> V" 5 j> w 5 >

2 co 2 co * 2 co *

1 LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIWS S3IBVaa
co 2 co — co

o “ W' 5 5 m

2 _j 2: -1 2 ' _i 2

ON NOlinillSNI NVINOSHIIWS S3IHVH8n LIBRARIES SMITHSONIAN INSTITUTION NOliniil!
2 <2 > z f- 2 r- 2

0 ro X - co /^J^X - 00 xToovSx o X^i^

^ x%% 3 t; fS4/;

£ « * (ti.Jl £ r

m xi^osHjx 2 m

in — in _ co

m LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIWS S3iaVHail L I B R A R
5 ^ ^ 2 co 2 <5 2 ,

^ .x -1 /sAWKn 7 xx =i yyy/,. 2 -1 wa///s, z

Fft o

■yjrrM*. “>
x 7>v j-

00 2 w •••*• 2 Jo 2 CO

ON NOlinillSNI NVINOSHIIWS S3IHVaan LIBRARIES SMITHSONIAN INSTITUTION NOliniil!
CO 2 v co — (O 2 CO

o

2

m LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIWS SBIHVHan LIBRAR

m qc>2 z m

co — co “ co __

ion NouniiiSNi nvinoshiiws S3iavaan libraries Smithsonian institution Nounur

CO 2 co 2. ...CO 2 CO

II 1 I 1 «g< I ^

9 «fl| 1 %% | § %% | *

> -W" 2 > W 5 ^ > xXjiis^X 2 ^ >

2 CO 2 CO A 2 00 k 2

1 LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIWS S3IHVaan LIBRARI
co 2 co — co

O \°iv pc/ “ O •x-CP' ds^ x^)t|iStx^ q

ion2 NouniiiSNi ""nvinoshiiws ^s 3 1 avae n~’u b rar i es^ Smithsonian institution Noiiniu:

2 1 ■ 2 * 2

to

m X- \^W ” X^VA$t£2 fT| X^VOSVO>2 2 m

co _ co _ co

311 LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIWS S3iaVaan LIBRAR

2 r CO

h

y-

ION NOlinillSNI NVINOSHIIWS S3IBVaan LIBRARIES SMITHSONIAN INSTITUTION NOlinillS

THSONIAN

CO * ^ co *

institution NouniiiSNi nvinoshiiws saiavaa

- m
> ^

Z CO

1 LIBRARIES SMITHSONIAN

2 .. 1/5

co .

i\ — cc

noshiiins s 3 i y v a a n libraries Smithsonian institution noiidiiisni nvinoshiiws S3i

^ 13 Z r- 2 r- v 2

° m XT2°*qX ° m /TSo^oX O “ \\ O

33

> |p|

33

rn _ m

— co _ co ± co _

THSONIAN INSTITUTION NOimillSNI NVINOSH1IWS S3IHVyail LIBRARIES SMITHSONIAN INS'

. Z * CO 2 CO 2 * CO 2-

I s% q 1 g | -

00 \v3«l3f|} CO y&OjW CO ^fc|] 5 Mat CO

X 2 vyjffi'' jE I F X if# 2

_ _ „ _ >

co ■** z co 2 co z 55

N0SH1IIAIS S3 iyvaan_ LIBRARIES SMITHSONIAN INSTITUTION N0I1 LSNI_ NVINOSHIIIMS S3

THSONIAN INSTITUTION NOIlfUllSNI NVINOSH1IIMS SaiUVUail LIBRARIES SMITHSONIAN INS'

NOSH1IINS S3IHVyan LIBRARIES SMITHSONIAN INSTITUTION NOlinilJLSNI NVINOSH1IINS S3

„ . > I " "V I W I I ^

CO CO * 2 CO

THSONIAN _ INSTITUTION NOIlfUllSNI NVINOSH1IWS S 3 I B V H 3 1 1 LIBRARIES SMITHSONIAN INS1

2 . . 60 ^ — co — co —

!2 If w ^ « w

O " _ Xj'ins^z o \CJJ0C2 “ ■cjjgp’-' o

'JOSHiiws S3 lavaan libraries Smithsonian-1 institution ^NouniiisNi^NviNOSHiitMs^ss

S f- Z r- 2 r~ v 2

r & *

m ^ x_ivas^2 rn

— CO ± CO _ co _

THSONIAN INSTITUTION NOIlfUllSNI NVIN0SH1IWS S3iaVHail LIBRARIES SMITHSONIAN INST

z », CO 2 CO 2 CO -r

< A S 2 ^55*v < S z£?R7D>\ <

Krffmlv

i
CO

— vW'*7 o

t '(My

> 2 X^osv'AZ > 5

sioshjli iajs saiavaan lirrarifr rm ITHQDM I fi N IWCTITIITiriM M D I I D I I I <? M I MHIMOCU I IIAIC^ C 1 I

VOLUME 9

Palaeontology

1966

PUBLISHED BY THE
PALAEONTOLOGICAL ASSOCIATION
LONDON

Dates of publication of parts in Volume 9

Part 1, pp. 1-181, pis. 1-27
Part 2, pp. 183-354, pis. 28-59
Part 3, pp. 355-522, pis. 60-82
Part 4, pp. 523-705, pis. 83-109

21 March 1966
18 July 1966
25 October 1966
23 December 1966

THIS VOLUME EDITED BY N. F. HUGHES, GWYN THOMAS, ISLES STRACHAN, AND M. R. HOUSE

© The Palaeontological Association, 1966

PRINTED IN GREAT BRITAIN

CONTENTS

Part Page

Alvin, K. L. Two cristate megaspores from the Lower Carboniferous of Scotland 3 488

Barker, D. See Kaye, P.

Barr, F. T. The foraminiferal genus Bolivinoides from the Upper Cretaceous of the

British Isles 2 220

Upper Cretaceous foraminifera from the Ballydeenlea Chalk, County Kerry,

Ireland 3 492

Bolton, T. E. Some late Silurian Bryozoa from the Canadian Arctic Islands 3 517

Brett, D. W. Fossil wood of Anacardiaceae from the British Eocene 3 360

Brunton, H. Predation and shell damage in a Visean brachiopod fauna 3 355

Bruton, D. L. A new odontopleurid trilobite genus from the Devonian of Bohemia 2 330

Churkin, ML, Jr. Morphology and stratigraphic range of the phyllocarid crustacean

Caryocaris from Alaska and the Great Basin 3 371

See Lenz, A. C.

Clarkson, E. N. K. Schizochroal eyes and vision of some Silurian acastid trilobites 1 1

Schizochroal eyes and vision in some phacopid trilobites 3 464

Craig, G. Y., and Jones, N. S. Marine benthos, substrate and palaeoecology 1 30

Donovan, D. T. The Lower Liassic ammonites Neomicroceras gen. nov. and Paracym-

bites 2 312

Glaessner, M. F., and Wade, M. The late Precambrian fossils from Ediacara, South
Australia 4 599

Grant, R. E. Late Permian trilobites from the Salt Range, West Pakistan 1 64

Habib, D. Distribution of spore and pollen assemblages in the Lower Kittanning Coal

Measures of western Pennsylvania 4 629

Holdsworth, B. K. Radiolaria from the Namurian of Derbyshire 2 319

Hubbard, J. A. E. B. Population studies in the Ballyshannon Limestone, Ballina Lime-
stone, and Rinn Point Beds (Visean) of NW. Ireland 2 252

Hughes, N. F., and Moody-Stuart, J. Descriptions of schizaeaceous spores taken from
early Cretaceous macrofossils 2 274

Jenkins, T. B. H. The Upper Devonian index ammonoid Cheiloceras from New South
Wales 3 458

Johnson, H. M. Silurian Girvanella from the Welsh borderland 1 48

Johnson, J. G. Middle Devonian brachiopods from the Roberts Mountains, central
Nevada 1 152

Parachonetes, a new Lower and Middle Devonian brachiopod genus 3 365

Jones, N. S. See Craig, G. Y.

Kaye, P., and Barker, D. Ostracoda from the Upper Tealby Clay (Lower Barremian)
of South Lincolnshire 2 208

Large, N. F. See Savage, R. J. G.

Lenz, A. C., and Churkin, M., Jr. Upper Ordovician trilobites from northern Yukon 1 39

Macurda, D. B., Jr. The Devonian blastoid Belocrinus from France 2 244

Medd, A. W. The fine structure of some Lower Triassic acritarchs 2 351

Moody-Stuart, J. See Hughes, N. F.

Oliver, W. A., Jr. Description of dimorphism in Striatporaflexuosa Hall 3 448

Ormiston, A. R. Occurrence of Australosutura (Trilobita) in the Mississippian of

Oklahoma, U.S.A. 2 270

IV

CONTENTS

Part

Page

Palframan, D. F. B. Variation and ontogeny of some Oxfordian ammonites: Tara-
melliceras richei (de Loriol) and Creniceras renggeri (Oppel), from Woodham,
Buckinghamshire 2 290

Pedder, A. E. H. The Upper Devonian gastropod Orecopia in western Canada 1 142

Pocock, Y. P. Devonian schizophoriid brachiopods from western Europe 3 381

Pollard, J. E. A non-marine ostracod fauna from the Coal Measures of Durham and
Northumberland 4 667

Rawson, P. F. A phylloceratid ammonite from the Speeton Clay (Lower Cretaceous)
of Yorkshire 3 455

Savage, R. J. G., and Large, N. F. On Birgeria acuminata and the absence of labyrin-

thodonts from the Rhaetic 1 135

Shergold, J. H. A revision of Acaste downingiae (Murchison) and related trilobites 2 183

Sdzuy, K. An improved method of analysing distortion in fossils 1 125

Shiells, K. A. G. A new productid brachiopod from the Upper Visean of Scotland 3 426

Stearn, C. W. The microstructure of stromatoporoids 1 74

Strusz, D. L. Spongophyllidae from the Devonian Garra formation. New South Wales 4 544

Tavener-Smith, R. The micrometric formula in the classification of fenestrate crypto-
stomes 3 413

Wade, M. See Glaessner, M. F.

Walker, C. A. Podocnemis somaliensis, a new pleurodiran turtle from the Middle

Eocene of Somalia 3 511

White, D. E. The Silurian rugose coral Microplasma lovenianum Dybowski from

Monmouthshire 1 148

Ziegler, A. M. Unusual stricklandiid brachiopods from the Upper Llandovery Beds
near Presteigne, Radnorshire 2 346

The Silurian brachiopod Eocoelia hemisphaerica (J. de C. Sowerby) and related

species 4

523

VOLUME 9 • PART 1

Palaeontology

MARCH 1966

PUBLISHED BY THE
PALAEONTOLOGICAL ASSOCIATION
LONDON

Price £3

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, etc., on payment of the
appropriate annual subscription :

Institute membership . . . . £7. (U.S. $20)

Ordinary membership . . . £3. 3s. (U.S. $9.50)

Student membership £2. 2s. (U.S. $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; on first
applying for membership, they should obtain an application form from the Secretary
or the Treasurer. Subscriptions are due each January, and should be sent to the
Treasurer, Dr. C. Downie, Department of Geology, The University, Mappin Street,
Sheffield 1, 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
1966 will receive Volume 9, Parts 1 to 4.

All back numbers are still in print and may be ordered from B. H. Blackwell,
Broad Street, Oxford, England, at the prices shown below (post free):

Vol. 1 (for 1957-8) in 4 parts at £2 or U.S. $6.00 per part.

Vol. 2 (for 1959) in 2 parts at £2 or U.S. $6.00 per part.

Vol. 3 (for 1960) in 4 parts at £2 or U.S. $6.00 per part.

Vol. 4 (for 1961) in 4 parts at £2 or U.S. $6.00 per part.

Vol. 5 (for 1962) in 4 parts at £3 or U.S. $9.00 per part.

Vol. 6 (for 1963) in 4 parts at £3 or U.S. $9.00 per part.

Vol. 7 (for 1964) in 4 parts at £3 or U.S. $9.00 per part.

Vol. 8 (for 1965) in 4 parts at £3 or U.S. $9.00 per part.

A complete set. Volumes 1-8, consists of 30 parts and costs £76 or U.S. $228.

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, Department of Geology, Sedgwick Museum,
Downing Street, Cambridge, England. A sheet of detailed instructions for authors
will be supplied on request.

© The Palaeontological Association, 1966

SCHIZOCHROAL EYES AND VISION OF SOME
SILURIAN ACASTID TRILOBITES

by E. N. K. CLARKSON

Abstract. The eyes of some British Silurian Acastinae are described and analysed functionally. Techniques are
given for measuring the size and spatial arrangement of the lenses of schizochroal eyes and the angular bearings
of their lens-axes, from which may be inferred the angular range of vision of the whole eye and the relative
visual acuity in different directions. In the post-larval ontogeny of the eye of Acaste downingiae (Salter) the
lens-number increases with size but the angular extent of the visual field remains constant. Two eye-variants
are distinguished and attributed tentatively to sexual dimorphism. Comparisons are made with the adult eyes
of Acaste downingiae macrops (Salter) and Acastoides constricta (Salter).

Variation of lens-size and strong differentiation between the relative acuity in the horizontal and vertical
directions of vision are perhaps the most significant features of the eyes of Acastinae, and the latter is inter-
preted as indicative of movement-perceiving eyes. Analogies are made with the compound eyes of recent
arthropods, and the ecological significance of the work is briefly discussed.

Trilobites are amongst the earliest arthropoda to appear in the geological record
and are the oldest fossil animals known to have possessed compound eyes. The mor-
phological similarity between the eyes of trilobites and those of other arthropods was
noted by the first students of trilobite morphology, including Quenstedt (1837) and
Burmeister (1846). Since then, trilobite eyes have been studied in detail by many
workers; in particular Barrande (1852), Clarke (1889), Lindstrom (1901), Brink (1951),
and Beckmann (1951). An excellent summary of current knowledge is given by Harring-
ton (1959) in the Treatise on Invertebrate Paleontology Part O (Arthropoda I), pp. O 87-
O 92. In the main these workers have contributed to our knowledge of the morphology
and development of trilobite eyes, and there has been little attempt to relate structure
to function.

I have made a functional study of the eyes of a number of species of trilobites.
This paper presents the results of the first part of this investigation and concerns a small
group of trilobites which have proved especially suitable for such functional analysis.

Within the Trilobita there are two distinct types of eyes, defined by Clarke (1889)
as ‘holochroal’ and ‘schizochroal’. Full descriptions of the two types are given by
Harrington, and it is sufficient to state here that holochroal eyes consist of many small
closely packed lenses overlaid by a single corneal membrane, through which the lenses
may usually be seen. Schizochroal eyes have a fairly small number of relatively larger
lenses, the maximum number recorded being 770 (Roy 1933). Each lens has its own
cornea, and is separated by interstitial material from its neighbours. Whereas holo-
chroal eyes are present in trilobites of all ages from Lower Cambrian to Middle Per-
mian, schizochroal eyes, which are to be found only in the Phacopina, a few Cheirurina
and, possibly, Harpina, do not occur before the Ordovician or after the Devonian.
They are consequently to be regarded either as a specialized development of the normal
holochroal pattern or as quite distinct organs of independent origin. As yet it is not
known which of the two possibilities is the more probable. The suborder Phacopina,
in which only schizochroal eyes occur, appeared in the early Ordovician after the main

[Palaeontology, Vol. 9, Part 1, 1966, pp. 1-29, pis. 1-3.]

B 6612

B

2

PALAEONTOLOGY, VOLUME 9

lines of trilobite descent had been established (Stubblefield 1959), and reached its acme
comparatively late in the history of the Trilobita when the class as a whole was declining.

Schizochroal eyes of Phacopina have been found to be particularly suitable for use in
the investigation of the nature of trilobite vision because of their large size and few
lenses. Several species of Phacopina are common in the English Wenlock Limestone
(Silurian); amongst them are Acastinae in a state of preservation suitable for study.
These are described in this paper.

TERMINOLOGY

The eyes of Phacopina have been frequently figured and Lindstrom’s (1901) magnifi-
cent monograph contains figures of some of these eyes in thin section. Their detailed
morphology has seldom been described. In the succeeding text descriptions, although
the terminology of the Treatise following Warburg (1925) has been used, it has been
necessary to supplement this with a number of new terms.

These terms, illustrated in text-fig. 1, are applicable to the eyes of Acaste and all other
Phacopina. The external surface of the eye which bears the lenses and is bounded by the
facial suture is known as the ‘ visual surface '. The limits of the visual surface nearest
to the sagittal fine are defined as the ‘ anterior ' and ‘ posterior edges'. The ‘ interlensar
sclera ’, defined by Clarke (1889), is that part of the visual surface which lies between the
schizochroal lenses and is not covered by cornea. It is generally identical in structure
with the rest of the cephalic exoskeleton.

All the lenses are invariably arranged conspicuously in hexagonal close packing
and thus form three intersecting rows, defined as the ‘ dorso-ventral files' and the
‘ ascending ' and ‘ descending diagonal rows'. The former term is intended to replace
the somewhat misleading expression ‘vertical rows’, as these files usually diverge ven-
trally and are not always vertical. Since the dorso-ventral files are more frequently
referred to in the text than the other rows, the term ‘ file ' by itself is normally used to
distinguish them. The files are numbered 1, 2, 3, . . . etc., from the anterior edge of the
eye. Lenses occur in alternate files at the same level from the front to the rear of the
visual surface, and lie in ‘ horizontal rows'. The ‘ upper horizontal row' is defined as that
which extends the full distance below the facial suture. Within this row the central
lenses are always smaller than the outer ones. In Acaste and other Phacopina there is
often an ‘ accessory upper horizontal row' lying above the latter. It is confined to the
posterior part of the visual surface and generally consists of a few lenses only. In some
cases the regularity of the pattern in which the lenses are distributed is disrupted by the
interpolation of one or two extra lenses between certain files. The files on either side of
these extra lenses are usually bowed outwards, but a short distance from the hiatus all the
files are quite straight and normal in appearance. This condition has been observed in a
small specimen of Acaste downingiae s. str. (Salter), SM A 28744 (text-fig. 4a, b, PI. 1,
figs. 1, 4, 5).

Discussion of the angular relationships of the lens-axes involves the use of the terms
‘ longitudinal ' and ‘ latitudinal axial angles' which refer to the measurable angles of
separation between the axes of adjacent lenses. Since the lenses in two neighbouring
files are situated at different levels and hence usually have incongruent latitudinal bear-
ings, the longitudinal axial angles are taken as half the angle between the axis of any one

E. N. K. CLARKSON: EYES AND VISION OF SILURIAN TRILOBITES

3

.fixigena

librigena

descending
diagonal row

accessory upper*,
horizontal row ' i"%..

upper .

horizontal raw"

occipital v

furrow . 1 ‘ j

glabellar furrows

posterior
marginaT
furrow /•

facial

suture

sclera 11 th dorso-
ventral file

L ( max)

L (min)

lower rim

palpebral

area

palpebral
furrow

posterior

edge

palpebral

lobe

- lateral
cephalic border

fixigena

W

(min)

W

(max)

text-fig. 1. Acaste downingiae (Salter) s. str. External morphology of eye and immediate environs
After SM A 28720 (eye-variant A); a. Lateral view. b. Dorsal view. Dimensions; L, W, H, length

width, height.

4

PALAEONTOLOGY, VOLUME 9

lens and that of the lens in the same horizontal row (i.e. at the same level) in the next file
but one. A latitudinal axial angle is the angle of separation between the axis of any
lens and that of the lens directly above or below it in the same dorso-ventral file.

The term ‘facet’ which has sometimes been substituted for ‘lens’ is not used here
since it has already been applied in trilobite terminology to designate the articulating
antero-lateral part of the thoracic pleurae in trilobites capable of enrolment.

TECHNIQUE OF INVESTIGATION

A simple apparatus was devised in order to measure accurately the angular bearings of the individual
lenses of schizochroal eyes. A set of such measurements of all the lenses of one eye, when plotted
graphically, gives the total angular range of vision, and the angular separations between the lens-axes.
This apparatus was also used for a detailed study of the sizes and the spatial arrangement of these
lenses, and their mode of development during postlarval ontogeny.

Apparatus. The apparatus (text-fig. 2) consists of two units : a stereoscopic microscope fitted to a long
arm stand, allowing the optical assembly to be either horizontal or vertical, and a turntable. (The
microscope illustrated is by W. R. Prior & Co. Ltd.) The specimen selected for analysis is fixed firmly
on the horizontal turntable, which can be rotated so that different parts of the visual surface of the
eye may be examined in turn.

The optical system used has a maximum magnification of X 40. The left eyepiece (A) may be fitted
either with a micrometer scale registering to mm. for measurement of lens diameters and spacing,
or with interchangeable graticules ruled in 0-5 mm. or 0-25 mm. squares. An aluminium collar (B)
with its pointer (C) was fitted to the left eyepiece. This can be locked by the screw (D). In determining
the angular bearings of the lens-axes the pointer (C) is aligned parallel with one of the sets of lines on
the graticule. The whole eyepiece assembly complete with collar and pointer may be rotated inside its
barrel. An immovable protractor (F) is held by a Perspex collar (G) on the barrel of the eye-piece
below the aluminium collar. The 90° bearing of the protractor is fixed horizontally, and upon rotation
of the eyepiece assembly any angle measured by means of the graticule will be shown by the pointer
(C) on the protractor. To avoid rotation of the graticule during measurement the eyepieces have been
adjusted for personal eye-spacing. They are locked by means of the clamp (E).

The rigid turntable consists of an immovable 6| in. wooden disc (H) on which is fixed a 6-in.
circular protractor (J). Above the protractor is a rotating upper assembly of Perspex. This consists
of three graduated discs, of which the lowermost (K) is the largest, and has a pointer (L) bearing on the
protractor (J). This disc rotates about a central screw, and has on its upper surface two parallel bars
grooved so that the dovetailed fixture (N) of the base of the next disc (M) may slide between them.
This disc (M) carries on its upper surface another dovetailed fixture at right angles to that on the
lower surface, engaging with grooved bars on the lower surface of the uppermost disc (P). The plane
surface of the top disc (P) is inscribed with two sets of lines at right angles, one parallel to the pointer
(L), the other perpendicular to it. If the specimen is mounted in plasticine on the upper disc, either
of its eyes may be brought into the field of view of the microscope by sliding the discs, and held level
whilst the stage is rotated.

Orientation of the trilobite. All measurements of the angular bearing of the lens-axes are made with
primary reference to the trilobite’s plane of bilateral symmetry, the sagittal line. This fine is orientated
parallel with one of the lines engraved upon the upper surface of the disc (P), so that its bearing can be
measured on the protractor (J) by means of the pointer (L).

The chief problem arises in the orientation of the cephalon in profile, i.e. with respect to the horizontal
plane of the apparatus. A conventional ‘horizontal’ orientation has been chosen which probably
approximates to the original attitude of the cephalon as it was in life (p. 27). This orientation is
defined as that in which the upper edge of the visual field, or at least the greater part of it, lies along
a single latitude. In this attitude the ‘horizontal’ rows of lenses in the eye are in fact aligned horizon-
tally, a feature which may be used in practice as a guide to the orientation of the specimen on the stage.

Size and spacing of lenses. It is desirable to make a graphical representation of the whole visual surface,
in order to show the number, size, and spacing of the lenses in every file, and to trace their ontogenetic

E. N. K. CLARKSON: EYES AND VISION OF SILURIAN TRILOBITES

5

text-fig. 2. Apparatus used in the investigation of vision in trilobites. a, b, perspective views from top

and side. Explanation in the text.

6

PALAEONTOLOGY, VOLUME 9

development and intraspecific variation. Although Barrande (1852, p. 514) had drawn up a table
showing intraspecific variation in the number of lenses per file in Phacops fecundus Barrande, Clarke
(1889), working with Phacops rana (Green), was the first to attempt to show diagrammatically the
number and arrangement of the lenses. He represented each lens as a dot equidistant from its neigh-
bours. The dorso-ventral files were shown as vertical, and the diagonal rows as inclined at 60° to the
vertical. Beckmann (1951) used a similar method, but showed the diameters of the lenses with greater
accuracy. Although this projection can give an accurate representation of the sizes of the lenses it
necessarily distorts their relative positions.

The problem, as in cartography, is to find an appropriate projection which will depict a portion of
an approximately spherical surface with minimal distortion. The projection chosen resembles a conical
equidistant projection in character though not in method of construction. This projection (text-fig. 4)
shows both the spacing of the lenses and their diameters with relatively little distortion. The method of
construction used is as follows (text-fig. 3). A long central file (A) is selected and observed in profile;
the distances between the lens centres may then be measured accurately by the eyepiece micrometer
and plotted on paper at an appropriate magnification (text-fig. 3c). The specimen is then turned so that
the visual surface is perpendicular to the observer, and the distances between the lens-centres of
adjacent files are measured by triangulation from the lenses of the marked file A whose positions have
already been recorded.

The lenses of files B and C are normally at equal distances from those of file A and one set of measure-
ments will therefore suffice for both. The positions of the lens-centres of further files are then plotted
in the same way giving a grid for all the lens-centres. Slight irregularities in spacing often occur near
the anterior and posterior edges of the visual surface and in such cases separate measurements must
be made for the spacing of each lens. When the grid is complete the measured diameters of all the lenses
may be drawn in.

Angular bearings of the lens-axes. Before measuring the angular bearings of the lens-axes, the eyes are
lightly coated with ammonium chloride and are illuminated against a dark background. The longi-
tudinal and latitudinal components of the angular bearings of each lens are measured separately.

The longitudinal component is measured with the microscope vertical so that the eye is seen directly
from above (text-fig. 3a). The left eyepiece assembly with its pointer (C) is rotated in its fixed barrel
until the set of lines on the graticule parallel with the pointer has the same angular bearing as the axis
of the lens. The other set of lines on the graticule are now parallel with the principal plane of the lens.
The pointer now records the longitudinal component on the protractor (F). Each lens is examined in
turn by this method.

The microscope body is restored to the horizontal position (text-fig. 2), in order that the latitudinal
components may be measured. For each lens, the turntable is rotated so that the pointer (L) shows the
already recorded longitudinal bearing of the particular lens under investigation and its profile may be
observed through the microscope as it projects from the curving surface of the eye (text-fig. 3b). The
graticule is then used for the determination of the latitudinal bearing of this and other lenses in the
manner already shown.

All the dorso-ventral files are brought into view in turn for study but with continued rotation of the
turntable, either the glabella or the high posterior border may move into the field of view and obscure
the lenses of the first or the last few files. The turntable is then reversed through 180°, and the lenses
are measured from the opposite direction.

In both sets of measurements, the close spacing of the lines in the graticule ensures considerable
accuracy. Repeated measurements on the same eye have shown that the margin of error is generally
less than 1° and never more than 2°. The ease of measurement with which this degree of accuracy can
be attained depends, however, on the detailed morphology of the lenses and the sclera.

In Acaste the individual lenses project above the surface of the interlensar sclera, and the junction
between lens and sclera is normally very distinct (text-fig. 1). In some other Phacopina, and particularly
in Phacops s.s. (text-fig. 3a, b) the sclera may be highly inflated so that each lens appears sunken in
a shallow crater-like depression. When one of these lenses is seen directly in profile, the lens/sclera
junction may be just obscured by the inflated sclera. The uppermost lenses may, furthermore, be set
obliquely to the surface of the sclera, and the lower rims of the depressions in which they lie may be
more strongly pronounced than the upper rims. All the axes of such lenses in consequence have a lower

E. N. K. CLARKSON: EYES AND VISION OF SILURIAN TRILOBITES

7

text-fig. 3. Technique of measurement of the spatial and angular relationships of the lenses, a, b.
Angular directions of lens-axes of Phacops latifrons Bronn. The two sets of perpendicular lines re-
present the central cross wires of the graticule, and parallel lines following the optic axis and principal
plane of the lens. Other graticular lines have been omitted for clarity. Lens A has latitudinal and
longitudinal axial bearings of 28° and 81° respectively; c, Triangulation of lens-centres in files B and
C from those of file A; d, Lambert equal-area net with notation adopted for recording bearings of
lens-axes. The bearing of the axis of lens A (vide text-fig. 3a and b ) is indicated.

latitudinal elevation than if they had been set flush with the surface, with their principal planes parallel
with the outside of the sclera. In making accurate estimations of the latitudinal bearing of such lens-
axes, the lens must be kept in focus as it is being rotated into position for measurement.

When the longitudinal and latitudinal components have been measured, the resultant bearings of
all the lens axes are plotted upon a Lambert equal-area stereographic net (text-figs. 5, 6). Text-fig. 3 d,
shows the notation adopted in the description of the axial bearings of the lenses.

The term ‘ visual field' is used here to include the area enclosed by the peripheral lens-axes. As each
individual ommatidium presumably had a certain visual field of its own, however, the actual field of

a

A 28744 R

A 28744 L

A 28764 R

A 28720 R

A 28743 R

j BU H 132 A28719R

(composite)

text-fig. 4. Projected visual surfaces of Acastinae. All X 10. a-f, Acaste downingiae (Salter) s. str.,
eye-variant A ; g-h, A. downingiae s. str., eye-variant B ; j, Acaste downingiae macrops (Salter), composite
figure from both eyes. The blank circles represent badly weathered lenses; k, Acastoides constricta
(Salter). The anterior edge is at the left-hand side in all these diagrams. Thus right eyes (R) are viewed
from within; left eyes (L) from outside. In SM A 28744 the visual surfaces are slightly damaged.
Irregularities in lens-distribution are marked by arrows.

E. N. K. CLARKSON: EYES AND VISION OF SILURIAN TRILOBITES

9

-fig. 5. Visual fields and axial bearings of lenses shown by stereographic projection, a-f Acaste
downingiae (Salter) s. str., eye-variant A.

10

PALAEONTOLOGY, VOLUME 9

text-fig. 6. Visual fields and axial bearings of lenses (cf. text-fig. 5) g, h, Acaste downingiae (Salter)
s. str., eye-variant B. j, Acaste downingiae macrops (Salter), composite figure from both eyes;

k, Acastoides constricta (Salter).

view of the trilobite eye as a whole probably extended some way outside the visual field shown on the
stereogram (p. 25). The term is retained for convenience of description.

FUNCTIONAL DESCRIPTIONS OF THE EYES OF ACASTINAE
Acaste downingiae (Murchison 1839)

This species occurs in the Wenlock Limestone (Silurian) of Dudley and the Welsh
Borderland. Murchison’s original description (1839, p. 655, pi. 14) is very brief, and
the fullest existing accounts of the species are those of Salter (1853, 1864) in which six
varieties are enumerated. In the definition of these varieties, Salter used various morpho-
logical criteria, including the external structure of the eye, which, as he observed,
varies considerably within the species. Although the following detailed studies of the
eyes of A. downingiae have given further information which might be of some taxonomic
value, it is not the purpose of this paper to attempt any further systematic interpretation
of the species, and Salter’s varieties, as revised by the Richters (1954), are retained,

E. N. K. CLARKSON: EYES AND VISION OF SILURIAN TRILOBITES 11

pending taxonomic revision of A. downingiae and related species by Shergold which is
now in press (personal communication).

The exoskeleton of this species appears to have been remarkably resistant to pressure,
as compared with some other Wenlockian Phacopina and Cheirurina which are fre-
quently found to be crushed. Hence 40 out of the 62 specimens of 'A. downingiae ’ in the
Sedgwick Museum, with cephala ranging in length from 3-5 mm. to 11 mm., are almost
perfectly preserved and show no trace of distortion. Most of the very best specimens
are enrolled and a few are partially silicified. Extremely small, though crushed, specimens
are known and in some of these the lenses can be distinguished.

Almost all the Sedgwick Museum specimens have been identified as A. downingiae
s. str. (Salter). Of Salter’s other varieties, no material showing intact eyes of A. downingiae
inflatus (Salter), A. downingiae cuneatus (Salter) or of Acastella spinosa (Salter) has been
available for examination. A single almost perfect specimen of A. downingiae macrops
(Salter) with well-preserved eyes was lent by the Geology Department of Birmingham
University and two specimens of Acastoides constricta (Salter) were in the Sedgwick
Museum collection. These latter species are described in a later section of this paper.

Acaste downingiae (Salter 1864) s. str.

Plates 1,2; Plate 3, figs. 1-3.

1839 Calymene ? downingiae Murchison, p. 655, pi. 14, fig. 3 a.

1851 Phacops downingiae (Murch. sp.); M‘ Coy, p. 160.

1853 Phacops downingiae Murchison; Salter, p. 1-12, pi. 1, figs. 1, 2, 5-13, 15.

1864 Phacops ( Acaste ) downingiae Murchison, var. a vulgaris Salter, p. 26, pi. 2, figs. 17-25.

1954 Acaste ( Acaste ) downingiae (Murchison 1839); R. and E. Richter, pp. 16-17, pi. 3, figs.

36-41.

A comparison of Salter’s descriptions and figures with the Sedgwick Museum speci-
mens indicates that within Salter’s ‘variety’ there are at least two distinct forms differing
in the size and position of the eye, the character of the visual surface, the degree of
inflation of the glabella, and the character of the genal angles.

The eyes of the two forms which he figured are quite distinct and do not have any
morphological intermediates. About three-quarters of the fully grown Sedgwick
Museum specimens have large eyes with more widely spaced lenses and rounded genal
angles as in Plate 1, fig. 8 (cf. Salter 1853, fig. 1). This form will be termed eye-variant A.
The rest of the specimens have smaller eyes with more closely spaced lenses and pointed
genal angles (ibid. fig. 2) and are referred to as eye-variant B. The two types of eye also
differ in the number and arrangement of the lenses within the visual surface.

The existence of two eye-variants, which are often associated with minor morpho-
logical differences in the cephalon, has also been detected in Phacops rana (Green), P.
latifrons (Bronn), P. breviceps Barrande, and P.fecundus Barrande by the present author.

In Acaste downingiae s. str. five fully preserved specimens of eye-variant A (SM A
28744, -64, -49, -20, and -43) and two of eye-variant B (SM A 28737, -41) have been
selected for analysis as representative examples of individuals of different sizes. Com-
parisons were made throughout with other specimens of equivalent dimensions. In
each eye-variant specimens of the same size normally have a very similar number of
lenses.

12

PALAEONTOLOGY, VOLUME 9

Eye-Variant A. Text-figs. \a, b; 4a-f; 5 a-f; la-e\ 9a-e; Plate 1, figs. 1-9; 2, figs. 1^1;
3, figs. 1-3.

Eyes of moderate size, not particularly prominent, occupying the central third of the
cephalic length (sag), and extending from glabellar furrow 3p almost to the posterior
marginal furrow. The anterior edge almost reaches the axial furrow, but the posterior
edge is free and lies further from the sagittal line. In profile (PI. 3, fig. 1) the eye-base
lies about half way between the genal angle and the crown of the glabella, but the palpe-
bral lobe does not quite reach the level of the latter. Curvature of visual surface in
dorsal aspect quite strong, profile curvature extremely low (text-fig. 1, PI. 1, fig. 8).
The visual surface rises abruptly from an indistinct eye-platform, is bounded ventrally
by a shallow groove, and projects slightly laterally. Both the upper and lower eye borders
are parallel in young specimens but in older ones the lower border is downwardly convex.
Palpebral lobe granulated, consisting of a narrow flattened outer strip running parallel
with the facial suture, and sloping first gently, then steeply to the palpebral furrow.
In some specimens a narrow line of shallow pits can be seen near the outer rim of the
palpebral lobe running parallel with the facial suture (PI. 2, fig. 3). Palpebral area very
granular, shelving gently towards the axial furrow, open posteriorly.

Lenses (see Table 1) normally disposed regularly in 22-24 files, but in the smallest
specimen (SM A 28744. Text-fig. 4 a-b, 5 a-b, la-b) there are two extra lenses in the right
eye and a hiatus in the centre of the left eye, which breaks up the regularity of the visual
surface. Such irregularities are uncommon. In the smallest specimen the angle at
which the files diverge ventrally is high, but in the largest specimens some of the central
files are subparallel. The lenses of small specimens are almost uniform in size (Table II
and text-fig. 9) and spaced at about the distance of a lensar radius from each other (cf.
Salter 1853, pi. 1, fig. 8). Subsequently, size differentiation appears, and in the larger
specimens the upper central region of the visual surface contains small lenses lying
between the horns of a crescent of larger lenses which increase in diameter downwards.
At the lower ends of the diagonal rows, the lenses are often small and immature. In
SM A 28744 (text-fig. 4 a-b) the lowest lenses are elliptical and set close in under the
lenses above them, but in larger specimens the corresponding lenses are never elliptical.
An accessory upper horizontal row is usually present, extending posteriorly from about
the centre of the eye.

The interlensar sclera may be flat or slightly inflated, and is covered with minute
granules. Ventrally they become continuous with the granules of the lower rim of the
visual surface.

EXPLANATION OF PLATE 1

Acaste downingiae Salter, s. str. Eye-variant A. Wenlock Limestone, Silurian. Dudley, England.

Figs. 1, 2, 4, 5, 7. Small specimen, SM A 28744. Right eye, lateral aspect, x20. 2. From above, x 5.
4, 5, 7. Right eye, lateral aspect, X 20.

Figs. 3, 6. SM A 28749. Left eye, lateral and dorsal aspects, X 7-5. In fig. 6 the eye of a neighbouring
specimen of the same species is visible on the left-hand side.

Fig. 8. SM A 28764. Right eye, lateral aspect, X 15.

Fig. 9. SM A 28720. Lenses near lower rim of right eye, X 25.

Palaeontology, Vol. 9

PLATE 1

CLARKSON, Eyes of acastid trilobites

E. N. K. CLARKSON: EYES AND VISION OF SILURIAN TRILOBITES

13

Eye-variant B. Text-figs. 4 g, h; 6 g, h; 8 a-b; 9fi g; Plate 2, figs. 5-8.

Eyes smaller than in eye-variant A, situated more towards the rear of the cephalon
and not reaching anteriorly quite as far as glabellar furrow 3p. Palpebral lobe more
highly inflated than that of eye-variant A, curving uniformly downwards to the axial
furrow, with less trace of a palpebral furrow. Similar shallow pits to those seen in eye-

table 1 — Distribution of lenses
Acaste downingiae (Salter) s. str.

Eye-variant A

Specimen

Number of lenses per file ( front to rear)

Total

Max. per
file

SM A 28744 (Rt)*

345

666

667

677

555

4 (2) 33

334

32

111

7

SM A 28744 (Rt)f

345

666

667

677

676

6 (2) 66

554

32

127?

7

SM A 28744 (Lt)*

345

666

667

677

7 (5) 66

654

333

22

121

7

SM A 28744 (Lt)f

345

666

667

677

7 (5) 66

666

554

32

130?

7

SM A 28764

345

666

767

676

776

766

543

2

122

7

SM A 28749

345

677

778

788

888

766

654

3

138

8

SM A 28720

345

677

777

787

777

767

665

43

140

8

SM A 28743

1456

788

889

889

888

877

665

43

159

9

N.B. In SM A 28744 the upper posterior parts of both eyes are slightly damaged.

* Number of lenses actually preserved. t Probable original number. ( ) Small extra lenses

interpolated between complete files and breaking up the regularity of the visual surface (p. 12).

Eye-variant B

SM A 28737

345

556

567

666

656

655

543

—

109

7

SM A 28741

345

677

778

787

777

676

654

32

136

8

Acaste downingiae macrops (Salter)

567

889

989

989

999

887

654

3

163

Acastoides constricta (Salter)

SM A 28719

456

666

676

666

565

543

—

—

98

7

variant A lie in a curve just inside the facial suture. Visual surface very compact, just
over twice as wide as it is high. All the dorso-ventral files diverge ventrally at quite a
high angle. Lenses (Tables 1, 2) arranged in a similar pattern to those of eye-variant A
but very much more closely packed. The largest lenses are often contiguous, even in
small specimens.

Character of visual field. As can be seen on the stereogram (text-fig. 6g, h), the longitu-
dinal visual limits are normally 0-180°, in both eye variants. The latitudinal limits are
somewhat variable, but are generally in the region of 10-20° in eye-variant A and 8-20°
in eye- variant B. The upper visual limit is fully latitudinal (text-fig. 6 g, h) except at
the extreme anterior and posterior edges where it rises slightly higher. The lower limit

14

PALAEONTOLOGY, VOLUME 9

is latitudinal anteriorly, but towards the rear the elevation of the lowest lens-axes rises,
so that a shallow postero-ventral ‘lacuna’ is formed. In some cases a smaller anterior
lacuna may be present.

table 2 — Dimensions of cephalon, eyes and lenses (in mm.)

Acaste downingiae (Salter) s. str.

Eye-variant A

Specimen

Cephalon

Eye

Lens diameters

*L(sag.)

W

L(max.)

L{min.)

Wimax.)

W(min.)

H

Large

Average

Small

SM A 28744

3-5

7-5

1-9

1-55

1-2

0-9

1

01

01

—

SM A 28764

5-5

12

2-5

2-1

1-5

1-25

1-3

014

0-125

—

SM A 28749

7-5

14

2-9

2-5

2-2

1-5

2

0-175

0-165

—

SM A 28720

10-5

18

3-5

3-1

2-25

1-5

2

0-25

0-23

0-1

SM A 28743

115

21

4

3-6

2-4

1-75

2-25

0-275

0-25

0-175

Eye-variant B

SM A 28737

5

9

1-75

1-5

1 25

1

1

0-14

0-125

—

SM A 28741

8

15

3

2-75

2

1-75

1-75

0-175

0-165

0-05

Acaste downingiae macrops (Salter)

BU H 132

6-5

11

3-5

2-5

2-5

2-0

2-25

0-25

0-2

—

Acastoides constrict a (Salter)

SM A 28719

7

10

3

2-3

2

1-5

1-7

0-2

0-175

0-075

* L, W, H = length, width, and height as in text-fig. 1.

The strong differentiation between plan and profile curvature of the visual surface
results in the concentration of the lens-axes into distinct and widely spaced ‘ visual strips’’ ,
which of course correspond to the dorso-ventral files and diverge radially from the
top to the bottom of the visual field. In the anterior and posterior visual strips of the
larger specimens (text-fig. 5 e,f) the lowest lens axes form a higher angle with the polar
meridian than the upper ones. In the region of file 8 each strip tends to lie in a single
longitude, but anteriorly and posteriorly they become more oblique and are often curved.
The average spacing of adjacent strips is in the region of 7° at the top and 10° at the
bottom, but posteriorly it may increase to 10° and 15° respectively.

EXPLANATION OF PLATE 2

Figs. 1-4. Acaste downingiae (Salter) s. str., eye-variant A. Wenlock Limestone. Silurian. Dudley,
England. 1-3, SM A 28720, right eye. 1, Lateral aspect, X7-5. 2. Lenses near the anterior margin,
X 25. 3, Dorsal aspect, X 7-5. 4, SM A 28743. Right eye, lateral aspect, X 7-5.

Figs. 5-8. Acaste downingiae (Salter) s. str., eye-variant B. Same locality and horizon. 5, 6, SM A
28737. Left eye, lateral and dorsal aspects, x20. 7, 8, SM A 28741. Left eye, lateral and dorsal
aspects, x7-5.

Palaeontology, Vol. 9

PLATE 2

CLARKSON, Eyes of acastid trilobites

r\v

E. N. K. CLARKSON: EYES AND VISION OF SILURIAN TRILOBITES 15

Angles separating adjacent lens-axes within the strips are very low. Even in the youngest
stages, they are only 1° or 2° in the long central files, and 3° or 4° in the outer files.

The visual field of both eye-variants is very similar, but in eye-variant B the visual
strips are more regular and less highly slanted.

Post-larval development of the visual surface. Ontogenetic studies of the eyes are compli-
cated by the existence of the two eye-variants. This has led to divergent interpretations
of ontogeny. For example, Clarke (1889) charted the number and arrangement of the
lenses in various specimens of Phacops rana (Green), and showed that there was a defi-
nite relationship between the number of lenses in the eye and the size of the specimen.
Equating size with age he concluded that the number of lenses increased from youth to
maturity, and then decreased from maturity to senility. He believed that the extra lenses
present in the smaller specimens had been later absorbed by sclerosis, due to the growth
of the interlensar sclera or the migration of the lenses below the surface of the palpebral
lobe. The strong inflation of the sclera in the older specimens was used as evidence in
support of this hypothesis.

Beckmann (1951), who worked with Phacops cf. breviceps and Phacops schlotheimi,
found no such apparent diminution in lens number with senility, but only an increase
until full maturity was reached. Further, using Clarke’s measurements of specimens of
P. rana, Beckmann produced a graph showing the number of lenses compared with the
width of the cephalon. This graph indicated that Clarke’s material was composed of
two distinct elements, in one of which the cephala were small and the lens number large,
and in the other the cephala were large and the lens number small. He suggested that
since Clarke’s material came from various localities in North America, two ecologically
distinct faunas were present which had probably been mixed prior to examination.
In the species which Beckmann investigated there was apparently only one fauna, and
hence only one type of eye.

The probability that Clarke had inadvertently included more than one type of eye in
his analyses is enhanced by Stumm’s (1953) systematic revision of P. rana. Stumm dis-
tinguished a number of subspecies using the character of the eye as a diagnostic feature.
The significance of Stumm’s work is to be discussed in a later paper.

Neither Beckmann nor Clarke referred to the work of Barrande (1852, p. 514, pi. 21)
who recorded the lens number and distribution in 12 specimens of Phacops fecundus
Barrande, noting that the largest specimens never had as many lenses as the smaller
ones. Barrande attributed this to the considerable degree of intraspecific variation
characteristic of this species. His figures (pi. 21, fig. 19) show that the longest file of the
largest specimen only has 5 lenses, whereas the maximum number recorded, in a smaller
specimen, was 9. In the largest specimens the interlensar sclera is inflated and the lenses
are rather widely spaced. Hence there is a strong resemblance between the condition
of P. rana and that of P. fecundus.

The present studies tend to confirm Beckmann’s view that two distinct elements were
present in Clarke’s material. But in Acaste downingiae s. str. the two variants commonly
occur in faunas which were collected at a single locality and are, therefore, unlikely to
be the result of ecological control. No evidence of lensar sclerosis or absorption by the
palpebral lobe has been collected, and Clarke’s hypothesis of decrease of lens number
with senility has been abandoned.

16

PALAEONTOLOGY, VOLUME 9

In the earlier growth stages the distinction between the two eye variants is less marked.
Immature specimens of eye-variant A may have small genal spines resembling those of
eye-variant B, which become progressively reduced and finally disappear. Thus in the
series referred to, A 28744 has a short genal spine, A 28764 has pointed genal angles,
A 28749 has a barely perceptible point, and in A 28720 and A 28743 the genal angles
are fully rounded, with no trace of a point.

During ontogeny the eyes in eye-variant A double in length and more than double
in height in their development from the youngest preserved stage. The relative increase
in height is due to an increase in number of lenses in the files. The angle at which
adjacent files diverge decreases through development. The size and spacing of the
lenses increase progressively from youth to maturity but remain approximately con-
stant relative to the size of the whole visual surface.

The conclusions of Clarke and Beckmann, that the lenses increased in number with
age, and in a particular order, have been confirmed in general, though it seems that the
eye of A. downingiae is rather more variable than the species which they investigated.

From Table 1 and the diagrams of the visual surface (text-fig. 4) we may draw the
following conclusions:

1. The full lens complement of the upper and accessory upper horizontal rows was
already developed from an early stage (SM A 28744) and no more lenses were added to
the upper ends of the dorso-ventral files; these rows remained complete and separate
entities. Beckmann showed that in P. cf. breviceps and in P. schlotheimi some lenses
were added at the upper ends of the files, but his youngest specimens seem to represent
an earlier stage in development than any yet seen in A. downingiae, where the youngest
specimen has already a maximum of seven lenses per file. It is possible that in earlier
stages in the development of the eye of A. downingiae lenses may have been added
immediately below the facial suture.

2. As Clarke has suggested, new lenses were added exclusively at the lower ends of the
files, i.e. at the lowest points of the visual surface, at least during the growth stages
studied. But as the species is more variable than those which Clarke and Beckmann
studied, it has not been possible to produce diagrams like those of Beckmann (Abb.
278/3), showing the order of development of the lenses. The newly developed lenses are
small and set closely below the preceding lenses. Generally they are circular but in the
young stages of eye-variant A they are elliptical. The immature lenses may be compared
with the corresponding lenses on larger specimens, and by comparison with the latter
it appears that only when they have grown to their full size do new lenses arise below
them. The manner of emplacement of new lenses appears to resemble that described by
Clarke for P. rana. He showed by sections that the 1 ommatidial cavities ’ arise by evagi-
nation of the sclera from below, and these cavities meet with the outer surface where

EXPLANATION OF PLATE 3

Figs. 1-3. Acaste downingiae (Salter) s. str., eye-variant A. Wenlock Limestone. Silurian. Dudley,
England. 1-3, SM A 28720. Lateral, dorsal, and anterior aspects of enrolled specimen, x 2.

Figs. 4-9. Acaste downingiae macrops (Salter). Same locality. BU HI 32, 4, 5, 6, 9. Lateral, anterior,
and dorsal aspects of enrolled specimen, X4. 7, 8. Right and left eyes, lateral aspect, X7-5.

Figs. 10-12. Acastoides constricta (Salter). Same locality. 10, SM A 28719. Right eye, lateral aspect.
11, 12, SM A 28719. Lateral and dorsal aspects of enrolled specimen, x3£.

Palaeontology, Vol. 9

PLATE 3

CLARKSON, Eyes of acastid trilobites

E. N. K. CLARKSON: EYES AND VISION OF SILURIAN TRILOBITES 17

there may be a corresponding inward invagination. Some of his sections showed im-
mature cavities which had not penetrated the outer surface. The size of all the lenses
increases progressively during ontogeny, remaining approximately constant relative to
the size of the whole visual surface. There are marked variations, however, in the size
of the lenses in different parts of the visual surface. The lenses along the upper border,

text-fig. 7. Latitudinal (a^ bu etc.) and longitudinal (a2, b2, etc.) angular separations of the lens-axes,
axial angles represented by contoured stereograms (cf. text-fig. 5). a-e. Acaste downingiae (Salter) s. str.,
eye- variant A. a, SM A 28744, left eye; b, SM A 28764, right eye; c, SM A 28749, left eye, d, SM A

28720, right eye; e, SM A 28743, left eye.

and to a lesser degree, the lower border remain relatively small, whereas those in the
antero-lateral and postero-lateral sectors become relatively large. The differentiation is
more pronounced in the later growth stages.

Changes within the visual field are represented by a series of stereograms (text-
figs. 7, 8) which show the angular separations of the lens-axes by contouring. Latitudinal
and longitudinal separations are indicated on separate diagrams. For latitudinal
separations 1° contours were found to be suitable, but owing to the irregular spacing of
the visual strips, 2° and 4° contours have been used to show longitudinal separations.

c

B 6612

18

PALAEONTOLOGY, VOLUME 9

It is apparent from these and from the original stereograms (text-figs. 5, 6), that
throughout the development of the eye, there are no major changes in the angular
extent of the visual field. Apart from some slight intraspecific variation, the angular
range of vision of an eye with 127 lenses differs little from one with 159. The longi-
tudinal extent of vision of specimens with 22, 23, or 24 files is virtually the same (180°)
and whether there is a maximum of 7, 8, or 9 lenses per file, the latitudinal extent is
always of the order of 11° in eye-variant A, or 13° in eye-variant B. The extra lenses
added with growth are orientated so that their axes remain within the prescribed visual

text-fig. 8. Latitudinal (au bu etc.) and longitudinal (a2, b2) angular separations of the lens-axes. Cf.
text-fig. 6. a, b, Acaste downingiae (Salter) s. str., eye-variant B. a, SM A28737, left eye; b, SM A 28741
right eye; c, Acaste downingiae macrops (Salter), BU H132, composite figure from both eyes;
d, Acastoides constrict a (Salter), SM A 28719, right eye.

field. As the number of lenses increases, their latitudinal separation decreases, particu-
larly (text-figs. lax, bx, cx, etc., 8al5 bx, q), in an elliptical region in the lower central part
of the visual field surveyed by the longest files (40-120° long.; 10-17° lat.). At all stages
during ontogeny zones of increasing latitudinal separation are concentric with this
region and the highest separations are found at the anterior and posterior visual
limits. The profile curvature of the visual surface remains virtually constant throughout
growth. The latitudinal elevation of the lowermost lens-axes must rise as new lenses
are added below them, hence the axial bearing of any one of these new lenses will be
similar to the original direction of the lens-axis now above it.

Changes in the longitudinal separation of the visual strips are less apparent, chiefly
because so few lenses are added at the anterior and posterior edges of the eye. The total
number of visual strips increases only from 22 to 24, and the pattern in which these
strips are arranged remains constant throughout (text-figs. 5, 6). The appearance of the
visual strips reflects the topographical situation of the lenses on the visual surface.

E. N. K. CLARKSON: EYES AND VISION OF SILURIAN TRILOBITES 19

When the eye of A. downingiae s. str. is examined from directly above (text-fig. lb),
each lens of the anterior files appears slightly offset on the curving visual surface with
regard to the lenses below and above it so that the axes of the lower lenses are directed
more anteriorly than those of the upper lenses of the same file. This angular difference
decreases posteriorly so that in the longest files all the lenses in one file have about the
same longitudinal axial bearing. In the succeeding files the reverse situation to that of
the anterior files is apparent.

Thus on the stereograms (text-figs. 5, 6) the separation of the visual strips correspond-
ing to these files increases vertically except for the central longitudinal strips, which
correspond to files 8 and 9. In young specimens the visual strips are almost straight.
In older specimens the new lenses which have been added to the lower ends of the
diagonal rows each have a slightly different longitudinal axial bearing from the lenses
about them (except in the longitudinal strips), so that the total longitudinal range of
the diagonal visual strips increases with age. On the stereogram the strips now become
more highly slanted and may be curved. The tendency for the axis of the top lens of any
one strip to have the same longitudinal bearing as that of the bottom lens of the ad-
jacent strips, in the anterior and posterior parts of the visual field, appears in the youngest
stages and becomes more pronounced during the later development of the eye. This is
apparently a device which enables the whole of the visual field to be covered by the
visual strips with maximum economy of lenses. The apparent complexity of the con-
toured stereograms showing longitudinal separations of the lens-axes (text-figs. 7 a2,
b2 , c2, etc.; 8 a2, b2) is the result of the variability in spacing of the visual strips, which
may range from 5° to 12° or more.

Comparison of the two sets of contoured stereograms (text-figs. 7 ax, a2, bx, b2, etc.,
8 aL, a2, bx, b2) indicates that regions of maximum or minimum longitudinal separation
do not necessarily coincide with like regions of latitudinal separation.

At this point it is interesting to consider the coverage of the visual field by lenses of
varying sizes. This is again best represented by contoured stereograms (text-fig. 9).
These stereograms indicate how, whilst the lenses increase in size during ontogeny, those
regions of the visual field covered by the largest and by the smallest lenses remain rela-
tively constant. The actual pattern of coverage differs in the two eye-variants. Thus in
eye-variant A, the largest lenses survey regions close to the anterior and posterior visual
limits, of which the anterior has the wider extent; whereas in eye- variant B, the lenses
are more uniform in size and the two regions are less well marked.

Finally, the above stereograms (text-fig. 9), may be compared with the paired stereo-
grams showing axial separations (text-figs. 7, 8). There is some degree of relationship in
this species between the size of the lenses and their latitudinal axial separations. The
largest lenses cover the regions where axial separations are high, i.e. at the extremities
of the visual field. The relationship between lens sizes and longitudinal axial separations
is somewhat less obvious owing to the greater variability of the latter.

Acaste downingiae macrops (Salter 1864)

Plate 3, figs. 4-9.

1864 Phacops ( Acaste ) downingiae Murchison var. fi macrops Salter, pp. 26-27, pi. 2, figs. 26-29.

Salter’s (1853) description of A. downingiae included a brief account and figure of the

20

PALAEONTOLOGY, VOLUME 9

text-fig. 9. Contoured stereograms showing coverage of different regions of the visual fields by lenses
of various sizes, a-e, Acaste downingiae (Salter) s. str., eye-variant A. a, SM A 28744, left eye;
b, SM A 28764, right eye; c, SM A 28749, left eye; d, SM A 28720, right eye; e, SM A 28743, left eye;
/, g, Acaste downingiae (Salter) s. str., eye-variant B . f SM A 28737, left eye; g, SM A 28741, right eye;
h, Acaste downingiae macrops (Salter), BU H 132, composite figure from both eyes; j, Acastoides

constricta (Salter), SM A 28719, right eye.

E. N. K. CLARKSON: EYES AND VISION OF SILURIAN TRILOBITES 21

‘large-eyed’ form of the species, in which he noted that the visual surface was almost
double the size of the normal type and possessed about 180 lenses. In 1864 he gave this
varietal status.

A single small enrolled specimen, BU H 132 from the Wenlock Limestone of Dudley
has been analysed. Both eyes are preserved undistorted but in each the visual surface is
slightly damaged and some lenses are missing. The surface of the left eye is the more
complete but a few of the posterior lenses have been destroyed. The posterior region
of the right eye is entire but anteriorly patches of lenses are absent. The diagram (text-
fig. 4 j) representing the projected visual surface is composite and is derived from
measurements of the existing parts of both eyes.

Morphology of eyes. Eye very large, prominent, set towards the rear of the cephalon,
and extending from glabellar furrow 3p to the posterior marginal furrow. Both the
anterior and posterior edges of the eye lie almost in the same exsagittal plane. In profile
the base of the eye is situated rather low on the cephalon, at less than half the distance
from the genal angle to the crown of the glabella, and the palpebral lobe stands at about
the same level as the latter. In plan the visual surface is parabolic and sharply flexed
about the transverse line; profile curvature moderate, strongest nearest the upper and
lower borders. The visual surface rises sharply from the indistinct eye platform, is
bounded ventrally by a shallow groove, and projects laterally near to the antero-lateral
border of the cephalon. Both eye borders are parallel. Palpebral lobe moderately broad
(tr), flat and smooth, declining very sharply adaxially to the sunken palpebral furrow,
which at its point of maximum flexure in the transverse line turns sharply through 90°.
Palpebral area smooth, shelving gently toward axial furrow and open posteriorly.

Lenses (see Table 1) disposed regularly in 22 dorso-ventral files.

All files diverge ventrally at a very low angle. Lenses quite closely packed, showing
comparatively little variation in size (see Table 2), though in the central files the largest
lenses are slightly smaller than those of the penultimate anterior and posterior files.
The first and last files have small lenses and the lowest lenses of the diagonal rows are of
about the same diameter. Very small lenses are present in the centre of the upper hori-
zontal row. An accessory upper horizontal row is present, which extends from files 13
to 17.

Interlensar sclera planar and finely granular. A hexagon of larger granules surrounds
each lens.

Character of visual field. The stereogram (text-fig. 6j) shows the visual field, which has
longitudinal visual limits of about 0-180°, with a small anterior overlap, and a maximum
longitudinal range of 0-28°. The upper visual limit is fully latitudinal, whereas the lower
is equatorial only between files 1 and 12, and rises posteriorly so that there is a fairly
deep postero- ventral lacuna with a maximum latitudinal extent of 0-18°.

The lens-axes form distinct visual strips, which contrast with those of A. downingiae
s. str. in several respects, most notably in that all except the few posterior strips are
longitudinal. As a result of the parabolic plan curvature of the visual surface, these
strips tend to be clustered in two regions, in front of and behind the polar meridian,
where they are normally separated by angles of less than 10° longitude (text-figs. 6j, 8c2).
Between the more widely spaced central strips the separation may be as high as 15°.

Differentiation in the profile curvature of the visual surface leads to strong variation

22

PALAEONTOLOGY, VOLUME 9

in latitudinal separation of the lens-axes within each strip (text-figs. 67, 8q). In a
longitudinal band lying between 8° and 20° lat. axial angles are low (2-3°), but above
and below this band the separation of lens-axes increased quite regularly to 6° or more.

The regions of minimal latitudinal and longitudinal separation almost coincide in
A. downingiae macrops. The composite text-fig. 9 h shows the coverage of the visual field
by lenses of both eyes. Although the latitudinal extent of vision is greater than in A.
downingiae s. str. a remarkably similar pattern is apparent. Likewise the sizes of the
lenses bear no relationship to their axial separations.

Acastoides constricta (Salter 1864)

Plate 3, figs. 10-12.

1864 Phacops ( Acaste ) downingiae Murchison, variety or subspecies e constrict us Salter, pp.

27-28, pi. 2, figs. 13-16.

1954 Acaste ( Acastoides ) constricta (Salter 1864); R. and E. Richter, pp. 17-18, pi. 3, figs.

42, 43.

Salter (1864) noted that the eyes of this species were small and set very close to the
glabella, having a total of 130 lenses each.

A single small enrolled specimen (SM A 28719) from the Wenlock Limestone (Silurian)
of Dudley, was analysed. The right eye is complete and perfectly preserved. The left is
damaged.

Morphology of eyes. Eyes small, not very prominent, lying close to glabella, situated in
the posterior part of the cephalon, and extending from between glabellar furrows (lp)
and (2p) to the posterior marginal furrow. Both the anterior and posterior edges lie
in the same exsagittal plane, the anterior being separated from the glabella by a narrow
space. In profile the base of the eye is set high upon the cephalon, almost two-thirds
of the distance between the genal angle and the crown of the glabella, and the palpebral
lobe is level with the latter. Plan curvature moderate, increasing behind the transverse
line, profile curvature about half the former, decreasing posteriorly. The visual surface
rises sharply from librigena, is bounded ventrally by a very shallow groove, and does not
project laterally as far as the antero-lateral cephalic border. Both eye-borders are paral-
lel. Palpebral lobe moderately broad (tr), inflated and somewhat tuberculate, bounded
adaxially by a row of shallow pits, and shelving first gradually, and then steeply to the
indistinct palpebral furrow. Palpebral area small, almost flat, and open posteriorly.

Lenses (text-fig. 4Ar and Table 1), disposed regularly in 18 dorso-ventral files. All files
diverge ventrally at a rather low angle. Lenses packed so closely that over most of the
visual surface they are hexagonal and only the smallest lenses are rounded. Within each
file they increase ventrally in diameter, though in the first and last two files their diameter
is constant. An accessory upper horizontal row of small round lenses extends from files
8 to 16. The lenses of the anterior two or three files are small.

Interlensar sclera hardly discernible between the hexagonal lenses, but where visible
below the facial suture it is planar and smooth.

Character of visual field. The maximum extent of the visual field (text-fig. 6k) is 6-174°
long. ; 0-27° lat. The upper visual limit anteriorly lies slightly above the 27° latitude and
the lower is truncated by a small anterior and a larger shallow posterior lacuna.

E. N. K. CLARKSON: EYES AND VISION OF SILURIAN TRILOBITES 23

The lens-axes are widely separated and form longitudinal visual strips which are
relatively indistinct because throughout the visual field longitudinal separations are
rarely more than twice the latitudinal. The variation in longitudinal axial angles (text-
figs. 6 k, 8d2) resembles that observed in A. downingiae macrops so that there are two
regions of minimal separation, anterior and posterior to the polar meridian. The latitu-
dinal axial angles, however, vary in a manner quite unlike those hitherto observed
(text-figs. 6k, 8 df. The region of minimal latitudinal separation lies directly above the
postero-ventral lacuna, and the axial angles increase regularly to the upper and anterior
regions.

The largest lenses (text-fig. 9j), unlike those of A. downingiae cover the lowermost
central part of the visual field, but this region of coverage does not really correspond
with the areas of maximal or minimal axial separation. Whereas the structure and visual
characters of the eyes of A. constrict a differ radically from those of A. downingiae,
similar features have been observed in Dalmanites caudatus (Brunnich) and in other
Phacopina.

COMPARATIVE VISUAL PHYSIOLOGY

A fuller interpretation of the physiology of vision in these trilobites will be deferred
until descriptions of the visual attributes of other schizochroal eyes have been published,
but the nature of vision in the trilobites studied here is briefly considered with reference
to that of recent arthropods.

Visual powers of single ommatidia. The fundamental units of which all modern compound
eyes are constructed are the radially arranged ommatidia. Each of these has the form of a
tapering cylinder whose upper end is capped by a lens capable of transmitting light to
the photoreceptive retinular cells below. In the eyes of trilobites only the lenses and
the interlensar sclera are ever preserved and the structure of the eye below this level is
unknown. It must be assumed that the underlying regions were similar in their organiza-
tion to those of recent arthropods, and that below each lens lay an ommatidium-like
unit.

Considering the remarkable variability in the separation of the lens-axes and pre-
sumably therefore of the ommatidia, it is of some importance from the physiological view-
point to know the angular receptivity to light of single ommatidia, and whether or not
the visual fields of adjacent photoreceptors could have overlapped. The possible con-
ditions in trilobites cannot be ascertained directly, but may be inferred from the com-
parable eyes of recent arthropods, though even in these the functions of single ommatidia
as individual units is hardly known.

Burtt and Catton (1954) working with Locusta, and Waterman (1954) with Limulus,
have shown that single ommatidia have a wide range of directional sensitivity. In
Locusta each ommatidium has a visual field of about 20° solid angle, and in Limulus
it has been shown that whereas high light sensitivity is centred within 10-20° of the
optic axis, responses were still given to reasonable light intensities up to 80° or 90° away
from it.

In such eyes the visual fields of adjacent photoreceptors must overlap considerably.
It is possible that in trilobites the cones of light-receptivity of neighbouring ommatidia
were wide enough to extend laterally into the spaces between their axes and even between

24

PALAEONTOLOGY, VOLUME 9

the visual strips at a short distance from the visual surface, and hence there may have
been no gaps in the object space. As in Limulus, however, high sensitivity was probably
limited to the region of the optic axis.

Types of perception in compound eyes. Von Buddenbrock (1935) classified compound
eyes in a scale of progressive efficiency as being capable of (1) fight and dark perception,
(2) direction perception, (3) movement perception, and (4) form perception. Only the
more advanced compound eyes are capable of the last, which as recently shown (Burtt
and Catton 1962; Rogers 1962) seem to operate by the production of successive dif-
fraction images, which can only work efficiently in an eye consisting of large numbers of
small lenses. In these eyes, visual acuity may be three or four times as high as the angular
separation of the lens-axes would indicate. The schizochroal eyes of trilobites with their
small number of large lenses, are of a lower grade of organization than those of many
insects. It is unlikely that their large lenses could form diffraction images and the strong
degree of astigmatism of the eyes of Acaste which would naturally result from the
extreme differentiation in lens-axis separations in different directions would tend to
preclude form perception.

Movement perception in Acastinae. The astigmatism, which in Acastinae must amongst
other factors have limited the possibility of image formation, nevertheless seems to have
been an intrinsic part of their visual organization, and could indicate that the eyes may
have been adapted for the perception of movement. Astigmatic eyes are known in other
arthropods. Amongst insects, strong differentiation in the surface curvature of the eye
and in the separation of the ommatidia resulting in a considerable degree of astigmatism
was noted by Exner (1891). It has been described in sections of the eye of the honey bee
(Baumgartner 1928) and of many other insects (del Portillo 1936). Some experimental
evidence (Hecht and Wolf 1929) indicates that in the bee visual acuity in different planes
varies with the degree of ommatidial separation. In these insects, however, the angular
separation of adjacent ommatidia in the horizontal axis, even at its maximum, is never
more than three times that in the vertical axis.

In Acaste , however, the equivalent angular separation of lens-axes may be up to
fifteen times greater in the horizontal than in the vertical plane. The visual strips thus
formed are established from a very early stage in ontogeny. Those visual strips, which
are distributed radially in such a manner as to cover the whole of the visual field leaving
no part of its total longitudinal extent unsurveyed by at least one lens-axis, are the most
striking and apparently the most important characteristic of vision in Acaste. Their
existence, however, necessitates the constriction of the overall angular range of vision
to an extremely narrow visual field which in A. downingiae s. str. may be of only 10°
latitude, and thus imposes a limit upon the efficiency of the eye as a whole. The visual
fields of A. downingiae macrops and A. constricta are twice as wide. In the former, well-
developed strips occur in a central longitudinal belt, but in the flanking belts of this
eye, and in the whole visual field of A. constricta visual strips are indistinct, and the
ratio of axial separation in horizontal and vertical planes is only 2:1, like that of many
insects.

The visual fields of both A. downingiae macrops and A. constricta are truncated
ventrally, both anteriorly and posteriorly, by quite deep lacunae, in contrast with the
shallow posterior lacuna of A. downingiae s. str. All such posterior lacunae may be

E. N. K. CLARKSON: EYES AND VISION OF SILURIAN TRILOBITES 25

interpreted as an adaptation which prevents vision in the posterior and ventral region
being obscured by the high posterior border of the cephalon.

The eye of Acaste downingiae s. str. could have detected movement by means of the
visual strips. Burtt and Catton (1962) have stated that ‘for perception of movement
of a discontinuous object there must be discontinuities in the image formed in the eye at
photoreceptor level’. Such discontinuities would be provided by the visual strips. For,
if, as has been already suggested, the highest sensitivity to light of the ommatidia was

text-fig. 10. Acaste downingiae (Salter) s. str., eye- variant A. Anterior view of cephalon showing
latitudinal limits of vision of 10° and 20°. An object of height h approaching the trilobite would not be
detected at A. At B it would come within the lower part of the visual field, and at C it would be sighted

by the uppermost lenses.

centred in the vicinity of the optic axis, then uniform light intensity surrounding the
trilobite would be registered by the trilobite’s eye as a series of lighter and darker bands
relating respectively to the regions covered by and in between the visual strips. Any dark
object moving horizontally within the visual field would pass across successive bands
(strips) and temporarily occlude them. A conception of the size, speed, and direction of
the moving object would be given by the changing pattern. Dr. Rudwick has suggested
how such an eye might also be able to detect the movement of an object towards the
trilobite. It has been shown that the lower limit of the visual field of A. downingiae
s. str. is directed at a few degrees above the equator (text-fig. 10). A distant object at A
would lie below the visual limit or on its periphery. The same object at B would lie
within the lower part of the visual field, and hence would be detected by the lower
lenses. If the object approached to C, it would come within the visual compass of the
upper lenses. Hence the horizontal approach of a predator along the sea floor would
be recorded by the eye as the progressive darkening of the visual strips from the bottom
to the top. To what extent the mutual overlap of the cones of vision of the individual
ommatidia would affect the eye’s efficiency for movement perception is uncertain.

It is also worth noting that the highest concentration of lens-axes per strip is found in
that region of the visual field immediately anterior to the polar meridian, and this
is apparently the area of maximum sensitivity. A comparable condition is known in the
eye of the honey bee ( Baumgartner 1928).

It is clear that if each photoreceptor had a high angular range of fight sensitivity, the
apparent visual field as recorded upon the stereograms must have been surrounded by a
peripheral zone or penumbra, from which visual impressions may have still been

26

PALAEONTOLOGY, VOLUME 9

received even if these were less direct. The outer zones of the visual field of A. downingiae
macrops, which are surveyed by fewer lens-axes than the central horizontal band,
presumably received less distinct visual impressions than the latter but they have the
effect of extending the overall visual field down to the equator and latitudinally upwards.
Within these zones vision appears to have been adapted less for the perception of move-
ment than for the perception of varying light intensity. These flanking zones which
are surveyed by relatively large lenses, cover those regions which in the eye of A.
downingiae s. str. would be blind or penumbric. In A. constricta the visual strips are
indistinct and were less efficiently adapted for movement perception than those of A.
downingiae but this is offset by the advantage of the extended visual field.

The differentiation in lens-size within the visual surface must affect the physiology of
the eye as a whole. The large lenses gathered most light but the small lenses of the upper
central region inevitably transmitted less bright illumination to the photoreceptive cells
below them. Thus the visual strips do not have uniform illumination along their whole
length, and as is shown in text-fig. 7, those strips in which the lens-axes have the lowest
angular separations are not necessarily those covered by the largest lenses. In A.
downingiae s. str. there is, as previously stated, a certain degree of relationship between
the sizes of the lenses in individual eyes and the latitudinal separations of their axes.
The axial separations of the largest lenses, which cover the extremities of the visual
field, are high. This might perhaps be interpreted as an adaptation which would tend
to even the illumination received by all the visual units of the eye, which would other-
wise be unequal because of the lens-size differentiation throughout the visual surface.
In A. downingiae macrops , and A. constricta, however, the areas of the visual field covered
by the largest lenses show little correspondence with the regions covered by the greatest
or least number of lens-axes.

Differentiation of lens sizes within the eyes is a feature common to all Phacopina,
and normally follows a similar pattern to that of Acaste. Acastoides is exceptional. It
may be concluded that throughout the evolution of Phacopina, varying degrees of
differential curvature resulting in different types of vision were superimposed upon visual
surfaces whose pattern of lens-size differentiation was already genetically fixed. This
variation in lens-size apparently imposes a limitation upon the efficiency of vision, but
may have originally been necessitated by the form of the ancestral eye. The eyes of A.
downingiae s. str. may be adapted to function with more uniform sensitivity over the
whole visual field in spite of the lens-size differentiation.

Ecological significance of vision in Acastinae. All the specimens studied were collected
last century from the Silurian Wenlock Limestone of Dudley. The sedimentary features
of this limestone have been described by Butler (1939, p. 56). Relatively shallow water
conditions prevailed throughout deposition and the environment was principally that
of a coral reef, with strong light and almost silt-free water, though at intervals the depth
of water increased and corals were unable to grow. The exact horizons of the fossils
studied and their locations are, however, unrecorded and it is unknown whether A.
constricta and A. downingiae macrops are ever found in the same beds as A. downingiae
s. str. The chief visual characteristics of A. downingiae s. str., i.e. the visual strips and
small lenses, suggest that the eyes were primarily adapted for the perception of moving
organisms in strong light, as would be expected in a shallow water environment.

E. N. K. CLARKSON: EYES AND VISION OF SILURIAN TRILOBITES 27

Whether any inference of a deeper water or otherwise different environment may be
drawn from the eyes of A. downingiae macrops and A. constricta is uncertain and further
purely ecological work is required before any more definite statements can be made.

The two distinct eye-variants in Acaste downingiae s. str. may be tentatively interpreted
as a feature indicative of sexual dimorphism. It has been suggested by previous authors
that sexual dimorphism is exhibited in certain trilobites and is generally apparent as
minor differences in the overall structure of the exoskeleton. More evidence relating to
eye morphology comes from a study of other Phacopina, but analogy may be made with
recent insects and crustaceans, where sexual dimorphism commonly affects the size
and shape of the eye.

Lastly, it may be stated that the results obtained here give a strong indication of the
normal life orientation of the phacopid cephalon. The visual fields of Phacopina and
of Acaste in particular are usually in the form of long narrow bands which in life were
undoubtedly orientated so that the trilobite had 360° vision in a horizontal plane.

When the cephalon is orientated so that the ‘horizontal’ lens-rows are actually aligned
horizontally, then the visual limits are normally latitudinal. If this cephalon is examined
from the side, the antero-lateral border (PI. 3, figs. 3, 8, 9, 12) appears inclined, and the
rear margin of the occipital ring forms a hood-like projection which would cover the
articulating half-ring of the first thoracic segment.

On turning the cephalon towards the observer, the border forms an ‘anterior arch’
(PI. 3, figs. 1, 6). This anterior arch is a constant feature of Phacopina and of many other
groups of trilobites also. In Cheirurina it is often made very evident by the presence of
hypostomes projecting below the antero-lateral border.

It is of interest to note the contrast between the cephalic attitude inferred here and
that which Whittington (1956) deduced as the normal profile orientation of cephala of
Odontopleuridae. In this specialized group he noted that the cephala seem to have been
adapted for resting upon a flat surface on the tips of almost vertical spines produced
from a horizontal antero-lateral border, or upon lateral extensions of the border itself.
A cephalon in this orientation is tilted forwards through about 30°, relative to the atti-
tude envisaged for phacopids, and it seems that the phacopid attitude is indicative of a
different mode of life. Whittard (1939), Warburg (1939), and Whittington and Evitt
(1954) have in turn touched upon the problem of the natural orientation of the cephalon
but further studies are still required. When this has been determined it will be possible to
deduce the manner of articulation of the thorax and pygidium relative to the cephalon.
From these observations the life attitudes and certain environmental adaptations of
such trilobites may possibly be inferred.

Acknowledgements. I would like to thank Professor O. M. B. Bulman for research facilities in the
Sedgwick Museum, Cambridge, and Mr. A. G. Brighton and Dr. C. L. Forbes for access to collections.
Dr. Forbes kindly cleaned many of the specimens. Dr. I. Strachan and Dr. G. R. Coope of Birmingham
University also made collections in their charge available. My thanks are particularly due to Dr. M. J. S.
Rudwick of the Sedgwick Museum, who first suggested this field of study, for advice on many prob-
lems connected with this work. I am grateful to Dr. E. T. Burtt of the Department of Zoology,
University of Newcastle upon Tyne, for a most informative discussion. Dr. G. Y. Craig of Edinburgh
University, and my wife ( nee C. M. Cowie) kindly criticized the manuscript and made many helpful
suggestions. I also wish to thank J. Shergold of the Department of Geology, University of Newcastle
upon Tyne, for taxonomic advice.

The work was carried out during the tenure of a D.S.I.R. Research Studentship.

28

PALAEONTOLOGY, VOLUME 9

REFERENCES

barrande, j. 1852. Systeme Silurien du Centre de la Boheme., Ier Partie. Crustaces; Trilobites, 1,
1-935, pi. 1-51. Prague and Paris.

Baumgartner, h. 1928. Der Formensinn und die Sehscharfe der Bienen. Z. vergl. Physiol. 7, 56-
144.

beckmann, h. 1951. Zur Ontogenie der Sehflache grossaugiger Phacopiden. Palaont. Z. 24, 126-41,
pi. 10.

brink, a. s. 1951. On the compound eye of an unusually large trilobite from the Bokkeveld beds south
of Steytlerville, Cape Province. S. Afr. J. Sci. 47, 162-4.
buddenbrock, w. von. 1935. Die Physiologie des Facettenauges. Biol. Rev. 10, 283-316.
burmeister, h. 1846. The organisation of trilobites deduced from their living affinities. Ray Society,
London.

burtt, E. t. and catton, w. T. 1954. Visual perception of movement in the locust. J. Physiol. 125,
566-80.

1962. A diffraction theory of insect vision. I. An experimental investigation of visual acuity and

image formation in the compound eyes of three species of insects. Proc. roy. Soc., B. 157, 53-82.
butler, a. j. 1939. The Stratigraphy of the Wenlock Limestone of Dudley. Quart. J. geol. Soc. Lond.
95, 37-74, pi. 3.

clarke, J. m. 1889. The structure and development of the visual area in the trilobite Phacops rana
Green. J. Morph. 2, 253-70, pi. 21.

exner, s. 1891. Die Physiologie der facettirten Augen von Krebsen und Insekten. Leipzig and Vienna.
Harrington, H. j. 1959. in moore, R. c., ed. Treatise on Invertebrate Paleontology, Part O Arthropoda
1. Lawrence, Kansas.

hecht, s. and wolf, e. 1929. The visual acuity of the honey bee. J. Gen. Physiol, 12, 727-60.
lindstrom, g. 1901. Researches on the visual organs of the trilobites. K. Svensk. Vetensk. Akad. Handl.
34, 1-86, 6 pi.

m‘ coy, f. 1851. British Palaeozoic Fossils in the Geological Museum of the University of Cambridge.

1-661, pi. 1-3. London and Cambridge.

Murchison, r. 1839. The Silurian System. London.

Portillo, j. del. 1936. Die Beziehungen zwischen den Offnungswinkel der Ommatidien, Krummung,
und Gestalt der Insektenaugen und ihrer funktionellen Ausgabe. Z. vergl. Physiol, 23, 100-45.
quenstedt, A. 1837. Beitrage zur Kenntnis der Trilobiten, mit besonderen Rucksicht auf ihr bestimmte
Gliederzahl. Arch. Naturgesch. 3, 337-52.

richter, r. and richter, e. 1954. Die Trilobiten des Ebbe-Sattels und zu vergleichende Arten (Ord.,
Goth. Dev.). Abh. senckenb. naturf. Ges. 488, 1-76, 6 pi.

Rogers, g. l. 1962. A diffraction theory of insect vision. II. Theory and experiments with a simple
model eye. Proc. Roy. Soc. B. 157, 83-98.

roy, s. K. 1933. A new Devonian trilobite from southern Illinois. Fieldiana: Geology, 6, 67-82.
salter, j. w. 1853. Figures and descriptions illustrative of British organic remains. Decade VII. Mem.
Geol. Surv. U.K.

1864-83. A monograph of British Trilobites. Palaeontogr. Soc. (Monogr.) 1-216, pi. 1-30.

London.

shergold, j. 1966. A revision of Acaste downingiae (Murchison) and related trilobites. Palaeontology,
9, pt. 2.

Stubblefield, c. J. 1959. Evolution in trilobites. Quart. J. geol. Soc. 115, 145-62.
stumm, e. c. 1953. Trilobites of the Devonian Traverse Group of Michigan. Contr. Mus. Paleont. Univ.
Mich. 10, 101-57, pi. 1-12, 1 map.

warburg, e. 1925. The trilobites of the Leptaena Limestone in Dalarne. Bull. geol. Inst. Univ. Uppsala,
27, 1-450, pi. 1-11.

1939. The Swedish Ordovician and Lower Silurian Lichidae. K. Svensk. Vetensk. — Akad. Handl.

17, 1-162, pi. 1-14.

waterman, t. h. 1954. Directional sensitivity of single ommatidia in the compound eye of Limulus.

Proc. Nat. Acad. Sci. Wash. 40, 258-62.

et al. 1961. The Physiology of Crustacea. 1 and 2. Academic Press.

E. N. K. CLARKSON: EYES AND VISION OF SILURIAN TRILOBITES 29

whittard, w. F. 1939. The Silurian Illaenids of the Oslo region. Norsk, geol. Tidsskr. 19, 275-95,
4 pi.

Whittington, h. b. 1956. Type and other specimens of Odontopleuridae. J. Paleont. 30, 504-20.

and evitt, w. r. 1954. Silicified Middle Ordovician Trilobites. Mem. geol. Soc. Amer. 59, 1-137,

33 pi.

E. N. K. CLARKSON

Grant Institute of Geology,
University of Edinburgh,

Manuscript received 30 October 1964 Edinburgh, 9

MARINE BENTHOS, SUBSTRATE AND
PALAEOECOLOGY

by g. y. craig and n. s. jones

Abstract. The distribution of marine benthos in the Irish Sea is influenced by the nature of the substrate. The
majority of epifaunal species are suspension feeders or carnivores and are associated with rocks or coarse-
grained sediments. The majority of infaunal species are deposit feeders and are associated with fine-grained
sediments. The relationships of the organic content of the sediment and the physical nature of the substrate to
the benthos are discussed. It is argued that epifauna, and benthos living in well-sorted sands, are more likely to
be transported after death than most infauna. Fossil epifauna are most likely to be found in situ at unconfor-
mities, disconformities, and bioherms. Over half of the species of epifauna but only one-third of the species of
infauna have hard parts and are preservable as fossils.

Geologists are accustomed to the association of a rock facies and its contained
fauna, and such terms as black-shale facies, sandy facies, and reef facies usually evoke
a picture of well-known depositional environments and their faunas, relatable to the
experience of the individual concerned. Yet in spite of the geologist’s dependence on,
and indeed intimate knowledge of sediments, he rarely pursues so far as he might the
relationship between sediment and fauna. We believe that a discussion of certain aspects
of the relationships between living organisms and their sediments may provide a useful
contribution to a subject which is perhaps more appreciated by ecologists than by
palaeoecologists. (A recent article by Purdie (1964) which appeared after this paper
was submitted shows that geological appreciation of this problem is indeed growing.)

An important aspect of the relationship between marine faunas and sediments was
demonstrated by Petersen (1913), who divided the benthos of the Danish seas into epi-
faunas and infaunas. He recognized as epifauna those animals living upon, or associated
with extraneous matter or vegetation resting on, the substratum forming the bottom,
and as infauna, animals living in the sediments forming the level sea bottom. Various
other terms, such as epibionts and endobionts (Schafer 1956) have been introduced to
overcome etymological objections and to include plants as well as animals. Epibenthos
and endobenthos are perhaps more satisfactory, but the original terms are entrenched in
the literature and will be used here. The epifauna may be divided into sessile forms,
occurring on rock, shells, weeds, and other objects on the sea floor but rarely on sedi-
ments, and active forms which can move freely over the substratum (Buchanan 1958).

Influence of sediment on benthos. Grain size, the most obvious feature of a sediment,
affects the organism both directly and indirectly. Feeding methods are related to grain
size; mode of attachment or movement depends on plasticity or the availability of
hard surfaces; organic content and grain size are frequently a reflection of the strength
of bottom currents; and the aerobic or anaerobic nature of the sediment is generally a
function of grain size, although it is basically dependent on the rate of water circulation.

Grain size and feeding methods. The distribution of the faunas of the central part of the
northern Irish Sea is well known (Jones 1940, 1951, 1952, 1956). From published and

(Palaeontology, Vol. 9, Part 1, 1966, pp. 30-38.]

CRAIG AND JONES: MARINE BENTHOS

31

unpublished data it is possible to calculate the numbers of species of invertebrates on the
various grounds off Port Erin, Isle of Man, in relation to the method of feeding. For
this purpose they can most conveniently be classified as :

1. Suspension feeders (sestonophages)

2. Deposit feeders (detritus collectors and soil swallowers)

3. Carnivores

4. Uncertain feeding habit

There is no sharp distinction between epifauna and infauna, apart from ecological
position, but infaunas are made up largely of species dependent for their food on detritus
or plankton; that is, they are deposit or suspension feeders, together with a proportion
of carnivores. Epifaunas, on the other hand, contain few or no deposit feeders but usually
a number of specialized feeders such as browsers and sponge eaters (Thorson 1957),
although these can in a broad sense be classed respectively as deposit feeders and
carnivores. Epifaunal animals are naturally not adapted to burrowing as are many
members of the infauna. They tend to be better protected than infaunal animals, with
thicker shells, tests or carapaces.

Table 1 is an analysis of the feeding methods of 79 of the more abundant species of
infauna. For each feeding method the species are classified by the grade of sediment

TABLE 1

% on

each type of deposit

Feeding

method

Mud

Muddy

sand

Fine

sand

Gravel /
Shell

Widely

distributed

Suspension

0

20-5

36-5

29

8-5

Deposit

62-5

61-5

54-5

33-5

58-5

Carnivores

37-5

9

9

29

16-5

Uncertain

0

9

0

8-5

16 5

Total number
of species

8

24

11

24

12

to which they are confined or centred upon. Where they occur more or less equally on
two or more kinds of sediment they are classified as ‘widely distributed’.

In Table 2 the feeding habits of 56 species of epifauna are analysed.

TABLE 2

% on

each type of deposit

Feeding

method

Mud

Muddy

sand

Fine

sand

Gravel

Shell

Widely

distributed

Suspension

0

0

0

31-4

33-3

Deposit

0

20

33-4

17-2

16-7

Carnivores

100

80

66-6

45-7

50

Uncertain

0

0

0

5-7

0

Total number
of species

1

5

3

35

12

32

PALAEONTOLOGY, VOLUME 9

In Table 3 all species (368 in number) found on each ground during 1947 (including
epifauna and infauna) have been analysed. No distinction is made in this case between
species confined to one ground and those occurring on several. Where definite informa-
tion is lacking a species is referred to its most probable feeding method.

table 3

0/

/o

on each type of deposit

Feeding

method

Mud

Muddy

sand

Fine

sand

Gravel!

Shell

Suspension

0

17-5

12-5

24-5

Deposit

56-5

52

54-5

26

Carnivores

43-5

305

33

49-5

Total number
of species

23

127

68

150

There is a marked increase in the proportion of suspension feeders with the increase
in grain size, from mud (where none was found) to gravel/shell deposits. Deposit feeders,
on the other hand, show a proportionate decrease from the mud and sand environ-
ments to the coarser sediments. Carnivores depend on the presence of other organisms
and may not be especially affected by grain size.

The grain size of the sediment therefore exhibits considerable control over the type
of feeding mechanism acceptable in the different environments and provides an indi-
cation of the kinds of mechanism that may have been characteristic of past environ-
ments. The relationship is well known (see, for example, Sanders 1956).

Biomass and organic content. Biomass is low in fine-grained muds and well-sorted sands,
and reaches maxima in muddy sand and shelly gravel deposits. Fine-grained sand,
because of its susceptibility to erosion (see Hjulstrom 1935, Inman 1949) and its low
organic content, is a relatively unsuitable medium for the support of benthos. But
with an increase in organic content (and with it almost certainly an increase in the
amount of mud) sand becomes an organically rich medium, ideal for benthonic life. The
resistance of the sediment to erosion is increased, moreover, as the organic material
and mud bind the sand grains together.

Zenkevitch (1963, pp. 89, 90) stated that as a rule in the Barents Sea the larger the
amount of the fine-grained fraction of the sediment, the richer its organic content. In
other words, in regions with favourable conditions for deposition of the fine-grained
fractions, large amounts of detritus are also deposited, but, on the other hand, these
regions are usually unfavourable for the development of bottom life, presumably be-
cause of a low oxygen content in the deposit. However, many regions with a sandy
bottom and a rich fauna may have a low content of organic matter. Good vertical and
horizontal circulation prevent the accumulation of organic matter on the bottom. Hence
though one may accept the rule that seas rich in life have more organic matter in their
sediment, in some a reverse relationship between the amount of bottom fife and organic
matter in the sea bed may be created.

Zenkevitch (p. 138) also quoted Idelson (1930) on the quantitative distribution of the
fauna on the Spitzbergen bank. On the middle part of the bank, where the bed is

CRAIG AND JONES: MARINE BENTHOS

33

washed clean, the fauna is very scarce. At the edges of the shallow, however, the biomass
increases sharply, some 95-99 percent, of it epifauna. Further on at the slope of the bank
the benthos biomass is again reduced and then, on the mud beds surrounding the bank,
it rises again. The main factor conditioning this biomass range is the distribution of
foodstuffs, mainly organic detritus. The high biomass at the edges of the shallow, con-
sisting mostly of epifauna, is conditioned by the presence of rich detritus washed out
from the central parts of the bank and brought by water as a solid suspension. Further
on, the reduction of the biomass is due to conditions unfavourable for the develop-
ment of epifauna and infauna. (This is apparently due, though not stated, to an increase
in suspended matter, causing reduction in epifauna, and a lower amount of organic
matter in the sediment, which is less favourable to infauna.) The last increase in biomass
is due to the infauna, which receives here, in a comparatively calm zone, an abundant
amount of sedimentary detritus.

The richest fauna of the Barents Sea grows on sandy silts and silty sand floors.
Epifaunal species are numerous on hard floors in the regions of strong currents. Areas
rich in infauna are usually poor in epifauna and vice versa. On the one hand, this is
explained by the properties of the sea floor since infauna cannot develop in rocky or
cliff floors. On the other hand, in some areas the floor could have given refuge to infauna,
but the abundant epifauna had taken all the food supplies; the bottom may contain
large amounts of sponge spicules and owing to mechanical factors may become unfit
for benthos habitation. This occurs on the Kildin bank, where finely cartilaginous
and sufficiently silted floors give support to a rich epifauna and are almost devoid of
infauna.

Savilov (1961) reported that in the northern Okhotsk Sea there is a replacement of one
ecological zone by another in proportion to distance from the coast and increase in
depth. In shallow coastal waters, rocks are occupied predominantly by a sessile epifauna
consisting largely of suspension feeders. In the next zone on sandy bottoms mobile
suspension feeders, mainly bivalves, predominate. This is followed by a wide zone of
muddy sand with an infauna consisting mainly of detritus feeders, bivalves being re-
placed in the deeper parts by ophiuroids. Bottom feeders (soil swallowers) occupy the
central deep part of the sea where there is a pronounced prevalence of sedimentation over
erosion. Finally, in the depths from 1 ,000 to 3,000 m. the predominant group in the fauna is
composed of immobile suspension feeders on soft bottoms. Carnivores cannot be allo-
cated to any particular zone.

Nature of surface and plasticity of sediment. Firm surfaces, not necessarily rock outcrops,
are suitable for attachment by such means as cementation, byssal or pedicle fixation,
adhesion, etc. They also provide certain organisms, such as gastropods, with a suitable
surface for locomotion. Many species are able to attach themselves to pebbles or shells
exposed on the surface of an unconsolidated sediment.

Soft substrates vary considerably in their fluidity, which is a reflection of grain size
and shape, and porosity (see Weller 1959). Relatively firm sediments may preclude the
development of burrowers, as Chapman and Newell (1947) have shown with Arenicola;
unstable sands do not encourage the development of life; and some muds are too soft
to permit the survival of benthos. Indeed it is the common experience of geologists
to find an absence of endemic life in originally fluid sediments, which have subsequently

B 6612

D

34

PALAEONTOLOGY, VOLUME 9

revealed their instability by flowage (e.g. bottom structures in mudstones and slump
structures in mudstones and sandstones).

Grain size and the distribution of epifauna and infauna. The varying abundance of epi-
fauna and infauna in different sediments of the sea floor of the Irish Sea is shown in
text-fig. 1. The proportion of epifaunal species decreases from gravel grounds to mud,
whereas the proportion of infaunal species increases. A similar trend would be shown if
the number of individual epifaunal and infaunal animals were plotted.

text-fig. 1. Histograms of the frequency distribution of infaunal
and epifaunal species on different sediments in the Irish Sea off
Port Erin, Isle of Man.

The factors which have determined these changing proportions are complex and
interrelated. Three of them have been discussed — feeding methods, organic content, and
physical nature of the sediment; all are related to grain size. A fourth, strength of bottom
currents, is also relevant. Different species are influenced by these factors to differing
degrees and it is not possible at this stage to categorize the species more closely than
epifauna and infauna.

The rock terms used in text-fig. 1 suggest more mature deposits than are actually
the case. The sediments are generally poorly sorted, including the two which yield the
greatest number of species and individuals — muddy sand and shelly gravel.

Physically at least, the terms muddy sand and shelly gravel, as used in marine eco-
logy, have much more in common than their names might at first imply. An organically
rich muddy sand because of its suitability as a medium for benthonic life will, in the
right conditions of temperature and salinity, support an abundant benthos. Under a
low rate of sedimentation such an environment could develop into a shelly sediment
in which the range of particle size (matrix + shells) could be as great as or even greater
than that of a gravel/shell bank. The major difference between the two environments is
that the shell/gravel sediment is formed in high energy conditions, whereas the shell
bed of muddy sand forms in low energy conditions — an obvious difference, but one
that has far-reaching implications so far as reworking and faunal type are concerned.

Sessile epifauna can live in either environment but shell/gravel grounds tend to support
more epifauna than muddy sand environments where there are generally fewer suitable
surfaces for attachment. Because of the different amounts of mud in the two environ-

CRAIG AND JONES: MARINE BENTHOS

35

ments, deposit feeders are more abundant in muddy sand, and suspension feeders more
abundant in shell/gravel grounds.

We are now in a position to define in general terms the different ecological classes of
animals to be found on, and in, different sea floors. Fine-grained sediments encourage the
development of an infauna which consists mostly of deposit feeders; coarse-grained
sediments support a community of infauna and epifauna which are mainly suspension
feeders; rock surfaces support an epifauna of suspension feeders. Fine-grained sand is a
poor medium for benthonic life and what little there is consists largely of infauna. Carni-
vores may inhabit all four environments.

Transportation. Work on the erosion and transport of sediments by Hjulstrom (1935) and
Inman (1949) has demonstrated the comparative resistance of sediments of different
grain sizes to erosion (see text-fig. 2). Fine-grained sand (of about 0-18 mm. diameter)
marks the transition zone between transport in suspension and transport by surface
creep. It is more easily moved than either the finer or the coarser grades of sediment and
so tends to be better sorted. Although practically nothing is known of the behaviour of
different grades of poorly sorted deposits in currents of different strengths, it is known
that turbulence is greater over the rougher surfaces and that selective scouring must
occur. Both Menard and Boucot (1951) and Johnson (1957) have discussed this problem
in terms of sand and shells, and we assume that muddy sands and other ill-sorted sedi-
ments with shells will show a somewhat greater tendency to erosion (and shell concen-
tration or removal) than the same sediments without shell material.

It is possible to generalize much of the information we have presented in the previous
section. Text-fig. 2 shows the relationship between the abundance of epifaunal and
infaunal species and the different grain sizes of sediment that they may inhabit. Since
epifaunal species are more common in the higher energy environments and more exposed
to currents, they are more likely to be transported after death. This is especially true of
sessile epifauna. Active epifaunal species which inhabit areas of deposition are less likely
to be transported after death. Only on sandy beaches are infauna more abundant than
epifauna in a high energy environment.

One final point is of particular concern to palaeontologists. We have estimated the
numbers of infaunal and epifaunal species which have hard parts and have therefore a
chance of preservation as fossils in the Irish Sea. Out of a total of 79 of the more abun-
dant species of infauna on the sea floor off the Isle of Man about 33 per cent, are pre-
servable: out of a total of 56 of the more abundant species of epifauna 58 per cent, are
preservable.

Influence of benthos on sediment. If a sediment is conducive to the establishment of life,
then the character of the sediment will change. Burrowing organisms will modify existing
sedimentary structures and create new organic structures by their burrowing activities
and through digestion by deposit feeders and deposition of faeces; further, the
increasing abundance of living and dead shelled animals will alter the texture of the
sediment so that in extreme cases the sediment becomes a shell bed. The presence of
organisms on the surface of the sediment affects its rate of accretion or erosion. At low
current velocities, sedimentary particles may be trapped, but at higher velocities erosion
may be prematurely induced or accelerated by the development of scouring around the
shells. Experiments with shells in a flume tank have demonstrated this process (Johnson

36

PALAEONTOLOGY, VOLUME 9

1957). Indeed it is possible that the appearance of shells on a previously stable substratum
could lead to its partial or even complete removal, without the need for any other change
in the environment.

At the same time the change in the nature of the sediment will have an effect on the
composition of the fauna. The larvae of those animals with a pelagic phase do not
simply settle wherever they happen to be when the time for settlement arrives but show,

text-fig. 2. Idealized frequency distribution of infaunal and epifaunal species
according to the nature of the substrate. The upper graph indicates the likeli-
hood of transport or destruction of the benthos after death on the different

substrates.

more or less, some power of selection. They respond to a suitable substrate and can
delay settlement for some time until one is found (Wilson 1952, Smidt 1951, Thorson
1957). One of the stimuli affecting the settlement of many species is the presence of
adults of the same species, and the successful settlement of some individuals may lead
to the establishment of a larger population of that species. The build-up in the sediment
of shells and their possible eventual transformation into a shell bed will have obvious
effects on the fauna and may even lead to the complete change from an infaunal to an
epifaunal community.

Palaeoecological implications. One aspect of the environment, the substrate, has been
taken and its influence and effects on the benthos examined. Geologists understandably

CRAIG AND JONES: MARINE BENTHOS

37

tend to overemphasize the importance of the sedimentary environment since the other
two important factors of salinity and temperature are not immediately obvious. The
preoccupation in this paper with the substrate and its effects on the benthos should not
blind the reader to the importance of these other factors.

We have shown that the chances of transportation of the shells of epifaunal species
must be greater than those of infaunal species. From this it could be presumed that the
chances of the destruction of epifaunal shells must also be greater. But epifaunal skele-
tons are usually stronger than infaunal skeletons and to this extent the balance is
redressed.

It is difficult to arrive at an accurate estimate of the proportion of epifaunal to in-
faunal species. Thorson (1957) has shown that the number of epifaunal and infaunal
species are much the same in the Arctic seas but that epifaunal species may greatly
outnumber infaunal species in tropical waters. He estimates that overall, epifaunal
species may be about four times as abundant as infaunal. The considerable reduction
in the proportion of epifaunal species in polar water is the result of the excessively cold
waters during the polar winters, and is not typical of the geological past. For this reason
it seems that there should have been even more epifaunal species in the geological past
and that they should form a high proportion of the preserved fossil record.

Epifauna may be found in situ in isolated cases in the geological record, principally
at unconformities or disconformities and in bioherms. They will probably be more
abundant than infauna (although reworked) on the rocks flanking and immediately
above such structures, and in the earlier rather than the later stages of infilling of a
basin. They will also occur more commonly in sediments of transgressing rather than
regressing seas.

Another factor which influences the picture of the fossil state is the evidence, derived
from the benthos of the Irish Sea, that even under favourable conditions of preservation
between one-half and two-thirds of the major species of the benthos are not preservable
as body fossils. More analyses of this kind are required from different environments but
the results underline what every palaeontologist knows empirically, and emphasize
all too clearly the difficulties to be faced in the reconstruction of the ecosystem of a
fossil community. It would seem to be a whimsical fact that the fossil communities
most likely to be preserved in situ — the infaunal — are most likely to have fewest
members of the original living community preserved as fossils.

Acknowledgements. We are indebted to Dr. A. Hallam and Mr. J. Wilson for reading and commenting
on the typescript.

REFERENCES

buchanan, j. b. 1958. The bottom fauna communities across the continental shelf off Accra, Ghana
(Gold Coast). Proc. zool. Soc. Lond. 130, 1-56.

chapman, g. and Newell, g. f. 1 947. The role of the body fluid in relation to movement in soft-bodied
vertebrates. 1, The burrowing of Arenicola. Proc. R. Soc. 134B, 431-5.
hjulstrom, f. 1935. Studies of the morphological activity of rivers as illustrated by the River Fyris.
Bull. geol. Instn Univ. Upsala, 25, 221-527.

idelson, m. 1930. A preliminary quantitative evaluation of the bottom fauna of the Spitzbergen Bank.
Trudy, morsk. nauch. Inst. 4, 3.

inman, d. l. 1949. Sorting of sediments in the light of fluid mechanics. J. sedim. Petrol. 19, 51-70.
Johnson, r. g. 1957. Experiments on the burial of shells. J. Geol. 65, 527-35.

38 PALAEONTOLOGY, VOLUME 9

jones, n. s. 1940. The distribution of the marine fauna and bottom deposits off Port Erin. Proc. Trans.
Lpool biol. Soc. 53, 1-34.

1951. The bottom fauna off the south of the Isle of Man. J. Anim. Ecol. 20, 132-44.

1952. The bottom fauna and the food of flatfish off the Cumberland coast. Ibid. 21, 182-205.

— — ■ 1956. The fauna and biomass of a muddy sand deposit off Port Erin, Isle of Man. Ibid. 25,
217-52.

menard, h. w. and boucot, a. j. 1951. Experiments on the movement of shells by water. Am. J.
Sci. 249, 131-51.

petersen, c. g. j. 1913. Valuation of the sea. II, The animal communities of the sea bottom and their
importance for marine zoogeography. Rep. Dan. biol. Stn. 21, 1-42.

purdie, e. g. 1964. Sediments as substrates. In imbrie, j. and Newell, n. d. Approaches to Paleoecology.
Wiley, New York.

sanders, h. l. 1956. Oceanography of Long Island Sound, 1952-1954. X, The biology of marine bottom
communities. Bull. Bingham oceanogr. Coll. 15, 345-414.

savilov, a. 1961. Ecologic characteristics of the bottom communities of invertebrates in the Okhotsk
Sea. Trudy. Inst. Okeanol. 46 (R).

Schafer, w. 1956. Wirkungen der Benthos-Organismen auf den Jungen Schichtverband. Senckenberg.
leth. 37, 183-263.

smidt, e. 1951. Animal production in the Danish Waddensea. Meddr Kommn Danm. Fisk, -og
Havunders., Ser. Fiskeri, 11 (6), 1-151.

thorson, G. 1957. Bottom communities (sublittoral or shallow shelf). In Treatise on Marine Ecology
and Paleoecology (ed. j. e. hedgpeth). Mem. geol. Soc. Am. 67, 461-534.

weller, j. m. 1959. Compaction of sediments. Bull. Am. /4m. Petrol. Geol. 43, 273-310.

wilson, d. p. 1952. The influence of the nature of the substratum on the metamorphosis of the larvae
of marine animals, especially the larvae of Ophelia bicornis Savigny. Annls Inst, oceanogr., Monaco,
27 (2), 49-156.

zenkevitch, l. 1963. Biology of the seas of the U.S.S.R. London.

G. Y. CRAIG

Grant Institute of Geology,
King’s Buildings,

West Mains Road,
Edinburgh 9

N. S. JONES

Marine Biological Station,
Port Erin,

Isle of Man

Manuscript received 30 October 1964

UPPER ORDOVICIAN TRILOBITES FROM
NORTHERN YUKON

by a. c. lenz and michael churkin, jr.

Abstract. The trilobites Robergia yukonensis sp. nov., Cryptolithoides, and Ampyxina salmoni Churkin are
described from the Road River Formation. They occur above the Ashgillian (Late Upper Ordovician) grapto-
lite zone of Dicellograptus complanatus var. ornatus and below beds containing the graptolite Climacograptus
scalaris var. normalis of possible earliest Silurian age. Robergia and Cryptolithoides, although widely distri-
buted geographically, have hitherto been reported from only Middle or presumed Middle Ordovician beds, and
as a result have been considered reliable indices of that series.

The trilobite genera Robergia , Cryptolithoides, and Ampyxina were collected in 1963
from a 10-foot interval within the Road River Formation of northern Yukon Territory.
Robergia and Cryptolithoides, although widely distributed geographically, have hitherto
been reported from only Middle or presumed Middle Ordovician beds, and as a result
have been considered reliable indices of that series. The Ampyxina, in association
with Toernquistia ? idahoensis and Primaspis sp., was described previously by Churkin
(1963a) from a late Middle Ordovician graptolite shale in Idaho. In the same paper he
noted that the presumably Middle Ordovician Caesar Canyon Limestone of Kay (1960)
in Nevada has the same trilobites (A. salmoni, T. ? idahoensis, and Primaspis) in associa-
tion with Robergia and Cryptolithoides.

In the southern Northwest Territories, closer to the present study area, Lenz and
Jackson (1964) listed Robergia, Ampyxina, Cryptolithoides, and other trilobites from
possible late Middle or early Upper Ordovician beds containing lOrthograptus quadri-
mucronatus (Hall). The trilobites were identified by R. J. Ross, Jr., who considered the
presence of Robergia and Cryptolithoides as good evidence for a Middle Ordovician
age assignment.

This paper first presents conclusive evidence, based on associated graptolites, of an
extension in range of Robergia, Cryptolithoides, and Ampyxina salmoni into the Ash-
gillian (late Upper Ordovician), and then gives a formal description of the trilobites.
The introductory section is by A. C. Lenz and the trilobite descriptions are by Michael
Churkin, Jr.

LOCATION, STRATIGRAPHY, AND FAUNAS

The trilobites were collected just west of the Snake River, Yukon Territory (65° 21 ' N.,
133° 30' W. ; fig. 1) from a 10-foot shale and calcareous shale band approximately
30 feet below the top of the Road River Formation, which in this area ranges in age
from Late Cambrian to Early Silurian. Graptolites were collected 10-20 ft. below, and
20-30 ft. above, the trilobites (fig. 2). The graptolites collected beneath the trilobites
consist of Dicellograptus sp. (rare), Orthograptus truncatus var. abbreviatus Elies and
Wood (fairly common), Climacograptus supernus Elies and Wood (fairly common),
Cl. hvalross Ross and Berry (common), Cl. hastatus var. martini Ross and Berry (rare),
Cl. cf. raricaudatus Ross and Berry (very rare), and Glyptograptus sp. In beds 20-30 ft.

(Palaeontology, Vol. 9, Part 1, 1966, pp. 39-47, pis. 4-5.]

40

PALAEONTOLOGY, VOLUME 9

LENZ AND CHURKIN: UPPER ORDOVICIAN TRILOBITES 41

above the trilobites, the graptolite Climacograptus scalaris var. normalis Lapworth
occurs in considerable numbers.

AGE AND CORRELATION

The graptolite fauna which underlies the trilobites correlates with the Dicellograptus
complanatus var. ornatus Zone of northern Yukon Territory (Jackson and Lenz 1962);
this is indicated by the presence of Orthograptus truncatus var. abbreviatus and Climaco-
graptus supernus, both typical representatives of the Ashgillian ornatus Zone of Yukon.
The fauna may also be correlated, at least in part, with the Ashgillian of Great Britain,
the Dicellograptus complanatus Zone of the Basin and Range area of western United
States (Ross and Berry 1963), and with the Dicellograptus anceps Zone of Idaho and
Utah (Churkin 19636). The overlying graptolite Climacograptus scalaris var. normalis
ranges from uppermost Ordovician to lower Lower Silurian in Great Britain and else-
where in northern Yukon, but because of its stratigraphic position in this section is
possibly lowest Silurian.

It is therefore evident that the Ordovician trilobite genera Ampyxina, Robergia, and
Cryptolithoides, while in this instance of Late Ordovician (Ashgillian) age, range through
Middle and Upper Ordovician beds of the North American Cordillera.

SYSTEMATIC PALAEONTOLOGY

Family remopleurididae Hawle and Corda
Genus robergia Wiman

Robergia yukonensis sp. nov.

Plate 4; Plate 5, figs. 1-5.

Material. Collection consists of twenty-two specimens, mostly isolated cranidia. Four specimens are
nearly complete, having thoracic segments and pygidia attached to cranidia. Holotype, G.S.C. no.
19864; paratypes, G.S.C. nos. 19865-71. Geological Survey of Canada, Ottawa.

Description. Opisthoparian trilobite characterized by glabella narrowing posteriorly,
expanding between eye lobes and with long (sag.) tongue, three pairs of lateral glabellar
furrows successively deeper posteriorly. Anterior glabellar furrows nearly straight.
Middle furrows slightly convex forward. Posterior furrows nearly straight but posterior
tips faintly deflected forward in holotype. Faint axial furrow in the frontal part of the
glabella, accentuated in some crushed specimens. Prominent palpebral lobes start
opposite the middle of the first glabellar lobes and end opposite the third glabellar
lobes. Free cheeks widest (trans.) opposite eyes, with raised convex borders and bearing
long (6-8 mm.+) tapering genal spines. Eye composed of many quadrate lenses. Occi-
pital ring bears mesial tubercule on several paratypes (PI. 4, specimen c; PI. 5, figs. 1, 2,
5); apparently not preserved on other specimens. Thorax narrow and gently tapering
posteriorly. Thorax of eleven segments. Pleurae transversely directed and terminating in
small posteriorly directed spines. Deep pleural furrows start near anterior portions of
pleurae and cross each pleuron diagonally. The second pleural furrow begins at the
inner posterior corner of each pleuron and intersects the longer diagonal furrow at about

42

PALAEONTOLOGY, VOLUME 9

two-thirds of the distance from the axial furrow to the outer tip of the pleuron. Articu-
lating half-ring, large, nearly equal to axial ring in size. Each axial ring, and to a lesser
extent each corresponding axial half-ring, separated by two furrows into three trans-
versely convex portions.

Pygidium length about equal to width. Axis extends close to posterior margin and
connected to margin by narrow post-axial ridge. Five axial rings seen on holotype;
pleural regions flat and with two pairs of broad ridges merging into the outer and middle
pair of pygidial spines. Inner pair of pygidial spines much shorter and apparently with-
out corresponding inter-pleural ridges.

Details of the external surface where preserved show very fine (about 11 ridges/mm.)
parallel ridges. The ridges parallel the general shape of the cephalon and pygidium
but run across the pleurae parallel to the exsagittal direction and form a transversally
parallel pattern across the axial region of the thorax.

Measurements of type specimens are given in Table 1.

table 1 — Robergia yukonensis sp. nov.

Holo-

type

Paratypes

G.S.C. no.

19864

19865

19866

19867

19868

19870

19871

Measurements in mm.
Total length (sag.)

33-4

33-7

Length of cephalon (sag.)

12-7

12-3

—

—

—

116

51

Length of thorax (sag.)

14-7

15-4

—

—

—

—

—

Length of pygidium (sag.)

60

60

—

—

—

—

—

Maximum width of cephalon at
base of genal spines (transv.)

17-6

_

16-5

_

Maximum width of cranidium
across palpebral lobes (transv.)

11-2

10-5

10-7

13-4

11-7

5 0

Length of palpebral lobes (exsag.)

3-7

3-6

3-4

3-7

40

3-4

1-8

Width of first axial ring (transv.)

—

5-6

5-4

—

—

6-2

—

Width of last (11th) axial ring
(transv.)

3-4

3-6

Length of first pleuron (transv.)

3-7

—

3-7

—

—

3-9

—

Length of last pleuron (transv.)

2-8

3-5

—

—

—

—

—

Discussion. The Yukon specimens broadly resemble Robergia major Raymond and
Robergia deckeri Cooper from the southern Appalachians (Cooper, 1953; Whittington,
1959) but differ in the following ways: the glabellar furrows of R. yukonensis are less
curved and are more posteriorly directed than in R. major, the middle glabellar furrow
of R. deckeri is also more curved than in R. yukonensis, and the anterior end of the
palpebral lobe lies behind the anterior glabellar furrow instead of directly opposite
the glabellar furrow as in the Appalachian species. In R. micropthalmus (Linnarsson), the
type species, the glabellar furrows are like those in the Appalachian species and differ

EXPLANATION OF PLATE 4

Robergia yukonensis sp. nov. x4. a, holotype, G.S.C. no. 19864; b, paratype, G.S.C. no. 19865; c,
paratype, G.S.C. no. 19866.

Palaeontology, Vol. 9

PLATE 4

LENZ and CHURKIN, Ordovician trilobites

LENZ AND CHURKIN: UPPER ORDOVICIAN TRILOBITES

43

2

<

cr

ID

_l

CO

cr

LU

$

o

C/imacograptus scalar is var. norma Us

<

o

>

o

Q

cr

o

Ampyxina, Robergia, CryptoUthoides

cr

UJ

CL

CL

z>

Orthograptus truncatus var abbreviates,
Climacograptus supernus, Dicellograptus,
etc.

Shafe

Calcareous Limestone
shale

Shaley

limestone

text-fig. 2. Section of part of Road River Formation, showing the relative occurrence of the

graptolites and trilobites.

44

PALAEONTOLOGY, VOLUME 9

from R. yukonensis in the same way. In addition R. yukonensis has a faint axial farrow
on its glabella that seems to be absent in other species of Robergia. Because of these
differences the Yukon species is perhaps a derivative of Robergia as suggested by H. B.
Whittington (personal communication) and could be considered a new genus.

In the thorax of R. yukonensis the pleural furrows, beginning at the inner posterior
corners of each pleuron, intersect the longer diagonal furrows about two-thirds of the
distance from the axial furrow, more than twice as far out on the pleuron from the
axial furrow as in R. decked. Thus in R. decked the two sets of pleural furrows form
small triangular nodes next to the axis, whereas in R. yukonensis most of the area of each
pleuron is bounded by the same furrows.

In rare specimens possessing free cheeks the angle between the genal spine and the
posterior border in R. yukonensis is more acute than that in the Appalachian species.

Occurrence. Robergia is a widespread genus that has been reported from North America, Europe, and
Asia (see fig. 3). Fortunately, graptolites found in most of the Robergia-beaxing sequences can accurately
date each occurrence.

In the United States Robergia has only been recorded from Middle Ordovician rocks, and the
National Research Council Ordovician Correlation Chart for North America (Twenhofel 1954)
shows it as an important index fossil for the Black River and lower Trenton Stages (approximate
equivalents to the lower half of the Caradoc Series in Great Britain).

The oldest reported occurrence of Robergia is from the upper part of the middle Table Head Forma-
tion of Newfoundland that is considered by Whittington and Kindle (1963) as lower Llanvirnian.

In Yukon Territory the genus occurs directly above Ashgillian graptolites possibly equivalent to the
zone of Dicellograptus anceps and thus extends the range of Robergia to include highest Ordovician.

Family trinucleidae Hawle and Corda
Genus cryptolithoides Whittington

Cryptolithoides sp. indet.

Plate 5, figs. 6-8

Material. Three fragmentary cephala.

Description. Cephalon subrectangular in outline and characterized by antero-lateral
angulation of cephalic margin. Glabella strongly convex, clavate, and smooth. Genae
(cheek lobes) without ornamentation. Fringe narrows anteriorly, increasing in width

EXPLANATION OF PLATE 5

Figs. 1-5. Robergia yukonensis sp.x\ov.,paxatypes. l,cranidium x4,G.S.C. no. 19867; 2,cranidium x4,
G.S.C. no. 19868; 3, a nearly complete cephalon lying on a thorax with the pygidium attached
X2, G.S.C. no. 19869; 4, cranidium with part of the thorax attached x2, G.S.C. no. 19870; 5, small
cranidium x4, G.S.C. no. 19871.

Figs. 6-8, Cryptolithoides sp. indet. 6, half of a cranidium showing the pattern of pits X 5, G.S.C. no.
19872; 7, incomplete cranidium x4, G.S.C. no. 19873; 8, external mould of cranidium showing
strongly convex and clavate glabella x4, G.S.C. no. 19874.

Figs. 9-12, Ampyxina salmoni Churkin. 9, nearly complete specimen with long genal spines and small
glabellar spine x4, G.S.C. no. 19875; 10, incomplete specimen with thorax and pygidium attached
to cephalon x4, G.S.C. no. 19876; 11, pygidium x2, G.S.C. no. 19877; 12, cranidium x4,
G.S.C. no. 19878.

Palaeontology, Vol. 9

PLATE 5

LENZ and CHURKIN, Ordovician trilobites

LENZ AND CHURKIN: UPPER ORDOVICIAN TRILOBITES

45

antero-laterally and postero-laterally. Pits in the fringe numerous. One row of pits
external to girder. The outermost row of pits (Ex) in concentric and radial arrangement
with the second row (Ij) except at the posterior lateral margin where the pits of the
second row (lx) decrease in size and are not paired with the outer pits. Anteriorly,

SYSTEM

SERIES

STAGES

NORTHEASTERN
U.S.S.R.-Kharkin-
dzhin and Darpir
Fms . (Chugaeva et
al. 1964, p. 18, 29,
tables 3 and 8)

EUROPE-Ogygio-
caris shales
(Whittington,
1950, p. 543)

NEWFOUNDLAND-
Table Head Fm.
(Whittington
and Kindle ,
1963)

SOUTHEASTERN
U.S. A. -Edinburg
Fm. and equiv-
alents

(Cooper, 1953)

SOUTH-CENTRAL
U.S. A. -Viola Ls.
(Decker, 1933)

WESTERN
U.S.A.-
Caesar
Canyon Ls.
(Churkin ,
1963a)

YUKON

TERRITORY,
CANADA-
Road River

Fm.

(this paper)

ORDOVICIAN

a

CO

Cm

Cm

G

O

O'

CO

3

s

ASHGILL

CARA DOC

1

LLANDEILO

LLANVIRN

1

ARENIG

text-fig. 3. Range of Robergia based on the ages of associated graptolites.

well-defined radial ridges separate the radially paired outer two rows of pits. Anterior
of the glabella two more rows of pits (I2 and I3) in radial arrangement with the outer two
pits but postero-laterally the pits inside the outer two rows irregularly arranged.

Discussion. Only fragmentary specimens were collected, but no significant differences
could be found between the Yukon material and C. ulrichi (Whittington, 1941). The
size of the Yukon specimens, length (sag.) 3-4 mm., width (trans.) 7 mm., falls within the
range of the smaller cephala described by Whittington.

Occurrence. C. ulrichi occurs in the Viola Limestone of Trenton age (equivalent to part of the upper
half of the Caradocian) in Oklahoma, and the genus is reported from Ordovician rocks in Texas
(Whittington 1941). More recently, Cryptolithoicies has been reported from the western U.S. where
it occurs in the Caesar Canyon Limestone of the Toquima Range, Nevada, (Kay 1960) and in the
Middle (?) and Upper Ordovician Hanson Creek Formation of the Seetoya Mountains, Nevada (Kerr
1962). In the Caesar Canyon Limestone, Crypt oiithoides is associated with Ampyxina salmoni, Robergia
cf. major, Toernquistia ? idahoensis, Primaspis sp. and Flexicalymene sp., a trilobite fauna much like that
from graptolitic shale of Trenton age in central Idaho that is correlated (Churkin 1963a) with the
Caradocian graptolite Zones (11) of Climacograptus wilsoni and (12) of Dicranograptus clingani.
Crypto lit hoides thus ranges from Trentonian (Caradocian) in the southern United States, its type area,
into Ashgillian in Yukon and may have a range of Caradocian through Ashgillian in the western
United States.

46

PALAEONTOLOGY, VOLUME 9

Family raphiophoridea Angelin
Genus ampyxina Ulrich

Ampyxina salmoni Churkin
Plate 5, figs. 9-12

1963a Ampyxina salmoni Churkin, pp. 424-5, pi. 51, figs. 1-10.

Material. Four complete specimens, six cranidia, and seven pygidia.

Description. The specimens under study agree in all essential respects with A. salmoni
originally described from Ordovician graptolitic shale in central Idaho (Churkin
1963u). The Idaho collection consists of several hundred specimens, none of them as
complete as those from Yukon, and the new material makes possible the following
additions to the original descriptions.

Free cheeks apparently very narrow, lying along steeply flexed lateral margins of
fixigenae. The long genal spines extend smoothly backward as continuations of curved
cephalic margins. The spines gently curve inward, taper slightly, and extend well past
the pygidium.

Thorax with five segments. Axis about one-fifth total width. Pleurae extend out
horizontally but flexed downward at distal ends. Interpleural furrows extend from
anterior edges of pleurae next to axial furrows and cross pleurae obliquely toward
lateral margins. Furrows become deeper and broader distally. Pleural furrow of first
segment curved convexly forward, on succeeding segments less curved and nearly straight
across fifth segment. Distal ends of pleurae separated from each other by small notches
that produce a serrated margin.

Dimensions. G.S.C. no. 19875: total length (sag.) 7-7 mm., length (sag.) of cephalon 2-9 mm., length of
thorax (sag.) 2-4 mm., length of pygidium (sag.) 2-4 mm., width of cranidium at posterior margin
6-9 mm., maximum width of glabella 2-3 mm., width (trans.) of pygidium at anterior end 6-6 mm.,
length of genal spine 5-3 mm. + .

Discussion. The Yukon specimens possess the straight pygidial ribs characteristic of
Ampyxina salmoni instead of backward flexed ribs as in the very similar A. bellatula
(Savage) illustrated by Whittington (1950, 1959) and A. powe/li (Raymond) described
by Cooper (1953) and Whittington (1959). In addition, the first interpleural furrow in
the Yukon specimens appears less curved forward than in A. bellatula or A. powelli.

Occurrence. Type Ampyxina salmoni is from the type locality of the Saturday Mountain Formation of
central Idaho where it is associated with graptolites equivalent in age to the Caradocian Zones 1 1 and
12 of Elies and Wood (Churkin 1963a). The writer has identified A. salmoni from the Caesar Canyon
Limestone of the Toquima Range, central Nevada (Kay 1960; Kay and Crawford 1964) where it is
found with Robergia cf. R. major and Cryptolithoides sp., an association similar to that in Yukon.
A. salmoni occurs in Middle Ordovician (Caradocian) rocks in both Idaho and Nevada and is found
in considerably younger Upper Ordovician (Ashgillian) rocks in Yukon.

Ampyxina is a widely ranging genus as it has been reported from the early Middle Ordovician
(Porterfield-Wilderness Stages) of Virginia, Alabama, Tennessee (Cooper 1953; Whittington 1959),
and the Upper Ordovician of Missouri and Illinois (Rowley 1909; Savage 1917).

LENZ AND CHURKIN: UPPER ORDOVICIAN TRILOBITES

47

REFERENCES

chugaeva, m. N., rozman, k. s., and ivanova, v. a. 1964. Comparative biostratigraphy of Ordovician
deposits in the North-East of the USSR (in Russian). Trav. Inst. geol. U.R.S.S. 106, 236 pp.
churkin, michael, jr. 1963o. Ordovician trilobites from graptolitic shale in central Idaho. J. Paleont.
37, 421-8, pi. 51.

19636. Graptolite beds in thrust plates of central Idaho and their correlation with sequences in

Nevada. Bull. Amer. Ass. Petrol. Geol. 47, 1611-23.
cooper, b. n. 1953. Trilobites from the lower Champlainian formations of the Appalachian Valley.
Mem. geol. Soc. Amer. 55, 69 pp., 19 pi.

decker, c. e. 1933. Viola limestone, primarily of Arbuckle and Wichita Mountain regions, Oklahoma.
Bull. Amer. Ass. Petrol. Geol. 17, 1405-35.

jackson, d. E. and lenz, a. c. 1962. Zonation of Ordovician and Silurian Graptolites of northern
Yukon, Canada. Bull. Amer. Ass. Petrol. Geol. 46, 30-45.
kay, marshall. 1960. Paleozoic continental margin in central Nevada, western United States.
Int. Geol. Cong., 21st sess., pt. xii, 94-103.

and crawford, j. p. 1964. Lower Paleozoic facies from the miogeosynclinal belt to the eugeo-

synclinal in thrust slices in central Nevada. Bull. geol. Soc. Amer. 75, 425-54.
kerr, j. w. 1962. Paleozoic sequences in thrust slices of the Seetoya Mountains, Independence Range,
Elko County, Nevada. Bull. geol. Soc. Amer. 13, 439-60.
lenz, a. c. and jackson, d. e. 1964. New occurrences of graptolites from the South Nahanni Region,
Northwest Territories and Yukon. Bull. Canad. Petrol. Geol. 12, 892-900.
ross, R. j. and berry, w. b. n. 1963. Ordovician Graptolites of the Basin Ranges in California, Nevada,
Utah, and Idaho. Bull. U.S. Geol. Surv. 1134, 177 pp., 13 pi.
rowley, r. R. 1909. Geology of Pike County. Missouri Bur. Geol. Mines, 2nd ser., 8, 122 pp., 20 pi.
savage, t. e. 1917. The Thebes Sandstone and Orchard Creek Shale and their faunas in Illinois.
Trans. Illinois Acad. Sci. 10, 261-75, 2 pi.

twenhofel, w. h. 1954. Correlation of the Ordovician formations of North America. Bull. Geol. Soc.
Amer. 65, 247-98, chart 2.

Whittington, h. b. 1941. The Trinucleidae — with special reference to North American genera and
species. J. Paleont. 15, 21-41, pi. 5, 6.

1950. Sixteen Ordovician genotype trilobites. J. Paleont. 24, 531-65, pi. 68-75.

1959. Silicified Middle Ordovician trilobites, Remopleurididae, Trinucleidae, Raphiophoridae,

Endymioniidae. Bull. Mus. Comp. Zool. Harv. 121, 371^196, 36 pi.

and kindle, c. h. 1963. Middle Ordovician Table Head Formation, western Newfoundland.

Bull. Geol. Soc. Amer. 74, 745-58.

A. c. LENZ

Department of Geology,
University of Western Ontario,
London, Ontario, Canada

MICHAEL CHURKIN, JR.

U.S. Geological Survey,
Menlo Park, California, U.S. A.

Manuscript received 9 November 1964

SILURIAN GIRVANELLA FROM THE WELSH

BORDERLAND

by HAZEL M. JOHNSON

Abstract. A general survey of the calcareous algae from Silurian limestones of the Welsh Border has yielded
many algal remains particularly of the genus Girvanella. Nine species of this genus are here described, eight of
which are new; and they are grouped according to their size and mode of growth. The Sarmenta group includes
three new species, G. sarmenta, G. fragila, and G. prolixa ; the Problematica group two new species, G. pusilla,
G. incompta as well as G. problematica and its var. lumbricalis; the Media group with one new species G. media,
and the Ramosa group two new species G. ramosa and G. effitsa. The basis for subdividing the genus is given and
the distribution of the specific forms in the varied lithology of the limestones are recorded.

A general survey of the calcareous algae present in the Silurian limestones of the
Welsh Border has shown that algae are more abundant and widespread than has hither-
to been recorded. Specimens were collected from numerous localities within the area
demarked approximately by Wenlock Edge to May Hill and Dudley to Old Radnor
(text-fig. 1).

Previous records of algae in this area are few and the belief that algae contributed to
the formation of reef limestones was largely unproved. From the Woolhope Limestone,
Garwood and Goodyear (1918) described algal limestones at Old Radnor and Nash
Scar; also Garwood (1931) mentioned an undescribed Solenopora at Woolhope.
Girvanella is not recorded by Squirrel and Tucker (1960) in an account of the geology
of the Woolhope Inlier.

For the Wenlock Limestone, Wethered (1890, 1893) described algae at May Hill,
Ledbury, and Purley (West Malverns). Chapman (1907) and Crosfield and Johnston
(1914) recorded algae or algal fragments at Dudley. Several authors have indicated the
presence of algal limestones as opposed to brachiopod, crinoid, or other facies at Wen-
lock Edge but no details of localities were given. Furthermore they supposed that
masses of algae together with corals and stromatoporoids were engaged in the building
of the Ballstones (Hill 1936, Butler and Oakley 1936, Whittard 1952). Likewise the Crog
Balls (found at Dudley) were believed by these authors to be the result of algal decompo-
sition. As far as is known there have been no published records of calcareous algae
obtained from the Bringewood Beds (Aymestry Limestone) of this area.

Only a very few of the published accounts mention the genus Girvanella. G. proble-
matica recorded by Wethered (1890, 1893) was asserted by Garwood (1931) to be
‘ Sphaerococlium ’ munthei which is now recognized as an association of Rothpletzella
and Wetheredella (Wood 1948). G. conferta (Chapman) appears to belong to the genus
Rothpletzella as it possesses dichotomous branching. The only other record was by
Crosfield and Johnston (1941) who noted the presence of algal fragments in the Crog
Balls of the Wren’s Nest, Dudley, but gave no evidence to show whether they belong
to the genus Girvanella.

When the present research began, there was little known about Silurian calcareous
algae of the Welsh Border and no Girvanella records were identified to the species. The

[Palaeontology, Vol. 9, Part 1, 1966, pp. 48-63, pis. 6-12.]

H. M. JOHNSON: SILURIAN GIRVANELLA

49

commonest algal genus, Girvanella, was found to vary so much that it did not appear
valid to retain all variants at generic level, or to regard them as related to the Ordo-
vician type, G. problematica. It seems possible that the species of Silurian filamentous
algae were numerous and also similar forms could associate together as do living

text-fig. 1. Reference map of the Welsh Border showing the
important areas for Girvanella. ca, Craven Arms; du, Dudley; h,
Hereford ; gm, Great Malvern ; lb, Ledbury ; lh, Longhope ; lu,
Ludlow; mf, Mordiford; mw, Much Wenlock; ns, Nash Scar;
o, Onibury; or, Old Radnor; p, Presteign; wh, Woolhope.

forms today. Slight dissimilarities in filament character would not necessarily be re-
flected in the fossils although several different types of filamentous algae have been
proved.

The genus Girvanella is subdivided here on filament size and mode of growth which
includes branching habit, encrusting and non-encrusting, attached and free forms, the
constancy of tube diameter, curvature of tubes, and the amount of adherence between
tubes. Several species have similar modes of growth and are grouped together. They can
be distinguished by their filament size. A full account of these and other algae can be
found in Green (1955, 1959).

B 6612

E

50

PALAEONTOLOGY, VOLUME 9

METHODS OF SAMPLING AND EXAMINATION

The presence of Girvanella was rarely found to be indicated by macroscopic structures
in the limestones investigated so a suite of specimens from each exposure was examined.
Specimens were also taken at intervals along bedding planes. Particular attention was
directed to beds in the neighbourhood of reef or reef-like structures where changes in
the character of the limestone were likely to be rapid. Thin sections were found to be the
most satisfactory for the study of the algae. Measurements on algal material were made
by using camera lucida drawings mostly at magnifications of x 350 and x 600. The
accuracy of this method was checked by measurements from photographs and also by
direct measurements using an eyepiece micrometer scale.

The method used to determine the diameter of algal filaments was to measure the
internal diameters of five tubes per specimen to give the range of variation and probable
mean. With indistinct groups, the variation is greater, due to increased error in measure-
ment. Longitudinal sections of the tubes were found to be more satisfactory in giving
the true diameter. These were compared with tubes seen in cross-section as in the
method used by Wood (1957, 1963), but Wood’s method was found to be less accurate
when measuring finer filaments of 5 ^ or less in diameter. Furthermore it was difficult
to distinguish transverse and oblique sections especially in fine tubes with thick walls
or curving filaments when applying Wood’s technique. In many sections the position
of the original wall is marked by the greatest concentration of calcite dust (Wood
1941) against the clear calcite of the lumen, and, except in indistinct groups, this darker
concentration of dust can be used to determine the original diameter of a filament. This
criterion has to be used with caution however, for the calcite dust of the wall may partly
obscure the clear lumen in some longitudinal sections.

Crushing of an algal filament is unlikely to occur before death owing to its internal
pressure. Flattening only appears to occur when tubes are adherent to each other and
nearly always in encrusting forms where the filaments are in a felted mass. Where crush-
ing has occurred, the clarity of the wall disappears and this gives a good indication of
the doubtful value of any measurements taken.

The number of each species of GirvaneUa examined by the above method is as
follows : 3 specimens of G. problematica var. lumbricalis, 40 G. effusa, 60 G. ramosa,
70 G.fragila, 85 G. media, 95 each G. sarmenta and G. prolixa, 115 G. pusil/a, 125 G.
incomp ta, and 160 G. problematica.

MORPHOLOGY AND SYSTEMATIC POSITION OF THE GENUS

Class porostromata Pia 1927 ex Harlan Johnson 1959
Genus girvanella Nicholson and Etheridge 1878

Type species. Girvanella problematica Nicholson and Etheridge.

EXPLANATION OF PLATE 6

Figs. 1, 2. Girvanella cf. prolixa sp. nov., showing septation and possible heterocyst. MF 1, Woolhope
Limestone, GSM. PF 2838/1. 1, X220. 2, enlarged view X 1000.

Fig. 3. G. prolixa sp. nov. with septate filaments. MF 1, Woolhope Limestone, GSM. PF 2837/5; x220.
Fig. 4. G. sarmenta sp. nov. Tubes seen in cross-section. WE 1 1, Wenlock Limestone, GSM. PF 2845/2;
X 180.

Palaeontology , Vol. 9

PLATE 6

JOHNSON, Silurian Girvanella

H. M. JOHNSON: SILURIAN GIRVANELLA

51

Original Diagnosis. ‘Nodules composed of microscopic tubuli, with arenaceous or
calcareous (?) walls, flexuous or contorted, circular in section, forming loosely com-
pacted masses. The tubes apparently simple cylinders, without perforations in their
sides, and destitute of internal partitions of a similar kind.’ Wood (1957) in his re-
description of the type species of Girvanella does not give a formal emended diagnosis
of the genus.

Description. The original diagnosis gave the tube diameters as 10 p and 18 p but later
descriptions have considerably extended this range of diameter (Wethered 1889-93,
Hoeg 1932, Wood 1957, Johnson et al. 1959). During this investigation Silurian
members of the genus Girvanella were observed to have a range in diameter from 2 to
31 p. All the filaments examined had the fine granular calcite ‘dust’ wall (Wood 1941),
which was occasionally iron stained, and enclosed a lumen of clear calcite crystals.
Aggregations of opaque iron ores including pyrite were found within the lumen in some
specimens. The mode of growth was found to be variable. The specimens investigated
ranged from parallel sets of filaments having each tube of even diameter throughout to
highly contorted filaments with variable diameters ; from a loose non-adherent arrange-
ment of tubes to tubes in contact or adherent to each other; from no branching to highly
branched forms; and from free living, possibly floating plants to attached forms, en-
crusting or associating in composite growth with other algal genera. Another feature not
recognized in the original diagnosis is one of possible septation (PI. 6, figs. 1-3).

Definite reproductive organs are unknown within the genus and this may also be the
result of indifferent preservation or that the main method of reproduction was by frag-
mentation of the filaments. A possible ‘ heterocyst ’ has been found in one instance with
a general rounding of a septate filament which did not appear to be a cut branch as no
fracturing of the clear lumen was visible. The ‘ heterocyst ’ was oval in shape (PI. 6,
figs. 1, 2) and appeared to be an integral part of the filament rather than a protuberance
from it.

From the evidence stated in the foregoing paragraphs it would appear that the formal
diagnosis of the genus Girvanella requires some revision. Since, however, it is the purpose
of this contribution to describe the Silurian species of Girvanella, a discussion of the
Ordovician type material is not included. The genus Girvanella is used in this work in
agreement with previous workers though it is suggested that later studies may well lead
to subdivision of the genus.

Lower Palaeozoic species and varieties. Hoeg (1932), in describing the Ordovician algae
from the Trondheim area, subdivided Girvanella problematica into three varieties by
differences in size and habit, viz. vars. typical is, moniliformis, and lumbricalis. In 1942
Lewis reviewed the various Ordovician Girvanella and concluded that they were all
varieties of G. problematica including his new variety G. spiralis from the Levis group,
Quebec, and G. ocellata which had been described by Seely (1885) under the genus
Strephochetus. Silurian Girvanella have not been similarly subdivided and the present
work has shown that they have a different range of form from their Ordovician counter-
parts. Girvanella from the Welsh Borderland was found to be divisible into eight distinct
growth types as follows :

1. Sets of parallel or sub-parallel filaments of apparently unlimited growth, at least

52 PALAEONTOLOGY, VOLUME 9

partly adhering to each other as in the G. problematica var. moniliformis Hoeg
(1932) or as faggot-like bundles.

2. A loose coiling of filaments as in G. problematica (redefined by Wood 1957), or
more specifically in a coil (as in G. problematica var. typicalis Hoeg 1932). Branch-
ing here is infrequent, absent, or in limited portions only.

3. Short curved loose filaments, sparsely branched, sometimes encrusting, correspond-
ing in form to G. problematica var. lumbricalis (Hoeg).

4. A spiral development of a single filament (or a few filaments) as in var. spiralis
(Lewis).

5. Tight coiling of the filaments within a group. This is usually the result of adhering,
much branching, curving filaments.

6. Highly branched, curved filaments with uneven diameter, not strictly in a coil but
matted together.

7. A ‘raft ’-like development, probably a variety of the parallel tubes growth form (1)
but where the filaments are of approximately equal length and appear to be of
limited growth. Tubes are adherent.

8. Encrusting forms which may be one or more of the above growth forms.

The tube diameter of these growth forms were also compared and from this data
subdivision into district species was found to be possible. Four species groups have been
erected for the Silurian Girvanella.

Group A sarmenta (growth form 1 and rarely 7)

Description. Groups of parallel filaments mostly adherent to each other. In longitudinal
section appearing as long tubes (tubes of indefinite length), mainly in contact, un-
branched or with scarcely any branching. In transverse view the circular cross-sections
are arranged in clusters.

Remarks. Normally, no indications of any attachment to substratum are visible.
Only in a few specimens do the tubes appear to arise from one end. Raft-like forms are
at present included as smaller specimens of this group. Division into species within the
group is on the internal diameter of the tube, which varies from 2 to 31 fx.

Species. Girvanella sarmenta sp. nov., G. prolixa sp. nov., G. fragila sp. nov.

Group B problematica (growth forms 2, 3, 4, sometimes 8)

Description. Loosely coiling or tangled groups. Tubes mostly not in contact with each
other. Branching infrequent or none. Tubes do not lie parallel or if they attain a sub-
parallel direction they are mostly not adherent to each other.

EXPLANATION OF PLATE 7

Figs. 1, 2. Girvanella sarmenta sp. nov. 1, Holotype; only part of specimen figured. Length of specimen
2-6 mm. Tubes elongate and of even diameter. WE 1 5, Wenlock Limestone, GSM. PF2846/4; X 160.
2, Paratype; WE 15, GSM. PF 2846/3; X 140.

Figs. 3-5. G. fragila sp. nov.; MF 1, Woolhope Limestone. 3, 4, Holotype; different sections of the
same specimen (figs, not in alignment); GSM. PF 2839/1. Length of specimen 2-5 mm.; X 140.
5, Paratype; specimen with slightly larger tubes, GSM. PF 2837/1 ; x 180.

Palaeontology, Vol. 9

PLATE 7

JOHNSON, Silurian Girvanella

H. M. JOHNSON: SILURIAN GIRVANELLA

53

Remarks. Internal diameter of tubes ranges between 2 and 28 p . Branching occurs more
frequently in the larger forms. Also in the larger forms, the tubes tend to be more
adherent to each other than in the finer-tubed specimens. A variation of this is a tight
coiling of filaments of even diameter over each other. The extreme variety of this is the
production of a spiral form usually by one filament.

Species. G. problematica, G. pusiUa sp. nov., G. incompta sp. nov. (probably including
G. problematica var. spiralis Lewis). G. problematica var. lumbricalis may represent a
distinct species (growth form 3).

Group C media (growth form 5, sometimes 8)

Description. Groups of filaments tightly coiling and adhering branching clusters.
Branching fairly frequent. Tubes even diameter throughout with apparently no con-
strictions. Internal diameter 8-18 p.

Species. G. media sp. nov.

Group D ramosa (growth form 6, sometimes 8)

Description. Forms characterized by highly branched adherent tubes usually in clusters.
Branching irregular. Tubes uneven in diameter and length between branches short.

Remarks. Constrictions are often visible. Groups are sometimes encrusting. The internal
diameter range is usually between 1 1 and 27 p, rarely more. The group is distinguished
from G. media by its irregular diameter and more frequent branching and by its irregular
type of branching from the dichotomous branching of a Rothpletzella.

Species. G. ramosa sp. nov., G. effusa sp. nov.

SPECIES DESCRIPTIONS
Group SARMENTA
Girvanella sarmenta sp. nov.

Plate 6, fig. 4; Plate 7, figs. 1,2; text-fig. 3 g

Diagnosis. A non-encrusting Girvanella with bundles of parallel, even diameter tubes
with no constrictions. Tubes seldom branch and are mostly adherent to each other.
Filament, or internal tube diameter 18-31 p, the majority having diameter between 19
and 22 p.

Holotype. GSM. PF 2846/4; Plate 7, fig. 1. Wenlock Limestone, Silurian. Wenlock Edge,
WE 15(5), SO 856455.

Paratype. GSM. PF 2846/3 ; Plate 7, fig. 2. Same locality and horizon.

Description. This species is one of the largest Girvanella found in these Silurian lime-
stones. Within the Sarmenta group, the narrowest diameter tubes of this species just
overlap the range for G. prolixa, but the mean value for the filaments is 5 p greater.
It also appears to correspond to the Ordovician G. problematica var. moniliformis
(Hoeg) in size and longitudinal view but differs from it in transverse section in having
tubes arranged in clusters and not in a single row.

54

PALAEONTOLOGY, VOLUME 9

There does not appear to be any marked tapering to one end of a cluster so there is
no evidence of this form being attached to a substrate.

Distribution. Woolhope and Wenlock Limestones of the Woolhope Inlier and Wenlock
Edge. It is infrequent in the Woolhope Limestone. Most specimens have been found at
the south-west end of Wenlock Edge although it has been found at the other localities
in the Wenlock Limestone on the Edge and at Fownhope.

Girvanella prolixa sp. nov.

Plate 6, fig. 3; Plate 8, figs. 2-4; text -fig. 3d

Diagnosis. Girvanella occurring in small groups of parallel tubes, mostly adhering,
sometimes free. Filaments long, with circular cross-section, even diameter and nearly
straight. Range of filament diameter 12-18 /x, occasionally more. Most filaments are
between 14 and 16 p. wide.

Holotype. GSM. PF 2840/2; Plate 8, fig. 2. Grey Measures, Wenlock Limestone. Silurian. Farley
Quarry, north of Much Wenlock, Wenlock Edge, Shropshire WE 3(7), SJO 15629.

Remarks. Specimens examined from the Wenlock Limestone at Fownhope were found to
have tube diameters mostly between 17 and 18 /x wide, which is slightly larger than aver-
age. The Fownhope district appears to have been favourable for calcareous algae so
it is possible that some algal growths may have been better developed than elsewhere.
Some specimens from Fownhope show possible septation in the tubes (PI. 6, figs. 1-3).

In a very few cases, the filaments partly converge at one end, which could indicate
that some of these plants may have been attached during life. Another indication of this
may be the ‘rafts’, shorter length groups which may have formed either part of a larger
group or an early stage. In some, the adherent tubes converge to a narrow region and
end either by the tubes joining or overlapping.

Distribution. The Woolhope and Wenlock Limestones of the Woolhope Inlier and the
Wenlock Limestone of Wenlock Edge. It is occasional in the Woolhope Limestone at
Mordiford, but is more common in the Wenlock Limestone. It is abundant in the Grey
Measures of Wenlock Edge near Much Wenlock and to the north-east although it has
been obtained from most of the exposures examined along the Edge.

Girvanella fragila sp. nov.

Plate 7, figs. 3-5; text-fig. 3 a

Diagnosis. Girvanella in which the tubes are parallel, of even diameter in bundles, mostly
adherent, occasionally partly free. Tubes long, straight or slightly undulating. Branching

EXPLANATION OF PLATE 8

Fig. 1. Girvanella media sp. nov. Small cluster branching tubes mostly adjoining tubes in contact. FH 2,
Wenlock Limestone, GSM. PF 2841/3; X250.

Figs. 2-4. G. prolixa sp. nov. 2, Holotype showing sub-parallel tubes, slight branching; WE 3, Wenlock
Limestone, GSM. PF 2843/1 ; X 200. 3, Specimen with sub-parallel tubes slightly larger than average;
FH 2, Wenlock Limestone, GSM. PF 2840/2; X 250. 4, Transverse section of tubes; MF 1, Woolhope
Limestone, GSM. PF 2837/6; x260.

Palaeontology , Vol. 9

PLATE 8

JOHNSON, Silurian Girvanella

H. M. JOHNSON: SILURIAN GIRVANELLA

55

700-i

text-fig. 2. Range of diameter of Girvanella tubes from Mordiford, Fownhope, and
Wenlock Edge. Five measurements were taken from each Girvanella specimen, thus
giving the variation for each group and also more overlap of sizes than if the mean
only is used, a, Mordiford, 148 specimens measured; b, Fownhope and c, Wenlock
Edge, 1 62 and 1 64 specimens respectively.

rare. Internal diameter of tubes range from 2 to 9 p, mostly 5-7 p, very occasionally to
10 and 11 p.

Holotype. GSM. PF 2839/1; Plate 7, figs. 3, 4. Woolhope Limestone, Silurian small exposure at
junction Littlehope Lane and Mordiford Road, Haugh Wood, Mordiford MF 1(2), SO 373578.

Paratype. GSM. PF 2837/1 ; Plate 2, fig. 5; same locality and horizon.

Description. The mean value of filament diameters (or the internal diameters of tubes)
was found to be 6 p for the Woolhope specimens but the only Wenlock Limestone

text-fig. 3. Comparative drawings of Silurian Girvanella all at magnification X 125, showing the size
and growth forms of the different species. (Opaque inclusions also indicated in g; tube diameter slightly

above average in d and i.)

a, G. fragila sp. nov. b, G. pusilla sp. nov. c, G. media sp. nov. d, G. prolixa sp. nov. from Fownhope.
e, G. problematica Nich. and Eth. f G. problematica var. lumbricalis Hoeg. g, G. sarmenta sp. nov.
h, G. incompta sp. nov. i, G. ramosa sp. nov. j, G. ejfusa sp. nov.

H. M. JOHNSON: SILURIAN GIRVANELLA

57

specimen had a tube range of 2-4 p. The tubes mainly have very dark grey granular
calcite walls, ‘algal dust’ (Wood 1941). Branching may occasionally occur but is not
readily visible.

Remarks. No processes of attachment were seen and as these forms were abundant
in bedded limestones away from reef structures, it is possible that this species may have
been a floating form. G.fragila and G. pusilla constitute the finest tubed Silurian species
found. G.fragila can be distinguished readily from G. pusilla by its different growth habit,
the latter being loosely coiled.

Distribution. This species has been found in the Woolhope Limestone around Mordiford
where it is abundant at various horizons from just above the Petalocrinus Limestone
to nearly the top of the Woolhope Limestone. It is extremely plentiful in the Scutterdine
Quarries (Pocock 1930), sometimes appearing to be the dominant fossil in thin section.
In the Wenlock Limestone the only specimen was found at Fownhope. No specimens
were found in the limestones above this horizon.

Group PROBLEMATICA

Girvanella problematica Nicholson and Etheridge 1878
Plate 9, fig. 1 ; text-fig. he

Neotype. Figured by Wood (1957), Plate 5, fig. 2. Slide BM(NH) V34566. Stinchar Limestone, Ordo-
vician, Tormitchell, Girvan, Ayrshire.

Description. Loosely coiling filament clusters with internal tube diameters 12-18 p,
occasionally more but mostly 15-16 p, from the Silurian have been included in this
species. This corresponds with Wood’s redescription of the type species having average
external tube diameters 21-22 ft ranging from 18 to 25 ft (rarely 30 ft) and internal
diameter average 15-16 ft ranging from 13 to 20 ft (rarely 22 ft). Girvanella having ad-
herent or tight branching or faggot-like groups are not here included. G. problematica
has been found both as loose non-encrusting coils and encrusting brachiopods, crinoid
ossicles, bryozoa, and tabulate corals.

Distribution. Woolhope, Wenlock Limestones, and the Bringewood Beds. This species
is fairly common in the bedded Woolhope Limestone at Mordiford and a few specimens
have been obtained from Old Radnor and Woolhope. It is fairly common to abundant in
the Wenlock Limestone of Dudley, Fownhope, and Wenlock Edge, especially near
Much Wenlock. It has been found also at May Hill, Ledbury, West Malverns, and Sollers’
Hope, and is present in the Bringewood Beds (Aymestry Limestone) of Ledbury, View
Edge (Craven Arms), and the Whitcliffe Beds, Ludlow.

Girvanella problematica var. lumbricalis Hoeg
Plate 9, fig. 2; text-fig. 3/

1932 Girvanella problematica var. lumbricalis Heeg, pi. 1, fig. 6.

Description. Short strongly curving tubes, sometimes encrusting, but if so, only a
small area of tube is in contact with the substrate (PI. 9, fig. 2). Tubes non-adherent,

58

PALAEONTOLOGY, VOLUME 9

loosely interlaced or isolate, rarely branched, and do not taper. Diameter of the fila-
ments varies from 12 to 18 p, occasionally to 20 average diameter 14-17 p.

Distribution. G. problematica var. lumbricalis has been found in the Wenlock Limestone
of Dudley, Fownhope, and Wenlock Edge.

Girvanella pusilla sp. nov.

Plate 10, figs. 1-2; text-fig. 3b

Diagnosis. A loosely coiled Girvanella with fine thread-like tubes. Tubes long, even
diameter, mostly unbranched, not usually in contact with each other except when
encrusting other organisms. Diameter of tubes varies from 2 to 9 p, very rarely to 1 1 p,
average 4-7 p.

Holotype. GSM. PF 2838/2; Plate 10, fig. 1. Woolhope Limestone, Haugh Wood exposure, junction
of Littlehope Lane and Mordiford Road, MF 1(2), SO 373578.

Remarks. Specimens obtained from the Woolhope Limestone were found to have a
slightly larger tube diameter (mostly 5-7 p), than those from the Wenlock Limestone
(2-7 ix, mostly 4-5 p). Specimens from the Wenlock Limestone of Fownhope and Dudley
sometimes show filaments encrusting, sometimes with Wetheredella, different organisms
including Coenites and crinoid ossicles. Filaments have also been seen in small growth
forms in association with RothpletzeJla and Girvanella problematica.

Distribution. G. pusilla occurs in the Woolhope and Wenlock Limestones of the Wool-
hope District; Woolhope, Mordiford, Fownhope, and Soller’s Hope; the Woolhope
Limestone of Old Radnor and the Wenlock Limestone of Wenlock Edge, Dudley,
Ledbury, May Hill, and the west flank of the Malverns. This species was less abundant
in the thin sections examined than G. fragila in the Woolhope Limestone, but more
abundant in the Wenlock Limestone. Owing to the lack of clarity of most specimens for
photographic purposes, the type specimen was selected from the Woolhope Limestone
material.

Girvanella incompta sp. nov.

Plate 9, figs. 3, 4; Plate 10, fig. 1; Plate 11, fig. 1; text-fig. 3/i
Diagnosis. Loosely coiled clusters of algal tubes. Tubes even diameter throughout and

EXPLANATION OF PLATE 9

Figs. 1, 2. DU 6, Wenlock Limestone. 1, Girvanella problematica Nich. and Eth., loosely coiled tubes,
X 160. GSM. PF 2842/1. 2, G. problematica var. lumbricalis FToeg, with short curving tubes encrusting
trilobite pleura, X 160. GSM. PF 2842/2.

Figs. 3, 4. G. incompta sp. nov. WE 11, Wenlock Limestone. 3, Holotype; large loosely coiled tubes,
some branching, X 140. GSM. PF 2844/1. 4, Paratype; branching and curved tubes, x250. GSM.
PF 2844/2.

EXPLANATION OF PLATE 10

Figs. 1, 2. Girvanella pusilla sp. nov. 1, Holotype; short coiling branched filaments; MF 1, Woolhope
Limestone, GSM. PF 2838/2; x250. 2, Paratype; loosely encrusting specimen, x250; Woolhope
Limestone MF 2(7); Author’s coll.

Fig. 3. G. media sp. nov. Holotype; curved coiling, branched filaments; FH 2, Wenlock Limestone,
GSM. PF 2840/1 ; X 145.

Fig. 4. G. incompta sp. nov.; a few coiled tubes; WE 11, Wenlock Limestone, GSM. PF 2845/3; x 250.

Palaeontology, Vol. 9

PLATE 9

JOHNSON, Silurian, Girvanella

Palaeontology , Vol. 9

PLATE 10

JOHNSON, Silurian Girvane/la

H. M. JOHNSON: SILURIAN GIRVANELLA

59

usually branched. Tubes slightly to completely adherent with each other. Internal
diameter of tube 18-28 p, mostly 19-21 p .

Holotype. GSM. PF 2844/1; Plate 9, fig. 3. Wenlock Limestone, Silurian. Wenlock Edge, WE 11(1),
SO 844447.

Paratype. GSM. PF 2844/2; Plate 9, fig. 4. Same locality and horizon.

Description and Remarks. This species is variable in form. The variation extends from the
typical very loose cluster of branching tubes to the occasional tight coiling of tubes or
rarely one or two tubes arranged in a spiral form. The frequent occurrence of branching,
the larger tube diameter, and the tendency of tubes to adhere to each other are the main
features which separate this form from G. problematica. It has been found occasionally
encrusting organisms such as brachiopods and bryozoa.

Distribution. Wenlock and Aymestry Limestones (Bringewood Beds). G. incompta was
found to be abundant in the Wenlock Limestone at the south-west end of Wenlock Edge,
but less common elsewhere. It also was found in thin sections taken from the Wenlock
Limestone at Fownhope and Wren’s Nest, Dudley and the Bringewood Beds of Aymestry
Limestone of View Edge, Craven Arms.

Group MEDIA
Girvanella media sp. nov.

Plate 8, fig. 1; Plate 10, fig. 3; Plate 11, figs. 2-4; text-fig. 3c

Diagnosis. Tubes of even diameter adhering together, branching frequently and some-
times tightly coiled. Diameter range 8-18 p, very rarely more, mostly 13-14 p.

Holotype. GSM. PF 2840/1 ; Plate 10, fig. 3. Wenlock Limestone, Silurian. The Old Quarries, Common
Hill, Fownhope FH 2(3), SO 348586.

Remarks. G. media differs from G. problematica in its adherent nature and frequent
branching and from G. incompta by mode of branching and filament size. Occasionally
it is found to have encrusted other organisms such as brachiopods and an ostracod.

Distribution. Woolhope, Wenlock, and Aymestry Limestones. G. media has been found
at Woolhope and more at Mordiford (Woolhope Limestones). It occurs at Wenlock
Edge, one specimen was found from Dudley but it is most abundant in the Wenlock
Limestone at Fownhope. It has also been obtained from the Bringewood Beds (Aymestry
Limestone) of View Edge, Craven Arms.

Group RAMOSA
Girvanella ramosa sp. nov.

Plate 12, figs. 1, 3, 5; text-fig. 3 i

Diagnosis. Highly branched, uneven diameter tubes grouped together. Tubes are com-
pletely adherent to each other and branching is irregular. Diameters at internodes vary
between 11 and 20 p, average 15-16 p.

60 PALAEONTOLOGY, VOLUME 9

Holotype. GSM. PF 2846/2; Plate 12, fig. 1. Wenlock Limestone, Silurian. Wenlock Edge WE 15(5)
SO 856455.

Pciratype. GSM. PF 2846/1; Plate 12, fig. 5. Same locality and horizon.

Description and Remarks. Most groups examined were small in size. The uneven diameter
of the tubes of this species is due to the mode of branching and constrictions and in this
they can be distinguished from the slightly smaller, branching even diameter tubes of
G. media. In both this and the subsequent species G. ejfusa , the apparent constrictions
might be related to either branching and/or cell length, if the original filaments were
septate. No direct evidence has so far been found to support septation. Both these
species are distinguished from Rotlipletzella spp. by their irregular modes of branching
and lack of the characteristic type of beading of a Rothpletzella (beading here represent-
ing cut branches).

Distribution. Woolhope and Wenlock Limestones. G. ramosa has been found in the
Woolhope Limestone at Mordiford and the Wenlock Limestone of Fownhope and also
Wenlock Edge where it is fairly common.

Girvanella ejfusa sp. nov.

Plate 12, figs. 2, 4; text-fig. 3 j

Diagnosis. Tubes uneven in diameter, with some constrictions and highly branched.
Branching is irregular and tubes are in contact with each other. Diameter of filaments at
internodes 18 to 27 p, mostly 22-23 p wide.

Holotype. GSM. PF 2840/3; Plate 12, fig. 2. Wenlock Limestone, Silurian. The Old Quarries, Common
Hill, Fownhope, Herefordshire, FH 2(3), SO 348586.

Description and Remarks. This species, like G. ramosa, has only short internodal lengths
of filaments. The filaments curve and have a high branching frequency. The ‘true’
diameter is more difficult to find in this and the last species because of this high branch-
ing frequency, constrictions, and adhering tubes, so single measurements may read
slightly lower than the limits shown for these forms. Graphs or histograms showing the
range of diameters for each group are valuable and the ‘peaks’ on the graphs give better
results than random measurements of odd filaments.

EXPLANATION OF PLATE 11

Fig. 1. Girvanella incompta sp. nov., coiled group. WE 11, Wenlock Limestone, GSM. PF 2844/3;
X 270.

Figs. 2-4. G. media sp. nov. 2, 3, branched specimens, tubes in contact; FH 2, Wenlock Limestone.
2, GSM. PF 2841/2; x 180. 3, GSM. PF 2841/1 ; X 100. 4, Less branched specimen, tubes short in
length; MF 1, Woolhope Limestone, GSM. PF 2837/3, x225.

EXPLANATION OF PLATE 12

Figs. 1, 3, 5. Girvanella ramosa sp. nov.; specimens with branched tubes of uneven diameter. 1, Holo-
type; WE 15, Wenlock Limestone, GSM. PF 2846/2; x 145. 3, WE 11, Wenlock Limestone, GSM.
PF 2844/4; X 180. 5, Paratype; WE 15, Wenlock Limestone, GSM. PF 2846/1 ; X 145.

Figs. 2, 4. G. effusa sp. nov. (larger diameter species than G. ramosa). 2, Holotype; FH 2, Wenlock
Limestone, GSM. PF 2840/3; X 160. 4, WE 1 1, Wenlock Limestone, GSM. PF 2845/1 ; X 160.

Palaeontology, Vol. 9

PLATE 11

JOHNSON, Silurian Girvanella

Palaeontology, Vol. 9

PLATE 12

JOHNSON, Silurian Girvanella

H. M. JOHNSON: SILURIAN GIRVANELLA

61

Distribution. Woolhope and Wenlock Limestones. Very few specimens have been ob-
tained from the Woolhope Limestone (Mordiford). This species is found in the Wenlock
Limestone at Fownhope and in the bedded sediments at Farley and south-western end
of Wenlock Edge.

Acknowledgements. The author acknowledges assistance from many quarters during the progress of
the work, in particular from Dr. F. W. Anderson, Dr. M. G. Calder, Professor P. W. Richards, and
Professor P. W. Wardlaw. The final stages have been completed in the Durham University Geology
Department thanks to the facilities kindly placed at my disposal by Professor K. C. Dunham; the
plates have been prepared by Mr. C. Chaplin and Mr. G. Dresser, and text-figure 1 by Miss M.
Lumley. A Durham University Research Fund grant towards the cost of preparing the paper is
gratefully acknowledged.

REFERENCES

butler, a. j. and oakley, k. p. 1936. The Dudley District. Proc. Geol. Assoc. Lond. 47, 133-6.
chapman, f. 1907. On the relationship of the genus Girvanella and its occurrence in the Silurian Lime-
stones of Victoria. Australasian Ass. Adv. Sci. Sec. 2, 377-86.
crosfield, m. c. and johnston, m. s. 1914. A study of Ballstone and Associated beds in the Wenlock
Limestone of Shropshire. Proc. Geol. Lond. 25, 193-228.
garwood, e. j. 1931. Important additions to our knowledge of the fossils Calcareous algae since 1913,
with specific reference to the Pre-Cambrian and Palaeozoic Rocks. Quart. J. geol. Soc. Lond. 87,
lxxiv-c.

and Goodyear, e. 1918. On the geology of the Old Radnor district with special reference to an

algal development in the Woolhope Limestone. Ibid. 74, 1-30.
green, h. m. 1955. The Calcareous Algae of the Woolhope and Wenlock Limestones from certain
localities in the Welsh Borderland. M.Sc. Thesis; University of Manchester.

1959. Calcareous Algae of the Silurian of the Welsh Border. Ph.D. Thesis; University of

Wales.

hill, p., butler, a. j., oakley, k. p., and arkell, w. j., 1936. ‘Coral Reef Meeting.’ Proc. geol. Ass.
Lond. 47, 130-2.

HiNDE, g. j. 1887. Review of Dr. J. G. Bornemann — The fossils of the Cambrian strata of the Island
of Sardinia. Geol. Mag., 34, 226.

hoeg, o. A., 1932. Ordovician algae from the Trondheim area. Skr. Norske Vidensk. Akad. i. Oslo,
Math.-Nat. 4, 63-96.

Johnson, j. h., konisht, k., and rezak, r. 1959. Studies of Silurian (Gotlandian) Algae. Quart. Colorado
School Mines, 54, 1-173.

lewis, h. p. 1942. On Girvanella in the ‘Shumardia Limestone’ of Levis, Quebec. Ann. Mag. Nat.
Hist. (11), 9, 49-55, pi. 1.

Nicholson, h. a. and etheridge, r. 1878. A Monograph of the Silurian Fossils of the Girvan District in
Ayrshire, vol. 1, fasc. 1. Edinburgh and London.
seely, h. m. 1885. A new genus of Chazy sponges, Strephochetus. Amer. Journ. Sci. 30, 355-7.
squirrel, h. c. and tucker, e. 1960. The geology of the Woolhope Inlier, Herefordshire. Quart. J.
geol. Soc. Lond. 116, 139-86.

wethered, e. 1889. On the microscopic structure of the Jurassic pisolite. Geol. Mag. 36, 196-220.

1 890. On the occurrence of the genus Girvanella in Oolitic rocks and remarks on oolitic structure.

Quart. J. geol. Soc. Lond. 46, 270-82.

1891. The Inferior Oolite of the Cotteswold Hills with special reference to its microscopical

structure. Ibid. 47, 550-70.

— — 1892. On the microscopic structure and residues insoluble in hydrochloric acid in the Devonian
Limestones of South Devon. Ibid. 48, 377-89.

1893. On the Microscopic structure of the Wenlock Limestone. Ibid. 49, 236-48.

whittard, w. f. 1952. A Geology of South Shropshire. Proc. Geol. Am. Lond. 63, 169-75.
wood, a. 1941. ‘Algal Dust’ and the fine grained varieties of Carboniferous Limestone. Geol. Mag.
78, 192-200.

62 PALAEONTOLOGY, VOLUME 9

wood, A. 1948. ‘ Sphaerocodium' a misinterpreted fossil from the Wenlock Limestone. Proc. Geol.
Ass. Loud. 59, 9-22.

1957. The type species of the genus Girvanella (Calcareous Algae). Palaeontology, 1, 22-28.

1963. The British Carboniferous species of Girvanella (Calcareous Algae). Ibid. 6, 264-73.

HAZEL M. JOHNSON

Department of Geology,
The University,

Manuscript received 12 November 1964 Durham

APPENDIX A

A Key to the Silurian species of Girvanella

1. Algal tubes less than 10 /a (rarely 11 p) in internal diameter; tubes even in width.

{a) Tubes straight or slightly undulating; indefinite length with a moderate contact between adjoin-
ing tubes; arranged in bundles, diameters mostly 5-7 ^ G.fragila (text-fig. 3 a)

lb) Tubes short, of a finite length; branching variable; tubes in loose coils or encrusting, with
adjoining tubes in contact; diameters 2-9 /u. mostly between 4 and Ip G. pusilla (text-fig. 3 b)

2. Algal tubes between 12 and 18 p (rarely 20 p).

(a) Tube diameter range (8) 12-18 p, average 13-14 p\ even in diameter.

(i) Tubes with frequent branching; adjoining tubes in contact; may or may not encrust; found
in small groups G. media (text-fig. 3c)

lb) Tube with average diameter over 15 p and under 18 p; even in diameter.

(i) Tube average diameter 15-16 p\ long, slightly curved, with occasional branching; loosely
coiling or meandering over a substrate; sometimes small amount of contact between tubes

G. problematica (text-fig. 3c)

(ii) Tube average diameter 15-16 p; very short, many strongly curved ; occasional contact between

adjoining tubes. Found encrusting G. problematica var. lumbricalis (text-fig. 3/)

(iii) Tube average diameter 15-16 p (in one locality recorded tubes average between 14 and 18 ^i);

indefinite length, straight or slightly undulating; adjoining tubes in contact and arranged in
faggot-like bundles G. prolixa (text-fig. 3 d)

(c) Tube diameter at internodes 11—18 /a. (20 p), average 15-16 /a; diameter variable.

(i) Tubes irregular, curved; adjoining tubes in contact; branching frequent and possible con-
strictions G. ramosa (text-fig. 3/)

3. Algal tubes over 18 p in diameter.

(a) Even diameter tubes in loose coils mostly not in close contact; branching occasional to frequent;

non-encrusting; diameter range 18-28 p, mostly 19-21 p G. incompta (text-fig. 3 It)

( b ) Tubes even, long, straight or slightly undulating, arranged in bundles. Few tubes in close contact,

branching rare; diameter range 18-31 p, mostly 19—22 /x G. sarmenta (text-fig. 3g)

(c) Tubes large, uneven in diameter; branching frequent; also possible constrictions; adjoining

tubes in contact; diameter at internodes 18-27 p, mostly 22-23 p G. effusa (text-fig. 3 j)

APPENDIX B

List of Girvanella localities to which reference has been made in the text and plates.
Woolhope Limestone

1. Old Radnor, Radnorshire; Dolyhir Quarries and Yat Farm exposures (quarries b, c,j, k, Gar-
wood and Goodyear 1918) OR 1 + 2, grid refs. SO 242582 and SO 246585.

2. Woolhope Village, Herefordshire; WH Exposure by road south of church, SO 357613.

3. Mordiford, Herefordshire; a, MF 1 Haugh Wood, several exposures; and junction with Littlehope
Lane (beds just above Petalocrinus Limestone); SO 372578. b, MF 2 Scutterdine Quarries, Littlehope,
SO 367581.

H. M. JOHNSON: SILURIAN GIRVANELLA

63

Wenlock Limestone

4. Wenlock Edge, Shropshire; WE 1-15, Various localities along its length from Farley; SO 015629
to SO 844447 and SO 856455 further south.

5. Dudley, Worcestershire; the Wren’s Nest (DU), various exposures, SO 917935.

6. Fownhope, Herefordshire; FH, The Old Quarries, Common Hill, SO 348586.

7. Soller’s Hope, Herefordshire; Lindell’s Farm, SO 330522.

8. Woolhope area, Herefordshire; a small roadside exposure, SO 340622.

9. Longhope, Gloucestershire; Hobb’s Quarry, SO 195695.

10. Ledbury, Herefordshire; Old Quarries by main Ledbury to Worcester road, SO 376715.

11. West Malverns; Disused quarry, Park Wood, SO 445763.

Bringewood Beds (Aymestry Limestone and the more calcareous Ludlovian facies)

12. View Edge, Craven Arms, Shropshire; SO 806426.

13. Ludlow, Whitcliffe; Old Quarry by River Teme, SO 742509.

LATE PERMIAN TRILOBITES FROM THE
SALT RANGE, WEST PAKISTAN

by RICHARD E. GRANT

Abstract. Three enrolled trilobites are the first reported from the Permian of the Salt Range. Two are
assigned to Ditomopyge fatmii sp. nov., the third to Kathwaia capitorosa gen. et sp. nov. Correlations point to a
Late Permian age (late Guadalupian or early Dzhulfian) making these among the youngest trilobites known.

Fossils have been known from the Productus Limestone of the Salt Range for more
than 110 years. The first were collected in 1848 by Dr. A. Fleming, and described by
Davidson (1862, p. 25). Since then the unusual abundance and excellent preservation of
invertebrates in this region has attracted the attention of several palaeontologists who
made large and detailed collections. The outstanding major work on invertebrates of
the Productus Limestone is a series of monographs by W. Waagen, published between
1879 and 1891, based on collections made by Waagen and by several geologists of the
Geological Survey of India. This was supplemented by several papers and a large mono-
graph by Reed (1944), based on collections by Reed and others, especially E. R. Gee
who spent several years mapping the geology of the Salt Range. Despite the extensive
search for fossils in these strata, however, no trilobites were reported from the Productus
Limestone. Waagen (1883, p. 402) does mention that ‘trilobites’ were reported from the
Productus Limestone by Theobald, but that these turned out to be fragments of lep-
totid brachiopods.

During the winter of 1963-4 the writer and A. N. Fatmi of the Geological Survey of
Pakistan had the opportunity to collect fossils from the Productus Limestone, and were
fortunate enough to find three well-preserved enrolled trilobites. These are described
here because of the importance of the Productus Limestone fauna to the Permian of
the world, and because their extreme rarity makes it unlikely that additional trilobite
specimens will be available soon.

Acknowledgements. It is a pleasure to express appreciation to the Smithsonian Institution which,
through G. A. Cooper, made possible the expedition to the Salt Range; to A. N. Fatmi, Geological
Survey of Pakistan, who collected the two specimens of Ditomopyge described; to C. Teichert (now
University of Kansas), M. G. White, and A. Davis (U.S. Geological Survey) all of whose help, both
geological and logistical, made the stay in Pakistan most pleasant; and to C. Teichert and A. R. Palmer
(U.S. Geological Survey) for criticism of the manuscript. The photographs are by D. H. Massie, and
the text-figure is by E. Stromberg (U.S. Geological Survey). This publication is authorized by the
Director, U.S. Geological Survey.

Stratigraphy of the Productus Limestone. The Productus Limestone is composed of
three units, traditionally designated the Lower, Middle, and Upper Productus Lime-
stones. Teichert (in press) has recognized that these are palaeontological units as cur-
rently defined, following usage by Waagen (1889) and subsequent workers. Flowever,
they also approximate the actual lithic breakdown of the Productus Limestone, and, in
fact, have been mapped as rock units by E. R. Gee (unpublished data).

[Palaeontology, Vol. 9, Part 1, 1966, pp. 64-73, pi. 13.]

R. E. GRANT: LATE PERMIAN TRILOBITES FROM THE SALT RANGE 65

The Lower Productus Limestone (i.e. the lower lithostratigraphic unit) is 200-500 ft.
thick, mostly sandstone, with the proportion of calcareous cement increasing upward.
Near the top it is sandy or silty, and in places argillaceous. It contains fusulinids, pre-
dominantly Parafusulina kattaensis (Schwager), in the lower three-fourths; these forms
are especially abundant in two rusty brown calcareous sandstone ledges that normally
occur about 40-80 ft. below the top. These fusulinids are considered by Dunbar (1933)
and Dunbar et al. (1960) to be Artinskian in age, near the middle of the Lower Permian
of the present classification (Sarycheva, ed. 1960, p. 10; Cohee 1960, p. 1578; Dunbar
et al. 1 960).

The Middle Productus Limestone is the only unit of the group that actually is pre-
dominantly limestone. It is a thick-bedded biosparite that forms a prominent cliff
ranging from about 200 ft. to more than 500 ft. in height. The lower beds are somewhat
clayey or silty, with bedding planes 3-6 ins. apart; the middle beds are much thicker,
up to 5 or 6 ft. ; the upper beds of the cliff-forming limestone are thinner, like those near
the base. At the top of the cliff a very fossiliferous unit of shaly limestone and cal-
careous shale or siltstone that normally weathers back from the cliff into irregular
nodular beds is included in the Middle Productus Limestone.

Above the nodular beds are about 40 ft. of weakly resistant green shale, or less
commonly fine-grained sandstone, forming the base of the Upper Productus Limestone.
This is followed by beds of sandstone, varying in grain size, amount of calcareous
cement, and abundance of fossils, which bring the aggregate thickness of the Upper
Productus Limestone to about 200 ft. The ammonoid Cyclolobus occurs in the upper
third of this unit, indicating a Late Permian (Dzhulfian) age (Glenister and Furnish
1961, p. 684).

Occurrence and age of the trilobites. All three specimens are from the uppermost beds
of the Middle Productus Limestone, weathered out of the unit of alternating calcareous
shale and nodular argillaceous limestone. This upper unit is 43 ft. thick at Zaluch Nala
where Ditomopyge was found (fig. 1, loc. 2) and 23 ft. thick in the vicinity of Kathwai
where Kathwaia was collected (fig. 1, loc. 1).

The uppermost beds of the Middle Productus Limestone are faunally similar to the
Upper Productus Limestone, and faunally distinct from the Lower Productus Lime-
stone (for faunal lists see Waagen 1889, pp. 180-242). Therefore the trilobites can be
considered late Guadalupian or early Dzhulfian, depending upon the range of Cyclolobus
and the significance attached to the faunal similarity of the uppermost Middle and the
Upper Productus Limestone. If this assignment is correct, the Salt Range trilobites are
among the latest known (Harrington et al. 1959, pp. O 399-0 403).

Two fragmentary pygidia from the Chitichun Limestone of Tibet (Diener, 1897, pp. 3-6)
were found with brachiopods that closely resemble many from the top of the Middle
Productus Limestone, and therefore may be the same age as the Salt Range trilobites.
However, Diener (1903) found the ammonoid Cyclolobus in the same unit as his trilo-
bites, which suggests that they may be slightly younger than the Salt Range specimens.

Classification. Meaningful classification of the few genera of Permian trilobites neces-
sarily depends upon correct understanding of the systematic affinities of their Carboni-
ferous forebears. Hessler (1963, p. 543) has undertaken such a study in conjunction with
his work on Mississippian Proetidae, and has indicated that he will offer a classification

F

B 6612

66

PALAEONTOLOGY, VOLUME 9

of this group in a forthcoming paper. Therefore, no suprageneric breakdown of proe-
tids or phillipsiids is attempted here.

Genus ditomopyge Newell 1931

1931 Diotomopyge Newell, p. 267, emend. 1935, Weller, J. M., p. 505.

1933 Cyphinium Weber, pp. 45, 81.

1937 Neophillipsia Gheyselinck, p. 56.

Type species. D. lansingensis Newell 1931, p. 268, pi. 31, figs. 31-32.

Diagnosis. Cephalon subtrigonal in outline, genal spines short. Glabella anteriorly
expanding, but not strongly inflated; three preoccipital lobes at posterior. Eyes promi-
nent, reniform, not elongate, located beside posterior third of glabella. Thorax with nine
segments. Pygidium nearly the same as cephalon in size and outline. Axis with 12-18
segments, pleural regions with somewhat fewer. Border distinct but without border
furrow; flat or concave, downsloping. Ornamentation weak, consisting of striations on
cephalic border and underside of pygidial border, and a transverse row of low pustules
or slightly elongated tubercles on each axial ring. Other ornamentation, such as low
granules or shallow pits on cephalon, may be present.

Discussion. The generic position of Cyphinium Weber has been discussed fully by Weller
(1935, p. 504; 1936, p. 711); it is to be regarded as a junior synonym of Ditomopyge.

The genus Neophillipsia Gheyselinck was established on a supposedly Upper Car-
boniferous specimen from Kansas, and Weller (1944, p. 320, pi. 49, fig. 3) demonstrated
beyond doubt that it belongs to Ditomopyge. In addition, he offered convincing evi-
dence that it was collected from beds in southern Kansas that now are considered to be
Lower Permian.

Mudge and Yochelson (1962, p. 96) report a specimen from the Lower Permian of
Kansas which they identify as Ditomopyge! decurtata (Gheyselinck). Their specimen was
collected from slightly higher in the section than where Weller surmised Gheyselinck’s was
obtained. They mention that the border continues around the anterior of the glabella and
thus differs from Weller’s (1936, p. 711) diagnosis. The Salt Range specimens also differ
in this feature, and perhaps are more closely related to the stratigraphically slightly
higher form described by Mudge and Yochelson; other characters resemble closely those
of D. scitula (Meek and Worthen) and D. decurtata (see PI. 13, figs. 4, 5).

Comparisons. Only Pseudophi/lipsia Gemmellaro (1890, p. 14) resembles Ditomopyge
sufficiently to warrant comparison here. To judge from reconstructions of Pseudo-
phiUipsia by Gemmellaro (1890, pi. 2, fig. 4) and Weller (1944, pi. 49, fig. 10; also in
Harrington et al. 1959, fig. 307, no. 4) and a systematic analysis by Goldring (1957), it
differs from Ditomopyge in its more anteriorly divergent facial sutures, smaller and
sharper basal pre-occipital lobes, strong granular ornamentation of the posterior of the
glabella, and its great number of pygidial segments.

Presence of a distinct cephalic border on the Salt Range species, and the Kansas
Permian specimens called Ditomopyge ? decurtata by Mudge and Yochelson (1962)
may be another indication of the close relationship of Ditomopyge with Pseudophillipsia
that was emphasized by Goldring (1957, pp. 197 et seq.). Goldring suggested that Dito-
mopyge as now constituted may involve two subgenera, one with a border, presumably

R. E. GRANT: LATE PERMIAN TRILOBITES FROM THE SALT RANGE 67

more closely related to Pseudophillipsia, and one lacking a border. Ditomopyge is dis-
tinguished from Pseudophillipsia by other features, important among which is its smaller
number of pygidial segments (Goldring 1957, p. 199).

Ditomopyge fatmii sp. nov.

Plate 13, figs. 2 a-cl, 3

Diagnosis. Cephalon with distinct border, slightly raised to form short flange; shallow
border furrow continuing around anterior of cranidium; genal spines short; pygidium
with seventeen axial segments and broad border.

Description. Carapace essentially smooth, ornamented only by few granules on posterior
part of glabella, weak striations on cephalic border and underside of pygidial border,
and by transverse row of low sagittally elongate tubercules on crest of each axial ring
of thorax and pygidium. Cephalon subtrigonal in outline, somewhat flattened on top,
then sloping steeply toward border, thus producing strong curvature of anterior end of
glabella, and of librigenae. Glabella elongate, anteriorly expanding but not inflated,
with three distinct pre-occipital lobes; lateral pair slightly longer than median one.
Glabellar furrows other than those outlining pre-occipital lobes, absent. Occipital ring
without median tubercle. Palpebral lobes small, crescentic, located posterior to mid-
length of glabella. Facial sutures strongly divergent anterior to eyes, then convergent across
border; posterior course cutting diagonally postero-laterally across posterior border.
Eyes prominent, outline nearly semicircular, not elongate. Librigenae trapezoidal,
posterior border furrow flexed posteriorly, distal edge of cheek produced to form short
genal spine. Thorax composed of nine segments; distal ends of pleurae bent gently down-
ward. Pygidium proportionately large: size and outline conforming to that of cephalon.
Axis with seventeen segments and terminal axial piece, crest flattened and nodular;
pleural regions with twelve slightly sigmoidal segments; interpleural furrows shallow,
pleural furrows indistinct. Border distinct, rather broad, gently concave, without border
furrow.

Measurements (in mm.)

Cephalon
Length Width

Thorax

Length Width

Pygidium
Length Width

Holotype, G.S.P.B. no. 88
Plate 13, figs. 2 a-d

8-8 11-5

10 8

10-2

8-8 10-7

Paratype, G.S.P.B. no. 89
Plate 13, fig. 3

7-9 c. 10

9-5

Locality. Both specimens were collected from the Middle Productus Limestone, on the divide between
Kala Wahan and Zaluch Nala, east of Pai Khel, Survey of Pakistan map 38 P/9 (text-fig. 1, loc. 2).
These beds can be traced along this divide between the two valleys, about one mile upstream from their
mouths. The holotype was found nearer Zaluch Nala, and the paratype nearer Kala Wahan.

Repository. The specimens are deposited in the collections of the Geological Survey of Pakistan, where
they have been assigned catalogue numbers G.S.P.B. 88 and 89.

Comparisons. Ditomopyge fatmii differs from D. scitula (Meek and Worthen) in its
more distinct cephalic border marked by a border furrow that continues around the

68

PALAEONTOLOGY, VOLUME 9

anterior of the glabella, its smaller eyes which are distinctly less than half as long as the
glabella, its greater number of pygidial segments, and probably also in its much shorter
genal spines. The spines are broken from both of the available specimens, but the shape
of the librigenae and the posteriorly directed course of the posterior border furrow are

text-fig. 1. Map of part of Salt Range, showing localities from which trilobites were obtained.

similar to those of D. decurtota (Gheyselinck), which has only short genal spines (PI.
13, figs. 4, 5). The Salt Range species differs from D. decurtata in its complete cephalic
border and border furrow, somewhat greater number of pygidial segments, and its
wider and more concave pygidial border.

Ditomopyge meridionalis Teichert (1944, p. 458) was described from the Wandagee
Formation of Western Australia, considered to be of Artinskian age (Teichert 1944, p.
456; Glenister and Furnish 1961, p. 679). This species differs from D.fatmii in its pro-
portionately shorter and more bulbous glabella, less prominent pre-occipital lobes,
and by the much smaller number of axial and pleural segments on the pygidium. In
addition Teichert (1944, p. 460) remarks on the unusual length, for a Permian species,
of the genal spines of D. meridionalis.

R. E. GRANT: LATE PERMIAN TRILOBITES FROM THE SALT RANGE 69

Among Lower Carboniferous species described by Weber (1933) none seem closely
similar to D. fatmii. The cephalic border of D. acanthicaudum (Weber) is very narrow,
according to Weber (1933, p. 84), and the pygidium has somewhat fewer axial and pleural
segments. The eyes seem to be proportionately larger than in D. fatmii, and the pre-
occipital lobe more prominent. Ditomopyge kumpani (Weber) is ornamented by con-
spicuous granules, and although the variety crassicrusta appears smooth, Weber (1933,
p. 85) says that it is exfoliated.

The Artinskian species D. artinskiense (Weber) differs from D. fatmii in its more
anteriorly expanding glabella, distinctly fewer pygidial segments, and more or less
transverse, rather than subcircular, median pre-occipital lobe. According to Weber
(1933, p. 89) the genal spines extended posteriorly nearly to the pygidium, much
further than in D. fatmii, and more like the Australian species D. meridionalis Teichert.
The trend toward decreasing length of genal spines in stratigraphically younger species
of Ditomopyge has been noted by Weller (1937, p. 346, text-fig. 4). The short spines of
D. fatmii argue further for its Late Permian age.

Genus kathwaia gen. nov.

Type species. Kathwaia capitorosa sp. nov.

Diagnosis. Cephalon semicircular in outline, genal spines absent. Glabella inflated,
protruding anterior to border, with two small but prominent lateral pre-occipital lobes.
Eyes small, near mid-length of cephalon. Thorax with nine segments. Pygidium short,
with border but no border furrow. Ornamentation granular, present on all parts but
strongest on cephalon.

Description. The genus is monospecific at present; therefore its characters are given
below in the description of the type species.

Comparisons. This genus shares features with several genera of phillipsioid trilobites.
It resembles Phillipsia Portlock in its general shape, granular ornamentation, and in-
distinctly delimited pygidial border. It differs from that Lower Carboniferous genus in its
lack of genal spines, its non-continuous cephalic border, smaller eyes located farther
anterior, absence of glabellar furrows (other than those outlining pre-occipital lobes)
and in its fewer pygidial segments. It is similar to Griffithides Portlock in its granular
ornament and lack of genal spines, but differs in its swollen glabella, non-continuous
cephalic border, semicircular rather than semi-oval or subtrigonal outline of the
cephalon, more prominent lateral pre-occipital lobes, and fewer pygidial segments.
Despite the Early Carboniferous age of Griffithides, it is the genus most similar to Kath-
waia.

Humilogriffthoides Inai, recently discussed by Endo and Matsumoto (1962, p. 159),
resembles Kathwaia in its swollen glabella and lack of genal spines. It differs from the
Salt Range genus in its smooth carapace, sharp genal angle, larger eyes, short lateral
glabellar furrows, and its well-defined pygidial border.

Among Permian genera only Neoproetus Tesch resembles Kathwaia in more than gen-
eral aspect. The glabella of Neoproetus is bulbous and overhangs the anterior margin as
in Kathwaia, similarly the ornamentation of the cephalon is granular, although Neo-
proetus also has some wrinkles on the anterior part of the glabella (Weller 1944, p. 325).

70

PALAEONTOLOGY, VOLUME 9

Kathwaia differs from Neoproetus in its semicircular rather than subtrigonal cephalon,
its proportionately longer glabella, more inflated lateral pre-occipital lobes, smaller
eyes, more rounded genal angle, divergent anterior course of the facial sutures, and in
the granular ornamentation that is present over the whole carapace, not confined to
the cephalon as in Neoproetus. The number of pygidial axial segments is 9 or 10 in
both species of Neoproetus, N. indicus Tesch and N. verrucosus (Gemmellaro) ; it is 8 on
the one known specimen of Kathwaia. In addition, the pygidium of Neoproetus is
described by Gheyselinck (1937, p. 76) and Weller (1944, p. 325) as having a slightly
segmented or crenulated border. This feature is not present in Kathwaia.

The Middle Permian genus Paraphillipsia Toumansky (1935) from the Crimea and
Himalaya differs from Kathwaia in its broader glabella, wider and more inflated pre-
occipital lobes, small eyes, and its proportionately short pygidium with laterally short
pleurae and very broad, granular, convex border.

Other Permian genera such as Pseudophillipsia Gemmellaro, Ameura Weller, and
Auisopyge Girty seem to belong to the group of trilobites containing Ditomopyge, and
do not strongly resemble Kathwaia or Neoproetus. Neogriffithides Toumansky from the
Permian of the Crimea has rounded genal angles similar to those of Kathwaia, although
according to Weller (1944, p. 326) these were reconstructed from disassociated libri-
genae, and may not belong to Neogriffithides. This doubt is reflected by Weller in his
reconstruction of that genus in the Treatise on Invertebrate Paleontology (Harrington
et al. 1959, fig. 306, no. 6). The multi-segmented pygidium of Neogriffithides, with its
flattened distinct border, is the chief distinction from Kathwaia.

Additional species. The granular species Proetus ? girty i Toumansky, from Crimean rocks
considered by Toumansky (1935, p. 38) to be Guadalupian in age, may belong to Kath-
waia. It differs rather strongly from the type species, K. capitorosa, in features that may
be considered generic. It has a proportionately broader axis, 10 rather than 9 thoracic
segments, stronger granular ornament, and differs markedly by possession of a cephalic

EXPLANATION OF PLATE 13

All figures x 3.

Figs. 1 a-d, Kathwaia capitorosa gen. et sp. nov., enrolled holotype from upper part of Middle Pro-
ductus Limestone, south of Kathwai (text-fig. 1, loc. 1), USNM 145320. a, dorsal view of cephalon
and part of thorax; b, dorsal view of pygidium and part of thorax; c, left lateral view of enrolled
specimen ; d, anterior view.

Figs. 2 a-d, Ditomopyge fatmii sp. nov., enrolled holotype from upper part of Middle Productus Lime-
stone at Zaluch Nala (text-fig. 1, loc. 2). GSPB 88. a, dorsal view of cephalon and part of thorax;
b, dorsal view of pygidium and part of thorax; c, right lateral view of enrolled specimen; d, anterior
view.

Fig. 3, D. fatmii sp. nov., incomplete enrolled paratype from locality 2 (text-fig. 1), GSPB 89. Dorsal
view of cephalon and part of thorax.

Figs. Aa-c. Ditomopyge decurtata (Gheyselinck), enrolled topotype from Permian Beattie Limestone
near Wichita, Kansas, for comparison with D. fatmii, USNM no. 145322. a, dorsal view of cephalon
and part of thorax; b, dorsal view of pygidium and part of thorax; c, left lateral view of enrolled
specimen.

Figs. 5 a-b, D. decurtata (Gheyselinck), enrolled specimen from Florena Shale Member of Beattie
Limestone, East of Grand Summit, Kansas, for comparison with D. fatmii, USNM no. 145321 ; a,
dorsal view of cephalon and part of thorax; b, dorsal view of pygidium and part of thorax.

Palaeontology, Vo I. 9

PLATE 13

GRANT, Permian trilobiles

R. E. GRANT: LATE PERMIAN TRILOBITES FROM THE SALT RANGE 71

border, which is said to be longitudinally striated. Aside from this possible additional
species, no species other than the type is known to belong to Kathwaia.

Kathwaia capitorosa sp. nov.

Plate 13, figs. 1 a-d

Diagnosis. Glabella strongly inflated, projecting anteriorly beyond cephalic border.
Occipital ring rather broad sagittally, with small median node. Pygidium with eight
axial segments plus terminal piece, length about two-thirds that of cephalon, border
flat to slightly convex, broad but indistinctly delimited from pleural areas, lacking
border furrow.

Description. Carapace with relatively strong granules randomly and rather densely
distributed on cephalon, and in apparent rows on axial rings; weaker granules on
pleural regions of thorax and pygidium. Edge of border to cephalon and pygidium
weakly striated. Cephalon approximately semicircular in outline, strongly convex.
Glabella (exclusive of lateral preoccipital lobes) strongly inflated to tear-drop shape,
extending anteriorly beyond border, interrupting anterior course of border except at
very margin. Lateral pre-occipital lobes proportionately large, inflated, each subtrigonal
in outline. Eyes small, hemispherical (not elongate); midlines located at furrows mark-
ing anterior ends of lateral pre-occipital lobes, about midlength of cephalon (exclusive
of anterior glabellar protrusion). Facial sutures nearly parallel to axis between eyes and
posterior border furrow, then strongly divergent posteriorly across border; anterior
course moderately divergent anterior to eyes, then convergent around glabella. Libri-
genae strongly convex laterally; sides nearly vertical; genal angle gently and evenly
curved, without genal spines. Occipital ring somewhat constricted laterally by posterior
edges of lateral pre-occipital lobes; low node on crest. Thorax with 9 segments; pleural
segments strongly flexed and sloping steeply to lateral margins. Pygidium about two-
thirds actual length of cephalon (excluding protrusion of glabella). Axis with 8 rings
and terminal axial piece; pleural regions with 7 pleurae, each subdivided by shallow
furrow increasing in depth laterally. Border broad, border furrow absent, proximal edge
marked only by terminations of pleural and inter-pleural furrows. Margin weakly
striated, outline conforming exactly to that of cephalon, forming tight seal during
enrolment.

Measurements (in mm.)

Cephalon

Thorax

Pygidium

Length

Width

Length

Width

Length

Width

Holotype, USNM no. 145320

Length of cephalon including protruding

8-5

120

16-7

12-0

6-5

10-7

glabella

90

Locality. Kathwaia capitorosa was collected from beds at the top of the cliff of Middle Productus Lime-
stone, 5-5 miles road-distance south of Kathwai, on the road to Kushab, Survey of Pakistan map 43
D/3 (fig. 1, loc. 1). The Middle Productus Limestone is faulted in such a manner that part of the cliff
and the uppermost beds are repeated three times. The road passes over the lowermost fault block ; the
trilobite was collected from the uppermost one, two fault ‘steps’ above the east side of the road, about
250 ft. above road level.

72

PALAEONTOLOGY, VOLUME 9

Discussion. With only one specimen available, it is uncertain which characters are diag-
nostic of the species. The diagnosis was composed with reference to species of the
seemingly most nearly related genus, Neoproetus, with the assumption that species
characters in Kathwaia are analogous.

Repository. The specimen is deposited in the United States National Museum.

REFERENCES

cohee, g. f. 1960. Series subdivisions of Permian System. Bull. Amer. Ass. Petrol. Geol. 44, 1578-9.
davidson, t. 1862. On some Carboniferous Brachiopoda collected in India by A. Fleming, M. D.

and W. Purdon, Esq., F.G.S. Quart. J. Geol. Soc. Loud. 18, 25-35, pi. 1-2.
diener, c. 1897. The Permocarboniferous fauna of Chitichun No. 1 .Palaeont. Indica, ser. 15, 1, pt. 3,
1-105, p. 1-13.

1903. Permian fossils from the central Himalayas. Ibid., pt. 5, 1-204, 10 pi.

dunbar, c. o. 1933. Stratigraphic significance of the fusulinids of the Lower Productus Limestone of
the Salt Range. Rec. Geol. Surv. India, 66, 405-13, pi. 22.

et al. 1960. Correlation of the Permian formations of North America. Bull. Geol. Soc. Amer. 71,

1763-1806.

endo, R. and matsumoto, e. 1962. Permo-Carboniferous trilobites from Japan. Saitama Univ. Sci.
Reports, B, 4, 149-72, pi. 8-10.

gemmellaro, G. G. 1890. 1 crostacei dei calcari con Fusulina della valle del Fiume Sosio nella provencia
di Palermo in Sicilia. Mem. Soc. ital. Sci. nat., ser. 3, 8, 1-40, pi. 1-5 (Naples, 1892).
gheyselinck, r. f. c. R. 1937. Permian trilobites from Timor and Sicily. Meded. Geol. Inst. Amsterdam,
73, i-xvi, 1-108, pi. 1-4.

glenister, b. f. and furnish, w. m. 1961. The Permian ammonoids of Australia. J. Paleont. 35, 673-
736, pi. 78-86.

goldring, r. 1957. Pseudophillipsia (Tril.) from the Permian (or Uralian) of Oman, Arabia. Sencken-
bergiana leth. 38, 195-210, pi. 1.

Harrington, H. j. et al. 1959, in r. c. moore, ed.. Treatise on Invertebrate Paleontology, Part O,
Arthropoda 1. Lawrence, Kansas.

hessler, r. r. 1963. Lower Mississippian trilobites of the family Proetidae in the United States, Part 1.
J. Paleont. 37, 543-63, pi. 59-62.

mudge, m. r. and yochelson, e. l. 1962. Stratigraphy and paleontology of the uppermost Pennsyl-
vanian and lowermost Permian rocks in Kansas. Prof. Pap. U.S. geol. Surv. 323, 1-213, pi. 1-17.
Newell, n. d. 1931. New Schizophoriidae and a trilobite from the Kansas Pennsylvanian. J. Paleont.
5, 260-9, pi. 31.

reed, f. r. c. 1944. Brachiopoda and Mollusca from the Productus Limestones of the Salt Range.

Palaeont. Indica, n.s. 23, Mem. 2, 1-678, pi. 1-65.
sarycheva, t. g. ed. 1960. Osnovi Paleont ologii. Izdatelstvo Akad. Nauk. S.S.S.R., Moscow. 3,
Bryozoa and Brachiopoda, 1-343, pi. 1-7, 1-75.
teichert, c. 1944. Permian trilobites from Western Australia. J. Paleont. 18, 455-63, pi. 77.

— — in press. Nomenclature and correlation of the Permian ‘Productus Limestone’, Salt Range, West
Pakistan. Mem. Geol. Surv. Pakistan.

toumansky, o. g. 1935. The Permo-Carboniferous beds of the Crimea, Part 2, The Permo-Carboni-
ferous trilobites of the Crimea. Trans, cent. geol. Inst., Leningr., pp. 1-63, pi. 1-12. (Russ, with Eng.
summary).

waagen, w. 1883. Salt Range fossils. Productus-Limestone fossils, pt. 1, fasc. 2. Palaeont. Indica,
ser. 13, 1, 391-546, pi. 29-49.

1889. Salt Range fossils. Geological results. Palaeont. Indica, ser. 13, 4, 1-242, pi. 1-8.

weber, v. n. 1933. Trilobites of the Donetz Basin (Russian with Eng. summary). Trans, geol. Prosp.
Serv., U.S.S.R. 255, 1-95, pi. 1-3.

1937. Trilobites of the Carboniferous and Permian system of U.S.S.R. (Russian with Eng. sum-
mary). Monogr. Palaeont. S.S.S.R. 71, 1-59, pi. 1-11.

R. E. GRANT: LATE PERMIAN TRILOBITES FROM THE SALT RANGE

73

weller, j. m. 1935. Adolescent development of Ditoinopyge. J. Paleont. 9, 503-13.

— 1936. Carboniferous trilobite genera. Ibid. 10, 704-14, pi. 95.

1937. Evolutionary trends in American Carboniferous trilobites. Ibid. 11, 337-46, figs. 1-4.

1944. Permian trilobite genera. Ibid. 18, 320-7, pi. 49.

RICHARD E. GRANT

U.S. Geological Survey,
Washington, D.C.

Manuscript received 21 December 1964

THE MICROSTRUCTURE OF STROMATOPOROIDS

by COLIN W. STEARN

Abstract. Fourteen microstructures are recognized in stromatoporoid tissue, but not all are features of the
tissue laid down by the organism and much fibrous, transversely porous, flocculent, pseudotubular, and melano-
spheric tissue has developed from other microstructures during the process of preservation. Irregular specks of
opaque matter less than 5 p across are present in the tissue of most stromatoporoids but some have uniformly
coloured tissue. During preservation these specks have moved, diffusing out into the galleries and concentrating
into spheres. Fibrosity caused by intercrystalline boundaries and by the alignment of the specks transverse to
the structural elements or in a water jet pattern is both an original microstructure and developed during pre-
servation. Compact tissue in which the specks are evenly distributed passes through stages of alteration due to
the concretionary tendency of the specks, in which it is successively flocculent, pseudotubular, and finally
characterized by dark spherules (melanospheres).

Tissue of the family Stromatoporidae is marked by subspherical voids or dark spherules. Originally the tissue
was cellular, filled with subspherical voids. The tissue marked by dark spherules is here called melanospheric
tissue and has been altered by the breakdown of the walls between the cellules and the concentration of the dark
specks in the regions between. A puzzling feature is the occurrence of other microstructures such as micro-
laminae, tubules, tripartite laminae, etc., in both black-on-white and white-on-black states interchangeably.

The microstructures of 22 common genera and of the family Labechiidae are reviewed and emended diagnoses
of genera provided. Clathrodictyon cannot be defined in terms of cysts placed side to side, as the holotype has
regular continuous laminae. The laminae of Stictostroma and Stromatoporella contain a single line of cellules
opening to the galleries on either side by a pore. Clathrocoilona has compact tissue and is closely allied to
Stictostroma. Actinodictyon and Pseudoactinodictyon have compact tissue and are similar to Clathrodictyon and
Anostylostroma respectively. Trupetostroma has vacuolate tissue but the vacuoles are not concentrated at the
peripheries of the pillars. The median light zone of its laminae is probably the result of the merging of a plane of
cellules during preservation. Most species of ldiostroma should be assigned to other genera as they have in
common little but external form and do not have the microstructure of the type species. Hennatostroma is
polyphyletic and includes species with compact and cellular tissue. It is characterized by marginally cellular
or vesicular pillars. The tissue of Stachyodes contains fine dark or light rods. Parallelopora may be distinguished
from other members of the family Stromatoporidae by its large cellules arranged in vertical series. The cysts of
the Labechiidae are a single thin layer of compact tissue and the enclosing layers of flocculent or striated tissue
are due to diffusion of specks into the galleries during preservation.

Palaeontologists of the early and middle parts of the nineteenth century described
stromatoporoids in terms of their external form and surface. Rosen (1867), Nicholson
(1886-92), and others discovered the internal structure of these fossils by examining
them in thin sections cut tangentially and perpendicular to the banding. The coenosteum
was revealed as composed mainly of laminae, pillars, cysts (or dissepiments), and
amalgamate tissue in various combinations. These, and a few others that occur rarely,
may be referred to as the structural elements of the stromatoporoid coenosteum and
their arrangement defines a structure. Nicholson showed that the stromatoporoids could
be divided into species, genera, and families on the basis of their structure. He also
described the microstructures of the tissue which is the substance of the structural
elements and distinguished non-porous, porous, and tubulated tissue. Heinrich (1914)
based a classification of the stromatoporoids on the microstructure of the tissue and
since that time students of this group have taken various stands on the nature of the
microstructures of various genera, the significance of microstructures for classification,
and the extent to which different microstructures are the result of the processes of
preservation. In one of the two recently published classifications (Galloway 1957) the

[Palaeontology, Vol. 9, Part 1, 1966, pp. 74-124, pis. 14-19.]

C. W. STEARN: THE MICROSTRUCTURE OF STROMATOPOROIDS 75

microstructure of the tissue is important in separating familes and genera; in the other
(Lecompte 1951-2) differences in microstructure are considered to be largely secondary.

The lack of agreement on the interpretation of the microstructure of the order is
well illustrated by a review of opinions on the tissue of the genus Stromatopora and the
family Stromatoporoidae. The tissue of the genus occurs in two states: either marked by
light dots on a darker background or by dark dots on a lighter background. Nicholson
(1886-92) described the tissue as porous and recognized the light dots as vacuities
that may be filled with bituminous matter in some states of preservation so that they
appear to be darker than the surrounding tissue. His concept of the nature of the tissue
is better expressed by the term cellular. Parks (1936) did not consider the Stromato-
poridae extensively but believed that their structure was like that of Actinostroma on a
much smaller scale; that is, the light tissue contained an open, three-dimensional lattice
of dark tissue. Yavorsky, in his extensive writings on this group, well summarized in
1963, followed Nicholson in describing the tissues as porous. Lecompte (1951-2, 1956)
described the tissue as ‘cellulaire’ or ‘alveolaire’ with the cells arranged in a reticulate
pattern. He believed that the fundamental microstructure of stromatoporoids whose
tissue others would describe as maculate, vacuolate, porous, or striated is microreticu-
late. He also believed that the study of microstructures is not far enough advanced to
provide a basis for classification.

Against the writers noted above Galloway (1957), and Galloway and St. Jean (1957)
maintained that the microstructure of Stromatopora is not porous or cellular. They
described it as marked by maculae which are typically dark dots with a light centre but
may appear as solid dark or light dots. They would also distinguish other types of
microstructure in the stromatoporoids, such as microreticulate, porous, vacuolate,
etc., which are not accepted as distinct from cellular by Lecompte. St. Jean (1963)
has recently reviewed the concept of maculate microstructure.

Such opposing views require reconciliation before a classification can be based on the
microstructure of this group. Such a reconciliation is attempted in this paper but no new
classification is presented.

Faced with difficulty in applying the schemes of classification based on microstructure
to the determination of stromatoporoids from western Canada, I tried to seek the basis
for the conflicts by studying the major collections during 1964. Those examined were:
parts of the Galloway and St. Jean collection at the University of North Carolina; the
Nicholson collection at the British Museum (Natural History); the type materials of
Bargatzky, Maurer, and Heinrich, assembled by E. Fliigel, attheTechnischeHochschule,
Darmstadt ; the Ripper collections of Australian stromatoporoids at the British Museum
and at the Sedgwick Museum, Cambridge ; and type specimens of Parks’s genera at the
Royal Ontario Museum, Toronto. Unfortunately the collections of Lecompte at Brussels
were unavailable for study.

The Mesozoic sphaeractinoids were not considered as stromatoporoids for the pur-
poses of this paper and their structure was reviewed by examination of the Hudson
collection at the British Museum (Natural History) only for purposes of comparison.

Thin sections in the Nicholson collection at the British Museum (Natural History)
are referred to by Nicholson’s number with the prefix ‘Nich. Thin sections at the
University of North Carolina are referred to with the prefix UNC and those at the
Royal Ontario Museum with the prefix ROM.

76

PALAEONTOLOGY, VOLUME 9

Acknowledgements. I am indebted to the following for grants in support of this research work: the
British Council, the Royal Society, the Committee on Research of McGill University, the National
Research Council of Canada, Imperial Oil Ltd., and the British American Oil Co. Ltd.

For hospitality while visiting universities and museums and for profitable discussion of the prob-
lems of the stromatoporoids I am deeply grateful to Dr. Joseph St. Jean (University of North Carolina),
Dr. H. Dighton Thomas and Dr. W. J. Rees (British Museum (Natural History)), Dr. Erik Fliigel
(Technische Hochschule, Darmstadt), and Dr. Colin Forbes (Sedgwick Museum, Cambridge). I am
doubly indebted to Drs. St. Jean, Fliigel, and Thomas for critically reading the manuscript and offering
suggestions for its improvement.

CLASSIFICATION OF MICROSTRUCTURES

The Scale of Microstructures. The structural elements are composed of tissue that is
commonly several hundreds of microns thick (wide as it appears in thin section).
To see microstructures which have dimensions in the range 0-5-50 /x this tissue must be
examined in suitably thin sections at magnifications of 50-100 times. Photographs at
magnifications less than 20 times do not show microstructures clearly. Very few papers
describing stromatoporoids present photographs of magnifications suitable for the
study of microstructures (exceptions are St. Jean 1960, 1962). In thick sections which
may show gross structures well, fine structures are obscured by the overlapping of
features in depth. Rarely even in thinner sections can particles less than 1 /x in diameter
be related to the microstructural pattern, for where the section is thin enough to allow
such particles to be clearly resolved, it is too thin to show the structure in which these
particles participate. Although sections of uniform thickness are generally considered
desirable, sections which taper to a feather-edge in at least one direction are more
suitable for the study of microstructures.

A magnification of 50-100 times has been adopted in this study as suitable for
showing microstructures clearly. None of the photomicrographs illustrating this paper
is retouched.

Crystallinity and Preservation. Most Palaeozoic stromatoporoids have been preserved
by the infiltration of the highly porous galleries by calcite-depositing water. Coenostea
are infiltrated in whole or part by calcareous mud but these are very rare. Specimens in
which the galleries are now empty have commonly been silicified or dolomitized and
have in the process lost all but their grossest structure. In those stromatoporoids that are
preserved in the greatest detail the crystal boundaries of the calcite mosaic show no
relationship to the tissue of the fossil. Crystals are usually large and their boundaries
cross pillars and laminae indiscriminately. The orientation of the crystals in this mosaic
is random and therefore growth did not take place in optical continuity with the calcite
laid down by the organism. The whole coenostea of these specimens must have been
faithfully replaced and filled as a unit by infiltrating lime-rich waters, probably soon
after burial when the passage of water through the sediment was at a maximum.
St. Jean (1962) drew attention to this form of preservation in Stictostroina from Ontario.

In other specimens crystallization of the calcite filling the galleries has been controlled
by the structure of the fossil. First a coating of small crystals was laid down on the
edges of the galleries. Later the central areas of the galleries were filled in with coarser,
clearer crystals of calcite. In still other states of preservation the galleries seem to have
been filled with a single generation of coarse crystals and the tissue replaced by a finer

C. W. STEARN: THE MICROSTRUCTURE OF STROMATOPOROIDS

77

crystal mosaic. The galleries of many Mesozoic sphaeractinoids are filled with a single
generation of calcite crystals and the edge of their fibrous tissue is the sharp boundary
of these crystals. Some of the specimens, however, seem to have a thin, clear line of
crystals between the fibres and the coarse crystals of the galleries. This thin zone is
optically related to the fibres of the tissue.

The calcite in stromatoporoids from structurally disturbed rocks is commonly a
coarse mosaic of crystals with twin lamellae on which the outline of the tissue is super-
posed. The microstructure in such specimens is almost always obliterated and the tissue
may be so diffused that even the gross structure is lost. Yet some such recrystallized
specimens retain the structure clearly outlined in even, very finely dusty, tissue. The
stromatoporoids from Devon show a great variation in this respect; some are com-
pletely obscured, others retain the structure and rarely traces of the microstructure.

rsiii'.n

r

COMPACT

M

TRANSVERSELY
FI BROUS

FLOCCULENT

jyp

fsrfts- vU

TUBULATE CELLULAR VACUOLATE

TRANSVERSELY

POROUS

MICRO-

RETICULATE

Jill

/i§\

w~

■tiariggg

/ •’ -

/■

MELANO-

SPHERIC

PERIPHERALLY

VESICULAR

STRIATED

TRIPARTITE

LAMINA

text-fig. 1. Stromatoporoid microstructures.

Summary of Microstructures. Descriptions of the microstructures of stromatoporoids
have generally been interpretive and not objective. Detailed description of the highly
magnified tissue is uncommon in papers on this group. The small size and indifferent
preservation of these microstructures seem to have discouraged objectivity and en-
couraged idealization so that they have been described in terms of what they might
have been rather than what they are now.

The several types of microstructure that have been recognized are listed below and
are briefly described. They are considered in more detail in following sections or under
the appropriate genera. The smallest elements of microstructure are dark particles of
irregular shape and unknown composition from 1 to 5 /x in diameter. These particles
are called ‘specks’ (after Nicholson 1886) and considered in more detail below. Through
the co-operation of Drs. Erik Fliigel, Heldur Nestor, and D. Le Maitre the equivalent
terms in German, Russian and French are included in this summation. These equivalent
terms were requested by circulating to these palaeontologists a sheet illustrating the
various microstructures in photographs and drawings, and briefly describing them in
English. French equivalents were also extracted by the writer from Fecompte’s (1951-2)
monograph. Most of these microstructures are illustrated diagrammatically in text-fig. 1.

78

PALAEONTOLOGY, VOLUME 9

1. Compact (German — kompakt; Russian — nAomaa ; French — compacte). Tissue composed of
evenly distributed specks or evenly coloured calcite.

2. Transversely fibrous (German — quer-fibros ; Russian — noncpcmioBOAOKHi-icxax ; French —
fibreuse). Tissue traversed by vague lines caused by the alignment of specks or by crystal boun-
daries.

3. Water jet (German — wasserstrahlartig or radialstruktur; Russian — nepucroBOAOKHiioTaH ;
French — en jet d'eau). Tissue marked by fibres that spray outwards and upwards from a central
zone.

4. Flocculent (German — flockig; Russian — xAonbe3iinHaa ; French — spongieuse). Tissue composed
of unevenly distributed specks or colouring.

5. Transversely porous (German — quer-poros; Russian — rionepemionopiicTaa ; French — poreuse).
Tissue traversed by pores which open into the galleries.

6. Tubulate (German — mit Rohren; Russian — Tpy6naTaa; French — tubuleuse [Le Maitre], tubulee
[Lecompte]). Tissue contains curved and branching tubes, commonly horizontal.

7. Cellular (German — zellig; Russian — aneiicTan or ueAAK>AflpHaa ; French — celluleuse [Le Maitre]
cellulaire, alveolaire [Lecompte]). Tissue filled with closely spaced subspherical voids. This term
has also been translated in German as ‘makulat’, equivalent to Galloway’s (1957) ‘maculate’.

8. Vacuolate (German — mit Vacuolen; Russian — KaBepH03Haa; French — vacuolaire). Tissue con-
tains subspherical voids, larger and more distinctly spaced than cellules.

9. Microreticulate (German — mikroretikulat; Russian — xoHKopemcxxaxax ; French — microreticulee).
Tissue contains cellules arranged in vertical and horizontal series.

10. Melanospheric (German — melanospharisch; Russian — iraxmicxaH or KparraaTas). Tissue filled
with dark subspherical groups of specks.

1 1 . Peripherally vesicular (German — randlich vesikular ; Russian — KpaeBony3Mpncxas ; French — a
zone externe vesiculaire [Le Maitre], caniculo-cellulaire marginale [Lecompte]). Structural elements
bordered by a layer of vesicles or a continuous membrane.

12. Striated (German — gestreift; Russian — noAocaxaa ; French — striee). Tissue filled with dark or
light, thin, rod-like bodies.

13. Tripartite laminae (German — dreigeteilte Lamina; Russian — TpexcAoimaji AaMnna; French —
trilaminaire [Le Maitre], Lamelle bipartite [Lecompte]). Lamina has a central light zone that
may break up into a line of cellules.

I propose to call the microstructure of laminae which are occupied by a single plane of cellules
‘ordinicellular’ (Latin: row of small chambers).

14. Meshed fibre (French — ? fibrillaire [Lecompte]). Lamina longitudinally fibrous, composed of
intermeshed fibres and dissepiments. (Used only by Lecompte (1956) for Synthetostroma).

In a recent letter Yavorsky indicated that he uses translations of Nicholson’s terms
to describe the microstructure of the stromatoporoids. He believes that the word ‘com-
pact’ adequately describes the microstructure of all the genera in Nicholson’s hydro-
actinoid group. For the microstructure of Stromatoporella he uses ‘thread-like’ or
‘stream-like’. Yavorsky (1962) recently illustrated his concept of the microstructure of
many stromatoporoid genera.

Specks and Speckled Tissue. When appropriately thin sections of many stromatoporoids
are magnified about 100 times the tissue can be seen to be crowded with dark irregular
particles which Nicholson (1886) referred to as specks. Nicholson believed these to be
fillings of minute pores or tubules and Lecompte (1951, p. 16) also favoured this inter-
pretation. Galloway (1957, p. 361) noted that their composition is unknown, that they
are probably deposited by infiltrating water and may consist of iron oxide, sulphide, or
organic carbon. St. Jean (1962) described the specks in Stictostroma in detail but referred
to them as flecks. He noted their movement and their tendency to clump into what he
called pseudomaculae. Lecompte (1951) believed them to be centres of carbonaceous
pigment.

C. W. STEARN: THE MICROSTRUCTURE OF STROMATOPOROIDS 79

The specks are highly irregular in shape but rarely do they seem to be tubular (PI. 14,
fig. 1). As they are only 0-5-5 p in diameter determination of their composition is
difficult. The electron probe may offer a hope for a solution of this problem. St. Jean
(1962) reported the probable presence of an amino acid in a Palaeozoic stromatoporoid.
Such a compound was probably located in the dark specks of the tissue but its position
remains conjectural as the analysis was made by dissolving the whole coenosteum.

The specks seem to be bodies of carbonaceous matter that have been concentrated
from the diffuse organic matter originally present throughout the calcite crystals
secreted by the organism. The aragonite hard parts of recent hydrozoans do not con-
tain much organic matter. However, Moseley (1891) quoted analyses of skeletal tissue
of Millepora showing several per cent, of organic matter and perhaps the stromato-
poroids incorporated more within their tissue than their modern representatives. Many
modern hydrozoans have plant parasites in their hard parts but the specks of stromato-
poroid tissue are too evenly and universally present to be the remains of such parasites.

Some well-preserved specimens show no specks but the tissue is distinguished from
the clear gallery fillings by its darker, usually brownish, colour or by a milkiness that
cannot be resolved into particles. Such tissue is common in Devonian stromatoporoids
from Germany and Canada. This darker tissue may have specks superposed upon it
or more commonly the concentration of specks alone distinguishes the tissue from the
galleries. The coloured, speckless tissue is believed to be close to the form in which it
was secreted before the specks emerged.

In different stratigraphic units the stromatoporoids are preserved differently. The
stromatoporoids from Devon are recrystallized ; those from the Silurian of Shropshire
have widely spaced dark specks and crystal mosaics independent of the structure; those
from the Devonian of the central United States are coarsely speckled and few show clear
cellular structures; those from the Devonian of western Canada either have very fine
specks closely set or brown coloured speckless tissue. In well-preserved stromatoporoids
the edge of the speckled tissue is sharply defined. The specks, however, are capable of
movement during such processes of preservation as the lining of cavities, infiltration of
voids, replacement, and recrystallization. Whether the specks themselves move or
whether the diffuse organic matter from which they eventually emerge does so is
difficult to determine now, but the latter seems more likely. The end result is that the
specks seem to have moved (PI. 15, fig. 2). As St. Jean (1962) has shown they move out
into the galleries, making the boundary of the tissue gradational and diffuse. They also
show concretionary tendencies and clump into what St. Jean (1962) has called pseudo-
maculae, but which are here called melanospheres and are discussed below.

Transverse Fibrosity and Water Jet Microstructure. Fibrous microstructure is widespread
in a wide variety of genera of stromatoporoids. This fibrosity locally seems to be caused
by the alignment of the specks across the elements of structure and locally by the de-
velopment of transverse intercrystalline boundaries. Although transverse fibrosity is
more marked in some genera such as Anostylostroma (PI. 14, fig. 4), Amphipora, and
Stictostroma, it may develop under appropriate conditions in almost any tissue. In
tangential section such fibrosity may seem to be radial in the pillars, or patches of
tissue may show a direction of fibrosity unrelated to neighbouring patches. In some
states of preservation the alignment of specks into a fibrous microstructure develops

80

PALAEONTOLOGY, VOLUME 9

over relatively large areas of tangential sections either regionally aligned in a certain
direction or in swirls. Tangential sections of Syringostroma from the central United
States commonly show this type of fibrosity.

The long rod-like pillars of some genera show fibres radiating upwards and outwards
from their axes. This produces a water jet or feather structure in longitudinal section
and a radial structure in tangential section. Such a fibrosity is particularly characteristic
of Mesozoic sphaeractinoids (Steiner 1932, Hudson 1958) but occurs in certain states
of preservation in many Palaeozoic genera and seems in the latter to be a microstructure
formed during preservation. It occurs in two positions: ( a ) within the tissue, and ( b )
within the gallery space. If it occurs in the tissue of the pillars it may be accompanied
by an axial dark line as in those specimens identified by Nicholson as Stromatopora
beuthii (not Hermatostroma beuthii (Bargatzky) of Lecompte, with which only one or
two of Nicholson’s specimens agree) or in Actinostroma clathratum Nicholson. I have
also observed water jet microstructure in some species of Taleastroma and Actinostroma.
One of Nicholson’s specimens of Stromatopora bucheliensis (Bargatzky) (Nich. 97) shows
(PL 14, fig. 2) that water jet fibrosity may form in cellular tissue completely unrelated
to the position of the cellules. A water jet structure has also been described by Stearn
(1961, 1962) in Taleastromal confertwn Stearn, Syringostroma bifurcum Stearn, and
Trupetostroma pycnostylotum Stearn, and may be seen in several poorly preserved
stromatoporoids in the Nicholson collection with thick, closely set pillars.

In all these species the fibrosity appears to be caused by the outward growth of fibrous
crystals from an axis. Galloway (1957, p. 464, pi. 31, fig. 15) illustrated a microstructure
like a water jet in Syringostroma densum Nicholson caused by the spray-like arrange-
ment of melanospheres in the pillars. This micro structure is not comparable with that
shown by specimens described above.

Water jet fibrosity in the galleries is well illustrated by specimens of Parallelopora
dartingtonensis (Carter) from Devon which Nicholson (1886) described as ‘reversed’.
In these the tissue is light coloured or clear and the calcite of the galleries and astro-

EXPLANATION OF PLATE 14

Fig. 1. Syringostroma subfuscum Galloway and St. Jean (UNC 294-88). Vertical section of paratype,
X 250, showing the specks made of irregularly distributed dark matter. Middle Devonian, 5 miles
E. of Logansport, Indiana.

Fig. 2. Stromatopora bucheliensis (Bargatzky) (Nich. 97). Vertical section of holotype, X 100, showing
cellular microstructure and suggestion of water jet fibrosity in the pillars. Middle Devonian, Btichel,
Germany.

Fig. 3. Anostylostroma arvense Parks (UNC 272-16). Vertical section, X 100, showing lack of fibrosity
in tissue surrounded by sediment. Middle Devonian, Marblehead, Ohio.

Fig. 4. Anostylostroma laxum (Nicholson) (UNC 282-52). Vertical section, X 100, showing transversely
fibrous laminae and darker pillars. Middle Devonian, 5 miles NW. of Columbus, Ohio.

Fig. 5. Stromatopora mononensis Galloway and St. Jean (UNC 278-21). Vertical section of holotype,
X 100, showing clear microlaminae traversing speckled tissue. Middle Devonian, 2 miles W. of
Monon, Indiana.

Fig. 6. Clathrodictyon vesiculosum Nicholson (Nich. 2184). Vertical section, X 100, showing compact,
speckled tissue. Middle Silurian, Much Wenlock, Shropshire.

Fig. 7. Stromatopora tvpica Rosen (Nich. 51a). Tangential section, X 100, showing transitional stage
from cellular to melanospheric tissue. Middle Silurian, Kaugatoma, Oesel Island.

Fig. 8. Stictostroma damnoniensis (Nicholson) (Nich. 355). Vertical section, X 100, showing flocculent
tissue with slight development of fibrosity. Middle Devonian, Teignmouth, Devon.

Palaeontology, Vol. 9

PLATE 14

STEARN, Stromatoporoid microstructures

C. W. STEARN: THE MICROSTRUCTURE OF STROM ATOPOROIDS 81

rhizal canals, which is normally clear, is darkly stained apparently by bituminous matter.
Nicholson’s suggestion that the filling is a recrystallized calcareous mud does not seem
likely. This dark filling shows a fibrous crystallinity in which the crystal boundaries
curve outwards from the axis of the galleries or canals and the margins of the tissue are
irregular. This kind of preservation is uncommon but its existence confuses the inter-
pretation of specimens common in the Devonian of Alberta (Stearn 1963, p. 667) whose
whole structure consists of contiguous water jets.

The examples given above establish that fibrosity due to crystalline boundaries
occurs in a variety of otherwise unrelated genera and may develop during the preser-
vation of the specimen. In some genera such as Anostylostroma, Stictostroma, and
Amphipora, transverse fibrosity is almost universally present and in such genera may be
of systematic importance. In some specimens of Anostylostroma in which the galleries
have been locally filled with calcareous mud rather than crystalline calcite (PL 14, fig. 3),
the transverse porosity common in the genus has not developed and the tissue is crypto-
crystalline and without specks. This suggests that transverse porosity even in these
genera may be a feature of preservation and should be used with care in taxonomy.
Further research may show that transverse porosity due to crystalline boundaries is a
phenomenon of preservation and transverse porosity due to alignment of specks is a
primary microstructure.

In Mesozoic sphaeractinoids the fibres curve outwards from an axial strand (that
may or may not be darker) to meet the border of the tissue perpendicularly (clinogonal
of Hudson 1958) or extend straight out from the axial strand (orthogonal of Hudson
1958). This fibrosity is invisible in plane light in certain states of preservation but under
crossed nicols it is revealed as due to the alignment of fibrous crystals. In places dark
specks may also be aligned with the fibrosity but it is fundamentally crystalline.

Flocculent Microstructure. In many stromatoporoids the speckled tissue is neither fibrous
nor homogeneous but the density of the specks and colour is not uniform. Such tissue
looks blotchy and has the texture of curdled cream (PI. 14, fig. 8). Galloway and St.
Jean (1957) referred to it as flocculent tissue. It is common in thick stromatoporoid
tissue such as occurs in Stromatoporella and is further discussed under that genus.

Cellular, Melanospheric, Maculate, and Microreticulate Microstructure. The tissue of
Stromatopora and related genera contains spherical clear areas or dark areas (PI. 18,
figs. 2, 4). Galloway (1957, p. 355) listed ten terms which writers have used to describe
the microstructure of these genera and all of them imply that the tissue encloses vacuities.
Nicholson (1886, p. 36) referred to the microstructure as ‘dotted or porous’ but leaves
no doubt elsewhere in his monograph that he understood that the tissue contained sub-
spherical voids. Galloway (1957), and Galloway and St. Jean (1957) substituted for
Nicholson’s word ‘dotted’ the term ‘maculate’, and described the tissue as not porous
or cellular but dotted either with light maculae on a dark field or vice versa. Galloway
wrote (1957, p. 363) ‘The maculae are not pores but dots: e.g. light or dark’, yet (p.
358) he referred to the tissue of Stromatopora as ‘spongy’, which seems to imply pores.
Galloway conceived the maculae as spheres in which dark matter is concentrated at the
periphery so that in thin section they typically have a light centre. He believed that the
expansion of the clear centres gave rise to clear maculae, that is, dark tissue with lighter
dots. None of the specimens of Stromatopora examined by me show distinct annular

G

B 6612

82

PALAEONTOLOGY, VOLUME 9

structures. In uncommon states of recrystallization specks have moved towards the
peripheries of crystals making them light in the centre and dark at the edge but this
is not the microstructure referred to by Galloway. Galloway (1957, p. 363) described the
maculae as being 02-0-3 mm. in diameter but this must be a misprint for 0-02-0-03 mm.
for below he stated that larger ones may be 0 03-0 06 mm. He described, however, struc-
tures as large as 0-13 mm. as maculae (1960, p. 634).

The maculate concept of microstructure seems to be based on examination of the
stromatoporoids from the central United States which commonly show the maculae as
dark dots against a light background. Well-preserved specimens from Europe of
Stromatopora and related genera show spherical vacuities or cellules in darker tissue.
However, spherical clusters of dark specks in lighter tissue do occur in the tissue of
many European specimens. Galloway has used the term maculate to apply to tissue
containing light and dark spheres but these are two different microstructures and each
needs a descriptive term. The dark spheres in lighter tissue are called melanospheres
(Stearn 1965) and the microstructure they define is melanospheric. The tissue with
spherical vacuities (the light maculae of Galloway and St. Jean) is called ‘cellular’ as
suggested by Lecompte (1952, 1956) and others.

That melanospheric and cellular microstructures can be different states of preser-
vation of the same original tissue is indicated by the following:

1. Different specimens of the same gross structure may be cellular or melanospheric.

2. Within the same coenosteum or thin section melanospheric and cellular tissue may
coexist.

3. In some states of preservation a single field under the microscope may be inter-
preted as either melanospheric or cellular. Such fields present an optical illusion.
If the observer concentrates on the light areas, they seem to be holes in the dark
background; if he concentrates on the dark areas, they seem to be dots on a light
field.

How then are the two microstructures related? Examination of the transition from
one to the other in the same specimen shows that the filling of the cellules by organic
matter does not produce melanospheric tissue. The tissue of the best preserved specimens
which approaches the microstructure of the tissue secreted by the organism contains
closely set subspherical vacuities (cellules). In such tissue the largest masses of dark
matter would be between four closely packed cellules and would be connected around
.the cellules by tenuous films of dark matter to adjacent large masses. In the process of
preservation the films of dark matter connecting the larger masses are either diffused as
the cellules enlarge or else their dark matter migrates to the larger intercellular masses
leaving a light tissue with subspherical dark masses of specks, the melanospheres.
Specimens in intermediate stages of this process can be found showing melanospheres
with small extensions towards neighbouring melanospheres (PI. 14, fig. 7). The result-
ing melanospheres do not occupy the positions of the cellules but intercellular positions
(text-fig. 2). The dark matter may be either structureless or resolved into specks.

The proportions of light and dark tissue in cellular and melanospheric microstructures
in a given section will depend on several factors, such as:

1 . The position of the plane of the section with respect to the microstructure. Different
sections of the same random microstructure will give different proportions.

C. W. STEARN: THE MICROSTRUCTURE OF STROM ATOPOROIDS 83

2. The original size of the cellules.

3. The amount of their secondary enlargement.

4. The relationship of the thickness of the section to the size of the cellules or melano-
spheres.

5. The arrangement of the cellules, whether random or regular.

The appearance of tissue that originally had a cellular microstructure may for these
reasons differ widely within sections, between specimens, and between different species.

^4$

'Wi

•& Ms

.wVi':.

-lah-

In

y ij:-'

*.v

r

4

’s

KUv

&/■*:

•ifcS*

Wr- ■

text-fig. 2. Transition from cellular to melanospheric tissue.

Unfortunately, melanospheres develop in tissue that was not cellular and are not an
infallible sign of originally cellular tissue (PI. 16, fig. 8). Melanospheric tissue is not rare
in such genera as Stromatoporella and Stictostroma and has been interpreted as cellular
tissue by some investigators. Even such compact genera as Anostylostroma may show
isolated areas of melanospheric tissue. These subspherical masses of specks in ‘non-
maculate’ genera have been called pseudomaculae by St. Jean (1962) but are regarded
as melanospheres in this paper (PI. 1 5, fig. 3). The only fallible index to cellular tissue is
the appearance of closely spaced spherical voids in well-preserved specimens but stroma-
toporoids whose tissue is evenly and extensively melanospheric are likely to have had
cellular tissue.

Melanospheric tissue seems to develop in compact genera through the concretionary
tendencies of the specks. Flocculent tissue may be an intermediate stage in the transition
from compact to melanospheric tissue. The tendency for the dark matter to gather into
balls, which is evident in the progressive changes taking place in both cellular and com-
pact tissue, has no obvious explanation.

In one type of cellular tissue the cellules are arranged in vertical and horizontal series
and the dark or speckled tissue between the cellules forms a crude three-dimensional
lattice. This microstructure is called microreticulate (PI. 18, fig. 3). Nicholson (1886)
used the term ‘continuously reticulate’ to refer to tissue in which pillars and laminae
blend to form a network but the term microreticulate refers to a feature of much smaller
scale. Microreticulate tissue gives rise to melanospheric tissue and may also give rise
to tissue containing rod-like bodies of dark matter as in Parallelopora. The various
manifestations of this tissue and its relationship to microlaminae are discussed below
under Stromatopora, Syringo stroma, and Parallelopora.

Other Microstructures. Other microstructures which are not as widespread as those dis-
cussed above are considered further under appropriate genera below, as follows:

Transversely porous — Stromatoporella and Stictostroma (PI. 16, fig. 5)

Tubulate — Stromatoporella (PI. 16, fig. 3)

Vacuolate — Trupetostroma (PI. 17, fig. 1)

84

PALAEONTOLOGY, VOLUME 9

Tripartite lamina and ordinicellular tissue — Stromatoporella, Stictostroma, Trupe-
tostroma (PI. 16, fig. 4)

Striated — St achy odes (PI. 18, fig. 8)

Peripherally vesicular — Hermatostroma (PI. 17, fig. 4).

TISSUE REVERSAL

In some specimens the galleries have been filled with dark sediment and the tissue
therefore is lighter than the galleries instead of the galleries being clear and the tissue
dark. Such tissue reversal is not difficult to understand. A more puzzling phenomenon is
the occurrence of certain microstructures in either a ‘white-on-black’ or a ‘ black-on-
white ’ form. Genera that are characterized by tripartite laminae include Stictostroma,
Trupetostroma, Stromatoporella, and Clathrocoilona. The central layer of the laminae is
normally light and the outer layers darker. Well-preserved specimens of Stictostroma
and Stromatoporella suggest that the central light zone is the result of the joining during
preservation of a line of cellules along the axis of the lamina. In all these genera the
central zone may also appear to be darker than the lateral ones. In some specimens of
Stromatoporella kirki (Galloway and St. Jean) different laminae in the same section may
show a dark or light axial fine. In Clathrocoilona restricta (Galloway and St. Jean)
laminae with a light central zone may have this replaced laterally by a dark line which
looks like bituminous matter filling a crack. In these specimens the dark axial line may
be explained as an organic filling of a porous zone but when one group of specimens
consistently has dark axial zones and another group light axial zones, doubts arise as to
whether these were the same structures originally.

In several species of Stromatopora and Syringostroma fine microlaminae are repre-
sented by dark lines in the tissue (PI. 17, fig. 7). In the species Stromatopora mononensis
(Galloway and St. Jean) (PL 14, fig. 5) and Syringostroma bicrenulata (Galloway and
St. Jean) the microlaminae are clear zones traversing a generally dense, speckled and
melanospheric tissue. Are the clear zones the same microstructure as the dark micro-
laminae in a different state of preservation? How were they formed when the tissue was
secreted? In one specimen of Labechia conferta from the Nicholson collection (Nich.
266 a) most of the cysts are black lines which have a zone into which the specks have
migrated on either side. In one group, however, the zones enclose clear tissue as if after
the specks had spread into the galleries the cyst itself had been replaced by clear calcite.
In some specimens that have regular laminae crossing pillars, the laminae between the
pillars are a dark speckled layer but where they cross the pillars they are represented
by a clear zone in the tissue. More work is needed on these phenomena before an
explanation can be offered.

Related to these unsolved problems of tissue reversal are problems of secondary and
primary tissue and the superposition or continuity of pillars. Galloway (1957) distin-
guished primary tissue which forms the darker, finer, core structures of laminae and
pillars from secondary tissue which is lighter in colour, less dense in texture, commonly
porous and coats the primary tissue, tending to fill in the galleries. For example, he
pictures the structure of Stromatopora as made up of primary microlaminae coated with
secondary maculate tissue which fills most of the galleries. Although the distinction
between two types of tissue may be valuable, the assumption that one type of tissue was

C. W. STEARN: THE MICROSTRUCTURE OF STROMATOPOROIDS 85

laid down first and the other followed perhaps at a later stage is unjustified. In addition,
the adjective ‘ secondary ’ carries the suggestion of tissues deposited by inorganic agencies
after the organism was dead. Where a microlamina is covered with lighter tissue it is
referred to as clothing or coating tissue in this paper.

In genera with distinct long pillars and laminae, systematic importance has been placed
on whether the pillars are interrupted by the passage of the laminae and composed of
superposed segments (e.g. Trupetostroma , Gerrono stroma) or whether they are continuous
through the laminae (e.g. Actinostroma). Doubt is thrown on the validity of this distinc-
tion by many thin sections in which pillars seem to be continuous through laminae in
parts and broken into segments by laminae in other parts. Although poor preservation
may obscure the interruption of the pillars by the laminae, variations in preservation
alone cannot explain the differences in the expression of this feature. In certain parts of
coenostea growth of pillars seems to have been interrupted by the deposition of a lamina.
In other parts the pillars seem to have grown continuously, perhaps secreted by cells
elevated above the general level of the sheet of living matter, while periodically a lamina
was deposited below and around them.

DISCUSSION OF GENERA

In the following section the microstructures of the major genera of stromatoporoids
are described and the definitions of the genera reviewed. Holotypes of the type species
of the following genera were examined in the course of this study: Actinodictyon,
Actinostroma, Anostylostroma, Clcithrodictyon, Cystostroma, Dendrostroma, Hammato-
stroma, Hermatostroma, Idiostroma, Labechia, LabechieUa , Lophiostroma, Parallelopora,
Pseudoactinodictyon , Rosenella, Stachyodes, Stictostroma , Stromatoporella, Stylodictyon,
Syringostroma, Taleastroma, Trupetostroma. Types of lesser value of the type species
of the following genera were examined as indicated: Amphipora (topotype), Aulacera
(topotype), Clathrocoilona (hypotype), Cryptophragmus (topotype), Gerronostroma ( para-
type), Stromatopora (topotype). Holotypes of the type species of the following genera
have not been seen by the writer and most of these genera are not included except briefly
in the following discussion: Atelodictyon Lecompte, Cambrostroma Vlasov, Clathro-
stroma Yavorsky, Clavidictyon Sugiyama, Ferestromatopora Yavorsky, Intexodictyon
Yavorsky, Korovine/la Khalfina, Paramphipora Yavorsky, Praeactinostroma Khalfina,
Synthetostroma Lecompte, Vicunostcichyodes Yavorsky, some recently described Russian
genera, and several Labechioids. No attempt is made to review synonymies of genera or
the history of their usage as both have been done by Galloway (1957) and Lecompte
(1951-2). Structures such as dissepiments, astrorhizal canals, and surface features which
are not generic characters have been eliminated from the diagnoses and discussions.
Only a few of the critical species of each genus are mentioned.

The highly idealized three-dimensional drawings of the structure and microstructure
of many of the genera (text-figs. 3, 4, 6-10, 12-15) represent my interpretation of the
patterns shown by the unretouched photographs at a magnification of x50. In genera
with well-defined laminae only a single lamina and interlaminar space is illustrated by
the stripping of adjacent laminae and their associated pillars. Where the continuity of
pillars is important in the structure, these are illustrated on both sides of the lamina.
The front and side surfaces are vertical sections. The top surface is a tangential section

86

PALAEONTOLOGY, VOLUME 9

which intersects the pillars near their tops at the front and converges on the lamina
toward the back, cutting it tangentially at the back of the diagram.

Problems of classification are outside the scope of this paper. The genera have been
arranged in an order which places related genera as close together as possible.

actinostroma Nicholson
Text-fig. 3

Type species'. A. clathratum Nicholson

Coenosteum composed of long continuous pillars which at intervals give off
horizontal rod-like radial processes. Processes join neighbouring pillars forming
net-like laminae. Tissue compact, in a single layer in the pillars, processes, and
laminae.

This genus is narrowly defined and was fully revised by Fliigel (1959). In general the
tissue is evenly speckled throughout both pillars and laminae but the specks of Silurian
species seem to be larger (approx. 1-5 p) than the Devonian ones (less than 1 p). The
specks in the tissue move out into the galleries and tend to become aligned perpendicular
to the edge of the tissue to produce a fibrous microstructure. When the fibrosity is well
developed the pillars may show fibres bending upwards and outwards from the axes
(water jet microstructure). Such a microstructure is shown by the type specimen of
A. clathratum Nicholson in which the fibrosity is outlined not only by the position of
specks but also by fibrous crystals following the same pattern (PL 1 5, fig. 1). The pillars in
this specimen, and not uncommonly in others of the genus, are lighter in colour than the
laminae. Pillars with water jet microstructure are radially fibrous in tangential section.

In parts of the type vertical sections the laminae appear to be continuous; in other
parts they appear as a series of dark granules depending on the direction in which the
network laminae are cut. Specimens of A. bifarium Nicholson and A. verrucosum (Gold-
fuss) also show water jet microstructure. The pillars of A. bifarium , A. clathratum, A.
expansum (Hall and Whitfield), and A. verrucosum show an axial dark zone where specks
are concentrated, or more rarely a light zone. Whether the axial zone is light or dark
does not seem to be significant, for some specimens of A. clathratum from the North-
west Territories of Canada show both patterns. Undoubtedly this axial concentration

C. W. STEARN: THE MICROSTRUCTURE OF STROM ATOPOROIDS 87

of dark matter is related to the water jet microstructure, a relationship that is evident
also in the Mesozoic sphaeractinoids. The tissue of Actinostroma may also be flocculent.
Ripper (1938) suggested that species of one group of Actinostroma (that of A. ste/lu-
latum ) may have porous pillars. One of her new varieties, A. stellulatum distans, does
show pores in the tissue of the pillars. However, these specimens do not belong to the
genus Actinostroma but are here placed in the genus Gerronostroma discussed below.

Radial arms are given off by the pillars to form a network which in tangential section
has been referred to as the ‘hexactinellid network’. This network is the sine qua non of
Actinostroma (and Ate/odictyon ) but unfortunately in many specimens it is destroyed or
modified during the process of fossilization. When the network is poorly preserved the
genus Actinostroma is easily confused with Gerronostroma, but the latter has continuous
laminae pierced by transverse pores and pillars that are commonly interrupted by the
laminae. Several species described as Actinostroma should be assigned to this genus and
several specimens in the Nicholson collection belong in it (see below under Gerrono-
stroma). Actinostroma resembles Ate/odictyon closely and differs only in the continuity
of its pillars through more than one interlaminar space. Actinostroma stellulatum
(Nicholson) is not typical of the genus. The lectotype (Nich. 170) shows pillars with
radial arms in tangential section; however, they do not unite regularly to form a con-
tinous network, but rather become very numerous in the laminae, forming a fibrous
mat. The laminae in vertical section are persistent and appear to be solid. The pillars
are generally confined to a single interlaminar space, but locally they seem to be either
superposed or pass through several laminae. The type specimen has a compact tissue
which is somewhat flocculent and shows a tendency to break down into melanospheres in
the pillars. The species is closely related to Ate/odictyon but its superposed pillars make
its retention in Actinostroma possible.

Fliigel (1958) placed Actinostroma astroites (Rosen) in a separate subgenus of Actino-
stroma called Densastroma. The structure of this species is so unlike that of other species
of Actinostroma and superficially like that of Stromatopora that its separation from the
rest of the species of Actinostroma is desirable. The laminae and pillars of topotypes in
the Nicholson collection are very thin and the tissue is finely speckled like that of many
Silurian stromatoporoids. The grid of pillars and laminae is so fine that the slightest
diffusion obliterates the structure entirely.

Nestor (1964) separated the group of Actinostroma intertextum Nicholson from the
rest of the genus on the basis of the lack of regularity in the giving off of the radial pro-
cesses. The microstructure of the new genus, Plectostroma Nestor, determined on the
basis of the type of A. intertextum (Nich. 188) is compact, with speckling of the tissue
well developed.

atelodictyon Lecompte

Type species: A. fallax Lecompte

Coenosteum composed of well-defined pillars and laminae. Pillars confined to
a single interlaminar space, commonly incomplete and branching, giving off at
the top horizontal processes which join as a network, or in chain-like groups to
form the laminae. Laminae fundamentally net-like, typically discontinuous in
vertical section. Horizontal processes may be thickened so that laminae appear
to be a sheet of tissue with round closely spaced foramina. Tissue compact.

88

PALAEONTOLOGY, VOLUME 9

In proposing this genus Lecompte (1951) included in it a group of Silurian and
Devonian species formerly described as Clathrodictyon and described three new
Devonian species. Both he and Galloway (1957) emphasized its relationship to Clathro-
dictyon, yet in the way the laminae are formed it is fundamentally different from the whole
of the Clathrodictyonidae and similar to the Actinostromatidae. I was unable to see the
type specimens of A.fallax but they are figured and described by Lecompte (1951) and
by Galloway and St. Jean (1957). In the type species the pillars are simple short columns
and commonly superposed but in other species, particularly those from North America
(A. inter calare Galloway and St. Jean, A. stelliferum Stearn, A. ordination Stearn), the
pillars tend to divide and branch, to be oblique and incomplete.

The laminae of A.fallax are apparently a single layer of tissue with bordering darker
zones in places. The laminae in vertical section appear to be solid, but such a structure
could result from the cutting of the net-like laminae shown in tangential section only if
the vertical sections pictured are thick. In A. intercalare , A. stelliferum , and A. ordination
the laminae break up in vertical section into a series of granules — the cut ends of the
radial processes which form the laminar network. In A. intercalare the aligned granules
form a single layer but in A. stelliferum they are coated on either side with lighter
coloured continuous tissue. The pillars of Atelodictyon are compact and, in vertical
section at least, not porous. However, in tangential sections of several species, including
the type species, an open space appears in the centre of each pillar where it splits into
horizontal processes. Apparently the pillar has a slight closed depression where the pro-
cesses diverge.

Galloway (1957) restricted this genus to the Middle Devonian but Lecompte (1951)
included several Silurian specimens in it. The species of Atelodictyon described as pos-
sibly Silurian by Ripper (1933), such as Clathrodictyon chapmani (Ripper) and Clathro-
dictyon regulare distans (Ripper) (probably = Atelodictyon distans ), are of Devonian
age (Ripper 1938). Several species of Clathrodictyon described by Boehnke (1915) from
the Silurian drift of northern Germany (such as C. alternans Boehnke, C. spatiosum

EXPLANATION OF PLATE 15

Fig. 1. Actinostroma clathratum Nicholson (Nich. 141). Vertical section of lectotype, X 100, showing
compact pillars and laminae; water jet fibrosity is present but does not show well on the photo-
graph. Middle Devonian, Gerolstein, Germany.

Fig. 2. Clathrodictyon striatellum (d'Orbigny) (Nich. 246). Vertical section, X 100, showing clear zone
separating the specks of the laminae from those migrating out into the galleries. Middle Silurian,
Dorrington, Shropshire.

Fig. 3. Anostylostroma hamiltonensis Parks (ROM 16536). Vertical section of holotype, X 100, showing
flocculent (approaching melanospheric) pillar and transversely porous lamina. Middle Devonian,
Alpena, Michigan.

Fig. 4. Pseudoactinodictyon vagans (Parks) (ROM 9376). Vertical section of lectotype, X 50, showing
compact pillars and cysts. Middle Devonian, Kelly’s Island, Ohio.

Fig. 5. Hammatostroma albertense Stearn (McGill Univ. 38-12). Vertical section of holotype, x 50,
showing transversely fibrous laminae and pillar tissue. Upper Devonian, Isaac Creek, Alberta.

Fig. 6. Stromatoporella granulata Nicholson (Nich. 329c). Vertical section of holotype, X 100, showing
ordinicellular laminae. Middle Devonian, Arkona, Ontario.

Fig. 7. Stromatoporella granulata Nicholson (Nich. 329a). Tangential section of holotype, X 100,
showing pattern of cellules in a ring pillar and lamina. Middle Devonian, Arkona, Ontario.

Fig. 8. Stromatoporella selwyni Nicholson (Nich. 330). Vertical section of holotype, X 100, showing
microreticulate character of laminae and pillars. Middle Devonian, Port Colbourne, Ontario.

Palaeontology, Vol. 9

PLATE 15

STEARN, Stromatoporoid microstructures

C. W. STEARN: THE MICROSTRUCTURE OF STROM ATOPOROIDS 89

Boehnke, C. dirschkeimiense Boehnke) have laminae which are net-like in cross-section
and seem to belong in Atelodictyon. Their Silurian age is open to doubt. I do not agree
with Lecompte’s suggestion that C. linarssoni Nicholson and C. camicum Charlesworth
should be assigned to Atelodictyon. As C. linarssoni has solid laminae and no suggestion
of a network structure there is little justification for assigning it to Atelodictyon (Lecompte
1951, p. 127). C. camicum is closer to Hammatostroma (see also Fliigel 1956, p. 145
footnote). Better preserved specimens of C. ostiolatum Nicholson than are in the
Nicholson collection from the Silurian of Ontario will probably show that this species
is an early representative of Atelodictyon. The extension of the range of the genus into
the Silurian needs confirmation.

clathrodictyon Nicholson and Murie

Text-fig. 4

Type species: C. vesiculosum Nicholson and Murie
Coenosteum composed of imperforate, continuous laminae, commonly undulant
or crumpled, and short pillars that are confined to an interlaminar space. The
pillars and laminae are a single uniform layer of compact, commonly speckled,
tissue. Where the laminae are bent, the pillars extend from their downward
inflexions and are cylindrical in form.

Although the definition of the genus Clathrodictyon on the basis of its gross structure
is controversial, its microstructure is unequivocal. The type species (C. vesiculosum ) and
all those close to it in structure are composed of tissues with evenly distributed specks
in both laminae and pillars. The tissue shows no sign of porosity or cellular micro-
structure. Fibrosity is not typical but may develop under certain conditions of preserva-
tion. In well-preserved specimens the boundaries of the tissue are sharp, but movement of
the specks out into the galleries, particularly from the pillars, is common (PI. 14, fig. 6).

Some specimens show fine dark lines in the laminae. One of Nicholson’s specimens
(Nich. 449) of C. linarssoni Nicholson, and one of C. regulare Rosen (Nich. 250) show
such a dark concentration of particles in the laminae, but it is apparently produced
during preservation as neither is well preserved. Boehnke (1915) described C. spatiosum
as having central dark lines but from his illustrations this species might better be referred
to Atelodictyon.

90

PALAEONTOLOGY, VOLUME 9

Lecompte (1951) reviewed early work on this genus and recorded the divisions into
groups that each investigator had made among the species without recognizing the
groups taxonomically. The heterogeneity of the genus may have begun with Nicholson,
who first (Nicholson and Murie 1878) defined the genus in terms of undulant laminae
giving rise to vesicules, and later (1886) in terms of distinct pillars producing the laminae
by giving off radial arms like those of Actinostromci.

Lecompte recognized two groups, one with vesicular tissue and one with distinct
pillars and laminae but he hesitated to divide these groups into different genera. How-
ever, Galloway (1957) restricted Clathrodictyon to those species in which the horizontal
elements are not laminae but a series of cysts placed end to end and in which the pillars
have no separate existence but are the downward continuations of the edges of the cysts.
His statement that the tops of the cysts are atypically in straight lines cannot be sup-
ported from the type specimens of C. vesiculosum from Ohio in which the tops of almost
all the convex parts of the laminae are in line and almost uninflected laminae occur
locally. Galloway would place those species in which the pillars and laminae are distinct
in Anostylostroma. However, he restricted this latter genus to the Middle Devonian
species with upwardly branching pillars and made no provision for Silurian species such
as C. regulare Rosen, C. striatellum (d’Orbigny) (PI. 15, fig. 2), and C. linarssoni Nichol-
son, which have well-defined laminae. Nestor (1964) agreed with Galloway’s diagnosis
of the genus but retained the species listed above within it. As these species are transi-
tional into C. vesiculosum they cannot be excluded from the genus and therefore the
generic diagnosis of Clathrodictyon proposed by Galloway cannot be accepted. Neither
is his diagnosis consistent with the dot-like appearance of the pillars in tangential sections
of the type and related species. Such dots cannot be derived from cutting a series of cysts.
Nevertheless, the genus Anostylostroma is a convenient one for Devonian species such
as those that Lecompte (1951) referred to his second group of Clathrodictyon.

The Silurian species described by Nicholson are all (with the possible exception of
C. ostiolatum) regarded here as belonging to the genus Clathrodictyon as defined above.
I cannot agree with Lecompte (1951, p. 167) that C. crassum is a Stromatoporel/a as it
has neither ring pillars, an axial line of cellules, nor cellular tissue. It is a Clathrodictyon
with thick, single-layered tissue in which melanospheres have developed locally. The
Devonian species described by Nicholson can be assigned to Anostylo-
stroma (C. retiforme, C. laxum), to Stromatoporclla (C. cellulosum),
or are indeterminate structures (C. confertum).

Tangential sections of some specimens of Clathrodictyon in which
the laminae bend down to the pillars in a ‘V’, such as C. striatellum
and C. regulare, show a rectilinear network of light lines in the darker
speckled tissue of the laminae. Each polygon so formed represents the
head of a pillar and commonly encloses a central clear area. As the
pillars of such species spread upwards to meet the lamina they become
hollow and encounter adjacent spreading pillars. The interference of
these pillars forms a hexagonal network much like that of a honey-
comb, for at a level just below the top of the lamina they do not
blend but their integrity is maintained by a thin clear line between
them. Each pillar is shaped as illustrated in text-fig. 5.

Nestor (1964) established the genus Ecclimadictyon for those species of Clathrodictyon,

text-fig. 5. Dia-
grammatic sketch
of a pillar of Clath-
rodictyon striatel-
lum showing the
expanded end and
central depression,
X 50 approx.

C. W. STEARN: THE MICROSTRUCTURE OF STRO M ATOPOROIDS 91

typified by C. fastigiatum Nicholson, in which the laminae are bent into chevron-like
folds. He described the microstructure of the type species as being compact but marked
by small dark dots that are conspicuous in tangential section. Nicholson’s specimen
(Nich. 240) from Estonia shows this texture well but the specimens of the same species
from Shropshire show only compact, evenly speckled tissue. The dark dots are here inter-
preted as melanospheres developing out of originally compact tissue.

anostylostroma Parks

Text-fig. 6

Type species: A. hamiltonense Parks

Coenosteum of regular, persistent laminae and short pillars. Laminae composed
of a single layer of compact tissue that in some states of preservation appears to
be transversely fibrous and in some to be transversely porous. Compact pillars
confined to an interlaminar space and divide at the top either simply or into
a network of tissue that spreads on to the underside of the laminae.

branched tops of

The holotype of A. hamiltonense (PI. 15, fig. 3) in the Royal Ontario Museum has
very dark tissue which stands out against the empty galleries lined with calcite crystals.
The specimen shows some silicification. The specks in the tissue have not migrated into
the galleries to any great extent. The laminae are a single layer of tissue which locally
breaks up into a series of granules. At x50 the laminae seem to have fine, closely
spaced canals running transversely through them but are not fibrous. The tissue of the
paratype also lacks fibrosity. Anostylostroma papillatum Parks has tissue which is much
the same but is locally flocculent and locally melanospheric. In tangential section the
laminae seem to be penetrated by fine transverse pores.

The pillars of species of Anostylostroma from the central United States are commonly
darker in colour than the laminae and locally their ends are embedded a short distance
in the lighter laminar tissue (PI. 14, fig. 4). In tangential section they appear as fibrous
dots or vermiform bars that look like hairy caterpillars. Complexly branching pillars
cut tangentially near the top form a network or irregular tangle of fibres. The fibrosity
of the tissue is largely due to the alignment of specks but may also have a crystalline

92

PALAEONTOLOGY, VOLUME 9

basis. In A. meshbergerense (Galloway and St. Jean) the pillars have a dark axis from
which the fibres spread out like those of a Mesozoic sphaeractinoid, but such micro-
structure is rare in Anostylostroma.

Many specimens of Anostylostroma show compact tissue only but the diverse develop-
ment of transversely fibrous and porous microstructures is difficult to interpret. Some
fibrosity seems to be secondary but the possibility that some reflects a primary micro-
structure cannot be eliminated. Transversely porous laminae, such as those of the type
species, are not common in other species of the genus.

Rarely a specimen of Anostylostroma shows an axial light or dark zone in the laminae
but these specimens are atypical and may be transitional to Stictostroma. The upwardly
branching pillars and single-layered laminae distinguish Anostylostroma from Sticto-
stroma.

hammatostroma Stearn
Type species: H. cilbertense Stearn

Coenosteum composed of persistent laminae and incomplete pillars confined to
a single interlaminar space. The laminae are a single layer of transversely fibrous
speckled tissue. The pillars arise from the laminae and form a tangled irregular
mass of strands between the laminae. The tissue of the pillars is speckled, not
cellular.

In the original description of this genus (Stearn 1961) the laminae were described as
transversely porous. Although variations in the density of the specks in the tissue suggest
transverse porosity, the tissue in all the specimens I have seen is fibrous (PI. 15, fig. 5).
This fibrosity is not obvious in the tangled pillar tissue although it seems to be derived
from the laminae. Yavorsky (1963, and personal communication) believes the tissue of
both the laminae and the pillars of Hammatostroma to be porous. He points out that
species assigned by Stearn to Hammatostroma largely on the basis of gross structure
include some with compact tissue (Clathrodictyon katavense Yavorsky, and C. tschusso-
vense Yavorsky) and one with porous tissue ( Stromatoporella undata Yavorsky). He
established the genus Intexodictyon to include largely Silurian but some Devonian species
with compact tissue and irregular, incomplete, and almost cystose pillars.

I have been unable to examine Yavorsky’s material but have seen specimens of
Intexodictyon from the Silurian of Baffin Island, Arctic Canada. In structure these
specimens are similar to Hammatostroma but their microstructure is not fibrous but
compact and like that of Clathrodictyon. Nestor (1964) examined the original material
of C. katavense and C. tschussovense and believed, as I do, that on the basis of micro-
structure they are best placed in Hammatostroma. He also believed that more evidence
is needed before the range of Intexodictyon can be extended into the Devonian. These
relationships suggest that similar gross structures were developed at different times in
the history of the stromatoporoids, Intexodictyon from the compact Clathrodictyon in
Silurian times, and Hammatostroma from the fibrous Anostylostroma in Devonian times.

dendrostroma Lecompte
Type species: D. occulatum (Nicholson)

Coenosteum dendroid, columnar or branching, typically with an axial tabulated

C. W. STEARN: THE MICROSTRUCTURE OF STRO M ATOPOROIDS 93

tube. Other large tubes radiate from the axial one. Laminae thick, clearly differen-
tiated from the pillars forming paraboloids around the axis of the coenosteum.
Pillars confined to an interlaminar space, not superposed. Microstructure of both
pillars and laminae compact, commonly transversely fibrous.

The Nicholson collection contains only one fasciculate specimen of Dendrostroma
occulatum from which 31 sections were made. The other two specimens from Devon
which were referred to this species are not conspecific. The tissue of Nicholson’s type
is speckled and much diffusion of specks into the galleries has taken place. The micro-
structure of the tissue is fibrous but like that of many compact stromatoporoids this
fibrosity may be secondary. In a few places an axial light line appears in the tissue but
this also seems to be secondary. In thicker parts of the sections the tissue appears to have
tubules in it and is distinctly flocculent, not unlike that of StromatoporeUa in some states
of preservation. Lecompte (1951) suggested that the internal structure of this genus is
comparable to that of StromatoporeUa.

Galloway (Galloway and Ehlers 1960) described two species of Dendrostroma from
Michigan. Of these I have seen only D. fibrosum, which as its name implies has a very
fibrous tissue. Minute tubes parallel to the fibres seem to penetrate the tissue. Where
not fibrous the tissue is flocculent.

Ripper (1937u) described some specimens of D. occulatum as being hemispherical
rather than cylindrical. Such a coenosteum would be difficult to distinguish from that
of Stictostroma.

STROMATOPORELLA Nicholson
Text-fig. 7

Type species: S. granulata Nicholson

Coenosteum composed of well-defined laminae and short pillars confined to one
interlaminar space. The laminae typically enclose a single series of subspherical
cellules but one species at least shows more than one series. In some states of
preservation the cellules may be expressed as a continuous axial clear zone in the
lamina, in others as a series of pores crossing the lamina transversely. Some
pillars are spool-shaped, others are formed by the upward inflection of the
laminae into hollow cylinders which are cut as rings in tangential sections (ring
pillars).

Two concepts of the structure of StromatoporeUa can be found in descriptions of the
genus. One concept set forth by Nicholson (1886) and followed by Lecompte (1951)
emphasizes the relationship of the genus to Stromatopora and ignores or denies the
existence of the distinctive ring pillars (but Nicholson in 1892 noted the distinctiveness
of the ring pillars). The other, set forth by Parks (1936) and followed by Galloway and
St. Jean (1957), emphasizes the relation of the genus to Clathrodictyon and recognizes
the ring pillars as its essential feature. Yavorsky ’s views (1963) seem to fall between these
extremes. Since the revision of Parks is based on the type species, it is accepted in this
paper, as the diagnosis above indicates.

Nicholson (1886) described the microstructure of StromatoporeUa as ‘porous or finely
tubulate’. Lecompte (1951) reviewed opinions on the microstructure and concluded that

94

PALAEONTOLOGY, VOLUME 9

it is no different from that of Stromatopora, that is, cellular or filled with small vacuities.
Lecompte believed that the tubules which Nicholson described were not part of the
original tissue but developed during preservation. He also stated (1951, p. 157) that a
line of cellules in the laminae is not typical of the genus nor found in its type species.

However, vertical sections of the holotype of the type species (Nich. 329) show a line
of cellules in the axis of the laminae, i.e. the laminae are ordinicellular (PL 15, figs. 6, 7).
These chambers are not tubular but are equidimensional and appear in both tangential

TEXT-FIG. 7. X 50.

and vertical sections as subcircular clear areas. The tissue in tangential section shows very
fine specks (less than 1 p) which have diffused only slightly into the galleries, and small
round clear cellules throughout. Where a ring pillar is formed by the upbending of a
lamina, the cellules may be streaked out into radially directed short unbranched tubes.
In parts of vertical sections of the tissue the microstructure of the laminae is preserved
as a line of subspherical cellules, but locally the boundaries between the chambers break
down and they merge into a central light zone. The spool-shaped pillars generally have
a clotted or locally melanospheric texture and may contain some obscure cellules which
are more prominent in tangential than in vertical sections.

Several other species in the Nicholson collection show ordinicellular structure.
Stromatoporella tuberculata (Nicholson) shows an axial line of cellules which are
replaced locally by transverse pores. The tissue of S. selwyni (Nicholson) approaches
the microstructure of Stromatopora , for several zones of cellules forming a micro-
reticulate structure are longitudinally aligned in the laminae and pillars (PL 15, fig. 8).
In tangential section the laminae are also microreticulate. A specimen from the Eifel
identified only as Stromatoporella sp. in the Nicholson collection (Nich. 371) also shows
the cellules in the laminae and their blending locally into an axial light line. Several
specimens of Stromatoporella in the St. Jean collection from Ontario which have not yet
been described have the typical microstructure. In S. columbusensis Galloway and St.
Jean the cellules are beautifully preserved, particularly where the galleries have been
filled with sediment. The type specimen of S. columbusensis and others of S. selwyni sug-
gest that the cellules opened into the gallery by a narrow pore on either side.

Ordinicellular microstructure is not preserved in many specimens or species (I have
seen it in S. granulata, S. selwyni, S. solitaria (Nicholson), S. tuberculata, S. columbusen-
sis, and S. cellulosum (Nicholson) [Clathrodictyon cellulosum auct.]). In the process of

C. W. STEARN: THE MICROSTRUCTURE OF STROM ATOPOROIDS 95

preservation the microstructure has been changed by the destruction of the wall between
the cellules so that they merge into a continuous axial light zone, or by the destruction
of the wall between the cellule and the gallery producing transverse pores. Species which
show such an axial light line (rarely dark by tissue reversal) include S. eifeliensis Nichol-
son, S. kirki Galloway and St. Jean, S. eriensis (Parks), and S. percmnulcita Galloway
and St. Jean. The first of these has been assigned to Clathrocoilona and is discussed under
that genus. The conversion of the cellules to transverse pores is particularly common in
American species and can be seen in specimens of S. cryptoannulata Galloway and St.
Jean, S. huronesis Parks, S. paraso/itaria Galloway and St. Jean, and a manuscript
species of St. Jean’s.

Unfortunately many species which in their thick tissue and ring pillars seem to belong
to StromatoporeUa do not show any of these microstructures but may develop two others.
In S. arachnoidea Nicholson the tissue of the pillars and laminae is thick and seems to
be penetrated by fine light tubules about 5 ^ in diameter (PI. 16, fig. 2). In vertical section
they seem to pass transversely through the laminae and in tangential section they wander
irregularly in the tissue. They occur also in S. morelandensis Galloway and St. Jean, in
S. damnoniensis Nicholson (which is here referred to Stictostroma ), and to a lesser
degree in S. granulata. The occurrence of these fine tubules in the tissue of some of these
species (and S. eifeliensis) was the basis for Nicholson’s conclusion that the tissue of the
genus was typically finely tubulate.

Nicholson (1886, pi. 1, fig. 5) illustrated a set of branching tubules in the tissue of
S. granulata. Neither Nicholson’s catalogue nor the notations on the slides indicate
specifically from which of the seven thin sections of this species the illustration was
drawn. Slide 329<? on which the notation ‘pi. 1, fig. 4, 5, 15’ appears is a tangential
section and could not have included fig. 5. Nicholson’s idealized drawing cannot be
matched with any of the vertical thin sections but then his figures are all interpretative
and commonly no part of the thin section exactly matches the drawing, even where the
specific area from which the drawing was made has been indicated on the section. I was
unable in the type slides to find a structure which corresponds to the axial canal with
systematic branches illustrated by Nicholson. Apart from the elongation of the cellules,
already described, where the ring pillars are formed, other species do not show fine
tubular structures.

In most of these species the microstructure can be explained (Lecompte 1951) as
formed during preservation and as associated with the clumping of fine specks into
flocculent tissue. These thin irregular voids in the tissue are here called pseudotubules.
Locally the pseudotubules look like shrinkage cracks in the tissue or appear to be related
to transverse porosity. The pseudotubular microstructure is a variant of flocculent tissue
and where the concentration of the specks has reached a further stage it may lead to
melanospheric tissue. The development of melanospheres is common in species of
StromatoporeUa from the central United States. The occurrence of melanospheres and
cellules in specimens of StromatoporeUa has suggested to some that the tissue is closely
related to that of Stromatopora and that various species of StromatoporeUa evolved into
species of Parallelopora and Syringostroma (Lecompte 1951). I believe that the single
line of cellules opening into the galleries is a different type of microstructure to that
shown by Stromatopora , where cellules are scattered throughout the amalgamate tissue.
The cellular tissue characteristic of Stromatopora , Syringostroma , and Parallelopora was

96

PALAEONTOLOGY, VOLUME 9

widespread by Middle Silurian times but genera such as Stictostroma and Stromato-
porellci which have a single line of cellules are predominantly Middle Devonian genera.

The type species indicates that the essential features of the genus are the ring pillars
and the ordinicellular laminae. I agree with Lecompte (1951) that tubules in the tissue
are not characteristic of this genus and that secondary processes have given the tissue
the appearance of having tubules in some specimens. There remain, however, some speci-
mens in which the tubules are difficult to explain as secondary phenomena. The possi-
bility that they may be the result of parasitic organisms such as infest modern corals and
hydrozoans should be considered. 1 do not agree with Lecompte that the tissue has the
overall cellularity of Stromatopora or that ring pillars are more characteristic of Clathro-
dictyon than of Stromatoporella. Galloway and St. Jean (1957) have also disagreed with
Lecompte on this genus and have reassigned most of the species he described as Stromato-
porella to the genera Stictostroma and Clathrocoilona. Of the species described by
Lecompte Stromatoporella decora , S. gracilis and S. pertabulata have characters recog-
nized as belonging to the genus here. Others have been assigned to Clathrocoilona (S.
saginata, S. obliterate, and S. spissa) and other genera.

stictostroma Parks

Text-fig. 8

Type species: S. mamilliferum Galloway and St. Jean

Coenosteum composed of continuous laminae and pillars confined to an inter-
laminar space. Laminae, where well preserved, enclose an axial line of cellules
but in more common states of preservation have either an axial lighter zone or
are transversely porous. Pillars simple, spool-shaped, composed of compact
tissue and not systematically superposed.

The validity of Stictostroma as a genus and the difficulties arising from Parks’s (1936)
unorthodox procedure of naming two type species have been reviewed by Lecompte

EXPLANATION OF PLATE 16

Fig. 1. Idiostroma caespitosum Winchell (Univ. Michigan 32401a). Vertical section of holotype, X 50,
showing cellular nature of tissue. Middle Devonian, Petoskey, Michigan.

Fig. 2. Stromatoporella arachnoidea Nicholson (Nich. 3856). Vertical section, X 100, showing pseudo-
tubules in flocculent tissue. Middle Devonian, Gerolstein, Germany.

Fig. 3. Clathrocoilona [ Stromatoporella ] eifeliensis Nicholson (Nich. 3426). Vertical section, X 60,
showing tubulate tissue. Middle Devonian, Gerolstein, Germany.

Fig. 4. Stictostroma mamilliferum Galloway and St. Jean (ROM 2151). Vertical section, X 50, showing
ordinicellular laminae, Middle Devonian, Gorrie, Ontario.

Fig. 5. Stictostroma sp. (UNC 282-75). Vertical section, X 100, showing transversely porous laminae.
2\ miles E. of Arkona, Ontario.

Fig. 6. Gerronostroma elegans Yavorsky (UNC 306-44). Vertical section of paratype, X 100, showing
transversely porous laminae. Middle Devonian, Kuznetsk Basin, Russia.

Fig. 7. Trupetostroma warreni Parks (UNC 282-26). Vertical section, X 50, showing suggestion of
division of clear laminae into cellules. Middle Devonian, near Fort Norman, Northwest Territories,
Canada.

Fig. 8. Ano stylo stroma maculosum (Parks) (ROM 13,215a) ( Trupetostroma maculosum of Parks).
Vertical section, X 50, showing melanospheres well developed in originally compact tissue. Upper
Devonian, Charles City, Iowa.

Palaeontology , Vol. 9

PLATE 16

STEARN, Stromatoporoid microstructures

C. W. STEARN: THE MICROSTRUCTURE OF STROMATOPOROIDS 97

(1951) and by Galloway and St. Jean (1957). The genus is a useful one for the reception
of species that have the microstructure of Stromatoporella but do not have ring pillars.
The microstructure of this genus has been the subject of a thorough study by St. Jean
(1962) and therefore needs little further discussion here.

The holotype of S. mamilliferum (= Stromatopora mammillata Nicholson not Schmidt)
was not located in Nicholson’s collection. Topotype material in the University of North
Carolina collection and at the Royal Ontario Museum from south-western Ontario, and
particularly from Ashton’s Quarry at Gorrie, shows the fine structure of the type species

TEXT-FIG. 8. x50.

described by St. Jean (PL 16, fig. 4). The laminae are relatively thin and have an axial
line of small subspherical cellules (ordinicellular). In both vertical and tangential sec-
tion these appear as subcircular clear areas in the speckled tissue. In the pillars the tissue
also seems to be porous but this may be the result of the aggregation of the specks into
melano spheres.

The laminar tissue seems to change in one of two directions during preservation. In
some specimens the boundaries that separate the cellules from the galleries break down
and a lamina with transverse pores passing from gallery to gallery is the result. This
microstructure is also accompanied by the development of fibrosity in the tissue (PI. 16,
fig. 5). In other specimens the boundaries between the cells themselves break down and
a continuous light zone appears along the axis of the lamina giving it a tripartite appear-
ance. In Stictostroma kayi (Parks) and S. elevatum (Parks) some specimens show an
axial dark, rather than light, zone and this has been described by St. Jean (1962) as due
to the alignment of crystal boundaries. The dark zone of S. elevatum is too persistent to
be explained as due to crystal boundaries and must be the filling of an axial void with
bituminous matter (see p. 84).

Where the tissue is thick it may appear flocculent and, as in Stromatoporella, pseudo-
tubules may appear between the concentrations of specks. Unfortunately many speci-
mens with the gross structure of Stictostroma have laminae which show no distinctive
microstructure but are a single layer of slightly fibrous tissue. In this state of preserva-
tion Stictostroma can be distinguished from Anostylostroma by the upwardly dividing
pillars of the latter. Several of Nicholson’s species of Stromatoporella are here referred
to Stictostroma because they lack the ring pillars of the first genus. These include
Stromatoporella socialis Nicholson, and S. damnoniensis Nicholson.

H

B 6012

98

PALAEONTOLOGY, VOLUME 9

Galloway and St. Jean’s (1957) assignment of a group of Lecompte’s (1951) species of
Syringostroma to Stictostroma is not justified. In this group tangential sections show a
network which cannot be reconciled with the structure of the type species of Sticto-
stroma. I agree that these species are not Syringostroma but would provisionally assign
them to such genera as Atelodictyon ( Syringostroma microfibrosum Lecompte), Ferestro-
matopora {S. percaniculatum Lecompte), and possibly Stromatopora (S. minutitextum
Lecompte) pending revision.

clathrocoilona Yavorsky
Type species: C. cibeona Yavorsky

Coenosteum composed of thick laminae and pillars which enclose small rounded
galleries. Laminae laterally persistent, divided by a narrow axial light zone into
three layers. Pillars confined to a single interlaminar space, spool-like, not com-
monly superposed. Galleries small and rounded, not as high as the laminae are
wide. Tissue of laminae and pillars compact.

In his original definition of this species Yavorsky (1931) emphasized the restriction
of the galleries to round openings and the compactness of the tissue. On the basis of
Yavorsky’s illustrations of the type species in tangential section and the examination of
a topotype, Galloway and St. Jean (1957) determined that the genus should be inter-
preted as maculate (cellular of this paper). They described four species from Indiana
including one specimen identified with the type species. I have, through the courtesy of
Dr. St. Jean, examined these and would describe the tissue as speckled, fibrous, and to
a minor extent flocculent, but none of the species is distinctly melanospheric (darkly
maculate) or cellular. The tissue in all is thick and the laminae are marked by a clear
fine. Galloway and St. Jean (1957) assigned thirteen species to Clathrocoilona , four of
which were described by Lecompte (1951) as Stromatoporella. These four species were
described by Lecompte as having traces of cellular tissue but in some (e.g. S. saginata
Lecompte) the tissue is not clear enough to make the determination of microstructure
positive. Of these I have been able to examine only the specimens in the Nicholson collec-
tion which Lecompte referred to his Stromatoporella saginata (Nich. 362) and S. crassi-
texta (Nich. 356). In the first no cellular structure occurs, but in the second, parts of
the tissue are cellular and Lecompte noted that the Belgian specimens show aligned
cellules.

The interpretation of this genus can be made only on the basis of Yavorsky’s original
descriptions, repeated in 1955 (p. 38) and his assertion that a second species described
by him ( Clathrocoilona solidulum Yavorsky 1955) is also compact. As so described, the
genus comprises species much like Stictostroma in which the tissue is so thickened as to
restrict the galleries. A convenient line between these two genera can be drawn where
the thickness of the laminae reaches the height of the galleries. With its spool-shaped
pillars and tripartite laminae Clathrocoilona is out of place amid genera with cellular
tissue and generally amalgamate structure, but is naturally related to genera with com-
pact tissue. Most of Lecompte’s species assigned by Galloway and St. Jean to Clathro-
coilona cannot be accepted in the genus as originally described because they have cellular
tissue; S. crassitexta Lecompte, and S. irregularis Lecompte may be species of Stromato-
pora. Such species as Stromatoporella spissa Lecompte, S. saginata Lecompte, and S.

C. W. STEARN: THE MICROSTRUCTURE OF STROMATOPOROIDS 99

obliterata Lecompte can be assigned to this genus with Yavorsky’s and Galloway and
St. Jean’s specimens of Clathrocoilona.

The position of Stromatoporella eifeliensis Nicholson, which Galloway and St. Jean
(1957) referred to Clathrocoilona , remains enigmatic as it is one of the very few species
which have tortuous, irregular, clear tubules (20-30 p wide) penetrating the thick tissue
(PI. 16, fig. 3). Some of the specimens included by Nicholson in this species are not con-
specific with the type (see also Lecompte 1951, pp. 175-6). The microstructure is not
cellular nor simply compact but the species may be retained in Clathrocoilona until its
peculiar microstructure is found to be more widespread.

actinodictyon Parks
Type species: A. canadense Parks

Coenosteum cylindrical in all known species, composed of irregularly crumpled
laminae and through-going pillars. The laminae resemble dissepiments and are
inflected irregularly up and down so that they are impersistent and abut against
each other as in some species of Clathrodictyon. Pillars thick, irregular in cross-
section. Laminae and pillars of compact tissue.

Although Galloway and St. Jean (1957) described the tissue of the type species as
‘obscurely maculate’, Parks (1909) insisted that the type and genus are characterized by
‘dense’ (i.e. compact) tissue. He correctly pointed out its resemblance to Clathrodictyon
and referred to the horizontal elements as laminae, although later writers (Galloway and
St. Jean 1957; Fliigel 1958) have referred to them as cysts. Examination of the frag-
ments of the type of A. canadense at the Royal Ontario Museum convinced me that
Parks’s interpretation is correct and that the relationship of the genus to Clathrodictyon
is close.

Fliigel (1958) suggested that of Parks’s four species, A. neptuni is the same as A.
canadense, A. keelei Parks is a Clathrodictyon, and A. lowi Parks is a valid species.
Examination of the slides of these species showed that the tissue of all is compact.
A. keelei is more like an Actinostroma than a Clathrodictyon but in any case does not
belong to this genus.

Nestor (1964) independently reached the conclusion that the microstructure of
Actinodictyon is compact and assigned Actinostroma intertextum suevicum Nicholson,
a compact species, to this genus as Actinodictyon suevicum (Nicholson).

PSEUDOACTINODICTYON Fliigel
Text-fig. 9

Type species: P. juxi Fliigel

Coenosteum composed of persistent thin laminae and short rod-like pillars, some
of which are confined to an interlaminar space, others cross several laminae.
High interlaminar spaces crowded with large overlapping dissepiments. Laminae
a single layer of compact tissue with isolated foramina. Pillars compact or
flocculent.

Fliigel (1958) established this genus for the type species, P.juxi, and P. vagans (Parks)
to separate these Devonian species with laminae from Silurian species without laminae

100

PALAEONTOLOGY, VOLUME 9

which Parks included together in Actinodictyon. I examined the type specimens of P.juxi
in the Senckenberg Museum, Frankfurt, and those of P. vagans in the Royal Ontario
Museum, Toronto.

Although Flugel (1958) described the tissue of P. juxi as ‘gefleckt’ (= maculate)
I believe that it is better described as flocculent or compact. The tissue of the laminae
is a single layer of flocculent tissue with widely scattered foramina that may be covered
with a convex plate. The pillars are round in cross-section and appear to be compact.

text-fig. 9. x 50.

Many of them are short and confined to an interlaminar space but some penetrate
several laminae. The shorter ones divide conspicuously at the top, much like those of
Anostylostroma, and may spread along the under surface of the laminae producing a
layer of vacuolate tissue.

Pseudoactinodictyon vagans (Parks) (PI. 15, fig. 4) has a similar structure of widely
spaced laminae and many convex dissepiments. Some of the pillars are long, others
short. Some divide upwards as in P. juxi. The tissue of the laminae, pillars, and dissepi-
ments is all similar and coarsely speckled. Where thick it seems to be composed of dark
granules about 10 /x across. The tissue is not cellular nor markedly melanospheric but
some vacuoles may appear in the pillars.

Galloway and St. Jean (1957) described a specimen of Pseudoactinodictyon vagans
from Indiana and redescribed Parks’s type material. I disagree with their interpretation
of the tissues of the type as maculate (in the sense that it is similar to that of Stroma to-
pora ). Neither the gross structure nor microstructure of this genus justifies its classifica-
tion with the family Stromatoporidae. Flugel (1958) suggested that the specimen
described by Galloway and St. Jean from Indiana as Actinodictyon vagans is a new
species. Its tissue is coarsely speckled and locally the specks are aggregated into dark
masses about 7 p across.

Pseudoactinodictyon norrisi Stearn was described (Stearn 1962) before I had seen the
type species and when I understood the tissue of the genus to be ‘maculate’. It is not
congeneric with P. juxi as its tissue is cellular and laminae poorly defined. It is now
referred to Ferestromatopora. P. atluibaskense Stearn has flocculent tissue and is closer
to the type species.

C. W. STEARN: THE MICROSTRUCTURE OF STROMATOPOROIDS 101

Species described by Lecompte (1951) and Le Maitre (1934, 1949) under Actino-
dictyon are further discussed by Fliigel (1958).

gerronostroma Yavorsky
Type species: G. elegans Yavorsky

Coenosteum composed of well-defined compact pillars and laminae. Laminae
persistent and transversely porous. Pillars spool-shaped, composed of porous
tissue and conspicuously superposed from one interlaminar space to the next.

The genus was proposed by Yavorsky (1931) for the reception of species like Actino-
stromci which do not show the radial arms and ‘hexactinellid network’ of that genus.
Gerronostroma is much more closely allied to Stictostroma and differs essentially only
in the superposition of the pillars. Yavorsky (1931, p. 6) described the microstructure
of the tissue as compact or massive; however, sections of a paratype in the collections
at the University of North Carolina (UNC 306-43-44) show that the laminae are trans-
versely porous in much the same way as some specimens of Stictostroma and Stromato-
porella, and that this porous structure continues into the pillars (Galloway and St. Jean
1957, p. 152; and PL 16, fig. 6). As in Stictostroma these pores do not show in all states
of preservation and in most places even in the paratype the tissue appears to be trans-
versely fibrous owing to the alignment of the specks.

The genus has been used extensively only by Russian palaeontologists. Galloway and St.
Jean (1957) established two new species of this genus and assigned Stictostroma insolitum
Parks to it. G. plectile Galloway and St. Jean is much like Anostylostroma and might
be included in that genus. Both G. excellense and G. insolitum show curious circular
structures in tangential section which are similar to, if not identical with, the ring pillars
in Stromatoporella. Both species have other features that make assignment to Stromato-
porella difficult. The laminae of these species seem flocculent or transversely fibrous and
only in isolated parts of G. cf. insolitum is there a suggestion of transverse porosity.
These species seem therefore to be transitional to Stromatoporella.

In G. excellense an axial light line may occur in the laminae suggesting that the trans-
verse porosity may have originated in a line of cellules in the axis of the laminae as in
Stictostroma. Although no direct evidence of such a structure has yet been described,
the analogy with the structure of Stictostroma and Stromatoporella is suggestive.

Several specimens which Nicholson assigned to Actinostroma are placed here in Gerro-
nostroma. Actinostroma malutinum Nicholson has no suggestion of radial arms. Although
the preservation of the specimens is poor and the possibility that it is a Trupetostroma
cannot be eliminated, the species is here provisionally assigned to Gerronostroma. Two
specimens from Gerolstein (Nich. 458, 460) grouped by Nicholson in a manuscript
species represent different species of Gerronostroma. The specimen from Gerolstein
referred to by Ripper (19376, p. 31, Sedgwick Mus. A 4651) as Actinostroma stellulatum
is a Gerronostroma, and specimen 174 of Nicholson’s collection also referred to this
species belongs to this genus too. A. contortum Ripper and A. stellulatum distans Ripper
are both better placed in Gerronostroma. The former has been assigned by Galloway and
St. Jean (1957) to Anostylostroma and renamed A. buchanense to eliminate homonymity
with Actinostroma contortum Gorsky (Fliigel 1958, p. 183).

Although Gerronostroma is close to Trupetostroma, its laminae do not normally show

102

PALAEONTOLOGY, VOLUME 9

three layers nor the large circular foramina of this genus. Both these features seem to be
due to the extension of the pillar tissue on to the laminae in Trupetostroma.

trupetostroma Parks

Text-fig. 10

Type species: T. warreni Parks

Coenosteum composed of persistent three-layered laminae consisting of an axial
light or dark zone covered irregularly with thicker tissue, and short spool-shaped
pillars. Pillars superposed from one interlaminar space to the next, forming with
the laminae a grid in vertical section. Laminae pierced by large circular foramina
between the pillars or reduced to a network. Tissue of pillars and outer zones of
laminae compact but may contain widely separated vacuities.

text-fig. 10. x50.

The type species has been described by Parks (1936), Galloway and St. Jean (1957),
Galloway (1960), and Stearn (1963). The three-fold nature of the laminae and the super-
position of the spool-shaped pillars are conspicuous elements of its structure.

Slides of the holotype of T. warreni at the Royal Ontario Museum (ROM 13197) show
that the pillars are compact in microstructure and very finely and evenly speckled (PI.
17, figs. 1, 2). Little or no migration of specks has taken place at the edges of the tissue
into the galleries. Vacuoles are inconspicuous in vertical sections but those present seem
to be randomly distributed and are not concentrated on the periphery of the pillars. In
several places they seem to be voids produced by the joining of adjacent pillars at the
bottom and top. Locally they appear in the tissue coating the clear axial zone of the
laminae. They are large (50 p), considerably larger than the cellules of Stromatopora.
The nature of the axial clear zone of the laminae is obscure. Although Parks (1936)
described the zone as having a continuous independent existence across the section, it
disappears completely in the area between the pillars and is defined only where bordered
by the darker tissue of the pillars. The type vertical sections do not show any structure
within this clear zone but Galloway’s (1960) specimen (UNC 282-26, which is not
a topotype as implied in Galloway 1957, p. 464, and Galloway and St. Jean 1957, p.
160) suggests but does not show clearly that the zone is divided into cellules by cross
partitions much as in Stictostroma (PL 16, fig. 7).

C. W. STEARN: THE MICROSTRUCTURE OF STROMATOPOROIDS

103

Tangential sections of the type are difficult to reconcile with vertical sections. Between
the laminae the pillars are round and may contain vacuoles. Towards the laminae they
are arranged in circles and coalesce at their bases into a network which encloses circular
foramina. The breaking of the network into discrete pillars marks the other side of the
lamina. The clear axial zone of the laminae, so conspicuous in vertical sections, cannot
be identified in tangential sections. If such a continuous planar open space existed in the
structure it should be represented by a clear line between the networks formed by
the coalescence of the pillars above and below (see text-fig. 1 1). Around the edges of the
foramina in tangential sections of the type is a series of small round vacuities (PL 17,
fig. 2). These have been interpreted by Galloway and St. Jean (1957) and by Parks (1936)

text-fig. 11. Alternative interpretations of Trupetostroma in vertical and tangential sections. In a and
b the clear zone in the lamina is assumed to be a planar open space. Such an assumption requires a
continuous clear zone in the centre of the lamina in tangential section ( b ). In c and d the clear zone is
assumed to be made up of a line of cellules. The tangential expression of such a microstructure is
represented in d and is closer to reality. AiA, and BJU are corresponding lines in vertical and tangential

sections.

as vacuoles in the tissue that has spread out from the pillars, but they are not con-
spicuous in this position in vertical section. They could be interpreted as cross-sections
of a sheet of cellules that occupies the axial light zone of the laminae.

If the laminae were originally ordinicellular as suggested above, the partitions between
the cellules must have been very fine for they are rarely preserved. In other species of
Trupetostroma the position of the axis of the lamina is commonly infiltrated by bitu-
minous matter and a dark line which traverses the foramina as well as the pillars results.

The type species indicates that the characters fundamental to the recognition of the
genus are as follows: (1) Laminae that are three-layered, vacuolate or cellular and
traversed by large foramina which may reduce them to a network; (2) short, spool-
shaped, superposed pillars; (3) compact, vacuolate tissue.

Parks’s assignment of a diverse group of species to this genus has hindered an under-
standing of its limits. The following species, other than the type, were placed by Parks
in Trupetostroma :

Trupetostroma iowaense Parks. Prominent marginal vacuoles and no laminae, only dissepiments.
Closer to Hermatostroma than to Trupetostroma.

Trupetostroma coralvillense (Thomas). Not like the type species; superposition of pillars poor; possibly
a Stictostroma. Assigned by Galloway and St. Jean (1957) to Stromatopora but has llocculent tissue.

104

PALAEONTOLOGY, VOLUME 9

Trupetostroma planulata (Hall and Whitfield). This species has been variously interpreted as Caunopora
(Hall and Whitfield 1873), Syringostroma (Fenton 1920), and Parallelopora (Galloway and St. Jean
1957). The figure of Parks (1936, pi. 9, fig. 3) is not a good representation of his thin section, which
shows little organized structure. Other representatives of this species examined suggest that it is best
placed in Parallelopora.

Trupetostroma maculosum Parks. Parks ’s specimen is a highly melanospheric form of Anostylostroma
(PI. 16, fig. 8).

Trupetostroma incrustans (Hall and Whitfield). Poorly organized structure, probably Clathrocoilona.
Trupetostroma erraticum (Hall and Whitfield). Poorly organized structure, probably Clathrocoilona.
Trupetostroma solidulum (Hall and Whitfield). Poorly organized structure. Parks's specimen looks
much like Parallelopora planulata.

Of these species none is conspecific with the type species as described here. Parks also
suggested (1936, pp. 54, 55) that Parallelopora dartingtonensis (Carter) might be assigned
to Trupetostroma. This difficult species is not a typical Parallelopora but because all the
specimens come from the deformed rocks of Devon and the microstructure is obscure,
its generic assignment is doubtful. The types described by Carter have not been located
but topotypes are in the Nicholson collection and a specimen from the Pit Park Quarry,
Dartington (Nich. 126, P 5746) is here designated the lectotype. It is the specimen illu-
strated by Nicholson (1891, pi. 24, fig. 15, and pi. 25, fig. 1). It is not a typical Trupeto-
stroma for in tangential section the pillars are not circular at any level of the coenosteum
but anastomose irregularly into an irregular network. The tissue is finely speckled and
has spread into the galleries, precluding the determination of whether it was originally
cellular, vacuolate, or compact. Until better material is available the assignment of this
species will be in doubt. Some specimens labelled P. dartingtonensis in the collection
from Teignmouth (e.g. Nich. 129, 130) are definitely Trupetostroma but they are not
conspecific with the Dartington material. The type specimen of Parallelopora darting-
tonensis fillitexta Nicholson (Nich. 137) is better preserved than P. dartingtonensis but
the tissue has been recrystallized and is now uniformly speckled. The specimen is not
a Trupetostroma but shows close affinities to the genus Pseudoactinodictyon.

The concept of Trupetostroma discussed above is that formulated by Parks and
accepted by Galloway and St. Jean (1957) but Lecompte (1952) interpreted the genus in
another light. From an illustration by Parks (1936, pi. 10, fig. 2) he concluded that the

EXPLANATION OF PLATE 17

Figs. 1, 2. Trupetostroma warreni Parks (ROM 13,197). 1, Vertical section of holotype, X 50, showing
compact tissue with few vacuoles. 2, Tangential section of holotype, x 50, showing compact tissue
of round pillars at left and lamina at right. Devonian, Great Slave Lake, Canada.

Fig. 3. Idiostroma roemeri Nicholson (Nich. 406c). Cross-section of holotype, X 60, showing compact
laminae and pillars with peripheral membranes. Middle Devonian, Hebborn, Germany.

Figs. 4, 5. Hermatostroma schliiteri Nicholson (Nich. 386 h and 386c/). Vertical and tangential sections
of holotype, X 60, showing compact tissue and marginal membranes. Middle Devonian, Hebborn,
Germany.

Fig. 6. Stromatopora concentrica colliculata Nicholson (Nich. 18). Tangential section, X 100, showing
melanospheric tissue. Middle Devonian, Teignmouth, Devon.

Fig. 7. Stromatopora divergens Galloway and St. Jean (UNC 294-30). Vertical section of holotype,
X 100, showing dark microlaminae and melanospheric tissue. Middle Devonian, 5 miles E. of
Logansport, Indiana.

Fig. 8. Hermatostroma beuthi (Bargatzky) (Nich. 66). Tangential section, X 50, showing melano-
spheric tissue bordered by a thin membrane. Middle Devonian, Hebborn, Germany.

Palaeontology, Vol. 9

PLATE 17

STEARN, Stromatoporoid microstructures

C. W. STEARN: THE MICROSTRUCTURE OF STROMATOPOROIDS 105

tissue of Trupetostroma is cellular, like that of Stromatopora , and that the cellules are
concentrated on the margins of the pillars. No other investigator having examined the
type material has described it as cellular, and Parks himself emphasized (1936, pp. 56,
57) that the microstructure of both T. warreni and T. iowaensis was not cellular or spongy.
The species of Trupetostroma described by Lecompte as having uniformly cellular tissue
are better removed to Stromatopora ( T . sublamellatum Lecompte, T. ruedemanni
Lecompte) or to Paralleiopora (T. tenuilamellatum Lecompte). A large group of
Lecompte’s species assigned to Trupetostroma (T. cimacensis, T. porosum, T. maillieuxi,
T. crassum, T. thomasi arduennensis ) show a progressive development of marginal
membranes on the pillars, a microstructure peculiar to H er mat o stroma. The interpreta-
tion of the vacuoles of the type species as the initial stage in the development of such
membranes is not justified. The structure and microstructure of these species is closer
to that of Hermatostroma than to that of Trupetostroma, but some may represent transi-
tional forms between the two genera. Lecompte’s species that are here retained in
Trupetostroma (T. bassleri, T. laceration, T. pingue, T. thomasi ) have features closer to the
type. In these species the pillars are circular in section for only a short distance between
the laminae and merge in the laminae into a network of amalgamate structure.

Other species that have been described as, or referred to Trupetostroma and do not
have compact tissue or pillars of circular cross-section between the laminae should be
assigned to Stromatopora. They include such species as T. macrostylum Le Maitre and
T. schelomense Yavorsky.

Several problems concerning the interpretation of Trupetostroma remain. One of these
is the difficulty of distinguishing vacuolate from cellular microstructure. Although they
are here accepted as different microstructures, this does not imply that one may not
grade into the other or that one did not give rise to the other. The possibility that cellular
structure arose in several lineages of stromatoporoids or even repeatedly should not be
overlooked. Determination of microstructure may be difficult in poorly preserved speci-
mens that do not show the vacuoles, and the separation of such genera as Trupetostroma,
Gerronostroma, and Hermatostroma may present problems.

idiostroma Winchell

Type species: I. caespitosum Winchell

Coenosteum dendroid, columnar, or fasciculate, typically with an axial branching
tabulate canal. Laminae and pillars well defined in the peripheral zone. Laminae
composed of a thin dark axial zone clothed in lighter tissue and arch upward
from periphery in paraboloids. Pillars superposed to continuous, diverge upwards
keeping perpendicular to the laminae, forming with them a network with open
galleries. Microstructure of the tissue is compact but may have vacuoles.

For many years the interpretation of Idiostroma was based on the descriptions by
Nicholson of Idiostroma roemeri Nicholson because the type species, Idiostroma
caespitosum Winchell, had not been adequately described. The laminae and pillars of
I. roemeri are equally developed and composed of compact tissue in which marginal
membranes and vesicles appear on some specimens. Nicholson (1886) described the
microstructure of Idiostroma as coarsely porous but did not note the peripheral
membranes in I. roemeri from Hebborn, Germany. The holotype of this species does

106

PALAEONTOLOGY, VOLUME 9

not show a conspicuous internal porosity, only the marginal vesicles (PL 17, fig. 3).
Nicholson’s views on the coarse porosity may have been owing to his inclusion in the
genus of Parallelopora goldfussi (1886, p. 100), which is coarsely porous, for neither
I. roemeri nor I. occulatum Nicholson (now Dendrostroma) are porous. On the basis
of his examination of I. roemeri , Lecompte (1952) believed that marginal vesicles
on the tissue are characteristic of the genus but because he recognized striations like
those of Stachyodes in the tissue of two species, he preferred not to base the genus on
microstructure but on gross structure. These two species, one of which had been first
described as Stachyodes ( S . verticellata irregularis Heinrich), are both placed back in
Stachyodes by Galloway and Ehlers (1960).

The redescription of Winchell’s types of Idiostroma caespitosum by Galloway and
Ehlers (1960) has opened the way to a fuller understanding of the genus. Through the
courtesy of Dr. E. C. Stumm I was able to examine slide W2-18 (Univ. of Michigan
32401a) cut by Galloway from Winchell’s type (PI. 16, fig. 1). The tissue is not con-
spicuously speckled but seems to be distinguished from the gallery filling largely on the
basis of its finer crystallinity. It is full of subspherical voids. Galloway and Ehlers (1960)
described the tissue as ‘not maculate but it is of variable compactness and is vacuolate,
light, dark and mottled, with abundant dust particles ’ (p. 65). However, in the subsequent
discussion of the similarity of /. caespitosum to Stachyodes gracilis Lecompte and S.
paralleloporoides Lecompte (pp. 65, 102) they imply that the pores are maculae. The
tissue is not like that of Idiostroma roemeri which is centrally compact and contains
peripheral vacuoles. The tissue of I. caespitosum is similar to that of Stromatopora and
could be described as cellular. In places vague suggestions of the striated microstructure
of Stachyodes occurin the tissue, as noted by several writers. Ripper (1937a) and Lecompte
(1952) have recognized that the genus embraces a variety of structures in a dendroid
form which might be placed in other genera. Lecompte separated I. occulatum Nicholson
in the new genus Dendrostroma and noted that in marginal vesicles other species of
Idiostroma resemble Hermatostroma and Trupetostroma. Ripper (1937a) described Idio-
stroma as being massive in some species, but this applies only to her description of
I. occulatum, which is further discussed under Dendrostroma.

If the form of the coenosteum is ignored, most of the species of Idiostroma may be
distributed to other genera, principally Trupetostroma. As the laminae and cellular tissue
of the type species make its assignment to any other genus difficult, the genus is retained.
The following reassignment of species of Idiostroma is suggested :

Idiostroma roemeri Nicholson — Hermatostroma

Idiostroma occulatum Nicholson — Dendrostroma

Idiostroma crassum Lecompte — Stachyodes

Idiostroma fililamellatum Lecompte — Trupetostroma
Idiostroma mclearni Stearn — Trupetostroma.

HERMATOSTROMA Nicholson
Text-fig. 12

Type species: H. schliiteri Nicholson

Coenosteum composed of distinct pillars and laminae forming a grid in vertical
section. Laminae have continuous axial microlamina and a lighter, discontinuous

C. W. STEARN: THE MICROSTRUCTURE OF STROMATOPOROIDS 107

coating of tissue spreading from pillars. Pillars spool-shaped, circular in cross-
section, interrupted by laminae but accurately superposed, composed of compact
tissue in the type species. Both pillars and laminae are covered by peripheral
membranes which are held a short distance away from them by small processes,
or by a series of marginal vesicles or vacuoles along the edges of the galleries.

The type species, H. schliXteri , is represented in the Nicholson collection by a single
specimen whose tissue is intensely but unevenly pigmented. Nicholson (1892, p. 217)
misinterpreted this dark staining as opaque matter injected into axial canals of the
laminae and pillars, but this view was discussed at length and rejected by Lecompte

(1952). The type specimen shows well-defined laminae and pillars, dark in colour and
certainly compact (PI. 17, figs. 4, 5). I found no evidence for the statement by Galloway
and St. Jean (1957, p. 219) that ‘the laminae and pillar tissues are obscurely but coarsely
maculate’. Neither did Nicholson (1892) nor Lecompte (1952). In tangential section the
pillars show no internal vacuities, although well preserved, but distribution of the opaque
matter is irregular in the laminae as shown by Nicholson (1886, pi. 3, fig 1). This is not
typical nor evidence that the tissue is cellular. Near the laminae the processes to the
peripheral membranes seem more numerous and in tangential section the peripheral
space seem to break up into cellules. The laminae are partly formed by the periodic
thickening and coalescing of neighbouring pillars to form a network, but this thickening
tissue is based on a thin microlamina that appears in vertical section to be continuous.
This gross structure is almost identical with that of Trupetostroma. In some species of
Hermatostroma the structural elements are coated by vesicles but in the type species
this marginal zone is not composed of overlapping cysts but is a continuous thin mem-
brane that tangential sections show is held off the pillars by small radial processes (PL
17, fig. 5).

The principal problems in the interpretation of this genus are (1) the inclusion of
cellular and compact species in a single genus, and (2) the relationship of the marginal
membranes to the vacuoles of Trupetostroma.

Cellular and Compact Species. Nicholson described the microstructure of the type species
of Hermatostroma as compact but that of H. episcopale, his other species, as porous.
With the inclusion of the latter he enlarged the limits of the genus to include species with

108

PALAEONTOLOGY, VOLUME 9

porous tissue. The preservation of the type species of H. episcopate is typical of the
stromatoporoids from Devon and neither the marginal membranes nor the micro-
structure of the pillars is satisfactorily preserved. Characteristically tangential sections
of this species include a few opaque dots of doubtful significance. One of the best pre-
served specimens (Nich. 392) shows flocculent tissue with melanospheres. No funda-
mental conclusions on the structure of the genus can be drawn from these specimens.

Galloway and St. Jean’s (1957) statement that the genus is coarsely maculate may have
been influenced by their description of Hermatostroma logansportense Galloway and St.
Jean. Its lack of continuous laminae and well-developed marginal membranes indicates
that this species is not a typical Hermatostroma. In addition the microstructure of the
type specimen (UNC 279-2, 3) did not seem to me to be conspicuously maculate
(melanospheric or cellular) but is so described by the authors. Galloway and Ehlers
(1960) also described a specimen of H. episcopate from Michigan but this does not show
the marginal membranes and the tissue is flocculent rather than melanospheric or
cellular.

Although these species do not demonstrate the association of both cellular and
compact tissues with the gross structure of Hermatostroma, other evidence indicates
that this association exists. The type specimens of Hermatostroma beuthii (Bargatzky)
(Nich. 63; Bargatzky specimens were also examined) have the gross structure of
Hermatostroma, but none of the slides shows cellular tissue or the marginal membranes
distinctly. However, a topotype specimen (Nich. 66) which has the structure of the
type shows both well (PI. 17, fig. 8). Other specimens that Nicholson assigned to
this species are a diverse group, some of which belong to Stroma topora (Lecompte
1952). Of the species of Hermatostroma described by Lecompte some are cellular
throughout and some have only marginal cellules. H. perseptatum has well-developed
peripheral membranes and internally apparently compact tissue. H. pustulosum has
a line of cellules on the edges of the pillars also. H. parksi has cellules largely at the edges
of the pillars but Lecompte (1952) states that the whole tissue is in places cellular. Some
of the species transferred here from Trupetostroma to Hermatostroma also show cellu-
lar microstructure throughout the tissue (e.g. T. porosum). If peripheral membranes,
marginal vesicles and vacuoles are taken as generic characters for Hermatostroma then
the genus contains two groups of species, one with cellular tissue and the other without.

Marginal Vesicles and Membranes. The feature of the type species that immediately
separates it from nearly all other stromatoporoids is the marginal membranes on the
pillars and laminae. Other species of stromatoporoids have marginal cellules or cysts
which may be analogous or ancestral structures. Lecompte (1952) and Galloway and
St. Jean (1957) assigned species with marginal cellules and cysts to Trupetostroma. But
as pointed out in the discussion of that genus, the vacuities in the tissue of the type
species are not on the margins of the pillars. Lecompte believed that Trupetostroma gave
rise to Hermatostroma by the enlargement of the vacuoles on the margins of the pillars
and their coalescence into vesicles and membranes. In the absence of such structures in
typical Trupetostroma, those species with cellules, vesicules, vacuoles, or membranes on
the margins of the pillars are here considered to belong to Hermatostroma if they have
spool-shaped superposed pillars and three-layered laminae. Transitional forms between
Trupetostroma and Hermatostroma are probable, but the boundary between the two

C. W. STEARN: THE MICROSTRUCTURE OF STROM ATOPOROIDS 109

genera should not be drawn where the laminae thicken to equivalence with the pillars, as
recommended by Lecompte (1952), but between those species with marginal vacuities
of any sort in the pillars and those without.

Conclusion. The genus Hermatostroma as presently constituted is polyphyletic and con-
sists of a group of species with cellular tissue throughout (e.g. H. beuthii, H. episcopale,
H. parksi, H. porosum, H. maillieuxi) and a group with marginal vesicles, vacuoles and
membranes but compact tissue (e.g. H. schliiteri, H. perseptatum, H. crassum, H.
pustulosum). The name Hermatostroma must be retained for the second group and a new
name should be proposed for the first group with a type species chosen from among
the new species in the list above.

AMPHIPORA Schultz
Type species: A. ramosa (Phillips)

Coenosteum columnar, rarely branching, with or without an axial canal, consist-
ing of an amalgamate network of compact tissue which commonly is fibrous.
Laminae cannot be distinguished and pillars diverging upwards and outwards
from the axis are obscure and irregular. Some coenostea have marginal cysts of
the same tissue.

The type specimen of Amphipora ramoas has not been located. Although some speci-
mens of Phillips’s collection were found at Oxford, my inquiries did not uncover a speci-
men comparable to his illustration of this species. A specimen (No. 52873u) was found
in the Geological Survey Museum, London, labelled ‘ Caunopora ramosa Phillips,
possibly syntype’. Although it cannot be positively identified with the type, since it has
been cut on nearly all sides and polished, it is of similar size. The specimen supposed to
be the other syntype is a tabulate coral. A well-preserved specimen from Devon, possibly
from the Nicholson collection, must be established as a neotype for this widely distri-
buted species when the genus receives revision. As nearly all specimens from Devon are
recrystallized, the type specimen is unlikely to throw any light on the microstructure of
the genus.

Most discussion of the genus has been centred around the extreme variability of its
gross structure, reviewed by Lecompte (1952). Its micro structure is generally agreed to
be fibrous with a dark line marking the axis of each structural element. In many of
Nicholson’s specimens from Germany, the central dark line is poorly defined and the
microstructure is more like a water jet.

Yavorsky (1955) established the genus Paramphipora for stromatoporoids like
Amphipora which do not have a dark axial line in the tissue. Other palaeontologists have
been cautious about accepting this new genus, and the obscurity of the microstructure
in many otherwise well-preserved specimens of the genus prompts me to withhold judge-
ment until the Russian material has been examined.

Lecompte (1952) believed that Amphipora is unique among the stromatoporoids both
in microstructure and in gross structure. However, the gross similarity of Amphipora
to many species of Stacliyodes is striking and its microstructure closely resembles that of
Dendrostroma and even that of Anostylostroma in certain states of preservation. These
similarities leave little doubt that it is a stromatoporoid and suggest that it should be
classified with these genera rather than in a separate family, or with the Idiostromatidae,
to which it shows little relationship except in external form.

110

PALAEONTOLOGY, VOLUME 9

STROMATOPORA GoldfuSS
Text-fig. 13

Type species: S. concentrica Goldfuss

Coenosteum composed of an amalgamate network of tissue in which neither
laminae nor pillars are easily distinguished. Vertical elements (pillars) domi-
nate the structure and are separated by high galleries called pseudozooidal tubes.
Microlaminae, composed of a line of dark tissue, or more rarely a clear path in
the tissue, are present in some species. Tangential sections show an irregular
continuous network of tissue with small galleries and few areas of isolated tissue.
The tissue is cellular or microreticulate.

text-fig. 13. x50.

Of the stromatoporoids, the genus Stromatopora has been by far the most difficult
to restrict and the time for its dissection is overdue. Over two hundred species have been
described and they are of such diverse form that no definition can cover them all.

The holotype of the type species, S. concentrica, is poorly preserved and has been
differently described by Nicholson (1886, pp. 2, 3) and Lecompte (1952, p. 271).
Nicholson interpreted the type, on the basis of topotypes (which he said were identical
in all ways with it), as having a dominantly vertical structure with distinct pseudozooidal
tubes. Lecompte stated that the type nowhere showed the dominantly vertical structures
illustrated by Nicholson (pi. 1 1, fig. 18) and interpreted it on the basis of specimens from
the Ardennes showing few or no pseudozooidal tubes. Galloway and St. Jean (1957)
rejected Lecompte’s interpretation and stated that his specimens from Belgium do not
belong to Stromatopora at all but are Ferestromatopora tyrganensis Yavorsky. Because
Nicholson’s interpretation is based on topotype material and has been accepted as
definitive for many years, it is here adopted.

The microstructure of the specimens (Nich. 1-5) from Gerolstein, Germany, on which
Nicholson based his interpretation is flocculent with local suggestion of cellules. The
tissue is not prominently speckled but, like much of the material from the Eifel, has an
overall irresolvable brown stain. Little movement of this dark matter into the galleries has
taken place. In the flocculent tissue clear areas appear commonly but the cellular nature
is not preserved evenly throughout. The cellular microstructure is much better preserved

C. W. STEARN: THE MICROSTRUCTURE OF STROM ATOPOROIDS 111

in those specimens from Gerolstein that Nicholson described as S. concentrica var.
colliculata (a Devon specimen is illustrated in PI. 17, fig. 6).

The cellular microstructure characteristic of Stromatopora and its change to melano-
spheric tissue is more fully discussed in the first part of this paper.

Microlaminae. Continuous laminar structures of three types are present in species of
Stromatopora.

1. A thin line of granules making a more or less continuous dark plate (S. divergens
Galloway and St. Jean, S. dubia Lecompte) (PL 17, fig. 7).

2. A clear zone running through flecked or melanospheric tissue (S. mononensis
Galloway and St. Jean) (PI. 14, fig. 5).

3. A series of darker and lighter zones caused by the regular arrangement of cellules
and melanospheres (e.g. S. concentrica colliculata Nicholson, S. eumaculosa
Galloway and St. Jean, S. laminosa Lecompte, S. submixta Galloway and St. Jean)
(PI. 18, fig. 3).

Galloway’s (1957) definition of this genus emphasized the microlaminae and he believed
that when they are not present [and they are not in such species as S. luipscliii (Bargatzky),
S. buclieliensis (Bargatzky), S. concentrica (Goldfuss) (sensu Nicholson)] they have been
destroyed during preservation or resorbed during the life of the colony. Amalgamate
tissue with marked lamination is characteristic of Ferestromatopora but the species
which Galloway (1957) removed from Stromatopora to this genus ( S . marpleae Galloway
and St. Jean, S. obscura Galloway and St. Jean, S. larocquei Galloway and St. Jean, and
S', dubia Lecompte) do not have the large galleries typical of Ferestromatopora. As the
cellular nature of the tissue of these species is obscure, they are better considered to
belong to C/athrocoilona.

Even if these species are eliminated from Stromatopora, species showing continuous
microlaminae remain. Where microreticulation is well developed, as in such species
as S. laminosa Lecompte, the dark matter between the horizontal rows of cellules may
form a distinct microlamina. This microstructure is shown well by Silurian species of
Stromatopora, such as S. typica Rosen. Microreticulate tissue might be the basis for
the separation of a group of species from the main body of Stromatopora species. In
some species with microreticulate tissue the microlaminae seem to be independent of
this microstructure.

The characters which separate Stromatopora from the related genera Syringostroma,
Parallelopora, Taleastroma, etc. are discussed under these genera.

ferestromatopora Yavorsky

Type species: F. krupennikovi Yavorsky

Coenosteum of amalgamate structure, horizontal structural elements pre-
dominating over vertical ones. Fibres of the tissue anastomose forming a fine
continuous network enclosing irregular galleries that are not superposed to form
pseudozooidal tubes. Microlaminae may be present. Microstructure is cellular.

Yavorsky (1955) compared the gross structure of Ferestromatopora with that of
Clathrodictyon confertum Nicholson (better referred to Stromatoporella on the basis of
the microstructure of the laminae) and in its thin, arched laminae and lenticular galleries
this species is much like F. krupennikovi. In typical species of Ferestromatopora the

112

PALAEONTOLOGY, VOLUME 9

galleries are large, the tissue fibres are delicate, and form a fine three-dimensional mesh
(e.g. F. krupennikovi Yavorsky, F. tyrganensis Yavorsky, F. jacquesensis Galloway, F.
contexta Stearn). In tangential section the amalgamate structure forms a continuous
network and no isolated pillars can be distinguished. The microstructure of typical
species of Ferestromatopora is cellular, much like that of Stromatopora. It shows a marked
tendency to become melanospheric during preservation.

Although typical specimens of Ferestromatopora are easily distinguished from typical
ones of Stromatopora on the basis of the absence of pseudozooidal tubes and the
dominance of horizontal elements, such species as S. divergens Galloway and St. Jean,
S. laminosa Lecompte, and S. eumaculosa Galloway and St. Jean, whose microlaminae
seem to be derived from microreticulate tissue, could be assigned to either genus. Such
species can be assigned to Stromatopora if the galleries are superposed forming pseudo-
zooidal tubes, and to Ferestromatopora if the tubes are missing. The splitting of this
group with microreticulate tissue away from both genera seems to offer a solution to this
problem that expresses the biological relationships more satisfactorily.

taleastroma Galloway
Type species: T. cummingsi (Galloway and St. Jean)

Coenosteum of amalgamate structure similar to that of Stromatopora with domi-
nant vertical structural elements, pseudozooidal tubes, and dissepiments;
traversed by short, slender, rod-like pillars circular in cross-section. Micro-
structure of amalgamate tissue and pillars is cellular.

Galloway (1957) established this genus for the reception of three species that he and
St. Jean (1957) had described as Stromatopora and two species described by Lecompte
(1952) as Stromatopora. These species were separated from Stromatopora on the basis
of their long pillars, which are compact axially and have borders of ‘dark maculae’. In
the type species, T. cummingsi, the melanospheres (‘dark maculae’) are well defined,
locally closely set, locally widely separated and concentrated on the margins of the
pillars (PI. 18, fig. 5). In some places half a melanosphere at the edge of a pillar appears
to be cut off by the boundary with the gallery. In tangential section the disposition of the
melanospheres is not as obvious as in vertical section.

EXPLANATION OF PLATE 18

Figs. 1, 2. Stromatopora carteri Nicholson (Nich. 31 d, 37). Vertical and tangential sections, XlOO,
showing cellular tissue of the pillars. Middle Silurian, Benthall, Shropshire.

Fig. 3. Stromatopora typica Nicholson (Nich. 59 b). Vertical section, X 100, showing microreticulation.
Middle Silurian, Oesel Island.

Fig. 4. Stromatopora typica Nicholson (Nich. 49). Tangential section, X 100, showing melanospheric
microstructure. Middle Silurian, Ironbridge, Shropshire.

Fig. 5. Taleastroma cummingsi Galloway and St. Jean (UNC 304-47). Vertical section of holotype,
X 100, showing pillars with peripheral melanospheres. Middle Devonian, 5 miles E. of Logansport,
Indiana.

Fig. 6. Syringostroma densum Nicholson (Nich. 311g). Tangential section of holotype, X 50, showing
cellular microstructure. Middle Devonian, Kelly’s Island, Ohio.

Fig. 7. Syringostroma densum Nicholson (UNC 306-8). Vertical section of topotype, X 100, showing
open cellular microstructure. Middle Devonian, Kelly’s Island, Ohio.

Fig. 8. Stachyodes verticellata (McCoy) (Nich. 394). Peripheral part of cross-section, X 50, showing
striated tissue. Devonian, Shaldon, Devon.

Palaeontology , Vol. 9

PLATE 18

STEARN, Stromatoporoid microstructures

C. W. STEARN: THE MICROSTRUCTURE OF STROMATOPOROIDS 113

In specimens of Taleastroma pachytextum (Lecompte) in the Galloway and St. Jean
collection, generally the tissue is flocculent but the edges of the pillars contain melano-
spheres. The holotype of T. pachytextum was not examined but the Nicholson collection
contains an interesting series of specimens of this genus identified by Nicholson as
Stromatopora beuthii Bargatzky, some of which (Nich. 69, 74) were compared by
Lecompte (1952) to Stromatopora pachytexta. Lecompte (1952) showed that Bargatzky’s
type of S. beuthii is a Hermatostroma but such specimens as Nich. 69, 71, 74, and 81, all
from south Devon, show close affinities to Taleastroma. Some of these specimens are
similar in microstructure to the American ones showing pillars with peripheral melano-
spheres and compact axes. Some pillars are marked with an axial dark line and water jet
microstructure. Detailed description of these specimens and assignment of a name to
them will have to await a complete revision of the stromatoporoid fauna of Devon.

The other two American species, T. conicomamillatum (Galloway and St. Jean) and T.
magnimamil/atum (Galloway and St. Jean), do not show the peripherally melanospheric
microstructure as clearly as the first two.

In founding this genus Galloway emphasized the peculiar microstructure of the
pillars, compact inside and ‘maculate’ outside. The genus, however, also differs from
Stromatopora in the possession of rod-like pillars distinct from the amalgamate network.
In this paper melanospheric tissue is regarded as a secondary microstructure produced
during preservation and of doubtful generic significance. This view allows several inter-
pretations of the microstructure of Taleastroma.

1 . The pillars were originally completely cellular, the cellular tissue has changed to
melanospheric, then the melanospheres were destroyed in the axes of the pillars.

2. The pillars were originally compact, and melanospheric alteration has affected
only their peripheries.

3. The pillars were marginally vacuolate or cellular and this zone has become melano-
spheric.

If the third interpretation is accepted, Taleastroma would have to be placed in synonymy
with Hermatostroma. The lack of marginal vacuoles or cellules in parts of Taleastroma
specimens where marginal melanospheric microstructure is not developed does not sup-
port this interpretation. The combination of compact pillars with cellular amalgamate
tissue demanded by the second interpretation is not found in other genera of stromato-
poroids, for the two types of tissue seem to be mutually exclusive. Therefore, although
it is not the simplest hypothesis, the first interpretation, that the peripherally melano-
spheric tissue has been derived secondarily from uniformly cellular pillars, is preferred.

As defined above the genus Taleastroma is a valuable one. It differs from Syringo-
stroma in the lack of large, lacy pillars penetrating the structure, from Parallelopora in
the lack of alignment of its smaller cellules, and from Ferestromatopora in its through-
going pillars.

syringostroma Nicholson

Text-fig. 14

Type species: S. densum Nicholson

Coenosteum an amalgamate structure, like that of Stromatopora, within which
thick persistent pillars and laminae can be distinguished. Laminae composed of

i

B 6612

114

PALAEONTOLOGY, VOLUME 9

thin dark or light microlaminae. Pillars thick, long, rod-like, traversing many
laminae. Microstructure cellular or microreticulate.

Nicholson’s choice of specimen (Nich. 311) as the type of Syringostroma densum was
unfortunate as it is poorly preserved and its vague structure is open to different inter-
pretations. The microstructure of the tissue is densely speckled and somewhat recrystal-
lized. The tissue is faintly fibrous generally. One tangential section (Nich. 311g) shows
coarsely cellular pillars (PI. 18, fig. 6). In other sections the tissue seems at low power
( x 10) to contain melanospheres, but the melanospheric microstructure is not clear at

higher power. Thick, persistent pillars are crossed by fine, dark microlaminae (the struc-
ture is not well represented in Nicholson's illustration, 1886, pi. 11, fig. 14). Between the
microlaminae the tissue is irregularly distributed, forming short vertical structures like
pillars or coating the microlaminae to form thick laminae. Penetrating this whole struc-
ture are large, rod-like pillars with vague water jet microstructure. The microlaminae
do not bend upwards into the pillars and may or may not cross them. In tangential
section the pillars are circular darker areas in the irregular network of the amalgamate
tissue.

Galloway and St. Jean (1957) have described a topotype (UNC 306-8 and 9 from
Kelley’s Island, Lake Erie), which is much better preserved than the holotype (PI. 18,
fig. 7). The amalgamate tissue is cellular and the speckled tissue between the cellules is
reduced to lace-like form by the abundant cellules. The cellular tissue forms thick, short,
spool-shaped pillars and locally is microreticulate. Some microlaminae are formed
between two rows of cellules in this microreticulate tissue. The large pillars which
traverse the coenosteum show neither cellular nor water jet microstructure.

Galloway (1957) believed that the distinguishing features of the genus Syringostroma
are its microlaminae, which bend up into the pillars, and its long, thick columnar pillars.
Several specimens at the University of North Carolina and at the British Museum
(Natural History) show a vague water jet microstructure in the pillars that Galloway
(1957, see pi. 31, fig. 15) apparently attributed to aligned melanospheres. As pointed out
in the first part of this paper, the water jet microstructure is not confined to this genus
but seems to develop under certain conditions in many genera that have columnar pillars.

pseudozooidal

tube

TEXT-FIG. 14. X 50.

C. W. STEARN: THE MICROSTRUCTURE OF STROMATOPOROIDS 115

In the examination of numerous sections of many species of this genus I found no evi-
dence that the water jet microstructure was caused by the alignment of cellules or
melanospheres. Ripper (1937a) and Lecompte (1952) noted the vague vertical structures
in Syringostroma and Lecompte believed that this fibrous striation might indicate that
the genus was related to Parcillelopora.

Two other species described by Nicholson reflect his interpretation of the genus.
Syringostroma ristigoucliense (Spencer) from the Silurian of New Brunswick is only
superficially similar to S. densum. It has large columnar pillars, but they are not asso-
ciated with an amalgamate tissue between them nor with microlaminae. Laminae are
formed by periodic processes extending from the pillars. The microstructure of the
pillars is flocculent in the type specimens and the alignment of the specks suggests an
outward spraying fibrosity. The gross structure of this species is much closer to that of
Parallelopora than to that of Syringostroma and its assignment to that genus seems more
appropriate. Syringostroma nodulatum (Nicholson) was originally described as Stroma-
topora (Nicholson 1891) but has been assigned by Galloway and St. Jean (1957) to
Parallelopora. Cellular tissue is well preserved in Nicholson’s specimens and a fibrous
water jet structure also occurs. This species is closer to Syringostroma than to Parallelo-
pora and does not show the coarsely cellular microstructure of the latter genus.

The views of Parks (1909) and Ripper (1937a, 1938) concerning the transition between
Syringostroma and Stromatopora were reviewed by Lecompte (1952). Lecompte adopted
a very different view of Syringostroma and believed that the essential features of the genus
are the bifid (tripartite) nature of the laminae and the superposition of short pillars from
one interlaminar space to another. He described the tissue as cellular (‘alveolaire’) but
does not believe the genus is closely related to Stromatopora. According to Lecompte it
is more closely related to Trupetostroma, to which it gives rise in the Givetian. This inter-
pretation of Syringostroma is not supported by the type species nor has it been used by
any other palaeontologist. Because Lecompte ’s concept of the genus differs so greatly
from theirs, Galloway and St. Jean (1957) have reassigned many of Lecompte’s species
to other genera, largely to Stictostroma. However, many of the species they reassigned
are unlike Stictostroma in tangential section for they show a continuous network of
tissue rather than pillars cut individually into circular or vermiform shapes. A revision
of this group of species and similar laminate species now included in Stromatopora is
badly needed.

In the species described by Galloway and St. Jean (1957) the specks in the tissue are
so large that they obscure the other microstructures. A few have cellular tissue but many
show only a vague fibrosity which seems to be formed by the alignment of the specks. In
such species as S. sanduskyense Galloway and St. Jean the reticulate arrangement of
melanospheres may form microlaminae. In S. superdensum Galloway and St. Jean the
fibrosity seems to be most distinct where the melanospheric nature of the tissue is least
well defined, but the fibrosity cannot be traced to the alignment of cellules or melano-
spheres. Some of the microlaminae in these species seem to be formed by the alignment
of melanospheres from originally microreticulate tissue, others seem to be plates of dark
specks independent of the microreticulation. In S. fuscum Galloway and St. Jean clear
microlaminae present the same problems of interpretation as those in some Stromatopora
species, such as Stromatopora mononensis Galloway and St. Jean.

Syringostroma can be distinguished from Stromatopora on the basis of the long

116

PALAEONTOLOGY, VOLUME 9

columnar pillars which penetrate the amalgamate network. It differs from Parallelopora
in the smaller size of the cellules in the tissue which do not reduce the pillars to a lattice.
The pillars of Taleastroma are smaller and shorter but otherwise the genus is very similar
to Syringostroma. The microlaminae characteristic of Syringostroma are found in several
other related genera but the two sizes of vertical elements, one in the amalgamate net-
work and the other as through-going pillars, is unique to Syringostroma.

stylodictyon Nicholson and Murie

Type species: S. columnare (Nicholson and Murie)

Coenosteum composed of fine continuous microlaminae locally grouped into
thicker laminae, regularly bent upward into and across prominent persistent
mamelon columns. The columns are solid tissue without internal structure and
form papillae on the upper surface of the coenosteum. The pillars are short, in-
complete, and expand upwards. The fine laminae are locally coated with diffuse
tissue which partly fills the galleries. The microstructure of the tissue is not well
enough preserved for determination.

Nicholson (1886, pp. 79, 80) was uncertain of the position of this genus in the Stroma-
toporidae or the Actinostromatidae. This uncertainty and the subsequent hesitation of
investigators to use the genus can be traced to the poor preservation of the type speci-
men (P6031, Nich. 340 in the British Museum (Natural History) labelled as the type in
Nicholson’s handwriting, with tangential and vertical sections). The laminae are of
variable thickness and are composed of several microlaminae about 15 p thick. These
may join with more diffuse tissue to make laminar elements up to ten times this thickness.
The pillars are composed of widely scattered specks. They are confined to an inter-
laminar space and spread out against the lamina above. The mamelon columns are of
speckled tissue of various densities showing water jet fibrosity. Some of the concentra-
tions of specks might be called melanospheres. None of the tissue is obviously cellular
or melanospheric.

In its short pillars combined with large through-going columns, in its laminae com-
posed of microlaminae, and in its regularly upturned lamina, this species is a Syringo-
stroma. Had the cellular nature of the tissue been preserved, no question of its generic
nature would have arisen. It is similar to Syringostroma tuberosum Galloway and St. Jean
but differs in the dimensions of internal structures and in external morphology of the
coenosteum. Unfortunately it cannot be matched with any species described by Gallo-
way and St. Jean (1957) from Ohio. Stylodictyon is a junior synonym of Syringostroma
and is not a valid genus.

stachyodes Bargatzky
Type species: S. ramosa Bargatzky

Nicholson (1886) placed S. ramosa Bargatzky in synonymy with Stromatopora ( Caunopora )
verticellata McCoy. Since then the type species has been considered to be S. verticellata, but
E. H. Fliigel is investigating the problem of whether these two species are identical.

Coenosteum dendroid, fasciculate, or rarely massive. Each column typically has
an axial canal or canals from which lesser canals branch towards the periphery.

C. W. STEARN: THE MICROSTRUCTURE OF STROMATOPOROIDS 117

Tissue is thick, filling most of the coenosteum apart from the canals, striated,
that is traversed by dark, fine rod-like concentrations of specks or tubes which
spray outwards and upwards from the axis of the column. Galleries small or
absent. Laminae thin lines of dark granules, asymptotic to the periphery,
restricted to the peripheral area, poorly developed, forming invaginating para-
boloids about the axis of the columns. Thick pillars recognizable only near the
periphery, oriented at right angles to the laminae, forming a continuous net-
work in tangential sections.

Nicholson (1886) considered the striated tissue of St achy odes to be diagnostic of the
genus but other writers have emphasized the indistinctness of the pillars and laminae of
Stachyodes as separating it from other dendroid stromatoporoids. Bargatzky’s type of
S. ramosa in the Bonn University collections is accompanied by two slides which are
small but show the striations in the tissue spraying outward from the axis of the column.
Laminae are absent in these two inadequate sections but growth stages are indicated
by variations in the density of the tissue.

Nicholson believed that the striations in the tissue are tubules that have been filled with
opaque matter, but in his specimens labelled S. verticellata (McCoy) from Shaldon
(Devon) they are rod-like concentrations (15 /x in diameter) of dark specks (1 /x across)
(PI. 18, fig. 8). In tangential section they are round dots of dark tissue on a light back-
ground. Locally the dots are arranged in a circle or form a rough network. Nicholson’s
collection also contains a specimen (Nich. 397) from Hebborn (West Germany) in
which the tissue is not speckled but a dark brown colour. In this specimen striated micro-
structure seems to be caused by light-coloured rods or tubules 30-50 y across rather
than by dark rods in a lighter tissue. Tissue reversal is thus found in Stachyodes as in
many other genera. The better preservation of the German specimens suggests that the
tubules represent the original state of the tissue and that the rods may be analogous to
melanospheres. Lecompte (1952) believed that the fundamental microstructure of
Stachyodes, like that of many other stromatoporoids, is microreticulate rather than
striated. He described specimens showing transverse lines which cut the longitudinal
striations, producing a microstructure that he compared with that of Parallelopora. The
description by Galloway and Ehlers (1960, p. 102) of the microstructure of S. parallelo-
poroides Lecompte as composed of ‘ superposed maculae much as in the genus Parallelo-
porcP may also refer to this texture. The microreticulate nature of the tissue and its
relationship to Parallelopora is not confirmed by either Bargatzky’s or Nicholson’s
material.

Lecompte does not agree that striated tissue is characteristic of Stachyodes alone
because two species that he places in Idiostroma also show it. The relegation by Gallo-
way and Ehlers (1960) of the two species in question (/. crassum Lecompte, 1. irregularis
(Heinrich)) to Stachyodes helps to maintain the homogeneity of both genera with respect
to microstructure, but the appearance of striated tissue in the holotype of Idiostroma
caespitosum shows that the distinctness of microstructure of the two genera may be
difficult to maintain.

Galloway and Ehlers (1960) described the tissue of Stachyodes as porous with pores
parallel to the axis in the axial regions and perpendicular to it in the peripheral region.
The pores referred to are presumably the tubules described by Nicholson.

118

PALAEONTOLOGY, VOLUME 9

Most species of Stachyodes are columnar or dendroid but in some, such as S. fasci-
culata Heinrich, the tissue is continuous between the columns, which are merely high
mamelons in a massive coenosteum. Such specimens illustrate the difficulty of basing
generic distinctions on the form of the coenosteum. The axial tube is not present in all
cross-sections of the columns of Stachyodes and its presence or absence is not a valid
generic character in this or other genera.

parallelopora Bargatzky

Text-fig. 15

Type species: P. ostiolata Bargatzky

Coenosteum of amalgamate structure with thick, continuous, long pillars and
thick laminae generally composed of microlaminae. Tissue microreticulate and
cellules large, comparable in diameter to thickness of pillars. Cellules of pillars
so large and closely spaced that they reduce darker speckled tissue to a tenuous
three-dimensional lattice of strands and rods or to subspherical melanospheres.

TEXT-FIG. 15. X 50.

The distinctive features of the genera Paral/e/opora, Stromatopora, and Syringostroma
have been a matter of discussion ever since late in the nineteenth century. Recent dis-
cussions of the problem can be found in Yavorsky 1963, Lecompte 1951-2, and Gallo-
way 1957. Lecompte reviewed the older literature and concluded that the separation
of Parallelopora from Stromatopora was of doubtful validity and the only feature which
might justify Parallelopora would be the dominance of vertical pillars in its structure.
His viewpoint depended on his interpretation of the type of Stromatopora concentrica
Goldfuss as not showing such vertical structures (see above). Yavorsky believed that
Parallelopora shows many differences from Stromatopora in the net-like structure of the
coenosteum, the predominance of the pillars, their coarser porosity, and the ampullae.
Galloway emphasized as typical of Parallelopora its pillars ‘ composed of small, parallel,
vertical tubules and vertical rods or columns of dark dots’ (1957, p. 450). Some species,
such as Parallelopora nodulata (Nicholson), have been at one time or another assigned
to all these genera. Yavorsky (1963, p. 18) stated that the type species, P. ostiolata, is not
typical of the genus!

C. W. STEARN: THE MICROSTRUCTURE OF STROMATOPOROIDS 119

The differences of opinion concerning Parallelopora have stemmed from the assign-
ment to this genus by palaeontologists starting with Nicholson of species which
resemble the type species only slightly. The type specimen and sections of P. ostiolata
are in the Bonn University collections but sections of the type are also in the Nicholson
collection at the British Museum (Natural History). Pillars are the most conspicuous
element of the structure and horizontal elements are restricted to dissepiments and
latilaminae. In most places the dark material of the pillars seems to be arranged as
bundles of rods and large melanospheres, but where preservation is better, both in
Nicholson’s and Bargatzky’s slides, the pillars can be seen to contain two or three
vertical rows of large cellules (PI. 19, figs. 8, 9). These are so large that they reduce the
dark material between them to a series of rods and subspherical areas of speckled tissue.
The cellules are commonly elongated vertically and may join in vertical series to make
tubules. The pillars are so reduced by the cellules that their outlines are obscure and the
cellules can be confused with the superposed galleries that form pseudozooidal tubes.
In tangential section the pillars join into a continuous network. In this tissue are large
melanospheres (30 /x across) composed of fine specks (less than 1-3 ^ across). These
melanospheres are largely discrete in Nicholson’s slide (Nich. 125) but in Bargatzky’s
slide the transition from tissue in which the melanospheres are not joined, to tissue in
which the darker areas enclose cellules, can be seen locally. The microstructure of
Paralle/opora is established by the type species as being essentially microreticulate like
other related genera, but is distinctive in the size of the cellules which reduce the pillars
to a three-dimensional lattice.

Bargatzky described three other species of Parallelopora. P. stellaris is represented in
his collections by a specimen and tangential section but no vertical section appears to
have been cut. The broken surface of the specimen, however, shows coarsely cellular
pillars with ragged borders and poorly developed laminae. The tangential section is
much like that of P. ostiolata but shows well-developed astrorhizae. Slides of the type of
P. goldfussi Bargatzky both in Bonn and London show that it is similar in microstructure
to P. ostiolata (PI. 19, fig. 3). It has thick pillars containing large cellules arranged in
vertical series which tend to join into tubules. Lecompte (1952) pointed out that the last
species described by Bargatzky, P. eifeliensis, is probably the same as Stromatoporella
eifeliensis Nicholson. Lecompte referred Stromatopora bucheliensis Bargatzky to
Parallelopora but its cellular microstructure is much finer than that which would qualify
it for inclusion in the genus as here defined.

Nicholson referred several specimens with poorly preserved microstructure from
Devon to Parallelopora. Those referred to P. ciartingtonensis (Carter) have already been
discussed under Trupetostroma. Those referred to P. capitata (Goldfuss) are a diverse
group, none of which belongs in this genus. Nicholson’s specimens of P. capitata from
Germany have been assigned by Lecompte to a new species, Trupetostroma thomasi
Lecompte.

The species of Parallelopora described by Galloway and St. Jean (1957) from the
mid-western United States are closely related to the type species. P. campbelli Galloway
and St. Jean and P. typicalis Galloway and St. Jean show a vertical alignment of large
cellules. In P. pulchra Galloway and St. Jean microreticulate tissue is well developed and
continuous microlaminae are prominent. P. nodulata (Nicholson) and P. eumamillata
Galloway and St. Jean do not have the generic characters of Parallelopora (the large

120 PALAEONTOLOGY, VOLUME 9

cellules and microreticulate tissue) as well developed, and show some resemblance to
Syringostroma.

Family labechiidae Nicholson

The genera of this family are considered together since their structures and micro-
structures are similar. Their coenostea are composed primarily of overlapping cysts,
which may be of small radius as in Cystostroma Galloway and St. Jean, or long and low
as in Rosenella Nicholson and Pseudostylodictyon Ozaki. The cysts may form laminar
coenostea as in these two genera or they may build a cylindrical coenosteum around
a line of larger cysts as in Aulacera Plummer (formerly Beatricia Billings) and Crypto-
phragmus Raymond. They may be penetrated by through-going pillars of lamellar form
as in Stromatocerium Hall. They may have minute denticles or villi on their surfaces
( Rosenella and Cystostroma). The microstructures of the cysts and the pillars are the
subject of the following discussion.

The type species of Labechia Edwards and Haime, and therefore the standard for
comparison in the family, is L. conferta (Lonsdale) from the Wenlock limestone (Middle
Silurian) of Shropshire. Lonsdale’s specimens were not located. In topotypes of L. con-
ferta from the Nicholson collection the pillars are a light brown colour and are full of
small specks, from 5 to less than 1 p across, which have diffused slightly into the galleries.
The pillars have zigzag edges and extend in a sort of buttress where each cyst joins (PI.
19, figs. 1, 2). The cysts have sharp boundaries, are about 30-45 p thick and are com-
posed of a single layer of speckled compact tissue. The labechiid stromatoporoids are
generally poorly preserved but the preservation of these specimens is exceptionally good.
In many of them the crystal boundaries of the infilling calcite are completely independent
of the structure of the stromatoporoid. In tangential section the pillars show a slightly
differently coloured central area or several vaguely defined concentric zones of different
density. The concentric zones are poorly defined in vertical section where they appear
as upward pointing, invaginating cones. They mark lines of growth of the pillars and
do not demonstrate that the pillars were hollow. In most states of preservation the

EXPLANATION OF PLATE 19

Figs. 1. 2. Labechia conferta (Lonsdale) (Nich. 264, 2646). Tangential and vertical sections, X 50,
showing compact tissue, concentric zones of the pillars, and single layered cysts. Middle Silurian,
Dudley, Worcestershire.

Fig. 3. Paral/elopora goldfussi Bargatzky (Nich. 106). Tangential section, X 100, showing large cellules
up to 50 p across. Middle Devonian, Steinbreche, Germany.

Fig. 4. Aulacera cylindrica (Foerste) (UNC 302-74). Vertical section of a large cyst, X 100, showing the
migration of specks out into the gallery and the independence of crystal boundaries. Upper Ordo-
vician, 2 miles SW. of Deatsville, Kentucky.

Fig. 5. Aulacera nodulifera (Foerste) (Nich. 289, Beatricea undulata of Nicholson 1886). Cross-section
of cysts, X 100, showing bursts of diffusing specks. Upper Ordovician, Kentucky.

Figs. 6, 7. Rosenella macrocystis Nicholson (Nich. 280). Two vertical sections in the same slide, the
first showing no diffusion of specks due to the filling of the galleries, and the second showing diffu-
sion into the galleries where they are filled with calcite. Middle Silurian, Gotland.

Figs. 8, 9. Paral/elopora ostiolata Bargatzky (Nich. 125, 1254). Tangential and vertical sections of holo-
type, X 100, showing large vertically aligned cellules. Middle Devonian, Biichel, Germany.

Palaeontology , Vol. 9

PLATE 19

STEARN, Stromatoporiod microstructures

C. W. STEARN: THE MICROSTRUCTURE OF STROM ATOPOROIDS 121

cysts look dark against the light background of the galleries but in some specimens the
cyst may appear lighter than its halo of diffusing specks.

In some species and specimens of Labechia, e.g. L. macrostyla Parks, L. huron-
ensis (Billings) (Nich. 211a), and in Labechiella serotina (Nicholson), the cysts have
the appearance of those in Labechia conferta; in most other species and in other genera
of this family their appearance has been altered. Commonly the cyst appears in vertical
section as a thin dark line of specks bordered on either side by thicker layers of speckled
tissue. The outer edge of these layers is abrupt but is not marked by a darker membrane
or suggestion of a tissue boundary (PI. 19, fig. 4). The speckled layer on the concave side
of a cyst is usually thicker than that on the convex side and not uncommonly is so thick
that it almost fills the gallery. The density of specks in these two belts is uneven
and transverse clear areas may cross them. The specks may be arranged in fibrous
groups (‘bursts’) which converge on the cyst surface (PI. 19, fig. 5). In tangential section
the groups are circular concentrations of specks whose fibrosity is unrelated to adjacent
concentrations.

The speckled layers have been interpreted by Galloway and St. Jean (1961) and
Nicholson (1886) as part of the original structure of the coenosteum. That these cysts
were originally single layers of tissue and that the layers are zones of diffusion of specks
from the cysts and therefore a result of the process of preservation is shown by the
following evidence:

1. The speckled layers are not shown by the type species of Labechia and several other species and
specimens which are much better preserved than the specimens which show the layers.

2. The diffuse layers of speckled tissue around the laminae and cysts are common to nearly all genera
of Palaeozoic stromatoporoids and have no morphologic significance.

3. In many specimens the ‘bursts’ of specks are related to the pattern of crystals in the calcite which
has infiltrated the fossil. Specimens of this family in which the crystal mosaic is independent of coe-
nosteal structure are rare ; in most specimens a line of small crystals has first developed along the edges
of the galleries and later the rest of the galleries have been filled with large clear crystals. The diffusion
of specks seems to have taken place during the first phase of gallery filling. In the very rare cysts where
the boundary of the speckled layers is marked by a dark line, the line seems to be the boundary of a
single crystal or a set of crystals. More commonly the specks end just within the boundary of the
crystals lining the galleries. Each diverging group or ‘ burst ’ of specks is limited to a single such crystal.
Why the specks are arranged in a diverging pattern and do not reach the edge of the crystal is yet to be
explained but their arrangement is easy to observe.

4. Transverse clear zones along the layers which were interpreted by Galloway and St. Jean (1961)
as pores, are related to crystal boundaries and are due to the division of the layer into diverging groups
of specks.

5. Where sediment fills the galleries of specimens, and it commonly does in this family, the layers
are missing and no migration of specks has taken place (PL 19, figs. 6, 7).

Although the diffusion layer on the concave side is generally thicker, comparable
‘bursts’ of specks may also extend from the convex side of cysts. Perhaps the focusing
of diffusion forces from the concave surface has concentrated them and moved the
specks further than on the convex side. Gravity is not likely to be an agent causing this
asymmetry, for the concave side is not the downward side in all labechiids, but in colum-
nar coenostea, like Aulacera, the concave side is inward and the convex side outward.

Nearly all cysts in the Labechiidae (except in the genus Forolinia Nestor) appear to be
without pores but one or two specimens in the Galloway and St. Jean collection show
a well-defined perforation in the central dark line of a cyst. Such a rare structure could

122

PALAEONTOLOGY, VOLUME 9

be pathologic. How the living surface of the animal maintained contact with the tissue
within the coenosteum in the general absence of such pores is a mystery. Perhaps the
living tissue occupied only the surface. In this family, more than in any other group of
stromatoporoids, sediment has infiltrated the coenosteum. Nicholson (1886) illustrated
specimens in which all the galleries have been filled with micrite and specimens showing
zones and even isolated chambers so filled are common. In Aulacera and Cystostroma
layers of sediment occur in nearly all coenostea. How sediment which is rare in other
stromatoporoid coenostea has infiltrated a structure which is less porous than all others
is an unsolved problem.

The pillars of this group are rarely well preserved and in some genera are nearly
always dissolved away during preservation. Nicholson (1886) found some in Aulacera
but most specimens do not show them. In several species of Stromatocerium (e.g. S.
canadensis Nicholson and Murie) they have been replaced or dissolved and are now areas
of crystalline calcite.

Nestor (1964) interpreted the pillars of Stromatocerium as hollow and developed by
superposed upturnings of the laminae or cysts. This may be true of some species but
species like Stromatocerium platypilae Galloway have solid pillars and those of S. leiper-
sense Galloway and Ehlers seem to be penetrated by tubules. Those of other species may
contain dark streaks that suggest remnants of original tissue. The crystalline texture of
the pillars in most specimens may be explained by the poor calcification or aragonitic
nature of the original material.

The preservation of Cystostroma , Aulacera, and Crypt ophragmus, where organized
tissue occurs only here and there in a crystalline or sedimentary groundmass, shows that
the earlier members of this whole family were poorly calcified.

Other pillar-like structures are found in the genus Rosenella Nicholson and in
Cystostroma simplex Galloway and St. Jean. In the former denticles arising from the cysts
are composed of speckled tissue like that of the cysts. In most specimens the upper
boundary of the cyst to which the denticles are attached shows more diffusion of specks
than the sharp boundary on the lower side of the cyst. The ‘villi ’ of Cystostroma simplex
which extend upwards from the cysts as little fingers into the gallery above, consist of
microcrystalline aggregates embedded in the coarsely crystalline gallery filling. Another
curious vertical structure was described by Nicholson (1886u) in Rosenella platyphylla
Nicholson. This species has been made the type of the new genus Forolinia by Nestor
(1964). In the holotype (Nich. 283) the low cysts are locally thickened by tissue with vague
vertical markings, but no solid pillars exist. If the structure that Nestor interprets as
superposed foramina in the cysts represent leached or recrystallized pillars, then the
genus would be difficult to distinguish from Labechia.

An anomalous structure was described by Nicholson (1886) on a single specimen of
Aulacera undulata (Billings) from Anticosti. He regarded it as a peripheral phase of
Aulacera but as it is not found on other specimens of the species it is here considered to
be an associated organism of unknown affinities.

CONCLUSION

Before any classification of the stromatoporoids can be based on microstructures a
clear distinction must be made between those microstructures that were secreted by the

C. W. STEARN: THE MICROSTRUCTURE OF STROMATOPOROIDS

123

organism and those imposed on the tissue during the process of preservation of the hard
parts. During this study some microstructures have been determined as original, some
as formed during preservation, and some as both original and preservational. The follow-
ing kinds of tissue are primary in the sense that they are believed to have been secreted
by the organism :

compact, cellular and microreticulate, vacuolate, or (Unicellular, striated, tubulate,
peripherally vesicular.

The following kinds of tissue are secondary in the sense that they have been produced
in whole or in part by the processes of preservation:

water jet fibrous, flocculent, pseudotubular, melano spheric.

The following kinds of tissue are both primary and secondary:
fibrous, tripartite laminae, transversely porous.

Fibrosity in such genera as Amphipora, Hammatostroma, and perhaps Anostylostroma,
seems to be primary but the fibrosity developed in Syringostroma, Stromatoporella, and
other genera has formed during preservation. Water jet microstructure is apparently
always secondary in Palaeozoic stromatoporoids. Transversely porous tissue definitely
forms secondarily from ordinicellular laminae but may also be a primary microstructure
in such genera as Gerronostroma. Laminae with an axial dark or light line may be formed
by the coalescence of a single line of cellules or may possibly also represent a primary
microstructure.

REFERENCES

boehnke, k. 1915. Die Stromatoporen der nordischen Silurgeschiebe in Norddeutschland und Holland.
Palaeontographica, 61, 147-90.

fenton, c. L. 1920. The Hackberry Stage of the Upper Devonian. Am. Midi. Nat. 6, 179-200.
flugel, E. 1956. Revision der devonischen Hydrozoen der Karnischen Alpen. Carinthia II. Mitt,
naturw. Ver. Kdrnten, 41-60.

1958. Pseudoactinodictyon n. gen. und Actinodictyon Parks (Stromatoporoidea). Senckenberg.

leth. 39, 135-51.

1959. Die Gattung Actinostroma und ihre Arten (Stromatoporoidea). Annin natnrh. Mas. Wien,

63, 90-273.

galloway, j. j. 1957. Structure and classification of Stromatoporoidea. Bull. Am. Paleont. 37, no.
164, 341-480.

1960. Devonian Stromatoporoids from the Lower Mackenzie Valley of Canada. J. Paleont. 34,

620-36.

and ehlers, g. m. 1960. Some Middle Devonian stromatoporoids from Michigan and South-
western Ontario. . . . Contr. Mus. Paleont. Univ. Mich. 15, 39-120.

and st. jean, j. 1957. Middle Devonian Stromatoporoidea of Indiana, Kentucky, and Ohio.

Bull. Am. Paleont. 37, no. 162, 24-308.

hall, j. and whitfield, r. 1873. Description of new species of fossils from the Devonian rocks of
Iowa. 23 rd Ann. Rep. Regents Univ. State of New York, 233-43.

HEINRICH, m. 1914. Ueber der Bau und des System der Stromatoporen. Zentbl. Miner. Geol. Paldont.
732-6.

Hudson, r. g. s. 1958. Actostroma gen. nov., a Jurassic stromatoporoid from Maktesh Hathira, Israel.
Palaeontology, 1, 87-98.

lecompte, m. 1951-52. Les Stromatoporoids du Devonien moyen et superieur du bassin de Dinant.
Mem. Inst. r. Sci. nat. Belg. 116 (1951), 1-215; S17 (1952), 216-369.

124 PALAEONTOLOGY, VOLUME 9

lecompte, m. 1956. Stromatoporoidea. in moore, r. c., ed., Treatise on Invertebrate Paleontology,
Geol. Soc. Am., FI 07-44.

le maitre, d. 1934. Etude sur la fauna des calcaires devonien du bassin d’Ancenis. Mem. Soc. geol.
N. 12, 254 pp.

■ 1949. Sur quelques genres de Stromatopores devoniens et leur microstructure. Bull. Soc. geol.

Fr. ser. 5, 19, 513-26.

moseley, h. n. 1881. Report on certain hydroid, alcyonarian, and madreporarian corals procured
during the voyage of H.M.S. Challenger in the years 1873-1876. Rep. scient. Res. Challenger Exped.,
Zool. 2, pt. 7, 11-101.

nestor, H. e. 1964. Stromatoporoidei Ordovika i Llandoveri Estonii. Geol. Inst. Uurim. 112 pp.

nicholson, h. a. 1886, 1889, 1891, 1892. A monograph of the Britisli stromatoporoids. Palaeonto-
graphical Soc. [ Monogr .] 39, 42, 44, 46, 1-234.

1886a. On some new or imperfectly known species of stromatoporoids. Ann. Mag. nat. Hist., ser.

5, 18, 8-22.

and murie, j. 1878. The minute structure of the skeleton of Stromatopora and its allies. J. Linn.

Soc., Zool. 14, 187-246.

parks, w. a. 1909. Silurian stromatoporoids of America. Univ. Toronto Stud., geol. Ser., no. 6, 1-52.

1936. Devonian stromatoporoids of North America. Ibid. no. 39, 1-125.

ripper, e. a. 1933. The stromatoporoids of the Lilydale Limestone. Pt. I, Actinostroma and Clathro-
dictyon. Proc. R. Soc. Viet. N.s., 45, 152-64.

• 1937a. Stromatoporoids of the Lilydale Limestone. Pt. II, Syringostroma, Stromatopora, and other

genera. Ibid. 49, 178-200.

19376. On the stromatoporoids of the Buchan District, Victoria. Ibid. 50, 11-38.

1938. Notes on the Middle Palaeozoic stromatoporoid faunas of Victoria. Ibid. 50, 221-43.

rosen, f. b. 1867. Ueber die Natur der Stromatoporen und iiber die Erhaltung der Hornfaser der
Spongien im fossilien Zustande. Verb. russ. Mineral. Ges., ser. 2, 4, 1-98.

st. jean, j. 1960. The widespread distribution of characteristic Devonian stromatoporoid microstruc-
tures and their stratigraphic significance. Int. geol. Congr. 21st sess., Norden, Pt. 21, 239-50.

1962. Micromorphology of the stromatoporoid genus Stictostroma Parks./. Paleont. 36, 185-99.

1963. Maculate tissue in Stromatoporoidea. Spec. Pap. geol. Soc. Am., 76, Abstr. for 1963,

p. 143.

stearn, c. w. 1961. Devonian stromatoporoids from the Canadian Rocky Mountains. J. Paleont. 35,
932-48.

1962. Stromatoporoid fauna of the Waterways formation (Devonian) of Northeastern Alberta.

Bull. geol. Surv. Can. 92, 1-22.

1963. Some stromatoporoids from the Beaverhill Lake Formation (Devonian) of the Swan Hills

Area, Alberta. J. Paleont. 37, 651-68.

1965. Maculate microstructure in the stromatoporoids. Spec. Pap. geol. Soc. Am. 82, Abstr.

for 1964, 194.

steiner, a. 1932. Contribution a l’etude des Stromatopores secondaires. Bull. Labs. Geol. Geogr.
phys. Miner. Univ. Lausanne, no. 50, 1-117.

yavorsky, v. i. 1931. Some Devonian stromatoporoids from the outskirts of the Kuznetsk Basin, the
Urals, and other localities. Izv. glav. geol.-razved. Uprav. 50 (94), 1387-415.

1955. Stromatoporoidea Sovetskogo Soyuza, Pt. I. Trudy vses. nauchno-issled. geol. Inst. N.s., 8,

1-173.

1962. Stromatoporoidea. In sokolov, b. s., ed., Osnovy Paleont ologii, 2, 157-68. Moscow.

1963. Stromatoporoidea Sovetskogo Soyuza, Pt. IV. Trudy vses. nauchno-issled. geol. Inst., N.s.,

87, 1-93.

COLIN W. STEARN

Department of Geological Sciences,
McGill University,

Manuscript received 8 January 1965 Montreal

AN IMPROVED METHOD OF ANALYSING
DISTORTION IN FOSSILS

by KLAUS SDZUY

Abstract. Fossils distorted by tectonic strain can be reconstructed to their original shape if the strength and
direction of the principal axes of strain are known. The ratio and orientation of the principal strains can be
calculated by measuring the distortion in the angle between the median and transverse line in bilaterally sym-
metrical fossils like brachiopods and trilobites by several methods:

a. They can be determined with exactitude if two such fossils are found on one bedding plane.

b. If a number of single fossil specimens is available, but the relative orientations of their median and trans-
verse lines from specimen to specimen are unknown, the amount of deformation can be calculated by Breddin’s
method, making use of the most strongly deformed former right angle in the collection.

c. When the principal strain ratios have been calculated, the orientations of the principal axes of strain can
be determined in any single specimen.

The different processes are greatly simplified by a special device described in the text.

In 1962 I described a method for determining the amounts and orientations of the prin-
cipal strains produced by homogeneous strain in bilaterally symmetrical fossils if two
such fossils are found on the same bedding plane. A method based on the same principle
was first employed by Lake (1943) who applied it to cases where one of the two fossils
had remained right angled. When I wrote my first paper on the subject I was unaware of
Lake’s work.

The methods described in 1962 give good results, but they are time-consuming. For
palaeontologists who have frequently to deal with distorted fossils, as well as for struc-
tural geologists who may wish to use distorted fossils to determine the amount of com-
pression of rocks, speedier methods are desirable. Rapid methods for calculating the
strains are presented in this paper. When the amount of distortion has been calculated
the original form of the fossil may be reconstructed using the techniques employed by
Lake (1943), and Sdzuy (1962), and other authors cited by the latter.

Distortion of fossils is rather the rule in geosynclinal strata. In Cambrian strata from
different regions of Spain, for instance, deformation of about 60-70% is frequent.

THE PRINCIPLE EMPLOYED

For determining the amount of distortion only the angle between median and trans-
versal direction in bilaterally symmetrical fossils need be measured. This angle may be
called yS; it is constantly 90° in undistorted specimens. In deformed fossils /3 may remain
90°, or it may be either increased or decreased, and it can be measured directly in the
distorted specimens (text-fig. 1). There are for each fossil two complementary angles j8,
enclosing together 180°. The tables, etc., in this paper are calculated for the obtuse
angle

D indicates the amount of deformation as expressed by the ratio of the shortest to the
longest axis of the strain ellipse. D = 50% could mean that the specimen has been com-
pressed to 50% of its original extent parallel to the short axis of the ellipse and remained

[Palaeontology, Vol. 9, Part 1, 1966, pp. 125-34.]

126

PALAEONTOLOGY, VOLUME 9

unchanged parallel to the long axis. The direction of the short axis may then be pre-
sumed to be identical with the direction of deformation. A ratio of 50% could also have
been produced if the specimen is stretched to 200% of its original extent parallel to the
long axis and left unchanged parallel to the short axis of the ellipse, and still other pro-
cesses are possible. The fossils restored by the methods described here will in any case
show the original ratio of length/width.

text-fig. 1 . Relationship between undistorted and distorted fossil.

Interrupted lines : Undeformed cranidium of a trilobite drawn into a circle. Its main axes (transverse
and median) include a right angle; the main axes include with the direction of later principal strain
two acute angles ax (only one of these complementary angles is indicated).

Full lines'. Effects of deformation (D = 50%). The circle is deformed into an ellipse. The right angles
between the main axes of the fossil are changed into angle /? (only obtuse angle /3 indicated). Angle ax
is changed into angle a2.

The size of /3 depends not only on the amount of deformation ( D ), but also on the
size of the angle between the sides of j8 and the direction of principal elongation (text-
fig. 1). If this angle is 0°, that is if one of the sides of the original right angle is parallel
to the direction of deformation, /3 will be a right angle regardless of the amount of
deformation. The angle /3 is made up of two angles, and 90— before deformation,
and a2 and ,8— a2 after deformation. The effects of deformation ( D ) upon angle ax

can be expressed as follows :

tan c^. 100

tan olo -

D

and tan(90— a2) =

tan(90— (%). 100

D

or tan(90— a2) =

cot a.1. 100

D

K. SDZUY : ANALYSING DISTORTION IN FOSSILS

127

At a given amount of deformation /3 is greatest if was 45°. Since tan 45° (and cot 45°)
equals 1, 100

tan a2(or tan §/3) = — .

Text-fig. 3 shows the combinations /3/a 2 at deformations ranging from D = 100%
to D — 35%. A similar figure was published by Breddin (1956, fig. 32).

Since the deformed angle /3 depends on both D and the size of angle a, different
combinations of these two factors may produce the same angle /3. Consequently, the
amount and direction of the principal strains cannot be deduced from a single angle /3.

text-fig. 2. An ellipse obtained by deformation of a circle, with two different angles /3 (former right
angles deformed by the same strain) drawn (one indicated by a dashed line, the other by a full line).
The two angles /3 are at different angles (a2) to the direction of principal strain, and consequently their
size also differs. The two different angles /3 and angle y can be measured in specimens; angle a2 (necessary
for reconstruction of the fossils) can then be calculated.

Two fossils on the same bedding plane which have two different angles /3 are, however,
sufficient to determine with exactitude D and a. The relations between two such
different angles /3, produced by the same process of deformation, may be expressed by
angle y (text-fig. 2), which is equivalent to the difference between the two corresponding
angles a2 and which may be measured in the specimen.

If two different angles and angle y are known, D and a can be calculated. The mathe-
matical process is, however, tedious to manipulate and will not be treated here because
the indirect methods to find D and a described below and in my 1962 paper are easier and
much faster to use. Slight inexactitudes inherent in these methods do not matter because
they do not exceed the errors in measuring the different angles in the deformed specimens.

METHODS TO DETERMINE D AND a

1. If angle y is not known. A single angle /3 does not allow determination of D or a.
If, however, a sufficiently large number of specimens showing a single angle /3 each are
at hand and if the mode of their occurrence (same bed and outcrop, homogeneous rock,
etc.) indicates that they have been subjected to the same deformation, a method described
by Breddin (1956, pp. 264-6) will allow the determination of D and a. It is only necessary
to measure the most strongly deformed angle /3 in the material. If the material consists

(] *l6u V

128

PALAEONTOLOGY, VOLUME 9

of a large number of specimens, one may suppose that in the fossils with the most strongly
deformed angle /?, angle ax was close to 45°, and angle a2 equals |/3. D may then be
estimated using text-fig. 3 (the apex of the curves representing D in the graph indicates
the most strongly deformed angle on the vertical scale of the graph), or it may be

a2 that make up the obtuse angle The lines ascending from the bottom line towards the upper right
represent the progressing deformation of angle ax into a2-

calculated : D = (100/tan 3/3). The same equation, but with different symbols and another
expression of D, was given by Breddin (p. 266). The device described below offers a
quick way of measuring /3, of discovering the most deformed angle /3 and computing the
value of D.

When D has been calculated, angle a2 can be found for any specimen showing a
deformed angle by using the device described below. The graph, text-fig. 3, can also
be used for that purpose. The intersection between the curve representing D and the
horizontal line representing angle as measured in the specimen will indicate angle a2

K. SDZUY: ANALYSING DISTORTION IN FOSSILS

129

on the horizontal scale. Therefore, if D has been determined for a single specimen (or
pair of specimens), all other specimens incorporating a former right angle and subjected
to the same deformation can theoretically be restored. Inpractice, however, a serious draw-
back to this method is the fact that it is not known from which side of angle j8 that angle
a2 has to be measured. If angle a2 is measured from the wrong side of /3, /3 will be a right
angle in the restored specimen, but the deformation concerning the ratio length/width
of the specimen will be increased instead of eliminated. This effect is nil if the angle cq
was originally 45° and it increases with decrease in the value of a1. Where angle ax was
small (and consequently angle /3 does not differ much from a right angle), experience
with the fossil in question will usually show which position of angle «2 is the right one.

2. If angle y is known. As already explained, two specimens showing two different angles
/3 on the same bedding plane are sufficient to determine D and a with an accuracy which
is only impaired by imperfections of the material. Some of the methods are described
elsewhere (Sdzuy 1962); those described here are easier to use.

For a quick determination of D and a, the graph (text-fig. 3) may be used as follows:
Measure the two angles /3 and angle y, then find in the graph the curve representing D,
on which the intersections with the horizontal lines representing the two angles /3 are
separated on the horizontal scale by angle y. One drawback to this method is that the
measuring of the several angles is somewhat tedious; another is that because the curves
representing D in the graph have been constructed at intervals of 5%, it is sometimes
difficult to locate precisely the two points in the diagram which satisfy the observed
angles.

The easiest way to determine the deformation in fossils is by making use of a device
which allows direct reading of D and a,. Such a device is described in the following
section.

DEVICE FOR DIRECT READING OF D AND a2

Description. The device is made of four discs of transparent material which can be
rotated around their common centre (m in text-figs. 5, 6). Two of the discs (named a for
convenience) bear the design shown in text-fig. 4: a set of parallel lines in the central
part and a set of curves.

The other two discs (b) bear the following design: one line through the centre and a
marker (in in text-figs. 5, 6). This marker has to be at a distance from the centre which is
just a little greater than the distance of the outer margin of the curve-design in disc a
from the centre. The connecting line between this marker and the centre of disc b must
be at right angles to the line through the centre. The design of disc b is indicated in
text-figs. 5 and 6.

In addition to discs A and b, a ruler is needed (text-figs. 4, 6), on the scale of which
the amount of D is indicated. It is convenient to fix this ruler, also moveable, to the com-
mon centre of the four discs.

The most important and also the most complicated part of the device is the design in
disc a. If one wishes to make a device as described here, one may do so by simply
reproducing text-fig. 3 photographically and enlarging it to a convenient size on sheets of
clear film. Care must be taken that the ruler (also shown in text-fig. 4) is reproduced at

K

B 0612

130

PALAEONTOLOGY, VOLUME 9

text-fig. 4. Disc A and ruler. Both must be reproduced on exactly the same scale. The mode of con-
struction of the curves in disc a is explained in text-fig. 5. The centre m is indicated by a cross in the
central part of disc a and on the ruler. On the scale in disc A the number of degrees of the acute angle
will be indicated by the marker of disc b (the corresponding obtuse angle fi equals 180— acute angle
]8; angle a2, as needed for the construction of the curves, is obtained by subtracting the acute angle (3

K. SDZUY: ANALYSING DISTORTION IN FOSSILS

131

exactly the same scale as disc a. The two sheets of film, corresponding to discs a, may
then be placed between two plates of glass or some stiff transparent plastic, into which
the simple design of disc b has been scratched. The ruler also can be made of transparent
plastic. A small screw through the centre of the discs and through the ruler will hold them
in place and allow them to rotate against each other.

Each of the curves in disc a represents for a given angle /3 the possible values of angle
a2 at deformations ranging from D = 35 % to D = 100 %. The mode of construction of
these curves is explained in text-fig. 5. The intersection of the two curves for two different
angles / 1 in their natural position (separated by angle y) will then show the correct angles
oi2 for the two deformed fossils.

Use of the device. 1. To determine D and a 2 from two specimens on one bedding plane,
proceed as follows (see also text-fig. 6):

a. Place the device over the bedding plane with two deformed specimens (or, better,
over a photograph of the same on which the two angles f3 have been drawn with ink).

b. Make the lines in the central part of one of the discs a parallel to the right side of one
of the acute angles (or to the left side of the corresponding obtuse angle j3). The line
in the central part of one of the discs b must be parallel to the left side of the same
acute angle /3. Do the same with the other two discs for the other angle /3, without
changing the position of the device.

c. Follow the two curves indicated on the two discs a by the markers of the correspond-
ing discs b to their point of intersection.

d. Place the ruler over this point of intersection, so that the centre m of the ruler
coincides with the centre of the discs. The direction of the ruler is then equal to the
principal axes of strain, and the point of intersection of the two curves will indicate
on the scale of the ruler the exact amount of deformation.

2. If Breddin’s method has to be employed, only one pair of discs a and b is needed.
Proceed as follows :

a. To find the most deformed specimen, the device serves as a convenient angle meter;
if the lines in the central parts of the discs are parallel to the sides of angle /3, the marker
of disc b will indicate the number of degrees on disc a.

b. The point of intersection between the curve representing oq = 45° on disc A and the
curve of the most oblique angle /3 indicates the approximate value of D on the ruler
(see text-fig. 7).

text-fig. 4 ( continued )

from 90°). The dashed line cutting across the curves in disc a represents angle = 45°; it helps to
find D by Breddin’s method.

For the sake of clarity, only those curves corresponding to angles /3 with a round number of degrees
have been drawn. To help avoid confusion of neighbouring curves, every second one has been drawn
as a dashed line. The short radial part at the left end of each curve serves to connect the marker of the
corresponding disc b with the curve. This radial part may be cut off in one of the two discs A needed
and the marker of the corresponding disc B may be placed accordingly nearer to the centre of the disc;
this will help to distinguish more easily the two superimposed sets of discs.

The set of parallel lines in the central part of disc A, which has to be parallel to the right side of the
acute angle /3, helps to measure specimens without rearranging the lines forming the two angles /j.

132

PALAEONTOLOGY, VOLUME 9

text-fig. 5. Mode of construction of the curves in disc a, exemplified by the curve corresponding to
angle /3 = 102°. It is advisable to draw this on a polar stereographic net, as used in structural geology.

A net is formed by concentric lines, corresponding to degrees of deformation D (indicated only in
the left side of the drawing), and by radii, corresponding to angle a2 (indicated only for a2 = 30°
and 45°). Into this net the horizontal line representing angle /3 = 102° in the graph, text-fig. 3, is trans-
ported. (For instance: in text-fig. 3, = 102° intersects the curve representing D = 75% in the vertical

line representing a2 = 30°; the curve constructed in text-fig. 5 has therefore to intersect the concentric
line representing D — 75% on the radius that corresponds to a2 = 30°.) Points found in this manner
are indicated by small circles.

Those points are sparse in the middle part of the curve. Additional points may be found here by
transferring the lines representing the changes of angle from the graph, text-fig. 3, into the net (indi-
cated by small points, all of them drawn for the line corresponding to ax = 45° only). The points of
intersection of those lines with the horizontal line representing angle /3 can then be transferred from
the graph into the net. Points found in this manner are indicated by larger circles.

On the outer circumference, beyond the concentric line corresponding to D = 35%, the intersection
with the radius corresponding to angle a2 = /S— 90° is marked. (Where /3 = 102° is chosen as example,
this is the radius corresponding to a2 = 12°.) This point is connected to the end of the curve which is
on the concentric line corresponding to D = 35%. This point will be aligned to the marker of disc b
if the lines in the central parts of the two discs include the angle /?, to which the curve corresponds.

The lines in the central part of disc a have to be parallel to the radius corresponding to a2 = 0°.

The design of disc b is indicated by dotted lines : a line, or a system of parallel lines, in the central
part and a ‘marker’ (m) at right angles to it. (The thin line connecting m and m is not necessary on
disc b; it has been drawn to show the position of m.) The lines in the central parts of discs a and b
include an angle of 102° (or 78°), and consequently ‘m’ is aligned with the starting point of the curve

corresponding to /3 = 102° of disc a.

K. SDZUY: ANALYSING DISTORTION IN FOSSILS

133

text-fig. 6. Use of the device to determine D and <x2, exemplified by two distorted cranidia of trilobites.

One cranidium, its main axes forming angle jS, and the elements of discs a and b corresponding to it,
are drawn with dashed lines. The other cranidium and elements of the discs corresponding to it are drawn
with full lines (one axis of this cranidium is transferred into m). The ruler is indicated by a heavy line.

The lines in the central parts of discs a and b (indicated by letters in the drawing) are parallel to the
axes of the fossils. The markers ‘m’ of the two discs b indicate a curve each on the corresponding
disc a. The point of intersection of the two curves indicates on the ruler a deformation of 50%; the
ruler, placed over the point of intersection, is parallel to the direction of deformation. Distortion of
the two fossils in text-fig. 5 would be eliminated after stretching them to twice their extent parallel

to the direction of the ruler.

TEXT-FIG. 7 TEXT-FIG. 8

text-fig. 7. Determination of D with specimen in which angle j8 is most strongly deformed (indicated
by a brachiopod). See text. Angle /3 of the specimen is 102° (or 78°); D is then about 81%, as can be

read on the ruler.

text-fig. 8. Distorted fossil. D is known (50% in this case). Two different positions of the ruler are
possible; one of them is parallel to the direction of principal strain. The marker of disc b is indicated

by an arrow in text-figs. 7 and 8.

134 PALAEONTOLOGY, VOLUME 9

3. If D is known, a2 can be found in any single specimen as follows (see also text-

fig. 8):

a. Make the lines in the central parts of a single pair of discs A and b parallel to the sides
of angle /3.

b. The point on the ruler corresponding to D is then placed on the curve of disc a
indicated by the marker. Of the two different positions of the ruler, one is parallel to
the direction of deformation.

REFERENCES

breddin, h. 1956. Die tektonische Deformation der Fossilien im Rheinischen Schiefergebirge.
Z. deutsch. geol. Ges. 106, 227-305, pi. 2-4.

lake, p. 1943. Restoration of the original form of distorted specimens. Geol. Mag. 80, 139-47, pi. 7.
sdzuy, k. 1962. fiber das Entzerren von Fossilien (mit Beispielen aus der unterkambrischen Sau-
kianda- Fauna). Paldont. Z. 36, 275-84, pi. 25.

KLAUS SDZUY

Paliiontologisches Institut der
Universitat Wurzburg,
Pleichertorstrasse 3,
Wurzburg,

Manuscript received 21 January 1965 Western Germany

ON BIRGERIA ACUMINATA AND THE ABSENCE
OF LABYRINTHODONTS FROM THE RHAETIC

by r. J. G. savage and n. f. large

Abstract. A maxilla of Birgeria acuminata from the Rhaetic of the Bristol Channel is described. Previous
records of labyrinthodonts from the British Rhaetic are shown to belong to this piscine genus.

Vertebrate remains from the Rhaetic Bone Bed around the Bristol Channel are
well known. The great majority of the small specimens comprise fish teeth, scales, and
spines while the larger specimens are attributable to reptiles, mainly plesiosaurs and
ichthyosaurs. The fish remains usually figure under the names Acrodus, Hybodus,

‘ Saurichthys ’, Nemacanthus, Sargodon, Gyrolepis , and Ceratodus. In addition to the
marine reptiles, Plesiosaurus and Ichthyosaurus, occasional terrestrial reptiles have been
found; these bones have been referred to dinosaurs but the material is not sufficiently
diagnostic for more precise identification. From specimens which have been described
and from material now in the University of Bristol Geology Museum, at least two
species of terrestrial reptile are clearly represented, a large species and a small species
about 2-3 ft. long. No pterosaurs or mammals are recorded from the Bone Bed, and there
remains for consideration only one other vertebrate group, the amphibia.

Miall (1875) reported that Metopias diagnostics von Meyer, a labyrinthodont from
the Keuper of Germany, was present in the Aust Bone Bed and the statement has been
often repeated in the literature. Reynolds (1946) stated that labyrinthodont remains
were common at Aust, usually very fragmentary cranial bones with ‘characteristic’
pitted surface and rarely jaw fragments with teeth. Reynolds further noted that two
types of teeth are found on these jaw fragments, and that some isolated teeth commonly
attributed to Saurichthys may be labyrinthodont.

The true identity of these ‘labyrinthodont’ specimens has remained unresolved, due
largely to lack of really good material. Reynolds illustrated a mandible fragment with
five teeth, but the specimen is highly pyritized and impossible to prepare. In 1954
Professor C. F. A. Pantin found a dentary which he kindly donated to the University
of Bristol Geology Museum (UBGM 19001); this specimen prepared well in acetic acid
and has four teeth. More recently one of us (N.F.L.) found a maxilla at Westbury-on-
Severn which contains seven major teeth and a labial row of smaller teeth. The speci-
men has been beautifully prepared by Mr. A. E. Rixon of the British Museum (Natural
History) and is described in detail below; it is patently Birgeria and indistinguishable
from other specimens previously referred to the labyrinthodonts, which can now be
placed in this piscine genus.

Family birgeriidae Berg 1940

Diagnosis. Body almost naked, scales only present on dorsal half of caudal pedicle, on
body axis of caudal fin, and anteriorly to the caudal pedicle in a single row along the
main lateral line. As in the Palaeoniscidae, two nasal openings on each side, one anterior

[Palaeontology, Vol. 9, Part 1, 1966, pp. 135-41, pi. 20.]

136

PALAEONTOLOGY, VOLUME 9

and one posterior to the postrostral. Sensory canal system differing from that in
Palaeoniscidae mainly in the following respects: supraorbital sensory canal ending
blindly in rostropremaxillary without communicating with ethmoidal commissure of
infraorbital sensory canals; dorsal part of preopercular sensory canal extending from
radiation centre of preoperculum in an almost anterior direction towards most dorsal
part of postorbital vertical section of infraorbital sensory canal. Very large unpaired
rostropremaxillary. Parietals (one or two on each side) separated by frontals. One (or
two?) extrascapulars on each side. Supratemporal independent. Intertemporal fused
with dermosphenotic and with a number of posterior supraorbitals. A larger or smaller
number of free supraorbitals. Neural endocranium not ossified in a single piece but in
separate bones (an unpaired occipital bone, a paired prootico-opistotic, a paired auto-
sphenotic, an unpaired sphenoid, and a paired ethmoidal bone). Posterior myodome
smaller than in Palaeoniscidae. Parasphenoid strong, extending backward below the
whole neural endocranium, and with a strong paired processus ascendens. Mouth much
larger than in Palaeoniscidae. Suspensorium very oblique. Teeth on jaws arranged in
a labial and a lingual row. Labial teeth small, lingual teeth very large. Operculum small,
suboperculum lobate much as in Polyodon. Spiracle present. Branchiostegal rays
numerous. Fin rays jointed, devoid of ganoine. Pelvic fin large, with about fifty lepido-
trichia. Dorsal and anal fins both situated far back, dorsal fin slightly anterior to anal
fin, both fins large and with more than fifty lepidotrichia. Fulcra present only on the
body axis of the caudal fin. Endoskeleton of dorsal fin represented by three series of
radial ossifications; an axonost, a baseost, and a distal series. Endoskeleton of anal fin
represented by two series of radial ossifications, probably an axonost and a baseost
series ; baseost series very short, supporting only posterior third of fin. Posterior elements
of axonost series of endoskeleton of both dorsal and anal fins fused to a fairly extensive
axonost plate. Notochord persistent. Axial skeleton with ossified dorsal and ventral
arcuals, in abdominal region moreover with supraneurals, in part of caudal region
possibly also with independent infrahaemals. Skeleton of pelvic fin consisting of an
ossified pelvic plate and one or two rows of radial ossifications.

Remarks. The family name was erected by Aldinger (1937) but without a diagnosis.
Nielson (1949) gave the above detailed amended diagnosis in the light of new Green-
land Triassic material. Nielsen bases the family on Birgeria and does not definitely
include in it any other genus.

Genus birgeria Stensio 1919

Diagnosis. Large or fairly large fishes. Trunk long and slender, oval in transverse section.
Head rather large, with big eyes and extremely long jaws. Neural endocranium narrow
both anteriorly and posteriorly, but very broad in its middle part and especially between
its strong postorbital processes. Neural endocranium in both small and large specimens
partly occupied by a number of independent bones. External dermal bones of the head
relatively thick and mainly ornamented with tubercles, to a smaller extent, however,
also with striae. Striae best developed on maxillary and dentalo-splenial, and on these
bones to some extent covered by more superficially situated tubercles of varying shape.
On the jaws a labial row of small, and a lingual row of large teeth. Parallel with these
two rows, a third row consisting of small to medium-sized teeth on the ventral or ventro-

SAVAGE AND LARGE: ON BIRGERIA ACUMINATA

137

lateral part of the oral faces of the ectopterygoid and the dermopalatines. Moreover,
large areas of oral faces of ectopterygoid, entopterygoid, dermopalatines, dermo-
metapterygoid, and prearticular covered by very small teeth. Small teeth also present
on ossifications of basibranchial series and branchial arches. Teeth of jaws and at least
larger teeth on dermal bones of palate pointed and conical with a two-edged enamel
cap, in the geologically older species on the whole smooth, in the geologically youngest
species equipped with vertical, elevated, sharp striae; basally to enamel cap the teeth of
all the species vertically striated, either with fine striae alone or with coarse striae which
in their turn bear the fine striae; plicidentine not present. Teeth of branchial arches
more slender than those of the jaws and with an enamel cap, which always seems to be
smooth. [Abbreviated from Nielsen 1949, pp. 280-4.]

Type species. Birgeria mougeoti (Ag.) from the Triassic of Europe and Spitsbergen.

Species. In addition to the type species, the following species have been recognized
by Nielsen as belonging to the genus:

Birgeria stensidi Aldinger 1931
Birgeria groenlandica Stensio 1932
Birgeria velox (Jordan) 1907
Birgeria acuminata (Ag.) 1844
Birgeria costata (Munster) 1839
Birgeria annulata (Winkler) 1880
Birgeria nielseni Lehman 1948

from Triassic of Europe
from Triassic of Greenland
from Triassic of California
from Rhaetic of Europe
from Rhaetic of Europe
from Triassic of Europe
from Triassic of Madagascar

Remarks. The diagnosis in Nielsen (1949, pp. 280-4) contains an account of the ana-
tomy of the genus. It has been shortened here to the statements apposite to the present
discussion. Lehman (1952) gives a full synonymy for the genus. Although eight species
are listed, most are poorly known. They have a wide geographic distribution and rela-
tively narrow stratigraphical range from the Lower Triassic to the Rhaetic.

Birgeria acuminata (Agassiz)

Plate 20

Abbreviated Synonymy.

1843 Saurichthys apicalis, J. E. Portlock (err ore), p. 470, pi. 14, fig. 19.

1844 Saurichthys acuminatus, L. Agassiz, p. 86, pi. 55a, figs. 1-5.

1844 Saurichthys longidens, L. Agassiz, p. 87, pi. 55a, figs. 17, 18.

1921 Birgeria acuminatus, E. A. Stensio, p. 150.

Remarks on the Synonymy. Without a detailed study of the German material it would
be pointless and possibly confusing to list the twelve specific names sometimes ascribed
to Saurichthys , among which some are probably synonymous with B. acuminata. Details
of these are to be found in Woodward (1895, pp. 21-23) and Boni (1937).

Diagnosis. Large fish with robust dentition. Basal portion of teeth below enamel cap
relatively shorter than in B. mougeoti. Enamel cap terminated proximally by prominent
collar; teeth with fine vertical striations, best developed immediately proximal to collar.

Remarks on Diagnosis. Agassiz (1844) noted the dental characters which differentiate
B. mougeoti from B. acuminata. Woodward (1895) restated some of these. Boni (1937,
pp. 584-5) listed characters including table of jaw lengths, which he regarded as

138

PALAEONTOLOGY, VOLUME 9

differentiating B. mougeoti , B. stensidi, B. groenlandica, and B. acuminata. Lehman (1952)
stated that B. groenlandica had two forms, a large and a small, the latter being comparable
with B. nielseni. Boni (1937, p. 596) in discussing B. costata from the Rhaetic of the
Imagna Valley in Italy, suggested that presence or absence of a keel on the tooth crown
is a variable which cannot be relied upon as a specific character. B. costata is very much
larger than B. acuminata. Stratigraphically only B. acuminata and B. costata among
Birgeria species have a Rhaetic range; whether these are merely stratigraphic species
or represent biological differences must await better material. B. acuminata is common
in the Rhaetic of north-west Europe and there are no known occurrences outside the
Rhaetic.

Holotype. The teeth illustrated by L. Agassiz (1844) on Plate 55, figs. 1-5 were described by him as
coining from the Muschelkalk of Aust Cliff. The specimens, then in the Bristol Institution Museum,
were later incorporated in the Bristol City Museum collections and registered collectively under the
catalogue number C 4578. Unfortunately all were destroyed when the Museum was bombed in 1941.
The horizon is undoubtedly the Rhaetic Bone Bed at Aust Cliff, 10 miles on 6° west of north from
Bristol (NGR 566898). The International Rules of Zoological Nomenclature do not encourage the
creation of neotypes unless circumstances are exceptional and to conform to this ruling no new type is
here established.

Description. The new specimen of Birgeria acuminata, BMNH P47287, is from the
Westbury Shales of the Rhaetic at Westbury-on-Severn, 8 miles WSW. of Gloucester,
(NGR 715132). The specimen comprises the middle portion of a left maxilla, and is
broken both anteriorly and posteriorly while the dorsal ramus is largely missing. The
external surface is completely covered with fine tubercles except along the antero-dorsal
margin where the orbital bones overlapped. The tubercles are densest along the dental
margin and near to the orbital region are extremely fine. The dorsal part of the bone is
insufficiently preserved to detect any pattern in the arrangement of the tubercles.
Internally the bone is smooth and extended along the ventral margin of the teeth.

Along the 12 cm. of tooth row preserved there are 14 alveoli for lingual teeth and in
8 of these the teeth are present. Lateral to these the maxilla margin carries a row of small
labial teeth, of which some 8 remain intact. The lingual teeth are conical, lie close
together in a row without any space between them at the level of the insertion into the
maxillary bone. When a tooth is missing an ovoid pit remains which extends on either
side to the base of the adjoining teeth. At the level of the ventro-lateral maxillary margin
the lingual teeth appear more spaced due to the effect of the tapering cone and the dis-
tance between the teeth is about equal to the tooth diameter at this level. The ventro-
lateral margin of the bone varies in thickness a little; is thickest between the lingual teeth
and thinnest around them. The distribution of the labial teeth is slightly irregular, but
the pattern is broadly a single row with on average 6 labial teeth to every one lingual
tooth. In size the labial teeth tend to be slightly larger between the linguals and smallest
when alongside the linguals. The lingual teeth also vary slightly in size, but without any
observable pattern; their size may reflect more the time of eruption than their position
in the jaw.

EXPLANATION OF PLATE 20

Birgeria acuminata (Agassiz), UBGM 19001, Rhaetic of Westbury-on-Severn, Gloucs. Left maxilla a,

lateral aspect; b, occlusal aspect, x T5.

Palaeontology, Vol. 9

PLATE 20

SAVAGE and LARGE, Birgeria acuminata from the Rhaetic

SAVAGE AND LARGE: ON BIRGERIA ACUMINATA

139

The lingual teeth are roughly conical in outline, the basal portion being flayed out for
attachment to the bone. The conical axis of the teeth is curved slightly inward, but there
is no detectable anterior or posterior curvature. Each tooth is terminated by an enamel
cap, which in newly erupted teeth is pointed but becomes blunted with wear. The tooth
cap is terminated proximally by a ring, collar, or cingulum. From this collar arise
anterior and posterior keels which do not quite reach the summit. Fine striations also
arise from the level of the collar, and extend both proximally and distally; they are
densest on the medial side of the tooth and are few or lacking on the lateral side. Those
rising on to the cap do not reach the summit and frequently unite. Those extending
proximally also disappear before reaching the base. The labial teeth appear to have
a similar build to the linguals though striations and keels are rarely detectable.

Remarks. The purpose of the present paper is to clear up some of the confusion prevailing
with respect to labyrinthodonts in the Rhaetic and a detailed appraisal of Birgeria is not
attempted. The Westbury beds of the Rhaetic yield vertebrates at several horizons, from
the shales and limestones as well as from the basal bone conglomerate. The vertebrate
fauna appears to be essentially the same at all levels though reptile bones are rare outside
the basal bone bed. Isolated conical teeth are abundant in the Westbury beds and these
can invariably be assigned to either Birgeria or Ichthyosaurus. The Ichthyosaurus teeth
differ from those of Birgeria in being larger, less tapering and hence more nearly parallel
sided, and in lacking a collar; keels on the caps are wanting and the proximal portion
of the teeth is infolded to a greater or lesser extent. While most teeth are recognizable
on external morphology, preliminary investigations suggest that histological structure
can be used to distinguish reptile and fish, and possibly even to identify genera. Careful
collecting from individual beds and use of histological techniques could considerably
advance our knowledge of these faunas.

Several other maxillae and mandibles are known beside the specimen described
above; though none is so well preserved they are all clearly Birgeria and not labyrin-
thodont. Reynolds (1946) described a mandible (UBGM 7976) as ‘ Metopias diagno-
s ficus' ; the specimen is unfortunately highly pyritized and cannot be prepared with acid,
but is undoubtedly Birgeria acuminata. Woodward (1889) described a left maxilla (BCM
C4579) as Saurielithys acuminatus; the specimen is closely comparable with that
described above, has two lingual teeth and the row of alveoli for the labial teeth is
distinguishable. Numerous other smaller pieces of maxilla and mandible exist in
collections. Another frequently found part is the symphysis of the mandible; this is very
thick and heavy (e.g. UBGM 18038, i-iv). Comparison of Birgeria with the living Lepi-
dosteus is particularly close on many anatomical features — the long jaws, the heavy
mandibular symphysis, the external ornament, the arrangements of the lingual and
labial teeth; striations and keel are present on the lingual teeth but differ in detail from
those of Birgeria.

A search of the Rhaetic exposures around the Bristol Channel and museum collections
has revealed no vertebrate material that cannot be assigned to either fish or reptile.
The previous records of ‘labyrinthodont’ are shown to be misidentifications for large
specimens of Birgeria. No evidence for amphibia remains. Stratigraphically this inter-
pretation clarifies the position regarding the labyrinthodonts. It can now be stated that
there are no known Rhaetic metaposaurid labyrinthodonts; their last appearance is in

140

PALAEONTOLOGY, VOLUME 9

the Upper Triassic beds of India and North America. Jain, Robinson, and Roy Chowd-
hury (1964) record metaposaurs from the Maleri beds of the Pranhita-Godavari valley,
India, and these beds they equate with the Carnian to Middle Norian zones of the
marine Upper Triassic.

Nilsson (1946) recorded the presence of a plagiosaur from the Rhaetic of Scania,
Sweden ; the material he identified as Gerrothorax and is well preserved in a coal-bearing
deposit of Middle Rhaetic age. Nilsson considered the plagiosaurs to form the type
family of an order of Triassic stegocephalians, distinct from labyrinthodonts. Panchen
(1959) writing on plagiosaur affinities, excluded the brachyopids from his order Plagio-
sauria and placed the plagiosaurs within the superorder Labyrinthodontia. Hence the
presence or absence of labyrinthodonts in deposits of post-Triassic age depends on
(n) whether the definition of the Labyrinthodontia includes or excludes the Plagiosauria
and ( b ) whether the definition of the Trias includes or excludes the Rhaetic. The Rhaetic
of Britain has yielded no amphibia; the Rhaetic elsewhere contains amphibia which
may be labyrinthodont.

Acknowledgements. We have had discussions and correspondence with the late Professor B. Peyer, Dr.
J. Griffith, Dr. D. Rayner, Dr. T. K. Roy Chowdhury, and Dr. B. Schaeffer; we are particularly grate-
ful to Mr. A. E. Rixon for preparing the fragile specimen. Through the paper the following abbrevia-
tions for museums have been used: BCM, Bristol City Museum; BMNH, British Museum (Natural
History); UBGM, University of Bristol, Geology Museum.

REFERENCES

agassiz, l. 1844. Recherches sur les poissons fossiles, 2. Neuchatel.

aldinger, h. 1931. Uber Reste von Birgeria aus der alpinen Trias. Neues Jb. Min. 66B, 167-80,
8 figs., 1 pi.

1937. Permische Ganoidfische aus Ostgronland. Medd. Gronland , 102, 1-392, 105 figs., 44 pi.

berg, l. s. 1940. Classification of fishes, both recent and fossil. Trav. Inst. Zool. Acad. Sci. URSS:
Leningrad. 5, 85-517, 190 figs.

boni, a. 1937. Vertebrati Retici Italiani. Mem. R. Accad. Lincei, 6, 521-719, 7 tables, 1 1 figs., 6 pi.
jain, s. l., robinson, p. l., and roy chowdhury, t. k. 1964. A new vertebrate fauna from the Triassic
of the Deccan, India. Quart. J. geol. Soc. Load. 120, 115-24, 3 figs.

Jordan, d. s. 1907. The fossil fishes of California, with supplementary notes on other species of extinct
fishes. Univ. Calif. Publ. geol. Sci. 5, 95-145, 33 figs., 2 pi.

Lehman, j-p. 1948. Sur la presence du genre Birgeria dans Eeotrias de Madagascar. C'.R. Acad. Sci.,
Paris, 226, 426-28.

1952. Etude complementaire des poissons de l’Eotrias de Madagascar. K. svenska Vetensk-Akad.

Handl. 2, 201 pp„ 129 figs., 48 pi.

miall, L. c. 1875. On the structure and classification of the labyrinthodonts. Rep. Brit. Ass. Advanc.
Sci. for 1874. 149-92, 4 pi.

munster, g. von. 1839. Ueber einige merkwiirdige Fische aus dem Kupferschiefer und dem Mus-
chelkalk. Beitr. Petrefactenk. 1. 114-22, 2 pis.

nielson, e. 1949. Studies on Triassic fishes from East Greenland. II. Australosomiis and Birgeria.
Medd. Gronland, 146, 1-309, 82 figs., 20 pi.

njlsson, t. 1946. A new find of Gerrothorax rhaeticus Nilsson, a plagiosaurid from the Rhaetic of
Scania. Acta. Univ. Lund. 42, 1-42, 13 figs. 2 pi.

panchen, a. l. 1959. A new armoured amphibian from the Upper Permian of East Africa. Phil. Trans.
242B, 207-81, 20 figs.

portlock, j. e. 1843. Report on the geology of the county of Londonderry, and parts of Tyrone and
Fermanagh. Dublin, H.M.S.O.

Reynolds, s. h. 1946. The Aust section. Proc. Cotteswold Nat. Fid. Cl. 29, 29-39.

SAVAGE AND LARGE: ON BIRGERIA ACUMINATA 141

stensio, E. 1919. Einige Bemerk ungen iiber die systematische Stellung von Saurichthys mougeoti Agassiz.
Senckenberg. leth. 1, 177-81, 3 figs.

1921. Triassic fishes from Spitzbergen, Pt. I. Vienna, xxviii+307 pp., 87 figs., 34 pi.

1932. Triassic fishes from East Greenland. Medd. Gronland, 83, vi+305 pp., 94 figs., 39 pi.

winkler, t. c. 1880. Description de quelques restes de poissons fossiles des terrains triasiques des
environs de Wurzbourg. Arch. Mus. Teyler, Haarlem. 5, 109-49, 5 pi.
woodward, a. s. 1889. On a maxilla of Saurichthys from the Rhaetic of Aust Cliff, near Bristol. Ann.
Mag. nat. Hist. 3, 297-302, 1 pi.

1895. Catalogue of the fossil fishes in the British Museum , Pt. III. xlii+544, 45 figs., 18 pi. London,

Trustees of the British Museum.

R. j. g. savage
Department of Geology,
University of Bristol
N. F. LARGE

St. John’s College,

Manuscript received 28 January 1965 University of Cambridge

THE UPPER DEVONIAN GASTROPOD ORECOPIA
IN WESTERN CANADA

by A. E. H. PEDDER

Abstract. The distribution of Orecopia is reviewed. Apart from a possible Famennian occurrence in Nevada,
the genus as presently known is confined to the Frasnian of western North America and eastern Europe and is
represented by five described species. Orecopia mccoyi and O. uchtensis are refigured and a new species, O.
cotei, is erected for specimens from British Columbia and the Northwest Territories.

It is suggested that the Tathlina Formation of the southern Northwest Territories may be late, rather than
middle Frasnian.

Gastropods now referred to Orecopia have been known from Timan and Latvia
since 1846 and from Nevada since 1884. Although subsequently numerous specimens
have been identified from eastern Europe and cordilleran U.S.A., the genus has only
once been reported in Canada, and then without either geographical or stratigraphical
location. The present paper confirms that Orecopia is present in Canada and describes
a new species of late Frasnian age from British Columbia and the Northwest Territories.

Acknowledgements. It is a pleasure to acknowledge help extended from several sources. E. G. Kauff-
man and E. L. Yochelson of the United States National Museum arranged for the loan of specimens
from Nevada, California, Latvia, and Timan, while further specimens from Nevada were provided by
R. H. Waines of the State University College, New Paltz. The Canadian specimens were collected
during field-work mounted by Triad Oil Co. Ltd., Calgary. The management of the Company has
granted permission to publish this work and has allowed the Company’s specimens to be transferred
to the Geological Survey of Canada, Ottawa.

SYSTEMATIC PALAEONTOLOGY

The collections containing types are abbreviated as follows:

GSC Geological Survey of Canada, Ottawa.

USNM United States National Museum, Washington.

Family omphalotrochidae Knight
Genus orecopia Knight 1945

Type species {original designation). Platyschisma ? mccoyi Walcott 1884, p. 188, pi. 17, fig. 1. ‘Central
portion of the Devonian limestone; Newark Mountain Eureka District, Nevada.’ In terms of current
stratigraphical nomenclature the type horizon probably lies in the upper part of the Hayes Canyon
Member of the Devils Gate Limestone (Nolan et al. 1956) and is early Frasnian. Merriam (1940, pi. 11,
fig. 3) has subsequently figured one of the syntypes and Knight (1945, pi. 80, fig. \a-d, g) has figured
the specimen, which he designated as lectotype, and also three paralectotypes.

Diagnosis. See Knight 1945, p. 586.

Distribution. At the present time five Frasnian species, including Orecopia cotei, described
below, are referred to the genus. These are distributed in western North America from
southern California to Great Slave Lake and in the Soviet Union from Timan and Latvia

[Palaeontology, Vol. 9, Part 1, 1966, pp. 142-7, pi. 21.]

A. E. H. PEDDER: THE UPPER DEVONIAN GASTROPOD ORECOPIA 143

to the Urals. In addition one of the species may be represented by derived fossils from
the North German Plain (Roemer 1885, p. 384, pi. 11, fig. 1 a).

Orecopia mccoyi (Walcott) and comparable forms are widely distributed in the Basin
and Range Province of western United States. They have been reported in the higher
parts of the Devils Gate Limestone in the Sulphur Springs Range (Carlisle et al. 1957,

text-fig. 1. Known distribution of Orecopia in North America. Occurrences of Orecopia mccoyi
(Walcott) indicated by a circle and those of Orecopia cotei sp. nov. by a triangle.

p. 2188), the Phillipsastrea Zone and Hayes Canyon Member of the Devils Gate
Formation of the Eureka district (Merriam 1940, p. 60; Nolan et al. 1956, p. 51), the
upper beds of the Guilmette Formation of the Gold Hill mining district (E. Kirk in
Nolan 1935, p. 21), the upper member of the same formation in the southern Egan
Range (Kellog 1963, p. 699), Devils Gate or Sultan equivalent in the Yucca Flat district
(Johnson and Hibbard 1957, p. 356), the Valentine Member of the Sultan Formation

144

PALAEONTOLOGY, VOLUME 9

in the Spring Mountains of the Goodsprings Quadrangle (E. Kirk in Hewett 1931, p.
16; Knight 1945, p. 586 and collections made by R. H. Waines), and in the same
formation in the Ivanpah Quadrangle (collected by Hewett, USNM loc. 2220). It is also
possible that the species occurs in the upper beds of the Lost Burro Formation of the
Ubehebe and Inyo Mountains of California (Merriam 1963u, p. 55; 1963Z?, p. 16). In
view of the considerable morphological variation in specimens of Orecopia from the
Basin and Range Province, particularly in regard to the height of the spire and umbilical
width, it is possible that more than one species is represented. Most occurrences are in
the early Frasnian argentarius Zone and Merriam (1963a, p. 55) regards the gastropod
as an especially useful indicator of the lower part of the zone. However, he previously
reported (1940, p. 60) a comparable species in the slightly younger Phillipsastrea Zone
northwest of Eureka, and there are sufficient Lime Creek species in the Valentine Mem-
ber of the Sultan Limestone in the Goodsprings Quadrangle (E. Kirk in Hewett 1931,
p. 16), to suggest that O. niccoyi may range into the Middle Frasnian.

Knight (1945, p. 586) alluded to undescribed specimens of Orecopia from the Devonian
of western Canada, but gave no information concerning their occurrence. Neither Dr.
Kauffman nor Dr. Yochelson (personal communication, October 1963) has been success-
ful in locating the material to which Knight was referring. The present paper describes
Orecopia from north-eastern British Columbia and south-western Northwest Territories
and apart from Knight’s paper, is the first report of the genus in Canada. Although there
is no doubt that the first of these occurrences is late Frasnian, the second, in the Tathlina
Formation, is less clearly dated. McLaren (1962, fig. 1) and the writer (in House and
Pedder 1963, text-fig. 2) have placed the Tathlina Formation high in the middle Frasnian.
The upper beds of the underlying Twin Falls Formation contain several species, includ-
ing MictophyUum modicum Smith, ThamnophyUum tructense (McLaren), and Atrypa
ciliipes Crickmay, which are otherwise known only in the late Frasnian. Their presence
below the Tathlina Formation and the presence of Orecopia cotei in it are probably
significant and suggest that the Tathlina Formation is late, rather than Middle Frasnian.
In this case the aibertcnsis Zone would be equivalent to the Twin Falls Formation only,
of the Hay River sequence, and the Tathlina would correlate with part of the Nisku
and Arcs Dolomites of Alberta.

Three species from the Soviet Union are now referred to Orecopia. Platyschisma
uchtensis , the first of these, was described by Keyserling (1846, pp. 263, 264, pi. 11, fig. 6)
and was said to occur in Devonian calcareous sandstone on the River Uchta (Ukhta),
a tributary of the Ishma. In the light of Nalivkin’s work (1947, p. 14) the type horizon is
probably in either the Sirachoy or Ukhta Series and is late Frasnian. Tschernyschew
(1884, p. 53, pi. 1, fig. 2; 1887, p. 35, pi. 5, fig. 4) extended the known geographical range
of the species by describing specimens from the Voronezh and southern West Urals
regions ; in both areas these specimens were associated with ‘ Spirifer ’ anossofi and would
now be regarded as late Frasnian. Besides reiterating these occurrences Nalivkin (1947,
pp. 9-21) reported the species in the Semiluki Formation of the Voronezh region, the
Svinord, Ilmen and Buregi Formations of the Main Devonian Field and in the Kultuban
Limestone of the southern Urals. Nalivkin regarded all these formations as middle
Frasnian. However, Malakhova (1960; 1961) has shown that the foraminifera of the
Kultuban Formation at Lake Kultuban indicate that the formation includes both
younger and older beds.

A. E. H. PEDDER: THE UPPER DEVONIAN GASTROPOD ORECOPIA 145

Platyschisma kirchholmiensis, the second of Keyserling’s (1846, p. 264, pi. 11, fig. 7)
species now referred to Orecopia, came from the Diina (Zap Dvina) River, near Kirch-
holm, in Livland (Latvia). Keyserling considered the age of P. uchtensis and P. kirch-
holmiensis to be ‘analagous’. Stuckenberg (1878, pp. 469, 470) placed Rotella micro-
stoma Eichwald (1860, p. 1163) in synonymy with P. kirchholmiensis and claimed a wide
distribution for the species in the southern part of the main Devonian Field and in the
Voronezh region. Little is published concerning the exact horizon of these occurrences,
but it seems likely from Pacht’s (1859) and other work that the species is middle Fras-
nian in the type region.

Without access to specimens, the generic identity of PlatyschismaP ambiguum (Wal-
cott 1884, pp. 188, 189, pi. 17, fig. 3) remains in doubt. Walcott regarded it as con-
generic with the type species of Orecopia, but the species was not mentioned by Knight in
his discussion of the genus. It was originally cited as coming from the ‘upper horizon
of the Devonian limestone, The Gate, Eureka District’. The wording suggests that the
type horizon is in the ‘ Cyrtospirifer ’ Zone of the Devils Gate Formation (Merriam 1940,
p. 51). This zone is a correlative of the Alexo and Sassenach Formations of Alberta and
is early Famennian, consequently if P. ambiguum is an Orecopia, it would be the youngest
of the species ascribed to the genus.

Orecopia cotei sp. nov.

Plate 21, figs. 2, 6-9, 11, 12, 14; text-fig. 2b
Name derivation. From L. F. Cote.

Material. Holotype, GSC 17633; paratypes 1-3, GSC 17634-6 respectively. The holotype and para-
types 1, 2 were collected by L. F. Cote in 1959 from an unnamed grey, medium-grained limestone,
140 ft. below the Besa River Formation, near Nabesche River, British Columbia, lat. 56° 17' N., long.
123° 23V W. The gastropods were accompanied by a typical late Frasnian assemblage including
Phillipsastrea exigua sensu Smith 1945 non Lambe 1901, P. vesiculosa Smith, Hexagonaria sp., Thanmo-
phyllum sp., ‘ PtychophyllunT kindlei Smith, Thamnopora sp., Devonopro ductus sp., Atrypa sp., Spina-
trypa sp., and Theodossia or Vandergrachtella sp. From small collections made 905 and 955 ft. lower in
the same section, D. J. McLaren has identified the early Frasnian rhynchonelloid Leiorhynchus russelli
McLaren.

Paratype 3 was collected by the author in 1958 from the Tathlina Formation on Hay River, opposite
mile 59T of the original Mackenzie Highway, Northwest Territories. It occurred in a fawn, fine-
grained limestone containing fine detrital quartz and was accompanied by Nervostrophia sp., Devono-
pro ductus sp., Atrypa ciliipes Crickmay, Atrypa sp., Tenticospirifer sp., and Theodossia or Vandergrach-
tella sp. As discussed above, the precise age of the fauna is in doubt, but is either late Middle, or early
Late Frasnian.

Diagnosis. Umboniform species of Orecopia having a mean spire angle of about 110°.
Periphery low in the whorl section and carinate. Growth-lines with a very marked for-
ward projection at the periphery.

Description. Shell umboniform, having a mean spire angle of about 1 10° after the
formation of five whorls. In early stages the umbilicus is relatively wide, but is reduced
during growth, so that the shell is narrowly, or even minutely phaneromphalus in late
stages. Protoconch unknown. Sutures very shallow. The whorl section is gently arched
from the suture to the periphery and immediately under the periphery, but is almost

L

B 6612

146

PALAEONTOLOGY, VOLUME 9

flat along much of the base; the periphery is marked by a distinct, although variable
carina; the adaxial region of later whorls is apparently much thickened, however the
interior of the sectioned paratype is poorly preserved, and the later whorls in text-fig. 2b
are liberally reconstructed in the adaxial region. Growth-lines are re-entrant below the
suture giving rise to a broad asymmetrical sinus on the upper abaxial surface of the
whorl. Towards the periphery they project forward prominently, but immediately under

text-fig. 2. Cross-sections of Orecopia. Both x2. a. Orecopia mccoyi (Walcott), GSC 17638.
Collected by R. H. Waines from the Sultan Formation in the Spring Mountains, Goodsprings Quad-
rangle, southern Nevada, b. Orecopia cotei sp. nov., paratype 1, GSC 17634. Unnamed late Frasnian
limestone, 140 ft. below base of the Besa River Formation, near Nabesche River, British Columbia.

the periphery are abruptly swept back again; another more gentle projection is present
near the umbilicus; thus over the base the growth-lines are sigmoidal. The largest speci-
mens seen have about five whorls and are between 20 and 25 mm. in diameter and
between 10 and 12 mm. high.

Comparisons. The new species resembles Orecopia uchtensis more than O. kirchhol-
miensis or O. mccoyi. The spire angle is similar in both species, but the whorl profile
and growth-lines are different; in O. cotei the periphery is more acute and carinate, is
also nearer the base, and the forward projection of the growth-line at the periphery is
much more pronounced.

Orecopia kirchholmiensis has a higher spire and more rounded whorl periphery.

Most specimens of Orecopia mccoyi differ considerably from the new species by having
a much higher spire, deeper sutures, and a well-marked angulation on the upper abaxial
surface. The specimen shown in text-fig. 2a is a low-spired example of O. mccoyi and yet
the mean spire angle of about 95° is distinctly less than that of O. cotei.

All figures X 2.

Figs. 1, 3, 4. Orecopia mccoyi (Walcott). GSC 17637, collected by R. H. Waines from the Sultan
Formation (early or middle Frasnian), Spring Mountains, Goodsprings Quadrangle, southern Nevada.

Figs. 2, 6-9, 11, 12, 14. Orecopia cotei sp. nov. 2, 8, 9, 12, holotype, GSC 17633, unnamed limestone
(late Frasnian), Nabesche River region, north-eastern British Columbia. 6, 11, 14, paratype 1
(before sectioning), GSC 17634, same horizon and locality as the holotype. 7, paratype 3, GSC
17636, lower part of the Tathlina Formation (middle or late Frasnian), Hay River, Northwest
Territories.

Figs. 5, 10, 13. Orecopia uchtensis (Keyserling). Topotype? USNM 60907, Ukhta River, Timan. An
early label indicates that the specimen was originally in the Museum of the Mining School at St.
Petersburg and gives the age, presumably mistakenly, as Lower Devonian.

a

b

EXPLANATION OF PLATE 21

Palaeontology , Vol. 9

PLATE 21

■_ ' -/A

/-'•jU

n

12

13

14

PEDDER, Upper Devonian Orecopia

A. E. H. PEDDER: THE UPPER DEVONIAN GASTROPOD ORECOPIA

147

REFERENCES

Carlisle, d., murphy, m. a., nelson, c. a., and winterer, E. l. 1957. Devonian stratigraphy of Sulphur
Springs and Pinyon Ranges, Nevada. Bull. Artier. Ass. Petrol. Geol. 41, 2175-91.

eichwald, c. e. 1860. Lethaea rossica oil paleontologie de la Russie, decrite et figuree, 1, Stuttgart (not
seen).

hewitt, d. F. 1931. Geology and ore deposits of the Goodsprings Quadrangle, Nevada. Prof. Pap.
U.S. geol. Surv. 162.

house, m. r. and pedder, a. e. h. 1963. Devonian goniatites and stratigraphical correlations in western
Canada. Palaeontology, 6, 491-539, pi. 70-77.

Johnson, m. s. and hibbard, d. e. 1957. Geology of the Atomic Energy Commission Nevada proving
grounds area, Nevada. Bull. U.S. geol. Surv. 1021-K, 333-84, pi. 32, 33.

kellog, h. e. 1963. Paleozoic stratigraphy of the southern Egan Range, Nevada. Bull. geol. Soc. Amer.
74, 685-708, pi. 1-4.

keyserling, A. 1846. Geognostiche Beobachtungen, in Wissenschaftliche Beobachtungen auf einer Reise
in das Petschora-Land, im Jahre 1843, 149-406. St. Petersburg.

knight, j. b. 1945. Some new genera of Paleozoic Gastropoda. J. Paleont. 19, 573-87, pi. 79, 80.

Malakhova, N. p. 1960. The problem of the age of the Koltuban Limestones in the southern Urals (in
Russian). Doklady Akad. Nauk. S.S.S.R., 132 (translated Amer. geol. Inst. 132 (1961), 486-8).

mclaren, d. j. 1962. Middle and early Upper Devonian rhynchonelloid brachiopods from western
Canada. Bull. geol. Surv. Can. 86.

merriam, c. w. 1940. Devonian stratigraphy and paleontology of the Roberts Mountains region
Nevada. Spec. Pap. Geol. Soc. Amer. 25.

1963 a. Paleozoic rocks of Antelope Valley, Eureka and Nye Counties, Nevada. Prof. Pap. U.S.

geol. Surv. 423.

19636. Geology of the Cerro Gordo mining district Inyo County California. Ibid. 408.

NALiVKiN, d. v. (ed.) 1947. Atlas of the guide forms of the fossil faunas of the U.S.S.R. 3, Devonian
(in Russian). Moscow and Leningrad.

nolan, t. b. 1935. The Gold Hill mining district Utah. Prof. Pap. U.S. geol. Surv. 177.

, merriam, c. w., and williams, j. s. 1956. The stratigraphic section in the vicinity of Eureka,

Nevada. Ibid. 276.

pacht, r. 1859. Der devonische Kalk. in Livland. Arch. Naturk. Liv-, Ehst-u. Kurt. ser. 1, 2, 249-98
(not seen).

roemer, c. f. 1885. Lethaea erratica oder Aufzahlung und Beschreibung der in der norddeutschen
Ebene vorkommenden Diluvial-Geschiebe nordischer Sedimentar-Gesteine. Palaeont. Abh. 2 (5),
250-450, pi. 1-11.

stuckenberg, a. a. 1878. The Devonian basin of European Russia (in Russian). Trud. St. Peter sbourg
Obshch. estestv. 9, 447-93, pi. 1, 2.

tschernyschew, T. n. 1884. Materialien zur Kenntniss der devonischen Ablagerungen in Russland.
Mem. Com. geol. St. Petersb. 1 (3).

1887. Die Fauna des mittleren und oberen Devon am West-Abhange des Urals. Ibid. 3 (3).

walcott, c. D. 1884. Paleontology of the Eureka District. Monogr. U.S. geol. Surv. 8.

A. E. H. PEDDER

Department of Geology,
University of New England,
Armidale, N.S.W.,

Manuscript received 29 January 1965 Australia

THE SILURIAN RUGOSE CORAL MICROPLASMA
LOVENIANUM DYBOWSKI FROM
MONMOUTHSHIRE

by D. E. WHITE

Abstract. Specimens of Microplasma lovenianum Dybowski 1874 from the Wenlock Limestone of the Usk
Inlier, Monmouthshire, are described and illustrated. This is the first record of the occurrence of Microplasma
in Britain. The diagnosis of the genus is re-stated and a diagnosis of M. lovenianum is also given.

Dybowski (1873, p. 340) erected the genus Microplasma for fasciculate rugose corals,
the septa of which are represented by thorn-like projections (trabeculae). The following
species, all of Silurian age, were described and figured (pp. 508-14, pi. 5, figs. 3-6):

Microplasma gotlandicum
Microplasma lovenianum
Microplasma sclimidti
Microplasma pectiniseptatum

from the Baltic islands of Gotland and Karlso
from Karlso
from Karlso
from Gotland

Lang and Smith (1927, p. 478, footnote), in rejecting the views of Schliiter (1889) and
Pocta (1902) concerning Cystiphyllum siluriense Lonsdale and C. cylindricum Lonsdale,
state that they ‘recognise Microplasma for compound Cystiphyllum ’.

Wedekind (1927, pp. 63-64) considered Microplasma to be a dwarf form of Cysti-
phyllum, in which only the central part of the internal structure is present. He selected
M. gotlandicum as the type species. This was the first valid selection.

Prantl (1940) rejected Wedekind’s views. He slightly amended Dybowski’s diagnosis
of Microplasma to include the presence of ‘spinose ridges on successive dissepimental
floors’ as well as in the peripheral stereozone. He wrongly considered it to be con-
generic with Aphyllum Soshkina 1937. The tabulae are complete and widely spaced in
Aphyllum, affinities of which are closer to the family Tryplasmatidae Etheridge 1907.
In the same paper, Prantl described and figured Microplasma flexuosum from the Upper
Wenlock, eastern part of the quarry ‘America’, near Budnany, Bohemia. He considered
this species to be ‘very closely allied with M. schmidti, with which it is perhaps con-
specific’. Prantl suggested that M. schmidti and M. gotlandicum ‘represent only the
different ontogenetical stages of growth of the same species’.

In October, 1961, during the resurvey of the One-Inch Newport (249) Sheet by the
Geological Survey, Mr. J. D. D. Smith and the author visited a Wenlock Limestone
quarry [National Grid Reference — ST/33989839], 550 yd. E. 13° S. of the church.
Common Coed-y-Paen, Monmouthshire. Among the fossils collected were two speci-
mens of parts of large fasciculate coral colonies, later identified as M. lovenianum. They
are described below. A brief note on this occurrence has already been published (White
1965).

[Palaeontology, Vol. 9, Part 1, 1966, pp. 148-51, pi. 22.]

D. E. WHITE: MICROPLASMA LOVENIANUM DYBOWSKI

149

SYSTEMATIC PALAEONTOLOGY

Family cystiphyllidae Milne Edwards and Haime
Genus microplasma Dybowski, emend. Prantl

Type species. M. gotlandicum Dybowski; Wedekind 1927

Diagnosis (after Prantl 1940): Rugose coral with phaceloid corallum of thin, cylindrical,
scolecoid corallites. Septa represented by trabeculae projecting internally from a narrow
peripheral stereozone ; also present on upper surfaces of horizontal skeletal elements.
Dissepiments and tabellae not always separately recognizable.

Microplasma lovenianum Dybowski 1874
Plate 22, figs. 1-11

1874 Microplasma lovenianum Dybowski, pp. 510-11, pi. 5, figs. 4, 4 a

Material. GSM 102583, 102583a-c, and PF 2903-4; parts of a colony from the lower part of the section;
GSM 102584 and 102584a-c, parts of a colony from a loose block. Both specimens are from a small
quarry, 550 yd. E. 13° S. of the church, Common Coed-y-Paen, Monmouthshire (N.G.R. ST
33989839). They are in the Geological Survey and Museum collections, London.

Diagnosis. Microplasma in which dissepiments and tabellae are numerous, closely
spaced, irregular, and of very variable size. Peripheral increase.

Description. Corallum phaceloid. Corallites upright, scolecoid, closely spaced, cylin-
drical to sub-cylindrical, with occasional deformation when in contact (PI. 22, fig. 3).
Length of protocorallites exceeds 40 mm.; maximum diameter 10 mm. Five hystero-
corallites, averaging 3-5 mm. diameter, forming in calice of protocorallites, at 8-5-10
mm. diameter, by characteristic peripheral increase (PI. 22, figs. 6, 7). Calice steep-sided,
depth variable up to 9 mm. Epitheca worn, not clearly seen, thin with trace of transverse
and vertical ribbing (PI. 22, figs. 10, 11). Peripheral stereozone, thickness 0-1-0-15 mm.,
from which trabeculae project internally (PI. 22, figs. 4, 5). Internal vesicular structure
of small, steeply dipping dissepiments, convex upwards, and incomplete tabulae
(tabellae), flat-lying, convex upwards, occasionally difficult to distinguish, with trabe-
culae present as small, thorn-like projections (PI. 22, fig. 4).

Discussion: The genus Microplasma has not previously been recorded from Britain. The
specimens from the Usk Inlier agree very closely with Dybowski’s illustrations of M.
lovenianum from the Baltic island of Karlso, which are reproduced on Plate 22, figs. 1, 2,
and with his description of the species, except that he did not mention trabeculae on the
horizontal skeletal elements. However, Wedekind (1927, pi. 19, fig. 6) figured a specimen
of Microplasma sp. with well-developed trabeculae in this position, and Prantl (1940,
p. 105), in his re-definition of the genus, mentioned ‘septa reduced to mere spinose
ridges on successive dissepimental floors and especially on the periphery ingulfed in
secondary tissue’. Therefore it is considered that the presence of trabeculae on the
internal vesicular tissue of the specimens from the Usk Inlier does not justify the erection
of a new species.

The characteristic peripheral increase of M. lovenianum is well seen. In GSM 102583
(PI. 22, fig. 6), neo-tissue has developed in two calices and the resulting five hystero-

150

PALAEONTOLOGY, VOLUME 9

corallites are present in each case in GSM 102583c (PI. 22, fig. 7). This process takes
place quite rapidly as the two sections are approximately 3 mm. apart. This type of
increase has been fully described and illustrated by Smith and Tremberth (1929, pp.
370-2, text-fig. 2).

From M. schmidti , M. gotlandicum, and M. flexuosum, the specimens from the Usk
Inlier differ in having a smaller internal vesicular structure. Dybowski’s remaining
species, M. pectiniseptatum, is a form more closely related to the genus Hedstroemo-
phyllum Wedekind 1927 than to Microplasma.

AGE OF THE GENUS MICROPLASMA

Species of Microplasma described by Dybowski (1874) were from the ‘Silurformation’
of Gotland and Karlso. No horizon details were given. Wedekind (1927, p. 84) mentioned
that Microplasma is abundant in his middle and upper divisions of the ‘Mittelgotlan-
dium’ of Gotland, from the upper division of which he figured Microplasma sp. (pi. 19,
fig. 6). M. flexuosum Prantl 1940 occurs in the ‘Motoly Beds, ea2 (Upper Wenlock)’,
near Budnany, Bohemia. Prantl (1940, p. 108) concluded that ‘ Microplasma seems to be
restricted to the Silurian (Wenlock-Lower Ludlow)’, but suggested the possibility of an
extension of its range into the Devonian.

The specimens of M. lovenianum from the Usk Inlier occurred within a few feet of the
top of the Wenlock Limestone, a horizon which probably corresponds closely to those
of the Gotlandian and Bohemian occurrences.

Acknowledgements. I record my thanks to the following: Dr. F. W. Anderson and Mr. M. Mitchell for
reading through the manuscript of the paper and help during its preparation; Dr. H. Dighton Thomas
for his helpful advice; publication is by permission of the Director, Geological Survey and Museum.

REFERENCES

dybowski, w. n. 1873-4. Monographic der Zoantharia sclerodermata rugosa aus der Silurformation
Estlands, Nord-Livlands and der Insel Gotland, . . . 257-532, pi. 1-5. Dorpat. Also published in

EXPLANATION OF PLATE 22

Microplasma lovenianum Dybowski 1874. Photographs by Mr. C. A. Friend, Geological Survey and
Museum.

Figs. 1, 2. Reproduction of Dybowski’s illustrations, 1874, pi. 5, figs. 4, 4 a (fig. 4a said to be X in
error, p. 530), from ‘der Silurformation’ of Karlso.

Figs. 3-8, parts of GSM 102583, from the Wenlock Limestone, Usk Inlier. 3, part of GSM 102583a.
Transverse section showing deformation of corallites in contact. 4, PF 2903. Transverse section
showing trabeculae on horizontal skeletal elements and projecting internally from peripheral
stereozone. 5, PF 2904. Longitudinal section showing disposition of horizontal skeletal elements
and trabeculae projecting from peripheral stereozone. 6, 7, part of GSM 102583 and of 102583a.
Transverse sections, approximately 3 mm. apart, showing the development of neo-tissue in two
calices (top middle of 6), and the resulting hysterocorallites, five from each calice (top middle of 7).
8, GSM 102583c. Longitudinal section showing hysterocorallites developing from calices of proto-
corallites.

Figs. 9-11, parts of GSM 102584. Horizon and locality as for figs. 3-8. 9, GSM 102584. As for fig. 8.
10, 11, parts of GSM 102584c. Fragments of epitheca, poorly preserved; faint longitudinal ribbing
and transverse wrinkles.

Figs. 1-3, 8, 9, x 1, figs. 4-7, x2, figs. 10, 11, x5.

Palaeontology , Vol. 9

PLATE 22

WHITE, Silurian rugose coral Microplasma lovenianum

D. E. WHITE: MICROPLASMA LOVENIANUM DYBOWSKI 151

Archiv Naturkunde Liv-, Ehst- und Kurlands, (1), 5, (3), 1873, 257-414, pi. 1, 2; 1874, 415-532,
pi. 3-5.

lang, w. d. and smith, s. 1927. A critical revision of the rugose corals described by W. Lonsdale in
Murchison’s ‘Silurian System’. Quart. J. geol. Soc. London, 83, 448-91, pi. 34-37.
pocta, p. 1902. Anthozoaires et Alcyonaires, in J. Barrande, Systeme Silurien du Centre de la Boheme,
8,(2), viii+ 1-347, pi. 20-118.

prantl, f. 1940. Some Silurian Rugose Corals from Bohemia. Ball, internat. Acad. Sci. Boheme.
105-8.

schluter, c. 1889. Anthozoen des Rheinischen Mittel- Devon. Ab/t. Geol. Specialkarte Preussen. 8,
269-465, pi. 1-16.

smith, s. and tremberth, r. 1929. On the Silurian corals Madreporites articulatus, Wahlenberg, and
Madrepora truncata, Linnaeus. Ann. Mag. nat. Hist., ser. 10, 3, 361-76, pi. 7, 8.
soshkina, E. d. 1937. Corals of the Upper Silurian and Lower Devonian of the eastern and western
slopes of the Urals. Trav. Inst. Paleozool. Acad. Sci. U.R.S.S. 6, (4), 112 pp., 21 pi. [In Russian with
English summary. ]

wedekind, r. 1927. Die Zoantharia Rugosa von Gotland (bes. Nordgotland) nebst Bemerkungen zur
Biostratigraphie des Gotlandicum. Sverig. geol. Unders., ser. Ca, no. 19, 1-94, pi. 1-30.
white, d. e. 1965. In Summ. Progr. geol. Surv. G.B. for 1964.

D. E. WHITE

Geological Survey and Museum,
Exhibition Road,

Manuscript received 2 Eebruary 1965. London, S.W. 7.

MIDDLE DEVONIAN BRACHIOPODS FROM THE
ROBERTS MOUNTAINS, CENTRAL NEVADA

by J. G. JOHNSON

Abstract. Middle Devonian stratigraphy in the Roberts Mountains is briefly reviewed and Middle Devonian
limestone strata below the Devils Gate Limestone are assigned to a brachiopod-rich formation, the Denay
Limestone (new name). Walcott’s Skenidium devonicum is made type species of a new dalmanellid genus
Vallomyonia and his Cryptonella ? circula is assigned to Leptathyris. The brachiopods of the Leplathyris circula
zone (new name) and the Pentamerella subzone in the lower part of the Denay Limestone are described and
illustrated and their palaeontologic evidence suggests an Eifelian assignment. Fifteen species are described,
fourteen from the Leplathyris circula zone in the Roberts Mountains. Three new species and one new subspecies
are Pentamerella wintered, Hadrorhynchia eurekaeitsis, Echinocoelia denayensis, and Warrenella kirki praekirki.
Cyrtinacea is proposed as a superfamily.

In 1940 C. W. Merriam published the results of his stratigraphic and palaeontologic
studies of the Devonian in central Nevada (text-fig. 1). The stratigraphic column known
to comprise Devonian rocks was divided into two formations, the Nevada Formation
below and the Devils Gate Limestone above, and both were zoned largely on the basis
of brachiopods and corals. The Nevada Formation has since been the subject of addi-
tional stratigraphic studies (Merriam and Anderson 1942; Nolan, Merriam, and Williams
1956) and the term is commonly applied to beds in Eureka County between the Lone
Mountain Dolomite and the Devils Gate Limestone.

Merriam’s zonal scheme (1940, p. 9) was incomplete in that he recognized the presence
of an unzoned gap between the ‘ Spirifer ’ pinyonensis zone and the higher Warrenella
(= Martinia ) kirki zone. Inasmuch as Merriam at that time considered the pinyonensis
zone to be Middle Devonian, the gap appeared to be an insignificant one within the
Middle Devonian.

With Merriam’s work as a background, the writer in 1957 and 1958 mapped and
collected fossils at significant Devonian sections in the northern Simpson Park Range.
Study of the brachiopods in these collections as well as in many obtained in the Roberts
Mountains by Professors E. L. Winterer and M. A. Murphy and, in addition, collections
from the Sulphur Spring Range made by Carlisle, Murphy, Nelson, and Winterer (1957)
led to the recognition of the new faunal zone, based on brachiopods, which filled the
gap between the previously defined pinyonensis and kirki zones (Johnson 1962a). The
fossiliferous beds were designated Athyris circula zone and Pentamerella subzone. The
occurrence of these two brachiopod assemblages directly on beds with the pinyonensis
zone fauna allowed a comparison for the first time and resulted in the recognition of an
almost complete faunal discontinuity between the pinyonensis zone below and the circula
zone above. These studies also led Johnson ( 1962a, p. 166) to conclude that the pinyonen-
sis zone is Lower Devonian (Emsian) rather than Middle Devonian as previously thought
(Merriam 1940, p. 53). The Lower Devonian assignment has since been corroborated by
a restudy of the pinyonensis zone ammonoids (House 1962, p. 253), one of which is
known to come from low in the zone. The brachiopod evidence, only touched upon
earlier (Johnson 1962a, p. 166), has been thoroughly documented and discussed in

[Palaeontology, Vol. 9, Part 1, 1966, pp. 152-81, pis. 23-28.]

J. G. JOHNSON: MIDDLE DEVONIAN BRACHIOPODS

153

a more extensive monograph (Johnson in preparation). Studies of conodonts also
indicate an Emsian assignment for the pinyonensis zone (Walliser in Johnson in
preparation).

text-fig. 1. Eureka County and environs, central Nevada. Shaded portion of the inset map of Nevada

delineates the area shown.

The faunal discontinuity recognized in the northern Simpson Park Range and in the
northern Roberts Mountains through collections made by the writer at these localities
proved to coincide with the lithologic break between argillaceous limestone of the
pinyonensis zone and overlying ledge-forming outcrops of siliceous, platy, and flaggy
grey limestone with the Leptathyris circula zone fauna. In a second paper (Johnson
19626) the writer attempted to trace the lithologic break eastward from the fossiliferous

154

PALAEONTOLOGY, VOLUME 9

sections. In the limestone sections the break had already proved to be a significant one
for geologic mapping and beds below it were named the McColley Canyon Member of
the Nevada Limestone by Carlisle and others (1957, p. 2181). Johnson (19626, p. 543)
recommended that in view of coincident faunal and lithologic breaks at the top of the
McColley Canyon that the surface of the discontinuity probably indicated unconformity
and recommended that the McColley Canyon be elevated to formational rank.

DENAY LIMESTONE

Application of the name McColley Canyon Formation to the Lower Devonian rocks
between the Lone Mountain Dolomite below and the Leptcithyris circula zone beds at
the base of the Middle Devonian section above, left unnamed Middle Devonian lime-
stones present in the northern Simpson Park Range and in the Roberts Mountains.
Neither the Sulphur Spring Range terminology of Union Mountain Sandstone and Tele-
graph Canyon Dolomite (Carlisle et al. 1957), nor Simonson Dolomite, extensively
applied to Middle Devonian rocks east of Eureka County, is applicable to these sections.
Neither is Nolan, Merriam, and Williams’s member division of the upper Nevada, into
Sentinel Mountain Dolomite, Woodpecker Limestone, and Bay State Dolomite appli-
cable in this area on either a member or formation basis. To eliminate this deficiency the
writer proposes the name Denay ( Day-nee) Limestone as a formation composing Middle
Devonian limestone strata in the northern Simpson Park Range and in the Roberts
Mountains (principally northern Roberts Mountains) that together with the underlying
McColley Canyon Formation constitute the Nevada Group in this area. The name
comes from Denay Valley which separates the north-eastern tip of the Simpson Park
Range from the north flank of the Roberts Mountains. According to Winterer (oral
communication 1961) the Denay Limestone interfingers with Simonson-like dolomites
in the southern Roberts Mountains. The type section of the new formation, approxi-
mately 1,000 ft. thick, is on the east flank of Willow Creek Canyon on the north side of
Roberts Mountains upsection from the point where the Lone Mountain-McColley
Canyon Formation contact crosses Willow Creek (elev. 6,570 ft.) due east to an elevation
of approximately 7,320 ft., west of the ridge that is capped by Devils Gate Limestone
east of Willow Creek. The contact between the top of the McColley Canyon Formation
and the base of the Denay Limestone is met at approximately 6,840 ft. in this section.
An excellent supplementary section is exposed in the J-D window of the northern
Simpson Park Range between Red Hill (sec. 17, T. 25 N., R. 50 E.) where the Denay is
in contact with the overlying Devils Gate Limestone, and a point about a mile to the
south-west where a small section of the upper McColley Canyon Formation is present
at its base. This is the same as the section of ‘unnamed limestone’ depicted in the column
for J-D Window in an earlier paper (Johnson 19626, fig. 3).

AGE AND CORRELATION

Before outlining the evidence from the brachiopods it will be useful to discuss the
occurrence of a goniatite-bearing collection from the Denay Limestone in the southern
Roberts Mountains that bears on the problem of the age of the Leptathyris circula zone.
According to Winterer (oral communication 1959), who has mapped the Palaeozoic rocks

J. G. JOHNSON: MIDDLE DEVONIAN BRACHIOPODS

155

in the Roberts Mountains, the stratigraphic unit here named Denay Limestone consists
of a lower well-bedded limestone member and an overlying thick-bedded limestone
member in its lower half as well as one or two higher members with which we are not
concerned here. So far as the writer has been able to observe personally in the field and
deduce from examination of collections made by others, the Leptathyris circula zone
includes beds that lie within the lower well-bedded unit. In 1 959, Winterer made a collec-
tion (loc. 4745) in the southern Roberts Mountains about 200 ft. above the lower well-
bedded unit that is rich in brachiopods, pelecypods, several types of gastropods, and in
addition several specimens of the goniatite Cabrieroceras aff. crispiforme (Kayser); the
latter identification being made for the writer by Dr. M. R. House (written communica-
tion 1963 and House 1965). According to House, C. crispiforme itself is the guide fossil
to the basal Givetian zone to which it gives its name and the occurrence of specimens in
the Roberts Mountains designated C. aff. crispiforme suggests that the containing beds
be regarded as basal Givetian. Thus, according to the goniatite evidence, the Leptathyris
circula zone lies, apparently in its entirety, below beds of earliest Givetian age and either
is Eifelian or is itself very early Givetian in age.

Twenty-three species of brachiopods are recognized in the L. circula zone including
its basal Pentamerella subzone and of these fourteen are described and illustrated in this
paper:

Vallomyonia devonica, ‘ Schuchertella ’ sp., Pentamerella wintered, Gypidula cf. re-
currens, Hadrorhynchia eurekaensis, Leiorhynchus sp., Spinatrypa sp. B, Vagrania? sp.,
Dubaria? cf. thetis, Leptathyris circula, Warrenella kirki praekirki, Warrenella cf.
columbina columbina, Echinocoelia denayensis, Cyrtina sp.

Species of Teichertina, Schizophoria, Mesopliolidostrophia? , Productella, Spinatrypa,
Anatrypa, Nucleospira, Meristina?, and Quadrithyris are represented by inadequate
material. The occurrence OH Vallomyonia devonica at present has no value for correlation
since the genus is unknown outside of Nevada. ‘ Schuchertella ’ is fairly common in Lower
and Middle Devonian rocks and is of no help in precise correlation. Pentamerella
wintered is most like species of Eifelian age in the Ural Mountains (Khodalevich 1951)
and differs quite strongly from the myriad of species of the Hamilton-Traverse rocks of
the eastern United States. Gypidula cf. recurrens is perhaps identical with Meyer’s species
which occurs in the Blue Fiord Formation of Ellesmere Island. According to McLaren
(in Fortier et al. 1963, p. 61) the Blue Fiord Formation is of Eifelian age. Hadrorhynchia
is typical of Stringocephalus-beaving beds in north-west Canada, but the species from
the Denay Limestone is a new one. The same may be said for Leiorhynchus since the
species apparently has not yet been reported beyond central Nevada. Spinatrypa sp. B
is probably a new species and is quite unlike typical Spinatrypa species that are fairly well
known in the Givetian of western Canada. Both Vagrania and Dubaria cf. thetis are
questionably assigned and so should be given little weight in correlation. If the latter
identifications could be placed on firmer ground they would probably indicate an
Eifelian age for the containing beds in Nevada since they are most common in Lower
Devonian and Eifelian in the Old World. Leptathyris, represented by L. circula, is a
recently defined athyridid most common in the Eifelian, but present in the Lower
Devonian and inasmuch as its definition is so recent its stratigraphic range is probably
insufficiently known on which to base a firm correlation. Warrenella kirki praekirki is
of little help in distinguishing Eifelian from Givetian. Similar forms are typically Middle

156

PALAEONTOLOGY, VOLUME 9

Devonian. Warrenella cf. columbina columbina compares very favourably with that sub-
species as illustrated by Havlicek (1959, pi. 28, fig. 3) from the Acanthopyge Limestone
(upper Eifelian) of central Czechoslovakia. Echinocoelia denayensis differs considerably
from the Tully species E. anibocoelioides Cooper and Williams, but is fairly close to E.
incurva Cooper and Williams, from the Mottville Member at the top of the Marcellus
(or base of the Skaneateles) in New York State. According to the goniatite evidence dis-
cussed by House (1962, p. 254) at least part of the Marcellus is Givetian and E. incurva
would be of earliest Givetian age. E. denayensis is also close to E? rogeri (= Anibocoelia
rogeri Drot 1955, p. 565, pi. 27, fig. 3 a-d) which is reported to be of late Eifelian (Cou-
vinian) age. Cyrtina is common throughout the Devonian and so is of no help in precise
correlation.

In summary, the brachiopods appear to have their closest relations to brachiopod
species from beds elsewhere assigned to the Eifelian, possibly high Eifelian, and some
lesser affinity with Givetian species. The goniatite occurrence also suggests that beds
bearing the L. circula fauna are older than early Givetian.

There seems to be little reason to pursue the possibility that beds within the L. circula
zone could be Lower Devonian inasmuch as the underlying ‘ Spirifer" pinyonensis zone
fauna is completely different and is very high Lower Devonian, Emsian, and possibly
upper Emsian (Johnson in preparation). It is suggested, therefore, that the combined
evidence is sufficient to indicate a tentative Eifelian assignment for the L. circula
zone.

The stratigraphic occurrence of the L. circula zone south of the area studied by the
writer in the Simpson Park Range and Roberts Mountains can be deduced from data
published by C. W. Merriam. Merriam (written communication 1958) indicated that
forms identical to those listed as L. circula in this paper occur in the Tower Nevada
Formation’ at the north end of the Antelope Range. Merriam (1940, p. 23; 1963, p. 45,
table 2) includes some 718 ft. of strata in his Tower Nevada’ unit at Lone Mountain.
However, he notes that the lower 530 ft. is much more fossiliferous than the remaining
part of the Tower Nevada’. Probably, beds between 530 and 718 ft. above the base of
the Nevada at Lone Mountain belong to the L. circula zone and taken in conjunction
with Merriam’s recognition of L. circula in his Tower Nevada’ in the Antelope Range,
indicates that the views of the writer and of Merriam are in essential harmony regarding
what part of the fossiliferous western sections is equivalent to the Oxyoke Canyon Sand-
stone to the east. Johnson (19626, fig. 3) has already concluded that the Union Mountain
Sandstone (= Oxyoke Canyon Sandstone) is at least partially equivalent to the L. circula
zone, an identical position as that which would be indicated by the presence of the L.
circula zone at the top of Merriam’s Tower Nevada’ (1963, p. 45, table 2).

Outside of central Nevada the only brachiopod faunas that appear to bear any
resemblance to that of the L. circula zone are reported from north-western Canada and
from the Canadian Arctic Archipelago. Few of the significant forms of the L. circula
zone have been reported in these areas although ‘ Athyris ’ circula was reported from the
upper Hume Formation of the MacKenzie Basin by Warren and Stelck (1962, p. 278).
The species Gypidula cf. recurrens may prove to be significant for correlation in the
Eifelian in western North America and in the Arctic since it is reported from the Blue
Fiord Formation of Ellesmere Island. According to Dr. A. W. Norris of the Geological
Survey of Canada (written communication, December 1964), the specimens referred

J. G. JOHNSON: MIDDLE DEVONIAN BRACHIOPODS

157

to here as Gypidula cf. recurrens compare most closely with a form present in the upper
Blue Fiord Formation (GSC loc. 26525). Crickmay’s report (1962, p. 3) of G. cf .pseudo-
galeata from the Bear Rock Formation (called by him Lower Devonian) may be an
occurrence of a similar form. Crickmay reported the species to be present with Fimbri-
spirifer, a form present in the Blue Fiord, but not present in Lower Devonian faunas of

N. ROBERTS MOUNTAINS

EUREKA DISTRICT

LlI

o

F-

GO

LjJ

I

>-

<

z

LlI

O

Warrene/la
oc ci dental is
Zone

Unzoned beds with
Schizophoria mcfar/anei
Stringocepha/us, Ladjia , /

Subrensse/andia, "Atrypa
cf. insquamosa, 8
Hadrorhynchia sandersom

Warrene lla kirki
Zone

Leptathy ris circula
Zone

a

Pentamere/ /a
Subzone

CT

<

CL

cr

LlI

Q.

CL

=)

<

o

<

>

LlI

Z

CD

o

CD

Cl

TD

O

O

Bay State
Dolomite with
S tring o cep ha /us

Beds with Stringocepha-
/us and " Rensse/andia"

Beds with Schizophoria
mcfarianei 8 large Airypa

Beds with Warrene/la kirki 8
Leiorhynchus nevadensis

Bedswith small W cf. kirki
8 small L. cf. nevadensis

Sentinel Mountain
Dolomite

Oxyoke Canyon
Sandstone

text-fig. 2. Suggested local correlation between the brachiopod-rich Denay Limestone on the north
side of the Roberts Mountains and the sequence of mainly dolomite members of the Middle Devonian
part of the Nevada Formation in the Eureka district (Nolan et al. 1956). ‘ Rensselandia ’ in the Eureka
district was examined on loan from Dr. C. W. Merriam after this paper went to press. Merriam’s
specimens are large and well preserved and upon preparation prove to belong to Subrensse/andia.

central Nevada studied by Johnson (in preparation). More adequate comparisons of
correlative faunas must await description and illustration of more of the oldest Middle
Devonian species in Canada.

The stratigraphic position of the L. circula zone and Pentamerella subzone is depicted
in text-fig. 2 which shows the relations of the assemblage zones and undefined faunal
units of the Denay Limestone on the north side of the Roberts Mountains in comparison
with the faunal sequence of the upper part of the Nevada Formation of the Eureka
District as reported by Nolan, Merriam, and Williams (1956, p. 47).

158

PALAEONTOLOGY, VOLUME 9

SYSTEMATIC PALAEONTOLOGY

Phylum BRACHIOPODA
Suborder orthoidea
Superfamily enteletacea Waagen 1884

[nom. transl. Alikhova 1960 (pa Enteletinae Waagen 1884)]

Family schizophoriidae Schuchert and LeVene 1929
Genus vallomyonia gen. nov.

Type species. Skenidium devonicum Walcott 1884, p. 116, pi. 13, fig. 4, 4a.

Discussion. The new genus is represented only by minute or small shells with a shallow
dorsal sulcus. Internally very small pedicle valves bear minute dental lamellae, but larger
specimens with the width of about a centimetre completely lack them. In the brachial
valve the structures are for the most part delicate. The cardinalia consist of thin, widely
divergent brachiophores to which are attached long rod-like brachiophore processes.
The brachiophores of very small specimens are supported by short, thin, brachiophore
supporting plates that are continuous anteriorly with ridges bounding the muscle field
laterally and short, thin, fulcral plates serve as the bases of the sockets. Larger specimens,
with a maximum width of about a centimetre, have more strongly differentiated muscle
bounding ridges that have become separate from the brachiophore bases and lie outside
of them — giving a greater relative width to the posterior adductor muscle scars. Larger
specimens also show the anterior adductor scars enclosed anteriorly and the whole of
the antero-lateral margin of the muscle field is slightly elevated by the well-developed
muscle bounding ridges. Both minute and small specimens bear a high, triangular median
septum that extends from near the base of the cardinal process to the anterior margin
of the shell.

The genus Prokopia Havlicek 1 953 was considered by the writer as a possible receptacle
for Walcott’s species and specimens of Prokopia bouskai from the Eifelian Trebotov
limestones kindly sent by Dr. Vladimir Havlicek have a relatively strongly convex
pedicle valve and appear to lack dental lamellae, however, the hinge teeth are relatively
ponderous and are supported postero-laterally by shell material deposited in the um-
bonal cavities, thus short dental lamellae, if present, could have been made obsolescent.
The ventral muscle field of Prokopia is broadly transverse and approximately semi-
circular in outline, separated from the interior of the valve along its anterior edge by a
small step or elevated platform similar to that developed in small specimens of Levenea.
The median septum of Prokopia is very high and triangular and apparently reaches
nearly to the anterior edge of the valve. Its adductor muscle field is bounded antero-
laterally by relatively strongly impressed margins that define a pair of triangular muscle
impressions much the shape as are present in Hypsomyonia, however, the anterior edges
of the muscle bounding ridges are not elevated above the floor of the valve as in the latter
genus. On the specimen of Prokopia at hand it is not possible to detect fulcral plates. The
sockets apparently are defined by shell material connecting the bases of the brachio-
phores laterally with the floor of the valve.

Vallomyonia resembles Monelasmina Cooper 1955 (see also Pedder 1959) in some
respects, but Cooper’s genus bears strong dental lamellae in the ventral valve and a

J. G. JOHNSON: MIDDLE DEVONIAN BRACHIOPODS

159

relatively long pair of diductor muscle impressions. The cardinalia of Monelasmina, in
addition, are of the Schizophoria type, that is, with well-developed brachiophore sup-
porting plates in contrast to the relatively minute ones seen only on very small specimens
of Vallomyonia. Vallomyonia seems closest to Cooper’s genus Hypsomyonia (1955) from
the Frasnian Independence Shale of Iowa and so far as their differences can be evaluated
the two genera can be distinguished by only slight elevation of the muscle bounding
ridges in the brachial valve of Vallomyonia in contrast to Hypsomyonia in which the
whole adductor muscle field is elevated above the floor of the valve. Vallomyonia is the
most logical ancestor to Hypsomyonia that has come to the writer’s attention.

Vallomyonia devonica (Walcott 1884)

Plate 23, figs. 1-17

Material. The holotype, USNM no. 13827, was borrowed from the U.S. National Museum. It is a
silicified specimen attached to a small slab of limestone and is from the same locality of Walcott’s
at Lone Mountain (553) as are the types of Leptathyris circula. There are several other specimens of
L. circula on the type slab.

Figured specimens. USNM nos. 145478-83, 145538, 145538A, 145538B.

Exterior. The shells are minute to small, transversely suboval to subquadrate in outline,
and unequally biconvex in lateral profile. The pedicle valve is relatively low and not
strongly convex and is faintly carinate. The brachial valve is nearly flat or gently convex
and markedly sulcate. The interarea of the pedicle valve is steeply apsacline to nearly
catacline, low, triangular, and nearly flat. The delthyrium is triangular and open,
enclosing an angle of about 40°. Among the larger specimens, approaching a centimetre
in maximum width, there is commonly a small concave plate inside the apex of the
delthyrium. The ventral beak is short. The interarea of the brachial valve is long, flat,
and anacline.

The ornament consists of numerous subangular costellae of several sizes which increase
in number anteriorly by bifurcation and by implantation. The shell substance is rela-
tively thin and delicate, and is endopunctate.

Interior of pedicle valve. The hinge teeth are small, stubby, and triangular with their long
edges parallel to the interarea of the valve. They are supported basally by very short,
thin, widely divergent dental lamellae which define a transverse, apically situated, ven-
tral muscle field. Larger specimens have the dental lamellae almost completely obsole-
scent. The muscle field is transverse and smooth, unfaceted, and undivided medially by
a myophragm. Its anterior edge is at the same level as the interior of the valve or is
slightly elevated in larger specimens. The interior is crenulated in the anterior half of the
shell or peripherally by the impress of the costellae.

Interior of brachial valve. The cardinal process is bilobate, short, and triangular with its
myophore face approximately parallel to the plane of the interarea. It is supported basally
by the posterior end of the median septum and postero-laterally by a pair of ancillary
struts (Williams and Wright 1963, p. 9). The latter structures, however, join laterally to
the brachiophores and not to brachiophore supports. The brachiophores are slender and
suboval in cross-section and are extended as long slender brachial processes toward the

160

PALAEONTOLOGY, VOLUME 9

ventral valve and with a slight anterior inclination. The sockets are defined between the
interarea and the brachiophores in very small shells by minute fulcral plates that are
elevated well above the floor of the valve. Brachiophore supporting plates on very small
specimens diverge moderately antero-laterally and recurve to become subparallel to the
midline and continue as ridges bounding the lateral edges of the unpartitioned adductor
muscle field. On larger specimens brachiophore supporting plates are obsolescent and
the muscle-bounding ridges become more widely set apart and distinct from the brachio-
phore bases. Also the posterior adductors become relatively wider and the whole
adductor muscle field becomes elevated along its margin. A high triangular median
septum is present from the base of the cardinal process almost to the anterior edge of the
valve. The interior is moderately to strongly crenulated by the impress of the costellae.

Occurrence. Vallomyonia devonica is a relatively common species in the Leptatliyris
circula zone in the exposures of the Denay Limestone at the north-eastern tip of the
Simpson Park Range, on the north side of the Roberts Mountains at Willow Creek,
and on the west side of the Roberts Mountains south of Hanson Creek and east of
Red Canyon. It is also present in the Pentamerella subzone at Birch Creek.

Suborder pentameroidea
Superfamily pentameracea M‘Coy 1844

[nom. transl. Schuchert 1896 (ex Pentameridae M‘Coy 1844)]

Family pentameridae M‘Coy 1844
Subfamily gypidulinae Schuchert and LeVene 1929

Genus gypidula Hall 1867

Type species. Pentamerus comis Owen 1852, p. 583, pi. 3a, fig. 4.

EXPLANATION OF PLATE 23

Figs. 1-17. Vallomyonia devonica (Walcott). 1-3, Pentamerella subzone, locality 4744, northern Roberts
Mountains. 4-17, lower Leptathyris circula zone, locality 4739. 1, Interior of pedicle valve X 6,
USNM 145538. 2, Interior of brachial valve X4, USNM 145538b. 3, Interior of brachial valve x4,
USNM 145538a. 4, Interior of pedicle valve X 5, USNM 145479. Note very short dental lamellae.
5, Oblique view of interior of fragment of brachial valve x4, USNM 145480. Note high triangular
median septum. 6-10, Ventral, dorsal, anterior, posterior, and side views X 5, USNM 145482.
11-15, Ventral, dorsal, anterior, posterior, and side views X4,USNM 145483. 16, Interior of brachial
valve X 8, USNM 145478. Note high triangular median septum. 17, Interior of brachial valve X 8,
USNM 145481. Note long slender brachial processes.

Figs. 18-22. ‘ Schuchertella' sp. 18-20, Exterior, interior, and posterior views of pedicle valve x2,
USNM 145485, Pentamerella subzone, loc. 4744. 21, 22, Interior and exterior views of brachial
valve X 3, USNM 145484, upper Leptathyris circula zone, loc. 4737.

Figs. 23-35. Pentamerella wintereri Johnson sp. nov. Pentamerella subzone, lower Leptathyris circula
zone, northern Roberts Mountains. 23, Exterior of pedicle valve X 3, USNM 145496, loc. 4741.
24, Exterior of pedicle valve x3,USNM 145497, loc. 4741. 25, Interior of pedicle valve X4,USNM
145494, loc. 4740. 26, Exterior of pedicle valve X 3, USNM 145495, loc. 4741. 27, Exterior of brachial
valve x2, paratype, USNM 145488, loc. 4744. 28, Exterior of brachial valve X2, paratype, USNM
145487, loc. 4744. 29, Exterior of pedicle valve X 2, holotype, USNM 145486, loc. 4744. 30, Interior
of brachial valve x4, USNM 145491, loc. 4740. 31, Interior of brachial valve x4, USNM 145492,
loc. 4740. 32, Exterior of brachial valve X3, USNM 145493, loc. 4741. 33, Exterior of brachial
valve X2, paratype, USNM 145490, loc. 4744. 34, 35, Interior and exterior of pedicle valve X 3,
paratype, USNM 145489, loc. 4744.

Palaeontology, Vol. 9

PLATE 23

JOHNSON, Devonian brachiopods from Nevada

J G. JOHNSON: MIDDLE DEVONIAN BRACHIOPODS

161

Gypidula cf. recur rens (Meyer 1913)

Plate 24, figs. 1-7

1913 Pentamerus pseudogaletus Hall, mut. nov. recurrens Meyer p. 34, pi. 7, figs. 5a-c.
Material figured. USNM nos. 145498-500.

Exterior. The shells are elongate subpyriform in outline and strongly unequally biconvex
in lateral profile. The pedicle valve bears a relatively broad and very strongly curved
umbo and the valve, along its midline, commonly curves through an arc of 180° or
slightly more. The brachial valve is subquadrate in outline and is only moderately convex
and cap-like. There is a low, poorly defined, elongate, narrow fold on the pedicle valve.
The brachial valve bears a shallow, broad, subrectangular sulcus at its anterior and
commonly projects as a short to relatively long tongue on the largest specimens. There
commonly are two or three low rounded plications, strongest at the fold and sulcus near
the anterior commissure, but extending along approximately the anterior two-thirds of
the length of the pedicle valve. The largest specimens bear a pair of shallow furrows
bordering the sulcus on the brachial valve and affecting only the posterior fourth or
fifth of the length of the brachial valve. Larger specimens bear a relatively strongly de-
veloped, moderately incurved, interarea on the pedicle valve. The pedicle valve does not
flare broadly toward the anterior, but maintains almost subparallel lateral extremities.

Interior of pedicle valve. There is an elongate, V-shaped, rhomboidal spondylium
originating beneath the delthyrium and projecting slightly more than half its length
anterior to the hinge line. It is supported along its entire length by a thin median septum,
but the septum retreats slightly basally, not attaining as great a distance anterior as
the distal end of the spondylium.

Interior of brachial valve. The brachial valve bears a short, flat, orthocline interarea
medially that equals slightly less than half the maximum width of the valve. It consists
of two long, low, flat, triangular areas with their bases at the posterior edge of the valve
and the two medial basal angles meeting at the beak. The sockets are defined by the inner
edges of the interarea and the postero-lateral edges of the inner plates. The inner plates
are triangular in outline, diverge antero-laterally, and converge slightly basely toward the
midline. The outer plates curve outward and then recurve medially in a cylindroidal
fashion, then join the interior of the valve along two discrete, slightly divergent tracks.

Occurrence. The writer has collected specimens from the west flank of the Willow Creek on the north
side of the Roberts Mountains and has examined several other collections made by Winterer and his
students from other places in the Roberts Mountains including the southern Roberts Mountains.

Genus pentamerella Hall 1867
Type species. Atrypa arata Conrad 1841, p. 55.

Pentamerella wintereri sp. nov.

Plate 23, figs. 23-35

Material. The specimen, USNM no. 145486, in fig. 29 of Plate 23 is the holotype. It is a free pedicle
valve, no articulated specimens being available among several hundred studied. The holotype measures
10-8 mm. in width, 7-6 mm. in length, and 4-7 mm. in thickness. The specimens in figs. 27, 28 and 33-35
of Plate 23 are paratypes. These include USNM nos. 145587-97.

M

B 6612

162

PALAEONTOLOGY, VOLUME 9

Diagnosis. Small, strongly biconvex Pentamerella, with strong, rounded plications.

Exterior. The shells are transversely suboval in outline and unequally biconvex in lateral
profile. Pedicle valves are about twice as strongly convex as brachial valves and are very
strongly incurved at the beak which, however, is short and stubby. No interarea is
developed on any of the specimens examined, the palintrope curves smoothly to join
the sides of the valves. The delthyrium is open and triangular. Many of the well-preserved
pedicle valves bear an apical plate filling a quarter or a third of the delthyrium and
covering, but not filling, the posterior end of the spondylium. The pedicle valve bears
a shallow sulcus extended anteriorly as an elongate projecting tongue and the brachial
valve bears a relatively well-defined median fold that accommodates the projecting
anterior tongue of the pedicle valve.

The exterior bears a few strong rounded plications on each valve. Commonly there
are three plications on the fold of the brachial valve and two plications in the sulcus of
the pedicle valve. However, on some specimens, the pedicle valve bears a single median
plication within the sulcus. On some larger specimens that bear a single medium plication
in the early growth stage, a second plication may be intercalated toward the anterior.
There commonly are two or three plications on each flank of the brachial valve and from
two to four plications on each flank of the pedicle valve.

Interior of pedicle valve. There is an elongate spondylium of rhomboidal outline situated
beneath the delthyrium and projecting anteriorly beyond the hinge line. It is supported
apically by a median septum, but the length of the septum appears to be subject to some
variability in its length along the base of the valve (compare figs. 25 and 34 of PI. 23).
The surface is crenulated by the impress of the plications.

Interior of brachial valve. The sockets are rudimentary, being formed by the edge of the
shell on the postero-lateral edge above the inner plates. The inner plates diverge antero-
laterally and converge toward the midline. The outer plates are elongate and subparallel,
in some shells meeting along a line medially and in others joining the surface of the valve
along a pair of discrete subparallel tracks. A number of variations may be seen between
structures which are easy to classify as one type or the other. The interior is crenulated
by the impress of the plications.

Occurrence. The new species is extremely abundant at some horizons within a few feet to 50 or 60 ft.
above the base of the Denay Limestone on the north side of the Roberts Mountains. The writer has

EXPLANATION OF PLATE 24

Figs. 1-7. Gypidula cf. recurrens (Meyer). Lower Leptathyris circula zone, loc. 4739, northern Roberts
Mountains. 1-5, Dorsal, anterior, posterior, side, and ventral views X 1, USNM 145498. 6, Interior
of brachial valve X2, USNM 145499. 7, Interior of pedicle valve x2, USNM 145500.

Figs. 8-23. Leiorhynchus sp. Lower Leptathyris circula zone and Pentamerella subzone, northern
Roberts Mountains. 8, Side view of interior of both valves x3, USNM 145503, loc. 4739, lower
Leptathyris circula zone. Note long crus, high triangular median septum, and short dental lamella.
9-13, Side, dorsal, ventral, anterior, and posterior views X 2, USNM 145501, loc. 4740, Pentamerella
subzone. 14-18, Dorsal, ventral, side, posterior, and anterior views x3, USNM 145502, loc. 4739,
lower Leptathyris circula zone. 19-23, Dorsal, ventral, posterior, anterior, and side views X 2,
USNM 145504, Leptathyris circula zone, loc. 4743, west slope of Roberts Mountains.

Figs. 24-28. Dubarial cf. thetis (Barrande). Lower Leptathyris circula zone, loc. 4739, northern
Roberts Mountains. 24-28, Ventral, dorsal, side, anterior, and posterior views x2, USNM 145520.

Palaeontology, Vol. 9

PLATE 24

JOHNSON, Devonian brachiopods from Nevada

J. G. JOHNSON: MIDDLE DEVONIAN BRACHIOPODS

163

collected slabs with abundant disarticulated specimens in both Willow Creek and Birch Creek Canyons
and in the small canyon immediately east of the mouth of Willow Creek along the front of the range.
Other collections have been made by E. L. Winterer in the southern Roberts Mountains in the vicinity
of Roberts Creek. Cooper (1942, p. 1773) mentions the presence of small Pentamerella at Lone Moun-
tain indicating that the species may be present there also. Another species of small Pentamerella is pre-
sent in a silicifled collection containing Stringocephalus and Subrensselandia in Red Canyon on the west
flank of the Roberts Mountains. However, the Stringocephalus zone specimens bear more and finer
plications. The species has not been recognized in the Denay Limestone exposures in the northern
Simpson Park Range.

Comparison. The new species most closely resembles P. suspecta Khodalevich (1951, pi.
7, fig. 7 d-d) and P. sosvaensis Khodalevich (1951, pi. 7, fig. 1 \a-d) from the Eifelian of the
Ural Mountains. P. sosvaensis may be distinguished by being more elongate and plicate
only on the anterior half of the shell. P. suspecta is very similar to the new species in size,
relative convexity, and in the size and number of the plications. However, the illustrated
specimen is somewhat more strongly rhomboidal in outline than P. wintered. P. arata
Conrad, P. pavilionensis Hall, and P. dubia Hall (Hall and Clarke 1895, pi. 71) are all
much larger than the new species and thus cannot be closely compared. All appear, how-
ever, to have relatively smooth umbonal regions on both valves so that specimens the
size of those typical of the Nevada fauna would be much less strongly plicate. P. obsole-
scens Hall (1867, pi. 58, figs. 24, 25) is a small species, but almost completely lacks
plications. P. thusnelda Nettelroth (1889, p. 51, pi. 31, figs. 26-28) is a much larger and
elongate species than is wintered. Branson (1923) named two new species from the
Middle Devonian of Missouri. P. wintered contrasts strongly with one of these, P.
missouriensis, in having a much shorter ventral beak and less prominent ventral umbo.
P. wintered differs from Branson’s second species, P. fidtonensis, in having plications
(Branson 1923, pi. 16, figs. 21-23; pi. 24, figs. 1-4). The new species differs from P.
liorhyncha Cooper and Cloud (1938, p. 447, pi. 54, figs. 12-16, 31-34) which is deci-
dedly more elongate and pyriform in outline.

P. athyroides (Winchell) as illustrated by Imbrie (1959, pi. 51, figs. 6-10) is a larger
species with plications not developed on the umbones. The same may be said of P. peto-
skeyensis (Imlay) and P. papilla Imbrie (Imbrie 1959, pi. 51, figs. 25-30, 18-24). P. aulax
Imbrie (1959, pi. 51, figs. 1 1-17) has much finer plications than does P. wintered. Other
named species of Imbrie, P. pericosta, P. lingua, P. aftonensis, P. tumida, and P. alpe-
nensis all differ from P. wintereri in being relatively large forms that lack plications on
the umbones. P. proteus Imbrie (1959, pi. 53, figs. 1-8), a small species with a few plica-
tions that originate close to the posterior, differs in having a transverse subrhomboidal
outline and a much smaller apical angle than do specimens of P. wintereri. The species
Spirifer davidsoni Schnur (1853, pi. 35, fig. 7 ; pi. 44, fig. 3), a Pentamerella according to
Havlicek (1952, p. 2, pi. 1, figs. 7, 9), is very similar to P. wintereri but has a consider-
ably more prominent beak in the pedicle valve. P. sclavus Norris from the Middle
Devonian Horn Plateau Formation of the Mackenzie District of north-west Canada
(McLaren and Norris 1964, p. 37, pi. 13, figs. 9-13) is larger and has a narrower and
more elongate ventral beak than does P. wintereri. The Canadian species Pentamerus
borealis Meek was assigned to Pentamerella by Schuchert (1897, p. 302), but it appears
that exception should be taken to the assignment since ‘iV borealis bears a ventral
fold and dorsal sulcus (Meek 1868, pi. 13, fig. 1 la, b, d). In any event ‘TV borealis bears
little similarity to Pentamerella wintereri.

164

PALAEONTOLOGY, VOLUME 9

Suborder strophomenoidea
Superfamily orthotetacea Waagen 1884
[nom. transl. Williams 1953 (ex Orthotetinae Waagen 1884)]

Family schuchertellidae Williams 1953
[nom. transl. Stehli 1954 (ex Schuchertellinae Williams 1953)]

Subfamily schuchertellinae Williams 1953
Genus schuchertella Girty 1904

Type species. Streptorhynchus lens White 1862, p. 28.

‘ Schuchertella ’ sp.

Plate 23, figs. 18-22

Material figured. USNM nos. 145484-5.

Discussion. Assignment to Schuchertella is not firm since pseudopunctae were not
observed.

Exterior. The species is small, averaging about a centimetre in width and is transversely
subsemicircular to shield-shaped in outline. The valves are unequally biconvex in lateral
profile with the pedicle valve more strongly convex than the brachial valve and com-
monly somewhat elevated with a moderately twisted beak. The ventral interarea is
apsacline, flat, and triangular. Its height varies and the delthyrium is covered by a gently
convex pseudodeltidium. The ornament consists of fine, rounded, slightly parvicostellae
costellae that increase in number anteriorly, principally by intercalation. The costellae are
crossed by a few irregularly spaced growth lines. A shallow sulcus may be developed
on the brachial valve. The maximum width of the valves is attained near midlength.

Interior of pedicle valve. The hinge teeth are relatively large and triangular in horizontal
section. They leave stout tracks on the inner edges of the delthyrium, but there are no
supporting dental lamellae. The interior of the valve is not crenulated by the impress of
the costellae.

Interior of brachial valve. The socket ridges are widely divergent antero-laterally and are
unsupported on their divergent inner edges. Two short prongs of shell material are
attached to the posteromedial edges of the socket ridges and project posteriorly. It was
not determined whether a chilidium was present or not. The interior is not crenulated by
the impress of the costellae.

Occurrence. This small species of ‘ Schuchertella ’ is present in one collection from high in the Lepta-
thyris circula zone in the southern Roberts Mountains and from the Pentamerella subzone on Birch
Creek in the northern Roberts Mountains.

Suborder rhynchonelloidea
Superfamily rhynchonellacea Gray 1848
[nom. transl. Schuchert 1896 (ex Rhynchonellidae Gray 1848)]

Family trigoniphynchiidae Schmidt 1965
Genus hadrorhynchia McLaren 1961

Type species. Pugnoides sandersoni Warren 1944, p. 115, pi. 2, figs. 5, 6.

J. G. JOHNSON: MIDDLE DEVONIAN BRACHIOPODS

165

Hadrorhynchia eurekaensis sp. nov.

Plate 25, figs. 1-23

Material. The specimen in figs. 6-10 of Plate 25, USNM no. 145511, is designated the holotype. It
measures 10-8 mm. in width, 9-8 mm. in length, and 8 0 mm. in thickness. The other illustrated speci-
mens are paratypes. They include USNM nos. 145505-10.

Diagnosis. Multicostate Hadrorhynchia, costae relatively strong on the umbonal regions.
Shape lenticular to subcuboidal, not reflexed.

Exterior. Most of the available specimens of this species are small or medium sized and
bear a pyriform to subpentagonal outline. Commonly, smaller specimens are somewhat
elongate and pyriform while the large ones develop slight irregularities in the shape
resulting in the subpentagonal outline. The valves are strongly biconvex in lateral profile
with the brachial valve about a quarter to a third more strongly convex than the pedicle
valve. Smaller specimens are more commonly somewhat lenticular, but a lesser number
of even very small specimens attain the subcuboidal outline more common to medium-
sized specimens. The beak of the pedicle valve is moderately incurved and projects
somewhat strongly anterior to the umbo of the brachial valve. There is a triangular del-
thyrium and low, apsacline, incurved interarea. The beak angle is commonly between
90° and 100° and is relatively constant for a distance equal to about a third of the length
of small shells and a fourth or fifth of the length of larger shells and then decreases
anteriorly to the place of maximum width which is anterior to midlength. On most
specimens the pedicle valve has a relatively flattened outline along the lateral flanks as
viewed from the side, then curves abruptly in a nearly geniculate fashion at the anterior
commissure. The sulcus, however, attains a more uniform curvature from the beak to
the anterior commissure. The pedicle valve bears a shallow median sulcus that originates
in the posterior portion of the valve and becomes relatively well marked anteriorly where
it is accentuated by the development of a tongue-like projection that is accommodated
by a subrectangular fold in the anterior portion of the brachial valve. The ornament
consists of numerous subangular costae which have smoothly rounded crests. The costae
expand in size anteriorly and do not bifurcate. Commonly, the median costae on the
brachial valve become relatively broad anteriorly and appear notably broader than the
adjoining lateral costae at the same distance from the beak which continue strongly
curving adjacent to the projecting tongue of the sulcus. Most commonly there are three
costae in the ventral sulcus and four costae on the dorsal fold, but specimens with lesser
and greater numbers are also encountered.

Interior of pedicle valve. The hinge teeth are small and delicate. They are supported
basally by thin dental lamellae of triangular outline which are situated relatively close
to the postero-lateral walls of the valve. The interior is crenulated by the impress of the
costellae.

Interior of brachial valve. The sockets are long and narrow and are cylindroidal in shape,
being defined postero-laterally by the inner edge of the shell and basely and ventrally
by the curved hinge plates. A pair of triangular horizontal plates connects the inner
edges of the outer hinge plates with the crura which arise medially on either side of a
narrow V-shaped septalium that is supported by a long high median septum. The crura

166

PALAEONTOLOGY, VOLUME 9

are T-shaped in cross-section and curve strongly ventrally, recurving with their convex
edges facing antero-laterally. The interior is fairly strongly crenulated over much of the
length of the valve by the costellae.

Occurrence. This species is the most common rhynchonellid in the Leptathyris circula zone on the north
and south sides of the Roberts Mountains. The species is also present in the Warrenella kirki zone over-
lying the L. circula zone in the J-D window of the northern Simpson Park Range. It appears that
Rhynchonella transversa of Meyer (1913, pi. 7, fig. la-e\ not Hall) from the Blue Fiord Formation of
Ellesmere Island could belong to H. eurekaensis. Unfortunately the Arctic species is unknown
internally.

Comparison. H. sandersoni (Warren) has much more poorly developed costae on the
posterior portion of the valve. In addition, the anterior sulcus appears to be deeper on
sandersoni and the shells are commonly relatively more transverse. Furthermore, maxi-
mum width of H. eurekaensis appears on almost all the shells to be developed at a some-
what greater distance anterior to the beak (McLaren 1962, pi. 7). H. eurekaensis appears
to be distinguishable from H. intermissa and H. vailorum Crickmay (1963, pi. 7) on the
same grounds that it may be distinguished from sandersoni.

Family leiorhynchidae Stainbrook 1945
[nom. transl. Crickmay 1952 (ex Leiorhynchinae Stainbrook 1945)]

Genus leiorhynchus Hall 1 860

Type species. Orthis quadricostata Vanuxem 1842, p. 168, fig. 2.

Leiorhynchus sp.

Plate 24, figs. 8-23, text-fig. 3
Material figured. USNM nos. 145502-4, 145521.

Exterior. The shells are variable in outline and convexity, commonly subpyriform to
somewhat pentagonal in outline. Most specimens are strongly biconvex with the brachial
valve more convex than the pedicle valve. A few specimens are relatively lenticular. The
beak angle is about 90° or a little more and the postero-lateral margins are relatively

EXPLANATION OF PLATE 25

Figs. 1-23. Hadrorhynchia eurekaensis Johnson sp. nov. Lower Leptathyris circula zone, loc. 4739,
northern Roberts Mountains. 1-5, Ventral, dorsal, side, posterior, and anterior views X2, paratype,
USNM 145509. 6-10, Ventral, dorsal, side, posterior, and anterior views X2, holotype, USNM
145511. 11-15, Ventral, dorsal, side, posterior, and anterior views X3, paratype, USNM 145510.
16, 17, Interior of fragment of pedicle valve and posterior of brachial valve x4, paratype, USNM
145505. The two valves belong to the same individual. In fig. 16 note shape of the dental lamella.
18, 19, Dorsal and ventral views X 3, paratype, USNM 145506. 20, 21, Dorsal and ventral views x 3,
paratype, USNM 145508. 22, 23, Ventral and dorsal views of small specimen X 3, paratype, USNM
145507.

Figs. 24-28. Spinatrypa aff. lata (Warren 1944). Near base of Telegraph Canyon Member of Nevada
Formation, loc. 3536, Union Mountain, Sulphur Spring Range. Horizon probably should be assigned
to Warrenella kirki zone. 24-28, Anterior, posterior, side, ventral, and dorsal views X 2, USNM
145514.

Figs. 29-35. Spinatrypa sp. B. 29, 30, Dorsal and ventral views X 3, USNM 145512, loc. 4738, Penta-
merella subzone, southern Roberts Mountains. 31-35, Ventral, dorsal, side, posterior, and anterior
views x2, USNM 145513, loc. 4741, Pentamerella subzone, northern Roberts Mountains.

Palaeontology, Vol. 9

PLATE 25

JOHNSON, Devonian brachiopods from Nevada

J. G. JOHNSON: MIDDLE DEVONIAN BRACHIOPODS 167

straight in the posterior third of the shell. Maximum width is commonly anterior to mid-
length. The beak of the pedicle valve is strongly incurved and bears a small apical fora-
men. The pedicle valve bears a broad flat-bottomed sulcus in its anterior half and the
brachial valve bears a corresponding subrectangular fold. In most specimens, the fold
is divided medially by a shallow narrow furrow and anteriorly may bear two to four low
rounded plications. Most larger specimens
bear a few very short plications adjoining
the fold and sulcus, but affect only a small
portion of the length of the shell.

Interior of pedicle valve. The hinge teeth are
small and delicate and are attached along
the inner edge of the postero-lateral shell
margin. They are supported by a pair of
short thin dental lamellae that converge
medially to join, forming a sessile spondy-
lium. No median septum is present. The
dental lamellae converge at an angle of
about 75-80°.

Interior of brachial valve. The dental sockets
are exceedingly narrow and diverge antero-
laterally, nearly parallel to the valve margin.

The crura originate at the beak and diverge
at a very small angle as they project antero-
ventrally. They are circular in cross-section with fluted lateral flanges that are roughly
parallel to the plane of commissure along their posterior portions, but which become
depressed below the level of the crura and extend anteriorly along planes nearly parallel
to the median plane (text-fig. 3). Anteriorly the bases of the crura and their lateral flanges
are connected to the inner edges of the hinge plates by a pair of thin, narrow, tri-
angular plates. Medially the dorsal edges of the crura are attached to a high median
septum. No septalium is present. The median septum is long and blade-like, reaching
about half way or slightly less to the anterior of the shell.

Occurrence. Leiorhynchus sp. is a relatively uncommon species in the Leptathyris circula zone on the
north side of the Roberts Mountains. A number of specimens have also been collected from the Denay
Limestone in the northern Simpson Park Range.

Suborder atrypoidea
Superfamily atrypacea Gill 1871

[no/n. transl. Schuchert and LeVene 1929 (ex Atrypidae Gill 1871)]

Family atrypidae Gill 1871
Subfamily atrypinae Gill 1871
Genus spinatrypa Stainbrook 1951

Type species. Atrypa hystrix var. occidentals Hall 1858 ( errore pro A. aspera var. occidentals Hall 1858,
p. 515).

text-fig. 3. Interior of posterior portion of
articulated shell of Leiorhynchus sp., USNM
145521. Sketch of articulating apparatus on the
right side is generalized because of poor pre-
servation. Note circular cross-section of the
crura and the prominent flanges that extend
ventrally from them.

168

PALAEONTOLOGY, VOLUME 9

Spinatrypa aflf. lata (Warren 1944)

Plate 25, fig. 24-28

1944 Atrypa borealis var. lata Warren, p. 122, pi. 3, figs. 7-9.

1962 Spinatrypa iota McLaren and Norris, pi. 8, figs. 25-27.

Materia l figured. USNM no. 145514.

Exterior. The shells are transversely suboval to subpentagonal in outline and subequally
biconvex in lateral profile. The valves are more or less lenticular and none are strongly
convex. A few of the smaller specimens are elongate with the maximum width near mid-
length or posterior to it, however, the relatively larger specimens are all transverse and
have their maximum width posterior to midlength at a distance approximately one-third
of the distance from the ventral beak. The ventral beak is small, pointed, and nearly
straight and is pierced apically by a small circular foramen. The foramen joins anteriorly
and dorsally with a broad, low, triangular delthyrium enclosing an angle greater than
90°, which is completely covered by deltidial plates that lie adjacent to the umbo of
the brachial valve that has its beak incurved beneath them. On most specimens the
brachial valves are of relatively even convexity across the midline, but pedicle valves
are mildly carinate, except anteriorly where a faint ventral sulcus appears or is suggested
by dorsal deflection of the anterior commissure.

The ornament consists of numerous, well-developed, rounded costae that increase in
number most commonly by bifurcation on the pedicle valve and by intercalation on the
brachial valve. They are crossed by numerous well-developed concentric growth lamellae
that occur at regular intervals from the posterior to the anterior margin of the valves.
The number and size of the radial costellae is variable, but in general the costae are
relatively fine for the genus.

Occurrence. This species comes from just above the Union Mountain Sandstone in the
northern part of the Sulphur Spring Range in beds that were incorrectly assigned to
the Heliolites horizon by Carlisle and others (1957, loc. 3536). Present in the same collec-
tion are numerous small ambocoeliids and some small athyroids close to Lcptathyris
circula. Species of this type from the Sulphur Spring Range are regarded as a possible
link between finely costellate Spinatrypa of the L. circula zone in the Roberts Mountains
as well as in the northern Simpson Park Range and with Spinatrypa cf. andersonensis
which is common in the typical Warrenella kirki zone.

Spinatrypa sp. B
Plate 25, figs. 29-35

Diagnosis. Transverse lenticular Spinatrypa with fine costellae and growth lamellae.
Material figured. USNM nos. 145512, 145513.

Discussion. This species has a size, outline, and convexity similar to Spinatrypa cf. lata
described above, however, it may be distinguished by its finer ornament. The radial
costellae are exceptionally fine for the genus and appear commonly to be somewhat
irregular and subject to numerous bifurcations and intercalations posteriorly. Costellae
remain essentially fine and narrow anteriorly. In some cases their width is less than some

J. G. JOHNSON: MIDDLE DEVONIAN BRACHIOPODS

169

of the main costellae posteriorly. They are crossed by numerous, very fine, imbricating,
concentric growth lamellae.

This appears to be a new species insofar as the writer has been able to find no closely
comparable form in the literature. However, the specimens available are relatively few
in number and are not exceptionally well preserved. Thus, pending a more extensive
collection of more adequately preserved specimens, this form will be referred to as
Spinatrypa sp. B. This is Spinatrypa sp. B of Johnson {\962a, p. 166).

Occurrence. Spinatrypa sp. B is a common atrypoid of the Leptathyris circula zone and Pentamerella
subzone in the northern Roberts Mountains and in the L. circula zone in the northern Simpson
Park Range.

Subfamily carinatininae Rzhonsnitskaya 1960
Genus vagrania Alekseeva 1959

Type species. Atrypa kolymensis Nalivkin 1936, p. 17, pi. 2, fig. 8.

Vagrania ? sp.

Plate 26, figs. 1-12

Material figured. USNM nos. 145515-19.

Discussion. Vagrania is a strongly costate or plicate atrypoid that lacks concentric growth
lamellae. Instead, as shown by Alekseeva (I960, pi. 10, fig. 3), the genus bears a fine
ornament consisting of numerous, very fine pustules scattered over the surface in a
quincunx pattern. Internally, large specimens carry diagnostic and unusual structures
(Boucot, Johnson, and Staton 1964, pi. 128, figs. 10-12).

The specimens described below from central Nevada are very small and do not
develop the structures which are known to be present in larger specimens of the genus,
however, as yet we have no knowledge of the development of these same structures in
small specimens of the Russian types. The exterior of the specimens, in addition, is a bit
too roughly preserved to ascertain the original presence or absence of the peculiar fine
ornament of the genus. In the absence of these features questionable assignment is made
on the basis of the coarse ribbing and lack of concentric ornament on the specimens
from the Roberts Mountains.

Exterior. The shells are elongate suboval to somewhat pyriform or subtrigonal in outline
and subequally biconvex in lateral profile. The beak of the pedicle valve is short and
pointed and projects posteriorly, straight or nearly straight. The hinge line is short and
strongly curved and the pedicle valve bears a relatively flat palintrope consisting of two
triangular flat surfaces on either side of the circular or suboval foramen present antero-
dorsally to the beak of the pedicle valve. The anterior commissure is rectimarginate.

The ornament consists of well-defined, rounded, radial costae that increase in number
anteriorly by bifurcation and by implantation ; the latter most commonly on the brachial
valve and the former on the pedicle valve (see figs. 5 and 6 of PI. 26).

Interior of pedicle valve. The hinge teeth are thin and plate-like and are approximately
perpendicular to the median line. They are attached directly to the inner edges of the
valve and are not supported by dental lamellae. The interior is moderately strongly
crenulated by the impress of the costae.

170

PALAEONTOLOGY, VOLUME 9

Interior of brachial valve. The sockets are relatively narrow and broadly divergent antero-
laterally. They bear a pair of crural lobes on their inner edges. The inner edges of the
crural lobes border the notothyrial cavity and diverge antero-laterally.

Occurrence. This small species of Vagrania ? sp. is relatively common in the Pentamerella subzone on
the north side of the Roberts Mountains at Willow Creek and in the small canyon immediately east
of Willow Creek along the north face of the Roberts Mountains. The species is also present with
Pentamerella in the southern Roberts Mountains east of Roberts Creek.

Family lissatrypidae Twenhofel 1914

[nom. transl. Boucot, Johnson, and Staton 1964 (ex Lissatrypinae Twenhofel 1914)]

Subfamily septatrypinae Kozlowski 1929
Genus dubaria Termier 1936

Type species. D. lantenoisi Termier 1936, p. 1266, pi. 17, 17 bis.

Dubaria? cf. thetis (Barrande 1847)

Plate 24, figs. 24-28

1847 Terebratula thetis Barrande, p. 349, pi. 14, fig. 5.

1879 Atrypa thetis Barrande, p. 54, pi. 86, case iv, figs. 1-11.

1951 Septatrypa{ ?) thetis Khodalevich, p. 84, pi. 21, figs. 6 a-d.

1962 Rhynchatrypa thetis Siehl, p. 200, pi. 28, figs. 1, 2; pi. 37, fig. 11 ; pi. 38, figs. 1, 2.
Material figured. USNM no. 145520.

Exterior. The shell is subtrigonal in outline and lenticular in lateral profile with the
pedicle valve nearly flat or reflexed and the brachial valve moderately convex. In the
pedicle valve, approximately the first centimetre of shell is low convex, but antero-
laterally the flanks flatten out on larger shells and the median portion becomes deflected
into a deep flat-bottomed sulcus. The brachial valve is evenly rounded from the umbo
to the antero-lateral flanks, but the median section is only faintly curved, nearly flat,
due to the elevation anteriorly of a well-developed, rounded to nearly flat fold. The hinge
line is short and moderately curved.

EXPLANATION OF PLATE 26

Figs. 1-12. Vagrania ? sp. Pentamerella subzone; figs. 1-4, 8-12 northern Roberts Mountains; figs. 5-7
southern Roberts Mountains. 1, 2, Interior and exterior of pedicle valve X3, USNM 145516, loc.
4740. 3, Dorsal view X 3, USNM 145518, loc. 4740. 4, Ventral view X 3, USNM 145519, loc. 4740.
5-7, Dorsal, ventral, and posterior views X3, USNM 145517, loc. 4738. 8-12, Anterior, posterior,
side, dorsal, and ventral views x2, USNM 145515, loc. 4741.

Figs. 13-22. Leptathyris circula (Walcott 1884). Lower Leptathyris circula zone, loc. 4739, northern
Roberts Mountains. 13-17, Dorsal, ventral, anterior, posterior, and side views X2, USNM 145522.
18-22, Posterior, anterior, dorsal, ventral, and side views X 2, USNM 145523.

Figs. 23-30. Echinocoeliadenayensis Johnson sp.nov. 23, Interior of brachial valve X 3, USNM 145524,
loc. 4743, Leptathyris circula zone, west slope of Roberts Mountains. 24, Interior of brachial valve
X 3, paratype, USNM 145525, loc. 4739, lower Leptathyris circula zone, northern Roberts Moun-
tains. 25, Exterior of brachial valve X 3, paratype, USNM 145526, loc. 4739, lower Leptathyris
circula zone, northern Roberts Mountains. 26-30, Ventral, anterior, posterior, side, and dorsal views
x3, holotype, USNM 145527, loc. 4739, lower Leptathyris circula zone, northern Roberts Moun-
tains.

Palaeontology, Vol. 9

PLATE 26

JOHNSON, Devonian brachiopods from Nevada

J. G. JOHNSON: MIDDLE DEVONIAN BRACHIOPODS

171

The ornament consists of a few poorly defined concentric growth lines developed in
the anterior two-thirds of the valves. No radial ornament is present.

Discussion. Unfortunately, only two specimens of this form are available and neither
show the internal structures, however, the external form and lack of radial ornament are
so distinctive it appears that these shells merit a provisional assignment to Duboria and
particularly to the broad lenticular form Dubaria thetis. So far as the writer knows, this
form is restricted to the Emsian and Eifelian in central Europe and in the Ural Moun-
tains, however, the genus is most common in latest Silurian and early Devonian rocks
in Bohemia, the Carnic Alps, and in European and Central Asian Russia.

Occurrence. Dubaria ? cf. thetis is present in a single collection in the northern Roberts Mountains on
the west flank of Willow Creek in the lower Leptathyris circula zone.

Suborder athyridoidea
Superfamily athyridacea M‘Coy 1844

[noin. transl. Williams 1956 (ex Athyridae M‘Coy 1844)]

Family athyrididae M‘Coy 1844
Subfamily athyridinae M‘Coy 1844

[nom. transl. Waagen 1883 (ex Athyridae M‘Coy, 1844)]

Genus leptathyris Siehl 1962
Type species. L. gryphis Siehl 1962, p. 212, pi. 36, fig. 1 , pi. 39, figs. 2, 3.

Leptathyris circula (Walcott 1884)

Plate 26, figs. 13-22; text-fig. 4

1884 Cryptonellal circula Walcott n. sp., p. 163, pi. 15, figs. 2, a, b.

1940 ICryptonella circula Merriam, pi. 11, fig. 8.

1962 Athyris circula Johnson 1962c;, p. 166.

Material. The holotype, USNM no. 13861, was borrowed from the U.S. National Museum and was
accompanied by twenty-four paratypes. They are small silicified shells with the valves tightly welded
together. However, one of the paratypes has the ventral beak broken away revealing the cardinalia in
the brachial valve. They exactly duplicate the structure described above from specimens collected by
the writer from the Roberts Mountains. The structure of this paratype is shown in text-fig. 4. The label
with the types indicates that they are from the ‘ Lower Devonian ’ at Lone Mountain.

Figured specimens. USNM nos. 145522, 145523.

Discussion. Siehl (1962, p. 212) proposed Leptathyris for small non-lamellose athyridids
with a thin, depressed, medially-crested, cardinal plate. These are diagnostic features of
‘ Athyris ’ circula, which is therefore assigned to Leptathyris.

Exterior. The species is small and subcircular to elongate suboval or short pyriform in
outline. In lateral profile the valves are unequally biconvex with the pedicle valve slightly
more convex than the brachial valve. The ventral beak is sub-erect. The palintrope is only
faintly rounded, nearly flat, but beak ridges are absent. The hinge line is short and curved.
There is a triangular delthyrium commonly enclosing about 70°. It is unmodified by
deltidial plates and in all the specimens examined there invariably is a minute ragged
opening at the apex of the delthyrium. Its ubiquitous presence suggests that it is in fact

172

PALAEONTOLOGY, VOLUME 9

a mesothyrid or permesothyrid foramen. Both valves curve evenly anteriorly and antero-
laterally with slightly more gentle curvature than at the umbos. There is no fold, but
some specimens develop a flattened area on the anteromedial part of the pedicle valve
or a faintly indented sulcus and the anterior commissure is gently deflected toward the
brachial valve.

The ornament consists of a few widely spaced, faint, concentric growth lamellae. No
radial ornament was observed.

Interior of pedicle valve. The hinge teeth are supported by a pair of thin dental lamellae
that are situated close to the inner postero-lateral walls of the valve. The anterior edges
of the dental lamellae converge slightly towards the midline and are continued for a short

text-fig. 4. Leptathyris circula. a. Interior of pedicle valve from the Roberts Mountains with fragment
of brachial valve retaining cardinal plate attached, b, Cardinalia of brachial valve of a paratype from
Lone Mountain showing the shape of the cardinal plate and the apical foramen.

distance as ridges that bound the muscle impression. These ridges terminate anteriorly
at about the place of constriction between the pedicle cavity and the diductor muscle
scars. The diductors are faintly impressed, elongate, and slightly divergent. Medially
there is a faintly elevated myophragm dividing the diductors into two more or less
separate tracks that blend anteriorly with the interior of the valve (text-fig. 4).

Figs. 1-5. Cyrtina sp. Lower Leptathyris circula zone, loc. 4739, northern Roberts Mountains. 1-5,
Anterior, posterior, side, ventral, and dorsal views x2, USNM 145528.

Figs. 6-9. Warrenella cf. columbina columbina (Flavlicek 1959). Lower Leptathyris circula zone, loc. 4739,
northern Roberts Mountains. 6-9, Anterior, side, dorsal, and ventral views x 3, USNM 145529.

Figs. 10-25. Warrenella kirki praekirki Johnson subsp. nov. Figs. 10-24, lower Leptathyris circula
zone, loc. 4739, northern Roberts Mountains; fig. 25, zone unknown, loc. 4742, northern Roberts
Mountains. 10-14, Anterior, posterior, side, dorsal, and ventral views Xl-5, holotype, USNM
145530. 15-19, Posterior, anterior, side, dorsal, and ventral views X2, paratype, USNM 145531.
20, Interior of brachial valve x4, paratype, USNM 145532. 21, Interior of posterior portions of
articulated fragments of both valves x 5, paratype, USNM 145533. 22, Interior of fragment of pedicle
valve X3, paratype, USNM 145534. 23, Interior of brachial valve x4, paratype, USNM 145535.
24, Exterior of fragment X 5, paratype, USNM 145536. 25, Interior of fragment of pedicle valve
X 2, USNM 145537.

a

b

EXPLANATION OF PLATE 27

Palaeontology, Vol. 9

PLATE 27

JOHNSON, Devonian brachiopods from Nevada

J. G. JOHNSON: MIDDLE DEVONIAN BRACHIOPODS 173

Interior of brachial valve. The sockets diverge antero-laterally and are unsupported by
crural plates. Their inner edges are joined by an apically perforate cardinal plate which is
slightly depressed dorsally across its width, but which is crested along the median line
with its crest ventrally directed. In plan view the anterior edge of the cardinal plate is
two sided, converging anteromedially. Neither median septum nor myophragm are
present. The spiralia have their apices directed laterally and appear to consist commonly
of six or eight turns.

Occurrence. Leptathyris circula is very common in the lower two or three hundred feet of the Denay
Limestone at a number of places in the Roberts Mountains and in the northern Simpson Park Range.
As reported previously (Johnson 1962a, p. 167) Merriam states that forms assignable to this species
are present at the north end of the Antelope Range.

Suborder spiriferoidea
Superfamily delthyriacea Phillips 1841

[nom. transl. Ivanova 1960 (ex Delthyridae Phillips 1841)]

Family reticulariidae Waagen 1883
[nom. transl. Ivanova 1960 (ex Reticulariinae Waagen 1883)]

Genus warrenella Crickmay 1953
Type species. W. eclectea Crickmay 1953, p. 596, figs. 1-5, 16-19.

Warrenella kirki praekirki subsp. nov.

Plate 27, figs. 10-25

Material. The holotype is USNM no. 145530, illustrated in figs. 10-14 of Plate 27 and measures 19-2
mm. in width, 18-2 mm. in length, and 11-4 mm. in thickness. It was collected at locality 4739 on the
west slope of Willow Creek Canyon in the northern Roberts Mountains. The remainder of the speci-
mens illustrated on Plate 27 are paratypes excepting fig. 25 which originates from another locality.
These include USNM nos. 145531-7.

Diagnosis. Includes relatively transverse, less strongly convex specimens of W. kirki in
populations lacking elongate individuals.

Exterior. The shells are transversely suboval in outline and unequally biconvex in lateral
profile with the pedicle valve about twice as strongly convex as the brachial valve. The
hinge line is equal to about two-thirds the maximum width of the valves which is attained
near midlength. The cardinal angles are rounded and obtuse. The pedicle valve bears
a low apsacline triangular interarea with a relatively strongly incurved beak. The inter-
area is divided medially by a triangular open delthyrium enclosing an angle of about
70°. The interarea of the brachial valve is long, narrow, and anacline. The pedicle valve
commonly bears an indistinct medial sulcus which may extend from the umbo or may
develop anterior to it. The sulcus of all the specimens examined is relatively shallow
and rounded, not angular. A fold is not developed on the brachial valve.

There are no radial plications. The concentric ornament consists of numerous very fine,
concentric growth lines and a few more strongly accentuated ones, more commonly on
the anterior of larger shells. The fine ornament consists of concentric rows of fine radial
striae along the anterior edge of the growth lamellae, but not continuing back to the
anterior edge of the previous growth lamellae.

174

PALAEONTOLOGY, VOLUME 9

Interior of pedicle valve. The hinge teeth are strong knob-like projections, directed
dorsally and converging slightly postero-medially. They are supported basely by stout
dental lamellae that initially converge slightly toward the midline and diverge slightly
from it along their anterior edges. The dental lamellae are very closely set together at
the umbo and constrict that space in most specimens making the pedicle cavity obsole-
scent. The ventral muscle scar is deeply impressed, triangular, and elongate with slightly
divergent sides laterally and a semicircular outline anteriorly where it blends with the
interior of the shell. The adductor muscle scars compose a pair of elongate raised ridge-
like tracks medially which are separated by a subangular myophragm. On some speci-
mens the adductor tracks appear to extend slightly anterior to the anterior edge of the
elongate oval anterior portion of the diductor impressions.

Interior of brachial valve. The hinge plates define relatively strong, deep, subconical, and
broadly divergent sockets that are attached beneath the inner edges of the dorsal inter-
area. Commonly, there is a small rim of shell material extending laterally from the distal
end of the socket beneath the interarea and parallel to it. Medially the hinge plates bear
a pair of triangular crural plates that are attached to the inner surface of the posterior
end of the valve, but which are free of the floor of the valve along most of their length.
The crural plates are relatively strongly concave on their inner (medial) faces and with
their dorsal sides thus attaining a position subparallel to the interior of the umbo of the
valve. Medially in the notothyrial cavity, the site of diductor attachment is composed
of a deeply striate, transverse, comb-like process. The adductor muscle scars comprise
two pairs of very elongate subelliptical impressions with the medial pair more deeply
impressed and separated from the lateral pair by thin rounded ridges. The medial pair
are divided by a thin low myophragm throughout most of their length.

Comparison. The new subspecies is distinguished from Warrenella kirki kirki by a more
transverse outline and a less well-developed median sulcus in the pedicle valve of the
former. A typical population of W. kirki kirki includes transverse specimens, but these
commonly bear a relatively strong subangular sulcus. Included with these are an almost
equal number of elongate suboval specimens. W. kirki praekirki appears to be the initial
subspecies in Nevada, subject to a trend from transverse to elongate specimens. Some
collections above W. kirki kirki contain an abundance of markedly elongate forms and
lack of any of the transverse ones typical of the early forms or which are also found in
the middle stratigraphic horizons. W. occidentalis is a very large transverse species with
a relatively deep median sulcus which may be relatively flat-bottomed in some forms. In
the writer’s opinion, the Canadian species W. franklini is very close to the transverse
forms of kirki kirki and may prove to be a senior synonym of that species. W. occidentalis
may have been derived from the later form at a time when the typical kirki kirki gave
rise to two species out of its diverse population.

Crickmay (1962, p. 10) has recently assigned his species W. timetea to the genus
Tingella Grabau, because of the presence of crural plates. Timetea is very close to occi-
dentalis of Nevada and may be conspecific, however, the writer has not studied large
collections of the Canadian species. Specimens of occidentalis from Nevada also bear
relatively large crural plates as do all the specimens assigned to kirki and its various
subspecies discussed here. However, the writer is not convinced that the information on
the structural details and external ornament of Tingella are well enough known to bring

J. G. JOHNSON: MIDDLE DEVONIAN BRACHIOPODS

175

that name into widespread usage in western North America. All of the reticulariid
genera in the Devonian that the writer has studied from Nevada and including a number
of species from both the Lower and Middle Devonian, bear crural plates that are
attached to the base of the valve along varying percentages of their total length. It seems
appropriate that the type species of Warrenellci be much more closely studied with
regard to the structure of the cardinalia to determine whether crural plates are present
and buried in secondary shell material, or present and not in contact with the base of
the valve, or are completely absent.

Grabau’s only figured internal mould of Tingella (1933, pi. 43, fig. 7), in addition to
showing long slots for the crural plates, also shows thin, discrete, widely spaced, dental
lamellae in the pedicle valve and a muscle field that is not impressed. W. kirki typically
bears dental lamellae that are closely set and adjoin umbonal cavities that bear con-
siderable secondary shell material. This suggests to the writer that W. kirki , despite its
well-developed crural plates, is not closely related to Tingella.

Warrenel/a appears to be a senior subjective synonym of Minatothyris Vandercammen
1957, which is based on a species from the Frasnian of western Europe ( Spirifer eury-
glossus Schnur), a species which is in fact very close to the type of Warrenel/a. The latter
genus, as viewed by the writer and discussed above, is also a senior subjective synonym of
Sinothyris Minato 1953, which is based on the west European Givetian species S. maureri
Holzapfel. This, on exterior considerations at least, appears to be a species relatively
close to the group of franklini, occidentalism and timetea. Vandercammen (1957) has
already included maureri in Minatothyris and it is evident that Minato’s conception of
Sinothyris based on Grabau’s illustration and description was erroneous since the type
species lacks a ventral median septum.

Occurrence. W. kirki praekirki is at present known only from the Leptathyris circula zone in Eureka
County, Nevada.

Warrenel/a cf. colwnbina colwnbina (Havlicek 1959)

Plate 27, figs. 6-9

1959 Eomartiniopsis colwnbina colwnbina Havlicek, p. 190, pi. 28, figs. 3-5.

Material figured. USNM no. 145529.

Exterior. The shells are transversely subquadrate in outline and strongly biconvex in
lateral profile with the pedicle valve more strongly so than the brachial valve. The beak
on the pedicle valve is relatively short and stubby. The hinge line is short and straight
medially and there is a narrow triangular apsacline interarea, slightly incurved near the
beak, and equal to a little less than half the maximum width of the valves. The interarea
is cleft medially by a triangular open delthyrium enclosing an angle 55-60°. The umbo
on the brachial valve is relatively prominent, strongly convex, and the beak is strongly
incurved. The cardinal angles are obtuse and strongly rounded. The maximum width is
posterior to the midlength and the antero-lateral margins slope forward and converge
slightly toward the midline. There is a deep subangular sulcus on the pedicle valve,
originating near the umbo and becoming decidedly stronger and flaring in width toward
the anterior commissure. The brachial valve bears a rounded fold that is faintly marked
off from the lateral portions of the valve.

176

PALAEONTOLOGY, VOLUME 9

The ornament consists of numerous fine concentric growth lines. The fine radial orna-
ment, if present, was not detected due to inadequate preservation.

Interior of pedicle valve. There are a pair of short thick dental lamellae in the apex of the
valve that diverge only slightly from one another and which are situated very close
together at the beak. The diductor scar consists of a triangular, faintly impressed area
which blends anteriorly with the interior of the valve. The umbonal cavities are free of
infilling secondary shell material.

Occurrence. The species appears at two localities in the Leptathyris circula zone on the north side of the
Roberts Mountains and in a single collection from Lone Mountain which also contains specimens of
Warrenella kirki praekirki.

Family ambocoeliidae George 1931
[nom. transl. Ivanova 1960 {ex Ambocoeliinae George 1931)]

Genus echinocoelia Cooper and Williams 1935
Type species. E. ambocoelioides Cooper and Williams 1935, p. 844, pi. 59, figs. 9, 13, 17, 21, 23.

Echinocoelia denayensis sp. nov.

Plate 26, figs. 23-30

Material. The specimen, USNM no. 145527, illustrated in figs. 26-30 of Plate 26 is the holotype. It
measures 9 0 mm. in width, 8 0 mm. in length, and 7-6 mm. in thickness. Other figured specimens in-
clude USNM nos. 145524-6.

Diagnosis. Echinocoelia with typical form, but with rudimentary, anteriorly free, crural
plates.

Exterior. The shells are subtrigonal in outline and strongly unequally biconvex, pyra-
midal in lateral profile. The brachial valve is only gently curved, nearly flat and cap-like.
The pedicle valve is deeply convex and strongly arched with a high, steeply apsacline
to nearly catacline, only slightly incurved interarea which is equal to about three-
quarters of the maximum width of the valves. It is cleft medially by a high, triangular,
open delthyrium that encompasses an angle of about 30°. The interarea on the brachial
valve is long, low, and anacline. Most specimens are transverse, but become less so with
an increase in size. The outline of larger specimens, especially brachial valves, is some-
what pentagonal with the maximum width well posterior to midlength and with the
sides sloping antero-medially and truncated by a short anterior commissure at right
angles to the median plane. No fold or sulcus is developed on the pedicle valve but a very
faint flattening or depression of the interior edge of the brachial valve causes the
anterior commissure to be deflected slightly ventrally.

The valves are without plication but develop a fairly strongly lamellose ornament of
concentric growth lines that are very numerous and closely set and which are more pro-
nounced at a few irregular intervals. Preservation of the fine ornament is imperfect, but
on the best preserved specimens it appears to consist of extremely fine radial striae
situated in concentric rows on the growth lamellae.

Interior of pedicle valve. The hinge teeth are small, thin, elliptical processes which project
dorsally from the base of the hinge line and have their long axes approximately midway

J. G. JOHNSON: MIDDLE DEVONIAN BRACHIOPODS

177

between the hinge line and the median plane. The hinge teeth leave small ridge-like tracks
on the inner edges of the delthyrium. There are no dental lamellae. The apex of the
delthyrium is closed by a small plate.

Interior of brachial valve. The sockets are broad, shallow, and widely divergent antero-
laterally. They are attached on their medial edges to relatively large triangular crural
plates which contact the interior surface of the valve posteriorly, but which are free
anteriorly. Medially there is a small knob-like cardinal process in the apex of the noto-
thyrium. Commonly there is a long, low, narrow myophragm running most of the length
of the valve. The adductor muscle scars, however, are not impressed. Commonly, on the
more lamellose brachial valves, the interior is strongly crenulated in the anterior half of
the shell by the concentric undulations of the more prominent growth lamellae.

Occurrence. Echinocoelia denayensis is a common species in the Leptathyris circula zone on the north
side of the Roberts Mountains as well as on the west slope, east of Red Canyon. This species is also
present in the L. circula zone in the northern Simpson Park Range.

Comparison. Echinocoelia ambocoelioides Cooper and Williams (1935, p. 844, pi. 59) has
a more pyriform outline than E. denayensis and long crural plates, lacking in the latter
species. The new species is very close to E. incurva Cooper and Williams (1935, p. 845,
pi. 59, figs. 20, 24, 25) but has a more nearly catacline palintrope and intermittantly
coarse concentric growth lines. The interior of E. incurva is unfortunately not known.
E. denayensis very closely resembles Ambocoelia rogeri Drot (1955, p. 565, pi. 27, fig.
3 a-d). Drot (p. 566) mentions the presence of a septum (= fmyophragm) in the brachial
valve, but excludes her new species from Echinocoelia which it resembles in most
characters. E. denayensis differs from E. ? rogeri in having a relatively flatter brachial
valve that bears a less well pronounced ventral deflection of its anterior commissure. The
palintrope of E. ? rogeri is less incurved at the beak and there are some minor differences
in surface ornament as a comparison of the illustrated specimens will indicate, but it is
not certain what value these may have.

Superfamily cyrtinacea Fredericks 1912
[nom. transl. Johnson herein]

Discussion. Inclusion of the cyrtinids in the Spirifereinacea has been suggested (Ivanova
1960, p. 278), however, such an assignment should not be taken to indicate more than
a morphologic grouping of the punctate Spiriferoidea. If phylogenetic implications were
assumed, this procedure would require Spiriferina and its allies to have been derived
from Cyrtina, a connexion regarded as improbable by the writer since Cyrtina possessed
a more specialized morphology, exemplified by its tichorhinum and stout apically per-
forate deltidium, long before the appearance of the more generalized spiriferinaceids.
Cyrtina appeared at or near the beginning of the Early Devonian and represented a
major new stock contrasting markedly with the then existing Cyrtiidae (the eospiriferids)
and the Delthyridae (principally Delthyris and Howellella).

Family cyrtinidae Fredericks 1912
[nom. transl. Stehli 1954 {ex Cyrtininae Fredericks 1912)]

Genus cyrtina Davidson 1858

Type species. Calceole heteroclite Defrance 1824, p. 306, pi. 80, figs. 3, 3 a.

N

B 6612

178

PALAEONTOLOGY, VOLUME 9

Cyrtina sp.

Plate 27, figs. 1-5

Material figured. USNM no. 145528.

Exterior. The shells are transversely subtrigonal to sub-semicircular in outline and
strongly unequally biconvex in lateral profile. The pedicle valve is strongly convex, sub-
pyramidal, and the brachial valve is gently convex and cap-like. The pedicle valve bears
a high, nearly flat, catacline, triangular interarea cleft medially by a long delthyrium
enclosing an angle of about 20°. The deltidium is not preserved. There is a shallow sub-
angular sulcus on the pedicle valve and a corresponding strong rounded fold on the
brachial valve and three or four rounded plications on each flank of the valves separated
by shallow U-shaped interspaces.

Occurrence. Cyrtina sp. is present, but not common in the Leptathyris circula zone on the north side of
the Roberts Mountains.

APPENDIX OF LOCALITIES

The numbers listed below are those of the locality register of the Department of Geology, University
of California, Los Angeles.

3536 North-east slope of Union Mtn., north-east of McColley Canyon, elev. 6,825, 900 ft. W., 3,350 ft.
N. of SE. cor. of sec. 20, T. 27N., R. 53E., Mineral Hill quad., Sulphur Spring Range, Elko Co.,
Nevada. Telegraph Canyon Member of Nevada Fm. (Carlisle et al. 1957).

4737 Approx. 4,500 ft. due E. of Roberts Creek on small spur on south slope of hill 7964, elev. 7,420,
sec. 1 (unsurveyed) T. 22N., R. 50E., Roberts Creek Mtn. quad., Roberts Mts., Eureka Co., Nevada.
Denay Ls., M. Dev., 20-50 ft. above top of McColley Canyon Fm. Collector: E. L. Winterer, 1959.

4738 Approx. 3,300 ft. due E. of Roberts Creek, elev. 6,920, eastward across canyon from hill 6963,
7,400 ft. almost due N. of NE. cor. sec. 24, T. 22N., R. 50E. Roberts Creek Mtn. quad., Roberts Mts.,
Eureka Co., Nevada. Denay Ls., M. Dev., about 300 ft. above top of McColley Canyon Fm. Collector:
E. L. Winterer, 1959.

4739 Approx. 3,200 ft. due W. of Willow Creek, elev. 7,520, at crest of ridge between Willow Creek and
Birch Creek, 4,200 ft. south of SE. cor. of sec. 22, T. 24N., R. 50E., Roberts Creek Mtn. quad.,
northern Roberts Mts., Eureka Co., Nevada. Denay Ls., M. Dev., approx. 170 ft. above top of McCol-
ley Canyon Fm. Collectors: Student group 11, 1958 and J. G. Johnson, 1959.

4740 East of Willow Creek, elev. 6,900 ft., 2,800 ft. E., 3,700 ft. S. of SE. cor. of sec. 15, T. 24N.,

R. 50E., Roberts Creek Mtn. quad., northern Roberts Mts., Eureka Co., Nevada. Lower Denay Ls.,
M. Dev. Collector: J. G. Johnson, 1959.

4741 Elev. approx. 6,800 ft., on east flank of Willow Creek, approx. 6,500 ft. upstream from Willow
Creek Ranch, 1,900 ft. S., 2,300 ft. E. of SE. cor. of sec. 22, T. 24N., R. 50E., Roberts Creek Mtn.
quad., northern Roberts Mts., Eureka Co., Nevada. Denay Ls., M. Dev. Collector: J. G. Johnson,
1959.

4742 South flank of Hanson Creek, north of Red Canyon, elev. 7,320 ft., 7,900 ft. due E. of SE. cor.
of sec. 1, T. 23N., R. 49E., west slope of Roberts Creek Mtn., Roberts Creek Mtn. quad., Eureka Co.,
Nevada. Denay Limestone, M. Dev. Collectors: Student group 14, 1958.

4743 South-west slope of hill 8919, east of Red Canyon, elev. 8,040 ft., 10,400 ft. E., approx. 1,200 ft.

S. of SE. cor. of sec. 12, T. 23N., R. 49E., Roberts Creek Mtn. quad., Eureka Co., Nevada. Denay
Limestone. M. Dev. Collector: unknown.

4744 Elev. 7,880 ft., head of Birch Creek, crest of ridge c. 5,500 ft. NNW. of Cooper Peak, and c.
9,100 ft. ESE. of Western Peak, T. 23N., R. 50E., northern Roberts Mts., Roberts Creek Mtn. quad.,
Eureka Co., Nevada. Denay Limestone, M. Dev. Collector: M. A. Murphy, 1957.

J. G. JOHNSON: MIDDLE DEVONIAN BRACHIOPODS

179

4745 Elev. 8,640 ft. east slope of saddle 500 ft. N. of hill 8788, approx. 3 miles W. and 3 miles N. of
Roberts Creek Ranch, T. 22N., R. 50E., southern Roberts Mts. west of Roberts Creek, Roberts Creek
Mtn. quad., Eureka Co., Nevada. Denay Limestone, M. Dev. Collector: E. L. Winterer, 1959.

Acknowledgements. The writer wishes to thank Dr. G. Arthur Cooper of the U.S. National Museum,
Washington, who kindly loaned the types of Walcott’s species Cryptonella ? circula, Skenidium
devonicum, and Leiorhynchus nevadensis for comparative purposes. Dr. Vladimir Havlicek, Geological
Survey of Czechoslovakia, generously made available specimens of Prokopia bouskai and these have
been deposited with the U.S. National Museum. The writer's work at Pasadena has been supported
by National Science Foundation Grant No. GP 2290.

REFERENCES

alekseeva, r. e. 1960. Devonian Atrypida of the Kuznetsk and Minusinsk Bassins and the east slope
of the north Ural. Akad. nauk. SSSR, Siberian Div., Inst. Geol. Geophy. 1-196, pi. 1-12, Moscow.

barrande, j. 1847. Uber die Brachiopoden der Silurischen Schichten von Bohmen, I. Naturwissen-
schaftliche Abhandlungen ( Haidingers ), 1, Wien.

1879. Systeme Silurian du centre de la Boheme, 5, Brachiopodes. 1-226, pi. 1-153, Prague, Paris.

boucot, a. j., Johnson, j. g., and staton, r. d. 1964. On some atrypoid, retzioid, and athyridoid
Brachiopoda. J. Paleont. 38, 805-22, pi. 125-8.

branson, e. b. 1923. General account of Missouri Devonian in The Devonian of Missouri. Missouri
Bur. Geol., Mines , 17, 2nd ser. 1-279, pi. 1-71.

Carlisle, d., murphy, m. a., nelson, c. A., and winterer, e. l. 1957. Devonian stratigraphy of Sulphur
Springs and Pinyon Ranges, Nevada. Bull. Amer. Ass. Petrol. Geol. 41, 2175-91.

conrad, t. A. 1841. Fifth annual report on the paleontology of the state of New York. N.Y. St. Geol.
Surv. Ann. Rept. 5, 25-57.

cooper, g. A. 1955. New Genera of Middle Paleozoic Brachiopods. J. Paleont. 29, 45-63, pis. 11-14.

et al. 1942. Correlation of the Devonian sedimentary formations of North America. Bull. geol.

Soc. Amer. 53, 1729-94, 1 pi.

and cloud, p. e. 1938. New Devonian fossils from Calhoun County, Illinois. J. Paleont. 12, 444-

60, pi. 54, 55.

- — — and williams, j. s. 1935. Tully formation of New York. Bull. geol. Soc. Amer. 46, 781-868,
pi. 54-60.

crickmay, c. h. 1953. Warrenella, a new genus of Devonian brachiopods. J. Paleont. 27, 596-600.

1960. The older Devonian faunas of the Northwest Territories. Published by the author. Imperial

OH Ltd., 1-21, pi. 1-11. Calgary.

1962. New Devonian fossils from western Canada. Published by the author. Imperial Oil Ltd.,

1-16, pi. 1-9, Calgary.

1963. Significant new Devonian brachiopods from western Canada. Published by the author,

Imperial Oil Ltd., 1-62, pi. 1-16. Calgary.

defrance, m. j. l. 1 824. Dictionnaire des Sciences Naturelles. 32.

drot, jeannine. 1955. Quelques brachiopodes nouveaux ou rares du Devonien Marocain. Bull. Soc.
geol. Fr., 5, ser. 6, 563-9, pi. 27.

FORTIER, Y. O., BLACKADAR, R. G., GLENISTER, B. F., GREINER, H. R., MCLAREN, D. J., MCMILLAN, N. J.,
NORRIS, A. W., ROOTS, E. F., SOUTHER, J. G., THORSTEINSSON, R., and TOZER, E. T. 1963. Geology of the
north-central part of the Arctic Archipelago, North-west Territories (Operation Franklin). Mem.
geol. Surv. Can. 320, 1-671, illus.

grabau, a. w. 1931, 1933. Devonian Brachiopoda of China. China Geol. Surv., Pal. Sinica, 3, ser. B,
fasc. 3, 1-545 (1931), pi. 1-54 (1933).

hall, j. 1858. Part 2, Paleontology, Rept. on the Geol. Survey of the state of Iowa, 473-724.

1867. Descriptions and figures of the fossil Brachiopoda of the upper Helderberg, Hamilton,

Portage, and Chemung Groups. New York Geol. Surv., Palaeontology, 4, 1-428, pi. 1-63.

and clarke, j. m. 1893, 1895 (1894). An introduction to the study of the genera of Palaeozoic

Brachiopoda. Palaeontology of New York, 8, pt. 2, 1-317 (1893), 319-94, pi. 21-84 (1895).

180

PALAEONTOLOGY, VOLUME 9

havucek, v. 1952. A Paleontological study of the Devonian of Celechovice-Brachiopods (Penta-
meracea, Rhynchonellacea, Spiriferacea). Shorn, ustr. Ust. geol. 18, 1-20, pi. 1-4.

- 1953. O nekolika novych ramenonozcich ceskeho a moravskeho stredniho devonu. Vestn. ustr.
Ust. geol. 28, 4-9, pi. 1, 2.

1959. Spiriferidae v ceskem siluru a devonu. Czechoslovakia , Rozpr. ustr. Ust. geol. 25, 1-275.

pi. 1-28.

house, m. r. 1962. Observations on the ammonoid succession of the North American Devonian. J.
Paleont. 36, 247-84, pi. 43-48.

1965. Devonian goniatites from Nevada. Neues Jb. Geol Paldont., Abh., 122, 337-42, pi. 32.

imbrie, j. 1959. Brachiopods of the Traverse Group (Devonian) of Michigan. Part 1. Dalmanellacea,
Pentameracea, Strophomenacea, Orthotetacea, Chonetacea, and Productacea. Bull. Anter. Mus. nat.
Hist. 116, 345-410, pi. 48-67.

ivanova, e. a. 1960. Order Spiriferida in sarycheva, t. g., Osnovi Pa/eonto/ogii , Bryozoa, Brachiopoda.
264-80, pi. 57-64.

Johnson, j. g. 1962n. Brachiopod faunas of the Nevada Formation (Devonian) in central Nevada.
J. Paleont. 36, 165-9.

19626. Lower Devonian-Middle Devonian boundary in central Nevada. Bull. Amer. Ass. Petrol.

Geol. 46, 542-6.

(in prep.). Great Basin Lower Devonian Brachiopoda.

khodalevich, a. n. 1951. Lower Devonian and Eifelian brachiopods of the Ivdel and Serov districts
of the Sverdlovsk region. Trud. Sverdlovsk min. Inst. 18, 1-169, pi. 1-30.
mclaren, D. j. 1962. Middle and early Upper Devonian rhynchonelloid brachiopods from western
Canada. Bull. Can. geol. Surv. 86, 1-122, pi. 1-18.

and norris, a. w. 1964. Fauna of the Devonian Horn Plateau Formation, District of Mackenzie.

Ibid. 114, 1-74, pi. 1-17.

norris, A. w., and mcgregor, d. c. 1962. Illustration of Canadian fossils — Devonian of Western

Canada. Pap. Can. geol. Surv. 62-4, 1-34, pi. 1-16.
meek, F. b. 1 868. Remarks of the Geology of the valley of Mackenzie River, with figures and descriptions
of fossils from that region, in the Museum of the Smithsonian Institution, chiefly collected by the
late Robert Kennicott, Esq. Trans. Chicago Acad. Sci., 1, 61-114, pi. 12-14.
merriam, c. w., 1940. Devonian stratigraphy and paleontology of the Roberts Mountains region,
Nevada. Spec. Pap. geol. Soc. Amer. 25, 1-114, pi. 1-16.

1963. Paleozoic rocks of Antelope Valley, Eureka and Nye Counties, Nevada. Prof. Pap. U.S.

geol. Surv. 423, 1-67, 2 pi.

and anderson, c. a. 1942. Reconnaissance survey of the Roberts Mountains, Nevada. Bull.

geol. Soc. Amer. 53, 1675-728.

meyer, oskar-erich. 1913. Die Devonischen Brachiopoden von Ellesmereland. Videnskabs-Selskabet
i Kristiania , Rept. 2nd Norwegian Arctic Exped. in the ‘ From ’ 1898-1902, 29, 1-43, pi. 1-8.
minato, m. 1953. On some reticulate Spiriferidae. Trans. Proc. Japan Palaeont. Soc., N.s., 11, 65-73.
nalivkin, d. v. 1936. The Middle Paleozoic faunas of the head parts of the Kolyma and Kandyga
rivers in The Paleozoic faunas of the Kolyma. Contr. to Knowledge of Okhotsk-Kolyma Land, ser. 1 ,
fasc. 4, State Trust Dalstroy, 1-28, pi. 1, 2.

nettleroth, h. 1889. Kentucky fossil shells; a monograph of the fossil shells of the Silurian and
Devonian rocks of Kentucky. Kentucky Geol. Surv. 1-245.
nolan, t. b., merriam, c. w., and williams, j. s. 1956. The stratigraphic section in the vicinity of
Eureka, Nevada. Prof. Pap. U.S. geol. Surv. 276, 1-77, 2 pi.
owen, D. D. 1852. Report of a geological survey of Wisconsin, Iowa, and Minnesota and incidentally of
a portion of Nebraska Territory. 1-638, Philadelphia.
pedder, a. e. h. 1959. Monelasmina besti, a new schizophoriid brachiopod from the Upper Devonian
of western Canada. Geol. Mag. 96, 470-2, pi. 16.

schnur, j. 1853. Zusammenstellung und Beschreibung sammtlicher im Uebergangsgebirge der Eifel
vorkommenden Brachiopoden. Palaeontographica , 3, 169-247, pi. 22-45.
schuchert, c. 1 897. A Synopsis of American fossil Brachiopoda including bibliography and synonymy.
Bull. U.S. geol. Surv. 87, 1-464.

siehl, a. 1962. Der Greifensteiner Kalk (Eifelium, Rhenisches Schiefergebirge) und seine Brachiopoden

J. G. JOHNSON: MIDDLE DEVONIAN BRACHIOPODS 181

fauna 1. Geologie; Atrypacea und Rostrospiracea. Palaeontographica, 119, pt. A., 173-221, pi.
23-40.

termier, H. 1936. Etudes Geologiques sur le Maroc central et le rnoyan Atlas septentrional. Notes
Serv. Min. geol. Maroc , 3, no. 33, Cinquieme partie (Paleontologie), 1088-421, pi. 1-23.
vandercammen, a. 1957. Revision de Spirifer euryglossus Schnur 1851, = Minatothyris nov. gen.

euryglossa (Schnur) (Brachiopoda, Devonien superieur). Senckenbergiana letli. 38, 177-93, pi. 1-3.
vanuxem, L. 1842. Geology of New York, Part III, comprising the survey of the third geological district.
1-306, Albany.

walcott, c. d. 1884. Paleontology of the Eureka District. Mon. U.S. geol. Surv. 8, 1-298, pi. 1-24.
warren, p. s. 1944. Index brachiopods of the Mackenzie River Devonian. Trans, roy Soc. Can. sec. 4,
105-30, pi. 1, 2.

and stelck, c. R. 1962. Western Canadian Givetian. Jour. Alberta Soc. Petrol, geol. 10, 273-91.

white, c. A. 1862. Description of new species of fossils from the Devonian and Carboniferous rocks
of the Mississippi Valley. Proc. Boston Soc. Nat. Hist. 9, 8-33.
williams, A. and wright, a. d. 1963. The classification of the ‘ Orthis testudinaria Dalman’ group of
brachiopods. /. Paleont. 37, 1-32, pi. 1, 2.

J. G. JOHNSON

Div. Geological Sciences,
California Inst. Technology,

Manuscript received 21 December 1964 Pasadena, California

THE PALAEONTOLOGICAL ASSOCIATION

COUNCIL 1965-6

President

+Dr. L. R. Cox, British Museum (Natural History), London
Vice-Presidents

Dr. W. S. McKerrow, University Museum, Oxford
Professor F. H. T. Rhodes, University College, Swansea

Treasurer

Dr. C. Downie, Department of Geology, The University, Mappin Street, Sheffield 1

Secretary

Dr. C. H. Holland, Department of Geology, Bedford College, London, N.W.l

Editors

Mr. N. F. Hughes, Sedgwick Museum, Cambridge
Dr. Gwyn Thomas, Department of Geology, Imperial College of Science, London, S.W.7
Dr. I. Strachan, Department of Geology, The University, Birmingham, 15
Dr. M. R. House, University Museum, Oxford

Other members of Council

Dr. C. G. Adams, British Museum (Natural History), London
Professor P. M. Butler, Royal Holloway College, Surrey
Dr. W. J. Clarke, British Petroleum Company, Sunbury-on-Thames
Dr. G. Y. Craig, The University, Edinburgh
Dr. T. D. Ford, The University, Leicester
Dr. B. M. Funnell, The University, Cambridge
Dr. J. M. Hancock, King’s College, London
Dr. G. A. L. Johnson, The University, Durham
Dr. F. A. Middlemiss, Queen Mary College, London
Mr. M. Mitchell, Geological Survey and Museum, London
Dr. W. D. I. Rolfe, Hunterian Museum, The University, Glasgow
Dr. A. J. Rowell, The University, Nottingham
Professor Scott Simpson, The University, Exeter
Dr. L. B. H. Tarlo, The University, Reading
Dr. H. Dighton Thomas, British Museum (Natural History), London

Overseas Representatives

Australia: Professor Dorothy Hill, Department of Geology, University of Queensland, Brisbane
Canada : Dr. D. J. McLaren, Geological Survey of Canada, Department of Mines and Technical
Surveys, Ottawa, Ont.

India: Professor M. R. Sahni, 98, The Mall, Lucknow (U.P.), India

New Zealand : Dr. C. A. Fleming, New Zealand Geological Survey, P.O. Box 368, Lower Hutt
West Indies and Central America-. Mr. John B. Saunders, Geological Laboratory, Texaco Trinidad,
Inc., Point k Pierre, Trinidad, West Indies

Eastern U.S.A. : Professor H. B. Whittington, Museum of Comparative Zoology, Harvard Univer-
sity, Cambridge 38, Mass.

Western U.S.A. : Professor J. Wyatt Durham, Department of Paleontology, University of California,
Berkeley 4, Calif.

f Dr. Cox died in August.

PALAEONTOLOGY

VOLUME 9 ' PART 1

CONTENTS

Schizochroal eyes and vision of some Silurian acastid trilobites. By e. n. k.

CLARKSON 1

Marine benthos, substrate and palaeoecology. By G. Y. craig and N. s. jones 30

Upper Ordovician trilobites from northern Yukon. By a. c. lenz and m.

CHURKIN 39

Silurian Girvanella from the Welsh borderland. By h. m. Johnson 48

Late Permian trilobites from the Salt Range, West Pakistan. By r. e. grant 64

The microstructure of stromatoporoids. By c. w. stearn 74

An improved method of analysing distortion in fossils. By k. sdzuy 125

On Birgeria acuminata and the absence of labyrinthodonts from the Rhaetic.

By r. j. g. savage and n. f. large 135

The Upper Devonian gastropod Orecopia in western Canada. By a. e. h.
pedder 142

The Silurian rugose coral Microplasma lovenianum Dybowski from

Monmouthshire. By D. e. white 148

Middle Devonian brachiopods from the Roberts Mountains, central Nevada.

By j. g. Johnson 152

PRINTED IN GREAT BRITAIN AT THE UNIVERSITY PRESS, OXFORD
BY VIVIAN RIDLER, PRINTER TO THE UNIVERSITY

VOLUME 9 * PART 2

Palaeontology

JULY 1966

PUBLISHED BY THE
PALAEONTOLOGICAL ASSOCIATION
LONDON

Price £3

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, etc., on payment of the
appropriate annual subscription:

Institute membership .... £7 (U.S. $20.00)

Ordinary membership . . . £3. 3s. (U.S. $9.50)

Student membership . . . £2. 2s. (U.S. $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; on first
applying for membership, they should obtain an application form from the Secretary
or the Treasurer. Subscriptions are due each January, and should be sent to the
Treasurer, Dr. C. Downie, Department of Geology, The University, Mappin Street,
Sheffield 1, 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
1966 will receive Volume 9, Parts 1 to 4.

All back numbers are still in print and may be ordered from B. H. Blackwell,
Broad Street, Oxford, England, at the prices shown below (post free):

Vol. 1 (for 1957-8) in 4 parts at £2 or U.S. $6.00 per part.

Vol. 2 (for 1959) in 2 parts at £2 or U.S. $6.00 per part.

Vol. 3 (for 1960) in 4 parts at £2 or U.S. $6.00 per part.

Vol. 4 (for 1961) in 4 parts at £2 or U.S. $6.00 per part.

Vol. 5 (for 1962) in 4 parts at £3 or U.S. $9.00 per part.

Vol. 6 (for 1963) in 4 parts at £3 or U.S. $9.00 per part.

Vol. 7 (for 1964) in 4 parts at £3 or U.S. $9.00 per part.

Vol. 8 (for 1965) in 4 parts at £3 or U.S. $9.00 per part.

A complete set. Volumes 1-8, consists of 30 parts and costs £76 or U.S. $228.

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, Department of Geology, Sedgwick Museum,
Downing Street, Cambridge, England. A sheet of detailed instructions for authors
will be supplied on request.

The Palaeontological Association, 1966

A REVISION OF ACASTE DOWNING IAE
(MURCHISON) AND RELATED TRILOBITES

by J. H. SHERGOLD

Abstract. Three species of Acaste, A. clowningiae (Murchison),/!, inflat a (Salter), and A. subcaudata ( M urchison) ;
two species of Acastocepliala gen. nov., A. macrops (Salter) and A. dudleyensis sp. nov. ; and Acastoides constrieta
(Salter) are described from rocks of Wenlockian age from the West Midlands, Welsh Borderlands, and South
Wales. Their relationships to later Silurian and early Devonian acastomorph genera are discussed.

The species Calymenel downingiae was first recognized by Murchison ( 1 839) who divided
it from Calymene macrophthalma Brongniart on account of its smaller eyes and well-
developed glabellar furrows. The species was subsequently placed in the genus Phacops
by Emmerich (1845) and by Salter (1848, 1853), M’Coy (1851), and Murchison (1859).
Goldfuss (1843) established the genus Acaste, which was regarded as a subgenus of
Phacops by Salter (1864). Acaste downingiae was the first of a group of eight species listed
by Goldfuss (1843, p. 563) and has been regarded by subsequent authors as the type
species although this was not definitely established until 1959 (Struve in Moore, 1959,
p. 0488). R. Richter (1909) and R. and E. Richter (1939, 1952, 1954) discussed the mor-
phology, relationships, and successions of acastomorph trilobites but concentrated their
attention mainly on Devonian genera.

Salter (1864) divided the species into six varieties which he distinguished by Greek
letters. At the present time these are assigned to the following taxa. After the revision
of R. and E. Richter (1954, p. 16) var. a, vulgaris becomes Acaste downingiae (Murchison)
s.s. and var. e, constrictus is assigned to the genus Acastoides Delo 1935, becoming
A. constrieta (Salter). Var. 8, spinosus was chosen by Reed in 1925 as the type species
of Acaste/la. Two further varieties are revised below. Var. y, inflatus has been retained
in the genus Acaste as A. inflata (Salter). For trilobites previously referred to var. /?,
macrops a new genus, Acastocepha/a, has been erected to include two species, A. macrops
(Salter), the type species and A. dudleyensis sp. nov. Salter’s var. £, cuneatus, based on
a poorly preserved cephalon from the Denbigh Grits of Llanrwst, is the only specimen
known from that area and until further material becomes available cannot be adequately
described.

The present paper is based mainly on materials from the Wenlockian of the West Midlands, Welsh Border-
lands, and South Wales in the following collections: the Sedgwick Museum, Cambridge (SM); Oxford Univer-
sity Museum (OUM); Birmingham University Museum (BU); the British Museum (Nat. Hist.) (BM); the
Geological Survey Museum (GSM); the Geological Society Collection (GSC); and the National Museum of
Wales (NMW). Additional material has been used from the Bristol City Museum; the Ludlow Museum; the
Hancock Museum, Newcastle upon Tyne; the Departmental Collections, University of Newcastle upon Tyne
(UNT), and from collections made by the author in South Wales.

Relationships of considered genera. Struve (19586) has used the term ‘acastomorphen’
for Phacopacea with dalmanitid tendencies as distinct from those with calmoniid ten-
dencies. The writer has here used the term in a somewhat different sense to include

[Palaeontology, Vol. 9, Part 2, 1966, pp. 183-207, pis. 28-32.]

C 3803

o

184

PALAEONTOLOGY, VOLUME 9

the genera Acaste, Acastoides, Acastella, AcasteUina R. and E. Richter 1954, Acastava
R. and E. Richter 1954, and Acastocephala gen. nov., feeling that these genera form
a closely related Siluro-Devonian complex. In Struve’s (19586, 1959) current classifica-
tion these genera are grouped into two subfamilies, Acastinae Delo 1935 and Acas-
tavinae Struve 1958. The former is included in the family Calmoniidae and the latter
in the Dalmanitidae. The Acastavinae ( Acastava , Acastella, and AcasteUina) are thought
by Struve to show an asteropyginid trend in certain characters, namely the shape of
the cephalon, the construction of the genal spines and the poorly segmented pygidium
with its marked tendency to develop denticulations along the lateral margins. The
Acastinae, while including such genera as Acaste and Acastoides, embraces also Scotiella
and Phacopina, subgenera in which the glabellar segmentation has become faint or
obsolete. The mucronations possessed by members of the Acastinae are thought by
Struve to be essentially similar in construction to the Calmoniidae rather than the
Dalmanitidae.

During the early holaspid ontogeny of Acaste downingiae and A. subcaudata, short
but distinct genal projections are observed on the posterolateral fixigenal margins. These
projections, in effect miniature spines, are rapidly resorbed in succeeding moults. In
their construction these spines differ from those produced by the Dalmanitinae and
Asteropyginae which are massive continuations of the cephalic margins but are more
similar to the small, elegant mucronations shown by such typical calmoniid genera as
Bainella, Schizostylus, and Paracalmonia or to those of the acastavinid genus, Acastella.
In their orientation these small acastid spines contrast with those of the Calmoniidae
which are deflected prominently outwards from the line of the cephalic margin, but
compare well with those of the Acastavinae, being most closely comparable with the
mucronations displayed by the adult of Acastella prima Tomczykowa (1962a, pp. 261-2,
pi. 1, figs. 2, 4, 5). If the early holaspid mucronations of Acaste downingiae were to be
continued into the adult stage the condition of Acastella prima would be closely approxi-
mated (text-fig. 3). The derivation of Acastella from Acaste through an intermediate
form such as Acastella prima becomes a strong probability (Tomczykowa 19626, p. 202).

Acastocephala dudleyensis sp. nov. also possesses small genal mucronations in its
early holaspid ontogeny. In the adult condition these are again resorbed and the genae
become posterolaterally angled or rounded. The construction and orientation of these
spines are essentially similar to the condition shown by the adult of Acastava atava
(W. E. Schmidt 1907). Further similarities are shown between Acastocephala and
Acastava, for the lateral pygidial margins of young holaspides in both Acastocephala
macrops and A. dudleyensis are provided with three or four pairs of denticulations on
the external surface of the shell, similar to those found on the internal mould of the
adult Acastava atava (text-fig. 5). As with the genal spines, these denticulations are
resorbed in the adult which possesses a typically entire margin. Young holaspides,
therefore, show the asteropyginid trend of Struve, while the adults with rounded genae
and entire pygidial margins do not. In other characters, notably the size and position
of the eye and the poorly segmented pygidium, adults of Acastocephala macrops are
closely comparable to Acastava atava. It appears, therefore, that the conditions shown
by the early holaspides of Acastocephala are retained into the adult stages of Acastava.
Accordingly Acastocephala is thought to be ancestral to Acastava and is classified with
the Acastavinae.

J. H. SHERGOLD: A REVISION OF ACASTE DO WNINGIAE 185

Two lines of development may thus be postulated among the acastomorph genera.
Acastava (Siegenian-Emsian) may be derived from Acastocephala (Wenlockian) and
Acastella (Middle Ludlow-?Siegenian) from Acaste (Wenlockian-L. Gedinnian). The
various Silurian and Devonian species of Acastoides (Wenlockian-U. Emsian) appear
to be a closely related and distinct group, isolated from other lines of development
from Wenlockian times onwards.

text-fig. 1. Parameters recorded on a typical acastomorph cephalon.

Measurements. Measurements based on the lengths and widths of the cephalon, thorax
and pygidium have been found to be extremely variable and to be largely dependent
on preservation. Selected measurements connected with the eye, on the other hand,
have shown constancy both intra-generically and intra-specifically. The position of the
eye on the cheek with relation to the cephalic margins, posterior border furrow and
glabella and the differing relationships between eye length and glabellar length are
important in this respect.

The measurements recorded are basically those of Struve (text-fig. 1). They are essen-
tially the eye length, A ; glabellar length, G ; occipital glabellar length, Gn ; and a measure-
ment, H, across the posterior fixigena between the back of the eye and the posterior
border furrow. In addition to Struve’s parameters, the maximum width. Pa, and the
posterior width, Pb, across the palpebral area have been recorded. Of these latter
measurements. Pa has been found to depend upon the definition of the palpebral furrow
and to be more variable than Pb (see text-fig. 1).

After Struve (1958a), several ratios have been considered important. A/G, the larger
eye index; A/Gn, the smaller eye index, and H/A have been used to give some idea of
the eye length in relation to glabellar length and the distance of the eye from the pos-
terior border furrow. Values of H/A have shown that in some species the eye remains
in a more or less constant position during holaspid ontogeny, while in others it appears
to migrate in a posterior direction (see section under Acastocephala dudley easts').

186

PALAEONTOLOGY, VOLUME 9

111 the text the following glabellar notation is used; preoccipital lobe, 1L; median
lateral lobe, 2L; anterior lateral lobe, 3L; IS, 2S, and 3S are the equivalent glabellar
side furrows.

Generic and Specific Differentiation. Characters considered to be of generic importance
are as follows: the shapes of both cephalon and pygidium, i.e. the geometry of outline;
the segmentation of the pygidium, both in axial and pleural elements, and the size of
the eye and its position on the cheek. To illustrate the position of the eye and its relative
size the proportionate relationships between A, H, and Pb have been plotted on tii-
angular-based graph paper (text-fig. 2). Further, the development during ontogeny of
eenal mucronations and pygidial denticulations is of fundamental importance.

Of importance specifically are the convexities (tr. and sag.) of the frontal lobe, 1L
and the occipital ring; the height to which the occipital ring is raised above the glabellar
side lobes; the shape, in plan view, of the glabella; the number of facets at the maximum
height of the visual surface; the convexity (tr.) of the pygidium, both pleural and axial
elements; and the width of the pygidial border.

J. H. SHERGOLD: A REVISION OF ACASTE DO WNINGIAE

187

SYSTEMATIC DESCRIPTIONS

Family calmoniidae Delo 1935
Subfamily acastinae Delo 1935
Genus acaste Goldfuss 1843
Acaste downingiae (Murchison 1839)

Plate 28, figs. 1-18; Plate 29, figs. 1-8

1839 Calymene ? Downingiae (n.s.) Murchison, p. 655, pi. 14, figs. 3n(BU 54) and 3 b (specimen
untraced).

1851 Phacops Downingiae (Murchison sp.); M'Coy, p. 160 (SM A28766).

1853 Phacops Downingiae Murchison; Salter, pp. 1-12, pi. 1, figs. 1 (BM 44252), 2 (untraced),
4 (GSM 19305), 5-13 (untraced), 15 (untraced).

1864 Phacops ( Acaste ) Downingiae Murchison, var. a, vulgaris Salter, p. 26, pi. 2, figs. 17 (BM
44407), 18 (BU 54), 19-22 (untraced), 23 (BM 44409), 25 (BU 129), also an un-
numbered ventral surface between figs. 25 and 34 (BM 44410).

1864 Phacops (Acaste) Downingiae Murchison, var. y, inflatus Salter, p. 27, pi. 2, figs. 31 (GSM
19305) and 33 (GSM 19306).

1954 Acaste (Acaste) downingiae (Murchison 1839); R. and E. Richter, pp. 16-17, pi. 3, figs. 36
(BM In36153), 37-41.

1965 Acaste downingiae (Murchison 1839); Clarkson, p. 1 1, pis. 1, 2, pi. 3, figs. 1-3.

Lectotype (here designated). Murchison 1839, pi. 14, fig. 3 a, BU 54, figured here, pi. 28, fig. 1. Wenlock
Limestone, Wren’s Nest, Dudley, Worcs.

Diagnosis. A species of Acaste with frontal lobe anteriorly depressed and glabellar lobes
in general with low convexity (both tr. and sag.). Fixigenae of young holaspides
with short, stout mucronate projections which are rapidly resorbed in the succeeding
moults to give prominent angulations, these becoming less apparent with further in-
crease in size. Visual surface of eye carrying columns of 8 (rarely 9) lenses alternating
with 7 (rarely 8) at the maximum height. Pygidium with 7-9 axial segments, 6-7 pleural
segments. In young, pleural field and border area more or less continuous, becoming
separated in adult specimens by a shallow furrow or flattening.

Description. Cephalon subtriangular in outline, with narrow, triangular area of librigena
projecting in front of the glabella. Surface of test with fine granules. Length of cephalon
approximately -|— § width.

Glabella anteriorly rounded, laterally sub-parallel, transverse width decreasing grad-
ually and evenly to posterior; the axial furrows curving very gently round 2L and 3L.
Glabella apparently more parallel-sided in large adults than in younger stages. Frontal
lobe rounded both anteriorly and antero-laterally, with low to moderate convexity when
viewed in lateral profile, large adult specimens having a somewhat greater convexity
than immature specimens. 2L and 3L approximately equal in size, with slightly less
convexity (trans.) than frontal lobe; abaxially fused. 1L about half as wide (exsag.) as
2L, with little greater convexity (tr.). 3S long, shallow, sigmoidal even in smallest
specimens, with a marked posterior median deflection becoming gradually less distinct
during ontogeny. 2S shorter than 3S, impressed to a similar depth, transverse or gently
curved, with a faint posterior median deflection most pronounced in young forms;
failing laterally to reach the axial furrows. IS wider and deeper than 2S, of similar

188

PALAEONTOLOGY, VOLUME 9

length, deflected both medianly and abaxially to the anterior, becoming more transverse,
less wide (exsag.) and less deep in later ontogenetic stages. All furrows reaching to within
approximately equal distance of the sagittal line. Short median furrow on sagittal line
typically present, extending from median extremities of 3S to about one quarter of the
length of the frontal lobe.

Occipital furrow abaxially deeper than IS, becoming shallow but remaining well
defined sagittally. Occipital ring narrow (sag.), with slightly less transverse width but
greater transverse convexity than 1L; in lateral profile raised slightly above glabellar
side lobes at all stages of growth.

Genae gently sloping to a poorly defined border which passes antero-laterally into
a narrow (sag.) triangular area of librigena forming the anterior margin of the cephalon.
Preglabellar furrow narrow (sag.), shallow. Postocular section of facial suture cutting
lateral cephalic margin opposite IS or mid 2L; preocular section dorsal intramarginal,
separating preglabellar furrow from anterior cephalic margin. Fixigenae in smallest
holaspides produced postero-laterally into short, stout mucronate projections which in

EXPLANATION OF PLATE 28

Figs. 1-18. Acciste downingiae (Murchison 1839). 1, 2. BU 54; figured Murchison 1839, pi. 14, fig. 3a;
Salter 1864, pi. 2, fig. 18. Wenlock Limestone, Dudley, Worcs. 1, General view, lectotype arrowed,
X 1. 2, Lectotype, cranidium, x2. 3. BU 129; cephalon possibly figured Murchison 1839, pi. 14,
fig. 3 a (discussed R. and E. Richter 1954, pp. 16-17); Salter 1864, pi. 2, fig. 25. Wenlock Limestone,
Dudley, Worcs. Dorsal view of cephalon, x3. 4-8. BM 44407; figured Salter 1864, pi. 2, fig. 17.
Wenlock Limestone, Dudley, Worcs. 4, General view, X 1. 5, Detail, dorsal view of cephalon, x2.
6, Dorsal view of pygidium and thorax, x 2. 7, Lateral view of cephalon, X 2. 8, Eye showing lens
cavities, x8. 9. BU 129; Eye, x8. 10-12. GSM 19306; figured Salter 1853, pi. 1, fig. 15 and 1864,
pi. 2, fig. 33. Wenlock Limestone, Eastnor, nr. Ledbury, Herefordshire. 10, Lateral view, x3.
1 1 , Adult pygidium showing border furrow, dorsal view, x 3. 12, View from posterior, X 3. 1 3-14.
GSM 19313; Adult pygidium, Wenlock Limestone, Dudley, Worcs. 13, Dorsal view showing border
furrow, x2. 14, View from posterior, x2. 15-17. BM 44410; ventral surface figured Salter 1864,
pi. 2, no figure number given. Wenlock Limestone, Dudley, Worcs. 15, General view, x2. 16,
Hypostome, x6. 17, Articulation of posterior five thoracic segments and pygidium (on left), x8.
18. BM 44252; dorsal exoskeleton, figured Salter 1853, pi. 7, fig. 1. Wenlock Limestone, Dudley,
Worcs. x2.

EXPLANATION OF PLATE 29

Figs. 1-8. Acaste downingiae (Murchison 1839). 1. BM 44409; cephalo-thorax figured Salter 1864, pi. 2,
fig. 23. Wenlock Limestone, Dudley, Worcs. x 2. 2-3. SM A28717; Wenlock Limestone, Dudley,
Worcs. 2, Dorsal view of juvenile pygidium without border furrow, x 3. 3, View from posterior, X 3.
4-6. SM A28744; Wenlock Limestone, Dudley, Worcs. 4, Dorsal view of young holaspid showing
genal projection, x4. 5, Lateral view, x4. 6, Genal angle, x8. 7, NMW G.391.2; Wenlock Lime-
stone, Dudley, Worcs. Enlargement of genal angle showing resorption of genal projection, the
stage succeeding that of fig. 6, X 8.

Fig. 8. Acastel downingiae (Murchison 1839). GSM 19305; glabella figured Salter 1853, pi. 1, fig. 4,
and 1864, pi. 2, fig. 31. Wenlock Limestone, Ledbury, Herefordshire. Dorsal view, X2.

Figs. 9-16. Acaste inflata (Salter 1864). 9-14, OUM C9; Lectotype; figured Salter 1864, pi. 2, fig. 30.
Wenlock Limestone, Ledbury, Herefordshire. 9, Dorsal view of cephalon, x3. 10, Lateral view,
x3. 11, Dorsal view of pygidium, X3. 12, Posterior view of glabella showing convexities of 2L,
1L, and the occipital ring, x 4. 13, Eye, X 8. 14, Posterior view of pygidium, x 3. 15-16, BM In36154;
Wenlock Limestone, Dudley, Worcs. 15, Dorsal view of pygidium, x3. 16, Posterior view of
pygidium, x3.

Palaeontology, Vol. 9

PLATE 28

SHERGOLD, Acaste

Palaeontology, Vol. 9

PLATE 29

15

16

14

SHERGOLD, Acaste

J. H. SHERGOLD: A REVISION OF ACASTE DO WNINGIAE 189

the succeeding stages become gradually resorbed to give the angled adult condition (see
below). Posterior border furrow transverse, deep but narrow (exsag.).

Eyes crescentic in plan, situated centrally between cephalic margin and glabella, on
low ocular platform; extending from middle of 1L to anterior edge of 3L or to 3S;
A/G, 46-67 per cent.; A/Gn, 37-54 per cent. Postocular area of fixigena relatively large,
H/A 12-30 per cent. In lateral profile top of eye sloping gently to anterior; in anterior
profile, well below surface of glabella. Palpebral lobes high; palpebral furrows generally
well defined; palpebral areas relatively wide, with low inclination to axial furrows.

Visual surface overhanging slightly the ocular platform; gently convex or plane
outwards; sloping more steeply outward-backwards than outward-forwards. Up to 168
lenses arranged in roughly alternating vertical columns containing at the maximum
height of the surface 7 (8) or 8 (9) individual lenses, typically rounded or polygonal in
outline and with a diameter of 0-2-0-255 mm. Interspaces granular, the granules arranged
in a rough hexagonal network.

Hypostome subquadrate, anteriorly rounded, antero-laterally extended into short
wings. Posterior margin straight, postero-lateral corners truncated. Lateral borders
narrow, flat; posterior border more extensive. Median body adventrally convex, without
distinct median furrow. Maculae forming small indistinct patches in anterior half of
median body, close to lateral borders. Surface granulose.

Pygidium subtriangular in outline, margin entire at all growth stages; in posterior
profile slightly vaulted; in lateral profile posterior margin curving gently addorsally.
Axis raised above axial furrows and pleurae, moderately convex; composed of 8 axial
segments in young, 8 (rarely 9) in adults, tapering to posterior and with a rounded,
unsegmented terminal piece; segments 1-3 well defined, with strong transverse furrows;
remaining segments becoming gradually less distinct to posterior. Pleurae moderately
convex; 5 or 6 in young, 6 (rarely a trace of a seventh) in adults. Pleural furrows strong,
interpleural furrows apart from the first, much weaker. Border wide at all growth
stages. In young and juvenile forms the border is a smooth, unfurrowed extension of
the pleural area but specimens with a pygidial length of 7-50 mm. and greater rapidly
develop a flattening or shallow furrow at the junction of border and furrowed zones.
Length: width of pygidium approximately f-f.

Remarks. Although restricted to the variety a, vulgaris of Salter, there remains con-
siderable variation within the species. This is mostly accounted for by changes taking
place during the ontogeny. These changes may be summarized as follows. The glabella
becomes apparently more parallel-sided; the frontal lobe slightly more convex; the
furrows straighter; the strong posterior median deflections of the younger stages of
both 3S and 2S gradually disappear; IS becomes more transverse, less deep and less
wide. In the pygidium 5 (6) pleurae are recognized in young holaspides; 6 (7) in large
adults; a ninth axial segment may sometimes be discerned in large individuals. While
in young and juvenile holaspid stages the border of the pygidium is merely an unfurrowed
continuation of the pleural zone, the two become separated in adults with a cephalic
length in excess of 7-50 mm. by a flattening or shallow furrow. Linally, and of importance
phylogenetically, is the presence of short genal mucronations in very young holaspides
with a cephalic length of up to 4-40 mm. Six specimens ranging in cephalic length
between 2-72 and 4-40 mm. have been observed to possess these characters. The

190

PALAEONTOLOGY, VOLUME 9

mucronations are short and stout and are thrown off from the genal angle in a similar
manner to the genal spines of Acastella prima Tomczykowa and A. spinosa (Salter)
(text-fig. 3). They are not continuations of the cephalic outline as in the Dalmanitinae.
Specimens with cephalic lengths within the range 4-50-7-50 mm. have very acutely
angled genae with points in place of projections (text-fig. 3c), i.e. ‘a tubercle only in

text-fig. 3. The development of the genal angle in Acaste downingiae (Murchison), compared to the
genal condition of Acastella prima Tomczykowa and Acastella spinosa (Salter), a. Small holaspid of
Acaste downingiae (cephalic length 3-2 mm.) showing genal projections; UN T5034. x4. b. Small hol-
aspid of A. downingiae (cephalic length 4-3 mm.) SM A28744. X 4. c, A. downingiae , juvenile holaspid,
projection resorbed but genae remaining distinctly angled (cephalic length 8-8 mm.); NMW G.391.2.
x4. d, Acastella prima Tomczykowa. Drawing after Tomczykowa 1962a (text-fig. lc, p. 262). x4.
e, Acastella spinosa (Salter), holotype, GSM 19412. x4.

place of a spine' (Salter 1864, p. 25). The specimen figured by R. and E. Richter (1954,
pi. 3, fig. 39 a) shows this stage in the development of its genal angles. Specimens with
cephalic lengths greater than 7-50 mm. have less acutely angled genae, ‘rounded off’
(Salter, op. cit., p. 25), but still clearly show the positions from which the original pro-
jections were resorbed.

Apart from ontogenetic changes, further variation is observed in Acaste downingiae.
Rarely, specimens are encountered in which 2S actually reaches and joins with the axial
furrows in the adult stages. Equally rare are specimens in which not only 2S but also IS
fails abaxially to reach the axial furrow.

When values of A and H are plotted for 93 specimens of A. downingiae, the points
are found to fall into five distinct groupings lying on a gentle curve (text-fig. 4). The
possibility of a sixth is indicated by the solitary point at the extreme end of the graph.
These groupings are interpreted as instars. Each group overlaps the preceding in values
of eye length. Further, within each group, with the exception of the first, there appears
to be a division into two fields, which is tentatively suggested as being the result of
dimorphism.

Relationships. The relationships of A. downingiae (Murchison) to A. inflata (Salter) are
dealt with under comments on the latter. R. and E. Richter (1954, p. 17) have made
comparisons with their species, A. dayiana. The convexity and degree of tapering of the
glabella and the number and distribution of the facets of the visual surface serve to
differentiate the two species. No details of the ontogeny of A. dayiana are given. R. and
E. Richter (1954, pi. 4, figs. 52-56) show, however, that when preserved with the shell.

191

J. H. SHERGOLD: A REVISION OF ACASTE DO WNINGIAE

the border of the pygidium and the furrowed pleural area are continuous, but in the
internal mould a flattening separates the two areas.

As mentioned above, the adult of AcasteUa prima Tomczykowa possesses the mucro-
nate projections characteristic of the smallest preserved holaspid stage of A. downingiae
(text-fig. 3). The close relationships of the genus AcasteUa to members of the Acastinae

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 l.l 1.2 1.3

Back of Eye Posterior Border Furrow.. H mm.

text-fig. 4. The eye length plotted against the distance of the eye from the posterior border furrow
for 93 specimens of Acaste downingiae (Murchison). The graph shows groupings into five instars, with

the possibility of a sixth.

are demonstrated further by Tomczykowa (19626, p. 202) who derives Scotie/la from it
during the Upper Silurian, both genera occurring together in the Rzepin Beds (Upper
Ludlow).

H. Hollard (1963, pp. 21-22, text-fig. 1 a-c) has recently described, from the internal
moulds of two pygidia, Acaste talebensis, a species of Acaste found near the junction of
the Monograptus uniformis and M. hercynicus zones in Morocco. Both pleurae and axis
of this species are more strongly convex (tr.) than the condition shown by A. downingiae.
As a result the pleurae curve more sharply to the posterior. The interpleural furrows
are more deeply incised than those shown on the internal moulds of A. downingiae and
there are traces of segmentation, shown by slight swellings, on the border. The axial
termination ends more abruptly and closer to the posterior margin. The two species
share a similar outline and segmentation.

Range and Distribution. Acaste downingiae (Murchison) s.s. appears to be confined to the Wenlock
Limestone of the Welsh Borderlands and West Midlands. Though it occurs typically in a calcareous
environment, specimens similar to A. downingiae, from the Buckover Section, near Tortworth, Glou-
cestershire (Curtis and Cave 1964) occur in purplish-red and brownish weathering siltstones and
yellow-brown fine-grained calcareous sandstone. Again, forms similar to A. downingiae occur in the
grey-green calcareous siltstones of the highest Black Cock Beds (Lower Bringewood Beds) in the Golden

192

PALAEONTOLOGY, VOLUME 9

Grove, Llandeilo, and may even pass into the lowest Middle Bringewood Beds in the Llandovery area
(J. F. Potter, personal communication). At Freshwater East, Pembs., A. downingiae is replaced by
A. subcaudata (Murchison). The species has not been recorded from the Pen-y-lan inlier.

A. downingiae s.s. is definitely recorded from the following localities; Walsall, Staffs. ; Dudley (Wren's
Nest and Castle Hill), Worcs.; Benthall Edge, Salop; Wenlock Edge, Salop (rare); Malvern, Worcs. ;
Ledbury, Herefordshire.

Acaste inflat a (Salter 1864)

Plate 29, figs. 9-16

1864 Phacops ( Acaste ) Downingiae Murchison, var. y, inflatus Salter, p. 27, pi. 2, fig. 30 (OUM
C 9), not figs. 31 (GSM 19305), 32 (GSM 19314) or 33 (GSM 19306).

Lectotype {here designated). Salter 1864, pi. 2, fig. 30; OUM C9, figured here, pi. 29, figs. 9-14.
Wenlock Limestone, Ledbury Railway Tunnel, Ledbury, Herefordshire.

Diagnosis. A species of Acaste characterized by its strongly convex (sag.) frontal lobe
and by the high convexity (tr.) of 1L which causes it to slope abaxially well below
the level of 2L; 2L and 3L are also more convex (tr. and sag.) than in A. downingiae,
the convexity of the lobes causing the furrows to appear more closely spaced than in
that species. Visual surface at maximum height with rows of 7 lenses alternating with 6.
Pygidium with 7-8 axial segments, 5 pleural segments.

Description. Cephalon subtriangular in outline. Narrow, triangular area of librigena
prominently projecting in front of glabella. Surface of test finely granular. Length of
cephalon approximately |— § width.

Glabellar plan as in A. downingiae. Axial furrows gently curving round 2L and 3L.
Frontal lobe rounded both anteriorly and antero-laterally, moderately to strongly
convex (tr. and sag.). 2L and 3L abaxially fused, roughly equal in size, with slightly less
convexity (tr.) than frontal lobe. 1L about half as wide (exsag.) as 2L, with much
stronger convexity (tr.) than 2L in posterior profile, sloping abaxially well below the
dorsal level of 2L. Side furrows 3S long, shallow, fairly wide, curving backwards with
posterior median deflection. 2S shorter than 3S, impressed to similar depth, transverse,
abaxially linear, medianly with marked deflection to posterior; failing laterally to join
with axial furrows. IS of similar length to 2S, wider and deeper, deflected both abaxially
and medianly to anterior. Short sagittal furrow between median extremities of 3S.
Side furrows appear closer together than in A. downingiae due to the greater convexity
of the lobes (the ‘crowded’ condition of Salter 1864, p. 27).

Occipital furrow abaxially a little deeper than IS, becoming shallow but still well
defined sagittally. Occipital ring narrow (sag.), as wide (tr.) as 1L and with similar
convexity (tr.); in lateral profile raised higher above the side lobes than in A. dow-
ningiae.

Genae strongly sloping to narrow flattened border. Preglabellar furrow shallow, very
narrow (sag.). Postocular section of facial suture cutting lateral cephalic margin opposite
middle of 2L; preocular section dorsal intramarginal, becoming marginal at antero-
lateral edges of frontal lobe. Fixigenae postero-laterally angled in adults; with deep
but narrow (exsag.) posterior border furrow.

Eye crescentic in plan, rather small; situated centrally between cephalic margin and

J. H. SHERGOLD: A REVISION OF ACASTE DO WNINGIAE 193

glabella, on high ocular platform which falls steeply to antero-lateral and lateral margins
of cephalon. In lateral profile top of eye sloping gently to anterior; in anterior profile,
well below level of glabellar surface. Eye extending from posterior margin of 1L to
anterior margin of 3L; closer to posterior border furrow than in A. downingiae’, H/A,
5-18 per cent.; A/G, 47-56 per cent.; A/Gn, 40-48 per cent. Palpebral lobes high; pal-
pebral areas relatively wide, with low inclination to axial furrows.

Visual surface overhanging ocular platform; gently convex outwards. Up to 130
closely packed lenses arranged in roughly alternating vertical columns containing at
maximum height 6 or 7 individual lenses, with a typical diameter of up to 0-2 mm. and
rounded or polygonal in outline. Interspaces with similar texture to surrounding genae.

Pygidium subtriangular in outline, margin entire; in posterior profile slightly vaulted.
Axis raised above axial furrows and pleurae; moderately to strongly convex; composed
of 7-8 segments and a rounded, unsegmented terminal piece, tapering gradually to
posterior; segments 1-3 well defined. Pleurae moderately convex; 5 in number.
Pleural furrows strongly impressed, interpleural furrows weak. Border wide, as in adult
A. downingiae , with a pronounced flattening separating the segmented and unsegmented
pleural areas. Length of pygidium §— f width.

Comments on figured material. All the specimens used by Salter were derived from the
Malvern-Ledbury area. His fig. 30 was obtained by Dr. R. B. Grindrod from the
Ledbury railway tunnel. Pigs. 31 and 32 were merely stated in 1864 to be from Ledbury,
no stratigraphic horizon being given. In 1853, however, Salter (p. 9) recorded the same
specimens from ‘the Wenlock limestones of the Malverns’ and from ‘the Wenlock
Limestone, Ledbury’ respectively. Pig. 33 is said to be from Eastnor, near Ledbury.

In the text (1864, p. 27), it is implied that only fig. 30 represents the variety with any
certainty. The diagnosis is given from this specimen as follows: ‘The glabella is greatly
swollen, so as to be very convex, instead of flattened, in front. All the furrows are
indeed distinct and in their proper situations, but from the inflation of the glabella they
appear crowded. The eyes are small.’ Accordingly this specimen is here selected as
lectotype for the species.

Of the other specimens figured Salter appears to have been less certain. Pig. 31, he
states, ‘probably . . . belongs to var. /3’. This specimen has a similar furrow pattern to A.
inflat a but differs from it in the rather depressed frontal lobe more typical of A. downin-
giae: the low convexity of the glabellar lobes; the width (trans.) and convexity of the
occipital ring and its failure to rise, in lateral profile, significantly above the glabellar
side lobes. As such its characteristics are more compatible with A. downingiae, with
which it is here included.

Pig. 32 is a cephalon of exceptional size (occipital glabellar length 20 mm., cephalic
width 32 mm.). The eyes are missing. Although the frontal lobe is strongly convex (sag.),
the side lobes lack the convexity typical of A. inflata. Purthermore, side furrows 3S and
2S are relatively shorter and the transverse convexity of 1L and the occipital ring are not
nearly so pronounced as in that species. No other specimens referable to A. inflata
approach anywhere near this specimen in size. Large specimens of A. downingiae have,
however, been observed from the Ledbury area and of Acastocephala macrops from
Pen-y-lan quarry, Cardiff. Due to its highly convex frontal lobe and deep side furrows
this specimen has been refigured here with the latter species.

194

PALAEONTOLOGY, VOLUME 9

Fig. 33, a large pygidium (length 1 1 mm., width 15 mm.), is here certainly referred to
the downingiae group. It has in common with that species 8 axial segments, 5 pleural
segments, and the typical convexity when viewed in posterior profile. It is also wider
than the typical pygidium of A. inflata.

Relationships. A. inflata appears to be closely related to A. dayiana R. and E. Richter
from the Kobbinghauser Dayia-Schichten (see R. and E. Richter 1954, pp. 14—16,
pi. 4, figs. 45-56). The frontal lobe of their species has a slightly greater convexity (sag.),
and there is a higher degree of tapering of the glabella. The general glabellar plan and
convexity of the lobes is very similar. At its maximum height the visual surface of
A. dayiana carries rows of 5 lenses alternating with 6, and the total number of lenses
borne by the surface is only 94. Although the two species have essentially similar cephalic
outlines, the pygidium of A. dayiana is more subsemicircular than subtriangular in
outline and is also apparently wider.

A. inflata is well differentiated from A. downingiae (Murchison) by virtue of its highly
convex glabellar lobation. The visual surface of inflata carries 6 lenses alternating with
7, while that of downingiae has 8 alternating with 9. The eye of inflata is also a little
closer to the posterior border furrow, H/A, 5-18 per cent, compared with 12-30 per
cent, for downingiae (see text-fig. 2). Both cephalon and pygidium of inflata are slightly
narrower than those of downingiae, the ranges of length/width for the cephala being
51-68 per cent, and 46-62 per cent, respectively and for the pygidia, 65-85 per cent, and
60-80 per cent. In general it may be said that the pygidia are too closely related to be
differentiated further. Compared with downingiae, the sample of inflata studied is small
and the detailed ontogenetic changes have not been set out as for the former as the
smallest holaspid stages have not yet been seen.

Range and Distribution. From the available material Acaste inflata appears to be confined, like Acaste
downingiae, to the Wenlock Limestone of the Welsh Borderlands and West Midlands. Specimens close
to A. inflata , however, occur in the siltstones and calcareous sandstones of the Buckover Section, near
Tortworth, Gloucestershire (Curtis and Cave 1964). A further similar species with strongly convex
glabellar lobes has been collected by the author from Lower Elton Beds outcropping near Upper
Millichope, Shropshire.

The species has been identified with certainty from the following localities: Dudley (Wren's Nest),
Worcs. ; Malvern (railway tunnel), Worcs.; Ledbury (railway tunnel), Herefordshire.

Additional Material. A. inflata is not a common trilobite. The writer has seen to date but 19 specimens,
the bulk of which are to be found in the Grindrod Collection, Oxford University Museum.

From the Wenlock Limestone of Ledbury; OUM C 559; specimens undocumented but probably
also from this locality and horizon; OUM C597, C610, C612-20. From the Wenlock Limestone,
Malvern; GSM 19301-3. From the Wenlock Limestone of Dudley; BM In36154, In36156;
NMW 27.1 10.G.998.3.

Acaste subcandata (Murchison 1839)

Plate 30, figs. 1-16

1839 Asaphus subcaudatus (n.s.) Murchison, p. 655, pi. 7, fig. 10.

1839 Asaphus Cawdori (n.s.) Murchison, p. 655, pi. 7, fig. 9.

1848 Phacops Downingiae Murchison; Salter, p. 336, pi. 5, figs. 2 (untraced), 3 (GSM 19377)
and 4 (untraced).

1853 Phacops Downingiae Murchison; Salter, p. 1 (both Asaphus Cawdori and A. subcaudatus
classed as junior synonyms of P. Downingiae).

1921 Phacops Downingiae (Murchison); Dixon, p. 21, footnote 6.

J. H. SHERGOLD: A REVISION OF ACASTE DOWNINGIAE 195

Holotype. Murchison 1839, pi. 7, fig. 10; GSC 6591. The holotype is ‘possibly from the Wenlock
Series’ (Dixon 1921, p. 21), Freshwater East, Pembrokeshire.

Diagnosis. A species of Acaste with frontal lobe more strongly convex (sag.) than that
of A. downingiae but less so than A. inflata; 3S more faintly impressed than in A.
downingiae. Fixigenae of young holaspides with postero-lateral projections, resorbed in
later moults, as in A. downingiae. Eyes small; visual surface narrow, similar to A.
dayiana , with fewer lenses than A. downingiae, bearing at the maximum height rows of
5 lenses alternating with 6. Pygidium rather long; 7 (8) axial segments; 5 (6) pleurae;
more strongly convex (tr.) than A. downingiae', border narrow, separated by marked
flattening from furrowed pleural area at all holaspid stages. Margin entire throughout
holaspid ontogeny.

Description (based on internal moulds). Cephalic outline subtriangular to ogival as in
A. downingiae. The narrow, triangular area of librigena anterior to the frontal lobe
which is common to other species of Acaste is also present in A. subcaudata.

Glabella anteriorly rounded, laterally more or less subparallel-sided, transverse width
decreasing evenly to posterior. Axial furrows linear, without the gentle curve round 2L
and 3L which characterizes A. downingiae, diverging at a slightly greater rate than
A. downingiae in small specimens, appreciably less in large individuals. Frontal lobe
anteriorly rounded, antero-laterally more parallel-sided than in A. downingiae. In lateral
profile frontal lobe has a moderate convexity (sag.) which is greater than that of A.
downingiae but less than A. inflata. Small individuals have a sagittal convexity similar
to or greater than the adult of A. downingiae. 2L and 3L of roughly equal size, abaxially
fused. 1L about half as wide as 2L; in young holaspides transverse convexity of 1L
is slightly greater than that of 2L but in adults is less apparently so. 3S sigmoidal, long,
more faintly impressed on the internal mould than in specimens of A. downingiae pre-
served with the shell, with a distinct median deflection to the posterior which appears
to be present at all holaspid ontogenetic stages. 2S shorter than 3S, impressed to a
similar depth, linear, transverse or with very faint posterior median deflection in some
individuals; failing laterally to reach the axial furrows. IS much wider and deeper but
also shorter than 2S or 3S, curving both medianly and abaxially to the anterior.

Occipital furrow as wide and deep abaxially as IS. Occipital ring narrow (sag.), with
similar transverse width and convexity to 1L; in lateral profile barely rising above the
level of the glabellar side lobes.

Fixigenae in young holaspides produced postero-laterally into short, stout projections
which appear to be longer than in specimens of A. downingiae of corresponding size. In
larger specimens the projection is resorbed but there remains a distinct point. Genal
projections have been observed in specimens with a cephalic length of up to 2-9 mm.
but a full range of sizes has yet to be collected. Adult genae are completely unknown,
but those of intermediate stages are distinctly angled.

Eyes rather small, similar in plan to those of A. inflata, situated centrally on the genae
on low ocular platform, as in A. downingiae', extending roughly from IS to confluence
of 3S and axial furrows; A/G, 45-63 per cent.; A/Gn, 37-46 per cent. In lateral profile
the top of the eye slopes gently to the anterior and fails to reach the level of the glabella
surface. Palpebral areas with low inclination to axial furrows. Fixigenal area between

196 PALAEONTOLOGY, VOLUME 9

back of eye and posterior border furrow similar in extent to A. downingiae, H/A, 12-27
per cent.

Visual surface known only from one juvenile specimen; rather narrow bearing fewer
lenses than in A. downingiae , at the maximum height of the surface arranged in alter-
nating columns of 5 or 6 (as in A. dayiana).

Outline of hypostome subrectangular; anterior margin rounded and antero-laterally
produced into short wings; lateral margins very gently curved abaxially; posterior
margin gently curved backwards; postero-lateral corners truncate, curved. Posterior
border narrow, separated from median body by wide, well-defined furrow; lateral
borders extremely narrow. Median body moderately convex adventrally. Maculae
situated halfway along body, close to lateral borders.

Pygidium subtriangular in outline; in posterior profile rather strongly convex (tr.).
Axis strongly convex (tr.), raised well above axial furrows and pleurae; typically
with 7 (8) segments, the eighth being discernible only in cases of exceptional preserva-
tion; narrow postaxial border; axial segments 1-3 are most distinct, separated by well-
defined transverse furrows; succeeding segments rather poorly defined towards the
posterior. Pleurae generally more strongly convex (tr.) than in A. downingiae, falling
off rapidly in postero-lateral direction; 5 (6) in number. Pleural furrows strongly defined,
wide and deep; interpleural furrows, with the exception of the first, rarely preserved on
internal moulds. Border narrow, separated from furrowed pleural area by a marked
flattening at all holaspid growth stages. Margin entire throughout holaspid ontogeny.

Type Material. Specimens, GSC 6591 and 6592 are labelled as having been used by Murchison (1839)
for his illustrations of Asaphus subcaudatus and A. cawdori. GSC 6591 is here selected as the holotype

EXPLANATION OF PLATE 30

Figs. 1-16. Acaste subcaudata (Murchison 1839). 1-2. GSC 6591 ; Holotype; figured Murchison 1839,
pi. 7, fig. 10. Probably Wenlock Series, Freshwater East, Pembrokeshire. 1, Dorsal view, internal
mould of pygidium, X3. 2, View from posterior, x3. 3-4. GSM 19377; figured Salter 1848, pi. 5,
fig. 3. Probably Ludlow Series, Freshwater East, Pembrokeshire. 3, Dorsal view, internal mould of
pygidium, x3. 4, View from posterior, x3. 5-6. GSM 102597; Ludlow Series, Freshwater East,
Pembrokeshire. 5, Internal mould of young holaspid cephalon with genal projection, x 4. 6, Oblique
view, X 4. 7. GSM 102605 ; Ludlow Series, Freshwater East, Pembrokeshire. Internal mould of pygi-
dium, X 3. 8. GSM 102604; Ludlow Series, Freshwater East, Pembrokeshire. Internal mould of eye,
x8. 9. GSC 6592; figured Murchison 1839, pi. 7, fig. 9 as Asaphus cawdori. Probably Wenlock
Series, Freshwater East, Pembrokeshire. Internal mould of pygidium, x 3. 10. GSM 102603 ; Ludlow
Series, Freshwater East, Pembrokeshire. Latex cast from internal mould of pygidium, X 3. 11. GSM
102602; Ludlow Series, Freshwater East, Pembrokeshire. Internal mould of cranidium, x3. 12.
GSM 102606; Ludlow Series, Freshwater East, Pembrokeshire. Internal mould of glabella, x3.
13-14. GSM 102600-1; Ludlow Series, Freshwater East, Pembrokeshire. 13, Internal mould of
cranidium, X3. 14, Latex cast from external mould showing side furrows 2S and 3S more deeply
incised, x3. 15. GSM 102599, Ludlow Series, Freshwater East, Pembrokeshire. Internal mould of
cranidium, X 3. 16. GSM 102598; Ludlow Series, Freshwater East, Pembrokeshire. Latex cast from
external mould of cranidium, x3.

Figs. 17-23. Acastoides constricta (Salter 1864). 17-20. BU 58; figured Salter 1864, pi. 2, fig. 15 a-c.
Wenlock Limestone, Dudley, Worcs. 17, Dorsal view of cephalon, x3. 18, Lateral view, x3. 19,
Eye, X 8. 20, View from anterior, x 3. 21. BM 58897; Wenlock Limestone, Dudley, Worcs. Hypo-
stome, X 6. 22. OUM C8; figured Salter 1864, pi. 2, fig. 14. Wenlock Shale, Malvern Tunnel, Worcs.
Dorsal view, x2. 23. OUM C7; Lectotype; figured Salter 1864, pi. 2, fig. 13, 13a. Wenlock Shale,
Malvern Tunnel, Malvern, Worcs. Dorsal view, X 2.

Palaeontology, Vol. 9

PLATE 30

SHERGOLD, Acaste, Acastoides

J. H. SHERGOLD: A REVISION OF ACASTE DO WNINGIAE

197

for Acaste subcaudata (Murchison). In its segmentation, its degree of convexity and in the nature of
its border, it is indistinguishable from the specimen (GSM 19377) figured by Salter (1848, pi. 5, fig. 3)
and from others collected by the author from Freshwater East. GSC 6592, the holotype of Asaphus
cawdori, is a specimen most difficult to evaluate, largely due to its distortion and indifferent preserva-
tion. There is also difficulty in equating it with Murchison’s figures. Murchison (1839, pi. 7, fig. 9)
originally figured a block showing two small pygidia but in 1872 (pi. 18, fig. 3) only one pygidium is
depicted. In both illustrations the specimens are complete and undistorted. GSC 6592, therefore, can
only doubtfully be identified with A. cawdori. Essentially similar to Acaste subcaudata in construction
it is here included with reserve in the synonymy of that species. Stubblefield (1938, p. 31) states that
Salter’s fig. 2 (pi. 5) has been possibly compounded from two specimens in the Geological Survey
Collections, GSM 19376 and 19378, both being incomplete cranidia and neither showing the full out-
line or extent of the cephalon as depicted by Salter. The specimen used for fig. 3 is thought to be GSM
19377. That used for fig. 4 is untraced.

Relationships. A. subcaudata differs from A. downingiae, to which it is closely related,
by its more convex (sag.) frontal lobe; by the tapering of the glabella; the depth of
impression of the side furrows, which would normally be expected to be deeper on the
internal mould (the furrows must be quite faint on the shell); the number of facets of
the eye and their distribution; the general shape of the hypostome and the convexity
and segmentation of the pygidium.

From A. inflata, A. subcaudata differs by the lower convexity (sag.) of the frontal
lobe; the lower convexity (tr.) of the preoccipital glabellar lobe; by the number and
distribution of eye facets and by the same pygidial characters as stated above for
A. downingiae.

A. dayiana has a more strongly convex (sag.) frontal lobe and a greater degree of
glabellar tapering. The convexity (tr.) of the side lobes would also appear to be stronger.
Pygidial segmentation and the presence of a border flattening on the internal moulds
are common to both species, as is the content of the visual surface. A common
environmental control may account for these similarities.

Range and Distribution. A. subcaudata is at present only known from strata occurring at Freshwater
East, Pembrokeshire, attributed by Dixon (1921) to the Wenlock and Ludlow Series. GSM 102597-606
were collected by the author from the Ludlow Series of the South side of Freshwater East Bay, from
rottenstone bands in group (4) of Dixon (1921, p. 14). These beds are accessible where the High Water
Mark shown on Dixon’s sketch map (1921, p. 23) is marked as meeting the cliff. Approximate Grid
Ref., SS 0168/9739.

Additional Material-. GSM 19376-8, 102597-606.

Genus acastoides Delo 1935
Acastoides constrict a (Salter 1864)

Plate 30, figs. 17-23

1864 Phacops ( Acaste ) Downingiae Murchison, variety or subspecies e, constrictus. Phacops
constrictus. Salter, pp. 27-28, pi. 2, figs. 13 (OUM C7), 14 (OUM C8), 15 (BU 58).
1954 Acaste ( Acastoides ) constricta (Salter 1864). R. and E. Richter, pp. 17-18, pi. 3, figs. 42, 43.
1965 Acastoides constricta (Salter 1864); Clarkson, p. 22, pi. 3, figs. 10-12.

Lectotype ( here designated). Salter 1864, pi. 2, fig. 13; OUM C7. Wenlock Shale, Malvern Tunnel
Malvern, Worcs.

198

PALAEONTOLOGY, VOLUME 9

Remarks. R. and E. Richter (1954, pp. 17-18) have given an account of this species
but have omitted to consider the hypostome. Accordingly no detailed description will
be given here. Instead, an extended diagnosis is given together with an account of the
hypostome and notes on the relationship of Acastoides constricta with other species
of the genus.

Extended diagnosis. A species of Acastoides characterized by a strongly convex (sag.)
frontal lobe which frequently extends to the anterior margin of the cephalon; 1L with
a greater transverse convexity than 2L, sloping abaxially below the level of 2L; 3S
shallow, wide, sigmoidal, with strong posterior median deflection; 2S long, transverse,
just reaching the axial furrow; IS and 3S tending to converge adaxially. Occipital ring
raised high above level of glabella, with axial node. Genae postero-laterally rounded off.
Preocular section of facial suture marginal or just dorsal intramarginal. Eyes small,
dose to glabella, extending from occipital furrow to anterior edge of 3L, raised on high
ocular platform nearly to the level of the glabella. Palpebral lobes high; palpebral
areas narrow, with high inclination to the axial furrows. Visual surface with 104+ lenses,
at maximum height bearing columns of 6 alternating with 7. Pygidium subsemicircular
in outline, posteriorly truncate, scarcely vaulted in posterior profile; margin entire.
Axis short, with low convexity (trans.), composed of 5 (6) segments and wide axial
terminal piece. Pleurae weakly convex (trans.), 4 in number; furrows weakly impressed,
pleural furrows being slightly stronger than interpleural furrows. Border flat and broad.

Elypostome outline subquadrate; anterior margin arcuate, antero-laterally extended
into short wings; lateral margins curving very gently abaxially; posterior margin short,
gently curved backwards; postero-lateral margins truncate. Lateral borders narrow;
posterior border wide (sag.). Median body strongly convex adventrally, without trace
of median furrow. Maculae large, ovoid and prominent, situated on lateral margins of
median body just inside the anterior half of the structure. Surface coarsely granular.

The hypostome of A. constricta (Salter) differs from that of Acaste downingiae in
its more highly convex median body with prominent maculae; in the arcuate outline of
the anterior margin and the wide (sag.) posterior border. Though similar in structure
to the hypostome of Acastocepha/a macrops (Salter), that of Acastoides constricta
contrasts strongly by its distinctive coarse ornament.

Relationships. In construction the cephalon of Acastoides constricta appears to be
closely related to A. henni henni R. and E. Richter 1952 described from the Wiltzer
Schichten, Upper Emsian. A. constricta may be differentiated by its more marginal
frontal lobe; greater degree of glabellar tapering; eyes larger and more extensive;
visual surface with a greater number of lenses and a maximum of 7 in any column
compared with 4 in A. henni henni ; the pygidial margin is entire and non-denticulate.

A. constricta is similarly distinct from A. henni posthuma R. and E. Richter 1952
from the Heisdorfer Schichten, high Upper Emsian, and from A. paeckelmanni
R. and E. Richter 1939, from the Pendik Schichten, Pendik, Bosphorus, high Upper
Emsian in age.

Range and Distribution. Acastoides constricta is recorded from the Wenlock Shale and Wenlock
Limestone of Dudley, Walsall and the Malvern Tunnel, Malvern, Worcs. It appears to be restricted
to the Wenlockian of the Welsh Borderlands.

J. H. SHERGOLD: A REVISION OF ACASTE DO WNINGIAE 199

Additional Material. In addition to the lectotype and paratypes the following material has been
located.

From the Wenlock Shale of the Malvern Railway Tunnel: OUM C561-7, C576, C609, C611;
BM 59040, 59048; GSM 19242-3. From the Wenlock Limestone, Malvern; BM 44251; GSM
19304; labelled Wenlock? Limestone, Malvern; BM 58897, 59048; labelled simply Malvern; GSM
49836. From the Wenlock Limestone, Walsall; GSM 22450. Unlocalized specimen; GSM 49838.

Family dalmanitidae Vogdes 1890
Subfamily acastavinae Struve 1958
acastocephala gen. nov.

Derivation of name. Trilobites with cephalon constructed in a similar manner to that of Acaste Goldfuss.
Type species. Acastocephala macrops (Salter 1864).

Other species. Acastocephala dudleyensis sp. nov.

Diagnosis. Cephalon typically acastid in construction but with larger eye; visual surface
containing, at maximum height, rows of 10 lenses in A. macrops (Salter), 9 in A. dud-
leyensis sp. nov. Pygidium wide, weakly furrowed; 4-5 pleural segments; 7 axial seg-
ments. During ontogeny of both known species smallest holaspides have a denticulate
lateral pygidial margin; denticulations resorbed in succeeding growth stages to give
adult with typically entire margin.

Range and Distribution. Wenlock Shale, Malvern, Worcs; ? Wenlock Shale, Pen-y-lan, near Cardiff;
Wenlock Limestone, Malvern and Dudley, Worcs.

Relationships. Acastocephala differs from Acaste Goldfuss in the following character-
istics. The eye is larger, carries a greater number of facets and is relatively closer to the
posterior border furrow. The pygidium has a lower convexity (tr.) and is more poorly
segmented. Both pleural and interpleural furrows are weaker. As in Acaste , the fixigenae
of the smallest holaspides are provided postero-laterally with short projections but these
are not nearly so long or distinct and appear to be more rapidly resorbed. In addition,
young holaspides of Acastocephala have a denticulate lateral pygidial margin which
contrasts with the entire margin of the adults, while the margin of Acaste is entire in
holaspides of all sizes. Posteriorly the margin is without trace of a caudal projection.
The hypostome is larger and more convex (tr. and sag.) than in Acaste and the
maculae are larger and more clearly defined.

Though closely related to Acaste , Acastocephala is equally close to Acastava R. and
E. Richter 1954. The glabellae are essentially similar in construction. Schmidt’s text-figure
of Acastava atava (1907, p. 10) shows side furrows 3S to be obliquely directed, without
a posterior median deflection and very similar to the condition of Acastocephala macrops.
Similarly, 2S is linear, transverse and only just failing abaxially to reach the axial furrows.
In Acastava atava the eyes are large but the visual surface at its maximum height con-
tains columns with only 6 lenses. A. atava also differs in retaining genal spines into adult
holaspid stages. When preserved with the shell the pygidial margin of A. atava is entire
but the internal mould possesses 5 pairs of lateral denticulations (see Schmidt 1907, p. 10,
text-fig. 1 and R. and E. Richter 1954, pi. 6, figs. 81 and 82) formed in exactly the same
manner as those produced in young holaspides of Acastocephala (text-fig. 5). Pygidial

C 3803

P

200

PALAEONTOLOGY, VOLUME 9

segmentation is also similar. Acastaval schmidti (R. Richter), like Acastocephala dud-
leyensis, possesses large eyes situated both close to the glabella and to the posterior
border furrow. Acastocephala is confined, in the British Isles, to the Wenlockian while
Acastava ranges in the Lower Devonian from Siegenian to Emsian. The affinities
between these two genera may prove to be significant phylogenetically.

a b c

text-fig. 5. A comparison of Acastocephala macrops (Salter) and Acastava atava (W. E. Schmidt).
a. Young holaspid of A. macrops showing lateral denticulations on pygidial margin of shell; EU 862.
X 2-f. b . Adult of A. macrops showing entire pygidial margin; BU 861. X 2. c. Adult of Acastava atava
showing denticulations along lateral pygidial margin on the internal mould. Drawing after Struve

1959, p. 0476, fig. 377, 1 a, lb. x2f.

Acastocephala is quite distinct from Acastella Reed 1925. Acastella is characterized
by slender genal spines and by a caudal mucronation throughout its ontogeny. Pygidial
segmentation and the size and position of the eyes in Acastella are essentially similar to
Acaste. Some species of Acastella have a smooth pygidial margin when preserved with
the shell or as internal moulds as, for example, in A. heberti heberti (Gossellet),
A. patula Hollard, and A. jacquemonti Hollard. Others, A. tiro R. and E. Richter and
A. heberti elsana R. and E. Richter, produce lateral denticulations on the internal
mould but not on the shell.

The genus Acastoides Delo 1935, is characterized by a subpentangular cephalic out-
line and eyes set high up on the genae. The shape, convexity and segmentation of the
pygidium differ considerably from Acastocephala. In some species, Acastoides henni
henni (R. Richter) and A. paeckelmatwi R. and E. Richter, there is a tendency to
develop spine-like denticulations along the pygidial margin but they differ from those
of Acastella , Acastava, and Acastocephala in that they are directed ventrally. Acastoides
constricta (Salter) and Acastocephala macrops (Salter) occur together in the Wenlock
Shale of Malvern.

The decision to erect a new genus to cover trilobites referred to Salter’s variety macrops
is based largely on the differences of these species from both Acaste and Acastava. The
larger number of eye facets and the lack of genal spines in the adult prevent inclusion
of these trilobites in the genus Acastava, while the pygidial margins of the young and

J. H. SHERGOLD: A REVISION OF ACASTE DOWNINGIAE 201

weaker pygidial furrowing preclude their inclusion in the genus Acaste. A further con-
sideration has been the possibility of a phylogenetic link between Acastocepha/a and
Acastava.

Acastocepha/a macrops (Salter 1864)

Plate 31, figs. 1-15, Plate 32, figs. 1-2

1864 Phacops ( Acaste ) Downingiae Murchison, var. /3, macrops, Salter, pp. 26-27, pi. 2, figs. 27
(BU 861), 29 (GSM 19319), 32 (GSM 19314), not figs. 26 and 28 (specimens untraced).

Lectotype ( here designated). Salter, 1864, pi. 2, fig. 27; BU 861. Wenlock Shale, Malvern, Worcs.

Diagnosis. A species of Acastocepha/a gen. nov., with moderately convex (sag.) frontal
lobe; 3S wide and deep, opening out abaxially, 2S short, linear, sloping gently abaxially
to posterior. Eyes large, situated centrally on the genae; extending anteriorly past the
confluence of 3S and the axial furrows. Visual surface at maximum height bearing 9
lenses alternating with 10. Pygidium with low convexity; 4 (5) pleurae, 7 axial segments.
Border undifferentiated from furrowed pleural area. Margin in small holaspides denti-
culate, entire in adults.

Description. Cephalic outline ogival, subtriangular or in large individuals, more rarely,
subsemicircular. In common with species of Acaste and Acastel/a there projects, anterior
to the glabella, a narrow triangular area of librigena, in this species not as extensive as
in Acaste downingiae or Acaste inflata. Surface of test finely granular. Length; width of
cephalon

Glabella anteriorly rounded, laterally subparallel. Angle of divergence of axial furrows
similar to A. downingiae. Degree of tapering in transverse width rather irregular, giving,
in some specimens, a distinctly waisted appearance (cf. the lectotype), which is less
apparent in specimens which have either lost the greater part of their shell thickness or are
preserved as internal moulds. Frontal lobe both anteriorly and antero-laterally rounded,
with moderate to strong convexity (sag.) in young holaspides and adults. 2L and 3L
laterally fused, more or less equal in size, slightly less convex (tr.) than the frontal
lobe. 1L narrow (exsag.), about half the size of 2L. Convexity (tr.) of 1L insig-
nificantly greater than 2L in specimens with shell but slightly greater in internal moulds.
3S long, sigmoidal, wide and deep, opening out and becoming deeper abaxially; with
distinct posterior median deflection in the majority of specimens. 2S shorter, less deep
than 3S, linear, transverse, sloping gently to the posterior abaxially, generally failing
laterally to join with axial furrows but in some examples very nearly attaining this
condition. IS only a little deeper than 3S, longer than 2S, deflected both medianly and
abaxially to anterior.

Occipital furrow impressed to similar depth as IS. Occipital ring typically a little
wider (trans.) than 1L and with similar convexity (tr.). In lateral profile failing to rise
significantly above level of glabellar side lobes.

Genae restricted in area due to size and position of eyes. Postocular section of facial
suture cutting lateral cephalic margin opposite the posterior edge of 2L; preocular
section dorsal intramarginal but may become almost marginal at antero-lateral margins
of glabella as in Acaste inflata. Genae postero-laterally angled in young holaspides,
rounded in adults.

202

PALAEONTOLOGY, VOLUME 9

Eyes large, subcrescentic in plan, occupying a large area of the gena, situated a little
closer to the cephalic margin than to glabella, especially in young specimens; ocular
platform low. In lateral profile top of eye sloping gently to anterior and failing to reach
the level of the top of the glabella. Extending from occipital furrow or mid 1 L to posterior
edge of frontal lobe. Palpebral lobes low but clearly defined; palpebral areas with low
inclination to axial furrows. Area of fixigena between back of eye and posterior border
furrow narrow; H/A, 6-17 per cent.; A/G, 53-70 per cent.; A/Gn, 45-65 per cent.

Visual surface overhanging slightly the ocular platform, gently convex outwards;
sloping more strongly outward-backwards than outward-forwards. Surface bearing
a greater number of lenses (in excess of 170) than in species of Acaste and Acastella.
Columns arranged in rows containing at maximum height of surface 9 or 10 lenses.
Interspaces non granular.

Hypostome subrectangular in outline; anterior margin rounded, produced antero-
laterally into rather long wings, behind which lie shallow notches; posterior margin
gently curved to anterior; postero-lateral corners truncated. Lateral borders of un-
determined extent; posterior border wide (sag.), bearing a transverse ridge. Median
body strongly convex adventrally, with faint median furrow. Maculae large, ovoid
tubercles situated laterally and in anterior half of median body. Fine granular ornament
overall.

Pygidium subtriangular or subsemicircular in outline; in posterior profile with low
convexity (trans.). Axis with low or moderate convexity (tr.), barely raised above
axial furrows and pleurae; composed of at least 7 segments of which the anterior
3 are most distinct and separated by well-defined transverse furrows, segmentation
becoming less distinct thereafter towards the posterior. Pleurae with low convexity (tr.)
in posterior profile; 4 (more rarely 5) in number. Furrows weakly impressed, pleural
furrows stronger than interpleural furrows, the latter being well defined only on the
first pleura or adaxially on the succeeding pleurae. Border wide, separated both in adults
and young by a very shallow flattening, more obvious in the former than the latter
and in internal moulds. Margin in young holaspides (with cephalic lengths of up to
5-45 mm. and pygidial lengths up to 4-40 mm.) provided laterally with 3 pairs of short
denticulations associated with the first 3 posterior pleural bands. In the adult these
denticulations are resorbed to give an entire margin. No trace has been seen at any
holaspid stage of a caudal spine or projection.

EXPLANATION OF PLATE 31

Figs. 1-15. Acastocephala macrops (Salter 1864). 1-5. BU 861; Leototype; figured Salter 1864, pi. 2,
fig. 27. Wenlock Shale, Malvern, Worcs. 1, Dorsal view of cephalon, x2. 2, Lateral view, x2.
3, Dorsal view of pygidium, X 2. 4, Posterior view of pygidium, X 2. 5, Eye, X 8. 6-7. GSM 19314;
figured Salter 1853, pi. 1, fig. 14; 1864, pi. 2, fig. 32. Wenlock Limestone, Ledbury, Herefordshire.
6, Lateral view of cephalon, x 2. 7, Dorsal view of cephalon, X 2. 8-9. BM 44319; Wenlock ?Shale,
Pen-y-lan, nr. Cardiff'. 8, Lateral view of complete specimen, x 2. 9, Dorsal view of cephalo-thorax,
x2. 10. BM 44323; Wenlock ?Shale, Pen-y-lan, nr. Cardiff. Internal mould of pygidium, x2.
11. GSM 19319; figured Salter 1864, pi. 2, fig. 29. Wenlock ?Shale, Pen-y-lan, nr. Cardiff. Internal
mould of cephalo-thorax, x2. 12-15. BU 862; Wenlock Shale, Malvern, Worcs. Enrolled dorsal
exoskeleton of young holaspid. 12, Dorsal view of cephalon, x4. 13, Lateral view, x4. 14, Dorsal
view of pygidium, with denticulate margin, x4. 15, Anterior view of cephalon, x4.

Palaeontology , Vol. 9

PLATE 31

SHERGOLD, Acastocephala

J. H. SHERGOLD: A REVISION OF ACASTE DO WNINGIAE

203

Type Material. Salter (1853, pi. 1, fig. 3) included among his illustrations of Phacops downingiae a large-
eyed variety from the Wenlock Limestone of Dudley. This specimen, unfortunately, remains untraced
but from the figure, the rather bull-nosed frontal lobe and the high occipital ring indicate that it should
now be classified as Acastocephala dudleyensis sp. nov. rather than A. macrops (Salter). In 1864, how-
ever, Salter separated the large-eyed Phacops downingiae under the variety macrops (pi. 2, figs. 26-29).
Figs. 26 and 28, from Dudley, again appear to belong to A. dudleyensis but these specimens also remain
untraced. Of the remaining specimens, fig. 27, from the Wenlock Shale of Malvern, is in the type col-
lection, Birmingham University Museum, No. BU 861, originally from the Ketley Collection. It has
been selected here as the lectotype for A. macrops on account of its superior preservation to fig. 29
(GSM 19319), an internal mould with damaged pygidium from Pen-y-lan, near Cardiff. Fig. 32 (GSM
19314), was formerly classified by Salter as var. y, inflatus but it seems more probably to be a large
specimen of A. macrops.

Range and Distribution. Specimens of A. macrops collected by Dr. R. B. Grindrod were derived from
the Malvern Railway Tunnel. The exact horizon is, therefore, uncertain. However, Symonds and
Lambert (1861, p. 158) and Salter (1861, p. 161, appendix to former paper) state that Grindrod’s
material is from the Woolhope Shales. The possibility of them coming from the Wenlock Limestone
or Lower Ludlovian is diminished in the light of further statements by Salter (op. cit.). ‘The Wenlock
Limestone of this section’, he states, ‘is not well developed and is cut out in many places by faulting.’
When comparing the faunas of the Woolhope Shales and those of the Lower Ludlovian Salter is
impressed by the fact that ‘In the fewness of Trilobites this shale differs materially from the Woolhope
Shale . . .’. It seems most probable, therefore, that all of Grindrod’s material is from the Woolhope
Shale, Lower Wenlockian.

A. macrops is also recorded from beds of Wenlockian age at Pen-y-lan, Cardiff, and more rarely
from the Wenlock Limestone of Dudley. Again the exact horizons at Pen-y-lan are uncertain. Sollas
(1879, p. 480) refers the strata of Pen-y-lan quarry to the lowest Wenlockian. The Dudley specimens
are generally associated with a blue-grey, calcareous siltstone matrix and are rarely if ever found in
the purer limestone bands. They may have been derived either from silty bands within the Wenlock
Limestone or from the uppermost Wenlock Shale.

Additional Material. From the Wenlock Shale, Malvern Tunnel; OUM C568, C574, C575, C577, C578,
C586, C593, C594, C596, C605, C606, C608; BU 862. From the Wenlock Limestone, Malvern; BM
4441 1, 58918. From the Wenlock Limestone, Dudley; BM In36155; GSM 19317, 19353; BU (Holcroft)
250. Labelled Wenlock Limestone, Pen-y-lan, Cardiff; BM 44319, 44232. Labelled Wenlock, Pen-y-lan,
Cardiff; NMW 14.281.G.14, 40.199.G.7.7. Labelled? Ludlow, Ledbury; NMW 00.312.

Acastocephala dudleyensis sp. nov.

Plate 32, figs. 3-16

Derivation of name. From the type locality of Wren’s Nest, Dudley, Worcs.

71853 Phacops Downingiae Murchison; Salter, p. 4, pi. 1, fig. 3 (untraced).

71864 Phacops ( Acaste ) Downingiae Murchison, var. /3, macrops, Salter, pi. 2, figs. 26 and 28
(specimens untraced).

1965 Acaste downingiae macrops (Salter 1864); Clarkson, p. 19, pi. 3, figs. 4-9.

Holotype. BU 863. Wenlock Limestone, Wren’s Nest, Dudley, Worcs. (ex Holcroft Collection).

Diagnosis. A species of Acastocephala with rather strongly convex (sag.) frontal lobe;
3S sigmoidal, similar to Acaste inflat a (Salter), much more faintly impressed than in
Acastocephala macrops (Salter); 2S curved both medianly and abaxially to the posterior,
longer than in A. macrops. Eyes large, situated close to the glabella in adult holaspides,
impinging on the posterior border furrows causing them to bulge posteriorly. Visual sur-
face as in A. macrops but with one less facet at the maximum height of the surface.

204

PALAEONTOLOGY, VOLUME 9

Genae in smallest holaspides with minute postero-lateral projections, rapidly resorbed
in succeeding stages. Pygidium as in A. macrops but the lateral denticulations are more
rapidly resorbed.

Description. Cephalic outline subtriangular, as in Acaste. Anterior librigenal projection
limited, as in Acastocephala macrops (Salter). Surface of test finely granular. Length of
cephalon -|-§ width.

Glabella anteriorly rounded, more or less laterally subparallel-sided. Degree of taper-
ing in transverse width irregular due to the laterally extensive frontal lobe and the con-
vexity (tr.) of 1L. Frontal lobe both anteriorly and antero-laterally rounded but
spreading out laterally; width in relation to 2L and 3L greater than in A. macrops ;
axial furrows deflected abaxially a little at their confluence with 3S instead of running
normally round the frontal lobe as a continuation of the line defining 2L and 3L, as is
the case in A. macrops', rather strongly convex (sag.) in all holaspid stages, bull-nosed,
the lateral profile falling steeply to the preglabellar furrow from a point on the sagittal
line halfway along the length of the frontal lobe. Outline of frontal lobe in young holas-
pides more similar to that of the adult macrops. 2L and 3L subequal in size, laterally
fused in adults but frequently separated in young holaspides. 1 L narrow (exsag.), about
half the size of 2L, with a greater transverse convexity than that of 2L so that the lateral
margins slope below those of 2L, this convexity becoming less in adults. 3S shallow,
long sigmoidal, with distinct posterior median deflection, similar to Acaste but contrast-
ing with A. macrops. 2S shorter, impressed to a similar depth, abaxially linear but
with strong median deflection to the posterior; in young holaspides reaching the axial
furrows but in adults just failing to do so. IS wider and deeper than either 2S or 3S,
curving both medianly and abaxially to the anterior.

Occipital furrow impressed to similar depth as IS. Occipital ring narrow (sag.) with
similar width and convexity (tr.) to 1L; in lateral profile raised strongly above
glabellar side lobes; in young holaspides often with axial node.

Genae, as in A. macrops, rather restricted in area due to size and position of eyes.
Postocular section of facial suture cutting lateral cephalic margin opposite mid 2L;
preocular section dorsal intramarginal, becoming marginal when skirting the antero-
lateral margins of the glabella. Genae postero-laterally provided with short projections
in small holaspides (with cephalic lengths observed up to about 4-00 mm.); adults with
rounded off genal angles.

EXPLANATION OF PLATE 32

Figs. 1-2. Acastocephala macrops (Salter 1864). BM 58918; Wenlock Limestone, Malvern, Worcs. 1,
Dorsal view of cephalon, X 2. 2, Hypostome, X 6.

Figs. 3-16. Acastocephala chuUey easts sp. nov. 3-7. BU 863; Holotype; Wenlock Limestone, Dudley,
Worcs. 3, Dorsal view of cephalon, x4. 4, Lateral view, X4. 5, Dorsal view of pygidium, x4.
6, Posterior view of pygidium, x 4. 7, Eye, X 8. 8-10. BU 864; young holaspid. Wenlock Limestone,
Dudley, Worcs. 8, Dorsal view, X 5. 9, Genal angle with short projection, x 15. 10, Lateral view,
x5. 11. BU 865; young holaspid. Wenlock Limestone, Dudley, Worcs. Dorsal view, X 5. 12-13.
BU 866; young holaspid. Wenlock Limestone, Dudley, Worcs. 12, Dorsal view, x5. 13, Pygidial
margin showing denticulations, X 15. 14. GSM 19311; Wenlock Limestone, Dudley, Worcs. Dorsal
view, X 3. 15. BU 867; Wenlock Limestone, Dudley, Worcs. Dorsal view, X 3. 16. BU 868; Wenlock
Limestone, Dudley, Worcs. Dorsal view, x3.

Palaeontology, Vol. 9

PLATE 32

SHERGOLD, Acastocephala

J. H. SHERGOLD: A REVISION OF ACASTE DO WNINGIAE 205

Eyes large, crescentic in plan, situated closer to the glabella than to the cephalic
margins, on low ocular platform. In lateral profile sloping gently to anterior and rising
to the level of the glabella. Extending in adults from occipital furrow or mid 1L to
posterior edge of frontal lobe; in young holaspides extending from mid 1L or 1 S to con-
fluence of 3S and axial furrows. Palpebral lobe well defined, high; palpebral areas
with high inclination to axial furrows. Fixigenal area, between back of eye and posterior
border furrow, varying with length of eye; in smallest holaspides H/A is relatively large,
11-27 per cent., but during subsequent growth the eye appears to extend backwards
towards the posterior border furrow, H/A decreasing in value, varying in the largest
individuals between 3-7 per cent. At this condition the eye abuts against the posterior
border furrow and causes it to bulge outwards to the posterior. Table 1 shows the range
interval of H/A for arbitrary growth stages, relative age being approximately equated
with cephalic length. Variation in A also affects the eye index ratios giving a larger range
than would normally be expected; A/G, 52-79 per cent.; A/Gn, 42-62 per cent.

table 1. The range interval for the ratio H/A showing the interval decreasing during the holaspid
growth of Acastocephala dudleyensis sp. nov., cephalic length being used as a standard.

CLASS

RANGE

mm.

Acaste

downinqiae

Acastocephala macrops

A. dudleyensis

Acastoides consfricta

No

h/a RANGE

No

H/A RANGE

No

h/a RANGE

No

H/a RANGE

1-6- 2-6

1

23-20

2- 6-3-6

3

24-14 — 24-81

3

II- II — 27-44

3-6-4-6

5

15-38 — 25-10

IO

6-40 — 13-66

4-6 -5-6

5

13-33 — 22-56

2

1 Id 1— 15-28

3

5-55 — 1 1-76

1

7-31

5-6 -6-6

7

15-57 —2057

2

8-10— 12-50

3

4-76— 8-59

5

7-31 -12-50

6- 6 - 7-6

10

12-59 —29-43

1

13-04

1 1

4-57—10-75

7- 6 -8-6

1 6

13-57 —24-92

1

1 1-72

-2

5-31 — 6-39

6

2-96—10-00

8-6 -9-6

ii

12-50 — 24-84

6

7-45—16-66

3

3-22 — 6-45

2

4-00 — 8-69

9-6-10-6

12

13-14 -28-57

7

5-66— 17-18

1

6 • 60

IO-6-1 1-6

IO

17-85 — 2 5-00

4

8-33— 12-98

1

5-50

II -6-12-6

9

12-20 — 29-62

4

8-11 —17-15

12-6-13-6

1

8-11

Visual surface overhanging slightly the ocular platform; gently convex outwards;
bearing a similar number of lenses to A. macrops ; at maximum height of the surface,
columns of 9 lenses alternate with 8.

Pygidium subtriangular in outline; in posterior profile gently arched. Axis with low
or moderate convexity, raised above both axial furrows and pleurae; composed of 7
segments, of which 1-3 are clearly defined, separated by deep transverse furrows and
the remainder becoming less distinct posteriorly; postaxial ridge short. Pleurae with
rather low convexity (tr.); 4 in number, with frequently a trace of a fifth. Pleural
furrows more strongly defined than interpleural furrows but typically weaker than in
Acaste. Border wide, more extensive in adult holaspides, without appreciable flattening

206

PALAEONTOLOGY, VOLUME 9

or furrow. Margin in very small holaspides (observed to a cephalic length of 3-50 mm.,
and a pygidial length of 1-50 mm.) showing up to 4 pairs of lateral denticulations asso-
ciated with the posterior bands of pleurae 1-4. These denticulations are rapidly resorbed
to give the typically smooth margin of the adult. No traces have been observed of a
caudal termination.

Relationships. Adult specimens of A. dudleyensis and A. macrops may be differentiated
by the following characteristics. The frontal lobe of dudleyensis is more convex (sag.)
than that of macrops and appears more laterally extensive ; the axial furrows in the former
do not diverge evenly; the side furrows of dudleyensis are more similar to species of
Acaste than to macrops , the latter has 3S wide and deep and opening out abaxially and
2S short and linear while those of dudleyensis are gently curved, reaching, in the adult,
nearly to the axial furrows. The eye of dudleyensis is closer to the glabella; the posterior
border furrow bulges outwards behind the eye; the palpebral areas are less wide and
more steeply inclined to the axial furrows; the occipital ring stands higher. Young
specimens are more difficult to separate. The occipital ring of dudleyensis often bears
a median node in the young. A comparison of the rates of resorption of the genal pro-
jections and lateral pygidial denticulations shows that these features persist in A.
macrops to a later ontogenetic stage.

Range and Distribution. A. dudleyensis is recorded only from Dudley, Worcs. All the specimens seen
by the author are labelled Wenlock Limestone. The matrix associated with all, except enrolled speci-
mens, is, however, always a smooth, bluish or greenish-grey calcareous siltstone, a lithology which
may occur as siltstone interleavings within the Wenlock Limestone or in the upper members of the
Wenlock Shale. Little can be said, therefore, on the range of this trilobite until the detailed palaeon-
tology and stratigraphy of Dudley is known.

Additional Material. The bulk of the available material is to be found in the Ketley and Holcroft
Collections deposited in the Birmingham University Museum. All specimens are from the Wenlock
Limestone of Dudley. BU (Holcroft) 48, 50, 102, 295, 440, 474, 541, 548; BU (Ketley) 243, 244, 251 ;
BU (Keeping) 19; also BU 864-8; SM A28693, A 28771; BM In36158-61, 1560; GSM 19311, 19312,
19316.

Acknowledgements. I would like to express my deep gratitude to Dr. J. Shirley for his interest and en-
couragement at all stages of this work and for his critical reading of the manuscript. I wish also to thank
Professor T. S. Westoll for making available the facilities which enabled the work to be completed.
I am most grateful to the following for allowing me to borrow specimens in their care and for providing
facilities to study their collections; Mr. A. G. Brighton of the Sedgwick Museum, Mr. J. M. Edmonds
of the Oxford University Museum, Dr. W. T. Dean and Mr. S. E. Morris of the British Museum
(N.H.), Mr. D. E. White of the Geological Survey Museum, Dr. I. Strachan of Birmingham Univer-
sity, Mr. J. Norton of the Ludlow Museum, Dr. M. L. K. Curtis of the Bristol City Museum, and
Dr. D. A. Bassett of the National Museum of Wales. I would also like to thank Dr. J. F. Potter of
Norwood Technical College for making available specimens from his personal collection.

The work has been made possible by a grant from the Department of Scientific and Industrial
Research.

REFERENCES

Clarkson, e. n. K. 1965. Schizochroal eyes and vision of some Silurian acastid trilobites. Palaeontology,
9, 1-29, pi. 1-3.

curtis, m. l. k. and cave, R. 1964. The Silurian-Old Red Sandstone Unconformity at Buckover,
near Tortworth, Gloucestershire. Proc. Bristol Nat. Soc. 30, 427-42, pi. VII.
delo, d. m. 1935. A revision of the Phacopid trilobites. J. Paleont. 9, 402-20.

1940. Phacopid Trilobites of North America. Spec. Pap. Geol. Soc. Amer. 29, 1-135, 13 pi.

J. H. SHERGOLD: A REVISION OF ACASTE DO WNINGIAE 207

dixon, e. E. l. 1921. The Geology of the South Wales Coalfield. Pt. xm. The Country around Pem-
broke and Tenby. Mem. geo!. Surv. Engld & Wales.

emmerich, H. f. 1845. Ueber die Trilobiten. Neues Jb. Min. Geol. Paldont. 18-52.

goldfuss, a. 1843. Systematische Uebersicht der Trilobiten und Beschreibung einiger neuen Arten
derselben. Neues Jb. Min. Geol. Paldont. 537-67.

hollard, H. 1963. Les Acastella et quelques autres Dalmanitacea du Maroc presaharien, leur dis-
tribution verticale et ses consequences pour 1'etude de la limite Silurien-Devonien. Notes Serv.
geol. Maroc, 176, 1-66, 6 pi.

m’coy, f. 1851. A Synopsis of the Classification of the British Palaeozoic Rocks, with a Systematic
Description of the British Palaeozoic Fossils in the Geological Museum of the University of Cambridge.
Fasc. I, 1-184.

murchison, R. i. 1839. The Silurian System. London.

1859. Siluria. 3rd ed. London.

1872. Siluria. 5th ed. London.

reed, f. r. c. 1925. Some New Silurian Trilobites. Geol. Mag. 62, 61-16, p\. ll.

richter, r. 1909. Beitrdge zur Kenntnis devonischer Trilobiten aus dem Rheinischen Schiefergebirge.
1-96. Inaugural-Dissertation. Marburg.

and richter, e. 1939. Trilobiten aus dem Bosphorus-Gebiet. Abh. preuss. geol. Landesanst.,

n.f. 190, 1-49, pi. 16-19.

1952. Phacopacea von der Grenze Emsium/Eiflium (Tril.). Senckenbergiana leth. 33, 79-

108, pi. 1-4.

1954. Die Trilobiten des Ebbe-Sattels und zu vergleichende Arten (Ordovizium, Got-

landium/Devon). Abh. senckenb. naturf. Ges. 488, 1-76, pi. 1-6.

salter, J. w. 1848. in Palaeontological Appendix to Prof. John Phillips’s Memoir on the Malvern
Hills compared with the Palaeozoic Districts of Abberley etc. Mem. geol. Surv. U.K., 2, Pt. 1.

1853. Figures and Descriptions Illustrative of British Organic Remains, Decade VII. Mem.

geol. Surv. U.K.

1861. Note on the Fossils found in the Worcester and Hereford Railway Cuttings. Quart. J.

geol. Soc. Lond. 17, 161-2.

1864. Monograph of the British Trilobites from the Cambrian, Silurian and Devonian Forma-
tions. Palaeontogr. Soc. (Monogr.), 1-80, pi. I-VI (issued Aug. 1864).

schmidt, w. e. 1907. Uber Cryphaeus in dem Siegener Schichten. Z. dtsch. geol. Ges. 59, Monatsber.,
9-12.

sollas, w. J. 1879. On the Silurian District of Rhymney and Pen-y-lan, Cardiff. Quart. J. geol. Soc.
Lond. 35, 475-507, pi. XXIV.

struve, w. 1958c/. Beitrage zur Kenntnis der Phacopacea (Trilobita). 1. Die Zeliszkellinae. Sen-
ckenbergiana leth. 39, 165-220, pi. 4.

19586. Beitrage zur Kenntnis der Phacopacea (Trilobita). 2. Acastavinae n. subfam. Sen-
ckenbergiana leth. 39, 221-6.

1959. (in Moore, R. C.). Treatise on Invertebrate Paleontology. Pt. O, Arthropoda 1.

Stubblefield, c. J. 1938. The Types and Figured Specimens in Phillips and Salter’s Palaeontological
Appendix to John Phillips’s Memoir on The Malvern Hills compared with the Palaeozoic Districts
of Abberley, etc.’ Summ. Progr. geol. Surv. U.K. for 1936, Pt. II, 27-51.

symonds, w. s. and Lambert, a. 1861. On the Sections of the Malvern and Ledbury Tunnels (Wor-
cester and Hereford Railway) and the intervening Line of Railroad. Quart. J. geol. Soc. Lond. 17,
152-60.

tomczykowa, e. 1962a. O trylobicie Acastella prima n. sp. Instytut Geologiczny. Kwart. Geol.,
Warszawa, 6, 260-6, pi. 1 (Polish, English Summary).

19626. O rodzaju Scotiella Delo z warstw rzepinskich Gor Swictokrzyskich (On the Genus

Scotiella Delo (Trilobita) from the Rzepin Beds of the Holy Cross Mts.). Acta geol. pol., Ksi^ga
Pam. J. Samsonowicza, 187-205, 2 pi. (Polish, English Summary).

J. H. SHERGOLD
Department of Geology,
University of Newcastle upon Tyne,
Newcastle upon Tyne, 1

Manuscript received 17 February 1965

OSTRACODA FROM THE UPPER TEALBY CLAY
(LOWER BAR REM I AN) OF SOUTH LINCOLNSHIRE

by p. kaye and d. barker

Abstract. Thirty-two species and subspecies of Ostracoda are recorded from the Upper Tealby Clay (Lower
Barremian) at Dalby Hill, South Lincolnshire. Three species, Cytherelloidea dalbyensis, Amphicytherura barten-
steini and Orthonotacythere problematica are considered new; two species are left under open nomenclature.
Abundance data are given, and the ranges of the species in other British Lower Cretaceous formations are
shown.

Since the latter part of last century the Tealby Series of the Wolds of Lincolnshire
has been divisible into three on a lithological basis (Judd 1867, 1870; Lamplugh 1896).
However, the relationship to the beds above and below is uncertain due to a paucity
of fossils and exposures. The three lithologic divisions of the Tealby Series are from the
base, the Lower Tealby Clay followed by the Tealby Limestone, and at the top the
Upper Tealby Clay. In this account we deal only with the ostracods from the Upper
Tealby Clay, the major exposure of which in the South Wolds at the present time is
the road cutting at Dalby Hill (G.R. TF/409696). In the course of field mapping during
the period 1963-4, J. Newton-Smith of Leicester University brought the roadside section
at Dalby Hill to our notice. The ostracods found in the washings of his samples prompted
the authors to collect further samples during April 1964. The Upper Tealby Clay is
generally a dark clay very rich in limonite ooliths; locally it is glauconitic and there
are occasional horizons rich in pyrite. Macrofossils obtained from this clay include
Liostrea, Oxyteuthis , and Aulacoteuthis. The Ostracoda obtained from this horizon are
of Lower Barremian age. The specimens described are all deposited in the collections
of the British Museum of Natural History (BMNH), London.

Acknowledgements. The authors thank Mr. J. Newton-Smith for his assistance, and are grateful to
other members of Reading and Leicester Universities for help during collection of the samples, and to
Mr. J. L. Watkins of Reading University for the photographs.

SYSTEMATIC DESCRIPTIONS
Order podocopida Muller 1894
Suborder platycopina Sars 1866
Family cytherellidae Sars 1866
Genus cytherelloidea Alexander 1929

Cytherelloidea ovata Weber 1934

1934 Cytherelloidea ovata Weber, 145, pi. 8, fig. 4.

1950 Cytherelloidea ovata Weber; Wolburg, 162.

1963n Cytherelloidea ovata Weber; Kaye, 115, pi. 20, figs. 18-21.

Material. A left valve BMNH Io.3048, from the Upper Tealby Clay, Dalby Hill, Lines.
[Palaeontology, Yol. 9, Part 2, 1966, pp. 208-19, pi. 33.]

KAYE AND BARKER: OSTRACODA FROM THE UPPER TEALBY CLAY 209

Cytherelloidea dalbyensis sp. nov.

Plate 33, figs 1, 2

Holotype. A right valve BMNH Io.3068, from the Upper Tealby Clay, Dalby Hill, Lines.

Pctratype. A right valve BMNH Io.3069, from the same horizon and locality.

Diagnosis. Large Cytherelloidea, strongly compressed laterally. Dorsal rib complex,
joined both anteriorly and posteriorly to the median rib. Shell surface strongly pitted,
almost reticulate.

Measurements. R.V. Holotype Io.3068; length 0-80 mm., height 0-30 mm.

Description. Valves elongate, rather thin and strongly compressed laterally. The dorsal
and ventral margins are subparallel but both show a weak concavity centrally in the
larger right valves. Anterior margin semicircular, posterior margin rather subquadrate.
The lateral surface bears a series of low ribs with shallow depressions between them.
Posteriorly the valve is more inflated having a swollen area from which the longitudinal
ribs run. Anteriorly there is a crescentic weakly-inflated marginal area equal in width
to about one-sixth of the length. This anterior marginal ‘strip’ merges with the long
margins, antero-dorsally and antero-ventrally. Three longitudinal ribs cross the lateral
surface, the ventral one being most prominent. This rib runs in a weakly arcuate path
a short distance above the ventral margin for the bulk of the length of the valve;
posteriorly it is joined to the inflated posterior area but anteriorly it is separated from
the marginal strip by a shallow groove. Ventrally from this rib there is a wide marginal
shelf. The median rib starts posteriorly at J height running horizontally for a short dis-
tance and then bifurcates at f length. A long ventrally convex portion runs below the
muscle pit terminating at \ length whilst a short upper portion runs dorsally to join
the dorsal rib at the mid-length of the valve. The dorsal rib is weak and short; it runs
from a position on the dorsal margin at mid-length of the valve obliquely to join the
anterior end of the lower branch of the median rib; it is entirely separate from the
anterior marginal strip. Between the ribs the valve surface is strongly pitted.

Remarks. Cytherelloidea dalbyensis shows strong affinities to C. elongata Kaye 1963a
from the Middle Barremian of East Yorkshire but is distinct in the pattern of the ribs
from any other described Mesozoic species of the genus. From C. elongata, C. dalbyensis
differs in shape and in being much less inflated with the ribs and depressions much less
well differentiated; there are also differences in the arrangement of the ribs. In C.
elongata the ventral rib runs along the margin and is joined anteriorly to a much
broader anterior marginal ‘strip’. The dorsal rib also runs to join this marginal strip
and not the anterior end of the median rib.

Suborder podocopina Sars 1866
Superfamily cytheracea Baird 1850
Family cytherideidae Sars 1925
Genus dolocytheridea Triebel 1938

210

PALAEONTOLOGY, VOLUME 9

Dolocytheridea intermedia Oertli 1958

1958 Dolocytheridea intermedia Oertli, 1505, pi. 3, figs. 63-74, pi. 4, figs. 75-82.

1963c Dolocytheridea intermedia Oertli; Kaye, 33, pi. 3, figs. 10-14.

Material. A right valve BMNH Io.3051, from the Upper Tealby Clay, Dalby Hill, Lines.

Genus clithrocytheridea Stephenson 1936

Clithrocytheridea brevis (Cornuel) 1846

1846 Cy there amygdaloides var. brevis Cornuel, 199, pi. 7, fig. 12.

1956 Clithrocytheridea brevis (Cornuel) Deroo, 1510, pi. 11, figs. 15-21.

Material. Five valves BMNH Io.3052, from the Upper Tealby Clay, Dalby Hill, Lines.

Genus schuleridea Swartz and Swain
Schuleridea cf. bernouilensis Grosdidier 1964

Plate 33, figs. 9, 10

71964 Schuleridea bernouilensis Grosdidier, 225, pi. 1, figs. 2 a-g.

Material. Two valves and one carapace BMNH Io. 3053-5, from the Upper Tealby Clay, Dalby Hill
Lines.

Remarks. Abundant specimens of a species very closely akin to the French Lower
Barremian form S. bernouilensis Grosdidier are found at Dalby Hill. Apart from a
much poorer development of the eye node in the British material the two forms seem
to be identical.

Schuleridea rhomboidalis Neale 1960

1960 Schuleridea rhomboidalis Neale, 210, pi. 2, figs. 1, 2, 5, 7, 8.

1963c Schuleridea rhomboidalis Neale; Kaye, 32, pi. 3, figs. 1-4.

Material. A right valve BMNH Io.3058, from the Upper Tealby Clay, Dalby Hill, Lines.

EXPLANATION OF PLATE 33

All figs. X 60; all specimens from the Upper Barremian of Dalby Hill, Lines.

Figs. 1, 2. Cytherelloidea dalbyensis sp. nov. 1, R.V. (Holotype), lateral view; Io.3068. 2, R.V. (Para-
type), lateral view; Io.3069.

Figs. 3, 4. Orthonotacythere problematica sp. nov. 3, L.V. (Holotype), lateral view; Io.3060. 4, Cara-
pace (Paratype), dorsal view; Io.3061.

Figs. 5-8. Amphicytherura bartensteini sp. nov. 5, Female L.V. (Holotype), lateral view; Io.3071. 6,
Male R.V. (Paratype), lateral view; Io.3073. 7, Male L.V. (Paratype); lateral view; Io.3072. 8,
Female R.V. (Paratype), lateral view; Io.3074.

Figs. 9, 10. Schuleridea cf. bernouilensis Grosdidier. 9, L.V., lateral view; Io.3053. 10, R.V., lateral
view; Io.3054.

Figs. 11, 12. Amphicytherura roemeri (Bartenstein). 11, Male L.V., lateral view; Io.3034. 12, Female
R.V., lateral view; Io.3035.

Fig. 13. Ranocythereis caistorensis Kaye. R.V. lateral view; Io.3033.

Fig. 14. Acrocythere hauteriviana laeva Neale. Male L.V., lateral view; Io.3046.

Fig. 15. Acrocythere hauteriviana hauteriviana (Bartenstein). Female R.V., lateral view; Io.3044.

Figs. 16, 17. Cythereis cf. geometrica Damotte and Grosdidier. 16, Male L.V., lateral view; Io.3063.
17, Male R.V., lateral view; Io.3064.

Palaeontology, Vol. 9

PLATE 33

KAYE and BARKER, Lower Cretaceous ostracods

KAYE AND BARKER: OSTRACODA FROM THE UPPER TEALBY CLAY 211

Genus apatocythere Triebel 1940
Apatocythere ellipsoidea Triebel 1940

1940 Apatocythere ellipsoidea Triebel, 171, pi. 2, figs. 20-26, pi. 9, figs. 104.
1956 Apatocythere ellipsoidea Triebel; Deroo, 1510.

1963c Apatocythere ellipsoidea Triebel; Kaye, 32, pi. 3, figs. 5-7.

Material. A left valve BMNH Io.3056, from the Upper Tealby Clay, Dalby Hill, Lines.

Apatocythere simulans Triebel 1940

1940 Apatocythere simulans Triebel, 170, pi. 1, figs. 14-19, pi. 9, fig. 103.
1960 Apatocythere simulans Triebel; Neale, 210, pi. 2, figs. 1, 2, 5, 7, 8.
1963c Apatocythere simulans Triebel; Kaye, 32, pi. 3, figs. 1-4.

Material. A right valve BMNH Io.3057, from the Upper Tealby Clay, Dalby Hill, Lines.

Genus euryitycythere Oertli 1959
Euryitycythere sp. B

Material. A carapace BMNH Io.3070, from the Lower Barremian, Dalby Hill.

Measurements. Carapace Io.3070; length 0-60 mm., height 0-32 mm.

Remarks. This extremely rare species is represented by a single closed carapace from
Dalby Hill. It is closely related to the Hauterivian species E. parisiorum Oertli 1959 but
differs in having a crest-like ridge running along the dorsal margin of the larger left
valve. This crest also partially overlaps the dorsal margin of the right valve. A further
difference is the stronger flattening of the lateral expansion ventrally forming a short
blade-like longitudinal rib.

This occurrence is the first record of the genus in the Barremian though a worn but
distinctly different specimen has been recorded (Kaye 1965c) from the Upper Aptian
of the Isle of Wight.

Holotype. A female left valve BMNH Io.3071, from the Upper Tealby Clay, Dalby Hill, Lines.
Paratypes. Six valves BMNH Io. 3072-5, from the same horizon and locality.

Diagnosis. Amphicytherura with strongly reticulate lateral surface and marked keel-like
ventral longitudinal rib.

Family cytheruridae Midler 1894
Genus amphicytherura Butler and Jones 1957

Amphicytherura hartensteini sp. nov.

Plate 33, figs. 5-8

Measurements.

Female L.V. Holotype Io.3071
Female R.V. Paratype Io.3074
Male L.V. Paratype Io.3072

Length Height
0-57 mm. 0-30 mm.
0-57 mm. 0-30 mm.
0-60 mm. 0-30 mm.

212

PALAEONTOLOGY, VOLUME 9

Description. Valves subquadrate in outline, strongly laterally compressed. Dorsal and
ventral margins straight and subparallel in the left valves. Anterior margin bluntly
rounded, slightly asymmetrical ventrally; posterior drawn out into an acute point at
f height. Lateral surface bearing a series of ribs. The most prominent is high keel-like,
ventrally arcuate, ventral longitudinal; this rib runs from the anterior margin at \
height across the lateral surface to terminate on the postero-lateral surface at mid-
height. A further longitudinal rib runs below this one on the ventral undersurface
following a ventrally identical path but reaching the margin postero-ventrally as well
as antero-ventrally. A low smooth eye tubercle occurs antero-dorsally bearing a series
of weak ribs on its upper surface. These ribs run weakly ventrally and antero-ventrally
for a short distance across the lateral surface. A short rib runs from the dorsal margin
at mid-length curving towards the posterior margin and terminating at mid-height. A
weak node or rudimentary median rib occurs postero-laterally in certain specimens.
The whole of the lateral surface is strongly reticulate. Duplicature narrow crossed by
few straight radial pore canals. Hinge strongly amphidont, with boss-like divided ter-
minal teeth and strongly divided median furrow in the right valve. Muscle group a row
of four oval scars slightly inclined posteriorly with a single small oval scar anteriorly
to them opposite the centre of the row.

Remarks. The strong reticulate ornamentation of this species and weak dorsal and
median ribbing make it distinct from other members of the genus. It is one of the most
abundant ostracods in the Upper Tealby Clay (Lower Barremian) of South Lines, and
is one of the few abundant ostracods from this formation not found in the Lower
Tealby Clay and Tealby Limestone (U. Hauterivian) of North Lines.

Amphicytherura roemeri (Bartenstein) 1956
Plate 33, figs. 11, 12

1956 Orthonotacy there roemeri Bartenstein, 531, pi. 3, figs. 76-77.

1965n Amphicytherura roemeri (Bartenstein), Kaye, 79, pi. 5, figs. 5-7.

Material. Two valves BMNH Io. 3034-5, from the Upper Tealby Clay (Lower Barremian), Dalby
Hill, Lines.

Remarks. This species, characteristic of the Upper Hauterivian in N. Lines, occurs
fairly abundantly in the Lower Barremian of S. Lines.

Genus eucytherura M tiller 1894
Eucytherura nettletonensis Kaye 1 9646

19646 Eucytherura nettletonensis Kaye, 321, pi. 55, figs. 5, 6, 8.

Material. Four valves BMNH Io.3040, from the Upper Tealby Clay, Dalby Hill, Lines.

Genus cytherura Sars 1866
Cytherura reticulosa (Chapman) 1894

1894 Cytheropteron reticulosum Chapman, 692, pi. 33, figs. 6 a-c.

1964 b Cytherura reticulosa (Chapman); Kaye, 318, pi. 55, figs. 7, 9.

Material. A left valve BMNH Io.3038, from the Upper Tealby Clay, Dalby Hill, Lines.

KAYE AND BARKER: OSTRACODA FROM THE UPPER TEALBY CLAY 213

Genus cytheropteron Sars 1866
Subgenus cytheropteron Sars 1866

Cytheropteron ( Cytheropteron ) reightonensis Kaye 1964#

1964a Cytheropteron (C.) reightonensis Kaye, 102, pi. 5, figs. 1-5.

Material. A left valve BMNH Io.3039, from the Upper Tealby Clay, Dalby Hill, Lines.

Cytheropteron (C.) rugosa Kaye 1965a

1965c Cytheropteron (C.) rugosa Kaye, 38, pi. 8, figs. 4-5.

1965 d Cytheropteron (C.) rugosa Kaye, Kaye and Barker, 379, pi. 48, fig. 6.

Material. A left valve BMNH Io.3037, from the Upper Tealby Clay, Dalby Hill, Lines.

Subgenus eocytheropteron Alexander 1933

Cytheropteron ( Eocytheropteron ) nova Kaye 1964a

1964a Cytheropteron ( Eocytheropteron ) nova, Kaye, 104, pi. 5, fig. 6.

1965a Cytheropteron ( Eo .) nova Kaye; Kaye, 75, pi. 5. figs. 18-20.

Material. Two valves BMNH Io.3062, from the Upper Tealby Clay, Dalby Hill, Lines.

Subgenus infra cytheropteron Kaye 1964a

Cytheropteron ( Infracytheropteron ) exquisita Kaye 1964a

1964a Cytheropteron ( Infracytheropteron ) exquisita Kaye, 105, pi. 5, figs. 9-10.

1965 Cytheropteron (/.) exquisita Kaye; Kaye and Barker, 380.

Material. A right valve BMNH Io.3036, from the Upper Tealby Clay, Dalby Hill, Lines.

Genus orthonotacythere Alexander 1934
Or thonotacy there blanda Kaye 1 963e
1963c Orthonotacythere blanda Kaye, 437, pi. 61, figs. 17, 18.

Material. A left valve BMNH Io.3043, from the Upper Tealby Clay, Dalby Hill, Lines.

Orthonotacythere inversa inversa (Cornuel) 1848

1848 Cythere inversa Cornuel, 244, pi. 1, figs. 12-14.

1963c Orthonotacythere inversa inversa (Cornuel); Kaye, 435, pi. 61, figs. 1-8, 12, 13.
Material. A left valve BMNH Io.3042, from the Upper Tealby Clay, Dalby Hill, Lines.

Orthonotacythere problematiea sp. nov.

Plate 33, figs. 3, 4

1965c Orthonotacythere sp. A. Kaye.

Holotype. A left valve BMNH Io.3060 from the Lower Barremian at Dalby Hill, Lines.

Paratypes. One valve and one carapace BMNH Io.3061, from the Lower Barremian at Dalby Hill;
4 valves and 1 carapace BMNH lo. 2070-74, from the basal Atherfield Clay (Lower Aptian) Ather-
field, Isle of Wight.

214

PALAEONTOLOGY, VOLUME 9

Diagnosis. Small Orthonotacy there with tuberculate lateral surface and ventral ridge
increasing in height posteriorly. The ventral ridge is drawn out into an alaeform spine
posteriorly.

Measurements. Holotype, L.V. Io.3060; length 0-47 mm., height 0-25 mm.

Remarks. This species originally described from the Lower Aptian of the Isle of Wight
is here named. The Upper Tealby specimens are a little larger than the Aptian ones
and have the ventral row of tubercles fused to form a prominent ridge.

Genus acrocythere Neale 1960
Acrocythere hauteriviana hauteriviana (Bartenstein) 1956
Plate 33, fig. 15

1956 Orthonotacy there hauteriviana Bartenstein, 532, pi. 3, fig. 81.

1965a Acrocythere hauteriviana hauteriviana (Bartenstein); Kaye, 78, pi. 5, figs. 12-13.

Material. Two valves BMNH Io. 3044-5, from the Lower Barremian, Dalby Hill, Lines.

Remarks. This characteristic Hauterivian and Lower Barremian subspecies is fairly
common at Dalby Hill. The most abundant member of this group in this formation
being A. hauteriviana laeva Neale, the reverse of the situation at Speeton.

Acrocythere hauteriviana laeva (Neale) 1960

Plate 33, fig. 14

1960 Orthonotacythere ( Acrocythere ) hauteriviana laeva Neale, 213, pi. 3, figs. 10 a-b, pi. 4,
fig. 13.

1965a Acrocythere hauteriviana laeva (Neale); Kaye 78, pi. 5, fig. 17.

Material Two valves BMNH Io. 3046-7, from the Upper Tealby Clay (L. Barremian), Dalby Hill,
Lines.

Remarks. This species is extremely abundant at Dalby Hill but is rare at Speeton.

Genus pseudobythocythere Mertens 1956
Pseudobythocythere ornata Kaye 1965#

Material Three valves BMNH Io.3067, from the Upper Tealby Clay, Dalby Hill, Lines.

Pseudobythocythere vellicata (Chapman) 1894

1894 Cvtheridea vellicata Chapman, 690, pi. 33, figs. 3 a-c.

1964 b Pseudobythocythere vellicata (Chapman); Kaye, 323, pi. 54, figs. 14, 17, 18, pi. 55, figs.

10, 11.

1965c Pseudobythocythere vellicata (Chapman); Kaye, 42, pi. 8, figs. 14, 15.

Material Four valves BMNH Io.3066, from the Upper Tealby Clay, Dalby Hill, Lines.

KAYE AND BARKER: OSTRACODA FROM THE UPPER TEALBY CLAY 215

Family progonocytheridae Sylvester-Bradley 1948
Subfamily progonocytherinae Sylvester-Bradley 1948
Genus neocythere Mertens 1956

Neocy there (N.) protovanveeni Kaye 19636
Material. Four valves BMNH Io.3050 from the Upper Tealby Clay, Dalby Hill.

Subfamily protocytherinae Mandelstam 1960
Genus protocythere Triebel 1938

Protocythere inornata Kaye 19646

Material. A right valve BMNH Io.3049 from the Upper Tealby Clay, Dalby Hill.

Protocythere hechti Triebel 1938

1938 Protocythere hechti Triebel, 189, pi. l.figs. 11-16.

1956 Protocythere hechti Triebel; Deroo, 1513.

1956 Protocythere hechti Triebel; Bartenstein, 530, pi. 3. figs. 73-75.

1962 Protocythere hechti Triebel; Neale, 446, pi. 9. figs. 5-7.

Material. A left valve BMNH Io.3059 from the Upper Tealby Clay, Dalby Hill.

Family trachyleberididae Sylvester-Bradley 1948
Genus cythereis Jones 1849

Cythereis blanda Kaye 1 963t/

Material. Two valves BMNH Io.3041 from the Upper Tealby Clay, Dalby Hill.

Cythereis cf. geometrica Damotte and Grosdidier 1963
Plate 33, figs. 16, 17

Material. Five valves and one carapace BMNH Io. 3063-5 from the Upper Tealby Clay, Dalby Hill.

Remarks. Specimens related to C. geometrica Damotte and Grosdidier 1963 occur
abundantly in the Upper Tealby Clay. The major difference between the two species is
the stronger development of the median longitudinal ridge in the Lower Barremian
forms. This feature shows similarities to C. geometrica fittoni Kaye 1965c from the
Upper Aptian of the Isle of Wight but the Lincolnshire specimens lack the strong surface
reticulation of the latter species. It is possible that the Lower Barremian forms are
ancestral to the Aptian species.

Genus ranocythereis Kaye 19656
Ranocythereis caistorensis Kaye 19656
Plate 33, fig. 13

Material. A right valve BMNH Io.3033 from the Upper Tealby Clay, Dalby Hill, Lines.

Q

C 3803

216

PALAEONTOLOGY, VOLUME 9

Remarks. This species, previously recorded from the Lower Tealby Clay and Tealby
Limestone (Hauterivian) of Nettleton,N. Lines., occurs moderately rarely at Dalby Hill.

CONCLUSIONS

The Upper Tealby Clay at Dalby Hill is rich in ostracoda, both in numbers of in-
dividuals, and in number of species. The relative abundances of the thirty-two species
and subspecies recorded are summarized as follows:

Abundant

Clithrocytheridea brevis (Cornuel); Eucytherura nettletonensis Kaye; Schuleridea cf.
bernouilensis Grosdidier; Amphicytherura bartensteini sp. nov. ; Acroeythere hauteriviana
laeva Neale; Pseudobythocy there ornata Kaye; Pseudobythoey there vellicata (Chapman);
Neocythere (N.) protovanveeni Kaye ; Protocythere hechti Triebel ; Cythereis blanda Kaye ;
Cythereis cf. geometrica Damotte and Grosdidier.

Common

Schuleridea rhomboidalis Neale; Apatocythere ellipsoidea Triebel; Amphicytherura
roemeri (Bartenstein); Cytheropteron ( Eo .) nova Kaye; Acroeythere hauteriviana hauteri-
viana (Bartenstein).

Frequent

Cytherelloidea ovata Weber; Apatocythere sinudans Triebel; Dolocytheridea intermedia
Oertli; Orthonotacythere inversa inversa (Cornuel); Protocythere triplicata (Roemer);
Ranocythereis caistorensis Kaye.

Rare

Cytherelloidea dalbyensis sp. nov.; Bairdia sp.; Cytherura reticulosa (Chapman);
Cytheropteron (C.) reightonensis Kaye; Euryitycy there sp. B.; Cytheropteron (C.) rugosa
Kaye; Cytheropteron [Infra.) exquisita Kaye; Orthonotacythere blanda Kaye; Orthono-
tacythere problematica; Protocythere inornata Kaye.

The known range of the Upper Tealby Ostracoda and their geographical occurrence
in England are shown in text-fig. 1 .

The Upper Tealby fauna is Lower Barremian in age; with such species as Orthono-
tacythere blanda , O. inversa inversa and, Pseudobythocy there ornata being only known
previously from beds at this horizon at Speeton, E. Yorks. The fauna as a whole shows
strong affinities to the lithologically similar Upper Hauterivian at Nettleton in N. Lines.,
seventeen of the species recorded at Dalby Hill being also known from Nettleton;
Lower Barremian strata are absent in N. Lines. The clays of Upper Hauterivian and
Lower Barremian age at Speeton differ markedly from those at Dalby, lacking the
glauconite and iron carbonate ooliths. Nine species found at Dalby Hill occur in higher
Barremian horizons at Speeton; the known range of one ( Cythereis blanda) being ex-
tended down into the Lower Barremian. Four Hauterivian species: Eucytherura nettle-
tonensis, Amphicytherura roemeri, Acroeythere hauteriviana laeva, and Ranocythereis
caistorensis have their known ranges extended up into the Lower Barremian whilst

KAYE AND BARKER: OSTRACODA FROM THE UPPER TEALBY CLAY 217

S PECI ES

UPPER

HAUTERI VIAN

BARREMIAN

APTIAN

UPPER

LOWER

M & U

L&u

UPPER

Nettleton
N . Lines.

Speeton
E. Yorks.

Dalby Hill
S . Lines.

Speeton
E . Yorks

Speeton
E .Yorks

South

East

England

Suttersby
S. Lines.

Cytherelloidea dalbyensis nov.

X

Cytherelloidea ovata Weber.

X

X

X

X

X

X

X

Bairdia sp.

X

Clithrocytheridea brevis (Cornuel).

X

Dolocytheridea intermedia Oertli.

X

X

X

X

X

Schuleridea cf. bernouilensis Grosdidier.

X

1

Schuleridea rhomboidalis Neale.

X

X

X

X

Apatocythere ellipsoidea Triebel.

X

X

X

X

X

Apatocythere simulans Triebel.

X

X

X

X

Euryitycythere sp. B.

X

Cytherura reticulosa (Chapman).

X

X

X

X

Eucytherura netllefonensis Kaye.

X

X

Amphicytherura bartenstelni nov.

X

Amphicytherura roemeri ( Bartenstein ).

X

X

Cytheropteron (C.) reightonensis Kaye.

X

X

X

Cytheropteron (C.)rugosa Kaye.

X

X

X

X

Cytheropteron (Eo)nova Kaye.

X

X

X

X

X

Cytheropteron (Infra) exquisite Kaye.

X

X

X

X

X

Orthonotacythere inversa inversa (Cornvel).

X

X

Orthonotacythere blanda Kaye.

X

X

Orthonotacythere problematica nov.

X

Pseudobythocythere ornata Kaye.

X

X

X

Pseudobythocythere vellicata (Chapman).

X

X

Acrocythere hauteriviana hautenviana(Bartenstein).

X

X

X

X

X

Acrocythere hauteriviana laeva Neale.

X

X

X

Neocythere (N) protovanveeni Kaye

X

X

X

X

Protocythere inornata Kaye.

X

X

Protocythere hechti Triebel.

X

X

X

X

Protocythere triplicato ( Roem er ).

X

X

X

X

X

Cythereis blanda Kaye.

X

X

X

Cythereis cf. geometica D&G .

X

X

Ronocyth ereis caistorensis Kaye.

X

X

text-fig. 1. Known distribution of Ostracoda found in the Upper Tealby Clay at Dalby Hill.

218

PALAEONTOLOGY, VOLUME 9

the ranges of Pseudobythocythere vellicata and Protocythere inornata are extended down
into the Lower Barremian. Two already described species, Clithrocytheridea brevis and
Scluderidea bernouilensis are recorded from England for the first time.

There is surprisingly little relationship between the fauna of the Upper Tealby Clay
and the Upper Aptian Sutterby Marl formation. Only five species are common to the
two, none of which are abundant at Sutterby. Stronger ties are found with the Aptian
of South-east England, eight species occurring there and at Dalby Hill. None of the
species extend into post-Aptian sediments.

REFERENCES

bartenstein, H. 1956. Zur Mikrofauna des englischen Elauterive. Senck. leth. 37, 509-33, 3 pi.
chapman, F. 1 894. The Bargate beds of Surrey and their microscopic contents. Quart. J. Geol. Soc.
Loud. 50, 677-92, pi. 33-34.

cornuel, J. 1846. Description des Entomostraces fossiles du terrain Cretace Inferieur du Departe-
ment de la Haute-Marne. Bull. Soc. geol. Fr. 1, 193-205, pi. 7.

1848. Description des nouveaux fossiles microscopiques du terrain Cretace Inferieur du Departe-

ment de la Haute-Marne. Bull. Soc. geol. Fr. 3, 241-6, pi. 1.
damotte, R. and grosdidier, e. 1963. Quelques Ostracodes du Cretace Inferieur de la Champagne
Humide. 2 — Aptien. Rev. Micropaleont. 6, 153-68, 3 pi.
deroo, g. 1956. Etude critique au sujet des Ostracodes marins du Cretace Inferieur de la Champagne
Humide et du Bas-Boulonnais. Rev. Inst. Frang. Petr. 11, 1499-539, pi. 1-5.
grosdidier, e. 1964. Quelques Ostracodes nouveaux du Cretace Inferieur de Champagne Humide.

Ill Barremian- — Hauterivian. Rev. Micropaleont. 6, 223-36, pi. 1-3.
kaye, p. 1963c. Species of the ostracod Family Cytherellidae from the British Lower Cretaceous.
Senck. leth. 44, 109-25, pi. 18-20.

19636. The ostracod genus Neocy there in the Speeton Clay. Palaeontology, 6, 274-81, pi. 41.

1963c. The interpretation of the Mesozoic ostracod genera of the family Cytherideidae Sars

1925. Rev. Micropaleont. 6, 23—40, pi. 1-3.

1963 d. Ostracoda of the subfamilies Protocytherinae and Trachyleberidinae from the British

Lower Cretaceous. Paldont. Z. 37, 225-38, pi. 18, 19.

- 1963c. The ostracod species Orthonotacythere inversa (Cornuel) and its allies in the Speeton Clay
of Yorkshire. Palaeontology, 6, 430-9, pi. 61.

1964c. Ostracoda of the genera Eucytherura and Cytheropteron from the Speeton Clay. Geol.

Mag. 101, 97-107, pi. 4, 5.

19646. Revision of the Ostracoda from the Bargate Beds in Surrey Palaeontology, 7, 317-30, pi.

54-55.

— — 1965c. Further Ostracoda from the British Lower Cretaceous. Senck. leth. 46, 73-81, pi. 5.

19656. Ranocythereis, a new Ostracod genus from the British Lower Cretaceous. Senck. leth. 46,

83-87, pi. 6.

1965c. Ostracoda from the Aptian of the Isle of Wight, England. Paldont. Z. 39, 33-50, pis. 6-8.

and barker, d. 1965. Ostracoda from the type Sutterby Marl (U. Aptian) of Lincolnshire.

Palaeontology, 8, 375-90, pis. 48-50.

neale, J. w. 1960. Marine Lower Cretaceous Ostracoda from Yorkshire, England. Micropaleon-
tology, 6, 203-24, pi. 1-4.

oertli, h. J. 1958. Les Ostracodes de l’albien-aptien d'apt. Rev. Inst, frang. petrole, 13, 1499-1537,
pi. 1-9.

1959. Euryitycythere und Parexophthalmocy there, zwei neue Ostrakoden-Gattungen aus der

Unterkreide Westeuropas. Paldont. Z. 33, 241-6, pi. 32.
swinnerton, H. h. 1935. The Rocks below the Red Chalk of Lincolnshire and their Cephalopod
faunas. Quart. J. geol. Soc. Lond. 91, 1-46.

1936. A monograph of British Lower Cretaceous Belemnites. Palaeontogr. Soc. [Monogr.\, 1-86.

KAYE AND BARKER: OSTRACODA FROM THE UPPER TEALBY CLAY 219

triebel, e. 1938. Ostracoden Untersuchungen. 1. Protocythere und Exophthalmocy there, Zwei neue
Ostracoden-Gattungen aus der Deutschen Kreide. Senckenbergiana , 20, 178-200, pi. 1-3.

1940. Die Ostracoden der Deutschen Kreide III Cytherideinae und Cytherinae aus der Unteren

Kreide. Ibid. 22, 160-227, pi. 1-10.

weber, h. 1934. Ostracoden aus dem Hauterive von Wenden am Mittelland-Kanal. Niedersdchs.
geol. Ver. 26, 139-49, pi. 8, 9.

P. KAYE

Burmah Oil Exploration Co.,
20 Esplanade,
Scarborough, Yorks.

D. BARKER

Paleoservices Ltd.,

162 High Street,

Manuscript received 8 January 1965 Watford, Herts.

THE FORAMINIFERAL GENUS BOL1VINOIDES
FROM THE UPPER CRETACEOUS OF THE
BRITISH ISLES

by F. T. BARR

Abstract. Nine species of Bolivinoides are described from the Upper Cretaceous (Santonian to Lower Maestrich-
tian) strata of the British Isles, and their stratigraphical ranges are recorded in relation to the classical mega-
fossil zonation. Three of these species, Bolivinoides hiltermanni, B. praelaevigata, and B. sidestrandensis, are
described as new and a lectotype is proposed for B. decorata (Jones). The phylogeny of various lineages of the
genus Bolivinoides is discussed.

During the last fifteen years, a series of important studies has been published on the
Upper Cretaceous-Lower Tertiary foraminiferal genus Bolivinoides Cushman. In Ger-
many, Hiltermann (1952, 1963) and Hiltermann and Koch (1950, 1955, 1960, 1962) have
published a number of valuable works demonstrating the stratigraphical value of Boli-
vinoides in strata of Santonian to Danian age. The usefulness of this group has been
shown by Edgell (1954) in Australia, Reiss (1954) in the Middle East, Pozaryska (1954)
in Poland, and Vassilenko (1961) in the Soviet Union. The stratigraphical ranges of
many species of this rapidly evolving group appear to be almost identical over much
of the world. Consequently, some lineages of Bolivinoides have proved to be extremely
useful in intercontinental correlations within the upper part of the Upper Cretaceous.

The Upper Cretaceous Chalk of the British Isles usually contains abundant and
diverse foraminiferal faunas. In strata of Santonian to Maestrichtian age, various species
of Bolivinoides are often an important element of this fauna. A number of short publica-
tions mentioned the presence of one or more of these species in the British Upper
Cretaceous, but no attempt had been made to monograph this group or to establish
their usefulness in the precise dating of the Upper Cretaceous Chalk in the British Isles.
Wright (1886) originally described Bolivinoides decorata from the Upper Campanian of
Keady Hill, Northern Ireland. In a short note, Chapman (1892) described B. strigillata
for the first time from the Santonian phosphatic chalk at Taplow in Buckinghamshire.
B. decorata was recorded by Heron-Alien and Earland (1910) from the interior of a
flint nodule found on the beach at Selsey Bill, Sussex. Williams-Mitchell (1948, p. 106,
pi. 9, fig. 3, pi. 10) illustrated a specimen of B. decorata from southern England which
he misidentified as B. strigillata (Chapman). In an important work on the Upper
Cretaceous Foraminifera from Northern Ireland, McGugan (1957) recorded and
illustrated specimens identified as B. decorata and B. draco (Marsson) from the Bally-
castle Pellet Chalk. Curry (1962) recorded several species of Bolivinoides from sub-
marine cores taken from the English Channel, but none of the specimens were illustrated
or described. Recently, Barr and Cordey (1964) re-examined Chapman’s (1892) material
and proposed a lectotype for B. strigillata.

The purpose of the present paper is to describe the various species of Bolivinoides
found in the British Isles and to record their stratigraphical ranges, which can be

Palaeontology, Yol. 9, Part 2, 1966, pp. 220-43, pis. 34-38.]

F. T. BARR: FOR A M INIFER AL GENUS BOLIVINOIDES

221

compared with the classical Upper Cretaceous megafossil zonation recognized in
western Europe. It is hoped that this study will aid in the precise dating and correla-
tion of the Upper Cretaceous strata of the British Isles and further our understanding
of the evolution of certain lineages of the genus Bolivinoides.

Deposition of types. The lectotype and syntypes of Bolivinoides decorata (Jones) are housed in the
Joseph Wright Collection in the Department of Zoology, Queen’s University, Belfast, Northern Ireland.
Illustrated topotype specimens of Bolivinoides decorata australis Edgell are located in the Common-
wealth Paleontological Collection, Bureau of Mineral Resources, Canberra, Australia. All other
holotypes, paratypes and hypotypes illustrated in this paper are deposited in the British Museum
(Natural History), London. Additional paratypes of B. hiltermanni, B. praelaevigata, and B. side-
strandensis will be deposited in the foraminiferal collections of the U.S. National Museum, Washington,
D.C.

Acknowledgements. 1 would like to thank Dr. Tom Barnard for his help in many aspects of this study;
Drs. H. Hiltermann and Z. Reiss for their helpful suggestions; Miss Anne Cameron, for the loan of
material from the Joseph Wright Collection; Dr. N. H. Fisher, for sending type material from Edgell’s
Collection; Dr. R. E. H. Reid, who furnished material from the Ballycastle Pellet Chalk and gave
helpful information concerning Irish stratigraphy and the location of old foraminiferal collections in
Northern Ireland; Dr. A. McGugan, who sent material from Northern Ireland; Mr. S. V. Bell for help
in collecting the Norwich and Sidestrand localities; Dr. W. A. Berggren, who drew my attention to
several important Russian studies; and also my wife Melza, for preparing the illustrations of fora-
minifera used in this paper.

STRATIGRAPHY

The Upper Cretaceous Chalk forms many prominent outcrops in southern England
and the excellently exposed sea cliff sections often represent continuous deposition from
Cenomanian to Upper Campanian times. Chalk of Maestrichtian age is present only
in a few small outcrops along the Norfolk coast (i.e. Trimingham, Sidestrand). Over
much of southern England, Upper Campanian strata have been truncated by a Lower
Tertiary unconformity. Samples used in the present study were collected from various
localities in southern England, Northern Ireland and Eire. Stratigraphic sections were
measured and collected at Culver Cliff, on the east coast of the Isle of Wight (see text-figs.
2, 3), and at Alum Bay, on the opposite side of the island. The stratigraphy of the Culver
Cliff section has already been described (Barr 1962). Most of the sample localities had
been previously dated by megafossils and placed in the classical zonation that has been
used in England and parts of western continental Europe for over half a century (see
text-fig. 1).

One of the earliest successful attempts to zone the Upper Cretaceous rocks of the
British Isles was made by Barrois (1876). He completed a comprehensive study of the
Upper Cretaceous rocks of England and Ireland and showed that it was possible to
recognize the same megafossil zones in the British Isles as were established in the
Upper Cretaceous of France, and that these zones could be grouped into the Ceno-
manian, Turonian and Senonian Stages proposed by d’Orbigny (1842, 1852). This
zonation has since been modified and refined by Jukes-Browne (1904, 1912), Rowe
(1908), Brydone (1914), Jeletzsky (1951), and others, but still forms the basis for the
present megafossil zonation of the British Upper Cretaceous.

Maestrichtian Stage. The Maestrichtian Stage was originally proposed by Dumont
(1849). Its type is section located in the E.N.C.I. Quarry at St. Pietersburg on the

222

PALAEONTOLOGY, VOLUME 9

outskirts of Maastricht, Netherlands. In recent years the literature on the type section
and correlation of the Maestrichtian Stage has been voluminous and sometimes contra-
dictory. However, it is now generally accepted that the Maestrichtian Stage should be

text-fig. 1. Chart showing megafossil zonation of the Upper Cretaceous strata

of the British Isles.

regarded as the uppermost Cretaceous stage, underlying the Danian Stage of Desor
(1846), which is now regarded as lowermost Tertiary, and overlying either the Senonian
Stage of d’Orbigny (1842, 1852) or the Campanian Stage of Coquand (1857). The
Maestrichtian Stage can be subdivided on the basis of belemnites into four or more
zones which can be recognized in its type area and over much of western Europe.

F. T. BARR: FORAMINIFERAL GENUS BOLIVINOIDES

223

The only chalk of Maestrichtian age found in England is in a few small outcrops
exposed along the Norfolk coast, including some large erratic blocks. Jukes-Browne
(1904) and Brydone (1906, 1908) published a series of papers on these exposures in
which they recognized a different fauna from that found in the Norwich Chalk or the
youngest chalk found along the south coast of England. They concluded that these
outcrops represented the youngest known Upper Cretaceous strata found in England.
Jukes-Browne (1904) proposed the Zone of Ostrea lunata to include the chalk at Trim-
ingham. Jeletzsky (1951) made a study of the belemnites from the Trimingham and

text-fig. 2. Location map of Culver Cliff on the east-central coast of the Isle of Wight
showing position of samples. Zones after Rowe (1908).

Norwich Chalk and concluded that the Trimingham Chalk was Lower Maestrichtian
in age and should be assigned to the Belemnitella lanceolata (sa.) and Belemnitella
lanceolata swnensis Zones (which are also recognized in the type area of the Maestrich-
tian Stage), whereas the Norwich Chalk contained Belemnitella mucronata and was
Upper Campanian. Jukes-Browne’s Zone of Ostrea lunata was found to be equivalent
to the combined zones of B. lanceolata (s.s.) and B. lanceolata sumensis.

Campanian Stage. Coquand (1857, 1858), in a study of the Cretaceous stratigraphy
of the northern Aquitaine Basin, proposed the erection of three stages, the Campanian,
Santonian and Coniacian, which are equivalent to the Senonian Stage of d’Orbigny.
The type locality of the Campanian Stage is in south-west France near Aubeterre,
while the type locality of the Senonian Stage is in the Paris Basin (d’Orbigny 1842,
1852). Through the years the Campanian, Santonian, and Coniacian have been regarded
as substages of the Senonian; however, over the last one or two decades, there has
been a tendency to use Coquand’s units as stages and to drop the broader term Senonian.
Unfortunately, the type sections of these stages are not well known and the exact posi-
tion of the boundaries between these stages is not always certain. This is partially

224

PALAEONTOLOGY, VOLUME 9

responsible for Coquand’s stages not being accepted in the Gulf Coast area of the
U.S.A., where the local stage names Navarro, Taylor, and Austin are in use.

The Campanian Stage is subdivided into two belemnite zones, the Belemnitella mucro-
nata Zone and the Actinoeamcix quadratus Zone, which are regarded as equivalent to the

if)

y

oc

LU

O

<

LU §_
= *0

(O

CO

LU

2

LITHOLOGY

LU CO
_J cc

£ 5

INSCRIPTION

LU

if)

h-

C/)

N£-

3

O

X

I-

< 3
CO 2

1 r

o

c

<3»

CO

k_

o

9

-C 10

z

3

£

—

-C II

Chalk, white
thick-bedded;

165'

r

3

o

100

-

&>

12

common thin

- i -

— C 13

dark grey flint

O

C.

<a>

nodule layers,

—

— C 17

spaced every

bJ

<D

CD

— C 18
— C 19

1 to 5 feet;
rare tabular

Z

flint beds*
rare marl seams.

O

if)

-

E»

<£>

— C 21

3

200

— C22

<

i i

— C23
-C24

<

TD

O

<£>

7

Q_

3

cr

1-

X73

<9

_

9

E

: l

— C26

LU

2

o o
o —

O h-

.

300

— C27

rr

<

E c
o ®

< m
©

in

9

-

10

\ r

— C28

O

O

9

— C29

0.1

9

«3»

400

t r

?

01

O

i r

U

z

o

9

<

<«»

m>

CL

—

500

i

z

m

•

Q_

o

© k-
«*- O

Chalk; nodular,
almost flintless;

3

h-

2

Cl

3 no
® 3

135'

common marl
seams.

<

° «

<3*

•

in

25 <D

600

*

-

&

text-fig. 3. Columnar section of upper part of the Upper Chalk at Culver Cliff,
Isle of Wight, showing stratigraphic position of samples.

Upper and Lower Campanian respectively. Jeletzsky (1951) presented cogent reasons
for placing the Maestrichtian-Campanian boundary at the top of the B. mucronata
Zone rather than within this zone as has sometimes been done (Barnard and Banner
1953, p. 207; Curry 1962, p. 181).

Santonian Stage. The Santonian Stage was erected by Coquand (1857, 1858), with the
type locality near the town of Saintes in the northern Aquitaine Basin. The limits of this

F. T. BARR: FORAMINIFERAL GENUS BOLIVINOIDES 225

stage have not been well defined, and consequently there is considerable disagreement
as to what constitutes the Lower Santonian in north-western Europe and the British
Isles. In Britain, the Santonian-Coniacian contact is usually placed at the base of the
Marsupites testudinarius Zone (Barnard and Banner 1953; Barnard 1963; Barr 1962;
Curry 1962). On the other hand, many continental stratigraphers (e.g. Gignoux 1955,
p. 392) include the Micr aster cor-anguinum Zone in the Lower Santonian and place the
Santonian-Coniacian contact at the base of the M. cor-anguinum Zone. This difference
can only be resolved when the type sections of the Santonian and Coniacian Stages
become better known and when exact correlations are possible between the Anglo-Paris
Basins and the northern Aquitaine Basin. One must keep this difference in mind when
comparing the stratigraphical ranges of fossils recorded from various parts of Europe.

PHYLOGENY

A number of papers have dealt with the phylogeny of certain species or lineages of
Bolivinoides. The works of Hiltermann and Koch (1950), Edgell (1954), Reiss (1954),
Hofker (1957, 1958u, 19586), Vassilenko (1961), and Hiltermann (1963) have shown
evolutionary development in this group to be similar in various parts of the world.
Consequently, the recognition of these evolutionary sequences within the genus
Bolivinoides can be extremely useful in the correlation and dating of Upper Cretaceous

STAGE

C A MPANIAN

UPPER

LOW ER

ZONE

B. mucrona t a

A. quadrotus

"-^SAMPLE

SPECIES''-^^^

o

o

o

OJ

O

ro

o

r-

o

CD

o

cx>

o

OJ

o

OJ

04

o

ro

OJ

o

sj-

CVJ

O

C26

b-

cvl

(_)

CO

CXI

(_>

a>

OJ

o

B . laevigata

o

o

B . decorate

o

B. praelaevigata

o

0

©

B. pustulata

o

o

o

O

8. hiltermanni

m °

O

I COMMON ® FEW O RARE

text-fig. 4. Chart showing distribution of Bolivinoides
in the Campanian chalk at Culver Cliff, Isle of Wight.

strata. However, there is not complete agreement as to the phylogeny of all Bolivinoides
lineages and it has been suggested (Hiltermann 1963) that subspecific differences may
exist within some evolutionary lineages in different geographic provinces. The Upper
Cretaceous Chalk of the British Isles provides an excellent opportunity to observe the
development of some lineages of Bolivinoides and it is hoped that these observations will
help clarify some of the phylogenetic relationships of this group.

Bolivinoides strigillata (Chapman) appears to be restricted to the Marsupites testudi-
narius Zone (Lower or Middle Santonian) and possibly the basal part of the Offaster
pilula Zone. It is the oldest species of Bolivinoides recognized in the British Isles or
elsewhere in the world, and from this species evolved the various lineages that are so
common higher in the Upper Cretaceous strata. B. strigillata (va.) is usually quite rare,
but the many forms which rapidly evolved from this ancestral species became increas-
ingly abundant and diversified later in the Upper Cretaceous. The genus Bolivinoides

226

PALAEONTOLOGY, VOLUME 9

reached its acme in the Maestrichtian and survived into the Danian, but with progres-
sively fewer individuals.

SANTONIAN

CAMPANIAN

MAESTRICHTIAN

LOWER!?) or MIDI?)

UPPER

LOWER

UPPER

LOWER

UPPER

Marsupites
testudl narius

Of faster
pilula

Actinocamax

quadratus

Belemnitella

mucronata

Belemnitella
lanceolata ( si)

B .

junior

B. casi -
mirovensls

B draco

B strigillata

B hitter manni

\ B . a us trot is

B pustutata

8 gig ant ea

B. praetoevigata

B laevigata

B peter ssoni

text-fig. 5. Diagram showing the stratigraphic ranges and suggested phylogeny of various

species of Bolivinoides.

The most common lineage of Bolivinoides observed in the British Isles commences
with B. strigillata and evolves through B. hiltermanni sp. nov., B. deeorata (Jones) to
B. miliaris Hiltermann and Koch. The end member of this lineage, B. draco (Marsson),
was not found, as the Upper Maestrichtian is not present in Britain, nor was the
B. australis-B. gigantea offshoot observed. The B. strigillata-B. gigantea lineage shows
several progressive changes in morphology: an increase in overall size; smaller length/
breadth ratios; and an increase in the number of ‘ornamental’ lobes on the final cham-
bers. The evolution of B. deeorata to B. miliaris to B. draco produces a progressively
more rhomboidal or kite-shaped test and the development of medial ribs. B. deeorata
has often been regarded as evolving directly from B. strigillata (Hiltermann and Koch
1950; Vassilenko 1961; Hiltermann 1963). However, in the well-developed Lower
Campanian section in southern England a transitional species, B. hiltermanni sp. nov.,
is found in considerable abundance. This species is intermediate in size and in strength
of ornamentation between B. strigillata and B. deeorata.

There is also an indication that B. hiltermanni is probably the ancestral species to the
B. pustulata Reiss-5, praelaevigata sp. nov. -5. laevigata Marie-5, peterssoni Brotzen

F. T. BARR: FORAMINIFERAL GENUS BOLIVINOIDES

227

lineage. B. hiltermanni has a thicker, less compressed test than species of the B.pustulata -
B. peter ssoni lineage; however, variants of this species found in the upper part of
the Actinocamax quadratus Zone may represent an intermediate form between
B. hiltermanni (5.5.) and B. pustulata, although the complete transition of forms was
not observed. The B. pnstnlata-praelaevigata-laevigata-peterssoni lineage is charac-
terized by a small compressed test with rather weakly developed ornamentation chiefly
restricted to the medial part of the test.

SYSTEMATIC PALAEONTOLOGY

Order foraminiferida
Superfamily buliminacea Jones 1875
Family bolivinitidae Cushman 1927
Genus bolivinoides Cushman 1927

Type Species. Bolivinoides draco (Marsson).

Description. Test completely biserial, gradually tapering to flaring, with greatest breadth
usually near apertural end, and circular or elliptical in cross section. Proloculus globular,
smooth. Chambers inflated, curved, uniformly and fairly rapidly increasing in size with
greatest height and thickness near longitudinal axis. Sutures often obscure, slightly
curved. Aperture a slit-like or loop-shaped opening near centre of apertural face, often
bordered by weakly developed lip and with internal tooth plate structure. Ornamentation
usually strongly developed, consisting of lobes and knobs running perpendicular to
sutures. Internal wall surface also contains semi-conical bulges. Wall calcareous, finely
perforate.

Remarks. Numerous specimens of Bolivinoides were serially sectioned and their internal
structure examined. The present study confirms many of Montanaro Gallitelli’s (1957,
pp. 145, 146) observations. She examined the internal structure of acid treated specimens
of Bolivinoides draco (Marsson) and observed that the internal surface of the test had
a tuberculate sculpture. The internal surface of the tests of all species sectioned in the
present study contained semi-conical bulges along the base of the chambers. In species
such as B. draco with strongly developed, but fine external ornamentation, the internal
circular bulges occupy most of the chamber walls in the earliest chambers and occupy
somewhat less in the final chambers. In most of the other species, however, the internal
bulges are restricted to about the lower quarter of the chamber surface. The maximum
breadth of the bulge is along the lower suture. These internal knobs correspond to the
depressed areas between lobes on the external surface. In fact, almost all of the external
‘ornamental’ lobes and knobs are reflected interiorly, producing a partially corrugated
test wall (see text-figs. 7, 8). These observations do not agree exactly with those of Rey-
ment (1959, p. 13) who stated that the tuberculated wall structure seen by Montanaro
Gallitelli was derived from cameral overlaps. The external sculpture, and concomitantly
the internal sculpture, of Bolivinoides therefore appears to represent structure more
fundamental than the more usual surface ornamentation of many genera, which is some-
times quite superficial.

An internal bolivine tooth plate structure was observed in Bolivinoides , similar to
that described by Hofker (1957) and Reyment (1959).

228

PALAEONTOLOGY, VOLUME 9

Bolivinoides strigillata (Chapman)

Plate 34, figs. 8, 9; Plate 37, figs. 7-9

Bolivina strigillata Chapman 1892, p. 515, pi. 15, fig. 10.

Bolivinoides strigillata (Chapman); Hiltermann and Koch 1950, pp. 614-23, text-figs. 2 (1-19),
3 (1-9), 5 (10).

Bolivinoides strigillata (Chapman); Edgell 1954, pp. 70, 71, pi. 13, fig. 8; pi. 14, fig. 9.

Bolivinoides strigillatus (Chapman); Bykova 1959, table 1, fig. 1.

Bolivinoides strigillata (Chapman); Vassilenko 1961, pp. 186-8, pi. 39, figs. la-c.

Bolivinoides strigillata (Chapman); Hiltermann 1963, pp. 209, 210, pi. 1, figs. 12-14, 16, 18, 19,
24.

Bolivinoides strigillata (Chapman); Barr and Cordey 1964, pp. 308, 309, pi. 49, figs. 1-3.

Description. Test biserial, elongate, sides gradually tapering, subelliptical to oval in
cross section; proloculus globular, followed by 6 to 8 pairs of indistinct, slightly inflated
chambers uniformly and gradually increasing in size; sutures slightly oblique, indistinct,
obscured by ornamentation; 2 or 3 weakly raised lobes on final chambers running per-
pendicular to sutures, ornamentation on earliest chambers less distinct, sometimes
surface of early part of test smooth; aperture a slit-like opening between lobes on final
chamber; wall calcareous, finely perforate; length/breadth ratio 1-7 to 2-2; breadth/
thickness ratio 1-4 to 1-9; approximate average length of adult specimen 0-40 mm.

Remarks. Bolivinoides strigillata is the most primitive species known of the genus
Bolivinoides, and the only species found in strata as old as the Marsupites testudinarius
Zone in the British Isles. It is one of the smallest species of Bolivinoides and appears
to be the ancestral species of several of the most common and stratigraphically im-
portant lineages of this genus (see text-fig. 5).

Chapman (1892, p. 515) originally described B. strigillata from the phosphatic chalk
at Taplow, Buckinghamshire. Barr and Cordey (1964, p. 308) collected additional
material from Taplow and re-examined Chapman’s original specimens, which are
deposited in the British Museum (Natural History). They redescribed B. strigillata and
proposed a lectotype for this species from syntypic specimens in the Chapman Collec-
tion (BM P44968).

A single specimen referred to Bolivina strigillata Chapman by Williams-Mitchell

EXPLANATION OF PLATE 34

All illustrations unretouched photographs.

Fig. 1. Bolivinoides australis Edgell. X 1 18. Topotype. C. Y. Creek, near Cardabia, on the west flank of
Giralia Anticline, north-west Australia; Edgell’s (1954) locality GC/304.

Figs. 2-6, 12. Bolivinoides decorata (Jones). X 1 18. Showing variation in test shape and ornamentation.
Upper Campanian, Belemnitella mucronata Zone. 2, Meudon, Paris Basin; BM P45711. 3-5, 12,
Alum Bay, Isle of Wight, Sample A2, 405 ft. above base of B. mucronata Zone. 3, BM P45712; 4,
BM P45713; 5, BM P45714; 12, BM P45716. 6, Culver Cliff, Isle of Wight, Sample CIO; BM P45715.

Fig. 7. Bolivinoides laevigata Marie. X 118. Upper Campanian, Belemnitella mucronata Zone, Culver
Cliff, Isle of Wight; Sample CIO, BM P45740.

Figs. 8, 9. Bolivinoides strigillata (Chapman). X 140. Topotypes. Santonian, Marsupites testudinarius
Zone, Phosphatic Chalk, Taplow, Buckinghamshire. 8, BM P45730; 9, BM P45731.

Figs. 10, 11. Bolivinoides sidestrandensis sp. nov. X 118. Lower Maestrichtian, Belemnitella lanceolata
Zone, Sidestrand, Norfolk. 10, Paratype, BM P45745; 11, Holotype, BM P45743.

*x

Palaeontology, Vol. 9

PLATE 34

BARR, Upper Cretaceous Bolivinoides

F. T. BARR: FORAMINIFERAL GENUS BOL1VINOIDES

229

(1948, p. 106, pi. 9, fig. 3) is located in the British Museum (Natural History). Barr and
Cordey (1964, p. 308) examined the Williams-Mitchell collection and found this speci-
men to be clearly conspecific with Bolivinoides decorata (Jones). Williams-Mitchell
(op. cit., pi. 10) lists the stratigraphical range of ‘ Bolivina strigdlata ’ (the only species
in his paper referable to the genus Bolivinoides ) as Marsupites testudinarius Zone to the
top of the Belemnitella mucronata Zone in the Upper Cretaceous strata of England.
This range undoubtedly represents the combined stratigraphical occurrence of several
species of Bolivinoides.

Occurrence. The type locality of Bolivinoides strigdlata is a small chalk pit, Lodge Pit,
on the outskirts of Taplow, Buckinghamshire. Chapman obtained his specimens from
about the middle of the phosphatic band in this pit. Osborne White and Treacher
(1905) placed this part of the section in the M. testudinarius Zone (Santonian). B.
strigdlata appears to be restricted to the M. testudinarius Zone in England, but may
possibly occur also in the Offaster pilula Zone.

Edgell (1954, pp. 70, 71) recorded B. strigdlata from the Santonian of north-western
Australia. Vassilenko (1961, p. 188, 297) reported this species from the Santonian and
basal Campanian of the Soviet Union.

Hiltermann (1963, p. 205) and Hiltermann and Koch (1960, p. 71) listed the strati-
graphic range of B. strigdlata in north-western Europe as Upper Santonian to Lower
Campanian. This range does not agree exactly with the range observed by me for this
species in England. The apparent difference in range for this species between Germany
and England may be due to the following reasons:

1. The M. testudinarius Zone is usually considered by British stratigraphers as Lower
Santonian (Barnard and Banner 1953; Curry 1962), whereas many continental strati-
graphers (Gignoux 1955, p. 392) place this zone in the Middle or Upper Santonian and
regard the underlying Micraster cor-anguinum Zone as the Lower Santonian.

2. Many specimens which were once grouped with B. strigdlata , or considered to be
transition forms between B. strigdlata and B. decorata , are now placed in the newly
erected species B. hiltermanni. Probably most of the specimens referred to B. strigdlata
from the Lower Campanian ( Actinocaniax cpiadratus Zone) should now be considered
as B. hiltermanni.

Bolivinoides hiltermanni sp. nov.

Plate 36, figs. 7, 8; Plate 37, figs. 1-3

Bolivinoides decorata (Jones) cf. delicatula Cushman; Edgell 1954, p. 71, pi. 13, fig. 7; pi. 14,
fig. 7.

Bolivinoides sp. [transition form between B. strigdlata (Chapman) and B. decorata (Jones)];
Hiltermann 1963, p. 209, pi. 1, figs. 2, 3, 7-9.

Description. Test completely biserial, elongate, tapering, greatest breadth and thickness
near apertural end, initial end acutely rounded, apertural end more broadly rounded,
elliptical in cross section; proloculus globular, succeeded by 7 to 8 pairs of tapering
chambers uniformly and fairly rapidly increasing in size; sutures oblique, slightly
curved, partially obscured by ornamentation; aperture a slit-like or loop-shaped opening
located near centre of apertural face, sometimes bordered by weakly raised indistinct

230

PALAEONTOLOGY, VOLUME 9

lip, and with internal bolivine tooth plate structure; 2 to 4, usually 3 distinct lobes on
final 2 or 3 pairs of chambers running perpendicular to sutures, less distinct lobes and
knobs present on earlier chambers and usually surface of earliest 1 to 3 pairs of chambers

BREADTH / THICKNESS

text-fig. 6. Diagrams showing comparison of length/breadth and breadth/thickness ratios of 100
specimens of Bolivinoides decomta (Jones) and 100 specimens of B. hiltermanni sp. nov. All specimens
of B. deco rata are from sample CIO, Belemnitella mucronata Zone, Culver Cliff. All specimens of B.
hiltermanni are from sample C29, Actinocamax quadratus Zone, Culver Cliff.

smooth; internal surface of chamber walls contains semi-conical knobs along top of
suture, corresponding to depressed areas between lobes on external surface ; proloculus
smooth; wall very finely perforate; length/breadth ratio 1-7 to 2-5, most commonly

EXPLANATION OF PLATE 35

All illustrations shaded camera lucida drawings, a = side view; b = terminal view.

Figs. 1-3. Bolivinoides australis Edgell. x90. Topotypes. C. Y. Creek, near Cardabia, on the west
flank of Giralia Anticline, north-west Australia; Edgell's (1954) locality GC/304.

Figs. 4, 5. Bolivinoides miliaris Fliltermann and Koch, x 95. Ballycastle Pellet Chalk, west side Bally-
castle Harbour, Northern Ireland. 4, BM P45722; 5, BM P45723.

Figs. 6-9. Bolivinoides decorata (Jones). X 95. Upper Campanian, Belemnitella mucronata Zone.
6, 7, Culver Cliff, Isle of Wight, Sample CIO; BM P45718, P45719. 8, Leicester Lime Co. Quarry,
Drayton, Norfolk; BM P45717. 9, Alum Bay, Isle of Wight, 405 ft. above base of B. mucronata
Zone, Sample A2; BM P45720.

Palaeontology, Vol. 9

PLATE 35

BARR, Upper Cretaceous Bolivinoides

F. T. BARR: FORAMINIFERAL GENUS BOLIV1NOIDES

231

1-9 to 2-3; breadth/thickness ratio 1-7 to 2-2; approximate average length of adult
specimen 0-50 mm.

Dimensions of holotype (BM P45724). Length 0-55 mm.; maximum breadth 0-25 mm.; maximum
thickness 0-14 mm.

Remarks. B. hiltermanni represents the transition form between B. strigil/ata (Chapman)
and B. decorata (Jones). Nevertheless, B. hiltermanni is quite distinct from both of these
species; its lower boundary with B. strigil/ata and its upper boundary with B. decorata
are usually sharply defined, and the transition between species takes place over a rela-
tively short stratigraphic interval. B. hiltermanni is distinguished from B. strigil/ata
by its: (1) larger size; (2) more elliptical cross section; (3) more rapidly tapering test;
(4) more broadly rounded apertural end; and (5) usually having more ornamental lobes
on its final chambers. B. hiltermanni differs from B. decorata by its: (1) smaller size; (2)
less flaring test; (3) larger length/breadth ratio (usually 1-9 to 2-3 compared with 1-5 to
1-8 most usual for B. decorata, see text-fig. 6); (4) less distinct ornamentation, especially
on the early part of the test; and (5) fewer ornamental lobes on the final chambers
(most often 3 compared with the usual 4 or 5 for B. decorata).

B. hiltermanni is also closely related to B. pustulata Reiss, but is usually larger and
has a more elliptical, less compressed end view. Furthermore, the ornamentation of
B. pustulata is not as strongly developed and tends to be restricted to the medial part of
the test. In the upper part of the Actinocamax quadratus Zone (upper part of the Lower
Campanian) variants of B. hiltermanni become quite similar to Reiss’s species. It is
suggested that B. hiltermanni gave rise to B. pustulata during the middle part of the
A. quadratus Zone. B. hiltermanni continued until the base of the B. mucronata Zone, at
which time it gave rise to B. decorata , whereas B. pustulata gave rise to the B. prae-
laevigata-laevigata lineage near the beginning of the Upper Campanian.

B. hiltermanni is named in honour of Dr. Heinrich Hiltermann of Hannover, in
recognition of his work on the genus Bolivinoides.

Occurrence. B. hiltermanni occurs in abundance throughout the A . quadratus Zone and
more rarely in the O. pilula Zone in southern England. Although this species is found in
abundance in the uppermost A. quadratus Zone, it does not survive into the B. mucronata
Zone. The A. quadratus Zone-5, mucronata Zone (Lower Campanian-Upper Cam-
panian) contact in southern England can be located with some accuracy at the strati-
graphic position where B. hiltermanni is replaced by B. decorata ( s.s .).

Bolivinoides decorata (Jones)

Plate 34, figs. 2-6, 12; Plate 35, figs. 6-9; Plate 36, figs. 1-5

Bolivina decorata Jones (in Wright 1875), pp. 87, 96, 97 (list only).

Bolivina decorata Jones (in Wright 1886), p. 330, pi. 27, figs. 7, 8.

Bolivina decorata Jones; Heron-Alien and Earland 1910, pi. 7, figs. 1, 2.

Bolivina decorata Jones; Macfadyen 1932, pi. 35, fig. 20.

Bolivinoides decorata (Jones); Dain 1934, pp. 33, 34, pi. 3, fig. 34.

Bolivinoides decorata (Jones); Marie 1941, p. 188, pi. 29, fig. 279.

Bolivina strigil/ata Chapman (partim); Williams-Mitchell 1948, p. 106, pi. 9, figs. 3 a, b.

Bolivinoides decorata decorata (Jones); Hiltermann and Koch 1950, pp. 606-10, text-figs. 2-4,
nos. 17-25, 27-31, 35-38, 42^45.

C 3803

R

232

PALAEONTOLOGY, VOLUME 9

BoJivinoides decorata decorata (Jones); Hagn 1953, p. 74, pi. 6, fig. 22.

Bolivinoides decorata decorata (Jones); Reiss 1954, p. 155, pi. 28, figs. 5-8.

Bolivinoides decorata decorata (Jones); Hiltermann and Koch 1955, p. 365, pi. 28, fig. 4.
Bolivinoides decoratus decoratus (Jones); McGugan 1957, p. 339, pi. 32, figs. 10-13, 15 ( non
fig. 14).

Bolivinoides decorata decorata (Jones); Bieda 1958, pp. 32, 33, text-figs. 6 o-c.

Bolivinoides regidaris Reiss; Bieda 1958, pp. 36-38, text-figs. 9 a-c.

Bolivinoides decorata decorata (Jones); Witwicka 1958, pp. 198, 199, pi. 9, figs. 9a, b.
Bolivinoides decoratus decorata (Jones); Bykova 1959, table 1, fig. 2.

Bolivinoides decoratus (Jones); Akimets 1961, pp. 188, 189, pi. 18, figs. 17, 18.

Bolivinoides decoratus decorata (Jones); Vassilenko 1961, pp. 189-90, pi. 39, figs. 8-10.
Bolivinoides decoratus decoratus (Jones); Hiltermann and Koch 1962, p. 315, table 19, pi. 46,
fig. 7.

Bolivinoides decoratus (Jones); Kaptarenko-Chernousova et al. 1963, p. Ill, pi. 25, fig. 4.

Description. Test biserial, pear-shaped in outline, elliptical in cross section; proloculus
globular, succeeded by 8 to 9 pairs of slightly inflated, tapering chambers uniformly and
rapidly increasing in size; sutures oblique, slightly curved, mostly obscured by orna-
mentation; aperture a slit-like or loop-shaped opening located near centre of apertural
face, sometimes bordered by slightly raised lip and with internal tooth plate structure;
3 to 6 thick distinct lobes on final chamber running perpendicular to sutures, lobes on
inner portion of chambers slightly thicker and more distinct than outer lobes, sometimes
giving suggestion of 2 weakly defined medial ribs, ornamentation on earliest 1 to 3
pairs of chambers less distinct consisting of weakly raised lobes or knobs, often surface
of earliest chambers smooth; proloculus smooth; wall calcareous, finely perforate;
length/breadth ratio 1-4 to 2-2, most commonly T5 to T8; breadth/thickness ratio T4
to 2-0, most commonly 1-6 to T9; approximate average length of adult specimen
0-58 mm.

Dimensions of lectotype. Length 053 mm.; maximum breadth 0-34 mm. ; maximum thickness 01 7 mm.

Remarks. In a short paper on some Upper Cretaceous Foraminifera collected from
the chalk at Keady Hill, County Derry, Northern Ireland, Wright (1886, p. 330, pi. 27,
figs. 7, 8) published the original description and illustrations of Bolivina decorata. How-
ever, this name had been used previously by T. Rupert Jones in unpublished work, and
therefore Wright credited Jones with authorship of Bolivina decorata, even though this
species was originally described in Wright’s paper from material collected by Wright.

In Wright’s collection of Foraminifera located in the museum of the Zoology Depart-
ment, The Queen’s University, Belfast, Northern Ireland, there is a small circular pillbox.

EXPLANATION OF PLATE 36

All illustrations shaded camera lucida drawings, a = side view; b = terminal view.

Figs. 1-3, 5. Bolivinoides decorata (Jones). X 100. Syntypic specimens from Joseph Wright Collection.
1, Lectotype. 3, 5, Specimens with badly worn surface. Upper Campanian, Belemnitella mucronata
Zone, Keady Hill, County Derry, Northern Ireland.

Fig. 4. Bolivinoides decorata (Jones). X 90. Upper Campanian, B. mucronata Zone, Culver Cliff, Isle
of Wight, Sample CIO; BM P45721.

Figs. 6, 9. Bolivinoides sidestrandensis sp. nov. X 95. Paratypes. Lower Maestrichtian, Belemnitella
lanceolata Zone, Sidestrand, Norfolk. 6, BM P45744; 9, BM P45746.

Figs. 7, 8. Bolivinoides hiltermanni sp. nov. X 95. Paratypes. Lower Campanian, Actinocamax quadratus
Zone, Culver Cliff, Isle of Wight. 7, Sample C28, BM P45727; 8, Sample C29, BM P45726.

Palaeontology , Vol. 9

PLATE 36

BARR, Upper Cretaceous Bolivinoides

F. T. BARR: FORAMINIFERAL GENUS BOLIVINOIDES 233

with celluloid cover, containing nine specimens of Bolivinoides, and is labelled as follows:
Bolivina decorata, Chalk powder, Keady Hill, Co. Derry 109.

No other specimens of Bolivinoides decorata have been found which might be regarded
as primary type material for this species. These specimens from the Wright Collection
are probably those originally described as Bolivina decorata Jones and are here regarded
as syntypes. Six of these specimens are broken or badly damaged and the remaining
three specimens are worn to varying degrees. Nevertheless, several of these specimens
are well enough preserved to show the morphology necessary in understanding the
species. A single specimen which appears to be very similar to Wright’s original illustra-
tion has been isolated from the syntypic series and is designated
the lectotype of Bolivina decorata (here figured PI. 36, figs. 1 a, b ).

Three other specimens from this syntypic series are also illustrated
(PI. 36, figs. 2, 3, 5).

Edgell (1954, pp. 71, 72) described a new subspecies of
Bolivinoides from Australia which he named B. decorata (Jones)
australis. It is distinguished from B. decorata ( s.s .) by: (1) having
a smaller length/breadth ratio; (2) having more ornamental lobes
on the final chambers (5-7 on the final chamber compared with
3-6 for B. decorata s.s.); (3) possibly being more commonly orna-
mented near its initial end; and (4) sometimes having a more
rhomboidal outline.

Ten topotypic specimens of B. decorata australis from C. Y.

Creek, west flank of Giralia Anticline, north-west Australia,
have been kindly loaned to me for examination by N. H. Fischer,

Bureau of Mineral Resources, Canberra, Australia. Four of these
specimens are here illustrated for comparison with B. decorata
s.s. (PI. 34, fig. 1 ; PI. 35, figs. 1-3). These specimens show B. decorata australis to be
quite distinct from B. decorata s.s. and to be closer morphologically to B. gigantea
Hiltermann and Koch. Possibly B. australis might best be considered a subspecies of
B. gigantea or even synonymous with this species. Nevertheless, although it may be
difficult to consistently distinguish B. decorata australis from B. gigantea, it appears to
represent a transitional form between B. decorata and B. gigantea.

The evolution of B. decorata to B. australis takes place in Western Europe and North
Africa in the earliest Maestrichtian(at the beginning of the Belenmitella lanceolata Zone).
Specimens of B. decorata become increasingly similar to B. australis in latest Campanian
times. The type specimens of B. decorata from Keady Hill are probably from the upper
part of the B. mucronata Zone and, consequently, two or three of the specimens of
B. decorata from the Wright Collection represent rather advanced forms of this
species.

Specimens referred to Bolivina decorata Jones by Heron-Alien and Earland (1910,
p. 409, pi. 7, figs. 1, 2) are located in the British Museum (Natural History). These
specimens have been re-examined and were found to be clearly conspecific with the
type for Bolivinoides decorata. These specimens are from chalk meal from the interior
of a cavernous flint nodule found on the beach at Selsey Bill, Sussex. The presence of
B. decorata in the fauna from this flint nodule would strongly suggest that the nodule
was originally from the B. mucronata Zone.

text-fig. 7. Terminal
view of small specimen
of Bolivinoides decorata
(Jones) with final cham-
ber mostly broken away
exposing corrugated
nature of lower part of
chamber wall. This view
shows that the external
‘ornamental’ lobes are
formed by folds in the
chamber wall rather
than by a thickening of
the test wall.

234

PALAEONTOLOGY, VOLUME 9

Occurrence. B. decorata occurs in abundance throughout the B. mucronata Zone in
southern England and appears to be restricted to this zone. The type locality for B.
decorata is at Keady Hill, County Derry, Northern Ireland, which is probably in the
upper part of the B. mucronata Zone. Wright (1875, pp. 96, 97) also recorded this species
from a number of other Upper Cretaceous localities in Northern Ireland.

McGugan (1957, p. 339, pi. 32, figs. 10-15) recorded B. decorata from the Ballycastle
Pellet Chalk of Northern Ireland. R. E. H. Reid, Queen’s University, Belfast, has kindly
sent me material from the Ballycastle Pellet Chalk from the main exposure on the
west side of Ballycastle Harbour. This material contained a number of specimens of
B. decorata along with rarer specimens of B. sidestrandensis sp. nov. and B. miliaris
Hiltermann and Koch. All specimens of Bolivinoides from this material were worn to
varying degrees. The benthonic fauna contained a mixture of Campanian and Maestrich-
tian species and represents largely, if not entirely, a derived fauna. McGugan (op. cit.,
p. 33) stated that the character of the exposures of the Ballycastle Pellet Chalk suggested
deposition of this formation in some kind of channel in the top of the underlying Upper
Cretaceous Chalk. R. E. H. Reid informs me that the principal exposure of the Bally-
castle Pellet Chalk contains blocks of Tertiary basalt. It appears that the Ballycastle
Pellet Chalk was derived by Tertiary erosion of an Upper Cretaceous surface (cf.
Charlesworth 1963, p. 369) and the specimens of Bolivinoides found in this formation
have also been derived from this nearby Upper Cretaceous Chalk.

I have found B. decorata in the newly discovered Ballydeenlea Chalk of County
Kerry, Ireland (Walsh 1960). The presence of this species along with the rest of the fauna
indicates that the Ballydeenlea Chalk is Campanian in age, not Middle Chalk (Turonian),
as had previously been suggested (Walsh 1960, p. 113). A description of the foraminiferal
fauna from this formation will soon be published.

Marie (1941, p. 188) recorded B. decorata from the B. mucronata Zone of the Paris
Basin. I have also found abundant specimens of B. decorata in the upper B. mucronata
Zone at Meudon, on the outskirts of Paris. These specimens are 10 to 15 per cent larger
than the average from England, but are very similar in all other characteristics.

I have found specimens of B. decorata in an advanced stage of development in material
from R. K. Olsson’s (1964, p. 160) locality NJK 134 in the Marshalltown Formation
of New Jersey. Olsson (op. cit., p. 163) recorded Globotruncana calcarata Cushman,
usually regarded as a reliable index fossil to the Upper Campanian, from this same
locality.

Bolivinoides miliaris Hiltermann and Koch
Plate 35, figs. 4, 5

Bolivinoides draco (Marsson) miliaris Hiltermann and Koch 1950, pp. 604-6, text-figs. 2-4,
nos. 32-34 (?), 39-41, 46-48; text-fig. 5, nos. 39 a-c.

Bolivinoides draco miliaris Hiltermann and Koch; Reiss 1954, p. 155, pi. 28, figs. 9-12, 14.

Bolivinoides draco miliaris Hiltermann and Koch; Pozaryska 1954, p. 254, text-fig. 4.

Bolivinoides miliaris Hiltermann and Koch; Hofker 1957, p. 267, text-fig. 3226.

Bolivinoides draco miliaris Hiltermann and Koch; Bieda 1958, pp. 44, 45, figs. 14 a-c.

Bolivinoides draco miliaris Hiltermann and Koch; Hiltermann and Koch 1960, p. 72.

Bolivinoides miliaris Hiltermann and Koch; Vassilenko 1961, pp. 200, 201, pi. 40, figs. 4 a-c,
pi. 41 , figs. 1 a-c.

Bolivinoides draco miliaris Hiltermann and Koch; Hiltermann and Koch 1962, p. 317, table 19,
pi. 46, fig. 9.

F. T. BARR: FORA MINI FERAL GENUS BOLIVINOIDES

235

Bolivinoides draco miliaris Hiltermann and Koch; Hiltermann 1963, p. 222, pi. 4, figs. 21-23.

Bolivinoides draco miliaris Hiltermann and Koch; van Hinte 1963, pp. 106, 107, pi. 13, figs. 7, 8.

Remarks. B. miliaris is morphologically similar to B. decorata (Jones) in many respects
and appears to have been derived from the latter during the late B. mucronata Zone
(uppermost Campanian). B. miliaris can be distinguished from B. decorata by having
a smaller length/breadth ratio. Its lateral outline is more rhomboidal than the more
elongate, pear-shaped outline of B. decorata ; the maximum breadth of B. miliaris is
nearer the mid point of test, whereas the maximum breadth of B. decorata is closer to
the apertural end. The ornamentation on B. miliaris is usually less regular than that of
B. decorata and often, on the earliest 30 to 50 per cent of the test,
the ornamentation consists of small knobs. The earliest 1 to 3 pairs
of chambers of B. decorata are ornamented by less distinct lobes or
knobs, or more often, these early chambers have a smooth surface.

B. miliaris differs from B. australis Edgell by having fewer orna-
mental lobes on its final chambers. B. miliaris usually has 4 and
sometimes 5, whereas B. australis has 5 to 7 on its latest chambers. B.
miliaris has a smaller size and usually has a more rhomboidal outline,
although variants of B. australis sometimes possess a similar rhom-
boidal outline.

B. miliaris gave rise to B. draco ( Marsson) during the upper part of
the Lower Maestrichtian. The earliest 50 to 60 per cent of the test
of B. draco is more flaring than that of B. miliaris , which has a less
angular outline. B. draco also has two very distinctive, well-developed,
parallel medial ribs, which are produced by the fusing of the inner
‘ornamental’ lobes. Hiltermann and Koch (1960, p. 73) listed the
stratigraphic range of B. draco in north-west Europe as the upper
part of the Lower Maestrichtian to the top of the Maestrichtian.

I have observed the same range for this species in western Europe
and North Africa; however, I have not found B. draco in the British
Isles. McGugan (1957, pi. 32, fig. 17) illustrated a specimen from the Ballycastle Pellet
Chalk of Northern Ireland which he referred to B. draco draco Hiltermann and Koch.
It is poorly preserved and there is some doubt as to whether this specimen should be
considered B. draco (s.s). Curry (1962, p. 180) also recorded a single specimen which
he referred to Bolivinoides cf. draco from a submarine core from the English Channel.

Occurrence. Rare specimens of B. miliaris were found in the Ballycastle Pellet Chalk
along the west side of Ballycastle Harbour, Northern Ireland. Curry (1962, p. 182, 184)
recorded B. miliaris from two submarine cores which were obtained from an area in
the English Channel halfway between the Isle of Wight and Cherbourg. The Upper
Cretaceous fauna of the Ballycastle Pellet Chalk, which consists of a mixture of late Cam-
panian and early Maestrichtian foraminifera, appears to be a derived fauna (see remarks
under B. decorata ) and therefore the age of these specimens is not precisely known.

The holotype of B. miliaris is from a bore hole near Niemburg, north-west Germany,
and is from the B. mucronata Zone. Hiltermann (1963, p. 205) and Hiltermann and Koch
(1960, p. 72) listed the range of B. miliaris in north-west Europe as uppermost Cam-
panian and Lower Maestrichtian.

text-fig. 8. Bisected
specimen of Bolivi-
noides draco (Mars-
son) showing tooth
plate structure and
internalsemi-conical
bulges along base
of chambers, corre-
sponding to depres-
sed areas between
lobes on external
surface.

236

PALAEONTOLOGY, VOLUME 9

Bolivinoides pustulata Reiss
Plate 37, figs. 4 a, b

Bolivinoides pustulata Reiss 1954, p. 156, pi. 29, figs. 9, 10.

Bolivinoides granulata Hofker 1957, pp. 250, 251, text-figs. 303, 310.

Bolivinoides pustulata Reiss; Hiltermann 1963, p. 214, pi. 2, fig. 17.

Bolivinoides granulata Hofker; Hiltermann 1963, p. 213, pi. 2, figs. 5, 11, 12.

Remarks. Topotypic specimens of B. pustulata have been compared with specimens
from southern England and they appear to be conspecific. B. pustulata is closely related
to B. hiltermanni (see p. 231) and it is suggested that the latter is the ancestral species.
It has a much greater stratigraphic range than B. pustulata , which is restricted to the
upper part of the A. quadratus Zone.

B. pustulata appears to be the direct ancestor of B. praelaevigata sp. nov., which is
distinguished from it by its smaller, more compressed, more gradually tapering test.
The ornamentation of B. praelaevigata is also more weakly developed and even more
restricted to the medial portion of the test than that of B. pustulata.

B. pustulata is morphologically very similar to B. granulata Hofker and the latter is
tentatively regarded as a junior synonym of B. pustulata. Unfortunately, however, the
original illustrations given by Hofker (1957, text-figs. 303-10) do not allow a very exact
comparison of all features and I have been unable to examine type material of B.
granulata. Final judgement on this point must be withheld until type material can be
studied.

EXPLANATION OF PLATE 37

All illustrations shaded camera lucida drawings, X 140. a = side view; b = terminal view.

Figs. 1-3. Bolivinoides hiltermanni sp. nov. Lower Campanian, Actinocamax quadratus Zone, Culver
Cliff, Isle of Wight. 1, Holotype, Sample C29, BM P45724. 2, Paratype, Sample C24. 3, Paratype,
Sample C29, BM P45725.

Fig. 4. Bolivinoides pustulata Reiss. Lower Campanian, A. quadratus Zone, Culver Cliff, Isle of Wight;
Sample 23, BM P45741.

Fig. 5. Bolivinoides aff. B. hiltermanni sp. nov. Transitional form between B. pustulata and B. hilter-
manni. Lower Campanian, A. quadratus Zone, Culver Cliff, Isle of Wight ; Sample C23, BM P45728.

Fig. 6. Bolivinoides sidestrandensis sp. nov. Paratype, BM P45745. Lower Maestrichtian, Belemnitella
lanceolata Zone, Sidestrand, Norfolk.

Figs. 7-9. Bolivinoides strigillata (Chapman). Topotypes. Santonian, M. testudinarius Zone, Phos-
phatic Chalk, Taplow, Buckinghamshire. Specimens 7 and 8 are also illustrated by unretouched
photographs on Plate 34. 7, BM P45729; 8, BM P45730; 9, BM P45731.

EXPLANATION OF PLATE 38

All illustrations shaded camera lucida drawings, X 140. a = side view; b — terminal view.

Figs. 1-5. Bolivinoides laevigata Marie. Series of specimens showing variation in test shape and
ornamentation. 1, 5, Lower Maestrichtian, B. lanceolata Zone, Sidestrand, Norfolk. 2-4, Upper
Campanian, B. mucronata Zone, Culver Cliff, Isle of Wight, Sample CIO. 1, BM P45735; 2, BM
P45736; 3, BM P45737; 4, BM P45738; 5, BM P45739.

Fig. 6. Bolivinoides peterssoni Brotzen. Final chamber broken. Lower Maestrichtian, B. lanceolata
Zone, Sidestrand, Norfolk; BM P45742.

Figs. 7-9. Bolivinoides praelaevigata sp. nov. Lower part of Upper Campanian, B. mucronata Zone,
Culver Cliff, Isle of Wight. 7, Holotype, Sample C13, BM P45732; 8, Paratype, Sample C13, BM
P45734; 9, Paratype, Sample Cl 8, BM P45733.

Palaeontology, Vol. 9

PLATE 37

BARR, Upper Cretaceous Bolivinoides

Palaeontology, Vol. 9

PLATE 38

BARR, Upper Cretaceous Bolivinoides

F. T. BARR: FORAMINIFERAL GENUS BOLIVINOIDES

237

Occurrence. B. pustulata is restricted to the upper part of the A. quadratus Zone (upper
part of the Lower Campanian). Reiss (1954, text-fig. 1) listed an identical range for this
species.

Bolivinoides praelaevigata sp. nov.

Plate 38, figs. 7-9

Description. Test biserial, elongate, very gradually tapering, greatest breadth near aper-
tural end, flatly elliptical in cross section; proloculus smooth, globular, succeeded by 7
or 8 pairs of slightly inflated, tapering chambers, uniformly and fairly rapidly increasing
in size; sutures oblique, slightly curved, apparent in peripheral portion of test, obscured
by ornamentation in medial area; aperture a slit-like or small loop-shaped opening
located near centre of apertural face, bordered by weakly raised lip and with internal
tooth plate structure; 2 or 3 weakly raised lobes on final chambers running perpen-
dicular to sutures, ornamentation on earlier chambers less distinct consisting of 1 or
2 weakly raised lobes; lobes on inner portion of chambers slightly thicker and more
apparent than outer lobes, sometimes giving appearance of 2 very weakly defined
medial ribs; wall calcareous, finely perforate; length/breadth ratio T9 to 2-5, most
commonly 2-0 to 2-3; breadth/thickness ratio 2-0 to 2-6; approximate average length
of adult specimen 0-35 mm.

Dimensions of holotype (BM P45732). Length 0-36 mm.; breadth 0-17 mm.; thickness 0 07 mm.

Remarks. B. praelaevigata is the direct ancestral species of B. laevigata Marie. It occurs
in the lower B. mucronata Zone and gives rise to B. laevigata in about the middle part
of this zone. No specimens of B. laevigata were found in the basal B. mucronata Zone
although B. praelaevigata was common. In the middle part of this zone, however, a
complete gradation was observed between B. praelaevigata and B. laevigata. The
length/breadth ratio of B. praelaevigata is most commonly 2-0 to 2-3, whereas that of
B. laevigata is usually T8 to 2-0. Furthermore, the test of B. praelaevigata has a smaller
breadth, is more compressed and often slightly smaller. The ornamentation of both
species, however, is quite similar.

Occurrence. B. praelaevigata is restricted to the lower part of the B. mucronata Zone
and appears to be a useful index fossil for this part of the Campanian. B. laevigata, on
the other hand, has a much greater stratigraphic range, extending into the Maestrich-
tian.

Bolivinoides laevigata Marie
Plate, 34, fig. 7; Plate 38, figs. 1-5

Bolivinoides decorata (Jones) laevigata Marie 1941, p. 189, pi. 29, fig. 281.

Bolivinoides decorata laevigata Marie; Hiltermann 1952, p. 63, text-figs. 4, 5.

Bolivinoides praecursor Reiss 1954, pp. 156, 157, pi. 30, figs. 4-8.

Bolivinoides sp., Reiss 1954, p. 157, pi. 29, fig. 11.

Bolivinoides laevigata Marie; Hiltermann and Koch 1955, pp. 367, 368, pi. 27, figs. 13-15.

Bolivinoides praecursor Reiss; Bieda 1958, pp. 41, 42, text-fig. 12.

Bolivinoides laevigata Marie; Hiltermann and Koch 1960, p. 72, table 3.

Bolivinoides laevigatus Marie, Hiltermann and Koch 1962, p. 316, table 19, pi. 51, fig. 10.

Bolivinoides laevigatus Marie; Hiltermann 1963, p. 213, pi. 2, figs. 6-10.

238

PALAEONTOLOGY, VOLUME 9

Description. Test biserial, gradually tapering, flatly elliptical cross section; proloculus
smooth, globular, succeeded by 7 to 9 pairs of slightly inflated, tapering chambers
uniformly and rapidly increasing in size; sutures oblique, slightly curved, apparent in
peripheral portion of test, obscured by ornamentation in medial area; aperture a slit-
like or small loop-shaped opening located near centre of apertural face bordered by
weakly raised lip and with internal tooth plate structure; 2 to 4 weakly raised lobes
present on final chambers near medial position of test, 1 or 2 lobes present on earlier
chambers near medial part of test, lobes become progressively weaker toward periphery;
wall calcareous, finely perforate; length/breadth ratio 1-5 to 2T, most commonly 1-8
to 2-0; breadth/thickness ratio 1-8 to 2-4; approximate average length of adult specimen
0-40 mm.

Remarks. B. laevigata was originally described as a subspecies of B. decorata (Jones)
(Marie 1941, p. 189). These species are quite distinct, however, and are not directly
related. I agree with Reiss (1954, p. 157) and subsequent authors who have regarded
these forms as separate species. B. laevigata is smaller, more compressed, and has much
more weakly developed ornamentation than B. decorata.

Reiss (loc. cit.) did not recognize B. laevigata in the Middle East, but described a form
which he believed closely related to this species and which he named B. praecursor. This
species has a similar length/breadth ratio to that of B. laevigata, and the stratigraphic
ranges of these two forms are identical. However, Reiss considered them distinct because
of ornamental differences. Unfortunately, when Marie (1941, p. 189) originally described
B. laevigata he illustrated only a single specimen and did not show or describe the
variation in ornamentation for this species. After examination of numerous specimens
of B. laevigata from the B. mucronata Zone of the Paris Basin and southern England,
I believe that B. praecursor falls within the range of variation of B. laevigata and should
be regarded as a junior synonym of that species.

B. laevigata is the direct ancestor of B. peterssoni Brotzen, giving rise to the latter
during the early part of the Maestrichtian. Consequently, these species are very similar
morphologically and are often difficult to distinguish consistently. B. peterssoni has
more distinct sutures, shorter and sometimes fewer ‘ornamental’ lobes and is often less
compressed than B. laevigata.

Occurrence. B. laevigata occurs in the B. mucronata Zone (Upper Campanian) of
southern England and in the Lower Maestrichtian chalk at Sidestrand, Norfolk. This
species was originally described from the B. mucronata Zone of the Paris Basin. Hilter-
mann and Koch (1960, p. 72) and Hiltermann (1963, p. 205) recorded the range of
B. laevigata in north-west Europe as Upper Campanian and Lower Maestrichtian.
Reiss (1954, p. 160) recorded a similar range for B. praecursor.

Bolivinoides peterssoni
Plate 38, fig. 6

Bolivinoides peterssoni Brotzen 1945, p. 49, pi. 1, fig. 10.

Bolivinoides peterssoni Brotzen; Pozaryska 1954, p. 256, text-fig. 8.

Bolivinoides peterssoni Brotzen; Reiss 1954, p. 157, pi. 30, figs. 12-14.

Bolivinoides peterssoni Brotzen; Hiltermann and Koch 1955, pp. 366, 367, pi. 28, figs. 7, 8.

Bolivinoides peterssoni Brotzen; Hiltermann and Koch 1960, p. 72, chart 3.

F. T. BARR: FORAMINIFERAL GENUS BOL1VINOIDES

239

Bolivinoides peterssoni Brotzen; Vassilenko 1961, pp. 198, 199, pi. 40, figs. 8o, b.

Bolivinoides peterssoni Brotzen; Hiltermann and Koch 1962, p. 317, pi. 50, fig. 16.

Bolivinoides peterssoni Brotzen; Hiltermann 1963, p. 213, pi. 2, figs. 1-3.

Occurrence. A few specimens of B. peterssoni were found in the Lower Maestrichtian
chalk at Sidestrand, Norfolk. Numerous specimens of this species were also found in
the Upper Maestrichtian chalk at Stevns Klint, Denmark. This species ranges through-
out almost the entire Maestrichtian, but appears to be most abundant in the upper
part of this stage. The type specimens of B. peterssoni were originally described by Brotzen
(1945, p. 49) from Upper Maestrichtian strata penetrated by a bore hole at Hollviken,
Scania, Sweden.

Bolivinoides sidestrandensis sp. nov.

Plate 34, figs. 10, 1 1 ; Plate 36, figs. 6, 9; Plate 37, fig. 6

Bolivinoides decora ta delicatula Cushman; Hiltermann and Koch 1950 (non Cushman 1927),
p. 612, fig. 5, nos. 65-67.

Bolivinoides decorata delicatula Cushman; Hiltermann 1952, pp. 61, 63, text-figs. 4, 5.

Bolivinoides decorata delicatula Cushman; Pozaryska 1954, pp. 255, 256, text-fig. 7.

Bolivinoides decorata delicatula Cushman; Witwicka 1958, p. 199, pi. 9, figs. 10o, b.

Bolivinoides delicatula Cushman; Bieda 1958, pp. 38, 39, text-figs. 10 a-c.

Bolivinoides decorata delicatula Cushman; Hiltermann 1960, p. 72, table 3.

Bolivinoides delicatulus Cushman; Vassilenko 1961, pp. 193-5, pi. 40, figs, la, b.

Bolivinoides decoratus delicatulus Cushman; Hiltermann and Koch 1962, p. 315, table 19,
pi. 47, fig. 5.

Bolivinoides delicatulus regularis Reiss; Hiltermann 1963 (non Reiss 1954), p. 205, pi. 3, figs. 7,

11, 12.

Description. Test biserial, elongate, gradually tapering with greatest breadth near
apertural end, flatly elliptical in cross section; proloculus smooth, globular, followed
by 7 or 8 pairs of slightly inflated, tapering chambers uniformly and rapidly increasing
in size; sutures oblique, slightly curved, usually obscured by ornamentation but some-
times fairly distinct in latest part of test; aperture a slit-like or small looped-shaped
opening located near centre of apertural face, bordered by weakly raised lip and with
internal tooth plate structure; 5 or 6 well-defined but small, uniformly arranged,
ornamental lobes on final chambers running perpendicular to sutures, ornamentation
on earlier chambers less distinct but arranged in same uniform manner; wall calcareous,
finely perforate; length/breadth ratio 1-8 to 2-5; breadth/thickness ratio 2T to 2-8;
approximate average length of adult specimen 0-45 mm.

Dimensions of bolotype (BM P45743). Length 0-48 mm. ; maximum breadth 0-20 mm. ; maximum thick-
ness 0 08 mm.

Remarks. Specimens which are now included in the species B. sidestrandensis have
been commonly recorded from various parts of Europe. However, they have usually
been considered to belong to either B. delicatula Cushman or B. regularis Reiss. Cush-
man (1927, p. 90) originally described B. decorata (Jones) var. delicatula from the Lower
Tertiary Velasco Shale near Vera Cruz, Mexico. Later Cushman (1946, p. 1 13, pi. 48,
figs. 10-14) recorded B. decorata delicatula from various formations of Navarro and
Taylor ages from the Gulf Coastal Area of the U.S.A. Reiss (1954, p. 158) elevated
Cushman’s variety to specific rank as he found no evidence to support the contention

240

PALAEONTOLOGY, VOLUME 9

that B. decorata delicatula was derived from B. decorata (Jones) or even from the
‘decorated group. Furthermore, Reiss regarded the specimens identified by Cushman
(loc. cit.) as B. decorata delicatula from formations of lower Navarro and Taylor ages
as being distinct from the typical form. He therefore proposed the name Bolivinoides
regularis for the Navarro-Taylor form, with a specimen figured by Cushman (1931,
pi. 35, fig. 13; 1946, pi. 48, figs. 12a, b) from the Saratoga Chalk of Howard County,
Arkansas, as the holotype. Reiss also concluded that the specimens figured as B.
decorata delicatula Cushman by Hiltermann and Koch (1950) and Hiltermann (1952),
although differing in some detail, were conspecific with his new species B. regularis.

I agree with Reiss that the specimens figured by Cushman (1946, pi. 48, figs. 10-14)
as B. decorata delicatula represent two distinct species. Cushman’s (op. cit., pi. 48, figs.
11, 12) specimens which Reiss placed in his species B. regularis are indeed very different
from the Danian-Paleocene specimens of B. delicatula ; however, they are quite similar
to B. decorata (Jones). The holotype of B. regularis , which is located in the U.S. National
Museum, Washington, D.C., has been examined and was found to have a greater
relative thickness than was shown by Cushman’s illustrations. The ornamentation
has a somewhat more uniform or regular appearance than is usually exhibited by B.
decorata , but in most respects it is very similar to this species. Only by the examination
of additional topotypic specimens will one be able to decide with assurance whether
B. decorata (Jones) and B. regularis Reiss are distinct or conspecific. It may be that
B. regularis would best be considered a geographic subspecies of B. decorata ; however,
final judgement on this taxonomic problem must be withheld until additional material
becomes available.

On the other hand, I do not concur with Reiss’s contention that the specimens illus-
trated by Hiltermann and Koch (1950, p. 612, fig. 5, nos. 65-67) and Hiltermann (1952,
pp. 61, 63, text-figs. 4, 5) from north-west Germany are conspecific with the holotype
of B. regularis from the Saratoga Chalk of Arkansas. The specimens from Germany
appear to be almost identical with the specimens from England which are placed in the
new species B. sidestrandensis. The latter differs from B. regularis by having a smaller
size, being more compressed, and having a less flaring test with a greater length/
breadth ratio. B. sidestrandensis has the same general shape as B. delicatula ( s.s .),
but is distinguished by its characteristic ornamentation, which is much more distinct
and uniform than that of B. delicatula.

Occurrence. B. sidestrandensis occurs in abundance in the Lower Maestrichtian chalk
exposed in the seacliff section at Sidestrand, Norfolk. Black (1964) described some of
the fossil nannoplankton from this locality. B. sidestrandensis was also found in the
Ballycastle Pellet Chalk from the main exposure along the west side of Ballycastle
Harbour, Northern Ireland. No specimens of this species were recovered from the
B. mucronata Zone of southern England. However, Hiltermann (1963, p. 205) recorded
the stratigraphic range of B. delicatula regularis (= B. sidestrandensis) as upper Upper
Campanian to lower Upper Maestrichtian.

Dr. Max Meijer kindly provided me with material from the Lower Maestrichtian
(B. lanceolata Zone) of Grindaal, north-east Belgium. This glauconitic chalk contained
numerous specimens of B. sidestrandensis associated with rare specimens of B. peter ssoni
Brotzen and B. australis Edgell.

F. T. BARR: FORAMINIFERAL GENUS BOLIVINOIDES

241

Hiltermann (1963, p. 217, pi. 3, fig. 11) illustrated a specimen identified as B. delica-
tulus regularis (Reiss) (= B. sidestrandensis ) from the Marshalltown Formation of New
Jersey. I have also found specimens of B. sidestrandensis in material from Olsson’s
(1964, p. 160) locality NKJ 134 in the Marshalltown Formation of New Jersey. This
locality is probably uppermost Campanian in age.

REFERENCES

akimets, v. s. 1961. Stratigrafiya i foraminifery verkhnemolovykh otlozhenii Belorussii. Paleontolo-
giya i Stratigrafiya BSSR , Izv. Akad. Nauk. beloruss. SSR, 1-245, pi. 1-19.
barnard, t. 1963. The morphology and development of species of Marssonella and Pseudotextulariella
from the Chalk of England. Palaeontology , 6, 41-55, pi. 7.

and banner, f. t. 1953. Arenaceous Foraminifera from the Upper Cretaceous of England.

Quart. J. geol. Soc. Load. 109, 173-216, pi. 7-9.

barr, f. t. 1962. Upper Cretaceous planktonic Foraminifera from the Isle of Wight, England.
Palaeontology, 4, 552-80, pi. 69-72.

— — and cordey, w. g. 1964. Some Upper Cretaceous Foraminifera from the Chapman Collection
(1892). J. Paleont. 38, 306-10, pi. 49.

barrois, c. 1876. Recherches sur le terrain cretace superieur de FAngleterre et de l’lrlande. Mem.
Soc. geol. N. 1-234.

bieda, e. 1958. Otwornice przewodnie i wiek kredy piszqcej Mielnika (Index Foraminifera and the
age of the Mielnik Chalk, Eastern Poland). Z Badan Mikropaleont. 3, Inst. Geol. Bull. 121, 17-89.
Warsaw (Russian and English summaries).

black, m. 1964. Cretaceous and Tertiary coccoliths from Atlantic seamounts. Palaeontology, 7,
306-16, pi. 50-53.

brotzen, F. 1945. De geologiska resultaten fran borrningarna vid Flollviken; Preliminar rapport —
Part I: Kritan. Sver. geol. Unders. Afh. (C), no. 465, 1-65.
brydone, r. m. 1906. Further notes on the stratigraphy and fauna of the Trimingham Chalk. Geol.
Mag. 53, 13-22, 72-78, 124-31, pi. 2-5, 8, 9.

1908. On the subdivisions of the Chalk of Trimingham (Norfolk). Quart. J. geol. Soc. Lond.

64, 401-12, pi. 47, 48.

1914. The zone of Ojfaster pilula in the south English Chalk. Geol. Mag., n.s., 1, 359-69, 405-1 1 ,

449-57 and 509-13.

bykova, n. k. 1959. K voprosy o zakonomernostyakh filogeneticheskogo razvitiya foraminifer v
ysloviyakh periodicheski izmenyayushscheisya sredy. Vop. paleont. biostrat., Trudy Vses. paleontol.
obshch., Moscow. 2, 63-75.

chapman, f. 1892. Microzoa from the Phosphatic Chalk of Taplow. Quart. J. geol. Soc. Lond. 48,
514-18, pi. 15.

charlesworth, J. k. 1963. Historical Geology of Ireland. Edinburgh and London.

coquand, h. 1857. Notice sur la formation cretace de la Charente. Bull. Soc. geol. Fr. (2). 14, 55-98.

1858. Description physique, geologique, pale ontologique et mineralogique du departement de la

Charente. Besangon.

curry, d. 1962. A Lower Tertiary outlier in the English Channel, with notes on the beds surrounding
it. Quart. J. geol. Soc. Lond. 118, 177-205.

cushman, J. A. 1927. American Upper Cretaceous species of Bolivina and related species. Contr. Cush-
man Lab. foramin. Res. 2, (4), 85-91, pi. 12.

1931. The Foraminifera of the Saratoga Chalk. J. Paleont. 5, 297-315, pi. 34-36.

1946. Upper Cretaceous Foraminifera of the Gulf Coastal Region of the United States and

adjacent areas. Prof. Pap. U.S. geol. Surv. 206, 1-241, pi. 1-66.
dain, l. g. 1934. Foraminifera of the Upper Jurassic and Cretaceous beds of the Djaksy-bai Oil
Field (Temir region). Trans. Geol. Oil Inst., Leningrad, ser. A, no. 43, 1-63, pi. 1-5.
desor, e. 1846. Sur le terrain danien, nouvel etage de la craie. Bull. Soc. geol. Fr., ser. 2, 4, 179-82.
dumont, a, 1849. Rapport sur la carte geologique du Royaume. Bull. Acad. r. Belg. Cl. Sci. 16, 351-73.

242

PALAEONTOLOGY, VOLUME 9

edgell, h. s. 1954. The stratigraphical value of Bolivinoides in the Upper Cretaceous of north-west
Australia. Contr. Cushman Fdn foramin. Res. 5 (2), 68-76, pi. 13, 14.
gignoux, m. 1955. Stratigraphic Geology. San Francisco.

hagn, h. 1953. Die Foraminiferen der Pinswanger Schichten (Unteres Obercampan). Ein Beitrag zur
Mikropalaontologie der helvetischen Oberkreide Siidbayerns. Palaeontographica, 104, 1-1 1 8, pi. 1-8.
Hancock, J. M. 1961. The Cretaceous System in Northern Ireland. Quart. J.geol. Soc. Loud. 117,
11-36.

heron-allen, e. and earland, a. 1910. On the Recent and fossil Foraminifera of the shore-sands
at Selsey Bill, Sussex. Part V. The Cretaceous Foraminifera. /. R. microsc. Soc. 401-26, pi. 6-1 1 .
hiltermann, h. 1952. Stratigraphische Fragen des Campan and Maastricht unter besonderer Beriick-
sichtigung der Mikropalaontologie. Geol. Jb. 67, 47-66.

1963. Zur Entwicklung der Benthos-Foraminifere Bolivinoides. In Evolutionary Trends in Fora -

mini/ era. Amsterdam.

and koch, w 1950. Taxonomie und Vertikalverbreitung von Bolivinoides- Arten im Senon Nord-

westdeutschlands. Geol. Jb. 64, 595-632.

■ 1955. Biostratigraphie der Grenzschichten Maastricht/Campan in Liineburg und in der

Bohrung Brunhilde. Part 2, Foraminiferen. Ibid. 70, 354-84, pi. 27-29.

1960. Oberkreide Biostratigraphie Mittels Foraminiferen. Int. geol. Congr., 21st Sess.,
Part 6, 69-76.

1962. Oberkreide des nordlichen Mitteleuropa. Lietfossilien der Mikropalaontologie, 299-

338, pi. 42-51. Berlin.

hinte, J. e. van. 1963. Zur Stratigraphie und Mikropalaontologie der Oberkreide und des Eozans des
Krappfeldes (Karnten). Jb. geol. Bundesanst., Wien. 8, 1-147, pi. 1-22.
hofker, J. 1952. Zur Fassung der Foraminiferengattung Bolivinoides Cushman 1927. Geol. Jb. 66,
377-82.

1957. Foraminiferen der Oberkreide von Nordwestdeutschland und Holland. Beih. geol. Jb.

27, 1-464.

1958«. Upper Cretaceous Bolivinoides guide forms. Micropaleontology, 4, 329-34, pi. 1, 2.

1 9586. Foraminifera from the Cretaceous of Limburg, Netherlands. Part 38, The Gliding Change

in Bolivinoides during time. Natuurh. Maandbl. 47, no. 11-12, 145-59.
jeletzky, j. a. 1951. The place of the Trimingham and Norwich Chalk in the Campanian-Maes-
trichtian succession. Geol. Mag. 88, 197-208.

jukes-browne, a. J. 1904. The Cretaceous Rocks of Britain. 3, The Upper Chalk of England.
Mem. geol. Surv. U.K.

1912. On the recognition of two stages in the Upper Chalk. Geol. Mag. 9, 304-13, 360-72.

KAPTARENKO-CHERNOUSOVA, O. K., GOLYAK, L. M., ZERNETSKII, B. F., KRAEVA, E. YA., and LIPNIK, E. S.

1963. Atlas kharakternikh foraminifer yury, mela i paleogena platformennoi chasti Ukrainy.
Trudy Inst. geol. Nauk, Kiev, ser. strut, paleont. 45, 1-200, pi. 1-47.
macfadyen, w. a. 1932. Foraminifera from some late Pliocene and glacial deposits of East Anglia.
Geol. Mag. 69, 481-96, pi. 34, 35.

marie, p. 1941. Les Foraminiferes de la Craie a Belemnitella mucronata du Bassin de Paris. Mem.

Mus. natn. Hist, nat., Paris, N.s. 12, 1-296, pi. 1-37.
mcgugan, a. 1957. Upper Cretaceous Foraminifera from Northern Ireland. J. Paleont. 31, 329-48,
pi. 31-35.

montanaro gallitelli, E. 1957. A revision of the foraminiferal family Heterohelicidae. Bull. U.S.
natn. Mus. 215, 133-54, pi. 31-34.

olsson, r. k. 1964. Late Cretaceous planktonic foraminifera from New Jersey and Delaware. Micro-
paleontology, 10, 157-88, pi. 1-7.

d'ORBiGNY, a. 1842. Paleontologie frangaise. Terrains cretaces, 2, Paris.

1852. Cours elementaire de paleontologie et de geologie stratigraphiques, 2 (2), 1-841, Paris.

osborne white, h. J. and treacher, l. 1905. On the age and relations of the Phosphatic Chalk of
Taplow. Quart. J. geol. Soc. Loud. 61, 461-94.

pozaryska, K. 1954. O przewodnich otwornicach z kredy gornej Polski srodkowej (The Upper
Cretaceous Index Foraminifers from Central Poland). Acta geol. pol. 4, 59-72, 249-76 (Russian and
English summaries).

F. T. BARR: FORAMINIFERAL GENUS BOLIVINOIDES 243

reiss, z. 1954. Upper Cretaceous and Lower Tertiary Bolivinoides from Israel. Contr. Cushman Fdn
foramin. Res. 5, 154-64, pi. 28-31.

reyment, R. A. 1959. The foraminiferal genera Afrobolivina gen. nov. and Bolivina in the Upper
Cretaceous and Lower Tertiary of West Africa. Stockh. Contr. Geol. 3, 1-57, pi. 1-7.
rowe, a. w. 1908. The Zones of the White Chalk of the English Coast. Pt. V, The Isle of Wight. Proc.
Geol. Ass. 20, 209-352, pi. 7-23.

vassilenko, v. p. 1961. Foraminifery verkhnogo mela poluostrova mangyshlaka. Trudy VNIGRI,
171, 1-487, pi. 1-41.

walsh, p. t. 1960. Cretaceous Chalk in the Killarney District, Eire. Proc. geol. Soc., No. 1581,
112-13.

williams-mitchell, e. 1948. The zonal value of Foraminifera in the Chalk of England. Proc. Geol.
Ass. 59, 91-112, pi. 8-10.

witwicka, e. 1958. Stratygrafia mikropaleontologiczna wiercenia w Chelmie (Micropalaeontological
stratigraphy of Upper Cretaceous of the Chelm Borehole, Lublin Upland). Z Badah Mikropaleont.
3, Bull. Inst. Geol. 121, 177-267, pi. 8-19. Warsaw (Russian and English summaries).
wright, J. 1875. A list of the Cretaceous Microzoa of the north of Ireland. Proc. Belf. Nat. Fid
Club, ser. 2, 1, 73-99.

1886. A list of the Cretaceous Foraminifera of Keady Hill, County Derry. Ibid. 2, 327-32,

pi. 27.

F. T. BARR

Oasis Oil Company of Libya, Inc.,
Oasis Oil Building,

P.O. Box 395

Manuscript received 3 February 1965 Tripoli, Libya

THE DEVONIAN BLASTOID BELOCRINUS
FROM FRANCE

by DONALD B. MACURDA, JR.

Abstract. The nature and validity of the Lower Devonian blastoid genus Belocrinus Munier-Chalmas from
north-western France has been in doubt since its original description. A complete specimen of the type species,
B. cottaldi (Munier-Chalmas) is described and illustrated. The genus Belocrinus is redefined; it is similar to, but
not synonymous with, the genus Cordyloblastus Fay. The number of anal deltoids and hydrospires in Belocrinus
could not be determined. The ontogenetic development of the plates of the calyx is also described.

In 1876, Munier-Chalmas briefly described a peculiar fossil from the Lower Devonian
of north-western France which he named Belemnocrinus cottaldi. He believed that it
was a crinoid, composed of five longitudinal pieces. Finding his generic name to be
preoccupied, he proposed the name Belocrinus for it in 1881. The description was
repeated by Oehlert (1882) and specimens were illustrated for the first time; in his plate
legend (p. 363) he described the calyx as being composed of three pieces. Etheridge and
Carpenter (1883) were the first workers to attribute this fossil to the Blastoidea. They
called attention to Oehlert’s figures and stated that they appeared to represent the
elongated basal cup of a Troosticrinus or Pentremitidea (p. 245). Munier-Chalmas
furnished them with a specimen which they described and illustrated in 1886; they
diagnosed the material as the large basal plates of a Metablastus. No additional material
other than the basals was known until Phillipot (1957) described a complete specimen
which was in the H. de Keravel collection at the Universite de Rennes. His description
furnished the first detailed information about the radials, deltoids, and ambulacra.
Unfortunately, the summit was obscured by matrix. On the basis of the available
evidence, he continued Etheridge and Carpenter's assignment of the species to Meta-
blastus.

Recent developments in preparatory techniques have made it possible to expose
hitherto obscured structures. A. Phillipot and M. Y. Milon kindly allowed the writer
to further develop the specimen described by Phillipot with an air abrasive machine. By
this means, it has been possible to determine that it has four spiracles and an anispiracle,
thus separating it from Metablastus , which has paired spiracles. It appears to represent
a valid genus, for which the name Belocrinus has priority. The writer studied the basal
plate in the British Museum (Natural History) which was figured by Etheridge and
Carpenter (E166). It is an isolated basal plate of the same type as is illustrated in Plate
39, fig. 7, and as are found in the complete specimen. These are all from Bois-Roux
in France and belong to the same species. No type specimen was ever designated for this
species; the complete specimen described and figured herein is designated the hypotype
for nomenclatural purposes. It is described below and the ontogenetic development of
the principal plates determined.

Acknowledgements. I would like to thank Andre Phillipot and M. Y. Milon (Institut de Geologie,
Universite de Rennes, Rennes) for allowing me to examine and prepare the specimen described in

[Palaeontology, Vol. 9, Part 2, 1966, pp. 244-51, pi. 39.]

D. B. MACURDA JR.: DEVONIAN BLASTOID BELOCRINUS FROM FRANCE 245

this paper; and R. P. S. Jefferies and Hugh Owen (British Museum (Natural History), London), for
their help while studying the museum’s blastoid collection. My studies in Europe were made possible
by grants from the Museum of Paleontology, Horace H. Rackham School of Graduate Studies, and
H. H. Power Education Trust for Faculty Travel, University of Michigan.

SYSTEMATIC DESCRIPTION

ClaSS BLASTOIDEA
Order spiraculata

Genus belocrinus Munier-Chalmas 1881

Type species. Belemnocrinus cottaldi Munier-Chalmas.

1876 Belemnocrinus Munier-Chalmas, p. 105 (junior synonym of White 1862).

1881 Belocrinus Munier-Chalmas, p. 503.

1938 Belocrinus Munier-Chalmas; Bassler, p. 52.

1943 Belocrinus Munier-Chalmas; Bassler and Moodey, p. 212.

Generic diagnosis. Spiraculate blastoids with five spiracles, one of which is an anispiracle
located between an epideltoid (?) and hypodeltoid (?), all deltoids quite visible in side
view, a single pore occurring between adjacent side plates along radial and deltoid
margins, lancet covered by side plates, calyx elongate club-shaped; large, massive
basals; ten hydrospire groups; number in each unknown.

Age. Lower Devonian, north-west France.

Remarks. The generic characters of this form are similar to those of the genus Cordylo-
blastus (see Fay 1964, p. 86) but present information indicates that they are separate
genera. They each have five spiracles; the number of anal deltoids in Belocrinus is not
clear, however, and they could be the same as in Cordyloblastus. The deltoids of these
two forms apparently differ, there being an extensive external aboral growth sector in
the former (see later). There also appear to be ambulacral pores along both the radials
and deltoids of Belocrinus ; Fay (1961, p. 54) described these as being absent along the
deltoids in Cordyloblastus (see PI. 39, figs. 9, 10). The basals of the former are also
unusually pronounced, giving the calyx an extremely elongate appearance. In view of
the uncertainty as to the configuration of the anal deltoids, however, it is recommended
that the generic name Belocrinus should not be extended beyond B. cottaldi until the
nature of the anal deltoids can be determined.

Belocrinus cottaldi (Munier-Chalmas)

Plate 39, figs. 1-8, 11

1876 Belemnocrinus cottaldi Munier-Chalmas, p. 105.

1877 Belemnocrinus cottcddi Munier-Chalmas; Delage, p. 81 (non vide).

1881 Belocrinus cottaldi (Munier-Chalmas); Munier-Chalmas, p. 503.

1882 Belocrinus cottaldi , (Munier-Chalmas); Oehlert, p. 362, pi. 9, figs. 3 a-e.

1883 Troosticrinus cottaldi ? (Munier-Chalmas); Etheridge and Carpenter, p. 245.

1883 Pentremitidea cottaldil (Munier-Chalmas); Etheridge and Carpenter, p. 245.

1886 Metablastus cottaldi (Munier-Chalmas); Etheridge and Carpenter, pp. 201-3, pi. 5,
fig. 22.

1899 Metablastus cottaldi (Munier-Chalmas); Bather, pp. 18, 19.

246

PALAEONTOLOGY, VOLUME 9

1938 Belocrinus cottaldi (Munier-Chalmas); Bassler, p. 52.

1943 Belocrinus cottaldi (Munier-Chalmas); Bassler and Moodey, p. 212.

1957 Metablastus cottaldi (Munier-Chalmas); Phillipot, pp. 64—68, text-figs. 1-5.

Description. Calyx elongate club-shaped in side view, pentagonal in oral view. Length
44-0 mm.; width 14-5 mm. Pelvis extremely long for a blastoid, massive, conical, flaring
out gradually beneath ambulacra. Vault parabolic. Vault 13-7 mm., pelvis 30-3 mm.;
pelvic angle 30°. Greatest width at ambulacral tips; cross-section pentagonal with
slightly concave interambulacral areas.

Basalia three, azygous basal in ab interambulacral area (for terminology see Fay 1961),
extremely massive, narrow, elongate. Azygous basal quadrate, slightly asymmetric;
lower edges very narrow V shape, distal edges broader V; length 24-2 mm., width
5-7 mm. Point of origin located very near proximal tip. Growth has occurred on both
distal and lateral edges. Plate has grown 20-3 mm. distally, 2-2 mm. laterally. When
br (distal) growth axis of azygous basal (see PI. 39, fig. 8) reached 5-0 mm. from origin,
interbasal (bb) 0-8 mm.; at 11-2 mm., bb 1-4 mm. If br+bb = 100% total growth at
a given increment, then br at 5-0 = 86% total growth, at 11-2= 88%, at 20-3 = 90%;
thus br accelerating relative to bb during ontogeny. Well-developed interbasal growth
axis; sharp, elevated sides to azygous basal which become reduced distally. Upper
surface very slightly convex. Fine growth-lines parallel to plate edges; 9 per mm. on br
axis.

Larger basals hexagonal; distal edges double V shape; lateral edges very narrow
V shape. Sides sharply elevated due to interbasal growth axis. Upper surface very
slightly convex. Length (to centre) 22-0 mm. ; to one of points 24-3 mm. ; width 9-3 mm. ;
distal (br) growth axis 21-4 mm., growth front 5-1 mm. Fine growth-lines parallel sutures.

Three basalia together have a trigonal rounded cross section, slightly concave along
sutures (PI. 39, fig. 2). They taper abruptly proximally near base to form a small round
facet for attachment of column (now chipped); attachment area 1-0 mm. in diameter;
no crenellar facets preserved. Each large basal has a crack extending proximally from

EXPLANATION OF PLATE 39

Figs. 1, 2. Oral and basal views of only complete specimen of Belocrinus cottaldi Munier-Chalmas;
anispiracle at 6 o’clock. Figs. 3-6, 8, and 1 1 of same specimen. X 2.

Fig. 3. Inclined oral view of B. cottaldi at bc interambulacral area, showing reflected radiodeltoid suture
and internal septum dividing spiracles. X 8.

Fig. 4. Lateral a ambulacral view of B. cottaldi. X 2.

Fig. 5. Enlarged oral view of B. cottaldi showing the four spiracles and the anispiracle (latter at
6 o'clock). Peristome in centre. X 8.

Fig. 6. Aboral end of b ambulacrum, showing brachiolar facets, pore furrows, side and main food
grooves, and cover plate lobes and furrows. X 8.

Fig. 7. Isolated basal plate of B. cottaldi, UMMP 51144. x2.

Fig. 8. Enlarged view of basal plates of B. cottaldi, showing growth lines of interbasal (bb — vertical)
and basal-radial (br — inclined) sectors. Origins at lower corners of figure. X 4.

Figs. 9, 10. Inclined b ambulacral and oral views of Cordyloblastus eifelensis (Roemer), USNM
S5086, showing four spiracles and anispiracle (at 6 o’clock in fig. 10). Note presence of pores only
along radials and presence of small deltoid body aboral to spiracles, x 8.

Fig. 11. ab interambulacral view of B. cottaldi, showing growth-lines of radiodeltoid (rd — upper
inclined), inter-radial (rr — vertical), and radial-basal (rb — lower inclined) sectors of radials. x4.

Palaeontology, Vol. 9

PLATE 39

MACURDA, Devonian blastoid Belocrinus

D. B. MACURDA JR.: DEVONIAN BLASTOID BELOCRINUS FROM FRANCE 247

the top centre. Each dies out before reaching origin of plate; growth-lines appear to
cross without interruption. Apparently not a primary suture; significance unknown.

Radials five, narrow, elongate (aboral portion V shaped; lateral sides almost parallel),
with deep narrow V for ambulacra (see PI. 39, fig. 4). Height 3-1 mm.; width at base
5-3 mm., at ambulacral tip 6-5 mm., at deltoids 4-9 mm. Radial grew outwards in three
directions from origin at aboral tip of ambulacrum (PI. 39, fig. 11); 12-3 mm. towards
radiodeltoid suture (rd axis), rd front (area along which calcite added) =1-5 mm.;
5-3 mm. towards interradial suture (rr axis), rr front 17-1 mm.; 8-9 mm. towards
radial-basal suture (rb axis), rb front 5-0 mm. Earlier ontogenetic stages: when
rr = 3-7 mm., rd = 8-2 mm. and rb = 5-1 mm. respectively; when rr = 2-0 mm.,
rd = 3-5 mm., rb = 2-7 mm. If rd+rr+rb = 100% of total growth at a given
instance then following percentages found :

RD

RR

RB

rr = 2 0 mm.

43%

24%

33%

= 3-7 mm.

48%

22%

30%

= 5-3 mm.

46%

20%

34%

Only rr shows a continuous trend, apparently slowing down ; rate of growth of rd and
rb appears to increase. Fine growth-lines parallel interradial and radial-basal sutures;
those parallel to radiodeltoid not as evident. Interradial growth area ornamented by
low, slightly irregular ridges parallel to interradial suture (PI. 39, fig. 11). Calcite of
plates an endoskeleton, secreted by mesoderm. Layer of tissue between radial and
basal plates secreted calcite on both edges simultaneously; rate of secretion by tissue
between radials and basals 2-3 times faster on basal br axis than on radial rb axis.
Rate of growth of br axis 3-8 times that of rr, 1-7 times that of rd.

Deltoids four, surround peristome (four regular plus anal deltoids). Length 3-8 mm.
Adoral edge 0-7 mm. wide, forms border of peristome. Large single spiracle divides
adoral portion from body (PI. 39, fig. 5). Spiracle ovoid, diameter 1-1 mm. Septum arises
0-5 mm. below adoral edge of spiracle, divides it internally, and connects adoral and
aboral portions of deltoid. Adoral edge of spiracle formed by adoral edge of deltoid,
adoral lateral edges by adjacent lancet plates, aboral lateral edges by side plates, and
aboral edge by adoral edge of aboral portion of deltoid. Latter part of plate is triangular,
3-0 mm. long; greatest aboral width at radiodeltoid suture 2-5 mm.; width just aboral
to spiracle 0-5 mm. Radiodeltoid suture peculiar shape (PI. 39, fig. 3). Portion adjacent
to interradial suture junction straight; half-way along radiodeltoid suture sharp bend
adorally. Each leg 0-8 mm.; base of triangle 1-5 mm. Significance of shape unknown.

Ornamentation of deltoids obscure; apparently faint growth lines parallel to radio-
deltoid suture.

Entire growth of deltoid 3-8 mm.; rate of growth of br 5-3 times this, and rd 3-2,
rr 1-4, rb 2-3 times respectively.

Anal spiracle an anispiracle, length 1-4 mm., width 1-3 mm. Number of anal deltoids
uncertain, apparently an epi- and hypodeltoid (see below). Epideltoid forms part of
edge of spiracle, width 1-0 mm. Internally, two prongs can be seen extending aborally
to divide anal hydrospire areas from anus but aboral ends obscured by matrix. Junction
with aboral part of anal deltoids not visible. Portion of anal deltoids aboral to anispiracle
presumably a hypodeltoid. Borders of anispiracle as for spiracles, except epideltoid

C 3803

s

248

PALAEONTOLOGY, VOLUME 9

adoral edge, hypodeltoid aboral edge. Latter 0-9 mm. wide here (greater width than
other deltoids to accommodate anus); greatest width at radiodeltoid suture 3-5 mm.
Configuration of hyporadial suture as for regular radiodeltoid suture; inner leg of
suture 1-2 mm., outer 0-5 mm.; length of base of triangle 1-6 mm. Length of hypodeltoid
3-7 mm. Full length of anal deltoids 4-7 mm. Ornamentation of hypodeltoid apparently
as for regular deltoids.

Ambulacra five, linear, narrow, tapering gradually; slightly convex lengthwise.
Length 13-7 mm., greatest width 2-2 mm. Lancet does not reach peristome except in one
ambulacrum (weathering); normally begin 0-3 mm. from peristome. Lancet exposed
only at most adoral end of ambulacrum. Two side plates per mm., 27 on one side of
an ambulacrum. Side plates elongate, obliquely disposed to centre of ambulacrum
(PI. 39, fig. 6). Outer edge embayed by an elongate outer side plate which reaches outer
edge of ambulacrum, and a pore which is formed by a notch on the adoral edge of the
outer side plate and aboral edge of the side plate. Abmedial edge of pore formed by
radial or deltoid. Cross-section of ambulacrum from main food groove to abmedial
edge convex. Side plates slope upwards from food groove for one third of distance,
then slope downwards remaining two-thirds. Abmedial edge about three times lower
in cross-section than main food groove. Brachioles were attached to large, elongate,
elliptical brachiolar facet, which was on the downward sloping outer edge of the side
plates; axis of facet inclined adorally, upper edge being closer to oral area. Brachiolar
facet located on side and outer side plates. Brachiolar food groove descended to am-
bulacrum at high point on cross-section of side plate, becoming a short side food groove
which joins the main food groove. Latter ascended adorally. Cover plate lobes and
furrows can be seen along main food groove; suggestion of presence along side food
grooves as well. A long, narrow, adorally directed arcuate pore furrow originates near
main food groove (but separate from it) admedially to crest of side plate, descends to the
pore via a depression between adjacent high points where the brachiolar food groove
meets the side food groove (PI. 39, fig. 6). After passing through the depression it rests
on the suture between the side plate and outer side plate. Judging from the brachiolar
facet, brachioles were relatively large and elongate in cross-section but no trace now
preserved.

Spiracles four plus an anispiracle; apparently ten hydrospire groups, but number in
each unknown.

Peristome pentagonal; width 1-5 mm.

Remarks. The above description is based on the only known complete specimen which
is in the collections of the Institut de Geologie, Universite de Rennes, Rennes (Ille-et-
Yilaine), France. The only other known material are basal plates.

The growth lines preserved on the plates of blastoids render them ideally suited for
ontogenetic study. Ontogenetic data are presented above for the principal plates but
the data are biased as there is no reflection of specific variation. However, ontogenetic
studies in the Mississippian blastoid Orophocrinus (Macurda 1966) appear to indicate
that ontogenetic curves derived from an individual or a series of individuals generally
reflect the ontogenetic development of the species under study. The growth of the
principal plates of blastoids proceeds by small increments which are added to the lateral
edges of the plate. If there are simultaneous additions on all plates, the number of growth

D. B. MACURDA JR.: DEVONIAN BLASTOID BELOCRINUS FROM FRANCE 249

lines along various growth axes should correspond with one another. This was analysed
in this specimen but the results were inconclusive.

The size of the basal plates relative to the other plates is very unusual. Normally
these are quite small, being confined to a relatively small area at the base of the calyx.
It is also unusual for a measurable interbasal axis to be developed in blastoids.

-RD

text-fig. 1. Schematic diagrams of two blastoids showing differences in growth
patterns in radials and deltoids in such genera as Belocrinus and Orophocrinus (fig. a)
and Phaenoblastus and Phaenoschisma (fig. b). Ambulacral sinus (those parts of
radials and deltoids which lie beneath ambulacra or slope steeply in toward them)
indicated by as. Ambulacra (a) proper have been removed; position indicated by
dashed lines. Fig. a shows the configuration of growth-lines (g) in genera in which
there is growth along the radiodeltoid suture external to the ambulacral sinus
(rd). Note reflexed growth-lines by ambulacral sinus and presence on aboral
portion of deltoid (d). Fig. b illustrates the case in which there is no growth
external to the ambulacral sinus. Note lack of reflexed growth-lines by sinus and
absence on aboral portion of deltoid, which is now a deltoid crest (dc). Basals

indicated by b.

Belocrinus appears to be very similar to Cordyloblaslus. In the latter genus, the deltoid
is not exposed in side view (but there is a small edge of the deltoid body exposed aboral
to the spiracle), while the deltoids of Belocrinus cottaldi are prominently developed
in side view. The exposure or non-exposure of the deltoid is controlled by the develop-
mental pattern of the plate; this is a reflection of the genetic make-up of the organism.
In blastoid genera such as Phaenoschisma and Phaenoblastus (text-fig. 1), the aboral
portion of the deltoid never grows outside the ambulacral sinus. Thus there are never
any growth lines of the radiodeltoid front of the radial exposed outside the ambulacral
sinus, growth occurring entirely within it. The growth-lines of the interradial front
extend up to the sinus and are never reflected as they are in genera such as Orophocrinus
(text-fig. 1), where the deltoid grows external to the ambulacral sinus. The growth-lines
in the radiodeltoid sectors of the radials in the ab interambulacral area of Belocrinus
cottaldi (PI. 39, fig. 11) appear to indicate that the deltoid has grown outside the am-
bulacral sinus throughout most, if not all, of its ontogeny, resulting in its exposure in
side view. The growth-lines in the radiodeltoid sectors are not as evident as those of the

250

PALAEONTOLOGY, VOLUME 9

interradial, but can be seen perpendicular to the ambulacrum for most of the length
of the radials which border the ambulacra. They are not evident near the aboral tip
of the ambulacra as on the e side of the a radial ; here the interradial growth-lines closely
approach the ambulacra. The lack of radiodeltoid growth-lines in this sector may be
a result of secondary secretions of calcite which have covered them, or of non-preserva-
tion.

The growth of the deltoid aboral to the spiracle appears to differ in Belocrinus and
Cordyloblastus. The aboral body of the latter is little developed. However, the specimen
of B. cottaldi described above is much larger than the average specimen of Cordyloblastus.
It is possible that the deltoids of the species in the latter genus may develop a more
prominent deltoid body later in their ontogeny.

The number of the anal deltoids in Belocrinus is unclear. A part of an anal deltoid
plate borders the adoral side of the anispiracle. Two prongs can be seen arising from the
aboral side of the plate. They are soon obscured by matrix and could not be developed
further. They may possibly lead to cryptodeltoids or continue as prongs of an epideltoid.
A large plate lies aboral to the anispiracle; its relation to the plate(s) described above
is unclear. These are tentatively interpreted as an epideltoid and hypodeltoid re-
spectively; further material is needed to determine the exact configuration of the anal
deltoids. If there are two cryptodeltoids, it would indicate a close similarity to Cor-
dyloblastus.

The ambulacral food grooves of Belocrinus were probably roofed over by cover
plates which are known in genera such as Pentremites and Orophocrinus. The peristome
is the highest topographic point on the calyx. There would have to have been some
propulsive mechanism to force currents up the main food grooves to the peristome.
This may have been accomplished by ciliary action. It should also be noted that the side
food grooves are offset from one another and enter the main food groove at an angle of
approximately 45 degrees (directed adorally) (PI. 39, fig. 6). This inclination would seem
to be the optimal hydraulic configuration. If they entered at a right angle, there would
be a decrease in current velocity ; additional energy would have to be expended to over-
come this. If they are inclined, water from the side food grooves would merge with the
currents in the main food groove with a minimum loss of energy. In addition, by having
each one offset from the preceding one, increases in the volume of water are distributed
over a greater area, rather than being concentrated at half as many points. Since the
side food grooves are obliquely oriented, the brachiolar facets are as well. Thus the
brachioles and brachiolar food grooves were obliquely oriented to the calyx. Whether
this conferred any physiological advantage upon the blastoid is unclear at present.

Phillipot (1957, pp. 66, 67) recorded isolated basal cones in the collections at Rennes
which reached a length of 35 to 40 mm. and noted that the complete specimens probably
reached a length of about 70 mm., a large size for blastoids.

Plaster casts of the complete specimen have been deposited in the British Museum
(Natural History), U.S. National Museum, and Museum of Paleontology, University
of Michigan.

Occurrence. The specimen described above was collected by H. de Keravel from the
limestones in the quarry at Bois-Roux, Gahard (Ille-et-Vilaine), to the north-east of
Rennes, France (Phillipot 1957). Isolated basal cones have also been found at this

D. B. MACURDA JR.: DEVONIAN BLASTOID BELOCRINUS FROM FRANCE 251

locality; it is now under water (Phillipot 1957, p. 64). Phillipot also recorded them from
Joue’-en-Charnie and Brulon in the department of the Sarthe (p. 64) and gave the horizon
for all these localities as middle Siegenien (Lower Devonian) (p. 67). He stated that the
Gres de Gahard horizon given by Etheridge and Carpenter (1886, pp. 200, 203) was
incorrect.

REFERENCES

bassler, r. s. 1938. Pelmatozoa palaeozoica: (Generum et Genotyporum Index et Bibliographic).
Fossilium Catcilogus, 1, Animalia, Pars 83.

and moodey, m. w. 1943. Bibliographic and faunal index of Paleozoic Pelmatozoan echinoderms.

Spec. Pap. geol. Soc. Am. 45.

bather, f. a. 1899. The genera and species of Blastoidea, with a list of specimens in the British Museum
( Natural History): London.

delage, M. 1877. Stratigraphic des terrains primairies d'llle et Vilaine. Rennes.
etheridge, r., jr., and carpenter, p. h. 1883. Further remarks on the morphology of the Blastoidea,
with descriptions of a British Carboniferous genus and some new Devonian species from Spain.
Ann. Mag. nat. Hist., 5th ser., 11, 225-46.

1886. Catalogue of the Blastoidea. Brit. Mus. Cat., 322 p., 20 pi. London.

fay, R. o. 1961. Blastoid studies. Paleont. Contr. Univ. Kansas, Echinodermata, Art. 3, 147 p., 54 pi.

1964. An outline classification of the Blastoidea. Okla. Geol. Notes, 24, no. 4, 81-90.

macurda, d. b., jr. 1966. The ontogeny of the Mississippian blastoid Orophocrinus. J. Paleont. 40,
92-124, pi. 11-13.

munier-chalmas, e. 1876. Mollusques nouveaux des terrains paleozoiques des environs de Rennes.

J. Conch., Paris, 3rd ser., 16, 102-9.

1881. Bull. Soc. geol. Fr., 3rd ser., 9, 503.

oehlert, d. 1882. Crinoides nouveaux du Devonien de la Sarthe et de la Mayenne. Ibid. 10, 352-63,
pi. 8, 9.

phillipot, a. 1957. Contribution a l'etude de Metablastus cottaldi (M. Ch.) Eth. et. Carp. Bull. Soc.
geol. miner. Bretagne, 1, 64-68.

white, c. a. 1862. Description of new species of fossils from the Devonian and Carboniferous rocks
of the Mississippi Valley. Proc. Boston Soc. nat. Hist. 9, 8-33.

DONALD B. MACURDA, JR.

Museum of Paleontology,
University of Michigan,
Ann Arbor, Michigan

Manuscript received 8 February 1965

POPULATION STUDIES IN THE BALLYSHANNON
LIMESTONE, BALLINA LIMESTONE, AND RINN
POINT BEDS (VISEAN) OF N.W. IRELAND

by J. A. E. B. HUBBARD

Abstract. A numerical comparison of the populations of the outcrops formerly ascribed to the Ballyshannon
Limestone and its lateral equivalents at Easky, Aughris, Serpent Rock, Streedagh, Kiln Port and Shalwy is
presented. The localities define the limits of a subtrapezoidal area some 300 square miles in extent, in which the
maximum distance between stations is 33 miles (Easky to Kiln Port) and the minimum 6-5 miles (Kiln Port to
Shalwy). The four first named Sligo stations bear close similarities representing coral colonies, thickets, and even
local coppices. The last two named Donegal stations have a different fauna which is not as dense and may be
argued to represent either a different faunal province or, more probably, a different stratigraphical horizon.
Aughrus (Co. Donegal), near the type area of the Ballyshannon Limestone, is alluded to. It shows a distinctive
fauna with affinities to the other Donegal stations though proportionally the assemblage is unique.

Stratigraphically the Donegal stations are regarded as belonging to the Ballyshannon Limestone (C2 Sx),
while the Sligo stations are tentatively ascribed to the Benbulben Shale — Glencar Limestone (S2DP transitional
phase.

Palaeoecologically all the faunas are regarded as having developed marginally between the dysphotic tropical
belt and shallow warm waters of the open shelf. Of the Sligo stations Aughris is regarded as the most seaward
facies of a coral biostrome. A niche biota faunal replacement of syringoporids by fasciculate lithostrotiontids,
michelininiids by cerioid lithostrotiontids, narrow caniniids by giganteid forms, and davisiellids by linoproduc-
tids is postulated.

The Ballyshannon Limestone was first named and described by Oswald (1955, p. 168).
He mentions that at Streedagh Point its thickly bedded argillaceous limestones and
shales are extremely fossiliferous (p. 170). He continues by quoting Wynne (1864,
p. 38) who describes the caniniids as \ . . like stumps in a cabbage garden, and one is
almost disappointed to find that one can not pull them up; some of them are from 18
inches to 2 foot long and 2 to 3 inches in diameter’. Oswald adds that the description
is apposite but corals are more numerous than it suggests. He mentions (p. 170) that
similar beds are found at Serpent Rock; and cites another conspicuously fossiliferous
outcrop between Fairies Bridge and Aughrus (Co. Donegal).

Bowes (unpublished Ph.D. thesis, 1957, p. 52) compares the Ballina Limestone with
the Ballyshannon Limestone on lithological and faunal similarities, adding that their
outcrops are virtually contiguous. He further describes the prolific faunas of the Ballina
Limestone at Easky and Aughris.

George (in George and Oswald 1957, p. 143) places the Rinn Point Beds of Donegal
high in the Ballyshannon Limestone, mentioning that they are better exposed in the
Largymore syncline than on St. John’s peninsula. Both outcrops are regarded as repre-
sentatives of a mud-zone environment (p. 152). In the highest part of these beds,
described in the eastern flank of the Largymore syncline, he draws attention to the very
rich coral layers with large caniniids in a profusion comparable with that of Streedagh
Point and Serpent Rock in the Lissadell Peninsula (Oswald 1955, p. 170).

To each of the above descriptions a fossil list giving generic and specific details is
appended.

[Palaeontology, Vol. 9, Part 2, 1966, pp. 252-69, pis. 40-41.]

J. A. E. B. HUBBARD: POPULATION STUDIES, VISEAN, N. W. IRELAND 253

METHOD

Seven stations have been selected from extensive coastal exposures, chiefly in the
intertidal zone, in which the partially silicified fauna is particularly well preserved.
These localities are constituent of the previously designated Ballyshannon Limestone
outcrop and its lateral equivalents. At each location, excluding Aughrus (Co. Donegal),
100 one-metre quadrants measured with a piece of string anchored by four pebbles were

text-fig. 1 . Outline map of the Sligo-Donegal coast to show the location of the stations
from which the population of the Ballyshannon Limestone has been sampled.

sampled. In each quadrant the total fauna, including entire and fragmental forms, above
the size of comminuted crinoid debris, was counted. In order to obtain a representative
sample bedding planes were studied at each locality.

At Aughrus (Co. Donegal), which is alluded to on account of its proximity to the
type area of the Ballyshannon Limestone, it was found impracticable to sample more
than 50 one-metre quadrants all of which are derived from the same bedding plane. Thus
restricted, with only entire forms counted, no direct correlation may be drawn with this
area though its significance is evident.

In the analysis of the fauna a specific identification has not been attempted. The

254

PALAEONTOLOGY, VOLUME 9
table 1. Field counts on 100 one-metre quadrants

SOLITARY CORALS
caniniids

EASKY

AUGHRIS

SERPENT

ROCK

STREEDAGH

KILN

PORT

SHAWY

a

b

a

b

a

b

a

b

a

b

a

b

355

29

323

19

748

82

1050

82

91

8

329

49

zaphrentids

1

4

3

11

4

5

clisiophy Ilids

6

7

28

COMPOUND CORALS
lithostroti ontids ^

25

30

31

26

29

144

5

87

lithostrotiontids ^

740

550

861

375

lithostrotiontids “t

11

19

3

25

2

1

syringoporids^ J

3

58

syringoporids

16

4

syringoporids.

32

auloporids

2

7

17

25

47

14

18

75

13

2

micheliniids

30

1

BRACHIOPODS

davisiellids

1

580

57

chonetids

2

linoproductoids

521

6

110

24

81

19

pustulids

10

1

2

1

productids

2

128

49

5

4

orthotetids

3

19

3

32

5

12

leptaemds

10

4

spiriferids

1

14

29

7

1

rhynchonellids

2

1

brachiopods+

6

28

117

69

174

39

70

3

2

BRYOZOA

fenestellids

2

22

17

32

23

5

5

1

trepostomes

3

10

7

51

1

TRILOBiTES

5

2

1

MOLLUSCS

gastropods^

2

6

8

gastropods-,

3

3

23

25

orthocones

1

1

ECHINODERMS

echinoids

1

1

27

crinoid debris

GEN

ERAL

LY

PRES

ENT

THR

OUGH

OUT

worm burrows

GEN

ERAL

LY

PRES

ENT

THR

OUGH

OUT

The total population recorded in the field at Easky, Aughris, Serpent Rock, Streedagh, Kiln Port
and Shalwy is tabulated.

Column (a) represents entire forms, column ( b ) fragmental forms.

Lithostrotiontids! fasciculate with corallites approximately 5 mm. in diameter.

Lithostrotiontids2 fasciculate with corallites approximately 12 mm. in diameter.

This lithostrotiontid form was at first mistaken for fragmental material. The figures employed are
the results of extrapolation. The majority of coralla were noticed to be composed of fewer than half
a dozen corallites, often only two or four. To approximate the fragmental quota 10 per cent was

J. A. E. B. HUBBARD: POPULATION STUDIES, VISfiAN, N. W. IRELAND 255

material involved, though prolific, is somewhat intractably set in a massive matrix
which precludes detailed laboratory studies. Specimens are often only partially visible
(Plate 41, fig. 2) and sometimes crushed by compaction. Thus brachiopod hinges are
often buried or removed by erosion which would not allow for accuracy even at generic
level though the profile available is adequate for the family or subfamily level of classifi-
cation. Likewise it is argued that the selective sampling of corals would add no greater
accuracy to this study when it is borne in mind that no less than 6,562 corals have been
enumerated.

The faunal lists made available during the last decade by Bowes (1957, pp. 45-48),
George and Oswald (1957, pp. 172-4), and Oswald (1955, p. 172) present the generic
and specific details which are grouped in the present account. Of these Bowes’ is the
most localized, specifying Aughris as a collecting locality. Oswald (1955) unfortunately
does not localize his material, of which much appears to have been collected from
Streedagh Point and Serpent Rock, thus detracting from the significance of the Bally-
shannon Limestone fauna quoted.

The Sligo outcrops, mentioned hereafter, consist of relatively thick, irregularly
bedded, light grey biomicrite (Folk 1959) with a prolific coral-brachiopod fauna, and
occasional thin shale partings. The fauna, which is scattered throughout, is particularly
rich along certain planes. There is a tendency for one form to predominate in each
stratum notably caniniids (Plate 40, fig. 1), lithostrotiontids (cerioid: Plates 40, figs. 2,
3, or fasciculate: Plate 41, fig. 1), or brachiopods; though a certain amount of mixing
is to be found (Plate 41, fig. 2). The Donegal outcrops, by contrast, appear to be less
fossiliferous, darker, more regularly bedded, somewhat impure biomicrites with a less
silicified fauna.

The location of the stations (Easky, Aughris, Serpent Rock, Streedagh, Kiln Port,
Shalwy, and Aughrus (Co. Donegal)) is indicated in text-fig. 1, with the total population
as tabulated in the field in Table 1 and text-fig. 2. To simplify interpretation the genera
were subsequently subdivided into five groups which seemed to have ecological sig-
nificance: solitary corals, compound corals, brachiopods, bryozoa, and others which
are plotted on histograms to show the varying proportions of entire to fragmental forms
as a measure of turbulence.

Text-fig. 3 which represents the total numerical population shows a distinctly greater
density of fragmental forms at Streedagh and Serpent Rock; these are closely followed
by Easky in fragmental forms suggesting that these were the more turbulent areas. Kiln
Port and Shalwy have a decidedly less dense fragmental population than the Sligo

discounted (a sum to which no statistical significance should be attached). The remaining 90 per cent,
was divided by three to produce an approximate ‘entire’ quota. Thus in subsequent calculations the
following numbers are used: Easky 222a 74b, Aughris 166a 55b, Serpent Rock 268a 86b, Streedagh
1 13u 37b.

Lithostrotiontids3 cerioid forms.

SyringoporidSi with corallites approximately 2 mm. in diameter.

Syringoporids2 with corallites approximately 3 mm. in diameter.

Syringoporids3 with corallites approximately 4 mm. in diameter.

Brachiopods, unidentified short hinged.

Gastropods! fusiform.

Gastropods2 helicoid.

256

PALAEONTOLOGY, VOLUME 9

stations and in ‘entire’ density are also comparatively sparse, indicating an indigenous
fauna in less favourable though calm seas.

In text-fig. 4 the total population is expressed in a percentage histogram in which
a similarity of distribution in forms is just discernible though less evident than in text-
fig. 3. It is also to be noted that most specimens are entire, and of the fragmental fossils
brachiopods and compound corals are generally dominant.

FIELD COUNTS

AUGHRUS (Co. Donegal )

SOLITARY CORALS

zaphrentids

287

COMPOUND CORALS

auloporids

3

micheliniids

2

BRACHIOPODS

davisiellids

1

chonetids

2

pustulids

339

productids

69

rhipidomeltids

319

worm burrows, crinoid.
echinoid and fenestellid
debris ubiquitous.

TOTAL

POPULATION

PERCENTAGE OF
POPULATION

PERCENTAGE OF
INVERTED
BRACHIOPODS

100 1

■■upright

C I INVERTED

text-fig. 2. The population of Aughrus (Co. Donegal). Field counts record the entire forms counted
in 50 one-metre quadrats restricted to one bedding plane. Histogram of the total numerical population
showing an anomalously high population which can be related to the selector bias of sampling only one
bedding plane. Percentage histogram of total population resembling Kiln Port (text-fig. 5).

1, solitary corals. 2, compound corals. 3, brachiopods. 4, bryozoa. 5, others.

In text-fig. 5 a similarity in proportions of fragmental population is seen at Easky and
Streedagh, while Shalwy and Serpent Rock show similar proportions for entire forms as
also seen in text-fig. 4.

In text-fig. 6 the percentage of fragmental material within each group illustrates the
abundance of fenestellid debris at Aughris, Serpent Rock, Streedagh, and Kiln Port.
Aughris and Serpent Rock show a similar general population reflecting similar con-
ditions of turbulence affecting the area despite initial differences in the faunal pro-
portions seen in text-fig. 4. Shalwy, as in text-fig. 3, appears to have a comparatively
small fragmental population in which the chief constituent is gastropods.

Inversion ratios shown in text-figs. 7 and 8 suggest mild turbulence in all areas, as
more than 75 per cent, of brachiopods are inverted at all stations; while the corals are
dominantly upright. Among the brachiopods no subfamily is present in all areas. In
order to approximate similar hydrodynamic properties the larger forms of comparable
profile were selected. Of necessity this resulted in the following forms being studied:
linoproductids at the four Sligo stations, davisiellids at Kiln Port and Shalwy, and

J. A. E. B. HUBBARD: POPULATION STUDIES, VISfiAN, N.W. IRELAND 257

pustulids at Aughrus (Co. Donegal). The state of preservation did not allow for the
determination of whether or not the valves were joined, but since adductor muscles

TOTAL FRAGMENTAL POPULATION

EASKY AUGHRIS SERPENT ROCK STREEDAGH KILN PORT SHALWY

TOTAL ENTIRE POPULATION

1100 -|

1000-
900-
800 -

ENTIRE

FRAGMENTAL

EASKY
AUGHRIS
SERPENT ROCK
STREEDAGH
KILN PORT
SHALWY

text-fig. 3. Histograms of the total numerical population, showing the dominance of entire over
fragmental forms. On the fragmental population histogram Streedagh, Serpent Rock and Easky, in
order of declining importance, show the greatest density. A similarity of proportions for the entire
population is to be seen at Shalwy and Serpent Rock.

1, solitary corals. 2, compound corals. 3, brachiopods. 4, bryozoa. 5, others.

contract on death it is unlikely that this would be of great significance. In relation to the
shell profile the brachiopods are described as upright when the pedicle valve is convex
downwards on the bedding plane. Aughrus (Co. Donegal) would appear to be unique

258

PALAEONTOLOGY, VOLUME 9

in this study (text-fig. 7). The significance of the data would seem to be related to the
number of sturdy spines for anchorage present. These are well developed only in the
pustulids, while the linoproductids and davisiellids are unattached forms more suscep-
tible to current traction. Of the compound corals the cerioid lithostrotiontid with the

PERCENTAGE OF TOTAL POPULATION

EASKY AUGHRIS SERPENT ROCK STREEDAGH KILN PORT

SHALWY

100

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

text-fig. 4. Percentage histogram for the total population showing the dominance of entire over
fragmental forms. In comparison to the data expressed in fig. 2 the significance of given forms is

accentuated.

1, solitary corals. 2, compound corals. 3, brachiopods. 4, bryozoa. 5, others.

wider apical angle and generally more massive corallum is more stable than the fascicu-
late lithostrotiontid. Similarly the syringoporids of Shalwy are seen to be comparable
with the fasciculate lithostrotiontids as a measure of stability.

The proportion of geniculate to straight caniniids shown in text-fig. 9 suggests that
attachment of these forms to a soft lime substratum would not be sufficient to support
the giganteid forms, which, while sinking, would tend to grow upwards towards the
light. Three horizons are given for comparison at Streedagh (text-fig. 9) showing
a uniformity which is not seen in the coral inversion ratios (text-fig. 8) for the same

J. A. E. B. HUBBARD: POPULATION STUDIES, VISfiAN, N. W. IRELAND 259

area. The relationship between the straight and geniculate caniniids would appear to
have no direct correlation with the size of the corallum or the species present.

The details of the Aughrus (Co. Donegal) (text-fig. 2) fauna show closest similarity
with those of Kiln Port (text-fig. 3). Though the faunal list makes a striking contrast

PERCENTAGE OF FRAGMENTAL POPULATION

PERCENTAGE OF ENTIRE POPULATION

1 2 3 A 5 1234 5 12345 12345 12345 12345

EASKY AUGHRI5 SERPENT ROCK STREEDAGH KILN PORT SHALWY

text-fig. 5. Percentage histograms of fragmental and entire populations. The reverse order of im-
portance of fragmental material is to be seen at Serpent Rock and Streedagh, and of entire material
at Streedagh and Kiln Port. Similar distribution for entire material is seen at Serpent Rock and

Shalwy.

1, solitary corals. 2, compound corals. 3, brachiopods. 4, bryozoa. 5, others.

this can, in part, be attributed to the selector bias involved in sampling only one bedding
plane. Caniniids and syringoporids, though not abundant, are to be found at this
locality. Likewise euomphallids and Conocardium, which are not listed, also occur at
the Donegal stations, though the latter is absent from the Sligo stations.

To attempt current analysis, orientations of crinoid stems from localities near
the stations were recorded as well as umbonal orientations in inverted productoid

260

PALAEONTOLOGY, VOLUME 9

FRAGMENTAL PERCENTAGE OF EACH GROUP

text-fig. 6. Percentage histogram of fragmental material expressed in terms of each group. A notable
abundance of fragmental bryozoan material is to be seen at all stations except Shalwy. Aughris and
Serpent Rock show similar population distributions.

1 , solitary corals. 2, compound corals. 3, brachiopods. 4, bryozoa. 5, others.

BRACHI0P0D INVERSION PERCENTAGES

INVERTED □ UPRIGHT

1007.1

90 -

80 -

70 -

60 4

EASKY AUGHRIS SERPENT STREEDAGH KILN SHALWY AUGHRUS

ROCK PORT (DONEGAL)

text-fig. 7. Inversion ratios of productoid brachiopods (linoproductids at Easky, Aughris, Serpent
Rock and Streedagh; davisiellids at Kiln Port and Shalwy; pustulids at Aughrus (Co. Donegal))
expressed in terms of percentage histograms. A dominance of inverted forms is to be noted at all

stations except Aughrus (Co. Donegal).

J. A. E. B. HUBBARD: POPULATION STUDIES VISfiAN, N. W. IRELAND 261

brachiopods, and the orientation of geniculate caniniids. The results are variable and
inconclusive but never suggest transport from the south, thus generally supporting the
palaeogeographical reconstruction presented by George (1958). This pattern is one
which might be expected in an area with such a uniform population density and lack of
evidence of channeling.

text-fig. 8. Inversion ratios of compound corals expressed as percentage histograms. Cerioid litho-
strotiontids shown for Easky, Aughris, and Serpent Rock are dominantly upright. Fasciculate litho-
strotiontids at Easky, Aughris, Serpent Rock and two horizons at Streedagh, and syringoporids of
Shalwy show a less pronounced dominance of upright forms.

ENVIRONMENT

The fauna outlined above may be described in terms of Hill’s (1948, p. 121) three coral
associations. She showed these to have different morphological characters which varied
according to the environment. Thus by analogy with present day faunas she reaches the
conclusions tabulated (Table 2).

In text-fig. 10 the populations of this account are plotted on a triangular diagram

262

PALAEONTOLOGY, VOLUME 9

whose apices represent Hill’s coral associations. This diagram supports her statement
that solitary Rugosa with dissepiments are found associated with the compound Rugosa,
but rarely with the Cyathaxonia fauna. In generalized terms the total fauna may be
ascribed to her intermediate fauna of solitary Rugosa. However, when the beds are
studied more closely, it would appear that there are alternations of the compound and

PERCENTAGE OF GENICULATE CANINIIDS

100 n

EASKY AUGHRIS SERPENT ROCK a b

STREEDAGH

text-fig. 9. Ratio of straight to geniculate caniniids expressed in terms of a percentage
histogram for Easky, Aughris, Serpent Rock, and three horizons at Streedagh. At all
stations except Easky geniculate forms are dominant, notably at Streedagh.

solitary Rugosa faunas within a few feet as illustrated in text-fig. 11, and Plates 40
and 41.

table 2. Hill’s (1948) three coral associations tabulated to show the relationship between the faunal
characteristics and the modern analogy and environment.

FAUNAL CHARACTERISTICS

MODERN ANALOGY AND ENVIRONMENT

I Cvathaxonia Fauna

small, solitary, non-dissepimented
forms.

small, solitary, corals
cold, deep or murky sea.

II Compound Ruaosa & Chaetetida

reef corals.

warm, shallow, pellucid sea

III Solitary Rugosa with dissepiments
(chiefly caniniids & clisiophyllids.)

intermediate between 1 & 11.

The palaeoecological conclusions quoted by Wells (in 1957, p. 774) have, until
recently, been accepted widely for the limitation of rugosan habitats. Thus by analogy
to present day, non-surface, essentially lagoonal reef corals he states that the ecological

J. A. E. B. HUBBARD: POPULATION STUDIES, VISEAN, N. W. IRELAND 263

niche has the following parameters: (1) a maximum depth of about 50 metres; (2) well
within the pellucid zone; (3) annual minimum temperature of 16-21° C; (4) well oxygen-
ated, gently circulating waters ; (5) on a substratum clear or relatively free from rapid
sedimentation, but not necessarily in clear non-turbid waters.

T • entire

text-fig. 10. The coral populations of Easky, Aughris, Serpent Rock, Streedagh,

Kiln Port, Shalwy, and Aughrus (Co. Donegal) are expressed in terms of Hill’s
three coral associations showing that the entire forms belong to the solitary Rugosa

association.

I. Cyathaxonia fauna: small, solitary, non-dissepimented forms.

II. Compound Rugosa and Chaeteticla.

III. Solitary Rugosa with dissepiments (chiefly caniniids and clisiophyllids).

Teichert (1958, p. 1064), however, reminds us of Dons’s (1933) and Shott’s (1926)
works in giving an account of cold deep water coral banks which are alive today off the
coast of Norway. Thus he divides the modern scleractinian corals into two major
ecological groups: the first of these, the Hermatypic, are dependent on the symbiotic
flagellates, the zooxanthellae, and consequently restricted to an annual minimum water
temperature of 18-5° C, and 300 ft. depth. Teichert’s second group of Ahermatypic
scleractinian corals are not dependent on zooxanthellae, and therefore are not restricted
to the pellucid zones. These Hermatypic conditions would appear to be similar to those
outlined by Wells (op. cit.). The degree of tolerance of the Ahermatypic group is wide,
members having been recorded at depths of up to 20,000 feet, and temperatures as low
as — 1T° C, though the majority occur at depths of 600 900 ft.

Teichert (1958, p. 1075) concludes that the best measure of depth is the presence
of calcareous algae which are not found below 150 ft. From comparison with the

C 3803

T

264

PALAEONTOLOGY, VOLUME 9

environments and biota tabulated (p. 1076) it would appear that the fauna described
herein probably belongs to Teichert’s category number 1. In this few species of frame
building coelenterates with rich invertebrate association and absence of calcareous algae
are recorded. This category is found in deep and cool waters, shelf or bay environments,
within the temperature range of the dysphotic zone of the tropical belt to the polar
regions. This contrasts with his category number 4, in which a rich assemblage of frame
building coelenterates, in association with a rich invertebrate fauna, with the addition
of abundant calcareous algae, is attributed to shallow warm waters in open shelf to near
shore environments of the tropical belt. Calcareous algae are not obvious in the Irish
coral limestone populations described here, though their presence has been noted else-
where in the succession both above and below this horizon. This would suggest that
the environment was probably marginal to the tropical dysphotic belt.

The development of conspicuously fossiliferous bedding planes alternating with less
fossiliferous bands has been attributed by Broadhurst (1964, pp. 866 and 867) to the rate
of sedimentation and the associated amount of bottom turbulence in the non-marine
Coal Measures. On this view each fossiliferous plane represents a phase of restricted
sedimentation. Teichert (1958, p. 1073) states that bioherms may show signs of sudden
extinction, in which the destruction of the coral frame results from the attack of boring
organisms, yet evidence of wave destruction is absent. He continues by saying that
though this might be regarded as evidence of subsidence and consequent ‘drowning’
of the reef, it might equally well result from the elevation of the sea floor above the
level of tolerance of the corals. In these beds there is evidence of extensive burrowing in
both horizontal and vertical directions. From the infill of the burrows it is apparent
that there was a variation in the nature of the sediment supplied. There is also abundant
evidence of auloporids encrusting brachiopods, caniniids, and even occasional syringo-
porids, not infrequently in the position of growth, at Shalwy, Kiln Port, and Easky.
This might be interpreted as a sign of a decline in sedimentational rate.

The fasciculate lithostrotiontid associations may chiefly be described as thickets,
though occasionally there is evidence of coppice (Plate 41, fig. 1) development (Squires
1964, pp. 904-6). An unusual relationship is seen in one place at Serpent Rock where
caniniids have been observed penetrating a fasciculate lithostrotiontid colony (Plate 41,
fig. 3); this appears to be a commensal association. A similar relationship between two
lithostrotiontids is seen high in the succession at Streedagh (text-fig. 1 1).

The cerioid lithostrotiontids, by contrast, seem to have an unsociable habit, commonly
occurring on separate bedding planes from the other faunas where there is a distinct
paucity of other fossils save trace fossils. The individuals thus appear as well separated
colonies (Plate 40, figs. 2 and 3) in bioturbated biomicrites. An exception to this rule is
seen near Easky pier where the coral brachiopod fauna is notably mixed, containing
lithostrotiontids (fasciculate and cerioid), pustulids, chonetids, productids, and even
rare caniniids which would suggest that the fauna may in part be transported.

The population is summarized in terms of the dominant ecologically significant
groups in text-fig. 12. In entire population close similarities are seen at Shalwy, Stree-
dagh, Serpent Rock characterized by a dominance of solitary corals, and few compound
corals and brachiopods, indicating calm well aerated shallow seas. Aughris and Easky
have fewer solitary corals but have more compound corals and brachiopods respectively,
suggesting that more stable forms were tolerant of a somewhat greater turbulence. Kiln

J. A. E. B. HUBBARD: POPULATION STUDIES, VISEAN, N. W. IRELAND 265

-g-L^f-ea-L

T 1 ( i

<321

I ithostrotiontid thicket

(TOP)

linoproductoids dominant
zaphrentids, auloporids,
trepostomes, lithostrotiontids
(inverted),
lithostrotiontids, caniniids,
linoproductoids, spiriferids,
trepostomes, phillipsiid.
occasional linoproductoid , leptaenid.
caniniid clusters, productids, ortho-
auloporids, spiriferids, tetids.

orthotetids, linoproductoids,
\lithostrotiontids (some inverted),
lit host ro t ion t id dominant phillipsid.
occasional zaphrentid, micheliniid,
auloporid trepostome spiriferid 8.
lithostrotiontid] productid.

thicket, rare, inverted coralla.
occasional inverted caniniids.
commensal association of lithos-
trotiontids j 'n thicket, rare
caniniid, linoproductoid, spiriferid
inoproductoids & trepostoitie,
prolific, auloporid encrusting
\caniniids, linoproductoid, leptaenids,
dense caniniid clusters,
prolific caniniids subsidiary lithos-
trotiontid2. linoproductoids,
pustulids, orthotetids, spiriferids,
^rhynchonellids & trepostomes.
caniniids, occasional ostracod,
linoproductoid, inverted lithos-
trotiontid], rare lithostrotiontid2-

caniniids dominant, ostracods,
crinoid ossicles, linoproductoids.

lithostrotiontid] thicket.

caniniids abundant , subsidiary
inverted lithostrontid j coralla.

caniniids dominant occasional
inverted lithostrotiontids coralla

"A>

caniniids dominant occasional
inverted lithostrostrotiontid.

"biostrome of actinoconchids
\lithostrotiontids, trepostomes

fusiform gastropods, euomphalid

linoproductoids & caniniids.

r72"

-60"

-48"

caniniids dominant

36"

24"

j/

J< %,

f ^ -J?

sparse fauna: (ithostrotiontid,
auloporid, euomphalid, zaphrentid,
productid, phillipsiid.

caniniids & lithostrotiontids.

linoproductoids & caniniids.

(BASE)

.ffitelto,* s^n fendiS;o^hPo°lhdTC^^^ fasciculate lithostrotiontids.

... , . . . phillipsiid

linoproductoids, ortho-

fasciculate li thostro t ion t id

.tetids, pustulids, £Plrif?dri^stlha||||'°P^-^ ^ cerioid lithostrotiontid.

cerioid lithostrotiontids, rare &= £ caninioid.

^ ? 9 zaphrentoid,

£ ^ <3 <3 <2 productids & chonetids.
© & 3 gastropods.

2-

pustulid & lithostrotiontid.
caniniids.

calcareous shale
shaly biomicrite

wavy bedded
biomicrite with
J shaly partings.

wavy bedded
biomicrite

text-fig. 1 1 . Diagrammatic representation of the lithology of the coral limestone horizon within the
Ballyshannon Limestone of Streedagh to illustrate the rapid alternation of Hill’s coral associations.

266

PALAEONTOLOGY, VOLUME 9

Port and Aughrus (Co. Donegal) are anomalous in showing a marked preponderance
of brachiopods over all other forms. The fragmental populations at Easky, Streedagh
and Serpent Rock have marked similarity with a decrease in the number of solitary

BRACHIOPODS

text-fig. 12. The population of the Ballyshannon Limestone summarized in terms
of the three dominant ecologically significant groups: brachiopods, compound
corals and solitary corals. A close similarity is seen between the entire forms of
Shalwy, Aughris, Streedagh, and Serpent Rock, while a secondary, less conspicuous
group is formed by Easky, Kiln Port, and Aughrus (Co. Donegal). Of the frag-
mental population Streedagh, Easky, and Serpent Rock are particularly similar,
with Kiln Port and Aughris showing a certain amount of similarity while Shalwy

remains anomalous.

EXPLANATION OF PLATE 40

Fig. 1. Portion of a bedding plane at Streedagh showing prolific partially silicified caniniids with
variable degree of geniculation, set in a biomicrite matrix. Scale: hammer measures 13 inches.

Fig. 2. Bedding plane at Serpent Rock showing well spaced cerioid lithostrotiontid colonies appearing
as lighter coloured patches in the jointed biomicrite matrix. Scale: hammer 13 inches long.

Fig. 3. Close-up of a cerioid lithostrotiontid colony in situ with horizontal burrow to the left of it,
from the same bedding plane as fig. 2 at Serpent Rock. Scale: coin measures one inch in diameter.

EXPLANATION OF PLATE 41

Fig. 1 . Fasciculate lithostrotiontid coppice showing fragmental lithostrotiontid corallites in the lower
portion of the stratum from which one large upright fasciculate lithostrotiontid corallunr pro-
trudes. Scale: hammer measures 13 inches.

Fig. 2. Mixed caniniid, fasciculate lithostrotiontids x & 2 and brachiopod fauna, worn smooth by wave
action; illustrating the impracticability of detailed identification of many intractable forms. Scale:
half the magnification of fig. 1 .

Fig. 3. Apparently commensal association of caniniids and lithostrotiontid thicket at Serpent Rock.
Scale : coin measures one inch in diameter.

Palaeontology, Vol. 9

PLATE 40

HUBBARD, Ballyshannon Limestone

Palaeontology, Vol. 9

PLATE 41

HUBBARD, Ballyshannon Limestone

J. A. E. B. HUBBARD: POPULATION STUDIES, VISEAN, N. W. IRELAND 267

corals present, possibly indicating that the solitary corals are more resistant on account
of their thicker skeletons. Aughris likewise shows a proportional decrease in solitary
corals. Kiln Port is once more anomalous in showing an increase in the number of both
compound and solitary corals present. This may represent a fauna which has been trans-
ported further from the source of coralline material. It may be suggested that Streedagh,
Serpent Rock, and Shalwy represent solitary coral biostromes in quiet seas while
Aughris was on the rougher seaward side sustaining a greater proportion of hydro-
dynamically more stable forms. Kiln Port may represent more muddy seas with a
vagrant benthonic fauna of brachiopods which developed on the landward side of the
coral biostrome where a more restricted circulation would not favour coral growth.

Easky, Serpent Rock, and Streedagh may thus be regarded as an inshore, and Aughris
a further seaward facies of a coral biostrome. These conclusions support the palaeo-
geography postulated by George (1958).

CONCLUSIONS

Variations in environment are clearly not the sole factor since there are faunal dif-
ferences between the Donegal stations and those of Sligo. Thus in general the Donegal
fauna consists of syringoporids, micheliniids, a relatively narrow caniniid and davisiel-
lids which are replaced by fasciculate and cerioid lithostrotiontids, giganteid caniniids
and linoproductids in Sligo. It is not yet known whether these represent niche biotas of
different faunal provinces or whether they are at different horizons, though the latter
is suspected. In Sligo the former association can be traced low in the succession.
Additional support for the latter argument has been reached independently on gross
lithological grounds by W. F. Hubbard and D. J. R. Sheridan (personal communication)
and on detailed foraminiferal studies by R. W. L. Oldroyd (personal communication).
However, until further detailed macropalaeontological and micropalaeontological
dating elucidates this problem, it could be argued that this faunal replacement results
from the diachronous northward migration of faunas.

It would appear highly probable that the Sligo outcrops are of approximately the
same age but of a different age to the Donegal outcrops. In support of this hypothesis
the dating of the underlying Carrowmoran Sandstone found in the ground between
Easky and Aughris is crucial. The Carrowmoran Sandstone has been shown by the
author (1966, in press) to correlate with the Mullaghmore Sandstone of Sligo, to which
an age between Sa-D! has been ascribed by Oswald (1955, pp. 173, 174, and 180) on
the basis of lithological comparison and ostracod dating. Furthermore the Ballyshannon
Limestone of the coastal regions of Sligo from which these population studies are
derived is shown to contrast strongly with that of the type area, yet it resembles higher
stratigraphical horizons only six miles away. In the type area about Ballyshannon the
Ballyshannon Limestone consists of massive crinoidal biomicrites, biosparites, and
oomicrites containing mainly fragmental or transported fossils among which cerioid
lithostrotiontids are not recorded. The lowest record of cerioid lithostrotiontids outside
Oswald’s (1955) area is in S2 beds of Garwood (1913, p. 547). The second lithological
comparison concerns Oswald’s (1955) transitional Benbulben Shale-Glencar Lime-
stone (SoDx) exposed on the south-western flanks of Benbulben Mountain, where four
recognizable units, one of which is the coral-brachiopod limestone mentioned, can be

268

PALAEONTOLOGY, VOLUME 9

correlated with similar units at Streedagh and Serpent Rock. Similarly material of com-
parable lithology to the type section of the Ballyshannon Limestone has been recorded
on bore-hole cores which penetrate the basement to the south of the outcrops herein
described. Thus it would appear that the Sligo outcrops under discussion are referable
to Oswald's (1955) Benbulben Shale-Glencar Limestone (SoD^.

Reference to the faunal lists of Bowes (1957), George and Oswald (1957) and Oswald
(1955) shows that the following corals are not present in the south-eastern flank of the
Donegal syncline but are to be found in Sligo. At Aughris: Caninia benburbensis Lewis
and Lithostrotion pauciradiale (McCoy). In Oswald’s (1955) Ballyshannon Limestone
list derived from the total outcrop: Zaphrentis ( Hapsiphyllum ) konincki (Edwards and
Haime), Lithostrotion affine Fleming, Lithostrotion martini Edwards and Haime, and
Lithostrotion portlocki (Bronn). To the south of this area Caldwell (1959, pp. 168, 180)
does not record these species below the Ballymore Beds which he correlates with the
Benbulben Shale.

In contrast to the situation in Sligo the Donegal outcrops are known to be relatively
close to the base of the Carboniferous succession and do not contain cerioid lithostro-
tiontids, even fasciculate lithostrotiontids being extremely scarce.

Environmentally it would appear that the Sligo fauna represents a similar ecological
niche to that of Donegal; having developed marginally between the dysphotic tropical
belt and shallow warm waters of the open shelf. The Sligo fauna is not present in the
Donegal assemblage, yet occasionally Donegal members are still present in the Sligo
fauna indicating that the dominant incoming fauna had not entirely excluded the earlier
forms. It is suggested that there is a general replacement of syringoporids by fasciculate
lithostrotiontids, narrow by giganteid caniniids, and davisiellids by linoproductids.
Each of these pairs, being analogous forms, would appear to occupy the same ecological
niche. The greater density and diversity of the Sligo fauna would suggest the attainment
of more favourable conditions during the later Visean, by which time the Carboniferous
marine transgression had become established.

Acknowledgements. The author is indebted to Dr. G. M. Dunlop and Professor T. Neville George
for the interest and encouragement shown during the course of the work; to Mr. J. W. Keith for
technical assistance; and to Dr. C. H. Holland and Professor B. C. King for critically reading the
manuscript, and the latter for supervision of the work which was carried out under the tenure of a
D.S.I.R. Studentship at Bedford College (University of London).

REFERENCES

bowes, g. 1957. Geology of the Ballina Syncline. Unpublished Ph.D. thesis, University of Glasgow.
broadhurst, f. m. 1964. Some aspects of the palaeoecology of non-marine faunas and rates of sedi-
mentation in the Lancashire Coal Measures. Amer. J. Sci. 262, 858-69.
caldwell, w. g. e. 1959. The Lower Carboniferous rocks of the Carrick-on-Shannon syncline.
Quart. J. geol. Soc. Lond. 115, 163-88, pi. Vi.

dons, c. 1933. ‘Zoologiske Notisser XVII. Om Breitstadtfjordens korallrev’. Kong. Norske Viden-
skab. Selsk. Forhandl. 5, 143-6.

folk, r. l. 1959. Practical petrographic classification of limestones. Bull. Amer. Assoc. Petrol. Geol.
43, 1-38.

garwood, e. J. 1913. Lower Carboniferous succession in the North-West of England. Quart. J.
geol. Soc. Lond. 68, 449-572.

J. A. E. B. HUBBARD: POPULATION STUDIES, VISEAN, N.W. IRELAND 269

george, t. n. and Oswald, d. h. 1957. The Carboniferous Rocks of the Donegal Syncline. Quart.
J. geol. Soc. Loud. 1 13, 137-79, pi. xiv-xvi.

1958. Lower Carboniferous Palaeogeography of the British Isles. Proc. Yorks, geol. Soc. 31,

227-318.

hill, d. 1948. The distribution and sequence of Carboniferous coral faunas. Geol. Mag. 85, 121-48.
hubbard, j. a. e. b. 1966. Facies patterns in the Garrowmoran Sandstone tVisean) of western County
Sligo, Ireland. Proc. geol. Assoc, (in press).

Oswald, d. h. 1955. The Carboniferous Rocks between the Ox Mountains and Donegal Bay. Quart.
J. geol. Soc. Loud. Ill, 167-86, pi. xi.

shott, g. 1926. Geographic des Atlantischen Ozeans. 2nd ed. xiv, 368 pp. 5 pi. C. Boysen, Hamburg.
squires, d. f. 1964. Fossil coral thickets in Wairarapa, New Zealand. J. Paleont. 38,904-15.
teichert, c. 1958. Cold and deep water coral banks. Bull. Amer. Assoc. Petrol. Geol. 42, 1064-82.
wells, j. w. 1957. In Treatise on Marine Ecology and Palaeoecology, edited by H. S. Ladd, Mem.
Geol. Soc. Amer. 67 (2), ‘Corals’, 773-4.

Wynne, a. b. 1864. On the geology of parts of Sligo, etc. Journ. Geol. Soc. Dublin, 33-41.

JULIA A. E. B. HUBBARD,

King’s College (University of London),
Strand,

Manuscript received 16 February 1965 London, W.C.2

OCCURRENCE OF AUSTRALO SUTURA
(TRILOBITA) IN THE MISSISSIPPIAN OF
OKLAHOMA, U.S.A.

by ALLEN R. ORMISTON

Abstract. A species of Australosutura is described from Mississippian rocks of Oklahoma regarded as being
Visean in age. It is most like the earlier-described Australian species, less like that from Argentina.

A trilobite collected by Dr. Kenneth W. Ciriacks in the course of his investigation
of the Mooreheld Formation of Oklahoma was identified by me as Australosutura. In
the company of Dr. Ciriacks I visited the locality from which the original specimen came
and was able to collect additional material. These Oklahoma specimens are the first
of the genus in North America.

So far as the writer is aware, Australosutura has previously been known to occur only
in Australia and Argentina. Amos, Campbell, and Goldring (1960, p. 229) erected the
genus for the single species Cordania gardneri Mitchell 1922, originally described from
the Carboniferous of Australia. A. gardneri was recognized by them in collections from
Argentina as well as from Australia. Another Australian occurrence of the genus was
reported by Campbell and Engel (1963, p. 110).

Acknowledgements. 1 am indebted to Pan American Petroleum Corporation for permission to publish
this paper. I thank Dr. K. W. Ciriacks of the Research Center of Pan American Petroleum Corporation
for making available to me the first Oklahoma specimen of Australosutura and for his help in collecting
more material. Dr. Gilbert Klapper, also of the Research Center, kindly provided age information
based on his studies of selected conodont faunas from the Moorefield Formation. My wife, Eleanor,
collected the only pygidium of Australosutura we have been able to find.

Stratigraphy. The north-eastern Oklahoma exposures of the Mississippian Moorefield
Formation were divided by Huffman (1958, p. 49) into four members which are in
ascending order: the Tahlequah, Bayou Manard, Lindsey Bridge, and Ordnance Plant
Members. All the specimens of Australosutura described in this paper come from a
6-in. bed 8-7 to 9-2 ft. below the top of the Bayou Manard Member exposed in the bed
of a dry stream 0-5 miles south of Welling, Oklahoma (exact locality information below).
At this locality the Moorefield Formation is incompletely exposed ; the lowest beds being
within the Bayou Manard Member. Approximately 30 miles north-west of the Welling
locality there is a complete exposure of the Moorefield Formation at Lindsey Bridge.
There the entire formation is 107 ft. thick and the Bayou Manard Member 53 ft. thick.
Although the Welling exposure is incomplete, the Bayou Manard Member appears to
be considerably thinner there than at Lindsey Bridge.

At Welling the Bayou Manard Member consists of dark gray to black, argillaceous,
micritic limestones, and calcareous shales.

Bayou Manard Fauna. The bed from which Australosutura was collected yields a
fairly diverse fauna including brachiopods, pelecypods, and trilobites. By far the most

{Palaeontology, Vol. 9, Part 2, 1966, pp. 270-3, pi. 42.]

A. R. ORMISTON: OCCURRENCE OF AUSTRALO SUTU RA (TRILOBITA) 271

abundant elements are Cleiothyridina sp. and the trilobite, Griffithides pustulosus Snider
1915. Less common are: Spirifer arkansanum Girty 1911 ; Neoehonetes sp.; two genera
of productids ; and ?Sphenotus sp.

Age. According to Dr. Gilbert Klapper (personal communication) the conodont fauna
present in the Australosutura- bearing bed of the Bayou Manard Member belongs to the
Gnathodus bilineatus-Cavusgnathus characta Zone of Collinson et al. (1962). This bed
of the Bayou Manard Member is thus Visean in terms of the European chronology
(see Collinson et al. 1962, p. 13, chart 5).

The Argentinian occurrence of A. gardneri (Mitchell) has also been suggested to be
Visean in age, whereas the Australian occurrences are considered Westphalian (Amos,
Campbell, and Goldring 1960, p. 229). The disparity in age between the Argentine and
Australian material of A. gardneri led Campbell and Engel (1963, p. 112) to suggest that
those occurrences represent separate species. Elsewhere in this paper are noted mor-
phologic differences which support this suggestion, but considering the Visean Okla-
homa Australosutura to be closely related to the Westphalian A. gardneri from Australia
implies that the species really is long ranging.

SYSTEMATIC DESCRIPTION

Family brachymetopidae Prantl and Pribyl 1951
Genus australosutura Campbell and Goldring 1960

Type species. Cordania gardneri Mitchell 1922.

Australosutura aff. A. gardneri (Mitchell 1922)

Plate 42, figs. 1-9

Material. USNM No. 145307, an incomplete cephalon lacking the left free cheek; USNM No. 145308,
a small partly exfoliated cranidium; USNM No. 145309, a fragmentary cephalon preserving the left
genal spine; and USNM No. 145310, a pygidium preserved as an internal mold.

Locality and Horizon. A bed 8-7 to 9-2 feet below the top of the Bayou Manard Member, Moorefield
Formation, bank of a dry stream which runs generally north-south through the pasture of Mrs. S.
Bracket’s farm, SE 1/4, SE 1/4, section 7, T. 16 N., R. 23 E., Cherokee County, Oklahoma.

Description. There are no consistent differences in the cephalon between the Oklahoma
and Australian specimens (Amos, Campbell, and Goldring 1960, p. 234, pi. 40) of
Australosutura with the possible exception of one minor character. The tubercles present
along the posterior slope of the anterior and lateral cephalic borders appear to be more
prominent on the Oklahoma specimens.

The Argentinian specimens assigned to A. gardneri by Amos, Campbell, and Goldring
(1960, p. 230) do, however, differ importantly from the Oklahoma material. They are
distinguished in having: (1) the cephalon more strongly inflated, (2) the anterior part
of glabella rising vertically in lateral profile, (3) the frontal area distinctly shorter (sag.),
and (4) the glabellar tuberculation finer than that elsewhere on the cephalon.

The single available pygidium is preserved as an internal mould. The description is

272 PALAEONTOLOGY, VOLUME 9

partly based on a latex cast of a fragmentary counterpart external mould of the same
specimen.

Pygidium wider than long. Axis with 15 rings, the last two of which are represented
by aligned tubercles. Each ring with 9 to 11 tubercles arranged in longitudinal rows,
median tubercles the most prominent. Ring furrows deep.

In lateral profile axis gently declining to position of sixth ring, more steeply declining
thereafter. Axis stops short of posterior margin and overhangs post-axial area. Anterior
axial rings inclined forward, posterior ones become progressively more backwardly in-
clined posteriorly.

In posterior profile axis rises high above pleural regions, semi-octagonal in cross
section. Pleural field rises gently from axial furrow but is flexed strongly downward at
about f total width of field from axial furrow. The fulcrum is marked by a longitudinal
(exs.) row of prominent tubercles.

Pleural region subdivided by ten prominent posterior pleural bands (Whittington,
1960, p. 406), the posteriormost two of which are nearly parallel to the axis. The posterior
bands persist to the pygidial margin where each is represented by a marginal tubercle.
The anterior pleural bands are visible only on the first five segments of the internal
mould. Posterior pleural bands about twice as long (exs.) as anterior bands and more
strongly convex. There are 5 to 8 evenly spaced tubercles on posterior bands 1 through 8.
Interpleural furrows very faint on internal mould.

There is no border furrow, the border being set off by a change in slope.

Dimensions (in mm.)

USNM 145307

USNM 145308

USNM 145309

cranidial length

80

6-9

—

glabellar length plus occip. ring

5-5

4-9

—

glabellar length to occip. furrow

4-7

41

—

max. glabellar width

5-0

40

6-3

beta-beta width of facial sutures

9-4

6-3

—

length of frontal area

2-6

20

—

cephalic width

(15-1)

—

(17-4)

USNM 145310

max. axial width

4-6

pygidial width

12-5

pygidial length

10-2

Discussion. Because of the widely different ages of the Australian and Argentinian
occurrences of A. gardneri, Campbell has proposed (Campbell and Engel 1963, p. 112)
that the Argentine specimens now be assigned to a separate species. I agree with this

EXPLANATION OF PLATE 42
Australosutura aff. A. gardneri (Mitchell).

Figs. 1-5. Cephalon lacking the left free cheek. 1, Dorsal view (x4-4). 2, Latex cast of fragmentary
counterpart mould of this specimen showing anterior branch of facial suture (x6). 3, Oblique
exterior view (x4 4). 4, Lateral view (x4-4). 5, Anterior view (x4-4). USNM No. 145307.

Fig. 6. Fragmentary cephalon showing genal spine, dorsal view (x4). USNM No. 145309.

Fig. 7. Cranidium, dorsal view (x5). USNM No. 145308.

Figs. 8-9. Exfoliated pygidium. 8, Dorsal view. 9, Lateral view (both x 3). USNM No. 145310.

Palaeontology, Vol. 9

PLATE 42

ORMISTON, Australosutura

A. R. ORMISTON: OCCURRENCE OF AUSTRALO SUTU RA (TRILOBITA) 273

suggestion but for morphologic rather than stratigraphic reasons. Several major charac-
ters which distinguish the Argentine specimens from the Australian topotypes as well
as from the Oklahoma material are discussed in the foregoing section on description.

Although the Oklahoma specimens appear to differ in no important respects from the
topotypes of A. gardneri, it is considered advisable to identify them for now as A. aft'.
A. gardneri because of their imperfectness and the fact that 1 have not examined the
topotype specimens.

A Tournaisian species of Australosutura from the Tulcumba sandstone of Australia
has been described as Australosutura sp. (Campbell and Engel 1963, p. 1 10). This species
differs in many respects from the Oklahoma specimens. It has: (1) a distinctly longer
and narrower glabella; (2) the glabella more strongly convex in transverse cross section;
(3) no enlarged tubercle opposite the 2p glabellar furrow; (4) a relatively shorter,
broader pygidium; and (5) the inner part of the pygidial pleural fields much flatter in
cross section.

The similarities between Australosutura and Cordania have already been pointed out
(Amos, Campbell, and Goldring 1960, p. 229), and Cordania has been suggested by
those authors (op. cit. p. 230) as a possible ancestor of Australosutura. 1 would suggest
that Mystrocephala Whittington (1960, p. 413) in having a broad, concave preglabellar
field; no broad, convex cephalic border; and the posterior pygidial segmentation
represented by posterior pleural bands is even closer to Australosutura than is Cordania.
Mystrocephala is also closer in age to Australosutura. The youngest representatives of
Mystrocephala are Givetian in age (Whittington 1960, p. 414), whereas Cordania does
not occur in beds younger than Lower Devonian (op. cit. p. 409). Thus Mystrocephala
is better qualified than Cordania as a possible ancestor of Australosutura.

REFERENCES

amos, a. J., Campbell, k. s. w., and goldring, r. 1960. Australosutura gen. nov. (Trilobita) from the
Carboniferous of Australia and Argentina. Palaeontology, 3, 227-36, pi. 39, 40.

Campbell, k. s. w. and engel, b. a. 1963. The faunas of the Tournaisian Tulcumba Sandstone
and its members in the Werrie and Belvue Synclines, New South Wales. J.geol. Soc. Aust. 10, 55-122,
pi. 1-9.

collinson, c., scott, a. j., and rexroad, c. b. 1962. Six charts showing biostratigraphic zones and
correlation based on conodonts from the Devonian and Mississippian rocks of the Upper Mississippi
Valley. Circ. III. geol. Surv. 328, 32.

Huffman, g. g. 1958. Geology of the flanks of the Ozark Uplift. Bull.Okla.Geol.Surv.il.
mitchell, J. 1922. Description of two new trilobites, and note on Griffithides convexicaudatus Mitchell.
Proc. Linn. Soc. N.S.W. 47, 535-40, pi. 54.

Whittington, h. b. 1960. Cordania and other trilobites from the Lower and Middle Devonian.
J. Paleont. 34, 405-20, pi. 51-54.

A. R. ORMISTON,

Research Department,

Pan American Petroleum Corporation,
P.O. Box 501,

Tulsa 2, Oklahoma

Manuscript received 18 February 1965

DESCRIPTIONS OF SCHIZAEACEOUS SPORES
TAKEN FROM EARLY
CRETACEOUS MACROFOSSILS

by N. F. HUGHES and JUDITH MOODY-STUART

Abstract. Full descriptions are given of spores taken from the original macrofossil material of Ruffordia
goepperti (Dunk. ), Pelletieria valdensis Seward, and Schizaeopsis americana Berry. All three plants are of Early
Cretaceous age, the first two from the Lower Wealden of England, and the last from the Patuxent Formation
of the Potomac Group of Virginia, U.S.A. The use of measurements in the erection of species in the dispersed
spore genus Cicatricosisporites is discussed in the light of these observations.

Dispersed spore palynological studies show that the spore type usually classified as
Cicatricosisporites Pot. and Gell. first appears in the Early Cretaceous or latest Jurassic,
and then diversifies rapidly. Unfortunately these spores have not yet proved to be of
great stratigraphical value in the Cretaceous period, and we think this is almost entirely
due to the effects of poor typification and description, and of uncoordinated nomen-
clature. Bolkhovitina (1961) has brought together most of the fossil records of this group
in a very useful paper, although her system of nomenclature is unfortunate and many
of the figures of other authors which she was obliged to reproduce were inadequate in
their original form.

It so happens that three separate fertile plant organ species, believed to belong to the
Schizaeaceae, were described with their spores in the period 1911-13. The illustrations
of the spores were generalized sketches, with in two cases rather inadequate photographs ;
and the descriptions do little more than allude to the probable fern family affinity,
a treatment which was no doubt considered adequate at the time. Couper (1958) pub-
lished some photographs of spores of Ruffordia goepperti which had been known
previously only from Seward’s (1913) sketch, but Couper regarded these spores as
comparable with his dispersed spore material which he placed in the Eocene species
Cicatricosisporites dorogensis. As far as we know neither the spores of Pelletieria valdensis
nor of Schizaeopsis americana have been adequately described.

All the material of Ruffordia and Pelletieria has been kindly loaned for the purpose
by the Department of Palaeontology, British Museum (Natural History). The necessary
pull from the Schizaeopsis specimen was taken during a visit (N.F.H. July 1959) to the
United States National Museum, Washington, and permission to figure the material
was kindly given later by Dr. F. M. Hueber of that institution. The relevant slide prepara-
tions will be sent to the institutions concerned. Potonie (1965, pp. 45-46) refers to this
same material which he examined when in Cambridge.

Preparations were made by brief oxidation (approx, two hours) in concentrated nitric
acid only, clearance in approx. 5 per cent, ammonia, and mounting unstained in
glycerine jelly. In each case all of the material removed from the plant was mounted;
it was handled only by pipette and was not centrifuged. Specimens recorded were all

[Palaeontology, Vol. 9, Part 2, 1966, pp. 274-89, pis. 43-47.]

275

HUGHES AND MOODY-STUART: SCHIZAEACEOUS SPORES

those present on certain selected traverses on the slide preparation. Stage co-ordinates
refer to Ortholux microscope no. 491249, Department of Geology, Cambridge.

30

A

KEY

Squares represent
one specimen each
in orientation
indicated

□ Oblique

□ Polar

IS Equatorial

text-fig. 1. Frequency distribution histograms of maximum diameter in microns of spores of Ruf-
fordici goepperti. a, New preparation (V2192b-f), 200 specimens; standard deviation 5-2 p, coefficient
of variation 10 per cent, b, Five point moving average from the data of A. c, Seward s preparation
(V2192a), 58 specimens (all available); standard deviation 6-1 /x, coefficient of variation 13 per cent.

DESCRIPTIONS
Family schizaeaceae
Genus ruffordia Seward 1894

Type species. Sphenopteris goepperti Dunker 1846 (Dunker's earlier reference not effective).

276

PALAEONTOLOGY, VOLUME 9

Remarks. Although the genus Ruffordia was well conceived, Seward should have erected
a new species for his relatively good fertile material, instead of relying on the inadequately
described holotype (sterile frond) of Dunker. In view of the remarks of Michael (1936,
p. 34) a new species may still be necessary, as she could not find in the German specimens
the variability which Seward reported. Nomenclatural difficulties would result from
any such change now.

Ruffordia goepperti (Dunk.) Seward 1894 (spores)

Plate 43, figs. 1-12

1913 Seward, p. 91, text-fig. 2a.
non 1921 Halle, p. 11, pi. 2, fig. 8.

1958 Couper, p. 109, pi. 17, figs. 4-6.

1962 Potonie, p. 104, pi. 10, fig. 268.

Holotype. We have located and figured (PI. 43, fig. 10) a spore on Seward’s slide V2192a which by
elimination appears to be the original of his text-fig. 2a (1913); we are considering this spore to be the
holotype. Seward’s preparation and mounting methods for this material are not known, and the spore
is now very pale in colour. The size range and mean of the fifty-eight specimens on his slide differ
from those of our preparation (see text-fig. lc). The spore figured by Couper (1958, pi. 17, figs. 4, 5)
is probably another specimen nearby on the slide (OR 48.1 126.1).

Locality. The specimen V2192a is clearly marked Rufford Collection, Ecclesbourne, nr. Hastings, in
spite of Seward’s (1913, p. 91) record of Fairlight. The rock is a hard red-brown sideritic siltstone of
maximum grain size 50 p. It is likely to have come from the base of the Ashdown Sands or from the
top of Fairlight Clay division e.

Material. Spores were taken by needle and by ‘puli' from what appeared to be sporangia on the middle
section of the frond of specimen V2192a, which according to Museum labels in Seward’s handwriting
was the specimen he used but did not specifically cite (1913) in making his slide (now V2192a).

Emended spore diagnosis. Spores trilete tetrahedral with strongly convex distal surface,
but with rounded triangular amb. Contact faces smooth (PI. 43, figs. 1-2). The lips are
simple membraneous 2-3 p high extensions of the exine (PI. 43, fig. 6) which has a
constant thickness of T5 p excluding sculpture. The inter-radial equatorial and distal
regions bear three sets of sub-parallel muri which occasionally join to form a partial
reticulum (PI. 43, figs. 4, 5, 8). The three sets of muri leave a triangle (PI. 43, fig. 10)
of varying size centred on the distal pole, with variable sub-parallel sculpture within
the triangle. Muri, of rounded profile, 1 -5-2-0 p high and wide, spaced on average

EXPLANATION OF PLATE 43

Figs. 1-12. Ruffordia goepperti (Dunk.) Seward; spores taken from specimen V2192a BMNH, X 1000.
1-3, Proximal aspect, high, mid and low focus respectively; prep. V2192c, OR 33.5 108.5. 4-5,
Equatorial aspect, mid and low focus; prep. V2192d OR 30.9 120.6. 6-7, Oblique proximal aspect,
mid and high focus; Prep. V2192d, OR 32.2 120.3. 8, Oblique aspect pattern of distal muri; prep.
V2192c, OR 46.4 117.9. 9, Proximal aspect, showing radial rib profile formerly diagnostic; prep.
V2192c, OR 46.4 112.4. 10, Holotype, proximal aspect, low focus (distal); Seward's prep. V2192a,
OR 48.2, 127.2. 11, Section perpendicular to polar axis, triplan, distal; prep. V2192g, OR 54 112.4.
12, Adjacent section, showing laesura with compressed lips and proximal face; prep. V2192g, OR
60.1 115.6. 13, Section parallel to polar axis, laesura open; prep. V2192h, OR 40.1 125.1.

Palaeontology, Vol. 9

PLATE 43

12

HUGHES and MOODY-STUART, Lower Cretaceous spores

HUGHES AND MOOD Y-STU ART: SCHIZAEACEOUS SPORES

277

2-3 p apart (PI. 43, fig. 8). Muri of adjacent sets do not coalesce (see text-fig. 2a-c) in
equatorial radial regions which are flexible and not thickened.

Dimensions. Diameter (200 specimens) 35— (51)— 63 /x (preparations V2192b-f); 80 per cent, of speci-
mens fall between 44 and 58 p. Percentage of specimens in polar view 28, equatorial view 13, and
oblique 59 as plotted in text-fig. 1a. Similar observations for Seward’s slide V2192a are given in text-fig.
lc; the lower values for diameter (58 specimens) 32-(46)-58 /x, 80 per cent, between 39 and 55 /x,
may be due to difference in maceration procedure.

text-fig. 2. Rujfordia goepperti (spores), diagrams x 1000; muri shown in black. A-c, from Seward’s
preparation (V2192a) to illustrate the configuration of the muri. a, Distal; b, Proximal faces of holo-
type, Plate 43, fig. 10, OR 48.2 127.2. c, Distal face of specimen figured by Couper (1958), OR 48.1
126.1. d. Section normal to laesura, half radial distance from pole, showing concave proximal face
and false margo. e, Specimen of Plate 43, fig. 12, showing folded proximal face and false margo.

Description. All spores are strongly compressed and usually folded. Possibly because
they were dried out before burial and became concave proximally, there is a tendency
for the parts of the exine adjacent to the laesurae to be pressed together; this raises the
lips, simulates a margo in some specimens, and causes some proximal rigidity (text-fig.
2a; PI. 43, figs. 6, 12, 13). The arrangement of muri in the radial region is comparable
with that shown by Dettmann (1963, text-fig. 4 p); the mode of preservation however,
always causes difficulty in observing details of the radial areas.

Comparison. The spores obtained from R. goepperti (Halle 1921) from an unspecified
horizon in Ussuri appear from poor illustrations to be of a different type.

Remarks. Groot and Penny (1960) combined the spores of Seward and Couper in the
dispersed spore genus Cicatricosisporites. Bolkhovitina (1961) erected Ruffordia araliea

278

PALAEONTOLOGY, VOLUME 9

for dispersed spores which have widely spaced high muri but differ from spores of R.
goepperti in overall smaller numbers of muri. Brenner (1963) placed all these spores
in synonymy in Cicatricosisporites aralica comb, nov., but none of this procedure is
admissible.

Genus pelletieria Seward 1913

Remarks. The available material of the single macrofossil species P. valdensis is more
extensive than that figured by Seward, but the necessary re-description cannot be
included in this paper. We assume that Seward intended the specimen V2329 (pi. xn,
fig. 12a) to be the holotype and herewith designate it as such in case there is any doubt.
The specimen of plate xn, fig. 12 b is on V2329 but has been partly destroyed in extracting
spores, presumably by Seward; on the same rock specimen are the originals of Seward’s
text-fig. 4a and g but 4b-e have not been located. The specimen of text-fig. 4f (now
damaged) is on slide V2329b; that of text-fig. 3b is on V2368, text-fig. 3a on V51843,
and text-fig. 3c on V51840.

Pelletieria valdensis Seward 1913 (spores)

Plate 44, figs. 1-8; Plate 45, figs. 1-7
1913 Seward, p. 91, pi. 14 fig. 5, text-fig. 2b.

Holotype. Although Seward's wording (191 3, p. 92, para. 2) may be taken to suggest that the specimen of
his text-fig. 3b (V2368 — Ruflord Coll.) provided the spores, it seems certain from all other evidence
that they came from V2329. We have found the original of Seward’s photograph (1913, pi. 14, fig. 5)
on slide V2329a and we re-figure it as the holotype (PI. 44, fig. 4). It has not been possible to identify
the obliquely viewed specimen of Seward’s text-fig. 3b.

Locus typicus. V2329 and V2368 are both labelled Ecclesbourne, nr. Hastings; V2329 is a coarse grey
siltstone, unbedded, with maximum grain size 150 p. These specimens must again come from the lowest
Ashdown sand or Fairlight clay e.

Material. Additional spores were taken from V2329 from an isolated spore mass very close to the
original of plate xii, fig. 12 b (Seward 1913).

EXPLANATION OF PLATE 44

Figs. 1-8. Pelletieria valdensis Seward; spores taken from specimen V2329 BMNH, X 1000. 1-2,
Sections parallel to polar axis, through laesura (2, near outer termination of laesura); prep. V2329j,
OR 33.3 119.9 and 31.5 114. 3, Distal aspect, high focus; prep. V2329i, OR 37.8 117.6. 4, Holotype,
proximal aspect; Seward's prep. V2329a, OR 28.7 118.9. 5-6, Equatorial aspects, prep. V2329f. 5,
Showing profile of muri; OR 42.6 117.9. 6, Laesura open; OR 35.3 115.3. 7, Small specimen, equa-
torial aspect with optical section of proximal exine; prep. V2329a, OR 38.7 125.8. 8, Equatorial
aspect; prep. V2329h, OR 35.7 117.8.

EXPLANATION OF PLATE 45

Figs. 1-7. Pelletieria valdensis Seward; spores taken from specimen, V2329 BMNH, X 1000. 1-2,
Sections perpendicular to polar axis; prep. V2329j, OR 29.4 113.9 and OR 51.5 107.7 1, Distal
(slightly oblique). 2, Proximal face showing depression bordering laesurae. 3, Proximal aspect,
high focus; prep. V2329g, OR 47.8 118.7. 4, Distal aspect; prep. V2329g, OR 40.7 121.7. 5-7,
Sections parallel to polar axis; prep. V23291, OR 50 111.8, OR 45 112.5, OR 42.3 112.3.

Palaeontology, Vol. 9

PLATE 44

HUGHES and MOODY-STUART, Lower Cretaceous spores

Palaeontology , Vol. 9

PLATE 45

HUGHES and MOODY-STUART, Lower Cretaceous spores

HUGHES AND MOOD Y-STUART: SCHIZAEACEOUS SPORES 279

Emended diagnosis. Trilete spores with rounded triangular amb. Laesurae simple and
with thin lips, of medium length flanked by variably deep and narrow areas of nega-
tive sculpture. Proximal and distal surfaces bear a regular canaliculate sculpture, best

A

90

.rJL

100

Mean! 73'9

S.D.

50

C

100

text-fig. 3. Frequency distribution histograms of maximum diameters in microns, of spores of Pelle-
tieria valdensis. A, New preparation (V2329f), 200 specimens; standard deviation 8-5 /x, coefficient of
variation 12 per cent, b, Five point moving average from the data of a. c, Seward’s preparation (V2329a),
100 specimens; standard deviation 7-7 /x, coefficient of variation 12 per cent.

described as of lumina 0-5 to 1-0 ju. wide and 4/z apart. The lumina, in three parallel
inter-radial sets, do not coalesce but merely inter-finger in the radial regions, maintain-
ing the latter as more rigid areas which affect the shape of the spore. Inter-radial exine
thickness 6-5 /x.

C 3803

u

280

PALAEONTOLOGY, VOLUME 9

Dimensions. Diameter (200 specimens) 52— (74)— 98 /z. (new preparation V2329f); 80 per cent, of
specimens have maximum diameter between 63 and 84 ^ (text-fig. 3a). Specimens in polar aspect
25 per cent., equatorial aspect 30 per cent. Similar observations on 100 specimens from Seward’s slide
V2329a give diameter 53-(66)-84 ^ (text-fig. 3c); 80 per cent, of values fall between 56 and 78 fi.
Specimens in polar aspect 26 per cent., equatorial aspect 23 per cent. Values for Seward's specimens
are again lower (see under Ruffordia).

text-fig. 4. Spores of Pelletieria valdensis; distribution histograms of measurements in microns with
one square per specimen. Total numbers of specimens differ as not all characters are observable on
every spore, a, Observed thickness of exine in radial equatorial region, 69 specimens, b. Inter-radial
exine thickness, 73 specimens, c, Span covered by four lumina (and four muri), 123 specimens.

Description. The symmetry of the distal sculpture varies from the extreme seen in 24
per cent, of the specimens in which the triangle is centred on the distal pole (PI. 45, fig. 4),
through various intermediates to the sub-parallel lumina of Plate 44, fig. 3 ; it is doubtful
whether this feature can have the taxonomic significance implied by Deak (1963). The
span covered by four lumina (the same as the usual four muri) is recorded on text-fig. 4c,
and plotted on text-fig. 5c. The resulting muri can be described as typically 3-(4)-5 p wide,
3-4 ^ high and flat-topped. The inter-radial exine thickness is recorded (text-fig. 4b)
and plotted (text-fig. 5); such variability is of course much more easily observed in
a thick exine such as in Pelletienia, than in one near the limit of optical resolution.
The measured radial and inter-radial variable exine thickness depends on the degree
to which the lumina are ‘open’ and this is affected by;

1. The directions of the lumina with respect to the amb (contrast PI. 44, fig. 5, with
a section, P). 45, fig. 7).

HUGHES AND MOOD Y-STUART: SCHIZAEACEOUS SPORES

281

g

“O

o

5 10

Radial equatorial exine thftkness

A

io.

g

“D

O

60

70 80

Maximum diameter

O

E

~D

c

O

o

c

'E

D

^r

o

c

o

a.

co

60

T70 ' So

Maximum diameter

B

90

100

C

"*90 ' ’ 100

text-fig. 5a-c. Pelletieria valdensis ; linear plots showing the possible relationships between some of the
characters recorded in text-figs. 3 and 4, scales in microns.

2. The distance of the measurement from the radial areas where the lumina close;
thus the diameter of the spore will affect this especially in small specimens.

3. The degree of inflation (in life?), and the compression of the specimen (in dia-
genesis).

On average (text-figs. 4, 5a) the radial thickness is fifty per cent, greater than the inter-
radial. The opening of lumina up to an observed maximum of 90° is calculated to reduce

282

PALAEONTOLOGY, VOLUME 9

to two-thirds the depth of a lumen (text-fig. 6a, c), but this reduction will not explain
the observed differences in exine measurements on many specimens. The simplest
assumption is that the exine suffers some distortion but no change of volume when it is
folded; the scale diagrams of text-fig. 6b-c show observed extremes of exine profile
and measurements; compare with Plate 45, fig. 4. Text-fig. 6a compares with exine seen
unfolded in Plate 45, fig. 7. By drawing on graph paper it was shown that all three

A

B

C

text-fig. 6. Pelletieria valdensis; diagrams of sections (x2000) of spore exine to explain
variable exine thickness, bars show exine thickness as would be measured by optical section
in polar view, a, Inter-radial exine, not folded; sections normal to direction of lumina.
b-c, Exine folded at amb; plane of section containing polar axis, b, Radial, c, Inter-radial.

diagrams have the same cross-sectional area. Since this area is taken normal to the
plane of folding (in text-fig. 6b, c), it will be proportional to the volume of a parallel
element of exine. The observed differential exine thickness may be explained thus with-
out postulating a taxonomically significant radial thickening. Since the orientation of
the lumina affects the spatial properties of the spore, any asymmetry in the distal sculp-
ture (76 per cent, of the specimens) often causes irregularity in the amb and in the exine
measurements.

The tapering lips project above normal exine level near the proximal pole only; they
are flanked by a variable negative area which closes or coalesces with a lumen when
the lip membrane merges into the exine at the apparent limit of a laesura. We suggest
that the negative sculpture, often deeper than a lumen, gives flexibility to an otherwise
rigid proximal face. From Plate 45, figs. 2 and 5, and Plate 44, fig. 2, it can be seen that
the laesurae extend lipless almost to the equator, well beyond what appears to be the
limit in proximal surface view.

Comparison. Couper (1958) suggested that these spores were similar to his dispersed
spore species Cicatricosisporites dorogensis but the size range and other features are

HUGHES AND MOODY-STUART: SCHIZAEACEOUS SPORES

283

entirely different. A distinction is more necessary between this spore and C. brevi-
laesuratus Couper (type from Barremian/Aptian) which is even larger (90 p) and is de-
scribed as having an unsculptured contact face. Bolkhovitina (1961 ), perhaps on account
of Couper’s comparison, has inadvisedly (and wrongly) placed a number of unrelated
dispersed spore species in this macrofossil genus.

The spores described from the Hastings Beds by Boodle (1895) were similar to those
of P. va/densis ; they occurred in a large spore mass but their provenance was not dis-
covered.

Botanical Affinity. Neither the macrofossil remains nor the spores of Pelletieria are very
close in detail to species of Ceratopteris, although the suggestion that the macrofossil
may have been a water plant and not necessarily related to normal Schizaeaceae is
worth exploring.

Genus schizaeopsis Berry 19116

Remarks. The genus is only valid from Berry’s second paper (19116), when he erected
the necessary new species for the single specimen concerned (now USNM 3209).

Schizaeopsis americana Berry 1911 (spores)

Plate 46, figs. 1-8; Plate 47, figs. 1-10

1911 a Berry, p. 193, pi. 12, figs. 2-6.

191 1 b Berry, p. 216, pi. 22, figs. 4-9.

Holotype. The original single photograph (Berry 191 1 b, pi. 22, fig. 4) is not recognizable; the drawings
(pi. 22, figs. 5-9) are too generalized to be used further.

Locus typicus. Patuxent formation, Frederiksburg, Virginia, U.S.A.

Material. The rock specimen (USNM 3209) is a medium brown non-calcareous siltstone with plant
fragments and the fairly large fertile frond described by Berry (191 1Z?, pi. 22, figs. 1-3). A small part
of the spore mass was taken from the tip of one of the central leaf lobes of the specimen; preparations
USNM 3209 g-j.

Emended diagnosis. Spores almost spherical; amb convex rounded triangular. Laesurae
trilete, short, simple, with raised thin (0-5 p) lips. Whole surface covered by uniform
narrow muri \k p wide, 1 p high, flat-topped; lumina 1 p wide. The whole proximal
face is covered by three inter-radial sets of muri; the equatorial region bears a single
set of five or six circular muri parallel with the outer members of the proximal inter-
radial sets. The distal surface is covered by a single set of parallel muri which converge
abruptly opposite two radial areas and swing round parallel to the equatorial set in the
third. Exine thickness 3-4 p in distal and equatorial regions, thicker immediately round
laesurae as a margo ( sensu Harris 1955, p. 26).

Dimensions. Diameter (200 specimens) 44-(71)-87 p (preparations USNM 3209 g-j); 80 per cent, of
specimens have maximum diameter between 61 p and 80 p. Specimens in polar aspect 16 per cent.,
equatorial 28 per cent. The records on text-fig. 7a, which are plotted on text-fig. 8a-c, show a
numerically small but statistically significant group of small spores in this preparation. The group was
distinguishable by eye, and proves to consist of small spores with thicker distal exines and narrower
muri and lumina which had never opened at the laesurae (PI. 46, figs. 5, 6).

284

PALAEONTOLOGY, VOLUME 9

' 1 ‘ 1

2 1 Mean | 713 1 2

t

text-fig. 7. Frequency distribution histograms of maximum diameters in microns of spores of
Schizaeopsis americcma. A, New preparations (USNM 3209 g-j), 200 specimens; standard deviation
7-4 p, coefficient of variation 10 per cent, b, Five point moving average from data of a.

EXPLANATION OF PLATE 46

Figs. 1-6 Schizaeopsis americcma Berry; spores (xlOOO) taken from specimen USNM 3209; each
preparation bears this number with the suffix here indicated. 1-2, Proximal aspect, mid and low
focus respectively; prep, g, OR 36.9 118. 3, Section parallel to polar axis, at end of laesura; prep. /,
OR 22.2 1 19.5. 4, Similar section, just beyond laesura; prep, k, OR 39.6 109.2. 5-6, Proximal aspect,
small specimen, mid and low focus; prep, j, OR 34.3 118.5. 7-8, Oblique aspect, high focus and
optical section through laesura ; prep. /;, OR 34.6 125.5.

EXPLANATION OF PLATE 47

Figs. 1-10. Schizaeopsis americana Berry; spores (x 1000) taken from specimen USNM 3209; each
preparation bears this number with the suffix here indicated. 1-2, Proximal aspect, low and mid focus;
prep, h, OR 34.9 120.7. 3-4, Proximal aspect, spore expanded showing lips (see text-fig. 10d), mid
and high focus; prep. OR 36.2 123.6. 5-6, Sections parallel to polar axis, open laesura; prep. /,
OR 40.4 119.6 and OR 27.5 120.4. 7-10, Successive sections parallel to polar axis, across laesura;
prep, m, OR 24.2 109.7, OR 37.3 108.9, OR 28.6 109.5, and OR 31.0 109.8.

Palaeontology, Vol. 9

PLATE 46

HUGHES and MOODY-STUART, Lower Cretaceous spores

Palaeontology, Vol. 9

PLATE 47

HUGHES and MOODY-STUART, Lower Cretaceous spores

HUGHES AND MOOD Y-STUART: SCHIZAEACEOUS SPORES

285

Description. Text-fig. 9a-c shows in diagrammatic section various states of the laesura
and margo as seen in normal spores (PI. 46, fig. 8; PI. 47, figs. 7-8). These observed
conditions may be understood as successive states on the model of an inelastic inner
layer of exine bound to an outer flexible sculptured layer, the whole spore gradually

-£ 5.

.3

~o

o

5 10

Span of 4 muri and lumina

o

a

to

Maximum diameter

o

T3

0

1

£>

c

: ;

Maximum diameter

text-fig. 8a-c. Schizaeopsis americana ; linear plots showing the possible relationships between the
characters indicated, on 38, 45, and 55 specimens respectively, scales in microns.

inflating. First the lower exine pulls apart at the laesura, then the upper exine parts,
and finally only the lips remain in contact.

The specimen in Plate 47, figs. 3-4, which is apparently dehisced, shows on fuller
examination closed laesurae at the bottom of a 10 p wide infolding of the proximal
exine. Our explanation is that text-fig. 9d shows the effect of deflation, perhaps on drying
out, on a spore which on the plant had reached the condition of text-fig. 9c. In surface
view this may give the appearance of muri truncated at high angles by the laesurae.

286

PALAEONTOLOGY, VOLUME 9

In these spores, as in Pelletieria, the laesurae and margo extend beyond the apparent
limit as seen in polar view (PI. 46, fig. 4; PI. 47, figs. 9-10).

Comparison. Berry's (1911) figures (drawings) were inadequate, and Potonie (1962,
pi. 10, fig. 270) unfortunately increased the confusion with a bad re-drawing. The dia-
meter was imprecisely quoted by Berry as 100 p, and this perhaps led Couper (1958,
p. 110) into comparing these spores with his large but quite different (wide muri)

text-fig. 9. Diagrammatic sections (x 1000) of Schizaeopsis spores, parallel to a polar axis, normal
to a laesura. a-c, Successive states during inflation of the spore, d. Probable effect of subsequent
deflation of a mature spore (as in drying out).

Cicatricosisporites brevi/aesuratus. Brenner (1963) perpetuates this error, while his
C. polomacensis (pi. 9, figs. 4, 5) probably is very close to Schizaeopsis americana. Groot
and Penny (1960) do not record such a spore, but Stover (1964) figures (pi. 1, fig. 20)
as C. dorogensis , a spore from the Arundel formation which again appears to be very
close to S. americana.

Remarks. Although the true nature of this spore was concealed by the early description,
the spore-masses and leaf were well illustrated by Berry (1911 a, b).

Validity of name. Berry (19116, p. 219) made taxonomic and nomenclatural changes in
erecting Schizaeopsis americana which may be suspect. We have followed his usage
because (n) Berry’s description is valuable while in this case Fontaine’s is not, (b) Berry’s
work has not been challenged and has been much quoted, and ( c ) Berry did amplify
the description of Acrostichopteris expansa (Fontaine), the material of which is not
fertile. We feel that it is just about possible to justify Berry’s procedure because no type
specimens were designated by Fontaine or by others before 1911, and that therefore on
balance it is simpler and more helpful to allow it to stand.

GENERAL MORPHOLOGY OF SCHIZ AEACEOUS SPORES

Detailed observation of (a) the pattern of junction of muri or lumina in the radial
areas, (6) the margin of the laesurae, and (c) the shape, spacing, and orientation of the
muri, together provide more characters than are available in other fossil trilete fern
spores. The sculptural configuration has diagnostic qualities at a species level, but the
distal pattern of muri can vary if one adjacent set is under- or over-'developed’ in num-
ber, and we do not believe that patterns 2 and 3 of Deak (1963, pi. 1) are fundamentally
different. The features mentioned in the first sentence of this paragraph determine the
flexibility of a specimen; the three spore species illustrated in this paper must differ
strikingly in this original property and thus in their tendency to fold in diagenesis. Their
attitude, however, under the partial compression in the sporangia or attached spore-
masses will have been random, and not subject to sedimentation factors as are dispersed
spores.

HUGHES AND MOOD Y-STU ART: SCHIZAEACEOUS SPORES

287

MEASUREMENT

Since the wavelength of light makes diffraction effects dominant below one micron,
we have omitted the possible inner layering of the Schizaeopsis exine from diagnosis
pending examination by electron micrographs.

We have chosen ‘maximum diameter’ as a principal measurement because it was
obtainable from all specimens, and in equatorial views it was measured along the equator.
It can be seen from text-fig. 1a that equatorial views of Ruffordia give lower values;
these will weight the lower end of the frequency histogram according to the relative
proportion of spores in equatorial aspect on a slide, but since this proportion and
any distinct size differences of the maximum diameter in equatorial view depend on the
structure of the spore it will enhance the diagnostic value of the histogram.

Variability is expected and is much more easily observed in a thick exine such as that
of Pelletieria than in a thin one, and for this reason the ranges of the minor Ruffordia
measurements have not been given. Similarly a dimensionless coefficient of variation
has been calculated (ratio: standard deviation/mean maximum diameter, expressed as
a percentage) for the distributions of maximum diameters, to make clearer comparison
between species.

We accept an estimated observation error of 2\ per cent for reasonable speed of
working, and thus validly submerge human partiality for round numbers in a five point
moving average (text-fig. 7a, b).

The span of four muri should be measured in a distal inter-radial area, and the shape
of muri in optical sections where the plane containing the murus is perpendicular to the
slide. Exine thickness should be measured inter-radially.

It is likely that spore ‘populations’ of varying preservational states, extracted from
diagenetically different rock samples may give different measurements; maceration
treatment may also cause differences. In these cases the difference from the holotype
and type-range (from the type locality) will be systematic and open to experimental
investigation and rationalization. Knowledge of these differences will widen the scope
of the taxon description, contributing information about the chemical and physical
behaviour of exines in different species. Similarly it will make specimens which are
other than ‘well preserved’ more precisely identifiable for stratigraphical use.

PURPOSE OF THIS STUDY

It is hoped that the straightforward presentation of measurements and of plots of
possible linear proportionality between pairs of characters for a large number of speci-
mens of these in situ spores may provide sufficient ground for confidence in subsequent
erection of stratigraphically useful dispersed spore species of Cicatricosisporites s.l. in
the Early Cretaceous. Extreme ‘lumping’ by Couper (1958) and others into the irrelevant
Eocene species C. dorogensis, and very imperfect original typification of that and of
many other dispersed spore species still being erected has led to a situation in which
virtually no stratigraphic use can now be made of spores of this type within the Lower
Cretaceous. With the small number of characters available successive species will clearly
overlap in morphography, but if a species is typified from one precise source or sample
with analysis of variation given on an adequate number of specimens it is hoped that
some further progress will be possible.

288

PALAEONTOLOGY, VOLUME 9

As it is clear that few characters can be observed in all (or even most) of the specimens,
an overall number of 200 is considered necessary for in situ spores, and 100 specimens
is a minimum for dispersed spores even when comparable observations for in situ spores
are available.

THE DISPERSED SPORE GENUS IC AT RICO S I S PO RITES

Range. Dettmann (1963, p. 74) has shown that Contignisporites may be separated in
morphography and in possible affinities from Cicatricosisporites. As Corrugatisporites
(Early Jurassic) may also be easily distinguished, it is not necessary to extend the range
of Cicatricosisporites down through the Jurassic as is done by Kedves and Sole de Porta
(1963) and others; the true range begins in the latest Jurassic (at about the top Kim-
meridge Clay of England), and the spores are only at all common from the earliest
Cretaceous onwards.

Comparison. We believe that in the earliest Cretaceous there is no distinct genus Plica-
tella , although by Barremian and Aptian times radial equatorial exine thickening may
well be diagnostic of distinct taxa. Study of Pelletieria has shown that occasional thicken-
ings occur in radial and proximal regions but that they are within the species ‘popula-
tion’. As none of the earliest Cretaceous records of Plicatclla ( Appendicisporites tri-
cornitatus etc.) are supported by more than a very few specimens (Couper 1958, Lantz
1958a, Hughes 1958), there is not yet any contrary evidence; the material of Lantz
19586, Pocock 1962, Brenner 1963, Singh 1964 is probably of Barremian or later age.

The establishment of a definite range beginning for Plicatella , and its separation from
the later Appendicisporites, must await the erection of properly based species, as in-
dicated above.

The separation of such spores from Cicatricosisporites appears to be a more complex
matter than measurement of the thickness of part of the wall (Pocock 1964, p. 162),
apart from the difficulty that the ‘wall layers’ do not appear in our sections (PI. 44, 45)
which thus fail to support the diagrams of Pocock (1962, p. 162, figs. 16, 17).

Use of the genus. We reject as undesirable and unworkable the proposal of Pocock (1964,
p. 156) to use Cicatricosisporities ‘as an organ genus of the ( extant ) genus Moliria and
morphologically related fossil species’. We believe that it should remain as a form genus
at least until the Cretaceous and Tertiary history of the plant groups involved has been
worked out with macrofossils or until the diagnoses of the extant genera ( Mohria etc.)
have been properly emended to include the fossils, as suggested by Hughes (1963).

Acknowledgement. The first-named author is indebted to the Science Research Council for a special
research grant (SR/447).

REFERENCES

berry, e. w. 191 In. A Lower Cretaceous species of Schizaeaceae from Eastern North America,
Ann. Bot., London, 25, 193-8, 1 pi.

19116. In lull, Clark, and berry. Systematic paleontology of the Lower Cretaceous deposits of

Maryland. Maryland Geol. Surv., Baltimore.

bolkhovitina, n. a. 1961. (Fossil and Recent spores of the family Schizaeaceae.) Trad. Geol. Inst.,
Acad. Nauk S.S.S.R., Moscow, 40, 1-176, 41 pi. (in Russian).

HUGHES AND MOODY-STUART: SCHIZAEACEOUS SPORES

289

boodle, L. A. 1895. Spores in a speciman of Tempskya. Ann. Bot., London, 9, 137-41.
brenner, g. J. 1963. The spores and pollen of the Potomac Group of Maryland. Bull. Maryland
Dept. Geol., Mines and Water Resources, Baltimore, 27, 1-215, 43 pi.
chandler, m. E. J. 1955. The Schizaeceae of the South of England in Early Tertiary times. Bull.

Brit. Mas. Nat. Hist. Geol., London, 2, 7, 291-314, 7 pi.
couper, r. a. 1958. British Mesozoic microspores and pollen grains. Palaeontographica, Stuttgart,
103b, 75-179, 17 pi.

dear, m. 1963. Quelques spores striees de l’etage Aptien. Rev. Micropaleont., Paris, 5 (4), 251-6, 2 pi.
dettmann, mary E. 1963. Upper Mesozoic microfloras from South-Eastern Australia. Proc. Roy.
Soc. Victoria, Melbourne, 77, 1-148, 27 pi.

groot, J. J. and penny, j. s. 1960. Plant microfossils and age of non marine Cretaceous sediments of
Maryland and Delaware. Micropaleontology, New York, 6, 225-36, 2 pi.
halle, t. G. 1921. On the sporangia of some of Mesozoic ferns. Ark. Botanik, Uppsala, 17, 1-28, 2 pi.
Harris, w. f. 1955. A manual of the spores of New Zealand Pteridophyta. Bull. N.Z. Dept. Sci. Ind.
Res., Wellington, 166, 186 pp.

hughes, n. F. 1963. The assignment of species of fossils to genera. Taxon, Utrecht, 12 (9), 336-7.
kedves, m. and sole de porta, n. 1963. Comparacion de las esporas del genero Cicatricosisporites
R. Pot. y Gell. 1933 de Hungria y Colombia. Bol. geol. Univ. Industr. Santander, 12, 51-76.
lantz, J. 1958a. Etude des spores et pollens d’un echantillon purbeckien de File d’Oleron. Rev.
Micropaleont, Paris, 1, 33-37.

19586. Etude palynologique de quelques echantillons Mesozolquesdu Dorset (Grande-Bretagne).

Rev. Inst. Fr. Petrole, Paris, 13, 917-43.

Michael, F. 1936. Palaobotanische und kohlen petrographische Studien in der nordwestdeutschen
Wealdenformation. Ab/t. Preufi. geol. Landesanst. Berlin, NF 166, 79 pp., 4 pi.
pocock, s. a. j. 1962. Microfloral analysis and age determination of strata at the Jurassic-Cretaceous
Boundary in the Western Canada plains. Palaeontographica, Stuttgart, 111B, 1-95, 15 pi.

1964. Pollen and spores of the Chlamydospermiidae and Schizaeaceae from Upper Mannville

Strata of the Saskatoon area of Saskatchewan. Grana Palyn., Stockholm, 5, 2, 129-209, 7 pi.
potonie, r. 1962. Synopsis der Sporae in situ. Beih. Geol. Jb., Hannover, 52, 1-204, 19 pi.

1965. Fossile Sporae in situ. Forsch Ber. Landes NRhein-Westf, Koln, 1483, 1-74.

seward, a. c. 1894. Catalogue of Mesozoic plants in the Dept, of Geology, British Museum: The
Wealden flora. Part I. London.

1913. Contribution to our knowledge of Wealden floras. Quart. J. geol. Soc., London, 69, 85-

116, 4 pi.

singh, chaitanya. 1964. Microflora of the Lower Cretaceous Mannville Group, East-Central Alberta.

Bull. Res. Council Alberta, Edmonton, 15, 1-239, 29 pi.
stover, l. e. 1964. Comparison of three Lower Cretaceous spore-pollen assemblages from Maryland
and England. In Symposium: ‘Palynology in Oil Exploration’. Spec. Publ. Soc. Econ. Palaeont.
Min., Tulsa, 11, 143-52, 2 pi.

N. F. HUGHES,

J. C. MOODY-STUART

Department of Geology,
Sedgwick Museum,

Manuscript received 2 March 1965 Cambridge

VARIATION AND ONTOGENY OF SOME
OXFORDIAN AMMONITES: TARAMELLICERAS
RICH El (DE LORIOL) AND CRENIC ERAS
RENGGERI (OPPEL), FROM WOODHAM,
BUCKINGHAMSHIRE

by D. F. B. PALFRAMAN

Abstract. Statistical and ontogenetic studies of Taramelliceras richei (de Loriol) and Creniceras renggeri
(Oppel) have shown identity in their early stages. Variation in protoconch size is consistent and small, as is the
diameter of the nepionic constriction. Divergence in shell form occurs only at the onset of maturity, which in
C. renggeri begins at about 8 4 mm. and in T. richei at about 18-8 mm. It is concluded that the two ‘species’
are a sexually dimorphic pair. The name C. renggeri has priority.

A collection of Lower Oxfordian ammonites was made from the Woodham Brick
Company’s Pit, with the objective of studying the variation and ontogeny of selected
species. The pit is situated almost midway between Bicester and Aylesbury, near Ake-
man Street Station, Buckinghamshire (Grid. Ref. 707186), and exposes some seventy
feet of Oxford Clay (Arkell 1939). A lithological unit ‘The Lamberti Limestone’, a bed
one foot thick, separates the lowermost forty feet of clay (Athleta Zone) from the upper-
most thirty feet (Mariae Zone).

The Mariae Clays, as the beds of the Mariae Zone are here named, furnished the
material studied. Both ‘species’ examined in this paper occur, together, throughout the
Mariae Clays at Woodham; they are commonest in the lower ten feet of the Mariae
Zone (generally pyritized), becoming rare in the upper twenty feet of beds (generally
limonitized). They are both found together at all levels at Woodham. There is no
evidence to suggest that either Taramelliceras richei (de Loriol) or Creniceras renggeri
(Oppel) occurs outside the Mariae Zone in this country. However, both species appear
to have arisen in the Upper Lamberti Zone, being found in Haute-Saone (Maire 1908).
They are both found together in the Renggeri Marls (Mariae and Cordatum Zones)
of the Jura Mountains, where they are fairly common (de Loriol 1 898-9, 1 900 and Arkell
1939). As far as the author is aware they are not recorded above the Cordatum Zone;
though, as Arkell (1939) says ‘. . . exact details of range [are] usually not available’.

All the specimens collected are preserved as internal moulds, pyritic near the base of
the zone becoming limonitic towards the top. No original shell material has been found
among the collection of several thousand ammonites. In a few individuals the innermost
whorls are calcified, but rarely to a diameter of more than 3-4 mm. Some, which are
otherwise completely pyritized or limonitized, were found to have a calcareous proto-
conch.

Very few specimens were found preserved as pyrite or limonite moulds above a maxi-
mum diameter of 2 or 3 cm. and it seems that there is a maximum size above which am-
monites were not pyritized or limonitized. Internal moulds composed of clay, seen in

[Palaeontology, Vol. 9, Part 2, 1966, pp. 290-311, pis. 48-52.]

D.F.B. PALFRAMAN: SOME OXFORDIAN AMMONITES

291

situ, showed that some ammonites must have had a diameter of not less than forty
centimetres (see PI. 48, fig. 1); often in the centre of such moulds were to be seen
pyritized or limonitized nuclei of familiar dimensions.

Nearly five hundred specimens of Taramelliceras richei (de Loriol) and over two hun-
dred specimens of Creniceras renggeri (Oppel) have been examined in this work. All
specimens collected by, or given to, the author are in the University Museum, Oxford
and are prefixed by OUM, those borrowed from the British Museum (Natural History)
by BM. The following abbreviations are also used throughout the text: D = diameter,
W = whorl width, HH = whorl height, U = umbilical diameter.

VARIATION AND ONTOGENY OF TARAMELLICERAS RICHEI

(DE loriol)

Due to the conditions of preservation, the most frequently occurring form in which
Taramelliceras richei (de Loriol) was found at Woodham, was as internal moulds of the
phragmocone. The type material of T. richei is based largely on internal moulds of
phragmocones (de Loriol 1898). In the author’s collection of nearly five hundred speci-
mens less than 10 per cent, have any part of the body chamber preserved, and in none of
these is it complete. The species is a typical oppelid, being compressed and involute with
a small umbilicus. Whorl height is approximately half the diameter, whorl width is
approximately quarter the diameter. Coiling is regular in the phragmocone, but on the
body chamber of adults the umbilical seam begins to uncoil.

Protoconch. Less than 5 per cent, of the specimens of T. richei available were found
to have protoconchs preserved; in all some eight protoconchs were dissected out. These
are beautifully preserved as smooth pyritic moulds, showing details of the prosuture
(text-fig. 1 and PI. 48, fig. 4). In all cases the protoconchs are barrel-shaped with a width
greater than the diameter. The largest recorded diameter of a protoconch of T. richei is
0-30 mm., the smallest 0-25 mm. (see Table 1). In one specimen (OUM J25230) the
caecal mould is visible (text-fig. 1a, c).

Phragmocone. Just as the surface of the protoconch internal mould is smooth, so also
is the first whorl of growth. The nepionic constriction occurs after one complete whorl
of growth from the proseptum, at a diameter of about 0-53 mm., or roughly twice that
of the protoconch (Table 1).

The nepionic constriction is well-defined, especially in the ventral and ventro-lateral
regions, fading on the flanks and almost imperceptible near the umbilicus (text-fig.
2a, b; PI. 48, fig. 6). In T. richei there is no ornamental difference, on the internal
mould, between the whorl preceding the nepionic constriction and that immediately
succeeding. In some ammonites, Kosmoceras and Garantiana, Makowski (1962) noted
that the ornamentation characteristics of the full-grown shell appeared after the nepionic
constriction. Ontogenetic features of note are: a decrease in the W/D ratio from the
protoconch to a diameter of 0-8-1 -0 mm. and a similar decrease in the U/D ratio to
about the same diameter. From a diameter of 0-8-1 -0 mm. onwards, growth is almost
isometric until the onset of maturity which occurs at a diameter of about 18-20 mm.
This is graphically represented here (text-figs. 3 and 4). From a diameter of 0-8-1 -0 mm.
until the onset of maturity the HH/D ratio shows almost isometric growth (text-fig. 5).

D E F

text-fig. 1. Protoconchs of Taramelliceras richei (de Loriol); Mariae Zone, Woodham, Bucks., Eng-
land. a and d, ventral view; b, inclined apertural view; c and f, apertural view; e, side view, a and c
based on OUM J25230; b, d, e, and F based on OUM J25228. All X 100.

text-fig. 2. Diagram of the innermost whorls of Taramelliceras richei (de Loriol); Mariae Zone,
Woodham, Bucks., England, a and b, showing the position of the nepionic constriction; e, protoconch.
a, b, c, and D, drawn from OUM J25038; e, drawn from OUM J25234. All X 50.

Ribbing is completely absent until a diameter of about 4 mm. ; at this diameter radial
umbilical swellings begin, strongly near the umbilicus but fading on the flanks. The
ribs are shallow and distant and are initially eight to ten per whorl. It is not until
a diameter of about 8-10 mm. that ribbing is differentiated into primaries and inter-
calatories (PI. 50, fig. 1 a, c).

D. F. B. PALFRAMAN: SOME OXFORDIAN AMMONITES

293

table I. Measurements in millimetres of the dimensions of protoconchs and the diameter of the
nepionic constriction in specimens of Taramelliceras richei (de Loriol). All specimens from University
Museum, Oxford (OUM). All specimens from the Mariae Zone, Woodham.

Specimen

Nepionic

Number

Protoconch

Constriction

D

W

D

J25024

0-53

J25038

0-54

J25040

0-28

0-40

0-54

J25042

0-54

J25045

0-27

0-38

J25046

025

0-40

0-56

J25228

0-30

0-42

0-53

J25230

030

0-42

0-53

J25231

0-28

0-41

0-52

J25232

0-27

041

0-52

J25234

0-28

0-38

0-53

Differentiated ribbing is undoubtedly a characteristic of T. richei , all specimens showing
this feature to some degree, with a variation from relatively coarse ribbing to fine (PI. 49,
figs. 1 a, b, 2a, c). On the final whorl of all adult specimens there are two series of ribs;
primaries and intercalatories. The primary ribs generally number ten to twelve per whorl,
developing initially as radial umbilical swellings then fading on the flanks and finally
becoming relatively sharp and distinct on the ventro-lateral area. They are slightly
sinuous with marked ventro-lateral prorsiradiation (PI. 48, fig. 2a; PI. 49, figs, la, 2a,
4a; PI. 50, fig. la). Ribs always fade on the venter of internal moulds, but may not do
so on the actual surface of the shell. In some cases the venter is smooth (see PI. 49,
fig. lb), whereas in most instances the ribbing simply becomes weaker in the ventral
region (PI. 49, fig. 2b). On the last quarter of the adult phragmocone whorl, nearly all
specimens develop feeble ventral tubercles, which become stronger adorally (PI. 49,
fig. 4a, b; PI. 50, fig. la, b). The tubercles coincide at the point at which opposing ribs,
on either side of the venter, both primaries and intercalatories, would meet had they
continued uninterrupted across the venter.

The intercalatory ribs, on the final whorl of adult specimens, generally number about
forty. They are restricted to the ventro-lateral area and are identical in form with the
primary ribs in this region, by way of amplitude, sharpness and direction, for each
particular specimen. Between each primary rib are three to four intercalatories (PI. 48,
fig. 2a; PI. 49, figs. 2a, 4a; PI. 50, fig. la).

The mean maximum diameter of the phragmocone, of more than two hundred
adult specimens, was found to be 18-8 mm., with a standard deviation (a) of T51. The
coefficient of variation (V = lOOcr/m, where m is the mean maximum diameter, ds,
of the adult phragmocones) is 8-02. The greatest diameter at which the septa are
approximated was found to be 27 mm., the smallest 14 mm. (text-fig. 6). The phrag-
mocone generally comprised 6^-7 whorls from the proseptum. Variation in the whorl-
shape is here shown diagrammatically (text-fig. 7 and PI. 48, fig. 2d).

The sutural ontogeny of T. richei develops in a way which, in Jurassic ammonites,
is beginning to appear fundamental (Schindewolf 1954). The prosuture is significantly

294

PALAEONTOLOGY, VOLUME 9

100

100

B

B

DIAMETER (mm) —

text-fig. 3. Graphs showing the relationship between diameter and whorl width of Taramelliceras
richei (de Loriol), and of Creniceras renggeri (Oppel). The two graphs are superimposed: the diameter
axis is common, the width axis is divided into A and B scales, the B scale being one log cycle behind
the A scale. Graph A (W/D for T. richei) is controlled by the A scale, graph B (W/D for C. renggeri) is
controlled by the B scale. The fine dashed line on graph B is that superimposed from graph A (approxi-
mate best-fit line). The areas denoted by P (coarse dashed line) enclose measurements made on the
protoconchs : M is the mean of these measurements. Encircled dots represent measurements made at the
maximum diameter of complete adults: asterisks — measurements made on adult body-chambers:
crosses — measurements made at the adoral end of adult phragmocones: dots — measurements made on

immature phragmocones.

Inset. The same graphs plotted on a linear scale and havjng both axes common, a . W/D for T. richei ;
b, W/D for C. renggeri. The arrowed points, ds ? and ds 3, mark the position of the mean septate
diameter of mature specimens of supposed female and male respectively. All specimens from the

Mariae Zone, Woodham.

D. F. B. PALFRAMAN: SOME OXFORDIAN AMMONITES

295

text-fig. 4. Graphs showing the relationship between diameter and umbilical diameter of Taramel-
liceras richei (de Loriol), and of Creniceras renggeri (Oppel). The two graphs are superimposed: the
diameter axis is common, the umbilical diameter axis is divided into A and B scales, the B scale being
one log cycle behind the A scale. Graph A (U/D for T. richei) is controlled by the A scale, graph B
(U/D for C. renggeri ) is controlled by the B scale. The fine dashed line on graph B is that super-
imposed from graph A (approximate best-fit line). Conventions as for text-fig. 3.

Inset. The same graphs plotted on a linear scale and having both axes common, a, U/D for T. richei;
b, U/D for C. renggeri. The arrowed points, as 9 and ds 3, mark the position of the mean septate
diameter of mature specimens of supposed female and male respectively. All specimens from the

Mariae Zone, Woodham.

X

C 3803

296

PALAEONTOLOGY, VOLUME 9

text-fig. 5. Graphs showing the relationship between diameter and whorl height of Taramelliceras
richei (de Loriol), and of Creniceras renggeri (Oppel). The two graphs are superimposed : the diameter
axis is common, the height axis is divided into A and B scales, the B scale being one log cycle behind
the A scale. Graph A (HH/D for T. richei ) is controlled by the A scale, graph B (HH/D for C. renggeri)
is controlled by the B scale. The fine dashed line on graph B is that superimposed from graph A
(approximate best-fit line). Conventions as for text-fig. 3.

Inset. The same graphs plotted on a linear scale and having both axes common, a, HH/D for T. richei ;
b, HH/D for C. renggeri. The arrowed points, ds $ and ds <$, mark the position of the mean septate
diameter of mature specimens of supposed female and male respectively. All specimens from the Mariae

Zone, Woodham.

D. F. B. PALFRAMAN: SOME OXFORDIAN AMMONITES

297

different from the primary suture, having a very large ventral saddle. The primary
suture (text-figs. 1 and 8) is divided into : external (or ventral), lateral, U2, Ul5 and dorsal
lobes, with corresponding saddles between. These first-formed lobes have been called
‘protolobes’ by Schindewolf (1954): subsequent works (Schindewolf 1960, 1962, and
1963) have shown that these basic protolobes are common to all Jurassic ammonites
so far investigated. Addition of successive sutural elements in T. richei occurs at the

text-fig. 6. Histogram showing the variation in the diameter at which septation ceases (ds) in mature
specimens of Taramelliceras richei (de Loriol) and in Creniceras renggeri (Oppel). The T. richei range
is from 14-28 mm., the C. renggeri range from 5-13 mm. All specimens from the Mariae Zone,

Woodham.

umbilical seam; this is the ‘U-type’ ontogeny of Schindewolf (1954). The final sutures
of adult specimens are greatly differentiated into several complex saddles and lobes
(text-fig. 8a). Sutural approximation, generally regarded as a feature of maturity, is
commonly to be seen in most specimens with a phragmocone diameter of 18-19 mm.
(PI. 48, fig. 2a; PI. 49, figs. 2a, 4a, 4b; PI. 50, fig. 1 a).

In the adult sutures the largest saddles are the first and second lateral and the largest
lobes the ventral and lateral. These major sutural elements show the greatest range of
variation and have been compared on either side of the venter in the same specimen, and
on the same side of the venter at similar diameters, in different specimens. Comparing
the same sutural elements, on either side of the venter for any single specimen, shows
remarkably little variation (text-fig. 9). The greatest difference recorded is in the frilling
of the most adapical accessory lobe: even so this difference is somewhat subtle and does
not compare with the wide intraspecific sutural variation shown by Morrisiceras morrisi
(Oppel) and Clydoniceras discus (Sow.) mentioned by Arkell (1957).

E F G H

text-fig. 7. Cross-sections through the phragmocone of Taramelliceras richei (de Loriol). A, OUM
J25041 ; b, OUM J25021; c, OUM J25010; d, OUM J25239; e, OUM J25241; F, OUM J25236;
G, OUM J25005; h, OUM J25238. All x4. All specimens from the Mariae Zone, Woodham.

D. F. B. PALFRAMAN: SOME OXFORDIAN AMMONITES

299

text-fig. 8. Sutural ontogeny of TarameUiceras richei (de Loriol). A, OUM J25013, X 9 (D = 15 mm.);
B, OUM J25036, X 15 (D = 8-7 mm.); c, OUM J25036, X 18 (D = 6-4 mm.); d, OUM J25036, X 20
(D = 4-6 mm.); e, OUM J25036, x25 (D 3-13 mm.); f, OUM J25036, X 30 (D = 2-13 mm.);
g, OUM J25036, X 35 (D = 1-53 mm.); H, OUM J25036, X 50 (D = 0-85 mm.); i, OUM J25040, X 50
(D = 0-32 mm.); j, OUM J25040, x 50 (D 0-28 mm.), i, primary suture. J, prosuture. All specimens

from the Mariae Zone, Woodham.

300

PALAEONTOLOGY, VOLUME 9

text-fig. 9. Sutural variation in Taramelliceras richei (de Loriol). a, OUM J25230 (D = 12-2 mm.);
B, OUM J25228 (D = 12 0 mm.); c, OUM J25229 (D = 12-1 mm.). All X 7-5. All specimens from the

Mariae Zone, Woodham.

text-fig. 10. Sutural variation in Taramelliceras richei (de Loriol). A, OUM J25029 (D = 12 mm.);
B, OUM J25028 (D = 12-4 mm.); c, OUM J25009 (D = 11-8 mm); d, OUM J25035 (D = 12 mm.);
E, OUM J25024 (D = 12 mm.); f, OUM J25026 (D = 12 mm.). All X'7-5. All specimens from the

Mariae Zone, Woodham.

Variation between the same sutural elements of different specimens, at approximately
the same diameter, shows relatively marked differences (text-fig. 10). These differences
are generally of size, however, the disposition of the elements being essentially the same.
The most marked differences are between the specimens OUM J25029 and OUM

EXPLANATION OF PLATE 48

Fig. 1. Internal clay mould of a large Peltoceras{ ?) sp. ; basal Mariae Zone (about 5 ft. above the
Lamberti Limestone), Woodham, Bucks. X -fg-.

Fig. 2. Taramelliceras richei (de Loriol, OUM J25571. Adult phragmocone. a, side view; b, aper-
tural view. Bothx 3. ' . -

Fig. 3. T. richei, BM C. 39588; Oxford Clay, Scarborough, Yorkshire. X 1.

Fig. 4. T. richei, OUM J25234 (lost); Mariae Zone, Woodham, Bucks. Protoconch. X 50.

Fig. 5. T. richei, OUM J25672. Adult specimen, a, side view; b, ventral view. The body-chamber
occupies about | of the final whorl. Both X 3. ,

Fig. 6. T. richei, OUM J25038. Innermost whorls showing the nepionic constriction. X40.

Fig. 7. T. richei, BM 39642; Oxford Clay, Scarborough, Yorkshire. Stereopair of adult specimen
showing the nature of the peristome and ventro-lateral spines, x 1 .

Fig. 1 — unwhitened; Figs. 2a, b, 3, 5a, b, and 7 — whitened with ammonium chloride; Figs. 4 and 6
— whitened with magnesium oxide. All specimens from the Mariae Zone, Woodham, Bucks.,
unless otherwise stated.

Palaeontology, Vol. 9

PLATE 48

PALFRAMAN, Oxford Clay ammonites

D. F. B. PALFRAMAN: SOME OXFORDIAN AMMONITES 301

J25024 (text-fig. 10a, e). Generally speaking it seems fairly safe to say that the intra-
specific sutural variation in T. richei is extremely small.

Body-chamber. The body-chamber of adult specimens of T. richei is quite different from
thephragmocone. The most marked change is in the development of ventral and ventro-
lateral spines. In some specimens spines develop quite suddenly after the last septum in
both positions. Specimen OUM J25676 shows the ventral spines beginning first, the
ventro-lateral spines developing later (PI. 49, fig. 3a, b). The ventro-lateral spines are
paired and alternate with those on the venter, relative to the median plane. Not all
body-chambers, of specimens which are otherwise identical, become spinose: they
merely retain the ventral tubercles of the adult phragmocone (PI. 48, fig. 5a, b). How-
ever, there is probably a whole range between tuberculate venters only, and spinose
venters and ventro-lateral regions: this is supported by a collection of specimens from
the Oxford Clay of Warboys, Hunts. (BM C.4333, C.10651, C.15960, C.16009, C.16010,
C. 19286, and C. 19296), which shows precisely this range.

Ribbing on the body-chamber appears to be weaker than on the phragmocone, but
the very poor preservation of the body-chambers of Woodham specimens may affect
this feature. Some of the better preserved specimens from Warboys show well-defined
ribs, both primaries and intercalatories, even on the flanks of the body-chamber (BM
C.4333 and C.15960). The phragmocones of these specimens are otherwise identical
with those from Woodham.

As would be expected in the final stages of adult specimens, the body-chamber uncoils
at the umbilical seam (PI. 49, fig. 3b and text-fig. 4). Among the Woodham specimens the
author was unable to find an adult which showed the nature of the peristome. However,
among three specimens from the Oxford Clay of Scarborough, Yorkshire (BM 39642,
C. 39588, and C. 39589), labelled Proscaphites oculatus (Phillips), is one specimen (BM
39642) which shows the nature of the peristome (PI. 48, fig. 7). These specimens are
almost certainly conspecific with T. richei, developing ventral and ventro-lateral spines
(PI. 48, figs. 3 and 7): unfortunately the phragmocone is not preserved and the body-
chamber of the specimens is crushed. The peristome of BM 39642 is fairly simple, its
outline coinciding with the course of the primary ribs. The diameter of the almost
complete specimen (45 mm.), which has a body-chamber of three-quarters of a whorl,
is in agreement with the predicted size of complete adults from Woodham, assuming
that the Woodham specimens have a body-chamber of about three-quarters of a whorl.
If this is so, the total number of whorls from the proseptum is from 7-7§ in T. richei.

Among the better preserved specimens from Woodham, the phragmocone is a dull
black colour, whereas that of the body-chamber is a golden reflective (if pyritic) mould.
The dull black colour extends just on to the body-chamber, and in many cases a well-
defined line can be traced between the black and golden parts of the body-chamber. The
line is almost co-incidental with the course of the primary ribs but with slightly greater
ventral prorsiradiation. This line may represent the annulus (Crick 1898): however,
no muscle-scars were recorded from any specimen of T. richei from Woodham.

VARIATION AND ONTOGENY OF CRENICERAS RENGGERI (OPPEL)

Creniceras renggeri is a smaller ‘species’ than Taramelliceras richei, complete adults
rarely attaining a diameter of more than 20 mm. It appears that because of its small

302

PALAEONTOLOGY, VOLUME 9

size C. renggeri is generally to be found complete, or nearly so, as it rarely exceeds the
optimum size for complete preservation as a pyritic or limonitic mould in the Oxford
Clay of Woodham.

text-fig. 11. Cross-sections through Creniceras renggeri (Oppel) A, OUM J25294; B, OUM J25378;
c, OUM J25378 (inner whorls of b); d, OUM J25298; e, OUM J25305; F, OUM J25322; g, OUM
J25272. All except fig. c are x 4; fig. c x 30. All specimens from the Mariae Zone, Woodham.

table 2. Measurements in millimetres of the dimensions of protoconchs and the diameter of the
nepionic constriction in specimens of Creniceras renggeri (Oppel). All specimens from the University
Museum, Oxford (OUM). All specimens from the Mariae Zone, Woodham.

Specimen

Number

Protoconch

Nepionic

Constriction

D

W

D

J25305

0-29

0-40

J25308

0-30

0-42

J25347

029

0-42

J25364

0-28

0-41

0-54

J25368

0-28

042

0-53

J25378

0-29

041

0-56

Protoconch. Ontogenetically C. renggeri goes through identical stages as T. richei until
the onset of maturity. Protoconchs are identical in shape to, and vary within almost
the same limits as, those of T. richei (Table 2, cf. Table 1).

EXPLANATION OF PLATE 49

Taramelliceras richei (de Loriol). All specimens are from the Mariae Zone of Woodham, Bucks.

and have been whitened with ammonium chloride. All X 3.

Fig. 1. OUM J25482. Adult phragmocone. a, side view; b, ventral view. Fig. 2. OUM J25617. Adult
phragmocone. a, side view; b, ventral view. Fig. 3. OUM J25676. Adult specimen; the body-
chamber occupies the last half-whorl; ventral and ventro-lateral spines are developed, a, ventral
view; b, side view. Fig. 4. OUM J25455. Adult phragmocone. a, side view; b, ventral view.

Palaeontology, Vol. 9

PLATE 49

PALFRAMAN, Oxford Clay ammonites

D. F. B. PALFRAMAN: SOME OXFORDIAN AMMONITES

303

text-fig. 12. Sutural ontogeny of Creniceras renggeri (Oppel). A, OUM J25244, X 12 (D]='10-61 mm.);
b, OUM J25364, X 15 (D = 8 03 mm.); c, OUM J25364, X 16 (D = 6-08 mm.); d,OUM J25364,
X 25 (D = 3-18 mm.); E, OUM J25364 x 50 (D = 1-43 mm.); F, OUM J25364, X 50 (D = 0-99 mm.);
G, OUM J25364, X 50 (D = 0-38 mm.); h, OUM J25364, X 50 (D 0-33 mm.); i, OUM J25364,
X 50 (D = 0-28 mm.), h, primary suture, i, prosuture. All specimens from the Mariae Zone,

Woodham.

304

PALAEONTOLOGY, VOLUME 9

Phragmocone. The nepionic constriction occurs at a diameter of between 0-53 mm.
and 0-56 mm., compared with 0-52-0-56 mm. in T. richei. However, only three measure-
ments of the nepionic constriction were available for C. renggeri, compared with ten
for T. richei. The whorl shape (text-fig. 11) and sutural ontogeny (text-fig. 12) of C.
renggeri vary within very similar limits to those of T. richei (text-figs. 7 and 8).

A complete lack of ornament occurs until a diameter of about 4 mm., where shallow,
distant, radial, umbilical swellings develop. These swellings fade on the flanks and
do not re-form on either the venter or ventro-lateral regions. These features are con-
sistent with those recorded in T. richei at comparable diameters. The W/D, U/D, and
HH/D ratios are identical to those of T. richei at similar growth stages (text-figs. 3, 4,
and 5).

The final whorl of the phragmocone is, in most cases, seen only on removal of the
body-chamber. Septation generally ceases at a diameter of 8-9 mm. The mean
diameter at which septation ceases, in over two hundred adults, is 8-4 mm.: standard
deviation (a) is 1-1: coefficient of variation (V) is 12-8 (text-fig. 6). Specimen OUM
J 25246, with a maximum septate diameter of 17 mm., is almost certainly a very large
adult C. renggeri (PI. 50, fig. 5). However, variation in over two hundred specimens
of C. renggeri is not continuous to such a large diameter, the range of continuous
variation being 5-13 mm. (text-fig. 6), and it seems that the specimen in question is an
extreme form.

The final whorl of the mature phragmocone is typically oppelid in character, being
compressed and involute and with a narrow umbilicus. There are 4 J — 5 T complete whorls
from the proseptum to the end of the phragmocone in adult specimens.

Radial umbilical swellings are developed, which fade on the flanks. These generally
number about ten per whorl and are comparable with those, in number, direction, and
nature, of T. richei at a similar diameter (PI. 50, fig. 2a, c and cf. PI. 50, fig. lc). The
ventro-lateral area is smooth, as is the venter, except on the final quarter whorl of adults,
where feeble ventral tubercles are almost always developed. The tubercles become well-
defined immediately before the beginning of the body-chamber (PI. 50, figs. 2a, 3;
PI. 51, figs. \a, 2a, b). Uncoiling of the umbilical seam in adults begins before the final
septa are deposited (PL 50, figs. 2, 4, 5; and PL 52, figs. 2b, 3, 5), becoming markedly
so on the body-chamber (PL 50, figs. 2a, 3; PL 51, figs. 2a, 3; PL 52, figs. 1, 3, and
text-fig. 13).

Sutures were drawn at a diameter of approximately 8 mm. from nine specimens,
showing the first and second lateral saddles and the ventral and lateral lobes. Differences
in the same suture on either side of the venter are decidedly small (text-fig. 14); whereas

EXPLANATION OF PLATE 50 ,

Fig. 1. Taramelliceras richei (de Loriol), OUM J25430. a, side view of adult phragmocone; b, ventral
view of adult phragmocone; c, preparation of the inner whorls.

Figs. 2-4. Creniceras renggeri (Oppel). Fig. 2. OUM J25244. a, side view of adult specimen; b, ventral
view of adult body-chamber; c, preparation of the inner whorls. Fig. 3. OUM J25360. Rostrate
adult specimen. Fig. 4. OUM J25282. Body-chamber with almost complete peristome, a, side view;
b, ventral view; c, apertural view.

Fig. 5. C. renggeri (?), OUM J25246. Giant adult male?

Specimens have been whitened with ammonium chloride. All from the Mariae Zone, Woodham,
Bucks. All figs. X 3.

Palaeontology , Vol. 9

PLATE 50

PALFRAMAN, Oxford Clay ammonites

D. F. B. PALFRAMAN: SOME OXFORDIAN AMMONITES

305

differences between sutures of different specimens, drawn at approximately the same
diameter, are more marked (text-fig. 15). The range of variation, qualitatively assessed,
appears to be similar to that shown by T. richei (cf. text-figs. 9 and 10). Sutural approxi-
mation can clearly be seen in a large number of specimens (PI. 50, figs. 2a, 3; PI. 51,
figs. 1 a, 3 a, 12; PI. 52, figs. 2a, b, 3, 4, 5, 6)(.

text-fig. 13. Diagrams of the umbilical seam of Creniceras renggeri (Oppel). The positions of the
last two sutures are indicated, c, position of venter and lappet indicated, a, OUM J20662; b, OUM
J20673; c, OUM J25364; d, OUM J20675; e, OUM J25377. All X 10. All specimens from the

Mariae Zone, Woodham.

Body-Chamber. In nearly all specimens of C. renggeri a large proportion of the final
whorl is body-chamber: in complete specipiens the body-chamber occupies two-thirds
to three-quarters of a volution (PI. 51, figs. 2a, 4; PI. 52, figs. 1-3, 5, 6). In complete
adult specimens the total number of whorls, counting from the proseptum, is from
5L-6L. Morphology apart, the most striking difference between the body-chamber
and phragmocone is that of colour. When pyritized the body-chamber is preserved as
a golden reflective internal mould, whereas the phragmocone is a dull black colour.
The body-chamber is frequently well preserved and some specimens have almost com-
plete apertures. ( J f

Spiral striae are developed on the body-chamber of all adult specimens: this feature
is particularly well seen on specimens OUM J25248, J25265 and J25392 (PL 51, fig. 3a;
PI. 52, figs. 2a, 6). The striation is only feebly developed on the phragmocone (PI. 51,
fig. 3a), if at all. f i

The ventral tubercles of the phraigmocone give rise to large truncated spines on the
venter of the adult body-chamber. A great deal of variation is to be seen between speci-
mens concerning the spines. At one extreme a few specimens have perfectly smooth
venters (PI. 52, figs. 2, 4), while others develop enormous spines (PI. 52, figs. 1, 5).
Among the Woodham population, however, a complete range from smooth venters to
coarsely spined ones is to be seen (PI. 51, figs. 5-12 and Arkell 1939, pi. 9, figs. 15-27).
The majority of specimens have spines similar to those of OUM J25392 (PI. 52, fig. 6).
In complete adults there are generally nine to twelve spines on the body-chamber;
however, the total number depends largely on the size of the spines: specimens with
finely denticulate venters having as many as twenty-four (PI. 51, fig. 1 1).

306

PALAEONTOLOGY, VOLUME 9

text-fig. 14. Sutural variation in Creniceras renggeri (Oppel). A, OUM
J25335 (D = 8-5 mm.); b, OUM J25327 (D = 8 0 mm.); c, OUM J25329
(D = 8-3 mm.). All x 15. All specimens from the Mariae Zone, Woodham.

text-fig. 15. Sutural variation in Creniceras renggeri (Oppel). A, OUM
J25367 (D = 7-8 mm.); b, OUM J25366 (D = 7-6 mm.); c, OUM J25326
(D = 7-8 mm.); d, OUM J25341 (D = 8-2 mm.); e, OUM J25345 (D =
8-2 mm.); F, OUM J25355 (D = 81 mm.). All X 15. All specimens from
the Mariae Zone, Woodham.

D. F. B. PALFRAMAN: SOME OXFORDIAN AMMONITES 307

The general inclination of the spines is slightly rursiradial (PL 51, fig. 1 a; PI. 52,
fig. 6); rectiradial direction is fairly common (PI. 51, fig. 2a) and prorsiradial direction
rare and, at best, very slight.

In all adult specimens which have a complete body-chamber the ventral spines fade
out completely just before the aperture (PI. 50, figs. 2 a, 3, 4a; PI. 51, figs. 2a, 4 and
PL 52, figs. 3, 6). Spines may not develop near the aperture, or, if they do, they are
formed as solid shell, a mould of which would not reflect their original shape. The final
spine of OUM J20695, an incomplete adult, is truncated near the base (PL 51, fig. 5);
obviously not broken, the spine may have formed as solid shell with only the lowermost
part hollow.

On well-preserved spines, fine ridges can be seen parallel to the edge of the spine
(PL 51, figs, la, d). Arising from the ventral spines in many individuals are sinuous
ventro-lateral ribs, which often fade on the flanks and may, or may not, reappear weakly
in the umbilical region (PL 51, fig. la). In specimens with a denticulate venter, these
shallow ribs are better defined in the ventro-lateral region, weakening on the flanks
and umbilical region (PL 51, fig. 11). Individuals with no spines rarely develop ribs; at
best they are very feeble and atypically weakest in the ventro-lateral position (PL 52,
figs. 2a, b, 4). Because of the shallowness and large width of the ribs, they are only seen
in well-preserved specimens.

The peristome of adult specimens of C. renggeri develops constrictions, flares, lappets
and rostra and is considerably more ornate than that of T. richei. Constrictions in adults
vary from feebly developed to quite strongly so (PL 51, fig. 2a; PL 50, fig. 4a). In the
case of well-developed constrictions the peristome is flared (PL 50, fig. 4a, c): however,
the conditions of preservation may have been unfavourable for faithfully recording
the true shape of both lappets and rostra in this species. Lappets are generally short,
close to the umbilical seam and triangular to elongate in shape (PL 51, figs. 2a, 4;
PL 52, figs. 2b, 3). Flattened specimens labelled Ammonites cristatus Sowerby, from the
Oxford Clay of Scarborough, Yorkshire, though poorly preserved, show the lappets
to be considerably more elongate than shown by the Woodham specimens (PL 52,
figs. 7, 8). The preserved part of the lappet of OUM J25325 (PL 51, fig. 4) is identical to
the early part of the lappet of BM C. 27545 (PL 52, fig. 8): the lappet of the former is
obviously broken and may well have been the same shape and size as that of the latter.
The two forms (A. cristatus and C. renggeri ) are probably conspecific, the Yorkshire
individuals giving a truer outline of the lappet. In some cases the Woodham specimens
have lappets which, in cross-section, are concave relative to the median plane (PL 50,
fig. 4c; PL 51, fig. 2b). The rostra of Woodham specimens (PL 50, figs. 3, 4a; PL 51,
figs. 2a, 4; PL 52, fig. 3) are almost identical with those of the Yorkshire specimens
(PL 52, figs. 7, 8) in both size and inclination.

As on T. richei a line appears on the body-chamber of C. renggeri immediately after
the last septum on adult specimens, possibly representing the position of the annulus
(Crick 1898). The course of this line is almost identical with that recorded for T. richei
(PL 51, figs. 10, 12). Faint muscle scars were seen on the venter of several individuals:
they are particularly prominent in a smooth-ventered specimen, OUM J25376, where
they occur immediately either side of the mid-ventral line in the earliest part of
the body-chamber and extend adorally for about 50° of a volution (PL 52, fig.
4a, b).

308

PALAEONTOLOGY VOLUME 9

DISCUSSION ON SEXUAL DIMORPHISM

De Blainville (1840) was the first author to point out the possibility of separate sexes
in the ammonites. More recent work (Callomon 1955 and 1963; Makowski 1962) has
shown that among synchronous large and small forms of initially identical ammonites,
there is sound reason for accepting de Blainville’s claim. By analogy with living cephalo-
pods, the large form is thought to be the female and the small form the male: these are
the macroconchs (M) and microconchs (m), respectively, of Callomon (1955 and 1963).

The present author accepts the criteria of Makowski (1962) relating to sexual dimor-
phism in ammonites :

(1) identical initial stages of ontogeny in both (large and small) forms, and identity
of their phylogeny,

(2) lack of intermediate forms in adult (gerontic) stages,

(3) presence of both forms in the same strata,

(4) numerical ratio of the supposed sexes (sex ratio), comparable to that observed in
living forms.

In the foregoing account it has been shown that T. richei and C. renggeri agree in
early ontogenetic stages but it has yet to be shown that they can be ‘phylogenetically
identified’. As Callomon (1963) has pointed out; "[Taramelliceras and Creniceras ]
both commence together at the base of the Oxfordian, basal Mariae Zone (Renggeri
Marls); both are there already very variable (see Arkell 1939), and I [Callomon] have
the feeling that they are merely dimorphic companions thence throughout the rest of

EXPLANATION OF PLATE 51

Creniceras renggeri (Oppel) from the Mariae Zone, Woodham, Bucks.

Fig. 1. OUM J25245. Adult specimen, a, side view; b, apertural view; c, ventral view; d, ridges on
ventral spines (x 10). Fig. 2. OUM J25249. Adult with almost complete peristome, a, side view;
b, apertural view. Fig. 3. OUM J25248. Adult specimen, a, side view; b, apertural view; c, ventral
view. Fig. 4. OUM J25325. Adult specimen with almost complete peristome. Fig. 5. OUM J20695.
Adult specimen showing truncated spine. Fig. 6. OUM J25252. Adult specimen. Fig. 7. OUM J25254.
Adult specimen. Fig. 8. OUM J20662. Adult specimen. Fig. 9. OUM J25250. Adult specimen.
Fig. 10. OUM J25250. Unwhitened adult specimen showing the course of the annulus (?). Fig. 11.
OUM J25253. Adult specimen. Fig. 12. OUM J25255. Unwhitened adult specimen showing the
course of the annulus (?). Specimens have been whitened with ammonium chloride, except figs. 10
and 12. Photographs by the author. All x 3, except Id which is x 10.

EXPLANATION OF PLATE 52

Figs. 1-8. Creniceras renggeri (Oppel), from the Mariae Zone, Woodham, Bucks.

Fig. 1. OUM J25247. Adult specimen. Fig. 2. OUM J25265. Adult specimen with almost complete
peristome, a, side view, showing spiral striae; b, side view, showing lappet. Fig. 3. OUM J25364.
Complete adult specimen showing lappet and rostrum. Fig. 4. OUM J25376. Incomplete adult
showing muscle scars, a, side view; b, ventral view. Fig. 5. OUM J25356. Almost complete adult
specimen. Fig. 6. OUM J25392. Almost complete adult specimen.

Fig. 7. BM 39644; Oxford Clay, Yorkshire. Complete specimen, probably an adult.

Fig. 8. BM C.27545; Oxford Clay, Yorkshire. Complete specimen, probably an adult.

Fig. 9. C. renggeri <3 and ?. Reconstruction: S based on OUM J25392 and BM C.27545 (see figs. 6
and 8); $ based on OUM J25660 and BM 39642 (see PI. 48, fig. 7). x 1. Specimens have been whitened
with ammonium chloride. Photographs by the author. All X 3, except fig. 9 which is natural size.

Palaeontology, Vol. 9

PLATE 51

I a

I b Ic

la

lb

5

6

7

8

9

10

II

12

PALFRAMAN, Oxford Clay ammonites

Palaeontology, Vol. 9

PLATE 52

PALFRAMAN, Oxford Clay ammonites

D. F. B. PALFRAMAN: SOME OXFORDIAN AMMONITES

309

the Upper Jurassic.’ Both Taramelliceras and Creniceras extend into the Kimmeridgian
(Arkell 1957, p. L125; see Arkell, Kummel and Wright 1957; Holder 1955 and Ziegler
1957 and 1959). As far as the author is able to determine, neither genus is found at
higher levels than the Kimmeridgian.

Among the Woodham oppelids there are certainly no intermediates between T. richei
and C. renggeri', there is a ‘morphological hiatus’ (Makowski 1962) of about one whorl of
growth and the largest adult C. renggeri is not as large as the smallest T. richei
(text-fig. 6). The two ‘species’ occupy the same strata, and, as far as can be determined,
both begin at the same level, basal Mariae Zone, at Woodham.

The ratio of the large to small forms, T. richei to C. renggeri , is two to one. This can
only be an approximate ratio as specimens from Woodham with a diameter of less
than 8-9 mm. cannot reliably be differentiated into large and small forms. By analogy
with living cephalopods, which show sexual dimorphism (Sweet 1964, p. K9; Makowski
1962, pp. 56—58), it is not unreasonable to contend that the two forms, T. richei and
C. renggeri, are sexual dimorphs: they show sexual differences, both size and sex ratio,
comparable with those displayed by living cephalopods. A reconstruction of the di-
morphic pair has been produced (PI. 52, fig. 9). It is to be inferred that the larger is
the female.

TAXONOMY

The taxonomic problems, which acceptance of the sexual dimorphism theory bring,
have been discussed at length by Callomon (1955 and 1963) and Makowski (1962).
The present author is of the opinion that sexual dimorphs of the same species should
have the same specific name (Makowski 1962), being differentiated by the zoological
symbols S and ?. The specific name accepted for a dimorphic pair should be that of the
first named ‘species’ (following the law of priority) and, as far as the author can assess,
there should be no more complication in ammonite taxonomy than in, say, ostracod
taxonomy where sexual dimorphs of the same species are given the same specific name.

Jeannet (1951) does not include the genus Creniceras Munier-Chalmas in the family
Taramelliceratinae Spath 1928. The genus, along with Bukowskites nov. and Popanites
Rollier, is classified by Jeannet as: Tncertae Sedis’, family Oppeliidae Bonarelli 1894.
As the present results show, C. renggeri and T. richei are sexual dimorphs of the same
species (= C. renggeri ) and on this basis the author agrees with Arkell, Kummel and
Wright’s (1957) classification in which the genus Creniceras is included in the subfamily
Taramelliceratinae.

SUMMARY AND CONCLUSIONS

Creniceras renggeri (Oppel) and Taramelliceras richei (de Loriol) from Woodham
are identical in every feature up to a diameter of about 8 mm. In the late stages of the
adult phrogmocone, both ‘species’ develop feeble ventral tubercles which become
ventral spines on the body-chamber. T. richei differs from C. renggeri in that ventro-
lateral spines may be developed on the body-chamber.

The variation shown by the sutures at diameters of 8 mm. and 12 mm., in C. renggeri
and T. richei respectively, is small and qualitatively similar. Sutural ontogeny is identical
throughout : the sutures of T. richei become more complex in the adult stages, but always
at larger diameters than are recorded for adult phragmocones of C. renggeri.

310

PALAEONTOLOGY, VOLUME 9

In adults the body-chamber of both ‘species’ uncoils at the umbilical seam. The peri-
stome of T. richei is relatively simple ; that of C. renggeri is ornate, developing flares,
constrictions, rostra, and lappets.

Recent work has added substance to de Blainville’s theory of sexual dimorphism in
ammonites. By analogy with living cephalopods it seems probable that the large form is
the female and the small form the male.

It is concluded that Creniceras renggeri (Oppel) and Taramelliceras richei (de Loriol)
are really male and female respectively of the same species, and that the specific name
of the dimorphic pair should be that of the earliest named form. It is finally considered
that the inclusion of the genus Creniceras in the subfamily Taramelliceratinae is correct.

Acknowledgements. For advice and discussion the author is very grateful to Dr. M. R. House. The
author is also indebted to Dr. J. H. Callomon for the gift of a large number of ammonites from
Woodham and to Mr. J. M. Edmonds of the University Museum, Oxford, and Dr. M. K. Howarth
of the British Museum (Natural History) for the loan of specimens in their care. The work has been
done during the tenure of a S.R.C. research studentship.

REFERENCES

arkell, w. j. 1939. The ammonite succession at the Woodham Brick Company's pit, Akeman Street
Station, Buckinghamshire, and its bearing on the classification of the Oxford Clay. Quart. J. geol.
Soc. Lond. 95, 135-222.

1957. Sutures and Septa in Jurassic Ammonite Systematics. Geol. Mag. 94, 235-48.

kummel, b., and wright, c. w. 1957. In Treatise on Invertebrate Paleontology, ed. R. C. Moore.

(L) Mollusca 4, Mesozoic Ammonoidea. L80-L437. Kansas Univ. Press.
blainville, m. h. d. de. 1840. Prodrome d’une monographie des ammonites. Extrait du Supplement
du Dictionaire des Sciences naturelles, 1-34, Paris, Bertrand.
callomon, J. h. 1955. The ammonite succession in the Lower Oxford Clay and Kellaways beds at
Kidlington, Oxfordshire, and the zones of the Callovian Stage. Phil. Trans. Roy. Soc. B 239, 215-64.

1963. Sexual dimorphism in Jurassic ammonites. Trans. Leic. Lit. & Phil. Soc. 57, 21-56.

crick, g. c. 1898. On the muscular attachment of the animal to its shell in some fossil Cephalopoda
(Ammonoidea). Linn. Soc. London Trans., ser. 2, Zoology, 7, 71-113.
holder, h. 1955. Die Ammoniten-Gattung Taramelliceras im Sudwestdeutschen Unter- und Mittel-
malm. Palaeontographica, Stuttgart, 106 A, 37-153.
jeannet, a. 1951. Die Eisen- und Manganerze der Schweiz. Stratigraphie und Palaeontologie des
oolithischen Eisenerzlagers von Herznach und seiner Umgebung. Beitr. Geol. Schweiz, ser. 13,
5, 1-240.

loriol, p. de. 1898-9. Etudes sur les mollusques et brachiopodes de l’Oxfordien inferieur ou Zone
a Ammonites renggeri du Jura bernois. Soc. Palaeont. Suisse Mem. 25-26, 1-197.

1900. Etudes sur les mollusques et brachiopodes de l’Oxfordien inferieur ou Zone a Ammonites

renggeri du Jura ledonien. Ibid. 27, 1-196.

maire, v. 1908. Contribution a la connaissance de la Faune des Marnes a Creniceras renggeri dans la
Franche-Compte septentrionale. 1 , Le Callovien et l’Oxfordien inferieur a Authoison (Haute-Saone).
Bull. Soc. Grayloise d'Emul. 11, 143-63.

makowski, h. 1962. Problem of sexual dimorphism in ammonites. Palaeontologia Polonica, 12, 1-92.
schindewolf, o. h. 1954. On development, evolution and terminology of the ammonoid suture line.
Harvard Univ. Mus. Comp. Zool., Bulk 112, no. 3, 217-37.

1960. Studien zur Stammesgeschichte der Ammoniten. Lief. I, Abh. math.-naturw. Kk Akad.

Wiss. Mainz, NR. 10, 635-745.

1962. Studien zur Stammesgeschichte der Ammoniten. Lief. II, Ibid. 8, 425-572.

1963. Studien zur Stammesgeschichte der Ammoniten. Lief. Ill, Ibid. 6, 285-432.

sweet, w. c. 1964. In Treatise on Invertebrate Paleontology, ed. R. C. Moore. (K) Mollusca 3,
Cephalopoda — General Features. K4-K13. Univ. of Kansas Press.

D. F. B. PALFRAMAN: SOME OXFORDIAN AMMONITES 311

zieglhr, b. 1957. Creniceras dentatum [Ammonitacea] im Mittel-Malm Siidwestdeutschlands. Neues
Jb. Geol. Palaontol., Mh. 1956, 533-75. Stuttgart.

1959. Evolution in Upper Jurassic Ammonites. Evolution, 13, 229-35.

D. F. B. PALFRAMAN

Department of Geology and Mineralogy,
University Museum,

Manuscript received 21 April 1965 Oxford.

C 3803

Y

THE LOWER LIASSIC AMMONITES
NEOMICROCERAS GEN. NOV. AND
PARACYMBITES

by D. T. DONOVAN

Abstract. Neomicroceras commune gen. et sp. nov., family Eoderoceratidae, is described and figured from the
Raricostatum Zone of the British Lower Lias. The genus Paracymbites Trueman and Williams 1927, from the
same zone, is revised from the original types and other material, and placed in the family Oxynoticeratidae.

Study of the Lower Liassic ammonites from the Geological Survey borehole at
Witney (Apley Barn), Oxfordshire, has led to the recognition of a new genus, here
named Neomicroceras , and to the revision of another, Paracymbites Trueman and
Williams, which was previously known only from a figure of the inner whorls. Both,
genera are now fully described and figured.

The following abbreviations are used: BM = Palaeontology Department, British
Museum (Natural History). GSM = Geological Survey and Museum, London.

SYSTEMATIC DESCRIPTIONS

Order ammonoidea
Family eoderoceratidae Spath 1929
Subfamily eoderoceratinae Spath 1929
Genus neomicroceras gen. nov.

Type species. Neomicroceras commune sp. nov.

Diagnosis. Evolute ammonites of the family Eoderoceratidae with simple, radial ribs
on the whorl-side. The ribs may bear small tubercles at their ventral ends. The suture-
line is very simple: the saddles are broad, rounded, and never constricted at their bases;
the external saddle is bifid or trifid, the other saddles are not normally divided. The
genus is a homeomorph of Crucilobiceras but is distinguished from it by the different,
much simpler suture-line and by the shorter smooth stage, up to 0-35 cm. diameter
compared with about 0-5 cm. in Crucilobiceras. The single known species is small,
never exceeding 20 mm. diameter.

Remarks. Neomicroceras is placed in the Eoderoceratinae on account of its similarity
to the inner whorls of the less strongly ornamented species of Crucilobiceras. Its range,
as at present known, lies entirely above that of Crucilobiceras and it could have evolved
from that genus. No intermediate forms are known. The suture-line of Crucilobiceras
passes through a stage similar to that of Neomicroceras during its ontogeny, as shown
by Schindewolf (1962, text-figs. 1076, c). By a diameter of about TO cm., however, the
Crucilobiceras suture-line has many more indentations than that of Neomicroceras and

[Palaeontology, Vol. 9, Part 2, 1966, pp. 312-18, pi. 53.]

D. T. DONOVAN: LOWER LIASS1C AMMONITES 313

the external and sometimes other saddles have constricted bases, a feature never observed
in the new genus.

Other genera of Eoderoceratinae in the British Raricostatum Zone ( Eoderoceras ,
Bifericeras ) show sutural development and complexity very like that of Crucilobiceras.

Neomicroceras commune sp. nov.

Plate 53, figs. 1-4, text-fig. 1

Holotype. GSM 87550 from the Lower Lias of the Apley Bam Borehole, near Witney, Oxfordshire.
Pcuatypes. GSM 87551, 87553-6, from the same borehole as the holotype.

Other material. GSM 87557-60 from the Witney (Apley Barn) Borehole. Three specimens from
the Dundry Borehole, Somerset, GSM 87561-3. Three specimens from Ballintoy, Co. Antrim, in the
Ulster Museum and Art Gallery, Belfast, on a tablet numbered 674-1931. Geological horizon: In the
Witney (Apley Barn) Borehole the genus ranges from 464 to 486J ft. and lies entirely within the range
of the ammonite Leptechioceras. It therefore occurs in the Macdonnelli Subzone of the Raricostatum
Zone as currently defined (Dean, Donovan, and Howarth 1961, p. 460). The specimens from the
Dundry Borehole are from the same subzone. The Irish specimens are not dated.

Description. The shell is evolute, with very little overlap between successive whorls.
The largest individuals, around 16 mm. in diameter, consist of five and a half whorls.
The first three whorls increase rapidly in size and the umbilicus is about 40 per cent, of
the diameter at a size of 5 mm. On the last two and a half whorls the increase is more
gradual and the umbilicus widens to about 52 per cent, at full size. The whorl height
is about 27 per cent, of the diameter, and the whorl thickness 25 per cent, at full
size.

The whorl side is smooth or faintly undulating to a diameter of about 35 mm. when
regular ribbing begins. The rib-frequency rises from about 25 to the whorl at a diameter
of 6 mm. to from 33 (holotype) to 40 (paratype no. 87554) at full size in the Witney
(Apley Barn) material. Two of the Ballintoy examples have closer ribbing. They are
provisionally regarded as individual variants as there is not enough material available
to demonstrate the existence of a distinct, more closely ribbed species. The variation in
rib-frequency within the Witney assemblage is normal for a Liassic ammonite species.
The ribs are very slightly convex forwards and bear small tubercles at the ventral ends.
All the Witney specimens are pyritized internal moulds but the Ballintoy examples
retain the shell and show that the tubercles, blunt on the internal mould, are sharp on
the shell. The Irish material also shows that the shell is striate between the ribs. In some
individuals the ribs continue very faintly across the arched venter.

The body-chamber is about three-quarters of a whorl. Individuals retaining the body-
chamber have the last two or three suture-lines approximated and are assumed to have
been adult.

Examples of the external suture-line, which is the distinguishing feature of Neomi-
croceras, are shown in text-fig. 1. The external saddle is divided into two or three
folioles. The other two saddles are normally broadly rounded but the larger sometimes
shows incipient, and occasionally more distinct subdivision. This feature varies between
successive suture-lines of the same individual (text-fig. la, c).

314

PALAEONTOLOGY, VOLUME 9

Family oxynoticeratidae Hyatt 1875
Genus paracymbites Trueman and Williams 1927

Type species. Paracymbites obsoletus Trueman and Williams by original designation.

The holotype of the type species (GSM 51464) is from the Lower Lias of Stanton
Fields cutting, about 1 0 miles north-east of Cheltenham, Gloucestershire (see Richardson,
1918, p. 17). The other fossils recorded from the cutting by Richardson indicate that the
Oxynotum, Raricostatum, and Jamesoni Zones were exposed in the cutting, but there
is no more precise information as to the horizon of the holotype.

text-fig. 1 . External suture-lines of Neomicroceras commune gen. et sp. nov. from the Lower Lias,
Raricostatum Zone, of the Witney (Apley Barn) Borehole, Oxfordshire ( a-c ) and from Ballintoy, Co.
Antrim (d). a, Paratype, GSM 87553; b , Paratype, GSM 87554; c, Holotype, GSM 87550, last three
suture-lines; d, Ulster Museum and Art Gallery, 674-1931. Arrows mark mid-ventral lines. All X 8.

Diagnosis. Small involute ammonites rarely exceeding 15 mm. in diameter. The earliest
whorls have a roughly semicircular section; on the last whorl the whorl-section becomes
parallel-sided with arched venter. Each whorl almost completely covers the preceding
one. The umbilicus is about one-quarter of the diameter, and may open out slightly
with the last whorl, ill-defined simple folds or ribs are present on the whorl-side of some
individuals. A very faintly marked keel may be present on the earlier part of the last
whorl. The aperture is plain with a blunt ventral rostrum. The body-chamber is probably
about three-quarters of a whorl. The external suture-line consists of three saddles each
of broadly arched form with slight indentations. Paracymbites differs from other Lower
Jurassic ammonites of broadly similar aspect as follows: Protocymbites (Hettangian) is
distinguished by its stronger and more regular ribbing; Cymbites (Sinemurian) by its
much more excentric coiling, the adult whorl-height being about 35 per cent, of the
diameter compared with 45 per cent, in Paracymbites ; Metacymbites (Pliensbachian)

EXPLANATION OF PLATE 53

Pigs. 1-4. Neomicroceras commune gen. et sp. nov., Lower Lias, Raricostatum Zone, of the Witney
(Apley Barn) Borehole, Oxfordshire. 1, 2, Holotype, GSM 87550, X 4; 3, Paratype, GSM 87555, x6;
4, GSM 87559, x4.

Figs. 5-12. Paracymbites dennyi (Simpson), Lower Lias, Raricostatum Zone. 5-7, GSM 51464, the
holotype of P. obsoletus Trueman and Williams, Stanton Fields cutting, Glos., x4; 8, GSM 87552,
Witney (Apley Barn) Borehole, Oxfordshire, X 4; 9, 10, GSM Zd 3034, Folly LaneBrickpit, Chelten-
ham, Glos., x 4; 11, 12, GSM Zd 3035, the holotype of P. undulatus Trueman and Williams, Folly
Lane Brickpit, Cheltenham, Glos., X 4.

Palaeontology, Vol. 9

PLATE 53

DONOVAN, Lias ammonites

D. T. DONOVAN: LOWER LIASSIC AMMONITES

315

by more excentric coiling, more depressed body-chamber, and suture-line with con-
stricted bases to the saddles; and Onychoceras (Upper Toarcian) by more excentric
coiling and stronger ornament, and more deeply incised suture-line.

Pavacymbites dennyi (Simpson)

Plate 53, figs. 5-12, text-fig. 2

Ammonites Dennyi Simpson 1843, p. 9.

Ammonites Dennyi Simpson 1855, p. 38.

Phylloceras Loscombi, Sowerby (pars)', Blake 1876, p. 296.

Ammonites Dennyi Simpson 1884, p. 66.

Oxynoticeras dennyi, Simpson sp. ; Buckman 1909, pp. lb, c, pi. 7, figs. 1, 1 a, 2, 3.

Cymbites sp. Richardson 1918, p. 17, pi. 2, fig. 31.

Oxynoticeras dennyi (Simpson) Buckman; Spath 1925, p. 111.

Paracymbites obsoletus Trueman and Williams 1927, p. 247, pi. 27, fig. 8 a-c.

Paracymbites undulatus Trueman and Williams 1927, p. 248.

Cymbites ? dennyi (Simpson); Howarth 1962, p. 104.
non Cymbites cf. obsoletus (Trueman and Williams); Schindewolf 1961, p. 214, pi. 30, figs. 14-16,
text-figs. 18, 19.

Types. Syntypes are four pyritized internal moulds in Whitby Museum on a tablet numbered 470.
Three were figured for the first time by Buckman (1909). The fourth, a nucleus 4 mm. in diameter still
attached to matrix, is probably conspecific with the others. It has not been figured. Buckman wrote
‘1 and 2 agree with simpson’s measurement; but fig. 1 seems to agree best with his description: there-
fore it is presumably the holotype’. The present writer disagrees; for example, Simpson wrote ‘[height
of the] outer whorl about ^ the diameter’, a statement which applies to Buckman's fig. 2 but not to
his fig. 1. Howarth (1962) accepts the original of Buckman’s fig. 1 as holotype or lectotype, and the
present writer prefers to regard it as lectotype designated by Buckman.

The holotype of Paracymbites obsoletus Trueman and Williams (1927), is in the Geological Survey
Museum, no. 51464. It is refigured in the present paper, Plate 53, figs. 5-7. The holotype of Paracym-
bites undulatus Trueman and Williams is GSM Zd 3035. It has not been previously figured and is now
shown in Plate 53, figs. 11, 12.

Other material. The species appears to be fairly common in the Lower Lias clays of Gloucestershire
and Worcestershire and examples from various localities are preserved in the British Museum (Natural
History) (nos. C6111o, C16666-9, C67060, and C67061) and in the Geological Survey Museum
(nos. 24629, Zd 2954 (figured by Richardson, 1918, pi. 2, fig. 31), Zd 3034, Zd 3048, Zd 3072-4.
The Witney (Apley Barn) Borehole yielded GSM 87552. The species does not appear to have been
found on the Dorset coast.

Geological age. The horizon of Paracymbites dennyi in Yorkshire, as recorded by Simpson, would fall
within the lower part of the Raricostatum Zone ; Mr. L. Bairstow kindly informs me that he has not
re-found the species during his work at Robin Hood’s Bay. The Gloucestershire and Worcestershire
specimens are not usually well dated, being finds from clay banks and spoil heaps. The Witney
(Apley Barn) Borehole specimen comes from the Raricostatum Zone.

Description. The shell form has been described in the generic diagnosis. The coiling of
the last whorl is excentric and variable; in some individuals, including the lectotype, the
umbilicus opens out suddenly at the beginning of the last half-whorl, but in others, such
as Buckman’s fig. 2, this opening out is more gradual. The lectotype represents one
extreme of variation in this character. The adult umbilicus varies between about 23 and
28 per cent, of the diameter (25 per cent, in the lectotype), and the whorl height just

316

PALAEONTOLOGY, VOLUME 9

behind the adult aperture is around 45 per cent, of the diameter. The shell maybe almost
smooth but is usually plicate, the folds— they are hardly ribs — having a slightly sigmoidal
course following that of the growth-lines, and dying out half to two-thirds of the way
from the umbilical suture to the mid-ventral line. The number of these folds varies

text-fig. 2. External suture-lines of Paracymbites dennyi (Simpson) from the Lower Lias, Raricostatum
Zone, from the north end of the Hunting Butts tunnel, Cheltenham, Glos. (a), Folly Lane brick pit,
Cheltenham ( b , c), and the Witney (Apley Barn) Borehole, Oxfordshire (d). a, GSM Zd 3048; b, GSM
Zd 3035, the holotype of P. undulatus Trueman and Williams; c , GSM Zd 3034; d, GSM 87552.

Arrows mark mid-ventral lines. Straight lines are guide lines. All X 8.

widely, from a few irregularly developed ones to over twenty to the whorl. The venter
is obtusely angular for the first half of the last whorl but usually becomes rounded on
the second half. On account of the frequency of specimens complete with body chamber
about 15 mm. in diameter, the change in coiling which precedes the attainment of this
size, and the absence of any larger specimens, a size of about 15 mm. was probably adult.
Suture-lines are often irregularly spaced so that it is difficult to judge whether or not
they are approximated before the body-chamber. The external suture-line (text-fig. 2)
has three saddles on each side of the venter. The external one is roughly parallel-sided,
and moderately indented; this character varies among individuals, and sometimes
between the two sides of the same individual (text-fig. 2d). The other two saddles vary
from being smooth (text-fig. 2d) to slightly indented (text-fig. 2b, c). The suture-line
ascends towards the umbilical suture so that the lateral lobes lie well above the guide-
line.

Synonymy. The holotype of P. obsoletus is wholly septate and only 8 mm. in diameter.
The shell-form, ornament, and suture-line are sufficiently close to P. dennyi for P.
obsoletus to be regarded as synonymous with Simpson’s prior species. The other species
placed in the genus by Trueman and Williams in 1927, P. undulatus, was not illustrated
by them at the time. The holotype is now figured (pi. 53, figs. 11, 12) and is regarded
as inseparable specifically from P. dennyi. Trueman and Williams did not say what they
considered to be the differences between their two species.

Relationships and systematic position. The present writer is impressed by the similarity
between Paracymbites and Cheltonia, an oxynoticeratid genus of the late Oxynotum
Zone. Cheltonia is small (usually 25 mm. or less) and its affinities are beyond doubt, for
it is almost indistinguishable from the inner whorls of Oxynoticeras except for slightly
excentric coiling and a modified body chamber. Cheltonia has in common with Para-
cymbites: change of whorl-section with growth, ornament on whorl side, aperture, and
angular venter giving place to rounded on the last half-whorl. The main differences lie

n

a

b

c

d

D. T. DONOVAN: LOWER LIASSIC AMMONITES

317

in the greater whorl thickness of Paracymbites, and the absence of the corrugated venter
which precedes the aperture in Cheltonia. On account of this close resemblance Para-
cymbites is placed in the family Oxynoticeratidae.

The similar genus Cymbites has been the subject of an exhaustive study by Schinde-
wolf (1961), who regards Paracymbites as a synonym of Cymbites. The present writer
disagrees with this view. In fairness it should be noted that Paracymbites is almost
impossible to interpret from Trueman and Williams’s original figures, and that the
body-chamber, which provides the clearest difference from Cymbites, was not figured
by them. Schindewolf figured three ammonites (1961, pi. 30, figs. 14, 15, 16 a, b) as
Cymbites cf. obsoletus (Trueman and Williams), but the present writer believes that they
are not related to that species. The reasons for this are the coiling, which closely re-
sembles that of Cymbites proper, and the ornament which passes undiminished in
strength across the whorl-side on to the venter, a character of Cymbites (cf. Schinde-
wolf's other figures of the genus) but not of Paracymbites.

Trueman and Williams did not place Paracymbites in a family, but suggested that it
might be ‘a dwarfed member of the Deroceratidae’ [recte Eoderoceratidae] (1927,
p. 248). In the Treatise (Arkell, 1957, p. L240) it was placed, as a subgenus of Cymbites,
in the subfamily Cymbitinae Buckman 1919 of the family Arietitidae. Schindewolf
(1961, p. 229) recognized a separate family Cymbitidae, Paracymbites being regarded
as a synonym of Cymbites. I have remarked above that I disagree with Schindewolf on
this point, and believe that his relegation of Paracymbites to the synonymy resulted
from misidentification with that genus of German fossils which really belong to Cym-
bites. The present revision has shown that the closest similarities of Paracymbites are
with Cheltonia and it is accordingly placed in the family Oxynoticeratidae and regarded
as a homeomorph of Cymbites. I support Schindewolf in regarding the Cymbitidae as
an independent lineage having its origin early in the Lower Lias; my earlier view (1957,
p. 416) that Cymbites was evolved from Eucigassiceras is not supported by evidence
from recent boreholes.

Acknowledgements. I wish to thank Dr. H. Ivimey-Cook of the Palaeontological Department, Geo-
logical Survey, for help in various ways. Dr. M. K. Howarth looked out material for me at the British
Museum (Natural History). The photographs are by Messrs M. Holliday and B. Nettleton. The paper
is published by permission of the Director of the Geological Survey of Great Britain.

REFERENCES

arkell, w. J. 1957. In R. C. Moore (ed.). Treatise on Invertebrate Paleontology, Part L: Mollusca 4.

Cephalopoda: Ammonoidea. New York: Geol. Soc. Amer.
blake, J. f. 1876. Cephalopoda, tate, r. and blake, j. f. The Yorkshire Lias. London, 261-330.
buckman, s. s. 1909. Yorkshire type ammonites. 1, (1). London.

dean, w. t., donovan, d. t., and howarth, m. k. 1961. The Liassic ammonite zones and subzones
of the north-west European province. Bull. Brit. Mas. (Nat. Hist.) (Geol.), 4 (10), 435-505, pis.
63-75.

donovan, D. T. 1957. Notes on the species Cymbites laevigatas (J. de C. Sowerby) and on the genus
Cymbites Neumayr. Geol. Mag. 94, 413-20.

howarth, m. k. 1962. The Yorkshire type ammonites and nautiloids of Young and Bird, Phillips,
and Martin Simpson. Palaeontology, 5, 93-136, pis. 13-19.

Richardson, l. 1918. The geology (Lias and superficial deposits) of the Cheltenham-Stratford-on-
Avon Railway (G.W.R.). Trans. Woolhope Nat. Fid. CL, vol. for 1914-17, 137-57, pis. 1-3.

318

PALAEONTOLOGY, VOLUME 9

schindewolf, o. h. 1961. Die Ammoniten-Gattung Cymbites im Deutschen Lias. Palaeontographica,
117, Abt. A, 193-232, pis. 29-31.

1962. Studien zur Stammesgeschichte der Ammoniten. Lieferung II. Ab/t. Akad. Wiss. Lit.

Mainz (Math.-naturw. Kl.) Jhrg. 1962, nr. 8, pp. 1 11-257.
simpson, m. 1843. A monograph of the ammonites of the Yorkshire Lias. London.

1855. The fossils of the Yorkshire Lias; described from nature. London and Whitby.

1884. The fossils of the Yorkshire Lias; described from nature. 2nd ed. London and Whitby.

spath, L. F. 1925. Notes on Yorkshire ammonites. I. — On the genus Oxynoticeras Hyatt. Naturalist,
Lond. April 1925 (no. 819), 107-12.

trueman, A. e. and williams, d. M. 1927. Notes on some Lias ammonites from the Cheltenham
district. Proc. Cotteswold Nat. Fid. Cl. 22 (3), 239-53, pis. 27, 28.

D. T. DONOVAN
Department of Geology,
The University,

Manuscript received 5 March 1965 Hull.

RADIOLARIA FROM THE NAM U RIAN OF

DERBYSHIRE

by B. K. HOLDSWORTH

Abstract. Albaillella pennata sp. nov. and Popofskyellum undulatum Deflandre are described from calcareous
concretions at the horizon of the goniatite Reticuloceras paucicrenulatum Bisat and Hudson. The diagnosis of
the genus Albaillella is emended. It is suggested that A. pennata and P. undulatum may possibly have been favoured
by depths of water greater than those necessary for the proliferation of Namurian spumelline Radiolaria, and
that depths favouring these two forms may have been attained only rarely during the deposition of Namurian
goniatite bands in Derbyshire and Staffordshire.

The occurrence of well-preserved Radiolaria associated with Namurian goniatites in
North Staffordshire and South-west Derbyshire has been reported (Holdsworth 1964).
The goniatites and Radiolaria are found in syndepositional calcite concretions — -
‘bullions’ — enclosed in shale, these bullions being closely comparable with the cal-
careous concretions of the Ohio Shale (Devonian) from which Foreman (1959, 1963)
described rich radiolarian faunas.

The Namurian bullions appear to contain some of the best-preserved Carboniferous
Radiolaria yet known, and other than the Ohio Shale occurrence, the faunas of some
lower Visean phosphatic nodules described by Deflandre (1946, 1952, 1958) and of some
upper Visean nodules (Demanet 1938), there are no records of Palaeozoic Radiolaria
which can be studied in comparable detail. Palaeozoic Radiolaria are most commonly
found in siliceous sediments and the faunas described only from thin-section studies
(cf. Hinde 1899, Aberdeen 1940). Though the problem of separating the Namurian
Radiolaria completely from the carbonate matrix has not been solved, nevertheless it has
been possible to study many hundreds of excellent specimens on limestone surfaces
etched with hydrochloric acid. In such preparations both external and internal struc-
tures can be determined more conveniently and with greater certainty than in thin-
section.

Radiolarian faunas at sixteen successive goniatite horizons have been examined (cf.
Floldsworth 1964), the zonal range being from low E2b.l to low R,b, Arnsbergian
Stage to mid-Marsdenian Stage, some two-thirds of the Namurian Series. The very
great majority of the faunas are largely composed of spumelline forms belonging to the
genera Entactinia, Entactinosphaera, and Polyentactinia. The detailed taxonomic work
on these spumellines is still incomplete, but it seems clear that there is no obvious,
regular change in species present from horizon to horizon comparable with the marked
and irreversible trend of the successive goniatite faunas.

Differences which are seen between radiolarian faunas at different horizons, and
between different collections from essentially the same horizon, are mainly ones of
population density and of relative abundance of a few rather well-marked species. The
form with the most notably fluctuating abundance, however, is the new species of the
genus Albaillella (suborder Albaillellina) described below.

[Palaeontology, Vol. 9, Part 2, 1966, pp. 319-29, pi. 54.]

320

PALAEONTOLOGY, VOLUME 9

Non-spumelline Radiolaria are usually rare members of the Namurian faunas.
Albaillella pennata sp. nov. has been found only at the horizon of Reticuloceras pand-
er emulation Bisat and Hudson, being abundant in two collections from a single locality
but absent in two other collections from different localities. A closely related form,
Albaillella aff. pennata, occurs at the lower horizon of Homoceras subglobosum Bisat.
Popofskyellmn undulatum Deflandre is very sparsely associated with A. pennata at the
horizon of R. paucicrenulatum but has not been seen elsewhere.

Both the genera Albaillella and Popofskyellum were first described from the Visean
of the Montagne Noire and Cabriere. Neither genus has previously been described
from other levels, nor from any British rocks.

SYSTEMATIC DESCRIPTIONS
Subclass radiolaria Muller 1858
Order porulosida Haeckel 1887
Suborder albaillellina Deflandre 1953

Diagnosis. Fossil Radiolaria with bilaterally symmetrical, elongate, siliceous shells,
having a thin and usually imperforate wall and possessing internally non-axial rays — -
‘columelles’ (columellae) of Deflandre — eventually merging with the shell wall. Outside
the shell is developed a skeleton, variable in detail.

Remarks. Deflandre (1952) suggested that a new order should probably be created for
the genus Albaillella , but did not erect the order Albaillellidea until the following year
(1953n). The order was defined fully by Deflandre in 1958 (p. 2278) and his diagnosis is
quoted above. The Albaillellids are omitted by Campbell and Moore (1954) from the
Treatise on Invertebrate Paleontology, Part D (cf. Deflandre 1960, p. 214), but in the
framework of the Treatise classification it appears that the group should be considered
as constituting a suborder rather than an order. In relegating the Albaillellidea to a sub-
order the present writer follows Foreman (1963, p. 285). (In Foreman (loc. cit.) the
words ‘Suborder Albaillellina ? Deflandre 1953’ should appear immediately above the
line ‘Family Albaillellidae ? Deflandre 1952’ (Foreman in litt. 1964).)

Two families, Albaillellidae and Lapidopiscidae (Deflandre 1958), are included in the
Albaillellina. Whilst recognizing a very broad similarity of form and structure between
the Albaillellidae, type family of the Albaillellina, and the Cyrtoidea, Deflandre (1952)
considered the bilateral symmetry of the Albaillellidae to exclude them from the
Cyrtoidea and from the Nassellina as a whole. This total separation of Albaillellina and
Nassellina is accepted here, but it may be noted that rudimentary bilateral symmetry is
expressed in the sagittal ring and spicule arrangement of some simple members of the
Nassellina (cf. Campbell, in Campbell and Moore 1954, fig. 8) and that there is at least
some analogy, possibly even homology, between the H-frame of Albaillella (cf. De-
flandre 1952, fig. 5) and the spicuiar structures of simple members of the Nassellina with
complete, unisegmental lattice shells. In the absence of demonstrable homologies, how-
ever, the Albaillellids are best considered as a separate suborder.

Family albaillellidae Deflandre 1952

Diagnosis. Fossil Radiolaria with thin, conical, bilaterally symmetrical, lamellar shells.
Shells built on a framework of which the principal part, formed by two longitudinal

B. K. HOLDSWORTH: RADIOLARIA FROM DERBYSHIRE 321

rods of dissimilar strength joined by a curved transverse bar (the H-frame), is exterior
to the shell and opposite the basal aperture.

Remarks. The most characteristic feature of the family is the externally developed
H-frame. The only known genus is AlbailleUa, though Foreman (1963) tentatively
referred Ceratoikiscum Deflandre and Holoeciscus Foreman to the family.

Genus albaillella Deflandre 1952
Type species. AlbailleUa paradoxci Deflandre.

Emended diagnosis. Albaillellina in which the conical lattice shell may or may not be
flattened bilaterally in the plane of the H-frame. Lattice shell may bear one or two
hollow wings in the plane of the H-frame or may lack wings. Lattice shell sometimes
shows a transverse segmentation or sometimes an oblique banding. Both rods of the
external H-frame are produced into the interior of the lattice shell as longitudinal rays
(columellae of Deflandre). They are of dissimilar strength, the stronger being developed
from the stronger ray of the H-frame. The weaker (ventral) columella is sometimes
indistinct. The stronger (dorsal) columella is usually attached to the inner surface of the
lattice by paired lateral apophyses — ‘trabecules’ (trabeculae) of Deflandre — and in some
cases both rays are thus attached.

Remarks. Deflandre’s several published accounts of forms attributed to AlbailleUa are
not sufficiently detailed to establish the full variation encountered within the Visean
material with which he deals, or to decide how closely individual species agree with the
original generic diagnosis. In the first publication on AlbailleUa (1952) a generic diagnosis
is given, but no species names are mentioned in the text. Three species, A. paradoxa,
A. cornuata, and A. undulata, and one variety, A. paradoxa var. gibbosa , are described
only by drawings and named only in captions. The specific descriptions have not
subsequently been amplified, though var. gibbosa was later given the status of a species
(1960, p. 216, caption to pi. 1, fig. 23).

It is unclear from the published descriptions whether Visean species of AlbailleUa
invariably possess only a single columella — as appears to be the case in A. paradoxa ,
the type species (Deflandre 1952, p. 872, and fig. 4) — or whether both rods of the H-
frame continue into the shell in the form of columellae, as is suggested by Deflandre
(19536, p. 408, fig. 307a, b). According to Deflandre (in lift. 1965) Visean forms of the
genus always, in fact, possess two columellae [as does the new Namurian species
pennata] but the weaker (ventral) columella ‘often shows a tendency to become effaced
by inclusion in the shell wall’. Deflandre (in litt.) states that in Visean species the
stronger (dorsal) columella is ‘often’ attached to the shell wall by trabeculae, and that
in ‘some individuals’ both dorsal and ventral columellae are so attached, the condition
found in A. pennata. It should be noted that the columellae are simple continuations of
the H-frame rods, not ‘branches’ from them as is suggested by the original description
(Deflandre 1952, p. 872).

Clearly, the organization of the Namurian species is sufficiently similar to that of the
Visean forms for all to be included in a single genus. Apertural spines and an obliquely
banded shell (cf. Deflandre 19536, pp. 431-2) should not be considered as generic

322 PALAEONTOLOGY, VOLUME 9

features since they are not invariably present in Visean species and are absent from the
Namurian species.

text-fig. 1. Albaillella pennata sp. nov.

a. Diagrammatic representation of longitudinally bisected specimen to show the main architectural
elements. a\, a2, dorsal and ventral rods of H-frame; b, cross-bar of H-frame; cl, c2, dorsal and ven-
tral internal columellae; d, d, trabeculae; e, segment;/, constriction; g, pore; hi, h2, dorsal and ventral

wings; /, wing rib; j, wing spine; k, apical spine; /, wall of lattice shell.

b, H-frame. Drawing from specimen, X 375.

Anatomical nomenclature of Albaillella. The main architectural elements of frame and
shell are illustrated diagrammatically in text-fig. 1. The side of the shell bearing the
stronger columella is termed ‘dorsal’, as in the bicolumellate Lapidopiscidae (Deflandre
1958) which lack the H-frame. In Albaillella pennata the lattice shell is divided by trans-
verse constrictions into a number of units and a similar, though less marked, partition
of the lattice appears to be present in A. undulata. Each unit is here termed a ‘segment’,
but the segmentation of the Albaillella lattice is not homologous with the segmentation
in the Nassellina. In the Nassellina the first segment contains the spicular structures, if
present, and these structures are thus isolated from all subsequent segments. There-
fore, if the internal skeletal elements of the Albaillellina are considered homologous with
the spicular structures and sagittal ring of the Nassellina (see above), then the entire
shell of Albaillella must be considered homologous with a single nasselline segment,
comparable with the unisegmental lattice of the Archipiliicae.

The segmented lattice of A. pennata differs from the banded lattices (Deflandre 1952,
p. 873; 19536, p. 432) of A. paradoxa and A. cornuata in that the bands of these latter
species are obliquely, not transversely, arranged. Thus portions of more than one band

B. K. HOLDSWORTH: RADIOLARIA FROM DERBYSHIRE 323

are present at the periphery of the basal aperture in banded shells. In segmented shells
the aperture is in a single segment.

The ‘wings’ of A. pennata are hollow portions of the lattice developed dorsally and
ventrally, similar to the single outgrowth of the shell in A. cornuata, and differing from
the ‘horns’ of Lapidopiscum (Deflandre 1958, p. 2279) in that they contain no portions
of the columellae. It should be noted that Deflandre uses ‘horn’ in a quite different sense
with respect to Popofskyellwn (see below). In this case the horns are slender, solid, or
hollow spines developed from the cephalis, not containing or continuous with columellae
or spicules. These structures in Popofskyellwn are therefore better termed simply
‘cephalic spines’.

The outer margins of the wings in A. pennata are usually thickened to form longi-
tudinal ‘wing ribs’ which may be produced beyond the tip of the wings as ‘wing spines’.
In his ‘hypothetical reconstruction’ of the A. cornuata frame Deflandre (19536, fig. 307a)
shows a lateral branch from the ventral columella in a position corresponding to the
wing of A. cornuata. It is now known that there is no such connection between out-
growth and columella (Deflandre, in litt. 1965). The A. cornuata wing is comparable to
the wings of A. pennata in that it is developed solely from the shell.

AlbaiUella pennata sp. nov.

Plate 54, figs. 1-3, 5-7

Winged nassellinid, Holdsworth 1964, p. 698.

Locality. Upper Dove Valley, South-west Derbyshire, England.

Horizon. Faunal band of Reticuloceras paucicrennlatum, Reticuloceras circumplicatile Zone (Ria),
Kinderscoutian Stage (Rff, Namurian Series, Carboniferous.

Diagnosis. The shell bears prominent dorsal and ventral wings developed in a single
plane, is conical from apex to the level of emergence of the wings, and tubular there-
after. In the most typical form the tubular portion of the shell constitutes about one-
third of the total length, the overall outline being essentially triangular, but elongate
variants occur in which the tubular portion may constitute half the length of the shell.
Slightly oblique transverse constrictions divide the shell into segments and nine seg-
ments can be detected in complete specimens, though the first three are small and
indistinct. A slender, slightly curved apical spine, hollow at the base, is developed from
the first segment and there is generally a sudden increase in width of the lattice shell at
the fourth segment. The wings arise most commonly at or about the level of the sixth
segment and usually bear weak, longitudinal wing ribs, sometimes produced into wing
spines. There is a general tendency for the dorsal wing to be developed slightly nearer
the apex than the ventral wing. Both longitudinal rods of the H-frame continue into
the shell interior as columellae, lying in the plane of the wings. The columellae are of
markedly dissimilar strength, the stronger (dorsal) arising from the stronger rod of the
H-frame. The columellae converge towards the apex of the shell, extending at least to
the third segment and possibly to the apical segment. Along their entire length both
columellae are attached to the adjacent lattice wall by pairs of small lateral branches
(trabeculae), members of a pair being slightly offset vertically. The branches contact

324

PALAEONTOLOGY, VOLUME 9

the lattice shell at the constrictions (text-fig. 2a), and commonly there is indication of
slight thickening of the shell along the apical sides of constrictions. The lattice shell is
imperforate except for the final segment, where a girdle of irregular pores may be
developed close to the constriction (text-fig. 2b). The transverse bar of the H-frame,
only very rarely preserved completely in the available material, is slightly curved and
ornamented with short spines (text-fig. 1 b).

text-fig. 2. Albaillella pennata sp. nov.

a , Diagram showing relationship between constrictions of lattice shell (dashed lines) and trabeculae
of the internal columellae.

b, Final and penultimate segments with portion of dorsal rod of H-frame, showing girdle of irregular
pores in final segment. Drawing from specimen, X 250.

Variation. Specimens assigned to A. pennata are somewhat variable. The most distinct
variant is an elongate form in which the tubular portion of the shell is up to half the
total shell length and the wing spines are very long, extending to or beyond the level of
the mouth. In the typical triangular form there is variation in the degree to which the
segments are defined, in the development of wing ribs and wing spines, and in the size
and position of the wings. In an extremely rare variant the development of the ventral

EXPLANATION OF PLATE 54

All figures X 250.

Figs. 1-3, 5-7. Albaillella pennata sp. nov. 1, Holotype, LZ 3651. 2, Topotype, LZ 4740, showing
segmentation and incomplete apical spine. 3, Topotype, LZ 4739, showing marked inequality in
strength of rods of H-frame and difference in level of the wings. 5, Paratype, LZ 3652, showing very
well-marked segmentation. 6, Topotype, LZ 4737, elongate form showing segmentation and portion
of wing spine. 7, Paratype, LZ 3653, elongate form with wing rib produced into long wing spine.

Figs. 4, 8. Popofskyellum undulatum Deflandre. 4, Very incomplete specimen showing one lateral
cephalic spine and the stronger of the two longitudinal ribs, LZ 4738. 8, Specimen showing the
widely conical, segmented lattice of undulose outline, one lateral cephalic spine, and the stronger
of the two longitudinal ribs, LZ 4736.

Palaeontology, Vol. 9

PLATE 54

HOLDS WORTH, Namurian radiolaria

B. K. HOLDSWORTH: R ADIOLARI A FROM DERBYSHIRE

325

wing is delayed until the final segment. Occasionally pores are present in both final and
penultimate segments, and some specimens lack pores completely.

Material. Deposited at the Geological Survey, Leeds. Holotype, LZ 3651 (PI. 54, fig. 1), typical trian-
gular form, but with segmentation rather weakly developed; length from tip of apical spine (probably
broken) to base of dorsal rod of H-frame 0-24 mm. LZ 3658 is closely comparable with the holotype.
Paratype, LZ 3652 (PI. 54, fig. 5) and LZ 3654 show typical triangular shape with well-developed
segmentation. LZ 3657, typical form, laterally broken, showing columellae internally. LZ 3655, triangular
form with ventral wing weak, delicately spined and developed closer than usual to mouth. Paratype, LZ
3653 (PI. 54, fig. 7), extreme elongate form with long wing spine; length from tip of apical spine
(probably broken) to damaged aperture 0-31 mm. LZ 3656, specimen with shape intermediate between
typical and elongate forms, having very weak ventral wing with strong, curved spine. Specimens cited
are on the etched surfaces of two small slabs of limestone, and detailed instructions for locating speci-
mens are deposited with the material. Figured specimens not cited above also have Geological Survey,
Leeds, registration numbers.

text-fig. 3. Albaillella aff. pennata sp. nov. Drawings from specimens, a. Broken
specimen showing apex of lattice shell with wings and one internal columella with
remnants of trabeculae, X 203. b, Broken specimen showing complete H-frame, two
internal columellae and wings, X 180. c. Broken specimen with only fragments of
lattice shell, showing part of H-frame, and two internal columellae bearing remnants

of trabeculae, x 164.

Remarks. A. pennata differs from the four Visean species of Albaillella so far de-
scribed in possessing two wings and a rather markedly segmented shell. It differs
also in that specimens invariably show a well-differentiated ventral columella bearing
trabeculae.

A. pennata has been seen only at the horizon of R. paucicrenulatwn, but poorly pre-
served specimens of a closely related form, determined as A. aff. pennata , are present
at the lower horizon of Homoceras subglobosum Bisat. In A. aff. pennata (text-fig. 3 a-c),
as in A. pennata, there are two columellae, both bearing trabeculae, but the shell is more
slender and less obviously segmented, and the two wings are smaller and set more
closely to the apex.

326

PALAEONTOLOGY, VOLUME 9

Suborder ? nassellina
Family popofskyellidae Deflandre 1964

Diagnosis. Fossil Radiolaria with a perforated, bell-shaped shell in which two opposed,
longitudinal, lateral rods are incorporated. Cephalis simple. Podome with two branches.

Remarks. Though the greater part of the shell has a bilateral symmetry Deflandre (1964)
excluded the family from the Albaillellina on the grounds that the arrangement of
cephalic spines lacks this symmetry. The family differs also in having a perforate shell
and lacking internal columellae, the longitudinal ribs being integral parts of the shell,
homologous with the wing ribs of A. pennata. In the Namurian examples seen these
longitudinal ribs or rods are of unequal strength.

Genus popofskyellum Deflandre 1964
Type species. Popofskyellum pulchrum Deflandre (1960, pi. 1, fig. 26).

Diagnosis. Shell perforated, bell-shaped or conical, with two opposed lateral ribs or
rods, usually projecting at the base. Podome with two branches. Cephalis with un-
branched spines, frequently with a short, conical, tricostate spine. Poorly known,
asymmetrical trabecular system basally.

Popofskyellum undulatum Deflandre
Plate 54; figs. 4, 8

Popofskyellum undulatum Deflandre 1964, pp. 3055-8; figs. 3, 4, 13, 14.

Remarks. All five incomplete Namurian specimens of the genus so far seen agree closely
in size and architecture with P. undulatum, described by Deflandre from the Visean.
The most complete specimen (PL 54; fig. 8) has a cephalis and five other rather well-
marked segments as in the holotype (cf. Deflandre 1964, fig. 14), giving the characteristic
undulose outline to the shell, which is finely perforated overall and of muslin-like
texture. The longitudinal rods of unequal strength arise in the region of the second
extra-cephalic segment and extend beyond the periphery of the mouth. The total number
of lateral cephalic spines cannot be determined, but there is at least one per specimen,
and also a slender apical spine.

Popofskyellum is very rare in the Namurian faunas studied, being seen only at the
horizon of R. paucicrenulatum, where it is associated with abundant specimens of A.
pennata. The genus also appears to be rare in the Visean faunas, for Deflandre’s four
species are based on a total of only about twenty specimens (Deflandre 1964, p. 3055).

ECOLOGY OF ALBAILLELLA PENNATA
Foreman (1963, p. 267) noted marked differences in relative abundance and mor-
phology of radiolarian species between three concretions studied, all collected at
‘approximately the same horizon’ and separated by a total horizontal distance of only
some four miles. The radiolarian faunas of the R. paucicrenulatum faunal band provide
an even more striking example of apparent lateral variation at a single horizon.

B. K. HOLDSWORTH: R ADIOLARI A FROM DERBYSHIRE 327

The band was studied at three localities. At locality 1 in the Upper Dove Valley (loc.
179 of Holdsworth 1963) two bullions with R. paucicrenulatum, at exactly the same
level, 1 foot apart horizontally, contained rich spumelline faunas abundantly associated
with A. pennata and with very rare P. undulatum. Both these forms were completely
absent from a bullion with the same goniatite at locality 2, 20 yards away and also from
a fourth bullion with the goniatite at locality 3, 1 mile away in the Upper Manifold
Valley (loc. 177 of Holdsworth 1963), which contained a rich but exclusively spumelline
fauna. The possibility that the bullions at the three different localities are parts of three,
vertically separated goniatite bands is remote. The goniatites present in all four collec-
tions are typical forms of R. paucicrenulatum (Holdsworth 1963, pp. 142-4) and in no
known succession is this species seen to occur more than once. It must be accepted that
the four bullions consist of sediment deposited during the single relatively brief time
span marked by the existence in the area of R. paucicrenulatum.

This is not to say, however, that the two bullions from locality 1 and those from
localities 2 and 3 constitute precisely contemporaneous sediment samples, even though
the majority of faunal bands in this portion of the succession average only a foot or
so in thickness. From the perfect three-dimensional preservation of goniatites in bullions
it is clear that these concretions grew very early in diagenesis, before the oozes and
muds had suffered any appreciable compaction. To judge by the total crushing of many
globose goniatites in Namurian shale bands, an inch of shale must represent many
inches of original sediment, and at the time the bullion concretions formed it seems
probable that many inches, even feet, of highly water-charged mud and ooze were
present below the sediment-water interface. As the R. paucicrenulatum bullions are
relatively small, approximately seven inches deep at maximum, it is possible that the
two bullions of locality 1 and those of localities 2 and 3 developed at three slightly
different levels in the single sheet of sediment containing R. paucicrenulatum.

Thus there is a high probability that the variation between faunas at the three localities
is a vertical, not a true lateral, variation. During the single time span marked by the
existence of R. paucicrenulatum there were probably at least two distinct periods, one
in which A. pennata coexisted abundantly with the goniatite and spumelline Radiolaria,
and one in which only the goniatite and spumellines were present in the area.

To suppose that this indicates an extremely brief stratigraphical range for A. pennata
is unrealistic in view of the evidently long ranges of the spumelline species and the
presence of the closely comparable A. aff. pennata at a single lower horizon. The most
reasonable explanation is that the occurrence of A. pennata was controlled by environ-
ment, and it is significant that Popofskyellum occurs only in association with A. pennata.
Though probably not closely related phylogenetically, there are obvious similarities
between the conical, bilaterally symmetrical genera Albaillella and Popofskyellum , and
Deflandre (1964, p. 3057) suggested an evolutionary convergence in morphology. If the
appearance of A. pennata was due to environmental change, then we must suppose, in
view of the rarity of faunas including Albaillella , that the change taking place during
'R. paucicrenulatum time’ was somewhat unusual in the histories of the goniatite bands
of Derbyshire and Staffordshire.

In considering the nature of this change it is significant that whilst A. pennata is
extremely abundant in the bullions of locality 1, all the common spumellines of other
horizons are also present in considerable numbers. The change in environment had no

C 3803

z

328

PALAEONTOLOGY, VOLUME 9

effect upon the spumellines or the goniatites. The appearance of the two non-spumelline
species could have been the result of several kinds of influence, of a minor change in
salinity or temperature, or possibly a slight alteration in the speed of water movement.
Such explanations presuppose that the spumellines were less sensitive to their immediate
environment than A. pennata and P. undulatum. A rather more satisfactory explanation
is that the two non-spumellines are relatively deep-water forms and were favoured by
depths only seldom attained during the deposition of the Namurian goniatite bands
studied. During the brief period of deeper water conditions favouring these two forms
the free-swimming goniatite and the spumellines floating at relatively shallow depth
could also flourish.

Potts (1960) adduced geochemical arguments for fluctuating sedimentation rates in
at least one Namurian goniatite band, and it is possible that such changes in rate of
sedimentation were linked to depth fluctuations of the kind now suggested for the
R. paucicrenulatum band. The suggestion that A. pennata and P. undulatum are indicators
of relatively deep-water conditions is at least in accord with the little yet known concern-
ing modern radiolarian ecology and functional morphology. Both Albadlella and
Popofskyelluni possess shells superficially similar to those of many members of the
Nassellina, and Haeckel (1887, p. cliv) believed nassellines to increase in abundance with
depth in modern oceans and to predominate in the abyssal zone. Delicate lattice shells
and elaborate spines amongst the Spumellina are probably adaptations to pelagic
existence (cf. Aberdeen 1940), and whilst the Namurian spumellines show many
examples of such structures (Holdsworth 1964), neither A. pennata with its dense, only
slightly perforate shell, nor P. undulatum have any features which might suggest that
they are highly adapted to a free-floating life in surface or immediately sub-surface
currents. Both may even have been benthonic forms. The possibility of deep-water
control is also in accord with the presence of Albadlella and Popofskyelluni in the
phosphatic Visean nodules studied by Deflandre, sediments likely to have formed at
considerable depth under conditions of reduced sediment accumulation.

Acknowledgements. I am indebted to Mrs. Helen P. Foreman (Oberlin College, Ohio) and Miss
Catherine Clark (Scripps Institute of Oceanography, La Jolla) for helpful discussion, and to Prof. G.
Deflandre (Laboratoire de Micropaleontologie, Paris) for unpublished information regarding
Albaillella in the Visean.

REFERENCES

Aberdeen, e. 1940. Radiolarian fauna of the Caballos formation, Marathon Basin, Texas. J. Paleont.
14, 127-39.

Campbell, a. s. and moore, r. c. 1954. Treatise on invertebrate paleontology. Part D, Protista 3.
Univ. Kansas Press and Geol. Soc. Amer.

deflandre, g. 1946. Radiolaires et Hystrichosphaerides du Carbonifere de la Montagne Noire.
C.r. Acad. Sci., Paris, 223, 515-17.

1952. Albaillella nov. gen., Radiolaire fossile du Carbonifere inferieur, type d’une lignee aberrante

eteinte. Ibid. 234, 872-4.

1953 a. Titres et Travaux Scientifiques de Georges Deflandre (Supplement 1949-53). Paris.

— 19536. Radiolaires fossiles. In Trade de Zoologie, ed. grasse, p. p., 1, (2), 389-436. Paris.

1958. Lapidopiscum nov. gen., type nouveau de Radiolaire viseen, famille des Lapidopiscidae

fam. nov., de l’ordre des Albaillellidae (sic). C.r. Acad. Sci., Paris, 246, 2278-80.

1960. A propos du developpement des recherches sur les Radiolaires fossiles. Revue Micro-

paleont. 2, 212-18.

B. K. HOLDSWORTH: RADIOLARIA FROM DERBYSHIRE 329

deflandre, g. 1964. La famille des Popofskyellidae fam. nov. et le genre Popofskyellum Defl.,
Radiolaires viseens de la Montagne Noire. C.r. Acad. Sci., Paris, 259, 3055-8.
demanet, f. 1938. La fauna des couches de passage du Dinantien au Namurien dans le synclinorium
de Dinant. Mem. Mas. r. Hist. nat. Belg. 84.

foreman, H. p. 1959. A new occurrence of Devonian Radiolaria in calcareous concretions of the
Huron member of the Ohio shale. J. Paleont. 33, 76-80.

1963. Upper Devonian Radiolaria from the Huron member of the Ohio shale. Micropaleon-
tology, 9, 267-304.

Haeckel, e. 1887. Report on the Radiolaria collected by H.M.S. Challenger during the years 1873-76.

Kept. Voy. Challenger, Zool. 18, i-clxxxviii, 1-1803.
hinde, G. J. 1899. On the Radiolaria in the Devonian rocks of New South Wales. Quart. J. geol.
Soc. Lond. 55, 38-64.

holdsworth, b. k. 1963. Unpublished Ph.D. thesis. University of Manchester.

1964. Radiolarian nature of the thicker-shelled goniatite faunal phase in some Namurian lime-
stone ‘bullions’. Nature, Lond. 201, 697-9.
potts, J. G. 1960. Unpublished Ph.D. thesis, University of Manchester.

B. K. HOLDSWORTH
Department of Geology,
The University,
Keele, Staffs.

Manuscript received 1 March 1965

A NEW ODONTOPLEURID TRILOBITE GENUS
FROM THE DEVONIAN OF BOHEMIA

by DAVID L. BRUTON

Abstract. A new odontopleurid genus Isoprusia (type I. mydlakia sp. nov.) is described and assigned to the
subfamily Miraspidinae R. and E. Richter, 1917. The new genus includes Odontopleura laportei Hawle and
Corda, Acidaspis Ursula Barrande, ? A. sperata Barrande, A. ( Ceratocephala ) sandbergeri R. and E. Richter,
Orphanaspis cornuticauda Erben, and Koneprusia pennata Liitke. Features of the pygidium of the new genus
enable it to be distinguished from Koneprusia Prantl and Pribyl 1 949, the type material of which cannot be traced.
Odontopleura subterarmata Barrande, is tentatively retained in the genus Koneprusia. The type, and where
possible, topotype material of these species is redescribed and figured. Evidence from the thorax of Isoprusia
indicates that, in the pleural field and spines, considerable morphological modification can exist between the
first formed (anterior) and last formed (posterior) segments. The relationship of the new genus to one or both
of the Miraspis-Ceratocephala root stocks has been briefly discussed. A reconstruction of Isoprusia mydlakia is
given.

During a visit in 1963 to the Geological-Paleontological Museum at the University
of Marburg, West Germany, I found amongst the collections some well-documented
odontopleurid material collected by the late Professor J. Jahn, from the Devonian
Limestones of Bohemia. It seems likely that the material came to Marburg to be
identified by the late Dr. Fr. Drevermann, who examined specimens for Professor Jahn
on other occasions (see Jahn, 1903, p. 23, footnote).

Preparation of the soft, yellow, argillaceous limestone, revealed numerous well-
preserved cranidia, pygidia, thoracic segments and hypostomata, either retaining the
entire exoskeleton, or partly exfoliated. All the exoskeletal parts are thought to repre-
sent one new species, Isoprusia mydlakia, the type selected for the new odontopleurid
genus Isoprusia.

The excellent preservation and abundance of the material has allowed a fairly full
diagnosis to be given, and because of this, it has been found that some of the species
formerly assigned to Koneprusia by Prantl and Pribyl (1949), are best placed in the new
genus as here defined.

Visits to Museums in Czechoslovakia (1962), West Germany (1963), and the United
States of America (1964) have allowed me to study types and additional material of all
these species.

Acknowledgements. I am indebted to the following persons for so kindly allowing me to study the
odontopleurid material in their care: Dr. Ronald Huckriede, Geological-Paleontological Institute,
Marburg University (Mbg.), Dr. VI. Zazvorka and Dr. Radvan Horny, National Museum, Prague
(N.M.P.), Professor O. H. Schindewolf, Geological-Paleontological Institute, Tubingen University
(T.U.), Dr. R. Heineck, Paleontological Department, Natural History Museum, Wiesbaden, and
Professor Harry B. Whittington, Museum of Comparative Zoology, Harvard University (M.C.Z.).
At Bonn in 1963, Dr. F. Liitke generously made available to me all of his recent collections from the
Harz region along with types on loan from the Geological-Paleontological Museum, Humboldt Uni-
versity, East Berlin (H.U.). Mr. Jin Vanek and Dr. Ivo Chlupac (Prague) have been most helpful in
personal correspondence with regard to stratigraphic problems, and both have sent casts of their
material. It gives me great pleasure in thanking Professors Whittington and P. C. Sylvester-Bradley

f Palaeontology, Vol. 9, Part 2, 1966, pp. 330-45, pis. 55-57.]

D. L. BRUTON: A NEW ODONTOPLEURID TRILOBITE GENUS 331

for reading the manuscript and for all their help and encouragement during its preparation. Mrs. E. K.
Harris (Leicester) kindly prepared the text-figure from my original reconstruction.

This work has been made possible with the help of a D.S.I.R. studentship which I gratefully acknow-
ledge.

Terminology and techniques. The terminology used is that recommended by Whitting-
ton (1956 a), but I have followed Jaanusson (1956, pp. 36-37) in using the terms rachis
(= axis) and dorsal furrow (= axial furrow); and in numbering and lettering the
glabellar lobes (L) and furrows (S), from the posterior forwards. Before making the
photographs, each specimen was coated with a dilute ‘Opaque’, and then lightly coated
with ammonium chloride in the manner described by Whittington (1963, ‘Explanation
of Plates’ following p. 118). The orientation of the specimens is that used by Whittington.

SYSTEMATIC DESCRIPTIONS

Family odontopleuridae Burmeister 1843
Subfamily miraspidinae R. and E. Richter 1917
Genus isoprusia gen. nov.

Derivation of name. Greek Isos, equal: prusia, suffix from Koneprusy, a small village in Bohemia.
Type species. Isoprusia mydlakia gen. et sp. nov.

Diagnosis. Cephalon rectangular in outline, cranidium trapezoidal; frontal glabellar
lobe expanded laterally. Two pairs of lateral glabellar lobes well defined. Occipital ring
strongly convex medially, with median spine which may rise straight upwards or curve
backwards distally. Posterior band may or may not be obviously separated by weak
furrow from remainder of ring. Eye ridge weakly convex, positioned on outer part of
inflated fixed cheeks. Palpebral lobe short, with well-defined palpebral furrow; eye
ovoid, non-stalked. Anterior suture gently curved forward outside and almost parallel
to eye ridge. Posterior suture directed outwards and backwards approximately in line
with anterior suture, and crossing posterior border so as to produce ear-like projection
of the latter which is widest at the suture.

Free cheek as in Ceratocephala, with straight anterior and sigmoidal lateral outline;
lateral margin considerably widened. Border with short peg-like spines directed out-
wards and downwards. Librigenal spine with origin inside posterior margin.

Hypostoma subquadrate, anterior margin gently curved convexly outwards; posterior
margin with deep narrow median notch and lateral ears. Median body shield-shaped,
gently convex; border furrow running from inside antero-lateral margins. Deepened
notches may be present at antero-lateral parts of median body. Small anterior wings,
lateral shoulder, and lateral notch.

Thorax of ? ten segments. Pleurae with flattened central ridge which may be weakly
furrowed or smooth. Band produced into two spines, the terminal downwardly directed
spine blade-like with secondary spines from edges, and the principal pleural spines
apparently directed outwards and upwards on anterior segments, becoming pro-
gressively curved upwards and backwards on posterior segments.

Pygidium semicircular in outline. Rachis not reaching posterior border, composed
of two obvious rings and sometimes a lower terminal portion. Principal border spine

332

PALAEONTOLOGY, VOLUME 9

connected to weakly raised ridge which curves backwards from first ring. Secondary
border spines lacking or numerous, small downwardly directed from postero-lateral
margin.

Exoskeleton granulate or finely spinose.

Discussion. The following species are now assigned to the new genus. Acidaspis Ursula
Barrande 1872, Odontopleura laportei Hawle and Corda 1847, Orphanaspis cornuticauda
Erben 1952, Acidaspis ( Ceratocephala ) sandbergeri R. and E. Richter 1917, ? Acidaspis
sperata Barrande 1872, and Koneprusia pennata Lutke 1965.

Geological range. The genus is known from the Lower Devonian (Pragian-Emsian) and Middle
Devonian (Eifelian) in Bohemia; Lower Devonian (lower-upper Emsian) and Middle Devonian
(Eifelian-Givetian) in Germany; Middle Devonian (Couvinian) in Poland, and Middle Devonian of
Kazakhstan, U.S.S.R.

Isoprusia mydlakia sp. nov.

Plate 55, figs. 1-20; Plate 56, figs. 1-16; Plate 57, fig. 13
Derivation of name. From the word mydlak, a Bohemian quarry term for soap-stone.

Hoiotype. Mbg. 398, Plate 1, figs. 1, 3, 4, 10, 11. Geological-Paleontological Institute, Marburg
University.

Paratypes. Mbg. 399, Plate 55, figs. 2, 5, 7; Mbg. 400, Plate 55, fig. 8, Plate 57, fig. 13; Mbg. 401, Plate
55, figs. 6, 14, 18; Mbg. 402, Plate 55, figs. 12, 13; Mbg. 403, Plate 55, fig. 19; Mbg. 404, Plate 55,
figs. 15, 20; Mbg. 405, Plate 55, fig. 16; Mbg. 406, Plate 55, fig. 17; Mbg. 407, Plate 55, fig. 9; Mbg.
408-16, Plate 56, figs. 1-16.

Material. In addition to the types, several cranidia, pygidia, one free cheek, and many thoracic seg-
ments.

Type locality and Horizon. The label accompanying the material reads, ‘Menanian Kalk (zone 9 mydlak),
Zlaty Kun; Professor Jahn 1904’. This information and the profile given by Jahn (1903, p. 21, fig. 4;
see also Chlupac, 1959, fig. 7, pp. 475-6), allows the material to be accurately placed in the Devonian
succession exposed at Zlaty Kun ; the zone 9 corresponding to a local lens of soft limestone near the
base of the Acanthopyge Limestone, i.e. at the base of the Upper Eifelian (Chlupac, personal com-
munication). The occurrence in the material of Phacops cf. breviceps also supports this conclusion.
Associated fauna listed by Chlupac (1957, p. 473; 1959, pp. 490-5).

EXPLANATION OF PLATE 55

Figs. 1-20. Isoprusia mydlakia gen. et sp. nov., Acanthopyge Limestone (Upper Eifelian), Zlaty Kun,
Bohemia. 1, 3, 4, 10, 11, Mbg. 398, hoiotype, cranidium; dorsal, right lateral, oblique side, oblique
posterior, and posterior views; x4. 2, 5, 7, dorsal view, posterior view showing posterior occipital
band and anterior view of Mbg. 399, cranidium with exoskeleton partly exfoliated ; X 4. 8, Mbg. 400,
slightly damaged cranidium with exoskeleton complete; anterior view; x5. 6, 14, 18, Mbg. 401,
free cheek and eye lobe; 6, left lateral view; x6. 14, anterior view; x4-5. 18, external surface of
eye lobe, oblique view; X 18. 12, 13, anterior and dorsal views of Mbg. 402, pygidium; x4. 19,
Mbg. 403, postero-lateral margin of pygidium showing secondary border spines; dorsal view; X 18.
15, 20, oblique left lateral and dorsal views of Mbg. 404, internal mould of incomplete hypostoma;
x 7-5. 16, Mbg. 405, internal mould of incomplete hypostoma showing small depression on antero-
lateral part of median body; dorsal view; x6. 17, Mbg. 406, small hypostoma, dorsal view; X 12.
9, Mbg. 407, complete eye lobe, oblique lateral view; x 18.

Palaeontology, Vol. 9

PLATE 55

BRUTON, Devonian odontopleurid trilobites

D. L. BRUTON: A NEW ODONTOPLEURID TRILOBITE GENUS 333

Description. Cranidium (PL 55, figs. 1,2; PI. 57, fig. 13) trapezoidal in outline, gently
convex (sag., tr.). Glabella ( PI. 55, figs. 3, 8, 11), gently convex (sag., tr.) maximum width
(tr.) at occipital ring tapering forward to slightly more than half this width at frontal
lobe; length (sag.) approximately equal to maximum width. Occipital ring of length
(sag.) slightly less than half total glabellar length, becoming narrower (exs.) where its
lateral parts fuse with inner area of fixed cheek. Outline of posterior margin of ring
strongly curved posteriorly; median part rather strongly convex (tr.) and bearing a slim
occipital spine (length unknown), which arises from inside posterior margin (PI. 55,
figs. 3, 10, 11). Spine inclined backwards at approximately 70-80°. Occipital furrow
broad, shallow and transverse medially, deepened as it becomes directed backwards and
outwards around base of Lx. Occipital lobes lacking. Posterior occipital band (PI. 55,
figs. 5, 10, 11) indistinctly separated from remainder of ring; widest and vertically
sloping beneath base of spine, becoming slightly narrower and turned inwards and
forwards laterally. Two pairs lateral glabellar lobes, Lx the larger; both weakly convex
and elongated (exs.). Si shallow, directed obliquely outwards and forwards, deepest
between lobes becoming very shallow adjacent to dorsal furrow. S2 shorter, very deep
and slot-like (PI. 57, fig. 13). Median lobe (PI. 55, figs. 8, 11) gently convex (tr.) almost
fiat longitudinally (PI. 55, fig. 3); frontal lobe sloping steeply downwards to anterior
border furrow. Median lobe parallel-sided and separated from lateral lobes by broad,
shallow longitudinal furrows which run into the curved S2 furrow thus delimiting the
broad lateral expansion of frontal lobe. Immediately in front of S2 a small prominent
swelling (PI. 55, figs. 2, 8; PI. 3, fig. 13), which may represent a third lateral lobe.
Dorsal furrow deepest between adjacent parts of fixed cheek and lateral lobes, elsewhere,
weakly defined and marked only by change in slope. Fixed cheek subtriangular, taper-
ing forward from maximum width (tr.) just posterior to palpebral lobe; moderately
to strongly convex (tr.), highest point on transverse line drawn midway through palpe-
bral lobe. Such a line passes just posterior to outer extremity of Si furrow and crosses
median glabellarlobe at approximately half length (sag.). When cranidium is viewed from
anterior (PI. 55, fig. 8), highest point of cheek above the level of median lobe. From
highest point, cheek slopes gently forwards and then more steeply downwards to eye
ridge; posteriorly cheek more steeply inclined and posterior part drops vertically down
to posterior margin. Palpebral lobe gently curved and sloping slightly outwards from
side of and below highest point of fixed cheek; lobe separated from cheek by broad
shallow palpebral furrow which runs into shallower furrow along inner side of eye
ridge (PI. 55, fig. 1). Eye ridge broad, weakly defined, and directed forward before
curving around outer anterior part of cheek to merge with smooth depressed area at
lateral expansion of frontal lobe (PI. 55, fig. 8). Anterior branch of facial suture curves
forwards outside and sub-parallel to eye ridge and crosses anterior border (PI. 55, fig. 8).
Latter straight, gently arched (tr.) narrow (sag.) separated from frontal lobe by narrow,
deep border furrow which becomes wider (exs.) and deepened just inside suture (PI. 55,
fig. 2). Posterior branch of suture continued outwards and downwards at some 45° to run
out and over posterior margin. Posterior border furrow shallow to lacking at inner corner
where fixed cheek and occipital ring fuse; laterally, furrow deepens and curves forwards
around base of fixed cheek separating it from gently convex border widest (exs.) at the
suture. Posterior margin (PI. 55, fig. 11) gently swept upwards at suture, curved
under with short downwardly projecting edge to doublure. External surface of cheeks.

334

PALAEONTOLOGY, VOLUME 9

fronto-median lobe, and lateral glabellar lobes covered with evenly, but widely spaced
spine bases; furrows, occipital ring, and posterior border smooth.

Free cheek (PL 55, figs. 6, 14) wider (tr.) than long (exs.), with approximately straight
anterior and sigmoidal lateral margins. Anterior border area near suture narrow and
convex, becoming flatter and broader (exs.) laterally, here occupying half cheek length.
A broad, smooth concave area separates lateral margin from convex part of cheek;
area becoming deeper and narrower (exs.) near anterior suture. Librigenal spine arising
from just inside border, its base merging with the lateral border and convex part of the
cheek. Spine present on one specimen (Mbg. 407) but incomplete; apparently slim and
directed straight upwards and backwards. One isolated spine curves distally at its tip.
Lateral margin deep and ventrally directed immediately beneath librigenal spine (PI.
55, fig. 6) becoming lower and rolled under towards fronto-lateral margin. At this point,
six small, short peg-like border spines (PL 55, fig. 14), directed outwards and downwards
from upper edge. Spines extend around outer lateral area as indicated by broken spine
bases, but they do not extend along the whole length of anterior margin. Eye (PL 55,
fig. 18) ovoid in plan, situated on highest point of cheek and sloping gently outwards
and downwards; non-stalked and arising directly off cheek surface.

Eye surface (PL 55, fig. 9) covered in small, closely packed hexagonal lenses each
facet convex externally. Convex part of cheek and flattened border with regularly spaced
spine bases which extend on to and down the length of librigenal spine.

Elypostoma (PL 55, figs. 15-17, 20) subquadrate, slightly wider (tr.) than long (sag.).
Anterior margin gently curved convex forwards; posterior margin with deep, narrow
notch. Median body shield-shaped, gently convex (sag. and tr.), delimited by deep
narrow border furrow which runs from antero-lateral margin convergently backwards
to circumscribe posterior margin of body. Short dorsally directed anterior wings; lateral

EXPLANATION OF PLATE 56

Figs. 1-16, Isoprusia mydlakia gen. et sp. nov. Thoracic segments; all specimens paratypes. I, Mbg.
408, dorsal view of assumed first segment, showing broken base of median rachial spine, pleural area
with faint furrow, and articulating flanges and sockets; x5. 3, 5, 11, Mbg. 409; first thoracic seg-
ment; 3, 5, dorsal and anterior views; X 6. 11, oblique left lateral view showing articulating flanges,
tubercle-like pleural spine and short barbed terminal spine; X 9. 2, 4, 9, 14, lateral, dorsal, oblique
lateral and anterior views of Mbg. 410; ? second thoracic segment; note absence of median rachial
spine, unfurrowed pleural area, tubercle-like pleural spine and short, downwardly directed terminal
spine; x4. 8, Mbg. 411; dorsal view of incomplete segment with upwardly directed pleural spine;
note well-developed pleural furrow; x4. 6, Mbg. 412; dorsal view of more posterior segment with
lengthened pleural spine curved backwards; note longer and curved terminal spine, also pleural
furrow; x4. 7, Mbg. 413; similar segment; oblique lateral view; x4. 10, 15 Mbg. 414; dorsal and
right lateral view of incomplete posterior segment; X4. 12, 13, Mbg. 415; assumed tenth thoracic
segment 13, dorsal view; X 4. 12, left lateral view of distal end showing the blade-like pleural spine
and the long, backwardly curved, barbed terminal spine; note also articulating flanges and presence
of articulating process only; X 9. 16, Mbg. 416; left lateral view of long terminal spine of a posterior
segment; x 9.

Fig. 17, Isoprusia sandbergeri (R. and E. Richter), Stringocephalus Limestone (Lower Givetian),
Villmar-Lahn, West Germany. Dorsal view of holotype (Nat. Hist. Museum, Wiesbaden); X 10.
The incomplete cranidium, an internal mould, has been compressed from the left-hand side along
the line of the longitudinal furrow so that the glabella has become partly pushed under.

Fig. 18, Isoprusia cornuticauda (Erben), Greifensteiner Kalk (Eifelian), Wiege near Greifenstein, Harz,
West Germany. Dorsal view of holotype T.U. Ar. 1010/40, a pygidium with damaged left pleural
region; X 10.

Palaeontology, Vol. 9

PLATE 56

BRUTON, Devonian odontopleurid trilobites

D. L. BRUTON: A NEW ODONTOPLEURID TRILOBITE GENUS

335

margin convex and narrow anteriorly with small pointed shoulder and shallow lateral
notch. Posteriorly, margin becoming wider, outwardly sloping and produced into
flattened ear-like projections separated by the posterior notch. Latter does not reach
border furrow. On largest hypostoma (PI. 55, fig. 16) a short ovoid depression occurs
on edge of median body opposite shoulder. Depression absent on the smallest specimen

text-fig. 1. Isoprusia mydlakia gen. et sp. nov. Reconstruction, a, dorsal view; b, anterior view of
cephalon; c, left lateral view. Number of thoracic segments unknown. Approximately X 2.

(PI. 55, fig. 17), slightly impressed on the next largest one (PI. 55, fig. 20). Median body
smooth to slightly granulated; lateral margin and posterior ears very finely granulated.

Number of thoracic segments not known but the reconstruction (text-fig. 1) has been
made assuming the number to have been ten. From the attitude of the pleural spines, it
has been assumed that those segments which have spines directed laterally outwards or
upwards and forwards belong anteriorly, and those on which the spines are directed
progressively more backwards, belong posteriorly. Such an assumption seems a valid
one since this is always the case in the Miraspidinae and Apianurinae Whittington
(1956a), two subfamilies with genera possessing similarly shaped thoracic pleurae.
Numerous fairly complete and some broken segments allow the following descriptions
and observations to be made. Rachis (PL 56, figs. 1, 2, 5, 14) moderately to strongly
convex and wide (tr.) occupying almost half entire width of segment. Posterior margin
gently curved convex forward (sag.); articulating halfring approximately half as long
(sag.) as rachial ring; anterior margin convex forward like articulating furrow, sloping
backwards and becoming narrower at dorsal furrow. Pleurae horizontal and flattened;
anterior segments (PI. 56, figs. 1, 3, 4) smooth or very weakly furrowed; posterior seg-
ments (PI. 56, figs. 6-8, 10, 13) with shallow but obvious pleural furrow. Anterior and
posterior margins straight with very narrow (exs.) flattened flanges. Small fulcral

336

PALAEONTOLOGY, VOLUME 9

articulating (anterior) and socket (posterior) processes developed (PI. 56, figs. 3, 11, 16).
Principal pleural spine on at least the first two anterior segments in the form of a small
upwardly and outwardly directed tubercle (PL 56, figs. 1-5, 9, 11, 14) from the pleural
surface. Immediately beneath tubercle (PI. 56, figs. 2, 9, 11), short, tapering terminal
spine. Latter downwardly directed, flattened dorso-ventrally and bearing row of small
spines along edges. Two pleural segments (PI. 56, figs. 1, 3, 5, 11) with a similar pleural
spine, have a rachis like the occipital ring of the cranidium including the median spine
(unfortunately this was broken during preparation, but the base can be seen). In these
segments, the rachis is slightly greater than twice the width (tr.) of the very short pleural
area and such proportions fit those of the occipital ring and posterior margin of the
larger cranidia. For these reasons, it is thought most likely that this type of segment is
the first. Segments (PI. 56, figs. 6-8, 10, 13, 15) from positions more posterior have
a rachis of the same configuration as those described for pleurae without median rachial
spine but in addition, the tubercle-like pleural spine is produced into long, slim pleural
spines, the lengths, curvature, and directions of which are indicated on the reconstruc-
tion. Articulating flanges and processes visible on pleurae with upwardly directed
pleural spine (PI. 56, fig. 8) while only flanges and anterior process visible on segments
in which pleural spine is directed backwards (PI. 56, figs. 7, 10, 15). Terminal spines as
on anterior pleurae but much longer (PI. 56, fig. 16) and curved backwards distally.
Pleural spines circular in cross section. Pleurae of type shown (PI. 56, fig. 13) regarded
as representing the tenth segment. Pleural area, rachis, and halfring identical to other
posterior segments, but pleural and terminal spine (PI. 56, fig. 12) differ in, (1) pleural
spine flattened in cross-section and blade-like; spine produced backwards and slightly
upwards from fulcrum, becoming incurved distally; (2) terminal spine similar to those
on other segments, but inclined backwards and almost horizontal.

On all segments, pleural area smooth or with odd spine bases; rachis with transverse
row of small spine bases which extend diagonally down the lateral parts. Terminal and
pleural spines apparently finely spined.

EXPLANATION OF PLATE 57

Figs. 1, 4, 6, 7. Isoprusia laportei (Hawle and Corda), ? Suchomasty Limestones (Upper Emsian-
Lower Eifelian), Menany near Koneprusy, Bohemia. Plaster cast of lectotype, N.M.P. CF639;
cranidium, part internal mould ; 1 , 6, dorsal and oblique views ; 4, posterior view showing posterior
occipital band; 7, anterior view; X 3.

Figs. 2, 3, 5, 8-12, 14, 16. Isoprusia Ursula (Barrande), Upper Koneprusy Limestone, Koneprusy area,
Bohemia. 2, 3, 9, plaster cast of holotype, N.M.P. CF635, incomplete cephalon, internal mould;
2, 3, oblique and dorsal views showing posterior position of eye and long anterior facial suture; 9,
anterior view; x3. 5, 16, M.C.Z. 4215, incomplete cranidium; 5, posterior view showing narrow
posterior occipital band; 16, anterior view; X 6. 8, 10, M.C.Z. 4216, incomplete cranidium, anterior
and dorsal view showing narrow (sag.) occipital ring and backwardly directed spine; X4-5. 11, 12,
14, M.C.Z. 4217, incomplete and partly exfoliated cephalon, anterior, dorsal, and oblique views
showing position of free cheek and librigenal spine; x3. Schary Collection.

Fig. 13. Isoprusia mydlakia gen. et sp. nov. Mbg. 400, incomplete cranidium, dorsal stereograph; x4.

Figs. 15, 17, 18. ? Koneprusia subterarmata (Barrande), Suchomasty Limestones (Upper Emsian-
Lower Eifelian), Menany near Koneprusy, Bohemia. 17, 18, plaster cast of lectotype N.M.P.
CF644a, incomplete pygidium, internal mould; 18, dorsal view; 19, posterior view; X 5. 15, plaster
cast of paralectotype, N.M.P. CF664b, incomplete pygidium, internal mould, dorsal view; x5.

Palaeontology , Vol. 9

PLATE 57

BRUTON, Devonian odontopleurid trilobites

D. L. BRUTON: A NEW ODONTOPLEURID TRILOBITE GENUS

337

Discussion of Pleural Modification

Accepting the theory that during ontogeny new segments are formed by growth to
the anterior border of the hind-most somite and then released forward (Stubblefield
1926), it follows that the most anterior segments are the first formed, and the posterior,
the last. Thus it is possible that the posterior segments resemble in character more
closely the original pleural condition, the anterior showing some modification. In Iso-
prusia mydlakia , the anterior segments either lack the pleural furrow or have it very
faintly defined, while on the posterior segments it is still obviously present. On segments
with an upwardly directed spine (PI. 56, fig. 8), a pleural furrow with point of origin at
the antero-lateral part of the rachis, is directed straight outwards towards the base of
the spine where it becomes obliterated. Thus the pleura is divided into two bands of
equal length (sag.), the spine base straddling both bands. In more posterior segments
(PI. 56, figs. 6, 10, 13), a pleural furrow runs diagonally across the pleural area and
distally follows the inner posterior part of the pleural spine base. Hence the pleura is
divided into two bands as before, but here the principal pleural and terminal spine both
belong to the anterior band. The degree of modification is taken further in the most
anterior segments (PI. 56, figs. 1, 2, 4, 9) where the terminal spine is a downward con-
tinuation of the full pleural width, and the pleural spine is reduced to a small tubercle.
Transverse sections of the segments suggest that the doublure does not extend beneath
the pleura, and only the curled-down edge of the pleura forms the articulating process
and socket. Such an arrangement is similar to that found in Ceratocephala laciniata
Whittington and Evitt (1954, p. 59, pi. 6, figs. 5, 6). It seems reasonable to suppose that
on the anterior segments, the reduction to a tubercle of the pleural spine, is an accom-
modation to ease of articulation and possibly enrolment, since an upwardly directed
spine would interfere with the librigenal spine. In view of the obvious presence of
a pleural furrow on the posterior segments, it seems unlikely that the anterior segments
represent the original pleural type from which the posterior segments have become
secondarily modified by the extreme elongation of the tubercle in the form of a spine.

To homologize the pleural spines in Isoprusia with those found on the pleurae of
other odontopleurid genera is more difficult. In genera ( Primaspis , Diacanthaspis,
Leonaspis, and others; see Whittington 1956n) with a recognizable pleural furrow, the
principal pleural spine, without exception, belongs to the posterior band, and a second
spine, if developed, to the anterior band. Even in Ceratocephala laciniata Whittington
and Evitt, the second spine (= barbed spine), belongs to the posterior part of the seg-
ment. In Isoprusia , it forms as a continuation of the full pleural width, or belongs to the
anterior band. For this reason I have called this spine in Isoprusia mydlakia terminal
(not terminal in the sense of Reed 1925; Prantl and Pribyl 1949; = anterior spine of
Whittington 1956n, p. 161). It has been shown that the pleural spines can belong to either
band and thus it is difficult to use them for comparative morphology.

Isoprusia laportei (Hawle and Corda, 1847)

Plate 57, figs. 1, 4, 6-7

1847 Odontopleura Laportii Hawle and Corda, p. 156.

1852 Acidaspis Laportei : Barrande, p. 750, pi. 39, fig. 22 only.

1926 Acidaspis ( Leonapis ) laportei: Kegel, p. 20, pi. 2, fig. 15.

338

PALAEONTOLOGY, VOLUME 9

1949 Acanthaloma ( Acanthalomci ) Iciportei: Prantl and Pribyl, p. 162, pi. 3, fig. 3 only.

1950 Ceratocephaln ( Leonaspis ) laportei : Erben, p. 281.

1952a Acanthaloma ( Acanthaloma ) laportei : Erben, p. 289, pi. 20, figs. 11-12.

1965 Koneprusia laportei laportei : Liitke, p. 223, pi. 22, figs. 9-11; text-fig. 35a; not Kone-
prusia laportei n. subsp. A.

non 1957 Leonaspis {Leonaspis) laportei : Osmolska, p. 69, pi. 3, fig. 8.

Lectotype (Prantl and Pfibyl 1949, p. 162). N.M.P. CF639, Plate 57, figs. 1, 4, 6-7, Cranidium in
limestone and still retaining part of exoskeleton.

Material. Several paralectotypes, N.M.P. CF640; the specimen of Kegel, H.U. T292; the original
cranidia of Erben, H.U. T343-4; plus topotype material from the Schary Collection. M.C.Z. 4213-14.

Locality. Menany near Koneprusy, Bohemia.

Horizon. Not known with certainty, but most probably from the Suchomasty Limestones (Upper
Emsian-Lower Eifelian).

Description. The species Isoprusia laportei (PI. 57, fig. 1), is closest to I. mydlakia (PL
57, fig. 13), but differs in the following features: (1) More strongly inflated fronto-
median glabellar lobe and lateral lobes. (2) The less strongly inflated fixed cheek which
is flatter and narrower (tr.) and slopes less steeply down to the posterior margin
(PI. 57, figs. 6, 7). (3) Highest part of cheek and that of median lobe at approximately
the same level (PL 57, fig. 7), when specimen is viewed from the anterior. In this respect
I. laportei resembles I. Ursula (PL 57, fig. 9), but in this species the position of the eyes
(PL 57, figs. 3, 10) and the configuration of occipital ring and posterior band (PL 57,
figs. 5, 10) are different. In I. mydlakia (PL 55, fig. 8), the median glabellar lobe is lower
than the adjacent fixed cheeks. The hypostoma which Liitke (1965, pi. 22, fig. 10)
figured as belonging to laportei , differs from that of I. mydlakia in having a narrower
posterior notch and deeper middle furrows. The free cheek, thorax, and pygidium of
I. laportei are not known.

Discussion. New collections made by Barrande (1852, p. 751, pi. 39, fig. 25) enabled
amplification of the original description, and the description of a pygidium considered
to belong to the species. Prantl and Pribyl (1949, pi. 3, fig. 4) also thought that the
pygidium belonged to the species, but it is quite unlike any known for the genus Iso-
prusia and is more likely to belong to a species of Leonaspis. Vanek (1961, p. 85, foot-
note) has noted the similarity to Leonaspis pigra (Barrande, 1872).

Prantl and Pribyl (1949, p. 162) assigned this species to Acanthaloma (= Leonaspis ).
In their remarks, they noted that the specimen figured by Barrande does not show dis-
tinctly The subtrigonal areas on both sides of the frontal lobe, nor the course of the
anterior branch of the facial suture, which deviates distinctly from the ocular ridge’.
Study of the type (PL 57, figs. 1, 6, 7) does not confirm this observation. The anterior
suture is curved convergently towards the anterior border and runs sub-parallel and
just outside the weakly defined eye ridge. In other respects, the glabella, frontal lobe
with lateral swellings, and shape of occipital rings are like those of the other species now
placed in the genus Isoprusia. Examination of the second cranidium figured by Prantl
and Pfibyl (1949, pi. 10, fig. 9), leads me to believe that this belongs to the genus
Leonaspis.

D. L. BRUTON: A NEW ODONTOPLEURID TRILOBITE GENUS

339

Horizon. The exact horizon in which the lectotype was found is still not known, and it seems likely
that Prantl and Pribyl were wrong in referring all of the reddish limestone material from Menany to
the Vinarice Limestone. At Menany, beds from the lowest Pragian to Upper Eifelian are exposed in
which the lithologies are very much alike, but careful collecting has shown the distinctiveness of faunas
from the Vinarice, Suchomasty, and Acanthopyge Limestones (see Chlupac and Vanek 1957; Chlupac
1959).

The specimen of Kegal from the Steinberger Kalk of Lindener Mark is exceedingly small, but it is
like those figured by Erben (1952#) and Liitke (1965, pi. 22, figs. 9, 1 1) from the Upper Emsian of the
Harz, and all seem to be indistinguishable from the Bohemian material of I. laportei (for remarks cf.
Liitke 1965, pp. 223-4).

Isoprusia Ursula (Barrande, 1872)

Plate 57, figs. 2, 3, 5, 8-12, 14, 16

1872 Acidaspis ursula Barrande, p. 84, pi. 16, fig. 28.

1949 Koneprusia ursula: Prantl and Pribyl, p. 201, pi. 3, fig. 25.

1957 Leonaspis (L) laportei: Osmolska, p. 69, fig. 4; pi. 3, fig. 8.

Holotype. N.M.P. CF635, PI. 57, figs. 2-3, 9. An incomplete cephalon with damaged occipital ring
and fronto-median lobe as internal mould.

Material In addition to the type, two cranidia and an incomplete cephalon from the Schary Collection,
M.C.Z. 4215-17.

Locality and Horizon. The type specimen was collected from the general area of Koneprusy, most
probably from the Upper Koneprusy Limestones.

Description. Cranidium (PI. 57, figs. 9, 1 1, 16) flat to gently convex (tr.) ; when specimen
is viewed from anterior, highest point of median glabellar lobe and fixed cheeks at
approximately the same level. Median glabellar lobe widest at occipital furrow, tapering
forwards (PI. 57, figs. 10, 12); dorsal furrow shallow. Lateral lobes suboval, and inflated
to same level as adjacent cheek. Occipital ring (PI. 57, figs. 5, 10) broad (tr.) very short
(sag.) and gently convex. Occipital furrow shallow and only change of slope marks off
ring from median lobe. Narrow posterior band (PI. 57, fig. 5) separated from lateral
part of anterior band by shallow furrow; medially, furrow shallow to lacking and
separation of two bands indistinct. Ring with stout spine which curves backwards in
almost horizontal plane from posterior margin. Eye ridge narrow (tr.) and disposed on
highest outer part of fixed cheek. Latter very narrow (tr.). Palpebral lobe far back and
directed upwards and slightly backwards (PI. 57, figs. 5, 16), so that mid-point lies on
transverse line through middle of Lj. Anterior facial suture very long (exs.), runs straight
forward to point opposite outer end of Si furrow, and then curves inwards to anterior
margin parallel to and outside eye ridge (PI. 57, figs. 2, 9, 14). Posterior suture (PI. 57,
fig. 12) at approximate right-angle to anterior suture at base of eye; from here, directed
outwards and downwards across cheek to border furrow before curving around base of
librigenal spine.

Free cheek as in I. mydlakia, but more convex and swollen around eye. On the holo-
type (PI. 57, fig. 3) only this part of the cheek is visible, the lateral margins being
obscured by matrix. The material from the Schary Collection at Harvard (PI. 57, figs.
11, 14) shows the sigmoidal outline and the lateral margin with at least five short
fringing spines. Librigenal spine directed upwards and backwards.

Fronto-median glabellar lobe with sub-symmetrical rows (both tr. and exs.), of coarse

340 PALAEONTOLOGY, VOLUME 9

granules or spine bases. Remainder of cephalon with granules or spine bases more
unevenly distributed.

Thorax, pygidium, and hypostoma not known.

Discussion. The species resembles Isoprusia laportei, but the narrow fixed cheeks, eye
position, shape of occipital ring, and nature of posterior band and spine in I. Ursula
are distinguishing features.

Osmolska (1957) noted that the glabellar proportions of the Couvinian specimens
from the Holy Cross Mountain area of Poland were different to those specimens from
Bohemia and Germany. Dr. Osmolska has kindly supplied additional photographs of
her material and study of these suggests a closer relationship to I. Ursula than I. laportei,
on account of the transverse profile and shape of occipital ring and spine. She notes
(p. 70) that the eye ridge and palpebral lobe is not discernible on the material, but study
of her figure (pi. 3, fig. 8) clearly shows a well-defined eye ridge and base of the palpebral
lobe; the latter positioned far back. The Polish material comes from a higher horizon
than Barrande’s type, but Osmolska (1957, p. 54) has shown that some of the species
associated with the Polish Couvinian fauna are known from the Lower Devonian of
both Bohemia and the Harz regions.

During a recent visit to the Central Geological Museum at Leningrad, I had the
opportunity of examining the specimen described from the Middle Devonian of Kazakh-
stan (Turkestan) by Weber (1932, p. 134, pi. 1, fig. 20 a-c) as Acidaspis cornuta. I have
no doubt that this is a species of Isoprusia which is closely related to I. Ursula.

Acidaspis cornuta Weber 1932 is a junior primary homonym of Acidaspis cornuta
Beyrich 1846, a species of Miraspis from the Upper Ordovician of Sweden. However,
I do not intend to designate a new name here, but leave this until the specimen is
redescribed by one of my Russian colleagues.

Isoprusia cornuticauda (Erben, 1952)

Plate 56, fig. 18

1952a Orphanaspis cornuticauda Erben, p. 314, pi. 20, fig. 13.

Holotype. T.U. Ar.1010/40, Plate 56, fig. 18. Incomplete pygidium.

Locality. ‘Wiege’ near Greifenstein, Harz, West Germany.

Horizon. Greifensteiner Kalk (Eifelian).

For full description, see Erben (1952a).

Discussion. The similarity between this species and the pygidium of I. mydlakia, indicates
that cornuticauda should be transferred to the genus Isoprusia.

This specimen of cornuticauda is very similar to the pygidium described and figured
by Liitke (1965, p. 226, text-fig. 36c) under the name pennata. Further material may
prove that there is no distinction. These pygidia differ from I. mydlakia in not having
the numerous secondary border spines. The cranidium of I. pennata and I. mydlakia
are quite distinct.

It should be pointed out that the free cheek of I. pennata has been reconstructed by
Liitke (op. cit., p. 226, text-fig. 36a) from broken material (cf. Liitke, pi. 22, fig. 13a, b ),
and I am not satisfied that the reconstructed outline is reliable. A free cheek with

D. L. BRUTON: A NEW ODONTOPLEURID TRILOBITE GENUS 341

a sigmoidal outline like that of I. mydlakia (PI. 55, figs, 6, 14) and I. Ursula (PI. 57, fig.
14) is more likely to be the correct one.

Isoprusia sandbergeri (R. and E. Richter, 1917)

Plate 56, fig. 17

1850 ‘Odontopleura fragmenta, speciei incertae (? novae)’ G. and F. Sandberger, p. 56, pi. 2,
fig. 5a, b.

1917 Acidaspis ( Ceratocephala ?) sandbergeri R. and E. Richter, p. 467, fig. 8 <7, b.

Holotype. Plate 56, fig. 17. Incomplete and distorted cephalon. Type deposited in the Natural History
Museum, Wiesbaden, West Germany.

Locality. Villmar-Lahn, West Germany.

Horizon. Stringocephalus Limestone (Lower Givetian).

Description. Despite the compression and distortion, the free cheek and eye lobe (PI. 56,
fig. 17) does not appear to have been displaced to any great extent (cf. R. and E. Richter
1917, p. 467).

The posterior position of the eye and the long, gently curved eye ridge, is reminiscent
of Isoprusia Ursula. The free cheek, although damaged along the anterior margin is more
like that of I. mydlakia , in so far as the cheek area is less convex than in I. ursula. The
lateral swelling of the frontal glabellar lobe is typical of the genus Isoprusia. The con-
vexity and outline of the glabella and lobes, and the depths of furrows have no doubt been
exaggerated owing to the distortion of the specimen.

Discussion. R. and E. Richter, assigned this species with some reservation to Ceratoce-
phala, and remarked on the unusually high stratigraphic position of the genus. Isoprusia
sandbergeri is, to date, the youngest known for the genus, and one of the youngest species
of the family Odontopleuridae Burmeister.

? Isoprusia sperata (Barrande, 1872)

1872 Acidaspis sperata Barrande, p. 423, pi. 32, figs. 20-21.

1 872 Acidaspis sparsa Barrande, p. 82.

1949 Koneprusia (?) sperata : Prantl and Pfibyl, p. 202.

Discussion. The type of Acidaspis sperata illustrated by Barrande is extremely misleading
and inaccurate when one compares the figure and actual specimen, a poorly preserved
internal mould of a pygidium. The rachis seems to have three rings; the small median
outgrowth at the posterior margin and the small secondary border spines are just
discernible (it is impossible to count the actual number; cf. Prantl and Pfibyl 1949,
p. 202). In addition, details of the pleural region, ornamentation and general outline
are not apparent. The holotype is the only known specimen and its preservation makes
it an inadequate basis for the species.

Barrande (1872, p. 424) gave the locality of Branik which most likely refers to the
former ‘Podolska cementarna’ (= cement works). Only the lowermost Dvorce-Prokop
Limestone (Pragian) is exposed at this locality (Chlupac, personal communication).

342 PALAEONTOLOGY, VOLUME 9

Barrande described both Acidaspis sperata and A. sparsa from the same locality and
horizon, and their respective descriptions seem remarkably similar. Further confusion
is caused by the fact that both species were said to be figured (Barrande 1872, pi. 32),
but only the figure of Acidaspis sperata appears on this plate. To my knowledge, no
specimens labelled A. sparsa have been found amongst the Barrande collections at the
National Museum, Prague.

Vanek (1961, p. 85) notes that former references to the occurrence of laportei (pre-
sumably of cranidia) from the Dvorce-Prokop Limestone, should in fact relate to
sperata, but this cannot be proved with the material available.

Genus koneprusia Prantl and Pribyl 1949
Type species (by monotypy). Acidaspis fuscina Novak 1883 (not traced).

Diagnosis. Pygidium differing from that of Isoprusia gen. nov., in being more triangular
with the development of a stout median border spine. In the type species, the remainder
of the border is apparently smooth, but one species tentatively assigned to this genus
shows the development of numerous small secondary border spines. Details of thorax
and cephalon not known.

Discussion. In erecting the genus Koneprusia, Prantl and Pribyl (1949, pi. 3, fig. 26;
pi. 5, fig. 2) reproduced Novak’s (1883, pi. 10, fig. 19 a) two figures of the type species,
an incomplete thorax and pygidium with very damaged cephalon attached. The type
specimen was not rephotographed nor was any documentary evidence published of
additional material. During my stay in Prague (Autumn, 1962) I was unable to trace
the holotype among the collections in the National Museum and both previous and
subsequent searches have also failed (Vanek, personal communication, 9 March 1964).
The type species of Koneprusia is thus only known from Novak’s (1883, pi. 10, fig. 19 a)
illustration.

I have regarded the type of pygidium Novak portrays, with a well-developed stout
median border spine, as sufficiently distinct to enable the separation of Koneprusia from
Isoprusia. The species K. fuscina (Novak) is very rare and the pygidium has not been
found during recent collections made by Dr. Chlupac and Mr. Vanek from the Acan-
thopyge Limestone. Thus, until an exoskeleton bearing the characteristic pygidium
figured by Novak for Acidaspis fuscina is known, it is not possible to define the characters
of thorax and cephalon in Koneprusia.

The description of the horizon and associated fauna listed by Novak (1883, p. 38;
Prantl and Pribyl 1949, p. 200) would seem to indicate that the type species of Kone-
prusia, like that of Isoprusia, was collected from the Acanthopyge Limestone near
Koneprusy. Mr. Vanek (personal communication) has recently collected two incomplete
cranidia from this horizon (I have seen casts of both specimens which have the strongly
inflated fixed cheek characteristic of Isoprusia mydlakia ) which he considers to be topo-
type material of Koneprusia fuscina. This cannot be proved because of the loss of the
holotype of this species and the inadequate illustration of it given by Novak. For the
same reasons, K. fuscina cannot be compared with the cranidium of Isoprusia mydlakia ,
but as outlined above, the pygidia are considered to be generically different.

D. L. BRUTON: A NEW ODONTOPLEURID TRILOBITE GENUS

343

? Koneprusia subterarmata (Barrande, 1846)

Plate 57, figs. 15, 17-18

1846 Odontopleura subterarmata Barrande, p. 18.

1847 Odontopleura impar Hawle and Corda, p. 157.

1852 Acidaspis subterarmata: Barrande, p. 749, pi. 39, figs. 16-17.

1949 Koneprusia subterarmata: Prantl and Pfibyl, p. 200, pi. 3, figs. 27.

1961 Koneprusia subterarmata: Vanek, p. 85 (footnote).

1965 Koneprusia subterarmata: Liitke, p. 222, text-fig. 34.

Lectotype (wrongly designated as holotype by Prantl and Pfibyl), N.M.P. CF644a, PI. 57, figs. 17-18,
one of Barrade’s two syntypes. An incomplete pygidium; internal mould.

Paralectotype. N.M.P. CF644b, Plate 57, fig. 15. Pygidium, internal mould.

Horizon. Not known with certainty, but most likely from the Suchomasty Limestone.

Locality. Menany near Koneprusy, Bohemia.

Description. Pygidium (PI. 57, figs. 15, 17-18), triangular, approximately three times as
wide (tr.) as long (sag.). Rachis of two obvious rings and a low, very narrow (sag.)
terminal portion marked off by weakly impressed transverse furrow. First ring broad
(tr.) moderately convex; second ring lower and narrower (sag.); rachis not reaching
posterior border. Articulating half ring as wide as first ring, very narrow (sag.); articulat-
ing furrow broad and shallow, curved forward medially. Dorsal furrow shallow along-
side first ring, deepened around second, but becoming shallow and lacking posteriorly.
Pleural ridge broad, weakly convex, runs outwards and curves backwards to be produced
into stout, slightly outwardly and upwardly directed major border spine (length un-
known). Stout median border spine (PI. 57, fig. 15), directed from posterior margin;
spine apparently of equal strength to major border spine. Secondary border spine bases
visible beneath the major border spines and at least eight short, downwardly directed
secondary spines (PI. 57, figs. 17-18) from lateral margin. Pleural area flat to gently
depressed between rolled lateral, and straight and flattened anterior margin. At least
four prominent spine bases along lateral margin and a pair each side of median spine.
Rachis with pair of spine bases on each ring.

Discussion. This species is tentatively assigned to Koneprusia on account of the triangular
outline of the pygidium and the presence of the stout median border spine.

Both Barrande and Hawle and Corda collected their material from the general area
of Menany, but the exact stratigraphic horizon remains uncertain. Prantl and Pfibyl
(1949, p. 120) recorded the species as having come from the Vinafice Limestone (Pragian),
but since then revision of the stratigraphy indicates that it may well have come from the
Suchomasty Limestone (Upper Emsian-Lower Eifelian). To date, the remainder of
the exoskeleton of this species is unknown.

ORIGINS AND RELATIONS OF ISOPRUSIA

The relationship of Isoprusia to other members of the Miraspidinae R. and E. Richter,
1917; emend. Whittington 1956« (= Ceratocephalidae of Prantl and Pfibyl 1949;
Erben 19526) is still obscure, but certain exoskeletal features found in Isoprusia also

a a

C 3803

344

PALAEONTOLOGY, VOLUME 9

occur in Dicranurus (Lower-Middle Devonian), Ceratocephala (Llanvirn-Lower
Middle Devonian; see Whittington 1963, p. 103, pi. 31, figs. 18, 21; pi. 32, figs. 1-3;
Prantl and Pribyl 1949, p. 183), and Miraspis ( ?Arenig-Llanvirn-Lower Devonian; see
Whittington and Bohlin 1958; Whittington 1956 b, p. 515, pi. 60, figs. 1, 4).

Dicranurus (see Whittington 1956a, p. 248, text-fig. 18), like Isoprusia, has a cranidium
which is trapezoidal in outline and the frontal glabellar lobe is expanded laterally. The
eye position and sutural directions are also alike, but Dicranurus has a conical eye raised
well off the cheek surface. The lengthened (sag.) and less convex (tr.) occipital ring of
Dicranurus is an important difference, although the configuration of the occipital spines
(paired and strongly curved in Dicranurus) is not considered of great taxonomic
importance. The pleural regions of Dicranurus, with well-marked pleural furrow
and ridge, points to a closer connexion with Miraspis, while the smooth or weakly
furrowed pleural band of Isoprusia resembles those found in Ceratocephala. The
hypostoma of Isoprusia, although having the shoulder and lateral notch (a family
characteristic, Whittington 1956a, p. 170), has a deep posterior median notch and
posterolateral ears quite unlike any odontopleurid hypostomata previously known.
A shallow, median notch is found in both Miraspis (see Barrande 1852, pi. 39, fig. 8)
and Ceratocephala (see Whittington 1956a, pi. 14, fig. 11; pi. 15, fig. 23) and may indicate
a relationship. The pygidium of Isoprusia bears no resemblance to those known for
Ceratocephala (see Barrande 1852, pi. 38, figs. 5, 18; Whittington and Evitt 1954, pi. 6,
fig. 8), and its semicircular outline sets it aside from the subtrigonal shape of Dicranurus.
The development or loss of secondary border spines or median spine, and the disposition
of the pleural ridge in pygidia of Isoprusia suggests a modification of a Miraspis- like
pygidium. The shortened (sag.) rachis is a characteristic of Miraspis.

The discovery of a Miraspis species in the early Ordovician of Sweden (Whittington
and Bohlin 1958, p. 42, pi. 3, figs. 1-4) now shows that at this time, Miraspis was distinct
from Ceratocephala (cf. Whittington 1956a, pp. 188-90, fig. 3). However, the fact that
Isoprusia has features of the exoskeleton which are shared by both Miraspis and Cerato-
cephala indicates that this genus must be either derived from one or other of these stocks,
or be derived independently from an unknown Cambrian root stock which also gave
rise to Miraspis and Ceratocephala.

A similar origin for Dicranurus seems likely, but whether this genus lies on the same
branch which gave rise to Isoprusia is not known.

REFERENCES

barrande, J. 1 846. Notice preliminaire sur le Systeme Silurien et les Trilobites de Boheme. Leipzic.

1852. Systeme Silurien du centre de la Boheme, I. Prague and Paris.

1872. Systeme Silurien du centre de la Boheme, Suppl. to I. Prague and Paris.

beyrich, E. 1846. Untersuchungen iiber Trilobiten. Berlin.
burmeister, H. 1843. Die Organisation der Trilobiten. Berlin.

chlupac, i. 1957. Facial development and biostratigraphy of the Lower Devonian of Central Bohemia.
Sbor. listr. Ust. geol. — Odd. geol, 23 (for 1956), 369-485, pis. 1-7. (In Czech with English summary.)

1959. Facial development and biostratigraphy of Daleje Shales and Hlubocepy Limestones

(Eifelian) in the Devonian of Central Bohemia. Sbor. listr. Ust. geol. — Odd. geol. 25 (for 1958),
445-511, pis. 1-5. (In Czech with English summary.)

and vanek, J. 1957. Nove nalezy fauny ve vrstvach Konepruskeho devonu. Sjezd cs Spol.

Mineral. Geol. 3, (2), 349-51.

D. L. BRUTON: A NEW ODONTOPLEURID TRILOBITE GENUS

345

erben, h. k. 1950. £lber die stratigraphischen Beziehungen zwischen bohmischem und rheinischem
Devon. Neues Jb. Geol. Pal. Mh. 9, 278-86, pi. 1.

1952n. Trilobiten aus dem Alteren Hercyn (Unterdevon) des Unterharzes. Neues. Jb. Geol. Pal.

Abh. 94, 150-362, pis. 17-64.

19526. Zur Gliederung der Ceratocephalidae Rud. und E. Richter, 1925, emend. Prantl and

Pfibyl, 1949 (Tril). Neues. Jb. Geol. Pal Mh. 7, 304-17.
hawle, I. and corda, A. J. c. 1847. Prodrom einer monographic der bohmischen trilobiten. Prague.
jaanusson, v. 1956. On the trilobite genus Cehnus Angelin, 1854. Bull. geol. Instn. Univ. Uppsala, 36,
35-49, pi. 1.

jahn, J. J. 1903. Geologische Exkursionen in alteren Palaeozoikum Mittelbohmens. Guide geologique,
9e Congr. geol. intern, de Vienne, 1-45.

kegel, w. 1926. Unterdevon von bohmischer Fazies (Steinberger Kalk) in der Lindener Mark bei
Giessen. Abh. preuft. geol. Landesanst. 100, 1-77, pis. 1-4.
lutke, F. 1965. Zur Kenntnis Herzynischer Trilobiten aus dem Unter- und Mitteldevon des Harzes.

Palaeontograpliica , A, 124, 151-236, pis. 17-22.
novak, o. 1883. Zur Kenntnis der Bohmischen Trilobiten. Wien.

osmolska, h. 1957. Trilobites from the Couvinian of Wydryszow (Holy Cross Mts., Poland). Ada
palaeont. polon. 2, 53-79, pis. 1-3.

prantl, f. and pribyl, a. 1949. A study of the superfamily Odontopleuracea nov. superfam. (trilo-
bites). Rozpr. listf. Ust. geol. 12, 1-221 , pis. 1-1 1 . (In Czech and English, with Russian summary.)
reed, f. r. c. 1925. The classification of the Acidaspidae. Geol. Mag. 62, 416-30.
richter, r. and richter, e. 1917. fiber die Einteilung der Familie Acidaspidae und tiber einige ihrer
devonischen Vertreter. Zbl. Mineral. Geol. Pal. Jahrg. 1917, 462-72.
sandberger, g. and sandberger, f. 1850-6. Die Versteinerungen des Rheinischen Schichtensystem
in Nassau. Wiesbaden. Plates separate (1850-5).

Stubblefield, c. J. 1926. Notes on the development of a trilobite Shumardia pusilla (Sars). J
Linn. Soc. Lond., Zool. 36, 345-72, pis. 14—16.

vanek, j. 1961. Predbezna zprava o paleontologickych vyzkumech nekolika lokalit v siluru a devonu
Velke Prahy. Zpr. geol. vyzk. v Race, 81-85.

weber, v. 1932. Trilobites of the Turkestan. Trudy geol. Kom. New series, 178, Leningrad, 1-157,
pis. 1-4. (In Russian with English summary.)

Whittington, H. b. 1956u. Silicified Middle Ordovician trilobites: the Odontopleuridae. Bull Mus.
Comp. Zool. Harv. 114, 155-288, pis. 1-24.

19566. Type and other species of Odontopleuridae (Trilobita). J. Paleont. 30, 504-20, pis. 57-60.

1963. Middle Ordovician trilobites from Lower Head, Western Newfoundland. Bull. Mus.

Comp. Zool. Harv. 129, 1-118, pis. 1-36.

and bohlin, b. 1958. New Lower Ordovician Odontopleuridae (Trilobita) from Oland. Bull.

geol. Instn. Univ. Uppsala, 38, 37-45, pis. 1-3.

and evitt, w. r. 1954. Silicified Middle Ordovician trilobites. Mem. geol. Soc. Amer. 59, 1-137,

pis. 1-33.

Manuscript received 27 March 1965

DAVID L. BRUTON,
Department of Geology,
University of Leicester

UNUSUAL ST RICK LAN DUD BRACHIOPODS
FROM THE UPPER LLANDOVERY BEDS NEAR
PRESTEIGNE, RADNORSHIRE

by A. M. ZIEGLER

Abstract. Both smooth and strongly ribbed stricklandiid brachiopods occur together in early Upper Llandovery
bedstQ-Gpat Presteigne. Aenigmastricklandia contortci gen. et sp. nov. is proposed for the ribbed species, while
the smooth species is assigned to Stricklandia lens aft', progressa Williams. Aenigmastricklandia appears to be
an early ribbed offshoot of the main stricklandiid line; a parallel development is seen in Costistricklandia, which
appeared toward the end of Upper Llandovery time (C5-C6).

One of the specimens which Davidson included on his plate illustrating the species
Stricklandia lirata was ‘An internal cast of a small, exceptionally shaped or malformed
specimen’ collected from Nash Scar, near Presteigne (Davidson 1864-71, pi. 20, fig. 13).
The specimen is preserved in the Geological Survey Museum and is unlike other
stricklandiids in having very irregular, coarse, contorted costae. It might be considered
an aberrant Costistricklandia lirata , but it was not a contemporary of C. lirata ; this is
indicated by the presence on the same block as the original specimen of several speci-
mens of Eocoelia identical with Williams’s subspecies Eocoelia (= Coelospira ) hemi-
sphaerica sefinensis (Williams 1951, p. 113). This Eocoelia occurs in Q beds at Llandovery
and evolved into another form which occurs in C4 beds at Llandovery (op. cit., p. 129).
An upper time limit of C3 may thus be established for Davidson’s specimen, whereas
Costistricklandia lirata with costae of equivalent strength did not appear until C6 of the
Upper Llandovery (op. cit., p. 129; St. Joseph 1935, p. 421).

To resolve this problem, bulk collections were made from all the fossil localities at
present exposed in the Upper Llandovery beds of the inlier near Presteigne (see Table 1).
Stricklandiids are nowhere abundant at Presteigne, but a few specimens were discovered
in five of the six samples collected; one of these samples (No. 10247) came from a sand-
stone outcrop above the limestone quarries at Nash Scar and is probably close to
Davidson’s locality. The Nash Scar collection yielded two pedicle valves similar to
Davidson’s specimen and two pedicle valves of the ordinary smooth Stricklandia lens.
Specimens like Davidson’s were also found at locality 10245, a mile to the east-north-
east. Thus, several specimens of Davidson’s peculiar stricklandiid are now known from
two, possibly three, localities. At locality 10247 they occur together with the distinctly
different smooth Stricklandia lens. A new name, Aenigmastricklandia contorta, is here
proposed for the peculiarly ribbed and distorted species.

A second, but possibly related, problem is posed by the other three collections. They
contain smooth specimens referred to Stricklandia lens aff. progressa, but the specimens
are small for this species (2-5 cm. long at the most) and some have asymmetric indenta-
tions (PI. 58, fig. 6). These specimens are thus similar in some respects to the new species,
Aenigmastricklandia contorta, but they do not have costae of any sort, and they have
very short hinge-lines in contrast to specimens of A. contorta which have hinge-lines

[Palaeontology, Vol. 9, Part 2, 1966, pp. 346-50, pi. 58.]

A. M. ZIEGLER: UNUSUAL STRICKLANDIID BRACHIOPODS 347

table 1. Localities of the Presteigne stricklandiid collections

VSNM No.

Field No.

Grid Ref.

Exposure

Location

Stricklandiids

10242

P-R-A

SO/3177 6343

loose blocks
beside secon-
dary road

125 yds. north of
The Folly’ and 275
yds. NW of Corton
House.

Strieklandia lens
aff. progressa

10243

P-Q-A

SO/3197 6334

very large dis-
used quarry

North of the B4362
road across from
Corton House and
170 yds. WSW of
railway bridge.

Strieklandia lens

aff. progressa

10244

P-N-C

SO/3160 6320

small disused
quarry

North of the B4362
road about 1 mile
ENE of Nash Scar
and 690 yds. WSW
of railway bridge.

Strieklandia lens

aff. progressa

10245

P-N-D

SO/3153 6316

roadside expo-
sure

North of the B4362
road about 1 mile
ENE of Nash Scar
and 690 yds. WSW
of railway bridge.

Aenigmastrieklandia

contorta

10247

P-S-A

SO/3018 6234

slumped out-
crop

Above the large
limestone quarries
at Nash Scar.

Aenigmastrieklandia

contorta

Strieklandia lens

that form, in some cases, the maximum width of the shell. The cardinalia are also
different ; the specimens referred to S. lens. aff. progressa possess outer plates, but these
are absent in the one brachial valve available of A. contorta. Thus, the specimens referred
to * S', lens aff. progressa are quite distinct from A. contorta despite some superficial
similarities.

SYSTEMATIC DESCRIPTION

Genus Aenigmastrieklandia nov.

Type species. Aenigmastrieklandia contorta gen. et sp. nov.

Diagnosis. Aenigmastrieklandia is a small stricklandiid characterized by very coarse
irregularly branching costae and by a deep fold in the brachial valve with corresponding
sulcus in the pedicle valve.

Comparison. Aenigmastrieklandia differs from its contemporary, Strieklandia, in having
costae and in having the ‘outer plates’ of the cardinalia atrophied; and from Costi-
stricklandia in the irregularity of its costae, the prominence of its fold and sulcus, and
the width of its hinge-line.

Aenigmastrieklandia contorta gen. et sp. nov.

Plate 58, figs. 8-16

Holotype. Strieklandia lirata pars. Davidson 1864-71, 159-61, pi. 20, fig. 13, but not figs. 1-12.

Description. Exterior. The pedicle valve is quite convex and the brachial valve less so.
The length-width dimensions are about equal, the greatest width being at the hinge-line

348

PALAEONTOLOGY, VOLUME 9

or just anterior to it. A prominent, rather narrow fold is present in the brachial valve
with a corresponding sulcus in the pedicle valve. Commonly, there are irregular folds
on other parts of the valves. Coarse uneven costae cross the valves and branch irregularly.

Interior of brachial valve. Only one specimen is known. It possesses cardinalia of
a type reminiscent of Stricklandia lens ultima and Costistricklandia lirata; that is, the
‘outer plates’ are atrophied, and the ‘inner plates’ occur laterally to the brachial pro-
cesses (Williams 1951, p. 103).

Interior of the pedicle valve. A spondylium is present but the median septum is
atrophied.

Distribution. This species is known only from two localities (10245 and 10247) from the early Upper
Llandovery beds of the small inlier to the south of Presteigne, Radnorshire.

Genus Stricklandia

Stricklandia lens aff. progressa Williams 1951

Plate 58, figs. 1-7

Description. Exterior. The pedicle valve is more convex than the brachial valve. The
valves are often pear-shaped; the hinge-line is very short and the valve expands
anteriorly. A fold is usually developed in the brachial valve with a corresponding sulcus
in the pedicle valve. Some valves are not symmetrical and may have irregular folds or
indentations.

Interior of brachial valve. The cardinalia are as figured by Williams (1951, pi. 5, figs.
1-3); outer plates, inner plates, and brachial processes are developed.

Interior of pedicle valve. The spondylium is present as usual and is supported by
a short median septum.

Comparison. Stricklandia lens aff. progressa from Presteigne is similar to specimens
from the type locality of S. lens progressa at Mandinam, Llandovery district (Williams

EXPLANATION OF PLATE 58

All figures x2. USNM — United States National Museum; GSM — Geological Survey and Museum.

Figs. 1-7. Stricklandia lens a ft. progressa Williams.

1,2, USNM Collection No. 10244, spec. 140421; 1, Posterior view of internal mould of brachial valve;
2, Internal mould of brachial valve.

3, 4, USNM Collection No. 10244, spec. 140422; 3, Posterior view of internal mould of brachial
valve; 4, Internal mould of brachial valve.

5, 6, USNM Collection No. 10244, spec. 140423; 5, Posterior view of internal mould of brachial
valve; 6, Internal mould of brachial valve.

7, USNM Collection No. 10243, spec. 140424; Internal mould of pedicle valve.

Figs. 8-16. Aenigmastricklandia contorta gen. et sp. nov.

8, 9, USNM Collection No. 10245, spec. 140425; 8, Posterior view of internal mould of brachial
valve; 9, Internal mould of brachial valve.

10-13, USNM Collection No. 10245, spec. 140426; 10, Posterior view of rubber replica of exterior of
pedicle valve; 11, Rubber replica of exterior of pedicle valve; 12, Internal mould of pedicle valve;
13, Anterior view of internal mould of pedicle valve.

14, USNM Collection No. 10247, spec. 140427; Posterior view of rubber replica of exterior of pedicle
valve.

15, 16, Holotype, GSM 13783, Nash Scar, Presteigne; 15, Internal mould of pedicle valve; 16, An-
terior view of internal mould of pedicle valve.

Palaeontology, Vol. 9

PLATE 58

ZIEGLER, Stricklandiid brachiopods

A. M. ZIEGLER: UNUSUAL STRICKLANDIID BRACHIOPODS 349

1951, p. 102), but the specimens are only about half as long, with irregularities in the
growth of the valves, and the hinge-line is unusually short.

Distribution. At Presteigne, S. lens aff. progressa occurs at localities 10242, 10243, and 10244 and

possibly at locality 10247.

Conclusions

There are at least three ways to account for one or both of the unusual stricklandiids
at Presteigne: (1) the individuals may have been diseased; (2) they may have been the
victims of particularly harsh environmental conditions; or (3) they may have become
genetically distinct from the parent stock, Stricklandia. The rocks are not deformed
in any way, so the possibility that the specimens were contorted after they were buried
may be ruled out. The first possibility is considered the least likely; a disease probably
could not have produced the strong ribs of Aenigmastricklandia, and though it might
have caused the deformities common among the stricklandiids described, it would be
necessary to assume that this disease was peculiar to Presteigne, and persisted for some
time (Collection 10245 came from at least 25 ft. stratigraphically above Collection
10244).

The second possibility, that of harsh environmental conditions, must be examined
carefully. The fact that the stricklandiids of Presteigne are rare, small, and usually
deformed suggests that environmental conditions unfavourable to stricklandiids per-
sisted. The situation at Presteigne is particularly marked when compared with some
surrounding areas, where large individuals of Stricklandia occur in profusion and
characteristically dominate the faunal assemblage. An example of an environmental
condition that might have caused the deformities in the stricklandiids of Presteigne is
fast current action; there is, in fact, a suggestion in the conglomeratic nature of the
sandstones that strong current action persisted at Presteigne. Under such conditions
the mantles of the stricklandiids might very well have become damaged, resulting in the
asymmetric, irregular form of the valves.

Environmental extremes are sufficient to account for many of the features of the
Presteigne stricklandiids, but Aenigmastricklandia possesses a variety of characters,
such as a prominent fold and sulcus, a wide hinge-line, and coarse costae, which easily
distinguish it from contemporary stricklandiids and point to an evolutionary history
separate from the main Stricklandia lens community. Alternatively, Aenigmastricklandia
may have been a sport, but it was clearly viable since several specimens are known. In
either case, Aenigmastricklandia would have been genetically distinct from Stricklandia.

In conclusion, the Upper Llandovery rocks of Presteigne are thought to represent
a difficult environment for stricklandiid brachiopods. Two distinct genera, Stricklandia
and Aenigmastricklandia , did, however, survive there, though specimens of each show
deformities. Probably the individuals referred to S. lens aff. progressa drifted, as larvae,
into the area from neighbouring S. lens progressa communities and either died in their
youth or became stunted. Aenigmastricklandia, on the other hand, may have been
endemic to the environment represented at Presteigne where it may very well have had
an evolutionary history separate from the common S. lens community.

In any case, the presence of coarsely ribbed stricklandiids in deposits which may be
confidently dated as early Upper Llandovery (Q-C3) on the basis of Eocoelia hemi-
sphaerica sefinensis and S. lens aff. progressa must induce a measure of caution in those

350

PALAEONTOLOGY, VOLUME 9

concerned with correlating Silurian deposits. Up to now it has been assumed that the
occurrence of strongly ribbed stricklandiids indicated a late Upper Llandovery age (C6),
particularly as Costistricklcindia lirata typica St. Joseph (C6) may be linked with Striek-
lanciia lens ultima Williams (C4) by a series of forms showing intermediate grades of rib
strength; these intermediates find taxonomic expression in the subspecies Costistrick-
landia lirata alpha St. Joseph. The development of ribbing in stricklandiids evidently
occurred more than once.

REFERENCES

davidson, T. 1864-71. Silurian Brachiopoda. Palaeontogr. Soc. [ Monogr .], 3, pt. 7.

st. Joseph, J. K. s. 1935. A critical examination of Stricklandia (= Stricklandinia) lirata (J. de C.

Sowerby) 1839 forma typica. Geol. Mag. 72, 401-22.
williams, a. 1951. Llandovery brachiopods from Wales with special reference to the Llandovery
District. Quart. J. geol. Soc. Loud. 107, 85-136.

A. M. ZIEGLER

Division of Geological Sciences,
California Institute of Technology,
Pasadena,

Manuscript received 5 April 1965 California, U.S.A.

THE FINE STRUCTURE OF SOME LOWER
TRIASSIC ACRITARCHS

by ALAN WILLIAM MEDD

Abstract. Acritarchs from the Lower Triassic of Western Australia are examined with an electron microscope.
As they are almost opaque to the electron beam, a modified replication technique is used to elucidate the
fine structure of their tests. This examination is shown to be of taxonomic value, and a new variant is described.

The electron microscope has proved a very useful tool in the study of foraminifera
(Hyde and Krinsley 1964), pollen (Pettit and Chaloner 1964), and coccoliths (Black and
Barnes 1961 ; Black 1963). The present paper describes its application to the acritarchs.
Examination of their replicas shows the presence or absence of even the smallest
structure on the surface of the acritarch test. Such an examination, together with one
using an optical microscope, has led the writer to revise several species of Triassic
acritarchs.

Material. This paper is based on a sample (No. 43305) of Lower Triassic (Scythian), prepared by
Balme (1963), from the Kockatea Creek No. 19 Bore, over 300 miles north of Perth, Western Australia.
The sample is rich in excellently preserved microfossils, particularly the acritarchs.

All the specimens and photographs described are now in the Archive Collection of the Department
of Geology, University of Reading.

Methods. Some of the material was mounted in glycerin jelly on glass slides and then examined with
an optical microscope.

A dilute suspension of this sample was dried on a Formvar membrane, which rested on a 200 mesh-
to-the-inch copper grid. The sample was then examined with a Philips EM75C electron microscope.
As the acritarchs are almost opaque to electrons, a satisfactory transmission-electron image of their
surface detail can only be obtained when the condenser lens is set for maximum intensity. Unfor-
tunately, the heat generated by the electron bombardment of the Formvar membrane at this intensity
is such as to break the membrane. Therefore, a lower electron intensity must be used and this is seldom
found to produce satisfactory micrographs because of the loss of most of the surface detail (PI. 59,
fig. 1).

Replica techniques overcome this problem as the carbon film, which has been deposited on the
specimen, exactly reproduces the surface detail and is stable under electron bombardment. The method,
adopted by Bradley and Williams (1957) for the study of spore morphology in the genus Bacillus, is
followed here with some refinements of their technique. The carbon film may break when the specimen
is dissolved, as there is a slight swelling of the specimen and also because the solvent slowly attacks
both carbon and copper. Other replica techniques have been developed to overcome these problems
(Bradley 1958; Takeoku and Stix 1963), but repeated attempts to use these techniques on the acritarchs
failed to obtain more than a few satisfactory replicas of the commonest species in the assemblage.

If the specimen is coated with a single thickness of carbon of 250-400 A, instead of 100-200 A as
recommended by Bradley, the extra thickness of film renders it more resilient to the later chemical
processes. It should not be so thick, however, as to mask the finer surface detail. The solvent used is
a freshly made 20% solution of potassium dichromate and potassium permanganate mixture in con-
centrated sulphuric acid, instead of the 10% solution as used by Bradley. The grid is slowly immersed
in this solution and held there in a vertical position for only a few seconds. The grid is then removed
and washed first in dilute sulphuric acid, and then in concentrated hydrochloric acid, 50% hydro-
chloric acid, 10% hydrochloric acid respectively, and finally twice in distilled water. After drying it is

[Palaeontology, Vol. 9, Part 2, 1966, pp. 351-4, pi. 59.]

352

PALAEONTOLOGY, VOLUME 9

ready for shadowing with platinum/palladium metal. With this series of washings after the solvent
treatment, there is usually a high percentage of the grid still covered by the film, and so every species
in the assemblage can be examined with the electron microscope. Electron micrographs were taken on
Ilford N60 plates and developed in Kodak D76.

SYSTEMATIC DESCRIPTIONS

Group Incertae sedis acritarcha Evitt 1963
Sub-group sphaeromorphitae Downie, Evitt and Sarjeant 1963
Genus micrhystridium Deflandre 1937 emend. Downie and Sarjeant 1963

Micrhystridium cf. breve Jansonius 1962
Plate 59, fig. 6

Remarks. Electron micrographs of the specimens referred to this species show that the
thin-walled test possesses large, irregularly arranged granules; the processes are short
(length about 2 ft) and are easily broken, when they leave bosses (diameter 0-5/x) on the
surface of the test. Although the specimens examined have a smaller test diameter
(about 1 1 p.) than that considered to be typical by Jansonius, they are otherwise similar
to one of his figured specimens: Imp. 3010-2-1 13-5 x 28-6.

Micrhystridium cf. fragile Deflandre 1 947
Plate 59, figs. 1 and 2a, b

Remarks. The electron micrographs show that the test and processes are thin-walled
and are covered by a regular arrangement of small granules (diameter about 500 A).
The long processes of this species are robust but flexible and occasionally develop from
expanded bases on the surface of the test.

Wall and Downie (1963) suggest that the only valid criteria for distinguishing M.
fragile Deflandre from M. stellatum Deflandre are that the former has delicate processes
without expanded bases, and that the latter possesses relatively rigid processes whose
bases are expanded. The Triassic specimens examined have some of the diagnostic
elements of both species: they have a test diameter of about 13 ft (about 25-29 /x includ-
ing processes), which is comparable to the range of measurements of M. stellatum as
given by Wall and Downie. The occasional basal expansion of the processes is also
characteristic of M. stellatum , whereas the dominantly spherical nature of the text and
the flexible processes are diagnostic of M. fragile. Wall and Downie also stated that :
‘Separation of the two species becomes artificial to some extent, especially in some

EXPLANATION OF PLATE 59

Figs. 1, la, b. Micrhystridium cf. fragile Deflandre. 1, E.M. 213, X3500. la, E.M. 291, X3000. lb,
E.M. 291, showing the granular test surface, X 10,000.

Figs. 3-5. Veryhachium reduction (Deunff) Jekhowsky. 3, Forma breve, E.M. 250, X 3300. 4, Forma
trispinoides, E.M. 204, x2500. 6, Form with a coarsely granular test, E.M. Ill, X 3200.

Fig. 6. Micrhystridium cf. breve Jansonius, E.M. 260, X 5000.

Figs. 1 and 4 are electron micrographs of the specimens; figs, la, b, 3, 5, and 6 are carbon replicas,
shadowed at 45° with platinum/palladium.

Palaeontology , Vol. 9

PLATE 59

MEDD, Triassic acritarchs

A. W. MEDD: FINE STRUCTURE OF SOME LOWER TRIASSIC ACRITARCHS 353

Jurassic strata, but on this basis the Permian forms must be regarded as belonging to
M. stellatum since their spine bases are always expanded even if the spines are not
always rigid.’ Agreeing with these remarks, the writer by contrasted reasoning considers
the Triassic specimens to belong to M. fragile', forms comparable with the holotype of
M. stellatum are not found in this sample.

text-fig. 1. Veryhachium reduction Deunff, forms possessing a fourth process.
a, S.A.M. 1-67 X 152. b, S.A.M. 2-54 X 1 19. c, S.A.M. 3-210 x 151.

Sub-group polygonomorphitae Downie, Evitt and Sarjeant 1963
Genus veryhachium Deunff 1954, emend. Downie and Sarjeant 1963

Veryhachium reduction Deunff 1954
Plate 59, figs. 3-5; text-fig. 1

Remarks. All of the morphological variants described by Jekhowsky (1961) can be
recognized in the sample, and there is no dominance of any of his types. Electron micro-
graphs show the test to be thin-walled and to have a surface irregularly covered by fine
granules. A further morphological variant which is occasionally found in the sample
has a coarsely granular test with convex sides, together with three short processes. This
is differentiated from V. reduction forma breve Jekhowsky by the more granulose nature
of the test. Although this is suggested by the optical microscope investigation, it is
clearly shown by the electron micrographs.

During an optical examination of several hundreds of specimens of this species, four
were found to possess four processes instead of the typical three. One of the corner
processes is bifid in two examples (text-fig. lc), and they are probably similar to the
type mentioned by Brosius and Bitterli (1961) for a specimen from the Trochitenkalk
(Upper Muschelkalk). The other two have a fourth process on a separate part of the
test, and all four processes are approximately of the same size.

Examination with an optical microscope of the tests of many specimens shows that
their surface is often irregular. The electron micrographs indicate that there is some
irregular folding of the test wall in about 40 per cent, of the specimens. This folding can
occur anywhere on the test, perhaps with breakage of the test wall. Such folding and
tearing occurs in other species in this sample, and could be interpreted as being collapse
structures, caused by the compaction of the sediment.

Jansonius (1962) stated that: ‘In Australian samples of Lower Triassic age forms
assignable to Wilsonastrum occur abundantly; these have in part a modest but dis-
tinctive ornamentation, and show the tiny bristle very clearly. These include specimens

354

PALAEONTOLOGY, VOLUME 9

of W. colonicum n. sp.’ None of the specimens of this sample, examined with either
the optical or the electron microscope, reveals the presence of this bristle, and so forms
of the genus Wilsonastrum Jekhowsky are absent from this sample.

Acknowledgements. The writer thanks Mr. B. E. Balme for his help and for the opportunity to examine
sample No. 43305 from the collection of the Department of Geology, University of Western Australia,
and to Mr. D. B. Williams for helpful discussion during the preparation of the manuscript. The work
was made possible by the award of a B.P. research grant.

REFERENCES

balme, b. e. 1963. Plant microfossils from the Lower Triassic of Western Australia. Palaeontology,
6, 12-40.

black, m. 1963. The fine structure of the mineral parts of Coccolithophoridae. Proc. Linn. Soc. Lond.
174, 1, 41-46.

■ and barnes, b. 1961. Coccoliths and Discoasters from the floor of the South Atlantic Ocean.

Jl. R. microsc. Soc. 80, 137-47.

Bradley, d. e. 1958. A ‘double-evaporation’ carbon replica technique for the electron microscope.
Mikroskopie, 13, 5/6, 180-6.

and williams, d. j. 1957. An electron microscope study of the spores of some species of the

genus Bacillus using carbon replicas. J. gen. Microbiol. 17, 75-79.
brosius, m. and bitterli, p. 1961. Middle Triassic Elystrichosphaerids from salt-wells Riburg-15 and
-17, Switzerland. Bull. Verein. schweiz. Petrol.-Geol. u. -Ing. 28, 74, 33-49.
deflandre, g. 1937. Microfossiles des silex cretaces. Part 2. Annl. Paleont. 26, 1-55.

1947. Sur quelques microorganismes planctoniques des silex Jurassiques. Bulk Inst, oceanogr.

Monaco, 921, 1-12.

deunff, J. 1954. Veryhachium, genre nouveau d'hystrichospheres du Primaire. C.R. Soc. geol.
France, 305-6.

downie, c. and sarjeant, w. a. s. 1963. On the interpretation and status of some hystrichosphere
genera. Palaeontology, 6, 83-96.

evitt, w. r. 1963. A discussion and proposals concerning fossil dinoflagellates, hystrichospheres and
acritarchs. Proc. natn. Acad. Sci. U.S.A. 49, 158-64, and 298-302.
hyde, p. and krinsley, d. 1964. An improved technique for electron microscopic examination of
foraminifera. Micropaleontology , 10, 491-3.

jansonius, J. 1962. Palynology of Permian and Triassic sediments. Peace River Area, Western
Canada. Palaeontographica, 110B, 35-98.

jekhowsky, b. de. 1961. Sur quelques hystrichospheres Permo-Triasiques d’Europe et d’Afrique.
Rev. de Micropal. 3, 207-12.

pettit, J. m. and chaloner, w. g. 1964. The ultrastructure of the Mesozoic pollen Classopollis.
Pollen et Spores, 6, 611-20.

takeoku, m. and stix, e. 1963. On the fine structure of the pollen walls in some Scandinavian Betu-
lacea. Gran, palynol. 4, 2, 161-88.

wall, d. and downie, c. 1963. Permian hystrichospheres from Britain. Palaeontology, 5, 770-84.

a. w. medd

Sedimentology Research Laboratory,
University of Reading

Manuscript received 8 April 1965

THE PALAEONTOLOGICAL ASSOCIATION

COUNCIL 1966-7

President

Professor T. S. Westoll, The University, Newcastle upon Tyne
Vice-Presidents

Dr. W. S. McKerrow, University Museum, Oxford
Professor F. H. T. Rhodes, University College, Swansea

Treasurer

Dr. C. Downie, Department of Geology, The University, Mappin Street, Sheffield 1

Secretary

Dr. C. H. Holland, Department of Geology, Bedford College, London, N.W.l

Assistant Secretary

Dr. J. M. Hancock, Department of Geology, King’s College, London

Editors

Mr. N. F. Hughes, Sedgwick Museum, Cambridge
Dr. Gwyn Thomas, Department of Geology, Imperial College of Science, London, S.W.7
Dr. I. Strachan, Department of Geology, The University, Birmingham, 15
Dr. M. R. House, University Museum, Oxford

Other members of Council

Professor P. M. Butler, Royal Holloway College, Surrey
Mr. M. A. Calver, Geological Survey Office, Leeds
Dr. G. Y. Craig, Grant Institute of Geology, Edinburgh
Miss Grace Dunlop, Bedford College, London
Dr. T. D. Ford, The University, Leicester
Dr. B. M. Funnell, Sedgwick Museum, Cambridge
Dr. R. P. S. Jefferies, British Museum (Natural History), London
Dr. G. A. L. Johnson, The University, Durham City
Dr. F. A. Middlemiss, Queen Mary College, London
Dr. W. D. I. Rolfe, Hunterian Museum, Glasgow
Professor Scott Simpson, The University, Exeter
Dr. A. H. Smout, British Petroleum Company, Sunbury-on-Thames
Dr, L. B. H. Tarlo, The University, Reading

Overseas Representatives

Australia : Professor Dorothy Hill, Department of Geology, University of Queensland, Brisbane
Canada: Dr. D. J. McLaren, Geological Survey of Canada, Department of Mines and Technical
Surveys, Ottawa, Ont.

India: Professor M. R. Sahni, 98, The Mall, Lucknow (U.P.), India

New Zealand: Dr. C. A. Fleming, New Zealand Geological Survey, P.O. Box 368, Lower Hutt
West Indies and Central America: Mr. John B. Saunders, Geological Laboratory, Texaco Trinidad,
Inc., Point k Pierre, Trinidad, West Indies

Eastern U.S.A.: Professor H. B. Whittington, Museum of Comparative Zoology, Harvard Univer-
sity, Cambridge 38, Mass.

Western U.S.A. : Professor J. Wyatt Durham, Department of Paleontology, University of California,
Berkeley 4, Calif.

PALAEONTOLOGY

VOLUME 9 • PART 2

CONTENTS

A revision of Acaste downingiae (Murchison) and related trilobites. By

J. H. SHERGOLD 183

Ostracoda from the Upper Tealby Clay (Lower Barremian) of South
Lincolnshire. By p. kaye and D. barker 208

The foraminiferal genus Bolivinoides from the Upper Cretaceous of the

British Isles. By F. T. barr 220

The Devonian blastoid Belocrinus from France. By D. b. macurda, jr. 244

Population studies in the Ballyshannon Limestone, Ballina Limestone,

and Rinn Point Beds (Visean) of N.W. Ireland. By julia a. e. b. hubbard 252

Occurrence of Australosutura (Trilobita) in the Mississippian of Oklahoma,

U.S.A. By A. R. ORMISTON 270

Descriptions of schizaeaceous spores taken from early Cretaceous macro-
fossils. By N. f. hughes and judith moody-stuart 274

Variation and Ontogeny of some Oxfordian ammonites: Taramelliceras
richei (de Loriol) and Creniceras renggeri (Oppel), from Woodham,
Buckinghamshire. By d. f. b. palframan 290

The Lower Liassic ammonites Neomicroceras gen. nov. and Paracymbites.

By D. T. DONOVAN 312

Radiolaria from the Namurian of Derbyshire. By B. K. holdsworth 319

A new odontopleurid trilobite genus from the Devonian of Bohemia.

By D. L. BRUTON 330

Unusual stricklandiid brachiopods from the Upper Llandovery Beds near

Presteigne, Radnorshire. By a. m. ziegler 346

The fine structure of some Lower Triassic acritarchs. By A. w. medd 351

PRINTED IN GREAT BRITAIN AT THE UNIVERSITY PRESS, OXFORD
BY VIVIAN RIDLER, PRINTER TO THE UNIVERSITY

*

72*

&7ac

c $ i v i

* Z to * z , , „ CO Z CO _

JLSNI NVIN0SH1IWS S3IBVaan LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHil

— CO 2 00 2 CO —

- -I - .w A- “ -c Att&tVtefX ~ /2kY»*

vvxa^y - v^-“^y “ S 2

_ z _ _ _

ES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IWS S3iaVH0ll L I B R A R I E s”~ SMITHSON

nvinoshiiws S3iavaan l

CO z

5 /CTz^qX — /fiffi'// ~

m

CO — CO _

BRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHil

Z CO 2:

I Jp#- O

° ‘0>'' x

z / V

> ^2^- 5 ^ >' 5

z co “ Z CO

ES SMITHSONIAN INSTITUTION NOlinillSNI _ NVINOSHil WS S3ldVaan LIBRARIES SMITHSON

co r; co -? \ co — to

1SN I NVIN0SH1IIAIS S3l0VHan LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHil

m S&Zo&x O -n 2 ~

'if /yjf va h&c t Avtf’fAk-'A 33

3-

co

m Xo/v ocjy'' W

co t: co _ co

ES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IWS S3ldVaa

C£ Z ,v CO z CO

< 2 <

03
33

> ^
33

<jr>iii.V$S m

CO

BRARIES SMITHSON

CO

> XL'uustii'

z _

1SNI NVIN0SH1IWS S3iaVdaiT LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1

5 ,v co ^ 5 ^ co - co

co ^ co X?S^X ^ .co ^

^ 5 tier 'fl© ^ tie" '4^ <^'4 ’ ~ ‘

c c?c ~ -/VssSM, c

o

o

ES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IWS S3iavaail LIBRARIES SMITHSON
z f~ z r- , z r- z

isni NVINOSHil ws saiavaan libraries Smithsonian institution NouniiiSNi nvinoshi

z co z , w z co z

< S x^y. < S < v. S <

> 5 '\s^- >' S

co z to z to z co

MES SMITHSONIAN INSTITUTION NOlinillSNI _ NVINOSHil WS S3IHVHan LIBRARIES SMITHSON
co — to \ co — to

"V

r4

i ^ >

, CO A Z CO _

avyan libraries Smithsonian institution NouniiiSNi nvinoshiiws S3 i ava a n_ li br

CO -y _ -y VV 00 _ 2

co xc^x w „ A X^?X - > * x^Dx <*

c is*. ,v^|j t *4sM„ .

nz&y O “ JW* 5 N<s^s>' ~ 5 ^ “ ^5^ o

ITUTION 2 NOliniliSNI^NVINOSHlIIAlS S3iavaail LIBRARIES SMITHSONIAN INSTITUTION NOIlf
z t - , z r- z ^

o „ xv S£ ra X^»2oX ^ /M?/'/. co r X ;

>

m ^ m l£ * m 's!^o^y !£

go - co £ co £

jvaan libraries Smithsonian institution NouniiiSNi nvinoshiiws saiavaan libr

N

S

ITUTION NouniiiSNi NVINOSHIIWS S3iavaan LIBRARIES SMITHSONIAN INSTITUTION NOlif

avaan libraries Smithsonian institution NouniiiSNi nvinoshiiws saiavaan libf

c/ 70

0'lj|isX/ m XOAtoC/ ^ X?

C o £ CO Z co

’ITUTION NouniiiSNi NVINOSHIIWS S3iavaail LIBRARIES SMITHSONIAN INSTITUTION NOli
,v- co z ^ ^ z .v co

= ^ 1 |

V Z CO Z CO “ Z CO' Z

avaan libraries Smithsonian institution NouniiiSNi nvinoshiiws saiavaan libf

5: „ ^ co - ^ co - ... co -

■*— ^,CA>V ill /Tu^Da/X til >^7icv,4n\ -*<—

o

z _j z _J z -x- _I Z

’ITUTION NOlinillSNI^NVINOSHlIINS^Sa I ava a 11 LIBRARIES SMITHSONIAN^INSTITUTION 2 NOli

° W 1% z XS&X ra I ro X^^x ? XT

30
>

30

m " YW X x^vasv'^z m XCyoTv^Z ^ m

go £ co £ co

avaan libraries Smithsonian institution NouniiiSNi nvinoshiiws saiavaan l

Z CO z * CO Z co z

- ^ - X^S&x - % E 5 zr^x E ^

2 § •%% 5 ff% § * «4

co co smut1, Ss/' co

g » s . XX"

5, s >' 5 >' 2

illSNI NVINOSHIUMS^SI I ava a ll^LI B RAR I ES^SMITHSONIAN INSTITUTION NOli

<S) ~ IS) Z co

)

\

X

^ -4° — i /tfc ^5A t: Jm*' jz/' —1

flUi™ S0NIAN INSTITUTION libraries

3 9088 01375 6598