at Berwick,
near Lewes, East Sussex. Another local unconformity has been noted at Kingsnorth,
near Ashford, Kent (Edmunds 1956, p. 32). In general, however, the relation of the
Lower Greensand to the Wealden Beds is that of disconformity rather than uncon-
formity. In the Isle of Wight, for example, the formation begins with a line of grit full
of fish-teeth and other debris washed from the top of the Wealden Shales; but here
the bedding of the two formations is strictly parallel and the absence of angular dis-
cordance is proved by the persistence of the Wealden Shales at the top of the Wealden
Beds both in the Isle of Wight and on the mainland to the east and to the west. The
old idea that the Oxford Wealden is a non-marine facies of the Lower Greensand has
been disproved (Arkell 1944), but it may now be shown that on the Dorset coast,
between Lulworth Cove and Swanage, Wealden conditions lingered on in the estuary
of a Lower Greensand river. And here the junction is gradational.

Palaeontology gives no indication of a lengthy break at this level. Although there is a
great change in fossils at the base of the Lower Greensand, the fauna of the Wealden
Shales and the Weald Clay shows an increasing saltwater influence as one ascends the
succession. The highest part of the Wealden is of near-marine facies, comparable with
that of the ‘Cinder Bed’ of the Purbeck, and contains foraminifera, echinoid spines,
and the molluscs Cassiope, Ostrea, Corbula, Nemocardium, and Filosina. The last is a
marine-brackish lamellibranch generally mistaken for the fresh-water-brackish Neo-
miodon (or ‘ Cyrena ’) and is common enough in places to be a rock-builder (Casey
1955a). It is found also in the Aptian of the Lebanon (‘ Corbicula' hamlini Whitfield)
and in the Upper Barremian of the Paris Basin (‘ Cyclas’’ neocomiensis Cornuel). Nemo -
cardiwn (Pratulum) ibbetsoni (Forbes) is another species common to the Upper Barre-
mian of the Paris Basin and the top beds of the English Wealden. Ammonites found
a few inches above the bottom of the Lower Greensand indicate an horizon just above
the base of the Aptian, and although no correlation of the Wealden Beds with the
marine succession can yet be made with certainty, what little evidence there is favours
an Upper Barremian age for the topmost beds.1 This supports Allen’s idea that the

1 Hughes (1958), working on plant spores, puts the Wealden Shales in the Aptian, but since he also
correlates them with part of the Fulletby Series it is clear that he is using the term Aptian in Spath’s
sense of including the recticostatus Zone, here reinstated in the Barremian. Derived Kimmeridgian
Paylovia occur both in the Lower Greensand and in the top of the Wealden Shales. One badly rolled

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 491

top of the Wealden Beds is not much older than the Lower Greensand (Allen 1955,
p. 272) and accords with the views of Strahan and Reid (in Bristow 1889, p. 19): ‘That
the change in sediment is such as might have been produced by the sudden conversion
of a partially land-locked estuary or lake into a bay open to the sea, whether by subsi-
dence or by washing away a barrier.’

Derived fossils in the Lower Greensand Woburn ( = Potton) Sands are supposed to
afford evidence of the former extension of the Wealden Beds north of the London
Ridge, this supposition dating from Walker’s ( 1 866a) discovery of water-worn Iguanodon
bones at Potton. No one would today accept these finds as proof of a Wealden origin:
Iguanodon was still living in Aptian times and worn fossils are commonplace in con-
temporary Lower Greensand deposits (see Keeping 1883, p. 40). Equally unsatisfactory
are the ‘Potton’ plants said to have originated in the Wealden. In an appendix to her
catalogue of the Lower Greensand flora Stopes (1915) described the following cycado-
phytes from the ‘Potton Sands’ as probable Wealden derivatives: Cycadeoidea yatesii
(= Yatesia morrisi ), C. buzzardensis, Bennettites inclusus, and Colymbetes edwardsi , the
last attributed to Potton with question. Said Stopes (1915, p. 295): ‘It is generally held
that Potton fossils of the colour and texture of these (rich red-brown limonite) are
derived from the Wealden.’ Teall (1875) made a study of the Potton fossils and had
concluded just the opposite: he thought the ferruginous fossils were indigenous and
specifically stated (p. 9) that the cycadophytes ought not to be regarded as derived.
The ‘tree-fern’ Tempskya (= Endogenites) erosa, which Teall looked upon as a Wealden
fossil at Potton, was included by Stopes in the native flora. There is a similar sharp
divergence of opinion about the pine-cone Pinostrobus cylindroides. Gardner (1886)
vouched for it as a Lower Greensand fossil, finding it ‘in excellent condition, certainly
not derived from any older beds, like so many of the Potton fossils’. Seward’s (1895,
p. 193) inspection of the same fossil led him ‘to unhesitatingly describe it as distinctly
worn and rolled, and imperfectly preserved . . .’. Stopes did not examine the original of
the unique Bennettites inclusus (in the York Museum) and the source of Colymbetes
edwardsi is unknown; of the species cited in her appendix, it is therefore on the Cyca-
deoidea that the question of provenance really hinges. Carruthers (1870) had mislead-
ingly described these as having been found in the same stratum as the cone Cycadeostro-
bus walkeri, i.e. the Potton nodule-bed. In fact they were obtained from Leighton
Buzzard, ‘near Leighton Buzzard’ or ‘sandpit just outside Leighton Buzzard’. I have
examined fifteen specimens in the British Museum (Natural History) and the Geological
Survey Museum so labelled. All are in heavy dark reddish-brown carstone, the largest
(BM. V 13238) weighing several pounds. They are not distinguishable in appearance
from the carstone concretions with fossil wood that occur near the top of the ‘Silver
Sands’ in the Leighton Buzzard pits in work today, from which horizon they may well
have originated (see Lamplugh and Walker 1903, p. 239). A century ago, before Bennet-
tites had been described from the Aptian, a cycad-like plant in the Lower Cretaceous
may have raised the presumption of a Wealden age. Today it is not so easy to believe
that these fossils are derived. Indigenous or derived, plant and reptile, none of these
fossils gives grounds for supposing that the Wealden Beds of south-east England

specimen was sent to me as an uncoiled ammonite, thereby suggesting a solution to the puzzling record
of Ancyloceras in the Wealden Shales (Judd 1871, p. 220). This should not deter anyone from searching
for drifted Barremian ammonites in the quasi-marine beds at the top of the Wealden.

492

PALAEONTOLOGY, VOLUME 3

stretched north into Bedfordshire. They suggest simply that sometime in the Lower
Cretaceous the northern slopes of the London Ridge supported lguanodon and a flora
with cycadophytes and ‘tree-ferns’, which is what one would expect whether the waters
that lapped against the Ridge were salt or fresh. The apparent absence of Wealden
fossils of aquatic type among the Potton derivatives favours the idea that these plants
and reptile remains were washed straight from the land into the sea.

Relations with other subjacent formations. Extending westwards beyond the Weald the
Lower Greensand oversteps the Wealden Beds and passes across various members of the
Jurassic System. North of the buried London Ridge it rests on Jurassic or marine
Neocomian rocks, and borings along the edge of the Ridge show that in places it laps
on to the Palaeozoic (e.g. at Lowestoft). In these regions there is no problem about
delimiting the base of the Lower Greensand.

Relations with the Gault. Fitton (1836) and Topley (1868) knew that in the field it is often
difficult to draw a rigid dividing-line between the Lower Greensand and the Gault. In
Norfolk and Lincolnshire, where the Gault may pass laterally into ‘ Red Chalk’, there
is a similar gradational junction. Later workers were impressed by the seemingly abrupt
introduction of Albian fossils in the transgressive beds at the base of the Gault and
thought there was an unconformity at this horizon, which in Britain had been taken as
the plane of division between the Lower and the Upper Cretaceous. Eventually Albian
fossils were found to range well down into the Lower Greensand, showing that the base-
line of the Gault does not mark any important break or boundary in the geological
time-scale; it is, in fact, diachronous and often arbitrary (Casey 1950). Generally speak-
ing the change from predominantly sandy to predominantly clayey sediment takes place
within a few feet of condensed strata of mammillatum Zone age; exceptionally, as in the
extreme east of Sussex, these passage-beds reach down to the top of the Aptian; com-
monly, as in the Isle of Wight, they extend up into the dentatus Zone. In these circum-
stances a workable boundary can be fixed only by palaeontology. For the purposes of
this paper, as in my Monograph of the Ammonoidea of the Lower Greensand , the upper
limit of the Lower Greensand is drawn at the top of the mammillatum Zone. As thus
defined, the formation corresponds exactly to the Aptian and Lower Albian stages of
international nomenclature.

ZONATION OF THE LOWER GREENSAND

The terms Aptian and Albian are anglicized versions of d’Orbigny’s ‘Aptien’ and
‘Albien’, the former named from the village of Apt (Basses-Alpes), the latter from the
district of the Aube, south of the Paris Basin (d’Orbigny 1840; 1842n). During the past
120 years numerous schemes have been proposed for subdividing these stages or for
altering their limits in a Procrustean manner to suit local requirements. Most of the
pioneer work in zonation was carried out in south-east France and north Germany and
it is from these two regions that much of present-day nomenclature derives.

Separation of the Aptian of south-east France into two broad divisions, an upper
portion typified by the marls in the neighbourhood of Apt, near Gargas (Basses-Alpes),
and a lower portion represented by the limestones of La Bedoule, near Marseilles, is of
long standing and was made already by Ewald (1850). These two divisions, the lower

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 493

characterized by Ammonites deshayesi and Ancycloceras matheronianum d’Orbigny, the
upper by Ammonites dufrenoyi, A. martini , and A. nisus d'Orbigny, were ranked as
substages by Dumas (1876) and it was to these substages that the terms Bedoulian and
Gargasian were subsequently applied by Toucas (1888) and Kilian (1887) respectively.
By the end of the nineteenth century a considerable body of information on the local
stratigraphy and sequence of faunas in the Aptian and Albian of south-east France had
accumulated, but it was not until the early years of the present century that this infor-
mation was co-ordinated into a scheme of zonation of general applicability. In a masterly
thesis on the Cretaceous strata of the French Alps and adjoining regions, Jacob (1907)
proposed the following classification:

/VIb

Subzone of Mortoniceras inflation and Turrilites bergeri

(Via

Subzone of Mortoniceras hugardianum

ALBIAN

V

Zone of Hoplites dentatus

IV

Zone of Hoplites ( Leymeriella ) tardefurcatus

III

Zone of Douvilleiceras nodosocostatum and D. bigoureti

( Gargasian

/ lib

Subzone of Douvilleiceras subnodosocostatum and D. buxtorfi

APTIAN

jlla

Subzone of Oppelia nisus and Hoplites furcatus

(Bedoulian

I

Zone of Parahopiites deshayesi and Ancy/oceras matheronianum

One of the notable features of this scheme was the dropping of Douvilleiceras mam-
millatum as a zone fossil, which had been used by Barrois (1874; 1875; 1878) and others
in northern France, in favour of Leymeriella tardefwcata.

Kilian and Reboul’s work (1915) on the Lower Aptian of the neighbourhood of
Montelimar (Rhone Valley), based on a collection of fossils from the gigantic limestone
quarries of l’Homme d’Armes, included a useful review of the Aptian ammonite suc-
cession in many parts of the world. They divided the Lower Aptian (Bedoulian) of this
area into an upper and a lower division and showed that ‘ Parahopiites’’ deshayesi
occurred only in the upper division. They recognized a lower zone of T.’ weissi and
‘ Doiiyilleiceras’’ albrechti-austriae (adopted from von Koenen 1902) and an upper zone
of ‘TV deshayesi.

Ganz (1912) gave a very full account of the Swiss Aptian and Albian and compared
the succession with that of France and of England, using Jacob’s zones.

In north Germany von Strombeck as early as 1856-61 had made out a faunal suc-
cession in the Aptian and Albian in which Ammonites martini, A. tardefurcatus, and A.
regularis figured as characteristic fossils. A great step forwards was made in the investiga-
tion of the German sequence in the early part of this century by von Koenen (1902;
1907) and Stolley (1908a, b ). Stolley, improving on an earlier scheme of von Koenen,
put forward the following classification of the North German Aptian:

Zone of Oppelia ( Adolphia ) trautscholdi and Parahopiites schmidti
Zone of Belemnites aff. ewaldi, &c. (no ammonites)

Zone of Hoplites deshayesi

Zone of Douvilleiceras albrechti-austriae and Parahopiites weissi
Zone of Hoplitides bodei

Stolley combined belemnites and ammonites in his zonal tables and in the continua-
tion of this work produced a similar scheme for the Lower Albian (styled ‘Middle
Gault’), naming species of Hypaeanthoplites, Nolaniceras (‘ Parahopiites ’) and Ley-
meriella (‘ Hoplites ’) as the index ammonites, as follows:

k k

B 6612

494

PALAEONTOLOGY, VOLUME 3

Zone of Hoplites regularis and Bel. strombecki mut. minor

Zone of Hoplites tardefurcatus, Pcircihoplites milletianus, and Bel. n. sp. aff. strombecki
Zone of Hoplites aff. tardefurcatus )

Zone of Parahoplites jacobi and Bel. strombecki ) Beds Wlth Desmoceras ketlhacki
Zone of Parahoplites nolani and Douv. cornuelianum

Careful collecting from brickpits and other artificial openings around Hanover
enabled Brinkmann (1937) to replace the top three zones by a more detailed scheme
based on the occurrences of Leymeriella, as follows:

Zones

Leymeriella regularis
Leymeriella tardefurcata
Leymeriella schrammeni

Subzones
Hoplites spp.

L. hitzeli

L. tardefurcata tardefurcata
L. tardefurcata anterior

L. schrammeni schrammeni
L. schrammeni anterior

Meanwhile, Spath (19236) had proposed the following zonation of the Aptian and
Lower Albian:

I regularis
milletianus
schrammeni

jacobi
nolani

aschiltaensis
nutfieldensis

tovi/ense
bowerbanki
hi/lsi

consobrinoides
hambrovi
weissi
bodei

bidentatus
rude

sparsicosta

At first sight this zonal scheme implies a great advance on the work of Jacob and
Stolley. It must be pointed out, however, that it was not based on the principle of
superposition as observed in the field, but on a perusal of the literature and examination
of museum specimens. Although expressly put forward as a means of correlating
British deposits, it was in fact not a zonal succession but a theoretical faunal sequence
with ‘index’ fossils drawn from areas as far afield as north Germany, southern England,
south-east France, and the Caucasus. It corresponds to nothing in Nature and its
proposition seems to have been made according to the principles followed by Buckman,
whose attempts at refined ammonite chronologies in the Jurassic have been the subject
of so much adverse criticism. It is much to the credit of Spath, however, that he applied
Buckman’s methods with greater caution than did the master, and although the scheme
reproduced above is unworkable in the field, contains much guesswork and some
questionable correlation, the species are listed in the right order, so far as is known.

Lower Albian

Upper Aptian

( Leymeriellan

fAcanthoplitan

(Parahoplitan

(subnodosocostatum zone)

Tropaeuman
( martini zone)

'Parahoplitoidan
( deshayesi zone)

Lower Aptian <

Parancyloceratan
( ( recticostatus zone)

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 495

The principal innovation in this scheme was the insertion of the ‘ Parancyloceratan
age’ in the Lower Aptian. As originally published (Spath 1923h), no zonal index was
chosen for the three subzones grouped in this ‘age’, but in 1924 he indicated that the
zone of Costidiscus recticostatus lay above the upper limit of the Barremian and in a
subsequent reproduction of this zonal scheme (Neaverson 1928) the word recticostatus
was added as the zonal index. This was endorsed by Spath (1930a) on the grounds that
varieties of Costidiscus recticostatus ranged up from the Barremian into the Lower
Aptian, as also did Maeroscaphites, another typically Barremian form. Spath showed
that a Zone of Douvilleiceras mammillatum was separable above the beds with Lev-
meriella tardefurcata and L. regularis and in a later publication (Spath 1941, p. 668)
he divided it into a Subzone of Douvilleiceras monile below and a Subzone of D. inaequino-
dum above. Contrary to the practice of the French, however, he placed this zone in the
Middle Albian. Another change in the 1923 table was the replacement of Hypacant-
hoplites milletianus by Lcymcriclla acuticostata for the index fossil of the middle part
of the tardefurcata Zone (Spath 1942, p. 673).

In 1947 an important review of the ammonite occurrences in the Albian of France
and England was published by Breistroffer. This author’s conclusions on zonation are
summarized in the following table:

Lower Albian

(Douvilleiceratian)

Upper Aptian
(Clansayesian)

Zone of Douvilleiceras monile and
D. orbignyi (Protohoplitan)

Zone of Leymeriella tardefurcata
and Hypacanthoplites trivialis
(Leymeriellan)

Zone of Diadochoceras nodosoco-
statum and Acanthoboplites
bigoureti (Acanthohoplitan)

(Horizon of D. inaequinodum (in
England)

Main level with Protohoplites puzo-
sianus, Sonneratia dutempleana,
Cleoniceras cleon

'Subzone of L. canteriata and L.

( EpUeymeriella ) hitzeli
Subzone of L. tardefurcata (and L.

acuticostata in Hanover)

Horizon of L. (ProleymerieUa)
, schrammeni (in Hanover)

{Subzone of Hypacanthoplites jacobi
and H. sarasini

Subzone of H. nolani, Parahoplites
grossouvrei, and Cheloniceras c/an-
sayense

The nodosocostatum Zone, or ‘Clansayes’ horizon, which forms a kind of buffer-state
between the classic Aptian and Albian, had been included in the Albian by Jacob,
Stolley, and Spath. Breistroffer now showed that certain species of Albian affinities had
been wrongly credited to this essentially Aptian horizon and that a more satisfactory
starting-point for the Albian was at the base of the tardefurcata Zone. The ‘Proto-
hoplitan’ ( mammillatum Zone of other authors) was reinstated by Breistroffer in the
Lower Albian, but Douvilleiceras mammillatum and Leymeriella regularis were both
replaced by other index species.

Attempts to correlate the Aptian-Albian series of Europe with that of Texas (Trinity
Group) by Scott (1940) and of south-east Arizona (Lowell Formation) by Stoyanow
(1949) have led to new conceptions of zonation, supposedly of world-wide significance.
Scott sought to interpose a Zone of Sonneratia trinitensis between the horizons of
Leymeriella regularis and Douvilleiceras mammillatum. Among the changes in the

496

PALAEONTOLOGY, VOLUME 3

‘standard’ zonal scheme advocated by Stoyanow (1949, p. 38) was the shifting down-
wards of the Sonneratia trinitensis Zone to a position between that of Leymeriella
tardefurcata and Hypacanthoplites jacobi and of the Parahoplites horizon (Spath’s
‘Parahoplitan age’ or subnodosocostcitum Zone) to the base of the Gargasian. He also
proposed to draw the boundary of the Aptian and Albian stages between the horizons
of H. nolani and H. jacobi.

The scheme of zonal classification here adopted is shown in Table 1 . This is based on
the order of succession seen in the Lower Greensand and can be demonstrated in the
field. In this scheme the Aptian stage begins with the entry of the ammonite family
Deshayesitidae. The zone of Costidiscus recticostatus, included by Spath in the Aptian,
is found only in the Mediterranean region and has always been regarded as part of the
Barremian. Coquand, who first proposed recognition of the Barremian Stage, regarded
C. recticostatus as one of its characteristic fossils, of equal rank with Macroscaphites
ivani (Coquand 1862). The various subzones included in the recticostatus Zone by
Spath were named from north German occurrences. They were regarded by their
originator, von Koenen, as Barremian, as also by Stolley (1908o) and Sinzow (1905).
Their index fossils belong to the Noith Sea Province and have never been found in
association with Costidiscus. Whatever relation these Mediterranean and Boreal ele-
ments have in time, there is no justification for removing them from the Barremian.

In the systematic part of this paper it is shown that the identification of Deshayesites
deshayesi in this country was at fault and that the true deshayesi Zone is somewhat
higher in the sequence than is indicated in Spath’s table. Four well-marked zones may
be recognized in the Lower Aptian based on occurrences of Deshayesitidae: ( \) fis si-
costatus Zone, with Prodeshayesites, (2) forbesi Zone, with early Deshayesites of the
forbesi type (= D. deshayesi Spath non d’Orbigny), (3) deshayesi Zone, with Deshaye-
sites s.s., (4) bowerbanki Zone, with Dufrenoyia. Tropaeum bowerbanki is chosen as
index-fossil for the highest zone of the Lower Aptian though Dufrenoyia furcata is
equally characteristic. Unfortunately a ‘ furcatus Zone’, based on misidentification of
d’Orbigny’s Ammonites dufrenoyi, has been widely used in European literature for what
is here called the martinioides Zone of the Upper Aptian. Of the various subzones of
the ‘ deshayesi Zone’ employed by Spath, that of D. weissi cannot be recognized in the
absence of the zonal ammonite, a German species. Roloboceras hambrovi has too long
a range and the Russian Deshayesites consobrinoides is less suitable for British strata
than Cheloniceras parinodum sp. nov.

To fix a boundary in line with the base of the Upper Aptian (Gargasian) of south-
east France, I have taken the appearance of Epicheloniceras as diagnostic. It is regretted
that the ‘'martini Zone’, so familiar to British geologists as the name for the lower part
of the Upper Aptian, must be abandoned, for reasons given in the systematic chapter.
The Russians (e.g. Sazonova 1958) have used Ch. ( Epicheloniceras ) tschernyschewi,
Ch. (E.) subnodosocostatum, and Gargasiceras gargasense as guide fossils for this part
of the succession. The first is exceedingly rare in Britain, the third is unknown in this
country, and the second has been used incorrectly for what is now called the nutfieldensis
Zone. Cheloniceras ( E .) martinioides sp. nov. is the best Lower Greensand substitute
for ‘ A . martini ’, for which indeed it has generally passed. Tropaeum hillsi and T. bower-
banki are Lower Aptian species that were misplaced in the ‘ martini Zone’ in Spath’s
table (Casey 1960#, p. 25 footnote) and the same author’s use of Ammonitoceras

table 1. Zonal classification and correlation of the Lower Greensand.

ZONES

SUBZONES

ISLE OF
WIGHT

EAST

KENT

WEST

KENT

SURREY

SUSSEX.

WESTERN

OUTLIERS

NORTHERN

BASIN

PROTOHOPLITES
HEMISONNERATIA)
PUZ OSI ANUS

SULPHUR

BAND

in

NODULE BEDS

NODULE BEDS

\ NOD

ULE BEDS

GLAUCONITIC

z

<

< 5

* 3
UJ £

d <

UJ -J

OTOHOPUTES

RAULINIANUS

\ BED

Wl T

JUNCTION

JUNCTION

NCTION

SANDS
AT BASE

CO

_1

<

_J d

d 2
> 2
^ <

CLEONICERAS

FLORIOUM

GAU

y/

GAULT

ITH

GAULT

OF GAULT

a 2

SONNERATIA

KITCHINI

\s. KITCHINI
\ BED

ON-SEOUENCE

01 \

2 g|=

1

UJ

5

O

_i

5$

LEYMERIELLA

REGULARIS

NON

SEQUENCE

z

O

1 OF
1 BED 3

P-

WOOD

ILEIGHTON
' NODULE

ISHENLEY
| LIMESTONE

z

U U
— CL
CL 3
UJ U-

~ o

HYPACANTHOPLITES
MILLETIOI OES

o

) BED 2

FO

LKESTONE

UPPER

PEBBLY

o

SANDS

z

35

FARNHAMIA

FARNHAMENSIS

NDROCK

NON - SEQUE

'ICE

S

BAND

CLAY-SILT

BAND

E

OF

(PARS )

1

10

UJ

H

3

HYPACANTHOPLITES

ANGLICUS

BED 1

BEDS

z

o

SILVER

UFFINGTON

ETC.

SANDS

z

O

0 CD

1 o

H O
Z <

HYPACANTHOPLITES

RUBRICOSUS

2

BASAL

C

z

<

0
<
0-
>

1

NOLANICERAS
NOLAN 1

CLAY BANO
OF GROUP XV

S

IN DGATE

\

SANDS

?

MAREHILL

CLAY

h-

Q.

<

in

UJ 1/1

t S
5? 9

PARAHOPLITES

CUNNINGTONI

GROUP XIV

MASS

s

PUTTENHAM

BEDS

">o"

SANDS

IRON SANDS
OF SEEND

UPWARE

DC

UJ

O d

5 ^
2 i

TROPAEUM

SUBARCTICUM

GROUP XIII

■

BEDS

1

BARGATE
& FULLERS
EARTH
BEDS

BARGATE

BEDS

FARINGDON
SPONGE GRAVELS

MARL

CL

CL

£ in
< UJ
* Q

CHELONICERAS

(EPICHELONICERAS)

BUXTORFI

A N D S

?

GROUPS XI & XII

!

BASAL

BOUGHTON

\ S

1 E

o o
z z

S H

CHELONICERAS

(EPICHELONICERAS)

GRACILE

GROUPS IX & X

m

NODULE

BED

TOP \ ^
CHERTsl |

Si

CHELONICERAS

(EPICHELONICERAS)

DEBILE

O

GROUP VIII

l

BEOS

blackdown\

ANO \

GODSTONE

H

THE

1 1
< g

CHELONICERAS

(CHELONICERAS)

MEYENDORFFI

E

GROUP VII

BEDS

BELOW

TOP HYTHE
PEBBLE BED

MID

UNSTANTON,

UTTERBY,

si

‘i

DUFRENOYIA
TRANSITORI A

GROUP VI

H Y T H E

X

SANDS

BEDS

7UPWARE

(DERIVED)

z

in

UJ —

t 13

in >.

DESHAYESITES

GRANDIS

GROUP V

BEDS

COALMAN

LANE

LOWER

<
1 —

^ 1
5 y

UJ Q
Q

CHELONICERAS

(CHELONICERAS)

PARINODUM

GROUP IV

STONE

SUTTER8Y,

UPWARE

(DERIVED)

D.

<

cr

m

UJ

H —

DESHAYESITES

CALLIDISCUS

w

UPPER

LOBSTER BED
& CRACKERS

ATHERFI ELD
CLAY

ATHERFIELD

"> UPWARE

(deriveo)

LLl

5

o

HAYESI

■ORBES

DESHAYESITES

in

LOWER
LOBSTER BEC

CLAY

ERFlELO

_i

in

UJ

Q

DESHAYESITES

FITTONI

1

HER F IELD

ATHERFI ELD
CLAY S.S.

CL»

jHAY -
ESITES
)STATUS

PRODESHAYESITES

OBSOLETUS

PERNA BEDS

\ PERNA
\ BED

PERNA BED

s a
g s

CL U_

PRODESHAYESITES

BODEl

HUNSTANTON,
UPWARE, POTTON,
SUTTERBY
(OERIVED)

498

PALAEONTOLOGY, VOLUME 3

tovilense as an index-fossil for part of this zone can scarcely be approved, this ammonite
being known by a single example undocumented as to horizon. The three divisions of
the martinioides Zone here recognized are characterized by species of Epicheloniceras,
the topmost being Ch. ( E .) buxtorfi , adopted as a guide-fossil from Jacob.

Arkell’s (1947a) proposal to drop Ch. ( E .) subnodosocostatum in favour of Parahop-
lites nutfieldensis for the zone fossil of the Upper Gargasian is here endorsed. Working
at great distance from the European museums and apparently handicapped by lack of
literature, Stoyanow (1949) reached the conclusion that European authors had mis-
apprehended the nature and stratigraphical position of the genus Parahoplites in its
restricted sense. He failed to see that far from being unique in North America his
Kasanskyella and ‘ Sinzowiella ’ (both analogous to Parahoplites) were congeneric, if
not conspecific, with Scott’s species of ‘ Sonneratia ’ from the trinitensis Zone of Texas
(compare Stoyanow, pi. 17, figs. 5, 6, and Scott 1940, pi. 66, fig. 2, and pi. 67, fig. 7)
and that in fact the trinitensis Zone and his own Zone of Parahoplites melchioris are
one and the same thing. Russian literature has always been consistent in placing P.
melchioris and its allies above, not below, Dufrenoyia (Sinzow 1909; Natsky 1918;
Sazonova 1958), which agrees with what is seen in Western Europe and in Texas. In
Arizona beds with Kasanskyella are followed directly by strata with Diadochoeeras or a
close ally (= Paracanthohoplites)\ the ‘ Dufrenoyia ’ from the overlying Joserita and
Cholla members are acanthohoplitids comparable with those described by Benavides-
Caceres (1956) from northern Peru as Parahoplites inta and P. c/uilla and placed at the
base of the Albian (‘Zone of Parahoplites nicholsonV). It is clear, therefore, that the
succession in Arizona is not very different from that of Eurasia, though having a greatly
expanded development of the ‘Clan sayes’ horizon. It is true that one of Sinzow’s species
of Parahoplites, or a form comparable therewith (P. cf. multicostatus), has been figured
and described from La Pena formation of Mexico (Humphrey 1949, pi. 12, p. 138),
of Lower Gargasian age, but the ammonite in question is a Colombieeras of the group
of C. alexandrinum (d’Orbigny) and quite properly associated with Dufrenoyia.

I have followed Breistroffer (1947) in extending the Aptian to take in the ‘Clansayes’
horizon, and in accordance with his views (though not his nomenclature) the Lower
Albian is understood to comprise the two zones of Leymeriella tardefurcata and Douvil-
leiceras mammillatum. The tardefurcata Zone seems to mark a new beginning for the
ammonites, in Europe at least (Casey 1957), and this is the level best fitted to start the
Albian. I have already pointed out (Casey 1950; 1957) that there is no need to follow
Breistroffer in replacing Leymeriella regularis as the guide-fossil for the topmost part
of the tardefurcata Zone and his objections to the name Douvilleiceras mammillatum
have also been removed (Casey 19546). It should be noted, however, that the mammillatum
Zone of my table corresponds only to the lower half of Spath’s mammillatum Zone, i.e.
his monile Subzone. Above this level there is a great change in ammonite fauna; Oto-
hoplites, Protohoplites, Hemisonneratia, Sonneratia, Pseudosonneratia, and Tetra-
hoplites disappear and are replaced by Hoplites, entry of which, in the Zone of Hoplites
dentatus, marks the commencement of the Middle Albian. It is now proposed to extend
the dentatus Zone downwards to include a Subzone of Hoplites (Isohoplites) eodentatus
sp. nov., corresponding more or less to Spath’s Subzone of Douvilleiceras inaequinodum.
The former is characteristic and widespread, being known from many localities in
south-east England and from northern France (Destombes 1958, p. 309), and is prefer-

RAYMOND CASEY: STRATIGR APHICAL PALAEONTOLOGY OF GREENSAND 499

able to D. inaequinodum , which is much too rare and has too long a range to be a suitable
horizon-marker for the base of the Middle Albian and the base of the English Gault.

DEPOSITIONAL HISTORY OF THE LOWER GREENSAND

Lower Greensand sediments were accumulated in two main basins, separated by the
relics of an Armorican mountain range that crossed the London area and extended
westwards over the south-east midlands of England. This area of Palaeozoic rocks,
now deeply buried, is one of the fundamental structural boundaries of Mesozoic
Europe and includes the so-called London Ridge or Platform, whose eastern prolonga-
tion breaks surface in the Ardennes of northern France and Belgium. The two Lower
Greensand basins thus marked off had different geological settings. In the south the
sea had long retreated and the old Jurassic trough of south-east England had become
the site of fresh- and brackish-water lagoons and delta swamps in which were deposited
the Purbeck and Wealden Beds. In the Northern Basin, on the other hand, the Cretaceous
sea never completely withdrew. At the southern end of the basin, in Cambridgeshire and
Bedfordshire, the Jurassic sea-floor was brought up within range of erosion, but north-
wards, in Norfolk and Lincolnshire, the sea maintained a footing, albeit precarious,
throughout the Neocomian.

In addition to this main structural feature which divided the Lower Greensand area
of deposition into two, there are other, minor ridges or axes of uplift that served to
demarcate sedimentary provinces within the two basins (text-fig. 1). From a study of
deep-boring records, Kent (1949) has inferred that the Sussex coast overlies a structural
nose that may be a continuation of the Paris-Plage ridge. King (1954) suggested that
this feature influenced Mesozoic deposition on an east-to-west axis stretching across
the Channel into southern England in the Beachy Head area. The attenuation of the
Lower Greensand in the Portsdown boring (Taitt and Kent 1958) is explicable on the
assumption that the westward prolongation of this ridge affected sedimentation in
Aptian times. It is here suggested that this ridge functioned as a boundary between
the Vectian and Wealden Provinces of the Southern Basin.

The Northern Basin is readily divisible into a Cambridge-Bedford Province and a
Lincolnshire-Norfolk Province, such division being apparent from the map in the
manner in which the outcrop is lost under the Fens north of Ely. Borings in this area
show that at depth the Lower Greensand has dwindled almost to nothing. At Upware,
near Ely, the Lower Greensand is banked against a ridge of Corallian rocks, but it is
unknown whether the dominant tectonic lines in Lower Greensand times ran east to
west, as I have tentatively indicated them in text-fig. 1. At the extreme north the Lower
Greensand outcrop terminates in the region of the Market Weighton upwarp. Beyond
this is preserved a portion of another Lower Cretaceous basin, that of the Speeton Clay,
outside the scope of the present work.

In the Southern Basin Lower Greensand deposition commenced with the Atherfield
Clay. The idea that the Atherfield Clay of the Isle of Wight is older than that of the
mainland (Spath 1930# ) and that the incoming Lower Greensand sea spread from south
to north is not supported by the present study of the ammonites. The same ammonite
fauna ( Prodeshayesites obsoletus and allies) is found in the Perna Beds, at the base of

500

PALAEONTOLOGY, VOLUME 3

the Atherfield Clay, in the Isle of Wight and in Surrey. On the other hand, in East Kent
there is no Perna Bed at the base of the Lower Greensand; there the succession com-
mences high in the forbesi Zone ( caJlidiscus Subzone). Commencement of Lower Green-
sand deposition in East Kent coincided with uplift in the Vectian Province, leading to
formation of the Crackers in the Isle of Wight and the Punfield Marine Band in Dorset.
A similarity between the fauna of the Punfield Marine Band (with the gastropod
Cassiope) and that of the Aptian of eastern Spain, first noticed by Judd (1871 ), has been
frequently claimed as evidence of a sea connexion between the two regions. It is tempting
to take this as supporting the notion of a westerly source for the invading Lower Green-
sand sea. Unhappily, as mentioned elsewhere, the fauna of the Punfield Marine Band
owes its distinctiveness to difference of facies: it is a marine-brackish deposit, and the
reason its fauna is not found in the Aptian of Prance is not because sea-routes did not
exist, but because conditions of normal salinity prevented the fauna using them at
that time. Too much has been made of the Spanish affinities of the Punfield fauna.
Cassiope lujani, C. Helvetica, and an assemblage of opisthobranchs comparable with
that of Punfield is found in the Upper Barremian of the Paris Basin (Gillet 1921).

Pollowing the deposition of the Atherfield Clay Series and coinciding with the arrival
of the new fauna of the deshayesi Zone, there was an influx of sandy sediment leading
to the formation of the Hythe Beds. In the Eastern Weald the sands are strongly cal-
careous, possibly owing to erosion of limestones on the London Platform (Kirkaldy
1939, p. 399). This change in fauna and sediment was probably accompanied by shrink-
age of the area of deposition, for in East Kent the Hythe Beds have a more restricted
distribution than the underlying Atherfield Clay.

The Upper Aptian was ushered in by movements more extensive than any that had
previously affected deposition of the Lower Greensand. In the Isle of Wight, in West
Kent, and in the Western Weald around Haslemere, sedimentation continued with only
minor interruptions, but elsewhere in southern England there was a period of retrench-
ment and of destruction of pre-existing deposits. In the Northern Basin the bowerbanki,
deshayesi, forbesi, and fissicostatus Zones were broken up and their rolled and phos-
phatized fossils now form a basal conglomerate to strata of nutfieldensis or later date.
Prom published evidence (Dutertre 1923, 1925; Corroy 1925) we may infer a syn-
chronous phase of movement in the Boulonnais and the Paris Basin. In the Kent coal-
field the nutfieldensis Zone rests unconformably on the forbesi Zone (Atherfield Clay)
or on remnants of the bowerbanki Zone (Hythe Beds), as it does in the Godaiming area
of Surrey. The passage from the marginal area of destruction to that of normal deposi-
tion takes place in the few miles of outcrop centred on Ashford, Kent, where the
martinioides Zone is represented by a band of phosphatic nodules and remanie fossils
at the base of the Sandgate Beds. In the Weald the martinioides Zone reaches its maxi-
mum thickness in the neighbourhood of Offham, between Maidstone and Sevenoaks,
and thins out when traced along the outcrop south-west and south-east from that area,
disappearing at Reigate and south-east of Little Chart respectively. A line connecting
these two points coincides almost exactly with the railway-line from Ashford to Redhill,
trending a few degrees north of west.

A renewed transgression commencing with the Sandgate Beds (nutfieldensis Zone)
carried the sea far into the West Country, passing over ground faulted probably at the
time of the martinioides retrenchment. The waters of the Southern and Northern Basins

RAYMOND CASEY: STR ATIGR APHIC AL PALAEONTOLOGY OF GREENSAND 501

now joined and for the first time there was interchange of Lower Greensand marine
fauna.

The Folkestone Beds saw another episode of shallowing of the basin and perhaps of
withdrawal from the newly won ground in the west. Vast quantities of quartz sand
poured into the basin and were swept around by strong currents before coming to rest.
Here and there the bedding structures and well polished sand-grains suggest that the
sands are in part re-sorted coastal dunes or near-surface sand-bars. Possibly the Western
Outliers, e.g. Seend and Caine, were oxidized at this time.

The change from Aptian to Albian time occurred while the Folkestone Beds were
being formed and at the margins of the basin in East Kent and East Sussex it was marked
by a break in the flow of sediment and winnowing of the sea-bed. In Surrey, where the
beds are thickest, the only physical expression of the stage boundary was an interval of
slow deposition and slack water. Deposition of the middle part of the tardefurcata Zone
(, milletioides Subzone) was followed by a period of instability during which the sedi-
ments were folded gently along east-to-west axes and eroded. A long phase of inhibited
deposition then ensued until the end of the Lower Greensand, this phase being charac-
terized by the formation of phosphatic nodule beds (often with radioactive enrichment)
of regularis and mammillatum age. Downwarped areas provided a maze of local troughs
or ‘dimples’ in which sediments of regularis age — sand, clay, or limestone — were laid
down; elsewhere, on the crests and flanks of the folds, erosion or oxidation proceeded
and sedimentation was not established until some time in the mammillatum Zone.
Strata above this mid -tardefurcata break thus show great lateral variation, the basement-
beds varying in age and lithology from place to place depending on the site of the
trough and the position they occupy in it. This is the most widespread gap in the
Lower Greensand; everywhere below the regularis Subzone in England there is a sharp
junction, either a plane of erosion, a bed of phosphatic nodules, or a sudden change in
lithology. In the Isle of Wight the critical level is at the junction of the Sandrock and
the Carstone; around Leighton Buzzard it is the iron-cemented top of the ‘Silver Sands’.
In East Kent the outcrop gives a natural section through a regular is-mammillatum
trough scooped out in milletioides and jacobi sediments. Proceeding north-eastwards
from East Cliff, Folkestone, where the regularis sandstones are preserved in the centre
of the trough, we may observe the various subzones wedging out, until at Quarrington
Wood, about 10 miles from East Cliff, the topmost part of the mammillatum Zone passes
into an ironstone at the rim of the trough. We are at present unable to define the extent
of this unconformity or disconformity in terms of ammonite chronology. The only
region where the tardefurcata Zone succession is known in detail is in north Germany,
and there the ammonites are of a different facies. Possibly the German Subzone of
Leymeriel/a acuticostata , which lies next below that of L. regularis , bridges this gap in
the Lower Greensand. The fact that neither L. acuticostata nor any other ammonites
that could fill this gap are known in France may indicate that the effects of mid-
tardefurcata movement were felt over a much wider area than the British Province.

STRATIGRAPHICAL ACCOUNT

Lower Greensand outcrops are shown on the map in text-fig. 1. These outcrops are
discontinuous and extend from Folkestone, Kent, westwards to Dilton, Wiltshire, and

502

PALAEONTOLOGY, VOLUME 3

from the southern tip of the Isle of Wight northwards to the neighbourhood of Market
Weighton, Yorkshire. This great expanse of Lower Greensand country may be divided
into the following geographical provinces, each a natural sedimentary trough :

Southern Basin Northern Basin

Vectian Province Cambridge-Bedford Province

Wealden Province Lincolnshire -Norfolk Province

SOUTHERN BASIN

Vectian Province

The Vectian Province comprises the Isle of Wight and a small part of the Dorset mainland
where a strip of Lower Greensand extends westwards from Swanage to Lulworth Cove.

Isle of Wight

The greater part of the southern or Cretaceous area of the Island is occupied by the Lower
Greensand, but the country is largely under cultivation and inland exposures are few and
insignificant. The formation is best seen in the sea-cliffs at Chale Bay and Compton Bay, on
the south-west side of the Island, and at Shanklin and Redcliff, on the south-east side. At
Redcliff the Lower Greensand is about 600 feet thick; at Chale Bay it has increased to over
800 feet, but at Compton Bay, about 16 miles west of Redcliff, the thickness is reduced to
400 feet. At Punfield, near Swanage, on the Dorset coast, 20 miles west of Compton Bay, it
is no more than 198 feet. This means that the direction in which the formation as a whole
thickens most rapidly lies a little east of south. An exception to this generalization is the
Carstone, at the top of the formation, which is thickest at Redcliff.

The Lower Greensand of the Isle of Wight was carefully examined by Fitton in the years
1824-47 and the results of his work appeared in a number of papers, chiefly in that published
in 1847. This paper gave a bed-by-bed description of the Atherfield (Chale Bay) section and
correlated it with strata in other parts of the Island, on the mainland, and in France. Fitton
employed a professional collector, Charles Wheeler, to help him collect fossils and most of
these were passed to J. Morris for naming, a few being done by T. Lonsdale and J. de C.
Sowerby. Fitton divided the Lower Greensand of the Isle of Wight into six major units
(lettered A-F), sixteen ‘Groups’ (given names and roman numerals), and fifty-five beds.
Ibbetson and Forbes (1845) had measured the section independently, but later authors were
content to paraphrase Fitton’s account and to add little or nothing new (e.g. Bristow 1862,
1889; Wright 1864; Norman 1887; Leriche 1905; Osborne White 1921; Chatwin 1935;
Kirkaldy 1939). Fitton’s scheme, emended by the Geological Survey (Strahan in Bristow 1889)
and with further small changes now introduced, is shown in Table 2 on p. 504.

In place of Fitton’s six major units the Geological Survey found it better to adopt a fourfold
division, as follows: (1) Atherfield Clay, (2) Ferruginous Sands, (3) Sandrock, and (4) Carstone.
The Atherfield Clay was extended to take in the Pema Beds below and the basal portion of
the Crackers Group (Lower Lobster Bed) above. Similarly, the Ferruginous Sands, though
roughly equal to Fitton’s division D (Groups IV to XIV), embraced also most of Group III
(Crackers) and a thick bed of sandy clay at the base of Group XV. Most of Fitton’s Group XV
and a portion of the overlying Group XVI were brought together by the Survey under the
name Sandrock, the term Carstone being reserved for the rest of Group XVI. As redefined
by the Survey the Atherfield Clay was considered equivalent to the beds of the same name on
the mainland; the Ferruginous Sands were correlated with the Hythe and Sandgate Beds of

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 503

the mainland; and the combined Sandrock and Carstone were taken to represent the Folke-
stone Beds.

text-fig. 1. Distribution of the Lower Greensand. Outcrops are shown solid black, provincial

boundaries by broken lines.

The only change in the Survey’s scheme now advocated is in the boundary of the Atherfield
Clay and the Ferruginous Sands. Fossils and lithology show that Fitton was right to draw the
line between his divisions C and D at the base of the Lower Gryphaea Beds (Group IV). The

504

PALAEONTOLOGY, VOLUME 3

Crackers rocks, included in the Ferruginous Sands by the Survey, are merely a local sandy
phase in an essentially argillaceous succession. Not only are the beds next above them (Upper
Lobster Beds) Atherheld Clay in the lithological sense, but it now appears that they are the
correlatives of much of the Atherheld Clay of the mainland. It is proposed, therefore, to take
the bottom of Group IV as the base of the Ferruginous Sands and to designate the beds below,
i.e. the Perna Beds, Atherheld Clay s.s., and the Crackers Group (with Lower Lobster Bed at
base and Upper Lobster Beds at top), the Atherheld Clay Series.

table 2. Divisions and subdivisions of the Lower Greensand of the Isle of Wight

Fitton 1847

Geological
Survey 1889

Present

author

XVI. Various Sands and Clays

F

Carstone

Carstone

XV. Upper Clays and Sandrock

E

Sandrock

Sandrock

XIV. Ferruginous Bands of Blackgang Chine
XIII. Sands of Walpen Undercliff
XII. Foliated Clay and Sand
XI. Cliff-end Sands
X. Upper Gryphaea Beds
IX. Walpen and Ladder Sands
VIII. Upper Crioceras Beds
VII. Walpen Clay and Sand
VI. Lower Crioceras Bed
V. Scaphites Beds
IV. Lower Gryphaea Beds

D

Ferruginous

Sands

Ferruginous

Sands

| Upper Lobster Beds
III. Crackers

1 Lower Lobster Bed

C

Atherheld
Clay Series

Atherheld

Clay

II. Atherheld Clay

B

I. Perna Beds

A

Atherfield ( Chaie Bay). By far the best section, and the most productive of fossils, is that of
Chale Bay. Here a shallow embayment of the coast extends from Atherheld Point south-
eastwards to Rocken End, a distance of 31- miles, in which the gentle dip of the strata brings
the whole thickness of the Lower Greensand into view. To conform with earlier literature the
Chale Bay exposure will be referred to simply as Atherheld.

The cliffs are high and precipitous, lapped by the waves at high water. Between Atherheld
Point and Rocken End ascent from the shore can be made only at one point, Whale Chine,
about a mile from Atherheld Point; it is essential, therefore, to examine the section on a
receding tide. Cliff falls are frequent and much of the section may be obscured by downwash,
and from time to time sand and shingle smother the foreshore exposures, the best places for
collecting fossils. Fossils occur mostly in hard concretions and serious work requires a sledge-
hammer.

Text-hg. 2 illustrates the zonation of the Atherheld section and its correlation with the Lower
Greensand of the Kent coast. The full thickness of the beds at Atherheld has been variously
estimated at 808 feet (Fitton 1847a), 833 feet (Ibbetson and Forbes 1845), and 752 feet 1 1 inches
(Simms 1845). Experience has taught me to take Fitton as the guide.

RAYMOND CASEY: STR ATIGR APHICAL PALAEONTOLOGY OF GREENSAND 505

Atherfield Clay Series. The basal few feet of the Lower Greensand at Atherfield form a well-
defined division, distinguished by Fitton as the Perna Beds in consequence of their containing
large numbers of the lamellibranch Mulletia mulleti, formerly called Perna mulleti. This divi-
sion appears at the top of the cliff about 300 yards south of Shepherd’s Chine and descends to
the beach 1 50 yards east of Atherfield Point, where it forms a ledge running out to sea. Fre-
quently this part of the section is concealed by slips and mudflows from the Atherfield Clay,
though fossil-collecting can always be done from the boulders on the beach. The fauna is
one of the richest in the Lower Greensand, the sandstone at the top being full of large lamelli-
branchs such as Mulletia mulleti , Isognomon ricordeanus, Gervillella sublanceolata, Gervillaria
alaeformis, Exogyra latissima, Prohinnites favrinus, Sphaera corrugata, Protocardia spltaeroidea,
Astarte obovata, Venilicardia protensa, Noramya forbesi, and Yaadia nodosa, together with
the nautiloid Cymatoceras radiatum, the gastropods Fossarus munitus and Globularia sublaevi-
gata, the brachiopods Sulcirhynchia hythensis and Sellithyris sella, knobs of coral ( Holocystis
elegans), and many other fossils. The hydrozoan Lonsda contortuplicata, once thought to be a
sponge, is found here. Indigenous ammonites are very rare. This is the only horizon in the
Lower Greensand where corals are abundant and the occurrence may be linked with the great
phase of reef-building that characterized the Barremian-Aptian deposits of the Tethyan belt
from southern Europe to Venezuela and Mexico.

Section of the Perna Beds at Atherfield Point

ft. in.

3. Grey-green calcareous sandstone, ironstained in patches. Very fossiliferous . 2 6

2. Dark greenish-blue sandy clay with pyrites. Many fossils, including Panopea

standing upright ........... 2 6

1. Line of grit, small pebbles, fish debris, and small black nodules, some rolled

bits of Kimmeridgian Pavlovia ........ 1

Sharp junction with Wealden Shales

Teeth of Hybodus, Acrodus, and the primitive myliobatid Hylaeobatis problematica occur
not only in the basal grit but also (more rarely) in beds 2 and 3; most are derivatives from the
Wealden, though some may belong to the native fish fauna, as apparently do the associated
teeth of Scapanorhynchus. Fitton called the basal grit and the clay above the ‘Lower Perna
Bed’ and said that it contained the ammonite A. furcatus. This is an impossible horizon for
the species, a form of Dufrenoyia, and I feel sure that what Fitton found was a derived Jurassic
Pavlovia. No ammonites have been obtained from bed 2, though at Sandown this bed has
yielded Prodeshayesites obsoletus gen. et sp. nov. This species occurs at both localities in
bed 3, the ‘Upper Perna Bed’ of Fitton, accompanied by specifically indeterminate Deshayesites,
and is the species recorded from here by him (1847, p. 296) and other authors as A. leopoldinus.
The identity of the other ammonites listed by Fitton from this bed, namely A. deshayesi, A.
furcatus, and A. inflatus, can only be surmised. Ammonites inflatus was recorded by him also
from the Atherfield Clay and is a puzzling addition to the faunal list, this species being an
Upper Albian Mortoniceras. In the Geological Survey Museum there is a specimen of ‘A.
inflatus' from ‘Atherfield’ which was originally presented by Fitton to the Geological Society
and which was cited by Forbes (1845, p. 355) (GSM Geol. Soc. Coll. 2296). It is a specimen of
Mortoniceras fissicostatum (Spath), of Upper Albian age, and is in a malmstone quite foreign
to the Lower Greensand, though typical of Potterne, Wiltshire, one of the principal sources
of the Mortoniceras fauna in this country. The citation of Ancyloceras matheronianum from
the Perna Beds of Atherfield is incorrect (Casey 1960a, p. 22).

Spath’s supposition that the Perna Beds are of bodei age (Spath 19236) was a close approxi-
mation to the truth. In my zonal scheme the Perna Beds represent the upper half of the

506

PALAEONTOLOGY, VOLUME 3

fissicostatum Zone ( obsoletus Subzone). The gritty seam at the base may mark an interval of
time during which the lower half of the zone (bodei Subzone) was laid down elsewhere.

The Perna Beds are followed upwards by 60 or 70 feet of brown- weathering, bluish-grey,
silty clay, to which Fitton originally applied the name Atherfield Clay. The clay is devoid of
lamination and abounds in flattened nodules of red or white clay-ironstone, red nodules
predominating in the lower part. In places the clay has the qualities of fuller’s earth. Wasting
of the Atherfield Clay is rapid and unceasing and a clear, measurable section is rarely seen
owing to cliff-founders and sludge-streams. Most of the ammonites here recorded from this
part of the succession were obtained in the few months following the great gale of October
1954, when the shore was stripped down to bedrock. The fauna is dominantly molluscan, the
lamellibranchs Nuculana scapha, Aptolinter aptiensis, Pseudoptera subdepressa. Pinna robin-
aldina, Panopea gurgitis, Resatrix dolabra, Parmicorbula striatula, and the gastropod Anchura
( Perissoptera ) robinaldina being locally common. Ammonites are less frequent and occur
either as crushed impressions in the clay or as clay-ironstone internal moulds, generally of the
body-chamber only. They include a few fragments of Prodeshayesites from the bottom 15 feet,
a single Roloboeeras from 20 feet above the Perna Beds, and various Deshayesites. Of these,
the zone fossil D.forbesi sp. nov. ranges almost throughout, but the chief form is the subzonal
index, D.fittoni sp. nov., which has been collected in numbers between 10 and 40 feet above
the Perna Beds. Spath’s ‘ Procheloniceras cf. pachystephanus (Uhlig)’ and ‘ Procheloniceras
? sp. indet.’, said to be the only two ammonites known from the Atherfield Clay (Spath 1930a,
pp. 422, 443), are distorted pieces of large Deshayesites. Spath (19236) correlated the Atherfield
Clay with the weissi Zone of the German Aptian, but Deshayesites weissi has not been found
here, as supposed by Neaverson (1928). Not infrequently bunches of branched tubes composed
of tiny ovoid pellets weather out from the clay; similar structures in the London Clay are
known as ‘ Granularia' and are thought to be the faeces of holothurians or annelids. Washed
samples of the clay give a large sand residue almost barren of microzoa.

The succeeding Lower Lobster Bed is an impure fuller’s earth, brownish or bluish-grey,
with white clay-ironstone nodules like those in the beds below. Near the top it has small
sandy concretions similar to those which occur on a large scale in the Crackers above. It crops
out on the shore north-west of the promontory of Crackers rocks but is seldom free of shingle.
In the cliff it is generally concealed by slipped material, though fallen blocks are at times
available. Fitton gives the thickness of the bed as 25 feet 6 inches; Ibbetson and Forbes as
29 feet. Fossils are much more abundant and better preserved than in the Atherfield Clay.
Many of the molluscs have the test, others are internal moulds in calcite, clay-ironstone, or
iron-pyrites. Commonly the larger fossils are coated with adherent oysters and some appear
water-worn. In the words of Fitton: ‘the fossils of this bed occur in thin clots or clusters, often
without any covering or crust, as if they had been just left upon a sand-bank at the bottom of
the sea’. The bed takes its name from the common occurrence of the prawn Meyeria magna
(= M. vectensis). A small crab, Mithracites vectensis, also occurs, but true lobsters, such as

EXPLANATION OF PLATE 77

Fig. 1. Whale Chine, Isle of Wight, looking north-west to Atherfield Point. The Lower Crioceras
Bed crops out by the posts at the mouth of the Chine. Concretions of the Upper Crioceras Beds
may be seen in ranges high in the cliff and as boulders on the beach.

Fig. 2. The Crackers, Atherfield. The upper line of concretions is seen in situ just below the middle of
the photograph, followed upwards by Upper Lobster Beds and, high in the cliff. Ferruginous Sands.
Fig. 3. Ladder and Walpen Chines, Isle of Wight, looking south-east to St. Catherine’s Point. Cliffs
of Ferruginous Sands. Group VII in the foreground.

Geological Survey and Museum photos. Reproduced by permission of the Controller, H.M. Stationery
Office. Crown copyright.

PLATE 77

CASEY, Lower Greensand , Isle of Wight

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 507

the nephropsid Homarus longimanus and the palinurid Linuparus carteri, are much less frequent.
The lamellibranch and gastropod fauna is more or less the same as that of the Crackers. The
ammonites include: Deshayesites forbesi sp. nov., D. kiliani, D. topleyi, D. punfieldensis, D.
spp. nov., Roloboceras hambrovi, R. perli, R. horridum, R. spp. nov., Megatyloceras sp. nov.
The subzonal index is Deshayesites kiliani; Roloboceras hambrovi is equally characteristic but
ranges into higher beds. It is here that Roloboceras reaches its acme, and its association with
the allied genus Megatyloceras is especially interesting. Megatyloceras has not been found in
Britain before and the only other known authentic occurrences of the genus are in Georgia,
U.S.S.R., and Yonne, France, both poorly localized in the succession. This is the horizon of
Heminautilus saxbyi, a discoidal nautiloid; also of Deshayesites punfieldensis, wrongly attri-
buted to the Punfield Marine Band.

Next above the Lower Lobster Bed is the Crackers, 20 feet of firm grey and brown clayey
sand with two lines of sandy calcareous concretions, exceptionally rich in fossil mollusca.
These concretions are large and rounded but of an irregular size and shape like boulders,
those in the lower tier reaching 6 or 7 feet in length and 2 feet in thickness. Some are cemented
clusters of a single species of shell, usually the lamellibranchs Gervillella sublanceolata or
Yaadia nodosa, or the ammonite Deshayesites forbesi. Smaller nodules, with stony crust only
2 or 3 inches thick, generally have a softer core from which fossils may be extracted in perfect
condition, horn-coloured and translucent. On account of their relative hardness, the Crackers
make a prominence in the cliff-line about 600 yards east of Atherfield Coastguard Station.
Here the sea hollows out the sand below the concretions and the rush of the waves in the
cavities, driving before them a volume of air, produces a sharp concussion, whence the name
‘Crackers’. Always a favourite horizon for the fossil-collector, the Crackers have furnished a
wealth of material for museums and private cabinets. They are the source of many of Forbes’s
and Woods’s type-specimens of mollusca and have provided material for description by
Withers (1945) of the minute crab Vectis wright i and the cirripede Virgiscalpellum wright i
and by White (1927) of the pycnodont fish Gyrodus atherfieldensis. Finds by Wright and Wright
(19426; 1950(3) also include the earliest known Cretiscalpellum and further examples of the
cephalopod Conoteuthis, shown by Spath (1939a) to be the phragmocone of a peculiar belem-
nite with a very short guard. Echinodermata are represented by Trochotiara fittoni and ossicles
of the starfish Lophidiaster. Brachiopods are seldom found. Some of the commoner or more
interesting mollusca are — Cephalopoda: Ancyloceras mantelli, Aconeceras nisoides, A. cf.
haugi (all very rare), Deshayesites forbesi sp. nov., D. callidiscus sp. nov., D. topleyi, Roloboceras
hambrovi, R. perli, Conoteuthis vectensis, C. dupiniana. Lamellibranchia: Aptolinter aptiensis,
Cucullaea fittoni, Gervillella sublanceolata, Brachidontes vectiensis, Yaadia nodosa, Thetironia
minor, Nemocardium ( Pratulum ) ibbetsoni, Protocardia anglica, Mactromya vectensis, Mediraon
sulcatum sp. nov., Venilicardia saussuri, V. anglica, Vectianella vectiana, Resatrix parva, R.

( Vectorbis ) vectensis, Scittila nasuta, Senis wharburtoni, Pholadomya gigantea, P. martini.
Gastropoda: Sulcoactaeon marginata, Tornatellaea aptiensis, Ovactaeonina forbesiana, Globu-
laria cornueliana, Anchura ( Perissoptera ) glabra, A. ( P.) robinaldina, Tessarolax moreausianum,
T. fittoni, Dimorphosoma ancylocltila, D. kinklispira, Uchauxia forbesiana, Mesalia ( Bathra -
spira) neocomiensis, Turritella ( Haustator ) dupiniana, Cassiope pizcuetana (very rare). Deshaye-
sites makes up the greater part of the ammonite fauna, there being many undescribed species.
Roloboceras is much less frequent and in my experience is restricted to the lower tier of concre-
tions. The Crackers and the Upper Lobster Beds are taken together as the topmost part of the
forbesi Zone with D. callidiscus as the subzonal index. This is the least rare of the ammonites
confined to this horizon, though the overwhelming dominance of D. forbesi itself is the chief
feature of this subzone.

The Upper Lobster Beds, 40 feet thick, were divided by Fitton into five beds (beds 6-10) of
approximately equal thickness. They consist of alternations of brown-weathering, grey silty

508

PALAEONTOLOGY, VOLUME 3

clay and grey sandy clay. Fossils are much less frequent than in the Crackers or the Lower
Lobster Bed, though the crustacean Meyeria magna and the echinoid Toxaster fittoni are
usually present. Washed samples of the clays yield a few foraminifera. Ammonites occur
crushed flat in the clay or as internal moulds in clay-ironstone or iron pyrites; the last are
prone to decomposition. Deshayesites forbesi is found in all the beds; other ammonites (many
undescribed), due either to accidents of collecting or natural restriction, are known only from
certain beds. Sanmartinoceras ( Sinzovia ) aptiana, Pseudosaynella cf. fimbriata, and P. aff.
undulata have been found in bed 10 and nowhere else in the Lower Greensand. At the base of
this bed are large Deshayesites encrusted with serpulae.

Ferruginous Sands. The Ferruginous Sands begin with Fitton’s Lower Gryphaea Group
(Group IV), which is divisible into three beds, as follows:

Section of the Lower Gryphaea Group of Atherfield

ft. in.

3. Firm dark reddish-brown sand with polished fragments of ironstone. The
top 2-3 ft. full of large Exogyra latissima and clusters of pebbles cemented by

black phosphorite . . . . . . . . . 10 0

2. Coarse brown sand, crowded with brachiopod shells ( Sellithyris sella and

Sulcirhynchia spp.) .......... 2 0

1. Grey-green, glauconitic, clayey sands, weathering brown, with rusty streaks;

portions of the sand are indurated into spherical nodules up to the size of a
football, some impregnated with phosphorite . . . . . .16 0

Total 28 0

The nodules of bed 1 are crowded with fossils in a good state of preservation, with many
small gastropods and lamellibranchs like those in the Crackers and the ammonites Deshaye-
sites deshayesi, D. consobrinoides, D. multicostatus , D. cf. involutus, Cheloniceras sp., Toxocera-
toides royerianus, T. cf. fustiformis. This fauna has much in common with that of the Argiles
a Plicatules of Bailly-aux-Forges (Haute-Marne), in the Paris Basin, the type locality of
Deshayesites deshayesi, and its discovery at the base of the Ferruginous Sands is a big step
forwards in the study of the Lower Greensand succession.

The ‘ Terebratula bed’ (bed 2) contains a profusion of brachiopods with a few Exogyra,
Gervillaria and polyzoa, and little else. Fitton noted the abundance of Pinna robinaldina at the
bottom of the overlying sand (bed 3), which also yields large examples of Exogyra latissima
(‘ Gryphaea ’) and Prohinnites favrinus. A few phosphatized and semi-phosphatized specimens
of Cheloniceras parinodum sp. nov. and Deshayesites cf. involutus have been collected and the
pebble-clusters at the top frequently enclose a sheaf of Serpula tubes. This Lower Gryphaea
Group is the parinodum Subzone of the deshayesi Zone of my classification. It was misnamed
the ‘grandis- bed’ by Spath in the belief that it was the source of the common Deshayesites
grandis.

The top of the Lower Gryphaea Group forms a ledge slanting across the beach about 350
yards west of Whale Chine and makes a good datum-line for identification of the next set of
beds, the Scaphites Group (Group V) of Fitton, detailed below.

Section of Scaphites Group of Atherfield, west of Whale Chine

ft. in.

4. Dark grey-green, glauconitic muddy sand with large Exogyra latissima at top c.21 0

3. Large red-stained calcareous concretions (up to 2 ft. in length) disposed

roughly in two lines. Much calcite in veins and covering fossils . . .50

2. Brown-weathering, grey-green, glauconitic sand. A few indurated nodules . 18 0

1 . Nodules of grey, argillaceous and sandy phosphorite, crowded with fossils . 2-4

Total about

50 0

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 509

The nodules at the base, not previously recorded, are rich in small fossils: juvenile gastropods
(especially Anchura ), tellinids, venerids, ammonite young and nuclei and serpulae. Hetero-
morph ammonites such as Toxoceratoides royerianus and T. cf. fustiformis are common,

AT HERFIELD

ZONES

LOWER ALBIAN

UPPER APTIAN

LOWER APTIAN

text-fig. 2. Comparative sections of the Lower Greensand of Atherfield, Isle of Wight, and of the

Folkestone-Hythe area of Kent.

together with Deshayesites grandis, species of Cheloniceras, and, rarely, Aconeceras cf. nisoides.
Carbonized leaf impressions of Weichselia reticulata also occur. Deshayesites grandis, Australi-
ceras gigas, Cheloniceras cornuelianum, and Ch. crassum are found in the nodules in bed 2,

L 1

B 6612

510

PALAEONTOLOGY, VOLUME 3

but the principal source of fossils in this Group is the large concretions of bed 3. When the
foreshore is clear of shingle they may be seen to crop out in a band from the foot of the cliff
to the water’s edge, each balanced on a pedestal of the underlying sand like so many giant
mushrooms. Split open, almost every one reveals a large ammonite as its nucleus of growth,
either the ancyloceratid Australiceras (‘ Scaphites'), or Deshayesites or Cheloniceras, from 1 to
2 feet across. Smaller fossils occur in the matrix of the larger ones, including the nautiloids
Cymatoceras pseudoelegans, Eucymatoceras plication, the lamellibranchs Panopea gurgitis,
Sphaera corrugata, Pterotrigonia mantelli anterior, Pseudaphrodina ricordeana, the echinoids
Tetragramma malbosi and Hyposalenia wrighti, and the crustacean Homarus longimanus. The
following ammonites have been identified: Australiceras gigas, A. sp. nov., Epancyloceras
hythense, Tonohamites decurrens, Cheloniceras cornuelianum, Ch. crassum, Ch. kiliani, Ch. spp.
nov., Deshayesites grandis, D. vectensis, D. spp. nov. The oft-quoted record of Tropaeum hillsi
in this bed or that above (bed 4) has not been confirmed. On account of the frequency of
Deshayesites grandis this Group is designated the grandis Subzone of the deshayesi Zone.

The Lower Crioceras Bed (Group VI), 16 feet 3 inches thick, contains concretions arranged
in irregular lines and embedded in grey-green muddy sand. With them are smaller, intensely
hard, concretions of dark-brown or black phosphorite veined with calcite, each a nest of
fossils. In the top of the Group ammonites occur incompletely phosphatized, the unphos-
phatized portions being crushed flat in the sand. The upper part of the Group comes down to
the shore to form ledges at the mouth of Whale Chine. Ammonites collected from this bed are:
Tropaeum bowerbanki. Id. var. densistriatum, Australiceras gigas, Tonohamites decurrens,
Cheloniceras cornuelianum, Ch. crassum, Ch. spp. nov., Dufrenoyia furcata, D. truncata, D.
lurensis, D. transitoria sp. nov. D. spp. nov. Part of this assemblage has ranged up from the
Scaphites Beds, but the ammonite fauna as a whole is quite distinct, characterized by the
incoming of giant crioceratitid Tropaeum and the replacement of Deshayesites by Dufrenoyia.
The rest of the fauna is similar to that of the Scaphites Beds, composed mainly of long-ranging
lamellibranchs and the ubiquitous Lower Aptian brachiopod Sellithyris sella. This is the
transitoria Subzone of the bowerbanki Zone.

Next above is the Walpen Clay and Sand (Group VII), divided as follows:

Section of Walpen Clay and Sand ( Group VII) at Whale Chine

ft. in.

2. Grey sandy clay with small nodules of phosphorite and pyrites . . . 33 0

1 . Dark greenish-grey muddy sands and sandy clay with phosphorite concretions 24 0

Total 57 0

The phosphorite concretions of bed 1 are from 2 to 6 inches in diameter and contain numerous
fossils. Near the bottom of the bed they are compact and difficult to crack open, like those in
the bed below; above they are less dense and fall to pieces under the hammer. Tropaeum
bowerbanki occurs at intervals through the sand and the nodules contain Cheloniceras cor-
nuelianum, Ch. crassum, Ch. meyendorff, Dufrenoyia furcata, D. lurensis, Tonohamites aequi-
cingulatus with many lamellibranchs, including Cucullaea cornueliana, Thetironia minor,
Panopea gurgitis, Chlamys robinaldina, Resatrix hythensis, and Area dupiniana. This lower part
of Group VII is accessible at the bottom of Whale Chine and forms the undercliff that runs
south-eastwards to Ladder Chine. Its junction with the overlying sandy clay of bed 2 is marked
by a line of water seepage. Fossils are much less numerous in bed 2 and are mostly phos-
phatized ammonite body-chambers. Dufrenoyia lurensis, Ch. meyendorff, and an obese variety
of Ch. kiliani have been found. Group VII is the upper half of the bowerbanki Zone ( meyendorff
Subzone) and marks the top of the Lower Aptian.

The Upper Aptian part of the succession commences with the Upper Crioceras Beds (Group

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 511

VIII), about 46 feet of grey, brown-weathering, clayey sand with four or more ranges of big
concretions. The top of the Group reaches the beach east of Walpen Chine, but the best
exposures are in Whale Chine, where one can climb the slopes to collect in situ or forage
among the tumbled blocks in the bottom of the Chine. Cheloniceras (Epicheloniceras) marti-
nioides and Ch. (E.) debile spp. nov. are the characteristic ammonites; Ch. (E.) tschernyschewi
occurs rarely. Giant ancyloceratids, formerly included in ' Crioeeras' bowerbanki, are repre-
sented by Tropaeum benstedi and an undescribed Ammonitoceras. Gervillella sublance olat a,
Panopea gurgitis, Yaadia nodosa, Linotrigonia ( Oistotrigonia ) ornata, Venilicardia sowerbyi,
and Anchura ( Perissoptera ) robinaldina are some of the other molluscs found here. Fossils tend
to lie in clusters in the rock, which is often impregnated with calcite and contains scattered
bits of drift-wood and fronds of Weichselia. These beds are the debile Subzone of the niartinioides
Zone.

The Walpen and Ladder Sands (Group IX) are 42 feet of greenish and grey sand with a line
of gritty calcareous concretions at the base, olive-green in colour. About 18 inches thick and
up to 4 feet long, these concretions each enclose a rounded mass of brown phosphorite full
of fossils. They come down to beach level east of Walpen High Cliff and break into a line of
boulders running out to sea. Heavy hammer and chisel are needed to extract the fossiliferous
cores, many of which are graveyards of small ammonites, usually young Epicheloniceras, or
colonies of Sellithyris. Secondary calcite invades all the cracks and cavities, lining the insides
of brachiopods and producing casts of ammonite phragmocones. Ammonites include Ammoni-
toceras sp. nov., Australiceras sp., Ch. ( E .) tschernyschewi, Ch. (E.) aff. volgense, Ch. ( E. ) gracile
sp. nov., Ch. (E.) spp. nov., Aconeceras cf. nisum, and an unnamed genus and species of
Aconeceratidae. The lamellibranchs Inoceramus neocomiensis, Resatrix parva. Modiolus
aequalis, Thetironia minor, the gastropod Chilodonta ( Agathodonta ) dentigera, the echinoids
Phyllobrissus fittoni and Toxaster fittoni, and a large Serpula are also characteristic. About
6 feet higher is a thin sandstone ledge with masses of intertwined serpulae, but the remaining
thickness of sand is poor in fossils. The whole Group, together with the overlying Group X,
comprises the gracile Subzone of the niartinioides Zone.

Group X, the Upper Gryphaea Beds, includes about 16 feet of ferruginous clayey sands
with bands of Exogyra latissima in the lower 12 feet. It crops out in a mural cliff-face and is
difficult to work. Fitton mentioned the presence of nodules with ''Ammonites martini ’ about
4 feet from the base; the only fossils I have seen at this level are badly preserved cheloniceratids
resembling those at the base of Group IX and a few long-ranging lamellibranchs. Plant debris,
with fronds of Weichselia, occurs throughout.

The succeeding 28 feet of glauconitic sands and clays that comprise Fitton’s Cliff-end Sand
(Group XI) is divisible into two beds of equal thickness. At the base of the lower bed are some
ferruginous gritty lumps embedded in a more argillaceous matrix and with much carbonized
plant debris; their poorly preserved fossils include Cheloniceras ( Epicheloniceras ) cf. buxtorfi
and tiny aconeceratids. Two feet higher is the thin clay seam from which Fitton recorded
Trigonia, and near the top of this lower bed are said to be concretions with Pinna. The upper
half of Group XI is chiefly remarkable for its cylindrical, branching concretions and lenses
of current-bedded greensand and is barren to the palaeontologist. No fossils have been found
in the next Group, the Foliated Clay and Sand (Group XII). This consists of 35 feet of inter-
laminated glauconitic sand and dark-blue pyritic clay, with some lenticular masses of coarse,
current-bedded, friable sandstone capped by 10 feet of white sand and sandstone, giving a
depressing preview of the great thickness of unfossiliferous sands encountered in the Sandrock
high above. These last two Groups are assigned with question to the buxtorfi Subzone, the
topmost part of the niartinioides Zone.

Glauconitic pale-green, yellow, and brown sands, about 90 feet thick, form the Sands of
Walpen Undercliff (Group XIII) and occupy the base of the cliff for about 700 yards below

512

PALAEONTOLOGY, VOLUME 3

Blackgang Chine. I agree with Kirkaldy (1939, p. 394) in drawing the lower limit of this Group
at the conspicuous pebble-bed (bed 42a of Fitton), the underlying 10 feet (‘First Sandrock’)
being better accommodated in Group XII. At about the middle of this Group, 200-250 yards
east of the cascade at Blackgang Chine, large fossiliferous nodules with moulds of rhyn-
chonellids, and Pterotrigonia mantelli, Thetironia minor, immature Parahoplites and other
molluscs, weather out at the foot of the cliff. Similar nodules are found fresh in the same Group
at Horse Ledge, Shanklin, where they yield a larger fauna of ammonites denoting the sub-
arcticum Subzone of the nutfieldensis Zone.

The Ferruginous Bands of Blackgang Chine (Group XIV) consist of about 20 feet of brown
and yellow sands with three ranges of iron concretions abounding in moulds and impressions
of fossil shells. These sands rise from the shore about midway between Blackgang Chine
and Rocken End and the topmost range of concretions is responsible for the cascade in the
Chine. Lengthy lists of fossils from this Group have been published by Fitton and Norman;
the latter author (1887, p. 47) records Iguanodon remains. Lamellibranchs and gastropods make
up the bulk of the fossiliferous masses, the following being especially characteristic: Thetironia
minor, Lucina cornueliana, Pterotrigonia mantelli, Cucullaea cornueliana, Senis wharburtoni,
Resatrix parva, Parmicorbula striatula, Globularia sublaevigata, and Anchura ( Perissoptera )
robinaldina. Drift-wood occurs and Norman mentions the presence of cycads. No cephalopods
have been recovered here from this Group despite the profusion of other molluscs, and its
position in the cunningtoni Subzone of the nutfieldensis Zone is inferred from finds elsewhere.
The overlying 40 feet of dark-grey sandy clay, included by Fitton in his Group XV, and
apparently equivalent to the Marehill Clay of the Pulborough region, is without recognizable
fossils. It may represent the nolani Subzone of the jacobi Zone.

Sandrock. This division here attains a thickness of 186 feet and is composed of white and yellow
quartz sand and sandrock. Apart from a little plant debris the beds are practically barren of
organic content, though Lamplugh (1901) recorded the discovery in the slopes south-east of
Blackgang Chine of a band of ferruginous concretions with casts of marine bivalves 10 feet
below the top of the division. Like Jackson (1939, p. 74) I have searched in vain for these
concretions, though the precipitous nature of the slopes, vegetation, and downwash from the
Gault prevent a critical examination of the section. This part of the succession must fall
within the jacobi and tardefurcata Zones.

Carstone. This division forms the top of the Lower Greensand, consisting of 12 feet of gritty
reddish-brown sands with pebbles and phosphatic nodules, and rests with sharp junction on
the sands below. It is accessible in the Undercliff near Blackgang and in tumbled blocks on
the beach far below. I have collected Anadesmoceras baylei from the pebbly basement-bed
and the Museum of Isle of Wight Geology possesses a fine example of Sonneratia kitchini
in a Carstone matrix picked up from the beach.

In the vicinity of St. Catherine’s Point the Lower Greensand is hidden beneath a mantle of
slipped Gault and Upper Greensand, but here and there the top of the Sandrock and the
Carstone may be seen in low cliffs and shore-ledges, and boulders of Carstone strew the floors
of the coves. In Reeth Bay, Puckaster Cove, and in Watershoot Bay the beach includes lumps
of phosphorite-cemented grit and pebbly sand from which Jackson (1939) recorded the dis-
covery of fossils by a local fisherman, Mr. G. R. Haynes. These nodules, which may now be
seen in situ in Reeth Bay, originate in the Carstone and have yielded a large fauna of mam-
millatum Zone age. The ammonites comprise several species of Sonneratia, together with
Anadesmoceras baylei, Beudanticeras dupinianum, Otohoplites sp., and Douvilleiceras mammil-
latum. Inoceramus coptensis sp. nov., Cuneolus lanceolatus, En folium orbicular e, Anthony a
cantiana, Senis wharburtoni, and Pinna robinaldina are among the lamellibranchs found here;

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 513

gastropods are represented by ClaviscaJa Clementina, Tessarolax fittoni, Gyrodes genti, Anchura
( Perissoptera ) cf. parkinsoni and Semisolarium moniliferum; echinoids by Toxaster murchi-
sonianus, Holaster ( Labrotaxis ) cantianus, and Polydiadema cf. wiltshirei. The nodules in
Reeth Bay yielded a unique dromiacean crab, Plagiophthalmus nitonensis (Wright and Wright
19506).

Compton Bay. At Compton Bay the Lower Greensand is not only much thinner than at
Atherfield but the beds have changed in character, so that precise correlation is impossible.
Fossils are very scarce above the Perna Beds and the only ammonites that have been obtained
from the main mass of the strata are insufficient to prove more than the presence of the
bowerbanki and martinioides Zones. From the pebble-bed at the base of the Carstone the
Wright brothers collected a small suite of rolled and phosphatized ammonite fragments,
including Sonneratia parenti and Cleoniceras morgani, both forms of the mammillatum Zone.

Redcliff. The whole thickness of the Lower Greensand is exposed at Redcliff, on the north
side of Sandown Bay, but the cliffs are deeply weathered, fossils are much rarer, and above
the Atherfield Clay it is impossible to make out the detailed zonal succession established at
Atherfield.

The Perna Beds and the lower part of the Atherfield Clay have for many years been accessible
in a large cliff-founder north of Yaverland Fort. The Perna Beds form a solid rib of rock at
the base of the cliff and break down into boulders over the beach. They are a rich source of
fossils and have contributed the following ammonites: Prodesliayesites obsoletus gen. et sp.
nov., P. sp. nov. aff. laeviusculus (v. Koenen), P. spp. indet., Deshayesites spp. nov. All these
were found in the grey-green calcareous stone at the top of the beds (bed 3 of Atherfield) and
a pyritic mould of P. obsoletus was found also in the underlying sandy clay (bed 2 of Ather-
field). In all other respects the occurrence is identical with that of the Perna Beds of Atherfield
and has furnished large quantities of fossils for museums.

The basal few feet of the Atherfield Clay contain species of Prodesliayesites, Roloboceras,
and Deshayesites, the last including D. forbesi and D. fittoni. Museum material suggests that
representatives of the Crackers and Lobster Beds are present here, though these beds cannot be
delimited in the section now visible. Black phosphatic body-chambers of Dufrenoyia similar to
those in the Lower Crioceras Bed at Atherfield have been found among the beach pebbles,
but their source has not been located. A conspicuous band of pebbles with derived Jurassic
fossils, mainly Kimmeridgian Pavlovia, occurs about 50 feet below the top of the Ferruginous
Sands, apparently on the same horizon as that seen at the base of Group XIII at Atherfield.
No fossils have been found in the Sandrock at this locality, and the Carstone, which here
reaches its maximum thickness of 72 feet, is similarly barren.

Shanklin. Under this heading will be considered the long stretch of Lower Greensand that
appears on the coast between Sandown and Bonchurch, near Ventnor.

The Perna Beds, formerly seen on the shore near Sandown Pier, are no longer exposed and
the Atherfield Clay is built over. Excavations made in April 1950 during extension of the
Trouville Hotel, 200 yards north-east of the Pier, yielded to Mr. J. Barker (1952) a few crushed
Prodesliayesites and Deshayesites from the bottom 4 feet of the Atherfield Clay.

No distinctive organic remains have been found in the lower part of the Ferruginous Sands,
displayed in the cliff south of Sandown, but at Lake Stairs, in division 1 of Osborne White
(1921, p. 37), Tropaeum bowerbanki, Cheloniceras cornuelianum, and Dufrenoyia furcata have
been found in ferruginous concretions, indicating the bowerbanki Zone of the Lower Aptian.
Another three-quarters of a mile south, near the slipway at Little Stairs, the top of his division
2 slants down to the shore. Here at low tide may be seen a band of discoidal concretions
with impressions of large Tropaeum and Epicheloniceras. Osborne White was probably right
to correlate this band with the Upper Crioceras Beds of Atherfield.

514

PALAEONTOLOGY, VOLUME 3

Forty-six feet above the last horizon and 4 feet below a tabular band of ironstone, 100 yards
south of the slipway, the sands contain rotted nodules with fossils, some being nests of small
Epicheloniceras like those in Group IX at Atherfield. Above this level the sands are riddled
with vacant moulds of Exogyra, strongly suggestive of the Upper Gryphaea Beds (Group X)
of Atherfield.

At Small Hope Chine may be seen glauconitic sand with clusters of Exogyra latissima
and Lopha diluviana, apparently the upper part of Osborne White’s division 3, identified by
Fitton with part of Group X. Blocks of orange-coloured ironsand fallen from division 5 lie
on the shore at Little Stairs Point and yield numerous Exogyra luberculifera and Lopha
diluviana, together with the echinoid Trochotiara fittoni and polyzoa. In the cliffs south of
Shanklin Chine Exogyra latissima is found in the sands singly and in bands. Running out from
the foot of Shanklin Point and forming the southern part of Horse Ledge is a prominent band
with Exogyra and knobs of speckled greensand full of white fossils, best seen at low tide.
Many of the knobs are a solid mass of brachiopoda or arborescent polyzoa ( Sipliodictyum
gracile, Chisma furcillata, Choristopetallum impar, &c.) or Serpula. This is the ‘Urchin Bed’,
probably the most important source of echinoids in the Lower Greensand, as the following
list indicates: Toxaster fittoni, T. renevieri, Phyllobrissus fittoni, Catopygus vectensis, Holaster
wrighti, Tetragramma rotulare, Trochotiara fittoni, Hyposalenia wrighti, H. stellulata. Lamelli-
branchs are represented by long-ranging species, Thetironia minor, Pterotrigonia mantelli, &c.,
but there are a few distinctive gastropods, such as Ringinella albensis, Dimorphosoma vectianum,
Confusiscala ischyra, and Claviscala ricordeana. Ammonites are found only in fragments or
in immature examples: Parahoplites cf. maximus, P. cf. nutfieldensis, P. sp. nov., and Tropaeum
subarcticum fix the horizon as the bottom part of the nutfieldensis Zone. Middlemiss (1959)
lists the following brachiopods from this bed: Rhombothyris extensa, Platythyris comptonensis,
Sellithyris sella shanklinensis, Cyrtothyris uniplicata, Praelongithyris praelongiforma, Oblongar-
cula oblonga, ‘ Ornithella ’ morrisi, ‘ O' celtica, ‘ O' tamarindus, ‘ O' wanklyni, ‘O' juddi ?,
Sulcirhynchia hythensis, ‘ Rhynchonella ’ parvirostris, and Lingula truncata.

The succeeding 20 feet of argillaceous greensand contains the ferruginous concretions for
which Shanklin has long been famed. This part of the sequence is the obvious correlative of
the Ferruginous Bands of Blackgang Chine (Group XIV), the concretions having an identical
fauna of lamellibranchia and gastropoda in the same mode of preservation. Fossils occur as
moulds, in subspherical masses, so tightly packed as to leave little room for matrix. At Shanklin
they have produced a few rare specimens of Parahoplites that indicate the cunningtoni Subzone
of the nutfieldensis Zone. This is also the type horizon and locality for the limpets Acmaea
formosa and Helcion meyeri Gardner (1877 a). As at Atherfield, no fossils have been found in
the thick band of clay that terminates the Ferruginous Sands.

The overlying Sandrock, nearly 120 feet thick, is very clearly displayed in Luccomb Chine
and Knock Cliff. Twenty feet above the base is a band of green clayey grit, 8 feet thick, with a
seam of phosphatic and pyritic nodules and fossil wood at the bottom. The bed is most readily
seen at the mouth of Luccomb Chine, where the beach is strewn with wood from the basal
nodular seam. This is the site of the famous plant discovery that enabled Carruthers (1870)
to diagnose an extinct order of Cycadophyta, the Bennettitales, more complex and specialized
than the living cycads. In addition to Bennettites gibsonianus and B. maximus, the flora con-
tains many conifers (Cupressinoxylon vectense, C. luccombense, Sequoia giganteoides, Cedro-
strobus leckenbyi, Podocarpoxylon gothani, P. solmsi, Pityostrobus jacksoni) and wood of
uncertain affinities ( Vectia luccombensis) (Carruthers 1869; Barber 1898; Stopes 1915; Creber
1956). The type specimen of the angiosperm Aptiana radiata almost certainly originated here.
Much of the wood is riddled with Terebrimya borings and appears to have been long adrift.
Discovery of rare Hypacanthoplites rubricosus in the nodules now fixes the horizon of this
important plant bed in the lower part of the jacobi Zone. The only other locality where I have

RAYMOND CASEY: STR ATIGRAPHIC AL PALAEONTOLOGY OF GREENSAND 515

found the rubricosus fauna is at the bottom of the Folkestone Beds of Folkestone. Lamplugh’s
view that the lower half of the Sandrock should be correlated with the upper part of the Sand-
gate Beds of the mainland (Lamplugh 1901, p. 119) is thus not supported.

On the shore at Dunnose Professor FI. L. Hawkins picked up a nodule containing a hollow
mould of Hypacanthoplites aff. trivialis, apparently derived from the top of the Sandrock.
This find is interesting in view of the record of fossiliferous nodules at the top of the Sandrock
at Blackgang Chine (Lamplugh 1901) and suggests correlation with the milletioides Subzone
of Sandling Junction. That the Sandrock is followed immediately by the mammillatum Zone is
proved by the occurrence of Sonneratia kitchini and allied species in the basement-bed of the
Carstone at Dunnose and in the few feet of grits above. Beudanticeras newtoni nom. nov. and
rare Otohoplites and Protohoplites have been collected from fallen blocks between Dunnose
and Bonchurch. Added to the finds around Reeth Bay, they show that all subzones of the
mammillatum Zone are present in the Isle of Wight. The junction with the Gault is gradational
and ammonites of the eodentatus Subzone of the Middle Albian still occur in a gritty Carstone
matrix.

Dorset Coast

The Lower Greensand of the Dorset coast is a condensed version of that seen in the Isle of
Wight. The best section is in Punfield Cove, at the north corner of Swanage Bay, where the
beds were first studied in detail by Judd (1871) and Meyer (1872). Judd thought they were
part of a transition series between fresh-water Wealden and marine Lower Greensand (or
‘Neocomian’), for which series he proposed the name Punfield Formation. This idea was
promptly contested by Meyer, who showed that, as in the Isle of Wight, the beds at Punfield
could be easily divided into Wealden Shales below and Lower Greensand above. The accuracy
of Meyer’s correlation is now generally acknowledged: it may now be shown that Judd, too,
was not wholly incorrect in his conception of a passage from one formation to the other, for
although the ‘Punfield Formation’ does not exist as a stratigraphical unit, the Lower Green-
sand does take on a brackish-water facies and merges into the Wealden when followed west-
wards along the Dorset coast.

The succession at Punfield is as follows (Strahan 1898; Arkell 1947A ; House 1958):

Lower Greensand at Punfield

ft. in.

Ferruginous Sands

15. Yellow sand, not well seen ...... about 10 0

14. Clay, dark, sandy, selenitic . . . . . . . 15 0

13. White sandstone with quartz pebbles . . . . . 20 0

12. Brown sandstone and yellow sandstone with shales . . . . 15 0

11. Interlaminated sand and clay; worm burrows? . . . . . 15 0

10. Ferruginous sand and hard irony sandstone with Nuculana . . .12 0

9. Interlaminated sand and yellow clay with some thicker beds of yellow and

white sand . . . . . . . . . . .610

Punfield Marine Band

8. Fossiliferous limestone with wavy seams of lignite . . . . 10

At her field Clay

7. Clay, reddish above, blue and very fossiliferous in lower part . . 28 0

6. Sandstone, soft yellow, with lamellibranchs . . . . .10

5. Clay, pale red, bluish in parts . . . . . . . .86

4. Sandstone in four hard grey bands . . . . . . .30

3. Clay, red ........... 6 0

516

PALAEONTOLOGY, VOLUME 3

Pebble Bed ft. in.

2. Sand, dark green, with small pebbles and grit . . . . .10

1. Pebbly clay, pale blue, sandy; small rolled bivalves, ammonites, &c., and

larger pebbles of sandstone, wood, &c., at base . . . . .20

Total 198 6

Strahan correlated the Pebble Bed with the Isle of Wight Perna Beds, but Arkell thought it
possible that the pebbly basement-bed of the Lower Greensand is diachronous, becoming
younger westwards. On the other hand, it is possible that the Pebble Bed of Punfield corresponds
only to the gravelly seam at the base of the Isle of Wight Perna Beds, that the clay above
(bed 3) is the clay of bed 2 of the Perna Beds of Atherfield, and that the sandstone referred to
the Atherfield Clay (bed 4) is really the calcareous sandstone at the top of the Perna Beds of
Atherfield. In the absence of indigenous ammonites nothing can be proved.

Arkell’s faunal list (19476, pp. 171-2) shows the Pebble Bed and Atherfield Clay to possess
many small lamellibranchs found also in the Atherfield Clay Series of the Isle of Wight, such
as Nuculana scapha, Aptolinter aptiensis, Pseudoptera subdepressa, Freiastarte subcost at a,
Panopea gurgitis, and Plectomya anglica. Mulletia mulleti and the other large molluscs do not
occur. The zone fossil Deshayesites forbesi ( D. deshayesi in Arkell) was found in abundance
in bed 7, apparently the equivalent of the Lower Lobster Bed of Atherfield. The chief palae-
ontological interest in this section, however, is in the Punfield Marine Band. This thin band of
limestone, full of lignite, contains a rich fauna, including the ammonites Deshayesites forbesi,
D. aff. cal/idiscus, D. sp. nov., and Roloboceras hambrovi. All previous determinations of
ammonites from this bed are incorrect (e.g. Arkell 19476, p. 172); Deshayesites punfieldensis
has not been found at Punfield and views as to the horizon and locality of this ammonite
expressed by Spath were misleading. The ammonites are all Crackers species and they confirm
Strahan’s correlation of the Punfield Marine Band with that bed. But whereas ammonites are
very abundant in the Crackers, they are exceedingly rare in the Punfield Marine Band. Con-
versely, the gastropod Cassiope, of brackish-water affinities and characteristic of the Punfield
fauna, is known from the Crackers only in one or two examples. The lamellibranch genus
Eomiodon, an indicator of marine-brackish conditions (Casey 1956), is present in the Punfield
Marine Band (e.g. GSM 86398) though unknown in any of the beds at Atherfield. Nemo-
cardium ( Pratulum ) ibbetsoni, a lamellibranch of very wide tolerance (being found both in the
Wealden Shales and the Atherfield Clay Series), is quite common at Punfield. There is no
doubt that the fauna of the Punfield Marine Band shows the influence of brackish-water and
the few ammonites found in this lignitiferous bed may have been drifted or washed in.

Support for this view is forthcoming from the next exposures. About 5 miles west of Punfield
Cove Geological Survey officers measured and collected from Lower Greensand exposed in a
cutting on the west side of Corfe Station. The Punfield Marine Band was found, full of shell
fragments and lignite, and the many fossils from this band, in the Geological Survey Museum,
were listed by Arkell (19476). Noteworthy features of the fauna are the absence of ammonites,
the greater abundance and variety of Cassiope, plentiful Nemocardium (P.) ibbetsoni, and the
presence of Eomiodon. Where the Lower Greensand comes down to the coast again, at Wor-
barrow Bay, about 6\ miles west of Corfe, the Punfield Marine Band has passed into a thin
fossiliferous ironstone (Arkell 19476, p. 176). Here Cassiope still occurs and is accompanied
by larger and more numerous specimens of Eomiodon ( Astarte obovata of Arkell) and an
abundance of a small bivalve identified by Arkell as Anthonya cornueiiana but here described
as Cuneocorbuia arkelli sp. nov. There are no ammonites. Finally, on the east side of Lulworth
Cove, 1 foot below the Gault, is an impersistent band of ironstone, 6 inches thick, from which
Strahan obtained ‘ Cyrena' , Exogyra, and a few gastropods and which he assigned to the
Wealden. Arkell, however, pointed out its resemblance to the fossiliferous ironstones of

RAYMOND CASEY: STR ATIGR APHIC AL PALAEONTOLOGY OF GREENSAND 517

Worbarrow Bay and he and Kirkaldy (1939, in discussion) agreed in placing it in the Lower
Greensand. A collection of fossils made recently by Mr. S. W. Hester from this ironstone was
kindly passed to me for study. Their matrix is indistinguishable from the ironstone of Wor-
barrow Bay and might have been expected to contain the same fauna. Instead the fossils were
all lamellibranchs and comprised Exogyra cf. tuberculifera, Eomiodon cf. libanoticus, and
Filosina gregaria. The last is the common ‘ Cyrena ’ of the Wealden Shales (Casey 19556),
while the species of Exogyra and Eomiodon are those found also in the Punfield Marine Band
farther east. Eomiodon and Filosina are known in association elsewhere only in the Aptian of
the Lebanon. Below this fossiliferous ironstone the beds at Lulworth Cove pass down without
break into the Wealden (Strahan 1898, p. 129).

The interpretation placed on the above facts is as follows: towards the end of forbesi times
a slight elevation of the land affected the easterly flowing river that discharged its sediment over
the present Dorset-Isle of Wight area, causing the estuary to move eastwards. On the south-
west coast of the Isle of Wight the movement was expressed by the interruption of the Ather-
field Clay Series by a bed of sand (Crackers). Proceeding westwards on the Crackers horizon
we travel up the estuary of the river, the fauna gradually changing as the waters become less
saline, until, at Lulworth Cove, the water was sufficiently diluted to support lamellibranchs of
Wealden facies. Here then is the passage of Lower Greensand into Wealden and, in some
measure, the vindication of Judd’s 'Punfield Formation’. That of all the Punfield beds, the
‘Marine’ Band should be the one to demonstrate a progressive brackish-water influence is
unfortunate; it is another example of the misnomers so replete in geological literature.

Wealden Province

The Wealden Province is here taken to comprise not only the Cretaceous area of Kent,
Surrey, and Sussex, which is the geological Weald proper, but also the areas of Lower Green-
sand deposition in Wiltshire and the adjacent counties. For convenience of treatment this
larger Wealden Province may be divided into five regions, as follows:

(1) East Kent (4) Sussex

(2) West Kent (5) The western outliers

(3) Surrey (with part of Hampshire)

Conforming with the anticlinal structure of the Weald proper the Lower Greensand crops
out in an elliptical band that continues from the coast at Folkestone through Kent, Surrey,
and Sussex, encircling the Wealden Beds and ringed in turn by the narrow outcrop of the Gault.
The beds were first studied in detail around Folkestone, where Fitton (1836) divided them into
three broad lithological units. Subsequently the presence of a fourth unit underlying those
recognized by Fitton was noted in Surrey by Austen (1843); this was later found to be wide-
spread and was correlated by Fitton (1847) with the Atherfield Clay of the Isle of Wight. In
the Geological Survey Memoir on the country between Folkestone and Rye the following
names for the four divisions of the Lower Greensand were used by Drew (1864):

(4) Folkestone Beds (2) Hythe Beds
(3) Sandgate Beds (1) Atherfield Clay

This nomenclature was given a Weald-wide application by Topley (1875) and despite the
fact that Topley himself was later a party to the proposal to merge the Folkestone and Sand-
gate Beds under the term ‘Shanklin Sands’ (Topley and Jukes-Browne 1888), it has remained
in use for the succession in the Weald.

East Kent

Within the East Kent region the Lower Greensand crops out in a belt some 2 to 3 miles
wide and about 18 miles long, extending from Folkestone in the east to Ashford in the west.

518

PALAEONTOLOGY, VOLUME 3

Previous work in this area has centred on the classic sections on and near the coast between
Folkestone and Hythe (Fitton 1836; Drew 1864; De Ranee 1868; Price 1874; Topley 1875;
Spath 19236, 1925, 1930a, 1935; Casey 1936, 1939, 1950). Gregory (1895) described the fauna
of the base of the Sandgate Beds at Great Chart, near Ashford, and the relations of these
beds to the underlying members of the Lower Greensand in East Kent were discussed by Kir-
kaldy (1937). Cornes and others (1925) contributed a brief account of the Lower Greensand
of the Ashford district and similar accounts for the whole of East Kent have been published by
Cornes (1928) and Kirkaldy (1939). The stratigraphy and petrographical characters of the
beds have been more fully described by Worrall (1954).

text-fig. 3. Regional divisions of the Wealden Province. 1, East Kent; 2, West Kent; 3, Surrey (with
part of Hampshire); 4, Sussex; 5, The Western Outliers.

Much was learnt about the underground extension of the Lower Greensand in East Kent
from borings and shafts put down in search for coal (Lamplugh and Kitchin 1911; Lamplugh,
Kitchin, and Pringle 1923). These show that the formation dwindles rapidly to the north and
to the east. Nearly 300 feet thick at the outcrop, it is reduced to 50 feet 8 miles farther north
and disappears altogether along a line running just north of the Stour Valley.

Atherfield Clay. This division consists of greenish-grey, brown, and blue silty clays, resting
with sharply defined base on the Wealden Beds. It is poorly exposed at outcrop and knowledge
of its palaeontological characters, thickness, and relations to the beds above and below has
been obtained chiefly from the Kent Coalfield shafts and borings. Re-examination of the
ammonites shows that in East Kent the Atherfield Clay represents the upper part of the forbesi
Zone ( callidiscus Subzone) and is the correlative of the Crackers and Upper Lobster Beds of
the Isle of Wight, not of the Atherfield Clay s.s. This is shown by the abundance of Deshaye-
sites forbesi, the presence of species of the callidiscus type, and the absence of any ammonite
diagnostic of the Lower Lobster Bed or Atherfield Clay s.s. There is no Perna Bed at the base.

In the Dover shafts the Atherfield Clay was proved between depths of 388 and 431 feet and
yielded a rich fauna, including many ammonites. The fossils were described by Kitchin (in
Lamplugh and Kitchin 1911, pp. 107-11) and are now in the Geological Survey Museum.
The long list of mollusca, echinoidea, and Crustacea is duplicated at Atherfield, and, as at

RAYMOND CASEY: STR ATIGRAPHICA L PALAEONTOLOGY OF GREENSAND 519

that locality, the base of the formation contains teeth of Hybodus and Acrodus apparently
derived from the Wealden Beds. The dominant ammonite is Deshayesites forbesi sp. nov.
(= Hoplites deshayesi of Kitchin), which occurs from top to bottom. Crushed and distorted
fragments of an undescribed species of Deshayesites found also in the Upper Lobster Beds
were collected at 410 and 415 feet ( = cf. Acanthoceras albrechti-austriae and Crioceras sp.
of Kitchin) and a few pieces of a feebly ornamented Deshayesites like D. topleyi or D. callidiscus
sp. nov. were recovered between depths of 418 and 431 feet ( Hoplites laeviusculus of Kitchin).
The ‘ Douvilleiceras martiniV recorded from 415 feet by Kitchin is an immature Roloboceras.

Deshayesites forbesi was found in numbers in the 25 feet of Atherfield Clay proved in the
Guilford Colliery shaft, situated 11 miles north-east of Lydden Church, near Dover (Geo-
logical Survey and Brigadier Bomford collections) and the same species was found to charac-
terize the Atherfield Clay in a boring recently put down at St. Margaret’s Bay, north of Dover
(Geological Survey collection). In the Brabourne boring, about 14 miles west of the Dover
shafts, a fragmentary Deshayesites, apparently D. forbesi, was found between 240 and 250 feet.
This is the ammonite recorded by Kitchin (ibid., p. 114) (as possibly Crioceras) from strata
assigned to the Sandgate Beds. Not only does the ammonite show this correlation to be false,
but its matrix and that of the associated fossils is identical with the Atherfield Clay in the
Smeeth railway-cutting. This explains the unusually great thickness of the ‘Sandgate Beds’
in the Brabourne boring (98 feet) and disposes of the suggestion, frequently repeated, that the
Hythe Beds have here passed into Sandgate Beds facies (Lamplugh and Kitchin 1911, p. 37).

A shaft sunk by Simms (1843) during the construction of the Saltwood railway tunnel,
near Hythe, proved a thickness of 49 feet 6 inches of Atherfield Clay between the Hythe Beds
and the Weald Clay. Fossils obtained from this shaft are in the Geological Survey Museum
and include Resatrix and other lamellibranchs, but not Muiletia mulleti, as stated by Simms.
The bottom of the clay yielded D. forbesi and the undescribed Deshayesites referred to above.
The Survey collections also contain the ammonites D. forbesi and Toxoceratoides cf. biplex
collected by H. B. Mackeson from the Atherfield Clay of Hythe. D. forbesi was also collected
by Mr. B. C. Worssam from exposures of the clay in the banks and bed of Brockhill Stream,
seven-eighths of a mile N. 65° W. of Hythe Church.

In 1925 a landslip in the railway-cutting half a mile west-north- west of Smeeth Station
exposed the junction of the Atherfield Clay and the Hythe Beds. Specimens collected here by
the Geological Survey show the Atherfield Clay as a pale-grey micaceous and silty clay with
pyritous threads and with the ubiquitous D. forbesi. This occurrence is in flat contradiction
to the statement of Cornes (1925, p. 260) that the slip revealed the Hythe Beds resting directly
on Weald Clay. Since he also said (ibid., p. 259) that in this district the upper part of the
Weald Clay contains ‘a definitely marine mollusc — Exogyra sinuata ’, a common Lower
Greensand species not otherwise recorded from the Wealden Beds, it may be suggested that
Atherfield Clay was mistaken for Weald Clay. It is mainly on the word of Cornes that subse-
quent authors have pinned their faith on the discontinuity of the Atherfield Clay in the Ashford
district (Kirkaldy 1939, p. 391; Worrall 1954, p. 187). Field evidence for the continuity of the
Atherfield Clay in this area being open to dispute (see discussion in Worrall 1954), it is impor-
tant to emphasize that in the critical section at Smeeth palaeontology puts the matter beyond
argument.

No fossils have been recorded from the marginal area of Atherfield Clay deposition in east
Kent, as for instance in the Harmansole boring, 3 miles south of Canterbury, where the
deposit is reduced to 2 feet of sandy clay full of phosphatic fragments.

Hythe Beds. In East Kent the Hythe Beds consist of alternating layers, generally about 6 inches
to 2 feet thick, of hard, sandy, grey or blue-grey limestone (‘ragstone’) and grey-green loamy
sand speckled with grains of glauconite (‘hassock’). The beds rise from the shore at Mill

520

PALAEONTOLOGY, VOLUME 3

Point, Folkestone, and strike inland to form an escarpment overlooking Romney Marsh.
Estimates of 60 feet for the thickness of the Hythe Beds at Hythe (e.g. Drew 1864, p. 7) are
excessive; they attain a thickness of 50 feet at the edge of the escarpment at the western end
of the region and diminish in thickness to the north and east. Thirty-five feet appears to be the
maximum in the Flythe district, but locally the beds are much thinner, as at Repton Manor,
three-quarters of a mile north-west of the centre of Ashford, where Worssam (in Worrall
1954) records only 15 feet. Borings show that the Hythe Beds become rapidly thinner and
disappear altogether a short distance north of the outcrop. This was seen very clearly in the
Brabourne boring, where the Hythe Beds were found to have vanished completely 2 miles
north of where they make a brave show at the surface. The Dover Colliery shafts and borings
farther north show that the Hythe Beds have a more restricted distribution than has the under-
lying Atherfield Clay and that they are overstepped by the Sandgate Beds.

The idea that the disappearance of the Hythe Beds north of their outcrop is due to facies-change or
to post-depositional changes in character and that the ‘rag and hassock’ beds pass into a Sandgate
Beds type of lithology underground has lately been revived by Worrall (1956). This author claims that
the ragstone bands of the Hythe Beds outcrop originated fairly recently, after removal of the impervious
Gault cover, and are the result of leaching of calcium carbonate from higher members of the Lower
Greensand and its subsequent precipitation in the Hythe Beds. The petrographical evidence for this
hypothesis seems to rest largely on the presence of a single blue tourmaline in the Sandgate Beds of the
St. Margaret’s Bay boring. In this boring the Sandgate Beds are no more than 40 feet in thickness and
they yielded near their base a distinctive little lamellibranch ( Freiastarte praetypica sp. nov.) (GSM
Bm 5165-6) that characterizes the Sandgate Beds and basal part of the Folkestone Beds of East Kent.
It has never been found elsewhere and its occurrence here thus supports the assumption that the strata
in question are of Sandgate Beds age. There are, however, more cogent reasons for rejecting this
hypothesis of secondary origin of the ragstone. Attention may be drawn to the fact that fossils in the
ragstone are always ‘solid’ or only slightly distorted, whereas in the hassock all but the stout calcite
belemnites and Exogyra are crushed flat. This means that originally the Hythe Beds were composed
mostly of hassock and that the consolidation of the ragstone must have taken place before vertical
pressure was exerted — certainly long before removal of the Chalk dome and the Gault had exposed
the Lower Greensand to meteoric water. That such exposure has now resulted in partial decalcification
of the ragstone is shown by the prevalence of fossil ‘cast-beds’ in the Hythe Beds. The presence of
ragstone as pebbles and rafts in the basement-bed of the Sandgate Beds at Mill Point, Folkestone, is
conclusively in favour of its primary origin.

The Hythe Beds carry a rich fauna belonging to the deshayesi and bowerbanki Zones of the
Lower Aptian. Fossils are locally abundant; some species, such as the trigoniid Linotrigonia
( Oistotrigonia ) ornata , the oyster Exogyra latissima, and the brachiopods Sellithyris sella and
Sulcirhynchia hythensis tend to occur in bands or nests. Other common fossils are — Lamelli-
branchia : Sphaera corrugata, Venilicardia inornata, Pseudaphrodina ricordeana, Resatrix
hythensis, Trigonia carinata, Pterotrigonia mantelli anterior, Yaadia nodosa, Gervillella sub-
lanceolata, Gervillaria alaeformis, Plicatula placunea, Pinna ( Stegoconcha ) cf. gervaisei. Gastro-
poda: Conotomaria gigantea. Cephalopoda : Australiceras gigas, Tropaeum hillsi, T. bowerbanki,
Cheloniceras cornuelianum, Ch. crassum, Dufrenoyia furcata, Cymatoceras radiatum, C. pseudo-
elegans, Eucymatoceras plication, Neohibolites ewaldi. Brachiopoda: Oblongarcula oblonga.
Echinoidea: Holaster benstedi, Discoidea decor ata, Tetragranona malbosi. Polyzoa: Chisma
furcillata, Reptomulticava fungiformis. Foraminifera, radiolaria, and ostracoda are found in
some of the beds but have not been studied. A series of enormous limb and pelvic bones,
collected by H. B. Mackeson (1840) from the Hythe Beds of Hythe and thought to belong to
the marine reptile Polyptychodon, were later diagnosed by Owen (1884) as belonging to a new
genus and species of dinosaur, Dinodocus mackesoni. Besides Conotomaria, the beds yield other
exceptionally large gastropods, such as the limpets Hipponyx neocomiensis and Brunonia

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 521

gigantea (Gardner 1877 a; 18776), the last being 4 inches in diameter. The quarrymen have
their own names for some of the fossils: internal moulds of Sphaera corrugata are called

‘bullocks’ hearts’ and detached shafts of the uncoiled ammonoids Australiceras and Tropaeum
are known as ‘hosepipes’: ammonites and nautiloids are ‘whirligigs’ and the large Conoto-
maria ‘screws’.

The old Hythe quarries, worked for building stone, are now defunct but there are good

522

PALAEONTOLOGY, VOLUME 3
Section exposed in Otterpool Quarry, June 1955

ft. in.

Sandgate Beds

34. Brown-weathered glauconitic loam, passing up into soil . . .20

33. Band of small white phosphatic nodules ...... 2-6

Hythe Beds

32. Dark-green calcareous hassock with doggers of green calcareous sandstone

and grey sandy limestone . . . . . . . .23

31. Dark-green indurated hassock, well laminated and weathering grey;
crowded with fossils ( Linotrigonia , Gervillella, Exogrya, &c. ; Ch. meyen-
dorffi group) ........... 4 9

30. Grey-green calcareous sandstone with scattered small black phosphatic
nodules; Exogyra bed at top ........ 1

29. Grey hassock with doggers of ragstone (Ch. meyendorffi, Dufrenoyia)

28. Brown ragstone with weathered-out shells ......

27. Grey hassock ...........

26. Brown, blue-hearted ragstone ........

25. Grey hassock ...........

24. Ragstone as 26 (Tropaeum bowerbanki, Dufrenoyia furcata, D. lurensis,

Ch. cornuelianum) .......... 1

23. Grey hassock ........... 1

22. Massive brown, blue-hearted ragstone, split by hassock veins into three
equal lanes; very fossiliferous, the fossils in the top lane (‘Green bed’)
having a green-dappled surface. Fauna as in 24 . . . . . 3

21. Hassock parting ..........

20. Ragstone as 26, with impersistent cast bed at top ( Tropaeum bowerbanki) 1

19. Grey hassock ...........

18. Ragstone as 26, with impersistent cast bed at top and nests of Sellithyris
sella and Sulcirhynchia hythensis. (Ch. cornuelianum, Ch. crassum, Dufre-
noyia furcata, D. transitoria) ........ 1

17. Hassock parting ..........

16. Ragstone as 26 .......... 1

15. Hassock parting ..........

14. Ragstone as 26 (Dufrenoyia furcata, D. truncata) .... 1

13. Grey hassock ...........

1 2. Ragstone as 26, locally forming a cast bed (Ch. cornuelianum, Ch. crassum)

11. Grey hassock ...........

10. Ragstone as 26 ......... 6-9

9. Hassock parting .......... 2

8. Massive pale blue-grey ragstone, glauconitic (Tropaeum hillsi) . .16

7. Grey-green hassock ......... 6-1 1

6. Ragstone as 8 (Austra/iceras gigas) . . . . . . .19

5. Grey-green hassock ......... 1 3

4. Ragstone as 8 .......... 9

3. Rubbly bed of blue-grey ragstone nodules in blue-green hassock . .10

2. Ragstone as 8 (Ch. parinodum, Deshayesites cf. involutus); numerous large

Exogyra on upper surface ........ 1 6

1 . Blue-green sandy clay ........ seen 6

(passing down into Atherfield Clay according to description supplied by
quarry manager)

Total of Hythe Beds about 35 feet

n oo I 0\ 00 OO

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 523

exposures inland where the ragstone is extracted for road metal. The best is at Otterpool
Manor (Folkestone Quarries, Ltd.), just south of the main Folkestone-Ashford road, a mile
west of New Inn Green cross-roads and about 3 miles north-west of Hythe. The whole of
this division and its junction with the Sandgate Beds are here seen in the quarry faces, as
detailed on p. 522 and illustrated in text-fig. 4. This section may be taken as a standard for
East Kent.

The blue-grey, glauconitic basal 9 feet of this section belong to the deshayesi Zone, the
presence of the parinodum Subzone being indicated by Ch. parinodum and Deshayesites cf.
involutus in bed 2, and of the grandis Subzone by Tropaeum hillsi in bed 8. The remaining
26 feet of strata (beds 9-32) are assigned to the bowerbanki Zone and yield abundant faunal
evidence of both its subzones, the transitoria Subzone below, the meyendorffi Subzone above.
The meyendorffi Subzone (beds 29-32) carries little ragstone, the beds consisting mainly of
green hassock, in places hardened to sandstone, with numerous guards of the belemnite
Neohibolites ewaldi. An Exogyra bed (bed 30) with small phosphatic nodules lies near the base.

The blue-grey ragstones of the deshayesi Zone, resting on blue-green sandy clay, are well
exposed in quarries on either side of the main Folkestone-Ashford road, half a mile south-
west of Willesborough, near Ashford. Here they have yielded the diagnostic fossils, Ch.
parinodum and Deshayesites of the involutus and grandis groups in the bottom two lanes of
ragstone, with Australiceras gigas, Lithancylus grandis, and Cheloniceras cornuelianum in a
‘cast-bed’ about 8 feet above the quarry floors. The same fauna has been collected from shallow
workings at Merstham and from the Handen Quarry, Clap Hill, Aldington, associated at the
latter locality with Tropaeum hillsi. From the basal few feet of the Hythe Beds in the railway
cutting half a mile west-north-west of Smeeth Station officers of the Geological Survey col-
lected a small fauna which included the ammonites Ch. aff. parinodum, Deshayesites deshayesi,
D. multicostatus, and D. consobrinoides.

A shallow working at Shepway Cross, Lympne, 2 miles west of Hythe, exposes 12 feet of
rag and hassock of the lower part of the bowerbanki Zone and the top of the deshayesi Zone.
Several feet of bright green glauconitic sandstone and hassock, with crushed Cheloniceras of
the meyendorffi type, were seen to underlie the Sandgate Beds in road-building operations at
the top of Bartholomew’s Lane, Hythe, and the same beds, with an underlying phosphatic
nodule-bed, may just be made out in the old quarry-site at Tanner’s Hill, on the east side of
Hythe. The eastward continuation of these green hassocky beds of the meyendorffi Subzone
cannot be followed. The site of Jeal’s Quarry, just north of the sluice gate at Seabrook, a mile
to the east, is now occupied by private gardens, and the old Horn Street Quarry, about half a
mile north of Jeal’s, is completely overgrown, judging by the Old Series map both lie close to
the junction with the Sandgate Beds, but the only ammonites preserved from these quarries
are of transitoria age. On the shore at Mill Point, Folkestone, another 2 miles to the east, there
is no sign of the meyendorffi Subzone, the top of the Hythe Beds consisting of brown ragstone
with an ammonite fauna similar to that of the top of the transitoria Subzone at Otterpool.
Its destruction prior to the next phase of deposition is shown by bits of greenish sandstone,
ragstone, and rolled phosphatic fossils in the basement-bed of the Sandgate Beds (see below).

Sandgate Beds. In East Kent the Sandgate Beds are composed of greenish, grey, and slate-
coloured loams and dark-grey silty clay. Exposures are poor and estimates of the thickness
of the beds vary considerably. Worrall (1954, p. 192) believes they reach as much as 120 feet
at Sellindge, but only 30 feet at Hinxhill. Seventy to eighty feet seems a reasonable figure for
their thickness in the Folkestone area. The Sandgate Beds are transgressive and everywhere in
East Kent their base is marked by a band of phosphatic nodules or by other signs of a pause
or break in sedimentation. In the Dover Colliery shafts and in the St. Margaret’s Bay boring
they were found to have overstepped the Hythe Beds and to rest on the bored top of the

524 PALAEONTOLOGY, VOLUME 3

Atherfield Clay. Farther north, at Walmestone and Ebbsfleet, borings show them in contact
with the Weald Clay.

At Folkestone, Price (1874) divided the Sandgate Beds into four beds, as follows:

4. Yellowish green sands, passing into brownish clayey sands upwards.

3. Black clayey sands, in part resembling the Gault.

2. Dark-green sands, passing up into yellowish-green sands.

1 . Zone of Rhynchonella sulcata. Black sands with nodules of iron pyrites.

This section was compiled from exposures in the Lower Sandgate road and on the foreshore
east of Folkestone Harbour. No thicknesses were given for the individual beds and for reasons
stated below it is believed that the section is very incomplete and unreliable. The principal
error is in the position of the ‘Zone of Rhynchonella sulcata ’, which lies not at the base of
the Sandgate Beds but near the top. This bed was formerly exposed at low-water spring tides
east of the Harbour but is now buried beneath modern shore deposits. Topley (1875), following
De Ranee (1868), took it for the base of the Folkestone Beds, but on the advice of Price
transferred it to the Sandgate Beds (Topley 1875, p. 138, footnote). The real base of the Sand-
gate Beds crops out between tide marks on the shore at Mill Point, a mile south-west of
Folkestone Harbour, and the north-easterly dip of the strata makes it impossible for it to
appear again at low water east of the Harbour.

Museum collections testify to the fossiliferous nature of the ‘Zone of Rhynchonella sulcata ’,
which is characterized chiefly by the lamellibranchs Resatrix ( Dosiniopsella ) cantiana, Eriphyla
striata, Freiastarte praetypica sp. nov., Anthony a cantiana, Cucullaea glabra, Parmicorbula
striatula, Lucina cornueliana, Gervillella sublanceolata, Yaadia nodosa, Pterotrigonia mantelli,
and an abundance of Lamellirhynchia caseyi ( Rhynchonella sulcata Auctt.). Bones of
Ichthyosaurus campylodon and the chimaeroid fish Edaphodon also occur. This is the type
horizon of Anthonya cantiana, credited by Woods (1906, p. 130) to the Folkestone Beds.

About 1,000 yards south-west of the Harbour the top part of the Sandgate Beds may be
seen in bare patches on the bank above the promenade. It consists of a few feet of pale yellow-
green, micaceous, and silty sand, passing down into dark-grey, micaceous, and more clayey
sand. Excavations made in 1956 in the adjoining gardens of the Lower Sandgate Road, 120
yards west of the site of the Victoria Pier, passed through these same beds and entered a dark-
green clayey sand with fossiliferous concretions, presumably the ‘Zone of Rhynchonella
sulcata'. In addition to poorly preserved lamellibranchia and Lamellirhynchia, material thrown
out of the trenches included the ammonites Nolaniceras aff". nolani and Nolaniceras sp. juv.
These are probably the ‘ Ammonites deshayesi' recorded from this horizon by Topley (1875,
p. 139) and are of great interest as establishing the presence of the nolani Subzone in the
Sandgate Beds.

Bare patches in the undercliff and gardens of the Lower Sandgate Road afford glimpses of
green and slate-coloured loams, but the prevalence of landslipping in the area makes it
hazardous to connect the exposures into a vertical succession. From 200 to 50 yards west of
Mill Point the top of the Hythe Beds appears at low water as a seaweed-covered reef and its
junction with the Sandgate Beds is sometimes seen after storms have scoured the beach.
The section given on p. 525 was seen in the summer of 1957, when the foreshore had been
cleared of shingle to an extent unprecedented in living memory.

The boxstones of bed 1 formed a red cobbled pavement on the broken ledges of Hythe Beds
exposed by the receding tide. This extraordinary bed has a complex history and it incorporates
three distinct elements: (1) debris from the destruction of the meyendorf) i Subzone at the top
of the Hythe Beds (sandstone and ragstone pebbles and black nodules, much rolled), (2) buff-
grey phosphatic nodules of buxtorfi age, perhaps contemporaneous with the grey calcareous
inclusions, and (3) an indigenous fauna of lamellibranchs and brachiopods of nutfieldensis

RAYMOND CASEY: STR ATIG R APHICAL PALAEONTOLOGY OF GREENSAND 525

age. Cheloniceras ( Epicheloniceras ) buxtorfi, here found in buff-grey phosphate, is an important
zonal ammonite, not previously known in Britain. It characterizes the nodule-bed of Luitere
Zug, in the Engelberger Valley, Switzerland, and was used by Jacob (1907) as an index-fossil
for the upper part of the Gargasian (his subzone lib). The indigenous fauna is chiefly remark-
able for its large brachiopods, Cyrtothyris cyrta, C. uniplicata, and Cyclothyris latissima,
found together elsewhere only in the Faringdon Sponge Gravels. Lamellibranchs are plentiful,
the commoner forms being: Area dupiniana, Limopsis dolomitica sp. nov., Entolium orbiculare.

Section of Sandgate Beds exposed at tow water at Mill Point,
Folkestone, August 1957

6. Dark-green glauconitic clayey sands ....... seen

5. Large concretionary masses of olive-green, brown- weathering calcareous
sandstone, with intertwined cylindrical bodies like stems of plants on the out-
side; each has a large brown phosphatic nodule (up to 12 in.) in the centre,
surrounded by bright green glauconitic clayey sand. Fossil wood.

4. Very dark (almost black) sandy glauconitic clay, full of burrows infilled with
sandier material, some apple-green in colour. Pyrites crystals at top. Exogyra
latissima ............

(Concealed; estimated gap 10 ft.)

3. Very dark glauconitic loam with hard doggers ..... seen

2. Pebble-bed; mainly black cherts up to \ in. in matrix of glauconitic loam with
brown, green, and mustard-coloured streaks ......

1. ‘Conglomeratic’ bed, composed of boxstones impressed into the top of the
Hythe Beds. Each boxstone has a mammillated ironstone rind, brick-red in
colour, which encloses rounded lumps of hard grey-green gritty calcareous
rock, dolomitic in places; small pebbles occur both inside the boxstones and
in the rind, mainly in clusters; some of the pebbles are rolled pieces of bone
and teeth of fish; black phosphatic nodules (including rolled moulds of lamelli-
branchs) scattered throughout; pieces of grey ragstone, greenish calcareous
sandstone, buff-grey phosphatic nodules and pale-grey calcareous inclusions
also occur in the boxstones. Eastwards the boxstones become larger and flatter
and hold a more sandy and shelly content. A single large raft of ragstone, 4 in.
thick, noted ............

ft. in.
1 0

2 0

2 0

1 9

2-4

4-9

Hythe Beds. Light-brown fossiliferous ragstone with carious upper surface . 5-9

Total about 8 0

Chlamys robinaldina, Acesta longa, Thetironia minor, Pseudocar dia sp. nov., Proveniella regu-
laris, Exogyra tuberculifera, Gryphaeostrea canaliculata. A few specimens of Myopholas cf.
semicostata were found in position of life, apparently bored into the top of the Hythe Beds.
In the Dover sinkings, between depths of 300 and 388 feet, the Sandgate Beds were found to
have a fauna similar to that of the ‘Zone of Rhynchonella sulcata', with Lamellirhynchia caseyi,
Resatrix ( Dosiniopsella ) cantiana, Parmicorbula striatula, &c., and with the boring shells
Girardotia and Panopea descending into the underlying Atherfield Clay (Lamplugh and Kitchin
1911; fossil-names revised).

The stone doggers in bed 3 were recognized as the source of the Parahoplites nutfieldensis
recorded from the base of the Sandgate Beds at this spot (Casey 1939, p. 368).

Drew (1864, p. 9) described the basement-bed of the Sandgate Beds at Mill Point (‘shore
near the turnpike between Sandgate and Folkestone’) as a ferruginous layer 6 inches thick and
the same description was applied to the bed once exposed in the Horn Street Quarry, Sea-
brook (‘hill side between Hythe and Shorncliffe’ in Topley 1875, p. 129). It now appears

M m

B 6612

526

PALAEONTOLOGY, VOLUME 3

doubtful whether these old records can be used as evidence of discontinuity of the nodule-
bed (e.g. Worrall 1954, p. 191). Rather it would seem that the nodules are made less conspicuous
by secondary formation of ironstone. They were found to be present at the junction with the
Hythe Beds in the Otterpool Quarry, described above, and were very clearly exposed in a
temporary road-cutting 100-200 yards south of Grove Bridge, Sellindge, where the following
section was measured in June 1953:

Section of Hythe-Sandgate Beds junction near Grove Bridge
Sandgate Beds ft. in.

5. Dark-green glauconitic loam with a line of incipient phosphatic nodules

18 in. above base ......... seen 4 6

4. Phosphatic nodule band. Compact glauconitic loam crowded with whitish
phosphatic nodules (seldom more than 1 in. long); many of the nodules
are internal moulds of mollusca; some are incompletely phosphatized

6 in. to 1 3

Hythe Beds

3. Bright green glauconitic hassock with two lines of indurated doggers, the
lower with a concentration of small Exogyra at top, the upper with an
impersistent reddish-brown coat . . . . . 2 ft. 6 in. to 3 0

2. Brown, blue-hearted ragstone . . . . . . .16

I . Grey-green glauconitic hassock ........ 2 0

Total about 12 0

The phosphatic nodule band (bed 4) was rich in fossils, lamellibranchia predominating,
with the venerids Pseudaphrodina ricordeana and Resatrix hythensis especially common. The
brachiopods Sellithyris sella var., Sulcirhynchia hythensis, Praelongithyris praelongiforma,
and Oblongarcula oblonga were present and rare specimens of Cheloniceras (E.) buxtorfi and
Ch. ( E .) sp. nov., the whole assemblage suggesting a condensed deposit equivalent to the upper-
part of the Hythe Beds ( martinioides Zone) of the Maidstone area. A similar phosphatized
fauna at the base of the Sandgate Beds in the neighbourhood of Great Chart, near Ashford,
has been described by Topley (1875, p. 129), Gregory (1895), and Kirkaldy (1937). Unfor-
tunately, the only ammonite recorded from here (as Cheloniceras cf. cornuelianum ) is too
immature to be identified closer than Cheloniceras sensu lato.

It is thus seen that the Sandgate Beds of East Kent span the martinioides, nutfieldensis, and
basal part of the jacobi Zone of the Upper Aptian, the first zone being represented in highly
condensed form in the basal nodule-bed.

Folkestone Beds. Within the East Kent region the Folkestone Beds undergo marked changes
in thickness and lithology. In the cliffs east of Folkestone Harbour they consist of about 60 feet
of coarse yellowish greensands with bands of calcareous and glauconitic sandstone. West-
wards they pass into uncompacted sands, more or less ironstained, current-bedded, and gener-
ally devoid of organic content. One hundred and eleven feet of such sands were encountered
in the Brabourne boring. In the Kent Coalfield area, under cover of the Gault, they are
reduced to a few feet of calcareous and glauconitic grit, resembling in condensed form the
beds as seen at Folkestone.

The zonal stratigraphy of the Folkestone Beds in the type region may be summarized as
follows (Casey 1939; 1950): in the coast section the beds belong mostly to the regularis Sub-
zone, the topmost part of the tardefurcata Zone, with a few feet of mammillatum Zone at the
top. Underlying the regularis Subzone is a remnant of the middle third of the tardefurcata
Zone ( mUletioides Subzone) and this in turn rests non-sequentially on a condensed basement-

RAYMOND CASEY: STR ATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 527

bed of middle and upper jacobi age. Traced westwards the jacobi and milletioides deposits
expand rapidly. Complementary to this expansion of the bottom beds, the regularis and lower
mammillatum strata wedge out beneath the transgressive top of the mammillatum Zone
(puzosianus Subzone) and disappear less than 5 miles inland from Folkestone. At Sellindge,
near Brabourne, 8| miles from the coast, the whole division has passed into sands of jacobi
age, capped by the nodule-beds of the puzosianus Subzone.

At East Cliff, Folkestone, the beds are admirably displayed in contact with the Gault for
a distance of half a mile. Due to the north-easterly dip they decline gently to the shore and are
lost beneath the tide-mark in East Wear Bay, just beyond Copt Point. Measured sections of
the cliff were given by Fitton (1836) and Hinde (1885), but these do not seem to fit any part of
the succession exposed today. Price (1874) gave a fuller description of the beds and divided
them into four and the present author gave these divisions zonal definition (Casey 1939;
1950).

The section on p. 528 was measured in 1939 at Baker’s Gap, East Cliff, about 30 yards
short of the eastern extremity of the present promenade.

The lowest 10 feet of the succession, obscured for many years, was made accessible in
1937-9 during the construction of a promenade at East Cliff. Reference has already been made
to the extraordinary composition of the basement-bed (Price’s bed 1) and the important
palaeontological information obtained from it in the course of these operations (Casey 1939;
1950), but since the bed is now permanently concealed by the promenade it is desirable to
place on record the fullest particulars of its occurrence. A representative set of specimens is
lodged at the Geological Survey.

The work of clearance and excavation along the foot of the cliff provided a continuous
section of nearly 200 yards in which it was possible to examine the basement-bed. Previous
authors have described it as a brown ferruginous sandstone: in the unweathered state it was
found to consist chiefly of a firm glauconitic sand, somewhat loamy in places and not always
sharply separable from the underlying silty greensands of the Sandgate Beds. Here and there
it contained pockets of a buff siliceous rock, almost devoid of glauconite and argillaceous
matter but highly charged with shell-debris, sponge-spicules, and minute echinoid-radioles.
The bed held an abundance of black phosphatic nodules and was sprinkled liberally with white
and green-veined quartz, black chert, and green sandstone in well-rounded and subangular
fragments up to an inch in length. The pebbles also included oval, flat-sided pieces of a soft
green stone, sometimes showing bedding — perhaps a brecciated glauconitic mud. Dark-brown
concretions of ferrugino-phosphatic rock, mostly spherical in shape and averaging 6 to 8 inches
diameter, were of more sparing occurrence. Phosphatic nodules and concretions were all
thickly coated with oysters, polyzoa, and other encrusting bodies. The nodules were mostly
shapeless lumps of calcium-phosphate-cemented sand, but some took the form of hollow
cylindrical structures with encrusting organisms both inside and outside; others were the
rolled remains of crustaceans, ammonite body-chambers, logs of wood, or aggregates of fossil
shells. Especially interesting were some large nodules riddled with ramifying perforations,
where arborescent polyzoa, since rotted away, had formed the nucleus of growth of the nodules.
Fish teeth and bone-fragments of larger vertebrates were also found in a phosphatized condi-
tion. Internal cavities in the nodules due to the disappearance of shelly material were often
lined with a film of tarnished pyrites. The ferrugino-phosphatic concretions were highly fossil i-
ferous, though many contained nothing but a small species of Pannicorbula, so densely packed
that the external moulds of the shell gave the rock a peculiar scoriaceous appearance. In the
most westerly part of the section, nearest the Harbour, these concretions occupied the lowest
part of the basement-bed, as described by Price, but farther east they were concentrated together
with the black nodules in the middle of the bed. In the most easterly excavation that touched
the Sandgate Beds no concretions were seen, but from the very bottom of the basement-bed

528

PALAEONTOLOGY, VOLUME 3
Section of Folkestone Beds at Baker's Gap, East Cliff, Folkestone
mammillatum Zone

35. ‘Sulphur Band’. An indurated layer of phosphatic nodules, the nodules
veined and encrusted with pyrites and embedded in a matrix of clayey
greensand, the whole coloured yellow and reddish-brown by decom-
position products. Two distinct concentrations of nodules recognizable.
Abundant fossil wood . . . . . 1 ft. to

34. Coarse grey sand, somewhat clayey at top . . . 1 ft. 6 in. to

33. Main mammillatum Bed. Seam of coarse yellow-green sand and grit with
clusters of phosphatic nodules . . . . . 6 in. to

32. Very coarse yellowish sand and grit . . . . . 1 ft. to

3 1 . Incoherent yellowish greensand .......

30. Hummocky band of indurated sand and grit with pockets of small
pebbles weathered out as knobs on the upper surface ....

29. Very coarse yellowish sand and grit with small pebbles and shell frag-
ments ............

28. Sonneratia kitchini Bed. Line of small black phosphatic nodules scattered
irregularly through coarse yellowish sand. Wisps of grey clay 4 in. to
tardefurcata Zone

27. Very coarse sand and grit as 32 .......

26. Hummocky band of indurated calcareous grit .....

25. Yellowish greensand with small ferruginous nodules scattered and in lines
24. Band of carious spicular sandstone, porcellanous and cherty in places .
23. Yellowish greensand with lenticles of sandstone as above. Comminuted
lamellibranch shells .........

22. Yellowish greensand .........

21. Sandstone as 24 . . . . . . . 6 in. to

20. Yellowish greensand .........

19. Impersistent sandstone as 24

18. Yellowish greensand .........

17. Tough grey calcareous sandstone .......

16. Yellowish greensand .........

15. Impersistent sandstone as 24 ........

14. Yellowish greensand .........

13. Hummocky tough grey calcareous sandstone . . . 9 in. to

12. Sandstone as 24 . . . . . . 9 in. to

1 1 . Yellow-green, slightly clayey sand with patches and wisps of iron-staining
10. Impersistent sandstone as 24

9. Sand as 11. Obscured by talus ...... estimated

8. Sandstone as 24

7. Greenish loamy sand, weathering brown ......

6. Bright-green loamy sand .........

5. Pockets of small phosphatic nodules and pebbles of black chert. Large
Exogyra numerous ..........

4. Tough grey-green, glauconitic, calcareous sandstone band .

3. Nodular bed as 5

2. Well-compacted clayey greensand with abundant shell fragments and very
small pebbles of black chert ........

jacobi Zone

1 . Firm glauconitic loamy sand, weathering brown, holding a great quantity
of pebbles and black phosphatic nodules, with large spherical concretions
of ferrugino-phosphatic rock at the base. Encrusting oysters common .
Sandgate Beds. Yellow-green silty sand

Total of Folkestone Beds about

ft. in.

1 3

2 0

1 0
1 6
3 0

1 3

2 2
8

2 2
1 2
10 0
9

1 2
3 9
10

1 2
0-5

2 5
1 4

11

0-3

3 0
1 0
1 0

4 4
0-3

7 0
6

1 3

1 4

0-2

1 9
0-2

2 0

1 0

60 0

RAYMOND CASEY: STRATIGR APHICAL PALAEONTOLOGY OF GREENSAND 529

the picks of the labourers uncovered lenticular masses (a few inches in thickness) of ferruginous
stone. The lenticles had an ironstone core without granular structure and graded outwards
into a sepia-coloured sandstone with pellets of apple-green clay, clusters of small quartz and
chert pebbles, and shell debris, all converted into a hard, gritty mass. Black phosphatic nodules
studded the skins of the lenticles, which were oxidized to a brick-red colour.

Some of the commoner fossils found in the ferrugino-phosphatic ( rubricosus ) concretions
are the lamellibranchs Parmicorbula striatula, Resatrix ( Dosiniopsella ) cantiana, Freiastarte
praetypica, Pterotrigonia mantelli, Thetironia minor, the gastropod Margarites ( Atira ) mirabilis,
the ammonites Hypacanthoplites rubricosus. Id. var. tenuiformis. Id. var. papillosus and H.
aff. jacobi, and the lobster Homarus longimanus. Nests of Lamellirhynchia caseyi are also found
in this type of preservation. The black ( anglicus ) nodules yielded chiefly Thetironia minor,
Cucullaea glabra, and Tortarctica similis, together with Homarus longimanus and the following
ammonites: Hypacanthoplites jacobi, H. anglicus. Id. var. audax, H. clavatus, H. elegans, H.
cf. sarasini, H. cf. hanovrensis, H. simmsi, H. cf. spatlii, H. cf. laticostatus, H. spp. nov. The
ferruginous stone contained Epicyprina harrisoni, Tortarctica similis, Spondylus striatus,
Resatrix ( Dosiniopsella ) cantiana, Acesta longa, and indeterminate vertebrate remains. Bones
of Ichthyosaurus campylodon and teeth of Isurus mantelli occurred loose in the sand. Lopha
diluviana, Ostrea cunabula, Diploschiza sp., and the polyzoans Proboscina crassa and Berenicea
gracilis had later used the nodules and bones for anchorage.

This basement-bed speaks of long exposure on a sea-floor free of sedimentation. During
this standstill in deposition at Folkestone the basal tardefurcata Zone ( farnhamensis Subzone)
was laid down elsewhere.

The succeeding 2 feet of glauconitic clayey sand (Price’s bed 2) is without ammonites but
contains much shelly debris and the following identifiable forms: Oxytoma pectination, Lopha
diluviana, Entolium orbiculare, Neithea quinquecostata, Serpula articu/ata. This bed is assigned
to the middle third of the tardefurcata Zone ( miUetioides Subzone) because its westwards
continuation (at Newington) contains Hypacanthoplites of the miUetioides type.

Price’s third division of the Folkestone Beds commences with a band of calcareous glauco-
nitic sandstone with clusters of small pebbles, phosphatic nodules, and Exogyra strung along
the top and bottom. Leymeriella regularis, L. pseudoregularis, Anadesmoceras sp., and a giant
undescribed Douvilleiceras are found either in the nodules or in the sandstone itself. The little
pteriid Oxytoma pectination is plentiful here. Though replete with fossils of other groups, the
succeeding 50 feet of sands and rock bands are very poor in ammonites. They have yielded
fragments of large Douvilleiceras and the single example of L. regularis recorded by Spath
(1933). This part of the succession contains seams of whitish, sinter-like ‘sponge-rock’ — largely
aggregates of sand-grains and sponge-spicules bound together by a calcareous matrix. Their
weathered surfaces provide some of the best fossil-collecting in the Folkestone Beds: Exogyra
latissima, Lopha diluviana, Entolium orbiculare, Aptolinter aptiensis, Tortarctica similis,
Cucullaea glabra, Pterotrigonia mantelli, Inoceramus coptensis, the echinoids Holaster ( Labro -
taxis) cantianus and Phyllobrisus artesianus, the annelid Serpula articulata, the brachiopod
‘ Rhynchonella' gibbsiana, and the polyzoan Siphodictyum gracile are of fairly common occur-
rence. Despite the myriads of spicules present in the rock, recognizable sponges do not occur.
Branching and intertwining cylindrical bodies commonly seen on the surfaces of the stone
doggers, thought by early writers to be some sort of sponge, are probably infilled lamellibranch
burrows. Exogyra shells are commonly infested with the tubular stolons and vesicules of the
boring polyzoan Graysonia. The types of the starfish Lophidiaster ornatus and the curious
jointed worm-tube Serpula articulata were both obtained from here and wrongly attributed
to the Upper Greensand.

The mammillatum Zone begins with the Sonneratia kitchini bed, a line of small phosphatic
nodules in clusters, 10 feet below the base of the Gault, with bits of Sonneratia and Douvil-

530

PALAEONTOLOGY, VOLUME 3

leiceras mammiUatum. The main bed of D. mammillatum is in the topmost 6 feet of sand
(Price’s fourth division), the fossils occurring in a band of nodules up to a foot in thickness
and from H to 4 feet below the top of the sand. Collecting is best done among the weed-
covered reefs and rocky pools east of Copt Point after the bed has been washed over by the
sea. Inoceramus salomoni, Panopea gwgitis var. plicata, Nanonavis carinata, CucuUaea glabra,
Thetironia minor, Resatrix ( Dosiniopsella ) vibrayeana, Pseudocardia tenuicosta var. constant i,
Pterotrigonia mantel/i, Linotrigonia fittoni, Entolium orbiculare, Neithea quinquecostata,
Exogvra Jatissima, and Gryphaeostrea canaliculata are the lamellibranchs most frequently met
with and Anchura ( Perissoptera ) parkinsoni, Tessarolax retusum, Eucyclus sp. nov., Meta-
cerithium trimonile, Mesalia ( Bathraspira ) tecta, Leptomaria gibbsi, and Gyrodes genti are the
chief gastropods. The nautiloid Eutrephoceras clementinum is not uncommon, but belemnites
are exceedingly rare. Ninety-five per cent, of the ammonites are species of Douvilleiceras and
Beudanticeras, usually in pieces, but the minority fauna is of great diversity, as the following
list shows: Douvilleiceras mammillatum, D. monile, D. orbignyi, D. spp. nov., Beudanticeras
newtoni, B. dupinianum, Uhligella subornata, Parengonoceras ebrayi, Hypacanthoplites cf.
milletianus, Otohoplites raulinianus, O. elegans, O. auritiformis, O. guersanti, O. spp. nov.,
Protohoplites (P.) latisulcatus, P. ( Hemisonneratia ) sp., Sonneratia dutemp/eana, S. aff. parenti,
Pseudosonneratia spp. nov., Cleoniceras (C.) cf. cleon, C. (C .) floridum sp. nov., C. (C.) janneli,

C. (C.) seunesi, C. (C.) quercifolium, C. (C.) spp. nov., C. ( NeosayneUa ) inornatum, C. (N.) sp.
nov., Tegoceras sp. nov., Oxytropidoceras alticarinatum, Handles praegibbosus, H. spp. nov.,
Protanisoceras raulinianum, P. cantianum, P. lardyi, P. blanched, P. acteon, P. vaucherianum,
P. cf. halleri, P. spp. nov., ' Prohelicoceras ’ anglicum, Gen. nov. (' Metahamites’) sp. nov.
Crustacea, polyzoa, and echinoidea are rare. There are isolated finds of teeth or bones of the
shark Isurus mantelli and the marine reptiles Polyptychodon and Ichthyosaurus and I have also
collected a vertebra of the dinosaur Acanthopholis horridus (GSM Zk 4775). A big reptilian
fauna is known at this horizon in the Ardennes.

The nodules, with their black and brown phosphatic fossils, are the remanie in place of the
floridum and raulinianus Subzones. Protohoplites, Sonneratia dutempleana, and Otohoplites
guersanti occur only in the matrix of the nodules, unphosphatized or incompletely phosphatized
and generally with their nacre. They are part of a later fauna belonging to the puzosianus
Subzone ; so too is the small zeilleriid ModesteUa modesta, which probably grew on the nodules.

The 'Sulphur Band’, described previously (Mackie 1856, 1860; Casey 1950), still lies in the
puzosianus Subzone, having yielded Inoceramus salomoni, fragments of Protohoplites and Pseudo-
sonneratia, Cleoniceras cf. quercifolium, large indeterminate Otohoplites, and the long-ranging

D. mammillatum, D. monile, and B. newtoni. Its washed residue contains sponge-spicules,
including ribbed spicules of Geodites, and glauconitic pseudomorphs of foraminifera. Fossil
wood bored by Terebrimya, Martesia, and Xylophagella is copious. Though generally taken
as the commencing point of the Gault, this 'junction-bed’ of the early authors is now put in

EXPLANATION OF PLATE 78

Fig. 1. East Cliff, Folkestone, looking eastwards to Copt Point, low tide. Folkestone Beds overlain
by Gault at their type locality.

Fig. 2. Copt Point, Folkestone. Junction of Folkestone Beds and Gault with waveworn blocks of
Folkestone Beds ( regularis Subzone and mammiUatum Zone) on the shore. The ‘Sulphur Band'
may just be made out as a thin ledge at the junction.

Fig. 3. Sandpit at Brabourne Lees, East Kent. Pale, current-bedded sands of the jacobi Zone ( anglicus
Subzone) are overlain unconformably by glauconitic loams of the mammillatum Zone ( puzosianus
Subzone), the whole capped by flint-drift.

Geological Survey and Museum photos. Reproduced by permission of the Controller, H.M. Stationery
Office. Crown copyright.

°alaeontology, Vol. 3

PLATE 78

CASEY, Lower Greensand and Gault , East Kent

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 531

the Lower Greensand to avoid having a local formational boundary in the middle of a zone.
The foot or two of dark sandy clay with phosphatic nodules underlying the ‘Sulphur Band’
in the Dover Colliery shafts is also of puzosianus age, as is denoted by the presence of Proto-
hoplites michelinianus, var. ( Hoplites cf. raulinianus in Lamplugh and Kitchin 1911, p. 100).

Westwards from East Cliff the basal beds of the Folkestone Beds (beds 1 and 2) expand
rapidly. Black phosphatic nodules with the anglicus- fauna were seen in a bare patch of cliff
near the bottom of Remembrance Road, about 300 yards west of the Harbour, but the rubri-
cosus concretions were absent and the underlying sand was found to be still of Folkestone
rather than Sandgate Beds type. Here was also observed the introduction of seams of siliceous
stone in bed 2. The tough glauconitic sandstone (‘bottom stone band’) at the base of the
regularis Subzone may be followed from this point through the undergrowth of the escarp-
ment above the Lower Sandgate Road until it emerges in a clear section at Mill Point, just
beyond the Toll Gate. Using the same enumeration for the beds as at East Cliff, we may
summarize the section as follows:

Summarized section of Folkestone Beds at Mid Point, Folkestone

ft. in.

Beds 6-27. Coarse yellowish greensand with seams of carious spicular sand-
stone and tough calcareous sandstone . . . . . 55 0

4-5. Tough, grey-green, glauconitic sandstone band, pebbly at top . 2 0

3. Band of phosphatic nodules (up to 1 in. long) .... 6

2. Compact green and brown loamy sand with bands and lenses of
siliceous stone from 1 ft. to a few inches in thickness. Nests of very
small lydite pebbles, phosphatic nodules, and shell debris, the phos-
phatic nodules commoner at the top, where they tend to lie in lines 16 0
1. Brown sandy clay with small phosphatic nodules, pebbles, and
rolled pieces of Homarus, the nodules concentrated in a band 1 ft.
above base .......... 3 0

Sandgate Beds. Pale, almost white, silty sand

Total of Folkestone Beds 76 6

As at Remembrance Road, there is no sign of the rubricosus concretions in bed 1 and the
only ammonites obtained from here are fragmentary Hypacanthoplites of the anglicus type,
together with large body-chamber portions referable to the same genus. One of these was
identified by Spath as Parahoplites nutfieldensis and recorded by me (Casey 1939, p. 368)
under that name.

Road-widening in the 1920’s in Upper Folkestone Road (Sandgate Hill), at the west end
of Folkestone, exposed the bottom stone band of the regularis Subzone with several gigantic
Douvilleiceras and Leymeriella, just as at East Cliff. The coarsely glauconitic stone (with an
ammonite) encountered 66 feet below the Gault in a well at Folkestone Waterworks (Whitaker
1908, p. 139) is almost certainly the same band.

About a mile and a quarter west of Mill Point, in the grounds of Encombe, Sandgate, the
basement-bed of the Folkestone Beds is seen to have expanded into several feet of loose sand
with a line of ferruginous nodules. Pieces of these nodules, with lamellibranchia and Hypa-
canthoplites, may be picked up on the beach at Sandgate. Fitton (1836, p. 122) thought that
this sand belonged to his second division of the Lower Greensand, i.e. the Sandgate Beds,
and he compared the nodules with those found at Shanklin and Parham Park, Sussex; those,
however, lie on a lower horizon ( nutfieldensis Zone). Nodules, with fossils of the Jacobi Zone,
were passed through in the construction of Saltwood railway tunnel and were again referred
to the ‘second division of the Lower Greensand’ by Simms (1843). Topley (1875, p. 128) also
attributed them to the Sandgate Beds. Fossils found here by Simms include the type of Hypa-
canthoplites simmsi (Forbes 1845, p. 353).

532

PALAEONTOLOGY, VOLUME 3

Where the Folkestone Beds turn inland in a north-easterly direction we find a rapidly diminish-
ing thickness of regular is-mammillatum strata. Thin slabs of cherty sandstone with Leymeriella
from close below the Gault were found during excavations for air-raid shelters in the playing
field of Morehall School, Cheriton, and similar slabs lie about the fields around St. Martin’s
Church, at the top of Horn Street, 2 miles north-west of Mill Point, Folkestone. This is the
farthest point west for the regularis Subzone in East Kent.

The mammillatum Zone could be seen for many years in the railway embankment of the
Canterbury branch-line, a quarter of a mile north of St. Martin’s Church. The section was
mentioned by Topley (1875, p. 147) but has become grassed over since the closing of the line
in 1952. As noted by Topley, it showed only two nodule-bands instead of the three seen at
East Cliff. Topley described the section as follows:

Section of Mammillatum Zone in railway embankment } mile north of
St. Martin s Church and \ mile north-east of Cheriton Church

ft. in.

(a) Sandy clay with phosphatic nodules . . . . . . .20

( b ) Yellowish- brown sand . . . . . . . . . .20

(c) Nodules in brown sand ......... 6

(d) White and buff sand with stone in places, false-bedded. 6 ft. seen [ ? milletioides
Subzone].

The top nodule-bed yielded species of Protohoplites, diagnostic of the puzosianus Subzone,
and the bottom nodule-bed, though yielding no hoplitids, contained D. mammillatum and B.
newtoni in sufficient numbers to warrant its assignment to the Main mammillatum Bed of
Copt Point, Folkestone. The S. kitchini bed, at the base of the mammillatum Zone, is absent.
The nodule-beds may be followed up-track on the north embankment of the main Dover-
London line for about 150 yards, due south of the Star Inn, Newington. In the most westerly
exposure the bottom nodule-bed (bed c) is missing and only 6 inches of yellow-brown sand
separate the top nodule-bed from the sandstone of bed d. West of this point all exposures of
the mammillatum Zone in East Kent show the puzosianus Subzone only.

A disused sandpit just south of the railway bridge at Newington, and about half a mile
west of the last locality, shows the junction of the jacobi Zone ( anglicus Subzone) and the
tardefurcata Zone (milletioides Subzone). The jacobi Zone consists of about 40 feet of loose
current-bedded sand with rare iron concretions, terminating upwards in a line of phosphatic
nodules. The nodules may be traced all round the pit-face and a few bespatter the lowest course
of stone doggers just above. The nodules are black, oyster- and serpulid-encrusted, and include
remanie Hypacanthoplites of the anglicus group. Shells of brachiopods that used the nodules
for anchorage lie broken in the matrix; among them is Terebrirostra arduennensis (— T.
incurvirostrum), known also from the tardefurcata Zone (Shenley Limestone) of Leighton
Buzzard, Bedfordshire. Above the nodule-bed are 25-30 feet of yellowish greensands with
doggers and bands of tough calcareous stone and seams of white spicular sandstone, very like
bed 2 of the Mill Point section. The sands are full of Chondrites, and fragments of straight-
ribbed Hypacanthoplites of the milletioides group have been found in the sandstones 12 and
20 feet above the nodule-bed. Nodules from the mammillatum Zone lie in the subsoil at the top
of the pit and it is estimated that only 3 feet of the total thickness of the beds above the jacobi
Zone are missing in this section.

The anglicus nodule-bed may be seen in a number of old sandpits between Newington and
Saltwood, but no good sections are met with until we reach Sandling Junction. Here, just above
the railway station, is a large working in Folkestone Beds, capped by an outlier of Gault.
The following section was measured in 1949:

RAYMOND CASEY: STRATIGR APH1CAL PALAEONTOLOGY OF GREENSAND 533

Section of Folkestone Beds exposed in Sandling Junction Sandpit,

Ik miles north-west of St. Leonard's Church, Hythe
mammillatum Zone ft. in.

16. Band of phosphatic nodules in a matrix of green sandy clay, weathering
reddish-brown. In places two lines of nodules may be made out; generally
they coalesce into a single band. Pebbles of grey-green quartz (up to | in.)
occur throughout, and flat-sided pieces of claystone (up to 1 in.) in the
bottom of the bed .......... 1 0

tardefurcata Zone

15. Grey-green sandstone with abundant Oxytoma. Thin vertical pipings of

dark-green clayey sand in upper half . . . .11

14. Grey-green sand .......... 3 4

13. Tough grey limestone band, passing laterally into white spicular sandstone

with sandy intercalations . . . . . . . . 1 10

12. Grey-green sand with low-angle current-bedding . . . .10

11. Tough grey sandy limestone band . . . . . . .22

10. Coarse yellowish greensand, striped by layers rich in glauconite; current-

bedded, the bedding contorted at the base . . . . . .30

jacobi Zone

9. Clusters of small black phosphatic nodules and pebbles disposed in a gently
undulating line. Occasional doggers of sandy limestone; matrix coarse
yellowish greensand .......... 1-3

8. Sharp yellow sand with lines of iron-staining . . .30

7. Chocolate-reddish-brown sandstone (Red Bed) . . . . .10

6. Yellowish sand with abundant small pebbles, partially indurated . .10

5. Very coarse sand with glauconitic and clayey laminae, steeply current-

bedded. Phosphatized and semi-phosphatized nodules at the base 18 in. to 2 9
4. Coarse sand with clayey streaks . . . . . . 3 ft. to 3 10

3. Sand as above but steeply current-bedded . . 12 0

2. Pale sands with wisps and pocks of bluish clay. Rotted ironstone concretions,

mainly in top 3 ft. . . . . . . . . . 15 0

1 . Sands as above but without concretions (seen in a temporary trench in the

pit floor) . . . . . . . . . . . .10 0

Total about 62 0

The lower part of the succession (beds 1-7) was first referred to the nolani Subzone on the
strength of ammonite determinations by Spath (Casey 1939, p. 369). Larger collections and
more detailed study of the ammonites now show that all the beds up to bed 9 belong to the
anglicus Subzone of the jacobi Zone and are a greatly expanded version of the anglicus nodule-
band at the base of the Folkestone Beds of East Cliff. The rotted ironstone concretions of
bed 2 contain Hypacanthoplites cf. laticostatus and other forms present in the anglicus nodules
at Folkestone. They are on the same horizon as the fossiliferous concretions found in the
nearby Saltwood Tunnel excavations (Simms 1843).

Concretions in bed 5 (horizon 3 of Casey 1939) are an important source of fossils, containing
a varied fauna of mollusca, polyzoa, echinodermata, and brachiopoda. ‘They appear to
represent aggregations of organic debris that accumulated in hollows on the sea-floor and were
cemented by syngenetic formation of calcium-phosphate, the shell substance of mollusca and
other carbonate being converted to collophane. Ammonites and gastropods are usually hollow,
and the preservation and mode of occurrence of the fossils suggest that the shells were buried
rapidly more or less where they died’ (Casey 19606, p. 273). Many of the nodules in this bed
are cylindrical and are phosphatized only on the outside; others enclose arborescent polyzoa. It
can be seen at a glance that these are the same nodules, but in an unrolled and unscoured

534

PALAEONTOLOGY, VOLUME 3

condition, that occur in the anglicus band at East Cliff, Folkestone. Thetironia minor, Ptero-
trigonia mantelli, Modiolus aequalis, Chlamys robinaldina, Limopsis albensis, Glycymeris
( Glycymerita ) sub/aevis, Palaeomoera inaequalis, and Tortarctica similis and other lamelli-
branchs occur clustered in the nodules, together with Lamellirhynchia and broken echinoids.
The small trochid gastropod Margarites ( Atira ) mirabilis is very common here and one remark-
able example was found to possess a mould of the intestines (Casey 19606). Another interesting

200-

100-

FOLKESTONE BEDS

SANDGATE BEDS
HYTHE BEDS
ATHERF I ELD CLAY

text-fig. 5. Comparative vertical sections of the Lower Greensand of East Kent and Maidstone.

feature of this fauna is the apparent symbiotic association of polyzoa and serpulids. Ammonites
are rather rare, but H. anglicus, H. simmsi, and undescribed allies have been found. The
‘Red Bed’ (bed 7) has contributed the same species of ammonites, and also yields Neithea
quinquecostata, Thetironia minor, Pterotrigonia mantelli, ‘ Rhynchonella ’ deluci, Lamelli-
rhynchia caseyi, and the echinoids Holaster ( Labrotaxis ) cantianus and Catopygus cf. colum-
barius as common fossils.

The black, oyster- and serpulid-encrusted nodules of bed 9 (horizon 5 of Casey 1939) are
highly charged with sponge-spicules and minute chips of shell and the enclosed sand-grains
are frequently coated with iron; they have yielded H. anglicus, H. cf . jacobi, H. aff. simmsi,
and the lobster Homarus longimanus. Hypacanthoplites cf. subelegans and H. milletioides
(— Douvilleiceras? , Casey 1939), indicative of the milletioides Subzone of the tardefurcata

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 535

Zone, occur rarely in the stone bands (beds 13 and 15) above the anglicus Subzone. Bed 13
contains silicified banks of the hexactinellid sponge Plocoseyphia, colonies of the polyzoan
Inversaria orbicularis, and terebelloid worms, and beds 11 and 13 have a fauna of terebratulids
and rhynchonellids not yet systematically studied. Oxytoma pectinatum occurs throughout
and is especially abundant in bed 15.

Resting on the bored top of the tardefurcata Zone is the phosphorite band of the mammil-
latum Zone (bed 16), containing at the base angular pieces of claystone and rare fragments of
Hypacanthoplites milletioides derived from some pre-existing bed in the zone below. Fossils
are invariably in a remanie state. The lamellibranchs Cucullaea glabra, Entolium orbiculare ,
Gryphaeostrea canaliculata, and Exogyra latissima are very numerous, the last generally having
a rotted shell with a network of infilled Cliona borings. The following ammonites have been
collected: Douvilleiceras mammillatum, D. monile, D. orbignyi, D. sp. now, Beudanticeras
newtoni, Sonne ratio dutempleana, Pseudosonneratia sp. nov., Protohoplites (P.) latisulcatus,
P. (P.) michelinianus, P. ( Hemisonneratia ) puzosianus, P. (H.) gallicus, P. (H.) sp. nov., Oto-
hoplites auritiformis, O. spp. nov., Cleoniceras cf. quercifolium, Protanisoceras raulinianum,
P. cantianum. The assemblage is of puzosianus age and shows that this mammillatum-bed is
approximately equivalent to the ‘Sulphur Band’ of Folkestone.

This important section not only proves the farnhamensis non-sequence at the base of the
tardefurcata Zone, but also demonstrates in a striking manner the disappearance of practically
all the Folkestone Beds seen in the cliffs east of Folkestone Harbour. In all, some 60 feet of
strata, comprising the regularis Subzone, Sonneratia kitchini bed, and Main mammillatum
bed, have been cut out from beneath the puzosianus Subzone.

North-west of Sandling Junction the plane of unconformity at the base of the puzosianus
Subzone is shown very clearly in sandpits south of Brabourne. In File’s Pit, at the top of Swan
Lane, Sellindge, a quarter of a mile south-east of Horton Priory and 2\ miles north-west of
Sandling Junction, the puzosianus Subzone, with characteristic ammonites, is split into three
lines of phosphatic nodules distributed through 2 feet of glauconitic, pebbly sands and loams.
This rests with sharp junction on pale, current-bedded sands with giant foresets; the top 12
to 16 inches of sand is patchily indurated into a yellowish sandrock and is riddled with the
same dark-coloured vertical pipings seen below the mammillatum Zone at Sandling Junction.
Four to ten feet below the top of the sand are sparsely distributed nodules with arborescent
polyzoa, exactly like those found in bed 5 of the jacobi Zone (anglicus Subzone) of Sandling
Junction. A similar succession is seen in the Granary Court sandpit, Brabourne Lees, just over
a mile and a half north-west of File’s Pit and a mile and a half north-east of Smeeth. Here the
polyzoan-bearing nodules lie immediately under the mammillatum Zone. In the Brabourne
area, therefore, the stone bands of the milletioides Subzone (already partly eroded at Sandling
Junction) and the topmost part of the anglicus Subzone have been cut out by the uncon-
formity.

Further evidence of pr ^-mammillatum erosion of the Folkestone Beds was provided by a
chance exposure in the underground workings of Chislet Colliery, about 6 miles north-east
of Canterbury. Three thousand and twenty yards N. 54i° E. of the North Pit (Downcast)
shaft, at a level of —1,016 feet O.D., the Gault was unexpectedly encountered, resting with
angular discordance on the Coal Measures. At the base of the Gault, below the benettianus
and eodentatus Subzones, was a conglomerate-bed, 9 inches thick, in which alongside the
normal phosphatized fauna of the mammillatum Zone were worn slabs (up to a foot in length)
of Folkestone Beds sandstone. Some of the slabs were composed of a green siliceous rock not
unlike the Ightham Stone (Geological Survey collection).

In Quarrington Wood, about 2 miles north-west of the pit at Brabourne Lees, the puzosianus
nodule-beds are replaced by an ironstone seam, similar to that found at the junction of the
Folkestone Beds and Gault in West Sussex (Worrall 1954).

536

West Kent

PALAEONTOLOGY, VOLUME 3

The area of Lower Greensand country considered under this heading extends from Ashford,
in the south-east, to the western border of the county at Westerham, 2 miles west of Sevenoaks.
From Ashford the outcrop continues its north-easterly trend to Maidstone, where the recession
of the Chalk escarpment at the Medway Gap has laid bare a broad triangular expanse of Lower
Greensand 6 miles wide. West of Maidstone the strike of the beds changes to WSW.-ENE. and
the outcrop steadily diminishes in width, being reduced to a mile and a half at the western end
of the region.

There are no fundamental works on the Lower Greensand of West Kent, though there is a
voluminous, scattered literature relating to local detail. Easy of access from London, the
district is a favourite one for student-parties and the Proceedings of the Geologists' Association
contain innumerable snippets of information on the Lower Greensand of this region, either in
short papers or in excursion reports. The ragstone quarries around Maidstone came under
the observation of Fitton (1836; 1845), Bensted (1860; 1862), and Topley (1875), and useful
information on the Lower Greensand exposed during the construction of the Sevenoaks rail-
way tunnel was contributed by Evans (1864; 1871). Among the more recent literature mention
may be made of papers by E. E. S. Brown (1941 ), who described the Folkestone Beds and basal
Gault in the Wrotham Heath area, by Dighton Thomas (in Wright and Thomas 1946), dealing
with the Hythe Beds of Dryhill, near Sevenoaks, and by Wells and Gossling (1947), who made
a special study of the pebble-beds in the Lower Greensand of East Surrey and West Kent.

Compared with East Kent, the present region shows an increase in thickness of the more
arenaceous di visions of the Lower Greensand, the Hythe Beds, and the Folkestone Beds. From
the viewpoint of zonal stratigraphy the most important changes are the westwards passage of
the Sandgate Beds basal nodule-bed into 60 or 70 feet of rag, hassock, and cherts of Hythe
Beds facies and the incoming at the western end of the region of the lower horizons of the
Atherfield Clay.

Atherfield Clay. This division crops out in a narrow tract along the foot of the Hythe Beds
escarpment, but is seldom exposed and in the field is difficult to distinguish from the Weald
Clay below. Over much of the outcrop its precise thickness is unknown. About 30 feet thick
at Maidstone, it expands southwards and may double this thickness on the escarpment
between Yalding and Linton. It consists mostly of silty clays, grey, blue, yellow, and reddish,
with a few calcareous and ferruginous claystone nodules. At the junction with the Hythe Beds
it is frequently glauconitic and sandy. Locally it contains seams of fuller’s earth.

The junction of the Atherfield Clay and the Hythe Beds may be seen in a pit formerly worked
by the Fuller’s Earth Union, a quarter of a mile north of Leeds Church, about 4 miles south-
west of Maidstone. The following section was measured in 1955 in steeply dipping strata:

Section of Atherfield Clay and Hythe Beds, j mile north of
Leeds Church , Kent

ft. in.

Hythe Beds

4. Alternation of rag and hassock .
3. Brown-grey fossiliferous ragstone

estimated 25 0
6-9

2. Grey-green glauconitic hassock with scattered pale phosphatic nodules;
impersistent hard band at base ........

10 0

Atherfield Clay

1 . Blue, slightly sandy clay, paler at top .

about 30 0

Total about 65 6

RAYMOND CASEY: STR ATIGR APHIC AL PALAEONTOLOGY OF GREENSAND 537

The clay has a good fauna of microzoa and from the topmost 10 feet were obtained crushed
specimens of the ammonite Deshayesites forbesi sp. nov.

The railway cutting at Teston, in the Medway Valley, at one time exposed the junction of
the Atherfield Clay and Weald Clay. Simms (1845) noted that ‘the beds resting on the Wealden
in this locality seem to be identical with the marine clays found at Hythe and at Atherfield in
the Isle of Wight. . . . There is also a bed of stone, not a continuous bed, but in concretionary
masses, just above the junction, from which I obtained fossils, and which, I consider, repre-
sents the Atherfield rocks.’ Unfortunately the fossils mentioned by Simms have not been
preserved, but the reference to fossiliferous concretionary masses at the base of the clays is
strongly suggestive of the Perna Bed. If confirmed, this would be the most easterly known
occurrence of the Perna Bed.

The best section of Atherfield Clay in this region was seen by Evans (1864; 1871) about a
century ago when the Sevenoaks railway tunnel was cut. His estimate of 50 feet for the
thickness of the beds included an upper portion of dark clayey sand containing ‘a vast amount
of water’ — almost certainly the basal sands of the Hythe Beds. From the greyish and blue-
coloured sandy clays overlying the Weald Clay he collected many fossils which were later
presented to the British Museum (Natural History). Other Atherfield Clay fossils from this
locality are in the Meyer Collection in the Sedgwick Museum. Most of Evans’s fossils were
obtained from cemented masses abounding in Mulletici [Perna] mulleti, and although the zonal
ammonite (always rare) was not found, the existence of the Perna Bed is itself proof that the
obsoletus Subzone of the fissicostalus Zone is present. The next higher forbesi Zone is denoted
by Ancyloceras mantelli in the Evans Collection and by Deshayesites forbesi in the Meyer
Collection, the latter labelled ‘Atherfield Clay, top’. The venerids Resatrix dolabra and Pseuda-
phrodina ricordeana are well represented in Evans’s Collection and the hinge structures of
these two species were first illustrated by some of his specimens (Casey 19526, pi. 9, figs. 1, 9).
Deshayesites forbesi was found in the Atherfield Clay samples from boreholes at Sundridge
and at Riverhead, near Sevenoaks, at depths of 198 and 250 feet respectively (Geological
Survey Collections). Clearly, at the western end of the region the Atherfield Clay has elements
of both fissicostatus and forbesi Zones and is probably a condensed version of the whole of the
Atherfield Clay Series of the Isle of Wight.

Hythe Beds. The Hythe Beds rise from the plain of the Weald Clay as a line of hills and sloping
cliffs cut by the valleys of the Medway, Len, Great Stour, Darent, and tributaries. About 45
feet thick in the Ashford district, they expand westwards, reaching a maximum thickness of
1 50 feet on the escarpment west of Sevenoaks. Over most of the outcrop the beds maintain a
‘ rag and hassock’ facies similar to that of East Kent, but west of Maidstone there is a gradual
change to a more sandy type of lithology. One important point of difference compared with
the East Kent region is the introduction of chert in the highest beds.

A large quarry at Little Chart, a quarter of a mile south-south-west of the Swan Inn and
about 4 miles north-west of Ashford, provided in 1949 a clear section of the greater part of the
Hythe Beds, as given on p. 538.

The presence of chert in the residual bed at the top and the absence of a phosphatic nodule-
bed at the junction with the Sandgate Beds are typical of the Hythe Beds throughout the whole
region. Bed 3 contains Ch. meyendorffi, and the association of black nodules and oysters is
reminiscent of bed 30 of Otterpool, near the base of the meyendorffi Subzone. At Little Chart
the nodules are larger and more numerous and it is possible that this bed marks a pause in
deposition equivalent to the whole of the meyendorffi Subzone of the Hythe district. The same
bed, with phosphatic nodules, oysters, and Ch. meyendorffi, was found by Mr. Worssam in a
small disused quarry on the eastern boundary of Surrenden Dering Park, a quarter of a mile
north-east of Rooting and about half a mile south-west of Little Chart. From the ragstone

538

PALAEONTOLOGY, VOLUME 3

bands below the meyendorffi horizon in the Little Chart quarry, mostly picked up loose on the
quarry floor, were obtained: Tropaeum bowerbanki, Australiceras gigas, Cheloniceras cor-
nuelianum, Ch. crassum, Dufrenoyia furcata, and D. lurensis, an assemblage indicative of the
transitoria Subzone of the bowerbanki Zone.

Summarized section of Lower Greensand in Little Chart Quarry, 1949

ft. in.

Sandgate Beds

5. Decomposed glauconitic loam ........ 1 0

Hythe Beds

4. Reddish-brown decalcified hassock with weathered slabs of chert and grey

ragstone ............ 4 0

3. Grey ragstone with black phosphatic nodules and abundant Exogyra

latissima ............ 1 0

2. Grey-green hassock with nodules of ragstone . . . .20

1 . Alternation of hard grey ragstone and grey-green hassock, estimated . 30 0

About 38 0

The next good exposures are in the Maidstone district. From very early times this town has
been the centre of a thriving ragstone-quarrying industry and it is surrounded by a number of
active and disused workings that give excellent sections of the Hythe Beds. The most famous
of all, now defunct, is the lguanodon Quarry, owned by W. H. Bensted, who in the last century
made many important finds in this formation. Fitton (1845) noted that the stone in the Maid-
stone quarries, especially at Boughton, in contrast to that of the other parts of the Kentish
Rag tract, assumes the form of continuous and uniform strata and he suggested for this part
of the Lower Greensand the term Boughton Group. Many of the courses of ragstone (locally
termed lanes) are traceable over a wide area and are given distinctive names by the quarrymen.
Thus, a lane just above the middle of the sequence, overlying a bed of hassock full of soft,
smutty phosphatic nodules, is called the Coalman, and another, nearer the base, underlying
a similar hassock bed, is known as Blackjack. Soft phosphatic nodules are disseminated to a
lesser extent through most of the hassock beds and were called ‘molluskite’ by Bensted (1860).
Above the Coalman the beds have lenses and nodules of chert and fossils are sometimes
chalcedonized. Some of the ragstones are saccharoidal and many have a high content of
microscopic organic debris. Fragments of the calcareous alga Girvanella intermedia have
been identified by Dr. F. W. Anderson, but none of the ragstones is a true algal limestone.

Ammonites are not common and when found by the quarrymen are often sold as garden
ornaments. Of those that have come into my hands, many have been found loose on tip-heaps
and others have been purchased from the men ; few have been localized precisely in the sections.
It is evident, however, that the deshayesi and bowerbanki Zones of the Lower Aptian and the
martinioides Zone of the Upper Aptian are all present in the Hythe Beds of the Maidstone
district. The Blackjack horizon, which in the easterly part of the district holds an abundance
of Exogyra, is the boundary of the deshayesi and bowerbanki Zones, and the Coalman Lane
is taken as the base of the martinioides Zone. It is impossible at present to fix the boundaries
of the different subzones. Since the quarries at Boughton work mainly the last zone, I have
elsewhere (Casey 1960c:, pp. 37-38) proposed to adopt Fitton’s term Boughton Group for
this upper part of the Hythe Beds of the Maidstone area, which in East Kent is represented by
a bed of phosphatic nodules at the base of the Sandgate Beds. The invertebrates of the Lower
Aptian portion are essentially the same as described in East Kent.

In the exposure near Leeds Church, mentioned on an earlier page, the beds above the Ather-
field Clay have yielded Deshayesites deshayesi and Cheloniceras sp. (bed 2) and a new species
of Deshayesites characteristic of the Scaphites Beds of Atherfield (bed 3), thereby proving the

RAYMOND CASEY: STR ATIG R A PH1C AL PALAEONTOLOGY OF GREENSAND 539

parinodum and grandis Subzones of the deshayesi Zone. Another typical grandis Subzone
ammonite, Tropaeum hillsi, was collected by Mr. Worssam from 1 foot 3 inches below an
Exogyra bed (Blackjack horizon) exposed on the north bank of Mill Pond, 900 yards N.
15° W. of Leeds Church. It was from the Maidstone district that Sowerby obtained some of
the specimens used in the original description of this species.

Spot Lane Quarry, Otham, sprawled over a large area of cambered Hythe Beds, has for the
past few years shown a good section of the beds in the vicinity of the Coalman Lane. From
the hassock just beneath the Coalman, associated with numerous Exogyra, Linotrigonia, and
the belemnite Neohibolites ewaldi, I collected Tropaeum bowerbanki, Cheloniceras meyendorffi,
and indeterminate Dufrenoyia. At Skinner’s Quarry, Brishing Court, near Boughton Mount,
south of Maidstone, almost the whole of the Boughton Group is exposed, overlying about
15 feet of bowerbanki Zone. Chert and sand, known locally as ‘callow’, form the top 18 feet,
and from between this and the Coalman Lane (called the Newington Lane in this quarry) I
have secured a large number of ammonites, mostly with the co-operation of the quarry fore-
man and the owner, Mr. Skinner. The list is as follows: Cheloniceras (Epicheloniceras) marti-
nioides sp. nov., Ch. (E.) aff. debile sp. nov., Ch. (E.) graci/e sp. nov., Ch. (E.)spp. nov., Tropaeum
benstedi, Ammonitoceras sp. nov. From these fossils it is possible to say that the 20 feet or so
of ragstone above the Coalman are the correlatives of Groups VIII, IX, and X of the Isle of
Wight, i.e. the 104 feet of strata from the base of the Upper Crioceras Beds to the top of the
Upper Gryphaea Beds. It is probable that the unfossiliferous ‘callow’ is the equivalent of
Groups XI and XII of the Isle of Wight, also very poor in fossils, and represents the buxtorfi
Subzone at the top of the martinioides Zone.

The quarries at Tovil, a southern suburb of Maidstone, have fallen into disuse and it is
not known which one furnished the type specimen of Ammonitoceras tovilense, described by
Crick (1916).

Very large workings in Hythe Beds are situated at the Coombe and Postley quarries, about
a quarter of a mile north-west of Hayle Place. At Coombe Quarry over 60 feet of Hythe Beds
are seen below a thin capping of Sandgate Beds loams. The Boughton Group ( martinioides
Zone) is here about 40 feet thick, this being perhaps little more than a third of the total thick-
ness of Hythe Beds in this neighbourhood. The zone fossil Cheloniceras ( E .) martinioides was
collected from the Chance Lane, just below a thick development of ragstone and chert (The
Flint) and about 8 feet above the Coalman. A specimen of Cheloniceras ( E .) aff. debile sp. nov.
was also found at the same general level. Twenty feet above the Coalman, in the Thrasher
Lane, a thick ragstone band with pockets of rusty-sand (‘snuff-boxes’), I collected Tropaeum
cf. rossicum.

The Iguanodon Quarry, 75 feet deep, was situated on the west side of Maidstone, south of
the main London road. The circumstances surrounding the discovery of the skeleton which is
the type of Iguanodon mantelli, now in the British Museum (Natural History), have been
narrated several times (Mantell 1834; Buckland 1836; Owen 1851; Bensted 1860, 1862;
Swinton 1951, &c.). Judging by Bensted’s description it was found in the bowerbanki Zone,
above the ‘molluskite hassock’ (Blackjack horizon) with frequent ‘ Nautilus elegans ’ (Cyma-
toceras pseudoelegans ) and below a thick cherty series (Boughton Group). The limestone was
said to abound in ammonites and sharks’ teeth (Buckland 1836). The British Museum col-
lections contain dental plates of the chimaeroid fish Ischvodus thurmanni and teeth of Hetero-
dontus sulcatus and Hybodus complanatus labelled ' Iguanodon Quarry’ and in a matrix identical
with that of the dinosaur. The types of Synechodus tenuis, labelled simply ‘Greensand, Maid-
stone’, have the same sort of matrix. Bensted found a tooth of the marine reptile Polyptychodon
continuus in the ‘molluskite hassock’, and some 15 feet below the Iguanodon level he discovered
the carapace of a large turtle, subsequently made the type of a new genus and species, Protemys
serrata (Owen 1851). The horizon of this last find must fall within the deshayesi Zone; Owen

540

PALAEONTOLOGY, VOLUME 3

remarked on the abundance of sponge-spicules in the matrix of the fossil, which in this and
other respects agrees with that of the lectotype of Tropaeum hil/si, also from Maidstone.

Not least of Bensted’s discoveries in the Iguanodon Quarry were beds rich in plant remains
in the Boughton Group. Coniferous wood from this quarry is described in Stopes’s Catalogue
under the names Pityoxylon benstedi, Pinostrobus benstedi, P. patens, Cedrostrobus mantelli,
Cedroxylon maidstonense, Abietites cf. solmi, and Cupressinoxylon cryptomeroides. Unfor-
tunately, the unique type specimens of the angiosperm Hythia elgari and the bennettitalian
Bennettites allchini are not localized closer than ‘Maidstone’ and it is not known if all were
one flora. Bensted’s most famous plant discovery, the ‘Dragon Tree’, excited great interest
for many years. Originally thought to be a monocotyledon, it was named Dracaena benstedii
by Konig and figured under that name by Mackie (1862). Seward (1896) later transferred it to
the cycads, giving it the generic name Benstedtia. Finally, Stopes (1911; 191 In) showed that
it was merely a rotted piece of the woody trunk of one of the higher conifers and commonplace.
Petromonile benstedi, an organic structure resembling a string of beads, once believed to be a
sponge, also occurred in the Boughton Group of this quarry.

A quarry about half a mile south-west of Allington Church, still worked by the Bensted
family, has yielded Cheloniceras ( E .) gracile sp. nov. in the highest beds exposed, apparently
equivalent to the Thrasher Lane of Coombe Quarry. The Blackjack Lane is present at the
bottom of the quarry with an overlying hassock crowded with the usual crushed fossils and
nodules. Cymatoceras pseudoelegans is the dominant cephalopod, both this nautiloid and the
belemnite Neohibolites ewaldi outnumbering the ammonites, here represented by Cheloniceras
of the cornuelianum type and a doubtful Australiceras gigas. This is one of the few horizons
in the Mesozoic where nautiloids have an ascendancy over ammonites.

The Town Mailing Quarry, East Mailing, whence came a specimen of Ch. ( E .) martinioides
in the British Museum (Natural History), is now overgrown and it has not been possible to
trace the provenance of some half dozen specimens of this species in the Maidstone Museum,
labelled simply ‘Maidstone’ or ‘Boughton’.

West of the Maidstone area the Hythe Beds increase in thickness and begin to partake of a
more sandy character. Large quarries just west of Off ham (Brown 1941) show 70-80 feet of
glauconitic and sandy ragstone alternating with gritty glauconitic hassock. The greater part
of this thickness belongs to the martinioides Zone, the only ammonites obtained being Tropaeum
benstedi and species of Epicheloniceras from near the base, both diagnostic of that zone. A
conspicuous bed of coarse sandy ragstone, 2 feet thick, with phosphatic nodules at the base
(Granny Lane), lies a few feet above the quarry floors and may be the equivalent of the Coal-
man Lane of the Maidstone area.

A rapid thinning of the martinioides Zone takes place west of Offham. In the large rambling
quarries at Basted House, between Ightham and Borough Green, about 3| miles west of the
last exposure, the top 70 feet of the Hythe Beds are displayed, of which only 45 feet can belong
to the martinioides Zone. This zone may in fact be confined to the topmost few feet in which
brown and pink chert (Sevenoaks Stone or ‘Shatter Rock’) is prevalent. Crushed Tropaeum
bowerbanki and Cheloniceras cf. cornuelianum occur in the hassock in the bottom 25 feet of the
workings. From a hassock bed near the floor of the quarry I collected a piece of a Kimmeridgian
Pavlovia, in black phosphatic preservation like those from the "rotunda- bed’ in the Warlingham
boring (Allen 1960, p. 161). Weathered surfaces of the ragstone at this locality are good for
collecting polyzoa.

A boring sunk to 290 feet, 850 yards S. 9° W. of the George Inn, Trottiscliffe, proved
97 feet of rag and hassock but just failed to bottom the Hythe Beds. The beds were coarse,
sandy and fossiliferous; a band between 273 ft. and 276 ft. 3 in. contained crushed Cymato-
ceras and partly phosphatized ammonites ( Cheloniceras and Dufrenoyia ) with pebbles and
Exogyra at the base, probably the Blackjack horizon.

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 541

From the large quarry near the Wheatsheaf Inn, West Mailing, in the martinioides Zone,
Brown obtained a petrified stem of the fern Protopteris fibrosa, known otherwise only by the
type-specimen, from the Turonian of Silesia (Whiteside 1956).

Roadstone quarries at Dryhill, Sundridge, 2 miles west of Sevenoaks, show 60 feet of sharply
folded and faulted Hythe Beds, briefly described by Dighton Thomas (in Wright and Thomas
1946). In the north face of the present working quarry is a faulted-down block of cherts and
coarse limestones of martinioides age, with Epieheloniceras and Ammonitoceras sp. nov. Else-
where the sandy rag and hassock contains the usual bowerbanki fauna of crushed mollusca,
with the ammonites Tropaeum bowerbanki, Cheloniceras cornuelianum and Dufrenoyia spp.
(= Deshayesites of Thomas). Among the nodules scattered through the hassock beds are
phosphatized pieces of Dufrenoyia, Cheloniceras, Sanmartinoceras (Sinzovia), Aconeceras
nisoides, and unnamed Aconeceratidae. The strong representation of the last-named family,
not otherwise known on this horizon in the Lower Greensand, gives point to my comments
on the curious sporadic distribution of this group of ammonites (Casey 1954c).

Sandgate Beds. Throughout West Kent the Sandgate Beds present a facies of glauconitic
loams and silts generally sterile for the palaeontologist. Though perhaps reaching a thickness
of 70 feet in the eastern part of the region, they become exceedingly thin in the Maidstone and
Sevenoaks areas, dwindling to as little as 4 feet in places. Records of fossils from the Sandgate
Beds at Aylesford (Himus 1939) have not been confirmed and may have been based on discards
from a nearby working in Folkestone Beds. At the Basted House quarries, where a few feet
of Sandgate Beds are let down into Cenozoic fissures in the Hythe Beds, the quarrymen dug
out the silicified trunk of a pine-tree, 12 feet long (Casey 1951c), portions of which are now in
the Geological Survey Museum.

Folkestone Beds. From about 110 feet at Ashford, the Folkestone Beds thicken westwards to
200 feet or more west of Sevenoaks. They are current-bedded, more or less ferruginous sands,
with a few pebbly or silty layers or seams of pipe-clay. Accumulation of the topmost beds in a
series of regularis-mammillatum troughs has resulted in a more varied lithology. Bands of
glauconitic or ferruginous sandstone appear locally close below the Gault; around Oldbury
Camp, near Sevenoaks, this part of the formation contains the well-known Ightham and
Oldbury Stones, beds of hard green chert and of brown quartzite respectively. Everywhere the
junction with the Gault is marked by a few feet of glauconitic sandy clays and clayey sands
with phosphorite nodules.

The main mass of the sands is almost completely devoid of fossils, though careful examina-
tion frequently reveals the presence of burrows and other structures indicating the work of
animals. In places, as at Wrotham, the type of bedding, wind-polished sand-grains and
absence of fossils, has raised the question of aeolian formation (Casey 1946). It is now known
that such an association does not exclude a marine environment of origin: dune-bedding may
be reproduced by the movement of sand-bars under water, and aeolian-type grains may be
blown or washed into the sea.

In Eastwell Lane, about a mile north of Ashford, the top of the Folkestone Beds may be
seen in sandpits on either side of the road. Just beneath the soil in the eastern pit is the basal
nodule-bed of the mammillatum Zone with S. kitchini, followed downwards by a few feet of
yellowish greensand and thin seams of cherty, spicular sandstone as in the regularis Subzone
at East Cliff. In the 6 miles of country between here and Brabourne Lees we seem to pass over
the crest of the anglicus-puzosianus unconformity and enter another regularis-mammillatum
basin. Little more can be learnt about this basin. A ditch 875 yards south-west of the Olive
Branch Inn, Westwell Leacon, about 4 miles west of Eastwell Lane, showed a second concen-
tration of nodules below the main bed with the puzosianus fauna, and the presence of the

B 6612

n n

542

PALAEONTOLOGY, VOLUME 3

raulinianus Subzone was confirmed by finding the index ammonite on the nearby ploughed
field. Ditches at Harrietsham and an old pit at the top of Weavering Street, Maidstone, showed
a mammillatum- bed of puzosianus age, but neither exposures were good enough for critical
study. Diggings made in 1958 for the new Maidstone bypass road entered the mammillatum
Zone at the southern end of Cottage Wood and just east of Longham Wood, on either side
of Chrismill Bridge, west of Hollingbourne, and at the ‘clover leaf’, north of the Chiltern
Hundreds Inn, on the north-east side of Maidstone. Below vivid green sandy clays of the
dentatus Zone the puzosianus nodule band was seen to pass down gradationally to running
sands of Folkestone type, but the intervening 4 feet of glauconitic and phosphatic passage-beds
failed to yield ammonites diagnostic of a subzone.

Ferruginous nodules picked up from the roadside and allotments at the top of Weavering
Street, Maidstone, about 200 yards north of the cross-roads by Birling House, were full of
moulds of the gastropod Anchura ( Perissoptera ) parkinsoni and the ammonites Hypacantliop-
lites clavatus and H. spp., indicative of the anglicus Subzone of the jacobi Zone. I could not
find exactly where the nodules came from, though the fine sand clinging to them is like that
found here in the middle of the formation.

Folkestone Beds have been extensively dug by the Aylesford Sand Company, north of the
village of Aylesford, 2\ miles north-west of Maidstone. The succession seen in 1948 is sum-
marized below.

Summarized section of Folkestone Beds in Aylesford Sandpits

ft.

in.

(Medway Gravels above)

4.

Ochreous sands, pebbly in places, with large lenticles of ferruginous
sands full of Exogyra conica ......

about

20

0

3.

Yellow sands with wisps of clay and clay balls up to 6 in. diameter

about

15

0

2.

Grey silt ..........

about

12

0

1.

Silver sands, conspicuously current-bedded in wedged-shaped
units resembling those of sand-dunes .....

about

45

0

Total about

92

0

The sequence Silver Sands-Silt Band-Coarser Sands is remarkably similar to that described
by Gossling (1929) in eastern Surrey, and although we are unable to follow these three divi-
sions through the intervening country, correlation with the Surrey succession may be correct.
The probability of the Clay-Silt Band being the Aptian-Albian boundary is discussed in the
account of Surrey.

Phosphorite-cemented lumps of grit, either loose or attached to large Ostrea cunabula,
occur at the base of the Gravels and show that a nodule-bed once existed at the top of the sands.
The lenticles in bed 4 suggest the fossilization in situ of oyster-banks and yield a disappointing
set of long-ranging molluscs. Poorly preserved lamellibranchs in iron concretions are found
on the same general horizon at the top of the sandpit by the railway line north of Wrotham
and Borough Green Station.

Sandpits in the lower third of the Folkestone Beds at Ivy Hatch, seven-eighths of a mile
south-south-west of St. Peter’s Church, Ightham, show current-bedded sands with seams of
pipe-clay and gravelly layers full of silicified valves of large thick-shelled lamellibranchs,
mostly Epicyprina harrisoni sp. nov., Yaadia nodosa, and Gervillella sublanceolata. Most of
the shells are broken and the whole deposit suggests a littoral, if not intertidal, environ-
ment. Epicyprina harrisoni, Pterotrigonia mantelli, Tortarctica similis, Panopea gurgitis, and
Neithea quinquecostata occur also in the Oldbury Stone. At Styants Bottom, west of Oldbury
Hill, loose blocks of stone derived from the Folkestone Beds contain chalcedonized polyzoa
and shell debris; and similar derivatives, with silicified sponges ( Plocoscyphia ), have been

RAYMOND CASEY: STR ATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 543

picked up at various localities between Sevenoaks and Ightham. Silicified wood from the
sands at Ightham provided Stopes (1915) with the types of the angiosperms Cantia arborescens
and the conifer Pityoxylon sewardi.

The rest of the exposures dealt with in this region belong to tire mammillatum Zone and it
will be convenient to start with that at Westerham, where the sequence is fully developed, and
to follow them eastwards.

Squerrye’s main pit, 500 yards east-north-east of Westwood Farm, Westerham, is a vast
opening in Folkestone Beds, plainly visible from the heights above Westerham. The junction
with the Gault is clearly exposed for about 200 yards along the northern side of the pit, where
the following section was demonstrated to the Geologists’ Association in July 1953:

Section of Folkestone Beds and basal Gault in Squerrye's main pit , Westerham

dentatus Zone (benettianns and eodentatus Subzones)

16. Grey, glauconitic clay with rusty streaks and iron-stained phosphatic ft. in.
nodules (Lyelliceras lyelli, Prolyelliceras sp., Beudanticeras laevigatum,

Hoplites benettianus, &c., in nodules) ..... seen 3 0

15. Grey, glauconitic clay with rafts of green sandy clay and large septarian
nodules, flying to bits when tapped; rusty streaks ( Hoplites baylei,

Lyelliceras , Isohoplites, &c., in nodules) . . . . . .40

14. Blue-green sandy clay with phosphatic nodules scattered throughout
and concentrated in a band at the base ( Isohoplites eodentatus, D.
inaequinodum) .......... 3 0

mammillatum Zone ( puzosianus Subzone)

13. Blue-green sandy clay with scattered putty-coloured phosphatic nodules 1 6

12. Band of putty-coloured nodules in matrix as above ( Otohoplites spp.

nov. in nodules) .......... 6

mammillatum Zone ( raulinianus Subzone)

1 1 . As bed 13. . . . . . . . . . . 16

10. As bed 12 ( O . raulinianus, D. mammillatum, B. newtoni in nodules, very

rare) ............ 4

mammillatum Zone (floridum Subzone)

9. Moss-green sandy clay with a line of putty-coloured nodules at base . 1 10

8. Moss-green sandy clay ......... 10

7. Band crowded with putty-coloured nodules in matrix as above (D.
mammillatum, B. newtoni, Cleoniceras floridum, Protanisoceras acteon,

&c., in nodules) .......... 2-4

6. Blue-grey, dicey clay, slightly sulphurous; incipient development of
nodules at top; much glauconite and arenaceous forams. in washed
residue . . . . . . . . 4 ft. to 5 6

5. Grey, very sandy clay with mauve and green streaks, passing up into

bed 6 . . . . . . . . . .15 in. to 26

mammillatum Zone ( kitchini Subzone)

4. Yellow-green clayey sands ........ 3 6

3. Band of white phosphatic nodules; densely packed; iron-stained in
places; abundant small pebbles ( S . kitchini. Cl. morgani, &c., in nodules,
rather rare) ........... 4-9

? tardefurcata Zone

2. Sharp white sand, current-bedded with giant foresets; small pebbles
tending to concentrate in lines; a conspicuous 6 in. pebble-bed 20 ft.
above base ........... 85 0

1 . Silt Band. Buff and grey silt ....... seen 6 0

Total about 120 0

544

PALAEONTOLOGY, VOLUME 3

No dividing line has been drawn between Folkestone Beds and Gault in this section. Above
the clean white sands of typical Folkestone facies there is a thick series of clayey sands and
sandy clays that grade upwards into Gault. If this section were considered on its own merits,
the base of the Gault would best be drawn at the bottom of the mammillatum Zone (bed 3).
This, however, is the correlative of the kitchini bed of East Cliff, Folkestone, unquestionably
part of the Folkestone Beds.

This pit is probably the most important in south-east England for studying the succession
of ammonite faunas in the mammillatum Zone and the lower part of the dentatus Zone, for
although the abundance of glauconite and contemporary phosphate indicates slow deposition,
the sequence is not so condensed and incomplete as it is at Folkestone. At the latter locality,
for example, th efloridum and raulinianus faunas lie together in the same remanie bed and the
benettianus fauna, with Lyelliceras and Prolyelliceras , is missing (though present in the
Chislet and Guilford Collieries). At Westerham the principal source of fossils is bed 7, just
above the thick clay bed, which has produced many species of Cleoniceras and Protanisoceras
besides the common forms of Douvilleiceras and Beudanticeras. Of the minority fauna the
forms most commonly met with are Cleoniceras floridum sp. nov., Cl. ( Neosaynella) inornatum,
and Protanisoceras acteon. Also present are Cl. cleon. Cl. spp. nov., Cl. (N.) sp. nov., P. blancheti,
P. cantianum, P. vaucherianum, P. sp. nov., and very rare Sonneratia. Heteromorphs are quite
a feature of this locality and horizon. Here also are found frequently the crabs Notopocorystes
stokesi, Eucorystes broderipi, and, rarely, Homolopsis edwardsi', lamellibranchs such as Ino-
ceramus salomoni, Leionucula ovata, Entolium orbiculare, Neithea quinquecostata, and Plicatula
gurgitis and the gastropods Gyrodes genti, Anchura ( Perissoptera ) parkinsoni, and Semisolarium
moniliferum are also typical . Several well-preserved belemnite phragmocones have been obtained
from this bed, though guards are unknown; what are generally mistaken for belemnites in
this bed are isolated lengths of ammonite siphuncle. The overlying beds of the mammillatum
Zone are poor in fossils and the presence of the raulinianus and puzosianus Subzones has been
proved only after many years collecting.

Throughout most of its length the pit-face shows a thick band of clay (beds 5-6) resting with
sharp junction on the white sands of bed 2, as noted by previous observers (Wells and Gossling
1947, p. 196). In 1954 extension of the pit in a north-easterly direction disclosed a wedge of
kitchini Subzone between beds 2 and 5, based by a band of white phosphatic nodules with
rare ammonites ( Sonneratia kitchini, S. sp. nov., Cleoniceras morgani, Anadesmoceras baylei,
and D. mammillatum). Followed westwards for about 75 yards this nodule-band (bed 3)
thinned away and was replaced by an impersistent seam of pebbly ironstone. The overlying
sands (bed 4) also thinned away rapidly, so that the ironstone was brought up to form the
junction of beds 2 and 5. This pit thus lies on the south-western flank of another regularis-
mammillatum basin and gives us a chance glimpse of one mammillatum Subzone overlapping
another.

Important information on the easterly extent of this basin was forthcoming from a Metro-
politan Water Board well, one-fifth of a mile south-east of Brasted Railway Station and 2|
miles north-east of Squerrye’s main pit, Westerham. Here, below the stiff grey clays of the
Gault, between depths of 59 and 77 feet, the strata set out below were entered.

The bright-green sandy clay (bed 8) is the obvious correlative of the basal dentatus Zone
and upper mammillatum Zone of Squerrye’s pit, and beds 6 and 7, containing crushed D.
mammillatum and Protanisoceras, could be recognized as a thinner and sandier version of
bed 6 of Squerrye’s. But the chief point of interest here is the passage of these clay beds down
into greensands and sandstones that have no counterpart at Westerham. This part of the
succession is much more like that of the regularis and basal mammillatum beds of East Cliff,
Folkestone, a similarity that would be strengthened if it could be shown that the bottom
phosphatic horizon (bed 3) is the S. kitchini bed. At all events, it is clear that sedimentation

RAYMOND CASEY: STR ATIGR APH ICAL PALAEONTOLOGY OF GREENSAND 545

in regularis or lower mammillatum times was relatively free in this area and that we are near
the centre of the basin whose flank is exposed at Westerham.

Strata in Metropolitan Water Board well at Brasted, Kent,
between depths of 59 and 77 feet

8. Bright green, glauconitic sandy clay with grey-green, black-hearted, phos-
phatic nodules; pyritic algal threads in top few feet .

7. Pale grey, dicey clay with sandy pockets ......

6. Clay as before, but more sandy, passing into ......

5. Bright green sandrock ..........

4. Coarse green sandstone with Entolium orbiculare .....

3. Yellow-green, somewhat clayey sand with a few phosphatic nodules and
phosphatized sponges ..........

2. Olive-green clayey sand with small pebbles ......

1 . Yellowish greensand ......... seen

ft. in.

8 0
1 0
1 0
1 0
1 0

2 0
2 0
2 0

Total 18 0

The next good exposures are around Wrotham, about 14 miles east-north-east of Wester-
ham, where the junction of the Folkestone Beds and the Gault may be studied in a number of
pits north of Wrotham and Borough Green Railway Station and near Wrotham Heath. Since
they all show a similar succession, the one most productive of fossils will be taken as standard.
This is a large opening made by the Rugby Cement Company east of Ford Place, on the lane
leading to Trottiscliffe, nearly three-quarters of a mile north-east of Wrotham Heath cross-
roads and adjacent to an older working (Olley’s pit) described by E. E. S. Brown (1941, p. 8).
Both give the following succession:

Basal Gault and Folkestone Beds exposed in sandpits at Ford Place,
Wrotham, Kent

Basal dentatus Zone

9. Very dark, glauconitic, sandy clay with brittle phosphatic nodules (rare
Hoplites and B. laevigatum) ........

mammillatum Zone ( puzosianus Subzone)

8. Band of dark, gritty phosphatic nodules in a matrix of dark-green, gritty
clay .............

7. Dark-green sandy clay with scattered black-hearted, gritty phosphatic
nodules ............

mammillatum Zone (raulinianus Subzone)

6. Band of white-skinned, dark-centred, gritty phosphatic nodules in a matrix
of brown clayey sand .........

5. Brown-weathering, glauconitic loam with scattered white-skinned, gritty
phosphatic nodules ..........

mammillatum Zone ( floridum Subzone)

4. Concretionary band of whitish, friable phosphatic nodules in a matrix of
reddish-brown loam ..........

3. Grey-brown, plastic sandy clay ........

2. Brown clayey sand with wisps of pure clay, scattered small pebbles, and
incipient phosphatic nodules. Near the base a few large pebbles (up to 4 in.)
of micaceous siltstone .........

(Sharp junction)

? tardefurcata Zone

1 . White and buff, coarse to medium grained sands, current-bedded with giant
foresets ........... seen

Total about

ft. in.
3 0
6

1 0

4

10

2-6

8

4 0

25 0
35 6

546

PALAEONTOLOGY, VOLUME 3

The current-bedded sands of bed 1 are succeeded abruptly by the mammillatum- beds, which
form a passage into the blue-grey clays of the Gault. Differential oxidation of the glauconite
in these passage beds causes a gradual colour-change upwards from rusty brown to dark green.
Both in lithology and fossils the succession is different from that of Westerham and Brasted
and these localities may belong to another basin of deposition. The lowest concentration of
nodules (bed 4) has yielded ammonites of the floridum Subzone, such as C. floridum and C. (N.)
inornatum, though the rarity of heteromorphs and the presence of Inoceramus coptensis and
species of Sonneratia denote a level below the floridum nodule-bed of Westerham. Common
fossils in this bed at Ford Place are D. mammillatum, B. newtoni, Nanonavis carinatus, Cucullaea
glabra, Tortarctica similis, Inoceramus salomoni, Entolium orbiculare, and Semisolarium
moniliferum. A few specimens of the coral Trochocyathus fittoni were found in the raulinianus
Subzone, which is otherwise poor in fossils, but collecting over the years has brought to light
an important set of ammonites in the topmost band of nodules ( puzosianus Subzone) (Casey
1959). Fossil wood, bored by Terebrimya, is fairly frequent in bed 8, and here Mr. J. Collins
collected several vertebrae and rib-fragments of the dinosaur Camptosaurus. Ammonite
occurrences are listed below under subzones.

floridum Subzone: Douvilleiceras mammillatum, D. monile, Beudanticeras newtoni, B. dupinianum,
Sonneratia spp. nov., Cleoniceras (C.) floridum, C. ( Neosaynella ) inornatum, Anadesmoceras? , Protani-
soceras acteon, Hamit es cf. praegibbosus. raulinianus Subzone: D. mammillatum, D. monile, B. newtoni,
Otohoplites raulinianus, Pseudosonneratia sp. nov. puzosianus Subzone: D. mammillatum, D. monile,
I), orbignyi, B. arduennense, Otohoplites elegans, O. spp. nov., Protohoplites (P.) archiacianus, P. (P.)
michelinianus. Id. var. nov., P. ( P .) latisulcatus, P. ( Hemisonneratia ) puzosianus, P. (H.) gallicus, Tetra-
hoplites cf. subquadratus, Sonneratia dutempleana, S. spp. nov., Pseudosonneratia spp. nov., Cleoniceras
sp. indet., Tegoceras gladiator, T. mosense, Protanisoceras cantianum.

The supposed differences in the fossils of the various mammillatum exposures in this area
which puzzled Brown (1941, p. 9) were largely the result of fortuitous and inadequate collect-
ing. Compared with Westerham, however, there is a great increase in the puzosianus fauna.
Many of the Protohoplites and Otohoplites are a foot or more in diameter, but the smooth
outer whorls are never completely phosphatized and fall to bits on extraction from the matrix.

Cuttings at the cross-roads at Parson’s Corner, Snodland, about 4\ miles north-east of
Ford Place, show a thin mammillatum-hed at the top of a 15-feet section (Bromehead 1924,
pp. 8, 9). Its nodules are similar to those in the puzosianus Subzone at Ford Place, but larger
and thickly studded with pebbles ; their fossils include large Otohoplites (‘ undescribed ammonite
representing a new genus’, Bromehead, ibid.) and numerous B. newtoni, suggesting that the
bed represents the raulinianus and puzosianus Subzones combined. Downwards the bed passes
into greenish clayey sands with incipient phosphatic nodules at the top and, below, first lenses
of light siliceous stone and then doggers of bright-green pebbly sandstone, up to 3 feet thick.
The siliceous stone is a felted mass of echinoid radioles, with moulds of rhynchonellids and
small Exogyra. Valves of the scallop Entolium orbiculare occur with shelly debris in the green
sandstone, the whole resembling the top of the Folkestone Beds in the Brasted well. Possibly
Parson’s Corner is near the centre of a regular is-mammillatum ‘dimple’ whose southern rim
lies south of Ford Place.

Surrey ( with part of Hampshire)

From the border of Kent at Westerham the Lower Greensand runs west-south-west through
Surrey to Farnham, in the north-west corner of the Weald, and then swings southwards
along the fringe of Hampshire to Petersfield, about 40 miles from Westerham as the crow flies.
Most of this region is covered by Geological Survey Memoirs (Dines and Edmunds 1929 ; 1933 ;
Osborne White 1910), and papers by Meyer (1868), Leighton (1895), Gossling (1929), Kirkaldy

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 547

(1932; 1933fi; 1 947a), Humphries (1956), and Knowles and Middlemiss (1958) deal with the
Lower Greensand in various parts. The region has made two important contributions to
knowledge of the ammonite succession. Owing to a change of facies in the Sandgate Beds the
nutfieldensis fauna has been preserved, and in the Folkestone Beds of the Farnham area are
ammonites of the basal part of the tardefurcata Zone (farnhamensis Subzone), known nowhere
else in Britain.

Atherfield Clay. In general this consists of 15-60 feet of brown and grey sandy clay and buff
loam with concretions of clay-ironstone or calcareous stone at the base (Perna Bed). Under-
ground, north of the outcrop, it may assume a more sandy facies. It is rarely seen. Most of
the fossils have been obtained from the Perna Bed (fissicostatus Zone), the only sign of the
forbesi Zone in the eastern part of the region being the presence of D.forbesi at Diana’s Well,
on the north side of Gibb’s Brook, Oxted, where the clay is brought up by a fault (Gossling
1936). Good exposures of the junction with the Weald Clay used to be seen in Brown’s Brick-
yard at Woodhatch, north-west of Earlswood Common, Reigate, now overgrown. Butler
(1922) and Chatwin (in Dines and Edmunds 1933) gave long lists of fossils collected here from
the Perna Bed (and now in the Geological Survey Museum). All the common Atherfield forms
were found, with hundreds of Mulletia mulleti and the ammonites Prodeshayesites obsoletus,
P. alf. laeviusculus, and P. sp. nov. A similar fauna, but without determinable ammonites,
was recorded from Brockham Brickfield (Gossling 1929, p. 218), and a series of lamellibranchs,
comprising Freiastarte subcostata, Parmicorbula striatula, Pseudolimea parallela, Resatrix
parva and Pseudoptera subdepressa, was obtained from a road cutting in the clay west of
Trashurst, near Dorking (Chatwin, ibid.). Fossiliferous nodules with a Perna Bed fauna were
dug up at Binscombe in 1935 and are now in the Godaiming Museum.

Rich collections of well-preserved fossils could once be obtained from the base of the clay
in the Pease Marsh and East Shalford district, south of Guildford. The best specimens were
found in hard grey and brown lumps, pink-shelled and lustrous. Meyer (1868) recorded up-
wards of 100 species of mollusca, many of which were figured by Gardner (1875; 1876) and
Woods (1899-1913) (see also Chatwin, ibid., 1929). Mulletia mulleti, Venilicardia protensa,
Sphaera corrugata, Yaadia nodosa, and a host of smaller clams (Anomia laevigata, Eonavicula
carteroni, Aptolinter aptiensis, Nuculana scapha, Freiastarte subcostata, Mediraon sulcatum
sp. nov., Senis wharburtoni, Scittila nasuta, Fenestricardita fenestrata, Camptonectes cot-
taldinus, Resatrix dolabra, &c.) and gastropods ( Ovactaeonina forbesiana, Ataphrus albensis,
Tessarolax moreausianum, Confusiscala cruciana, & c.), together with the usual Perna Bed corals,
Holocystis elegans and Discocyathus orbingyanus, a few echinoids (Toxaster fittoni, T. com-
planatus) and brachiopods ( Sellithyris sella, Sulcirhynchia hythensis. Lingula truncata) make up
the bulk of the fauna. The types of Fossarus munitus (Forbes 1845), Dimorphosoma pleurospira
(Gardner 1875), and Pseudaphrodina elongata (Casey 19527?) came from Pease Marsh and the
type of Scalaria meyeri (Gardner 1876) from East Shalford. ‘The very distinct species from
Peasemarsh, resembling Ammonites leopoldinus d’Orbigny’ (Forbes 1845, p. 355) is Prodeshaye-
sites obsoletus.

The whole of the Atherfield Clay, about 60 feet thick, was seen when the railway was cut
at Haslemere. Salter (in Topley 1875, p. 114; in Bristow 1889, p. 48, footnote) noted that here
the junction with the Weald Clay lacked the usual concretions but was marked by ‘abundant
tracks of marine worms, and the Panopaea vertical in their old burrows, within an inch or
two of the dark marls. A great Perna, a coral (Holocystis elegans), and numerous other fossils,
occur in plenty just above these.’ A good set of fossils, including Prodeshayesites alf. obsoletus
and many of the common Perna Bed types, was collected from the old Nutbourne Brickworks,
Shottermill, and is now in the Haslemere Educational Museum (Kirkaldy and Wooldridge
1938, pp. 138-9). Deshayesites fittoni and the venerids Resatrix parva and R. (Vectorbi.

548

PALAEONTOLOGY, VOLUME 3

vectensis denote the presence of the forbesi Zone. The arcticid Proveniella rosacea sp. nov.
gives local character to these clays around Haslemere.

Hythe Beds. In this region the Hythe Beds take on a distinctly arenaceous facies, consisting in
general of sand and sandstone with some beds of chert. They thicken westwards from about
160 feet to nearly double that in the Hurt Wood area, south of Shere; but at Guildford they
are again only 160 feet thick. Dines and Edmunds (1929, p. 18; fig. 5, p. 24) explained these
variations in thickness by supposing that the Hythe Beds were folded and eroded before deposi-
tion of the Sandgate Beds. This idea was contested by Kirkaldy ( 1 933a) but now receives
palaeontological support, as mentioned below. Around Godstone, at the eastern end of the
region, Gossling and Bull (1948) and Gossling (in Kirkaldy 1947a, p. 186) made out the
following succession in the Hythe Beds :

Hythe Beds succession in the Godstone area of Surrey

ft. in.

5. Top Hythe Chert Bed. Massive development of chert in the upper, more

lenticular development in the lower part . . . . . 34 0

4. Upper Hythe Pebble Bed. Yellow coarse sand with small pebbles . 4 6

3. Mid Hythe Sand. Fine to medium sands, glauconitic, with much curvilinear

ironstone and often current-bedded . . . . . . 72 0

2. Lower Hythe 'Stone' . Reddish-brown sand containing layers of soft stone . 45 0

1. Lower Hythe Sand. Greyish tine sands . . . . . . 10 0

Total 165 6

Gossling (ibid., p. 187) recorded Tropaeum cf. hillsi from the Lower Hythe ‘Stone’; in Out-
ward Lane, leading south from Bletchingley, exposures of this part of the Hythe Beds have
yielded Cheloniceras cornuelianum and Ch. parinodum. This means that both the parinodum
and grandis Subzones of the deshay esi Zone are present in the ‘Stone’. The presence of the
bowerbanki Zone in the Mid Hythe Sand is shown by the occurrence of the zone fossil in
lane-side exposures in Mid Street, Nutfield (author’s collection), and in the pits at Cockshot
Hill, and Bell Street, Reigate, and at Taylor’s Hill, Godstone. Pits at the last locality, south-
east of Godstone Green, present a steep 70-feet face of sand with silicified and iron- and
phosphorite-cemented lumps containing fossils. A list of fossils was published by Gossling
(1936) and further collections have been made by C. W. Hobley, A. G. Davis, R. V. Melville,
C. W. and E. V. Wright, and the author, from which the following are selected — Lamelli-
branchia: Barbatia cf. baudoniana. Area sanctae-crucis, Cucullaea cornueliana, Glycymeris
marullensis. Modiolus aequalis, Anotnia pseudoradiata. Cephalopoda: Cymatoceras pseudo-
elegans, Tropaeum bowerbanki, Australiceras gigas, Dufrenoyia sp., Cheloniceras kiliani, obese
var., Neohibolites ewaldi. Echinoidea: Phyllobrissus fittoni, Holaster benstedi. Brachiopoda:
Sulcirhynchia hythensis, Cyrtothyris uniplicata, C. cf. cantabridgiensis, ‘ Ornithella' celtica,
Oblongarcula oblonga. Anthozoa: Oculina liobleyi. Porifera: Plocoscyphia sp., Doryderma sp.,
Geodites cf. wrighti. Reptilia: Pliosaurus sp. (tooth). This is the only known source of Oculina
hobleyi, a surprisingly early occurrence of this genus of corals, which is not otherwise known
before the Tertiary (Thomas 1947). Another unique feature of the fauna is the absence of
gastropods. Most of the fossiliferous lumps are siftings from the dug sand and are not precisely
localized in the section, but the obese variety of Ch. kiliani, a meyendorffi Subzone type, occurs
near the top. Another Lower Aptian Cheloniceras was collected by A. G. Davis from the
Upper Hythe Pebble Bed at Cockshot Hill, Reigate. At this locality Quenstedtoceras mariae,
derived from the Oxford Clay and first recorded by Gossling (ibid., p. 183), occurs in the top
of the Mid Hythe Sand.

The chert beds are often massive and spiculiferous and it was from such beds at Tilburstow-

GO DALMING

o

o

w>

"I

o

o

o

o

o

o

o

»-»

7”

o

o

o

o

text-fig. 6. Comparative vertical sections of the Lower Greensand of the Weald.

550

PALAEONTOLOGY, VOLUME 3

hill and Haslemere that Hinde (1885) obtained specimens to demonstrate the association of
sponge spicules with chalcedonic chert. He founded no less than twenty-eight species of
Axinella, Dirrhopalum, Mastosia, Dory derma, and especially Geodites, on isolated spicules,
and many of them have since been identified in the cherts of Leith Hill and in washed residues
from a clay seam in the Mid Hythe Sand of Bell Street, Reigate (Chatwin, ibid., 1933, pp.
117-18). The last locality and horizon yielded the ostracod Cytheropteron umbonatum. The
ammonite Cheloniceras parinodum, from the lower part of the Hythe Beds, was picked up in
the bed of a stream in Vannmoor, about 5 miles east of Hambledon.

The fossils show clearly that the greater part of the Hythe Beds of the Godstone area are of
Lower Aptian age and that there is room for the martinioides Zone only in the Top Hythe
Chert Bed. This agrees with the picture seen in West Kent, where the cherty martinioides Zone
is thinning west of OfTham. The continued westerly dwindling of these top cherty beds is well
shown in a series of sections between Godstone and Reigate illustrated by Kirkaldy (1947a,
pi. 6). At Redhill and Reigate the Upper Hythe Pebble Bed comes to lie close below the Sand-
gate Beds, with no intervening chert. At Godaiming, about 20 miles west-south-west of Reigate,
the basal nodule-bed of the Sandgate Beds contains derived Oxfordian ammonites similar to
those found in the top of the Mid Hythe Sand at Cockshot Hill and phosphatized ammonites
of the bowerbanki Zone. The latter include a specimen of Dufrenoyia furcata from Holloway
Hill, Godaiming, in the Meyer Collection in the Sedgwick Museum, and another specimen of
the same genus was obtained from an exposure of the nodule-bed in an old quarry 1 mile
west of Bramley Church, east of Godaiming. No clearer proof could be required for the fact
that in this area the Sandgate Beds rest on the eroded top of the Hythe Beds and that the whole
of the martinioides Zone has disappeared. This bears out the observations of Meyer (1868)
and Dines and Edmunds (1929), who believed the Hythe-Sandgate junction to be unconform-
able in this area. That the line of uplift had an east-to-west trend is suggested by the reappear-
ance of the martinioides Zone when the outcrop turns south into the Haslemere district, the
presence of this zone being proved by Cheloniceras ( E .) cf. martinioides from a well at Black-
down (Haslemere Museum). The recurrence of chert beds at Leith Hill, some 3 miles south of
Godaiming, may be part of the same pattern of distribution.

Sandgate Beds. In this region the Sandgate Beds have two distinct facies; around Nutfield,
near the eastern border of the region, they consist of bands of fuller’s earth and cherty and
glauconitic sandstones and limestones, with glauconitic loamy sands above (maximum thick-
ness 80 ft.) ; west of Dorking they can be subdivided into a lower unit characterized by bands
and doggers of calcareous stone (Bargate Beds) and an upper unit of ferruginous loams
(Puttenham Beds). The western facies has been described in detail by Kirkaldy (1933 b).

From early times the fuller’s earth facies has been exploited in the area between Reigate
and Godstone and the pits at Nutfield furnished J. Sowerby (1815) with material for the first
description of a Lower Greensand ammonite, Ammonites nutfieldensis. The variations in
lithology and thickness of the beds are dealt with at length by Gossling (1929) and Dines
and Edmunds (1933). Fossils, mostly mollusca, are found chiefly in the calcareous bands
within the fuller’s earth and include the ammonites Parahoplites nutfieldensis, P. maximus,
P. sp. nov., Tropaeum subarcticum, the nautiloid Anglonautilus undulatus, the large gastropod
Pleurotomaria anstedi, and the following lamellibranchs: Cucullaea cornueliana, Freiastarte
subcostata, Resatrix parva, Eriphyla striata, Inoceramus neocomiensis, Pseudolimea parallela,
Modiolus aequalis, Panopea gurgitis, P. mandibula, Ensigervilleia forbesiana, Linotrigonia
(Oistotrigonia) upwarensis, Pterotrigonia mantelli, and Entolium orbicular e. The echinoid
Toxaster fittoni and the brachiopods Arenaciarcula fittoni and Trifidarcula trifida also occur.
Pieces of coniferous wood are common and the fuller’s earth itself yields microzoa (Davies
1916). Bundles of phosphatic tubes 18 inches or more in length traverse the limestone bands;

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 551

they belong to an organism of problematical affinities, Hallimondia fasciculata gen. et sp. nov.
Parahoplitids have been found in the brashy sandstone at the very bottom of the beds at
Limpsfield, presumably the source of specimens attributed to the top of the Hythe Beds by
Gossling and Hare (in Kirkaldy 1939, p. 392). Most other ammonites have been found loose,
but P. nutfieldensis itself was collected in situ in the limestone capping the main seam of fuller’s
earth in the Priory pit, together with Tropaeum subarcticwn, and in the limestone underlying
the same seam in the Copyhold pit.

The Bargate Beds around Godaiming and Guildford have long been of interest, both for
their indigenous fossils, which link them with the Lower Greensand of Upware and Faringdon,
and for their derivatives, which afford evidence of contemporaneous movement and erosion
of Jurassic strata along the edge of the London Platform. Chapman (1894) figured many
species of foraminifera and ostracoda and mentioned the occurrence of calcareous algae.
Exposures at Guildford provided the types of the corals Trochosmilia meyeri and Isastraea
morrisi (Duncan 1870) and the cirripede Cretiscalpellum aptiense, based on a complete capi-
tulum (Withers 1935). But the chief native forms are brachiopods, of which many species
have been described and figured by Meyer (1864fi), Davidson (1874), and Middlemiss (1959),
such as: Rhombothyris extensa, Platythyris comptonensis, Sellithyris sella var., Cyrtothyris
cantabridgiensis, C. seeleyi, Praelongitliyris praelongiforma, Terebratulina elongata, Gemmarcula
aurea, Trifidarcula trifida, Arenaciarcula fittoni, ‘ Ornithella' juddi, and " Rhynchonella ’ anti-
dichotoma. Lamellibranchs are represented by a few oysters and pectens, gastropods by a
solitary Pleurotomaria. Cephalopods include the large ammonites Parahoplites nutfieldensis,
P. maximus, and Tropaeum subarcticwn (Casey 1960n, p. 40, text-fig. 12), the nautiloid Anglo-
nautilus undulatus and the belemnite Neohibolites ewaldi. The echinoid Cidaris faringdonensis,
columnals of Isocrinus, and bones of the dinosaur Iguanodon mantelli have also been recorded.
The whole fauna was reviewed by Chatwin in 1929 (ibid., pp. 68-71).

Derived fossils found in the pebble-beds at the base of the formation contain a high pro-
portion of fish teeth and ammonites. Arkell (1939) found that 90 per cent, of the ammonites
belonged to the mariae Zone of the Oxford Clay and from their distribution inferred the
existence of an inter-Aptian fault underlying the Hog’s Back. This fault is now seen as part
of the movements that terminated the Lower Aptian phase of deposition in various parts of
Britain and is directly linked with the local disappearance of the martinioides Zone.

In general, the only signs of organisms in the Puttenham Beds are Chondrites- type borings.
Middlemiss, however, found an exposure on the west bank of the River Wey, half a mile
west of Headleywood Farm, north of Headley, Hampshire, where the loams contain fossili-
ferous ironstone lenticles (Knowles and Middlemiss 1958, p. 221). The fossils include echinoids
(Holaster, Catopygus ), Parahoplites cunningtoni sp. nov., and many other molluscs, the chief
being Limatula tombeckiana and a gastropod allied to Margarites ( Atira ) mirabilis. The ammo-
nite shows that the horizon is the same as that of the Iron Sands of Seend, i.e. the cunningtoni
Subzone of the nutfieldensis Zone.

Folkestone Beds. In this region the Folkestone Beds present their typical inland facies of loosely
coherent sand with pebbly and clayey seams and veins and doggers of ironstone (‘carstone’).
False-bedding is prevalent, especially in the upper beds, and the sands have been subjected to
varying degrees of iron-staining. As in West Kent, the sands in places have bedding and other
characteristics remarkably like sand-dunes (Gossling 1929). Thicknesses vary from about 180
feet in the east to a maximum of 260 feet in the Farnham area. Around Reigate Gossling (1929)
made out the following sequence:

4. Upper Pebbly Sands (50-60 ft.)

3. Clay Band
2. Silver Sands
1. Basal Pebbly Sands

(10 ft.)
(100-120 ft.)
(10-20 ft.)

552

PALAEONTOLOGY, VOLUME 3

To these must be added as a fifth and topmost unit the glauconitic loams with phosphatic
nodules which form a passage into the Gault. Throughout most of the region fossils are con-
fined to this unit.

The junction with the Gault may be seen in the Coney Hill (Priory) sandpit at Barrow Green,
Oxted (Wright and Wright 1948). The succession is similar to that in the western end of
Squerrye’s main pit, Westerham, little more than a mile to the east, and shows clean, white
current-bedded sands followed abruptly by a thick band of clay with the floridum nodules
above. Exposures around Merstham mentioned by Gossling are now obscured, but that of
Colley Lane sandpit, Buckland, west of Reigate (Gossling 1929, p. 249), has improved in
recent years. Here the succession is more like that of Ford Place, Wrotham, having the principal
concentration of nodules in the puzosianus Subzone, but with the raulinianus and floridum
Subzones also present and yielding their characteristic ammonites. Below the basal line of
small white nodules (floridum Subzone) are 18 inches of brown clayey sands (— bed 2 of Ford
Place) which form a sharp line of contact with the underlying current-bedded sands, the con-
trast between the two sets of beds being heightened when wet. A similar sequence was seen
during excavations for the Shere by-pass road.

Lenticles with the oysters Exogyra conica , E. tuberculifera, and Lopha diluviana were found
at the base of the Folkestone Beds at Abinger Hammer by Harper and Wilson (1938). No other
exposures of palaeontological interest are met with westwards until we reach the Farnham
area.

Palaeontological interest in the Folkestone Beds of the Farnham area dates from the dis-
covery by Shepherd in 1934 of ammonites and other fossils in the main mass of the sands,
long thought to be barren. Subsequently Wright and Wright (1942a) made further discoveries
of ammonites, both at Coxbridge, west of Farnham, the site of the original finds, and at High
Mill and Runfold, east of Farnham. Originally believed to be Aptian Parahoplites of the
nutfieidensis Zone, and then acanthohoplitids and desmoceratids of the jacobi Zone, these
ammonites are now known to be of basal Albian age (Casey 1954a). The fossils occur in a
band, here designated the farnhamensis Subzone, that forms the bottom of the tardefurcata
Zone.

The Coxbridge pit, just off the Alton road, shows about 30 feet of buff, current-bedded
sand with irregular masses of carstone. The sands also contain small scattered pebbles, a little
glauconite, and an appreciable amount of clayey matter, the last in thin seams or in large
buff-grey, slightly phosphatized concretions, which occur in a broad band about 10 feet thick
running through the middle of the section. The concretions are found loose in the sand or in
the centre of a carstone block and are the main source of fossils.

The standard of preservation is unusually high. Excepting specimens that occur near the
outside of a carstone block, shells retain the original test, faintly nacreous when freshly
extracted. It is a common occurrence for an ammonite to form the nucleus of a concretion,
and when this is split open the chambered whorls of the shell are found to be hollow. Oysters,
polyzoa, and other organisms invariably coat the ammonites, both on the outside and on the
inner walls of the body-chamber. The latter is also the repository of smaller shells, sponges,
pieces of drift-wood, and other sweepings of the sea-floor. The majority of the shells have the
appearance of being undamaged at the time of burial. The finds at Coxbridge include —
Lamellibranchs: Area dupiniana, Anomia pseudoradiata, Limatuia sabuiosa sp. nov., Acesta
longa, Oxytoma pectination, Gryphaeostrea canaiicuiata (attached to ammonites), Glycymeris
( Giycymerita ) umbonata, Thetironia minor, Modiolus aequaiis, Resatrix ( Dosiniopseiia ) sp.
Gastropods: Margarites ( Atira ) mirabiiis, Giobuiaria arduennensis. Cephalopods: Farnhamia
farnhamensis F. spp. nov., Hypacanthopiites spp. nov., Anadesmoceras sp. nov., Eutrephoceras
sp. nov.? Annelid : Serpuia cf. adnata (attached to ammonites). Echinoid : Phyilobrisus artesianus.
Polyzoan: Heteropora michelini. Porifera: Plocoscyphia cf. iabrosa, Stelletta cf. inciusa,

RAYMOND CASEY: STRATIG RAPHICAL PALAEONTOLOGY OF GREENSAND 553

Doryderma sp. Coniferous wood, bored by Martesia prisca, is plentiful and in an unusually
good state of preservation, microscopic sections showing the cells permeated by fungi. Farn-
hamia, a primitive genus of hoplitids reaching nearly 2 feet in diameter, is the special feature
of this district and horizon and has been found nowhere else. The long-ranging genus Hypa-
cantlioplites is represented by many new species, some equally large. Another noteworthy
feature of this fauna is the presence of recognizable sponges (as opposed to spicules) and the
absence of Exogyra latissima, Gervillella sublanceolata, Tortarctica similis, Cucullaea glabra,
and the other thick-shelled bivalves so common in the coarser, marginal deposits.

Their great thickness and the way they are bedded suggest that in general the Folkestone
Beds of this district were laid down rapidly and without any big breaks in the succession. It
must be assumed, however, that when the farnhamensis fauna existed there were minor pauses
in deposition, and that these pauses were long enough for empty shells to lie undisturbed on
the sea-bottom and gather an epifauna, though not long enough for complete phosphatization
and radioactive enrichment (the usual concomitant of inhibited deposition in the Lower Green-
sand) to manifest themselves.

At Coxbridge the Farnhamia band is about 25 feet below the Gault, the formational boun-
dary being visible in a small opening near the eastern end of the main pit. In the Jolly Farmer
and Princess Royal pits at Runfold, about 2\ miles north-east of Coxbridge, the band lies
near the bottom of a 50-feet face of sand, 200 yards south of the mapped boundary and at
least 80-90 feet below the Gault. About half a mile south-west of Coxbridge, in the Hyde-
Crete pit, off the Alton road, it is based by a pebble-bed, 6 inches thick, overlying a vivid
green seam and is followed by 40 feet of sand with no sign of the Gault. Exposures at Wreccles-
ham, six-tenths of a mile south of Coxbridge, show that the varying thicknesses of sand between
the Farnhamia band and the Gault may be attributed to a period of folding and erosion in
mid -tardefurcata times.

Pits showing the junction of the Gault and the Folkestone Beds have been worked inter-
mittently at the village of Wrecclesham (formerly Wracklesham) for over a century. They were
mentioned by Murchison in 1826 and were fully described by Paine and Wey in 1848. Drew
(in Topley 1875, p. 142) measured a section near Wrecclesham Church, which was quoted by
Jukes-Browne (1900, pp. 96-97) and Dines (in Dines and Edmunds 1929, p. 40). The section
given on p. 554 was demonstrated to the Geologists’ Association in July 1949 (Casey 1951o).

Bed 10 contains a condensed floridum-raulinianus fauna and besides the usual lamellibranchs
yields D. mammillatum, D. monile, D. orbignyi, Beudanticeras newtoni, B. dupinianum, Oto-
hoplites raulinianus, Cleoniceras floridum , C. sp. nov., Sonneratia parenti, S. sp. indet., Pra-
ia nisoceras raulinianum, P. acteon, and Hamites praegibbosus. From the main nodule-bed of
the regularis Subzone (bed 5) I have obtained : Pictetia depressa, Leymeriella regularis, L. tarde-
furcata, Id., var. intermedia, Id., var. densicostata, L. pseudoregularis, L. rudis, L. cf. renascens,
L. consueta, Id., var. magna, Cleoniceras sp. nov., Anadesmoceras strangulatum, A. subbaylei,
A. spp. nov. Many of these species also occur in the underlying sand (bed 4). Smooth shards,
each supporting a tangled mass of filaments (PI. 79, fig. 1), are found in bed 5 and are inter-
preted as phosphorite infillings of Exogyra shells that were infested with Graysonia. This pit
is the best collecting ground for the regularis ammonite fauna in England and the main source
of Anadesmoceras.

In 1955-6 the pit was extended westwards and the beds exposed for nearly 100 yards; the
relationship of the regularis sediments to the sands below were then seen with unprecedented
clarity. The basal bed of the regularis Subzone (bed 2) was found to wedge out to the west and
to the north and to be replaced by a line of pebbles and phosphatic nodules that passed with
angular discordance over the sands of bed 1. Bands of iron-staining, up to one foot thick,
mark the topsets of the current-bedded units of these sands and form a parallel series of
datum-lines across the pit-face. A particularly conspicuous band lies 8 feet below bed 2 at the

554

PALAEONTOLOGY, VOLUME 3

entrance to the pit and another 14 feet below that. In the western face, nearly 100 yards
distant, the base-line of the regularis Subzone oversteps the first band and is brought to within
4 feet of the one next below, the overstep taking effect most rapidly northwards. In the small
pit at Coxbridge mentioned above, just over half a mile north of Wrecclesham Church, the
regularis Subzone is condensed into a single band of pebbles and hard phosphatic nodules
with Anadesmoceras and a few Leymeriella. Two miles north-east of Wrecclesham, in an old
pit at the east end of the Farnham by-pass road, opposite High Mill, this band may still be

Gault-Folke stone Beds junction-beds about 50 yards south-west of Wrecclesham Church

ft. in.

dentatus Zone (pars)

13. Blue, somewhat sandy clay . . . . . . . .30

12. Mixture of blue clay and sand with scattered phosphatic nodules (Hoplites

sp.) ............ 2 6

mammillatum Zone (condensed)

11. As 12 but sandier and with fewer nodules . . . . . .16

10. Seam of whitish phosphatic nodules ....... 3-6

9. Grey clayey sand with thinly scattered phosphatic nodules ... 9

8. Buff sand ........... 1 6

7. Grey clayey sand .......... 1 0

6. Coarse buff sand .......... 9

tardefurcata Zone ( regularis Subzone)

5. Grey phosphatic nodules in an ill-graded matrix of coarse buff sand and

grey clay ........... 2-4

4. Coarse clayey sand with sparse phosphatic nodules . . .36

3. Impersistent seam of soft brick-red sand . . . 6 in. to 1 0

2. Coarse buff sand with occasional friable phosphatic nodules .40

tardefurcata Zone (? rnilletioides Subzone)

1. Coarse buff sand, streaky iron-staining, small-scale current-bedding,

some carstone near base . . . . 40 0

Total about 60 0

traced, though the pebble-studded nodules are more thinly scattered. Nodules with Anades-
moceras, Leymeriella, and Inoceramus coptensis may be picked out from under the gravel in
another old pit by the roadside at Wiley Mill, three-quarters of a mile west of Wrecclesham.
Evidently a regular is-mammillatum basin was centred just east of Wrecclesham and these
exposures show the first stages of wedging out of the regularis sediments towards the margins
of the basin.

The section set out on p. 555 was measured in 1947 in the old pit on the Farnham by-pass
just mentioned.

The various mammillatum beds are better separated than at Wrecclesham, though still
condensed. Bed 4 seems to be layers of sand and clay reworked; the clay is often in small
strips as though parts of a sheet broken up. The white nodules are like those in bed 10 at
Wrecclesham; the black-centred ones in the same bed were apparently derived from a clay
seam now sorted in with the sand.

South of the Farnham area the nodule-beds below the Gault have yielded only fossils of the
mammillatum Zone, as in the roadside pit at the cross-roads three-quarters of a mile south-
west of Kingsley Church and in the old disused sandpit 160 yards south-west of the cross-
roads at Blackmoor. A feature of these exposures is the increased pebbly content of the beds.
A mile and a half farther south a sandpit 200 yards west of Aldersnap Farm, Petersfield,
described by White (1910, p. 15), and Kirkaldy (1935, p. 531) shows the junction of the Folke-
stone Beds and Gault as a thin seam of glauconitic grit with pebbles and phosphatic nodules.

RAYMOND CASEY: STRAT1GRAPHICAL PALAEONTOLOGY OF GREENSAND 555

White (1910, p. 19) recorded a section, no longer visible, near Stroud Farm, a mile and a quarter
west of Petersfield, in which the junction-bed contained pebbles but no phosphatic nodules.
Between Farnham and Petersfield the main mass of the sands, frequently striped with clay
at the top, have yielded only driftwood and obscure traces of shells.

Attempts to trace the band with Farnhamia farnhamensis beyond the Farnham area have
been unsuccessful. I am of the opinion, however, that this band, with its argillaceous content
Gault-Folkestone Beds junction-beds, east end of Farnham by-pass

ft. in.

dentatus Zone (? bennetianus and dentatus-spathi Subzones)

7. Inky-blue clay with phosphatic nodules, passing up into grey clay with

Hoplites ........... seen 4 0

dentatus Zone ( eodentatus Subzone)

6. Ill-graded sand and clay with many blue-black-hearted phosphatic nodules.

Isohoplites, Douvilleiceras . . . . . . . . .36

mammillatum Zone ( puzosianus Subzone)

5. Yellow-grey sand with scattered small pebbles and black-hearted, slightly

pyritic, phosphatic nodules, getting clayey upwards. P. (H.) puzosianus . 3 0

mammillatum Zone (? raulinianus Subzone)

4. Ill-graded coarse sand and clay (weathering brown) with white- and

black-hearted phosphatic nodules. D. mammillatum , B. newtoni (abundant) 1 0

mammillatum Zone (Ifloridum Subzone)

3. Buff coarse sand, clayey towards top, with scattered small pebbles; phos-
phatic nodules dotted throughout and in a line at top. D. mammillatum, B.
newtoni ............ 5 0

tardefurcata Zone (regularis Subzone)

2. Band of small pebbles and pebble-studded hard phosphatic nodules.

Leymeriella . . . . . . . . . 6 in. to 9

tardefurcata Zone (? milletioides Subzone)

1. Buff coarse sand ......... seen 10 0

Total 27 3

and evidence of restricted deposition, has its sedimentary expression in the Clay Band of the
Reigate area. Heavily charged with glauconite and silt, this Clay Band marks an episode of
slow, tranquil deposition, in contrast to the periods of rapid accumulation of sandy detritus
indicated by the Silver Sands below and the Upper Pebbly Sands above. Its position in the
sequence, about two-thirds of the way up, supports this correlation. East of Reigate the Clay
Band is replaced by a slightly coarser facies, the Silt Band, which may be followed into the
Westerham district, on the county boundary. The reappearance of a conspicuous silt band
at the same general stratigraphical level at Aylesford, north of Maidstone, has been com-
mented on in the section on West Kent. Since this Clay-Silt Band runs parallel with the edge
of the London Platform, i.e. the old shore-line, it is unlikely to be diachronous. As a strati-
graphical aid to division of the Folkestone Beds the Clay-Silt Band has proved invaluable
in the country between Reigate and Westerham (Kirkaldy 1947tf). If I am correct in correlating
this Band with the farnhamensis Subzone of Farnham, it now acquires an added significance,
for its lower limit is the dividing-line between the Aptian and the Albian.

In the marginal areas of the basin, at Folkestone and Eastbourne, there was a definite break
in sedimentation at the beginning of the Albian during which deposits of Jacobi age were
reduced to a remanie.

Sussex

The Lower Greensand outcrop swings sharply round Petersfield and continues through
Sussex in an east-south-easterly direction towards Eastbourne. At the western end of the

556

PALAEONTOLOGY, VOLUME 3

region, around Midhurst, the formation is about 580 feet thick and has the same broad litho-
logical divisions as in Kent and Surrey. At Washington, about 17 miles south-east of Mid-
hurst, it is reduced to 350 feet; followed thence to the coast it becomes progressively thinner
and the lower divisions, the Atherheld Clay and the Hythe Beds, eventually lose their identity.
Finally, under cover of the alluvium around Eastbourne, the whole series passes into glauconitic
loams with phosphatic nodules at the base of the Gault. Much of the region is covered by
three Geological Survey Memoirs (Reid 1903; White 1924, 1926) and other contributions have
been made by Kirkaldy (1933h; 1935; 1937), Kirkaldy and Wooldridge (1938), Casey (1950),
and Humphries (1956). Less is known about the palaeontology of the Lower Greensand of
Sussex than of any other part of the Weald.

Atherfield Clay. Chocolate-brown, blue, and grey silty clays, perhaps reaching a thickness of
60 feet, have been described from the western part of the region. No doubt they have a similar
fauna to that found at Shottermill, just over the county boundary. Unfortunately, the exposures
at Harwoods Green, about a mile and a half north-west of Pulborough, seen by Fitton (1836,
p. 156) and from which Martin (1828) obtained a large fauna, including ammonites, no longer
exist. In the extreme east of Sussex, in the Cuckmere Brick Company’s pit by Berwick Rail-
way Station, a thin seam of small grey nodules and rolled fossils at the base of the Lower
Greensand has yielded Prodeshayesites sp. and Deshayesites forbesi (? = /). deshayesi, Kirkaldy
1937, p. 119), suggesting that it is a highly condensed remanie of the whole Atherfield Clay
Series.

Hythe Beds. About 220 feet of sandstone with chert beds at the top are present in the Petworth
area (Kirkaldy 1939, p. 393), the chert beds, with seams of spicular stone, descending to lower
and lower horizons eastwards. East of the Arun there is a change to a calcareous facies. At
Thakeham, 10 miles south-east of Petworth, the Hythe Beds are represented by a greatly
diminished thickness of rag and hassock, not easily separable from the Bargate Beds. Attenua-
tion of the beds, accompanied by progressive loss of ragstone, is evident as the formation is
followed still farther eastwards and east of Streat they appear to be overstepped by the Sand-
gate Beds (Kirkaldy 1937, p. 107).

Hinde (1885) thought that the beds of spicular stone and chert were derived from great
banks of siliceous sponges whose skeletons were broken up by currents before burial, and
Kirkaldy and Wooldridge (1938) explained the distribution of these beds around Midhurst
and Haslemere by picturing the sponge-banks spreading northwards during the deposition
of the Hythe Beds. Humphries (1956) believed that the chert was of primary origin and not
linked with occurrences of sponges. No ammonites have been found in the Hythe Beds of
this region, but from their conformable relations and lithological continuity with the Sandgate
(Bargate) Beds it may be inferred that the martinioides Zone is present in the topmost beds in
the western part of the region.

Sandgate Beds. Between Midhurst and Pulborough the Sandgate Beds are about 150 feet
thick and have a lower unit of calcareous stone (Bargate Beds) similar to that of west Surrey.
Above are ferruginous loams with sporadic fossiliferous concretions, followed by pale micaceous
sandstone (Pulborough Sand Rock) and dark silty clay (Marehill Clay). Eastwards the whole
series passes into glauconitic loams of rapidly diminishing thickness.

Easeboume Quarry, north-east of Midhurst, shows the lower parts of the Bargate Beds and
has yielded a small fauna comprising the annelid Rotularia concava, the echinoid Toxaster
fittoni, brachiopods (' Ornithella' juddi, Trifidarcula trifida), lamellibranchs ( Panopea mandibula,
Entolium orbiculare, Venilicardia sowerbyi), a gastropod (Pleurotomaria anstedi), and the
following cephalopods: Anglonautilus undulatus, Parahoplites nutfieldensis, P. maximus, P.
sp. nov., and Tropaeum subarcticum (Kirkaldy and Wooldridge 1938, p. 142; and A. H.

RAYMOND CASEY: STR ATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 557

Gunner Coll.). The Brydone Collection in the Sedgwick Museum contains P. cf. maximus
and P. sp. nov. from Upperton, near Petworth, and other parahoplitids from the Bargate
Beds of Pulborough are in the London museums. The latter include a distorted nucleus found
by Fitton (1836, p. 157) in ‘the lowest members of the sands’ near Pulborough and described
by Spath (1930a, p. 441) as a new species, P. sussexensis (GSM 46131).

Ironstone concretions in the overlying loams have long been known to yield fossils from
occurrences first described at Parham Park (Mantell 1822, p. 72; Martin 1828, p. 31). Other
localities are Park Lane, Pulborough, a lane 300 yards west of Muttons Farm, near Ashington,
June Lane, Midhurst, and the river cliff east of Habin Bridge, near Rogate (reached only by
wading or by boat). The fossils, which occur as clustered moulds, were listed by Kirkaldy
(1937, p. 118) and are revised herein. Characteristic species are — Lamellibranchia: Nuculana
scapha, Cucullaea cornueliana, Pterotrigonia mante/li, Cuneolus lanceolatus, Senis wharburtoni,
Chlamys robinaldina , Entolium orbiculare, Neithea syriaca, Gervillella sublanceolata, Ensiger-
villeia forbesiana, Oxytoma pectinatum, Resatrix parva, Thetironia minor, Nemocardium ( Pratu -
him) ibbetsoni, Venilieardia sowerbyi, Lucina cornueliana, Parmicorbula striatida, Panopea
gurgitis. Gastropoda: Ringine/la albensis, Dimorphosoma vectianum, Anclmra ( Perissoptera )
robinaldina, Tessarolax moreausiana, Eulima melanoides, Atresius fittoni, Mesalia ( Bathra -
spira ) neocomiensis, Gyrodes genti, Acmaea sp. Brachiopoda: Lamellirhynchia cf. caseyi,

‘ Rhynchonella ’ parvirostris. Echinoidea: Toxaster fittoni. The rudist Toucasia lonsdalii, the
cucullaeid Cryptochasma ovale gen. et sp. nov., and the gastropods Nerinea sp. nov. and Nerita
sp. nov., also occur, linking the deposit with the Iron Sands of Seend. Cephalopoda are
unknown in these loams, though their stratigraphical position and similarity to the beds with
ironstone concretions in Group XIV at Shanklin leaves little room for doubt that they belong
to the cunningtoni Subzone. Fresh samples from depth frequently show the loams themselves
to contain fossils, as at Ashington (Kirkaldy 1937, p. 116) and in the Hopton Wood borings,
described below. The succeeding Pulborough Sand Rock and Marehill Clay have yielded only
plant remains, including fronds of Weichselia reticulata. In the eastern half of the region the
Sandgate Beds are apparently devoid of organic content.

Folkestone Beds. Over most of the region the Folkestone Beds consist of white, yellow, or
reddish sands, strongly current-bedded, and with seams of pebbles and clay. They are involved
in the general easterly dwindling of the Lower Greensand of this region. Over 200 feet thick
at Petersfield, they are reduced to half that thickness at Washington, some 25 miles farther
east. Beyond Washington they are apt to change to an argillaceous and glauconitic facies and
near Eastbourne the whole division is assimilated into the basement-beds of the Gault. Change
of facies may take place with surprising rapidity, as in the Small Dole area, where the upper
part of the division passes into an argillaceous unit, here termed the Flopton Wood Clay.
Fossils have been found only in the glauconitic and clayey developments.

The southwards transition of the mammillatum Zone from its normal facies of glauconitic
loams with phosphatic nodules to a thin band of pebbles may be seen in the Petersfield area,
as mentioned above. Now cemented into an iron-grit, this band of pebbles may be followed for
nearly 20 miles through Sussex, forming a knife-sharp junction with the Gault. Such con-
tinuity is remarkable when it is considered how variable are these beds elsewhere. It would
seem that this part of the outcrop coincides with the rim of a regularis-mammillatum trough
and that the strike of the beds, ESE.-WNW., is the axis of a mid-tardefurcata fold. Immediately
the outcrop is deflected south of this axis the regularis-mammillatum sediments reappear. The
incoming of the normal mammillatum facies may be seen in two old sandpits on either side of
the Horsham road by West Winds Poultry Farm, a mile north of Steyning (Kirkaldy 1935,
pp. 526-7), which show wisps of dark glauconitic sandy clay and phosphatic nodules with
D. mammillatum and B. newtoni.

B 6612

o o

558

PALAEONTOLOGY, VOLUME 3

A big step forwards in the study of the Folkestone Beds of this region was made in 1956
when the British Portland Cement Manufacturers put down a series of exploratory borings
in the fields east of Hopton Wood, Small Dole, a mile and a half north-east of Upper Beeding.
Thanks to the good offices of Mr. Clements of the British Portland Cement Manufacturers
I was able to keep drilling operations under close observation and to examine the samples
fresh from the core-barrel. Since all the borings showed a similar succession, it will suffice
to give details of one, boring number 9a, situated on the eastern edge of Hopton Wood.

British Portland Cement Manufacturers' No. 9a boring, Hopton Wood, Small Dole

Depth in feet

Gault

dentatus Zone ( dentatus-spathi Subzone)

Brown-weathered clay with rotten fossils ...... 0-10

Dark grey, slightly brown-tinged, micaceous clay ( H . dentatus) . . 10-25

dentatus Zone ( benettianus Subzone)

Dark grey, slightly silty and micaceous clay with algal filaments, becoming
more silty and glauconitic downwards and passing into bed below ( Hoplites ,
Lyellieeras, Eubrancoceras, Beudanticeras, and numerous Protanisoceras
moreanum, all with iridescent test) ....... 25-40

Hard grey-green glauconitic sandy clay with pockets and channels of sand ;
algal filaments and a few dark phosphatic nodules .... 40-48

dentatus Zone ( eodentatus Subzone)

Hard dark-green glauconitic loam with rafts of clay and pockets and
channels of coarse sand; sandy phosphatic nodules and small pebbles;
pyritic nodules at top; hard pebbly band at 54 ft.— 54 ft. 6 in. ( Hoplites or
Isohoplites at 53 ft.) . . . . . . . . . . 48-56J

Folkestone Beds

? mammillatum Zone

Band of dark, gritty phosphatic nodules and small pebbles in glauconitic
loam ............ 56L-57

tardefurcata Zone ( regularis Subzone)

Hopton Wood Clay. Dark-grey, non-calcareous clay with hard, flat,
whitish nodules, especially at top, a few pyritic nodules and numerous
algal filaments; some threads of glauconitic sand; washed residue full of

glauconite and mica, a few forams. Aconeceras and Leymeriella with iri-
descent test; crustacean limbs fairly common ..... 57-69

? tardefurcata Zone ( milletioides Subzone)

Bright-green glauconitic sandy clay with phosphatic nodules, some gritty,
some not ............ 69-72

? jacobi Zone

Grey silt with threads of white sand, becoming clayey downwards and
passing into ........... 72-84

Bright-green glauconitic sandy clay with pink powdery traces of fossils,
passing into ........... 84-85

Clayey glauconitic sand, weathering white ...... 85-92J

Dark-green glauconitic loam with sandy pockets; line of small pebbles and
crushed fossils at base ......... 921-95i

Sandgate Beds

Grey-green sandy clay . . .

Bright-green glauconitic loam with rotted fossils, mostly Exogyra
Grey-green clayey sand and sandy clay, passing into
Bright-green glauconitic loam with traces of fossils

95J-106

106-115

115-136

136-137

A clay bed, 10 to 13 feet thick, was proved in all the borings below the presumed mammil-

RAYMOND CASEY: STRATIG R APHICA L PALAEONTOLOGY OF GREENSAND 559

latum nodule-bed; its existence was unexpected, there being no sign of it at the outcrop,
barely a quarter of a mile to the north. Dark grey in colour and with iridescent fossils, it
seemed indistinguishable from the Gault save in its non-calcareous property. Its most sur-
prising feature, however, is its fauna, the dominant fossil being Aconeceras sp. nov., with other
ammonites, Leymeriella cf. regularis and Anadesmoceras. Apart from two tiny scraps recovered

CHALVINGTON- FOLKESTONE

SELMESTON (EAST CLIFF )

FEET

6 0 r

40r

30 -

i°L

0L

H. den talus
SULPHUR BANO
I D.mammillatum
O.roulinianus
. C. floridum

S kitchini
D.mammillatum

L regularis

L . regularis

H. Jacobi t H ang/icus
H rubricosus

text-fig. 7. Comparative vertical sections of the basal Gault and the Folkestone Beds of the Chalving-
ton-Selmeston area of Sussex and of Folkestone, Kent (modified from Casey 1950).

from the nodule-beds at Leighton Buzzard and another from the Shenley Limestone of the
same locality, no aconeceratids have been seen before on this horizon. Moreover, the nearest
place where the regularis Subzone is known in a clay facies is Hanover, north Germany. The
geographical extent of this remarkable bed, here designated the Hopton Wood Clay, is un-
known. Hudson’s Red Sand Pit, Hassocks, 5b miles north-east of Small Dole, shows a return
to the iron-grit facies at the junction with the Gault. The presence of the Hopton Wood Clay
farther south is strongly suggested by a record of ‘brown clay, not effervescing with acid as
the rest of the Gault does, with hard white nodules (? phosphatic)’. This was found between
seams of greensand at the base of the Gault at a depth of 1,275 feet in a well at Warren Farm

560

PALAEONTOLOGY, VOLUME 3

Industrial School, 2\ miles north-north-east of Rottingdean (Edmunds 1928, p. 194). It may
be identified more positively at the surface near Willingdon, as mentioned below.

The whole of the Folkestone Beds in these borings at Small Dole had an unusually high
clay content and the boundary with the Sandgate Beds was not easy to fix. Vast quantities
of clayey sediment poured into the area during the deposition of the Gault; the borings proved
no less than 165 feet of dentatus Zone, with an unusually thick and fossiliferous benettianus
Subzone, not previously recorded in Sussex, though obvious enough in Hudson’s Red Sand
Pit, Hassocks, where white phosphatic Lyelliceras and Beudanticeras laevigatum occur about
7 feet above the iron-grit.

The next fossiliferous exposures of the Folkestone Beds are east of Lewes. A large sandpit
at Manor Farm, 300 yards north-west of Chalvington Church, formerly exposed the junction
with the Gault for a distance of nearly 200 yards and was described by Kirkaldy (1935, p. 520)
as follows :

Section formerly seen in sandpit at Manor Farm, Chalvington

ft. in.

'Gault'

8. Green glauconitic sandy clay with lenticles of reddish-brown clay 1 ft. to 1 6

7. Discontinuous layer of white phosphatic nodules and pieces of bored wood 0-3
6. Green glauconitic sandy clay . . . . . 1 ft. 6 in. to 3 0

5. Continuous layer of nodules and wood ...... 3-6

Folkestone Beds

4. White medium sands faintly speckled with grains of glauconite

1 ft. 6 in. to 3 0

3. Brown sand with scattered quartz pebbles up to a quarter of an inch in

diameter ............ 2 0

2. White sand as in bed 4. . . . . . . . .60

1 . Greyish, slightly clayey sand with a few soft nodules of ironstone . seen 2 0

Total 18 0

A small section was still visible in 1950 and furnished me with diagnostic fossils (Casey
1950). Those from bed 5 included the brachiopod Lamellirhynchia caseyi and several species of
Hypacanthoplites, including H.jacobi, H. clavatus, and H. elegans, showing that the bottom of
the ‘Gault’ here lies in the anglicus Subzone of the jacobi Zone and is on the same horizon as
the base of the Folkestone Beds at East Cliff, Folkestone. The upper bed of nodules (bed 7)
yielded an entirely different set of ammonites — Leymeriella tardefurcata, L. regularis, Son-
neratia aff. parent!, S. cf. sarasini, Anadesmoceras aff. baylei, D. aff. mammillatum, B. newtoni —
indicating a remanie of the regularis and kitchini Subzones comparable with that of Band II
of Leighton Buzzard. Since there must have been a long period of time during which the
anglicus nodules were scoured out and rolled on the sea-bed, we may infer a missing interval
at the base of the tardefurcata Zone. This means that the glauconitic bed 6 is probably of
milletioides age and equivalent to the greensand bed (bed 2) overlying the anglicus nodules at
Folkestone. The two ammonite horizons were traced in a number of old diggings around
Chalvington, Selmeston, and Berwick (Casey 1950), but a different succession was seen in the
roadside banks opposite Willingdon Mill, half a mile south of Polegate Railway Station.
Excavations on this site revealed the basal nodule-bed with Hypacanthoplites, followed up-
wards by olive-green glauconitic sandy clay and then a thin sandy development of the Hopton
Wood Clay, with phosphatic Leymeriella. Indications of the mammillatum Zone were found
still higher, not mixed with the tardefurcata fossils (Casey 1950, p. 280). The anglicus nodule-
bed abounds in fossil wood, as noted by Mantell (1822), p. 76) and this is almost certainly the
horizon of the pine cone Pinistrobus [ Zamia ] sussexiensis, obtained by Mantell (1843, p. 34)

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 561

from Selmeston, and of the coniferous wood Protopiceoxylon edwardsi, which Stopes (1915,
p. 81) described from Berwick Green. The Luccomb Chine plant-bed, in the Isle of Wight,
is only slightly earlier ( rubricosus Subzone).

It is doubtful if the underlying sands of the Folkestone Beds persist south-eastwards much
beyond Willingdon. They had disappeared already at Willingdon Laundry, 150 yards north of
Hampden Park Station, where a well proved only loams of Sandgate Beds type between the
Gault and the Weald Clay (Kirkaldy 1937, p. 107). Eight feet of yellow sand, presumably
representing the Folkestone Beds, are reported at Hydneye, 650 yards north-north-east of the
last locality, but the deep wells at Eastbourne have furnished no evidence of this division of
the Lower Greensand. Here it may be assumed that the process of assimilation of the Folke-
stone Beds by the Gault is complete (Casey 1950, p. 286).

The Western Outliers

For 50 or 60 miles west of the Weald proper the Lower Greensand is hidden beneath a cover
of younger strata and when it emerges again it is found only in patches, widely separated,
extending from near Aylesbury and Oxford, through Berkshire and Wiltshire, to the neigh-
bourhood of Shaftesbury. This marginal area of Lower Greensand is essentially the record of
two overspills from the Wealden Basin that carried the sea westwards over the folded and
faulted Jurassics. These two transgressions occurred in nutfieldensis and mammillatum Zone
times and it will be convenient to describe the beds under those headings.

Nutfieldensis Zone. Deposits of nutfieldensis age which are strewn along the forward edge of
the Gault in Berkshire and Wiltshire probably represent a number of separate marine embay-
ments that existed there during the Lower Greensand period and their present distribution is
not necessarily the result of subsequent denudation. Each of these groups of outliers has a
lithological and faunal character of its own and the manner in which they are entrenched into
the Jurassic and the nature of the faunas suggests that the shore-line did not lie far west of the
present outcrop.

The best known of these outliers is at Faringdon, Berkshire, described by Mantell (1839;
1844), Austen (1850), Sharpe (1854), Meyer (1864u), and Davey (1874). An excellent up-to-
date account of them is given by Arkell (1947u), who used the following stratigraphical
divisions :

The sponge gravel, where present, generally rests on Kimmeridge Clay, though in places it
overlaps on to Coral Rag. Pieces of Coral Rag occur in the gravel, together with many derived
Kimmeridge Clay fossils. The maximum thickness of the beds is between 160 and 190 feet.

The Sponge Gravels are current-bedded sand-and-gravel banks, composed largely of fossils,
chiefly calcareous sponges and polyzoa, resembling in appearance the Crag deposits of East
Anglia. Brachiopods and echinoids are also numerous, but except for oysters and large nauti-
loids in the Red Gravel, mollusca are comparatively rare. ‘The way in which the Sponge
Gravels are overlapped confirms the impression that they were accumulated in pre-existing
hollows, possibly excavated and probably at least scoured out by submarine currents. If the
valleys were of subaerial origin, soil, gravel, or detritus might be expected, but none seems
to exist. The abundance of derived nodules and fossils, especially in the pebble bed, suggests
that the higher ground of Kimmeridge Clay between the valleys was being reduced by marine

4. Sands with chert and ironstone
3. Sandy clays

2. Red gravel )

[pebble-bed at junction] The Sponge Gravels
1 . Yellow gravel J

562

PALAEONTOLOGY, VOLUME 3

planation while the Sponge Gravels were accumulating in submerged depressions a little way
off shore’ (Arkell 1947a, p. 160). Elliott (1947) examined 500 specimens of the brachiopod
Gemmarcula aurea and found that 84 per cent, were worn disconnected valves or mere pieces
and 16 per cent, were whole and undamaged. To account for this seemingly anomalous mixture
he pictured a turbulent marine environment with brachiopods and other sessile organisms
living in the crevices of the sponge-banks; some were buried where they grew, others lost
their anchorage and were pounded to bits by the currents. Later he abandoned this idea and
regarded the whole as a current-accumulation (Elliott 1956). This view accords better with the
absence of Exogyra latissima, Gervillella sublanceolata, Yaadia nodosa, and all the other heavy
molluscs built for life in current-swept waters but not easily moved after death. Arkell believed
that the sponge-banks were not of littoral origin but had accumulated in a clear, shallow,
neritic sea, and he compared their ecological assemblage with that of the Inferior Oolite
( parkinsoni Zone) of Shipton Gorge, Dorset. A much closer parallel is found in the ‘gompho-
lite ’ of Blangy, in the Ardennes of northern France, described long ago by Barrois (1878,
pp. 248-57). Here, similar sponge- and polyzoa-gravels of Aptian age occupy the bottom of a
channel in Silurian schists and quartzites.

Sowerby (1811) seems to have been the first to draw attention to the faunal peculiarities of
the Faringdon Sponge Gravels, though many of the fossils were first described by Sharpe
(1854), who thought the odd assemblage a Danian one. Meyer (1864 a) replaced it correctly
in the Lower Greensand. Hinde (1883) described and figured the sponges and Gregory (1899-
1909) and Canu and Bassler (1926) dealt with the polyzoa, of which there are no less than thirty
genera and sixty-two species. Almost as many foraminifera were listed by Davey (1905) and
Wright (1905). The following are characteristic fossils of the Sponge Gravels, found chiefly
in the Little Coxwell pit, which has contributed to almost every museum in the country:

Calcareous sponges: Raphidonema farringdonense , R. macropora, R. porcatum, R. contortion, R.
postulation , Corvnella foraminosa, Svnopella pulvinaria, Oculospongia dilatata, Elasmocoelia crassa, E.
mantelli, Barroisia anastomosans, B. irregularis, B. ciavata, Peronidella raniosa, P. giliieroni, P. pro-
lifera. Corals: Smilotrochus austeni, Astrocoenia sp. Echinoids: Hyposalenia wrighti, H. lardyi, H.
stelluiata, Tetragramma rotulare, ‘ C idaris' faringdonensis, ‘C’. coxweilensis, Plagiochasma faringdonense,
Goniopygus delphinensis. Brachiopods: Seliithyris coxweilensis, Cyrtothyris cyrta, C. uniplicata, C.
cantabridgiensis, Praelongitliyris praelongiforma, Gemmarcula aurea, Arenaciarcuia fittoni, 'Ornithella'
juddi, Cyclothyris latissima, ‘ Rhynchonella' depressa. Bifolium faringdonense. Polyzoa: Proboscina
crassa, P. coarctata, Berenicea faringdonensis , B. (Reptomultisparsa) tenella, Reptoclausa hagenowi,
Cellulipora spissa, Meliceritites cunningtoni, M. semiclausa, Siphodictyon gracile, S. irregulare, Petalo-
pora cunningtoni, Reptomulticava fungiformis, Ceriopora faringdonensis, C. collis, C. dimorphocella,
Heteropora ciavata, Diaperoecia orbifera, Laterocavea dutempleana. Multigalea canui, Zonatula
brydonei, Seminodicrescis nodosa, Stomatopora calypso, Tliolopora virgulosa, T. thomasi, Tretocycloecia
densa. Lamellibranchs : Lopha diluviana, Gryphaeostrea canaliculata, Exogyra conica. Cephalopods:
Eutreplioceras sublaevigatum, Anglonautilus undulatus.

The fauna also contains a species of Burgundia (R. F. Wise Coll., B.M.), the only known
occurrence of a stromatoporoid in the Lower Greensand, and the whole assemblage is dated
as the lower half of the nutfieldensis Zone ( subarcticum Subzone) by rare Parahoplites nut-
fieldensis and P. maximus in the Red Gravel (L. Treacher, C. W. Wright, and R. V. Melville
Coll.). Lamplugh’s (1903) correlation of these deposits with the upper part of the brunsvicensis-
beds of Speeton was based on belemnites misidentified as B. speetonensis, apparently worn
examples of Neohibolites ewaldi (see Swinnerton 1955, p. xxxiii).

Excavations for a reservoir on Faringdon Folly, to the east of the town, exposed fossiliferous
sands and pebbly sandstones, apparently belonging to the topmost of the four divisions of the
Faringdon Lower Greensand recognized by Meyer (bed 4 of Arkell 1947). The sandstone,
estimated to lie 50 feet or more above the Sponge Gravels, yielded to Mr. R. V. Melville and

RAYMOND CASEY: STRATIGRAPHIC AL PALAEONTOLOGY OF GREENSAND 563

others a small suite of fossils, including Parahoplites cf. nutfieldensis and an allied form known
also from Group XIV at Shanklin, the latter suggesting a position in the cunningtoni Subzone.
The lamellibranchs Ptychomya robinaldina and Isocyprina sedgxvicki and the gastropod
Bright onia turns gen. et sp. nov. also occurred; the last two are not known elsewhere in the
Southern Basin and link the locality with Potton and Upware.

South of Aylesbury, Buckinghamshire, near the northern border of the province, there are
a few patches of ferruginous sands of dubious relationship. Some are unquestionably of
Wealden age; others are probably vestiges of the Lower Greensand (Bishopstone Beds of
Davies 1899), but the only place where marine fossils have been recorded is a pit, long filled in,
south of the Bugle Inn, Hartwell. Here Morris (1867, p. 456) is said to have found Exogyra
latissima and other lamellibranchs, together with derived blocks of Wealden sandstone.

To the south-west of Oxford, on Boars Hill and round Culham and Clifton Hampden, the
Kimmeridge Clay is overlain by ferruginous pebbly sands, in parts much resembling the sands
seen in the Faringdon reservoir excavations. Pringle (1926, p. 100) recorded a small fauna of
molluscs and brachiopods, but no ammonites, from these sands (Toot Baldon Beds of Davies
1899). In the Summary of Progress of the Geological Survey for 1900, p. 120, cherty sandstone
with marine fossils is recorded as having been found east of Marsh Baldon by J. H. Blake.
The fossils were identified by E. T. Newton and included ‘ Ammonites sp. Three fragments,
one of which may be A. nutfieldensis, but is too imperfect for identification.’ The report then
goes on to say that ‘ Ammonites Desliayesi and Terebratula sella ’ had been found in a roadside
cutting at Toot Baldon by Professor Ramsey, Mr. Etheridge, and Mr. Hull during a tour of
inspection about the year 1860 or 1861. None of these fossils has survived, though their recorded
presence offers promise for future investigation.

Better evidence of the nutfieldensis Zone is afforded by the outliers around Caine and Devizes
in Wiltshire. Here the Lower Greensand oversteps the Portland Beds on to faulted Kimmeridge
Clay and Corallian. The most important outlier is at Seend, about 3| miles west of Devizes,
where the sands are so strongly impregnated with ferruginous matter that they were at one
time worked for iron-ore. No fossils have been forthcoming from Seend for many years, but
in the last century the sands produced a large fauna, now known largely by the Cunnington
and Davey Collections in the Geological Survey and Oxford University Museums. The fauna
is quite different from the sponge-polyzoa assemblage of Faringdon and is composed mainly
of molluscs and brachiopods. A short list of fossils was published by Cunnington (1850), but
the fauna was never studied systematically. Woods seems to have dealt with only a few of the
lamellibranchs and many of the common forms ( Myoconclia delta sp. nov., Cryptochasma
ovalis gen. et sp. nov., Pachythaerus tealli, Linearia cf. ole a) were either not mentioned or not
described in his monograph. The following list is also incomplete, there being many forms too
poor for determination:

Brachiopoda: Gemmarcula aurea, Arenaciarcula fittoni, Oblongarcula oblonga, ‘ Ornithella ’ juddi,
Sellithyris coxwellensis, Cyclothyris latissima, ‘ Rhynchonella' depressa. Polyzoa: Entalophora ramosis-
sima. Echinoidea: Cidaris sp. Lamellibranchia: Area dupiniana, Barbatia marullensis, Aptolinter
aptiensis, Cryptochasma ovale gen. et sp. nov., Nuculana scapha, Nucu/a meyeri, Septifer sublineatus,
Myoconcha delta sp. nov., Lopha diluviana, Panopea gurgitis, Senis wharburtoni, Linotrigonia ( Oisto -
trigonia) upwarensis, Sphaera corrugata, Venilicardia sowerbyi, Nemocardium ( Pratulum ) ibbetsoni,
Pachythaerus tealli sp. nov., Seendia saxoneti, Cardita upwarensis, Trapezicardita squamosa, Opis
( Trigonopis ) neocomiensis, Linearia cf. ole a, Protodonax minutissimus, Litliophaga spp., Chlamys
robinaldina, Neithea quinquecostata, N. atava, Acesta longa, Pseudolimea faringdonensis, Toucasia
lonsdalii. Gastropoda: Anchura ( Perissoptera ) robinaldina, Gyrodes genti, Conotomaria seendensis,
Nerinea sp. nov., Nerita sp. nov., Scurria calyptraeformis, S. depressa, Acmaea formosa, Loxotoma
neocomiense. Cephalopoda: Eutrephoceras sublaevigatum, Parahoplites nutfieldensis, P. cunningtoni
sp. nov., P. spp. nov.

Keeping (1883, pi. 51) mentions the occurrence of the fish Sphaerodus neocomiensis and

564 PALAEONTOLOGY, VOLUME 3

various undetermined reptilian bones. Cunnington’s ‘small corals’ are polyzoa. An interesting
feature of the gastropod community is the dominance of limpets ( Scurria , Acmaea, Loxotoma),
the species of Scurria being endemic. Several of the lamellibranchs are characteristic Upware
species (Opis neocomiensis, Trapezicardita squamosa, Cardita upwarensis) ; Pachythaerus tealli
is found also at Potton; Seendia saxoneti and Myoconcha delta are known nowhere else in the
Lower Greensand, and the genus Protodonax is a new record for Europe. Crypts of Lithophaga
of unusually large size extend into the limestones of the Kimmeridge Clay. The whole assem-
blage is slightly later than the Faringdon Sponge Gravels and is on the same horizon as the
Iron Sands of Pulborough, the Puttenham Beds of Surrey, and Group XIV of the Isle of
Wight.

Ferruginous Sands similar to those at Seend were once exposed at Stock Orchard, south of
Caine, and were found to contain a colony of the rudist lamellibranch Toucasia lonsdalii.

Mammillatum Zone. The remaining datable strata of Lower Greensand age in this region belong
to the mammillatum Zone. They consist of green and brown, glauconitic and ferruginous loams,
sometimes hardened into stone, and form the basement-beds of the Gault. Some of the sands
running parallel with the outcrop of the Gault, such as the strip of red sand near Uffington,
in the White Horse Vale, Berkshire, may be of earlier Albian age, but in the absence of fossils
the question must be left open.

As early as 1836 Fitton (1836, p. 258) recorded a mammillatum Zone ammonite (A. monile)
from Crockerton, in the Vale of Wardour, Wiltshire. The specimen is in the Geological Survey
Museum and is Douvilleiceras mammillatum. No exposures of this zone exist at Crockerton
today. Elsewhere in Wiltshire the mammillatum Zone has been found at Dinton, also in the
Vale of Wardour, and at Dilton Marsh, near the north-west border of the county (Casey 19556).

A well sunk north-east of the church at Dinton in 1890 gave the following (summarized)
section:

Gault clay

Gault basement beds

3.

2.

U-

Lower Greensand

Hard grey ferruginous sandy rock; fossils .
Reddish-brown sandstone with scattered pebbles, fossils,
and fragments of wood ......

Layer of small pebbles ......

ft. in.

5 8

2 6
6

This was recorded by Jukes-Browne in 1891 and a further description of the section, with
lists of fossils from the different beds, was published in 1900 (Jukes-Browne 1900, p. 228).
Both Jukes-Browne and Reid (1903, p. 32) thought that the ‘Gault’ basement beds were
younger than the mammillatum Zone, but re-examination of the fossils showed that they were
referable to the kitchini Subzone, species of Sonneratia, Inoceramus coptensis, and other
molluscs of early mammillatum age being included (Casey 19556).

A section close to the middle of the old working face of the Bremeridge pit, near Dilton
Marsh, north Wiltshire, was examined by Mr. G. A. Kellaway in 1943 and the following
description is summarized from his notes :

ft. in.

Gault clay (weathered) .......... 5 6

(2. Earthy and pebbly ironstone with septarian nodules

and scattered limonitic ooliths, passing into . .30

1. Sandy and ferruginous clays, dark bluish-grey, with
pebbles and lumps of day. A layer of re-sorted clay
with broken Ostrea delta at base . . . .26

Kimmeridge Clay . . . . . . . . . . .12 0

Westbury Ironstone ...........60

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 565

The sandy and ferruginous clay (bed 1) yielded a fauna very similar to that of Dinton,
including poorly preserved Sonneratia, and from the overlying bed (bed 2) Mr. Kellaway
obtained large arborescent polyzoa ( Ceriopora ) and portions of a gigantic species of Douvil-
leiceras, comparable with that found in the regularis Subzone of the Folkestone Beds of
Folkestone and in the kitchini Subzone at Westerham (Casey 19556, p. 233). It is concluded
that the ferruginous basement beds of the Gault at Dinton and near Dilton Marsh are the
correlatives of the Sonneratia kitchini band of Folkestone.

A considerable interval of time must have elapsed before the sea spread into the adjacent
area of north Dorset, for several distinct faunal assemblages, such as that of the floridum,
raulinianus, and puzosianus Subzones, existed between the period of deposition of the basal
mammillatum Zone and the basal dentatus Zone, to which is now referred the base of the
Albian at Okeford Fitzpaine, north Dorset. Here the Gault rests on Kimmeridge Clay and
contains Douvilleiceras inaequinodum and species of Hoplites (Newton 1897; Spath 1925a,
p. 75).

A few miles north-east of Okeford Fitzpaine a narrow band of glauconitic and ferruginous
loam intervenes between the Gault and the Kimmeridge Clay, extending in the direction of
Shaftesbury. These beds were first noted by Jukes-Browne (1891) and were subsequently
termed Bedchester Sands by White (1923, pp. 42-44). No fossils have been found in them and
they have been claimed variously as representatives of the Hythe Beds (White 1923) and of the
Sandgate Beds (Kirkaldy 1939, p. 402). Their lithology and the way in which they run parallel
with the Gault suggests that they are a continuation of the mammillatum Zone observed farther
north.

A pebble-band separating the Gault and Kimmeridge Clays at Culham, Oxfordshire, has
also been assigned to the mammillatum Zone (Treacher 1908, p. 549; Spath 1923c, p. 71;
Pringle 1926, p. 101; Arkell 1947a, p. 170); but the evidence for this is spurious. The fine
specimen of Douvilleiceras mammillatum figured by Spath from ‘Culham’ (Spath 1925a,
pi. 5, fig. 1) is, in fact, Fitton’s original ‘ Ammonites monile ’ from Crockerton, Wiltshire, now
preserved in the Geological Survey Museum (GSM Geol. Soc. Coll. 1713). The specimen of
‘D. mammillatum ’ which Pringle (1926, p. 102) said he had found in this bed, also in the
Geological Survey Museum, is a derived Kimmeridgian Pavlovia. I have not seen the example
of the zone ammonite found by White (Treacher 1908, p. 549), but its association with ‘ Ammo-
nites beudanti ... of very large size’ suggests the basal dentatus Zone, where Beudanticeras
laevigatum reaches a diameter of 6 or 8 inches. Another example of ‘ Douvilleiceras mammil-
latum’ from Culham, in the Cunnington Collection in the British Museum, is too small for
specific determination and may be of either mammillatum or dentatus age. Its mode of preserva-
tion, with parts of the nacreous shell attached, does not indicate the pebble-bed but the over-
lying clays, from which a rich fauna of early dentatus age has been obtained.

NORTHERN BASIN

The Lower Greensand deposits north of the London Ridge were laid down in a different
basin from the typical Lower Greensand of south-east England and the succession is relatively
thin and incomplete. In Lincolnshire and Norfolk the formation succeeds a marine facies of
the Neocomian, but when it extends southwards into Cambridgeshire, Bedfordshire, and
north-east Buckinghamshire and northwards to the fringe of Yorkshire it comes to rest on an
eroded surface of Jurassic age. In the southern part of the Basin deposition seems to have been
influenced by movements along axes of Charnian trend (Rastall 1919; 1925). The Northern
Basin is the sole source in the Lower Greensand of the bodei fauna, at the very bottom of the
Aptian, which occurs as a derived or remanie element, generally in a basal nodule-bed.

566

PALAEONTOLOGY, VOLUME 3
Cambridge-Bedford Province

North of Aylesbury there is an interval of 10 miles before the Lower Greensand reappears
in Bedfordshire, forming a thickness of 200 feet of predominantly yellow sands in the Woburn
and Leighton Buzzard districts. These deposits are known as the Woburn or Potton Sands
and they have long been famed as a source of derived Jurassic fossils and a large indigenous
fauna of brachiopods, lamellibranchs, and sponges, best known from the old ‘coprolite’
workings at Little Brickhill and Potton. A similar mixture of derived and native fossils was
found in the Lower Greensand of Upware, Cambridgeshire. Most of the fossils were obtained
as a by-product of ‘coprolite’ extraction, being picked out of the siftings by the workfolk.
The classic exposures disappeared with the decline in the home phosphate industry towards
the end of the last century, but we are fortunate in having contemporary accounts by Teall
(1875) and Keeping (1883) of Cambridge. Loss of the Potton and Upware exposures was
counterbalanced by the discovery of new fossiliferous horizons at the top of the sands around
Leighton Buzzard (Lamplugh and Walker 1903).

At Little Brickhill, 2\ miles east of Bletchley, Buckinghamshire, 30 feet of sand with scattered
phosphatic nodules was seen resting on Oxford Clay; the lower part consisted of greenish-
grey shelly sand, in places cemented into layers like Bargate Stone (Teall 1875, p. 43). Indi-
genous fossils were obtained only from the lower sands and comprised a few long-ranging
lamellibranchs, some of the calcareous sponges and echinoids found at Faringdon ( Rhaphi -
donema porcatum, Barroisia anastomosans, B. clavata, Peronidella ramosa, Hyposalenia wrighti,
Tetragramma rotulare) and above all brachiopoda. The following list bears out Keeping’s
(1883, p. 21) comment: ‘Brickhill was the metropolis of the Brachiopoda, in Cretaceous times.’
Rhombothyris extensa, R. microtrema, R. conica, Platythyris comptonensis, P. minor, Sellithyris
upwarensis, Cyrtothyris cyrta, C. uniplicata, C. cantabridgiensis, C. seeleyi, C. dallasi, Prae-
longithyris praelongiforma, P. lankesteri, ‘ Ornithella ’ juddi, ‘ 02 pseudojurensis, ‘O.’ tamarindus,
‘ O.' wanklyni, Zeilleria woodwardi, Kingena rhomboidalis, Gemmarcula aurea , Arenaciarcula
fittoni, Oblongarcuia oblonga, Trifidarcula trifida, Terebratella keepingi, T. davidsoni, Tere-
bratulina elongata, Cyclothryris latissima, LameUirhynchia cf. caseyi, ‘ Rhynchonella ’ upwarensis,
\R: cantabridgiensis, ‘ R.’ antidichotoma, ‘ R.' depressa. The hexactinellid sponge Plocoscyphia
pertusa, the echinoid Salenia hieroglyphica, and a few polyzoa were also found.

At Potton, near Sandy, Bedfordshire, indigenous fossils occurred in ferruginous layers and
included the gastropod Bathrotomaria ferruginea, the lamellibranchs Isocyprina sedgwicki,
Goniochasma dallasi, Pterotrigonia mantelli, Chlamys robinaldina,' Exogyra latissima, and
several species of the brachiopod Cyrtothyris. Derived material in the nodules near the base
comprised many Upper Jurassic fossils, blocks of Neocomian sandstone, bones of Iguanodon,
and the Lower Aptian ammonites Prodeshayesites fissicostatus and Australiceras gigas.

Cuttings for the London-Yorkshire Motorway half a mile north 45° east of All Saints
Church, Ridgmont, Bedfordshire, revealed coarse yellow sand resting on Ampthill Clay. At
the base of the sands was a band of pebbles and nodules, 18 inches thick, crowded with rolled
Pavlovia, Hartwellia, and other fossils derived from the Hartwell Clay. In the same bed,
probably having used the pebbles and nodules for anchorage, were indigenous brachiopoda,
including Platythyris comptonensis.

An important Aptian flora has been obtained from the Woburn Sands, described from drift-
wood and cones found mostly in the basal nodule-beds and in a band of fuller’s earth several
feet above (Carruthers 1866-70; Stopes 1912, 1915). At Potton and Sandy the nodule-bed has
yielded the cones Pinostrobus cylindroides, P. pottoniensis, Kaidacarpum minus, and Cycadeo-
strobus walkeri, coniferous wood Cedroxylon pottoniense, the ‘tree-fern’ Tempskya erosa, and
the cycadophyte Bennettites inclusus. Some of these have been regarded, without real evidence,
as Wealden derivatives. Woburn itself is the source of the conifers Pityoxylon woodwardi,
Cupressinoxylon hortii, Taxoxylon anglicum, Podocarpoxylon woburnense, P. bedfordense, and

RAYMOND CASEY: STR ATIG R A PH I C A L PALAEONTOLOGY OF GREENSAND 567

the angiosperms Woburnia porosa and Sabulia scottii. At Leighton Buzzard, probably from
the ‘Silver Sands’, were obtained the types of the cycadophytes Cycadeoidea yatesi (= Yatesia
morrisi ) and C. buzzardensis.

The fossiliferous beds at the junction of the Woburn Sands and the Gault near Leighton
Buzzard have been described by Lamplugh and Walker (1903), Kitchin and Pringle (1921;
1922h), Lamplugh (1922), Wright and Wright (1947), and Hancock ( 1958). The beds show rapid
lateral variation, as pointed out by Toombs (1935); of the three principal exposures now
available two show the familiar facies of glauconitic loams and phosphatic nodules, and the
third (Munday’s Hill) displays the lenticles of fossiliferous limestone (Shenley Limestone)
which first attracted attention to this locality. For many years a controversy raged between
Lamplugh and his Survey colleagues, Kitchin and Pringle, as to whether this limestone was
in place, the latter maintaining that it was of Cenomanian age and, together with the Gault,
had been turned upside-down by glacial action. Lamplugh’s straightforward reading of the
section now commands a more general acceptance than it did in his lifetime.1 In Arnold’s
pit (formerly Pratt’s pit), Billington Crossing, south-east of Leighton Buzzard, Wright and
Wright made out the following sequence:

Gault-Lower Greensand Junction- Beds at Arnold's Pit, Billington Crossing,

Leighton Buzzard

Grey clays of dentatus-spathi Subzone

Sandy-brownish clay with four bands of phosphatic nodules, as under:

Band IV (4 ft. 6 in. above base of clay). Scattered nodules less pebbly and
smoother than those below.

Band III (3 ft. 4 in. to 3 ft. 10 in. above base of clay). Sparse, irregular nodules,
perhaps lying in two beds.

Band II (2 ft. to 2 ft. 6 in. above base of clay). Abundant, irregular, round,
elongated or flattened nodules, blackish inside with a grey outer surface
studded with pebbles.

Band I (9 in. to 1 ft. above base of clay). Fairly smooth, dark-brown nodules
with pale-brown crusts.

Thin bed of indurated pebbly sand, with fragments of carstone, the whole some-
times phosphatized.

(Sharp junction)

Current-bedded ‘Silver Sands'.

ft-

5

0

Band I contains a regularis Subzone fauna with the ammonites Leymeriella regularis, L.
tardefurcata, L. rudis, L. consueta, Anadesmoceras sp. nov., the lamellibranchs Thetironia
minor, Cucullaea glabra, Pseudocardia tenuicosta , and the gastropods Claviscala Clementina ,
Gyrodes genti, Leptomaria billingtonensis, and many other small molluscs. Band II is a con-
densed deposit in which species of both the regularis Subzone and the kitchini Subzone of
the mammillatum Zone occur side by side; the former horizon is indicated by L. tardefurcata,
L. regularis, L. renascens, L. diabolus, L. consueta, L. pseudoregularis, Anadesmoceras baylei,

A. subbaylei, Aconeceras sp. nov., and Eogaudryceras shimizui; the latter by D. mammillatum ,

B. newtoni, and S. kitchini. Associated with these are many of the molluscs commonly found
on these horizons in the south : Inoceramus saiomoni, I. coptensis, Panopea gurgitis, Cucullaea
glabra, Thetironia minor, Resatrix ( Dosiniopseila ) vibrayeana, Pseudocardia tenuicosta, Entolium
orbiculare, Neithea quinquecostata, Gyrodes genti, Leptomaria gibbsi, Semisolarium monili-

1 For the sake of historical accuracy it must be pointed out that the debate did not end with the
discovery of Leymeriella in the limestone, for even Spath (1925d) was prepared to admit that it may
have been derived. It was the presence of Lower Gault ammonites in the lower part of the clays above
the limestone, showing the succession to be normal, that put the matter to rest (Spath 1930h, p. 271).

568

PALAEONTOLOGY, VOLUME 3

ferum, Tessarolax retusum, &c. Band III is another condensed horizon, containing elements of
the kitchini Subzone and the basal part of the dentatus Zone. Here are found D. mammillatum,
D. monile, B. newtoni, S. kitchini, S. perinflata, S. spp. nov., apparently mixed with Hoplites
and Isohoplites. Specimens of Protanisoceras acteon and Cleoniceras floridum picked up loose
suggest that there is also a sparse representation of the floridum Subzone in this band. Wright
and Wright believed that this band may contain two distinct concentrations of nodules, but it
has not been possible to sort out the faunas stratigraphically. Band IV is of early dentatus age.

The principal fact that has emerged from restudy of the Leighton Buzzard ammonites is
that only the lower part of the mammillatum Zone is present between the tardefurcata Zone and
the dentatus Zone. Although D. mammillatum occurs in both Bands II and III all the associated
ammonites of mammillatum age belong to the kitchini or floridum Subzones. The absence of
the raulinianus and puzosianus Subzones with their distinctive assemblages of Otohoplites,
Protohoplites, Hemisonneratia, and Pseudosonneratia is surprising; in the south the puzosianus
Subzone is a transgressive deposit and is the last to disappear on the crests of the regularis-
mammillatum troughs. The fact that the mammillatum Zone sequence is here out of phase with
that of the Southern Basin may have some tectonic meaning.

Chamberlain Barn pit, on the north side of the town, shows a similar succession to that of
Billington Crossing but exact correlation of the nodule-bands is difficult. The basal pebble-
bed is cemented into a conspicuous carstone breccia in which lumps of Shenley Limestone are
found occasionally, proving that the limestone was formed contemporaneously with the
breccia or before it. Lenticular masses of iron-cemented sand occur in the underlying ‘Silver
Sands’; they have yielded pieces of wood and, very rarely, lamellibranchs ( Acesta longa );
they have already been mentioned as the probable source of Cycadeoidea.

North of Chamberlain Barn, on the lower slopes of Shenley Hill, the phosphatic-nodule
facies of the regularis Subzone is replaced by the famous Shenley Limestone, at present exposed
only in the south-west corner of Munday’s Hill pit. The limestone occurs in lenticles up to
2 feet thick and several yards across and is remarkably varied both in lithology and in fossil
content. Sandy or pebbly, yellow or pink, crowded with brachiopods or lamellibranchs, each
lenticle has a character of its own. Laterally they are replaced by carstone breccia, the whole
resting on a guttered surface of phosphatized iron-grit. The fauna of the limestone is unique
and remarkable. Not only does it contain a rich and distinctive set of brachiopods, but its
assemblage of echinoids and Crustacea is also unmatched elsewhere. Hawkins (1921a; 1921Z?)
studied the echinoidea and was struck by the abundance of Pyrina; he thought that this prob-
ably indicated littoral conditions since Echinoneus, the modern representative of the family,
inhabits tidal flats. Apparently the brachiopods and other sessile benthos grew on the craggy
surface of the iron-grit, an environment too rough and shallow for ammonites, whose shells
are exceedingly rare in the limestone, though characteristic of the nodules a few hundred
yards away. Brachiopods are well preserved and constitute the bulk of individuals, the com-
monest being ‘ Terebratula ’ capillata, which in some lenticles may make up 90 per cent, of the
fossils (Hancock 1958, p. 39). The following list is by no means complete but includes the
commoner and more important forms:

Brachiopods: ‘ Terebratula ’ capillata , ‘ TV dutempleana , ‘7? gigantea, Rectithyris depressa, R.
shenleyensis, Terebratulina triangularis, Zeilleria convexiformis, Modestella sp. nov., Magas latestriata,
M. orthiformis, Terebrirostra arduennensis, Gemmarcula menardi, Id. var. pterygotos, Kingena lima ,
K. arenosa, K. newtoni, ‘ Rhynchonella' shenleyensis, ' Rr grasiana, ‘R.' lineolata, ‘ R .’ mirabilis, ' R.'
leightonensis, ‘RI dimidiata, ‘i?.’ antidichotoma. Lamellibranchs: Septifer sublineatus. Modiolus
reversus, Oxytoma pectinatum, Chlamys robinaldina, Neithea quinquecostata, Plagiostoma globosum,
Acesta longa, Limatula sabulosa sp. nov., Plicatula inflata. Gastropods: Claviscala Clementina, Con-
fusiscala dupiniana, Batlirotomaria leightonensis, Tectus cf. huoti, Eucycloscala mulled, Sipho gaultinus,
Neptunella cf. espaillaci. Ammonites: Leymeriella tardefurcata, L. regularis, Aconeceras sp. nov.
Echinoids: Pyrina desmoulinsii, Conulopyrina anomala, Hyposalenia studeri, Salenia rugosa, Toxaster

RAYMOND CASEY: STRATIG R APHIC AL PALAEONTOLOGY OF GREENSAND 569

murchisonianus, Nucleolites lacunosus, Catopygus columbarius, Holaster ( Labrotaxis ) cantianus.
Crinoids: Isocrinus fittoni, Torynocrinus sp. Crustaceans: Goniodromites scarabaeus, Cyphonotus
incertus, Diaulax carteriana, Enoploclytici tuberculata, Cretiscalpellum unguis, Pycnolepas rigida.

Scarcely less remarkable than the Shenley Limestone are the wedges of greensand seen by
Lamplugh (1922, pp. 10, 49) to be interposed locally between the carstone breccia and the
Gault. This greensand does not seem to have been exposed in recent years, though a set of
specimens are in the Geological Survey Museum. It is full of phosphatic fragments, small
pebbles, guards of the belemnite NeohiboJites minimus, and sharks’ teeth ( Lamna appendiculata,
Scapanorhynchus subulatus, S. raphiodon ? and Apateodusl). The oysters Ostrea vesiculosa and
Gryphaeostrea canaliculata, Serpula antiquata, cirripede valves, and a nautiloid were also
recorded by Lamplugh. The Red Clay immediately above the Shenley Limestone includes
lenticles crowded with columnals of Isocrinus and valves of the cirripedes Pycnolepas rigida
and Cretiscalpellum unguis ; its zonal position is unknown.

Since the statement that the Shenley Limestone is known only at Shenley Hill has been
repeated (Hancock 1958), it is necessary to draw attention again to the old pit a quarter of a
mile north of Long Crendon, Buckinghamshire, where Tumps of calcareous stone’ were
found between the base of the Gault and the Purbeck Beds. Both in its fauna and lithological
characters this stone is indistinguishable from the Shenley Limestone, as pointed out by
Lamplugh (1922, p. 41). Although the occurrence was discounted by Kitchin and Pringle
(1922 b), anyone who examines the specimens of this stone in the Geological Survey Museum
will surely admit that an outlier of Shenley Limestone exists north of Long Crendon. The
former extension of the regularis Subzone some 12 miles south-west of Leighton Buzzard is
thus indicated.

At Upware, near Cambridge, a few feet of Lower Greensand were found banked against
a folded ridge of Kimmeridge Clay and Corallian limestone (Keeping 1883, p. 4). Supposed
indigenous fossils occurred in two seams of nodules and pebbles at the base, mollusca mostly
in the lower seam, brachiopoda mostly in the upper. The fauna included a large series of
brachiopoda practically duplicating that of Brickhill, some calcareous sponges found also at
Faringdon ( Rliaphidonema porcatum, R. macropora, Barroisia anastomosans, B. clavata, & c.),
and polyzoa. Mollusca were much better represented than at Brickhill, Potton, or Faringdon:
the distinctive forms being the lamellibranchs Nucula meyeri, Barbatia marullensis, Eonavicula
carteroni, Cucullaea cornueliana, Cryptochasma ovale gen. et sp. nov., Glycymeris ( Glycymerita )
sublaevis, Trapezicardita squamosa, T. arcadiformis, Opis neocomiensis, Astarte cantabrigiensis,
Eriphyla upwarensis, Isocyprina sedgwicki, and the gastropods Pleurotomaria campichei, Brigli-
tonia turris gen. et sp. nov., Gymnocerithium tumidum, Tessarolax gardneri, Tridactylus walked,
Nododelphinula reedi, Eucyclus upwarensis, Ooliticia cantabrigensis, and O. varicosa. The
ammonites, now in the Sedgwick Museum, are redetermined as follows: Colombiceras sp.
nov. cf. tobleri (? nutfieldensis Zone), ? Cheloniceras crassum var. nov., Tropaeum keepingi
(? bowerbanki Zone), Deshayesites multicostatus, D. cf. consobrinoides, Toxoceratoides cf.
royerianus ( deshay esi Zone, parinodum Subzone), Deshayesites cf . forbesi (Iforbesi Zone),
Prodeshayesites fissicostatus, P. spp. nov. (fissicostatus Zone, bodei Subzone).

In Spath’s zonal table (19236, p. 148) the Upware deposit is shown as spanning the 'con-
sobrinoides'’ and ‘ hillsi ’ Subzones (= desliayesi Zone of the present classification) and as
equivalent to the Hythe Beds of East Kent. In fact there is a much wider zonal representation,
as indicated above. The vast majority of the ammonites are of Lower Aptian age; the only
exception is a single example of Colombiceras, which certainly belongs to the Upper Aptian,
probably the nutfieldensis Zone. Both this and the unique Tropaeum keepingi have portions
of the shell preserved in calcite and have a different aspect from the rest of the assemblage.
Whether any of them are truly indigenous is difficult to say. At all events, the Lower Greensand
of Upware, as shown by its ammonites, is an epitomized version of the greater part of the

570

PALAEONTOLOGY, VOLUME 3

Aptian stage. Some of the other mollusca ( Crypt ochasma , Isocyprina, Opis, Bright onia ) are
known elsewhere in the Lower Greensand only in the cunningtoni Subzone.

Beyond Upware and Ely the Lower Greensand is lost under the Fens. Borings show it
dwindling almost to vanishing point a few miles north of Ely.

Line olnshir e-Norfolk Province

In Lincolnshire the Lower Greensand crops out in a narrow strip running parallel to the
Chalk south-eastwards from the Humber to the southern end of the Wolds. It consists of a
few feet of ferruginous sand and grit (Carstone) overlying the Neocomian and is in turn over-
lapped by the Red Chalk, a facies of the Gault peculiar to the eastern border of the Northern
Basin. Its underground extension is proved by borings at Skegness (Woodward and others
1904, pp. 155-9). North of the Humber, in east Yorkshire, it oversteps the Neocomian and
becomes a discontinuous deposit of variable thickness, apparently filling shallow depressions
in an erosion surface. It may be seen, only a few inches thick, at Goodmanham, about a mile
north-east of Market Weighton, resting on Lower Lias (Boer, Neale, and Penny 1958, p. 178).
Strahan (1886) studied the Lincolnshire Carstone and showed that earlier views on its uncon-
formable relations with the Red Chalk (Judd 1867; 1870) were incorrect, the junction being in
fact gradational. Although there is no doubt that in the north there is a plane of erosion at the
base of the Carstone, we owe to Swinnerton (1935) the discovery that at the southern end of
the Lincolnshire Wolds the succession is more complete. In this area the Carstone is underlain
by a few feet of grey and yellow marls (Sutterby Marl), first detected in borings at Alford and
Maltby-le-Marsh and subsequently proved at outcrop east of the hamlet of Sutterby, about
2\ miles north-west of Fordington.

A rich cephalopod fauna was obtained from the Sutterby Marl of Sutterby, including
ammonites identified as Deshayesites fissicostatus, D. aff. laeviusculus, D. multicostatus,
Aconeceras nisoides, A. sp., Cheloniceras, and Tonohamites? and numerous belemnites. While
the belemnites occurred throughout the whole thickness of the marls, the ammonites were
almost entirely restricted to a phosphatic nodule layer near the base. On the basis of the
ammonite determinations the phosphate layer was assigned to the bodei Subzone, the base of
the Aptian as here understood (Swinnerton 1935, pp. 24-25). Elsewhere (Swinnerton 1937,
p. xxix) the Sutterby Marl has been equated with the bodei Subzone.

Through the kindness of Professor H. H. Swinnerton I have been able to examine his
collection of Sutterby Marl ammonites. They are determined as follows: (1) phosphatic nodule
band, Aconeceras nisoides, Sanmartinoceras (The ganec eras) cf. falcatum, Prodeshayesites
fissicostatus, P. aff. bodei, P. laeviusculus ?, P. spp. nov., Deshayesites cf. deshay esi, D. multico-
status, Dufrenoyia fur cat a, D. transitoria, (2) crushed in the marl, Colombiceras sp., Tropaeuml sp.

The ammonites in the phosphatic nodule bed belong to three different zones of the Lower
Aptian; the species of Prodeshayesites, and possibly the aconeceratids, represent the bodei
Subzone of the fissicostatus Zone, at the bottom of the Aptian; the Deshayesites belong to the
lower half of the deshayesi Zone ( parinodum Subzone); while the species of Dufrenoyia indicate
the bowerbanki Zone, the top zone of the Lower Aptian. The last may include only the Subzone
of Dufrenoyia transitoria (— Deshayesites aff. laeviusculus of Swinnerton). The ammonites
crushed in the marls, presumably part of the indigenous fauna, include only one form that is
generically determinable. This is a species of Colombiceras, a genus diagnostic of the Upper
Aptian, and its occurrence is of great interest since the only other British Colombiceras known
is the Upware specimen mentioned on an earlier page. There is, in fact, a very close agreement
between the ammonite horizons of the Sutterby Marl and those of the Lower Greensand of
Upware.

It is now apparent that the phosphatic nodule bed at the base of the Sutterby Marl is a
highly condensed remanie and that the Sutterby Marl itself is of Upper Aptian age. This

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 571

explains some anomalies in the belemnite fauna. Discussing the occurrence of Neohibolites
ewaldi, the dominant belemnite in the Sutterby Marl, Swinnerton observed that at Speeton
and in north Germany the species characterized strata above the bodei Subzone, only rare
examples of N. cf. ewaldi having been recorded by Stolley from the bodei Subzone itself.
From the association of N. ewaldi with ammonites of this subzone in the Sutterby Marl,
Swinnerton concluded that faunal failure was probably responsible for the absence of this
belemnite from the bodei Subzone of the Speeton and north German successions. Now that
the bodei ammonites in the Sutterby Marl are known to be derived the discrepancy in the
distribution of the belemnites disappears.

The plane of erosion which in the country to the north lies at the base of the Carstone is
here present at the base of the Sutterby Marl. Rolled fragments of Prodeshayesites have been
found in the Carstone of the north and central Wolds, but the only other locality in the Pro-
vince that has yielded Aptian ammonites in numbers is Hunstanton, Norfolk.

The existence of a phosphatic nodule-bed full of ammonites at the base of the Carstone
exposed on the foreshore at Hunstanton, Norfolk, has long been known and lists of fossils
from this bed have been given by Wiltshire (1869) and Keeping (1883). Keeping was of the
opinion that these fossils were derived, but this was denied by Lamplugh (1899, p. 142), who
considered that the fauna was on its proper horizon. Spath (1930a, p. 422) thought that it
contained ammonites of several horizons ranging from the bodei Subzone to the beginning of
the Upper Aptian. Examination of all available collections of Hunstanton Carstone ammonites,
supplemented by my own field work, discloses that the assemblage is composed of two faunas
only, both Lower Aptian in age. The faunal list is as follows: bowerbanki Zone ( transitoria
Subzone), Tropaeum bowerbanki. Id., var. densistriatum, T. drewi, T. sp. indet., Australieeras
gigas, Tonohamites (?) sp. nov., Cheloniceras ( Ch .) cornuelianum, Ch. ( Ch .) crassum. Id., var.
//or., Ch. (Ch.) spp. nov., Dufrenoyia furcata, D. truncata, D. transitoria sp. nov., D. sp. nov.
fissicostatus Zone (bodei Subzone), Ancyloceras cf. varians, Prodeshayesites fissicostatus, P.
bodei, P. laeviusculus, P. spp. nov.

Species of Prodeshayesites, chiefly P. fissicostatus (‘A. deshayesi ’ of early authors), make up
about 90 per cent, of the fauna. Fossils other than ammonites are rare, though the lamelli-
branch Mulletia mullet i has been found. The ‘peculiar dark grit’ mentioned by Keeping from
Hunstanton and which Kirkaldy (1939, p. 408) describes as occurring as derived blocks in
the base of the Carstone are nodules from the underlying Snettisham Clay and contain
Barremian ammonites of the genus Paracrioceras (‘ Hamit es or Ancyloceras, small species
with a double row of spines along the back’, Keeping 1883, p. 33).

The only other ammonitiferous deposits of Lower Greensand age in Norfolk occur at West
Dereham, between Stoke Ferry and Downham, at the southern extremity of the outcrop.
Here old phosphate workings at the junction of the Gault produced a large fauna in the last
century. The beds were fully described by Teall (1875) and Whitaker, Skertchley, and Jukes-
Browne (1893). Douvilleiceras mammillatum, Sonneratia kitchini, Hamites sp. nov., and other
forms characteristic of the kitchini Subzone of the mammillatum Zone were found in the
phosphatic nodules. The representation of the kitchini Subzone to the exclusion of the higher
parts of the mammillatum Zone compares with the Leighton Buzzard sequence.

Neither in Lincolnshire nor in Norfolk has the Carstone yielded indigenous ammonites,
though the way in which it passes up into the Red Chalk suggests that it is of Albian age.
I agree with Versey and Carter that it is probably represented at Speeton by the greensand
seam with phosphatic nodules and Leymeriella (Bed A 4) (Versey and Carter 1926).

572

PALAEONTOLOGY, VOLUME 3
PALAEONTOLOGY

PLANTAE

The Lower Greensand flora, first described as a unit by Stopes (1915), contains one
of the world’s earliest assemblage of angiosperms or flowering plants. In a public broad-
cast entitled ’The mystery of flowering plants’ (reproduced in The Listener) Professor
T. M. Harris called in question the authenticity of the angiosperms as Lower Green-
sand fossils (Harris 1956), mentioning specifically those from the Woburn Sands of
Bedfordshire. Harris’s scepticism is justified in so far as these plant-species are all based
on old museum specimens, some inadequately labelled, but whatever problems their
presence in the Lower Greensand poses to the palaeobotanists, the assumption that they
are spurious raises questions even more difficult to answer. The following are my notes
on the type specimens of the woods in question. With the exception of that of Hythia
elgari, all are in the British Museum (Natural History).

Aptiana radiata Stopes. Labelled ‘Lower Greensand, ?Luccomb Chine’. Matrix coarse glauconitic
sand with bits of whitish phosphate, absolutely typical of the Luccomb Chine plant-bed near the base
of the Sandrock. Parts of the ‘Sulphur Band’, at the top of the Lower Greensand at Folkestone,
produce a similar lithology. Both horizons are replete with fossil wood.

Cantia cirborescens Stopes. Sand from one of the boreholes in the specimen agrees with that of the
Folkestone Beds of Ightham, Kent, the stated provenance of the specimen.

Hythia elgari Stopes. The type block, in the Maidstone Museum, was examined by me some years ago
before I was aware of any doubts about its authenticity. From its matrix I had then accepted it as being
correctly labelled as from the Hythe Beds of Maidstone. This is now reaffirmed from examination of a
section of the specimen in the British Museum.

Woburnia porosa Stopes and Sabulia scotti Stopes. Both said to be from the Lower Greensand of
Woburn. No matrix or adherent grains to check.

These observations offer no support for the assumption that the specimens did not
originate in the Lower Greensand. Moreover, it is difficult to believe that two museums
could both have obtained undescribed fossil angiosperm wood in various Lower
Greensand-type matrices and have made the same blunder in labelling them.

Phylum COELENTERATA
Class HYDROZOA
Family milleporidae
Genus lonsda de Laubenfels 1955
Lonsda contortuplicata (Lonsdale)

Small calcareous growths with microscopic spongiform surface were described from
the Lower Greensand of Atherfield by Lonsdale (1849, pp. 55-66, pi. 4, figs. 1-4) as a
new genus and species of sponge (‘ Amorphozoa’), Conis contortuplicata. The generic
name Conis having been used previously by Brandt in 1835, de Laubenfels (1955,
pp. E86, 94) replaced it by Lonsda and treated the organism as a hyalosponge of uncer-
tain affinities. Sections cut from one of Lonsdale’s syntypes (GSM Geol. Soc. Coll.
1968) were examined by Dr. Kenneth Oakley, who reported (in lift. 16.1.48) that the
organism is not a sponge but a hydrozoan similar to Millepora lobata Roemer from the

RAYMOND CASEY: STR ATIGRAPHIC AL PALAEONTOLOGY OF GREENSAND 573

Neocomian of north Germany. According to Boschma (1956, p. F94) modern species
of Millepora are found commonly on coral reefs, generally at depths not exceeding
30 metres, which seems to be correlated with dependence of the colonies on symbiotic
unicellular algae that need light for their processes of assimilation. Lonsda contortu-
plicata occurs in the Upper Perna Bed at Atherfield and Sandown, where the coral
Holocystis elegans is also found in great numbers.

Phylum polyzoa
Class GYMNOLAEMATA
Family ascodictyidae
Genus graysonia Stephenson 1952

In Graysonia the zoarium is represented by a compound system of tubular stolons
and vesicles embedded in the shells of marine molluscs, comparable with that of the
Palaeozoic BascomeUa. The genus is monotypic, the type species being Graysonia
bergquisti Stephenson of the Cenomanian of Texas. A similar organism is found in the
shells of Exogyra in the Folkestone Beds, though there seems to be no previous descrip-
tion of it in British literature.

Graysonia anglica sp. nov.

Plate 79, figs. 1, 2

Holotype. GSM 98600, Folkestone Beds, regnlaris Subzone (bed 5), Wrecclesham, Surrey (Author’s
Coll.).

Diagnosis. Similar in size and form to G. bergquisti (Stephenson 1952, p. 53, pi. 9, figs.
2-6, pi. 10, figs. 27, 28), but stolons less arched and more frequently branching.

Plate 79, fig. 2 illustrates the typical mode of occurrence of the stolon-system of
Graysonia anglica in Exogyra. The holotype is part of the phosphatic infilling of an
Exogyra shell that was subsequently dissolved away, leaving Graysonia in relief. This
specimen shows a meshwork of stolons, a few widely scattered vesicles, and part of
another organ, possibly the zoecium. The last is a thin- walled tube or sac, rising above
the stolon-bearing surface, about 6 mm. in diameter and with an incomplete length of
12 mm. It is roughly elliptical in cross-section, bent about the middle, and presents an
irregular blistered surface covered with microscopic parallel striations. Stolons com-
municate with it at the base. Zoecia have not been described in this primarily Palaeozoic
family and I know of no other structure with which it could be compared.

Phylum BRACHIOPODA
Class ARTICULATA
Family zeilleriidae
Genus modestella E. Owen nov.

Type species. Modestella rnodesta gen. et sp. nov., Lower Albian, southern England.

Diagnosis. Small biconvex zeilleriids of terebratuloid aspect. Anterior commissure
rectimarginate, ligate, or strangulate. Test thin, finely punctate. A shallow median
sulcus between faint ridges in each valve. Beak suberect; beak-ridges sharp; foramen

pp

B 6612

574

PALAEONTOLOGY, VOLUME 3

large, mesothyrid; deltidial plates conjunct, concave. Hinge-plates fused; hinge teeth
wedge-shaped, supported by strong convergent dental lamellae. No cardinal process.
Septalium acute, angular, forming a V-shaped hinge trough which is supported by a
strong brachial septum extending two-thirds the valve-length. Crural bases and zeilleri-
form brachial loop given off dorsally.

Modestella modesta E. Owen gen. et sp. nov.

Plate 83, fig. 6 a-c

1874 Terebratula moutoniana Price, p. 140 {non T. moutoniana d'Orbigny).

Holotype. GSM Zk 4733, Folkestone Beds, main mammillatum bed, Copt Point, Folkestone, Kent
(Author’s Coll.).

Diagnosis. Modestella about 12 mm. long, 10 mm. wide, and 7 mm. thick. Outline of
pentagonal tendency; anterior commissure strangulate; interarea broad, slightly con-
cave; foramen subcircular; hinge-margin subterebratulid; growth-lines prominent.

The name Modestella is proposed for a group of small zeilleriids found in the Lower
Albian of southern England. The group is under investigation by Mr. E. Owen, from
whose notes I have been permitted to take the above diagnoses. Clusters of M. modesta
occur in the matrix of the main mammillatum bed nodules at Copt Point and may repre-
sent the fossilization more or less in situ of colonies that grew on the nodules. Isolated
internal moulds are found sporadically throughout the mammillatum Zone of Kent
and an allied species occurs in the Shenley Limestone.

Family terebratellidae
Genus terebrirostra d’Orbigny 1847
Terebrirostra arduennensis d’Orbigny

This brachiopod was described by d’Orbigny (1847, pi. 519, figs. 6-10) from the
Albian of Grandpre (Ardennes) and was listed by Barrois (1878, p. 275) as a fossil of
the Ardennes ‘ mammillatum Zone’, which I have shown (Casey 1957) to include not
only the restricted mammillatum Zone of southern England but also the underlying
regularis Subzone of the tardefurcata Zone. Middlemiss (1959, p. 140) quotes it as a
Lower Aptian form. Corroy (1925, p. 295) described it as an Albian form which Peron
had recorded from the Upper Aptian of Grandpre. Admittedly, the fauna listed by
Corroy has been regarded as Aptian since the time of Barrois, but for many years now

EXPLANATION OF PLATE 79

Figures natural size unless otherwise stated.

Figs. 1-2. G ray sonia anglica sp. nov., Folkestone Beds ( regularis Subzone). 1, Holotype, phosphatized
infilling of Exogyra shell showing stolons, vesicles and a possible zooecium (top right corner),
bed 5, Wrecclesham, Surrey. (GSM 98600.) 2, Stolons in situ in Exogyra shell, East Cliff, Folkestone,
Kent. (GSM Zm 24.) Both author's coll., x 3.

Fig. 3. Hailimondia fasciculata gen. et sp. nov., holotype, Sandgate Beds, Copyhold Pit, Redhill,
Surrey. (GSM Zk 3960-1 ; A. G. Davies coll.)

Fig. 4. Limopsis dolomitica sp. nov., holotype, base of Sandgate Beds, shore at Mill Point, Folkestone,
Kent. (GSM Zm 2137; author’s coll.) x2.

Figs. 5a, b. Anthonya woodsi sp. nov., side (a) and dorsal (b) views of holotype, Atherfhd 1 Clay Series
(Crackers), Atherfield, Isle of Wight. (GSM 98592; author’s coll.)

Palaeontology, Vol. 3

PLATE 79

CASEY, Lower Greensand fossils

RAYMOND CASEY: STR ATIG R APHICAL PALAEONTOLOGY OF GREENSAND 575

this age determination has rested on an ammonite described by Jacob (1905, p. 411,
pi. 13, fig. 3) as Parahoplites milletianus var. peroni. Thanks to the good offices of Dr.
P. Destombes, I have been permitted to examine a suite of ammonites from the type
horizon of Jacob’s form (Minerai de Bois-des-Loges). They are all Hypacanthoplites
of the tardefurcata Zone and include a specimen of H. cf. miUetioides sp. nov. (PI. 83,
figs. 1 a-b). It would appear, therefore, that in its type locality T. arduennensis is known
definitely to occur only in the tardefurcata Zone but possibly may range into the mam-
mil/atum Zone. Judging from specimens of T. arduennensis in the Paris museums, this
is the same species which Lamplugh and Walker (1903) described as Terebrirostra lyra
var. incurvirostrum and which is found at the base of the miUetioides Subzone at Newing-
ton, near Folkestone, and in the regularis Subzone (Shenley Limestone) at Leighton
Buzzard.

Phylum mollusca
Class LAMELLIBRANCHIA
Family parallelodontidae
Genus aptolinter nov.

Type species. Area aptiensis Pictet and Campiche 1866, Lower Aptian, Europe.

Diagnosis. Like Nanonavis Stewart but longer, less angular, with relatively subdued
umbonal region and more delicate radial sculpture on mid-shell. Hinge slender, anterior
teeth short (text-fig. \\a).

Following Woods (1899, p. 35), the group of Lower Cretaceous species centred on
Area aptiensis has been referred to Barbatia. Moulds of the hinge of A. aptiensis are
preserved in examples from the Perna Bed of Earlswood Common, Surrey (GSM
Zb 3393-401), showing that the species is not an arcid but a parallelodontid close to
Nanonavis. Other species of Aptolinter are: Area raulini , A. neocomiensis d’Orbigny,
and A. cymodoce Coquand. Gilbertwhitea Crickmay, 1930, is an allied genus with the
shape of Eonavieula carteroni (d’Orbigny).

Family cucullaeidae
Genus cucullaea Lamarck

Cueullaea tealli nom. nov. (= Peetunculus obliquus Keeping 1883, p. 116, pi. 6, fig. 1,
non Defrance 1826, nee Andrzejovski 1832, nee Lea 1833, nee Munster 1835, nee Reeve
1843, nee Brown 1845).

Genus noramya nov.

Type species. Area forbesi Pictet and Campiche 1866, Lower Aptian, south-east England.

Diagnosis. Subtrapezoidal or subtrigonal, thick-shelled cucullaeids with sharp posterior
carina and strongly incurved unibones. Ligamental area very large; hinge long and
narrow, with few horizontal teeth but numerous perpendicular teeth, best developed
anteriorly. Myophoric septum prominent. Surface with both concentric and radial
sculpture, the radial element strongest on the anterior half and in the young.

Noramya differs from Cueullaea and Idonearea in hinge and surface sculpture and
includes Area gabrielis Leymerie, A. dilatata Coquand, A. gresslyi de Loriol, and
Cueullaea tumida Matheron, all of Aptian or Neocomian age. The South African

576 PALAEONTOLOGY, VOLUME 3

Megacucullaea kraussi (Tate), also from the Lower Cretaceous, has a much bolder
radial costation.

Genus cryptochasma nov.

Type species. C. ovale sp. nov. (= Cucullaea sp. ?, Keeping 1883, p. 115, pi. 5, fig. 8, holotype; =
Cucullaea cf. cornueliana Kirkaldy 1937, p. 118), Upper Aptian, England.

Diagnosis. Small, elongate cucullaeids with faint radial ornament; interior with myo-
phoric septum and a ridge on the umbo; hinge area narrow, teeth parallel with the
hinge-margin.

The internal ridge on the umbo and the elongate shape, approaching the Parallelo-
dontidae, are the chief features of this genus, the type species of which is a characteristic
fossil of the cunningtoni Subzone. The umbonal ridge is reproduced as a cleft on internal
moulds (Plate 82, figs. 6a, 6b). In the Rhaetic-Jurassic genus Catella Healey the internal
ridge is much stronger and corresponds to a constriction of the surface; the hinge-plate
is broader, with the anterior teeth set obliquely across it.

Family limopsidae
Genus limopsis Sasso 1827
Limopsis dolomitica sp. nov.

Plate 79, fig. 4

Holotype. GSM Zm 2137, base of Sandgate Beds, Mill Point, Folkestone, Kent (Author's Coll.).

Diagnosis. Limopsis averaging 13 mm. in length, narrower and more oblique than L.
albensis Woods, with shorter, less rectilinear hinge line, and apparently no radial lines.

Limopsis is rare in the Lower Cretaceous and it is surprising to find it a common
fossil in the base of the Sandgate Beds at Folkestone. The specimens are preserved in
gritty dolomite, partly decorticated, and do not show the hinge. Also known from the
Ferruginous Sands at Shanklin.

EXPLANATION OF PLATE 80
Figures natural size unless otherwise stated.

Figs. 1, 2. Scittila nasuta gen. et sp. nov., Atherfield Clay Series (Crackers), Atherfield, Isle of Wight.
1, Hinge of holotype (left valve) (overhanging valve-margin below figure number should not be
mistaken for posterior lateral tooth). (SM B 12778.) 2, Hinge of right valve. (GSM 98608; author’s
coll.) Both X 3.

Fig. 3. Icanotia pennula sp. nov., holotype. Upper Perna Bed, Redcliff, Isle of Wight. (BM L16284.)
X 1-5.

Fig. 4. Epicyprina harrisoni sp. nov., holotype, Folkestone Beds, Ivy Hatch, near Ightham, Kent.
(GSM 98599; author’s coll.) X0-8.

Figs. 5a, b. Proveniella rosacea sp. nov., right side (a) and dorsal (b) view of holotype, Atherfield Clay,
Nutbourne Brickworks, Shottermill, near Haslemere, Surrey. (GSM 98590; J. F. Kirkaldy coll.)
Figs. 6, 7. Pachythaerus teal/i sp. nov. 6a, b. Side and interior of holotype, Lower Greensand, Potton,
Bedfordshire. (GSM 98593; author's coll.) 7, Right valve, Iron Sands, Seend, Wilts. (BM 88836;
W. Cunnington coll.) Both X 1-7.

Figs. 8a, b. Pterotrigonia mantelli sp. nov. ‘Vinagel’ squeeze from holotype-mould showing left side
(a) and escutcheon ( b ), Sandgate Beds (Iron Sands), Parham Park, Sussex. (BM 9140; Mantell coll.)
Fig. 9. Tortarctica similis (J. de C. Sowerby), hinge of right valve, Albian, St. Florentin, Yonne,
France. (BM. 41696.)

Fig. 10. Deshayesites cal/ic/iscus sp. nov., topotype, Atherfield Clay Series (Crackers), Atherfield, Isle
of Wight. (SM B 27054; Wiltshire coll.)

Palaeontology, Vol. 3

PLATE 80

CASEY, Aptian molluscs

RAYMOND CASEY: STR ATIG R A PH IC AL PALAEONTOLOGY OF GREENSAND 577

Family trigoniidae

Genus pterotrigonia van Hoepen 1929
Pterotrigonia mantelli sp. nov.

Plate 80, figs. 8 a, 8 b; text-fig. 8 b-d

Holotype. BM 9140, Sandgate Beds, Parham Park, Sussex (Mantell Coll.).

Diagnosis. Like P. vectiana (Lycett) but larger, the ribs less strongly crenulated and the
narrow areas bounding the escutcheon denuded of ribs except near the umbo.

text-fig. 8. Ornament of the escutcheon in Pterotrigonia. a, P. vectiana (Lycett), fissicostatus Zone.
b, P. mantelli anterior sp. nov., subsp. nov., bowerbanki Zone, c, d, P. mantelli s.s., nutfieldensis

Zone, X 1 .

Lower Cretaceous Pterotrigonia of the vectiana-aliformis group form an evolutionary
series, the principal line of progression being reduction of ribbing on and around the
escutcheon, starting at the posterior end of the shell. P. vectiana (lectotype here selected:
GSM 27075, figured Lycett 1875, pi. 24, figs. 10, 10 a, b ) is a Perna Bed species and has
strong transverse ribbing that spreads from the escutcheon over the whole of the
area, leaving only the posterior third of the area bare in mature shells. In P. mantelli
the inner portion of the area, bounded externally by a groove, is bare for the posterior
half of its length, the outer portion for about the posterior three-quarters. This species
ranges through the whole of the Upper Aptian and the Lower Albian. An earlier form,
tending towards P. vectiana, occurs in the deshayesi and bowerbanki Zones of the
Lower Aptian, and may be regarded as a chronological subspecies, P. mantelli anterior
subsp. nov. (type: the original of Pictet and Renevier 1847, pi. 14, figs. 2 a-c, from the
Aptian of Perte du Rhone, France; figured as Trigonia aliformis).

Family myoconchidae
Genus myoconcha J. de C. Sowerby 1824
Myoconcha delta sp. nov.

Plate 81, figs. 3, 4 a-b

Holotype. GSM 20084, Iron Sands of Seend, Wiltshire (Cunnington Coll.).

Diagnosis. Similar to M. cretacea d’Orbigny, but with posterior end symmetrically
convex, umbonal cavity not overhanging the adductor scar, and with a marginal posterior
lateral tooth in the left valve.

578

PALAEONTOLOGY, VOLUME 3

There are ten specimens of this species in the Cunnington Collection in the Geological
Survey Museum, either internal moulds or decorticated shells. It is not possible to say,
therefore, whether the surface ornament agreed with that of M. cretacea, the only
other member of the genus recorded from the British Cretaceous.

Family limidae

Genus limatula S. V. Wood 1839

Limatula sabulosa sp. nov.

Plate 83, fig. 5

1942 Limatula dupiniana (d'Orbigny); Wright and Wright, p. 86.

Holotype. GSM 98606, Folkestone Beds ( Farnhamia horizon), Coxbridge pit, Farnham, Surrey
(Author’s Coll.).

Diagnosis. Small, subelliptical, with symmetrically rounded ventral margin; posterior
margin only slightly more convex than the anterior margin. Ears equal. Median part
of shell with about twenty very narrow radial ribs, separated by broad depressions,
more closely spaced on the posterior slope.

Limatula dupiniana (d’Orbigny) has fewer ribs and they are placed asymmetrically
on the shell. L. tombeckiana (d’Orbigny) and L. fittoni (d’Orbigny) have fewer ribs,
narrow interspaces, and more prominent ears. The species occurs sporadically in the
tardefurcata and mammillatum Zones.

Family requieniidae
Genus toucasia Munier-Chalmas 1873

Toucasia lonsdalii (J. de C. Sowerby). The majority of specimens of this rudist, including
the type, were obtained from iron sands at Stock Orchard, south of Caine. Others have
been found at Lockswell Heath, near Caine, at Seend (GSM 44662), Parham Park,
Sussex (GSM 52029), and near Headleywood Farm, Hampshire (GSM 98598). This is
the only Lower Greensand representative of a group of lamellibranchs more typical
of the Tethyan region and the fact that all the occurrences noted above are in the
cunningtoni Subzone of the nutfieldensis Zone suggests an isolated penetration to the
British Province.

EXPLANATION OF PLATE 81
Figures natural size unless otherwise stated.

Figs. 1 a, b. Cheloniceras ( Epicheloniceras ) gracile sp. nov., side and venter of holotype, Ferruginous
Sands (nodules near base of Group IX), shore below Walpen High Cliff, Atherfield, Isle of Wight.
(GSM Zm 1953; author’s coll.)

Figs. 2a, b. Deshayesites forbesi sp. nov., side and venter of holotype, Atherfield Clay Series (Crackers),
Atherfield, Isle of Wight. (GSM 30918.)

Figs. 3, 4. Myoconcha delta sp. nov.. Iron Sands, Seend, Wilts. 3, Holotype, partly decorticated.
(GSM 20074.) 4a, b, Side and dorsal views of internal mould. (GSM 20087.) Both W. Cunnington
coll.

Fig. 5. Protodonax minutissimus (Whitfield), internal mould of right valve. (GSM 44617.) Locality,
horizon, and collector as before, x 1-5.

PLATE 81

Palaeontology, Vol. 3

CASEY, Aptian molluscs

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 579

Family mactridae

Genus geltena Stephenson (in Vokes) 1946
Geltena meyeri sp. nov. (= Mactra sp., Woods 1907, p. 177, pi. 27, figs. 17, 18). Holo-
type : original of Woods’s pi. 27, figs. 17a, b), Ferruginous Sands (‘Urchin Bed’),
Shanklin, Isle of Wight.

Family astartidae
Genus freiastarte Chavan 1952

Freiastarte praetypica sp. nov. (= Astarte sp., Woods 1906, p. Ill, pi. 15, figs. 3, 4).
Holotype : original of Woods’s pi. 15, fig. 3. A characteristic species of the jacobi Zone.
The matrix of Woods’s originals in the Sedgwick Museum shows that they were obtained
from Price’s bed 1 of the Sandgate Beds, not the Folkestone Beds as now understood.

Genus eriphyla Gabb 1864

Eriphyla pseudostriata (d’Orbigny) (= Astarte pseudostriata d’Orbigny 1850, nom. nov >
for A. substriata Leymerie 1842, non Bronn 1835). Recorded from the Lower Greensand
by Forbes, Fitton, and Morris under the name Astarte substriata Leymerie, but apparently
missed by Woods. I have found it only in the bowerbanki and nutfieldensis Zones. An
example from Shanklin in the Sedgwick Museum (B 13742) is in ‘Urchin Bed’ matrix.

Family crassatellidae
Genus pachythaerus Conrad 1869
Pachytliaerus tealli sp. nov.

Plate 80, figs. 6, 7

Holotype. GSM 98593, Lower Greensand, Potton, Bedfordshire.

Diagnosis. Shell small (up to 15 mm.), subtrigonal, height and length about equal,
moderately inflated, with posterior diagonal angulation, beaks a little anterior. Postero-
doisal margin very feebly convex, antero-dorsal margin long and almost straight,
anterior extremity low, posterior extremity truncated vertically. Surface with concentric
ridges, lamellose on the beak and behind the angulation. Hinge typical of the genus;
margins crenulate internally.

Common at Seend; a single valve from the regularis Subzone (bed 6) of East Cliff,
Folkestone. Not described by Woods.

Genus disparilia Chavan 1953

Disparilia disparilis (d’Orbigny). This primitive crassatellid, typically Neocomian,
occurs as a great rarity in the Perna Bed of Surrey.

Genus seendia nov.

Type species. Crassatella saxoneti Pictet and Roux, 1847, Albian, France.

Diagnosis. Oblong, inequilateral, umbo anterior, thick-shelled, compressed, lunule
narrow, circumscribed by an incised line. Surface with concentric ridges and faint

580

PALAEONTOLOGY, VOLUME 3

radial lines. Adductor and pedal scars deeply impressed, the posterior adductor mounted
on a projecting plate. Margins crenulate internally. Hinge-plate narrow, with two
cardinal teeth in each valve, 3 b much larger than the others. Ligament sunk between the
valves, apparently as in Disparilia.

This genus has the aspect of a Jurassic Prorokia without lateral hinge teeth. The type
species is characteristic of the Iron Sands of Seend, whence Woods (1906, p. 104, pi. 14,
figs. 2a, b, 3) figured an incomplete mould and an example with shell under the name
Astarte elongata d’Orbigny. The latter is a Valanginian species of Seendia, less flat-
sided than S. saxoneti, and with a sunken lunule and prominent umbo (see Pictet and
Campiche 1886, pi. 124, figs. 8, 9). A piece of shell is chipped from the lunular region
in the original of Woods’s pi. 14, fig. 2a, making the umbo appear unnaturally acute.

Family scambulidae
Genus anthonya Gabb 1 864

Anthonya cantiana Woods (1906, p. 130, pi. 19, figs. 4, 5). The types of this species were
attributed to the Folkestone Beds, though their matrix is that of Price’s Bed 1 of the
Sandgate Beds (cf. Freiastarte praetypica sp. nov.).

Anthonya woodsi sp. nov. (= A. sp.. Woods 1906, p. 131, pi. 19, fig. 6). Holotype:
GSM 98592, a bivalved example collected by the author (PI. 79, fig. 5). The species is
now known by several examples, all from the Crackers of Atherfield.

Genus mediraon Vokes 1946
Mediraon sulcatum sp. nov.

1906 Astarte sinuata d'Orbigny; Woods, p. 104, pi. 14, figs. 7-9.

Holotype. The original of Woods 1906, pi. 14, fig. 7, from the Crackers of Atherfield, Isle of Wight.

Diagnosis. More equilateral than M. sinuatum (d’Orbigny), with pointed umbo and a
less excavated lunular region.

There seems to be a complete gradation from the concentrically ribbed, equilateral,
sharply trigonal shells of the type of Astarte subacuta d’Orbigny to the oblong shells
with divaricate ribbing typical of Mediraon (type species M. divaricatum Vokes). The
hinge-structure of Mediraon is seen both in M. sulcatum and in the Barremian form
attributed to A. sinuata by Gillet (1921, pi. 1, figs. 13, 14; cf. Vokes 1946, pi. 6, figs. 6,
1 1). Scambula Conrad is intermediate in shape between Mediraon and Anthonya.

Family carditidae
Genus fenestricardita nov.

Type species. Venus Ifenestrata Forbes 1845, Lower Aptian, south-east England.

Diagnosis. Small, oblong shells with umbo well forward and not rising much above
the hinge-line. Lunule deeply sunk but with margin of left valve pouted above tooth
2b. Surface with strong reticulate sculpture; posterior area flattened, with two or more
nodular carinae. Hinge-teeth as in Xenocardita; margins crenulate internally.

The genus includes Cardita tricarinata d’Orbigny of the Cenomanian.

RAYMOND CASEY: STRATIGRAPHIC AL PALAEONTOLOGY OF GREENSAND 581

Genus trapezicardita nov.

Type species. Cypricardia squamosa Keeping 1883, Aptian, England.

Diagnosis. Small, rounded oblong, inflated shells with ventral and dorsal margins
straight and nearly parallel. Umbones terminal; lunule cordate; surface angulated
posteriorly and with narrow radial ribs and periodic concentric lamellae. Hinge as in
Praeconia', margins crenulate internally.

The genus includes Cypricardia arcadiformis Keeping. Both of these species were
referred by Woods to Trapezium with question, though Keeping had correctly com-
pared T. squamosa with Cardita. The latter species shows a curious resemblance in
hinge-structure and shape to the Middle Jurassic Praeconia rhomboidalis (Phillips).

Family tellinidae
Genus linearia Conrad 1860

Linearia cornea sp. nov. (= Tellina ( Linearia ) sp., Woods 1907, p. 175, pi. 27, fig. 9).
Holotype. BM 48626, the original of Woods’s pi. 27, fig. 9, from the Crackers of Ather-
field. A characteristic forbesi Zone species. I have excavated the hinge of a left valve
of my own collecting (GSM 98597) and find a single grooved triangular cardinal tooth
under the beak and mere vestiges of anterior and posterior lateral teeth.

Family icanotiidae nov.

Diagnosis. Equivalve, closed, compressed, elongate and fragile tellinaceans. Outline
subelliptical tending to oblong, anterior end narrowly rounded. Surface may be almost
smooth, but usually with ribs or threads radiating from the beak, strongest on, or
confined to, the posterior slope. Ligament external, opisthodetic, seated on nymphs.
Hinge lucinoid, without lateral teeth, formula 3 a, 2bj2b, 4b, the teeth entire, 2b and 3 b
prominent, triangular, 3 a and 4b subject to reduction or elimination. Large, deep,
rounded pallial sinus. Habitat marine.

This nominal family is proposed for reception of the two genera Icanotia Stoliczka
and Scittila gen. nov., discussed below. Species placed in these two genera have been
generally referred to the Veneridae and the Tellinidae respectively, but are here regarded
as closely allied derivatives of the Jurassic Tancredia adapted to life in a burrow.
Although retaining the simple cardinal dentition and some of the external features of
the Tancrediidae, their fragile shells, elongate form, deep pallial sinus, lack of lateral
teeth and tendency to develop strong radial sculpture, combine to exclude them from
that family. The Gariidae (= Psammobiidae), an essentially Tertiary and Recent group,
differ from the Icanotiidae in having, typically, subequilateral shells, prominent nymphs,
opisthogyral beaks and bifid principal cardinal teeth. Text-fig. 9 illustrates the line of
evolution of the Icanotiidae. Radial sculpture, which eventually spread over the whole
surface of Icanotia, had already appeared in Tancredia, being present behind the umbo
in well-preserved examples of T. donaciformis Lycett from the Upper Lias (e.g. GSM
LD 1536).

582

PALAEONTOLOGY, VOLUME 3
Genus icanotia Stoliczka 1870

The nominal type species of Icanotia is Psammobia impar Zittel 1865, by original
designation, but there has been some confusion as to the taxon to which this name
applies (Casey 1953). Psammobia impar was proposed by Zittel as a substitute combina-
tion for Capsa elegans d’Orbigny 1844, which became a secondary homonym of Solen
elegans Matheron 1842, when he transferred both species to the genus Psammobia.
Under the Rules d’Orbigny’s species, from the Cenomanian of Le Mans, is the taxonomic
type of Icanotia and its correct name is Icanotia impar (Zittel). The form from the Gosau
formation of Austria described and illustrated by Zittel as Psammobia impar is here

text-fig. 9. Evolution of shell-form in the Icanotiidae. a, Tancredia donaciformis Lycett, Lower and
Middle Jurassic, b, Scittila nasuta group, gen. et sp. nov.. Lower Cretaceous (Hauterivian-Lower
Aptian), c, S. nasuta var. radiata var. nov., L. Cretaceous (L. Aptian, fissicostatus-forbesi Zones).
d, Icanotia pennula sp. nov., L. Cretaceous (L. Aptian, fissicostatus Zone), e, /. siliqua sp. nov., L.
Cretaceous (Upper Albian). /, I. zitteli sp. nov., Upper Cretaceous (Senonian).

designated I. zitteli sp. nov. (holotype: the original of Zittel 1865, pi. 2, fig. 4). Stoliczka
(1870, p. 145) treated Icanotia as a subgenus of the venerid Baroda (= Legumen Conrad
1858), being influenced by similarities in external form and supposed identity of hinge-
structures. It is evident from the text that Stoliczka had only imperfect specimens to go
on and it now seems that his drawings of the hinge of Icanotia are inaccurate restorations.
The hinges of I. impar (as seen in topotypes in the Museum d’Histoire Naturelle, Paris),

RAYMOND CASEY: STR ATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 583

I. pennula sp. nov., and I. siliqua sp. nov. agree with that of Scittila , as does that of
I. pulchra, figured photographically by Wade (1926, pi. 29, fig. 5). The genus has an
almost world-wide distribution and ranges from Aptian to Maestrichtian but is never
common. In the two Aptian species I. studeri (Pictet and Renevier) and I. pennula sp.
nov. the beaks are more prominent and not so far forward as in later members of the
genus, features which link them with Scittila.

Icanotia pennula sp. nov.

Plate 80, fig. 3; text-figs. 9 d, 1 1c

Holotype. BM L 16284, Atherfield Clay Series, Upper Perna Bed, Sandown, Isle of Wight.
Diagnosis. Like I. studeri but anterior end more produced, posterior end truncated and
ventral margin without pronounced upward sweep.

The holotype is a unique bivalved shell with the valves displaced so as to show the
hinge.

Icanotia siliqua sp. nov.

Text-fig. 9e

1913 Tapes (Icanotia) sp., Woods, p. 431, pi. 62, figs. 14 a, b.

Holotype. BM L 3379, Upper (Blackdown) Greensand, Blackdown, Devon, figured by Woods (1913,
pi. 62, figs. 14 a-b) as Tapes ( Icanotia ) sp.

Diagnosis. Oblong elliptical Icanotia with beak distance seven-tenths and maximum
height and thickness at mid-length. Lunule narrow and deeply sunk. Postero-dorsal
and antero-dorsal margins straight, converging on the inconspicuous umbo at an angle
of 150°. Area of coarse radial sculpture covers a sector of about 25°; anteriorly the radii
become closely spaced and weaker, are almost obsolete on the mid shell, but rejuvenate
slightly at the anterior end.

I have located six specimens of this rare species in the British Museum and the
Geological Survey Museum, all from the Upper Albian greensands of Blackdown,
Haldon, and Seaton, Devon. An example from the Red Bed (anglicus Subzone) of Sand-
ling Junction, near Hythe, Kent (GSM Zm 667), is too poor for certain determination.
The anterior end is more produced than is indicated in Woods’s restored figure.

Genus scittila nov.

Type species. S. nasuta sp. nov., Lower Aptian, south-east England.

Diagnosis. Very compressed Icanotiidae with no lunule and only an incipient escutcheon.
Posterior margin obliquely truncated and posterior slope carinated. Umbo subcentral.
A shallow furrow between umbo and middle of ventral margin. Radial sculpture may
be obscure. Range: Hauterivian to Aptian.

Scittila nasuta sp. nov.

Plate 80, figs. 1,2; text-figs. 9b, 9c

1907 Tellina carteroni d’Orbigny; Woods, p. 171, pi. 26, figs. 15, 16.

Holotype. The original of Woods's pi. 26, figs. 16 a-c, from the Crackers of Atherfield, Isle of Wight.

The combination Tellina carteroni was proposed by d’Orbigny (1845, p. 420) as a
substitute for Tellina ? angulata Deshayes (in Leymerie, 1842, pp. 3, 24), this being a

584

PALAEONTOLOGY, VOLUME 3

homonym of T. angulata Linne. D’Orbigny illustrated T. carteroni by a specimen from
the Neocomian of Marolles, France, which, judging from the figures, is a different
species from that of Deshayes, which came from the Neocomian of Vendeuvre. Woods
noted this discrepancy in the figures and assumed that it was due to imperfect preserva-
tion of the originals. Since the original of Deshayes’s figure is lost and d’Orbigny’s
illustration is known to be restored, this view can be neither refuted nor confirmed.
Whichever specimen is taken to represent d’Orbigny’s species its identity with the
English Lower Greensand forms can be assumed only on the premiss that it is incorrectly
figured. Apart from the discrepancies in the figures, there are good reasons for regarding
this assumption as unsafe. Stoliczka (1870, p. 123) stated that casts of T. carteroni
show impressions of both cardinal and lateral teeth, and Gillet (1924, p. 136) in describ-
ing the hinge of this species alluded to its long lateral teeth A II and P II and to its
bifid cardinal tooth 2b. The English species has no lateral teeth (PI. 80, figs. 1, 2) and
the cardinal teeth are undivided. Better agreement with d'Orbigny’s figure is shown by
a specimen of ‘ Tellina carteroni ’ from the Hauterivian of Sainte Croix in the Sedgwick
Museum, an internal mould without impressions of lateral teeth. In view of the uncer-
tainty as to the characters of T. carteroni and since it is desirable that the type species
of a genus should be free from such uncertainty, it is proposed to apply the combination
Scittila nasuta to the Lower Greensand form previously described and figured by Woods
as T. carteroni. Radial ornament may be seen faintly under magnification in the typical
S. nasuta and is much more conspicuous in the var. radiata nov. (type: the original of
Woods, 1907, pi. 26, fig. 17).

Family donacidae?

Genus protodonax Vokes 1945

Protodonax minutissimus (Whitfield). Small wedge-shaped shells, up to 13 mm. long,
from the Iron Sands of Seend are referable to this species, originally described from the
Aptian of the Lebanon and refigured by Yokes (1945, pi. 46, figs. 16-18; 1946, pi. 9,
figs. 26-28). The Cunnington Collection in the Geological Survey Museum contains a
cluster of moulds (GSM 44617) and an isolated left valve (GSM 44618). Apart from the
Lebanon occurrence, Protodonax is well represented in the Cretaceous of the North
American interior, though it does not seem to have been noted previously in western
Europe.

Family solenidae
Genus senis Stephenson 1952

S. wharburtoni (Forbes). This common Lower Greensand species was referred to
Solecturus by Forbes (1845, p. 237) and to Pharus by Woods ( 1909, p. 217) and is known
only in closed shells or moulds. The valves of a specimen from the Crackers (GSM
3041) were prized apart, mounted in plaster, and cleaned out, thereby revealing an
edentulous hinge, a narrow internal ridge extending obliquely forwards and downwards
from the beak, and another, fainter ridge, closer to the shell margin, extending back-
wards from the beak. These are the characters of the genus Senis (type species S. elon-
gatus Stephenson) recently described from the Cenomanian of Texas (Stephenson 1952,
p. 120, pi. 31, figs. 8-13).

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 585

Family neomiodontidae
Genus eomiodon Cox 1935

Eomiodon cf. libanoticus (Fraas). A form conspecific with or allied to E. libanoticus of
the Aptian of the Middle East was an unexpected find in the Lower Greensand of Dorset
(Punfield Marine Band and westward equivalents). Eomiodon is a marine-brackish
genus allied to the Purbeck-Wealden Neomiodon (Casey 1956), is characteristically
Jurassic, and has not been noted previously in the British Cretaceous. Large internal
moulds of E. cf. libanoticus from Worbarrow Bay (GSM Rh 2438, 2439, 2453, 2456)

text-fig. 10. Eomiodon cf. libanoticus (Fraas), Punfield Marine Band, Dorset, a. Internal mould,
Worbarrow Bay (Bed 7) (GSM Rh 2438), X 1-3. b, Fragmentary juvenile, Corfe Castle Station (GSM
Rh 2811), x2. c. Ftinge of right valve, Punfield (GSM 86398), x7.

were recorded by Arkell (1941b, p. 171) as Astarte obovata J. de C. Sowerby, juveniles
from Corfe Castle Station (GSM Rh 2811, 2823) as Astarte subcostata d’Orbigny.
From a slab of ‘Marine Band’ collected at Punfield Cove 1 isolated a juvenile showing
all the fine details of hinge-structure (GSM 86398; text-fig. 10c), just like E. fimbriatus
(Lycett) of the Forest Marble, and Mr. S. W. Hester found another in the ironstone at
Lulworth Cove (GSM 86653).

Family arcticidae
Genus tortarctica nov.

Type species. Isocardia similis J. de C. Sowerby 1826, Lower Albian, south-east England.

Diagnosis. Large trigonal-ovate, well-inflated shells with prominent, spirally enrolled
beaks. Lunular region depressed, escutcheon limited by blunt carinae. Hinge cyprinoid,
formula A I, III, 1, 3 a, 3b, P I/A II, 2a, 2b, 4b, P II. A I and A II vestigial; A III pustu-
lar; 1 spoon-shaped; 3b bifid, united with 3 a into a single curved, strongly opisthocline
tooth lying almost horizontal; 2 a and 2b nodular, separated by a constriction; posterior
laterals close behind the nymph.

Tortarctica is an arcticid related to Venilicardia and Epicyprina in which the hinge
has been modified in correlation with spiral enrolment of the beaks, thus simulating
the Recent Glossus (= Isocardia), in which there is a much more radical alteration of

586

PALAEONTOLOGY, VOLUME 3

hinge-structure. I have discussed elsewhere (Casey 1952, pp. 1 46—50) the homologies
of the hinge-teeth of other Mesozoic arcticids wrongly assigned to the Glossidae
(= Isocardiidae). Sowerby (1826, p. 27) had attributed his Isocardia similis to the
greensand of ‘Sandgate, near Margate’; Woods (1907, p. 152) correctly identified the
matrix as that of the mammillatum Zone. The species ranges through the whole of
the Folkestone Beds. There is a particularly fine example of the right valve in the British
Museum (BM 41696) from the Albian of St. Florentin, Yonne, France, labelled Cyprina
cordiformis d’Orbigny. Though preserved in hard glauconitic sandstone, it furnished
the hinge-preparation illustrated in PI. 80, fig. 9. The hinge of the left valve is best seen
in GSM Zm 846. D’Orbigny’s Cyprina cordiformis, also found in the English mammil-
latum Zone, is also referable to Tortarctica.

Genus epicyprina Casey 1952
Epicyprina harrisoni sp. nov.

Plate 80, fig. 4; text-fig. 11 d

Holotype. GSM 98599, a silicified right valve, Folkestone Beds, Ivy Hatch, near Ightham, Kent
(Author’s Coll.).

Diagnosis. Large Epicyprina (averaging 115 mm. long), subtrigonal, very inequilateral;
inflation moderately strong but uneven, the shell flattening in a postero-ventral direction.
Posterior ridge very faint. Umbo well recurved, prosogyrous, placed at the anterior
quarter of the length. Lunular region well excavated, the lunule depressed, obscurely
circumscribed, the margin straight. Postero-dorsal margin long, convex, falling steeply
to a short, subvertical posterior margin. Dorsal margin sweeping up in a continuous
curve with the narrowly rounded anterior extremity. Hinge typical of the genus.

This is the Cyprina angidata of Harrison (Gossling 1929, p. 255). The sharp decline
of the postero-dorsal margin, narrowly rounded anterior extremity, and the deep lunular
area are the chief distinguishing features compared with Epicyprina angidata (J. Sowerby)
of the Upper Greensand. E. harrisoni occurs sporadically through the jacobi and tarde-
furcata Zones, but nowhere in greater numbers than around Ightham.

Genus proveniella Casey 1952
Proveniella rosacea sp. nov.

Plate 80, figs. 5a, 5b

1938 Cyprina sedgwicki (Walker); Kirkaldy and Wooldridge, p. 139.

Holotype. GSM 98590, Atherfield Clay, Nutbourne Brickworks, Shottermill, near Haslemere, Surrey
(J. F. Kirkaldy Coll.).

Diagnosis. Smaller, more rotund, and less elongate than Proveniella meyeri (Woods),
the hinge slender, with posterior lateral tooth P III tucked under the shell-margin.

Characteristic of the Atherfield Clay of the Haslemere district. One of Dr. Kirkaldy’s
specimens (GSM 98588) was dissected to expose the hinge, on which generic determina-
tion depends. Isocyprina sedgwicki (Walker) has a steeply falling, convex postero-
dorsal margin, a circumscribed lunule, and a different hinge.

RAYMOND CASEY: STR ATIGR APHICAL PALAEONTOLOGY OF GREENSAND 587

Genus venilicardia Stoliczka 1870

Venilicardia sowerbyi (Woods). Lectotype here selected: the original of Woods, 1907,
pi. 21, fig. 8, from the Hythe Beds of Hythe, Kent (Sedgwick Museum). Shells collected
by Fitton from the Flythe Beds around Folkestone and Hythe were identified by J. de C.
Sowerby as Cyprina angulata (Fitton 1836, p. 128). D’Orbigny (1850, p. 78) renamed them
Cyprina sowerbyi, but gave no description, figure, or indication. Nevertheless, Woods
(1907, p. 138) used the combination Cyprina sowerbyi d’Orbigny for a common species
of Venilicardia of the Lower Greensand. D’Orbigny’s use of the name being nude, the
combination Cyprina sowerbyi must be attributed to Woods, with Woods’s examples
as types. Neither d’Orbigny nor Woods seems to have consulted Fitton’s originals. Some
of them (GSM 52030, 18862) are here identified as Venilicardia inornata (d'Orbigny).

Family corbiculidae
Genus filosina Casey 1955

Filosina cf. gregaria Casey. Internal moulds indistinguishable from those of the common
Weald Clay and Wealden Shales Filosina occur in the Lower Greensand ironstone of
Lulworth Cove, associated with Eomiodon and Exogyra. GSM 86652 shows the diag-
nostic features of the hinge.

Family pinnidae
Genus pinna Linne 1758
Subgenus stegoconcha Bohrn 1 907

The large Stegoconchas of the Lower Greensand are a curious omission from earlier
literature. There are at least three species, apparently undescribed, some attaining nearly
a foot in length. The largest is comparable with P. (S.) iburgensis Weerth and occurs
in the Perna Bed of the Isle of Wight (e.g. BM 32584; also Sandown Museum); another,
similar to P. (S.) hombresi Pictet and Campiche, is found in the Crackers (e.g. BM
48626). The Hythe Beds of East Kent yield moulds of a Stegoconcha up to 10 inches
long, here listed as P. (S.) cf. gervaisii Dumas (e.g. GSM Zm 2212; also British Museum
and Folkestone Museum).

Family isognomonidae
Genus inoceramus W. Smith 1816
Inoceramus coptensis sp. nov.

Plate 82, fig. 5

1900 Inoceramus sp., large; Jukes-Browne, p. 228.

1939 Inoceramus ? neocomiensis d’Orb.; Jackson, p. 76.

1941 Inoceramus cf. anglicus Woods; Brown, p. 1 1.

19496 Inoceramus sp. nov.; Casey, p. 225.

1955c Inoceramus sp. nov.; Casey, p. 232.

Holotype. GSM Zm 26, a left valve, Folkestone Beds, regularis Subzone, bottom stone band, near
Copt Point, Folkestone, Kent (Author’s Coll.).

588

PALAEONTOLOGY, VOLUME 3

Diagnosis. Shell inequivalve, very inequilateral, of moderate inflation, longer than high,
greatest length from umbo to postero-ventral extremity. Left valve tumid, with greatest
convexity about one-third the distance from umbo to ventral margin; posterior and
postero-dorsal regions compressed ; anterior area fairly small, steeply turned, excavated
near the umbo; ventral margin convex, forming a parabolic curve together with the
posterior margin; hinge-line about three-quarters the length of the shell, making rather
more than a right angle with the anterior margin, which is nearly straight ; umbo anterior,
pointed, incurved, salient above the hinge-line. Right valve considerably less tumid
than the left, with a much smaller and less incurved umbo. Surface with narrow con-
centric ribs and growth-rings of asymmetrical curvature; on the internal mould the
ribs are sharper and separated by wide, shallow concave interspaces.

Characteristic of the regularis Subzone and the base of the mammillatum Zone in
southern England, this species is apparently ancestral to I. salomoni of the mammil-
latum Zone, with which it overlaps in range. It is relatively longer than I. salomoni,
less inequivalve, has a less inflated, non-sulcate left valve, longer hinge-line, and smaller
anterior area. The Aptian I. neocomiensis d'Orbigny is shorter, has more evenly curved
ribs, and is less tumid below the umbo.

Inoceramus salomoni d’Orbigny

One of the commonest fossils in the mammillatum Zone of Europe and may be
collected in thousands at Copt Point, Folkestone. Yet every example recorded, figured,
or described under this name is a left valve. Search for the missing right valve at Copt
Point drew attention to some flat, operculum-like shells previously identified as
"Lima' montana Pictet and Roux; these eventually proved to belong to the present
species by discovery of two examples with the valves joined (GSM Zk 4564, 4565) (text-
fig. 116). I am of the opinion that the original Lima montana of Pictet and Roux (1853,
pi. 43, figs. 1 a, b ), from the Albian of Saxonet, is also based on a right valve of I.
salomoni. Since other lamellibranchs in the mammillatum Zone are commonly found in a
bivalved condition, the rarity of double-valved I. salomoni is presumably due to excep-
tionally weak attachment of the valves. With the two halves of the shell so different in
shape, current-sorting seems an obvious explanation of the rarity of isolated right valves.

EXPLANATION OF PLATE 82

Figures natural size unless otherwise stated.

Figs. 1 a, b. ParahopJites cunningtoni sp. nov., holotype, Iron Sands, Seend, Wilts. (OUM K 184;
E. C. Davey coll. )

Figs. 2, 3. Prodeshayesites obsoletus gen. et sp. nov. 2a, b. Nucleus of holotype, Perna Bed, Woodhatch,
Surrey. (BM C 36944.) 3, Phragmocone, Upper Perna Bed, Atherfield, Isle of Wight. (Museum of
Isle of Wight Geology, Sandown, no. 88.) xO-5.

Figs. 4 a, b. Cuneocorbula arkelli sp. nov., holotype, bed 7, Worbarrow Bay, Dorset. (GSM Rh 2466.)
a X 2 , b X 3.

Fig. 5. Inoceramus coptensis sp. nov., holotype, Folkestone Beds ( regularis Subzone), East Cliff,
Folkestone, Kent. (GSM Zm 26; author’s coll.)

Figs. 6a, b. Cryptochasma ovale gen. et sp. nov., left side (a) and dorsal (b) views of internal mould.
Iron Sands, Seend, Wilts. (GSM 18676; W. Cunnington coll.) X 1-5. (The straight posterior margin
is due to breakage.)

Palaeontology, Vol. 3

PLATE 82

..ST^r

CASEY, Lower Greensand molluscs

text-fig. 11. Lower Greensand lamellibranchs. a. Aptolinter ciptiensis (Pictet and Campiche), hinge of
right valve reconstructed from natural impressions (GSM Zb 3396, 3400, Perna Bed, Earlswood
Common, Surrey), x 4. b, b', Inoceramus salomoni d’Orbigny, diagrammatic sketch of right side ( b ) and
posterior end ( b '), based mainly on GSM Zk 4565, main mammillatum bed, Copt Point, Folkestone,
Xl.r, c', Jcanotia pennula sp. nov., hinges of left (c) and right ( c ') valves of holotype, X 2. d, d\ Epicy-
prina harrisoni sp. nov., interior and hinge (d) and dorsal view ( d ') of holotype (right valve), X 0-66.

B 6612

Qq

590

PALAEONTOLOGY, VOLUME 3
Family ostreidae
Genus exogyra Say 1820

Exogyra latissima (Lamarck) (= Gryphaea latissima Lamarck 1801, = Gryphaea
eoirioni Defranc 1821, = Gryphaea sinuata J. Sowerby 1822, = Gryphaea aquila
Brongniart 1822). This synonymy has been pointed out by several authors, for example
Pervinquiere (1912, p. 176) and Renngarten (1926, p. 60). English authors, in defiance
of the Rules of Priority, have preferred the combination Exogyra sinuata (J. Sowerby).

Family corbulidae
Genus cuneocorbula Cossmann 1886
Cuneocorbula arkelli sp. nov.

Plate 82, figs. 4a, 4b

1941b Anthonya cormieliana (d'Orbigny); Arkell, p. 171.

Holotype. GSM Rh 2466, Lower Greensand (bed 7 of Arkell 19476, p. 176), Worbarrow Bay, Dorset.

Diagnosis. Shell small (up to 16 mm. long), elongate, ovate-trapezoidal, compressed,
not strongly inflated, produced posteriorly, umbo small, placed far forward. Anterior
extremity broadly rounded, forming a continuous curve with the convex ventral margin;
posterior-dorsal margin straight or feebly concave; posterior margin straight, inclined
strongly forwards, angular where it meets the dorsal and ventral margins. A sharp
carina extends backwards from umbo to postero-ventral angle, and another from umbo
to postero-dorsal angle, the area between them slightly concave. Surface with fine
concentric ridges which do not cross the umbonal carina to the posterior area. Fhnge
imperfectly known; right valve with a triangular cardinal tooth below the beak, margins
of the valve grooved.

This is the small lamellibranch (‘ Anthonya ’) which Arkell (\9Alb, p. 176) described
as a special feature of the Worbarrow ironstone (bed 7), correlated with the Punfield
Marine Band. Specimens are internal and external moulds with impressions of part of
the hinge. A possibly allied form occurs in the top of the Wealden Shales in the Isle of
Wight (e.g. GSM 98601). Resemblance to the scambulid Anthonya is very superficial.

Family pholadidae
Genus xylophagella Meek 1876

Xylophagella zonata sp. nov. (= Turnus sp., Woods 1909, p. 234, pi. 38, figs. 16, 17).
Holotype'. BM L 4996, the original of Woods’s pi. 38, figs. 16a, b. A wood-borer,
typically Gault, but found also in the mammillatum Zone.

Class gastropoda
Family pseudomelanidae
Genus brightonia nov.

(Mr. A. G. Brighton, Curator of the Geological Collections, Sedgwick Museum, Cambridge)

Type species. Brightonia turris gen. et sp. nov., Upper Aptian, southern England.

RAYMOND CASEY: STR ATIG RAPHIC AL PALAEONTOLOGY OF GREENSAND 591

Diagnosis. Tall, conical, many-whorled Pseudomelanidae; whorls concave, shouldered

at the sutures; base flat, angular or sharply turned at the periphery;
aperture trapezoidal; columella lip slightly arcuate, reflected. Surface
with fine spiral striations and crescent-shaped growth-lines.

!\

Brightonia turns gen. et sp. nov.

I t

1883 Nerinea sp.; Keeping, p. 94, pi. 3, figs, 7, la.

Holotvpe. The original of Keeping’s pi. 3, fig. 7, from the Lower Greensand of
Upware, Cambridgeshire.

Diagnosis. Brightonia 90-100 mm. long, with apical angle of about
18°. Whorls fifteen or more, concave, well shouldered; base angular,
keeled; spiral striations four to 1 mm. on final whorl.

In addition to Keeping’s originals a suite of partly crushed speci-
mens from Faringdon Folly (GSM Zk 4991-4) is now available. A
true Pseudomelania occurs in the Atherfield Clay Series and is prob-
ably that referred to by Keeping when describing this species.

Brightonia sandlingensis gen. et sp. nov.

Text-fig. 12

Holotype. GSM 98605, mammillatum Zone, Sandling Junction sandpit, near
Hythe, Kent (Author’s Coll.).

Diagnosis. Slender Brightonia 60-70 mm. long, with apical angle
about 12°. Whorls twelve or more, slightly concave, feebly shoul-
dered, the lower shoulder the stronger; base flat, narrowly rounded
at the periphery. Spiral striations five to 1 mm. on last whorl, produc-
ing a microscopic reticulation with the growth-lines.

The holotype is the only example with shell, though there are a
number of internal moulds which apparently belong to this species.

Class CEPHALOPODA
Order ammonoidea

I \

i 1

TEXT-FIG. 12.
Brightonia sand-
lingensis gen. et
sp. nov., holotype,
mammillatum bed,
Sandling Junction
sandpit, near
Hythe, Kent. GSM
98605, X 1-5.

Family desmoceratidae
Genus beudanticeras Hitzel 1905

Beudanticeras newtoni nom. nov. for Ammonites ( Desmoeeras ) beudanti var. Iigatus
Newton and Jukes-Browne (in Jukes-Browne 1900, p. 443) ( non Ammonites iigatus
d’Orbigny 1 841). It is not possible to locate any of Newton and Jukes-Browne’s originals
of this very common mammillatum Zone ammonite and Spath’s (1923c, p. 58) designa-
tion of a Tectotype’ of his own collecting is invalid. This specimen (BM C 28827,
figured Spath, ibid., pi. 3, figs. 3 a, b) is here designated neotype of A.(D) beudanti var.
iigatus Newton and Jukes-Browne and so becomes automatically the type of B. newtoni
nom. nov.

592 PALAEONTOLOGY, VOLUME 3

Family deshayesitidae
Genus prodeshayesites nov.

Type species. Ammonites fissicostatus Phillips 1829, p. 129, pi. 2, fig. 49, Speeton Clay, Yorkshire.

Diagnosis. Like Deshayesites but with flatter, loosely coiled whorls, and ventral ribs in
the form of chevrons; suture-line with relatively low elements.

text-fig. 13. Prodeshayesites obsoletus gen. et sp. nov., whorl-section and suture-line of holotype,
Pema Bed, Woodhatch, near Reigate, Surrey. BM C 36944, X 1.

Prodeshayesites obsoletus gen. et sp. nov.

Plate 82, figs. 2, 3 ; text-fig. 1 3

1845 Ammonites ? resembling A. leopoldimis d’Orbigny; Forbes, p. 355.

1847 Ammonites leopoldimis d’Orb.; Fitton, p. 296, table facing p. 289.

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 593

1887 Ammonites leopoldinus; Norman, pp. 29-30.

1889 Ammonites leopoldinus d’Orb. ; Bristow, p. 266.

1922 Parahoplites spp. n. cf. laeviusculus (v. Koenen); Butler, p. 316 (pars).

1923 c Parahoplitoides laeviusculus (v. Koenen); Spath, p. 66 (pars).

1930« Deshayesites aff. laeviusculus (v. Koenen); Spath, p. 434 (pars).

1933 Deshayesites aff. laeviusculus (v. Koenen); Chatwin (in Dines and Edmunds), p. 117.
Holotype. BM C 36944, Atherfield Clay Series, Perna Bed, Woodhatch, near Reigate, Surrey.

Diagnosis. More compressed than P. laeviusculus, the phragmocone more involute and
with mere traces of ribbing.

Characteristic of the Perna Bed and easily recognized by the smooth, desmoceratid-
like phragmocone.

Genus deshayesites Kasansky 1914
Deshayesites forbesi sp. nov.

Plate 81, figs, la, lb

1845 Ammonites Deshavesi Leymerie; Forbes, p. 354, pi. 21, fig. 2.

1930n Deshayesites deshavesi (Leymerie MS.) d’Orbigny sp.; Spath, p. 424 (pars).

Holotype. GSM 30918, Atherfield Clay Series, Crackers, Atherfield, Isle of Wight.

Diagnosis. Phragmocone resembling that of D. deshayesi but with an oblique umbilical
wall and with a more feebly ribbed nucleus that lacks the smooth flat ventral band of
that species. Ribs close up on last three-quarters of whorl so that rib-count for final
whorl is 50-60 compared with 45 in D. deshayesi.

This common species has been generally mistaken for D. deshayesi and has a long
synonymy. It ranges through the whole of the Atherfield Clay Series above the Perna
Bed, reaching its maximum in the callidiscus Subzone, and is eminently suited as a
zone fossil. The true deshayesi, as represented by d’Orbigny’s types from the Argiles
a Plicatules of Bailly-aux-Forges (Haute-Marne) (lectotype here selected : the original
of Wright 1957, p. L387, fig. 505), occurs in the lower part of Group IV of Atherfield
and in the base of the Hythe Beds.

Deshayesites fittoni sp. nov.

Plate 84, figs. 4 a, 4b

Holotype. GSM Zm 1843, Atherfield Clay, 25-30 feet above Perna Bed, Atherfield, Isle of Wight
(Author’s Coll.).

Diagnosis. Similar to D. latilobatus (Sinzow) (= Hoplites deshayesi Neumayr and
Uhlig, pi. 45, figs. 1, la, b) but smaller and with more strongly flexed ribbing.

Not uncommon in the middle part of the Atherfield Clay of the Isle of Wight, though
generally crushed or in body-chamber fragments only. Costation is irregular and
variable, some densely ribbed forms approaching D. weissi (e.g. GSM Zm 1688). The
fittoni Subzone is the most likely correlative of von Koenen’s weissi Zone of the German
succession.

594

PALAEONTOLOGY, VOLUME 3
Deshayesites callidiscus sp. nov.

Plate 80, fig. 10

1930a Deshayesites aff. latilobatus (Sinzow); Spath, p. 425.

1 930a Deshayesites kiliani Spath, p. 429 (pars).

Holotype. BM 48836, Atherfield Clay Series, Crackers, Atherfield, Isle of Wight.

Diagnosis. Like D. kiliani Spath, but with a more rectangular whorl-section and more
numerous ribs that do not form periodic bulges around the umbilicus.

Found in both the Crackers and the Upper Lobster Beds. Not common, though well
represented in the museums.

Genus dufrenoyia (Burckhardt MS.) Kilian and Reboul 1915
Dufrenoyia transitoria sp. nov.

Plate 83, figs. 3 a, 3 b

1930a Deshayesites aff. grandis Spath, p. 427, pi. 17, fig. 1.

1935 Deshayesites aff. laeviuscuhis (v. Koenen); Swinnerton, p. 31.

Holotype. BM C 29617, base of Carstone, Hunstanton, Norfolk.

Diagnosis. Similar to Deshayesites grandis Spath, but with the clavate ventral margins
of Dufrenoyia in the young and a greater tendency to smoothness in mid-life.

Characteristic of the lower half of the bowerbanki Zone and one of the commonest
ammonites in the ledges of Lower Crioceras Bed at the mouth of Whale Chine, Ather-
field. Found also as a remanie fossil in the Sutterby Marl.

Family douvilleiceratidae
Genus cheloniceras Hyatt 1903
Subgenus cheloniceras s.s.

Cheloniceras ( Cheloniceras ) parinodum sp. nov.

Plate 84, fig. 1 ; text-fig. 14a

Holotype. An example collected by Professor T. Matsumoto from the top of Group IV, Atherfield.

Diagnosis. Whorls suboctagonal, depressed-coronatiform, widest at the umbilical
tubercle. Umbilicus about 35 per cent, of the diameter, with high, steep wall, rounded
at the rim. Primary ribs pass straight up the flank, every third or fourth producing a
bifurcating secondary; tertiary ribs interposed irregularly in ones, twos, and threes
(mostly in twos) between the primaries and mostly ending at mid-flank, some reaching
the umbilical margin. All ribs in equal relief on the venter. Primary ribs bear umbilical
and lateral tubercles, represented by radially elongated nodes on the internal mould,
the two rows of nodes being of equal strength. Twenty-five ribs (9 primaries) to the half-
whorl at 60 mm. diameter. With further growth the ribs become increasingly thick,
close, and blunt, tubercles swell into prominent bullae, and umbilicus widens. Thirty-
eight ribs (15 primaries) at 120 mm. diameter. Lateral tubercle lost at about 180 mm.
diameter; costation eventually simplified to alternately long and short ribs.

This is the earliest species of Cheloniceras known in the Lower Greensand, its appear-
ance marking the base of the deshayesi Zone. In its evolute coiling, thick, simple ribbing

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 595

and equalization of the umbilical and lateral tubercles it shows affinities with Pro-
cheloniceras, an earlier development of the Douvilleiceratidae.

Subgenus epicheloniceras Casey 1952
Cheloniceras ( Epicheloniceras ) martinioides sp. nov.

Plate 84, figs. 2a, 2 6; text-figs. 14c/, 14<?

1847 Ammonites Martini ; Fitton, p. 307 (pars).

1930a Cheloniceras martini (d’Orbigny); Spath, p. 452 (pars).

Holotype. GSM 98603, Hythe Beds (Boughton Group), Skinner's Quarry, Boughton Mount, near
Maidstone, Kent (Author’s Coll.).

Diagnosis. Similar to Ch. ( E .) tschemyschewi (Sinzow) to 40 mm. diameter, after which
the costation becomes relatively coarser, with fewer tertiary ribs, and strong lateral
and ventral tubercles are retained longer.

A typical fossil of the Boughton Group and especially common in the Upper Crioceras
Beds of the Isle of Wight, which have provided many of the specimens of ‘ Ammonites
martini ’ in the museums and cited in the literature. D’Orbigny’s original A. martini,
from the Gargasian of south-east France, is now represented in the d’Orbigny Collection
in Paris by a few indeterminable nuclei and I have found no specimens or subsequent
illustrations that agree with the protographs. Furthermore, to my knowledge none of
the French Gargasian species to which the name martini could conceivably be linked
agrees with those of the German and British Provinces, wherein a ‘ martini Zone’ has
been employed since the time of von Strombeck (1861).

Cheloniceras ( Epicheloniceras ) debile sp. nov.

Plate 84, figs. 3a, 36; text-fig. 146

Holotype, GSM Zm 1952, Ferruginous Sands, Upper Crioceras Beds, below Walpen Chine, Chale
Bay, Isle of Wight (Author’s Coll.).

Diagnosis. An early tschemyschewi stage, with coronate-polygonal whorl-section,
ventral sulcus, prominent ventral and lateral tubercles, and groups of three to four
intermediary ribs, established at 20 mm. diameter. Thereafter ventral tubercles degener-
ate into ill-defined nodes, the venter becomes flat, and finally, at about 70 mm. diameter,
broadly rounded. Ribbing rather blunt. After 20-25 mm. diameter each group of inter-
mediary ribs differentiates into an anterior secondary, feebly bullate on the venter and
connected firmly to the lateral tubercle, and two or three tertiaries which mostly reach
to the umbilicus. Lateral tubercle lost between 70 and 90 mm. diameter; close, rounded
ribs then supervene, radial or slightly reclined on the flank, broadening a little as they
pass over the venter, and either bifurcating from the umbilical tubercle or ending freely
at the umbilical margin. Thirty-six ribs and ten lateral tubercles at 60 mm. diameter;
about forty-two ribs at 120 mm. diameter.

This species characterizes the middle and upper parts of the Upper Crioceras Beds
and is known in isolated examples from the base of the Upper Crioceras Beds and from
the Boughton Group of Kent. Its outstanding feature is early degeneration of the ventral
tubercles.

596 PALAEONTOLOGY, VOLUME 3

Cheloniceras ( Epicheloniceras ) gracile sp. nov.

Plate 81, figs. 1 a, 1 b\ text-fig. 14c

Holotype. GSM Zm 1953, Ferruginous Sands, Group IX (bed 35 a), Walpen High Cliff, Chale Bay,
Isle of Wight (Author’s Coll.).

Diagnosis. The tschernysehewi stage ends at 20-25 mm. diameter. Ventral and lateral
tubercles then disappear, whorl-section becomes subquadrate-depressed, venter flat, and
costation simplifies to low, closely spaced, rounded ribs, most prominent on the venter.
Ribs slightly arcuate on the flanks, straight on the venter; primaries connected to a
small umbilical tubercle; two or three intermediaries to every primary, but irregularly
distributed, some ending at mid-flank, others join the umbilical tubercle. Forty-five
ribs and fifteen tubercles at 45 mm. diameter. With further growth the venter becomes
broadly rounded, the ribbing finer, denser, and more subdued, degenerating at about
130 mm. diameter into low flattened bands with superimposed coarse striae. Ribbing
rejuvenates at very large diameters. Suture-line of normal Cheloniceras pattern, but
composed of unusually long and deeply dissected elements; median saddle in the princi-
pal lobe exceptionally slender and almost severed at the middle.

This densely ribbed species of Epicheloniceras is unlikely to be mistaken. It occurs
in the Boughton Group around Maidstone and is localized more precisely in the Isle
of Wight, where it forms part of the rich ammonite fauna of the Group IX nodules
(bed 35 a).

Family parahoplitidae
Subfamily parahoplitinae
Genus parahoplites Anthula 1899
Parahoplites cunningtoni sp. nov.

Plate 82, figs, la, lb

1850 Ammonites Nutfieldensis Sow.; Cunnington, p. 454 (pars).

1 930a Parahoplites aff. campichei (Pictet and Renevier); Spath, p. 439 (pars).

Holotype. OUM K 184, Iron Sands of Seend, Wiltshire (E. C. Davey Coll.).

EXPLANATION OF PLATE 83

Figures natural size unless otherwise stated.

Figs. 1, 2. Hypacanthoplites milletioides sp. nov. la, b, Phragmocone, Minerai de Bois-des-Loges,
Grandpre (Ardennes), France. (Museum National d’Histoire Naturelle, Paris.) 2, Holotype, crushed
body-chamber, milletioides Subzone, Sandling Junction, near Hythe, Kent. (GSM 70559; author’s
coll.) x0-5.

Figs. 3 a, b. Dufrenovia transitoria sp. nov., holotype, base of Carstone, Hunstanton, Norfolk. (BM
C 29617.)

Figs. 4a, b. Hoplites (Isohoplites) eodentatus sp. nov., holotype, Gault-Lower Greensand junction beds
(Band III), Arnold’s pit, Billington Crossing, Leighton Buzzard, Bedfordshire. (GSM 98602;
author’s coll.)

Fig. 5. Limatula sabulosa sp. nov., holotype, Folkestone Beds, Farnhamia horizon, Coxbridge pit,
Famham, Surrey. (GSM 98606; author’s coll.) X 5.

Figs. 6a-c. Modestella modesta E. Owen gen. et sp. nov., ventral (a), dorsal (b), and side (c) views
of holotype, Folkestone Beds, main mammillatum bed, Copt Point, Folkestone, Kent. (GSM
Zk 4733; author’s coll.) x 1-5.

Palaeontology, Vol. 3

PLATE 83

CASEY, Aptian and A Ibian fossils

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 597

Diagnosis. Similar to P. nutfieldensis (J. Sow.) but with narrower unbilicus, more tardy
compression of the whorls, primary ribs more closely set and with primary and secondary
ribs in regular alternation to at least 215 mm. diameter.

text-fig. 14. Whorl-sections and a suture-line of Cheloniceras. a, Ch. ( Ch .) parinodum sp. nov.,
holotype. b, Ch. ( Epicheloniceias ) debile sp. nov., holotype. c, Ch. ( E .) gracile sp. nov., holotype. d, e,
Ch. (E.) martinioides sp. nov., topotype (GSM Zm 1728). a-d xl,f x 2.

Especially characteristic of Seend, there being several well-preserved examples in
the Geological Survey Museum, British Museum, and the Oxford University Museum,
all contributed by W. Cunnington and E. C. Davey. A fragment from the Puttenham
Beds of Headleywood Farm, north of Headley, Hampshire.

598 PALAEONTOLOGY, VOLUME 3

Subfamily acanthohoplitinae
Genus nolaniceras nov.

Type species. Hoplites nolani Seunes 1887, p. 564, pi. 13, figs. 4 a, b, Clansayes horizon, France.

Diagnosis. Similar to compressed Hypacanthoplites, with close, flexuous ribbing, but
venter rounded, with only a trace of flattening in the young, and no tubercles at the
margins at any stage. Lateral tubercles represented only by microscopic pustules in
early youth. Suture-line as in Hypacanthoplites.

Nolaniceras is an important horizon-marker for the base of the jacobi Zone ( nolani
Subzone) and comprises a group of species that lie in the direct line of ancestry of the
compressed, closely ribbed Hypacanthoplites of the overlying rubricosus Subzone. It is
represented in Russia by Parahoplites uhligi Ant hula and Acanthohoplites bigot i Sinzow
(1907, pi. 4, figs. 19, 20; non Seunes); in Algeria by Parahoplites ouenzaensis Breistroffer
and P. (?) rigidus Breistroffer; and in Madagascar by forms described by Collignon
(1937) under the following names: Parahoplites cf. grossouvrei Jacob, P. aff. grossouvrei,
P. hourcqi Collignon, Acanthoplites nolani var. pygmaea Sinzow, A. nolani var. sub-
rectangulata Sinzow. Immunitoceras Stoyanow is an allied genus with distinctly flat,
angular venter in youth and may not be separable from the early Hypacanthoplites of
the rubricosus-subrectangulatus group.

Genus hypacanthoplites Spath 1923
Hypacanthoplites milletioides sp. nov.

Plate 83, figs. 1, 2

1875 Ammonites Milletianus d’Orb.; Barrois, p. 243 (pars).

Holotype. GSM 70559, Folkestone Beds (milletioides Subzone), Sandling Junction, near Flythe, Kent
(Author's Coll.).

Diagnosis. Aspect of H. trivial is Breistroffer, but larger, with coarser, less rigid ribbing.
About forty-five ribs at 75 mm. diameter.

I am indebted to Dr. P. Destombes for communicating an assemblage of Hypacant-
hoplites from the Bois-des-Loges horizon, described by Barrois (1875), and which also
furnished the type of H. peroni (Jacob). These show that d’Orbigny was right to assign

EXPLANATION OF PLATE 84

All figures natural size.

Fig. 1. Cheloniceras ( Cheloniceras ) parinodum sp. nov., holotype, Ferruginous Sands, top of Group IV,
Atherfield, Isle of Wight. T. Matsumoto coll.

Figs. 2a, b. Cheloniceras ( Epicheloniceras ) martinioides sp. nov., holotype, Hythe Beds, Boughton
Group, Skinner’s Quarry, Boughton Mount, Maidstone, Kent. (GSM 98603; author's coll.)

Figs. 3a, b. Cheloniceras (Epicheloniceras) debile sp. nov., nucleus of holotype, Upper Crioceras Beds,
Atherfield, Isle of Wight. (GSM Zm 1952; author’s coll.)

Figs. 4a, b. Deshayesites fittoni sp. nov., holotype, Atherfield Clay, 25-30 feet above Perna Bed,
Atherfield, Isle of Wight. (GSM Zm 1843; author's coll.)

Figs. 5a, b. Deshavesites forbesi sp. nov., topotype, Atherfield Clay Series (Crackers), Atherfield,
Isle of Wight. (SM B 27067.)

Figs. 6, 7. Cleoniceras (Cleoniceras) floridum sp. nov., Folkestone Beds, main mammillatum bed,
Copt Point, Folkestone, Kent. Both author’s coll. 6, Side view of holotype. (GSM 70401.) 7, Ventral
view of topotype. (GSM Zk 4866.)

Palaeontology, Vol. 3

PLATE 84

CASEY, Lower Greensand ammonites

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 599

the fossils of this horizon to the Albian rather than to the Aptian as did later authors
(e.g. Barrois 1875; Corroy 1925; Breistroffer 1947). A comparable assemblage is found
in the Folkestone Beds in the middle of the tardefureata Zone or as derived fossils in the
mammillatum Zone.

Family hoplitidae
Subfamily hoplitinae
Genus hoplites Neumayr 1875
Subgenus isohoplites Casey 1952
Hoplites ( Isohoplites ) eodentatus sp. nov.

Plate 83, figs. 4 a, 4 b

Holotype. GSM 98602, Lower Greensand/Gault junction Beds, Band III, Arnold's pit, Billington
Crossing, Leighton Buzzard, Bedfordshire (Author’s Coll.).

Diagnosis. Similar to H. dentatus (J. Sow.) in side view but with the ventral aspect of
H. (I.) steinmanni (Jacob).

Together with its varieties and allies, this species occurs at the base of the Gault at
Folkestone, Chislet Colliery, Westerham, Kent; Reigate, Surrey; Leighton Buzzard,
Bedfordshire; Bonchurch, Isle of Wight; and also in the Pay-du-Bray, northern France.
Though rare, it is the ammonite best suited as guide fossil for the base of the Middle
Albian in the Anglo-French Province.

Genus anahoplitoides nov.

Type species. Saynella splendens (J. Sowerby) var. gigas Sinzow (1915, p. 20) (= Leymeriella revili
Jacob, Sinzow 1909, pi. 1, figs. 1-4), Lower Albian, mammillatum Zone, Mangyshlak, Russia.

Diagnosis. Inner whorls like a costate Anahoplites but with the ventral tubercle paired
as in Isohoplites ; outer whorls like those of Farnhamia.

This genus is represented in Britain by a unique specimen from th tfloridum Subzone
of Oxted, Surrey, Anahoplitoides cf. gigas (Sinzow), recorded by Wright and Wright
(1948, p. 85) as Anahoplites sp. nov.

Subfamily cleoniceratinae
Genus cleoniceras Parona and Bonarelli 1896
Subgenus cleoniceras s.s.

Cleoniceras ( Cleoniceras ) floridum sp. nov.

Plate 84, figs. 6, 7

1936 Cleoniceras cf. quercifolium (d'Orb.); Casey, p. 446.

1941 Cleoniceras aff. cleon (d’Orb.); Brown, p. 10.

1942 Cleoniceras aff. quercifolium (d’Orbigny); Spath, p. 674.

1943 Cleoniceras aff. quercifolium (d’Orbigny); Spath, p. 736.

1947 Cleoniceras aff. quercifolium ; Breistroffer, p. 25.

1948 Cleoniceras sp. nov.; Wright and Wright, p. 85.

Holotype. GSM 70401, Folkestone Beds, main mammillatum bed, Copt Point, Folkestone, Kent
(Author’s Coll.).

600

PALAEONTOLOGY, VOLUME 3

Diagnosis. Phragmocone high, narrowly arched, sides gently convex, widest at the
inner third. Umbilicus nearly one-fifth diameter, with low, perpendicular wall, narrowly
rounded at the rim. Ten to eleven low, droplet-shaped bullae surround the umbilicus,
from each of which radiate fan-wise a bundle of about four falcoid ribs. Ribs form
sharp crescents on outer half of sides, but obscured at middle of sides and on siphonal
line by zones of incipient smoothness. Body-chamber with scaphitoid tendency; sub-
rectangular in section, with broadly rounded venter and coarsened ribbing which forms
chevrons on the venter. Suture-line with shallow, asymmetrical first lateral lobe.

Well represented at Folkestone, Westerham, and Oxted; characteristic of an horizon
intermediate between the kitchini and raulinianus Subzones of the mammillatum Zone.

PROBLEMATICA

Genus hallimondia nov.

Type species. Hallimondia fasciculata gen. et sp. nov., Upper Aptian (Sandgate Beds), Surrey.

Diagnosis. Bundles of split tubes or troughs apparently growing upwards from a common
origin ; each trough about J inch in diameter and with walls up to inch thick composed
of concentric wavy laminae of a weakly birefringent material, probably originally car-
bonate. Habitat marine.

Hallimondia fasciculata gen. et sp. nov.

Plate 79, fig. 3

Holotvpe. GSM Zk 3960-61, Sandgate Beds, Cockley Quarry, Nutfield, Surrey (A. G. Davis Coll.).

Diagnosis. Troughs averaging 12 mm. in diameter, roughly semicircular in cross-section,
and 40 cm. or more in length. Straight, approximate and parallel throughout most of
their length, but diverging slightly at the distal end and curving under at the proximal
end and apparently tapering to a common origin, like a bunch of bananas.

Briefly described by Dines and Hallimond (in Dines and Edmunds 1933, pp. 63, 64),
this organism is not uncommon in the limestones associated with the fuller’s earth
seams in the Nutfield quarries, where it forms the nuclei of phosphatic nodules. The
structure resembles a sheaf of curled rushes but does not appear to be of plant origin.
Phosphatization has destroyed any possible clues that the original mineralization gave
to its systematic position.

Genus petromonile nov.

Type species. Siplionia (or Spongites) benstedii Bensted, Upper Aptian (Hythe Beds, Boughton Group),
Maidstone, Kent. Lectotype here selected: the original of Bensted 1862, pi. 18, fig. 3, in the Maidstone
Museum.

Diagnosis. Irregularly branched stems, averaging 10 mm. diameter, periodically lobed
so as to resemble a string of beads.

These curious structures were described by Bensted (1862, pp. 335-6, pi. 17, 18) as
sponges, partly, it seems, because spicules were seen to be concentrated in parts of the
organism. Examination of a set of specimens in the Maidstone Museum, including
the originals of some of Bensted’s figures, shows a random distribution of spicules in
both organism and matrix and the sponge affinities must be considered doubtful. A
fucoid origin is a possibility.

RAYMOND CASEY: STR ATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 601

Conclusion of the descriptive part of this paper brings us to the point where we may
take stock of the fauna and flora of the Lower Greensand and attempt to set it out on a
zonal basis. Material for the following lists has been brought together from many
sources, but their nucleus is the collections in the British Museum (Natural History), the
Geological Survey Museum, and the Sedgwick Museum, Cambridge. In addition I have
worked over the material in a score of provincial museums and private collections: the
most important in the former category is the Museum of Isle of Wight Geology,
Sandown; in the latter category, the collection of the Wright brothers. The results of
my own personal field-work are incorporated in the collections of the Geological Survey
Museum. In matters of nomenclature I have undertaken original revisionary work
only in the Mollusca; the rest has been culled from the literature, British and foreign.
Microzoa are not listed. No work has been done on foraminifera and ostracoda since
Chapman (1894) and Wright (1905) and the mere repetition of their lists (which include
many names of Recent species) without critical examination of the originals is felt to
be pointless. Many species are here recorded from the Lower Greensand for the first
time: doubtless there are as many omissions. The following abbreviations are used for
the zones:

FAUNAL AND FLORAL LISTS

M = martinioides
B = bowerbanki
D = deshayesi
f = forbesi

m = mammillatum
T = tardefurcata
J = jacobi
N = nutfieldensis

F = fissicostatus

PLANTAE
ClaSS ALGAE

Sub-daSS CONIFERALES

Bennettites gibsonianus Carruthers J.

— allchini Stopes ?M.

— maximus Carruthers J.

— - inclusus Carruthers ?F.

Cycadeoidea yatesi Carruthers (= Yatesia morrisi

■ — buzzardensis Stopes ?T.
Cycadeostrobus walkeri Carruthers ?F.
Colymbetes edwardsi Stopes ?

Protop ter is fibrosa (Stenzel) M.

Girvanella intermedia (Wethered) B, M.

Class FUNGI

Unidentified ascomycetes parasitic in coniferous
wood T.

Tempskya erosa (Stokes, Webb & Mantell)

Class FILICINAE

Weichselia reticulata (Stokes & Webb) (= Lon-
chopteris mantelli Brongniart) D-N.

Carruthers) ?T.

(= T. schimperi Auctt.) ?F.

ClaSS GYMNOSPERMAE
Sub-class CYCADOPHYTA

Sequoia giganteoides Stopes J.
Protopiceoxylon edwardsi Stopes J.
Pityoxylon sewardii Stopes ?J, ?T.

— yvoodwardi Stopes ?N.

— sp. B, M.

Pseudoaraucaria benstedii (Stopes) M.
Pityostrobus sussexiensis (Mantell) J.

— benstedi (Mantell) M.

— patens (Carruthers) M.

— cylindroides (Gardner) ?F, ?N.

— pottoniensis (Gardner) ?F, ?N.

— jacksoni Creber J.

Kaidacarpum minus Carruthers ?F, ?N.
Cedrostrobus leckenbyi (Carruthers) J.

— mantelli (Carruthers) M.

Cedroxylon maidstonense Stopes M.

— pottoniense Stopes ?F, ?N.

Abietites cf. solmsi (Seward) M.
Cupressinoxylon vectense Barber J.

• — - luccombense (Stopes) J.

— cryptomerioides Stopes M.

— hortii Stopes N.

— sp. ?N.

Taxoxylon anglicum Stopes ?N.
Podocarpoxylon woburnense Stopes ?N.

— bedfordense Stopes ?N.

602

PALAEONTOLOGY, VOLUME 3

Podocarpoxylon gothani Stopes J.

— solmsi Stopes J.

Vectia luccombensis Stopes J.
Unidentified coniferous wood F-m.

Class ANGIOSPERMAE

Cantia arborescens Stopes ?J, ?T.
Woburnia porosa Stopes ?N.

Sabulia scottii Stopes ?N.

Hythia elgari Stopes ?M.

Aptiana radiata Stopes ?J.

INVERTEBRATA

Phylum porifera
Class DEMOSPONGIA

Mastosia neocomiensis Hinde B, M.
Chenendopora sp. B, M.

Undetermined Hallirhoidae B, M. T.
Dorydenna cf. benetti Hinde T.

— sp. B.

Stelletta cf. inclusa Hinde J, T.

Reniera gracilis Hinde B, M.

— zitteli Pocta B, M.

Axinella gracilis Hinde B, M.

— dispersa Hinde B, M.

— stylus Hinde B, M.

Spirastrella neocomiensis Hinde M.
Monilites haldonensis Carter M.
Dirrhopalum neocomiensis Hinde B, M.
Geodites carteri Hinde B, M. T, m.

— robustus Hinde B, M, T, m.

— audax Hinde B, M.

— obtusus Hinde T, m.

— politus Hinde B, M, T.

• — - pusillus Hinde B, M.

— haldonensis Carter T.

— divergens Hinde B, M.

— wrighti Hinde B, M, ?N.

— planus Hinde B, M.

Stellettites ? T.

Tethyopsis haldonensis Carter B, M, T.
Pachastrella quadriradiata Carter M.
Cliona cf. cretacea (Portlock) D-m.

— cf. microtuberum Stephenson J-m.

— cf. retiformis Stephenson J-m.

— sp.nov.l J.

Class HEXACTINELLIDA

Plocoscyphia pertusa Geinitz N.

— cf. labrosa (T. Smith) T.

— cf. fenestrata (T. Smith) T.

— sp. J, T.

Stauractinella sp. ? M.

Tremabolites sp. T, m.

Class CALCAREA

Peronidella ramosa (Roemer) N.

— gillieroni (de Loriol) N.

• — ■ prol if era (Hinde) N.

Bcirroisia anastomosans (Mantell) N.

— irregularis (Hinde) N.

— clavata (Keeping) N.

Elasmocoelia crassa (Fromentel) N.

— mantelli Hinde N.

Corynella for amino sa (Goldfuss) N.

Synopella pulvinaria (Goldfuss) N.

Oculospongia dilatata (Roemer) N.
Raphidonema contortion Hinde N.

— porcatum (Sharpe) N.

— pustulatum Hinde N.

— macropora (Sharpe) N.

— farringdonense (Sharpe) N.

— sp. T.

Phylum COELENTERATA
Class HYDROZOA

Lonsda contortuplicata (Lonsdale) F.

Burgundia sp. N.

Class ANTHOZOA
Oculina hobleyi Thomas B.

Holocystis elegans (Lonsdale) F.

Turbinoseris defromenteli Duncan F.
Astrocoenia sp. N.

Isastraea morrisi Duncan N.

— sp. F.

Thamnasteria sp. F.

Placosmilia neocomiensis (de Fromentel) F.
Smilotrochus austeni Edwards & Haime F.
Discocyathus orbignyanus (Edwards & Haime) F.
— fittoni (Edwards & Haime) m.

Trochocyathus meyeri Duncan N.

— conulus (Phillips) m.

— cf. harveyanus Edwards & Haime m.

Phylum ECHINODERMATA
Class CRINOIDEA
Isocrinus fittoni (Austen) T.

— - sp. f-J.

Torynocrinus rugosus Seeley T.

ClaSS OPHIUROIDEA
Ophiurites sp. T.

Class ASTEROIDEA

Lophidiaster ornatus Spencer T.

— sp. f.

Comptonia sp. f.

Gen. et sp. indet. D, B, N, J, T.

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 603

Class ECHINOIDEA

‘ Cidaris ’ faring donensis Wright N.

- — coxwellensis Hawkins N.

— sp. T, m

Tetragramma rotulare (Agassiz) N.

■ — malbosi (Agassiz & Desor) D, B.

Trochotiara fittoni (Wright) f, N.

Polydiadema cf. will shire i (Wright) T, m.

Sale nia rugosa d’Archiac T.

— hieroglyphica Keeping N.

- — prestensis Desor N.

Hyposalenia wright i (Desor) F-N.

— lardyi (Desor) N.

— stellulata (Agassiz) N.

— studeri (Agassiz) T.

— sp. F.

Goniophorus lorioli Lambert & Thiery T.
Discoidea decorata Desor D, B.

— cf. subuculus Leske T, m.

Pyrina desmoulinsii d’Archiac T.

Conulopyrina anomala Hawkins T.

Catopygus vectensis Wright N.

— columbarius (Lamarck) N-T.

— switensis Desor B.

Goniopygus delphinensis Gras N.

Plagiochasma faringdonense (Wright) N.

■ — coxwellense Melville N.

Toxaster murchisonianus (Mantell) T, m.

— complanatus Agassiz F.

— ( Pliotoxaster ) fittoni (Forbes) F-N.

— ( — •) renevieri (Wright) N.

Phyilobrissus fittoni (Wright) M, N.

— artesianus Hawkins T, m.

— circeleti (Desor) m.

Nucleolites lacunosus Goldfuss T.

Hemiaster sp. T.

Holaster benstedi (Forbes) B, M.

■ — ■ wrighti Lambert N.

— (Labrotaxis) cantianus Casey J-m.

Phylum ANNELIDA

Serpula antiquata J. de C. Sowerby D-m.

— filiformis J. de C. Sowerby F-m.

— articidata J. de C. Sowerby T.

• — plexus J. de C. Sowerby M, m.

— gordialis (Schlotheim) m.

— cf. adnata Wade T.

- — - spp. F-m.

Rotidaria po/ygonalis (J. de C. Sowerby) D-N.

— concava (J. Sowerby) N-m.

‘ Terebelkd sp. T.

Phylum polyzoa

Stomatopora calypso d’Orbigny N, J.

Cellulipora spissa (Gregory) N.

Proboscina crassa (Roemer) N.

- — • — var. divaricata (d’Orbigny) N.

— radiolitorum d’Orbigny N.

■ — • ricordeauana d’Orbigny N.

— virgula d’Orbigny N.

— depressa d’Orbigny N.

— - coarctata Canu & Bassler N.

— ziczac d’Orbigny N.

■ — faringdonensis Canu & Bassler N.

- — filifera Canu & Bassler N.

— grandipora Canu & Bassler N.

- — parvida Canu & Bassler N.

— pulchella de Loriol N.

— ( Reptomultisparsa ) tenella (de Loriol)

Clinopora quadripartita Canu & Bassler
Heteropora nummularia Canu & Bassler

— keepingi Gregory N.

— clavata Kade N.

— michelini (d'Orbigny) ?N, T.

■ — • buskana (de Loriol) N.

Multicrescis wammillosa Canu & Bassler N.
Seminodicrescis nodosa d’Orbigny N.

Ceriopora farringdonensis Gregory N.

— collis (d’Orbigny) N.

— confusa (de Loriol) N.

— ranmlosa (Michelin) m.

■ — dimorphocella Canu & Bassler N.

— spongioides Canu & Bassler N.
Reptomulticava fungiformis Gregory B, N.

— lobosa (Keeping) N.

— nodosa (Keeping) N.

Neuropora micropora Canu & Bassler N.

— tenuinervosa Canu & Bassler N.

Neuroporella hemispherica Canu & Bassler N.
Microecia cornucopia (d’Orbigny) N.
Trigonoecia haimeana (de Loriol) N.

Cardioecia faringdonensis Canu & Bassler N.

— pauper Canu & Bassler N.

Notoplagioecia faringdonensis Canu & Bassler N.
Cea granulata Canu & Bassler N.

Diaperoecia (?) simplex Canu & Bassler N.

— orbifera Canu & Bassler N.

Plethopora aptensis Canu & Bassler N.
Multigalea canui (Gregory) N.

— marginata Canu & Bassler N.

Tholopora virgulosa (Gregory) N.

— - colligata (Gregory) N.

— thomasi Pitt N.

Radiopora tuberculata (d’Orbigny) N.

— neocomiensis (d’Orbigny) N.

Lobosoecia semiclausa (Michelin) N.
Meliceritites haimeana (d’Orbigny) N, J, T.

— transversa Canu & Bassler N.

— cunningtoni (Gregory) N.

604

PALAEONTOLOGY, VOLUME 3

Meliceritites semiclausa Gregory N.

— (?) upwarensis Keeping N.

Chisma furcillata Lonsdale f-N.

Choristopetalum impar Lonsdale F-J.

Clausa cranei Canu & Bassler N.

— zonifera Canu & Bassler N.

Reptoclausa denticulata Canu & Bassler N.

— • hagenowi (Sharpe) N.

Tretocycloecia ( ?) multiporosa Canu & Bassler N.
Berenicea gracilis (Milne-Edwards) F.

— densa Canu & Bassler N.

Laterocavea dutempleana d'Orbigny N.

- — intermedia Canu & Bassler N.

Petalopora cunningtoni Gregory N, J.
Siphodictyon gracile Lonsdale F, D-T.

— - ir regal are Canu & Bassler N.

Sparsicavea irregularis d’Orbigny N.
Homoesolen pinnatus (Roemer) m.

Echinocava raidini (Michelin) N, m.

Inversaria orbicularis Gregory T.

Zonatula brydonei Gregory N.

Multizonopora arborea (Kock & Dunker) N.
Discocavea cf. neocomiensis d'Orbigny N.
Semimulticavea variolata Gregory N.

Graysonia anglica sp. nov. T.

Phylum BRACHIOPODA
Class INARTICULATA

Lingula truncata J. de C. Sowerby D-J.

— sp. T.

Discinisca sp. nov. B.

Bifolium faringdonense (Davidson) N.

Class ARTICULATA

Rhombothyris extensa (Meyer) N.

— microtrema (Walker) N.

— meyeri ( Walker) N.

— conica Middlemiss N.

Platythyris comptonensis Middlemiss M, N.

— minor Middlemiss N.

Sellithyris sella (J. de C. Sowerby) F-M.

shanklinensis Middlemiss N.

- — upwarensis (Walker) N.

— - coxwellensis Middlemiss N.

Cyrtothyris cyrta (Walker) ?F, B-N.

— uniplicata (Walker) B-N.

— - cantabridgiensis (Walker) ?B, N.

■ — - seeley i (Walker) N.

— - dallasi (Walker) N.

Praelongithyris praelongiforma Middlemiss ?B,
M, N.

— lankesteri (Walker) N.

Rectithyris depressa (Lamarck) T.

Rectithyris shenleyensis (Lamplugh & Walker) T.
‘ Terebratula ’ capillata d' Archiac T.

— dutempleana d'Orbigny T, m.

— gigantea Lamplugh & Walker T.

• — moutoniana d’Orbigny var. T.

• — boubei d’Archiac T.

— ovata J. Sowerby T.

‘ Ornithella ’ juddi (Walker) N.

— tamarindus (J. de C. Sowerby) N.

— cf. tamarindus (J. de C. Sowerby) T.

- — pseudojurensis (Auctt. non Leymerie sp.) N, T.

— wanklyni (Walker) N.

— morrisi (Meyer) N.

■ — celtica (Morris) B-N.

Aulacothyris woodwardi (Walker) N.

Zeilleria convexiformis Lamplugh & Walker T.
Modestella modesta Owen gen et. sp. nov. m.

— sp. nov. T.

Magas latistriata Lamplugh & Walker T.

— - orthiformis (d’Archiac) T.

‘ Terebratella ’ hercynica (Schloenbach) T.

- — ■ davidsoni Meyer N.

— keepingi Walker N.

Gemmarcula aurea Elliott M, N.

— menardi (Lamarck) T.

var. pterygotos (Lamplugh & Walker) T.

Arenaciarcula fittoni (Meyer) M. N.
Oblongarcula oblonga (J. de C. Sowerby) D-N.
Trifidarcula trifida (Meyer) N.

Terebrirostra arduennensis d’Orbigny (=7’. lyra
var. incurvirostrum Lamplugh & Walker) T.
Kingena lima (Defrance) T.

— arenosa (d’Archiac) T.

— newtoni Lamplugh & Walker T.

— spinulosa (Morris) m.

Terebratulina triangularis Etheridge T.

— elongata Davidson N.

Cyclothyris latissima (J. de C. Sowerby) N-T.
Sulcirhynchia hythensis Owen F-M.
Lamellirhynchia caseyi Owen N, J.

‘ Rhynchonella ’ gibbsiana (J. de C. Sowerby) T.

— leightonensis Lamplugh & Walker T, m.

— shenleyensis Lamplugh & Walker T.

— grasiana d’Orbigny T.

• — lineolata (Phillips) T.

— - carteri Davidson T.

— mirabilis Lamplugh & Walker T.

- — dimidiata (J. Sowerby) T.

— antidichotoma Buvignier N, T.

— depressa (J. de C. Sowerby) N.

- — parvirostris (J. de C. Sowerby) N.

- — cantabridgensis Davidson N.

— upwarensis Davidson N.

— deluci (Pictet) J.

• — nuciformis (J. de C. Sowerby) N.

RAYMOND CASEY: STRATIGRAPHIC AL PALAEONTOLOGY OF GREENSAND 605

Phylum mollusca
ClaSS LAMELLIBRANCHIA
Nuculana scapha (d'Orbigny) F, f.

— spathulata (Forbes) f.

— solea (d'Orbigny) m.

Mesosaccella mariae (d’Orbigny) T, m.

Nucula meyeri Gardner F, f, N.

- — ( Pectinucula ) pectinata J. Sowerby m.

— ( — ) arduennensis d’Orbigny J.

— ( Leionucula ) planata Deshayes F-B, J.

— - ( — ) albensis d’Orbigny m.

— ( — •) ovata Mantell m.

Acila ( Truncacila) bivirgata (J. de C. Sowerby) m.
Anomia pseudoradiata (d’Orbigny) F-J.

— laevigata J. de C. Sowerby f-M.

— convexa J. de C. Sowerby N.

— - sp. (Woods) f.

Area dupiniana d'Orbigny F-T.

— sanctae-crucis Pictet & Campiche F-N.
Eonavicula carteroni (d'Orbigny) F, N.

Barbatia marullensis (d'Orbigny) N, J.

- — cf. baudoniana (Cotteau) B.

Scaphula ? austeni (Forbes) F, f.

Nanonavis carinata (J. Sowerby) N-m.
Aptolinter aptiensis (Pictet & Campiche) F-T.
Cucullaea nana Leymerie m.

- — tealli nom. nov. ( Pectunculus obliquus
Keeping) N.

- — fittoni Pictet & Campiche f.

— cornueliana d'Orbigny f, B, N.

— ( Idonearca ) glabra Parkinson N-m.

— (• — ■) obesa Pictet & Campiche m.
Cryptochasma ovale gen. et sp. nov. N.
Noramya forbesi (Pictet & Campiche) F.

— gabrielis (Leymerie) F.

Isoarca obesa (d’Orbigny) m.

Glycymeris marullensis (Leymerie) B, N.

— ( Glycymerita ) sublaevis (J. de C. Sowerby)

N-T.

— ( — ) umbonata (J. Sowerby) T.

Limopsis albensis Woods J.

- — dolomitica sp. nov. N.

Trigonia carinata Agassiz F-B.

Pterotrigonia veetiana (Lycett) F.

— mantelli sp. nov. (s.s.) M-m.

anterior subsp. nov. D, B.

— caudata (Agassiz) F, f.

— - etheridgei (Lycett) F, f.

Linotrigonia ( Linotrigonia) fittoni (Deshayes) m.

— ( Oistotrigonia ) ornata (d’Orbigny) F, D-M.

— ( — ) upwarensis (Lycett) N, J.

■ — ( — ) archiaciana (d’Orbigny) m.

Yaadia nodosa (J. de C. Sowerby) F-J.
Myoconcha delta sp. nov. N.

Mytilus cf. tornacensis d’Archiac N.

Modiolus aequalis J. Sowerby F-m.

— reversus (J. de C. Sowerby) T, m.

— ligeriensis (d’Orbigny) F.

— subsimplex (d’Orbigny) F-m.

• — rugosus ( Roemer) f.

■ — undulatus ( Forbes) f.

Brachidontes vectiensis (Woods) F, f.

Arcoperna bella (J. de C. Sowerby) F-J.

Sept if er sublineatus (d’Orbigny) F, D-m.
Cuneolus lanceolatus (J. de C. Sowerby) F-m.
Spondylus roemeri Deshayes F.

— guttatus (Sharpe) N.

- — gibbosus d’Orbigny m.

— • striatus (J. Sowerby) N-T.

Plicatula placunea Lamarck D, B.

■ — carteroniana d'Orbigny B-N, T.

— aequicostata Keeping N.

Plicatula inaequidens Sharpe N.

- — gurgitis Pictet & Roux m.

— inflata J. de C. Sowerby T, m.

Diploschiza sp. J-m.

Entolium orbiculare (J. Sowerby) F-m.
Camptonectes cottaldinus (d’Orbigny) F, D, B.

— striato-punctatus (Roemer) F, T.

Chlamys elongata (Lamarck) m.

— robinaldina (d'Orbigny) F-m.

— subacuta (Lamarck) T.

Neithea ( Neitheops ) quinquecostata (J. Sowerby)
F-m.

• — ( — ) syriaca (Conrad) (= Pecten morrisi Pictet
& Renevier) F-B.

- — • ( — ) atava (Roemer) N.

Eopecten rhodani (Pictet & Roux) m.
Prohinnites favrinus (Pictet & Roux) F, D, B.
Plagiostoma globosa (J. de C. Sowerby) T, m.

— albensis (d’Orbigny) J, T, m.

— cf. orbignyana (Matheron) N.

Acesta longa (Roemer) N-T.

Pseudolimea parallela (J. de C. Sowerby) F,
D-m.

— gaultina (Woods) T, m.

- — farringdonensis (Sharpe) N.

— elongata (J. de C. Sowerby) m.

— cf. cantabrigiensis (Woods) J.

Ctenoides cf. rapa (d’Orbigny) T.

Limatula tombeckiana (d'Orbigny) D, B, N.

— dupiniana (d'Orbigny) F, D-N.

- — sabulosa sp. nov. T, m.

Oxytoma pectination (J. de C. Sowerby) D-m.

— cornuelianum (d'Orbigny) N.

Pseudoptera subdepressa (d’Orbigny) F, f.
Aucellina sp. T.

Gervillia linguloides Forbes f.

Bakevellia rostrata (J. de C. Sowerby) N, J.

R r

B 6612

606

PALAEONTOLOGY, VOLUME 3

Gervillella sublanceolata (d'Orbigny) F-J.
Gervillaria alaeformis (J. Sowerby) F-B.
Ensigervilleia forbesiana (d’Orbigny) F-J, m.
Mulletia mulleti (Deshayes) F.

Isognomon ricordeanus (d'Orbigny) F.

— raulinianus (d'Orbigny) m.

Inoceramus neocomiensis d'Orbigny D-N.

- — salomoni d'Orbigny m.

• — coptensis sp. nov. T, m.

Pinna {Pinna) robinaldina d’Orbigny F-J.

— (Stegoconcha) cf. iburgensis Weerth F.

— ( — ) cf. hombresi Pictet & Campiche f.

— (■ — ) cf. gervaisii Dumas D, B.

Astarte cantabrigiensis Woods N.

— ( Neocrassina ) obovata J. de C. Sowerby F, D.
Freiastarte subcostata (d'Orbigny) F, f, D.

- — praetypica sp. nov. J.

Eriphyla striata (J. de C. Sowerby) J.

— concinna (J. de C. Sowerby) J.

— up war en sis (Woods) N.

— pseudostriata (d’Orbigny) B, N.

Op is ( Trigonopis ) neocomiensis d’Orbigny N.

— ( — ) cf. dubisiensis Pictet & Campiche F.
Ptychomya robinaldina (d'Orbigny) F-N.
Seendia saxoneti (Pictet & Roux) N.

Disparilia disparilis (d’Orbigny) F.

Pachythaerus tealli sp. nov. N, T.

Anthonya cantiana Woods N-m.

— woodsi sp. nov. f.

Mediraon subacutum (d’Orbigny) F, f.

— sulcatum sp. nov. F, f.

Protodonax minutissimus (Whitfield) N.

Linearia subconcentrica (d'Orbigny) F, f.

— cornea sp. nov. f.

— cf. olea Vokes N.

— rauliniana (d’Orbigny) m.

Palaeomoera inaequalis (J. de C. Sowerby) J, T.
Scittila nasuta gen. et sp. nov. F, f.

— — var. radiata nov. f.

Icanotia pennula sp. nov. F.

— cf. siliqua sp. nov. J.

Senis wharburtoni (Forbes) F-m.

Geltena meyeri sp. nov. N.

Fenestricardita fenestrata (Forbes) F, f.
Trapezicardita squamosa (Keeping) N, J.

— arcadiformis (Keeping) N.

Pseudocardia cf. dupiniana (d’Orbigny) f.

— tenuicosta (J. de C. Sowerby) T, m.

— — var. constanti (d’Orbigny) m.

- — - — var. rotundata (Pictet & Roux) m.

— sp. nov. N, J.

Cardita upwarensis Woods N.

— sp. nov. (Woods, pi. xviii, fig. 6) F.

Lucina cornueliana d’Orbigny D-T.

— tenera (J. de C. Sowerby) J.

Lucina arduennensis d’Orbigny T.

Sphaera corrugata J. Sowerby F-N.

Mutiella canaliculata (J. de C. Sowerby) T.
Mactromya vectensis (Woods) F, f, N.

— compressa (Woods) f.

— ringmeriensis (Mantell) T, m.

Thetironia minor (J. de C. Sowerby) F-m.

— laevigata (J. Sowerby) m.

Protocardia anglica Woods f.

— sphaeroidea (Forbes) F.

Nemocardium (Pratulum) ibbetsoni (Forbes) F,
f, N.

Granocardium cf. proboscideum (J. de C. Sowerby)
F.

Cardium cottaldinum d’Orbigny N.

Eomiodon cf. libanoticus (Fraas) f.

Venilicardia saussuri (Brongniart).

— protensa (Woods) F, f.

- — anglica (Woods) f.

— sowerbyi (Woods) D-N.

— inornata (d’Orbigny) D, B.

Epicyprina harrisoni sp. nov. J, T.

Proveniella meyeri (Woods) F.

— regularis (d’Orbigny) N.

- — quadrata (d’Orbigny) m.

— ervyensis (d’Orbigny) m.

— rosacea sp. nov. F, ?f.

Vectianella vectiana (Forbes) f.

Isocyprina sedgwicki (Keeping) N.

‘ Cyprina ’ obtusa Keeping N.

Tortarctica similis (J. de C. Sowerby) J-m.

— cordiformis (d’Orbigny) m.

Lithophaga phosphatica (Keeping) N.

— oblonga (d’Orbigny) M.

— spp. N.

Resatrix {Resatrix) dolabra Casey F, f.

— ( — ) neocomiensis (d'Orbigny) F.

- — (■ — ) woodsi Casey f.

— ( — ) parva (J. Sowerby) f-N.

- — ( — ) hythensis Casey D, B, M.

— ( Dosiniopsella ) cantiana Casey J.

— ( — ) vibrayeana (d'Orbigny) m.

- — ( Vectorbis ) vectensis (Forbes) f.
Pseudaphrodina ricordeana (d’Orbigny) F-N.

— brongniartina (Leymerie) F, f.

— elongata Casey F.

Filosina cf. gregaria Casey f.

Toucasia lonsdalii (J. de C. Sowerby) N.
Gyropleura sp. N, T.

Parmicorbula striatula (J. de C. Sowerby) F-m.

— elegantula (d’Orbigny) f, J.

Caestocorbula olivae (Whitfield) f.

— ■ sp. (Woods pi. 34, fig. 13) f.

Cuneocorbula arkelli sp. nov. f.

Panopea gurgitis (Brongniart) F-m.

RAYMOND CASEY: STR ATIGR APHICAL PALAEONTOLOGY OF GREENSAND 607

Panopea gurgitis var. plicata J. de C. Sowerby
F-m.

■ — - — var. neocomiensis (Leymerie) F-B, m.

— mandibula (J. Sowerby) F-m.

• — cf. meyeri Woods D.

Martesia prisca (J. de C. Sowerby) B-m.
Opertochasma constrictum (Phillips) D-M, J, T.
Goniochasma dallasi (Walker) N.

Terebrimya gaultina (Woods) T, m.
Xylophagella zonata sp. nov. m.

Plectomya cinglica Woods f.

— marullensis (d’Orbigny) f.

Cercomya gurgitis Pictet & Campiche f.
Ceratomya (?) ornata (Forbes) ?F.

Thracia rotundata (J. de C. Sowerby) D, B.
Periplomya simplex (d'Orbigny) T, m.

— - robinaldina (d’Orbigny) F, f.

Pholadomya cornueliana (d'Orbigny) F, f.

- — gigantea (J. de C. Sowerby) F-D.

— martini (Forbes) F-N.

— fabrina d'Orbigny m.

Myopholas cf. semicostata (Agassiz) f, N.
Girardotia sp. N.

Goniomya archiaci (Pictet & Renevier) f.

— sp. J, T.

Lopha diluviana (Linne) F-m.

Ostrea ( Liostrea ) leymerii Leymerie F, D, m.

— - ( — ) cunabula Seeley T.

— - (• — ) vesicidaris Lamarck T, m.

— ( — ) walked Keeping N.

Gryphaeostrea canaliculata (J. Sowerby) N-m.
Exogyra latissima (Lamarck) (= E. sinuata

J. Sow. sp.) F-m.

- — ■ conica (J. de C. Sowerby).

— tuberculifera Koch & Dunker F-J.

- — ■ arduennensis (d'Orbigny) m.

ClaSS GASTROPODA
Acmaea formosa (Gardner) N.

— sp. T.

Helcion meyeri Gardner N.

Anisomyon vectis Gardner f.

Scurria calyptraeformis Gardner N.

■ — - depressa Gardner N.

Scurriopsis (Dietrichiella) plana (Gardner) F.
Loxotoma neocomiense (d’Orbigny) F, f, N.

— valangiense Pictet & Campiche F.

• — puncturellum Gardner F, f.

Brunonia gigantea (Gardner) B.

Hipponyx neocomiensis Gardner B.

Ringinella albensis (d’Orbigny) N.

— subalbensis (d’Orbigny) f.

— lacryma (Michelin) m.

— inflata (d’Orbigny) m.

Ovactaeonina forbesiana (d’Orbigny) F, f, D.

Snlcoactaeon marginatus (Deshayes) f.
TornateUaea aptiensis (Pictet & Campiche) F,
f,D.

— marullensis (d'Orbigny) F, f, D.

Actaeonella ( Trochactaeon ) cf. oliviformis

Coquand f.

— ( CylindriteUa ) cf. verneuilli (Vilanova) f.
Cinulia globosa (d'Orbigny) F.

Anchnra carinella (d’Orbigny) m.

— ( Perissoptera ) parkinsoni (Mantell) m.

■ — • ( — ) cf. parkinsoni (Mantell) N, J.

- — ( — -) glabra (Forbes) F, f.

— ( — ) robinaldina (d'Orbigny) F-N.

• — • ( — ) spartacus (Coquand) f.

Dimorphosoma cal carat um (J. de C. Sowerby) m.

— ancylochila (Gardner) f.

— - kinclispira (Gardner) f.

— vectianum (Gardner) f.

— pleurospira (Gardner) F.

Tessarolax moreausianum (d'Orbigny) F, f, N.

- — fittoni (Forbes) f.

— becklesii (Mantell) f.

— gardneri (Keeping) N.

— retusum (J. de C. Sowerby) m.

Tridactylus walker i (Gardner) N.

Pterocel/a macrostoma (J. de C. Sowerby) F.
Arrhoges (Monocuphus) dupinianus (d’Orbigny)

F.

Tylostoma rochatianum (d’Orbigny) F, f.

— fittoni (d’Orbigny) f.

— gaultinum Pictet & Campiche m.

Confusiscala dupininiana (d’Orbigny) T, m.

— cruciana (Pictet & Campiche) F.

— ischyra (Gardner) N.

Claviscala Clementina (Michelin) T, m.

— canaliculata (d'Orbigny) F.

— ricordeana (d'Orbigny) N.

‘ Scalar ia ’ keepingi Gardner N.

— kalospira Gardner ?N.

- — - meyeri Gardner F.

— • cerithioides Gardner N.

Cassiope lujani (de Verneuil) f.

- — • var. crassa Coquand f.

— helvetica (Pictet & Renevier) f.

- — pizcuetana (Vilanova) f.

— — - var. cf. renevieri Coquand f.

Uchauxia forbesiana (d’Orbigny) (= Cerithium
plu/lipsi Forbes non Leymerie) F, f.

- — albensis (d’Orbigny) F.

Mesalia ( Bathraspira ) tecta (d’Orbigny) m.

— ( — ) neocomiensis (d’Orbigny) F, f, M, N.
Metacerithium trimonUe (Michelin) m.

— campichei Cossmann f.

Gymnocerithium nostradami (Coquand) f.

— tumidum (Keeping) N.

608

PALAEONTOLOGY, VOLUME 3

Gymnocerithium marollinum (d’Orbigny) N.
Atresius lallierianus (d’Orbigny) m.

— fittoni (d'Orbigny) F, f, N.

Circocerithium aptiense (d’Orbigny) D.
Nerineopsis excavata (Buvignier) m.

‘ Cerithium ’ aculeatum Forbes F.

— subattenuatum d’Orbigny (= C. attenuatum

Forbes non J. de C. Sowerby sp.) f.

— turriculatum Forbes f.

- — clementinum d’Orbigny f.

Pseudomelania melanoides (Deshayes) N.

— sp. nov. f.

Brightonia turns gen. et sp. nov. N.

— sandlingensis gen. et sp. nov. m.

Eulima albensis d’Orbigny F.

Columbellina neocomiensis (d'Orbigny) F.

— sp. m.

Neptunella cf. espaillaci (d’Orbigny) T.
Buccinofusus clementinus (d'Orbigny) m.

Sip/io gaultinus (d’Orbigny) T, m.

Globularia cornueliana (d’Orbigny) F, f, N.

— sublaevigata (d'Orbigny) F, f.

• — arduennensis (d'Orbigny) T, m.

Gyrodes genti (J. Sowerby) N-m.

‘ Natica' pradoana Vilanova f.

Neridomus cf. alcibari (Coquand) f.

Nerita sp. nov. N.

Neritopsis robineausiana (d’Orbigny) N.

— sp. f.

Turritella ( Haustator ) vibrayeana d’Orbigny m.

— ( — ) rauliniana d’Orbigny m.

- — (■ — ) dupiniana d’Orbigny f, N.

— ( — ) sp. nov. J.

Nerinea sp. nov. N.

Pleurotomaria anstedi Forbes M, N.

— toulmini Cox D, B.

— shenleyensis Cox T.

Leptomaria gibbsi (J. Sowerby) m.

— billingtonensis Cox.

Bathrotomaria leightonensis Cox T.

— atherfieldensis Cox.

— ferruginea (Keeping) N.

Conotomaria gigantea (J. de C. Sowerby) D, B.

— seendensis Cox N.

— lamplughi Cox T.

Cimolithium astierianum (d'Orbigny) m.
Semisolarium moniliferum (Michelin) m.

— astierianum (d'Orbigny) T, m.

Proconulus esqueriae (de Verneuil & de Loriol) f.
Metriomphalus mantelli (Leymerie) F.
Chilodonta ( Agathodonta ) dentigera (d'Orbigny)

M.

Eucycloscala mulled (d'Archiac) T.

Tectus cf. huoti (d'Archiac) T.

‘ Littorina' conica (J. de C. Sowerby) N.

Eucyclus upwarensis (Keeping) N.

■ — ■ sp. nov. cf. albo-aptiensis (Sinzow) m.
Ooliticia cantabrigensis (Keeping) N.

— varicosa (Keeping) N.

— sp. nov. f.

Ataphrus albensis (d'Orbigny) F, f, D.
Delphinula cf. mailleana (d’Orbigny) F.
Nododelphinula reedi (Keeping) N.

Fossarus munitus (Forbes) F, f.

‘ Solarium ’ minimum Forbes F, f, D.
Margarites ( Atira ) mirabilis Casey J, T.

— (— ) sp. D, N.

Class SCAPHOPODA

Dentalium cylindricum J. Sowerby f, B, N.

— jeffreysi Gardner f.

- — ( Fissidentalium ) decussation J. Sowerby J-m
ClaSS CEPHALOPODA

Pictetia depressa (Pictet & Campiche) T.
Eogaudryceras shimizui Breistroffer T.
Ancyloceras mantelli Casey f.

— cf. various d’Orbigny ?F

Tropaeum bowerbanki J. de C. Sowerby B.
var. densistriatum Casey B.

— hill si (J. de C. Sowerby) D.

— drewi Casey B.

— benstedi Casey M.

— subarcticum Casey N.

Tropaeum keeping i Casey ?D.

— cf. rossicum Casey M.

Australiceras gigas (J. de C. Sowerby) D, B.

— sp. nov. D.

— sp. M.

Ammonitoceras tovilense Crick M.

— spp. nov. M.

Epancyloceras hythense Spath D.

— sp. nov. D.

Lithancylus grandis (J. de C. Sowerby) D.

— sp. nov. D.

Toxoceratoides royerianus (d’Orbigny) D.

- — proteus (Spath) D.

— cf. fustiformis (v. Koenen) D.

— cf. biplex (v. Koenen) f.

Tonohamites decurrens Spath D, B.

— aequicingulatus (v. Koenen) B.

— spp. nov. B.

Elamites praegibbosus Spath m.

— spp. nov. m.

Protanisoceras raidinianum (d’Orbigny) m.

— lardy i (Pictet & Renevier) m.

— cantianum Spath m.

— blanched (Pictet & Campiche) m.

— acteon (d’Orbigny) m.

RAYMOND CASEY: STR ATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 609

Protanisoceras vaucherianum (Pictet) m.

• — • cf. halleri (Pictet & Campiche) m.

— spp. nov. m.

" Prohelicoceras ’ anglicum Spath m.

Gen. nov. (‘ Protanisoceras ’) sp. nov. m.
Aconeceras nisoides (Sarasin) F, f, ?D, B.

■ — ■ cf. nisum (d’Orbigny) M.

— cf. haugi (Sarasin) f.

— sp. nov. T.

Sanmartinoceras ( Sinzovia ) aptianum (Sarasin) f.

— ? ( — ?) sp. nov. T.

— ( Theganeccras) cf. fa lea t tun (v. Koenen) ?F.
Gen. nov. (‘ Aconeceras ’) sp. nov. M.
Beudanticeras newtoni nom. nov. (= B. ligatum

Newton & Jukes-Browne sp., Spath) m.

— arduennense Breistroffer m.

■ — - dupinianum (d’Orbigny) m.

— laevigatum (J. de C. Sowerby) m.

Uhligella subornata Casey m.

Pseudo say nella cf. fimbriata Imlay f.

— afif. undulata (Sarasin) f.

Roloboceras hambrovi (Forbes) f.

— horridum (Spath) f.

— perli (Spath) f.

— spp. nov. f.

Megatylocercis sp. nov. f.

Cheloniceras ( Cheloniceras ) parinodum sp. nov.
D.

— ( — ) cornuelianum (d'Orbigny) D, B.

— ( — •) kiliani (v. Koenen) B.

— ( — ■) crassum Spath D, B.

— ( — ) meyendorffi (d’Orbigny) B.

— ( — ) cf. gottschei (Kilian) B.

— ( — ) spp. nov. D, B.

— ( Epicheloniceras ) tschernyschewi (Sinzow) M.

— ( — ) martinioides sp. nov. M.

— ( — ) debile sp. nov. M.

— ( — -) gracile sp. nov. M.

— ( — ) buxtorfi (Jacob) M.

— ( — •) aff. volgense (Vasilievsky) M.

— ( — ) spp. nov. M.

Gen. nov. (‘ Cheloniceras ’) sp. nov. M.
Douvilleicems mammillatum (Schlotheim) m,

— monile (J. Sowerby) m.

— orbignyi Flyatt m.

— spp. nov. m.

Deshayesites deshayesi (d’Orbigny) D.

— consobrinoides (Sinzow) D.

— multicostatus Swinnerton D.

— grandis Spath D.

— vectensis Spath D.

— involutus Spath D.

— kiliani Spath f.

— topleyi Spath f.

— punfieldensis Spath f.

Deshayesites forbesi sp. nov. f.

- — ■ fittoni sp. nov. f.

— callidiscus sp. nov. f.

— spp. nov. F, f, D.

Prodeshayesites obsoletus gen. et sp. nov. F.

— fissicostatus (Phillips) F.

— bodei (v. Koenen) F.

— laeviusculus (v. Koenen) F.

- — • spp. nov. F, f.

Dufrenoyia furcata (J. de C. Sowerby) B.

■ — • lurensis (Kilian) B.

- — • truncata Spath B.

— transitoria sp. nov. B.

■ — -spp. nov. B.

Parahoplites nutfieldensis (J. Sowerby) N.

— - maximus Sinzow N.

■ — • cunningtoni sp. nov. N.

— spp. nov. N.

Colombiceras sp. nov. cf. tobleri (Jacob) ?N.
Nolaniceras aff. nolani (Seunes) J.
Hypacanthoplites jacobi (Collet) J.

— - anglicus Casey J.

var. audax Casey J.

— simmsi (Forbes) J.

— rubricosus Casey J.

— — var. tenuiformis Casey J.

var. papillosus Casey J.

— elegans (Fritel) J.

— • cf. hanovrensis (Collet) J.

— clavatus (Fritel) J.

• — cf. sarasini (Collet) J.

— cf. laticostatus (Sinzow) J.

— cf. spat hi (Dutertre) J.

— trivialis Breistroffer T.

- — - milletioides sp. nov. T.

— cf. millet ianus (d'Orbigny) m.

— spp. nov. J, T.

Parengonoceras ebrayi (de Loriol) m.
Farnhamia farnhamensis Casey T.

— spp. nov. T.

Tetrahoplites cf. subquadratus (Sinzow) m.
Sonneratia dutempleana (d’Orbigny) m.

— kit chini Spath m.

— parenti Jacob m.

— perinflata Breistroffer m.

— spp. nov. m.

Pseudosonneratia spp. nov. m.

Protohoplites ( Protohoplites ) archiacianus

(d’Orbigny) m.

— ( — ) latisulcatus (Sinzow) m.

— ( — ) michelinianus (d’Orbigny) m.

— ( — ■) sp. nov. m.

— ( Hemisonneratia ) puzosianus (d’Orbigny) m.

— ( — ) gallicus (Breistroffer) m.

— ( — ) sp. nov. m.

610

PALAEONTOLOGY, VOLUME 3

Otohoplites raulinianus (d'Orbigny) m.

— elegans (Spath) m.

■ — guersanti (d'Orbigny) m.

— auritiformis (Spath) m.

■ — - spp. nov. m.

Anahoplitoides cf. gigas (Sinzow) m.
Anadesmoceras strangulation Casey T.

— subbay lei (Spath) T.

— baylei (Jacob) m.

— spp. nov. T.

Cleoniceras ( Cleoniceras ) cleon (d'Orbigny) in.

■ — ( — ) seunesi Bonarelli m.

— ( — ) quercifolium (d'Orbigny) m.

• — ( — ) janneli (Parent) m.

— ( — ) morgani Spath m.

— ( — ) floridum sp. nov. m.

- — ( — ) spp. nov. T, m.

- — ( Neosaynella ) inornatum Casey m.

— ( — ) sp. nov. m.

Levnieriella (Leymeriella) tardefurcata (d'Orbigny)
T.

— - ( — ) — var. intermedia Spath T.

— (■ — ) — var. densicostata Spath T.

— ( — ) regularis (Bruguiere) d’Orbigny sp. T.

— ( — ) consueta Casey T.

■ — ( — ) — var. magna Casey T.

■ — ( — ) pseudoregularis Seitz T.

— (■ — ) diabolus Casey T.

— ( — ) rudis Casey T.

— ( — ) renascens Seitz T.

— (■ — ) cf. revili (Jacob) T.

— ( Epileymeriella ) cf. hitzeli (Jacob) T.
Tegoceras gladiator (Bayle) m.

- — mosense (d'Orbigny) m.

— sp. nov. m.

Oxytropidoceras alticarinatum (Spath) m.
Cymatoceras radiation (J. Sowerby) F, D-J.

- — pseudoelegans (d’Orbigny) F, D-N.

— albense (d’Orbigny) m.

Eucymatoceras plication (J. de C. Sowerby) F,
D, B.

Eutreplioceras sublaevigatum (d’Orbigny) N.

— clementinum (d’Orbigny) m.

- — sp. nov.? J, T.

Anglonautilus undulatus (J. Sowerby) (= Nautilus
farringdonensis Sharpe) N.

Heminautilus saxbyi (Morris) f.

Conoteuthis dupiniana (d’Orbigny) f.

— vectensis Spath f.

Neohibolites ewaldi (v. Strombeck) D-N.

- — minimus (Miller) T, m.

• — strombecki (Muller) ?N.

■ — spicatus Swinnerton ?N.

— wollemanni Stolley ?N.

— sp. J.

Hibolites minutus Swinnerton ?F.

Oxyteuthis depressus Stolley ?F.

Phylum ARTHROPODA
Class CRUSTACEA
Sub-class MALACOSTRACA
Notopocorystes stokesi (Mantell) m.

Eucorystes broderipi (Mantell) m.

Goniodromites scarabaeus Wright & Wright T.
Goniodromites ? M.

Plagiophthalmus nitonensis Wright & Wright m.
Cyphonotus incertus Bell T.

Diaidax carteriana Bell T.

Mithracites vectensis Gould F, f.

Homolopsis edwardsi Bell m.

Gen. nov. (‘ Homolopsis ’) cf. depressa Carter f, D.
Vectis wrighti Withers f.

Oriothopsis cf. bonneyi Carter J.

Meyeria magna M'Coy f.

G/yphea vectensis Woods F.

Homarus longimanus G. B. Sowerby F-m.
Enoploclytia tuberculata (Bell) T.

Linuparus carteri (Reed) f.

Sub-class CIRRIPEDIA

Cretiscalpellum unguis (J. de C. Sowerby) T, m.
- — aptiense Withers N.

— sp. nov. f.

Pycnolepas rigida (J. de C. Sowerby) T.
Scalpellion ( Arcoscalpelhon ) simplex Darwin M.

— ( — ■) accumulation Withers N, T.

— ( — ) comptum Withers f, B.

Virgiscalpellum wrighti Withers f.

VERTEBRATA

Phylum pisces

Hybodus complanatus Owen B.

— basanus Woodward F.

— obtusus Agassiz N.

— subcarinatus Agassiz F.

Acrodus ornatus Woodw'ard F, f.

— lac vis Woodward m.

Synechodus tenuis Woodward B.

Heterodontus sulcatus Woodward B.

Sp/iaerodus neocomiensis Agassiz N
Gyrodus atherfieldensis White f.

Mesodon coidoni (Agassiz) N, m.
Scapanorhynchus subulatus Agassiz T.

■ — raphiodon Agassiz? T.

— sp. F, f.

Apateodus sp.l T.

Lamna appendiculata Agassiz T, m.

Isurus mantelli (Agassiz) J-m.

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 611

Unidentified elasmobranch teeth and vertebrae
F-m.

Ischyodus thurmanni Pictet & Campiche M.
Edaphodon sp. J.

Hylaeobatis problematica Woodward F.

Phylum reptilia

Pelorosaurus (= Dinodocus) mackesoni (Owen)
B.

Iguanodon mantelli Meyer B, N.

Camptosaurus sp. m.

Acanthopholis horridus Huxley m.

Ichthyosaurus campylodon Carter J, m.
Polyptychodon interrupt us Owen m.

Polyptychodon continuus Owen B.
Cimoliosaurus latispinus (Owen) B.

- — planus (Owen) ?F.

Cimoliosaurus bernardi (Owen) ?F.
Pliosaurus sp. B.

Protemys serrata Owen D.

Unidentified thalassemyd turtle f.

PROBLEMATICA

Hallimondia fasciculata gen. et sp. nov. N.
Petromonile benstedii (Bensted) M.
Granularia spp. f, J. T.

Chondrites targionii (Brongniart) D-N.

— spp. F-m.

REFERENCES

allen, p. 1955. Age of the Wealden in North-Western Europe. Geol. Mag. 92, 265-81.

— - — 1960. Strand-line pebbles in the mid-Hastings Beds and the geology of the London Uplands.

General features. Jurassic pebbles. Proc. Geol. Assoc. 71, 156-65.
arkell, w. j. 1939. Derived ammonites from the Lower Greensand of Surrey. Proc. Geol. Assoc. 50,
22-25.

1944. Stratigraphy and structures east of Oxford. Part II: The Miltons and the Haseleys. Quart.

J. Geol. Soc. London, 100, 45-60, pi. 4.

1 947a. The Geology of Oxford. Oxford.

19476. Geology of the country around Weymouth, Swanage, Corfe and Lulworth. Mem. Geol.

Surv. Gt. Brit.

austen, r. a. c. [Godwin-]. 1843. On the geology of the south-east of Surrey. Proc. Geol. Soc. 4,
167-73.

• — — 1850. On the age and position of the fossiliferous sands and gravels of Farringdon. Quart. J.
Geol. Soc. London, 6, 454-78.

barber, c. A. 1898. Cupressinoxylon vectense, a fossil conifer from the Lower Greensand of Shanklin,
in the Isle of Wight. Ann. Bot. 12, 329-61, pi. 23-24.
barker, j. 1952. In Geological records: Hampshire. South-eastern Naturalist and Antiquary, 57, 21.
barrois, c. 1874. Sur le Gault et sur les couches entre lesquelles il est compris dans le bassin de Paris.
Ann. Soc. geol. Nord, 2, 1-61.

■ 1875. L’age des ‘Folkestone Beds’ du Lower Greensand. Ibid. 3, 23.

1878. Memoire sur le Terrain Cretace des Ardennes et des regions voisines. Ibid. 5, 227-487.

benavides-caceres, v. e. 1956. Cretaceous System in Northern Peru. Bull. Amer. Mus. Nat. Hist.
108 (4), 359-493, pi. 31-66.

bensted, w. h. 1860. On the Kentish Ragstone as exhibited in the Iguanodon Quarry at Maidstone.
Proc. Geol. Assoc. 1, 57.

1862. Notes on the geology of Maidstone. Geologist, 5, 294, 334, 378, 447.

boer, G. de, neale, J. w., and penny, l. f. 1958. A guide to the geology of the area between Market
Weighton and the Humber. Proc. Yorks. Geol. Soc. 31, 157-209, pi. 11-12.
boschma, H. 1956. Milleporina and Stylasterina. In R. C. Moore (ed.) : Treatise on invertebrate
paleontology. F, Coelenterata, F90-F106. Univ. Kansas Press.
boswell, p. g. h. 1929. Cretaceous. In Evans and Stubblefield (eds.) : Handbook of the geology of Great
Britain, 383-410. London.

breistroffer, M. 1947. Sur les zones d’ammonites de 1’Albien de France et d’Angleterre. Trav. Lab.
geol. Grenoble, 26, 1-88.

brett, d. w. 1957. On Pseudoaraucaria Fliche emend., a genus of fossil pinaceous cones. Ann. Bot.,
n.s., 21, 33-51, pi. 2, 3.

brinkmann, r. 1937. Biostratigraphie des Leymeriellenstammes nebst Bemerkungen zur Palao-
geographie des Nordwestdeutschen Alb. Mitt. Geol. Staats-Inst. Hamburg, 16, 1-18.

612

PALAEONTOLOGY, VOLUME 3

BRISTOW, h. w. 1862. Geology of the Isle of Wight. Mem. Geol. Surv. Gt. Brit.

1889. Idem. 2nd edit, revised by Reid and Strahan.

bromehead, c. e. n. 1924. Iii Dewey, Bromehead, and others: The geology of Dartford. Mem. Geol.
Surv. Gt. Brit., 8-10.

brown, e. e. s. 1941. The Folkestone Sands and base of the Gault near Wrotham Heath, Kent. Proc.
Geol. Assoc. 52, 1-15.

buckland, w. 1836. Geology and mineralogy, considered with reference to natural theology. Bridgewater
Treatise, 6, 2 vols. London.

butler, G. w. 1922. On the Pema Bed and the Weald Clay at Reigate. Proc. Geol. Assoc. 33, 313-18.
canu, f. and bassler, r. s. 1926. Studies on the cyclostomatous bryozoa. Proc. U.S. Nat. Mus. 67,
art. 21, 1-124, pi. 1-31.

carruthers, w. 1866. On some fossil coniferous fruits. Geol. Mag. 3, 534-46, pi. 20-21.

1867. On Cycadoidea Yatesii, a fossil Cycadean stem from the Potton Sands, Bedfordshire.

Ibid. 4, 199-201, pi. 9.

1868. British fossil Pandaneae. Ibid. 5, 153-6, pi. 9.

• — — 1869. On some undescribed coniferous fruits from the Secondary rocks of Britain. Ibid. 6, 1-7,
pi. 1-2.

1870. On fossil Cycadean stems from the Secondary rocks of Britain. Trans. Linn. Soc. 26,

675-708, pi. 54-63.

casey, r. 1936. Recent additions to the Albian ammonoid faunas of Folkestone. Geol. Mag. 73,
444-8.

1939. The upper part of the Lower Greensand around Folkestone. Proc. Geol. Assoc. 50, 362-78,

pi. 22.

— — 1946. The Folkestone Beds: aeolian or marine? South-eastern Nat. & Ant. 51, 5 pp., pi.

1949 a. Field meeting at Folkestone and Sandling. Proc. Geol. Assoc. 60, 223-5.

— — 1949 b. The ammonite genus Uhligella in the English Albian. Geol. Mag. 86, 333-45, pi. 19,
text-fig.

1950. The junction of the Gault and Lower Greensand in East Sussex and at Folkestone, Kent.

Proc. Geol. Assoc. 61, 268-98, pi. 14, 6 text-figs.

1951a. The zonal position of the Gault-Lower Greensand junction beds at Wrecclesham, Surrey.

Ibid. 62, 95-99.

19516. Note on the provenance of the type of Lopliidiaster ornatus Spencer (Asterozoa). Ann.

Mag. Nat. Hist. (12), 4, 463^4.

1951c. Kent: in geological records. South-eastern Nat. & Ant. 55, xxiv-xxv.

— — - 1952. Some genera and subgenera, mainly new, of Mesozoic heterodont lamellibranchs. Proc.
Malac. Soc. 29, 121-76, pi. 7-9.

— — - 1953. The application to be given a trivial name which, when first published, was both applied
to a particular species or to particular specimens and also stated to be a substitute name for some
previously published trivial name or is clearly implied to be such a substitute. Bull. Zool. Nom.
10, 174-6.

1 954c7. New genera and subgenera of Lower Cretaceous ammonites. J. Washington Acad. Sci.

44 (4), 106-15, figs. 1-10.

19546. Proposed use of the plenary powers to designate (i) a neotype for the nominal species

‘ Ammonites mammillatus ’ Schlotheim, 1813, and (ii) a type species for the genus ‘ Douvilleiceras ’ de
Grossouvre, 1893 (Class Cephalopoda, Order Ammonoidea). Bull. Zool. Nom. 9, 250-4.

1954c. Fa/ciferella, a new genus of Gault ammonites, with a review of the family Aconeceratidae

in the British Cretaceous. Proc. Geol. Assoc. 65, 262-77, pi. 7.

1955a. The pelecypod family Corbiculidae in the Mesozoic of Europe and the Near East. J.

Washington Acad. Sci. 45 (12), 366-72.

19556. Notes on the base of the Gault in Wiltshire. Proc. Geol. Assoc. 66, 231-4.

1956. The Neomiodontidae, a new family of the Arcticacea (Pelecypoda). Proc. Malac. Soc. 31,

208-22, pi. 11.

- — — ■ 1957. The Cretaceous ammonite genus Leymeriella, with a systematic account of its British
occurrences. Palaeontology, 1, 29-59, pi. 7-10.

1959. Field Meeting at Wrotham and the Maidstone By-pass. Proc. Geol. Assoc. 70, 206-9.

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 613

casey, R. 1960a. A monograph of the Ammonoidea of the Lower Greensand. Part 1. Palaeontogr.
Soc. i-xxxvi, 1^44, pi. 1-10.

19606. A Lower Cretaceous gastropod with fossilized intestines. Palaeontology, 2, 270-6, pi. 41.

— — 1960c. A new echinoid from the Lower Cretaceous (Albian) of Kent. Ibid. 3, 260-4.
chapman, f. 1894. The Bargate Beds of Surrey and their microscopic contents. Quart. J. Geol. Soc.
London, 50, 677-730, pi. 33-34.

chatwin, c. p. 1935. Guide to geological excursion to the Isle of Wight and the mainland opposite.
Geological Survey Centenary. London.

collingnon, m. 1937. Paleontologie de Madagascar, xxii. Les ammonites pyriteuses de l’Aptien
d'Antanatanamirafy. Ann. Paleont. 26, 28 pp., 3 pi.
conybeare, w. d. and Phillips, w. 1822. Outlines of the geology of England and Wales, etc. London.
coquand, H. 1862. Sur le convenance d’etablir dans le groupe inferieur de la formation cretacee
un nouvel etage entre le Neocomien proprement dit (couches a Toxaster complanatus et a Ostrea
Couloni) et le Neocomien superieur (etage Urgonien d’Alc. d'Orbigny). Bull. Soc. geol. France (2),
19, 531-41.

cornes, h. w., and others. 1925. The geology of the Canterbury district. Proc. Geol. Assoc. 36, 257-84.
cornuel, j. 1860. Sur le groupe du gres vert inferieur du bassin de la Seine, sur sa division d’apres
les oscillations du sol et les caracteres geologiques et stratigraphiques et sur les rapports, assise
par assise, avec les diverses parties du gres Wealdien et du Lower Greensand d’Angleterre. Bull.
Soc. geol. France (2), 17, 736-89.

corroy, G. 1925. Le Neocomian de la bordure orientale du bassin de Paris. Doctorate thesis. Univer-
sity of Nancy. 334 pp., 1 1 pi. Nancy.

cox, l. r. 1960. The British Cretaceous Pleurotomariidae. Bull. Brit. Mus. (Nat. Hist.) Geol., 4, no. 8.
creber, g. t. 1956. A new species of abietaceous cone from the Lower Greensand of the Isle of Wight.
Ann. Bot., N.s., 20, 375-83, pi. 15.

crick, c. c. 1916. On Ammonitoceras tovilense from the Lower Greensand of Kent. Proc. Malac.
Soc. 12, 118-20, pi. 6.

cunnington, w. 1850. On a section of the Lower Greensand near Devizes. Quart. J. Geol. Soc.
London, 7, 453-4.

DAVEY, E. c. 1874. The ‘ Sponge-Gravel ’ Beds at Coxwell, near Faringdon, etc. Wantage.

1905. The Neocomian sponges, bryozoa, foraminifera and other fossils of the Sponge-gravel beds

at Little Coxwell, near Faringdon. London, Bath, and Faringdon.
davidson, t. 1851-86. British fossil Brachiopoda. 6 vols. Palaeontogr. Soc. London.
davies, A. M. 1899. Contributions to the geology of the Thame Valley. Proc. Geol. Assoc. 16, 15-57,
pi. 2.

davies, g. m. 1916. The rocks and minerals of the Croydon Regional Survey area. Trans. Croydon Nat.
Hist. Soc. 8, 53-96.

de rance, c. e. 1868. On the Albian or Gault of Folkestone. Geol. Mag. 5, 163-71.
destombes, j-p. and destombes, p. 1938. Note sur le Gault de Wissant. Ann. Soc. geol. Nord. 62,
98-113.

destombes, p. 1958. Revision de l’Albien de la region du Havre. Deductions paleogeographiques sur
le NW du Bassin Parisien au Cretace moyen. Bull. Soc. geol. France (6), 8, 305-13.
dines, h. G. and Edmunds, F. h. 1929. The geology of the country around Aldershot and Guildford.
Palaeontology by C. P. Chatwin. Mem. Geol. Surv. Gt. Brit.

— — 1933. The geology of the country around Reigate and Dorking. Ibid.

drew, f. 1864. Geology of the country between Folkestone and Rye, including the whole of Romney
Marsh (Sheet 4). Lists of fossils by Robert Etheridge. Ibid.

1875. See topley 1875.

dumas, E. 1875-6. Statist ique geologique, mineralogique, metal/urgique et paleontologique du Depart-
ment du Gard. Pt. 1, i-lxxi, 73-284 (1875); pt. 2, 1-735, pi. i-ix (1876). Paris and Nimes.
duncan, p. m. 1866-91. A monograph of the British fossil corals. Second series, being a supplement
to the monograph of British corals by Milne Edwards and Haime. Palaeontogr. Soc.
dutertre, a. p. 1923. Note sur le Cretace inferieur du Bas-Boulonnais. Ann. Soc. geol. Nord, 48, 35-74.

1925. Remarques sur le Cretace inferieur du Bas-Boulonnais et du Sud-est de l’Angleterre. Ibid.

49, 237-50, pi. iv.

614 PALAEONTOLOGY, VOLUME 3

Edmunds, f. h. 1928. Wells and springs of Sussex. Mem. Geol. Surv. Gt. Brit.

■ 1956. In Summary of progress of the Geological Survey of Great Britain and the Museum of

Practical Geology for the year 1955, 32.

elliott, g. f. 1947. The development of a British Aptian brachiopod. Proc. Geol. Assoc. 58, 144-59,
pi. 3-4.

— — 1948. Palingenesis in Thecidea (Brachiopoda). Ann. Mag. Nat. Hist. (12) 1, 1-30, pi. 1-2.

1956. Post-palaeozoic brachiopod ecology: a re-assessment. Geol. Mag. 93, 196-200.

■ — — 1959. Six new genera of Mesozoic Brachiopoda. Ibid. 96, 146-8.

evans, c. 1864. On fossils from the railway-cuttings in the vicinity of London (Sevenoaks). Proc.
Geol. Assoc. 1, 347-51.

1871. On the strata exposed by the line of the railroad through Sevenoaks Tunnel. Ibid. 2, 1-4.

ewald, J. 1850. Ueber die Grenze zwischen Neocomien und Gault. Zeit. deutsch. geol. Gesell. 2,
440-78.

fitton, w. h. 1824. Inquiries respecting the geological relations of the beds between the Chalk and
the Purbeck Limestone in the South-east of England. Ann. Phil. 24, n.s. 8, 365-458. (Reprinted in
4to in 1833.)

1836. Observations on some of the strata between the Chalk and the Oxford Oolite in the South-
east of England. Including: Appendix A, Descriptive notes respecting the shells figured in pis. 1 1-
23, by J. de C. Sowerby. Trans. Geol. Soc. (2), 4, 103-390, pi. 11-23.

1843a. Observations on part of the section of the Lower Greensand at Atherfield, on the coast

of the Isle of Wight. Proc. Geol. Soc. 4, 198-203.

18436. Comparative remarks on the Lower Greensand of Kent and the Isle of Wight. Ibid.

208-10.

- — — 1844. Observations sur le Lower Greensand de File de Wight. Bull. Soc. geol. France (2), 1,
438-53, pi. 8-9.

■ — — 1845. Comparative remarks on the sections below the Chalk on the coast near Hythe, in Kent,
and Atherfield, in the Isle of Wight. Quart. J. Geol. Soc. London, 1, 179-89.

— — 1846. Stratigraphical account of the section from Atherfield to Rocken-end in the Isle of Wight.
Ibid. 2, 55-56.

1847a. A stratigraphical account of the section from Atherfield to Rocken-end, on the south-west

coast of the Isle of Wight. Ibid. 3, 289-327, pi. xii, table.

18476. On the arrangement and nomenclature of some of the sub-Cretaceous strata. Rep. Brit.

Assoc. Adv. Sci., 1846, 58.

forbes, e. 1845. Catalogue of Lower Greensand fossils in the Museum of the Geological Society
with notices of species new to Britain. Quart. J. Geol. Soc. London, 1, 237-50; 345-55, pi. 1-4.

ganz, e. 1912. Stratigraphie der mittleren Kreide (Gargasien und Albien) der oberen helvetischen
Decken in den nordlichen Schweizeralpen. Neue Denkschr. Schweiz. Naturf. Ges, 47, Abh. 1, i-vii,
1-148.

Gardner, J. s. 1875. On the Cretaceous Aporrhaidae. Geol. Mag. 12, 392-400, pi. 12.

1876. Cretaceous Gastropoda. Ibid. 3, 75-78, 105-14, 160-3, pi. 3^1.

1877a. On British Cretaceous Patellidae and other families of patelloid Gastropoda. Quart. J.

Geol. Soc. London, 33, 192-206, pi. 7-9.

18776. Notes on Cretaceous Gastropoda. Geol. Mag. 14, 556-7, pi. 16.

1886. On fossil flowering or phanerogamous plants. Ibid. (3), 3, 495-503.

gillet, s. 1921. Etude du Barremian superieur de Wassy (Haute-Marne). Bull. Soc. geol. France (4),
21, 3-47, pi. 1-3.

1924-5. Etudes sur les lamelli branches neocomiens. Mem. Soc. geol. France, n.s., 1, fasc. 3-4,

1-224, pi. 7-8; 2, fasc. 1, 225-339.

gossling, f. 1929. The geology of the country around Reigate. Proc. Geol. Assoc. 40, 197-259, pi. 5-23.

— — 1936. Field Meeting at Oxted and Godstone. Ibid. 47, 322-7.

and bull, a. J. 1948. The structure of Tilburstow Hill, Surrey. Ibid. 59, 131-9, pi. 11-12.

Gregory, J. w. 1895. On a collection of fossils from the Lower Greensand of Great Chart in Kent.
Geol. Mag. 32, 97-103, 187-9.

1 899. Catalogue of the fossil bryozoa in the Department of Geology, British Museum ( Natural

History). The Cretaceous Bryozoa, 1 , London.

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENSAND 615

Gregory, j. w. Idem, 2.

Hancock, J. m. 1958. The Lower Cretaceous near Leighton Buzzard. In W. S. Pitcher and others:
The London Region. Geologists' Association Guides, No. 30: 36-40. Colchester.
harper, J. c. and wilson, v. 1938. A new fossiliferous locality in the Folkestone Sands at Abinger
Hammer, Surrey. Proc. Geo/. Assoc. 49, 58.

Harris, t. m. 1956. The mystery of flowering plants. The Listener, 26 April 1956, 514-16.
hawkins, h. l. 1921a. Note on a collection of echinoids from the limestone-lenticles in the sand-pits
of Shenley Hill. Geol. Mag. 58, 57-60.

19216. Morphological studies on the Echinoidea Holectypoida and their allies. XI. Conulopyrina

anomala, a new type of the Echinoneidae. Ibid. 420-6, pi. 7.
himus, G. w. 1939. Field Meeting at Aylesford. Proc. Geol. Assoc. 50, 68-71.

hinde, G. J. 1883. Catalogue of the fossil sponges in the Geological Department of the British Museum
{Natural History). London.

1885. On beds of sponge-remains in the Lower and Upper Greensand of the South of England.

Phil. Trans. Roy. Soc. 176, 403-53, pi. 40-45.

holmes, s. c. h. 1959. In Summary of progress of the Geological Survey of Great Britain and the
Museum of Practical Geology for the year 1958, 28.
house, m. r. 1958. The Dorset Coast from Poole to the Chesil Beach. Geologists' Association Guides,
No. 22. Colchester.

hughes, n. f. 1958. Palaeontological evidence for the age of the English Wealden. Geol. Mag. 95,
41-49.

Humphrey, w. E. 1949. Geology of the Sierra de los Muertos area, Mexico (with descriptions of Aptian
cephalopods from La Pena formation). Bull. Geol. Soc. Amer. 60, 89-176, pi. 4-18.

Humphries, d. w. 1956. Chert: its age and origin in the Hythe Beds of the Western Weald. Proc. Geol.
Assoc. 67, 296-313, pi. 15-16.

ibbetson, l. l. b. and forbes, E. 1845. On the section between Blackgang Chine and Atherfield Point.
Quart. J. Geol. Soc. London, 1, 190-7.

jackson, J. F. 1910. The Rocks of Hunstanton and its neighbourhood. London.

• 1939. Notes on the discovery of fossiliferous nodules in the Carstone near Niton. Proc. I. Wight

Nat. Hist. Arch. Soc. 3, pt. 1 (1938), pp. 73-80.

Jacob, c. 1905. Etude sur les ammonites et sur l’horizon stratigraphique du gisement de Clansayes.
Bull. Soc. geol. France (4), 5, 339-432, pi. 12-13.

1907. Etudes paleontologiques et stratigraphiques sur la partie moyenne des terrains cretaces

dans les alpes frangaises. Trav. Lab. geol. Grenoble, 8, 280-590, 6 pi.
judd, j. w. 1867. On the strata which form the base of the Lincolnshire Wolds. Quart. J. Geol. Soc.
London, 23, 227-50.

1868. On the Speeton Clay. Ibid. 24, 218-50.

1870. Additional observations on the Neocomian strata of Yorkshire and Lincolnshire, with

notes on their relations to the beds of the same age throughout Northern Europe. Ibid. 26, 326-48.

1871. On the Punfield Formation. Ibid. 27, 207-27, table.

jukes-browne, a. j. 1875. On the relations of the Cambridge Gault and Greensand. Ibid. 31,
256-314.

1886. On the application of the term Neocomian. Geol. Mag. 23, 311-19.

1891. On the geology of Devizes with remarks on the grouping of the Cretaceous deposits.

Proc. Wilts. Nat. Hist. Soc. 25, 371 and ff. ; Proc. Geol. Assoc. 12 (1892), 254-66.

1900. Cretaceous Rocks of Britain. Vol. 1, The Gault and Upper Greensand of England. With

contributions by W. Hill and E. T. Newton. Mem. Geol. Surv. Gt. Brit.

191 1. The building of the British Isles. 3rd ed. London.

■ 1912. The student's handbook of stratigraphical geology. 2nd ed. London.

keeping, w. 1883. The fossils and palaeontological affinities of the Neocomian deposits of Upware
and Brickhill. Sedgwick Prize Essay for 1879. Cambridge.
kent, p. e. 1949. A structure contour map of the surface of the buried pre-Permian rocks of England
and Wales. Proc. Geol. Assoc. 60, 87-104.

kilian, w. 1887. Description geologique de la Montagne de Lure (Basses-Alpes). Ann. Sci. Geol.
19-20, 1-458, 8 pi., 1 map.

616

PALAEONTOLOGY, VOLUME 3

kilian, w. and reboul, p. 1915. Contributions a l’etude des faunes Paleocretacees du Sud-est de la
France, 1, La faune de l’Aptien inferieur des environs de Montelimar (Drome). Mem. Explic. carte
geol. det. France, 221 pp., 9 pi.

king, w. b. r. 1954. The geological history of the English Channel. Quart. J. Geol. Soc. 110, 77-101,
pi. 4.

kirkaldy, J. F. 1932. The geology of the country around Hascombe, Surrey. Proc. Geol. Assoc. 43,
127-51.

• 1933o. The tectonic development of the Western Weald in Lower Cretaceous times. Geol. Mag.

70, 254-68.

■ — 19336. The Sandgate Beds of the Western Weald. Proc. Geol. Assoc. 34, 270-311.

1935. The base of the Gault in Sussex. Quart. J. Geol. Soc. London, 91, 519-37.

■ - - 1937. The overstep of the Sandgate Beds in the Eastern Weald. Ibid. 93, 94-126.

1939. The history of the Lower Cretaceous period in England. Proc. Geol. Assoc. 50, 379-416,

pi. 23-26.

— — 1947n. The work of the late Frank Gossling on the stratigraphy of the Lower Greensand between
Brockham (Surrey) and Westerham (Kent). Ibid. 58, 178-92.

- — — 19476. The provenance of the pebbles in the Lower Cretaceous rocks. In Wells and others:
Studies of pebbles from the Lower Cretaceous rocks. Ibid. 223-41.

1958. Geology of the Weald. Geologists' Association Guides, No. 29. Colchester.

and Wooldridge, s. w. 1938. Notes on the geology of the country around Haslemere and Mid-
hurst. Proc. Geol. Assoc. 49, 135^17.

kitchin, f. L. and pringle, J. 1921. On an inverted mass of Upper Cretaceous strata near Leighton
Buzzard, Beds., and on an overlap of the Upper Gault in that neighbourhood. Geol. Mag. 57,
4-15; 52-62; 100-13; 285-6.

1922a. On the overlap of the Gault in England and on the ‘Red Chalk’ of the Eastern

Counties. Ibid. 59, 156-66; 194-200.

19226. The Gault and Lower Greensand near Leighton Buzzard. Ibid. 283-7.

knowles, l. and middlemiss, f. a. 1958. The Lower Greensand in the Hindhead area of Surrey and
Hampshire. Proc. Geol. Assoc. 69, 205-38.

koenen, a. von. 1902. Die Ammonitiden des Norddeutschen Neocom. (Valanginien, Hauterivien,
Barremien und Aptien.) Abh. k. preuss. geol. Landes., n.f., no. 24, 1-451, Atlas of 55 pi.

— — 1907. Uber das Auftreten der Gattungen und Gruppen von Ammonitiden in dem einzelnen
Zonen der Unteren Kreide Norddeutschlands. A us. den Nach. d. k. Gesell. Wiss. Gottingen. Math.-
phys. Klasse, 1907, 1-10.

lamplugh, g. w. 1899. In Summary of Progress of Geological Survey for 1898, 141-3.

1901. Idem for 1900, 115-19.

1903. Belemnites of the Faringdon ‘Sponge-gravels’. Geol. Mag. (4), 10, 32-34.

1922. On the junction of the Gault and Lower Greensand near Leighton Buzzard. Quart. J.

Geol. Soc. London, 78, 1-80.

— — -and kitchin, f. l. 1911. On the Mesozoic Rocks in some of the Coal explorations in Kent-
Mem. Geol. Surv. Gt. Brit.

■ and pringle, J. 1923. The concealed Mesozoic Rocks in Kent. Ibid.

— — and walker, j. f. 1903. A fossiliferous band at the top of the Lower Greensand near Leighton
Buzzard. Quart. J. Geol. Soc. London, 59, 234-65, pi. 16-18.

laubenfels, m. w. de. 1955. Porifera. In R. C. Moore (ed.): Treatise on invertebrate paleontology.
E, Archaeocyatha and Porifera, E21-E122. Univ. Kansas Press.

leighton, T. 1895. The Lower Greensand above the Atherfield Clay of East Surrey. Quart. J. Geol.
Soc. London, 51, 101-24.

leriche, m. 1905. Observations sur la geologie de l’llede Wight. Ann. Soc. geol. Nord. 34, 16-42.

leymerie, a. 1841-2. Memoire sur le terrain Cretace du Department de l’Aube, contenant des con-
siderations generates sur le terrain Neocomien. Mem. Soc. geol. France (1), 4 (1841), 291-364,
pi. 16-18; 5 (1842), 1-34, pi. 1-18.

— — 1845. Observations on a communication made by Dr. Fitton to the Geological Society of France
at the Meeting of May 20, 1844, on the Lower Greensand of the Isle of Wight. Phil. Mag. (3), 26,
281 ; see also Bull. Soc. geol. France (2), 2, 41-47.

RAYMOND CASEY: STR ATIG R APHIC AL PALAEONTOLOGY OF GREENSAND 617

mackeson, h. b. 1840. Note on the discovery of some portions of a large Saurian near the bottom of
the Lower Greensand in the vicinity of Hythe. Proc. Geol. Soc. 3, 325.
mackie, s. J. 1856. A handbook of Folkestone. Folkestone and London.

1860. Geology of Folkestone. Geologist, 3, 41-45, 81-90, 121-31, 201-7, 281-4, 321-7, 353-7,

393-6.

1862. The ‘Dragon Tree’ of the Kentish Rag. Ibid. 5, 401-4, pi. 22.

mantell, G. 1822. The fossils of the South Downs. London.

1833. The geology of the south-east of England. London.

1834. Discovery of the bones of the Iguanodon in a quarry of Kentish Rag fa limestone belonging

to the lower greensand formation) near Maidstone, Kent. Ed. New Phil. J. 17, 200.

1839. The wonders of geology. 2 vols., ed. 3, vol. 2. London.

1843. Description of some fossil fruits from the Chalk formation of the south-east of England.

Proc. Geol. Soc. 4, 34—35.

1844. The medals of creation; or first lessons in geology and in the study of organic remains.

2 vols. London.

1851. Geological excursions round the Isle of Wight and along the adjacent coast of Dorsetshire.

2nd ed. London.

martin, P. J. 1828. A geological memoir on a part of Western Sussex; with some observations upon
Chalk-basins , the Weald denudation , and outliers of protrusion. London.
meyer, c. J. A. 1864a. Three days at Farringdon. Position of the Sponge Gravel. Geologist, 7, 5-11.

18646. Notes on brachiopoda from the pebble-bed of the Lower Greensand of Surrey; with

descriptions of new species, and remarks on the correlation of the Lower Greensand Beds of Kent,
Surrey and Berks., and of the Farringdon Sponge Gravel, and the Tourtia of Belgium. Geol. Mag.
1, 249-57, pi. 11-12.

1866. Notes on the correlation of the Cretaceous Rocks of the south-east and west of England.

Ibid. 3, 13-18, pi. 2.

1868. On the Lower Greensand of Godaiming. Proc. Geol. Assoc. 1868, 1-20.

1872. On the Wealden as a fluvio-lacustrine formation, and on the relation of the so-called

‘Punfield formation’ to the Wealden and Neocomian. Quart. J. Geol. Soc. 28, 243-55.

— — 1873. Further notes on the Punfield section. Ibid. 29, 70-76.

middlemiss, f. A. 1959. English Aptian Terebratulidae. Palaeontology, 2, 94-142, pi. 15-18.
morris, J. 1867. On the Ferruginous Sands of Buckinghamshire, with remarks on the distribution of
the equivalent strata. Geol. Mag. 4, 456-62.

muller, s. w. and schenk, h. g. 1943. Standard of the Cretaceous System. Bulk Amer. Assoc. Petrol.
Geol. 27, 262-78.

Murchison, R. 1826. Geological sketch of the north-western extremity of Sussex and the adjoining
parts of Hants and Surrey. Trans. Geol. Soc., London (2), 2, 97-107.
natsky, a. d. 1918. Stratigraphie der unteren Kreide Mangyschlak. Mat. Geol. Russ. 26, livr. 1, 287 pp.,
9 pi.

neaverson, e. 1928. Stratigraphical palaeontology. London.

newton, r. b. 1896. On the identification of the Acanthoceras mammillatum- and the Hoplites inter-
ruptus- zones at Okefield Fitzpaine, Dorset. Geol. Mag. 33, 198.

1897. An account of the Albian fossils lately discovered at Okeford Fitzpaine, Dorset. Proc.

Dorset Nat. Hist. F.C. 18, 66-99, pi. 1-3.
norman, m. w. 1887. A popular guide to the geology of the Isle of Wight. Ventnor.
orbigny, a. d'. 1840-2. Paleontologie Frangaise. Terrains Cretaces. 1, Cephalopodes. 662 pp., pi. 1-148.
Paris.

— — 1842u-3. Idem. 2, Gasteropodes. 456 pp., pi. 149-236. Ibid.

1844-7. Idem. 3, Lamellibranches. 807 pp., pi. 237-489. Ibid.

1847o-51. Idem. 4, Brachiopodes. 390 pp., pi. 490-599. Ibid.

1 850«. Prodrome de paleontologie stratigraphique universelle. Vol. 2, 428 pp. Ibid.

owen, e. F. 1956. The Lower Cretaceous brachiopods ‘ Rhynchonella' gibbsiana (J. de C. Sowerby)
and Sulcirhynchia hythensis sp. nov. Ann. Mag. Nat. Hist. (12), 9, 164-72, pi. 3.

1960. A note on ‘ Rhynchonella ’ sulcata (Parkinson) from the Lower Cretaceous of Great Britain.

Ibid. (13), 2, 248-56, pi. 5.

618 PALAEONTOLOGY, VOLUME 3

owen, r. 1851-64. A monograph of the fossil Reptilia of the Cretaceous formations. Palaeontogr. Soc.
1884. A history of British fossil reptiles. Vol. 2, p. ix. London.

Paine, J. m. and way, j. t. 1848. On the phosphoric strata of the Chalk Formation. Journ. Roy.
Agric. Soc., London (1), 9, 56-78.

pervinquiere, L. 1912. Carte geologique de la Tunisie. Etudes de paleontologie tunisienne. II, Gastro-
podes et lamellibranches des terrains cretaces. 352 pp., atlas of 23 pi. Paris.
pictet, f. j. and campiche, g. 1858-67. Materiaux pour la Paleontologie Suisse. Description des fossiles
du terrain Cretace des environs de Ste. Croix. Vols. 1-3; 380+752+558 pp., 139 pi. Geneva.

— — and renevier, e. 1854-8. Idem. Description des fossiles du terrain Aptien de la Perte-du-Rhone
et des environs de Ste. Croix, I, no. 1, 184 pp., 23 pi. Ibid.

and roux, w. 1847-53. Description des mollusques fossiles qui se trouvent dans les Gres Verts

des environs de Geneve. Mem. Soc. phvs. et hist. nat. Geneve, 11, pt. 2 (1847), 257-412; 12 (1849),
157-287; 13 (1854), 73-173, 489-558, pi. 1-51.

price, f. g. h. 1874. On the Lower Greensand and Gault of Folkestone. Proc. Geol. Assoc. 4, 135-50.

1875. On the Gault of Folkestone. Quart. J. Geol. Soc. London, 30, 342-66.

■ 1876. Excursion to Sandgate and Folkestone. Proc. Geol. Assoc. 4, 554-6.

1879. The Gault. London.

pringle, J. 1926. Geology of the country around Oxford. 2nd ed. Mem. Geol. Surv. Gt. Brit.
rastall, r. h. 1919. The mineral composition of the Lower Greensand Series of Eastern England.
Geol. Mag. 56, 211-20; 265-72.

1925. On the tectonics of the southern Midlands. Ibid. 62, 193-222.

reid, c. 1903. The geology of the country near Chichester. With contributions by G. W. Lamplugh
and A. J. Jukes-Browne. Ibid.

renevier, e. 1854. Memoir geologique sur la Perte-du-Rhone et ses environs. Nouv. Mem. Soc. Helv.
Sci. Nat. 14, 1-71, 3 pis., 1 map.

renngarten, v. 1926. La faune des depots cretaces de la region d’Assa-Kambileevka, Caucase du
Nord. Mem. Com. geol. Russ., n.s., livr. 147, 132 pp., 9 pi.
sazonova, i. g. 1958. Lower Cretaceous deposits of the central regions of the Russian Platform.
In O. V. Flerovoi: Mesozoic and Tertiary deposits of the central regions of the Russian Platform
[in Russian], Vses. Nauchno-Issled. Geo/.-Razved. Neft. Inst. Moscow, 31-136, pi. 1-23.
scott, g. 1940. Cephalopod from the Cretaceous Trinity Group of the south-central United States.
Univ. Texas Pub!., no. 3945, 969-1106, pi. 55-68.

seward, a. c. 1895. Catalogue of the Mesozoic plants in the Department of Geology, British Museum
(Natural History). The Wealden flora. Pt. II, Gymnospermae. London.

1896. Notes on the geological history of Monocotyledons. Ann. Bot. 10, 205-20, pi. 14.

sharpe, d. 1854. On the age of the fossiliferous sands and gravels of Farringdon and its neighbour-
hood. Quart. J. Geol. Soc. London, 10, 176-98.

shepherd, w. b. 1934. Some observations on the Folkestone Beds around Farnham. Proc. Geol.
Assoc. 45, 85-114.

simms, f. w. 1843. Account of a section of strata between the Chalk and the Wealden Clay in the
vicinity of Hythe, Kent. Proc. Geol. Soc. 4, 206-8.

1845. On the thickness of the Lower Greensand beds of the south-east coast of the Isle of Wight.

Quart. J. Geol. Soc. London, 1, 76-77 .

sinzow, i. 1905. Uber einige evolute Ammonitiden-Formen aus dem oberen Neokom Russlands.
Mat. Geol. Russ. (2), 22, 291-348, pi. 15-22.

1909. Beitrage zur Kenntnis des siidrussischen Aptien und Albien. Verb, russ.-kais. Min. Ges.

St. Petersb. (2), 47, 1-48, pi. 1-4.

sornay, j. 1957. Cretace (France). In Lexique stratigraphique international. Vol. 1, Europe, fasc. 4a,
403 pp. Int. Geol. Congr. Comm. Strat. Paris.

sowerby, j. 1811. Extract from the minute book of the Linnean Society of London, Nov. 7, 1809.
Trans. Linn. Soc. HO, 405-6.

and sowerby, j. de c. 1812-46. The Mineral Conchology of Great Britain. 7 vols., pi. 1-383 (1812-

22) by J. Sowerby; pi. 384-648 (1823-46) by J. de C. Sowerby. London.
sowerby, J. de c. 1836. See fitton 1836.

spath, l. f. 1921. In Stamp: Excursion to Tilburstow Hill and Nutfield. Proc. Geol. Assoc. 32, 30-32.

RAYMOND CASEY: STR ATIG R APHICAL PALAEONTOLOGY OF GREENSAND 619

spath, l. f. 1923(7. Excursion to Folkestone, with notes on the zones of the Gault. Ibid. 34, 70-76.

19236. On the ammonite horizons of the Gault and contiguous deposits. Appendix II, in Summ.

Progr. Geol. Surv.for 1922, 1 39 — 49.

1923c. A monograph of the ammonoidea of the Gault. Pt. 1. Palaeontogr. Soc.

1924. On the ammonites of the Speeton Clay and the subdivisions of the Neocomian. Geol. Mag.

61, 73-89.

— - — 1925a. A monograph of the ammonoidea of the Gault. Pt. 2. Palaeontogr. Soc.

1925 6. Notes on the ammonites of the Lower Greensand and Gault, Folkestone and neighbour-
hood. In J. W. Walton: Folkestone and the country around, 31-36. Folkestone.

■ — - — ■ 1925c. A monograph of the ammonoidea of the Gault. Pt. 3. Palaeontogr. Soc.

1925 d. Cretaceous ammonites [letter on], Geol. Mag. 62, 336.

■ 1930(7. On some ammonoidea from the Lower Greensand. Ann. Mag. Nat. Hist. (10), 5, 417-64,

pi. 14-17.

19306. A monograph of the ammonoidea of the Gault. Pt. 7. Palaeontogr. Soc.

- 1933. Review of Spath, 1930a. Palaeont. Zentra/hl. 3, no. 765, 266-7 .

— — 1935. Field Meeting at Folkestone, Kent. Proc. Geol. Assoc. 46, 429-31.

1939a. On a new belemnoid ( Conoteuthis rennet) from the Aptian of the Colony of Mozambique.

Bol. Serv. Ind. Min. Geol. Mozambique, no. 2, 1939, 13-16, 2 figs.

19396. Problems of ammonite nomenclature. V: On Acanthohoplites jacobi (Collet) and the

Jacobi Zone of the Folkestone Sands. Geol. Mag. 76, 236-9.

1941. A monograph of the ammonoidea of the Gault. Pt. 14. Palaeontogr. Soc.

1942. Idem. Pt. 15.

1943. Idem. Pt. 16.

stephenson, l. w. 1952. Larger invertebrate fossils of the Woodbine formation (Cenomanian) of
Texas. U.S. Geol. Surv. Prof. Paper, 242.

stoliczka, p. 1870-1. The Cretaceous fauna of Southern India. 3, Pelecypoda. Mem. Geol. Surv.
India, Palaeont. Indica, Ser. 6.

stolley, e. 1908a. Die Gliederung der norddeutschen Unteren Kreide. Centralbl. Min., &c., 107-24,
140-51, 162-72, 211-20, 242-7.

19086. Zur Kenntniss der Unteren Kreide Norddeutschlands. Ibid. 753-61.

stopes, m. 1911a. The ‘Dragon-tree’ of the Kentish Rag, with remarks on the treatment of imperfectly
petrified woods. Geol Mag. 48, 55-59.

19116. The name of the ‘Dragon-tree’. Ibid. 468-9.

- — — 1912. Petrifactions of the earliest European angiosperms. Phil. Trans. Roy. Soc. 203B, 75-100,
pi. 6-8.

- — — 1913. Catalogue of the Mesozoic plants in the British Museum ( Natural History). The Cretaceous
flora. Part 1. London.

1915. Idem. Part 2. The Lower Greensand ( Aptian ) plants of Britain. London.

stoyanow, a. 1949. Lower Cretaceous stratigraphy in southeastern Arizona. Geol. Soc. Amer. Mem.

38.

strahan, a. 1886. Notes on the relations of the Lincolnshire Carstone. Quart. J. Geol. Soc. London, 42,
481-93.

• — ■ — 1896. Ammonites (Acanthoceras) mammiJlatus in the Isle of Wight. Geol. Mag. 33, 287.

1898. The geology of the Isle of Purbeck and Weymouth. Mem. Geol. Surv. Gr. Brit.

strombeck, a. von. 1853. Uber den Gault im subhercynen Gebirge. Zeit. deutsch. geol. Gesell. 5,
501-15.

1854. Schichtenfolge und Gliederung der Unteren Kreideformation in Braunschweig. Ibid. 6,

264-70.

1856. t)ber das Alter des Flammenmergels im nordwestlichen Deutschland. Ibid. 8, 483-93.

- — — 1857. Beitrag zur Kenntniss des Gaults im Norden vom Harze. Neues Jahrb. Geol. Petr., &c.,
1857, 641-78.

1861. Uber den Gault und insbesondere die Gargasmergl im nordwestlichen Deutschland.

Zeit. deutsch. geol. Gesell. 13, 20-63.

swinnerton, h. h. 1935. The rocks below the Red Chalk of Lincolnshire and their cephalopod
faunas. Quart. J. Geol. Soc. 91, 1-46, pi. 1-5.

620 PALAEONTOLOGY, VOLUME 3

swinnerton, h. h. 1937. A monograph of British Lower Cretaceous Belemnites. Pt. 2. Palaeontogr.
Soc.

1955. Idem, Pt. 5.

swinton, w. e. 1951. Gideon Mantell and the Maidstone Iguanodon. Notes & Rec. Roy. Soc. London,
8, 261-76.

taitt, A. H. and KENT, p. E. 1958. Deep boreholes at Portsdown (Hants) and Henfield (Sussex). British
Petroleum Company Ltd., London.

teall, j. j. h. 1875. The Potton and Wicken phosphatic deposits. Sedgwick Prize Essay for 1873.
Cambridge.

thomas, h. d. 1947. Some English Cretaceous and Eocene hexacorals. Quart. J. Geol. Soc. London,
103, 163-70, pi. 8-9.

toombs, H. A. 1935. Report of a Field Meeting at Leighton Buzzard. Proc. Geol. Assoc. 46, 432-6.
topley, w. 1868. On the Lower Cretaceous beds of the Bas-Boulonnais; with notes on their English
equivalents. Quart. J. Geol. Soc. London, 24, 472-83.

• 1875. The geology of the Weald (with contributions by F. Drew, and others). Mem. Geol. Surv.

Gt. Brit.

and jukes-browne, a. j. 1888. Report of sub-committee no. II. Cretaceous. In Reports of sub-
committees on classification and nomenclature, pp. 61-78. 4th Internat. Geol. Congr. Cambridge.
toucas, a. 1888. Note sur le Jurassique superieur et le Cretace inferieur de la vallee du Rhone. Bull.
Soc. geol. France (3), 16, 903-27, table.

treacher, l. 1908. Excursion to Culham and Abingdon. Proc. Geol. Assoc. 20, 548-52.
versey, h. c. and carter, c. 1926. The petrography of the Carstone and associated beds in Yorkshire
and Lincolnshire. Proc. Yorks. Geol. Soc. 20, 349-65.
vokes, h. 1945. Protodonax, a new Cretaceous molluscan genus. J. Palaeont. 19, 295-308, pi. 46-47.

1946. Contributions to the palaeontology of the Lebanon Mountains, Republic of Lebanon.

Part 3, The pelecypod fauna of the ‘Olive Locality’ (Aptian) at Abeih. Bull. Amer. Mus. Nat. Hist.
87, 139-216, pi. 1-10.

wade, b. 1926. The fauna of the Ripley Formation on Coon Creek, Tennessee. U.S. Geol. Surv. Prof.
Paper, 137.

walker, J. f. 1866m On the fossils contained in a Lower Greensand deposit of phosphatic nodules
in Bedfordshire. Ann. Mag. Nat. Hist. (3), 18, 31-32.

1 8666. On a phosphatic deposit in the Lower Greensand of Bedfordshire. Ibid. 381-6, pi. 13.

■ 1867. On some new Terebratulidae from Upware. Geol. Mag. 4, 454-6, pi. 19.

1868. On the species of Brachiopoda, which occur in the Lower Greensand at Upware. Ibid.

5, 399-406, pi. 18-19.

- — — 1870. On secondary species of Brachiopoda. Ibid. 7, 560-4.

webster, t. 1825. Reply to Dr. Fitton's paper entitled: ‘Inquiries respecting the Geological Relations
of the Beds between the Chalk and the Purbeck Limestone in the Southeast of England'. Ann. Phil.
25, n.s. 9, 33.

wells, a. k. and gossling, f. 1947. A study of the pebble beds in the Lower Greensand in East Surrey
and West Kent. Proc. Geol. Assoc. 58, 194-222, pi. 7-9.
whitaker, w. 1886-7. Borings in Kent. Contributions to the Deep-seated Geology of the London
Basin. Quart. J. Geol. Soc. London, 42 (1886), 26-47, pi. 3; 43 (1887), 197-205.

1908. The water supply of Kent. Mem. Geol. Surv. Gt. Brit.

and jukes-browne, a. J. 1899. The geology of the borders of the Wash, including Boston and

Hunstanton. Ibid.

skertchly, s. b. j., and jukes-browne, a. j. 1893. The geology of south-western Norfolk and of

northern Cambridgeshire. Ibid.

white, e. i. 1927. Two new species and a new genus of Cretaceous Pycnodonts from the South of
England. Ann. Mag. Nat. Hist. (9), 20, 186-91.

white, h. j. osborne. 1910. The geology of the country around Alresford. Mem. Geol. Surv. Gt. Brit.

■ 1921. A short account of the geology of the Isle of Wight. Ibid.

1923. The geology of the country south and west of Shaftesbury. Ibid.

1924. The geology of the country near Brighton and Worthing. Ibid.

1926. The geology of the country near Lewes. Ibid.

RAYMOND CASEY: STRATIGRAPHICAL PALAEONTOLOGY OF GREENLAND 621

Whiteside, l. m. p. 1956. A petrified fern stem, Protopteris fibrosa Stenzel, from the Lower Cretaceous
of Kent. Ann. Mag. Nat. Hist. (12), 9, 81-85.

Wiltshire, t. 1869. On the red chalk of Hunstanton. Quart. J. Geol. Soc. London, 25, 185-91.
withers, t. 1935. Catalogue of fossil Cirripedia in the Department of Geology, British Museum ( Natural
History). II, Cretaceous. London.

1945. New Cretaceous Cirripedes and Crab. Ann. Mag. Nat. Hist. (11), 12, 552-60, pi. 2.

woods, h. 1899-1913. A monograph of the Cretaceous Lamellibranchia of England. 2 vols., 104 pi.
Palaeontogr. Soc.

woodward, h. b., and others. 1904. The water supply of Lincolnshire from underground sources:
with records of sinkings and borings. Mem. Geol. Surv. Gt. Brit.
worrall, g. A. 1954. The Lower Greensand in East Kent. Proc. Geol. Assoc. 65, 185-202, pi. 4.

1956. The mineralogy of some Lower Greensand borehole samples. Ibid. 67, 138-41.

wright, c. w. 1957. In W. J. Arkell, Bernard Kummel, and C. W. Wright: Mesozoic Ammonoidea.
In R. C. Moore (ed.) : Treatise on invertebrate paleontology, Pt. L, Mollusca 4, Cephalopoda,
Ammonoidea. pp. L80-L490. Univ. Kansas Press.

and thomas, h. d. 1946. Notes on the geology of the country around Sevenoaks, Kent. Proc.

Geol. Assoc. 51, 315-21, pi. 24.

and wright, e. v. 1942a. Some new sections and fossils from the Folkestone Beds of the Farnham

district. Ibid. 52, 86-87.

19426. New records of Cretaceous fossils from the Isle of Wight. Proc. I. O. Wight Nat.

Hist. Arch. Soc. 1941, 283-7.

1947. The stratigraphy of the Albian Beds at Leighton Buzzard. Geol. Mag. 84, 161-8.

- — — -■ — — 1948. Note on two exposures of the base of the Gault in Surrey. Proc. Geol. Assoc. 59,
84-86.

1950a. Further records for Cretaceous fossils from the Isle of Wight. Proc. I. O. Wight Nat.

Hist. Arch. Soc. 1949, 123-6.

19506. Some Dromiacean crabs from the English Cretaceous. Proc. Geol. Assoc. 61, 13-27,

pi. 1.

wright, j. 1905. Lower Greensand foraminifera from Little Coxwell, near Faringdon. Geol. Mag.
42, 238-9.

wright, t. 1864-82. A monograph on the British fossil Echinodermata from the Cretaceous forma-
tions. 1. Palaeontogr. Soc.

zittel, k. a. von. 1865-6. Die Bivalven der Gosaugebilde in den nordostlichen Alpen. Denkschr. k.
Akad. Wiss. Wien, 24, 105-78, pi. 1-10; 25, 77-198, pi. 11-27.

RAYMOND CASEY

Geological Survey and Museum,

Manuscript received 3 May 1960 London, S.W.7

Note added in proof. Recent collecting has produced the following important addition to the list of
Lower Greensand ammonites on pages 608-10: Cymahoplites sp. nov., Folkestone Beds, main mam-
millatum bed, Copt Point, Folkestone (GSM 99240; author's coll.). This is a boreal genus of ammon-
ites otherwise known only by its type species, C. kerenskianus (Bogoslowsky), from the Albian of
Central Russia.

S s

B 6612

THE PALAEONTOLOGICAL ASSOCIATION

Extracts from the Annual Report of the Council for 1959

Membership. On 31 December 1959 there were 724 members (417 Ordinary and 307 Institutional).

Finance. The Accounts and Balance Sheet for 1959 are given below. A new donation was received
from the Kuwait Oil Co. and previous donations from other oil companies were renewed for another
year. The Council wishes to thank the managements of these companies for their continued support;
their donations have enabled the Association to publish considerably larger parts of ‘Palaeontology’
than would have been possible on subscription income alone. Thanks are also due to several authors
and their universities or companies for donations towards the cost of publishing particular papers.

These donations, together with the increase in membership to over 700, have enabled the rate of
publication to be increased to four parts in 1960. This will cost over £3,000, about double the 1959
subscription income. Since many donations are not promised more than one year in advance, it is
necessary to increase the present subscription income in order to ensure that this rate of publication
is maintained. Although this can be done to a certain extent by increasing the membership, the Council
considered that an increase in the subscription would not be out of place in relation to the publications
to be received by members.

Constitution. A special general meeting was convened on 20 January 1960 in order to consider pro-
posed changes in the Constitution. These were designed (a) to allow additional Officers to be appointed
to cope with the increasing volume of the Association’s work; ( b ) to allow an increase in the annual
subscription so that four parts a year of Palaeontology could be published from 1960 onwards without
disturbing the Association’s financial stability; and (c) to introduce a new category of student member-
ship. It was resolved:

(i) That the second sentence of Rule 3 of the Constitution be amended to read as follows: ‘The

Council shall consist of twenty-four members, of which not more than eleven shall be Officers.
The Officers shall consist of a President and, at least, two Vice-Presidents, a Treasurer, a
Secretary, and an Editor, and such other Officers as the Council may from time to time
determine.’

(ii) That, effective from 1 January 1961, Rule 2 of the Constitution be amended to read as follows:

‘There shall be Ordinary members. Institutional members, and Student members, and their
annual subscriptions shall be Three Guineas, Five Guineas, and Two Guineas respectively.
Each subscriber shall be considered a member of the Association, but Institutional members
shall not be eligible to take part in the government of the Association.’

‘Palaeontology.’ Volume 1, part 4, and Volume 2, part 1 were published during the year. They
contained 21 papers and 3 notes.

Plans were made during the year to publish four parts of ‘Palaeontology’ each year from 1960
onwards, in order to eliminate delays in publication of the increasing number of papers being sub-
mitted. Three Editors will be appointed in 1960-1.

Meetings. Five meetings were arranged during 1959-60, all of which were successful and well attended.
The Association is grateful to the Council of the Geological Society of London, the Director of the
British Museum (Natural History), Dr. K. S. Sandford (Oxford), Prof. F. W. Shotton (Birmingham),
and Prof. D. Williams (Imperial College) for generously granting facilities for meetings; and to the
Local Secretaries for their efficient services.

a. The Second Annual General Meeting was held in the Rooms of the Geological Society of London
on 1 1 March. The Annual Report of the Council for 1958 was adopted, and the Council for 1 959—
60 elected. Professor Gerhard Regnell (Lund, Sweden) delivered the second Annual Address on
‘The Lower Palaeozoic Echinoderm faunas of the British Isles and Balto-Scandia’.

b. A Demonstration Meeting was held in the Department of Geology and Mineralogy, University

THE PALAEONTOLOGICAL ASSOCIATION 623

Museum, Oxford, on 9 May. There were fifteen exhibits. Dr. W. S. McKerrow was Local Secre-
tary.

c. A Demonstration Meeting on ‘Palaeontological Techniques’ was held in the British Museum
(Natural History) on 17 October. There were thirty exhibits and several special demonstrations,
visits, and discussions. Over 200 persons attended. Mr. A. E. Rixon was Local Secretary.

d. A Discussion Meeting was held in the Department of Geology, The University, Birmingham, on
18/19 December. The subject was ‘Lower Palaeozoic Faunas’. Nine papers were read during
two sessions, and there were twenty-two exhibits. A dinner was held in the University Union on
18 December. Dr. J. D. Lawson was Local Secretary.

e. A Discussion Meeting on ‘The Teaching of Palaeontology’ was held in the Department of Geo-
logy, Imperial College, London, on 20 January. Ten speakers introduced various topics during
two sessions and over thirty members contributed to the discussion. The Chairman was Professor
T. Neville George.

Council. The following were elected for 1959-60 at the Annual General Meeting on 1 1 March 1959:
President : Dr. R. G. S. Hudson. Vice-Presidents'. Dr. L. R. Cox, Mr. T. F. Grimsdale. Treasurer'.
Dr. W. S. McKerrow. Secretary: Dr. Gwyn Thomas. Editor: Dr. W. H. C. Ramsbottom. Other
members: Dr. T. Barnard, Prof. O. M. B. Bulman, Mr. M. A. Calver, Dr. W. G. Chaloner, Mr. H. V.
Dunnington, Dr. F. E. Eames, Prof. T. N. George, Dr. C. H. Holland, Mr. N. F. Hughes, Prof. L. R.
Moore, Prof. F. H. T. Rhodes, Dr. I. Strachan, Dr. C. J. Stubblefield, Prof. P. C. Sylvester-Bradley.

ACCOUNTS FOR THE YEAR ENDING 31 DECEMBER 1959

General Account

Income

Expenditure

£

s.

d.

£

x.

d.

Balance from 1958

909

10

8

Palaeontology , Vol. 1, Part 4 (ex-

Subscriptions ....

1,599

16

0

eluding plates)

937

17

3

Sale of separate parts

302

2

6

Palaeontology, Vol. 2, Part 1

1,166

1

8

Miscellaneous ....

54

8

3

Officers’ expenses

87

12

3

Sale of reprints

120

17

7

Insurance of back stock

3

2

6

Interest .....

61

19

4

Geological Society (A.G.M.)

1

16

0

Donations: Apex (Trinidad) Oil-

Provision for Palaeontology, Vol.

fields, Attock Oil Co., British

2, Part 2 ... .

850

0

0

Petroleum Co., Burmah Oil Co.,

Transfer to Reserve Fund .

1,000

0

0

Iraq Petroleum Co., Kuwait Oil

4,046

Co., Shell International Petro-

y

0

leum Co., and Ultramar Co. .

1,150

0

0

Balance carried to 1960

408

14

2

Donations via authors (British

Petroleum Co., Iraq Petroleum

Co., Queen Mary College, Uni-

versity of Hull, University of

Sheffield, Yale University)

256

9

6

£4,455

3

10

£4,455

3

10

Reserve

Fund

Income

Expenditure

£

s.

d.

£

s.

d.

Balance from 1958

1,000

0

0

Transfer from General Account .

1,000

0

0

£2,000 0 0

£2,000 0 0 Balance

624

PALAEONTOLOGY, VOLUME 3

Liabilities

Balance of General Account
Balance of Reserve Fund .

Balance Sheet

£ s. d.
408 14 2
2,000 0 0

Assets

£ s. d. £ s. d.
5% Defence Bonds . . . 2,000 0 0

Barclays Bank Ltd.

Deposit account . 915 15 11
Current account . 342 18 3

£2,408 14 2

1,258 14 2
Less provision for

Palaeontology, Vol.

2, Part 2 . . 850 0 0

408 14 2

£2,408 14 2

Auditors : D. Nichols, W. J. Wadsworth.

INDEX

Pages 1 to 128 are contained in Part 1 ; pages 129 to 244 are in Part 2; 245 to 396 in Part 3; and 397 to 622
in Part 4. Figures in Bold Type indicate plate numbers.

A

Acanthoclymenia, All.

Agnostids, 410, 412.

Amos, A. J., Campbell, K. S. W. and Goldring, R.
Australosutura (Trilobita) from the Carboniferous
of Australia and Argentina, 227.

Anahoplitoides, 599.

Anarcestes, 131, 24.

Ancyrospora grandispinosa, 55, 14.

Antevsia extans, 350, 58; zeilleri, 347, 58.

Anthonya, 580; woodsi, 79.

Anthraconaia, 106; A.? kirki, 141, 25.

Anthraconauta, 106.

Anthracosia, 106.

Anthracosphaerium, 1 06.

Anulatisporites, 82, 20; amdatus, 20.

Aparchites sinuatus, 100, 23.

Apiculatisporis elegans, 30, 11.

Aptolinter, 575.

Argentina: Australosutura, Carboniferous trilobite,
227.

Arthropoda. See Eurypterids, Ostracods, ‘Scorpions’,
Trilobites.

Auroraspora, 49; aurora, 51, 14; macromanifest us, 50,
14; micromanifestus, 5 1 .

Australia: Australosutura, Carboniferous trilobite,
227; clymeniid from Wocklumeria zone, 237;
Hoegisporis, Cretaceous form genus, 485; Middle
Palaeozoic favositids, 186; Upper Devonian
Foraminifera, 397.

Australosutura, 229; gardneri, 230, 39, 40.

B

Barnard, P. D. W. Calathospermum fimbriatum, a
Lower Carboniferous pteridosperm cupule, 265.

Bennison, G. M. Lower Carboniferous non-marine
lamellibranchs from East Fife, Scotland, 137.

Benniescorpio tuberculatus, 322, 56, 57.

Beudanticeras newtoni, 591.

Beyrichia {Neobey richia) equilatera, 98; kochii, 98,
23; maccoyiana, 99, 23; var. sulcata, 99, 23; B.
(Nodibeyrichia) pustulosa, 96, 23.

Biharisporites submamillarius, 33, 11, 12.

Bollandites castletonense group, 18.

Bollandoceras micronotum, 17.

Brachiopods: Carboniferous rhynchonellid, Pugnoides
triplex (M‘Coy), 477; feeding mechanism of
Prorichthofenia, 450; Lower Greensand, 573, 604;
Silurian Dinobulus from Canada, 242; Upper
Devonian, from Canada, 208.

Brett, D. W. Fossil oak-wood from the British
Eocene, 86.

Brightonia, 590; sandlingensis, 591 ; turns, 591.

Bryozoa: Cretaceous Pyripora and Rhammatopora,
370; Lower Greensand, 573, 603; Ordovician
bifoliate, 1 ; Upper Silurian from Wales, 69.

Butcher, N. E. and Flodson, F. Review of Carboni-
ferous goniatite zones in Devon and Cornwall, 75.

Buthiscorpius buthiformis, 277 , 46-48; major, 300, 51,
52.

Bythocypris phillipsiana, 101, 23.

C

Calathospermum fimbriatum, 273, 45.

Cambrian: alimentary caeca of trilobites, 410.

Campbell, K. S. W. See Amos, A. J.

Canada: Devonian brachiopods, 208; Devonian
spores, 26; Silurian Dinobulus, 242; Silurian
ostracods, 93.

Carbonicola, 106, 137; antiqua, 139, 142, 25; elegans,
140, 147, 25.

Carboniferous : ammonoids and trilobites from Corn-
wall, 153; Australosutura, trilobite, 227 ; goniatite
zones in Devon and Cornwall, 75; miospore wall
structure, 82; non-marine lamellibranchs, 104, 137;
pteridosperm cupule, 265 ; Pugnoides, rhynchonellid,
477; ‘scorpions’, 276.

Casey, R. New echinoid from the Albian of Kent,
260; stratigraphical palaeontology of the Lower
Greensand, 487.

Casts of fossils, 124.

Centropleura sp., 426, 68.

Cephalopods: abnormal growths, 129; ammonoids
from Cornwall, 153; clymenid from Australia, 237;
goniatite zones in Devon and Cornwall, 75; Lower
Greensand, 591, 608; supposed clymenid Acantho-
clymenia, 472.

Chaunoproetus aff. carnicus, 178, 29.

Cheloniceras ( Cheloniceras ) parinodum, 594, 84; C.
(Epicheloniceras) debile, 595, 84; gracile, 596, 81;
martinioides, 595, 84.

Circumsporites melvillensis, 34, 12.

Cleoniceras ( Cleoniceras ) floridum, 599, 84.

Clymenids, origin of, 474; from Australia, 237.

Coelenterates : Lower Greensand, 512, 602; Middle
Palaeozoic favositids, 186.

Colonammina imparilis, 405, 65.

Convolutispora flexuosa forma minor, 34, 12.

Cookson, I. C. Hoegisporis, a new Australian Creta-
ceous form genus, 485.

Copeland, M. J. Ostracoda from the Upper Silurian
Stonehouse Formation of Arisaig, Nova Scotia,
Canada, 93.

Corals, Middle Palaeozoic favositids, 186.

Corvaspis kingi, 217, 37, 38.

Cosmosporites, 52 ; microspinosus, 53, 14 ; velatus, 52, 14.

626

INDEX

Costaclymenia muensteri, 157, 26.

Crespin, I. Upper Devonian Foraminifera from
Western Australia, 397.

Cretaceous: Australian form genus, Hoegisporis, 485;
Bryozoa Pyripora and Rhammatopora, 370; echinoid
from Kent, 260; lamellibranch Pseudavicula, 392;
ostracods from Yorkshire, 439; stratigraphical
palaeontology of the Lower Greensand, 487.

Cristatisporites conannulatus, 60, 14; mediconus, 60,
14; orcadensis, 58, 14.

Cryptochasma, 576; ovale, 82.

Cucullaea tea Hi, 575.

Cuneocorbula arkelli, 590, 82.

Cyclogranisporites ampins, 29, 11.

Cymaclymenia borahensis, 239, 41; catnerata, 27;
constricta, 166, 27; constrict a var. A, 167, 27; con-
stricta var. globosa, 167, 27; cordata, 27; striata
var., 166, 27; tetragona, 27.

Cyrtina lapidea, 214, 35, 36.

Cyrtoclymenia tetragona, 165, 27.

Cyrtosymbole (? Calybole ) cf. nepia, 180; C. (Macro-
bole) aff. blax, 182, 29; drewerensis, 180, 29;
duodecimae, 181, 29; C. (? Macrobole) aff. bergica,
180, 29; sp., 183, 29; C. (IVaribole) aff. conifera,
179, 29; aff. italica, 179, 29; C. (? IVaribole)
dunhevedensis, 179.

Cytherellina siliqua, 101, 23.

D

Dekayella megacanthopora, 71, 16; ramosa, 72, 16.

Densosporites, 36, 57, 82; crassus, 36, 13; devonicus,
57, 14; loricatus, 20; sphaerotriangularis, 20;
spongeosus, 20; striatus, 20.

Deshayesites callidiscus, 594, 80; fittoni, 593, 84;
forbesi, 593, 81, 84.

Devonian: abnormal growths in goniatites, 129;
ammonoids and trilobites from Cornwall, 153;
brachiopods from Canada, 208; clymenid from
Australia, 237; Foraminifera from Australia, 397;
ostracoderms, 217; spores, 26, 45; supposed
clymenid Acanthoclymenia, 472; Devonian? favo-
sitids, 186.

Dickins, J. M. The Permian leiopteriid Merismopteria
and the origin of the Pteriidae, 387; characters and
relations of the Mesozoic pelecypod Pseudavicula,
392.

Dinobulus sp. cf. D. conradi, 242, 41.

Discoclymenia cornwallensis, 174, 28; aff. cornwallensis,
175, 29; cucullata, 28.

Disparilia, 579.

Dufrenoyia transitoria, 594, 83.

Douvillinella, 211; D. ? crickmayi, 212, 36.

E

Echinoids: Lower Cretaceous, 260, 603.

Entogonites grimmeri, 17.

Eocene: oak wood from Britain, 86.

Eomiodon, 585.

Epicyprina harrisoni, 586, 80.

Epiwocklumeria, 161; dunhevedensis, 162, 26.

Eriphyla, 579.

Escharopora recta, 1 7, 6.

Eury dicta montifera, 12, 3.

Eurypterids: the median abdominal appendage of
Slimonia, 245.

Exogyra, 590.

F

Favosites grandiporus, 196, 32-34; moonbiensis, 31,
33; squamuliferus, 197; squamuliferus forma aus-
tralis, 200, 30, 33, 34; squamuliferus forma bryani,
197, 31, 33; squamuliferus forma nitidus,' 199, 30, 34;
squamuliferus forma ovatiporus, 200, 32; squamuli-
ferus forma stelliformis, 199, 30, 33; squamuliferus
forma £, 201, 31; squamuliferus forma rj, 201, 30,
33; squamuliferus forma 6, 201, 32-34.

Fenestricardita, 580.

Filosina, 587.

Fistulipora strawi, 70, 16; umbrosa, 69, 16.

Foraminifera: Upper Devonian, from Western

Australia, 397.

Freiastarte praetypica, 579.

G

Gastrioceras circumnodosum, 19.

Gastropods, Lower Greensand, 590, 607.

Gattendorfia crassa, 176; occlusa, 176, subinvoluta,
176; tenuis, 176.

Geltena meyeri, 579.

Girtyoceras burhennei, 18.

Glyptagnostus, 428; reticulatus, 430, 70; stolidotus,
432, 70.

Glvptoscorpius, 331, 53.

Goldring, R. See Amos, A. J.

Goniatites. See cephalopods.

Goniatites bisati, 18; concentricus/ striatus group, 17;
crenistria, 18; falcatus, 18; aff. granosus, 18;
sphaericostriatus, 18.

Gonioclymenia (Kallodymenia) biimpressa, 159, 26;
f rechi, 159; wocklumensis, 159.

Graptodictya elegant ula, 21, 8, 9; perelegans, 19, 7, 8.

Graysonia anglica, 573, 79.

H

Hallam, A. Kulindrichnus langi, a new trace-fossil
from the Lias, 64.

Hallimondia, 600; fasciculata, 600, 79.

Heterostraci, development of dermal plates, 222.

Hodson, F. See Butcher, N. E.

Hoegisporis lenticulifera, 485, 76.

Holaster (Labrotaxis) cantianus, 261, 44.

Homoceras beyrichianum, 19; magistrorum, 19;
undulation, 19.

Hoplites ( Isohoplites ) eodentatus, 599, 83.

House, M. R. Abnormal growths in some Devonian
goniatites, 129; Acanthoclymenia, the supposed
earliest Devonian clymenid, is a Manticoceras,
472.

Hymenozonotriletes inaequus, 37, 13.

Hypacanthoplites milletioides, 598, 83.

Hyperammina devoniana, 406, 64.

Hystricosporites, 31; delectabilis, 32, 11.

INDEX

627

I

lcanotia pennula, 583, 80; siliqita, 583.

Icanotiidae, 581.

? Imitoceras ornatissimus, 17.

Inoceramus coptensis, 587, 82; salomoni, 588.

J

Jefferies, R. P. S. Photonegative young in the Triassic
lamellibranch Lima lineata (Schlotheim), 362.

Jenkins, T. B. H. Non-marine lamellibranch assem-
blages from the Coal Measures (Upper Carboni-
ferous) of Pembrokeshire, West Wales, 104.

Jurassic: ostracods from England, 439; trace-fossil
from the Lias, 64.

K

Kenseyoceras, 169; K. ( Keynseyoceras ), 170; rostrata,
171, 28; A. ( Mayneoceras ), 171; nucleus, 172, 28;
sinuconstricta, 173, 28.

Kilyeni, T. I. See Neale, J. W.

Kloedenia wilkensiana, 99, 23.

Kosmoclvmenia bisulcata, 27; linearis , 26; pattisoni.

165, 27.

Kulindrichnus langi, 64, 15.

L

Labrotaxis, 260.

Lagenammina ampullacea, 404, 66.

Lamellibranchs: Carboniferous, non-marine, 104, 137;
Lower Greensand, 575, 605 ; Mesozoic Pseudavicula,
392; Permian leiopteriids and the origin of the
Pteriidae, 387; young of Triassic Lima, 362.

Larwood, G. P. See Thomas, H. D.

Latosporites (?), 38, 13.

Leioclema explanation, 72, 16.

Leiotrilites confertus, 27, 11; dissimilis, 27, 1 1 ;
marginalis, 28, 11; microdeltoides, 28, 11.

Lepidopteris martinsii, 345 ; stormbergensis, 340, 58.

Lichnophthalmus pulcher, 331.

Lima lineata, 362, 59.

Limatula sabulosa, 578, 83.

Limopsis dolomitica, 576, 79.

Linearia cornea, 581.

Lonsda contortuplicata, 572.

Lopha sp., 59.

Lower Greensand, stratigraphical palaeontology, 487 ;
faunal and floral lists, 601; palaeontology, 572;
stratigraphical account, 501 ; zonation, 492; in Isle
of Wight, 77; in Kent, 78.

Lycospora magnifica, 35, 12, 13; magnifica forma
endoformis, 36, 12; pallida, 36, 12, 13.

M

Mandelstamia angulata, 443, 71; maculata, 444, 71;
rectilinea, 440, 71; sexti, 446; triebeli, 442, 71;
sp. 1, 446, 71.

Manticoceras, 472; neapolitanum, 75.

Marsipella sp., 401, 65.

Mazoniscorpio mazonensis, 290, 49-51.

McGregor, D. C. Devonian spores from Melville
Island, Canadian Arctic Archipelago, 26.

Meade, M. J. See Rixon, A. E.

Mediraon sulcatum, 580.

Merismopteria, 387, 63.

Merocanites cf. henslowi, 17: aff. applanatus, 17.

Modestella, 573; modesta, 574, 83.

Monotrypa flabellata, 72, 16.

Myoconcha delta, 577, 81.

N

Naiadites, 106.

Neale, J. W. and Kilyeni, T. I. New species of Mandel-
stamia (Ostracoda) from the English Mesozoic,
439.

Nervostrophia borealis, 209, 35; maclareni, 210, 36.

Nolaniceras, 598.

Noramya, 575.

Norford, B. S. Dinobulus from the Middle Silurian of
British Columbia, Canada, 242.

O

Oak wood from the British Eocene, 86.

Old Red Sandstone. See Devonian.

Olenellids, 418.

Opik, A. A. Alimentary caeca of agnostid and other
trilobites, 410.

Ordovician: bifoliate Bryozoa, 1; caeca of trilobites,
410.

Ostracoderms, Devonian, 217.

Ostracods: Canadian Upper Silurian, 93; English
Mesozoic, 439.

Owen, D. E. Upper Silurian Bryozoa from Central
Wales, 69.

P

Pachydictya robust a, 14, 4, 5.

Pachythaerus tealli, 579, 80.

Papyriaspis lanceola, 68, 69.

Paragoniatites newsomi, 18.

Parahoplites cunningtoni, 596, 82.

Parawocklumeria laevigata, 167, 27; laevigata var.
obesa, 169, 28; sp. 169, 28.

Parkinson, D. The Carboniferous rhynchonellid
Pugnoides triplex (M‘Coy), 477.

Pedder, A. E. H. Brachiopods from the Upper
Devonian of Western Canada, 208.

Pelecypods. See Lamellibranchs.

Peltaspermum thomasi, 356, 58.

Pericyclus ( Rotopericyclus ) aff. homoceratoides, 17:
? sp. 17.

Permian: feeding mechanism of Prorichthofenia, 450;
leiopteriid lamellibranchs, r387 ; pteridosperms, 333.

Perotrilites sp. 35, 12.

Petromonile, 600.

Phacops ( Cryphops ?) ensae, 177; wocklumeriae, 178;
P. ( Dianops ) sp., 178, 29; P. ( Phacops ) accipitrinus
accipitrinus, 1 77.

Philip, G. M. Middle Palaeozoic squamate favositids
of Victoria, 186.

Phillips, June R. P. Restudy of types of seven Ordo-
vician bifoliate Bryozoa, 1.

628

INDEX

Pickett, J. W. A clymeniid from the Wocklumeria
zone of New South Wales, 237.

Pinna ( Stegoconcha ), 587.

Planisporites dilucidus, 30, 11; minimus, 29, 11.

Plants: Lower Greensand, 572, 601; and see also:
Pteridosperms, spores.

Platyclymenia ( Platyclymenia ) bicost at a, 163; patti-
soni, 164, 26; valida, 163, 26; ? Platyclymenia, 41.

Polyzoa. See Bryozoa.

Postglatziella carinata, 162, 26.

Primitia mundula, 101, 23.

Prodeshayesites , 592; obsoletus, 592, 82.

Prolecanitid, 17.

Prorichthofenia, 450; permiana, 72-74; uddeni, 73, 74.

Proteonina sp., 404, 65.

Protodonax, 584; minutissimus, 81.

Proveniella rosacea, 586, 80.

Pseudavicula cf. papyracea, 392, 63.

Pteridosperms: Calathospermum fimbriatum. Lower
Carboniferous, 265; Peltaspermaceae, Permian and
Triassic, 333.

Pteriidae, origin of, 390.

Pterotrigonia mantelli, 577, 80.

Ptilodictya gracile, 74, 16.

Ptychopariids, 419, 68, 69.

Pugnoides triplex. All , 76.

Punctatisporites arcticus, 28, 11; putaminis, 29, 11;
scabratus, 29, 11.

Pyripora anglica, 375, 60, 61 \faxensis, 379; filimargo,
380; filum, 378, 61; laxata, 371, 60, 61; parvicauda,
380; texana, 379, 61.

Q

Quercinium, 87; pasanioides, 89, 21, 22; porosum, 89,

21, 22.

R

Redlichia, 418.

Reticuloceras bilingue, 19; aff. gracile, 19; nodosum
19; reticulatum, 19; superbilingue, 19.

Rhabdammina virgata, 400, 64, 65.

Rhabdosporites, 53; langi, 54, 14.

Rhammatopora gasteri, 384, 62; gaultina, 382, 62.

Rhombopora minima, 73, 16.

Richardson, J. B. Spores from the Middle Old Red
Sandstone of Cromarty, Scotland, 45.

Rixon, A. E. and Meade, M. J. Glass fibre resin casts
of fossils, 124.

Rudwick, M. J. S. The feeding mechanism of the
Permian brachiopod Prorichthofenia, 450.

S

Saccammina glenisteri, 401 , 65.

Scittila nasuta, 583, 80.

‘Scorpions’, Carboniferous, 276, 46-57.

Seendia, 579.

Sellanarcestes, 131, 24.

Selwood, E. B. Ammonoids and trilobites from the

Upper Devonian and lowest Carboniferous of the
Launceston area of Cornwall, 153.

Senis, 584.

Silurian: appendage of eurypterid, 245; Bryozoa from
Wales, 69; Dinobulus from Canada, 242; ostracods
from Canada, 93.

Slimonia acuminata, 245, 42, 43.

Smith, A. H. V. Structure of the spore wall in certain
miospores belonging to the series Cingulati Pot.
and Klaus 1954, 82.

Sobolewia, 129, 24.

Sorosphaera adhaerens, 403, 66.

Sporadoceras orbiculare, 174, 29.

Spores: Devonian, 26,45 ; Hoegisporis, new Cretaceous
form genus, 485; structure of miospore wall, 82.

Stictopora fenestra ta, 1, 1 ; nicholsoni, 9, 1,2.

Stictoporella interstincta, 23, 10.

Sudeticeras aff. ordination, 17.

T

Tarlo, L. B. Downtonian ostracoderm Corvaspis
Kingi Woodward, with notes on the development of
dermal plates in the Heterostraci, 217.

Terebrirostra arduennensis, 574.

Tholosporites densus, 37, 13; punctatus, 38, 13; tenuis,
38, 13.

Thomas, H. Dighton, and Larwood, G. P. The
Cretaceous species of Pyripora d'Orbigny and
Rhammatopora Lang, 370.

Tolypammina helina, 407, 67 ; nexuosa, 407, 67.

Tortarctica, 585 ; similis, 80.

Toucasia, 578.

Townrow, J. A. The Peltaspermaceae, a pteridosperm
family of Permian and Triassic age, 333.

Trace-fossil, Lias, 64.

Trapezicardita, 581.

Traquairaspis Campbell i, 38.

Triassic: pteridosperms, 333; young of the lamelli-
branch Lima, 362.

Trilobites: alimentary caeca in agnostids and others,
410; Australosutura from the Carboniferous, 227;
Devonian and Carboniferous from Cornwall, 153.

V

Venilicardia, 587.

Verrucosisporites grandis, 31, 11; variabilis, 30, 11.

W

Waterston, C. D. The median abdominal appendage
of the Silurian eurypterid Slimonia acuminata
(Salter), 245.

Wattisonia, 318; coseleyensis, 319, 56.

Wills, L. J. External anatomy of some Carboniferous
‘scorpions’, Part 2, 276.

Wocklumeria sphaeroides, 159, 26; W. zone, 237.

X

Xylophagella zonata, 590.

THE PALAEONTOLOGICAL ASSOCIATION

COUNCIL 1960

President

Professor O. M. B. Bulman, Sedgwick Museum, Cambridge
Vice-Presidents

Dr. R. G. S. Hudson, Trinity College, Dublin
Professor T. N. George, The University, Glasgow
Treasurer: Professor P. C. Sylvester-Bradley, Department of Geology,

The University, Leicester

Assistant Treasurer: Dr. T. D. Ford, Department of Geology,

The University, Leicester

Secretary: Dr. Gwyn Thomas, Department of Geology,

Imperial College of Science, London, S.W. 7
Assistant Secretary: Dr. C. H. Holland, Department of Geology,

Bedford College, London, N.W. 1

Editors

Dr. W. H. C. Ramsbottom, Geological Survey Office, Ring Road, Halton, Leeds, 15
Mr. N. F. Hughes, Sedgwick Museum, Cambridge
Dr. W. S. McKerrow, University Museum, Oxford

Other members of Council

Dr. D. V. Ager, Imperial College of Science, London
Dr. F. T. Banner, British Petroleum Company, Sunbury on Thames
Mr. M. A. Calver, Geological Survey Office, Leeds
Dr. W. G. Chaloner, University College, London
Dr. A. J. Charig, British Museum (Natural History), London
Dr. L. R. Cox, British Museum (Natural History), London
Dr. R. H. Cummings, The University, Glasgow
Dr. J. C. Harper, The University, Liverpool
Professor F. Hodson, The University, Southampton
Professor L. R. Moore, The University, Sheffield
Dr. Dorothy H. Rayner, The University, Leeds
Mr. J. D. D. Smith, Geological Survey and Museum, London
Dr. I. Strachan, The University, Birmingham
Dr. C. J. Stubblefield, Geological Survey and Museum, London

Overseas Representatives

Australia: Professor Dorothy Hell, Department of Geology, University of Queensland, Brisbane,
Queensland

Canada: Dr. D. J. McLaren, Geological Survey of Canada, Department of Mines and Technical
Services, Ottawa

New Zealand: Dr. C. A. Fleming, New Zealand Geological Survey, P.O. Box 368, Lower Hutt
Pakistan: Dr. M. Haque, Geological Survey of Pakistan, P.O. Box 15, Quetta, West Pakistan
West Indies and Central America: Dr. L. J. Chubb, Geological Survey Department, Kingston, Jamaica
Eastern U.S.A.: Professor H. B. Whittington, Museum of Comparative Zoology, Harvard Uni-
versity, Cambridge 38, Mass.

Western U.S.A.: Dr. J. Wyatt Durham, Department of Paleontology, University of California,
Berkeley 4, Calif.

PRINTED IN GREAT BRITAIN
AT THE UNIVERSITY PRESS, OXFORD
BY VIVIAN RIDLER
PRINTER TO THE UNIVERSITY

'S'-U