ree if My
peste
by ay htye

teks
HON

oe

5:

saat)
oni
Hal

Sogeem

oy ha ote

Be
ae

t} »
Siting ihe te 5
ita he mua!
Vee t
+ Vik

:
J

eyssrsey

a8

;
‘i
sh iy

ae

*

i)
is
U

un

Hs

Lpran
"Sten

Pere
NES

ie
aby
ssf

iF
th

HNN

eee
ed HYIsH > ‘
BRAS MARS aihite

SORE bs Pate 4 ars

aaboy, Bib atpinie’s & ' 9) ay % nied
Shirt eeat§ SS MD 5 e , 4
ards ie i ? i. if Os

Ay
Ph ESF ys
pois ible kath

u

Hie
Hie
Myotis

=

a

»
ai
:
a
. 7
ik
105%
10)
*
-

i)

Pat,

’ aa ip : ; =

f

i
j

Wi

;

Bulletin of the
British Museum (Natural History)

Geology series Vol 33. 1979-1980

British Museum (Natural History)
London 1980

Dates of publication of the parts

INOW: : : ; : : : : : : 20 December 1979
No2 . : : : : : ; : ; ‘ 31 January 1980
INOS! : ‘ ; ; : : ; 27 March 1980
No4 . ; : ; : : : 27 March 1980
INoiS) = : ; ; : ‘ ‘ : j : 27 March 1980

ISSN 0007-1471

Printed in Great Britain by Henry Ling Ltd, at the Dorset Press, Dorchester, Dorset

No 1

No 2

No 3

No 4

No 5

Contents
Geology Volume 33

An account of the Ordovician rocks of the Shelve Inlier in west Salop
and part of north Powys.
The late W. F. Whittard, F.R.S. (Compiled by W. T. Dean)

Miscellanea

A new, possibly algal, microproblematicum from the Lower
Carboniferous of England.
G. F. Elliott

Acanthopleurella Groom 1902: origin and life-habits of a miniature
trilobite.
R. A. Fortey & A. W. A. Rushton

Pleistocene bird remains from Tornewton Cave and the Brixham
Windmill Hill Cave in south Devon.
C. J. O. Harrison

The succession of Hyracotherium (Perissodactyla, Mammalia) in
the English early Eocene.
J. J. Hooker

Salenia trisuranalis sp. nov. (Echinoidea) from the Eocene (London
Clay) of Essex, and notes on its phylogeny.
D. N. Lewis & R. P. S. Jefferies ‘

Tertiary and Cretaceous brachiopods from Seymour, Cockburn
and James Ross Islands, Antarctica.
E. F. Owen

Revision of the rugose coral Diphyphyllum concinnum Lonsdale,
1845 and historical remarks on Murchison’s Russian coral collection.
B. R. Rosen & R. F. Wise

Neuroptera (Insecta) in amber from the Lower Cretaceous of
Lebanon.
P. E. S. Whalley

The Caradoc faunal associations of the area between Bala and Dinas
Mawddwy, north Wales.
M. G. Lockley

Fossil insects from the Bembridge Marls, Palaeogene of the Isle of
Wight, southern England.
E. A. Jarzembowski

The Yorkshire Jurassic fern Phlebopteris braunii (Goeppert) and its
reference to Matonia R. Br.
T. M. Harris, F.R.S.

Page

73

79

91

101

115

123

147

157

165

2351.

295

‘ e{ pte a & pegs we ’
fi wunde® eyes. | “es

Sie
ni fi Fa eer ce Holt ary ty [eae
wit tron Yo Tea.

x { 1.4 @ iw) 2.4. —_— a -:
a pay ar
ta

> yet peaagitemelharpewtives Lael. ay an P
am = . “gett andi P

| =: —_ = u
‘ « aaitote 40 nfs aiun i cH rigyai 9

= ponte

of bin it? eniveeneT. owe ‘caine
i ee) end al we) iat i
“poet 2

elated soley nck! vaw iia | »

ol) weeot te) mgt eben). on, ay
Yerayeterty url om RO
4 , TA 7 |
> Tumieivent ent) ‘sbaepaisoert wo
, ke pivzretok
a7 _ - a = ba
p> G pesca rmdiiapag furl s -

i oat s\ eo Reem & ‘noni me

43
if, ash secs) a oc) invest
) ; B
Waokt bax alall jecovne scl peng a Teh t
al . : | |
lis Gel wt ori! oi
4 <ri9
oa
> oa
) Why my qtres) Sra |
‘ios, “ ies @ rh ene

. S6/).0.

Bulletin of the
_ British Museum (Natural History)

‘An account of the Ordovician rocks of the
s helve Inlier in west Salop and part of

i. G jeology series Vol 33 Nol 20 December 1979

a ~~ “Ae
ca Rs i

ee ec)

The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four
scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology, and
an Historical series.

Papers in the Bulletin are primarily the results of research carried out on the unique and
ever-growing collections of the Museum, both by the scientific staff of the Museum and by
specialists from elsewhere who make use of the Museum’s resources. Many of the papers are
works of reference that will remain indispensable for years to come.

Parts are published at irregular intervals as they become ready, each is complete in itself,
available separately, and individually priced. Volumes contain about 300 pages and are not
necessarily completed within one calendar year. Subscriptions may be placed for one or more
series. Subscriptions vary according to the contents of the Volume and are based on a forecast
list of titles. As each Volume nears completion, subscribers are informed of the cost of the
next Volume and invited to renew their subscriptions. Orders and enquiries should be sent to:

Publications Sales,
British Museum (Natural History),
Cromwell Road,
London SW7 S5BD,
England.

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

© Trustees of the British Museum (Natural History), 1979

ISSN 0007-1471 Geology series
Vol 33 No | pp 1-69

British Museum (Natural History)

Cromwell Road
London SW7 5BD Issued 20 December 1979

An account of the Ordovician rocks of the Shelve
Inlier in west Salop and part of north Powys

The late W. F. Whittard, F.R.S.
(Compiled by W. T. Dean, Department of Geology, University College, Cardiff, CF1 1XL)

Contents

Synopsis 1
Introduction and Weknowledgements F : ; ; : ; 3
Geographical location and Previous researches : : : : ; : 4
Classification of the Shelve Ordovician rocks ; : ; ; : 8
Lithostratigraphic subdivisions : ; : : ; : ; : 8
The Shineton Shales : : : : : 5 10
Lithostratigraphic description of the Shelve Ordovician rocks 5 : 5 : 10
Shelve Formation . ; ; ; k : ; : : . : 10
Stiperstones Member ; , : ‘ : : : : ‘ 11

Mytton Member . ‘ : ; ; 3 : F : 5 16

Hope Member : : : ; ; ; ; ‘ ; 24

Stapeley Volcanic Member : ; : : : : 4 : 29

Stapeley Shale Member . : 3 ; : ; é 5 : 35
Middleton Formation . : 3 ; : , : ; : : 38
Weston Member. : : : : ; : ‘ : ; 38

Betton Member ; : ; : : : : : F . 41
Meadowtown Member . : : : : ; : : j 43
Rorrington Member ; : : : 5 : 5 , : 45
Chirbury Formation , 5 5 ; , : ; 3 ’ : 49

Spy Wood Member : : ; ; é : , : : 49

Aldress Member ; : 3 : ; : ; : : 50

Hagley Volcanic Member : 5 : j , 5 : : 51

Hagley Shale Member : : : ‘ A : : ; , 53
Whittery Volcanic Member ; , : ; ; : ‘ : 54
Whittery Shale Member . : : : ; : : : : 58
Miscellaneous notes é 3 ; 5 : 59
Correlation with neighbouring Ordovician successions E : ; : 59
Intrusive igneous rocks . : . ‘ : ; ; : f ‘ 61
Silurian rocks ; F : : ; : : ; 62
Geological structure of the Shelve Inlier : ; ; ; ; ; ‘ 62
References . : 5 : j j ¢ : : : : ‘ : 63
Index : : : : ‘ 3 ; 65
Geological Map of Shelve Inlier : ‘ . ; : : 2 in rear ar pocket

Synopsis

On the basis of the late Professor W. F. Whittard’s field maps and notes, a geological map, scale 1 : 32 000,
is presented for the Ordovician rocks of the Shelve Inlier. These strata, excluding the Shineton
(or Habberley) Shales of Tremadoc age, range from lower Arenig Series to lower Caradoc Series. The
history of the lithostratigraphic classification is reviewed, and a scheme of three formations and fifteen
members proposed on the basis of already-named subdivisions. Descriptions are accompanied by en-
larged sketch maps showing outcrops and relevant fossil localities. A slightly revised correlation table is
included, together with a table showing characteristic trilobites. Geological structure and igneous rocks
(all known outcrops of which are shown on the map) are briefly discussed.

Bull. Br. Mus. nat. Hist. (Geol.) 33 (1): 1-69 Issued 20 December 1979

NO

W. F. WHITTARD

The late Professor W. F. Whittard, F.R.S.
1902-1966

SHELVE INLIER 3
Introduction and Acknowledgements

The task of mapping and describing the Ordovician rocks and faunas of the Shelve Inlier was
begun by Professor W. F. Whittard (frontispiece) in 1930 as a sequel to work on the Lower
Silurian rocks of the same area. From then until his untimely death early in 1966 the Shelve Inlier
occupied the greater part of his research time and was frequently visited by him with mapping
classes from the University of Bristol. Although the Ordovician rocks of the Inlier have yielded
large numbers of fossils — and Whittard’s beautiful monograph (1955) on the trilobites, published
by the Palaeontographical Society, is a classic of its kind — they are rarely obviously fossiliferous
to the casual visitor, who may find collecting both difficult and unrewarding. Much of Whittard’s
time on the Inlier, particularly during the early years, was spent collecting fossils in an attempt
to define the faunal content of the numerous lithostratigraphic subdivisions established earlier,
principally by Lapworth & Watts (1894), and it was not until the appointment of Mr T. R. Fry
as Whittard’s research assistant that the time-consuming task of fossil-collecting was eased. Fry
displayed a remarkable aptitude for obtaining faunas from even the most unpromising strata
and his contribution received appropriate recognition when the University of Bristol awarded
him the honorary degree of M.Sc. in 1970. In later years he was ably assisted by Mr M. White
and both had trilobite species named after them by Whittard.

SHREWSBURY

o.Pontesbury

Snailbeach

is Stretton

(e)
Ludlow

Fig. 1 Sketch map of part of the Welsh Borderland showing the position of the Shelve Inlier in
relation to various towns and topographic features.

4 W. F. WHITTARD

Whittard’s mapping of the Ordovician rocks in the Shelve Inlier formed a natural sequel to
his well-known mapping and highly original interpretation of the Lower Silurian strata of the
region. He noted that the lack of cleavage affecting the Ordovician rocks, together with their
relatively abundant fossils, not always readily apparent during brief visits, showed the palae-
ontological promise to be much greater than previously published accounts might have led one to
believe. Furthermore, the rich trinucleid trilobite faunas suggested to him that the Shelve suc-
cession might become the standard for the Lower and Middle Ordovician of the Anglo-Welsh
area, a sentiment expressed earlier with equal confidence by Lapworth (1887 : 663), while the
not infrequent intermixing of shelly fossils and graptolites held out hope for the solution of
numerous correlation problems.

At the time of Professor Whittard’s death the Shelve trilobite monograph was almost complete,
lacking only a final chapter (since completed, Dean 1967) on the affinities of the trilobites; more
recently the brachiopods have been described by Williams (1974) and the graptolites are being
studied by Dr Isles Strachan. A large proportion of the mapping had been completed at a scale
of 6in to | mile (1 : 10 560) and the results entered upon Quarter-Sheets mounted on cloth.
Unfortunately several gaps remained on this version of the geological map, particularly in
complex areas where more than one interpretation is possible, and it has been my object to try to
fill these gaps utilizing Whittard’s field slips, some of which are at a scale of 25in to | mile
(1 : 2534-4). No systematic account of the Shelve Inlier’s geology had been prepared but notes
on a large number of Ordovician localities are contained in Whittard’s meticulous field notebooks
and I have drawn extensively upon these in preparing the Ordovician stratigraphic account.
Consequently most of the geological observations and conclusions in the present paper are
attributable to Professor Whittard but in addition numerous publications are quoted in an
attempt to formulate a coherent account of the Inlier, and any shortcomings in drawing the
two sets of data together are my own responsibility. It had clearly been Whittard’s intention to
make his description of the Shelve Inlier as comprehensive as possible, including chapters on
geological structure, igneous activity and mineralization, but as no such accounts had been
prepared I have preferred to keep these sections to a minimum.

All the measurements are given in British Imperial units, as originally made by Whittard;
metric equivalents are appended in every case.

The small geological maps reproduced as figures in the text are at a scale of 4 in to | mile
(1: 15 840); grid squares are indicated. The photographs are from colour transparencies taken by
Whittard; all the places named may be found with reference to points shown in the large Map.

The numbered localities refer to Whittard’s manuscript notes.

Geographical location and Previous researches

The Ordovician rocks of the Shelve Inlier form an irregular area (Fig. 1) of some 43 mile?
(111 km?) centred approximately on Shelve, an otherwise obscure hamlet situated c. 21 km
(13 miles) SW of Shrewsbury. Most of the inlier is located within the bounds of west Salop
(formerly Shropshire), but the area in the south-west, comprising Corndon Hill, Todleth Hill and
Church Stoke (Fig. 2), forms part of the adjacent Welsh county of Powys. Topographically the
area is one of rapidly-changing relief, drained by the Rivers Camlad and West Onny together
with numerous streams which exhibit a suggestion of a radial pattern. The highest points are
found at Corndon Hill (1683 ft, =513 m) and the Stiperstones (up to 1762 ft, =537 m), the latter
bounding the sombre, marshy area of The Bog and forming an impressive rampart which
achieved some literary recognition in Mary Webb’s novel ‘Gone to Earth’. The region is sparsely
inhabited and villages and hamlets are relatively few, the most notable being Chirbury, Hope,
Old Church Stoke, Priestweston, Rorrington and Wotherton. In Roman times the area around
Shelve and Ritton Castle was the site of important lead mines and lead mining continued until
early in the 20th century; it has since been discontinued but exploitation of barytes in the old tip
heaps continues sporadically at the present day. Notable view-points are Bromlow Callow (Map)
and the top of Priest Weston Bank from which can be seen a panorama that includes Plynlimmon,

SHELVE INLIER 5)

Maddox's }*’..: Habberley
Coppice /--

\ Snailbeach
a :
Lordshill

2 ' -7@ \y
RonAingenn ae Meadowtown iw A
et) Aeers /Mytton
Common / Batch
Wotherton 4 sentley

Middleton ¢

U
/
/

pf Whitegrit /
th “=x
La
Priestwes ton
FRO

A
CORNDON
HILL
gD/ Brithdir
ca e
Roundton

Hyssington

Es} Todleth

Fig. 2. Outline map of the Shelve Inlier to show the location of principal place-names mentioned
in the text.

Cader Idris, the twin Peaks of the Arenigs, Berwyn and, to the north, the Cheshire Plain with its
girdle of Bunter Sandstone (Trias) hills such as Grinshill, Hawkstone and Nesscliff. From the
Stiperstones all these are visible and, in addition, the Uriconian hills of the Church Stretton area
showing over the top of the Longmynd, the isolated Uriconian mass of the Wrekin to the north,
and the successive Silurian escarpments running away from view to the south.

Work on the Ordovician rocks of the Inlier dates from the time of Murchison and a brief
chronicle of subsequent researches was given by Whittard (1931 : 332; 1952) and, with particular
reference to the trilobites, in the first part of his monograph (1955: 1). As he noted, the first
geological maps of the region were provided at a scale of | in to 1 mile (1 : 63 360), although a
detailed 6 in to 1 mile (1 : 10 560) map by Lapworth & Watts remained unpublished except for a
reduced and ‘slightly modified’ version which appeared in the Geological Survey’s regional guide
to the Welsh Borderland (Pocock & Whitehead 1948: pl. 11; Earp & Hains 1971: pl. 4). The
frequent alternation of resistant bands of the Whittery, Hagley and Stapeley Volcanic Members,
and quartzites of the Stiperstones Member, with soft shaly bands that are also often present
within the volcanics themselves, gives rise to a ribbed topography which quickly shows any
displacement caused by faulting. The innumerable breaks in the run of these hard rocks, associated

6 W. F. WHITTARD

LAPWORTH, 1916

WATTS, 1925

Whittery
Shales
Whittery

Ash

LAPWORTH
& WATTS, 1910

Whittery
Shale

Whittery

LAPWORTH
& WATTS, 1894

LAPWORTH, 1887

Marrington or
Whittery Shales

Whittery Ashes &
overlying shales

Whittery Ash

Whittery
Shales

Hagley Volcanic
Ashes & Shales

Marrington Group
Marrington Group
Marrington Group

Ash =
°
&

Hagley Hagley
Shales Fy Hagley: shales Shales
®
oO
=

Hagley

H Ash

Ries

Shales with Di-
ctyonema, etc

Chirbury Series
Chirbury Series
Chirbury Series
Chirbury Series
Aldress Marm
H
Ee oe SEE a

Aldress
Shale

Aldress Grapt-
olitic Shale

Aldress Grapt-
olitic Shale

Dictyonema Band, L.
sericea Bands, Clima-
co. & Dicrano. Beds,

Beyrichia Grits

Group

Aldress

Spy Wood Calc-

areous Grit Beyrichia Grit

Spy Wood Gp
Spy Wood Gp

Leptograptus

Leptograptus Beds
ptograp Beds

Rorrington
Shales

ton Cp

Rorrington Group Rorrington Flags

Nemagraptus
Beds

Nemagraptus Beds

Ogygia Shales Ogygia Shales
Me adowtown
Mount Bed
Calcareous Beds (Mount Beds) sts & Flags
or Stage Meadowtown Lst, Flags Ogygia &

Middleton Group Asaphus Flags

D. murchisoni
Beds
Grits & Shale

Middleton Group Ogygia Flags

Didymograptus
murchisoni Beds

Betton or D. murch-—
isoni Shales

Middleton Series
Middleton Series

Meadowtown Series
Middleton Series

L. & U. Weston Grits
Weston Shales

Weston Flags
& Shales

lest-| Bett—

We
on St}on St

Weston Group Weston Group

West-—|Bett—| Meadow- | Rorring-

on Gp|on Gp} town Gp

Upper

> > :
a Stapeley Shales o Aleernating
Stapeley F eee ee a 8 Perey % &| sequence of
Volcanic Group tapeley Stapeley Ashes a. 0 as 3 lavas and
Rate ae 3 shales
_ |
3 a g |e L, M & U Hope Shales ch [te eee
7 Hope Shale Group i Hope Shales S|) eS & Chi . Y Rein dial - oa
in 8 iB fs inastone As Cin ee & Chinastones
n n 2
uv
o > @) Tankerville 2 Tankerville
Ladyyel UCzOur 4 g 2 (ey Flags & Shales a | Flags
a Mytton Group na Mytton Flags be iS a |
° Ladywell, Snailbeach fA Shelve Church
§ Sh., Lord's Hill Beds e Beds, etc.
3 ra
F = ; >
Stiper Group Stiper SEOs Stiper Quartzite = Stiper Stones = Stiperstones
Quartzite (Quartzite) Quartzite

Habberley Shales

Habberley Shales
(conformable)

Shineton Shales
(conformable) (conformable)

Cambrian Shineton Shales

Fig.3 Series of tables showing the progressive development in Ordovician

with high dips approaching the vertical, soon demonstrated that the inlier is traversed by tear-
faults which, in the present paper, are shown to possess a remarkably uniform pattern. Whittard’s
notes emphasize that most of these tear-faults, which constitute such a dominant structural
feature of the inlier and were reviewed by him (Whittard 1952: 186, et seqq.), had not been
recognized by Lapworth & Watts. A point reiterated in the notes is that such structures could
only be mapped satisfactorily on the basis of a detailed knowledge of the stratigraphical palae-
ontology, so that it became possible to differentiate between sets of outcrops which, although
lithologically distinct (and therefore true lithostratigraphic subdivisions) when clear sections were
available, nevertheless posed considerable problems of identification when seen only in the form
of small, weathered outcrops of shattered rock.

Geologically the inlier is bounded on the east by the Pre-Cambrian (Western Uriconian and
Longmyndian) rocks of the Longmynd and adjacent areas, which are separated from the
Shineton Shales of the Habberley Valley by a major structural line, the Linley—Pontesford Fault.
The latter was not shown on Whittard’s manuscript map but for the sake of giving a more
complete geological picture is included here together with certain details of the Linley and
Pontesford districts taken from papers by James (1956) and by Dean & Dineley (1961). The
southern boundary is formed by the overstep at the base of the unconformable Lower Silurian

WHITTARD, 1931

Marring-
ton St.

Whittery
Volcanic Group
Hagley Hagley
Stage Shales
Hagley
Volcanic Group
Aldress
Shales
| Aldress
Stage Spy Wood
Grit

Rorrington
Beds

Whittery
Shales

Meadowtown
Beds

Betton
Beds
Weston Grits
& Flags
Stapeley
Shales
Stapeley
Volcanic Group
Hope Shales &
Chinastone Ash

Mytton Flags
Stiperstones
Quartzite

Habberley Shales

Betton
Stage

|

WHITTARD, 1940

Rorrington Beds

Meadowtown Beds

Mytton Flags

Stiperstones Quartzite

Habberley Shales

SHELVE INLIER 7

WHITTARD, 1952

Whittery Shales

Whittery Volcanic Gp.

Hagley Shales

Hagley Volcanic Gp

Aldress Shales

Spy Wood Grit

1955 1966 PRESENT PAPER

WHITTARD,

WHITTARD,

Whittery Shales

Whittery Volcanic Gp.

Whittery

Whittery Shales Shale Member

with
Volcanic Gp.

Hagley Shales
and

Volcanic Gp.

Aldress Shales

Spy Wood Grit

Whittery
Volcanic Member

Hagley

HEGUSy BUNCE Shale Member

Hagley
Volcanic Member

Aldress Member

Spy Wood
Member

Hagley Volcanic Gp

CHIRBURY FORMATION

Aldress Shales

Spy Wood Grit

Rorrington Beds

Meadowtown Beds

f _neton moe

Weston Beds

Stapeley Shales

Stapeley Volcanic Gp

Hope Shales

Mytton Flags

Stiperstones Quartzite

Tremadocian

ae

Rorrington
Member

Rorrington Beds

Rorrington Beds

Meadowtown
Member

Betton Member

Weston Member

Meadowtown Beds Meadowtown Beds

Betton Beds Betton Beds

MIDDLETON FORMATION

Weston Beds Weston Beds

Stapeley

Stapeley Shales Shale Member

Stapeley Shales

]

Stapeley Volcanic Gp

Stapeley
Volcanic Member

Hope Member

Stapeley Volcanic Gp

Hope Shales with
Chinastone Ash

Hope Shales

Tankerville Flags

Mytton Member

SHELVE FORMATION

Shelve Church &
Lord's Hill Beds

Stiperstones Quartzite

Habberley Shales

Flags

Stiperstones
Member

Habberley Shales Shineton Shales

fae tatigraphic terminology for the Shelve Inlier from 1887 to the present.

rocks. Similar Silurian strata form the northern boundary but extend farther across the Ordovician
outcrop and are in turn overstepped by the unconformable base of the Coal Measures of the
Hanwood Coalfield. The relationships of these overlying Silurian and Carboniferous rocks are
much obscured by glacial deposits and by alluvium of Habberley Brook, River West Onny,

H River Camlad, and Aylesford and Rea Brooks. The western boundary of the inlier is still more
| obscured and it is possible that additional, and younger, Ordovician strata may underlie the
| cover of superficial deposits. Whittard’s unpublished notes on this area include the following

paragraphs:

Excluding the exposures of Hagley Shales in the village of Church Stoke, there are no localities on the
west of the Church Stoke — Chirbury road and south of the road from Alport to Cross House, Rhiston

and Upper Grwarhlow where any solid rock can be examined. Much of this area is about 500 ft (152 m)
O.D. and topographically it is composed of an eastward-trending ridge which is aligned at right-angles

much boulder-clay over the region, this feature and the absence of solid exposures can reasonably be
attributed to a thick glacial accumulation; this may be morainic in origin and may be associated with
the mass of ice which led to the formation of the Marrington Dingle overflow channel.

In one field two wells have been dug near Whitepits and both provided glacial lake clays and some

|
to any known strike in the Ordovician rocks or expected strike in the Silurian rocks, and since there is

8 W. F. WHITTARD

boulder-clay; the depth of one well was 36 ft (11 m) and solid rock had not been reached. Trial holes dug
to the north and south-west at Poundbank were also abortive, producing nothing other than boulder-clay.
The delineation of the western boundary of the Ordovician is accordingly most inexact and the very fact
that no feature is attributable to the Whittery Volcanic Group, which invariably is expressed in the topo-
graphy, almost certainly shows the Ordovician is covered by glacial deposits almost as far as the Church
Stoke — Chirbury road.

Classification of the Shelve Ordovician rocks

Lithostratigraphic subdivisions

The stratal subdivisions upon which Professor Whittard’s map is based are lithostratigraphic
units and represent the culmination of researches which began with Lapworth in 1887. The
original terms were subsequently modified and expanded, particularly by Lapworth & Watts
(1894, 1910), Lapworth (1916), Watts (1925) and finally Whittard (1931, and, to a lesser degree,
1940, 1955 and 1966). The manner in which the terminology evolved is shown by the comparative
stratigraphical columns in Fig. 3. The stratigraphic terminology employed by Whittard and his
predecessors does not conform with present-day requirements and does not follow the recom-
mendations of either the Stratigraphic Committee of the Geological Society of London or the
American Commission on Stratigraphic Nomenclature; consequently some appropriate modifica-
tions are proposed in the present paper. In an earlier publication, which contains his most
detailed account of the Shelve Ordovician rocks (excluding the Habberley (now Shineton) Shales,
of Tremadoc age), Whittard (1931) utilized a sequence of seventeen subdivisions, e.g. Mytton
Flags, Aldress Shales, etc., which clearly correspond to what would nowadays be termed litho-
stratigraphic or rock-stratigraphic terms. Following Watts (1925 : 340) these were grouped into
ten so-called ‘Stages’, an inappropriate method which utilized what is strictly a chronostrati-
graphic term for what were in fact groupings of lithostratigraphic subdivisions. These ‘Stages’
were then variously assigned to the ‘Arenigian, Llanvirnian, Llandeilian and Caradocian’, terms
which in this form cannot be employed as Series of the Ordovician System. These last were
replaced in the descriptive text of the same paper by the more appropriate usage of Arenig,
Llanvirn, Llandeilo and Caradoc Series, though without comment. The same ‘Beds’ and ‘Stages’
were utilized in the 1948 (second) edition of the Geological Survey’s handbook on the geology
of the Welsh Borderland (Pocock & Whitehead 1948 : 42) but abandoned in the latest (third)
edition of the same work by Earp & Hains (1971 : 41). Eventually Whittard (1952 : 156-165,
table III) omitted all mention of the above ‘Stages’, while most of the lithostratigraphic terms of
the 1931 paper were used and correlated with the various Series of the Ordovician, though the
Hope Shales were shown as undivided.

In the first part of his Shelve trilobite monograph (Whittard 1955: 5) the Tankerville Flags,
previously regarded as the topmost of four subdivisions of the Mytton Flags that had fallen into
disuse, re-entered the stratigraphical succession without comment as a separate subdivision
between the Mytton Flags and the Hope Shales. The Tankerville Flags were clearly regarded by
Whittard as meriting separate status as their trilobite fauna was listed by him later (1966 : 299)
in the distribution tables for the monograph, but in the same work (1966 : 303) he pointed out
that they could be distinguished in only a restricted part of the outcrop. The Tankerville Flags
had obviously not proved a satisfactorily mappable unit and were not differentiated as such
by Whittard on any of his 6 in to | mile field maps, where only the Mytton Flags occupy the
position between Stiperstones Quartzite and Hope Shales.

For present purposes it has been deemed appropriate to divide the Shelve Ordovician (excluding
Tremadoc) rocks into three Formations, names introduced by Lapworth, and subdivided in turn
into members that correspond to the lithostratigraphic terms used by Whittard and earlier
workers. As far as possible the type locality or stratotype for each member is listed and illustrated
by extracts from Whittard’s unpublished maps; additional data have been compiled from his
manuscript notes and, where necessary, published papers. Among the unpublished notes was a
proposed zonal sequence founded upon trilobites, the vertical distribution of which was shown in
detail in Part 8 of the monograph (Whittard 1966 : 299-302), and this has been incorporated in

ead

SHELVE INLIER

GRAPTOLITE
ZONE

Onny Shales or
Onnia Beds

Acton Scott Beds

Cheney
Longville Flags

Alternata Limestone

Horderley
Sandstone

Glenburrell
Beds

Smeathen Wood Beds
& Harnage Shales

Coston Beds &
Hoar Edge Grits

ABSENT

Shineton Shales

MIDDLETON FORMATION

2
o
=I
iS}
z
z
:
n

CHIRBURY FORMATION

Pleurograptus

ASHGILL SERIES NOT REPRESENTED IN SHROPSHIRE linearis Zone

Onnia gracilis

Onnia? cobboldi

Platylichas laxatus &
Calyptaulax actonensis

ACTONI AN

Broeggerolithus
transiens

ABSENT MARSHBROOKT AN

Broeggerolithus
longiceps
LONGVILLI AN

Broeggerolithus
globiceps &
B, nicholsoni

Broeggerolithus
soudleyensis

Broeggerolithus
broeggeri

Broeggerolithus
ulrichi

Salterolithus
caractaci

Reuscholithus
reuschi

Costonia ultima

Aldress Member

HARNAGT AN

Spy Wood Member COSTONI AN

Marrolithoides arcuatus

Rorrington Member & Spirantyx calvarina

Lloydolithus lloydi &
Cryptolithus inopinatus

Meadowtown Member

Betton Member

Mytton Member

Trinucleus acutofinalis
& Bettonia

Platycoryphe vulcani

Stapeleyella inconstans
& Ampyx linleyensis

Bergamia rhodesi

ill-defined interval

Myttonia confusa

Neseuretus grandior
Stiperstones

Menber No fossils found

Shineton Shales

questionably
present in
Shropshire

Dicranograptus
clingani

Zones
ill-defined

CARADOC

Diplograptus
multidens

Nemagraptus
gracilis

Glyptograptus
teretiusculus

Didymograptus
murchisoni

Didymograptus
bifidus

Didymograptus
hirundo

Didymograptus
extensus

Fig. 4 Table showing lithostratigraphic subdivisions for the Ordovician rocks of the Shelve Inlier
and the Caradoc District of Salop, with the corresponding Series and Stages (where available).
Since this paper was submitted, the name Cryptolithus inopinatus has been changed to Whittard-

aspis inopinata.

10 W. F. WHITTARD

the correlation scheme for the Inlier shown in Fig. 4. The lithostratigraphic sequence given below
is followed in the descriptive account of the Shelve Ordovician rocks.

Whittery Shale Member ; F 1000 ft+ (305 m+)
Whittery Volcanic Member . : 300 ft (91.5 m)
: ; Hagley Shale Member . : ; c. 1000 ft (305 m)
Chibusy Eomnation Hagley Volcanic Member : F 350 ft (107 m)
Aldress Member . ; ; J c. 1000 ft (305 m)
Spy Wood Member ; : ; 300 ft (91-5 m)
Rorrington Member , 3 1000 ft (305 ft)
MiddictonJEonnution ( Meadowtown Member . : F c. 1300 ft (396 m)
| Betton Member . ; : ? 600 ft (183 m)
Weston Member
Stapeley Shale Member . 3600 ft COT sa)
Stapeley Volcanic Member
Shelve Formation Hope Member ; : 5 : 800 ft (244 m)
Mytton Member . F ; ; c. 3000 ft (914 m)
Stiperstones Member . : . estd 400 ft (122 m)

The Shineton Shales

These strata, cropping out along the Habberley Valley between the Stiperstones on the west and
Pontesford Hill and The Longmynd on the east, were termed Shineton Shales by Callaway
(1878 : 333) on the basis of their lithology and stratigraphical position. Later Lapworth (1916 : 37)
introduced the term Habberley Shales for them, a name retained by Watts (1925 : 339) as there
was then no faunal evidence available to prove their equivalence to the Shineton Shales. Palae-
ontological proof was subsequently provided by Stubblefield & Bulman (1927 : 116-118), whose
faunal lists indicated a close correspondence with the strata of the Wrekin District. Whittard
(1931 : 323, et seqq.) continued to use the term Habberley Shales in an account (1931 : 324, fig. 43)
of an excavation, made in the vicinity of Granham’s Moor Farm, near Habberley, which demon-
strated what he described as an unconformity, marked by a 3 in (76 mm) band of conglomerate,
at the base of the overlying Stiperstones Quartzite. In view of the correspondence of dip below
and above the basal conglomerate of the Quartzite, the latter may perhaps be better regarded
as disconformable rather than unconformable upon the Shineton Shales. A brief mention of
this locality prior to excavation and a few notes of other localities constitute the only records of
the Shineton Shales in the Whittard documents. A later account by Whitehead (in Pocock et al.
1938 : 72) thought it unnecessary to retain the name Habberley Shales and commented on
Whittard’s excavation to the effect that the Stiperstones Quartzite/Shineton Shales junction was
conformable though ‘not necessarily without a break in deposition’. In the first part of his
monograph on the Shelve Ordovician trilobites Whittard (1955 : 5) continued to use the term
Habberley Shales, but his last published note on the subject (1960: 144) regarded them as
synonymous with the Shineton Shales. Whether one should assign these Tremadoc strata to the
Cambrian or the Ordovician is beyond the scope of the present work, in which the Shineton
Shales are excluded from the Shelve Formation.

Lithostratigraphic description of the Shelye Ordovician rocks

Shelve Formation

First USAGE. As Shelve Series, by Lapworth (1887 : 662), composed of the Stiper Group, Ladywell
Group and Stapeley Volcanic Group, in ascending order. The succession was later modified by
Lapworth (1916 : 37) to comprise Mytton Group, Hope Group and Stapeley Group, though the
lower and upper limits of the Shelve Series remained unchanged.

TYPE LOCALITY. Presumably the vicinity of Shelve hamlet, but no type section was designated.

SHELVE INLIER 11

/ yp Aye ok Nig REERARRY | | oe “py es ae on aa

Fig. 5 Eastern margin of Shelve Inlier, showing Pontesford Hill (Western Uriconian) at left, and
hills formed by Longmyndian rocks. View looking NE from above Upper Yessons, 0-9 mile
(1:5 km) east of Snailbeach and situated on the outcrop of the Shineton or Habberley Shales.

Stiperstones Member

First USAGE. As Stiper Group, by Lapworth (1887 : 662), overlain by Ladywell Group. The
original usage stated that the Group included ‘the well-known Stiper Quartzites’ and the term
‘Stiper Stones (Quartzite)’ was later listed by Lapworth & Watts (1894 : 317). Since then the
name Stiperstones Quartzite has come into general use and is replaced by Stiperstones Member,
though employed in the same sense.

TYPE LOCALITY. The escarpment of the Stiperstones, but no specific type section was designated.

DESCRIPTIVE NOTES. Quartzites of the Stiperstones Member form a structurally complicated
outcrop extending NE from the vicinity of the River Camlad, 1-5 km (0-94 mile) SE of Disgwylfa
(Squilver) Hill in the south, over the Heath Mynd to Black Rhadley where it is crossed by an
important oblique fault, beyond which minor faulting imposes a zig-zag outline upon the western
boundary as far as The Bog. Here the outcrop is reintroduced on the northern side of another
major oblique fault and runs through broken ground, each faulted mass tending to produce a
crag, so typical of the Stiperstones and culminating in the Devil’s Chair (Fig. 7). Succeeding
to the NNE are Scattered Rock and Shepherd’s Rock, beyond which the outcrop, though still
faulted, is not so intensely shattered and jointed. The strata can now be followed northwards
fairly readily, there being a significant change in strike north and south of Blakemoorgate,
but near Granham’s Moor an oblique fault, approaching the quartzite outcrop from the west,
curves into the strike and by repetition of the beds creates an increased breadth of outcrop as
far as just beyond Nill’s Hill. North of here the breadth is once more reduced, the strike swings
NE, and the outcrop terminates against the Coal Measures and glacial deposits of the Pontesbury
district after having covered a distance of more than 10 miles (16 km).

The Stiperstones Member is composed of hard, resistant rocks and forms an eastern edge to
the Shelve Inlier which attains a maximum height slightly in excess of 1725 ft (526 m). All along
the escarpment the quartzite tends to weather into crags but nowhere is this type of topography

12 W. F. WHITTARD

Fig. 6. Outcrops of Stiperstones (Quartzite) Member. View looking north from point south of
Manstone Rock, between Devil’s Chair and Cranberry Rock.

so well seen as between Cranberry Rock and the Paddocks, and The Rock and Nipstone Rock,
which significantly are the two lengths where the outcrop shows the most complex fault-pattern.

Reliable figures for the thickness of the Stiperstones Member are difficult to determine. The
angles of dip vary quickly and range from 40° to vertical, whilst in some places, as on both sides
of the River West Onny and at Nill’s Hill, the dip is reversed to the east. Nowhere is there a
section through the Member, neither the upper nor the lower limit is visible in more than a few
places, and only indirect evidence of the thickness can be utilized. Topography is a most uncertain
guide because the escarpment is continually being degraded, crags are actively being reduced in
height and enveloped in scree-clitter and, by virtue of bedding and of jointing parallel to the
strike, rocks fall on both dip and scarp slopes. The marked change of slope on the scarp side
suggests the outcrop of the softer Shineton (—Habberley) Shales, Tremadoc Series, but this can
be deceptive because at several locations undoubted exposures of quartzite occur in what, on
this assumption, should have been Tremadoc ground. Consequently the Stiperstones Member
may at times be below the surface but farther down the scarp than is expected, and its failure to
crop out is due to ‘scaling-off’ of the quartzites along the joints. On the dip slope, exposures are
hardly ever seen owing to the blanket of quartzite blocks, heather and whinberry. How far west
the top of the Member extends is frequently difficult to determine, and topography reveals little
because the quartzites show a zone of gradation into the slightly less resistant Mytton Member.
Thus the thickness of the Stiperstones Member as shown in the craggy exposures appears to be
not more than 100 ft (30-5 m), which generally is probably correct because the crags are eroded
from faulted blocks in which the top or bottom, or sometimes both boundaries, of the sub-
division are cut out. The thickness of the Stiperstones Member probably does not exceed 550 ft
(167-6 m) in the north of the inlier between The Paddock and Lord’s Hill, but apparently increases
SW to c. 650 ft (198 m) on the south side of Heath Mynd. The beds are probably strongly dia-
chronic but palaeontological evidence is insufficient to prove this. The Member is invisible at the
surface in the Shelve Anticline, but the estimated depth at which it has been supposed to occur
is 1200 ft (366 m) (Hall 1919 : 12; Dines 1958 : fig. 2).

The dominant rocks of the subdivision are massive, compact and usually moderately well

SHELVE INLIER 13

Green Hill

Die

ss Perkins Beach

Fig. 7 Geological map of the area between Myttonsbeach and The Bog. Adjoins Fig. 10. 3=
Shineton Shales; 4=Stiperstones Member; 5=Mytton Member; 6=Hope Member; black
outcrop D=dolerite. (SO 39 and SJ 30)

bedded quartzites which are white, light grey, bluish-white or liver-coloured; they are sometimes
colour-banded in alternations of bluish-grey and light fawn. The banding accentuates false
bedding as seen on weathered joint faces, for example at the northern end of the Manstone
Rock; occasionally the bedding is lenticular and produces lenses up to 2 ft (60 cm) long and a few
cm thick; at other times it is undulose, the wavelength being 2-3 ft (about 75 cm) and the ampli-
tude a few cm; symmetrical ripple-marks are uncommon, as in the Nill’s Hill quarries. Hard,
shiny, black bitumen was noted on joint faces north of The Hollies, Snailbeach and at Nill’s Hill.
Bedding surfaces show burial preservation examples of Chondrites sp., animal tracks referred to
Cruziana semiplicata Salter (1866: 292) and fucoidal markings, but the commonest remains are
worm-tubes infilled with sand which is coarser-grained than that of the surrounding quartzite.
The tubes are at right angles to the bedding, frequently show paired openings as though they are
U-shaped, and vary in diameter from } to | in (3-25 mm). In regions where dips are high and the
quartzites severely jointed, as at the Devil’s Chair, the occurrence of these infilled worm-tubes is a
distinct aid to the determination of bedding. Angular fragments of blue-black shale and sandy
mudstones, as much as 2 in (5 cm) across and indistinguishable from the Mytton Member, lie
in the bedding planes; they have often been removed from the quartzite by weathering, leaving
characteristically-marked surfaces with flat-bottomed, vertical-sided, irregularly rectilinear
depressions. Much of the compact quartzite is secondarily silicified and in extreme cases is
converted into chert-like rocks; rarely the quartzites are felspathic.

The most frequently seen conglomerates are pale in colour, occur at any horizon, and are
normally composed of a quartzite matrix in which small pebbles of quartz, chalcedony, quartz
aggregate and cryptocrystalline felsite, rarely spherulitic, are included, and they seldom exceed
1 in (2:5 cm) in maximum diameter. Noteworthy is the absence of granite types and of feldspar,

14 W. F. WHITTARD

ES)

ity Wings

aes, Vase jeans = Rie ‘aati ANC estas NS ais il ets, Fo as 238
Fig. 8 View of the Stiperstones, with Tankerville Hollow on left, looking east from Round Hill,
0:7 mile (1:1 km) ENE of Shelve hamlet.

unless some of the fine-grained micaceous pebbles are highly weathered feldspar in which all
traces of twinning are lost. The basal conglomerate, 3 in (7:5 cm) thick, exposed in an excavation
by the side of a derelict concrete tank opposite Granham’s Moor quarries, is much darker in
colour owing to ferruginous and chloritic fragments, and interstitial argillaceous matter. The
provenance of the conglomerates, and also of the quartzites, is unknown. If the derivation had
been from the east, i.e. the direction of the immediate land-region of those times, the content of
Uriconian and Longmyndian material would have been much higher than it is; consequently,
an along-shore drift of sediment, presumably from the SW, is not unlikely. Also possibly bearing
on this problem is the existence at Pontesford, only 1-2 miles (about 2-5 km) away, of repre-
sentatives of the Caradoc Series which rest directly upon Pre-Cambrian. There are many geo-
logical anomalies here; the unconformity at Pontesford is undeniable yet there is no evidence of
such an important hiatus in the corresponding succession of the Shelve Inlier. Important faults
occur between the Inlier and the Ordovician outcrop of Pontesford, and two geological regions
which were at one time widely separated may have been abruptly collocated by strike-slip faulting
in which the movement was essentially horizontal, while downthrow may have brought to about
the same present topographical level two deposition surfaces which originally were separated,
relative to one another, by a large vertical interval.

The entire thickness of the Stiperstones Member is not, however, made up of quartzite and
conglomerate. Towards the base are interbedded silvery, greenish-white, soft, soapy shales which
have yielded dendroid graptolites and are only a few cm thick, whilst at Nills Hill thin papery
shales and sandstones are present in bands up to 10 in (25 cm) thick. Elsewhere thin, micaceous,
dark grey shales alternate with layers of coarse but friable quartzite, and these are included within
massive quartzites. In the mid-thickness of the subdivision both soapy shales and fucoidal flags
similar in lithology to the Mytton Member occur.

As revealed under the microscope, much of the compact quartzite has been secondarily
silicified. Within narrow limits the grains are generally of uniform size and form an interlocking

SHELVE INLIER 15

mosaic: the clear secondary silica has grown in optical continuity with the quartz of the original
grains, which are frequently dusty from inclusions and are rounded, the ghost of their outlines
being clearly evident in many samples. At other times, the original outline is lost and secondary
silicification is then recognized by the irregular sutural contacts made between individual quartzite
components. Secondary growth of silica does not usually form around quartz aggregates.

The normal sedimentary base of the Stiperstones Member has not been seen in any natural
exposure, but the contact with the Shineton Shales was found to be exceedingly sharp in an
excavation made some years ago near a disused concrete tank on the opposite side of the valley
to the lower Granham’s Moor Quarry. At its eastern extremity the excavation showed olive-green
shales and micaceous flags; 5 ft (1-5 m) below the succeeding conglomerate Dr C. J. Stubblefield
in 1931 collected Shumardia pusilla (Sars) and Lingulella nicholsoni Callaway, thus proving their
Tremadoc age, Shumardia pusilla Zone (Whittard 1931 : 344). Resting upon, and inclined at the
same angle as, these Shineton Shales is a 3 in (7:5 cm) band of dark, distinctive conglomerate
which makes a clean-cut junction and is of a type unknown from the main mass of the Stiper-
stones Member. The conglomerate is succeeded by a thin band of broken shale, unlike the
Shineton Shales, and finally by flaggy quartzites which definitely belong to the Stiperstones
Member. The excavation was terminated by blocks of quartzite set in a sandy matrix, but these
were not in situ. The junction between the Stiperstones Member and the Shineton Shales was
taken at the base of the conglomerate because (i) no conglomerate is known in the Shropshire
Tremadoc rocks; (ii) the conglomerate consists mainly of angular quartz grains, interstitial
ferruginous and chloritic material, and rounded pebbles of quartz and quartz-aggregate, all of
which can be matched in undoubted Stiperstone quartzite; and (ili) the Stiperstones Member is
disconformable to the Shineton Shales as the younger of these Tremadoc deposits belong to the
Shumardia pusilla Zone, whilst representatives of the higher Apatokephalus serratus Zone are
absent. The disconformity can only be inferred on palaeontological and lithological evidence,
because there is no discordance in angle of dip, neither is there any detectable difference in strike;
but with such easily-deformed strata as make up the Shineton Shales, the mass of the overlying,
thick Arenig beds has probably brought the inclination of their upper layers into coincidence
with that of the quartzites.

The upper limit of the Stiperstones Member is different from the lower because there is no sharp
line of demarcation; indeed, the condition is that found in passage beds where one lithology is
superseded by another through a thickness of alternating types. Stratigraphically above the
dominantly quartzitic beds is an interval, of unknown magnitude, wherein the light-coloured,
secondarily silicified, clean quartzites alternate with striped, micaceous, rusty-weathering, bluish-
grey shales and siltstones of an aspect like rocks found in the Mytton Member; these are included
in the Stiperstones Member. Still higher, however, the clean quartzites are replaced by dirty, dark
quartzites, usually but not always thin-bedded, carrying much argillaceous matter, alternating
with sandy shales and flags; these are placed in the Mytton Member and clearly represent a
sedimentary transition. Whereas the lower limit of the Stiperstones Member is precise, the
upper can be drawn on the map only arbitrarily.

The commonest indications of organic life in the quartzite are innumerable worm-tubes of
many sizes, | in (2:5 cm) and less in diameter, frequently but not always U-shaped, and filled with
coarser-grained material than the enclosing sediment. Salter (1866 : 243, 256, 291-293, fig. 2;
pl. 11B, fig. 27) compared the burrows with Arenicolites linearis (Hall). No corrugated skin has
been detected in the many tubes which have been examined in the quartzites. The only other
fossils are two pygidia of a large trilobite, Neseuretus grandior Whittard, and a sponge-spicule
of which four axons are identifiable in thin-section. Until recently these were all the known
fossils from the Stiperstones Member but now a few dendroid graptolites and some anomalous
and structureless remains, which may be algal, have been recovered from silvery and soapy shales
interbedded with quartzite in the quarry at the NW corner of Poles Coppice, north of Granham’s
Moor Farm. Unfortunately Neseuretus grandior and the dendroids are valueless for detailed
correlation. Neseuretus is found in the Lower Ordovician, is particularly common in parts of the
Mytton Member, and there is no reason to consider the Stiperstones Member to be other than
Arenig in age.

16 W. F. WHITTARD

Fig. 9 Mytton Batch (=Myttonsbeach) and Snailsbeach, formed by rocks of the Mytton Member.
View looking NNE from above Bergam, 1-3 miles (2:1 km) NNE of Shelve.

The quartzites accumulated as coarse-grained, worm-ridden sands with pebbly layers, and
comprise a well-sorted deposit generally washed clean of mud. The fact that animal tracks were
impressed and preserved suggests that the water was shallow and, indeed, the conditions may
have been littoral. The sands were covered by muds for short intervals, particularly during the
earlier periods of deposition, but mud and sandy mud became dominant towards the top of the
subdivision as Mytton Member conditions became established.

Mytton Member
TYPE LOCALITY. Mytton Batch, near Snailbeach (Whittard 1960: 194).

FirsT USAGE. As Mytton Group, by Lapworth & Watts (1894 : 316), later renamed Mytton Flags
by Lapworth & Watts (1910: 752). The term Mytton Group as employed later by Lapworth
(1916 : 37) was equivalent to the Stiper Quartzite and Mytton Flags of previous accounts; it was
subsequently renamed Mytton Stage by Watts (1925 : 341), a term employed temporarily by
Whittard (1931 : 323), and is now no longer used. Whittard’s (1931 : 323, 325) usage of Mytton
Flags corresponds to the present Mytton Member. He noted also Lapworth’s subdivision of the
unit into the following:

4. Tankerville Flags

3. Shelve Church Beds

2. Ladywell & Snailbeach Grits and Flags

1. Lord’s Hill Beds.
The lowest of these was said to be unfossiliferous but brief notes were given on the faunas of the
other three. Later Whittard (1940 : 154; 1952 : 157, table on p. 165; 1953 : 239, 240) used Mytton
Flags in a similar sense without further subdivision and, noting Lapworth’s thickness of 1600 ft
(490 m) as being far too small, gave a figure of c. 3000 ft (915 m). He appears to have been
uncertain how to treat the problem of further subdivision because in the first part of his trilobite
monograph (1955 : 5) he reverted to the fourfold subdivision of Mytton Flags. In Part 8 of the
same monograph, his final published contribution to the subject (1966 : 299-304), he rejected
the term Shelve Church Beds (and presumably also the Lord’s Hill, Ladywell and Snailbeach
Beds) but retained the Tankerville Flags as a subdivision between the Mytton Flags and the Hope

SHELVE INLIER 17

Shales, though acknowledging that the distinctive lithology (comprising c. 100 ft (30-5 m)
of ‘bluish-grey flaggy shales with but little content of arenaceous material’) and fauna could be
recognized ‘only over a restricted part of the outcrop’. The field mapping completed by Whittard
shows only undivided Mytton Flags (now Mytton Member of the Shelve Formation), and
nowhere was a separate Tankerville unit mapped. Consequently it is preferred to disregard the
latter for most practical purposes although, in view of Whittard’s 1966 remarks, the topic is
discussed further with reference to the Tankerville area.

DESCRIPTIVE NOTES. Whittard (1931 : 325; 1960: 194) noted the rocks only briefly as rusty-
weathering, olive-green, ribbed, flaggy, sometimes quartzitic siltstones and bluish-grey, flaggy
shales with a total thickness of c. 3000 ft (915 m). Lapworth & Watts’ map (in Pocock & White-
head 1948: pl. 11) showed ‘Mytton Flags & Shales’ occupying two large outcrops within the
Shelve Inlier: an elongated area NW of the Stiperstones and forming part of the eastern limb of
the Ritton Castle Syncline and a subelliptical area centred approximately on the Shelve District
and forming the core of the Shelve Anticline. This distribution corresponds broadly to the one
shown in the present work.

In general the rocks of the Mytton Member comprise rusty-weathering, massive and flaggy,
grey-green and blue-black siltstones and subsidiary shales. Although less resistant to erosion than
the quartzites of the Stiperstones Member, the beds are generally more resistant than the shales
of the succeeding Hope Member and form the conspicuous area of high ground extending from
mid-way between Pontesbury and Minsterley SSW to Pennerley, beyond which the outcrop
narrows markedly at the marshy area of The Bog. As Whittard (1931 : 325) emphasized, the
rocks of the Mytton Member are important in that they contain the majority of the principal
mineral veins (galena, blende, barytes) of the district and the remains of old workings are
commonplace over much of the outcrop. Mining in the Shelve Inlier has since been reviewed in
some detail by Dines (1958), who demonstrated that fault fissures receptive to mineral deposition
are better developed in ‘harder’ formations such as the flags of the Mytton Member and die
out in shales, where the veins are barren. The quartzites of the Stiperstones Member, althcugh a
hard formation, contain virtually no mineral deposits but the latter are common in rocks of the
Mytton Member owing to the sealing effect of the overlying shales of the Hope Member, which
prevented the upwards migration of mineral-bearing solutions. In the relatively few cases where
mineralizing solutions escaped through the Hope Member shales to higher stratigraphic horizons
in the Stapeley Volcanic and Hagley Volcanic Members, the occurrences are of barytes with only
a few traces of galena and blende.

The striking unwooded valleys which drain the west side of the high ground underlain by the
Stiperstones and Mytton Members are known locally as ‘batches’ or ‘beaches’. They are remark-
able physiographic features and in the case of Mytton Batch, usually written as Myttonsbeach on
present-day maps, the slope of the head of the valley, covered by heather and whinberry, is 35°.
Whittard’s notes speculate that although it has been suggested that some of the batches are due
to marginal overflows formed during the glacial period, they may in fact have originated from
springs of which the flow passed down into the much lower-lying country occupied by the
regressive Hope Member. If the difference in level was maintained between country formed of
Mytton Member and that made of Hope Member, the velocity of the stream would be high and
would give it considerable power to cut back quickly. The batches do not show devious courses —
they are almost straight.

Mytton Batch, type locality for the Mytton Member, is a valley (Fig. 10) running east-west on
the north side of Green Hill, some 3 miles (4-8 km) SSW of Minsterley. The rocks are particularly
well exposed along the north side of the batch forming the south flank of Oak Hill. Those exposed
in the section between the Minsterley — Pennerley road, near the one-time Stiperstones Inn, and
the area east of Myttonsbeach cover the greater part of the succession, but the lowest strata are
not visible and the first exposures recorded there are at MS Locs 832, where massive, brecciated
siltstone, veined with quartz, is seen, and 831, where the rocks dip 46° at 315° true. The latter
locality shows massive siltstones which in no way resemble shales, and this type of lithology is
found elsewhere, including the heads of the batches, e.g. Tankerville Hollow (see p. 13). Generally

18 W. F. WHITTARD

N65

Crowsnest
\ Dingle LAY

/ Gorstybank
Samat <a

N55
p.

Oak Hill

Not
Loc
“a> * 801 / /
/ The,/Paddock
/ /
Loc 831 SO

+
Myttonsbeach Ve Loc 4

+ 399
SS

Green Hill

Fig. 10 Geological map of the Mytton Member at and north of the type locality. Adjoins Figs 7, 11.
3=Shineton Shales; 4=Stiperstones Member; 5=Mytton Member; 6=Hope Member; black
outcrop D=dolerite. (SJ 30)

speaking, the softer siltstones and shales, particularly at the top and near the base of the Mytton
Member, have proved fossiliferous, but the middle part of the succession, constituting the greatest
thickness, comprises coarse siltstones, some of them very massive, and fossils are rarely found
therein. From here westwards, there are numerous exposures in the area just east of the head of
Myttonsbeach, including fossiliferous flags (Loc. 821) with Myttonia confusa Whittard, Neseuretus
and Ogygiocaris selwynii (Salter), but these end at the valley, though the section continues along
the top of its northern slope. Very massive flags make the main edge-feature of Mytton Batch,
coarse siltstones of a type encountered in many places throughout much of the thickness of the
Mytton Member. Looking downhill along the strike of these rocks, one can see three distinct
ridges separated by hollows, features which indicate the alternation of hard and softer rocks. A
similar alternation is applicable to most of the thickness of the Member as seen elsewhere in the
district. Not only is there a coarse alternation involving thicknesses of up to 100-200 ft (30-5—
61 m), but also within these thicknesses there are minor hard and soft strata alternating and in
this respect the rocks of the Mytton Member are not unlike those of the Weston Member (see
p. 38). This, however, is not the condition of the lowest beds, which are usually softer, silvery-
grey weathering and frequently fossiliferous. The latter are the so-called ‘Lord’s Hill Beds’ of
Lapworth, and if we exclude these and the softer strata of the ‘Tankerville Flags’ from the total
thickness, we are left with by far the greatest proportion of the Mytton Member made up of
more massive rocks which presumably are Lapworth’s ‘Ladywell and Snailbeach Grits’.

On the south side of Green Hill the valley called Perkins Beach (Fig. 7, p. 13) contains numerous
exposures of the Mytton Member, though the section is less extensive than that of Myttonsbeach.
Loc. 805, beside an old adit, is important as it has yielded what is apparently the oldest fauna
known from the Member, comprising M. confusa, O. selwynii and Neseuretus preserved in silver-

SHELVE INLIER 19

grey weathering, micaceous, silty shales and flags. An interesting feature of this area is the
occurrence of doleritic intrusions, which appear at the surface in crags and then disappear
uphill; when followed along their strike, their place is taken by Mytton rocks which are sometimes
indurated, sometimes not. One has the impression that the dolerites have only been exposed by
the excavation of the valley, without which it is doubtful whether they would have broken surface.

The manner in which the main thickness of the Mytton Member is made up largely of greyish-
weathering, blue-hearted siltstones is evident when working along Perkins Beach and the valley
known as Tankerville Hollow, which runs ESE from Tankerville Mine (Fig. 7). The name
Tankerville, incidentally, derives from an old term for an oven-pipe or organ-pipe, and refers to
the shape of some of the mine shafts, which were narrower at the base than at the top. The
alternation of predominant siltstones, which are usually exposed but yield few fossils, with shales,
which may be fossiliferous but are rarely exposed, has made it impossible to subdivide the strata
of the Mytton Member on the basis of their faunal content.

Bergam Quarry (Loc. 783), sited c. 1000 ft (305 m) NNE of Tankerville Mine and excavated in
the highest portion of the Mytton Member (Fig. 7), is of interest as being the first locality at which
species common to this horizon and the succeeding Hope Member were discovered. The rocks
are bluish-grey, micaceous, flaggy shales with but little content of arenaceous material, and are
hardened by a dolerite intrusion some 14 ft (4:3 m) wide. The shales on the north side of the
intrusion are pale grey through a width of 11 ft (3-4m), but those on the south side are not
visibly altered. Fossils are hard to come by nowadays, but in the past the locality has yielded
Didymograptus hirundo and numerous trilobites including Bergamia rhodesi Whittard, Ectillaenus,
Geragnostus, Placoparia and Pricyclopyge.

To the south of Tankerville Hollow exposures of Mytton Member become fewer and an
east—west fault running through Pennerley Mine (see Fig. 7) offsets the outcrop eastwards adjacent
to the swampy area of The Bog. In the vicinity of The Bog lead mine no rocks are now visible
in situ, though the tip heaps are composed of Mytton Member débris, but c. 1000 ft (305 m) to
the ENE Loc. 279 yielded a single example of Ogygiocaris and a dubious Tetragraptus.

The outcrop of the Mytton Member narrows markedly to the south of The Bog and, accom-
panied by an elongated series of small faulted ‘blocks’ of Stiperstones Member, runs in a narrow
strip bounded by two convergent faults which meet just south of the River West Onny and east
of Cefn Gunthly (Map). From there to its termination about half a mile (0-8 km) east of Snead
the outcrop is not well exposed and the western boundary is faulted against rocks of the Hope
Member. The rocks are seen only intermittently on the heath-covered ground in the vicinity of
Heath Mynd, and make up the eastern part of the hill feature of Roveries Wood, to the south
of which they are overlain unconformably by Lower Silurian beds.

Turning to the area north of Myttonsbeach, the valley of Crowsnest Dingle shows numerous
exposures of the more massive siltstone beds but, again, these die out eastwards so that the
lowest strata of the Member are not exposed. Half a mile (0-8 km) farther NE the area east of
Snailbeach (Fig. 11) is more informative. Immediately west of the mapped base of the Mytton
Member is a building called Eastridge, just west of which, at Loc. 858, are exposures of typical
flags with a preponderance of much harder layers of quartzitic rock. It is evident that the beds
represent a transition from those of the Stiperstones Member, and similar strata follow successively
to the west where, at Loc. 860, apparently the youngest beds of this sequence are still quartzitic
and include a band 7 ft (2:13 m) thick. Whittard clearly contemplated separating these un-
fossiliferous, transitional rocks under the name Eastridge Beds, but their outcrop was shown only
sketchily in the vicinity of Eastridge and he appears not to have recognized them elsewhere. In
the same valley, however, and only a short distance south along the strike from Loc. 860, is an
exposure (Loc. 158) of buff-weathering, bluish-grey hearted ‘grits’ which Whittard, in his notes,
considered to be probably the ‘Snailbeach Grits’ of Lapworth. This conclusion seems arguable
as the rocks occur below the shaly strata of Lordshill (see p. 18), and Lapworth (1916: 37) placed
his ‘Ladywell and Snailbeach Grits and Shales’ above his ‘Lord’s Hill Beds’. However, to use
either of these possibly synonymous terms, which cannot be employed with confidence outside a
relatively restricted area, is to risk nomenclatorial confusion and it would be better to allow both
to lapse.

20 W. F. WHITTARD
37 38

Maddox's Coppice! Eastridge

4
rae Loc 889 Sy Wood

i

1

H
ve 893 5
1

Js: "Ne
1894 ,Loc 891

Sanilieesen
Coppice
— +Loc 896
ba

+ LOC 897

ux
\
i)
\
i}
\
\
\
i)
i)
\
\

--._
---.

1

{
Loc 145 + H
Loc |

\ ‘
1 858
te _— 7100 860+ +s, Eastridge
]
pe Ew Lordshill 9 +boc1s8 180
Farm€ Chapel
Loc157 +

Loc 148 ++ 75;
Yewtree
| eis
Level wie j
oie Hollies
37 Ded hey by Sei 5 pRB,
Fig. 11 Geological map of the area around Snailbeach. arene Fig. 10. 3=Shineton Shales;
4—Stiperstones Member; 5=Mytton Member; 6=Hope Member; 20=Silurian rocks; black
outcrop D=dolerite. (SJ 30)

————
~—----—.

t
t
a]
)
1
i)
i 4
1
!
1
yi

Although Whittard’s notes state that the Lord’s Hill Beds “must not be confused with the
fossiliferous beds of Lord’s Hill so often mentioned by Salter’, he appears to have used the terms
in at least partly similar sense. South of the Baptist Chapel at Lordshill, towards Yew Tree Level,
and to the north, in the vicinity of Perkins Level, shaly and occasionally flaggy beds overlie the
transition beds noted above and have proved fossiliferous at a few places (Locs 145, 148, 897)
along the strike; the fauna includes Bergamia rhodesi, Myttonia confusa, Neseuretus parvifrons
(Salter), Ogygiocaris selwynii and Redonia anglica Salter.

Successively higher beds are encountered westwards along the road from Lordshill to Snail-
beach and, to an even greater degree, NW from Perkins Level through Snailbeach Coppice. A dry,
right-angled ‘gutter’ (Loc. 157), possibly a glacial overflow channel, centred c. 250 ft (76 m) west
of Lordshill Farm, shows a large exposure in green-weathering, greyish, resistant, gritty flags.
Across the valley and at a slightly higher horizon, a small pathside quarry (Loc. 896) c. 500 ft
(150 m) NW of Perkins Level shows hard, blocky, jointed siltstones which are cut by a small
dolerite intrusion immediately to the south. The beds here and in the adjacent spoil heap at one
time yielded fossils including trinucleids, Neseuretus and Ogygiocaris. Other exposures (Fig. 11) in
Snailbeach Coppice merit little detailed description but Loc. 891 shows massive quartzitic
siltstones, while flags at Loc. 893 are disturbed by doleritic intrusions. The rocks at Loc. 894,
a sunken cart-track, are apparently almost at the top of the Mytton Member but do not resemble
the lithology of the ‘Tankerville Flags’ discussed earlier. A note here mentions that ‘Tankerville
Flags’ were last seen in the roadside section (Loc. 846) just south of Crowsnest Dingle (Fig. 10,
p. 18), but the field notes for that place mention only ‘Mytton Flags’. The development of
‘Tankerville Flags’ in this area seems inadequately documented, though flags exposed by the
side of the old railway at Loc. 854, c. 700 ft (210 m) NE of Crowsnest (Fig. 11), were described as

SHELVE INLIER vA

fi Pontesbury J

Fig. 12 Geological map of the area south of Pontesbury. 3=Shineton Shales; 4—Stiperstones
Member; 5=Mytton Member; 6=Hope Member; 21 =Carboniferous rocks; black outcrop D=
dolerite. (SJ 30)

‘slightly more shaly than usual’. The wooded area of Maddox’s Coppice, immediately NE of
Snailbeach Coppice, contains few exposures but an old adit (Loc. 889) yielded fossils in silty
shales and mudstones which occur at approximately the same horizon as similar strata near
Perkins Level.

The Mytton Member’s outcrop extends NNE (Fig. 12) to within about half a mile (0-8 km)
of Pontesbury, where it is overlain unconformably by Carboniferous strata. Adjacent to the
faulted western boundary is a large quarry (Loc. 888), excavated in the NE part of Callow Hill
and approximately three-quarters of a mile (1:2 km) east of Minsterley. The rocks there are cut
by dolerite intrusions which extend into the adjacent outcrop of the Hope Member; they show
little variation and consist for the most part of massive, rusty-weathering, blue-grey siltstones
which sometimes show thin, argillaceous layers and rod-like fucoidal structures.

All the outcrops of the Mytton Member so far considered form part of the eastern limb of the
Ritton Castle Syncline. The remainder constitute an inlier, c. 24 by } to ? mile (4c. 1 km) and
elongated ENE, which forms the core of the Shelve Anticline. Truncated at its SW end by one of
the Shelve Inlier’s set of principal north-south tear-faults, the overall outcrop is divided into two

22

W. F. WHITTARD

+ Loc 703

l
|
|
|
l
|
|
|
|

|

|

A
m+ Loc 669

‘ ~\\ Whitegrit

|
|
|

|

I

|

|

| .
| Mine
|

25 a

Kee Oe Jseea
6 ay

Fig. 13 Geological map of the area SW of Shelve hamlet. Overlaps Fig. 14. 5=Mytton Member;
6—=Hope Member; 7=Stapeley Volcanic Member (7a=interbedded shales); black outcrop D=

dolerite. (SO 39)

subequal parts by a second, subparallel fault which runs through Roman Mine and just west of
Shelve hamlet. The SW part is centred broadly on Shelve Hill (Fig. 13), an elongated feature
which stands notably higher than the surrounding, often badly-drained ground underlain by
regressively-weathering shales of the Hope Member. Scattered exposures are most common
towards either end of Shelve Hill, where three disused mines (Whitegrit, Oldgrit and Ladywell)
are situated, but good sections are generally lacking. An old quarry (Loc. 670) 450 ft (137 m)
NW of Whitegrit school exposes sandy, micaceous flags, weathering yellowish-grey, from which
the type material of Dictyonema cobboldi Bulman (1928 : 33), together with D. irregulare Hall,
was obtained. The locality was quoted by Whittard (1931 : 327) as being 330 yd (300 m) SSW of
Whitegrit Mine, a misprint for SSE. About 450 ft (137 m) east of the same school, a collapsed
mine shaft and its associated tip heaps of blue-hearted, shaly flags were noted by Whittard
(1953 : 240) as having proved fossiliferous. The fauna includes Ogygiocaris selwynti, Monobolina
plumbea (Salter), Redonia anglica, hyolithids and rare Didymograptus. In this area the linear
arrangement of shafts and tip heaps running ENE between Whitegrit and Oldgrit suggests a

SHELVE INLIER 23

poor, Aajjues

300

Loc

| Church sie me

: be *s 6381
Shelve Loc Va
4 a
De VE
33 34

Fig. 14 Geological map of the area north of Shelve hamlet. Overlaps Figs 13, 15, 16. 5=Mytton
Member; 6=Hope Member, including ‘chinastone’ tuff bed (dotted); 7=Stapeley Volcanic
Member; 8=Stapeley Shale Member; black outcrop D=dolerite. (SO 39, SJ 30)

line of at least brecciation if not faulting. In the same general area Loc. 669, 800 ft (244 m)
at 034° true from Whitegrit school, produced only two specimens of possible ogyginid trilobites
and a fragmentary graptolite, but the rocks are of interest as being softer than the typical lithology
and probably lie very close to the top of the Mytton Member. At the NE end of Shelve Hill the
rocks are not well exposed, and spoil heaps of shale (Locs 677, 703) in the vicinity of Ladywell
Mine (Figs 13-14) nowadays provide the best opportunity for collecting fossils. The latter include
Ampyx salteri Hicks, Neseuretus, Ogygiocaris selwynii and Didymograptus.

At the northern extremity of the SW half of the Mytton Member outcrop, beds very close to
the boundary with the Hope Member are reported from the neighbourhood of Loc. 716, just
west of Roman Mine (Fig. 14). As the hill is descended there appears to be an increase in shales
approaching the condition of those forming the Hope Member and in this area there is thus at
least a suggestion of a gradational boundary, a situation contrasting with that described in
notes on part of the Hope Valley where a gradation was not definitely observed (see p. 24).

East of the north-south tear fault already noted as passing through Roman Mine, the NE
half of the Mytton outcrop is roughly rhombic in outline, its SE boundary faulted against the

24 W. F. WHITTARD

Hope Member. At Shelve hamlet the best-known rocks are those known at one time as Shelve
Church Beds; they were once clearly exposed (Fig. 14) in the road beside All Saints Church but
that outcrop is now much overgrown or built over and the beds are visible only in an adjacent
field (Loc. 720). Excavation there exposed blue-black, shaly flags interleaved with silver-grey
weathering, rusty-coated, blue-hearted shales in which the mainly graptolitic fossils occur matted
together in thin bands. The fauna, though much quoted (for example by Whittard, 1931 : 326),
has not received modern descriptive treatment but is famous for dendroids such as Dendrograptus,
Dictyonema and Desmograptus, while graptolites including Didymograptus extensus (Hall) and
Glyptograptus dentatus (Brongniart) also occur. Shelly fossils are in a distinct minority. Whittard’s
rejection of the term ‘Shelve Church Beds’ was noted earlier. In the relatively small area between
Shelve Church and the faulted boundary against Hope Member to the SE, typical though un-
fossiliferous rocks of the Mytton Member were recorded at Loc. 681 (Fig. 14) while nearby, at
Loc. 682, the ‘Shelve Church Beds’ lithology was seen, only a few metres from shales of the
Hope Member which have been hardened by a dolerite intrusion. No other exposures of note
occur in the immediate vicinity of Shelve but to the NE the outcrop is invaded by a large dolerite
intrusion and metamorphosed ‘flags’ were at one time quarried at Locs 736 and 737, respectively
WSW and SW of Shelfeld Farm.

From here to the NE apex of their subrhombic outcrop, the rocks of the Mytton Member
are intruded by three additional masses of dolerite, the largest of which is elongated roughly
north-south and underlies the western half of Santley Wood (Fig. 14). In the area immediately
west of these intrusions the boundary between the Mytton and Hope Members is not actually
visible but must lie near the base of the physiographic feature formed by the Mytton rocks
near the SE side of Hope Valley and midway between Roman Mine and Hope village. Beds high
in the Mytton Member were recorded at various points in the vicinity of an old quarry in dolerite,
and Whittard’s notes for his Loc. 938 state specifically that he had ‘not definitely noticed any
gradation in softness between Mytton Flags and Hope Shales’.

Between the two principal outcrops of Mytton Member in the Shelve district lies a small,
triangular outcrop, bounded to east and west by faults which diverge northwards at an acute
angle (Fig. 14). The area is of some interest as it contains numerous exposures and the remains of
several workings including Roman Mine which, as the name suggests, was one of the original
workings and dates from the second century (Dines 1958). A few typically Mytton fossils such
as Monobolina plumbea and Ogygiocaris selwynii were found at Loc. 711 in micaceous, flaggy,
blue-hearted shales which weather light grey. Beds in the highest part of the succession and
almost immediately SE of the mapped upper boundary of the Mytton Member were seen at
Loc. 927, where they include rocks said to resemble closely the “Tankerville Flags’ of the Stiper-
stones district, though the latter strata are generally regarded as not being developed in the
Shelve district. Similar beds were seen also at Loc. 931. The rocks of the Mytton Member in this
area were said to differ from the adjacent Hope Member in several respects: they are harder and
more resistant, the contained mica flakes are usually larger, the bedding planes show flatter
surfaces, and the edges of flaggy beds are angular and sharper. Spoil heaps (Loc. 934) of material
taken from East Roman Gravels Mines, although composed of blue-black flags from the Mytton
Member, may cause confusion as they are located on the mapped outcrop of Hope Member.
Loc. 929, a short distance to the east, is of interest as it exposes strata of a lithology resembling
that at Shelve Church, though no comparable graptoloids have been found. Whittard (1966 : 303)
noted the occurrence of this apparently impersistent facies, which he considered to be probably
older than that at Shelve, and advocated the rejection of Shelve Church Beds as a stratigraphical
subdivision.

Hope Member

First usAGE. As Hope Shales or Hope Shale Group by Lapworth & Watts (1894 : 316, 317),
underlain by Mytton Group and overlain by Stapeley Ashes, a usage approximating apparently
to that in the present paper. Lapworth’s (1916 : 36) subdivision of the rocks into Lower, Middle
and Upper Hope Shales did not pass into general usage.

SHELVE INLIER 25

Hope Loc
Rectory p %41 7 ™~

=

Fig. 15 Geological map of the Hope Valley. Overlaps Figs 14, 16. 5=Mytton Member; 6=Hope
Member, including ‘chinastone’ tuff beds (dotted); 7=Stapeley Volcanic Member; 8 =Stapeley
Shale Member; 20=Silurian rocks; black outcrop D=dolerite. (SJ 30)

TYPE LOCALITY. Hope Valley, near Minsterley (Whittard 1960: 156), where according to an
earlier paper ‘practically the whole thickness can be examined’ (Whittard 1931 : 327).

DESCRIPTIVE NOTES. Whittard (1931 : 327, 328) described the rocks as comprising nearly 800 ft
(244 m) of blue-black, rusty-weathering shales containing bands of so-called “‘Chinastone Ash’,
volcanic horizons noted later by him (1960: 156) as ‘occasional bands of andesitic tuffs’. The rocks
were said to be recognizable only in the Shelve Inlier and were correlated with the Didymograptus
bifidus Zone of the Llanvirn Series.

Outcrops of the Hope Member are the most extensive of all the subdivisions of the Shelve
Ordovician succession, but the rocks weather regressively so that good exposures are not common
and the beds often underlie areas of boggy, poorly-drained ground. Their role in the mineraliza-
tion of the Inlier was noted earlier. The most important outcrops are those of the type area in
the Hope Valley (Fig. 15) where the succession includes bands of andesitic tuffs at several
levels. Blyth (1938: 397) pointed out that ‘dust-tuff’ is a more appropriate term for these
extremely fine-grained tuffaceous horizons, which have proved locally useful in demonstrating
the form of the Shelve Anticline (Whittard 193la : 340). The subject had apparently not been
considered in detail by Whittard at the time of his death, and for the present account it is
preferred to retain, in spite of its admitted imperfections, the name ‘chinastone’, which is
strongly entrenched in accounts of west Salop geology.

The Mytton/Hope Member boundary near the SE side of Hope Valley was noted earlier and
exposures in this area are few, but at Loc. 914 the southern extremity of a thin bed of shattered
‘chinastone’ was seen to be interbedded with blue, micaceous shales. A section in blue-black
shales stratigraphically below this tuff horizon is exposed in the bed of the stream flowing

26 W. F. WHITTARD

it eee e
ego a Bentlawnt

ine 909
Ai Le Fath / |

¢

owy + a
of Loc 9083 6
7

Fig. 16 Geological map of the area between Hope and Overton’s Rough. Overlaps Figs 14, 15, 17,
27. 5=Mytton Member; 6=Hope Member, including “chinastone’ tuff beds (dotted); 7 =Stapeley
Volcanic Member (7a=interbedded shales); 8=Stapeley Shale Member; 9=Weston Member;
black outcrop D=dolerite. (ST 30)

through Hope Valley c. 1500 ft (457 m) SSE of Hope Rectory, where Loc. 142 yielded Pricy-
clopyge and pendent didymograptids (Fig. 15). Farther upstream, shales above the same ‘china-
stone’ are seen in a fairly long section which extends through most of the beds up to the next,
and best-known, tuff horizon; fossils, including Ectillaenus hughesi (Hicks), Pricyclopyge binodosa
and Placoparia, were found at Locs 834E, G, H and I.

The ‘chinastones’ have been found with certainty only in the vicinity of Hope hamlet, where
there are two major bands (and also smaller, unimportant ones) around the NE-plunging nose of
the Shelve Anticline. The first, and higher, band occurs as the faulted backbone of Oakedge
(Figs 15-16) and continues down and across the valley to the disused roadside quarry (Loc. 941)
adjacent to Ash Cottage and sited 450 ft (137 m) ESE of Hope Rectory. This is the often-visited,
so-called ‘Contorted Ash Quarry’ (see Whittard 1931 : 328) in which the rocks are predominantly
bluish-grey, exceptionally fine-grained tuffs with occasional subsidiary bands of shale. From the
quarry, the outcrop extends northwards where it is first faulted and eventually ends against the
outcrop of unconformable Silurian rocks. The second, lower band is thinner than the first and

SHELVE INLIER 27

can be traced in the ridge NW of Bank Farm, into Hope Valley and thence up the northern flank
where it makes a pronounced feature which apparently underlies unconformable Silurian strata
forming an outlier in the vicinity of the 900 ft (274m) contour. Both bands of ‘chinastone’
appear to wedge out to the SW and the upper one also thins to the NE. The large number of
exposures of the ‘chinastones’ clearly indicates that the angle and direction of dip of these rocks
are so variable that readings are of only general significance. The quarry near Ash Cottage gives
a good impression of the structural condition of the ‘chinastones’, and their contortions and
fracturings are probably a reflection of the structural complication of the shales which contain
them. Thin-bedded competent rocks such as these behave as incompetent strata when contained
in easily deformed beds such as shales, and the angles and direction of dips in shales of the Hope
Member are unquestionably complex. It is thought that the latter rocks behaved as plastic
material during the folding of the Shelve Anticline, but the deformation is better indicated by the
‘chinastones’ as their readings are less prone to modification by hill-creep than are the shales.

Successively higher horizons in shales overlying the upper ‘chinastone’ are encountered
upstream along Hope Brook and the tributary stream running into it from the WSW (Fig. 15,
p. 25). Several exposures, including Locs 834L, M and N, are recorded, yielding occasional
trilobites (Barrandia) and graptolites, and displaying great variation of dip and strike. Down-
stream from Hope one encounters the succession of shales and ‘chinastones’ forming part of the
southeastern limb of the Shelve Anticline. Fosiliferous shales in Hope Brook at Loc. 86 and the
unconformable Ordovician/Silurian junction at Hope Quarry (Loc. 82) were noted by Whittard
(1958 : 11).

From Hope Valley the outcrop of the Hope Member extends northwards in a broad strip
until it terminates beneath Recent deposits just north of the Brockton — Minsterley road (B.4499).
The outcrop (Fig. 17) is cut by conspicuous dolerite intrusions which are irregularly elongated
ENE. Exposures are generally uncommon but both the igneous rocks and shales with occasional
flags are seen in a stream valley which runs north almost along the centre of the outcrop, where
Locs 57-60 yielded trilobites and graptolites.

South-west from Hope Valley, a large outcrop is mapped along the NW flank of Shelve Hill,
forming part of the western limb of the Shelve anticline and flooring the poorly-drained areas of
Black Marsh and The Marsh. Along the SE side of the Mytton outcrop in the Shelve area, a
narrower outcrop of Hope Member forms part of the eastern limb of the anticline, and débris
from an old mining level (Loc. 680 — see Fig. 13, p. 22) 300 ft (90 m) north of Shelve Pool yielded
a few fossils including ///aenopsis and Pricyclopyge. The two limbs of the anticline converge to
form a single, wide outcrop near the well-known eminence of Corndon Hill, the geology of which
was described by Blyth (1944 : 178; pl. 20) whose boundaries for the dolerite phacolith were
incorporated by Whittard in the present geological map.

Beyond Corndon, several minor intrusions (Map) of dolerite occur as well as the larger mass
of the Cwm Mawr picrite (Fig. 19) described in detail by Blyth (1944: 187). Apart from the
igneous rocks, the most notable place-name in this area is Brithdir farm (Locs 635, 640) where
shales of the Hope Member at one time yielded numerous fossils, some of which were illustrated
in Whittard’s trilobite monograph. The outcrop of the Hope Member terminates at the Silurian
unconformity midway between Church Stoke and Snead, but shales were seen in the stream
sections at and SW of Llanerch (Locs 603, 605 and 609; Fig. 19) and a few typical fossils were
found.

About a mile (1-6 km) ESE of Llanerch lie the southernmost outcrops of the Hope Member
in the eastern limb of the Ritton Castle syncline. The rocks, generally poorly exposed, are
intruded by the elongated doleritic mass of Disgwylfa (now Squilver) Hill, described by Blyth
(1944 : 171) and termed originally the Pitcholds intrusion, after a farm sited farther east and on the
outcrop of Mytton Member. Disturbed shales in the highest part of the Member were seen just
west of The Hollies farm (Fig. 20) but from there, NE along the strike, the strata are little in
evidence, though an old trial shaft in the vicinity of the River West Onny produced a tip-heap
(Loc. 254, Fig. 21, p. 32) of poorly fossiliferous, slightly hardened shales.

Beyond the river a narrow strip of Hope Member extends NE alongside the feature formed
by the Stiperstones and Mytton Members as far as the south side of The Bog, where the outcrop

28 W. F. WHITTARD

Af Leigh, 0C , Loc 68

eB

5
High|\ Plantation

pe Nee

Knotmoor ! i
Plantation ; 4
Was

Luekley! + —QOverton’s
i i ; Rough
Bary 9 [7 } 7a >

i

Fig. 17 Geological map of the area between Leigh Manor and Leigh Hall. Overlaps Figs 15, 16, 17.
6=Hope Member; 7=Stapeley Volcanic Member (7a=interbedded shales); 8=Stapeley Shale
Member; 9=Weston Member; 20=Silurian rocks; black outcrop D=dolerite. (SJ 30)

suddenly widens and is bounded on its SW side by a fault which cuts out the Stapeley Volcanic
and Shale Members. Most of the vicinity of The Bog lacks exposures but a stream flowing west
and SW near Ritton Farm and Ritton Castle (Fig. 22, p. 33) shows several exposures of shales,
flags and occasional tuffs finally mapped by Whittard as Hope Member, though his notes at one
time attributed them to the Stapeley Volcanic Member.

The outcrop around The Bog and, immediately to the north, Pennerley (Map) is shown as
being cut medially by an almost north-south fault which then curves slightly NNE coincident
with the elongated dolerite intrusion of Buxton Hill (Fig. 7, p. 13) and is intersected by another
fault extending from Shelve hamlet NE to Maddox’s Coppice (Map and Fig. 11). North-east from
Buxton Hill an elongated outcrop has been mapped along the eastern limb of the Ritton Castle
syncline, but the junction with the subjacent Mytton Member is not well exposed, though out-
crops of the latter are relatively abundant to the east. A quarter of a mile (0-4 km) west of Crows-
nest, a stream runs north through Josey’s Wood (see Fig. 10, p. 18) and eventually joins Hope
Valley Brook. Just SE of the point where the stream cuts the unconformable Silurian rocks
(Map), massive, blue-black shales of the Hope Member exposed in the stream valley produced
graptolites and trilobites (mostly Pricyclopyge binodosa (Salter) but occasionally Corrugatagnostus
and Stapeleyella), a fauna considered at one time by Whittard (1931 : 327) to be especially
characteristic of what were then termed ‘Lower Hope Shales’, a term now superseded.

Little of note was recorded from the remaining outcrop, which extends for about another two
miles (3-2 km) before it disappears beneath Recent deposits at a point half a mile (0-8 km) east of
Minsterley. The rocks are much obscured by glacial deposits, which attain a thickness of 20 ft
(6m) in a stream section 2000 ft (600m) NW of Maddox’s Coppice, but reappear along the
line of the disused Snailbeach railway as it nears Callow Hill (Fig. 12, p. 21), a locality noted

SHELVE INLIER 29

St a i ER ae OU eet
5am -

aa ‘ = ee eae

a wnt on

Fig. 18 Corndon Hill, 3 miles (4:8 km) ESE of Chirbury, formed by a resistant dolerite phacolith
intruded into regressive Hope (Shale) Member. View looking SW from Squilver.

earlier in connection with rocks of the Mytton Member quarried close by. Doleritic intrusions
cut the Mytton rocks just south of the quarry but most of them form a complex area within the
outcrop of the Hope Member.

Stapeley Volcanic Member

First USAGE. By Lapworth (1887 : 662) as Stapeley Volcanic Group, the topmost of three sub-
divisions of his Stiper Group (Fig. 3, p. 6). In this original usage the term corresponded apparent-
ly to the combined Stapeley Volcanic and Stapeley Shale Members of the present paper. The
separation and use of the terms Stapeley Volcanics and Stapeley Shales dates from Whittard
(1931 : 329), but the latter strata had been differentiated earlier, as Upper Stapeley Shales (now
superseded), by Lapworth (1916: 37). This last work recognized the existence within what is
now the Stapeley Volcanic Member of a sequence of three volcanic horizons, separated by
‘Lower Shales’ and “Middle Shales’. For these volcanics the names Lower (Hyssington) Ash,
Middle (Tashkar) Ash and Upper (Todleth) Ash were introduced, but they did not pass into
general use.

TYPE LOCALITY. Stapeley Hill, near Minsterley, according to Whittard (1960 : 254).

DESCRIPTIVE NOTES. Rocks of the Stapeley Volcanic Member crop out along the western limb of
the Shelve Anticline and are even more extensively developed in both limbs of the Ritton Castle
Syncline. The former area includes the type area of Stapeley Hill, 14 mile (2 km) SE of Rorrington,
where a single group of shales separates a lower and an upper set of volcanics, each of which is
intruded by a dolerite sill (Map). The succession there dips approximately NW at from 54° to 66°
and much of the ground is not well exposed, but mapping of even small faults is facilitated by the
features formed by the resistant volcanic rocks.

In the area SW of Stapeley Hill the shales and upper set of volcanics are intruded by the NE
extension of the large mass of andesite which forms the greater part of the hill Lan Fawr, three-
quarters of a mile (1-2 km) west of Corndon Hill. An additional much faulted area occupied by

30 W. F. WHITTARD

330 31

>)
Brithdir

la
pS

~S

i

mil

Fig. 19 Geological map of the area west of Hyssington and south of Corndon Hill. Overlaps Fig. 20.
6—Hope Member; 7=Stapeley Volcanic Member (7a=interbedded shales); 20—Silurian rocks;
black outcrop D=dolerite; vertically shaded outcrops =Cwm-Mawr picrite. (SO 29, SO 39)

the Stapeley Volcanic Member extends south beyond Lan Fawr and another, slightly larger one
runs south along the eastern flanks of the intrusive andesite masses of the hill called Roundton
(sometimes written as The Roundtain) and Todleth Hill.

Faulting at the NE end of Stapeley Hill cuts out most of the Stapeley Volcanic Member until
one reaches the vicinity of Bromlow Callow and The Park, about a mile (1-6 km) east of Meadow-
town (Figs 16, 27, p. 40), where lithic tuffs (including Locs 908-9) are folded into a syncline,
truncated to the north by a sigmoidal east—west fault which skirts the south side of Luckley Hill.
From there the rocks of the Stapeley Volcanic Member form a strip-like outcrop, ‘stepped’ at
intervals by small faults, which extends (Fig. 17, p. 28) to the northern boundary of the Inlier just
north of Leigh Hall (not to be confused with Leigh Manor, a building situated 2620 ft (799 m)
at 329° true from the Rectory at Hope). Whittard’s map shows the succession within the Volcanic
Member north of Luckley Hill to be similar to that of Stapeley Hill, that is with a single shale
horizon separating two sets of volcanics. On the other hand at Luckley Hill itself, a fault-bounded
area shows apparently a more complex sequence of two shales separating three volcanic horizons,
the topmost one of which contains a further subsidiary bed of shale. The apparent incoming of

SHELVE INLIER 31
32 94

31

oOo)
N

eli |

j

Fremes ?

45
7

Mex

Hyssington’

}

at @

|

-------"-

J

te
ce
UN

Lie

oder —-
aes -
“=

Yo EO ay Sat
~S

-<---"
=---"

¢

Am mm
a
—_————_

He
i

-<-==
-
-
-

-O---

ere--,
---"-" == -e==
=
--"
=

93 31 32
Fig. 20 Geological map of the area SE of Hyssington, in the SE part of the Shelve Inlier. Overlaps
Fig. 19. 6=Hope Member; 7=Stapeley Volcanic Member (7a=interbedded shales); 8 =Stapeley
Shale Member; 20=Silurian rocks; black outcrop D=dolerite. Note particularly also the large

dolerite of Disgwylfa (Squilver) Hill, shaded diagonally, and dips indicating the Ritton Castle
Syncline. (SO 39)

another volcanic horizon coincides with a southerly widening of the mapped outcrop of the
Volcanic Member. In the roadside section just east of Leigh Hall (Locs 67, 68) ‘ashes’ in the
lower part of the Volcanic Member alternate rapidly with rusty-weathering, occasionally fossili-
ferous, blue-black shales (Fig. 17, p. 28). The evidence of the rocks indicates that the vulcanicity
was submarine, probably in short, sudden bursts separated by periods of quiescence whose
duration exceeded that of the vulcanicity, so that a benthic fauna could become established.

In the vicinity of Knotmoor Plantation, three-quarters of a mile (1-2 km) NE of Luckley Hill,
is a fault-bounded series of outcrops of Stapeley Volcanic Member which according to Whittard’s
mapping probably include a sequence of at least two volcanic and two shale horizons, the
possibility of any higher horizon having been eliminated by faulting (Map). An account by
Blyth (1938), based on mapping and excavation of the section NW of Leigh Manor (Figs 16-17),
gave an estimated thickness of 417 ft (127 m), made up of two pyroclastic divisions and an
intervening group of shales with thin tuffs. Whittard’s map shows a similar succession, but the
upper pyroclastic rocks form a feature which is succeeded to the NW by a hollow from which
the presence of, probably, another shale horizon is inferred. The relationships of the rocks in
this complex area are not clear and the above interpretation must be regarded as tentative.

In the SE of the Shelve Inlier, the Stapeley Volcanic Member is encountered first in Roveries
Wood, half a mile (0-8 km) NE of Snead (Fig. 20, above) where it forms an elongated, partly
fault-bounded outcrop immediately adjacent to the well-known dolerite intrusion which extends
NE to form Squilver Hill. The principal area of volcanics commences a short distance to the NW
and can be traced to the NE as two parallel series of outcrops, separated by Stapeley Shale

32 W. F. WHITTARD

1
1
I NiG0
/ i i 1 if
7H / i De i
t
\4 BL AITITT al
ay OF NR
i i) H!
U i 4

U

U

U
U

Fig. 21 Geological map of the valley of the River West Onny near Tasgar and Nind. Adjoins
Fig. 22. Note dips showing the form of the Ritton Castle Syncline. 3=Shineton Shales; 4=
Stiperstones Member; 5=Mytton Member; 6=Hope Member; 7=Stapeley Volcanic Member
(7a=interbedded shales); 8=Stapeley Shale Member; black outcrop D=dolerite. (SO 39)

Member forming the core of the Ritton Castle Syncline. The dips of the volcanic rocks show
clearly the form of the syncline, but the interbedded shales are less well exposed. Resistant features
formed by the volcanics have aided the mapping of numerous faults, including north-south tears
and complementary shears which are responsible for a gap in the continuity of the outcrop of the
Stapeley Shale Member SE of Hyssington. In this area (Fig. 20) the SE-dipping succession
comprises three volcanic horizons separated by shales. Several exposures of the lowest volcanics
occur in the immediate vicinity of Hyssington, and rocks of the middle horizon were quarried
(Loc. 251) at the south end of Fremes Wood, a quarter of a mile (0-4 km) east of the village,
where both tuffs and interbedded shales were seen. Still farther NNE a similar sequence of
three volcanic horizons was mapped beyond the valley of the River West Onny at Runnis Bridge,
an area in which the most noteworthy locality (Fig. 21) is Tasgar (sometimes Tasker or, rarely,
Tashkar) Quarry (Loc. 270), named for the nearby farm. The quarry, once well-known as a fossil
locality but now practically worked out, is cut in massive tuffs of the Volcanic Member and the
fossils came from subsidiary, interbedded shales.

At the southern end of the eastern limb of the Ritton Castle syncline a sequence of three
volcanic horizons and two intervening sets of shales is not readily apparent at first owing to
faulting near the north end of Squilver Hill. The sequence is demonstrable along the hill named
Cefn Gunthly (or Cefn Gwynlle), ‘stepped’ at intervals by small transverse faults, and continues
across the valley of the River West Onny. A third set of shales mapped at the top of the sequence
beside the north end of Cefn Gunthly apparently was regarded as distinct from the Stapeley
Shale Member, the outcrop of which is separated by a fault immediately to the west.

Just north of the West Onny, in Nind Wood and some 400 ft (122 m) SSE of Nind (sometimes

SHELVE INLIER

33

34 35
/ U 1 ' ‘x
6 ’ ‘ 1 ' ‘
A 4 i / u 20 \
hh nod ’ 8 , ! 6
/ 7 if t So
n 1 ! J
aN! 6 ay
i! X (2 Can
/ ‘ $ iv |
o~ r 4
eau 8 Ritton Gases /20/
/ Farm £20) ] \U7
/ Wo Com
' 29/ 98
> mer ot
lo
8 ) P
7 a =
ye 500) } \20\
/ Ritton - ey?
<Castle’. Bae THE
Oe o BOG
4 wv ¢
7 NSS i
l,” / a
Aa Ae A
be 7 KE Ki
Pant “ a me 6 |
a > Ze Ke 6
4 7 V. v2
4B ve We y
: Né ya ;
63s Je Teg
/ 7
ea fe is an
’ ¢
/ 4) i ole :
* Loc ‘ yy 7h Brookshill oF
7 G hf 4
ee Nee 7 Marsh
‘ / of 4
/ A Bs & pa Ta
Xs a 7 x je
Brooks ,~ ¥ oa 6
¢ . ¢ i
ta Hill / v a
me a He J 5
lee 7 int ae le
662," / a ii / 6 a
o / ae Ae == 5 Al
34 35

Fig. 22 Geological map of the synclinal area around Ritton Castle and The Bog. Adjoins Fig. 21.
Note the small Silurian outliers. 4=Stiperstones Member; 5 = Mytton Member ; 6=Hope Member;
7=Stapeley Volcanic Member (7a=interbedded shales); 8=Stapeley Shale Member; 20=

Silurian rocks. (SO 39)

Nynd or Neint) Farm, is Nind Quarry (Fig. 21, Loc. 263). The section, in the highest of the three
volcanic horizons, shows tuffs which sometimes form large, nodular masses, weathering to a
gingerbread colour, and may possibly have been calcareous. Interbedded shales at one time
yielded numerous fossils, but collecting is not now profitable.

The volcanics of the Nind area are truncated a short distance to the east by one of the Inlier’s
conspicuous north-south tear-faults which runs between Cefn Gunthly and Black Radley Hill
and skirts the western margin of Heath Mynd. Beyond the fault-line, at and near Brooks Hill,
three-quarters of a mile (1-2 km) NE of Nind Quarry, Whittard’s mapping (Fig. 22) shows
the outcrop of the Stapeley Volcanic Member to be decidedly broader, with four volcanic
horizons separated by shales. Even this appears to be a somewhat generalized picture and the
notes on Loc. 660, 530 ft (162 m) at 348° true from the summit of Brooks Hill, emphasize the
difficulty of distinguishing a detailed sequence of shale bands and tuffs in badly-exposed terrain.
At this point the dip-slope of the Volcanic Member, instead of presenting a uniform appearance,
exhibits two narrow ‘benches’ which can only be assumed to mark shaly horizons, though the

34 W. F. WHITTARD

ae vé fof f Tei 8
¢ fy a

Priestuetop ®. ‘Loc a
LN ta pa

/

70

1 Aldress |

D if
' i /

\

‘

\ Loc 952
by,

i
Cwmdwla

=

Fig. 23 Geological map of the Priestweston area and the andesitic intrusion of Lan Fawr. Adjoins
Fig. 24; overlaps Figs 30, 32. 7=Stapeley Volcanic Member (7a=intertedded shales); 8 =Stapeley
Shale Member; 9=Weston Member, including ‘grit’ horizons (dotted); 10—Betton Member;
11=Meadowtown Member; 12—Rorrington Member; 13=Spy Wood Member; 14=Aldress
Member; black outcrop D=dolerite; main black outcrop=andesite. (SO 29)

hillside is bracken-covered and there is little evidence of rock except for widely dispersed blocks of
tuff. The observation is recorded to emphasize the importance within the volcanics of shale
bands which may range in thickness from several yards (metres) to 1—2 in (2:-5—5 cm) in extreme
cases. The note goes on to speculate that it may eventually prove necessary not to separate (what
is now) the Stapeley Shale Member from the Volcanic Member, but no such action was ever
taken by Whittard. Loc. 662, on the SW side of Brooks Hill and 1400 ft (427 m) at 217° true
from the summit, is of particular interest as Whittard described the low, cliff-like section there as
showing interbedded, coarse-grained lithic tuffs, crystal tuffs and ashy shale, varieties of rock
which recalled the beds near Leigh Manor described in 1938 by Blyth (see p. 31). The wide
outcrop of the Volcanic Member extends NE of Brooks Hill for another half mile (0-8 km) and
is then truncated by a prominent NNW-SSE fault.

The Stapeley Volcanic Member in the western limb of the Ritton Castle syncline extends NE
to a point half a mile (0-8 km) NE of Shelve hamlet. The outcrop is truncated by faulting but in

SHELVE INLIER 35

addition the sequence is much attenuated and only a single volcanic horizon was mapped there,
overlain by the Stapeley Shale Member. After a gap of 1460 yd (1:34 km) along the strike, the
outcrop reappears in a small area NE of Lower Santley, SE of Hope. There also a single volcanic
horizon was mapped, succeeded apparently by the Stapeley Shale Member (though poorly
exposed and its outcrop largely inferred), the two subdivisions together being in the form of a
small syncline, unconformably overlain by the Lower Silurian strata of the Venusbank district
(Fig. 15, p. 25). In this area the map of Lapworth & Watts (in Pocock & Whitehead 1948 : pl. 11)
shows a very large outcrop of Stapeley Volcanics on ground that is now known to be occupied
largely by Silurian rocks.

Stapeley Shale Member

First usAGe. As Upper Stapeley Shales, by Lapworth (1916: 37). Subsequently modified to
Stapeley Shales by Whittard (1931 : 323) and so used by him in later publications.

TYPE SECTION. Chosen by Whittard (1960: 254), who described the beds as usually unfossili-
ferous; Holywell Brook, NW of Stapeley Hill and 1-75 miles (2-4 km) NNW of Shelve hamlet.
Earlier Whittard (1952: 159) noted an upward gradation from the Stapeley Shales into the
Weston Beds at Holywell Burn (—Brook), so that an arbitrary boundary between the two had
to be chosen.

DESCRIPTIVE NOTES. The strata of the Member are typically ‘bluish-grey, rusty-weathering, soft
shales’ (Whittard 1931 : 330) and crop out within two elongated areas which form the western
limb of the Shelve Anticline and the core of the Ritton Castle Syncline. Good continuous sections
through the Shale Member are rare and even Holywell Brook, the type section, exhibits only the
upper portion clearly, although the underlying Stapeley Volcanic Member crops out not far
away (Fig. 25, p. 37). At this point Holywell Brook arises on the NW side of the north end of
Stapeley Hill in an area of swampy ground with no solid outcrop, and it is not until the stream
meets the west side of Valley Knoll that exposures of the Stapeley Shale Member are encountered.
Here and in the stream section separating Valley Knoll from Rorrington Hill are seen strata
interpreted as the highest of the Shale Member, dipping NNW at 50°-60° and succeeded by
Weston Member. Whittard’s arbitrary selection of the boundary has already been noted and was
taken near the western end of an outcrop (Loc. 462) elongated east-west beside the northern
bank of Holywell Brook, just west of its confluence with the tributary stream flowing from the
SSE (Fig. 25, p. 37). At the western end of this outcrop were seen flaggy beds that could be
assigned without much hesitation to the succeeding Weston Member, but farther east the rocks
are more shaly with bands up to 2 in (5 cm) thick of micaceous shale resembling the typical
lithology of the Stapeley Shale Member. The latter also show an anomalous increase of dip which
is thought to be due to more than landslipping, and an almost continuous section (Loc. 463) along
part of Holywell Brook farther east also exhibits variations in dip and strike. It is possible that
such discrepancies may be the result of a fault running through the badly-exposed area of Rorring-
ton Covert, and Whittard’s notes show that he was considering an alternative interpretation
which is shown in Fig. 26, p. 37. The lithological transition from Stapeley Shale Member to
Weston Member is reflected to some degree in the type of fauna present; ogyginid trilobites are
present in both subdivisions but the Weston Member contains a bivalve and lingulid brachiopod
fauna which is almost unknown in the underlying Shale Member.

Some 1-62 miles (2:6 km) SSW of Holywell Burn, a road section (Loc. 767) NE of Priestweston
(Fig. 23) cuts through part of the highest third of the Member and is of interest as it demon-
strates the gradual appearance of more flagstone beds until the arbitrary base of the Weston
Member is reached, in this instance at a point closer to Priestweston.

In the area between Priestweston and Holywell Brook the boundary of the Stapeley Volcanic
and Shale Members runs NE along the edge of Stapeley Hill but is not generally visible. Of
particular interest, therefore, is a section in the lower portion of the Shale Member at an old
mining adit (Loc. 453) situated 2400 ft (732 m), 282° true from Stapeley Farm (Fig. 24). The beds
of the Shale Member here are rusty-weathering, dark, bluish-grey shales, with occasional
spheroidal concretions, and an apparently unfaulted thickness of 198 ft (60-35 m) was measured

36 W. F. WHITTARD
30 31

N
| “ }' Loc
FoR 4 453 + y
Lean 1
ye Midd leto n y /
/ Hill
7

Fig. 24 Geological map of the area NE of Priestweston. Adjoins Fig. 23; overlaps Figs 25, 29.
6=Hope Member; 7=Stapeley Volcanic Member (7a=interbedded shales); 8=Stapeley Shale
Member; 9=Weston Member, including ‘grit’ horizons (dotted); 10—Betton Member; 11=
Meadowtown Member; 12=Rorrington Member; 13=Spy Wood Member; black outcrop D=
dolerite; black outcrop A=andesite of Lan Fawr. (SO 29, SO 39)

along the adit. A normal contact with the underlying Volcanic Member was demonstrated by a
junction bed 5 ft (1-52 m) thick, comprising an alternation of shales and tuffs and indicating a
gradual passage between the two members.

Outcrops of the Stapeley Shale Member extend intermittently southwards from Priestweston
for 2 miles (3:2 km) until they are terminated by glacial deposits just south of the village of
Hurdley. An elongated hollow between Todleth Hill and Roundton (Map) has been mapped in
part as a NW extension of the outcrops of Shale Member which run along the eastern margin of the
intrusive masses of andesite. The rocks are exposed at various points along the northerly-trending
stream NW of Hurdley and graptolites of ‘tuning-fork’ type were collected at Loc. 624. Exposures
are poor along the eastern margin of Todleth Hill, but shales belonging to the Member were
found in an excavation for a water-pipe near the north end of the hill.

Still forming part of the western limb of the Shelve Anticline, the outcrop of the Stapeley
Shale Member extends NE for about 23 miles (4 km) before ending beneath the unconformable
cover of Lower Silurian strata which marks the northern boundary of the Shelve Inlier in the
vicinity of Leigh Hall, a mile (0-6 km) south of Worthen (Fig. 17, p. 28). The outcrops in this
structurally more complex area are discontinuous and the rocks are not well exposed.

In the eastern part of the Shelve Inlier, rocks of the Stapeley Shale Member form an elongated
outcrop, often broken by faults, which extends ENE along the axis of the Ritton Castle Syncline
from a point 0-62 mile (1 km) north of Snead (Map) for a distance of 44 miles (7-24 km) to Round
Hill, near Pennerley. The rocks are indifferently exposed and no noteworthy sections or localities
are recorded. A further outcrop is largely inferred to the NW of Lower Santley, 0-42 mile (0:67 km)

SHELVE INLIER 37

Loc 463 _
\-==
-= eo

Valley Knoll

White a

J House

Fig. 25 Geological map of the area around Holywell Brook, SE of Rorrington. Overlaps Figs
24, 27, 28, 31. 6=Hope Member; 7=Stapeley Volcanic Member (7a=interbedded shales);
8=Stapeley Shale Member; 9=Weston Member, including ‘grit’ horizons (dotted); 10—Betton
Member; 11—=Meadowtown Member; 12—Rorrington Member; 13=Spy Wood Member;
black outcrop D=dolerite. (SO 39, SJ 30)

300

5)

Fig. 26 Alternative interpretation of the area around Holywell Brook; beds as in Fig. 25. Heavy
mineralization of rocks at Loc. 592 suggests the proximity of a fault. (SO 39, SJ 30)

38 W. F. WHITTARD

SSW of Venusbank (Fig. 15, p. 25). The shales there are underlain by folded and faulted tuffs of
the Stapeley Volcanic Member and overlain by unconformable strata of Llandovery age.

Middleton Formation

First USAGE. As Middleton or Llandeilo Series, by Lapworth & Watts (1910 : 752, 753), underlain
by Shelve Series and overlain by Chirbury Series. Made up of successive subdivisions of Weston
Flags and Shales (or Stage), Betton or Didymograptus murchisoni Shales, Meadowtown Calcareous
Beds (or Stage) and Rorrington Shales, the term replaced without comment the earlier-named
Meadowtown Series of Lapworth (1887 : 662).

TYPE LOCALITY. Presumably the vicinity of the hamlet of Middleton, 23 miles (4 km) due west of
Shelve, but no type section was designated.

Weston Member

First USAGE. As Weston Group, by Lapworth (1887 : 662), apparently in the same sense as now
employed.

TYPE LOCALITY. ‘Priestweston hamlet’ was designated by Whittard (1960: 281) in the Lexique
Stratigraphique International, where he described the constituent rocks as being ‘divided into a
lower and upper group of massive greenish-brown grits, separated by rusty-weathering, blue-
hearted shales’. This description almost exactly matched one given earlier by him (1931 : 330)
except that the ‘shales’ were listed as ‘flags’. In his notes Whittard referred to the old quarry at
Priestweston as being the type locality, but with reservations that are discussed later.

DESCRIPTIVE NOTES. The rock types within the Weston Member include siltstone, flagstone and
tuff, while shale is only rarely developed. One of the characteristics of the more argillaceous
strata is that they seldom present a surface sufficiently flat for measuring the dip. The almost total
absence of soft shales is a useful factor in distinguishing the rocks from those of the Stapeley Shale
Member and the smoothness of the flat, fresh surfaces of the latter beds finds no parallel in the
Weston Member. Consequently, even though the boundary of the two members is arbitrarily
taken, and there is a gradation through c. 20-30 ft (6-9 m) of rock, it is fairly easy to distinguish
those belonging to the Weston Member. The strata appear to represent ill-sorted, extremely
shallow water deposits which might even be near estuarine in origin. In places they were appar-
ently laid down under turbulent conditions, as shown by the remarkably rapid and irregular
alternations of wisps of argillaceous matter within siltstones. Furthermore, the rocks are not
generally fossiliferous, but in places they are richly so. Some fossil localities are in the more shaly
beds, others are less richly fossiliferous in thin flagstones, others include the occasional occur-
rences in massive beds, and yet others are associated with tuffaceous layers in which graptolites
are not infrequent. The latter may be taken to indicate other than near-estuarine conditions,
but it was not unusual for graptolites to enter shallow coastal waters and if they were epi-
planktonic no doubt detached masses were sometimes washed into littoral zones. What is more
significant is that whenever the beds are found to be fossiliferous, by far the most important
elements in the fauna are the bivalves, horny brachiopods and, less commonly, gastropods. Other
forms are trilobites, mainly Ogyginus corndensis (Murchison), a few nautiloid cephalopods, rare
crinoids and some articulate brachiopods. This fauna is not one typical of the shallow waters of
the Ordovician sea, and its association with ill-sorted sediments suggests estuarine conditions.

Within the whole outcrop of the Weston Member there are subsidiary escarpments, sometimes
pronounced. These scarps are not so much owing to the alternation of soft rocks with much
harder ones as to variation in the proportions of grits, flags and shales developed in an inter-
bedded series. Even in the less resistant divisions, alternation of rock-types persists. In some
cases the main resistant horizons can be mapped for a moderate distance across country, but in
others a single horizon may quickly be replaced by more than one. It is evident that the Weston
Member, on both the large and the small scale, is composed of constantly-changing, lenticular
strata, a view not inconsistent with deposition in a very shallow water environment.

SHELVE INLIER 39

Outcrops of the Weston Member are confined to the western limb of the Shelve Anticline,
where they occupy a strip of country up to about half a mile (0-8 km) wide which extends NNE
from a point three-quarters of a mile (1-2 km) ESE of Church Stoke, by way of Priestweston, to
terminate at the northern boundary of the Inlier between Betton and Leigh Hall (Map). The
most southerly section occurs in an old quarry (Loc. 406), situated near Hoarstone Farm and
4350 ft (1326 m) at 114° true from St Nicholas’s church, Church Stoke. In this area the rocks are
flaggy, with a high percentage of coarse-grained volcanic dust, and in places there are thin beds
of flinty, vitric tuffs and occasional intercalations of soft, friable tuffaceous material. The outcrop
narrows northwards between the andesitic mass of Todleth Hill, to the east, and the Quaternary
deposits of the Church Stoke area, but widens to the north of Old Church Stoke and underlies
the flat ground south and east of Upper Aldress (Map).

A little farther north is Cwm-Dwla (sometimes Cwmdulla) farm, just west of the SW end of
Lan Fawr hill (Fig. 23, p. 34), which is of importance as it provides one of only two places (the
other is Holywell Brook — see p. 35) where the passage from Stapeley Shale Member to Weston
Member can be studied. The section (Loc. 952), situated just behind and north of the farm, shows
the Shale Member comprising light-grey weathering, silky shales, dark grey when fresh and inter-
bedded with flags and tuffs which vary in thickness from ‘less than an inch (25 mm) up to 10 in
(250 mm). These harder beds appear to increase in importance higher in the succession and
eventually give place to the typical Weston Member lithology, with its appearance of argillaceous
matter whisked up with silty material. There is no indication of a sedimentary break and, as at
Holywell Brook, the succession is undoubtedly continuous. The area immediately north and
east of Cwm-Dwla is interpreted as part of a fault-bounded, angular outcrop which includes also
Stapeley Shale Member and is truncated to both north and south by transverse faults, the
northerly of which coincides with a dolerite intrusion elongated east-west.

Farther north again lies the hamlet of Priestweston, said to be the type locality for the Weston
Group and consequently presumed to be that for the Weston Beds (now Weston Member),
although in practice only a fraction of the succession is found there. Whittard noted that the
old quarry (Loc. 766) at Priestweston might be taken as the type-section but observed that in a
sense this would be misleading since it exposes no more than 90 ft (27:5 m) of rock and is excavated
in the hardest, flaggy beds, most suitable for local building. There is no indication of the less
resistant beds which go to make up so much of the succession elsewhere and which are not
infrequently fossiliferous, whereas the massive beds seldom provide fossil remains and probably
accumulated under conditions inimical to animals. For the most part the rocks are flaggy, highly
micaceous, sometimes ripple-marked, and interleaved with thin siltstones. On the eastern, dip
face the surface is literally covered with fucoid markings where the rock face is markedly
micaceous. The fucoids are compressed, unbranched, rod-like structures; they show no particular
alignment but may cross one another.

The rocks at Priestweston Quarry were mapped by Whittard as part of the upper of two ‘grit’
horizons which occur in approximately the lower half of the Member and can be traced NNE for
almost 2 miles (3-2 km) before apparently dying out in the vicinity of Rorrington Hill Covert,
near Holywell Brook (Fig. 25). The ‘grit’ of Priestweston Quarry crops out c. 1700 ft (518 m) NE
of the hamlet, in the sides (Loc. 764) of the subsidiary road running NE from the Methodist
chapel, and thence to and just beyond the small valley of Cwm Dingle. Farther NE the outcrop is
cut by small, transverse faults; it and the overlying strata are intruded by two east-west dolerite
dykes about 1800 ft (549 m) WSW of Middleton Hill. The lower ‘grit’ horizon crops out just SE
of Priestweston, in the vicinity of what was once the Miners Arms inn, and continues parallel to
the outcrop of the higher ‘grit’, though extending a little farther (Map), and forms a feature to
the SE of Rorrington Hill Covert. The argillaceous strata between the two ‘grit’ horizons, though
infrequently exposed, were seen in a pathside exposure (Loc. 769) 1850 ft (564 m) at 046° true
from Priestweston Quarry (Fig. 23, p. 34).

Between Rorrington Hill Covert and Rorrington Hill runs the valley of Holywell Brook, to
the north of which (Fig. 25) is a low, cliff-like outcrop (Loc. 591) formed by rocks in the lower
part of the Weston Member. The section is so variable that detailed measurements are not
practicable, but c. 40 ft (12:2 m) of rock are exposed in the cliff, composed of coarse, tuffaceous

40 W. F. WHITTARD

Fig. 27 Geological map of the area around Betton Hill and Luckley Hill, east of Meadowtown.
Adjoins Fig. 28; overlaps Figs 16, 17, 25. 6=Hope Member; 7=Stapeley Volcanic Member
(7a=interbedded shales); 8 =Stapeley Shale Member; 9 = Weston Member; 10=Betton Member;
11=Meadowtown Member; 20=Silurian rocks; black outcrop D=dolerite. (SJ 30)

flags, finer tuffaceous flags showing curvilinear banding, and knobbly-weathering silty shales.
It would seem that the massive beds of the Weston Member owe their massiveness to included
volcanic débris, without which the rocks would probably have been deposited as muds and silts.
Immediately west of the cliff, an old quarry adds c. 40 ft (12:2 m) to the thickness exposed there
and is stratigraphically beneath it. The quarry face is mainly built up of fine-grained massive
rocks in which small lenses of apparently more tuffaceous material weather more readily.

From Rorrington Hill the outcrop of the Weston Member narrows northeastwards as the
result of faulting in the area east of Meadowtown, and then widens again between Lyde and
Bromlow (Fig. 27) before being terminated by unconformable Lower Silurian strata between
Betton and Leigh Hall. The passage from Weston Member to Betton Member in the stream
section near Lyde Cottage (Fig. 27) is assumed to be normal and gradational though the actual
junction seems somewhat arbitrary. The angle formed by this stream and a subsidiary one at a
point 150 ft (45-7 m) SSE of Lyde Cottage exposes a section in brown-weathering, nodular strata
of the Weston Member. This exposure was particularly noted by Whittard (1958 : 13) as showing

SHELVE INLIER 41

shales which elsewhere are interbedded with massive flags and which here are unusually fossili-
ferous.

Betton Member

First USAGE. As Betton or Didymograptus Murchisoni Shales by Lapworth & Watts (1910 : 752,
753). The rocks were said to yield an abundance of D. murchisoni and Orthis, and to be best
exposed SE of Rorrington. They were later termed Didymograptus murchisoni Beds, equivalent
to a Betton Group or Stage, by both Lapworth (1916 : 36) and Watts (1925 : 340), but the name
Betton Beds was introduced by Whittard (1931 : 323) and retained by him thereafter.

TYPE SECTION. Betton Dingle, NNE of Meadowtown (Fig. 27), according to Whittard (1960 : 52).

DESCRIPTIVE NOTES. The beds were described by Whittard (1931 : 331; 1960 : 52) as micaceous,
blue-black shales and flags with a thickness of 600 ft (183 m), a figure notably larger than the
200 ft (61 m) give earlier by Watts (1925 : 340). Junctions with the underlying Weston Member
and overlying Meadowtown Member were said to be apparently gradational.

Lapworth & Watts’ geological map of the Shelve Inlier (in Pocock & Whitehead 1948 : pl. 11)
indicated the outcrop of the ‘Betton Shales and Flags’ as forming a narrow, unbroken, band-like
outcrop which extended SSW from the vicinity of Betton farm to reach the southern boundary
about three-quarters of a mile (1:2 km) SE of Church Stoke. Whittard’s mapping demonstrated
that the outcrop is, in fact, broken into three discrete portions as a result of the north-south tear-
faults which affect the Inlier.

The northernmost outcrops of Betton Member occur (Fig. 27) in the vicinity of Betton Dingle,
which runs NNE almost along the strike in the area SSE of Betton farm. Only here is the suc-
cession complete, with both upper and lower boundaries unaffected by faulting. A conspicuous
feature of the outcrops immediately south of the margin of the Inlier, here comprising unconform-
able Silurian strata, is the presence of narrow, subparallel dolerite intrusions, members of a group
which are elongated east-west in the area between Overton’s Rough and Lower Wood. The
predominantly strike section in Betton Dingle exhibits little variation in lithology. The general
rock type is a micaceous, rusty-weathering, blue-hearted shale occasionally interbedded with
more siliceous layers which pass into flags, not infrequently colour-banded blue-grey and light
grey. At other times volcanic detritus is obviously present, producing coarser-grained rocks.

From here the outcrop continues SSW until level with Meadowtown hamlet, where it is
obliquely truncated by a fault so that the outcrop narrows and finally disappears half a mile
(0-8 km) east of Rorrington. Exposures of Betton Member are recorded (Fig. 28) from the lane
(Loc. 577) 600 ft (183 m) SSE of Meadowtown Quarry, where pendent didymograptids were
found, as well as in the fields SW of this point. Loc. 505, 1300 ft (396 m) SW in the stream
skirting the north side of Rorrington Hill, shows an exposure in blue-black shales which yielded
ogyginid and trinucleid trilobites. Whittard noted that the section here (assuming a uniform dip)
contains a thickness of 440 ft (134m) of Betton Member, a figure comparable with that in
Holywell Brook; this is considerably less than the 600 ft (183 m) given elsewhere by Whittard
for the Member, but the succession here is incomplete owing to faulting.

The southern continuation of this outcrop is ‘stepped’ westwards by a fault and the rocks are
seen next in the small valleys cut by Holywell Brook and Whitehouse Brook, the stream running
north along the western margin of Rorrington Hill Covert (Fig. 25, p. 37). Immediately north of
(i.e. downstream from) the junction of the two is a section (Loc. 437) in shales, flaggy and
tuffaceous shales, and thin, 1-3 in (2-5—7-5 cm) tuffs, the last containing angular fragments and
abundant pyrite. The rocks yielded a varied fauna comprising Didymograptus of murchisoni type
and trilobites (Ogyginus, Bettonia and Trinucleus). The apparent thickness of 400 ft (122 m) is
again low because of faulting, as a result of which the next set of outcrops is found farther SW,
underlying a wedge-shaped area to the west and, especially, north of Priestweston (Map). The
area is not well exposed but at Loc. 307 (Fig. 29), in the stream-section 990 ft (302 m) WSW of
Little Weston, may be seen blue-hearted flaggy beds with some very thin, shaly partings; these
contain a fauna similar to but less abundant than that at Loc. 437.

42

W. F. WHITTARD

=
So
=

‘

1
4
1
t
4
6
‘
4
4
!
1

rs

Fig. 28 Geological map of the area NW of Meadowtown, including Lower Wood Brook. Adjoins
Fig. 27; overlaps Fig. 25. 9=Weston Member; 10=Betton Member; 11=Meadowtown Member;

12—Rorrington Member; 13=Spy Wood Member; 14—Aldress Member; 20=Silurian rocks;
black outcrop D=dolerite; black outcrop A=andestic intrusion of Lower Wood (SJ 30)

The sole remaining outcrops of Betton Member occur in two small, immediately adjacent and
largely fault-bounded areas centred about half a mile (0-8 km) north of Church Stoke, where the
southerly continuation is terminated by glacial deposits forming the southern boundary of the
Inlier. The original interpretation of this area was subject to some later revision by Whittard,
and the present map represents his final version.

The uppermost (i.e. easterly) reaches of Spy Wood Brook expose sandy, argillaceous, flaggy
beds with much volcanic dust and pyrite at Loc. 382 (Fig. 30). Fossils are not common but
include Ogyginus, trinucleids, fragments of inarticulate brachiopods and poorly-preserved
biserial graptolites. About 750 ft (231 m) farther south, the last outcrops of Betton Member are
seen in the small valley known locally and in Whittard’s Monograph and notes as Deadman’s
Dingle, which runs westwards into Spy Wood Brook. Several exposures, including Loc. 385, of

micaceous, flaggy shales occur, some of which contain a typical fauna of graptolites and trilo-

bites, while Loc. 384, in the lower part of the succession, shows c. 18 ft (5-5 m) of rusty-weathering

shales rich in volcanic detritus.

SHELVE INLIER 43

Fig. 29 Geological map of the area between Lower Ridge and Little Weston, including the upper
reaches of Coed Brook. Adjoins Figs 23, 34; overlaps Fig. 24. 9=Weston Member; 10=—Betton
Member; 11=Meadowtown Member; 12—Rorrington Member; 13=Spy Wood Member;
14=Aldress Member; black outcrop D=dolerite. (SO 29)

Meadowtown Member

First USAGE. As Meadowtown Calcareous Beds or Meadowtown Stage, by Lapworth & Watts
(1910 : 752, 753). Equivalent to the Meadowtown Group of Lapworth (1916 : 36), later renamed
Meadowtown Beds by Whittard (1931 : 323). Further subdivisions of the Meadowtown Group
listed by Lapworth (1916 : 36) have not survived. The term should not be confused with ‘Meadow-
town Series’ as used by Lapworth (1887 : 662), a much larger subdivision which is equivalent to
the present Middleton Formation.

TYPE LOCALITY. ‘Meadowtown hamlet’ according to Whittard (1960: 183), no specific type
section being designated.

DESCRIPTIVE NOTES. The rocks were described briefly by Whittard (1931 : 331; 1960: 183) as
comprising blue-black shales and flags with some developments oi limestones, calcareous flags
and bedded tuffs. A thickness of c. 1300 ft (396 m) was given and the contact with the underlying
Betton Member described as gradational. Whittard’s later lists (1966 : 299-307) of trilobites
from the Shelve Inlier showed the trinucleids Whittardaspis [Cryptolithus| inopinata (Whittard)
and Lloydolithus lloydii (Murchison), and the ogyginid Ogygiocarella debuchii (Brongniart), to
be especially abundant and characteristic.

The boundaries of the Member as mapped by Whittard at various periods demonstrate changes
of opinion but his final word on the subject (1966 : 298) defined the lithostratigraphic boundary
with the overlying Rorrington Member as occurring at the point where ‘bluish-grey shales with
occasional thin, bluish limestones and sometimes bedded tuffs’ (in which shelly faunas areabundant
and graptolites rare except in a few thin bands) pass upwards into ‘massive sooty-black graptolitic
mudstones and shales’, described as monotonously uniform. The change in fauna was said to be
‘much more obvious than the lithological one’, but there can be little doubt that one is dealing
with a true lithostratigraphic boundary. The thickness of the Meadowtown Member as so
interpreted was described by him as being much greater than he originally thought (i.e. in the
early parts of the same monograph), and this was reflected also in the modified version of his
field slips, on which the outcrop is considerably wider.

Although Meadowtown hamlet was given as type locality it is sited within the outcrop of the
lowest third of the Meadowtown Member and for practical purposes the section there is confined
to the vicinity of the classic but long-disused quarry, knowledge of which dates from the time of
Murchison. The section in Meadowtown Quarry (Loc. 160) was measured by Whittard when it

44 W. F. WHITTARD

—
Pree ae

=~
ie)
RO
hO

ca
1 Sic
!

—
ie)

Brynkin

—-—---—~_

a
"

a

Fig. 30 Geological map of the faulted area around Spy Wood, NNE of Church Stoke. Overlaps
Figs 23, 32. 9=Weston Member; 10=Betton Member; 11—Meadowtown Member; 12=
Rorrington Member; 13=Spy Wood Member; 14=Aldress Member; 15=Hagley Volcanic
Member; 16=Hagley Shale Member; 17=Whittery Volcanic Member. (SO 29)

was reopened in 1932 and listed by him (1952 : 160). He noted that the c. 32 ft (9:75 m) of rock
exposed could be divided into three parts:

3. Shales, 12 ft (3-66 m),

2. Mainly limestone, nearly 10 ft (3 m),

1. Flags, shales and limestones, nearly 10 ft (3 m).
To the east of the quarry and in the nearby roadway (Loc. 161) an underlying succession of
flaggy, calcareous beds with bands of ashy material added a further 66 ft (20 m) of strata. West
of the quarry and along the road leading towards Rorrington (Locs 164-6) successively higher
levels were mapped. Near the point where the same road crosses Lower Wood Brook, a tributary
of Desert Brook (Fig. 28, p. 42), and immediately NW along the stream, are exposures of blue-
black mudstones (Loc. 144) with layers up to 3 in (76 mm) thick of small, black, nodular concre-
tions about the size of a pea. These strata were originally mapped by Whittard as Rorrington
Member but were later reinterpreted as Meadowtown Member. SE of the same road/stream
junction, exposures throughout much of the Member are marked sporadically along Lower Wood
Brook; the apparently lowest beds visible are marked at Loc. 504, where they comprise hard
mudstones with massive bands of fine-grained tuffaceous material, and the lower boundary of the
Member is drawn between this point and the nearby Loc. 505 which exposes blue-black, micace-
ous shales of the Betton Member.

The outcrop of the Meadowtown Member extends NNE from the eponymous hamlet as a
broad strip that is ‘stepped’ westwards slightly at each of three small east-west faults, two of
which coincide with dolerite intrusions, and ends beneath uncomformable Lower Silurian strata a
short distance north of Betton Wood farm (Map). The northern half of this area is not well
exposed but several outcrops are recorded in the vicinity of Mincop, and between there and
Meadowtown.

SW of Meadowtown and beyond Lower Wood Brook, the Meadowtown Member extends SW
to Holywell Brook at a point c. 1000 ft (305 m) SE of Rorrington, where the outcrop is shifted

SHELVE INLIER 45

westwards on the south side of an arcuate, approximately ENE-WSW fault. The brook section
SE of Rorrington (Fig. 25, p. 37) is of interest because, as a result of the faulting, one finds within
the comparatively short distance of c. 450 ft (137 m) both Meadowtown/Rorrington and Betton/
Meadowtown boundaries. The latter was not actually observed at the point where the mapped
line is drawn (south of the arcuate fault), but the highest part of the Betton Member comprises
the rocks at Loc. 437, described on p. 41. North of the fault, Loc. 431 marks the highest Meadow-
town Member, and the adjacent section of Loc. 420 is assigned to the lowest part of the Rorrington
Member.

The outcrop of the Member continues to the SW (Fig. 24, p. 36), so that shales and flags are
exposed in the lane section (Loc. 470) near and SE of Middleton vicarage, to the south of
which it narrows between two converging faults. The westerly of these is one of the large north—
south tear faults which cut the Inlier, and to the west of it the outcrop of the Meadowtown Member
appears narrower as the dips are steeper. At the top of the wooded area c. 1200 ft (366 m) NNW
of Little Weston (Loc. 323) are small outcrops of usually tuffaceous shales and sandy, flaggy beds
together with interbedded tuffs, some containing nodules, possibly calcareous (Fig. 29, p. 43).
A short distance along the strike is the lane leading west from Little Weston, where Loc. 314
showed bluish-black shales with interbedded, thin, up to 3 in (76mm) bands of limestone,
decalcified marginally; the beds are almost vertical, but there is some hill creep. The outcrops of
this area are truncated to the south by an east—west fault, to the south of which Meadowtown
Member was shown in Whittard’s original mapping, an interpretation subsequently modified
to show a large outcrop of Rorrington Member underlying the feature of Spy Wood Member
running along part of the upper reaches of Coed Brook (see p. 43).

Beyond the fault only two other outcrops of Meadowtown Member are encountered. The first,
centred on the stream section about quarter of a mile (0-4 km) SW of Priestweston, is sub-rhombic
in outline, divided into two unequal parts by the north-south tear fault noted above. In the
eastern part, flags and sandy shales at Loc. 354 (Fig. 23, p. 34) yielded fragmentary fossils
including Ogygiocarella debuchii and Whittardaspis sp. The western part is less well exposed but a
small outcrop (Loc. 944) in a ditch, not noted in the earlier mapping, contained O. debuchii and
Lloydolithus lloydii.

About half a mile (0-8 km) SSW and located between the same north-south tear fault and a
subparallel fault sited a short distance to the west, is a small outcrop of Meadowtown Member
centred c. 1500 ft (457 m) NW of Upper Aldress (Fig. 30, p. 44). The two halves of the outcrop
are separated by one of the complementary shear faults associated with the tear faults. The
northern half extends as a narrow strip between the stream sections of Deadman’s Dingle (see
description of Betton Member, p. 42) and the upper reaches of Spy Wood Brook. In the former
section, Loc. 388 exposed dark, micaceous, blocky shales with numerous O. debuchii, rare
Primaspis and inarticulate brachiopods; the rocks are faulted against those of the Betton Member,
which may be seen a short distance upstream. The latter section (Loc. 379) exposes blue, fine-
grained limestones and sandy shales with interbedded ashy bands which contain abundant
pyrite and volcanic material mixed with argillaceous material; shale pellets and irregular, phos-
phatic nodular masses also occur. The southern half of the outcrop, triangular in outline, is cut
by a small tributary stream which runs NW to Spy Wood Brook. The section (Loc. 393) shows
rocks of the Rorrington Member at its NW end, cut by calcite veins and separated by a fault
from dark-grey limestones and greenish-grey shales, both of which contain volcanic material,
assigned to the Meadowtown Member. A few fossils are reported, including O. debuchii, in-
articulate brachiopods and Dicellograptus. Immediately SE of the stream section, the Ordovician
outcrops are terminated by glacial deposits forming the southern boundary of the Shelve Inlier.

Rorrington Member

First USAGE. As Rorrington Group by Lapworth (1887 : 662), a term subsequently modified to
Rorrington Flags (Lapworth & Watts 1894: 316, 318) and Rorrington Shales (Lapworth &
Watts 1910 : 752) but with no apparent difference in meaning. A change to Rorrington Group or
Stage, subdivided into successive Nemagraptus Beds and Leptograptus Beds (Lapworth 1916:

46 W. F. WHITTARD

361; Watts 1925: 340) was abandoned by Whittard who (1931: 323, 332 and subsequent papers)
preferred the term Rorrington Beds, corresponding to the present Member.

TYPE LOCALITY. Lower Wood Brook, near Rorrington (Whittard 1960 : 237).

DESCRIPTIVE NOTES. The rocks were described as sooty, blue-black shales c. 1000 ft (305 m) thick
by Whittard (1931 : 332; 1960: 237) who assigned them to the Nemagraptus gracilis Zone and
listed numerous species of graptolites and trilobites. As pointed out elsewhere (Dean in Williams
et al. 1972 : 41), further refinement of the correlation between both Meadowtown and Rorrington
Members and the corresponding graptolite zones must await description of the appropriate
faunas.

In general, the outcrop of the Rorrington Member forms a broad band running SSW across
the Inlier parallel to that of the Meadowtown Member. In an earlier version of Whittard’s map the
Rorrington outcrop was almost three times as broad as that of the Meadowtown Member, but
following his later revision of the boundary between the two the outcrops were shown as of
subequal breadth. The northernmost extension of the outcrop meets the superficial deposits
forming the northern boundary of the Inlier in the vicinity of Lower Wood farm, where the
beds are intruded by both dolerite dykes and a mass of dolerite (Map), and to a small extent are
unconformably overlain by Lower Silurian strata. At the SW end of the quarry near Lower
Wood farm the igneous rock, which was worked to a depth of 70 ft (21-3 m), is thrust over shales
of the Rorrington Member. One small Dice/lograptus ? and two minute ogyginid trilobites were
found in the shales, of which a section (Fig. 28, p. 42) of 24 ft (7-3 m) was seen (Loc. 881), and
the beds show no sign of metamorphism other than a hardening which might be due to mineraliza-
tion along the contact rather than to heat. The igneous rock, which appears to be in the form of a
boss, shows no chilled margin and on intrusion caused little induration of the shale.

The valley of Lower Wood Brook runs SSW from the eponymous farm for about three-
quarters of a mile (1-2 km) and intermittent exposures of Rorrington Member appear. Un-
fortunately the stream runs approximately along the strike so that the exposures cover little more
than a quarter of the total thickness. However, the valley of an unnamed tributary stream, which
runs NW to join Lower Wood Brook a short distance upstream from the building named Desert,
provides important information and may be regarded as a supplementary type locality as it shows
several exposures in what is almost a dip section through the lower half of the Member (Fig. 28).
Beds of the typical lithology were recorded at Loc. 516 though those at the eastern end were said
to be a little more thickly bedded. Ogygiocarella debuchii was listed from the lowest fossiliferous
strata there, and was recorded in association with Marrolithoides arcuatus Whittard slightly
higher in the sequence. A short distance SSW along the strike, strata apparently at about the
same horizon as the highest exposed at Loc. 516 were shown at Loc. 514, but containing a more
varied fauna which included Dice//ograptus, ostracods, inarticulate brachiopods and the trilobites
Primaspis whitei Whittard and Spirantyx calvarina Whittard in addition to O. debuchii.

Slightly more than three-quarters of a mile (1:2 km) SSW, the valley of Holywell Brook also
provides a dip section through the Rorrington Member, though it lacks exposures in the middle
portion such as are found in Lower Wood Brook. On the other hand, it does show the lowest
strata and, supplemented by the section in its tributary stream Grey Grass Dingle (Fig. 31),
much of the highest part of the succession, though the boundary with the Spy Wood Member is
obscured by a small fault. As noted earlier, Loc. 431 in Holywell Brook (Fig. 25, p. 37) exposes
highest Meadowtown Member, with O. debuchii, Cnemidopyge granulata Whittard and ostracods.
The closely-adjacent Loc. 420 shows shales with sandy shales or siltstones, and higher beds
become progressively less resistant to weathering. Only inarticulate brachiopods were recorded
there but the succeeding strata at Loc. 429 yielded O. debuchii, C. granulata, S. calvarina and
M. arcuatus, while the rocks were said to include thin, up to 9 in (22:9 cm) bands of fine-grained
tuffs in addition to shales of the usual type. The stream section continues as far as Loc. 427, but
succeeding strata were seen only at Grey Grass Dingle, a short distance SW. Whittard’s notes
draw attention to the fact that these higher beds are poorly fossiliferous and speculate that the
same may be true of all the higher beds of the Rorrington Member. Although fossils are so scarce

SHELVE INLIER 47

Fig. 31 Geological map of the area SW of Rorrington. Overlaps Figs 25, 28. 9=Weston Member;
10=Betton Member; 11—Meadowtown Member; 12—Rorrington Member; 13=Spy Wood
Member; 14—Aldress Member; black outcrop D=dolerite. (SO 29, SO 39, SJ 20, SJ 30)

at Grey Grass Dingle, Loc. 334 is of interest as having yielded Nemagraptus in addition to
O. debuchii and M. arcuatus.

The remaining outcrops of the Rorrington Member occupy a not inconsiderable area but,
owing to the regressively-weathering nature of the rocks, reveal few exposures. In the area west
of Little Weston, at and around the head-waters of Coed Brook, no exposures are documented
on Whittard’s maps and the outcrop is apparently delimited on the basis of the Spy Wood and
Meadowtown Members, both of which are more clearly recognizable.

South of the ENE-WSW fault already noted near Little Weston, p. 45, the outcrop is shifted
eastwards and its eastern boundary is faulted against the Weston Member of the Priestweston
area. The mapped outcrop of the Rorrington Member is fairly extensive but, again, sections are
few though a small tributary stream of Coed Brook shows exposures of rocks in the highest part
of the Member c. 1850 ft (564 m) at 101° true from Hagley farm. From west to east there are good
exposures at first of fine-grained, medium-grey, tuffaceous grit, followed by blue-hearted, rusty-
weathering shale. Then, at Loc. 309, there are more massive beds traversed by numerous calcite
veins. The shaly partings at this locality yielded Marrolithus bilinearis Whittard, a species generally
found in the Spy Wood Member, and a single specimen of Platycalymene duplicata (Murchison)
was recorded (Whittard 1960qa : 156). Farther east in the same section, Loc. 310 exposed fossili-
ferous, blue-hearted, micaceous shale interbedded with bands of tuff up to 9 in (23 cm) thick;
recorded fossils include rare Spirantyx calvarina and abundant ostracods, and some of the
specimens were said to be preserved in pyrite.

About 1200 ft (366 m) NNE of Lower Aldress (Fig. 32) another east-west tributary of Coed
Brook shows several exposures of shales together with some massive flags in the western part of
the section, where the beds are faulted against the Aldress Member. Few fossils were found but
Leptograptus and Nemagraptus were recorded at Loc. 351.

The southernmost exposures of this large, elongated outcrop occur in the faulted area of Spy

48 W. F. WHITTARD

} iS Loc 301
Marrington : ISS

i
Las:
es
aviwvo Y3AI8
—_—

ee
SS

(9 SS

\ 17 Kingsw

3

Kee \ 17
Dingle Ti Wvso

House*))\\ _ \

=

e@

Aldress Ding ij

G t iy 1)
sik i4 Spy Wood |

f oe Cottage zi

27 28
Fig. 32 Geological map of the area east of Marrington. Overlaps Figs 23, 30, 33. 10=Betton
Member; 11—Meadowtown Member; 12=Rorrington Member; 13=Spy Wood Member;
14—Aldress Member; 15=Hagley Volcanic Member; 16—Hagley Shale Member; 17=Whittery
Volcanic Member; 18=Whittery Shale Member. (SO 29)

Wood Brook and part of its small tributary Deadman’s Dingle, where a fairly continuous section
in the lower half of the Member can be pieced together (Figs 30, p. 44, and 32). In the Spy wood
Brook, rusty-weathering shales about the middle of the Rorrington Member occur at Loc. 377,
where they yielded S. calvarina and P. whitei. The locality is just east of the fault separating the
rocks from others forming the highest part of the same Member in which several localities,
including 373-5, occur only a short distance below the base of the Spy Wood Member and contain
a few fossils including Marrolithoides arcuatus. Higher upstream, and consequently lower in the
succession, are poorly fossiliferous micaceous shales, weathering dark brown, underlain by
nodular, greenish-grey siltstones. In Deadman’s Dingle, rocks of the Meadowtown Member at
Loc. 388 (see p. 45) are succeeded by shales of the Rorrington Member. Loc. 389, only slightly
higher in the succession, proved poorly fossiliferous but yielded inarticulate brachiopods and a
few graptolites (Nemagraptus ? and Dicellograptus) and was followed by beds of typical lithology
at Loc. 390 containing O. debuchii and S. calvarina. Strata in the highest part of the Rorrington
Member are seen also in Brynkin (=Bryncyn) Dingle, where O. debuchii and S. calvarina were

SHELVE INLIER 49

recorded from shales at Loc. 943. These are the most southerly outcrops of the Rorrington
Member, and only a short distance to the south and SE the rocks are overlain by glacial deposits.

Chirbury Formation

First USAGE. As Chirbury Series (Lapworth 1887 : 662), the highest of three ‘Series’ of the Shelve
Ordovician rocks and underlain by Meadowtown (later Middleton) Series. The term remained in
constant use by both Lapworth and Watts until 1925 and was then abandoned without explana-
tion.

TYPE LOCALITY. Presumably the area east of Chirbury village, but no type section was designated.

Spy Wood Member

First usAGE. As Spy Wood Calcareous Grit, the lower of two subdivisions of the Aldress Group
proposed by Lapworth (1887: 662). Later abbreviated to Spy Wood Grit by Lapworth & Watts
(1894 : 316), the lower half of a newly-introduced Spy Wood Group (=Aldress Group of 1887).

TYPE LOCALITY. Spy Wood Brook (known alternatively as Spy Wood Dingle, Burn or Bourn),
NNE of Church Stoke (Whittard 1960 : 253).

DESCRIPTIVE NOTES. Whittard (1931 : 332) described the rocks as flaggy, calcareous sandstones
with interbedded greyish shales, weathering olive-green, which assumed a greater importance
higher in the succession. The sandstones, when decalcified, were said to contain well-preserved
fossils and there was an imperceptible passage upwards into the Aldress Shales (now Aldress
Member). In a later paper (Whittard 1960 : 253) the thickness was given as 300 ft (91-4 m) and
the strata were said to belong to the Dip/ograptus multidens Zone.

The southernmost extension of the Spy Wood Member is in the vicinity of Upper Brynkin and
Brynkin Green (Fig. 30, p. 44), in which area it forms a feature that is covered to the south by
glacial deposits forming the margin of the Shelve Inlier. To the north it is ‘stepped’ NE by two
small faults which converge east and bound a small, triangular ‘block’, exposed (Loc. 394) in the
small tributary stream of Coed Brook which runs NW from Brynkin Green and is shown in the
manuscript notes, though not on the 6 in to | mile (1 : 10 560) O.S. map, as Brynkin Dingle.

More extensive exposures occur at several points in and near the stream which runs through
Spy Wood and c. 450 ft (137 m) NE of Spy Wood Cottage (Fig. 30). Almost in the centre of the
stream section is a good exposure (Loc. 368) of a small anticlinal fold pitching SW at 35°. The
rocks in this vicinity are massive, highly micaceous, rusty-weathering flagstones, up to | ft (30 cm)
thick, interbedded with sandy shales carrying globules of pyrite. Farther east along the same
stream, the boundary of the Spy Wood Member and underlying Rorrington Member is exposed
in the south bank (Loc. 370). Graptolitic strata of normal Rorrington lithology (Loc. 371) are
followed by 9 ft (2:74 m) of micaceous, bluish-grey shale with interbedded thin bands of grey,
micaceous, unfossiliferous flagstone. These are succeeded in turn by 12 ft (3-66m) of more
massive beds of typical Spy Wood lithology and it is at the base of these strata that the base of the
Member is drawn. The sequence is apparently a normal one with no evidence of a basal conglo-
merate or angular discordance, and the presence of these unfossiliferous passage-beds at the top
of the Rorrington Member was described briefly by Whittard (1952 : 163).

The western end of the same stream section demonstrates the transition between the Spy Wood
and overlying Aldress Members. Although no continuous section is exposed the story, again, is
one of gradual lithological change, with no evidence of a sharp lithological break at which to
draw the boundary. Loc. 365 (Fig. 30) was considered by Whittard to be in the Aldress Member,
and the boundary with the underlying Spy Wood Member was drawn at a point 52 ft (15:85 m)
upstream (i.e. to the east). On this interpretation the upper beds of the Spy Wood Member
comprise flagstones alternating with soft shales and flaggy, sandy shales, all showing various
tones of grey and weathering to a dark, rusty colour which is characteristic also of the Aldress
Member. The beds are much less massive than those in the lowest part of the Member, and the
subdivision may be looked upon as forming the arenaceous base to the Aldress Member.

50 W. F. WHITTARD

The Spy Wood Member is mapped as a feature NNE from Spy Wood Brook for c. 2000 ft
(610 m), when it is truncated by a north-south tear fault running through Lower Aldress farm
(Fig. 32, p. 48). It is not seen again until the vicinity of Lower Ridge farm, 14 miles (2:4 km) east
of Chirbury, where it reappears on the east side of the same north-south fault and forms a small,
elongated outcrop along part of Coed Brook (Loc. 833 — Fig. 29, p. 43). A short distance north of
Lower Ridge, a further subparallel outcrop recommences and continues NE intermittently to
Rorrington Hall. There is little of note in this stretch, but the outcrop is ‘stepped’ by occasional
faults of relatively small throw and (Fig. 31, p. 47) at Loc. 327, 2410 ft (735 m) at 045° true from
Kinton farm, a mile (1-6 km) SW of Rorrington Hall, yielded two well-preserved brittle-stars in
a bed less than 1-57 in (4 cm) thick within a succession that includes mainly sandy, micaceous,
brown shales. At Loc. 331, 300 ft (91:4 m) south of Rorrington Hall and just inside the gate on
the footpath to Kinton, beds low in the Member are exposed. They comprise micaceous, blue-
black shales but, as the succession 1s ascended, sandy beds and flagstones appear alternating with
the shales, some of which are sandy. It was here that an impersistent bed 14 in (38 mm) thick
yielded several starfish. Still higher in this dip section there is a progressive increase in sandstones
until one reaches the typical so-called ‘Beyrichia grits’ of earlier authors, in which ostracods
commonly occur in bands but are comparatively rare in the intervening strata. It was from this
section that the type material of Tetradella salopiensis Harper (1947 : 351) was obtained.

The stream section immediately behind and NE of Rorrington Hall is not straightforward,
owing to faulting, but probably represents the approximate thickness of the Spy Wood Member.
None of the beds are true shales; all contain some sandy material, the more sandy ones pass into
tough siltstones or fine-grained sandstones, and many are quartzitic in appearance. The SE
end of the section would appear to indicate an upward passage from the Rorrington Member,
with a progressive increase in flaggy beds. It is important to note that the rocks seem virtually
unfossiliferous because they are not sufficiently weathered; when weathered, they may be richly
fossiliferous in certain bands, and this condition was noted also in other sections, for example
Spy Wood Brook.

From Rorrington Hall the outcrop of the Spy Wood Member extends (Fig. 28, p. 42) NNE,
broken occasionally by small faults, to terminate at the northern boundary of the inlier about half
a mile (0-8 km) west of Betton Wood Farm, where the rocks are overlain unconformably by
Upper Llandovery strata which are themselves succeeded by Quaternary and Recent deposits.
The beds form a mappable feature and débris rather than exposure tends to be the rule, but one
outcrop (Loc. 415), 2900 ft (884 m) at 026° true from Rorrington Hall, is of particular interest as
some of the rocks there (not those which are argillaceous or micaceous) yielded fossils which
include the zonal trilobite Costonia ultima (Bancroft), indicating an age high in the Costonian
Stage of the type Caradoc Series.

Aldress Member

First USAGE. As Aldress Graptolithic (sic) Shale, the upper of two subdivisions of the Aldress
Group, by Lapworth (1887 : 662). No description was given and the term subsequently came into
general use as Aldress Shale or, usually, Shales.

TyPeE LOCALITY. Aldress Dingle, 1} miles (2-4 km) NNE of Church Stoke, west Salop.

DESCRIPTIVE NOTES. Brief accounts by Whittard (1931 : 333; 1960: 32) described the rocks as
grey or greenish-brown, soft shales with occasional bands of tuffaceous material in which the
fossil assemblage is different from that found in the shales. The thickness was said to be c. 1000 ft
(305 m), and the beds were assigned to the Diplograptus multidens Zone.

In the north of the Shelve Inlier, rocks of the Aldress Member underlie a strip of country, with
a maximum width of some 0-7 mile (1-13 km) where affected by faulting but generally much less,
which extends to a point half a mile (0-8 km) NE of Wilmington and is bounded to east and west
by, respectively, the features of the Spy Wood and Hagley Volcanic Members. Striking physio-
graphic features of this northern area are formed by elongated dolerite intrusions within the shale
outcrop between Kinton and Rorrington Lodge, and to the east of Wilmington (Map), but the
shales themselves are not well exposed.

SHELVE INLIER 51

The stream section running north almost along the strike from Rorrington towards Wilmington
is excavated mainly in glacial deposits, but shows intermittent outcrops of shales and tuffs in the
lowest third of the Member. A better section, in the highest strata and noted briefly by Whittard
(1931 : 333), is that cut by the stream in Ox Wood (= Ox Wood Dingle), mid-way between
Wotherton and Rorrington (Map), where several localities, some graptolitic, are marked on the
manuscript map.

No notable sections are shown in the continuation of the overall outcrop SSW to the vicinity
of Hagley, though small, scattered outcrops are found, particularly in stream sections. From
Hagley southwards, however, there is an increase in the degree of availability of the rocks,
particularly along the type section of Aldress Dingle (Fig. 32, p. 48), though the latter suffers
from the disadvantage of being excavated almost along the strike and cuts mainly the middle and
upper parts of the Member. Loc. 344 (Fig. 30, p. 44), 1750 ft (533 m) at 079° true from Calcot, is
of interest as having provided the type material of Dictyonema fluitans Bulman (1928 : 35).
Two dip sections along streams running east-west in the area immediately north of Lower
Aldress farm show the lower beds of the Member, but the boundary with the underlying Spy
Wood Member is seen satisfactorily only in Spy Wood Dingle, north of Spy Wood Cottage, noted
on p. 49. The rocks in the lowest part of the Member here (Locs 364, 365) are medium-grey in
colour but many comprise micaceous siltstones and flags rather than shales, and bedding planes
are not always clearly developed (Fig. 30).

A further supplementary section is found in Brynkin Dingle, a term used by Whittard for a
tributary of Aldress Dingle (Fig. 30). The lowest beds exposed (Loc. 395) are shales with grey,
micaceous flagstones up to a foot (30-5 cm) thick, below which is a thin, inch (2:5 cm) band of
white, shaly material, possibly a bentonite and perhaps one of the occurrences noted for the
Aldress Member by Whittard (1952 : 164). Shales with flags in the highest part of the Aldress
Member were noted in Aldress Dingle only a short distance upstream from the mapped base of
the Hagley Volcanic Member, but no detailed account of the boundary was given. From Aldress
Dingle to the southern margin of the Inlier, the only outcrops of Aldress Member shown on the
manuscript map are some exposures of shale along the stream section which runs westwards to
join the Camlad about 500 ft (152m) NE of Alport. The outcrop was not numbered and no
notes were made.

Hagley Volcanic Member

FirST USAGE. As ‘Hagley volcanic ashes and shales’ (Lapworth 1887 : 662), a term equivalent to
the present Hagley Volcanic Member and Shale Member combined. ‘Hagley Ash’ was separated
by Lapworth & Watts (1894 : 316, 318) and later passed into general use as the Hagley Volcanic
Group (Whittard 1931 : 323), equivalent to the present Member.

TyPeE LOCALITY. Hagley Quarry, near Chirbury (Whittard 1960: 145).

DESCRIPTIVE NOTES. In a brief note on the subdivision Whittard (1931 : 333) stated: ‘This contains
massive crystal and lithic tuffs, which are often brecciated and agglomeratic. The normal rock is of
a pale greenish colour, brindled with darker green rings’. Later (Whittard 1952: 164) he gave a
thickness of 350 ft (106-7 m) and noted that at Hagley Quarry graptolitic evidence of the c/ingani-
linearis Zones had been found. The latter interpretation has now been superseded and the
Member is known to belong to the Lower Soudleyan Stage of the Caradoc Series, equated with
part of the Diplograptus multidens Zone.

The northernmost limit of outcrop is in the area between Wotherton and Wilmington, where
volcanics form a wide feature at Big Cuckoo Nest and Little Cuckoo Nest, just west of Ox Wood
Dingle (Map and Fig. 34). The strip-like outcrop here runs SSW through the eastern half of
Crest Wood, displaced by small east-west faults, but owing to strike faulting it becomes markedly
narrower towards the south end of the wood where it is truncated by a prominent north-south
tear fault which displaces the outcrop dextrally. On the west side of this fault the outcrop is
shifted NW as far as Rockabank farm and from there to the south forms a feature which extends
to Hagley farm (Figs 33-34).

52 W. F. WHITTARD
27 28

Heightleys2 ~~~“

Whittery’

2.

Whittery x50

NS

Walkmill &

- Marrington
\dall g

27 28
Fig. 33. Geological map of the Whittery area, east of Chirbury. Overlaps Fig. 32. 14—=Aldress

Member; 15=Hagley Volcanic Member; 16—Hagley Shale Member; 17=Whittery Volcanic
Member; 18=Whittery Shale Member. (SO 29)

Hagley Quarry (Loc. 301), the type locality, lies just north of Hagley farm and by the side of
the Chirbury—Priestweston road. The area is poorly exposed and although the map shows the
upper boundary of the Volcanic Member almost coincident with the western margin of the
roadside quarry, the stratigraphical base is not visible here owing to a fault which runs from
Upper Ridge farm to Hagley farm and intersects the strike at an acute angle. Hagley Quarry is
of particular interest for having yielded graptolites, which were collected from water-deposited
tuffs by members of the Geological Survey. The fauna, listed by Whittard (1931 : 333-334),
was originally quoted as indicating the topmost c/ingani or basal /inearis Zone, a widely-quoted
but mistaken interpretation that has since been modified to multidens Zone (see above). At a
point 1000 ft (305 m) just east of north from Hagley farm, and on the western edge of the same
volcanic feature, are the remains of an old adit (Fig. 33) from which, according to Whittard’s
notes, both barytes and (at depth) galena were at one time obtained (Loc. 300).

Apart from a small break just south of Hagley, the feature of volcanic rocks extends southwards
to Church Stoke village. Tuffs and agglomerates are seen at Kingswood farm (Fig. 32, p. 48),
and an old quarry (Loc. 356) 300 ft (91:5 m) south of the building shows a good exposure in
greyish-yellow weathering, lithic tuff with interbedded agglomerate. ‘Throstle-breasted’ and
other types of tuff crop out (Loc. 382A) 3200 ft (975 m) at 190° true from Kingswood (Fig. 32,
p. 48), near the ford in Aldress Dingle and a short distance from the River Camlad, but otherwise
there is little of note until one reaches the roadside quarry (Loc. 399) 450 ft (137 m) north of
Church Stoke Hall. At this point tuffs, interbedded with shales and showing false ripple-marking,
yielded the trinucleid trilobites Broeggerolithus cf. broeggeri (Bancroft) and Sa/terolithus caractaci
(Murchison), indicating a Lower Soudleyan age within the Caradoc Series. Intermittent outcrops
of tuffs continue south for c. 1500 ft (457 m) before disappearing beneath glacial deposits (Fig.
3S p.55))

SHELVE INLIER 53
28

15 3
‘13
Crest |woce! oe

28

Fig. 34 Geological map of the area south of Wotherton and NE of Chirbury. Adjoins Fig. 29.
13=Spy Wood Member; 14=Aldress Member; 15=Hagley Volcanic Member; 16=Hagley
Shale Member, including tuff beds (dotted); 17=Whittery Volcanic Member; 18=Whittery
Shale Member. (SO 29, SJ 20)

Hagley Shale Member

First USAGE. As Hagley Shales by Lapworth & Watts (1894 : 340, 342), a term which has since
remained in almost constant use.

TYPE LOCALITY. No specific section was designated, though the general vicinity of Hagley farm,
a mile (1:6 km) SE of Chirbury was presumably intended. The subject is discussed in more detail
below.

DESCRIPTIVE NOTES. Whittard (1931 : 333; 1960: 144) described the strata briefly as ‘rusty-
weathering, bluish shales, with harder olive-green shales tending to be nodular’. At that time they
had not yielded fossils but a few have now been found (see p. 54). The thickness was said to be of
the order of 1000 ft (305 m).

Much of the neighbourhood of Wotherton village (Map and Fig. 34) is underlain by rocks of
the Member but whereas outcrops of shales and flagstones are rare here, bands of tuffs and
agglomerates within the shales are more resistant and, consequently, available for examination.

54 W. F. WHITTARD

In particular a thin band of coarse agglomerate was seen in the sides of the road (Loc. 876) leading
to Rorrington, c. 600 ft (183 m) ENE of Wotherton. About 1200 ft (366 m) south of Wotherton
(Loc. 874) are exposures of volcanics which are unusual in that the surfaces show what appears
to be flow banding; if this is correctly interpreted, then we are dealing with one of the few lavas
known from the area.

At Loc. 340 in the stream-section 1250 ft (381 m) at 084° true from Marrington farm (Fig. 32,
p. 48) nearly all the succession, apart from one short length covered by felled trees, is visible.
Whittard’s notes did not specify this as type section for the Member, but it might be used as at
least a supplementary type section. The rocks comprise bluish-grey hearted, greenish-grey
weathering shales which on the outside are altered to a yellowish brown. Some are soft and contain
little mica. Interbedded with the shales, and dominant in the lower part of the succession, are
greenish-grey weathering, flaggy micaceous siltstones; less commonly there are more arenaceous,
light grey flags. A striking feature of all these rocks is the general paucity of fossil remains,
and the most abundant specimens so far found are from interbedded tuffs. In most respects the
rocks of the Hagley Shale Member are comparable with those of the Whittery Shale Member and
clearly belong to the same deposition, the Hagley and Whittery Volcanic Members representing
short-lived episodes of vulcanicity which nevertheless resulted in a considerable thickness of
deposits. The more westerly outcrops (Loc. 339) of the Shale Member in this stream section are
deeply-weathered, friable shales in which amplexoid-like graptolites have been found. Between
these rocks and the main section is a large bed of tuff (Loc. 362) c. 7 ft (2:13 m) thick, with some
interbedded shales.

Further exposures of the Hagley Shale Member are seen in the vicinity of the River Camlad
both NE and SE of Calcot (Fig. 32, p. 48). At Loc. 746 a presumably interbedded tuff shows
occasional graptolites in relief and contains the trinucleid trilobites Broeggerolithus broeggeri
(Bancroft) and Salterolithus caractaci (Murchison), indicating a Caradoc, Lower Soudleyan age.
A good section in the lane immediately NE of Alport (Loc. 744) shows 130 ft (39-6 m) of shales
(Fig. 30, p. 44) with flagstone beds dipping approximately 50° at 290° true. Nearly 30 bands of
flags occur, ranging from an inch (2:5 cm) to a foot (30 cm) in thickness and totalling some 6 ft
(1-83 m). A small shelly fauna again indicates a Lower Soudleyan age. Whittard’s notes suggest
that the section might well be ranked as a type section for the Member.

Shales of the Member crop out intermittently along the Chirbury road just outside the village
of Church Stoke (Fig. 35), particularly at Loc. 742, 1550 ft (472-4 m) north of St Nicholas’ church.
The exposure there includes several beds, up to 2 in (5 cm) thick, of tuffaceous flags, horizons
which suggest that the volcanics of the succeeding Member are being anticipated. Within Church
Stoke village, the road section 500 ft (152-4 m) at 323° true from St Nicholas’ church (Loc. 739)
shows rusty-weathering, greenish-grey, micaceous shales with thin, seldom more than 2 in (5 cm)
bands of micaceous flags, beds which yielded B. broeggeri and S. caractaci. Loc. 738, 400 ft
(122 m) at 212° true from St Nicolas’ church, shows greenish-grey, micaceous flags and shales
which constitute the last exposure of the Hagley Shale Member before glacial deposits forming
the southern boundary of the Inlier are reached.

Whittery Volcanic Member

First USAGE. As Whittery Ashes, by Lapworth (1887 : 662). The name was listed in a strati-
graphical table and no further details were given.

TYPE SECTION. According to Whittard (1960 : 283) it is “quarry in Marrington Dingle, east of
Whittery, west Shropshire’. This appears to be a misprint because no such quarry exists east of
the farm named Whittery, and Marrington Dingle lies to the south of both Whittery and Whittery
Wood. It seems likely that the intended type section was the one in Whittery Quarry, just over
a quarter of a mile (0-4 km) SSE of Whittery and at the southern end of Whittery Wood. There
only the upper portion of the volcanic member is exposed, together with the basal part of the
succeeding Whittery Shale Member (see p. 59).

DESCRIPTIVE NOTES. Published descriptions of the Whittery volcanic rocks are both few and brief.
Lapworth’s introduction of the term in 1887 gave no description. Lapworth & Watts (1894 : 316,

SHELVE INLIER 55)

27 28

—s
oO
TEAR

ae

«Long Leasow

---._

SS
i

als

1

!
1
|
/

27 28

Fig. 35 Geological map of the area between Church Stoke and Todleth Hill, in the SW corner of
the Shelve Inlier. 9=Weston Member; 14=Aldress Member; 15=Hagley Volcanic Member;
16=Hagley Shale Member; horizontally shaded outcrop A=andestic intrusion of Todleth Hill.

(SO 29)

318-319) noted the rocks as ‘Whittery Ash’ and listed them in a stratigraphical table. They
cited Whittery Quarry and Walk Mill Quarry (presumably the quarry c. 450 ft (137 m) SE of
Walkmill; Fig. 33, p. 52) as being places where the beds were ‘splendidly exposed’. The volcanics
were said to consist of ‘andesitic and rhyolitic breccias and conglomerates, fine ashes with curious
spherulitic or pisolitic structures, and bands of shale often fossiliferous’. Beds showing ripple
markings were also recorded.

Whittard (1931 : 333) stated merely ‘The Whittery Volcanic Group repeats the general litho-
logical and petrological characters shown by the rocks of the Hagley Volcanic Group’; later
(1952 : 164) he noted the rocks only briefly and gave the thickness as 300 ft (91-5 m). Subse-
quently (1960 : 283) he described them briefly as ‘crystal and lithic tuffs, agglomerates and
breccias with thin interbedded shales. Thickness 300 ft.’

The northernmost outcrops of the Whittery Volcanic Member occur in the vicinity of
Wotherton (Fig. 34, p. 53), where part of the outcrop is faulted against the Hagley Shale Member
to the east and the beds, which dip 40° bearing 280° true, were once quarried at Rock Coppice
(Loc. 296). The nature of the junction between the Whittery Volcanic and Shale Members at this
locality is noted in more detail under the description of the latter Member.

Of particular interest in this district is the outcrop 500 ft (152 m) SW of the house called
Rockabank, 2250 ft (686 m) south of Wotherton (Fig. 34). At this point (Loc. 295) there occurs
what Whittard described as an ‘amazing development’ of volcanic agglomerate with boulders
up to 3 ft (0-9 m) in diameter, though most are half that size or smaller, in a matrix of coarse
lithic tuff (Fig. 37). Those blocks which were examined showed the rock to be a lava, possibly a
hornblende andesite, and the boulders are so numerous, and in places so tightly packed together,

56 W. F. WHITTARD

Fig. 36 Topography of ridges formed by Whittery Volcanic Member, here faulted, with inter-
vening hollows excavated in regressive, interbedded shales. View looking south from Rockabank,
600 yd (550 m) south of Wotherton, towards Heightley Barn, 600 yd (550m) NE of Whittery.

that the matrix is not easy to find. There is no evidence, either north or south of the Rockabank
outcrop, of volcanic material which approaches this agglomerate in coarseness, though good
outcrops of agglomerate showing fragments up to 3-4in (75-100 mm) occur nearby. The
question remains: have we at last a focus for part of the Caradoc vulcanicity in the Shelve
Inlier? Mapping fails to show that the rock is in the form of a neck, although the general shape
of the outcrop tends to suggest this. On the other hand, the rock appears to be interbedded; but
here one should remember that, assuming the neck was originally vertical, the sides of the neck
would now be approximately at 45°.

The volcanic rocks of the Whittery Volcanic Member in this area form a feature which runs
SSW, its position ‘stepped’ by intermittent dip faults, but occasional indentations run along the
strike owing to differential erosion of shale beds in the sequence, and it is clear that not incon-
siderable masses of shale occur interbedded in the marine volcanics. On a hillside such as the
east side of Marrington Dingle (Fig. 32, p. 48), where the top and bottom of the Volcanic Member
can be mapped with reasonable confidence, shale fragments have been brought to the surface by
rabbits at several points mid-way up the slope. A shale depression was mapped north of Heightley
New Barn, 1650 ft (S503 m) NE of Whittery, and may represent a similar condition appearing
farther south (Fig. 33, p. 52).

The large quarry (Loc. 299) at the southern end of Whittery Wood (Fig. 33) is the original
Whittery Quarry and therefore the probable type section for the Volcanic Member, the rocks of
which were once extensively quarried there. The section shows also part of the overlying Whittery
Shale Member, the beds of which include flags and bands of volcanic rocks, indicating waning
vulcanicity as at Loc. 296 (see p. 59). Some of the bands may be up to 2 ft (0-61 m) thick and one
may once more appreciate the possibility of misidentifying such rocks as part of the Volcanic
Member when seen in isolated exposures. Six hundred feet (183 m) to the south is a smaller quarry
(Loc. 875), almost certainly the Walk Mill Quarry of Lapworth & Watts (1894 : 318), which shows

SHELVE INLIER Sy

Fig. 37 Volcanic agglomerate in Whittery Volcanic Member at Rockabank, 600 yd (550 m) south
of Wotherton, 1-9 miles (3 km) NE of Chirbury.

normal volcanics but is of interest because at the top there is a band showing flow-casting on its
lower surface, where it rests on shale.

Four hundred feet (122 m) south of Walk Mill Quarry a variety of tuffs, both lithic and the
so-called ‘throstle-breasted’, is shown (Fig. 33, p. 52) in a small disused quarry (Loc. 358),
300 ft (91-4 m) SE of which is a hillock where good exposures (Loc. 337) of agglomeratic tuff are
found, with well-rounded boulders, some attaining a length of a foot (30 cm). Northwards the
agglomeratic character of the rock remains evident but the agglomerate is obviously interbedded
with crystal and lithic tuffs. The hillock is peculiar in that it is followed to the west by a flat
region before the steep flank of Marrington Dingle is reached, and it is not clear whether one is
dealing with an isolated outcrop or one that is faulted from the main outcrop of Volcanic Member.
The hillock is separated by a small valley from the outcrop of Volcanic Member as traced from
the north and it strikes southwards into open country.

Good examples of agglomerate occur a little farther south on another hillock (Loc. 338)
upon which Bank Farm, 900 ft (274 m) east of Marrington Hall (Fig. 33) is built. This should
not be confused with the building called Bank which is situated 1060 ft (323 m) SE of Whittery
(Fig. 33).

The faulting of the Whittery Volcanic Member in the area parallel to Marrington Dingle and
south of Whittery Wood is noteworthy because, by the usual methods of determination, the
downthrow is to the north. However, if the upper surface of the volcanics is traced in relation to
its O.D. level, the fault block to the south is at a lower level than its fellow to the north, so that
the downthrow is apparently to the south. It would appear that the south side is really the down-
throw, but that tear-faulting has resulted in a greater horizontal than vertical movement. Conse-
quently the horizontal shift (to the east on the north side of each fault) is greater than the dis-
placement due to a southerly downthrow.

At a point 210 ft (64 m) ENE of the building called Marrington (Fig. 32, p. 48), the east bank of
the River Camlad is intersected almost at right angles by an unnamed stream,the upper reaches of
which show an almost continuous section (Locs 360-1) through the Whittery Volcanic Member.

58 W. F. WHITTARD

= be eae Sd ee 1! a * ; a a ay

Fig. 38 Flow casts in Whittery Volcanic Member. Quarry 220 yd (200m) south of Whittery
Quarry, 0-83 mile (1:34 km) east of Chirbury. (Coin 28 mm diam.)

The stratigraphically lower half of the sequence (Loc. 361) contains much more shale than the
higher half and the tuffaceous rocks are much finer grained, some of them yielding graptolites.
At this point volcanics seen just above a bed of shale 14 ft (46 cm) thick occur in large, con-
cretionary masses which are pillow-shaped, concentrically banded and a minimum of 4 ft (1:22 m)
in diameter. These are demonstrably depositional and not tectonic structures. Some anomalous
easterly dips seen a little farther upstream may indicate strike faulting and folding, thought to be
of a very minor character.

Farther SW, 440 ft (134 m) SE of Marrington, a section (Loc. 363) in the Camlad shows a
slight anticlinal flexure whereby volcanics near the top of the Member occur in the river and are
succeeded by beds of the Whittery Shale Member, seen in the west bank only a short distance
south.

The southernmost outcrops of the Whittery Volcanic Member are seen at an old quarry (Loc.
747; Fig. 32, p. 48) just north of Calcot and by the west bank of the Camlad. A minor syncline
is developed there and the rocks, which include tuffs interbedded with shales at the eastern end
of the quarry, are much broken. Southwards beyond this point the outcrops are obscured by
glacial deposits.

Whittery Shale Member

FIRST USAGE. By Lapworth & Watts (1894 : 316, 319). The name was listed as Whittery Shale in a
stratigraphical table and as Whittery Shales in the text, but no description was given.

TYPE LOCALITY. None was designated by Lapworth & Watts, though the beds were said to ‘form
the ground between Marrington and Chirbury’. The base of the Member is seen at Whittery
Quarry (see p. 56).

DESCRIPTIVE NOTES. The rocks of the Whittery Shale Member resemble those of the Hagley Shale
Member, and comprise predominantly blue-grey, micaceous shales which appear light rusty brown
or olive-green after weathering. Other, but subsidiary, rock-types include bluish green, heavily

SHELVE INLIER 59

micaceous, flaggy bands and both fine- and coarse-grained interbedded tuffs. The tuffs in par-
ticular are liable to be found fossiliferous, with both shelly and graptolitic faunas.

Whittery Quarry (Loc. 299), at the southern end of Whittery Wood, near Chirbury (Fig. 33,
p. 52), exhibits the boundary of the Whittery Shale Member with the underlying Whittery
Volcanic Member. The rocks there dip 45° bearing 292° true and show flags and volcanic bands
interbedded with the shales. One important feature there is that within the shales there may be
bands of volcanics up to 2 ft (0-61 m) thick, and small, isolated exposures of such rocks might
lead one to map such outcrops mistakenly as the Whittery Volcanic Member.

The junction between the Volcanic and Shale Members is seen also in the western part of the
old quarry (Loc. 296) at Rock Coppice, 333 yd (305 m) SW of Wotherton (Fig. 34). Above the
massive tuff there are alternations of thin tuff, mostly | in (25 mm) but at maximum 2 in (51 mm)
thick, and shales up to 6 in (152 mm) thick, totalling about 6 ft (1-83 m), after which point normal
shale conditions became established. The junction between the two members is thus a normal one,
with a gradual passage from volcanics to shale, demonstrating that the end of the vulcanicity was
marked by a few, probably explosive, eruptions. This was the last of the vulcanicity in the Ordo-
vician of west Salop except for tuffaceous bands at several levels in the Whittery Shale Member,
indicating short and unimportant recrudescences in the vulcanicity.

Streams flowing westwards into the River Camlad and down the dip of the Whittery Shale
Member show almost continuous sections at points 2600 ft (792 m) NE (Loc. 292) and c. 3800 ft
(1158 m) NNE (Loc. 293) of Heightley (Figs 33-34), near Chirbury. The predominant rock type
is shale but bands of flagstone are developed; a peculiar and so far characteristic rock is a
tuffaceous shale which contains stony fragments up to half an inch (13 mm) in diameter. The
latter rock type occurs also at Loc. 294, c. 250 ft (76 m) SE of, and upstream from, Loc. 293,
and has yielded fossils which include Salterolithus caractaci (Murchison) and Broeggerolithus
broeggeri (Bancroft), characteristic of the Glenburrell Beds (Lower Soudleyan) of the Onny
Valley in SE Salop.

Beds high in the Shale Member crop out along the stream south of Chirbury at Locs 272 and,
especially, 273 (Fig. 33, p. 52). The latter shows a good section in which shale is the predominant
element though there are some flaggy layers. A few graptolites have been obtained from this
locality.

The most westerly exposure of the Ordovician rocks of the Shelve Inlier occurs at Loc. 752,
a quarter of a mile (0-4 km) NE of Trimberth. Thirty years ago the strata of the Whittery Shale
Member were better exposed there than nowadays but the lithology is typical and there is no
mistaking their age. Of passing interest there is a small quarry where the shales are no longer
exposed but could be examined thirty years ago; on that occasion the dip was found to be high
and to the SE. The significance of this change-round from a prevailing westerly dip cannot be
determined as there are no other exposures in the immediate neighbourhood, where the only
material to be seen is the ubiquitous boulder-clay.

Miscellaneous notes

Correlation with neighbouring Ordovician successions

The correlation of the Shelve Ordovician rocks with those of the Caradoc District, the Breidden
Hills and the Pontesford district was summarized by Whittard (1952 : table III) in a table which
listed the corresponding graptolite zones. The Stiperstones Member and Mytton Member were
equated with the combined Didymograptus extensus and D. hirundo Zones; the Hope, Stapeley
Volcanic, Stapeley Shale and Weston Members with the D. bifidus Zone; the Betton Member with
the D. murchisoni Zone; and the Meadowtown Member questionably with the G/yptograptus
teretiusculus Zone. The Rorrington Member was equated with the Nemagraptus gracilis Zone
and part of the Diplograptus multidens Zone; the Spy Wood Member questionably with part of
the D. multidens Zone; and the Aldress Member with part of the D. mu/tidens Zone and part of
the Dicranograptus clingani Zone; the remainder of the succession was equated with part of the
D. clingani Zone and the Pleurograptus linearis Zone.

60 W. F. WHITTARD

The problem of how to account for this supposed correlation of the highest Ordovician strata
of the Shelve Inlier troubled Whittard for several years and appears to have arisen from the
Geological Survey’s discovery of graptolites in tuffs of the Hagley Volcanic Member at the
type locality of Hagley Quarry. This find was reported by Whittard (1931 : 334) and the speci-
mens, identified by G. L. Elles, were interpreted as indicating ‘a horizon about the base of the
Zone of Pleurograptus linearis or at the very top of the Zone of Dicranograptus clingani’. However,
it had for some years been customary to regard the Shelve Ordovician rocks as constituting a
virtually complete succession. For example, Watts (1925 : 340, 343) placed the top of the Caradoc
Series immediately above the Whittery Volcanic Member and regarded the Whittery Shale
Member as possibly representing the Ashgill. Consequently no major policy change was required
in order to relegate the Whittery Shale Member to a slightly lower horizon, supposedly in the
highest Caradoc Series.

Correlation of the Shelve rocks of Caradoc age has been facilitated by use of the ‘shelly’ zones
and stages established by B. B. Bancroft, though their original application was often inaccurate
and only relatively recently has a more convincing picture emerged. Bancroft’s (1933) correlation
tables followed the previous practice (e.g. Lapworth & Watts 1894 : 316) of equating the Hoar
Edge Grits of the Caradoc district with the Spy Wood Member (then Spy Wood Grit), claiming
that the latter contained ‘Marrolithus (Costonian type)’. Whittard’s (1966 : 281) later discovery of
the South Salop Costonian index trilobite Costonia ultima (Bancroft) — a species originally assigned
to Marrolithus — in the Spy Wood Member (see p. 50) showed this part of Bancroft’s table to be
essentially correct, though it had probably been made for the wrong reasons, as Bancroft never
recorded any South Salopian species from the Shelve Costonian and his collection contained no
such material. Bancroft did record a trinucleid assemblage of Broeggerolithus broeggeri (Bancroft)
and Sa/terolithus from the Hagley Shale Member as being indicative of the Soudleyan Stage, to
which subdivision he also assigned the Whittery Volcanic Member, but the remainder of his
succession showed the Whittery Shale Member ranging in age from Soudleyan to Onnian and he
did not recognize that most of the Caradoc Series is missing from the Shelve Inlier.

Some of Whittard’s 1952 conclusions were subsequently revised by him and in the first part of
the Shelve trilobite monograph (Whittard 1955: 5) the Whittery Volcanic and Shale Members
were said to extend no higher than the D. mul/tidens Zone and, questionably, the basal D. clingani
Zone, a conclusion more in keeping with evidence from the Caradoc Series of the type area. Ina
later part of the monograph (Whittard 1966 : 297) both the Betton Member and the underlying
Weston Member were placed in the D. murchisoni Zone, thus reverting to a correlation suggested
much earlier (Whittard 1931 : 331).

Correlation of the shelly and graptolitic faunas in the Llandeilo and lowest Caradoc Series
proved difficult and Whittard’s (1966 : 298) final published pronouncement differed from his
earlier interpretations, particularly with reference to the boundary between the Meadowtown and
Rorrington Members and the stratigraphic value of such trilobites as Lloydolithus lloydii (Murchi-
son). The latter was shown to be confined to the Meadowtown Series which, in turn, had pre-
viously been equated with the whole of the Llandeilo Series and the G/yptograptus teretiusculus
Zone. More recent work on the Ordovician of south Wales by Addison (in Williams et a/. 1972 :
35-36) clarified the situation by showing that the Meadowtown Member, with its abundant and
characteritic L. //loydii, corresponds only to the Lower Llandeilo of the type area, and demon-
strated that the base of the Nemagraptus gracilis Zone must be extended downwards to include,
at least approximately, the Middle and Upper Llandeilo Series. We may note here that a revision
(Dean 1967: 317-319) of the correlation between the Caradoc stages and corresponding graptolite
zones placed the highest Costonian of the Caradoc in the D. multidens Zone by analogy
with the Shelve succession, while the underlying Costonian strata contain a shale band yielding
Nemagraptus gracilis. At Shelve the Spy Wood Member, at least partly Costonian in age, is
underlain by shales of the Rorrington Member, a subdivision said by Whittard (1955: 5) to
include both the N. gracilis and D. multidens (in part) Zones, so that correlation of shelly and
graptolitic faunas in this part of the succession is not yet precisely defined.

The Ordovician rocks of the Breidden Hills, c. 9 miles (14:5 km) NNW of Shelve, were described
by Watts (1925 : 340, 343) and later by Wedd (1932), who erected a number of local lithostrati-

SHELVE INLIER 6]

graphic names, some of which may represent strata duplicated by faulting. The relatively re-
stricted succession commences with a series of ‘Lower Shales’, base not seen but thickness at least
1770 ft (540 m), which are known also as Nemagraptus Shales as they contain the eponymous
genus and other graptolites suggesting an assignment to the N. gracilis Zone, and thus to the
lowermost Caradoc or Middle/Upper Llandeilo Series. The shales are succeeded by the so-called
‘Black Grit’, a flaggy, black sandstone 20 ft (6 m) thick which corresponds in position approx-
imately to the Spy Wood Member of Shelve and part of the Hoar Edge Grits of the Caradoc
district. The sandstone is followed by a sequence 1870 ft (570 m) thick comprising flags and
shales with two groups of volcanic rocks, which proved partially fossiliferous and from which
Standring (Dean in Williams et al. 1972: 41) obtained the Lower Soudleyan trilobites Broeg-
gerolithus broeggeri and Salterolithus caractaci. This record confirms to a large degree Watts’
(1925 : 343-344) correlation of the Breidden volcanics with the Hagley and Whittery rocks of
Shelve, though Whittard (1952 : table 3) at one time regarded them as being somewhat older.

Only three-quarters of a mile (1:2 km) east of the north end of the Shelve Inlier lies the small,
isolated area of Ordovician rocks beside the eastern flank of Pontesford Hill (Map). Redescribed
by Dean & Dineley (1961), these are known now to include a basal conglomerate which rests with
marked unconformity on Pre-Cambrian (Uriconian) rocks and is followed successively by grey
shales with Diplograptus multidens and then grey-green mudstones containing both graptolites
of the D. mu/tidens Zone and trilobites of the Lower Soudleyan Stage. These last two sets of strata
correspond respectively to the Harnage Shales sensu stricto of the northern Caradoc district and
the Glenburrell Beds of the Onny Valley. The rocks bear a remarkably close resemblance to those
of the Caradoc district but not to those of the Shelve Inlier, from which they are separated by an
important structural line—the Linley—Pontesford Fault—and it is this fault-system rather than
the Church Stretton fault-system which marks the most profound change in development of the
Ordovician rocks in Salop (Dean 1964 : 37; in Williams et al. 1972 : 39).

Intrusive igneous rocks

Lapworth & Watts (1894 : 337-343) divided the igneous rocks of the Shelve Inlier into two groups,
namely ‘the interbedded andesitic ashes and lavas of Arenig and Bala ages and the intrusive
dolerites and picrites’. The latter, both dykes and sills, were said to consist chiefly of hypersthene
dolerite and were described as being of post-Llandovery age as they ‘come into contact with and
somewhat alter the Pentamerus limestones’. This last view, reiterated by Watts (1925 : 335), was
subsequently shown by Whittard & Blyth (in Blyth 1944: 178) to be untenable and the Shelve
volcanics and intrusives are generally accepted as predating the local Upper Llandovery strata.
In a very brief account Watts (1905: 177) noted that the Shelve vulcanicity occurred in two
maxima (one of Arenig, one of Bala age) and that the rocks were ‘almost exclusively andesites in
the form of lavas and tuffs’, some of which were said to be fossiliferous, having been deposited
under marine conditions. The last stages of the vulcanicity were stated to have been ‘marked by
the intrusion of sills, laccolites and dykes, along surfaces of movement due to folding, faulting
and torsion’. A later description by Watts (1925 : 341-343) placed the two principal volcanic
outbursts in the Llanvirn (Stapeley Volcanic Member) and Caradoc (Hagley and Whittery
Volcanic Members), and the water-deposited andesitic tuffs and ashes of the former were stated
to have been ‘invaded by massive sills of andesite and by dolerites of the Llanfawr type as well
as by the normal type’.

The principal paper dealing with this aspect of the Shelve Inlier is that of Blyth (1944), who
showed that the basic intrusions, ranging from picrite to alkali-rich andesite, formed a single
co-magmatic suite having affinities with rocks of plateau-basalt or olivine-basalt type. Some of
Blyth’s work was carried out in collaboration with Whittard and the results of his mapping of the
dolerites of Squilver Hill and Corndon Hill, as well as of several smaller intrusives and dykes, is
evident in the present work, though some of the geological boundaries were subject to later, minor
revision. Watts (1925 : 354; in Whittard 193la: 340) had earlier expressed the view that the
Corndon intrusion was ‘a laccolite intruded along an anticline which has sagged in the middle’, an
interpretation modified slightly by Blyth (1944 : 178-183; see also Whittard 1952 : 164), who

62 W. F. WHITTARD

considered the rock to be ‘a faulted phacolith’ intruded into the crest of a pre-existing anticlinal
axis and having a thickness not necessarily more than c. 200 ft (61 m).

Although almost all Whittard’s notes on the igneous intrusives are confined to brief field
observations, a somewhat longer section deals with the andesite of Lan Fawr (Map and Figs 23,
p. 34; 24, p. 36). The outline of the andesite, which appears to show little petrological variation,
was mapped in detail and a surprising feature is that the shales into which it is intruded are little
altered, show no spotted condition, and are at the most only slightly indurated. This andesite
therefore has less effect on the country rock than the Corndon dolerite and many of the minor
doleritic intrusions, which convert the rock at least into an adinole. The other feature which
emerged gradually during mapping is the post-faulting age of the andesite. No faulted outcrop of
the andesite can be proved, and yet well-established faults in the local sedimentary rocks trend
towards the andesite and might well emerge on the other side. If this claim is justified, and no
other answer has been found to explain the outcrops of the sedimentary rocks, the intrusion of the
andesites was probably almost the last phenomenon to occur in the Ordovician history of the
inlier. Certainly dolerites were intruded along fault planes but some, such as that of Corndon, are
evidently cut by faults. The dolerites may thus be claimed as approximately equivalent in age to
most of the faulting, though sometimes slightly older and often slightly younger. The andesite,
however, is unaffected by faulting and there is no known case where andesite and dolerite are in
juxtaposition. Nevertheless, the Lan Fawr andesite is jointed, and in places shattered, by tectonic
events of uncertain age but almost certainly pre-Upper Llandovery. Excluding tectonics, there is a
certain amount of mineralization with galena and this mineralization may be the last Ordovician
event of which we have any record, the intrusion of the andesite being the penultimate one.

Silurian rocks

The stratigraphy and distribution of the Llandovery rocks in South Salop were the subject of early
researches by Whittard whose 1932 account gave maps and descriptions of the outcrops around
the Shelve Inlier. A review by Ziegler et al. (1968 : 742) introduced three new lithostratigraphic
names (Bog Quartzite, Venusbank Formation and Minsterley Formation) for Whittard’s more
generalized terms and claimed that their “only major re-interpretation’ of his work had been to
connect the so-called ‘Venusbank outlier’ (Whittard 1932 : pl. 61A) with the large outcrops lying
to the north in the Hope Valley. During his later mapping of the area Whittard had, in fact, come
to the same conclusion and a revised boundary for these Silurian rocks at Venusbank was
published by him (1955 : fig. 1) as part of an outline map of the Shelve Inlier which matches that
of the present work (Fig. 15, p. 25), though he did not draw attention to the fact. A further
minor modification, described on p. 272 of the manuscript notes, involved the outlier of Silurian
rocks at Round Hill, 1900 ft (579 m) WSW of Tankerville Mine. The small area described
originally by Whittard (1932: 878, fig. 2) was subsequently extended, particularly to the NW,
and the revised boundary is shown in the Map.

The several outliers approximately centred on Ritton Farm (Fig. 22, p. 33) are difficult to
map in detail and the indication of a particular outline is adduced from the abundance of flags of
sandstone which are unmistakable. In some cases there may be no solid rock left as a cap to the
hill, but the very presence of a marked hill composed of soft shale is indicative of a protective
covering which is still retained or has been removed only recently.

Geological structure of the Shelve Inlier

Perhaps the most obvious structural features of the inlier are the Ritton Castle Syncline and,
immediately to the west of it, the Shelve Anticline. The axes of both these folds run approximately
NNE and in general the regional strike of the Ordovician members follows suit. The existence of
these structures has long been appreciated and was noted by Lapworth & Watts (1894 : 314, 317)
whose then unpublished map (used later in Pocock & Whitehead 1948: pl. 11) shows them
clearly. A later account by Watts (1925 : 343) stated: ‘In addition to this folding [the syncline
and anticline] the rock structures are often cut by faults, the more important of which are
longitudinal but tend to swing across the strike’. He noted that ‘combined folding and faulting

SHELVE INLIER 63

has produced remarkable outcrops in the Stapeley Stage at Bromlow Callow and the Hagley
Stage near Wotherton’, and observed that the faults often coincided with dolerite and picrite
intrusions or, in the case of small transverse faults, with mineral-bearing lodes.

Unfortunately the Whittard manuscript notes contain no comprehensive account of the
geological structure, allusions to which are confined to a few scattered comments. However,
some of Whittard’s main conclusions were expressed in his A geology of South Shropshire (1952 :
185-189), in which paper he noted particularly the ‘powerful northerly-trending tear-faults which
were accompanied by complementary shears running not quite at right-angles to one another
and each making an angle usually in excess of 45 degrees with the main tear-faults’. Whittard
considered that his recognition of the tear faults (which may readily be seen in the Map) provided
the main difference between his and earlier interpretations of the structure of the Inlier. Examina-
tion of his map shows that a few of the tear faults coincide approximately with less extensive
dislocations shown on the map of Lapworth & Watts (in Pocock & Whitehead 1948: pl. 11),
and it is likely that the latter structures are the longitudinal faults noted by Watts in 1925 (see
p. 62). The main folding and faulting were ascribed by Whittard (1952 : 186-187) to the Taconian
Orogeny and are certainly of pre-Upper Llandovery age. As far as the Shelve Inlier is concerned,
they are no older than the Soudleyan Stage of the Caradoc Series, and may well be post-Caradoc
in view of the lack of evidence for disturbances of this magnitude up to the top of the Onnian
Stage in southeast Salop.

One can really only confidently determine the northerly-trending tear fault pattern by being
able to feel sure of the relative stratigraphical ages of beds of somewhat similar but not identical
lithologies when they are brought into juxtaposition; for example, Aldress, Rorrington, Meadow-
town and Betton Members and some beds of the Weston Member. Although by no means the
same, their lithologies in small exposures may be similar enough to cause much indecision regard-
ing which subdivision they belong to. Consequently knowledge of the faunal sequence was of
great importance in confirming the lithostratigraphic assignment of many localities. The more
obvious faults are the complementary shears because they run across the strike, displacing ridges
formed of hard rocks, instead of running with the strike as do the major tear faults, in regions
occupied by softer rocks. Without the aid of stratigraphic palaeontology the pattern of faults
developed through shear might well not have been recognized, with regard to the dominant
northerly-trending members.

A major feature of South Salop geology, emphasized by Whittard (1952 : 185) and reiterated in
a later review by Dineley (1960 : 6), is the large number of unconformities, often of major pro-
portions, which range in age from Pre-Cambrian to Mesozoic. These are particularly noticeable
in the Caradoc district, some 7 miles (11-3 km) to the east, where beds of only Caradoc age are
found, resting with profound unconformity upon Pre-Cambrian rocks and succeeded unconform-
ably by Upper Llandovery strata. As far as the Shelve Inlier is concerned, however, few breaks
are documented. Following the probably relatively small disconformable gap between the
Stiperstones Member and Shineton Shales (the base of which is not seen here), the Ordovician
succession continued virtually unbroken, except by periodic outbursts of volcanic activity, from
the Lower Arenig Series to the Lower Soudleyan Stage of the Caradoc Series. The Ordovician
rocks of the Caradoc district were deposited along the margin of the so-called Midland Block
and were separated by the line of the Linley-Pontesford Fault from the subsiding region farther
west in which the Shelve rocks were deposited and which was connected with other areas of marine
deposition in southern Europe, including especially Bohemia.

References

Bancroft, B. B. 1933. Correlation tables of the stages Costonian—Onnian in England and Wales. 4 pp.,
3 tables. Blakeney, Glos. (privately printed).

Blyth, F. G. H. 1938. Pyroclastic rocks from the Stapeley Volcanic Group at Knotmoor, near Minsterley,
Shropshire. Proc. Geol. Ass., London, 49 : 392-404, figs 66-68.

— 1944. Intrusive rocks of the Shelve Area, West Shropshire. Q. J/ geol. Soc. Lond., 99 : 169-204,
pl. 20.

64 W. F. WHITTARD

Bulman, O. M. B. 1928. A monograph of British dendroid graptolites, 2 : i-xxxii, 29-64, pls 3-6. Palae-
ontogr. Soc. (Monogr.), London.

Callaway, C. 1878. On a new area of Upper Cambrian rocks in South Shropshire, with a description of a
new fauna. Q. J/ geol. Soc. Lond. 33 : 652-672.

Dean, W. T. 1964. The geology of the Ordovician and adjacent strata in the southern Caradoc District
of Shropshire. Bull. Br. Mus. nat. Hist., London, (Geol.) 9 : 257-296, 2 pls.

1967. Relationships of the Shelve trilobite faunas. Jn Whittard, W. F., The Ordovician trilobites of
the Shelve Inlier, West Shropshire, 9 : 308-320, fig. 10. Palaeontogr. Soc. (Monogr.), London.

— & Dineley, D. L. 1961. The Ordovician and associated Pre-Cambrian rocks of the Pontesford
District, Shropshire. Geol. Mag., London, 98 : 367-376, pl. 20.

Dineley, D. L. 1960. Shropshire geology: an outline of the tectonic history. F/d Stud., London, 1 : 1-23,
10 figs.

Dines, H. G. 1958. The West Shropshire mining region. Bull. geol. Sury. Gt Br., London, 14 : 1-43.

Earp, J. R. & Hains, B. A. 1971. British Regional Geology. The Welsh Borderland. (3rd ed.) xi+118 pp.,
11 pls. London.

Hall, T. C. F. 1919. Report on the Shropshire Mining District. 35 pp., 3 tables. London (privately printed).

Harper, J. C. 1947. Tetradella complicata (Salter) and some Caradoc species of the genus. Geol. Mag.,
London, 84 : 345-353, pl. 10.

James, J. H. 1956. The structure and stratigraphy of part of the Pre-Cambrian outcrop between Church
Stretton and Linley, Shropshire. Q. J/ geol. Soc. Lond., 112 : 315-337, pl. 15.

Lapworth, C. 1887. The Ordovician rocks of Shropshire. Rep. Br. Ass. Advmt Sci., London, 56 (Birm-
ingham 1886), C : 661-663.

— 1916. In Mem. geol. Surv. Summ. Prog., London, 1915 : 36-38, 3 figs.

—— & Watts, W. W. 1894. The geology of South Shropshire. Proc. geol. Ass., London, 13 : 297-355,
pls 8, 9.

— —— 1910. Shropshire. Jn Monckton, H. W. & Herries, R. (eds), Geology in the Field (4): 739-769,
pls 24, 25. London (Geologists’ Association Jubilee Volume).

Pocock, R. W. & Whitehead, T. H. 1948. British Regional Geology. The Welsh Borderland. (2nd ed.)
iv+80 pp., 11 pls. London.

——, ——, Wedd, C. B. & Robertson, T. 1938. Shrewsbury District, including the Hanwood Coalfield
(One-inch Geological Sheet 152 New Series). xii+297 pp., 8 pls. Mem. geol. Surv. U.K., London.

Salter, J. W. 1866. On the fossils of North Wales. Jn Ramsay, A. C., The geology of North Wales.
Mem. geol. Surv. U.K., London, 3 : 239-381, pls 1-26.

Stubblefield, C. J. & Bulman, O. M. B. 1927. The Shineton Shales of the Wrekin District: with notes on
their development in other parts of Shropshire and Herefordshire. Q. J/ geol. Soc. Lond., 83 : 96-146,
pls 3-5.

Watts, W. W. 1905. On the igneous rocks of the Welsh Border. Proc. Geol. Ass., London, 19 : 173-183,
oll, Se

—— 1925. The geology of South Shropshire. Proc. Geol. Ass., London, 36 : 321-363, figs 26-40.

Wedd, C. B. 1932. Notes on the Ordovician rocks of Bausley, Montgomeryshire. Summ. Progr. geol.
Sury. Lond., 1931 (2) : 49-55, 1 fig.

Whittard, W. F. 1931. The geology of the Ordovician and Valentian rocks of the Shelve country, Shrop-
shire. Proc. Geol. Ass., London, 42 : 322-339, pls 10, 11.

—— 193la. Easter Field Meeting (Extension) to Ministerley. April 8th to 11th, 1931. Proc. Geol. Ass.,
London, 42 : 339-344.

1932. The stratigraphy of the Valentian rocks of Shropshire. The Longmynd-—Shelve and Breidden

outcrops. Q. JI geol. Soc. Lond., 88 : 859-902, pls 58-62.

1940. The Ordovician trilobite faunas of the Shelve-Corndon District, West Shropshire. Part I.
Agnostidae, Raphiophoridae, Cheiruridae. Ann. Mag. nat. Hist., London, (11) 5: 153-172, pls 5, 6.

— 1952. A geology of South Shropshire. Proc. Geol. Ass., London, 63 : 143-197, 6 tables.

1953. Report of Summer Field Meeting in South Shropshire, 1952. Proc. Geol. Ass., London, 64:

232-250, | fig.

1955. The Ordovician trilobites of the Shelve Inlier, West Shropshire, 1 : 1-40, pls 1-4. Palaeontogr.

Soc. (Monogr.), London.

1958. Geology of some classic British areas: Geological Itineraries for South Shropshire. Geol. Ass.

Guide, Colchester, 27 : 1-24, 14 figs.

—— 1960. England, Wales & Scotland. Ordovician. Lexique Strat. int., Paris, 1 (3alV) : 1-296.

1960a. The Ordovician trilobites of the Shelve Inlier, West Shropshire, 4 : 117-162, pls 16~21.

Palaeontogr. Soc. (Monogr.), London.

SHELVE INLIER 65
— 1966. The Ordovician trilobites of the Shelve Inlier, West Shropshire, 8 : 265-306, pls 46-50.

Palaeontogr. Soc. (Monogr.), London.

Williams, A. 1974. Ordovician Brachiopoda from the Shelve District, Shropshire. Bull. Br. Mus. nat. Hist.,

London, (Geol.) Suppl. 11 : 1-163, 28 pls.

— etal. 1972.A correlation of Ordovician rocks in the British Isles. Spec. Rep. geol. Soc. Lond., 3: 1-74,

10 figs.

Ziegler, A. M., Cocks, L. R. M. & McKerrow, W. S. 1968. The Llandovery transgression of the Welsh
Borderland. Palaeontology, London, 11 : 736-782, 14 figs.

Index

The page numbers of the principal references are in bold type; an asterisk (*) denotes a figure. Page
numbers in brackets refer to place-names appearing on the small diagrammatic maps.

Aldress Dingle 50-2 (44, 48)
Aldress Member 10, 34, 42-4, 47-9, 50-1, 52, 55,
59, 63
Aldress Shales 8, 49-50
Alport 7, 54
Ampyx linleyensis 9
salteri 23
Apatokephalus serratus Zone 15
Ape Dale (3)
Arenicolites linearis 15
Arenig 1, 5, 8, 15, 61
Ash Cottage 26-7
Aylesford Brook 7

Bala 61
Bank (52)
Farm 27, 57 (25, 48, 52)
Barrandia 27
Bentlawnt (26)
Bergam (5)
Quarry 19
Bergamia rhodesi 9, 19-20
Berwyn 5
Betton 39-40, 45 (5)
Beds 41
Dingle 41 (40)
Hill 40*
Member 10, 34, 36-7, 40, 41-2, 43-5, 47-8,
59-60, 63
Shales 38
Bettonia 9, 41
Big Cuckoo Nest 51
Black Hill 33
Black Marsh 27
Black Rhadley 11; see Rhadley
Blakemoorgate 11
Blue Barn (28)
Bog, The 4, 11, 13*, 17, 19, 27-8, 33* (5, 33)
brachiopods 4
Breidden Hills 59-60
Brithdir 27 (5, 30)
Brockton 27
Broeggerolithus broeggeri 9, 52, 54, 59-60
globiceps, longiceps, nicholsoni, soudleyensis,
transiens, ulrichi 9

Bromlow 40

Callow 4, 30, 63 (40)
Bromlowhall Farm (40)
Brooks Hill 33-4 (33)

Marsh (33)
Brynkin Dingle 48-9, 51 (44)
Brynkin Green 49 (44)
Bunter Sandstone 5
Buxton Hill 28 (13)

Cader Idris 5
Caebitra, River (5)
Calcot 54, 58 (34, 48)
Callow Hill 21, 28 (5, 21)
Callow Wood (26); see Bromlow Callow
Calyptaulax actonensis 9
Cambrian 10
Camlad, River 4, 7, 11, 51-2, 54, 57-9 (5, 44, 48)
Caradoc Series 1, 8, 14, 50-2, 54, 59-61, 63
Cefn Gunthly (Gwynlly), Cefngunthly 19, 32-3
(5, 32)
Chapel (20, 34)
Cheshire Plain 5
Chirbury 4, 7-8, 29, 49-53, 57-9 (5, 52)
Formation 10, 49-59
Series 38, 49
Chondrites sp. 13
Church Stoke 4, 7-8, 27, 39, 41-2, 44, 49, 52, 54,
55* (5, 55)
Church Stretton 5, 61 (3)
Clun, Clun Forest (3)
Cnemidopyge granulata 46
Coal Measures 7, 11
Coed Brook 43*, 45, 47, 49-50 (43)
Contorted Ash Quarry 26
Corndon Hill 4, 27, 29*, 30, 61 (3, 5)
Corrugatagnostus 28
Corve Dale (3)
Costonia ultima 9, 50, 60
Cranberry Rock 12
Craven Arms (3)
Crest Cottage (53)
Crest Wood 51 (53)
Cross House 7

66 W. F. WHITTARD

Crowsnest 20, 28 (18) Heath Mynd 11-12, 19, 33 (5)
Dingle 19-20 (18) Heightley 59 (52)

Cruziana semiplicata 13 Barn 56

Cryptolithus inopinatus 9 Wood (53)

Cwm Dingle 39 High Plantation (28)

Cwm-Dwla (Cwmdulla, Cwmdwla) 39 (34) Hoarstone (55)

Cwm Mawr 27, 30 (30) Farm 39

Hoar Edge Grits 60-1

Deadman’s Dingle 42, 45, 48 (44) Hockleton (53)

Dendrograptus 24 Hollies, The 13, 27 (20, 31)

Desert (42) Holywell Brook 35, 37*, 39, 41, 46 (37, 47)
Brook 44 Hope 4, 26* (3, 5)

Desmograptus 24 Brook 27-8

Devil’s Chair 11-13 (5, 13) Common (5, 26)

Dicellograptus 45-6, 48 Group 10

Dicranograptus clingani (& Zone) 9, 59-60 Member 10, 13, 17-23, 24-9, 30-3, 36-7, 40, 59

Dictyonema cobboldi 22 Quarry (5)
fluitans 51 Rectory 26 (25)
irregulare 22 Shales 8, 16, 24

Didymograptus 22-3 Valley 23-4, 25*, 27, 62 (25, 26)
bifidus (& Zone) 9, 25, 59 Hurdley 36
extensus (& Zone) 9, 24, 59 Hyssington 30*, 31*, 32 (5, 31)
hirundo (& Zone) 9, 19, 59 Ash 29
murchisoni (& Zone) 9, 39, 41, 59-60

Dingle House (48) Illaenopsis 27

Diplograptus multidens (& Zone) 9, 49-51, 59-61 intrusive igneous rocks 61-2

Disgwylfa (Squilver) Hill 11, 27, 31 (5, 31)

Easn@nuy Riven) Josey’s Wood 28 (18)

Eastridge 19 (20)

Wood (20) Kingswood 52 (48)
Ectillaenus 19 Kinton 50
hughesi 26 Knotmoor Plantation 31 (26, 28)
Fremes Wood 32 (31) Ladywell Grits & Flags 16, 18-19
Ladywell Group 10-11
Ganderbeach (40) Ladywell Mine 22-3 (22)
Geragnostus 19 Lan Fawr 29-30, 34*, 36, 39, 61-2 (34)
Glenburrell Beds 59, 61 Leigh Hal! 28*, 30-1, 36, 39-40 (5, 28)
Glyptograptus dentatus 24 Leigh Manor 28*, 30-1, 34 (5, 26)
teretiusculus Zone 9, 59-60 Leptograptus 47
Gorstybank (18) Beds 45
Granham’s Moor 11, 14-15 Lingulella nicholsoni 15
Farm 10, 15 Linley 6 (5)
graptolites 4, 14 Hill (3, 5)
Green Hill 17-18 (13, 18) Linley—Pontesford Fault 6, 61, 63
Grey Grass Dingle 46—7 (47) lithostratigraphic terminology 6-7*, 9*
Grinshill 5 Little Aldress (48)
Little Cuckoo Nest 51
Habberley (5) Little Weston 41, 43*, 45, 47 (36, 43)
Brook 7 Llandeilo Series 8, 38, 60-1
Shales 1, 6, 8, 10-12 Llandovery 38, 61-3; see Upper Llandovery
Valley 10 Llanerch 27 (30)
Hagley 47, 51-2, 60 (48, 52) Llanvirn Series 8, 25, 61
Shales Member 7, 10, 44, 48, 52, 53-4, 55, 58, 60 Lloydolithus lloydii 9, 43, 45, 60
Volcanic Member 5, 10, 17, 44, 48, 50, 51-3, Long Leasow (55)
54-5, 60-1 Long Mountain (3)
Hanwood coalfield 7 Longmynd 5-6, 10 (3)

Hawkstone 5 Longmyndian 14

SHELVE INLIER 67

Lord’s Hill (Lordshill) 12, 19-20 (5)
Beds 16, 18-20
Farm (20)
Rough (20)
Lower Aldress 47, 50-1 (34, 48)
Lower (Hyssington) Ash 29
Lower Ridge 43*, 50 (43)
Lower Santley 35-6 (25)
Lower Silurian 3-4, 6, 19, 36, 40, 44, 46
Lower Soudleyan Stage 51-2, 54, 61, 63
Lower Wood 41, 46
Brook 42*, 44, 46 (42)
Farm (42)
Luckley Barn (26, 28)
Luckley Hill 30-1, 40* (5, 40)
Ludlow (3)
Lyde 40 (5)
Cottage (40)
Lydham (5)

Maddox’s Coppice 21, 28 (5, 20)
Manstone Rock 12*, 13
Marrington 48*, 54, 57-8 (5, 48)
Dingle 7, 54, 56-7
Hall (48, 52)
Marrolithoides arcuatus 9, 46-8
Marrolithus 60
bilinearis 47
Marsh, The 27
Marsh Pool (22)
Marton (5)
Meadowtown 30, 40-1, 42*, 43, 45 (5, 42)
Beds 38, 43
Member 10, 34, 36-7, 40-2, 43-5, 46-8, 59-60,
63
Series 43, 49
Mesozoic 63
Middle (Tashkar) Ash 29
Middleton 38, 45 (5, 36)
Formation 10, 38-49
Hill 39 (36)
Vicarage (47)
Midland Block 63
Minsterley 17, 21, 25-6, 29, 62 (5)
Monobolina plumbea 22, 24
Mytton Batch 16*, 17-18 (5); see Myttonsbeach
Mytton Flags 8, 16, 24
Mytton Group 10, 16, 24
Mytton Member 10, 12-15, 16-24, 18*, 26-8,
32-3, 59
Mytton Stage 16
Myttonsbeach 13*, 16*, 17-19 (18)
Myttonia confusa 9, 18, 20

Nemagraptus 47-8
Beds 45
gracilis Zone 9, 46, 59-61
Shales 61

Neseuretus 18, 23
grandior 9, 15
parvifrons 20

Nesscliff 5

New House (34)

Nill’s Hill 11-14 (21)

Nind 32* (5, 32)
Wood 32

Nipstone Rock 12

Oak Hill 17 (5, 18)
Oakedge 26 (23, 25)
Ogyginus 41-2
corndensis 38
Ogygiocarella debuchii 43, 45-8
Ogygiocaris 19
selwynii 18, 20, 22-4
Oldgrit 22 (22)
Onnia? cobboldi, gracilis, superba 9
Onnian Stage 60, 63
Onny River, see East Onny, West Onny
Valley 59, 61
Orthis 41
Overton’s Rough 26*, 41 (26, 28)
Ox Wood 51

Paddock, The 12 (18)
Park, The (26)
Pellrhadley Hill (32)
Pennerley 17, 19, 28, 36 (13)
Mine (5)
Pentamerus limestones 61
Perkins Beach 18-19 (13)
Perkins Level 20-1
picrite 30
Pitcholds 27
Placoparia 19, 26
Platycalymene duplicata 47
Platycoryphe vulcani 9
Platylichas laxatus 9
Pleurograptus linearis (& Zone) 9, 59-60
Plynlimmon 4
Poles Coppice 15 (21)
Polesgate (21)
Pontesbury 11, 17, 20, 21* (3, 5, 21)
Pontesford 6, 14, 59
Hill 10, 11*, 61 (5, 21)
Poundbank 8
Pre-Cambrian 6, 14, 61, 63
Pricyclopyge 19, 26-7
binodosa 26, 28
Priestweston 4, 34*, 35, 36%, 38-9, 41, 45-6, 52
(5, 34)
Primaspis 45
whitei 46, 48

Rea Brook 7 (5)
Redonia anglica 20, 22
Reuscholithus reuschi 9

68 W. F. WHITTARD

Rhadley (32); see Black Rhadley
Rhiston 7
Ritton Castle 4, 28, 33* (33)
Syncline 17, 21, 27, 29, 31-2, 34-6, 62
Ritton Farm 28, 62 (33)
Rockabank 55-6 (53)
Rock Coppice 55, 59
Rock House (44)
Roman Mine 22-4 (23)
Rorrington 4, 29, 37*, 41, 44-6, 47*, 50-1, 54
(5, 37, 42)
Hall (47)
Hill 35, 39-41 (37)
Covert (37)
Lodge (47)
Member 10, 34, 36-7, 42-4, 45-9, 50, 59-60, 63
Shales 38, 45
Round Hill 30, 36 (5)
Roundwood Cottage (30)
Roveries Wood 19, 31 (31)
Runnis Bridge 32 (32)

Salterolithus caractaci 9, 52, 54, 59-61
Santley (5)
Wood 24 (23)

Scattered Rock 11 (13)

Severn, River (3)

Shelfeld (Farm) 24 (23)

Shelve 4s 14175224) O3\*5 34=5. 389 (355, 22,23)
Anticline 12, 17, 21, 25-7, 29, 35-6, 39
Church Beds 16, 24
Formation 10-38
Hill 22-3, 27
Series 38

Shepherd’s Rock 11

Shineton Shales 1, 6, 10-13, 15, 18, 20-1, 32, 63

Shrewsbury (3)

Shumardia pusilla 15

Silurian 5, 7, 20, 26-8, 30-1, 33, 40, 42, 62; see

Lower Silurian

Snailbeach 11, 13, 15, 16*, 19, 20*, 28 (3, 5, 20)
Coppice 20-1 (20)

Flags 16
Grits 16, 19

Snead 19, 27, 36 (5)

Soudleyan Stage 60, 63; see Lower Soudleyan

Spirantyx calvarina 9, 46-8

Spring Coppice (53)

Spy Wood 44* (44)

Brook 42, 45, 48-50 (44)
Cottage 49 (48)
Member 10, 34, 36-7, 42-5, 47-8, 49-50, 51,
53, 59-61
Squilver 29 (22)
Hill 31-2, 61; see Disgwylfa Hill

Stapeley Group 10

Stapeley Hill 29-30, 35 (5, 36, 37)

Stapeley Shale Member 10, 26, 28-9, 31-4, 35-8,

39-40, 59

Stapeley Volcanic Member 5, 10, 17, 22-3, 26,
28, 29-35, 36-8, 40, 59, 61
Stapeleyella 28
inconstans 9
Stiper Group 10-11, 29
Stiper Quartzites 8, 11, 16
Stiperstones 4, 11, 14*, 17, 24, 27 (3, 5, 13)
Member 5, 10, 11-16, 12*, 17-21, 27, 32-3, 59,
63
Quartzite 8, 11, 16

Taconian Orogeny 63
Tankerville 17
Flags 8, 16, 18, 20, 24
Hollow 14*, 17, 19 (13)
Mine 19, 62 (13)
Tasgar (Tasker) 32* (5, 32); see Middle Ash
Quarry (32)
Tetradella salopiensis 50
Tetragraptus 19
Todleth Ash 29
Todleth Hill 4, 30, 36, 39, 55* (5, 55)
Todlith House (55)
Tremadoc 1, 8, 10, 12, 15
Trimberth 59
Trinucleus 41
acutofinalis 9

Upper Aldress 39, 45 (44)
Upper Brynkin 49 (44)
Upper Grwarhlow 7
Upper Llandovery 50
Upper (Todleth) Ash 29
Upper Yessons 11
Uriconian 5, 14, 61

Valley Knoll 35 (27)
Venusbank 35, 38, 62 (5, 25)

Walkmill 55 (52)

Walk Mill Quarry 55-7

Welsh Borderland 3*, 8

Welshpool (3)

Wenlock Edge (3)

Wentnor (5)

Weston Flags 38

Weston Member 10, 18, 26, 28, 34-7, 38-40,

41-4, 47, 55, 59-60, 63

West Onny, River 4, 7, 12, 19, 27, 32* (5, 32)

Whitegrit 22-3 (5)
Mine (22)

White House (37)

Whitehouse Brook 41

Whitepits 7

Whitsburn Hill (28)

Whittardaspis [Cryptolithus] inopinata 9, 43
sp. 45

SHELVE INLIER 69

Whittery 52*, 54-9 (52)
Bridge (52)
Shale Member 10, 48, 52-4, 58-9, 60
Volcanic Member 5, 8, 10, 44, 48, 52-3, 54-8,
56*, 57*, 58*, 59-61
Wood (52)
Wilmington 50-1

Woodgate (30)

Worthen 36 (5)

Wotherton 4, 51, 53*, 54-7, 59, 63 (3, 5, 53)
Wrekin 5, 10

Yew Tree Level 20 (20)

Accepted for publication 14 June 1978

Sop, ch ARN
wi goo > 1.
a Py Wie @ btmy 2

rn

= 2

+ : “ @

_
a

ou

Bh ate ty of

at), -a wae oe

Salop and part of north hee By the ate gt V7]
(Compiled by W. T. Dean) | ek og

Miscellanea

The Caradoc faunal associations of the’ area
Dinas Mawddwy, north Wales. By M. G.

Fossil insects from the Bembridge Ma
Wight, southern pees By, E. Re

ulletin of the
ritish Museum (Natural History)

‘series Vol 33.No2 31 January 1980

The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four
scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology, and
an Historical series.

Papers in the Bulletin are primarily the results of research carried out on the unique and
ever-growing collections of the Museum, both by the scientific staff of the Museum and by
specialists from elsewhere who make use of the Museum’s resources. Many of the papers are
works of reference that will remain indispensable for years to come.

Parts are published at irregular intervals as they become ready, each is complete in itself,
available separately, and individually priced. Volumes contain about 300 pages and are not
necessarily completed within one calendar year. Subscriptions may be placed for one or more
series. Subscriptions vary according to the contents of the Volume and are based on a forecast
list of titles. As each Volume nears completion, subscribers are informed of the cost of the
next Volume and invited to renew their subscriptions. Orders and enquiries should be sent to:

Publications Sales,
British Museum (Natural History),
Cromwell Road,
London SW7 5BD,
England.

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

© Trustees of the British Museum (Natural History), 1980

ISSN 0007-1471 Geology series

Vol 33 No 2 pp 71-164

British Museum (Natural History)
Cromwell Road

London SW7 5BD Issued 31 January 1980 i

Miscellanea

Contents
A new, possibly algal, microproblematicum from the Lower Carboniferous of England.
By G. F. Elliott

Acanthopleurella Groom 1902: origin and life-habits of a miniature trilobite. By R. A.
Fortey & A. W. A. Rushton ; : ‘

Pleistocene bird remains from Tornewton Cave and the Brixham Windmill Hill Cave in
south Devon. By C. J. O. Harrison

The succession of Hyracotherium (Perissodactyla, Mammalia) in the English early Eocene:
By J. J. Hooker

Salenia trisuranalis sp. nov. (Echinoidea) from the Eocene (London Clay) of Essex, and
notes on its phylogeny. By D. N. Lewis & R. P. S. Jefferies .

Tertiary and Cretaceous brachiopods from Seymour, Cockburn and James Ross Islands,
Antarctica. By E. F. Owen

Revision of the rugose coral Diphyphyllum concinnum Lonsdale, 1845 and historical
remarks on Murchison’s Russian coral collection. By B. R. Rosen & R. F. Wise

Neuroptera (Insecta) in amber from the Lower Cretaceous of Lebanon. By P. E. S. Whalley

Page

73

79

91

101

115

1123

147
157

A new, possibly algal, microproblematicum from ule
Lower Carboniferous of England
G. F. Elliott

Department of Palaeontology, British Museum (Natural History), Cromwell Road, London
SW7 SBD

Synopsis

Hypocaustella cartimanduae gen. et sp. nov. is described from the Lower Carboniferous (Viséan D,, D.)
of northern England. It consists of small monostromatic growths of calcareous-walled hexagonal-prismatic
cells, with occasional larger rounded-lenticular cells. The evidence for its possible interpretation as an
early coralline alga, and the evolutionary implications of this, are discussed.

Introduction

The little organism described below was discovered during examination for algae of two Viséan
limestone samples. One is from the Great Scar Limestone of Ingleton, North Yorkshire; the other
from the Oxford Limestone of Dunsall, Northumberland. The former is a biosparite with
subordinate pyrites, rich in small foraminifera and microproblematica, with frequent ostracoda,
occasional larger debris of brachiopods, bryozoa and serpulids, and rare fragments of the algae
Koninckopora and Ungdarella. The other sample is also pyritic and foraminiferal.

Systematic description
Microproblematicum, incerta sedis
Genus HYPOCAUSTELLA noy.

DIAGNOsiIs. Small curved monostromatic growths of tiny hexagonal-prismatic calcified cells;
growths presumed originally attached or partly encrusting, and sometimes showing branching; single
larger rounded-lenticular cells occasionally present. Lower Carboniferous, Viséan D,, D,;
northern England.

TyPE SPECIES. H. cartimanduae sp. nov.

GENERIC NAME. From a fancied resemblance of the cells to the hollow tiles of a Roman hypocaust
or heating system.

Hypocaustella cartimanduae sp. nov.
Figs 1-7

DESCRIPTION. In numerous thin sections this organism appears as thin curved plate-like mono-
stromatic growths, sometimes showing evidence of branching or division. All are under | mm in
diameter, and a majority are 0-5 mm or less. None have been found attached to shell or other
solid objects.

In vertical section the cells are square to high-rectangular. They vary in dimensions between one
specimen and another, the smaller being square, the larger rectangular (Fig. 8). The basal calcifi-
cation is usually well developed; the very thin lateral walls thicken at their outer terminations.
An outer or upper calcification is only occasionally present, though the cells usually have a
different crystalline infilling to that of the matrix. Some examples of apparent cell-roofing may

Bull. Br. Mus. nat. Hist. (Geol.) 33 (2): 73-77 Issued 31 January 1980
73

74 Gabo EELIOEE

possibly be due to diagenesis at the infilling-matrix surface, but some are the horizontal continua-
tion of the lateral cell-walls, as seen in vertical section. In transverse section the cells are uniformly
rounded-hexagonal, giving a honeycomb appearance. Oblique sections through the walls show
no trace of intercommunicatory pores.

Some specimens show single examples of larger cells occurring among the normal cells. They
are rounded-lenticular in outline, and are about three times the diameter of the ordinary cells of
the particular growth in which they occur. In height they are the same or up to one and a half
times that of the ordinary cells (Fig. 8). All these special cells are clearly roofed, unlike many of
the other cells.

TRIVIAL NAME. Commemorates Cartimandua, first-century Romano-British Queen of the Bri-
gantes, from whose former territories the fossil comes.

MATERIAL. 7ypes. British Museum (Natural History), Dept of Palaeontology, registered numbers
V.21205d (Holotype: Fig. 7); V.21205b, V.21205c, V.21205d (Paratypes: Figs 2, 1, 3). From the
Great Scar Limestone (Lower Carboniferous, Viséan D,) of Mealbank Quarry, Ingleton, North
Yorkshire, England.

Other material. Numerous examples in eight further thin sections from the same sample,
V.21205 (H. P. Lewis Coll., 1930). Also examples in four thin sections of the Oxford Limestone
(Viséan D,) of Dunsall Old Quarry, Berwick on Tweed, Northumberland, V.54331 (E. J. Gar-
wood Coll.).

Discussion. This little organism occurs free and often apparently incomplete amongst the for-
aminifera, etc. of the parent rock. None have been seen clearly attached to solid objects; one
doubtful example is probably an accidental juxtaposition. Nevertheless, the morphology suggests
attachment, possibly to non-calcified algae, fixed or floating, presumably in shallow water. The
North Yorkshire rock sample comes froma limestone quarry which shows evidence of emergence
in a thin coal-seam with underclay (Kendal & Wroot 1924 : (I) 59; Hudson & King 1933 : 434),
and the areal stratigraphy (Hudson 1933 : 247) indicates shallow marine conditions compatible
with the habitat suggested. The Northumberland limestone does not conflict with this: it is
similarly rich in foraminifera and microproblematica.

An animal origin, e.g. some lowly-organized colonial growth such as a simple extinct hydro-
zoan, seems unlikely in view of the very small cells and the absence of intercommunicatory pores
through the walls. If, however, the cells are regarded as originally open on the upper or outer
surfaces, the occasional closed examples being ascribed to either diagenesis or original abnorm-
ality, then this could support animal origin of some kind, with a tentative reproductive function
for the special cells. There is a superficial resemblance and approximate size-correspondence to
the Precambrian Papillomembrana (Spjeldnaes 1963, 1967), but this problematicum was probably
carbonaceous not calcareous, cylindrical not plate-like, and the fringing projections are papillae,
not wall-sections of polygonal mesh. If, however, Hypocaustella is considered as a small calcareous
alga, size and lack of pores are in keeping, though there is still some uncertainty.

Monostromatic crusting corallines (Rhodophyta; family Corallinaceae) referable to the genus
Lithoporella Foslie are common at the present day and in the Cenozoic (Lemoine 1974, 1976),
usually attached to solid objects, organic or inorganic. The high-rectangular cells are squarish in

Figs 1-7 Hypocaustella cartimanduae gen. et sp. nov. Lower Carboniferous, Great Scar Limestone
(Viséan D,); Mealbank Quarry, Ingleton, North Yorkshire. Thin sections, «120. V-numbers are
registered numbers, Dept of Palaeontology, British Museum (Natural History). Fig. 1, rectangular
cells, some roofed, in vertical cut; V.21205c. Fig. 2, small-celled individual showing branching;
V.21205b. Fig. 3, polygonal cells in transverse cut; V.21205d (same thin section as holotype).
Fig. 4, part of large-celled individual in vertical cut; V.2120Se. Fig. 5, polygonal cells, slightly
oblique view of walls in the thickness of the preparation; V.21205a. Fig. 6, special cell in centre of
ordinary cells; V.21205f. Fig. 7, holotype. Apparently complete individual showing ordinary cells
(left and right), special cell (top centre); twisting plane of organism gives polygonal view of cells
(top left); V.21205d.

LOWER CARBONIFEROUS MICROPROBLEMATICUM

76 G. F. ELLIOTT

transverse section; the conceptacles are large blister-like objects rising above the thin thallus.
Calcification of the box-like cells is usually complete in the fossils.

In the Mesozoic, Thaumatoporella Pia (Trias—Cretaceous: syn. Polygonella Elliott, Lithoporella
spp.) shows high-rectangular cells which are polygonal in cross-section: the calcification is
sometimes incomplete, especially that of the upper cell-surfaces, but in numerous examples
examined by me this appears to be due to the varying preservation. The conceptacles are rarely
found, but are similar to those of Lithoporella (Emberger 1958). Like Lithoporella, Thaumato-
porella seems to have been attached to solid objects.

Example 1 2 3 4 5 6 7 8 SY i wW 1251S eel 16

Cell—height 30 30 36 40 40 45 45 45 54 54 63 63 72 80 80 108

Cell- width 20 30 36 20 40 30 36 45 36 40 36 36 45 40 45 54

Conceptacle- 45 45 63 63
height

Conceptacle- 63 63 80 100
width

Fig. 8 Tables of cell-dimensions (jm) in different examples of Hypocaustella.

Hypocaustella, then, is possibly to be regarded as a Palaeozoic algal growth of similar coralline
nature from its vegetative thallus, the frequent absence of calcification on the upper or outer cell
surfaces being due to non-preservation of an original thin calcification, and the attachment-
habit being to non-calcified algae. If the larger cells are interpreted as conceptacles, they are clearly
different from the very conspicuous structures seen in Lithoporella and Thaumatoporella. Can
Hypocaustella be regarded as an ancestral monostromatic coralline alga?

Of other possible Palaeozoic claimants for this, the Permian Archaeolithoporella (Endo 1959 :
182) was accepted as such until recently. However, Babcock (1974 : 38) has shown that, although
probably but not certainly algal, it is laminar but not cellular, and certainly not coralline; the
significance of this is emphasized by Wray (1977). The peculiar Litostroma oklahomense (Penn-
sylvanian: Mamay 1959), whilst a monostromatic alga, possibly a rhodophyte, does not show
conceptacles and does not seem to be a coralline. European records of this species, given in
structural detail and reviewed by Homann (1972 : 170), who includes L. europaea Kochansky-
Devidé (1970) as a synonym, yield no real evidence to make its taxonomic position more precise.
Stenophycus (Devonian: Fenton 1943; Wray 1967) is a composite, doubtfully “primitive coralline’
structure which does show blister-like conceptacles but is not technically monostromatic. Eolitho-
porella dawsoni Johnson (1966), from the Canadian Mississippian, is perhaps the nearest to
Lithoporella, but is based on very limited material and conceptacles are not known.

All of these have considerably larger thalli than Hypocaustella, and although of Upper Palaeo-
zoic age, are very clearly spreads of roofed cells or laminae, with only occasional loss in preserva-
tion. Hypocaustella is curiously distinctive in this character, so its algal nature cannot be regarded
as certain.

Pre-Cretaceous genera referred, often doubtfully, to the Corallinaceae are a scattered assem-
blage whose individual members seem to have become extinct. Archaeolithophyllum (not mono-
stromatic: Johnson 1956) is the most convincing. If Hypocaustella is referable to this group, and
this is not certain, it must be regarded as similarly individual, and not in the main evolutionary
stream, since its ‘conceptacles’ are not of the characteristic large blister-like pattern. It would be
unwise at present to extrapolate further from this isolated little organism, whose systematic
position is so doubtful.

LOWER CARBONIFEROUS MICROPROBLEMATICUM 77
Acknowledgements

I thank Mr P. V. York, of the BM(NH) Photographic unit, for coping with the photography of
these obscure little organisms: also Mr M. Crawley for preparing the table, Fig. 8.

References

Babcock, J. A. (1974). The réle of algae in the formation of the Capitan Limestone (Permian, Guadalupian),
Guadalupe Mountains, West Texas-New Mexico. 241 pp. Acad. Thesis, Univ. Wisconsin.

Emberger, J. 1958. Lithoporella elliotti nov. sp., mélobésiée nouvelle du Jurassique supérieur des Monts
des Oulad-Nail, Atlas Saharien (Algérie). Bull. Soc. géol. Fr., Paris, (6) 7 : 625-629.

Endo, R. 1959. Stratigraphical and paleontological studies of the later Paleozoic calcareous algae in Japan,
XIV. Fossil algae from the Nyugawa Valley in the Hida Massif. Sci. Rep. Saitama Uniy., Urama, (B)
3 : 177-207.

Fenton, C. L. 1943. A new Devonian alga from Western Australia. Am. Midl. Nat., Notre Dame, Ind.,
30 : 112.

Homann, W. 1972. Unter- und tief-mittelpermische Kalkalgen aus den Rattendorfer Schichten, dem
Trogkofel-Kalk und dem Tressdorfer Kalk der Karnischen Alpen (Osterreich). Senckenberg. leth.,
Frankfurt a.M., 53 : 135-313.

Hudson, R. G. S. 1933. The scenery and geology of north-west Yorkshire. Proc. Geol. Ass., London, 44 :
228-255.

— & King, W. B. R. 1933. The Yorkshire Dales: the Summer Field Meeting, 1933. Proc. Geol. Ass.,
London, 44 : 428-440.

Johnson, J. H. 1956. Archaeolithophyllum, a new genus of Paleozoic coralline algae. J. Paleont., Menasha,
30 : 53-55.

— 1966. New Mississippian algae from Alberta. J. Paleont., Menasha, 40 : 1385-1387.

Kendal, P. F. & Wroot, H. E. 1924. Geology of Yorkshire. Anillustration of the evolution of northern England.
xxli+ 995 pp., 2 vols. Leeds (privately published).

Kochansky-Devidé, V. 1970. Die Kalkalgen des Karbons vom Velebit-Gebirge (Moskovien und Kassi-
movien). Palaeont. jugosl., Zagreb, 10. 32 pp.

Lemoine, M. 1974. Contribution a l’étude du genre Lithoporella (Corallinacées). Revue algol., Paris,
11 : 42-57.

— 1976. Le genre Lithoporella Foslie (algue Rhodophycée Corallinacée) au Tertiare et au Quaternaire.
Bull. Soc. géol. Fr., Paris, (7) 18 : 773-787.

Mamay, S. H. 1959. Litostroma, a new genus of problematical algae from the Pennsylvanian of Oklahoma.
Am. J. Bot., Lancaster, Pa., 46 : 283-292.

Spjeldnaes, N. 1963. A new fossil (Papillomembrana sp.) from the Upper Precambrian of Norway. Nature,
Lond. 200 : 63-64.

— 1967. Fossils from pebbles in the Biskopasen Formation in southern Norway. Norg. geol. Unders.,
Oslo, 251 : 53-82.

Wray, J. L. 1967. Upper Devonian calcareous algae from the Canning Basin, Western Australia. Prof:
Contr. Colorado Sch. Mines, Golden, 3. 76 pp.

1977. Late Paleozoic calcareous red algae. /n Fliigel, E. (ed.), Fossil algae: recent results and develop-

ments : 167-176. Berlin, Heidelberg and New York.

- ee A ft | VON IM) ee anal

: i ee pr sort: (a eer Se

t 15 aur :) Aiiegtth@ det, Caras iid
‘ Aout aheH 2270! walle” eg All bee
=

; 7 ., gin © = act Pe la “lp vy al-wes'h - :

a ; we = Le ; +. Trai re tee bere Sine)
| Suk. Gates A eile a oe Mig ee dey? Set eel
E . i 7 : an ees | A. ‘

om ie ( : a bled : ae 1 ie es te te

»
é
: - ; f rime a) oo.) GxteA ore di ees *
=i, = et ee ithbatenen {
. ef re = = ‘= <€*S5 Wy
Lu j ‘A

oan TT. or 9a ig
ot ? i pe ; ly iT la LL OR, Sa bx pe | ee La | .
nae P at ‘oro st el > sin

- ob yarn ir Ot herb ot dine wrote ony
:

mi i ’
= . , : 4 - ‘ _ a it a
—— "© eS OE Oe | a iA ihe
a ® a
7 = Ota) Sogn alot heernteiy rly
: - fi . ~ = “ vir

cS => 67)
e vw Ws
“ ¥
_— a i Th =
= i
= a i wi
a ‘ ©
oe rs

te pore eee)
j ; eS e%

Acanthopleurella Groom 1902: origin and life-habits
of a miniature trilobite

R. A. Fortey

Department of Palaeontology, British Museum (Natural History), Cromwell Road, London
SW7 5BD

A. W. A. Rushton
Institute of Geological Sciences, Exhibition Road, London SW7 2DE

Synopsis

The Tremadoc genus Acanthopleurella includes the smallest of mature trilobites. The type species, A.
grindrodi, is revised and a new species A. stipulae described. They both have four thoracic segments and
were derived neotenously from a shumardiid ancestor by suppression of the release of two (?) segments.
They were specialized, secondarily blind, burrowing deposit feeders. The extremely small size of A. stipulae
enabled it to co-exist with Conophrys salopiensis, inhabiting the same substrate.

Introduction

In 1902 Professor Theodore Groom described the tiny trilobite Acanthopleurella grindrodi from
the Tremadoc rocks of the Malvern Hills, but his material consisted only of internal moulds in an
unsatisfactory state of preservation. Despite observations by Stubblefield (1926) and Lake (1946),
and a figure by Bulman & Rushton (1973), the genus has remained little known. We have now,
however, studied several good complete specimens, and not only confirm that Acanthopleurella
is a valid genus with at least two species, but consider that it includes the smallest known adult
trilobites. Its morphology is quite complex and it is possible to draw some functional conclusions
concerning its mode of life and to investigate the possible causes of miniaturism im trilobites, a
topic briefly considered by Robison & Campbell (1974).

In the form of its cranidium and glabella, the structure of the thoracic pleurae and the pygidium,
Acanthopleurella is clearly related to Shumardia Billmgs, and in the discussion below we make
many comparisons with the Shineton Shales species Conophrys salopiensis Callaway, a species
formerly often referred to as Shumardia pusilla (Sars); this is not the place to discuss the validity
of Conophrys salopiensis, but in using Callaway’s name we intend to refer primarily to the well-
known Shineton Shales form.

Miniaturized trilobites

Size. Although tiny, Acanthopleurella is not based on the young stages of larger trilobites. The
only similar-looking forms are shumardiids, but Stubblefield (1926) showed that the form named
A. stipulae here could not be regarded as normal growth-stages of C. salopiensis. Furthermore, the
known size range of holaspids of A. stipulae (with four free thoracic segments) is considerable,
from 1:07 mm to 1:50 mm. This indicates a greater size range than in any of the meraspid degrees
of C. salopiensis described by Stubblefield (1926).

With a sagittal length of 1:07 mm to 1:50 mm A. stipulae appears to be the smallest trilobite
known. Agnostids generally enter the holaspid condition at a length of 3mm or more; the
eodiscid Opsidiscus is 2 mm or more in length (cf. Westergard 1946). A small complete Schmalen-
seeia is 2:75 mm long (Westergard 1922: pl. 1, fig. 19). The holotype of Hospes clonograpti
Stubblefield (in Stubblefield & Bulman 1927 : 129) is 2-51 mm long; smaller specimens seem not

Bull. Br. Mus. nat. Hist. (Geol.) 33 (2): 79-89 Issued 31 January 1980
79

80 R. A. FORTEY & A. W. A. RUSHTON

to be holaspides. Thoracocare minutum Robison & Campbell (1974 : 278) reaches the holaspid
state at a length of 1-7 mm and Is in the meraspid condition at 1-15 mm (1974 : fig. 1B). Shumar-
diids are generally considerably larger, but Stubblefield recorded a C. salopiensis ‘meraspis of
degree 6’ —a holaspis in more general if less precise terms — only 1-81 mm long; and Shumardia
nericiensis Wiman (1905) is only about 2 mm long.

Origin. Shwmardia and Conophrys are in any case small and most species have six thoracic seg-
ments (but some species have five or seven). Acanthopleurella has only four, and part of the
miniaturization of the genus is produced by reduction of the number of segments. But it is not
as simple as that. Acanthopleurella holaspides with four segments are larger than degree 4 mer-
aspides of Conophrys salopiensis (Fig. 12). Stubblefield has shown that in Conophrys the fourth,
macropleural, segment, after being released from the transitory pygidium, migrates forwards in
successive moults as the posterior two segments are added. In Acanthopleurella the process stops
after the release of the fourth segment, which is presumed to be homologous with the macropleural
fourth segment in Conophrys. The third thoracic segment on Acanthopleurella is also spinose, and
on the meraspis three pairs of spines are present (Fig. 19). Hence during its ontogeny segments
bearing the first two spines migrate forwards to be released as the two segments with spinose
pleurae, while the posterior, third pair is aborted.

It is reasonable to assume that the Conophrys condition, generally with six segments, is primi-
tive, and the Acanthopleurella condition, with four, is derived. The six-segment condition is also
persistent, since the much younger species Shumardia granulosa Billings (redescribed in Whitting-
ton 1965) also has six segments. There may have been a five-segment stage between a Conophrys-
like ancestor and Acanthopleurella (C. ?curta Stubblefield has five segments), but since there are

Figs 1-4 Acanthopleurella grindrodi Groom. All x20. Fig. 1, IGS TW454A. Latex cast of dorsal
exoskeleton. Figd Bulman & Rushton (1973 : pl. 6, fig. 6). Combe Abbey no. 1 Borehole, near
Coventry, at 76m depth. Fig. 2, IGS Zg184. Latex cast of dorsal exoskeleton. Bronsil Shales,
south of Chase End Hill, near Malvern. Fig. 3, Holotype, University Museum, Oxford, A7.
‘Dictyonema-shales’ of the Malvern Hills. Imperfectly preserved internal mould, partly enrolled.
Mould of the macropleural spine on the left appears as dark hole. Figd Groom (1902 : fig. 3)
and Lake (1907 : pl. 4, fig. 3). Fig. 4, Paratype, University Museum, Oxford, A8. Internal mould
of thorax and pygidium, loc. as for Fig. 3. Figd Groom (1902 : fig. 4).

Figs 5-10, 13-15, 19 Acanthopleurella stipulae sp. noy. All x20 except Fig. 19. Shineton Shales,
Sheinton Brook, loc. ‘RR2’ of Stubblefield & Bulman (1927 : pl. 5). Fig. 5, Latex cast of holotype,
complete exoskeleton, with associated cranidium. IGS Zs6184. Note genal spines. Fig. 6, IGS
Zs6186, cranidium showing anterolateral lobes. Fig. 7, IGS GSM 58712. smallest holaspis, figd
Lake (1907 : pl. 4, fig. 4). Fig. 8, BM(NH) 1In26816a, holapsis with damaged axis and pyritized
spines. Figd Stubblefield (1926 : pl. 15, fig. 16). Fig. 9, IGS Zs6187, pygidium, poorly preserved
internal mould but with pyritized border showing granules. Fig. 10, IGS Zs6182, pygidium showing
large tubercles on pleural fields and well-developed anterior segment. Fig. 13, IGS Zs6185, meraspid
of degree 3, showing pygidial doublure and pyritized spines on third thoracic segment and anterior
of transitory pygidium. Fig. 14, IGS GSM 48713, largest holaspis, figd Lake (1907 : pl. 4, fig. 2).
Fig. 15, Latex cast of BM(NH) In 26816b, showing large articulating half-ring on third thoracic
segment. Fig. 19, Scanning electron micrograph of latex cast from meraspis of degree 0 on same
rock chip as holotype. Note marginal spines and suture between cephalon and protopygidium.
« 90.

Figs 11-12, 16-17 Conophrys salopiensis Callaway. Shineton Shales, Sheinton Brook, loc. ‘RR2’ of
Stubblefield & Bulman (1927 : pl. 5). Fig. 11, BM(NH) In26819, x9, internal mould of complete
exoskeleton in oblique side view to show facial suture. Figd Stubblefield (1926 : pl. 16, fig. 20).
Fig. 12, BM(NH) In26815, » 20, internal mould of meraspis degree 4 showing macropleural spine
on fourth segment. Figd Stubblefield (1926 : pl. 15, fig. 15). Figs 16, 17, IGS Zs6188 (Stubblefield &
Bulman coll.), latex cast of exceptionally well preserved complete exoskeleton, dorsal and oblique
lateral views, «9, showing facial suture and free cheeks, possibly conjoined.

Fig. 18 Conophrys sp. IGS RU5885. Entire exoskeleton showing seven thoracic segments, x 15.
Lower Tremadoc Series, tributary of Nant-y-Gist-faen, near Arennig Fawr, North Wales.

81

MINIATURE TRILOBITE

82 R. A. FORTEY & A. W. A. RUSHTON

examples (e.g. Ceraurinella, see Whittington & Evitt 1954) where several thoracic segments are
released simultaneously into the thorax from the transitory pygidium, it seems perfectly possible
to us that elimination of the release of two segments may have occurred as a single evolutionary
step. Since Acanthopleurella had to be sexually mature, even at this small size, what essentially
happened was a movement backwards (ontogenetically speaking) of sexual maturity — the process
usually known as paedomorphosis.

A possibility that has to be reviewed is that Acanthopleurella was a sexual dimorph of Conophrys.
This is improbable. In the first place the two forms are not invariably found together. For example,
in the Amnodd Shales (near Arennig Fawr, Gwynedd) Conophrys salopiensis is common, but no
example of Acanthopleurella has been discovered. Conversely at Malvern we have not noted a
Conophrys from the same beds as Acanthopleurella grindrodi. If of opposite sex they would evi-
dently have encountered one another very rarely! One might also expect, generally, Shumardia-
like species would in that event tend to come in ‘pairs’, one male and one female morph occurring
together. In our experience this is not so. In the early Ordovician of Spitsbergen, for example,
where three species of shumardiid occur, they are found in sequence rather than together. Simi-
larly the Shumardia Limestone in Quebec seems to contain only a single species, and that in great
abundance.

Life habits

The present example of miniaturization is not unique. Robison & Campbell (1974) have described
a Cambrian corynexochoid trilobite Thoracocare with only two thoracic segments in the holaspid
stage. It is likely that a similar process of miniaturization was involved in the genesis of the
Agnostacea. Jell (1975) has discussed polymorphism in Pagetia, m which two- or three-segmented
forms are recognized within a single species.

Robison (1972) and Jell (1975) concur that trilobites of the Agnostida (sensu Salter 1864) were
pelagic or planktonic in habit. The same habits were deduced for Thoracocare (Robison & Camp-
bell 1974). It should be noted, however, that some authors have proposed different life habits for
the agnostids: for example, Bergstr6m (1973) has suggested that they may have been parasitic
on some (unspecified) free-swimming organism whereas Pek (1977) has presented evidence that
some forms may have been epipelagic upon some floating seaweed, and Jago (1973) has suggested
that they show some community behaviour related to former sea-bottom conditions, and hence
may have included some benthic forms.

Thoracocare has large eyes and a very large pygidium, and is not really comparable to Acantho-
pleurella in its morphology. Agnostids were blind, like Acanthopleurella, but beyond this and
their small size morphological comparisons cannot be extended in detail. Powerfully spinose
agnostids are unusual, and we can only find one species in which the pleurae are spinose in the
manner of Acanthopleurella (Lejopyge laevigata armata Westergard, 1946 : pl. 13, figs 32-36).

At the small adult size of Acanthopleurella genal and thoracic spimes would have been effective
in increasing frictional drag during descent through the water column. Many protaspides, which
were supposedly planktonic, are well equipped with spines which would have inhibited sinking,
as are Recent crustacean larvae. However, in this case we do not believe that the spies functioned
as frictional brakes. In the first place they lie more or less horizontally alongside the thorax, so
that in dorsal view they are not disposed in a way which would have been particularly effective in
inhibiting sinking. If this had been their principal purpose one would have expected them to have
been splayed out in the manner of odontopleurids, such as Miraspis, to present as large a surface
area as possible. Second, the mid-thoracic spine would have had no influence on frictional drag.
Third, Acanthopleurella seems to have had a very restricted distribution, being confined to Tre-
madoc shales of the Welsh borderlands and the English Midlands, not even extending over the
whole region from which Conophrys salopiensis has been recovered. Fortey (1974 : fig. 4) has
shown that pelagic trilobites tend to have an exceptionally elongate exoskeleton, whereas Acantho-
pleurella, the transverse width of which at mid-thorax is about three-quarters the length (sag.), 1s
not unusual in this regard. The balance of the evidence suggests that Acanthopleurella was benthic,

MINIATURE TRILOBITE 83

and the fact that the meraspis degree 0 already had assumed the morphology characteristic of the
adult further suggests that this habit was acquired from an early stage in ontogeny.

Some specimens (Figs 8, 15) show well-developed articulating half-rings, extending to almost
the entire length of the preceding thoracic axial ring on the segments anterior to the macropleural
segments. This implies considerable longitudinal flexibility. Given this fact, the lack of pleural
spines on the first two thoracic segments can be explained: had spines been present on the first
two segments any downward flexure of the cephalon (resulting in an upward swing of the genal
spines) would have been prevented. Fig. 5 shows well how closely the genal spines approach the
third thoracic segment.

Assuming benthic habits, we can extend the analysis of functional aspects of Acanthopleurella
to enrollment. It is extremely unlikely that enrollment would have occurred by movement of the
thorax relative to the cephalon, as this would have involved dragging the relatively enormous
thoracic spines through the sediment. On the other hand, enrollment could have been achieved
relatively easily by tucking the cephalon down and backwards beneath the pygidium; the short
anterior thoracic pleurae permit unimpeded downward flexure of the cephalon. Most of the
flexure seems to have been accommodated at the forward margin of the first, second and third
segments, where, as we have noted, remarkably broad (sag.) articulating half-rings are developed.
The resulting enrolled position might well have utilized the long backward-directed spines on the
third and fourth thoracic segment as a means of support on the sediment surface during enroll-

ment.
Both Conophrys and Acanthopleurella lack eyes, and, as in all trilobites, the blindness was

secondary, as shown by the librigenal remnants figured in Figs 20a, b. We can eliminate great
oceanic depth (Clarkson 1967) as a possible cause for this blindness because almost all the trilo-
bites found in association with these shumardiids had normal eyes (except the agnostids), and
were not different in eye development from other ‘shelf’ assemblages of the Tremadoc.

Blindness is explained by inferring that Acanthopleurella burrowed superficially in the surface
layer of the argillaceous sediments with which it is associated, a life mode which rendered eyes
superfluous. Reasonable support for this hypothesis is derived from a consideration of extant
infaunal, minute crustacea. Even at 1-2 mm long Acanthopleurella is too large to be considered as
a possible interstitial (meiofaunal) inhabitant, like the Mystacocarida, although visual surfaces
are generally reduced in those forms. The best, and rather close, analogy is perhaps with cephalo-
carids (Sanders 1963). Hutchinsoniella macracantha has an adult length of 3 mm (excluding caudal
spines) with a broadly hemispherical cephalon. It is blind, but secondarily so. It is benthic from
early in its ontogeny, which, like that of shumardiids (but not of all trilobites), is a simple,
progressive addition of segments without any profound morphological change. The principal
morphological difference is the vermiform and flexible posterior segments of Hutchinsoniella. As
we have inferred for Acanthopleurella, Hutchinsoniella lives in soft, muddy sediments. Hutchin-
soniella extracts edible particles from detritus by using its limbs as a kind of filtration device.
Whether or not Acanthopleurella used its limbs in the same way, it seems reasonable to suppose
that it also extracted organic particles from the sediment. It is also worth noting that shumardiids
occur in association with ostracods, with valve dimensions similar to the transverse cephalic
dimensions of Acanthopleurella and Conophrys (Stubblefield & Bulman 1927; Bulman & Rushton
1973 : pl. 6). Benthic ostracods are also frequently deposit feeders, and it is not unreasonable to
suppose that these arthropods were also concerned with exploiting the same source.

The occipital ring of Acanthopleurella stipulae is long (sag.) and produced backwards. This
feature acquires significance if we assume that Acanthopleurella adopted an active feeding
position with the cephalon tilted through an angle of perhaps 20° (Fig. 21c), in a manner similar
to that shown by Clarkson (1969) for the Silurian odontopleurid Leonaspis deflexa. In this
attitude the prolonged occipital ring serves to cover the ‘gap’ which would otherwise be created
between thorax and cephalon. The genal spines then tilt upwards past the first two thoracic
segments. Note that in Leonaspis deflexa the anterior thoracic pleural spines are (for an odonto-
pleurid) also relatively abbreviated.

As a final comment on the function of the macropleural spines Ingham (1968 : fig. 3) has
suggested that in Cybeloides girvanensis, a species equipped with long macropleurae, the spines

84 R. A. FORTEY & A. W. A. RUSHTON

a

Fig. 20 (a) Reconstruction of Acanthopleurella grindrodi Groom, 30. (b-d) A. stipulae sp. nov.;
(b) reconstruction, 30; (c) meraspid of degree 0, «60, reconstructed from Fig. 19; (d) pygidium
of meraspid degree 3, » 60, based on an internal mould on the same rock-chip as the holotype.

functioned as an aid in burial, and that the long macropleural spine had a stabilizing function
“preventing excessive lateral rolling during arching of the thorax’ while burial was in progress.
Acanthopleurella may have spent a large part of its life buried at least within the semi-fluid surface
layer of sediment. The vertical spine from the middle of the third thoracic segment, which we
have not so far discussed, seems ideally placed to have acted as a sensor to make sure that burial
did not proceed too far. At the minute size of Acanthopleurella we do not consider that the spines
would have had a particular defensive function.

Fig. 21 (a) Lateral view of Acanthopleurella grindrodi, » 30; (b), (c) lateral view of A. stipulae, x 30,
in relaxed attitude (b) and in ‘active’ feeding or grubbing attitude (c). Length of dorsal axial spines
and anterior course of facial sutures conjectural.

MINIATURE TRILOBITE 85

Summary of life habits. Acanthopleurella was derived from a Conophrys-like ancestor by suppres-
sion of release of two segments during ontogeny. It was a blind, benthic animal, but did not inhabit
a particularly deep water environment. It was probably a burrowing deposit feeder, using its
long macropleural spines as a support at or just below the sediment surface, during enrollment,
and as an aid to burial.

Resource partitioning and the relation of Conophrys to Acanthopleurella

Robison (1975) was the first to examine the effect of competitive exclusion in controlling the
numbers of species and adult size of sympatric agnostids. Where related sympatric species utilize
the same resource they can coexist without mutually exclusive competition if they exploit the
resource in different ways, for example by taking different particle sizes if they are deposit feeders.
May (1973) has given the mathematical basis for this kind of resource partitioning. Applications
of this theory to marine benthos living on soft bottoms are relatively limited, but it has been
applied to the small (c. 3 mm) benthic gastropod Hydrobia by Fenchel(1975). Thisisanappropriate
example in the present case because Hydrobia is a small deposit feeder, and may exploit a resource
and particle size not so very different from that utilized by shumardiids. Differences in particle-
size utilization are reflected more or less directly in differences in adult size of the exploiting
animal. Robison (1975) showed that the ratio between lengths of ‘neighbouring’ agnostids
commonly approached the theoretically predicted value of 1:3 (and up to about 1:5). We do not
have anything like the number of specimens of Acanthopleurella that Robison used for his
agnostid example, but we can compare the length (sag.) of largest A. stipulae with average
holaspid length of Conophrys salopiensis obtained from Stubblefield (1926). The length ratio of
the latter to the former is about 2:2, which indicates that the size-difference between the two forms
was more than adequate to allow them to exploit the same resource without direct competition.
This ratio is far larger than the theoretical minimum. There is another small blind trilobite in the
British Tremadoc, Hospes clonograpti Stubblefield (in Stubblefield & Bulman, 1927) which
almost exactly serves to bridge the size gap: the length ratio of Hospes to Acanthopleurella is
about 1-7, and that of Conophrys to Hospes is about 1-3. It is possible that Hospes, too, exploited
the same environment. However, it has not so far been found in the beds that have yielded
Acanthopleurella and Conophrys.

The shumardiid example is particularly interesting because we know the mechanism by which
the morphological difference necessary to avoid competition was achieved, in this case the
shortening of the thorax. There are other families of trilobites in which the number of thoracic
segments is highly variable, for example the Olenidae, and similar adaptive reasons may prove to
be behind otherwise inexplicable differences in thoracic development.

Systematic descriptions
Family SHUMARDIIDAE Lake, 1907
Genus ACANTHOPLEURELLA Groom, 1902
TYPE SPECIES. Acanthopleurella grindrodi Groom, 1902, by monotypy.

D1AGnosis. Minute, blind, opisthoparian shumardiid trilobites. Glabella without lateral furrows,
but faint anterolateral lobes may be developed. Occipital ring long sagittally but not spined.
Genal spines present on free cheeks. Thorax of four segments, the third bearing an axial spine, the
third and fourth having long pleural spines. Pygidium transverse, with border, pleural fields flat;
anterior segment well defined, segmentation obscure posteriorly.

REMARKS. Acanthopleurella differs from Conophrys most obviously in the shorter thorax and the
presence of macropleural spines on the third as well as the fourth segment. Conophrys salopiensis
Callaway (= Shumardia pusilla auctt.), from the Shineton Shales, has six thoracic segments of
which the fourth only has macropleural spines. Other Conophrys may have seven segments (e.g.
C. sp. of Bulman & Rushton (1973 : pl. 6, fig. 5) and another undescribed species (Fig. 18) from

86 R. A. FORTEY & A. W. A. RUSHTON

the lowest Tremadoc of North Wales) but in these only the fourth segment is macropleural. C.?
curta Stubblefield has five thoracic segments and again only the fourth is macropleural. No
Conophrys are known to have thoracic axial spines.

The cephalon of Acanthopleurella is less distinctive: the anterolateral lobes are weaker than in
C. salopiensis but are not weaker than those of C.? curta or the C. sp. of Bulman & Rushton, or
the cranidia referred to Shumardia alata Robison & Pantoja-Alor (1968 : 797). Whittington
(1965 : 329) showed that S. granulosa has free cheeks and suggested that the Shineton Shales ‘S.
pusilla might also have them, although the evidence he cited, Lake’s pl. 4, fig. 1, is based on a
figure of a young Asaphellus (Lake 1942 : 324). Nevertheless, we find that Whittington’s sug-
gestion is supported by Shineton Shales specimens (Figs 11, 16, 17). Unlike Acanthopleurella, C.
salopiensis does not show genal spines, although a possible Conophrys species, ‘Shumardia’
oelandica Moberg (1900 : pl. 14, fig. 4a), is illustrated with apparently fixigenal spines, and ‘S.’
nericiensis Wiman seems to have librigenal spines (Wiman 1905 : pl. 2, fig. 14).

The pygidia of Conophrys and Shumardia species vary greatly in shape and in the characters of
the border and doublure. Acanthopleurella pygidia are short and wide with flat pleural fields and
tend to be slightly emarginate behind the axis. In this they differ strongly from S. granulosa but
resemble some Conophrys.

Acanthopleurella grindrodi Groom, 1902
Figs 1-4, 20a, 21a
1902 Acanthopleurella grindrodi Groom : 70, figs 1-4.
1907 Shumardia pusilla (Sars); Raw in Lake : 42 (pars); pl. 4, fig. 3 only [figures holotype and places
it in synonymy of S. pusilla).
1946 Acanthopleurella grindrodi Groom; Lake : 341 [revives Acanthopleurella].
1973 Acanthopleurella sp.; Rushton in Bulman & Rushton : 25; pl. 6, fig. 6.

TYPE SPECIMENS. Holotype OUM A7 (Fig. 3), the original of Groom’s (1902) fig. 3. Paratype
OUM Asés (Fig. 4), the original of Groom’s fig. 4. Both are from the ‘Dictyonema-shales of the
Malvern Hills’, from beds correlated by Stubblefield & Bulman (1927) with parts of the Dictyo-
nema flabelliforme Zone or the Transition Beds of the Shineton Shales.

OTHER MATERIAL. IGS Zg183, Zg184, J. C. Harper coll. From Bronsil Shales (Tremadoc) south
of Chase End Hill, Malvern Hills. IGS TW454A from Tremadoc shales in the Combe Abbey no. 1
Borehole, 6 km east of Coventry; the age of these beds is a matter of some doubt (Bulman &
Rushton 1973 : 7-8, 11).

DiaAGnosis. Acanthopleurella without distinct anterolateral glabellar lobes; occipital ring about
one-third the width of cranidium; posterior macropleura relatively massive, nearly as long as
sagittal length of trilobite. Pygidial axis with four segments, pointed behind; pleural fields of
thorax and pygidium with rows of many but inconspicuous granules.

MEASUREMENTS. IGS Zg184: sagittal length = 1-70 mm, length of cephalon(maximized) =0:92 mm,
width of cranidium (estimated) = 1-42 mm, length of fourth segment = 1-5 mm, length of pygidium
=0:55mm, width of pygidium=1-01 mm. TW454A: sagittal length=2-03 mm. The holotype
has a pygidium 0-48 mm long and 0:92 mm wide; the paratype pygidium is 0-60 mm long and
1-25 mm wide.

REMARKS. The type specimens are rather poorly preserved and do not show several features
taken here to be characteristic of A. grindrodi (cf. Rushton in Bulman & Rushton 1973 : 26). We
interpret the species by reference to IGS Zg184 (Fig. 2) and do not see any significant primary
differences between that specimen and Groom’s types. The holotype does not show the macro-
pleura on the third segment, probably because the shale has split at a level lower than that at
which the spine was preserved, but the mould of the spine on the fourth segment is clear (Fig. 3).
We see no evidence for the four pairs of macropleurae shown in Groom’s (1902) fig. 1. Groom’s
interpretation of the cephalon is based on its posterior half only, the anterior part of the cephalon
being folded under and not clearly visible.

MINIATURE TRILOBITE 87

A. grindrodi was described by Rushton (in Bulman & Rushton 1973) on the basis of one
specimen which resembles Zg184 in almost all particulars (Fig. 1). Rushton’s description should
be amended because he described the cranidium as a cephalon. IGS Zg183 shows small free
cheeks with a genal spine about half as long as the cranidium. The punctate surface doubtfully
observed by Rushton cannot be confirmed by other specimens. In Zg184 the thoracic axis is a
little wider than the occipital ring. Pairs of small granules are placed exsagittally on the pleural
and pygidial axial rings, and rows of small granules are present on the anterior bands of all the
thoracic and pygidial pleurae as well.

This species is contrasted with the new form A. stipulae below.

Acanthopleurella stipulae sp. nov.
Figs 5-10, 13-15, 19, 20b-d, 21b, c
1907 Shumardia pusilla (Sars); Raw in Lake : 42 (pars); pl. 4, figs 2, 4.

1926 Shumardia pusilla variety ?; Stubblefield : 257-258; pl. 15, fig. 16 [distinguished from S. pusilla].
1946 Acanthopleurella; Lake : 341 (pars) [revives Acanthopleurella].

HototyPe. IGS Zs6183—4, counterparts (Fig. 5), Stubblefield & Bulman coll. Shineton Shales,
Shumardia pusilla Zone, Sheinton Brook, south of Sheinton, Salop. Locality “RR2’ of Stubble-
field & Bulman (1927 : map, pl. 5).

PARATYPES. BM(NH) In.26816, complete specimen; IGS Zs6181—2, 6185-7, 4 complete speci-
mens (2 immature), 4 cranidia, 3 pygidia (1 immature), all Stubblefield & Bulman coll. IGS
GSM 48712, 48713, Rhodes coll. All are from the same locality as the holotype.

DiaGnosis. Acanthopleurella with distinct anterolateral lobes; occipital ring wider than fixed
cheeks; posterior macropleura about half sagittal length of trilobite; pygidial axis short and
blunt; pleural regions of thorax and pygidium with few but conspicuous granules.

Name. ‘Of stubble’, with reference to Sir James Stubblefield.

DESCRIPTION. Cranidium convex, length two-thirds of width. Cephalic axis four-fifths or more
of cranidial length and two-fifths of its width. Glabella bluntly pointed in front, defined by distinct
axial furrow; weak but distinct anterolateral lobes mark its widest part, behind which the sides
of the glabella are slightly concave; no glabellar furrows. Occipital furrow weakly marked, nearly
transverse. Occipital ring about as wide as widest part of glabella and nearly half as long, pro-
duced backwards in a broad curve, with a faint median node on its posterior edge. Fixed cheeks
curved down frontally and laterally, without eyes or eye-ridges. Posteriorly the fixed cheek is two-
thirds of the width of the occipital ring, abaxially curves somewhat backwards and downwards
to the rounded posterolateral corner where the facial suture cuts off the genal spine. No border
except posteriorly, where a short pleuroccipital furrow defines a narrow depressed border near
the glabella, but dies out before reaching the facial suture. Facial suture appears to be nearly
marginal for most of its course but cuts off the genal spines at the posterolateral corners. Genal
spine depressed below the level of the first two thoracic pleurae but the tip curves up to clear the
third pleura. Exoskeleton smooth.

Adult thorax of four segments. Axis of first two segments nearly twice as wide as pleura. Axis
of third ring narrower and bears a large backwardly- and upwardly-directed spine of unknown
length. The first and fourth axial rings bear pairs of small exsagittally-placed granules; the second
and third axial rings have two pairs of such granules. First and second pleurae short, third and
fourth pleurae have long spines, the macropleura on the fourth segment being over half the length
of the rest of the trilobite. Stubblefield (1926 : 358) interpreted the specimen here refigured
(Fig. 8) as having a spine on the second as well as the third and fourth pleurae. Our interpretation
is that the genal spines have been displaced slightly backwards, perhaps during moulting, so that
the spine has a somewhat anomalous position alongside the second and first segments. We
believe that we can recognize the sutural margin on the spine that identifies its genal origin. The
‘trace of a genal spine’ also noted by Stubblefield cannot now be observed on the specimen. The

88 R. A. FORTEY & A. W. A. RUSHTON

pleurae each have two granules on the anterior pleural band, the granule nearer to the axial
furrow being the larger and comparatively conspicuous.

Pygidium with length about half the width, outline posteriorly faintly emarginate. The axis
occupies just over one-third of the greatest width, is shorter than wide and truncate behind;
composed of one well-defined ring, one or possibly two fainter ones behind and a terminal piece;
the anterior ring carries a pair of small exsagittally-placed granules. Pleural fields flat, composed
of one well-defined segment with a trace of one or two more behind. Conspicuous granules on the
pleural fields continue the exsagittal arc of conspicuous granules seen on the thoracic pleurae.
Posterior border weakly defined, furnished with small granules (Fig. 9). Width of doublure
about one-fifth of pygidial length.

MEASUREMENTS. Holotype: sagittal length of trilobite=1:32 mm, length of cranidium (maxi-
mized) =0-70 mm, width of cranidium (estimated) = 1:1 mm, length of fourth segment =0-78 mm,
length of pygidium=0-32 mm, width of pygidium=0-65 mm. Length of BM(NH) In.26816 is
1-40 mm if allowance is made for the crushed front of the cephalon. IGS GSM 48713 and 48712
are 1:50 mm and 1:07 mm long respectively.

REMARKS. A stipulae differs from A. grindrodi in its relatively wider axis. The glabella has more
definite anterolateral lobes than in A. grindrodi and the occipital ring has a faint median node.
In A. stipulae the pleural regions have fewer but larger granules than in A. grindrodi and the axes
of the second and third segments have an extra pair of exsagittal granules. The macropleura on
the fourth segment is not as massive as that of A. grindrodi but appears to have a sharper tip.
The pygidial axis has a shorter, blunter axis with one or two fewer rings. The available material
suggests that A. stipulae is a smaller species than A. grindrodi.

Stubblefield (1926 : 357-358) discussed the differences between the form named here as A.
stipulae and meraspid specimens of Conophrys salopiensis (“Shumardia pusilla’) with four thoracic
segments (Fig. 12). The principal differences include the larger anterolateral glabellar lobes of
C. salopiensis, its single macropleural segment and the stronger segmentation of the pygidium.
We would emphasize also the large occipital ring and the axial spine on the thorax of A. stipulae.
Meraspids of degree 4 reach 1:06 mm in length (Stubblefield 1926 : 356), and thus almost over-
lap the lower size-range of A. stipulae holaspides.

The presence of weak anterolateral lobes in A. stipulae recalls the condition of immature
specimens of the associated shumardiid C. salopiensis (cf. Stubblefield 1926 : pl. 14, fig. 6; pl. 15,
fig. 15). The only shumardiid known to be associated with A. grindrodi is the Conophrys sp.
figured by Bulman & Rushton (1973 : pl. 6, fig. 5) which, like A. grindrodi, shows scarcely a
trace of anterolateral lobes even though the specimen is apparently full-grown.

DEVELOPMENT. A meraspid of degree 0 is associated with the holotype of A. stipulae. It is 0-45 mm
long and has a comparatively narrow axis. It has three pairs of spines which spring from the edge
of the dorsal side of the transitory pygidium (Figs 19, 20c).

Also associated with these specimens is a transitory pygidium of a meraspid degree 3 (Fig.
20d). It is an internal mould with a sagittal length of about 0:22 mm, the ‘true’ pygidium being
about 0-15 mm long. The long macropleural spine is that of the fourth thoracic segment, not yet
released; behind this is a smaller macropleural segment (reconstructed here from the base of the
spine on the internal mould and a trace of the external mould of the free spinose tip). It repre-
sents the third pair of spines seen in the meraspid of degree 0, partly reduced such that in the
holaspid stage the pleural region of the pyridium is no longer spinose. The specimen shown in
Fig. 13 isa complete meraspid, degree 3. The transitory pygidium has a sagittal length of 0-30 mm,
the ‘true’ pygidium being about 0:22 mm long, and shows the bases of the reduced third pair of
spines. It also shows the broad pygidial doublure. The smallest holaspis (Fig. 7) has a pygidium
0:25 mm long.

Acknowledgements

We thank Mr H. P. Powell (Oxford University Museum) for lending us Groom’s type specimens,
Mr J. M. Evans for taking several of the light-photographs and Mrs B. E. Coleman for scanning

MINIATURE TRILOBITE 89

electron microscopy. Dr Rushton’s contribution is by permission of the Director, Institute of
Geological Sciences.

References

Bergstrom, J. 1973. Organisation, life and systematics of trilobites. Fossils Strata, Oslo, 2 : 1-69, 5 pls.

Bulman, O. M. B. & Rushton, A. W. A. 1973. Tremadoc faunas from boreholes in Central England. Bull.
geol. Surv. Gt Br., London, 43 : 1-40, pls 1-7.

Clarkson, E. N. K. 1967. Environmental significance of eye-reduction in trilobites and Recent arthropods.
Mar. Geol., Amsterdam, 5 : 367-375.

1969. A functional study of the Silurian odontopleurid trilobite Leonaspis deflexa (Lake). Lethaia,
Oslo, 2 : 329-344.

Fenchel, T. M. 1975. Character displacement and coexistence in mud snails (Hydrobiidae). Oecologia,
Berlin, 20 : 19-32.

Fortey, R. A. 1974. A new pelagic trilobite from the Ordovician of Spitsbergen, Ireland and Utah.
Palaeontology, London, 17 : 111-124, pls 13, 14.

Groom, T. T. 1902. On a new trilobite from the Dictyonema-shales of the Malvern Hills. Geol. Mag.,
London, (4) 9 : 70-73.

Ingham, J. K. 1968. British and Swedish species of Cybeloides (Trilobita). Scot. J. Geol., Edinburgh,
4 : 300-316, pls 1, 2.

Jago, J. B. 1973. Cambrian agnostid communities in Tasmania. Lethaia, Oslo, 6 : 405-421.

Jell, P. A. 1975. Australian Middle Cambrian eodiscoids with a review of the superfamily. Palaeonto-
graphica, Stuttgart, (A) 150 : 1-97, 9 pls.

Lake, P. 1906-1946. British Cambrian trilobites. 2 (1907) : 29-48, pls 3, 4; 13 (1942) : 307-332, pls 44-46;
14 (1946) : 333-350, pl. 47. Palaeontogr. Soc. (Monogr.), London.

May, R. M. 1973. Stability and complexity in model Ecosystems. Monogr. Popul. Biol., New Jersey, 6.
265 pp.

Moberg, J. C. 1900. Nya bidrag till utredning af fragen om grénsen mellan undersilur och kambrium.
Geol. Fér. Stockh. Forh. 22 : 523-540, pl. 14.

Pek, I. 1977. Agnostid trilobites of the Central Bohemian Ordovician. Sb. geol. Véd. Praha, P 19 : 7-44,
12 pls.

Robison, R. A. 1972. Mode of life of agnostid trilobites. 24th Int. Geol. Congr., Montreal, 7 (Palaeont.) :
33-40.

—— 1975. Species diversity among agnostoid trilobites. Fossils Strata, Oslo, 4 : 219-226.

— & Campbell, D. P. 1974. A Cambrian corynexochoid trilobite with only two thoracic segments.
Lethaia, Oslo, 7 : 273-282.

— & Pantoja-Alor, J. 1968. Tremadocian trilobites from the Nochixtlan region, Oaxaca, Mexico.
J. Paleont., Tulsa, 42 : 767-800, pls 97-104.

Sanders, H. L. 1963. The Cephalocarida. Functional morphology, larval development, comparative ex-
ternal anatomy. Mem. Conn. Acad. Arts Sci., New Haven, 15 : 1-80.

Salter, J. W. 1864. A monograph of British trilobites. 1 : 1-80, pls 1-6. Palaeontogr. Soc. (Monogr.),
London.

Stubblefield, C. J. 1926. Notes on the development of a trilobite, Shumardia pusilla (Sars). J. Linn. Soc.,
London, (Zool.) 36 : 345-372, pls 14-16.

—— & Bulman, O. M. B. 1927. The Shineton Shales of the Wrekin District: with notes on their develop-
ment in other parts of Shropshire and Herefordshire. Q. J/ geol. Soc. Lond. 83 : 96-146, pls 3-S.

Westergard, A. H. 1922. Sveriges olenidskiffer. Sver. geol. Unders. Afh., Stockholm, (Ca) 18. 205 pp., 16
pls.

—— 1946. Agnostidea of the Middle Cambrian of Sweden. Sver. geol. Unders. Afh., Stockholm, (C) 477.
140 pp., 16 pls.

Whittington, H. B. 1965. Trilobites of the Ordovician Table Head Formation, western Newfoundland.
Bull. Mus. comp. Zool. Hary., Cambridge, Mass., 132 : 281-441, 68 pls.

& Evitt, W. R. 1954. Silicified Middle Ordovician trilobites. Mem. geol. Soc. Am., New York, 59 :
1-137, 33 pls.

Wiman, C. 1905. Ein Shumardiaschiefer bei Lanna in Nerike. Ark. Zool., Stockholm, 2 (11) : 1-20, 2 pls.

sim 6 Nl gE a of

ae - f° 7 oo hone
if E t 4) cad ~~ T EL eT a : J » “me G ete
1.4 ght ; 7 : . (Ne -
cas adie? oe
‘
i eh wo
_ af wel
My
aE an)
aan > ale) PAL
¢ Se AY ENS
' iy “Pp aww
<a come
» ah . den 4 faae"t
bow: a) rs e- 7;
7k Wl, 2
7 a = ar eee

a ny or a ee wt)
(ar .

Sierr's ‘j bd oar Ff elaeou @ Gee “i
’ ’ a ee a Ee ea eaitils % aoe

Pleistocene bird remains from Tornewton Cave and
the Brixham Windmill Hill Cave in south Devon

C. J. O. Harrison
Subdepartment of Ornithology, British Museum (Natural History), Tring, Herts

Synopsis

Avian fossil material has been found in Tornewton Cave and the Windmill Hill Cave at Brixham, both
in south Devon, England. The Tornewton material originates from strata deposited during an early cold
phase, followed by an Ipswichian warm phase and by cold Devensian phases. The Brixham Cave material
is Devensian in age. Twenty-two bird species are identified, including a new species of small partridge
Alectoris sutcliffei. The abundant remains of the Shelduck Tadorna tadorna are thought to have been prey
of the White-tailed Sea-eagle Haliaeetus albicilla, and their absence from Tornewton in the Devensian may
be because of the absence of the predator rather than the prey species. Because many species are migratory,
birds are often poor indicators of climatic conditions. Apart from the presence of Ptarmigan Lagopus
mutus in the Devensian, the tundra species are absent. It is suggested that parts of the area may have car-
ried a more boreal flora during the cold periods, and that the area may have benefited from the proximity
to the sea which may have had some modifying effect on climatic conditions.

Introduction

The caves of south-west England have yielded a large quantity of late Pleistocene remains of
mammals and birds, with the former in greater abundance. Unfortunately in some caves an
accidental admixture of earlier deposits with Recent material during the later stages of deposition
or during animal or human occupation, and the collecting of some specimens in the last and the
early part of the present centuries without due regard to stratification of the cave floor, have
made the contents of these caves of limited value in the study of faunal succession at different
periods.

Tornewton Cave is an exception in that careful collecting of material during nearly two decades
makes it possible to study a succession of specimens dating from the last, Devensian glaciation
back through a warm period of the Ipswichian interglacial to a preceding cold period (Sutcliffe
& Zeuner 1962). This cave provides the main mass of material, and the more limited material
from the Windmill Hill Cave has been included because this is from the last glaciation, of Deven-
sian age, and supplements the information provided by the Tornewton specimens for this period.
All the material consists of dissociated bones, and mainly of the ends of the larger limb bones.

All the specimens referred to are in the collection of the Department of Palaeontology of the
British Museum (Natural History); numbers prefixed with A or without prefix. The Windmill
Hill Cave at Brixham was excavated by W. Pengelly in 1858-9 and the specimens collected were
presented, mainly by the Royal Society, in 1876. The Museum purchased from F. H. Butler
part of the collection made by J. L. Widger at the Tornewton and other Torbryan Caves (Walker
& Sutcliffe 1967) in 1892, and the eagle claws from this have been included. The remaining exten-
sive Tornewton material was collected in the period 1944-69 and presented by the late Mr S. R.
Willing.

The Brixham Cave specimens were identified by Lydekker (1891). Some of the earlier Tornewton
material had been identified by Miss P. Lawford in 1955, and some Glutton Stratum bones by
D. Bramwell in 1961. The material from this cave had also been examined more recently by Dr
J. Janossy of Hungary. For the present paper the material was critically re-examined and com-
pared with Recent osteological specimens in the Sub-department of Ornithology, British Museum
(Natural History) and reidentified where necessary.

Bull. Br. Mus. nat. Hist. (Geol.) 33 (2): 91-100 Issued 31 January 1980
91

92 C. J. O. HARRISON

Sites and strata

Tornewton Cave is one of a series of caves in the limestone on the west side of the Torbryan
Valley in south Devon. It is about 250 ft (76 m) above sea level, and about five miles (8 km) from
the sea at Torbay to the east of it, and about six and four miles (9:5 and 6-5 km) from the upper
ends of the estuaries of the rivers Teign and Dart respectively. The Windmill Hill Cave at Brixham
is on a hillside bordering the coast on the south side of Torbay, the cave now being within the
town of Brixham. The position of the coastline would have differed from that of the present day
during the various cold and warm periods of the Pleistocene. In cold periods it would have been
much further away, and maps reconstructing the distribution of the icefields sometimes tentatively
indicate a coastline from south-western Ireland to beyond Brittany. From the present altitude of
the surrounding areas, in warm periods the coast might not have been much closer to the cave
than it is at present. However, the preponderance of Shelduck bones in most strata of the caves
implies that conditions suitable for this species were present no further away than the distance
which a predator was likely to transport them. Unless their ecological requirements differed
significantly at this period, Shelduck would have required saline or brackish mudflats on which to
feed, such as usually occur along estuaries or around lagoons with fluctuating water levels.

The sequence of deposits in Tornewton Cave has been studied and named with reference to
the mammal fauna (Sutcliffe & Zeuner 1962). Both the mammal remains and the nature of the
deposits indicate that the latter were laid down in several different climatic periods. They com-
mence in an earlier cold period, which is overlain by a more typically warm Ipswichian deposit,
and the most recent of the Tornewton deposits are from the last, Devensian, glaciation. The
whole series of deposits referred to in the present paper, beginning with the earliest, is as follows.

Tornewton Cave

Glutton Stratum (? late Wolstonian or later cold phase). This deposit, so named because it
contains remains of the Wolverine or Glutton, Gulo gulo, was laid down during a cold period with
a severe climate and also contains remains of other northern mammals such as Reindeer, Rangifer
tarandus. This may have been the later stage of the Wolstonian (penultimate) glaciation, but
since there may have been one or more cold stages in the last interglacial interval (Shackleton
1969, 1977; Sutcliffe 1975, 1976; Woillard 1978) this cannot be established with certainty.

Bear Stratum (? late Wolstonian or later cold phase). This has a similar fauna to the Glutton
Stratum, but with some evidence from the nature of the deposits of an ameliorating climate,
cool but not as cold as the previous one.

Otter Stratum (?). This stratum occurs in a small side chamber of the cave. The otter referred to
is the extinct clawless otter Cyrnaonyx antiqua, also occurring in the Glutton Stratum but not
otherwise known in Britain. The stratum appears to consist of a mixture of two deposits, one
warmer and the other cooler, and to have been deposited in a period between the Glutton Stratum
and Hyaena Stratum depositions, with elements of both (Sutcliffe & Kowalski 1976).

Hyaena Stratum (Ipswichian Interglacial). The fauna is that of a warm interglacial period.

Reindeer Stratum (Devensian). The mammalian fauna of this period indicates a return to glacial
conditions; Man was present. An earlier Devensian cold stage is represented at the cave by ‘the
Head’, a deposit from which no bird material has been collected.

‘Eboulis’ (Devensian). The position of this deposit in the sequence of strata is not completely
certain, but it may represent another later phase.

Windmill Hill Cave, Brixham
(Devensian). The material from this cave has not been assigned to a series of strata.

The bird species identified, dating from the Devensian glaciation back to an earlier cold phase,
have been tabulated by strata (Table 1); to aid comparison the occurrences for the earlier and
later cold periods have, in addition, been grouped in single columns on either side of the list
for the single warm period between.

93

PLEISTOCENE BIRD REMAINS

= = x UdARY
= x = MOID UOLLUeD
= = = SUI[IRIS
= - x Ifqssor)
a e 7 oresplsla
is = = udig 921,
z * = YAPIAAS
= = x IMO 2[38eg
a a z pleysng [34]
= = x (‘Aou ‘ds) osprijieg U1a}sa\\
= = = asnoIH MOTIIM
= = = uPSIWIe}g
= = x [211895]
= = = piezzng uoWWIOD
<b = = Q[3v9-BIG Pd|Ie}-SIY MA
= = x Japursooy
hz ¥ = [PoL
= a = UOdSIMA
x x x yonpjeys uowUul0D
5 = = yonppeys Appny é
= = = asOOyH JUusIg

x x x YIOIS OYA
(plod) = (pjod)_~— (Pyoo) (PIO9) (waeM) (PIO2) (2) (009) = (pjod)_— (pyoo)
aAey) WwNnjeNg wnjens 3ye7T winjels Ajreo wunjeng wnjens Ieog winjens sa1oads
weyxlg sljnoqq, Joopulsy poulquiog euseAH] pourquiog JI3N0 Ieog /uoWN[H uoWN[H

Bat Ee eel ee ee eee

poynuapr saiseds yo aoueIInd9Q CT: aIqUL

94 C. J. O. HARRISON
Systematic list of material

? WHITE STORK, Ciconia ? ciconia. Two very similar species of stork occur in the Palaearctic
today, the White Stork and the Black Stork C. nigra. Both occur in marshy habitats, the former
in more open country and the latter in forest. Osteologically they are very similar, but with some
differences in the proportional lengths of limb bones. Since the fossil specimens are broken ends
of bones these differences cannot be used. Three Recent C. ciconia specimens and two of C. nigra
were available for comparison. The fossil specimens agree with those of C. ciconia in having on
the distal end of the tibiotarsus a broader anterior tendinal canal and a narrower tubercle on the
tendinal bridge.

The material comprises parts of carpometacarpi, radii, tarsometatarsi and tibiotarsi from the
Glutton Stratum, a distal end of a tibiotarsus from the Glutton/Bear Stratum transitional zone
and part of a radius from the Bear Stratum. A4095, A4110, A4117, A4118, A4135, A4143,
A4166.

BRENT Goose, Branta bernicla. Bone fragments of this species are sometimes difficult to separate
from those of Tadorna spp., but the present specimens can be assigned to it. The material com-
prises parts of tarsometatarsus and humerus from the Otter Stratum and parts of carpometacar-
pus, ulna and radius from the Hyaena Stratum. A4107, A4127, A4154, A4157.

COMMON SHELDUCK, Tadorna tadorna. Bones referable to this species form the major part of the
whole collection. When Lydekker (1891) examined the Brixham material he does not appear to
have used a specimen of this species for comparative purposes, although one was present in the
Museum collection at the time. He compared the bones with those of the Ruddy Shelduck T.
ferruginea and the South African Shelduck 7. cana. He seemed unaware that there is a strong
sexual dimorphism in the present species, the female being smaller, and he suggested that the
smaller specimens represented a new species. Possibly because of these comments most of the
shelduck material in the collection of the Department of Palaeontology had been tentatively
assigned to T. ferruginea. Skeletons of three T. tadorna and five T. ferruginea were available for
comparison. Differences in the limb-bones are slight, but in general 7. tadorna has shorter and
stouter bones and, with one exception, where differences were apparent the material was referred
to this species. Humerus, tarsometatarsus and to a lesser degree the femur are the most useful
bones for comparative purposes.

From the Tornewton Cave the material comprises tarsometatarsi, tibiotarsi, femora, coracoids,
humeri, radii, ulnae, carpometacarpi, anterior ends of sterna, and furcula from the Glutton
Stratum; tarsometatarsi, tibiotarsi, coracoids, humeri, ulnae, carpometacarpi, sacrum and verte-
brae from the Glutton/Bear Stratum transitional zone; tarsometatarsi, femora, coracoids and
ulna from the Bear Stratum; tarsometatarsi, tibiotarsi, femora, coracoids, humeri, radii, ulnae
and carpometacarpus from the Hyaena Stratum; and tarsometatarsi, tibiotarsi, coracoids, ulnae
and carpometacarpi from the Otter Stratum. A4091—3, A4096-7, A4099-4101, A4108, A4113-6,
A4136, A4155-7, A4162-4, A4170, A5S033—4, A5036.

From the Windmill Hill Cave, Brixham, we have tarsometatarsi, tibiotarsi, femora, coracoids,
scapulae, humeri, ulnae, radii, carpometacarpi, metacarpal, anterior ends of sterna, furcula and
vertebrae. 48911, 48915, A113, A3138-48, A3150-74, A3179-81.

? RUDDY SHELDUCK, Tadorna ? ferruginea. As mentioned above, there are difficulties in separat-
ing the bones of the two Tadorna species. The present specimen is a femur from the Hyaena Stra-
tum, A5037, which is longer and proportionally more slender than any in the available skeletons
of T. tadorna; it matches that of a male T. ferruginea. Specimens collected at the same time from
the same stratum are referable to T. tadorna.

WIGEON, Anas penelope. Coracoid and part of a carpometacarpus from the Hyaena Stratum.
A4111, A4124.

TEAL, Anas crecca. Femur, carpometacarpus and part of a humerus from the Hyaena Stratum.
A4104, A4121, A4122.

PLEISTOCENE BIRD REMAINS 95

GOOSANDER, Mergus merganser. Parts of humeri from the Glutton Stratum and parts of tibio-
tarsi and coracoids from the Otter Stratum. A4198, A4138, A5035.

WHITE-TAILED SEA-EAGLE, Haliaeetus albicilla. Claws, A233, collected by J. L. Widger and re-
ceived with other Widger specimens, are merely labelled ‘Torbryan Caves’ but are very probably
the ones referred to in his account (Walker & Sutcliffe 1967 : 80) as ‘claws of an immense bird’
and collected in the Bear Stratum of Tornewton Cave. There is also a single claw, A4141, from
the Tornewton cave but from unstratified material.

COMMON BUZZARD, Buteo buteo. It is possible that the Rough-legged Buzzard B. /agopus may
have occurred in this region in colder periods in view of its typical climatic and habitat prefer-
ences, but the present material is referable to the smaller B. buteo, the specimens matching those
of the smaller male of this dimorphic species. The material comprises parts of humerus, ulna,
carpometacarpus and pelvis from the Windmill Hill Cave, Brixham. 48913, 48917, A3149.

KesTREL, Falco tinnunculus. Parts of ulna and tarsometatarsus from the Glutton Stratum;
coracoid, carpometacarpus and ulna from the Hyaena Stratum. A4133, A4149-51, A4159.

PTARMIGAN, Lagopus mutus. Several bones of this species are present, but the characteristic
tarsometatarsi are a little smaller than those of Recent specimens. The material comprises parts
of tarsometatarsi and coracoid from the Reindeer Stratum outside the cave, and part of a carpo-
metacarpus from the “Eboulis’ deposit. A4125, A4147, A4148.

WILLOW Grouse, Lagopus lagopus. A tarsometatarsus from the Reindeer Stratum outside the
cave. A4106.

WESTERN PARTRIDGE, Alectoris sutcliffei sp. nov. See p. 96.

LITTLE BUSTARD, Otis tetrax. Part of a carpometacarpus from the Reindeer Stratum outside the
cave. A4112.

EAGLE OwL, Bubo bubo. Although incomplete the specimen agrees in size with the larger northern
Recent form of this species, whereas that from the earlier Cromer Forest Bed Series of East
Anglia (Harrison 1979) is similar to the smaller southern Recent form. The material is an ungual
from the Glutton Stratum. A4169.

SKYLARK, Alauda arvensis. An ulna, humeri and coracoid from the Hyaena Stratum; humeri,
coracoids and carpometacarpi from the Reindeer Stratum. A4102, A4103, A4134, A4152.

Tree Pipit, Anthus trivialis. There is little osteological variation among the pipits. Material is
part of a humerus from the Hyaena Stratum. A4123.

FIELDFARE, 7urdus pilaris. A coracoid from the Reindeer Stratum. A4192.

CrossBiLL, Loxia cf.curvirostra. The specimen resembles the corresponding bone of L. curvirostra,
and is smaller and less stout than that of L. pytyopsittacus. L. scoticus is intermediate in general
size between these two, and the Asiatic L. /eucoptera is smaller than L. curvirostra, but compara-
tive osteological material is not available. The breeding distribution of northern European cross-
bills is broadly linked with the occurrence of either Spruce or Pine trees, though birds may at
times take seeds of either as well as of other conifers. Nethersole-Thompson (1975) has suggested
that crossbills speciated from an ancestral form when their original principal foodsource, the
Spruce Picea abies, ceased to occur, to produce the Scottish Crossbill L. scoticus and the Parrot
Crossbill L. pytyopsittacus in Scandinavia and northern Russia, both specializing mainly on the
seeds of the Scots Pine Pinus sylvestris. Further east the Common Crossbill L. curvirostra per-
sisted as a specialist mainly on Spruce. The latter tree did not re-invade Europe after the last
glaciation until after the isolation of the British Isles, where neither tree nor bird is native,
although the first has been introduced and the latter has recently re-invaded and now breeds.
From the evidence of pollen deposits Spruce did occur in Britain through the Pleistocene until
the Lower Wolstonian, although in smaller proportions than other forest trees, and later as a
scarce and probably localized tree in Ipswichian to Lower Devensian times, but not subsequently

96 C. J. O. HARRISON

(West 1977). The present specimen might therefore have been co-extant with Spruce in south-
west England, and its normal existence was certainly correlated with the presence of conifers.
The material is part of a humerus from the Glutton Stratum. A4128.

STARLING, Sturnus vulgaris. On average these bones are slightly larger than those of Recent
specimens. Two of the humeri are atypical in lacking the anconal fossa near the proximal end of
the shaft. Carpometacarpi, ulnae, humeri, coracoid and femora are from the Hyaena Stratum;
femora, coracoids, humeri, ulnae and carpometacarpi are from the Reindeer Stratum. A4105,
A4130, A4139, A4140, A4144-6.

CARRION Crow, Corvus corone. The bones are towards the larger end of the size range of the
species and the tibiotarsus shaft is stout. We have part of a tibiotarsus from the Glutton/Bear
transitional zone and part of a femur from the Reindeer Stratum outside the cave. A4109, A4126.

RAVEN, Corvus corax. Parts of humeri, tibiotarsus and tarsometatarsus from the Glutton Stratum;
ulna and parts of coracoids, tibiotarsus and tarsometatarsus from the Hyaena Stratum. A4031,
A4044, A4119, A4120, A4132, A4142, A4160, A4161.

Systematic description
Order GALLIFORMES
Family PHASIANIDAE

Genus ALECTORIS Kaup, 1829
Alectoris sutcliffei sp. nov.
Western Partridge; Fig. 1

DIAGNOsIS. Small, distal end of tibiotarsus smaller than that of any known Alectoris species.

Name. After Dr A. J. Sutcliffe, who has worked extensively on Pleistocene mammal remains and
was responsible for assembling most of the present material.

MATERIAL. Holotype the distal end of a left tibiotarsus, collected by A. J. Sutcliffe in 1960;
BM(NH) no. A4165. Tentatively referred specimen a distal end of a tibiotarsus (illustrated as a
left-hand bone, but described in the caption as from the right) described by C. Mourer-Chauviré
(1975) as “Alectoris sp.’.

OccuRRENCE. Holotype from the Upper Middle Pleistocene; Glutton Stratum, Tornewton Cave,
Torbryan, Devon, England. Tentatively referred specimen from ? Gunz glaciation, Early Middle
Pleistocene; Mas Rambault a Frontignan, Hérault, southern France.

DESCRIPTION AND DISCUSSION. The specimen is the distal end of a tibiotarsus of a small phasianid
gamebird. It had tentatively been identified as that of a Common Partridge Perdix perdix but it
is stouter with less anteriorly-prominent condyles and a wider and more anteroposteriorly-
flattened shaft. It matches the tibiotarsi of Alectoris species but is smaller. Its width is 6-5 mm
and that of the French specimen, calculated from the scale provided, is c. 6-75 mm. For the Red-
legged Partridge, A. rufa, comparable measurements are 7-2—7:3 mm (5 specimens) and for the
Rock Partridge A. graeca, Barbary Partridge A. barbarae and Chukar A. chukar they range from
7-4 to 8-5 mm (11 specimens). The species is therefore distinctly smaller than others of this genus.
Mourer-Chauviré (1975) recognizes A. graeca and A. barbarae from cave deposits in southern
France back to the Mindel glaciation, but finds no evidence of A. rufa which is now the native
species of most of southern France. It is possible that the latter is derived from an Iberian isolate
of fairly late glacial origin. The western zone of the range of Alectoris, including the southern
site of A. sutcliffei, is now occupied by A. rufa, but the extinct sutcliffei occurred further north,
and in a colder period, and must have been more tolerant of low temperatures.

PLEISTOCENE BIRD REMAINS 97

MEASUREMENTS (in mm). Length of specimen 12:4; distal width 6:5; anteroposterior thickness
of internal condyle 6-4, of intercondylar groove 3-8, of external condyle 5-8; width and thickness
of shaft 4-5 x 2-7; distoproximal width of tendinal bridge 2:1.

The avian predators

The range of prey species found in caves is determined by the species of predator which fed
there, and from the specimens examined in the present study it seems likely that avian predators
were responsible for introducing most or all of the bird remains; they tend to leave bird bones
relatively complete whereas mammalian predators tend to splinter and destroy them. Avian
predators also tend to have a preferred prey size which may limit their range of prey significantly,
and they may concentrate on some easily available species.

t WY
Fig. 1 Alectoris sutcliffei sp. nov. Holotype, distal end of left tibiotarsus, BM(NH) no. A4165.
(Left to right) distal, internal, external, anterior views.

The predominant bird bones found were those of the Common Shelduck. This species presents
two problems; it is a large and heavy prey for an avian predator, and a bird of estuaries and
coasts. The coast is at present several miles from Tornewton Cave and is unlikely to have been
much nearer the cave in the past. The Shelduck feeds on tidal mudflats. Although it may fly
inland to utilize concealed nest sites, it is doubtful if this would explain the number of fossil
remains. The presence at the same period of the Brent Goose, a similar large species of tidal
mudflats, and possibly of the Wigeon which also occurs on this type of habitat in winter, tends to
support the suggestion of a predator hunting along the shore.

Foxes and wolves were present at the time, but even if they could produce the type of specimens
found, it seems unlikely that mammalian predators of this type would be persistently successful
in killing Shelduck, or would have transported them for long distances overland. Two avian
predators large enough to have taken such prey are known from the Tornewton Cave — the
Eagle-owl and the White-tailed Sea-eagle. The former is unlikely to have moved very far from the
cave to hunt, and may have concentrated mainly on mammalian prey. The Sea-eagle, however,
usually hunts and scavenges on seacoasts, taking numbers of ducks and larger waterfowl. The
Shelduck, which is conspicuous at a distance, feeds on mudflats where it cannot dive to avoid a
winged predator, and is fairly slow to take flight, would be a natural prey for the eagle. Sea-eagles
are not normally cave-nesters and it is not obvious how they would have utilized the Tornewton
site. It is possible that the rock platform at the top entrance might have formed a nest site from
which food and pellets could have fallen into the cave.

Shelduck remains, although numerous from the Glutton to the Hyaena Strata at Tornewton,
are suddenly absent in the Devensian. This could suggest a change in species distribution, but
since prey remains in caves are determined by the species of raptor present it could alternatively
indicate a change in the latter. The Brixham Devensian material consists almost entirely of
Shelducks, indicating that the species was certainly present for at least part of this period. An
absence of the Sea-eagle from Tornewton Cave in the Devensian period, possibly coincident with
the use of the cave by man, is likely to have been why the Shelduck bones are absent, rather than

98 C. J. O. HARRISON

a temporary absence of the latter species from this region. Since the larger prey are typical of the
Sea-eagle it may be significant that during the Devensian the only waterfowl remains at Tornewton
are those of the very small Teal.

The Common Buzzard of the Brixham Cave may have been there by accident or as a prey
item rather than a predator. It is just possible that this species might have utilized a cave in a very
bare and windswept biotope, but in inland areas it would normally have used trees where these
were available, or the ledges of larger rock outcrops, for roosting and nesting.

The Kestrel was present at Tornewton in the earlier cold period and the interglacial and may
well have been responsible for the presence of small passerine bird remains as well as those of
small mammals. Other species of falcon tend to concentrate almost entirely on birds. Other
raptors which produce small bird bones in prey remains are the small to medium-sized owls, of
which there is no evidence in the material from Tornewton, although the presence of the remains
of prey without those of the predator is not unusual where birds are concerned. There are no
Kestrel remains from the Devensian strata but the presence of small passerine remains from this
period suggests that this or a similar smallish raptor was present.

The species list as evidence of climate

Since there is evidence of climatic variation from one stratum to another in this cave material,
and since this is reflected in the mammal species present, similar changes in the bird fossils could
be expected. In fact, within this group the situation is more complicated because of the greater
mobility of birds. Resident bird species give some indication of the conditions at the time, but
within this region the majority of birds are partly or wholly migratory. The presence during a cold
period of a species which can tolerate the warmer summer conditions, combined with the presence
in a warmer period of cold-adapted species wintering in the area, can seriously confuse the general
picture which bird fossil may provide. There is some evidence of this problem in the present
material.

Of the species present in the earlier cold Glutton to Otter Strata, the White-tailed Sea-eagle,
Kestrel, Eagle-owl, Crow and Raven are now resident from the temperate zone at least as far as
the northern edges of the Boreal forest zone and may extend into wooded tundra (climatic zones
as in Voous 1960). The Brent Goose and Goosander have a similar range but make a more
southerly shift to warmer boreal and temperate parts in winter. The Shelduck has a mainly warm
temperate distribution at the present time but on the eastern Atlantic seaboard extends north to
northern Norway. This is correlated with the presence of the warm North Atlantic Drift current
modifying sea and shore temperatures, although in the Scandinavian part of its range the species
is only a summer visitor. The Western Partridge is a member of a genus of gamebirds normally
found in warm temperate to temperate zones, although the Rock Partridge and Chukar will
ascend to fairly high altitudes.

The two stork species are now summer visitors to Europe, nesting as far north as northern
Denmark and the Gulf of Finland, and Voous (1960) suggests that the northern limit of their
range is correlated with the July isotherm of 62 °F (16:7° C). The final species recorded for this
period is the Crossbill, and the occurrence of this species, unless it was a dying stray from some
irruptive movement, would indicate the nearby presence of conifer trees.

If it is reasonable to view this material as an entity, it suggests that during this early cold period
the surroundings of the cave might have resembled present-day southern Scandinavian boreal
conditions, with the presence of some conifer trees and a mid-summer temperature of c. 62 °F.
However, the winters might have been more severe (Sutcliffe & Zeuner 1962).

The species present in the Hyaena Stratum in the warmer interglacial are disappointing as
climatic indicators. The list could be one indicating birds now present in a temperate zone in
winter or a boreal zone in summer. The only exception is the tentatively-identified Ruddy Shel-
duck. This is mainly a species of drier steppe regions. The Starling may have roosted and nested
in the cave. The presence of both the latter species and the Skylark suggests that some open
grassland was present nearby.

PLEISTOCENE BIRD REMAINS 99

For the Devensian period, the presence of the Arctic/Alpine Ptarmigan indicates more definitely
a colder climate. The Willow Grouse might have been the typical bird of cold willow and birch
scrub, or possibly a Red Grouse form adapting to heather moorland. The grassland Starling and
Skylark were still present. Of the remaining species, Teal, Fieldfare and Crow are all species
that might occur from temperate to subarctic conditions, but are not tundra birds. The excep-
tional occurrence is the Little Bustard. This is a warm temperate species, extending into temperate
regions in summer, and at present just reaching northern France. It is, however, partly a migrant,
and it is possible that the Tornewton specimen was a stray individual that had overshot its usual
range at that time. At Brixham, the combination of Common Buzzard and Shelduck suggests
boreal rather than tundra conditions.

Sutcliffe & Kowalski (1976) have over 2000 specimens of rodents from Tornewton Cave and
with this more abundant material have produced a table of the relative proportions of the various
groups of rodents, part of which is shown in Table 2. It will be seen that they also found evidence
of boreal and forest species in the colder periods. Except for the presence of the Ptarmigan in the
Devensian, the tundra element cannot be confirmed among the birds.

Table 2 Relative proportions of the various ecologic groups of rodents
found in Tornewton Cave (from Sutcliffe & Kowalski 1976: 66)

Layer Tusnalee Boreal Forest Neutral
and steppe

Reindeer Stratum 22-9 8-7 5:5 62:8

Hyaena Stratum 0 0 2:3 97-7

Bear Stratum 25-8 48-4 9-6 16:1

Glutton Stratum 22:7 64:2 0:7 12:4

It is possible that the higher ground of Dartmoor and similar moorland areas could have
carried a tundra-type fauna and flora when more sheltered areas of lower ground might have had
a more boreal flora with some trees, as in parts of northern Scandinavia today. It also seems
possible that this area of south-west England, on the southern side of higher ground and close to
an oceanic shoreline, might have experienced, even under severe conditions and at different
stages of sea-level, a more favourable climate compared with sites further inland on the conti-
nental landmass. These opposite tendencies could have resulted in a more varied fauna and flora
at these earlier periods in this area.

Acknowledgements

I am very grateful to my colleague Mr C. A. Walker who helped me to trace and assemble this
material; to Dr A. J. Sutcliffe who provided much of the background information, and to Mr
A. P. Currant and Dr C. B. Stringer who helped with further information and discussion.

References

Harrison, C. J. O. 1979. Birds of the Cromer Forest Bed Series of the East Anglian Pleistocene. Trans.
Norfolk Norwich Nat. Soc. 25 : 277-286.

Lydekker, R. 1891. Catalogue of the fossil birds in the British Museum (Natural History). xxvii+ 368 pp.,
75 figs. London.

Mourer-Chauviré, C. 1975. Les oiseaux de Pléistocene Moyen et Supérieur de France. Docums Lab.
Géol. Fac. Sci. Lyon 64 : 1-624; 22 pls, 72 figs, 89 tabs.

Nethersole-Thompson, D. 1975. Pine Crossbills. 256 pp., 17 pls, 18 figs, 17 tabs, 3 maps. Berkhamsted.

Shackleton, N. J. 1969. The last interglacial in the marine and terrestrial records. Proc. R. Soc., London,
B 174 : 135-154.

100 C. J. O. HARRISON

— 1977. The oxygen isotope stratigraphical record of the late Pleistocene. Phil. Trans. R. Soc., London,
B 280 : 169-182.

Sutcliffe, A. J. 1975. A hazard in the interpretation of glacial-interglacial sequences. Quaternary Newsl.,
Glasgow, 17 : 1-3.

—— 1976. The British glacial-interglacial sequence — reply to R. G. West. Quaternary Newsl., Glasgow,
18 : 1-7.

—— & Kowalski, K. 1976. Pleistocene rodents of the British Isles. Bull. Br. Mus. nat. Hist., London,
(Geol.) 27 : 33-147; 31 figs, 13 tabs.

& Zeuner, F. E. 1962. Excavations in the Torbryan Caves, Devonshire. 1. Tornewton Cave. Proc.
Devon archaeol. Explor. Soc., Torquay, 5 : 127-145. 2 pls.

Voous, K. H. 1960. Atlas of European birds. 284 pp., 355 text photos, 419 maps. London.

Walker, H. H. & Sutcliffe, A. J. 1967. James Lyon Widger, 1823-1892, and the Torbryan Caves. Rep.
Trans. Devon. Ass. Advmt Sci., Plymouth, 99 : 49-110.

West, R. G. 1977. Pleistocene geology and biology. 2nd Edn. xi+ 440 pp., 16 pls, 151 figs, 48 tabs. London.

Woillard, G. M. 1978. Grand Pile Peat Bog; a continuous pollen record for the last 140,000 years.
Quaternary Res., New York & London, 9 : 1-21.

The succession of Hyracotherium (Perissodactyla,
Mammalia) in the English early Eocene
J. J. Hooker

Department of Palaeontology, British Museum (Natural History), Cromwell Road, London
SW7 5BD

Synopsis
In the light of recently-collected material and the documentation of old stratigraphical information, the

English species of Hyracotherium are re-evaluated with a better knowledge of variation and diagnostic
characters. A phylogeny is suggested and some correlations within Europe are proposed.

Introduction

The morphology of British Hyracotherium has been exhaustively treated by previous authors
(Owen 1841, 1842, 1858; Cooper, C. F. 19326; Simpson 1952) but two types of data, one old and
one new, put it in a different perspective and produce somewhat different taxonomic and evolu-
tionary conclusions. These data are a good knowledge of the stratigraphic position of most of the
specimens, which had not previously been fully appreciated, and an increase in the number of
specimens known from the Blackheath Beds and the Suffolk Pebble Beds by recent collecting,
giving a better idea of the degree of variation and more reliable dental characters. Since Simpson’s
(1952) revision of British material new specimens have been collected, mainly from Abbey Wood,
by S. A. Baldwin, D. Bone, J. Collins, J. Cooper, A. Gale, W. H. George, P. R. Gurr, K. J. Hall,
R. A. D. Markham, A. R. G. Packman, P. R. Payne, D. J. Ward, K. Wright and the author.
The registered numbers of specimens in the various institution collections are indicated by the
following prefix letters: British Museum (Natural History), London (M or no prefix); Sedgwick
Museum, Cambridge (C); Ipswich Museum, Suffolk (IM); Muséum National d’Histoire Natur-
elle, Paris (MU, AV or AL); University of California Museum of Paleontology, Berkeley (UC).
Symbols used in synonymies as in Matthews (1973).

Stratigraphy

Simpson (1952 : 202), referring to the holotypes of H. /eporinum and H. vulpiceps, stated: ‘their
relative levels within the London Clay have not been, and perhaps cannot be, established’. The
horizon of the latter is recorded by Owen (1858 : 55) as the Roman Cement bed at Harwich
(=Harwich cement stone (Davis & Elliott 1951 : 331) and the Harwich Stone Band (Elliott
1971)) which occurs 6 m above the base of the London Clay. This is now known to be an import-
ant ash marker (Elliott 1971) and is correlated with a Greenland volcanic episode dated at
53 Ma (Fitch et al. 1978). The lithology of the adherent matrix shows the quoted provenance to
be correct. The age-equivalent deposit in Denmark, the Mo Clay, belongs to the calcareous
nannoplankton zone 10.

The horizon of the holotype of H. /eporinum is less accurately known but details of its collection
from the cliff at Studd Hill, near Herne Bay, were given by Richardson (1841). Remains of the
crinoid Jsselicrinus subbasaltiformis (Miller) were formerly to be found (Cooper, J. 1977 : 173;
Gamble 1979) about half-way up the cliff before the section was obscured by sea defences. This
species has a restricted range in the London Clay between about 27 m and 37 m above the base
(= Division 2 of Wrigley, 1924; King 1970). Its main occurrence at Studd Hill is restricted to
31-32 m above the base (Gamble 1979). The holotype of Hyracotherium leporinum evidently
came from the vicinity of this horizon and thus occurs significantly higher than the holotype of H.

Bull. Br. Mus. nat. Hist. (Geol.) 33 (2): 101-114 Issued 31 January 1980
101

102 J. J. HOOKER

vulpiceps. A second specimen of H. /eporinum from the Herne Bay foreshore (C. B. Brown collec-
tion), now in the Sedgwick Museum, Cambridge, is most likely to be from the same horizon.

An imperfect cranium of H. vulpiceps (38801) is not, as quoted by Cooper, C. F. (19326 : 433)
and Simpson (1952 : 195), from Sheppey but, according to the British Museum (Natural History)
Palaeontology Department register, ‘from the Crag (wreck of London Clay)’; i.e. from the Red
Crag basement bed of Suffolk. This is of late Pliocene age but contains much material derived
from the London Clay. The dark reddish, highly polished surface of the adherent matrix shows
the quoted provenance to be correct. The original horizon in the London Clay is, of course,
unknown but today only the lower part is preserved in Suffolk below the unconformity at the
base of the Red Crag.

Further British specimens that have been attributed to Hyracotherium are two mandibular
fragments without tooth crowns (38939, 47986; see Lydekker 1886 : 11) from the London Clay
of Sheppey. They are generically indeterminable but the horizontal ramus (38939) is very similar
to that of the holotype of H. vulpiceps in size and morphology. Further specimens would be very
interesting in view of the later Cuisian age indicated by Costa & Downie (1976 : 600, 603) for
this part of the succession.

Specimens, here identified as H. aff. vulpiceps, from Abbey Wood, occur in the Abbey Wood
member of the Blackheath Beds which underlie the London Clay. The associated fauna is as
follows:

Charlesmooria childei Kiihne, 1969

? Didelphodus sp.

Didelphodontinae indet.

? Eochiromys sp.

? Macrocranion sp.

Adapisoricidae indet.

‘Adapisorex’ anglicus C. F. Cooper, 1932a
Phenacolemur cf. fuscus Russell, Louis & Savage, 1967
Teilhardina sp.

Pelycodus eppsi (C. F. Cooper, 1932a)
Esthonyx sp.

Microparamys sp.

Paramyidae indet.

? Prototomus sp.

Oxyaena sp.

Arctocyonidae indet.

Hyopsodus wardi Hooker, 1979
Lessnessina packmani Hooker, 1979

? Phenacodus sp.

Coryphodon eocaenus Owen, 1846
Hyracotherium aff. vulpiceps (Owen, 1858)
Protodichobune sp. 1|

The specimens of H. cuniculus from Kyson came from the Suffolk Pebble Beds, which imme-
diately underlie the London Clay in Suffolk, and further specimens are here recorded for the first
time from the same beds at Bramford and Ferry Cliff. The Suffolk Pebble Beds have often been
synonymized or equated in age with the Blackheath or Oldhaven Beds because of their identical
position below the London Clay and their superficial lithological similarity. However, their
mammal fauna is different from that of Abbey Wood, its overall aspect being more primitive and
rather similar to those of Dormaal and Erquelinnes, Belgium (Quinet 1969; Gingerich 1976 :
text-fig. 18). The fauna is as follows (locality initials indicated for Bramford, Ferry Cliff, Harwich
and Kyson):

Leptictidae indet. K, FC
Paschatherium dolloi (Teilhard de Chardin, 1927) FC

ENGLISH EOCENE H YRACOTHERIUM 103

Plesiadapis aff. remensis Lemoine, 1887 K

P. aff. tricuspidens Gervais, 1877 Re
Omomyidae indet. FC
Pelycodus sp. K
Paramys sp. K, Fe
Meldimys sp. Gat

? Prototomus sp. K
Creodonta indet. K
Mia-idae indet. K
Landenodon sp. 1K JC, Jal
Microhyus musculus Teilhard de Chardin, 1927 K, FC, H
Hyopsodus cf. wardi Hooker, 1979 K

? Coryphodon sp. FC
Hyracotherium cuniculus Owen, 1842 Kaas
Protodichobune sp. 2 K

Boswe!! (1916 : 571) stated: ‘the lithology and petrology of the pebble bed suggests that it
should be grouped with the Reading Beds, but its fauna is a typical London Clay assemblage’;
but also (1916 : 566) ‘unfortunately the species [molluscs and fish] are of rather wide range,
and are not sufficient to form the basis of any correlation’. The Suffolk Pebble Beds (White
1931 : 40) are here accepted as distinct from the Blackheath or Oldhaven Beds, and somewhat
older. The stratigraphical evidence for the English Hyracotherium succession is shown in the
geological columns in Fig. 6, p. 111.

Morphology

Most authors have agreed on the separation of at least two English Hyracotherium species on the
basis of size (smaller H. cuniculus and larger H. leporinum and vulpiceps), but opinions on the
validity of the separation of the two larger ones have varied (see synonomy lists). Simpson
(1952 : 202) found the teeth to vary so much that ‘no other specimens show exactly the same
distinction as these types [the holotypes of H. /eporinum and H. yulpiceps] and none can be
referred with real clarity to one species rather than the other’. In the stratigraphically-controlled
assemblages mentioned above certain characters do in fact show a distinctness which can be
used for specific separation. On the other hand, additional specimens from Abbey Wood streng-
then Simpson’s (1952 : 202) view of the variability of other characters suchas the lingual cingulum.
The characters found to be of low variability and useful in separating species and recognizing
trends are as follows:

1. Configuration of upper molar centrocrista (and to a certain extent the presence of a meso-

style)

2. Structure and size of M; hypoconulid lobe

3. Shape and structure of P*®

4. Overall size of teeth.

Features of the skull discussed in detail by Simpson (1952 : 196-200) and Savage et al. (1965 :
49-50) are not considered here because they are known for only two specimens (one each of H.
leporinum and H. vulpiceps) and their variability cannot be assessed.

Systematic descriptions

Family EQUIDAE Gray, 1821
Genus HYRACOTHERIUM Owen, 1841
1841 Ayracotherium Owen : 203-208.
1846 Macacus Desmarest; Owen : 1-10.

1858 Pliolophus Owen : 54-71; pls 2-4.
1876 Eohippus Marsh : 401.

104 J. J. HOOKER

1880 Lophiodochoerus Lemoine : 589 (fide Teilhard de Chardin 1922 : 67).
1896 Protorohippus Wortman : 92, figs 14-15.

TYPE SPECIES. Hyracotherium leporinum Owen, 1841.

REMARKS. The most recent generic diagnosis (Kitts, 1956) is evidently unsatisfactory as the type
species tends to have a weak upper molar mesostyle and H. vulpiceps either has no P+—P3 dias-
tema or this feature is intraspecifically variable. A revised diagnosis should await a full revision
of the North American species.

Hyracotherium cuniculus Owen, 1842
Fig. la-c

vy 1839 Macacus; Wood : 444-445, text-fig. 57.

v 1839 Macacidae; Owen : 446-448, text-fig. 58.

v* 1842 Hyracotherium Cuniculus Owen : 1-2, text-figs 2, 5.

vy 1846 Macacus eocaenus Owen : 1-10, text-figs 1, 3.

vy 1927 Hyracotherium sp.; Teilhard de Chardin : 27-28, text-fig. 28c; pl. 5, fig. 22.

vy 19326 Hyracotherium cuniculus Owen; C. F. Cooper : 438-441, text-figs 1A, C, 2B, E, F; pl. 51,
figs 6-9.

vy 1952 Hyracotherium cuniculus Owen; Simpson : 196-204, text-fig. 4B.

Types. The left M*, 36569 (Owen 1842 : text-fig. 2, reversed) was chosen as lectotype by Simpson
(1952 : 196). One of the figured paralectotypes, the right M!/2 M29709 (re-registered from 36569;
Owen 1842 : text-fig. 5, reversed), is correctly identified. The distal half of a left M!/? M14111
(reregistered from 36572) may be Owen’s unfigured paralectotype. Owen’s supposed upper
premolar (1842 : text-fig. 4, third paralectotype) looks most like M29699, which is a right M?
of Landenodon.

OTHER MATERIAL. This is listed in the table of measurements, Table 1.

TYPE HORIZON AND LOCALITY. Suffolk Pebble Beds of Kyson, near Woodbridge, Suffolk (Nat.
Grid TM 270475; see Prestwich 1850 : 272-273).

RANGE. Suffolk Pebble Beds of Bramford (TM 130477; Cooper, J. 1976) and Ferry Cliff near
Woodbridge (TM 278486; George & Vincent 1976 : 25), both Suffolk. Also the Sables d’Erque-
linnes of Erquelinnes (=Jeumont), Hainaut, Belgium (Rutot 1881).

EMENDED DIAGNOSIS. Small species of Hyracotherium (see Table 1), length of P?-M® estimated
35mm. Upper molars with straight centrocrista and no trace of a mesostyle. Lower molar
metalophids tend to join trigonids nearer to protoconids than to metaconids. M, hypoconulid
lobe small, tending to be buccal in position. P® with small protoconule; centrolingually placed
protocone; straight mesiolingual margin; no postprotocrista.

Discussion. Five individuals with lower molars and specifically four M,;s, four upper molars
and two P%s are available from the Suffolk Pebble Beds to demonstrate the constancy of the
diagnostic characters. The right M, figured by Teilhard de Chardin (1927 : pl. 5, fig. 22) from
Erquelinues fits H. cuniculus in size and morphology as this author noted. According to Quinet
& Verlinden’s (1970) measurements, the Erquelinnes mandible figured first by Rutot (1881) is
too big for H. cuniculus, being about the size of H. vulpiceps, but the metalophid joins the trigonid
just buccal to the midpoint, more like H. cuniculus than H. vulpiceps. The small diastema between
P, and P, is probably of little taxonomic value.

Hyracotherium vulpiceps (Owen, 1858)
Fig. 3a—c

v* 1858 Pliolophus vulpiceps Owen : 54-71; pls 2-4.
vy 1865 Hyracotherium leporinum Owen; Owen : 340; pl. 10, fig. 2.
y 1886 Hyracotherium leporinum Owen; Lydekker : 11.

ENGLISH EOCENE H YRACOTHERIUM 105

Table 1. Maximum length (1) and width (w) measurements (mm to the nearest 0-1) of English
Hyracotherium. The following localities are indicated by their initials: Bramford, Ferry Cliff, Kyson,
Abbey Wood, Harwich, Herne Bay, Suffolk (Red Crag basement bed). L=left, R =right.

Hyracotherium cuniculus

Number Locality tooth ] Ww Number Locality tooth ] Ww
M14112 K LP®* (4.4) 5:2 M29701 K LM1;2 - 3-9
36572 K Lee 5-1 5-6 RM, 7:3 $-1
M29709 K RM? (63) 8-1 Mles6 Be ie (8:8) 46
M14111 K LM?/2 _ (6:4) M36495 FC RMij23 — 5:0
M36496 FC LM? 6:6 - M29710 K 3 8-7 47
IM 1971-169 B LM?2 1S - M14113 K RM; 8-5 49
36569 K LM? 6:4 7-6 IM 1951-28-25 K RM; 8-4 48
Hyracotherium aff. vulpiceps
a RP®* 6:0 6:8 M29646 AW LDP? 5-4 -
RP? 6-7 8-1 RDP, 6:9 4:5
ee SRNL 7-65 99 N18 7163=4 als CAWile RIMaolal @:200005e3
RM? 8-3 10-0 RM, 719 6:0
LDP? Wil 8-1 RM, 72 5:1
Bize5Gt ee alee TF 92 M29719-21 AW RM, 83 56
4 = : . .
M25129 AW { LDP 6-6 8-1 RM; 9:8 5-1
‘ULM WES 8-9 M26906 AW RM, 76 53)
M32142 AW IL) P= 673 8:0 ; ‘RM; 10-5 5:5
M29716 AW RM‘? 77 = (8:8) M29360 AW RP; 35) 307)
M29754 AW RM? ee. 8-7 M29715 AW IER, 5:9 43
M15140a AW RM?/2 7-6 - M32127 AW RM1;2 8-0 5:6
M15139 AW LM? 78 9:8 M15143 AW RM;,;2 8-2 5:6
M13762 AW LM? 79 9-2 M29744 AW LM; 10-3 5:4
M15138 AW LM? 8-4 (8-8) M15130 AW RM; 10-6 55)
M15140b AW RM?/2/3 - ~ M13765 AW RDP; - 3:6
M29231 AW DP? 6:8 6:2 M29645 AW RDP, 7:3 4-6
Hyracotherium vulpiceps
RP? - - RP, (3-8) (2:2)
RP? 56 43 RP, (6:0) (2:7)
RP® 6-1 6:6 RP; (6:3) (3:8)
44115 H RP? 66 79 44115 H LP, 6:6 4-6
RM? qP22 9-0 LM, 7:4 35)
RM? 89 10-1 LM, 8:5 (6-0)
LM? 7:5 8-9 RM; (10:5) 61
|b) ee (6:5) (3-5)
RP* 6-4 6:7
RP* V2 8-3
38801 S RM? 7-6 9-5
RM? 8:2) 10:2
RM? 7-6 9:5
Hyracotherium leporinum
RP? 5:9 3-5 RIP 69 4-1
ILI 6:8 TS EB 6:7 7:0
M16336 HB LP? 6:9 8:5 C21361 HB [jee 7:6 8:8
RM?! 1k 97 LM? 8:0 10:3
RM? 8-6 10-5 LM? 85 11-0

RM* 8:5 10-0 RM° 8-1 10-4

106 J. J. HOOKER

vy 1901 Hyracotherium leporinum Owen; Depéret : 200-201; pl. 4, fig. 1.

Vv 19326 Hyvacotherium vulpiceps (Owen); C. F. Cooper : 432-436, text-figs 2D, 3A; pl. 50, figs 1-3;
pl. 51. figs 1-3.

vy 1952 Hyracotherium vulpiceps (Owen), Simpson : 195-206; pl. 37; pl. 38, figs a—d.

vy 1952 Hyracotherium sp.; Simpson : pl. 40, fig. a.

HovortyPe. The skull and incomplete skeleton 44115 (also registered on 44115a and M10657-61)
still partly embedded in matrix. Simpson (1952 : 196) has described the history of this specimen.
A subsequent misfortune has been the disappearance of the almost complete right mandibular
ramus which appears to have happened about 1960.

OTHER MATERIAL. Damaged cranium with left P?-M? and right P?-M3, 38801.

TYPE HORIZON AND LOCALITY. The Harwich Stone Band (= Roman cement bed), lower London
Clay of Harwich, Essex.

RANGE. 38801 is from the Red Crag basement bed, assumed derived from the London Clay.

EMENDED DIAGNOSIS. Medium-sized species of Hyracotherium (see Table 1), length of P?-M#
41 mm in holotype. Upper molars with very slightly buccally flexed centrocrista and a very faint
rib in the mesostylar position, separated from the buccal cingulum. Lower molar metalophids
tend to join trigonids equidistant between protoconids and metaconids. M, hypoconulid lobe
relatively large and with a median main cusp. P? with large protoconule; distolingually placed
protocone; slightly convex mesiolingual margin and little or no postprotocrista.

DOUBTFULLY REFERRED SPECIMENS. There are four other specimens the size of H. vulpiceps in the
Ipswich Museum from the Red Crag basement bed, presumed to be derived from the London
Clay. A palate (IM 1935-64a: E. C. Moor collection) with all the cheek tooth crowns broken off
has a P? with typical vulpiceps outline but, unlike the holotype, a short diastema between P! and
Pp’. A Pi-P2 diastema was used as a generic character by Kitts (1956). A right maxillary fragment
with P*-M? (IM 1935-64b) has upper molar centrocrista with buccal flexure intermediate between
the condition in H. vulpiceps and that in H. leporinum, and P* with a slight distal protoconal crest
trending distobuccally. A right mandibular fragment with M,_; (IM 1935-64c) has a typically
long M, hypoconulid lobe. The fourth specimen (IM unnumbered, old collection) is from Falken-
ham, Suffolk, and is a palate with apparently left and right DP?—-M? very abraded. The last two
specimens could, on the basis of size, be referred to either H. vulpiceps or H. leporinum.

Hyracotherium aff. vulpiceps
Fig. 2a—c

v 1931 Ayracotherium sp.; White : 18, 25.
v 1932b Hyracotherium vulpiceps (Owen); C. F. Cooper : 433, 437-438; pl. 51, figs 4-5.
v 1952 AHyracotherium sp.; Simpson : pl. 40, figs b—d.

MATERIAL. See Table 1.

RANGE. Abbey Wood member of the Blackheath Beds, Abbey Wood, London Borough of
Bexley (TQ 4801 7864). Also possibly the Blackheath Beds of Bean, Kent (TQ 590717) based on
an upper molar hypoconal fragment (M32179, J. Cooper collection).

Discussion. The Abbey Wood specimens show a very close resemblance to the holotype of
H. yulpiceps in size and morphology, with the exception of the unique P* (M13761). This differs
in having a small protoconule, centrolingually placed protocone and straight mesiolingual margin.
It thus resembles P? of H. cuniculus in morphology, but is larger. The combination of very small
numbers and different morphology of P* in the H. vulpiceps/aff. vulpiceps complex can be treated
in two different ways. We may combine them into one species on the basis of molar identity and

ENGLISH EOCENE H YRACOTHERIUM 107

assume that the P¥s are variable, or we can consider the two assemblages as different on a typo-
logical basis, but closely related. The second approach is here preferred as it is less subjective and
in particular because the P* of H. cuniculus and of H. leporinum are distinct, contrary to Simpson’s
statement (1952 : 203).

The constancy of molar characters in H. vu’»iceps and H. aff. vulpiceps is supported by the upper
molars of eleven individuals, the lower molars of eight individuals and specifically the M,s of five
individuals. The lei: M*® (M13762, figured C. F. Cooper 19326: pl. 51, fig. 4 lower) shows a
straight centrocrista probably as a result of the tapering of the end of the tooth row, which has
also resulted in a very lingual metacone.

Hyracotherium leporinum Owen, 1841
Fig. 4a—b

v* 1841 Ayracotherium leporinum Owen : 203-208; pl. 2, figs 1-4.

vy 1846 Ayracotherium leporinum Owen; Owen : 419-423, text-figs 165-167.

v 19326 Hyracotherium leporinum Owen; C. F. Cooper : 431-446, text-fig. 3B; pl. 49; pl. 50, fig. 4.
vy 1952 Hyracotherium leporinum Owen; Simpson : 195-206; pl. 39, fig. a.

vy 1952 Hyracotherium sp.; Simpson : pl. 39, fig. b.

HovoryPe. Facial part of cranium with alveoli or roots of both canines, both P's, left P?~%, left
M®? and crowns of right P?-M? and left P*-M? (M16336).

OTHER MATERIAL. Palate with left and right P?-M® (C21361).

TYPE HORIZON AND LOCALITY. Approximately Division 2 of the London Clay; Studd Hill near
Herne Bay, Kent (TR 153677). C21361, labelled ‘foreshore Herne Bay’, is probably also from
here:

EMENDED DIAGNOSIS. Medium-sized species of Hyracotherium (see Table 1), length of P?-M?
43 mm in holotype. The two available specimens have slightly larger upper premolars and slightly
wider upper molars than H. vulpiceps (see Fig. 5c-d). Upper molars with strongly buccally flexed
centrocrista and weak mesostyle. (Lower molars unknown.) P® with small protoconule; centro-
lingually placed protocone; slightly concave mesiolingual margin; strong postprotocrista present.

Discussion. Of the diagnostic characters only the degree of development of the mesostyle is seen
to be slightly variable between the two known upper dentitions; im C2136] it joins the buccal
cingulum. In particular the degree of buccal flexing of the centrocrista does not vary and is distinct
from H. vulpiceps and H. cuniculus. The fact that P? has a postprotocrista and a slightly concave
mesiolingual margin differentiates it from H. cuniculus (and from H. vulpiceps). Simpson’s
(1952 : 203) statement that they are very similar is incorrect and his text-fig. 4 is inaccurate.
Detail of the postprotocrista in the two specimens is different, it being single and sharp in C2136]
and a double structure in the holotype. In the holotype, P* and P* have a distal protoconal crest

?=postprotocingulum) which fades before reaching the distal cingulum; it is missing on C21361.

Evolutionary trends

Knowing the stratigraphical occurrences of most of the English specimens of Hyracotherium,
one can see a progression of morphologies through time which shows evolutionary trends, if not
the evolution of a single lineage. Figs 1-4 show the changes in P?, M!? and M, in four stages.
There is a size increase from stage | to 2 and from stage 3a to 4a (the slight width increase of the
upper molar from stages 3 to 4 is demonstrated in Fig. 5d, p. 109); the upper molar morphology
remains the same from stage 2 to 3; the M; hypoconulid lobe enlarges, lengthens and becomes
more median in position from stage | to 2 (and remains unchanged from 2 to 3). At the same
time the attachment of the metalophid to the trigonid shifts lingually to a more median position.
The actual lengthening of the hypoconulid lobe is demonstrated in Fig. 5a by measuring distances
between cusp tips. H. aff. vulpiceps also shows a relative lengthening of the whole Mg over H.
cuniculus (Fig. 5b). The specimen of H. aff. vulpiceps which overlaps greatly with H. cuniculus

108

J. J. HOOKER

Mz UNKNOWN

Fig. 1 A yracotherium cuniculus, from the Suffolk Pebble Beds of Kyson, Suffolk : (a) left P* (reversed)
(36572); (b) right M*/2 (M29709); (c) right M, (M29710). All x3.

Fig. 2 H. aff. vulpiceps, from the Blackheath Beds of Abbey Wood, London: (a) right P® (M13761);
(b) left M! (reversed) (M25129); (c) right Ms (M26906). All = 3.

Fig. 3H. vulpiceps, from the Harwich Stone Band, London Clay, Harwich, Essex, holotype (44115):
(a) right P*; (b) composite left and right M! (shown as right); (c) right M, (drawn from photograph

in Cooper, C. F. 19326: pl. 51, fig. 1). All x3.
Fig. 4 H. leporinum, from approximately Division 2, London Clay, Herne Bay, Kent; holotype

M16336): (a) right P®; (b) right M*. All x3.

ENGLISH EOCENE H YRACOTHERIUM 109

in hypoconulid length (M15130) is nevertheless typical in its overall length/width proportions
and appears to have achieved this by relative elongation of the talonid.

Other variable characters such as cingular development and outline shape of upper molars as
well as different wear stages (specimens figured chosen as far as possible to show similar wear)
make the specimens shown in Figs 1-4 appear more different from one another than their specific
diagnoses indicate.

|_| H.cuniculus

J H.vulpiceps & aff.

Fig.5 (a) Histogram of perpendicular distance from hypolophid to hypoconulid (a), as a percentage of
hypolophid length (b), in M; of Hyracotherium cuniculus, H. vulpiceps and H. aff. vulpiceps; (b-e)
Scatter diagrams of maximum length (|) and width (w), in mm, of M3, P*, M!~* and M® respectively.
Symbols: H. cuniculus (0), H. aff. vulpiceps (Q), H. vulpiceps (©) and H. leporinum (4). Symbols
in brackets indicate estimates; T indicates a plot of the primary type of a species; in Fig. 5d pre-
ultimate molars identified as M! or M? are indicated by solid left or right half symbol, respectively.

C. F. Cooper (19326 : 442-443), unaware that H. leporinum occurred above H. vulpiceps,
considered the latter to be more advanced because its P? resembled that of North American late
Wasatchian forms such as H. venticolum (Cope), which were molarizing their premolars in the
direction of Orohippus. In England, and probably the rest of Europe, the initial trend of molar-
ization of P? by the North American equid method (i.e. by enlargement of the protoconule and
distal shift of the protocone) changed to the European palaeothere method (i.e. by addition of a
hypocone behind the protocone, as occurred in the molarization of P* in both North American
and European equoids) in the evolution of Hyracotherium vulpiceps to H. leporinum. It is possible
that it was at this time that the Ypresian transgression finally isolated the North American
and European Eocene mammal faunas, whereupon the palaeothere and equid characteristics
established themselves.

Potential complications to the evolutionary scheme envisaged here are the vulpiceps-sized
mandible occurring alongside the M, of H. cuniculus at Erquelinnes, and a poorly preserved
palate (IM, unnumbered), with no data but in preservation identical to those from the Red
Crag basement bed, the size of H. cuniculus. Nevertheless, the English sequence of species is a
useful standard which might be recognized elsewhere in Europe and be used in biostratigraphic
correlation. At the moment, however, the limits of the stiatigraphical ranges of the various

110 J. J. HOOKER

species are too poorly known to erect legitimate biozones for them. Some correlations, however,
ure suggested belew. The occurrences of H. cuniculus in the London and Belgian Basins probably
mark the time of the initial migration into Europe of the Snparnacian mammal fauna. Hyraco-
therium species from the French Sparnacian, especially the higher parts, are not so easily com-
parable with the English species, and it is possible that the transgression of the Ypresian sea had
by this time isolated the two areas, allowing evolution to occur independently.

Foreign specimens attributed to English species

Savage eft al. (1965) described Hyracotherium from three Paris Basin Sparnacian sites (Pourcy,
Mutigny and Avenay). At each the size variation for the few isolated teeth known exceeds that
for the English species. In fact the measurements span those of H. cuniculus and either H. vulpiceps
or H. leporinum. This fact was recorded by Savage et al. (1965) only for the Avenay locality and
is reflected in their ‘cf. cuniculum’ (sic) identification of the smaller species there. In morphology
all these French specimens differ from both H. cuniculus and H. vulpiceps in the frequent, but
not constant, presence of buccally flexed centrocristae and development of mesostyles on the
upper molars. Their figures of the upper molars from Pourcy (1965 : text-fig. 3b-d) show slight
buccal flexing of the centrocrista. The smaller M?/? (1965 : text-fig. 3b) appears to have a fairly
strong mesostyle whilst the larger M®s (1965 : text-figs 3c-d) appear to have a weaker one some-
what like H. vulpiceps. The two M?’s from Mutigny (1965 : text-figs 2f, 21) appear to have little
flexure of the centrocristae and little or no sign of a mesostyle. However, another M* (MU12371)
and an original (MU12329) and a cast (MU L-218) of two M'!/*s, which I was able to see when
in the Muséum National d’Histoire Naturelle, Paris, have the same centrocristal flexure as H.
leporinum and a slightly greater mesostyle development. The smaller type from Avenay (referred
to by Savage et al. (1965 : 11) as H. cf. cuniculum) differs morphologically from H. cuniculus.
The M!/? AV4790 (1965 : text-fig. 4c) has a strongly buccally flexed centrocrista and has a meso-
style, the P* (1965 : text-fig. 4b) has a weak fold on the distal protocone wall, like the holotype
of H. leporinum, and the P® (1965 : text-fig. 4a) appears to have a well-developed postproto-
cingulum. The M’!” of the larger type from Avenay (1965 : text-fig. 4f) appears to have a weak
centrocristal flexure and no mesostyle. All the M,s from these three French Sparnacian sites have
a relatively long hypoconulid lobe as in H. vulpiceps and unlike H. cuniculus.

AL5198, the left mandibular fragment with P,-M, from Lemoine’s Ageian fauna (a mixture
from both the type Sparnacian and the younger Sables a Unios et Térédines), was originally
described by Teilhard de Chardin (1922 : 70, text-fig. 33C; pl. 3, fig. 30) as Propachynolophus
sp. nov. or sp. ind. Savage ef al. (1965 : 13, text-fig. 4h) referred it to Hyracotherium and suggested
that it was older than Propachynolophus (i.e. Sparnacian) and morphologically intermediate
between the two. Its short hypoconulid lobe is reminiscent of H. cuniculus but there is no lingual
accessory cusp and the animal was closer to H. vulpiceps in size. The apparently primitive M,
suggests that this specimen might be from earlier in the Sparnacian than those from Pourcy,
Mutigny and Avenay and it is possible that affinities might lie with the vu/piceps-sized mandible
from Erquelinnes.

Teilhard de Chardin’s (1922 : 52) fig. 26 indicates three teeth of a Hyracotherium the size of
H. cuniculus from the Conglomérat de Meudon, near Paris (at base of the Argile Plastique,
considered Sparnacian in age). His discussion, however, indicates an animal the size of H.
leporinum and the M, hypoconulid lobe is indeed longer than that of H. cuniculus.

Correlation with continental Europe

Unfortunately, the Paris Basin Hyracotherium material is still too poorly known to identify
specifically, but the specimens from Mutigny and Avenay appear to be as advanced as the
stratigraphically highest English species (H. /eporinum.). This conforms with Gingerich’s (1977 :
fig. 8) proposed succession of Sparnacian localities based on the phylogeny of adapid primates.
It also suggests, however, that the French Sparnacian localities Mutigny and Avenay are equiva-
lent to the early Ypresian of England. Erquelinnes can be correlated with the Suffolk Pebble

ENGLISH EOCENE H YRACOTHERIUM 111

sTaGes|Np M2!S ERNE BAY LONDON SUEEOER Sarr ia
KAN IVAN}
s ape eral si 3 sa Se 7| ena tar ete fai =
a E 5 E E 10
= T}) js < 5 Z| < 2 <
zie} i411] el} [se a 3 3
ie} ov
< |< isl Ee = = 2 = 3 metres
w |, |< ® gs caee Olena ili) a = — gaa as Ss fe akOre a
Ww j— ;
ac |j> = = =~ = 0
a 2c = =e
o o o
= [Rei sie 3 z 3
= 2 ore One O€
a
c
> |> = fa) a a
Sele 2
oc jo 3 z z z
< |<| |19| 2
Ww iw o ro) O° °
«| |& Hanwgh Sfane Hvulpiceps
EA al al al |
g
z)- Cee ee,
7 7RGILES
See = = 2 DPRESB |
iy
z= © J
< a c OLDHAVEN ;
= eal BEDS lH.aff. |
Zz c BEDS : ailpicens
w ) A a, |
(ay || Ss 5) /
< - /, hi,
Bh) at o Li SABLES
3 WOOLWICH SUFFO ees Hicuniculus D’ H.cuniculus
ie We MO), ae LL
= =
=< o <
4 WI BEDS mS
= Whiphy-4b
WOOL WICH <
UBEDS se = [il pre ts el gpa es lle
(THANET (THANET CHALK
BEDS) BEDS) { )

Fig. 6 Stratigraphic columns showing occurrences of English Hyracotherium and correlation with
Belgium. The Werzelie/la and standard calcareous nannoplankton zones (NP) are after Costa &
Downie (1976) and Costa et al. (1978). Oblique hatching indicates absence of deposits. The range
given for Isselicrinus subbasaltiformis follows King (1970).

Beds localities on the common occurrence of H. cuniculus, and is here considered to be early
Sparnacian. Recently Costa et al. (1978) correlated the French Sables de Cuise (type Cuisian)
with the English London Clay, already correlated with the Belgian Argiles d’ Ypres (type Ypresian)
by Costa & Downie (1976), and the French Sparnacian strata with the English Woolwich/
Reading Beds and Blackheath/Oldhaven Beds, on the basis of a succession of dinoflagellate
zones. This would appear to contradict the evidence from the mammals. However, the Sables
de Sinceny and Pourcy (‘Upper Sparnacian’ of Costa et a/. 1978), which immediately underlie
the Sables de Cuise, have been recognized above an erosion plane in the middle of the type
Sparnacian in the region of Epernay (Feugueur 1963 : 288-291). The mammaliferous horizons
of Mutigny and Avenay appear to lie above this level (Michaux 1964; Louis 1970 : 58-60) and
could therefore both be equivalent to lower levels of the Sables de Cuise; the higher Sables a
Unios et Térédines of the Epernay district (classically considered sole Cuisian of the area) would
then be equivalent to only the upper part of the type Cuisian. In that case the whole of the
Sparnacian of Costa et al. (1978), being based on the Soissonnais area, would refer only to the
lower part of the type Sparnacian.

According to these correlations, the two species of the plesiadapid primate Platychoerops

ae J. J. HOOKER

(P. daubrei (Lemoine, 1880) from Pourcy, Mutigny and Avenay and P. richardsoni Charles-
worth, 1855, from Herne Bay), considered by Gingerich (1976 : text-fig. 19) to have an ancestor—
descendant relationship, are approximately contemporaneous.

Comparison with North American material

Few valid comparisons can be made with the North American species of Hyracotherium until
they are thoroughly revised in a detailed stratigraphic context. Kitts (1956) evidently reduced
the number of valid species too drastically without sufficiently good stratigraphic control (see
comments by McKenna 1960 : 117-119, Delson 1971 : 355-356, and Guthrie 1967 : 42-45).
A few general points can be made. The buccal flexure of the centrocrista and development of
the mesostyle on upper molars had hardly begun before the late Wasatchian (e.g. Wind River
and Lost Cabin faunas). Moreover, the enlargement of the P? protoconule, heralding the pre-
molar molarization of Orohippus, had not taken place until this time. English and continental
European Hyracotherium species appear to have been more precocious in their evolution, as
were the later more advanced European palaeotheres over their contemporaneous North Ameri-
can equid relatives.

As in Europe, most of the North American records of Hyracotherium are from the early
Eocene, but two have been attributed to the Palaeocene (Morris 1968 : 1, Jepsen & Woodburne
1969). The latter record has been discredited by Gingerich (1976 : 51-53) as being in an area of
downwarped younger (Wasatchian) early Eocene strata amongst generally older (Palaeocene)
strata. The former record is of a small species (H. seekinsi Morris, 1968) from the Tepetate
Formation of Baja California, Mexico, associated with barylambdid pantodonts and Esthonyx
sp. The species is based on a few isolated upper molars and appears scarcely more primitive
than typical early Wasatchian species, e.g. H. angustidens (Cope). Gingerich (1976 : 56-58)
correlated the Clarkforkian (the stage immediately preceding the Wasatchian in North America)
with the lower Sparnacian of Europe, based on the evolution of plesiadapid primates. There
appear to be no records of Hyracotherium specifically from the Clarkforkian and so, with
the possible exception of H. seekinsi, the early Sparnacian H. cuniculus is the oldest species
of the genus. One of its features which appears to be primitive amongst the European species is the
small size of the M; hypoconulid lobe. M;s of this morphology, however, appear to occur ran-
domly in Wasatchian populations (e.g. Savage et al. 1965 : text-fig. 5c). This may be another
reflection of the generally slower evolution of the North American species compared with the
European ones, leading to retention of primitive characters in some individuals. The slightly
earlier appearance of Hyracotherium in Europe than in the North American Rocky Mountain
region might point to an immigration from the east rather than from Central America as postu-
lated by Sloan (1970) and Gingerich (1976 : 86-88).

Acknowledgements

I would like to thank Dr D. E. Russell, Dr A. J. Rundle and the collectors for making specimens
available; Dr D. E. Russell (Muséum National d’Histoire Naturelle, Paris), Dr M. Godinot
(M.N.H.N.), Dr C. Forbes (Sedgwick Museum, Cambridge) and Mr R. A. D. Markham (Ipswich
Museum) for access to and loan of specimens in their care; Dr A. W. Gentry for critically reading
the text and Dr M. E. Collinson for helpful discussion and typing the manuscript.

References

Boswell, P. G. H. 1916. The stratigraphy and petrology of the Lower Eocene deposits of the north-eastern
part of the London Basin. Q. J/ geol. Soc. Lond. 71 : 536-591.

Charlesworth, E. 1855. On several new vertebrate fossils. Rep. Br. Ass. Advmt Sci., Liverpool, 24 : 80.

Cooper, C. F. 19324. On some mammalian remains from the Lower Eocene of the London Clay. Ann.
Mag. nat. Hist., London, (10) 9 : 458-467, pls 11-12.

ENGLISH EOCENE H YRACOTHERIUM 113

— 1932b. The Genus Hyracotherium. A revision and description of new specimens found in England.
Phil. Trans. R. Soc., London, B 221 : 431-448, pls 49-51.

Cooper, J. 1976. Sections recorded by the Tertiary Research Group at Bramford, 20.11.1971. Bull.
Ipswich geol. Grp 16 : 11-12.

1977. The palaeontology of the London Clay (Lower Eocene) of the Herne Bay coastal section, Kent,
England. Proc. Geol. Ass., London, 88 : 163-178, 1 fig.

Costa, L. I., Denison, C. & Downie, C. 1978. The Paleocene/Eocene boundary in the Anglo-Paris Basin.
J. geol. Soc. Lond. 135 : 261-264, 2 figs.

—— & Downie, C. 1976. The distribution of the dinoflagellate Werzeliella in the Palaeogene of north-
western Europe. Palaeontology, London, 19 : 591-614, 6 text-figs, pl. 92.

Davis, A. G. & Elliott, G. F. 1951. The London Clay of coastal Suffolk and Essex. Geol. Mag., London,
88 : 329-338.

Delson, E. 1971. Fossil mammals of the early Wasatchian Powder River local fauna, Eocene of northeast
Wyoming. Bull. Am. Mus. nat. Hist., New York, 146 : 305-364, 24 figs.

Depéret, C. 1901. Révision des formes européennes de la famille des Hyracotheridés. Bull. Soc. géol. Fr.,
Paris, (4) 1 : 199-225, pls 4-S.

Elliott, G. F. 1971. Eocene volcanics in south-east England. Nature phys. Sci., London, 230 : 9.

Feugueur, L. 1963. L’Yprésien du Bassin de Paris. Essai de monographie stratigraphique. Mém. Serv.
Carte géol. dét. Fr., Paris. 568 pp, 107 text-figs, 8 pls.

Fitch, F. J., Hooker, P. J., Miller, J. A. & Brereton, N. R. 1978. Glauconite dating of Palaeocene—Eocene
rocks from east Kent and the time-scale of Palaeogene volcanism in the North Atlantic region. J. geol.,
Soc. Lond. 135 : 499-512, 14 figs, 2 tabs.

Gamble, H. J. 1979. Observations on some London Clay coastal localities in and around Whitstable,
Kent. Tertiary Res., London, 2 (4) : 151-160, 2 text-figs.

George, W. & Vincent, S. 1976. Some river exposures of London Clay in Suffolk and Essex. Tertiary Res.,
London, 1 : 25-28.

Gervais, P. 1877. Enumération de quelques ossements d’animaux vertébrés recueillis aux environs de
Reims par M. Lemoine. J. Zool., Paris 6 : 74-79.

Gingerich, P. D. 1976. Cranial anatomy and evolution of early Tertiary Plesiadapidae (Mammalia, Pri-
mates). Pap. Paleont. Mus. Paleont. Univ. Mich., Ann Arbor, 15. 140 pp., 42 text-figs, 12 pls.

— 1977. New species of Eocene primates and the phylogeny of European Adapidae. Folia primatol.,
Basel, 28 : 60-80, 8 figs.

Guthrie, D. A. 1967. The mammalian fauna of the Lysite member, Wind River Formation (early Eocene)
of Wyoming. Mem. sth. Calif. Acad. Sci., Los Angeles, 5 : 1-53.

Hooker, J. J. 1979. Two new condylarths (Mammalia) from the early Eocene of southern England.
Bull. Br. Mus. nat. Hist., London, (Geol.) 32 (1) : 43-56, 15 figs, 3 tabs.

Jepsen, G. L. & Woodburne, M. O. 1969. Paleocene hyracothere from Polecat Bench Formation, Wyoming.
Science, N.Y. 164 : 543-547, 1 fig., 2 tabs.

King, C. 1970. The biostratigraphy of the London Clay in the London district. Tertiary Times, Bromley,
1: 13-15.

Kitts, D. B. 1956. American Hyracotherium (Perissodactyla, Equidae). Bull. Am. Mus. nat. Hist., New
York, 110 : 1-60, 7 pls, 10 figs.

Kiihne, W. G. 1969. A multituberculate from the Eocene of the London Basin. Proc. geol. Soc. Lond.
1658 : 199-202, 1 fig. :

Lemoine, V. 1880. Sur les ossements des terrains tertiaires inférieurs des environs de Reims. C. r. Ass. fr.
Ayanc. Sci., Paris, 8 : 585-594.

—— 1887. Sur le genre Plesiadapis, mammifére fossile de l’Eocéne inférieur des environs de Reims.
C. r. hebd. Séanc. Acad. Sci. Paris 1887 : 190-193.

Louis, P. 1970. Note préliminaire sur un gisement de mammiféres de |’Eocéne inférieur situé route de
Broyes a Sézanne (Marne). Ann/s Univ. ARERS, Reims, 8 : 48-62.

Lydekker, R. 1886. Catalogue of the fossil Mammalia in the British Museum (Natural History), 3. xvi+ 186
pp., 30 figs. British Museum (Natural History), London.

McKenna, M. C. 1960. Fossil Mammalia from the early Wasatchian Four Mile Fauna, Eocene of north-
west Colorado. Univ. Calif. Publs geol. Sci., Berkeley, 37 (1). 130 pp., 64 text-figs.

Marsh, O. C. 1876. Notice of new Tertiary mammals. Am. J. Sci., New Haven, 12 : 401-404.

Matthews, S. C. 1973. Notes on open nomenclature and on synonymy lists. Palaeontology, London, 16
(4) : 713-719.

Michaux, J. 1964. Age des Sables 4 Unios et Térédines (Eocéne inférieur) d’Avenay (Marne) et leurs
relations avec les Sables du méme nom des gisements classiques d’Epernay. C. r. somm. Séanc. Soc.
géol. Fr., Paris, 1964 (3) : 103.

114 J. J. HOOKER

Morris, W. J. 1968. A new early Tertiary perissodactyl, Hyracotherium seekinsi, from Baja California.
Contr. Sci. Los Angeles 151 : 1-11, 4 figs, 1 pl.

Owen, R. 1839. Description of the Fossil mentioned in the preceding letter. Mag. nat. Hist., London, Ns
3 : 446-448, text-fig. 58A—C.

— 1841. Description of the Fossil Remains of a mammal (Hyracotherium leporinum) and of a bird
(Lithornis vulturinus) from the London Clay. Trans. geol. Soc. Lond. (2) 6 (1) : 203-208, pl. 21.

1842. Description of son.e Molar Teeth from the Eocene Sand at Kyson in Suffolk, indicative of a
new species of Hyracotherium (Hyr. Cuniculus). Ann. Mag. nat. Hist., London, 8 : 1-2, text-figs 1-5.

— 1846. A History of British Fossil Mammals and Birds. 560 pp., 237 text-figs. London.

— 1858. Description of a small Lophiodont Mammal (Pliolophus vulpiceps Owen) from the London
Clay near Ha: wich. Q. J/ geol. Soc. Lond. 14 : 54-71, pls 2-4.

— 1865. On a new genus (Miolophus) of mammal from the London Clay. Geol. Mag., London, (1)
2 : 339-341, pl. 10.

Prestwich, J. 1850. On the structure of the strata between the London Clay and the Chalk in the London
and Hampshire Tertiary Systems. Q. J/ geol. Soc. Lond. 6 : 252-281.

Quinet, G. E. 1969. Apport de l’étude de la faune mammalienne de Dormaal a la stratigraphie générale
du Paléocéne supérieur européen et a la théorie synthétique de la molaire mammalienne. Mém. Inst.
r. Sci. nat. Belg., Brussels, 162. 188 pp., 7 pls.

& Verlinden, W. 1970. Sur l’Hyracotherium dErquelinnes (Jeumont). Bull. Inst. r. Sci. nat. Belg.,
Brussels, 46 (34) : 1-10, 3 pls.

Richardson, W. 1841. Observations on the locality of the Hyracotherium. Trans. geol. Soc. Lond. (2) 6
(1) : 211-214.

Russell, D. E., Louis, P. & Savage, D. E. 1967. Primates of the French early Eocene. Univ. Calif. Publs
geol. Sci., Berkeley, 73. 46 pp., 14 text-figs.

Rutot, A. 1881. Sur la position stratigraphique des restes de mammifeéres recueillis dans les couches de
l’Eocéne inférieur de Belgique. Bull. Acad. r. Belg., Brussels, (3) 1 : 506-547, 3 pls.

Savage, D. E., Russell, D. E. & Louis, P. 1965. European Eocene Equidae (Perissodactyla). Univ. Calif.
Publs geol. Sci., Berkeley, 56. 94 pp., 1 pl., 42 text-figs.

Simpson, G. G. 1952. Notes on British Hyracotheres. J. Linn. Soc., London, (Zool.) 42 : 195-206, 4 figs.

Sloan, R. E. 1970. Cretaceous and Paleocene terrestrial communities of Western North America. Jn
van Valen, L. (ed.), Evolution of Communities : 427-453, 8 figs. Lawrence, Kansas (Proc. N. Am.
paleont. Cony. E).

Teilhard de Chardin, P. 1922. Les mammiféres de l’Eocéne inférieur francais et leurs gisements [cont’d.]
Annls Paléont., Paris, 11 : 1-108, text-figs 3-42, pls 1-8.

1927. Les mammifeéres de Il’Eocéne inférieur de la Belgique. Mém. Mus. r. Hist. nat. Belg., Brussels,
36. 33 pp., 6 pls, 29 text-figs.

White, E. I. 1931. The vertebrate faunas of the English Eocene. 1. 123 pp., 1 pl., 162 text-figs, 4 tabs.
London, British Museum (Natural History).

Wood, S. V. 1839. Letter from S. V. Wood Esq., late curator to the Geological Society, announcing the
discovery of Fossil Quadrumanous remains, near Woodbridge, Suffolk. Mag. nat. Hist., London, Ns
3 : 444445, text-fig. 57.

Wortman, J. L. 1896. Species of Hyracotherium and allied perissodactyls from the Wahsatch and Wind
River beds of North America. Bull. Am. Mus. nat. Hist., New York, 8 : 81-110.

Wrigley, A. 1924. Faunal divisions of the London Clay illustrated by some exposures near London. Proc.
Geol. Ass., London, 35 : 245-259.

Salenia trisuranalis sp. nov. (Echinoidea) from the
Eocene (London Clay) of Essex, and notes on its

phylogeny
D. N. Lewis and R. P. S. Jefferies

Department of Palaeontology, British Museu:n (Natural History), Cro nwell Road, London
SW7 5BD

Synopsis
The new species Salenia trisuranalis, with three suranal plates, is described from the English Eocene, and
its phylogenetic position with respect to other echinoids discussed in the light of Hennigian methodology.

The presence of suranal plates tessellated into the apical disc is seen as likely to be a synapomorphy with
irregular echinoids, and suggests they are probably the sister group to the Salenioida.

Introduction

Salenioids, after their great expansion in the Cretaceous, are uncommon subsequently, and in
the Eocene are very rare indeed. Salenia trisuranalis sp. nov. 1s the first record of a Salenia
from the British Cainozoic. Two continental species from the Eocene — S. pellati from La Gourépe
near Biarritz, and S. garciae from Callosa, Alicante, have certain similarities, but neither has
more than one suranal plate.

Salenia trisuranalis is now described, and an attempt is made to assign it to its phylogenetic
position with respect to other echinoids.

Systematic description

Superorder ECHINACEA Claus, 1876
Order SALENIOIDA Delage & Hérouard, 1903
Family SALENIIDAE L. Agassiz, 1838
Subfamily SALENIINAE L. Agassiz, 1838

Genus SALENIA Gray, 1835
Salenia trisuranalis sp. nov.
Figs 1-4
DIAGNOsIs. The species is characterized by having three suranal plates.

MATERIAL (Figs 1, 2). The unique holotype is well preserved, slightly crushed and flattened in
the ambital plane. One spine is preserved, an interambulacral, on the oral surface. The specimen
is in the Palaeontology Department of the British Museum (Natural History), number E76505.
It is from the London Clay (Lower Eocene) of Walton-on-the-Naze, Essex, England and was
collected by Mr William George.

SHAPE. The test is small, circular in outline at the ambitus, with a maximum diameter of 8:25 mm,
minimum 7:5 mm; the exact measurement is uncertain because of the damaged state of the test.
The oral surface is flat; the apical surface is irregular owing to plate displacement.

APICAL SYSTEM. This is dicyclic with three suranal plates. (Notation follows Lovén 1874).

Ocular Plates. The inner sutures are rounded, and the outer suture is a single arc. The ornament
consists of fine ridges arranged approximately perpendicular to each inner suture. The centre of

Bull. Br. Mus. nat. Hist. (Geol.) 33 (2) : 115-121 Issued 31 January 1980
115

116

Figs 1-4 Salenia trisuranalis sp. nov. Holotype. Fig. 1, adapical view, 5. Fig. 2, adoral view, 5.
Fig. 3, close-up of periproct to show the three suranal plates, x 15. Ornament of fine ridges and
granules, found over the whole of the apical system, can be seen on the suranal plates. Fig. 4,
close-up of ambulacra II and III and interambulacrum 2, x 13. A small spine is visible, attached to

interambulacrum 1, innermost margin (arrowed).

SALENIA TRISURANALIS 7

each ocular plate has granular ornament, and the outer edge has a knobbly ridge the whole of its
length. Each ocular plate has a slightly raised, broad v-shaped notch for the ocular tentacle,
in the centre of the ridge.

Genital Plates. Genital plate 1 is polygonal, with a shallow notch which is ornamented by fine
ridges and which forms part of the slightly raised rim around the periproct. The plate ornament
consists of fine ridges approximately perpendicular to the margins, a granular central ornament
and a knobbly ridge, concave to the outer margin and in continuation with the ridges on ocular
plates I and II. External to this ridge more granules ornament the surface.

Genital plate 2 is octagonal, with ornament similar to that of genital plate 1, comprising fine
ridges perpendicular to the inner five margins; the central part of the plate is divided into an
inner portion with an irregular surface, and radiating from this an outer portion with granular
ornament. There is no continuation of the knobbly ridge from adjacent oculars. However, the
granular ornament ceases where the ridges of the oculars meet the plate margin. Connecting the
outer two sutures is a convex strip of unornamented plate. There is no external evidence of the
madreporic pores.

Genital plate 3 is heptagonal, with a similar ornament of fine ridges around the inner five
margins. The centre has granular ornament which radiates towards the outer two sutures and,
as in genital plate 2, has an unornamented arcuate strip.

Genital plate 4 is displaced and therefore the boundary is not wholly visible. The plate is
polygonal, with ornament like that of genital plates 2 and 3.

Genital plate 5 is polygonal but the outer margin is not visible. A deep notch forms part of the
rim around the periproct. Ornament is of fine ridges along the margins as on the other genital
plates, but the centre has granular ornament and a knobbly ridge meeting the adjacent oculars,
as on genital plate 1.

Suranal Plates (Fig. 3; for discussion see Duncan & Sladen 1887). There are three suranal plates,
two along the axis connecting ambulacrum III with interambulacrum 5, and one to the right of
this axis, towards ocular plate II. One suranal plate, towards genital plate 5, is larger than the
other two and has five straight sides with ornament of fine ridges, and one curved side that
makes up part of the periproct rim. The centre of this suranal plate has fine granular ornament,
with fine ridges converging into this ornament. The second suranal plate has four straight sides
and one smoothly curved side which, again, forms part of the rim of the periproct. The ornament
is similar to that of the larger suranal plate. The third suranal plate is partly obscured by genital
plate 4 which has overridden it in crushing; however it is pentagonal, with ornament like that of
the other two suranal plates.

Periproct. This is bounded by genital plates | and 5, and by two suranal plates. The outline is
smooth and kidney-shaped, and lies right of the anteroposterior axis (ambulacrum III — inter-
ambulacrum 5) and between ocular plate I and two of the suranal plates. The ornament consists
of short fine ridges perpendicular to the edges.

Notes on the ornament. The whole of the apical system is ornamented by granules and fine ridges,
and in addition by larger knobbly ridges. The fine ridges are continuous from one plate to those
adjacent. The genital plates are not perforated by genital pores, suggesting that the individual
may be immature.

AMBULACRA (Fig. 4). These are straight and wide. At the peristome they are the same width as
the interambulacrals. At the ambitus they are half as wide as the interambulacrals, while at the
apical system they are slightly less than half as wide as the interambulacrals.

The plating is difficult to determine except where displaced along the plate boundaries by
crushing, but it appears to be bigeminate, with pore pairs oblique.

The pore pairs are surrounded by a smooth deep wall, higher on the aboral side and incomplete
on the adoral side. The pores are separated by a single granule situated only slightly towards the
adradial suture.

There are 18 primary ambulacral tubercles in each ambulacrum. Each tubercle is large, smooth
and in contact with the perradial suture and also close to the pore pairs, with each tubercle

118 D. N. LEWIS & R. P. S. JEFFERIES

covering at least half the width of the plate. The smallest primary tubercles are the first two or
three pairs nearest to the peristome, and the largest are developed at the ambitus. Small inner
tubercles are also present.

INTERAMBULACRA (Fig. 4). These widen rapidly from the peristome. The interambulacrals are
twice as wide at the ambitus as they are at the peristome, and have ten primary tubercles, with one
on each plate. The tubercles are situated close to the adradial sutures. In the interradial position
there are numerous smaller tubercles, two per larger primary tubercle. Very occasional granules
are also present. Plate boundaries are difficult to determine except where dislocated.

The primary tubercles are non-perforate, and have 8-11 crenulations according to size. All
smaller tubercles are non-crenulate.

The bosses of the primary tubercles are convex, rising from a rounded basal terrace. The
mamelons have slightly undercut necks. Smaller tubercles have convex bosses arising from
shallow, rounded basal terraces, and have very short, slightly undercut necks.

PERISTOME. This is large, round, about 4 mm in diameter, and Is slightly distorted (post mortem).
Remains of the raised rim of the gill slits can be seen on ambulacra III and IV and on inter-
ambulacra 1, 2, 4 and 5, the others having been broken away.

SPINE (Fig. 4). One spine is present on interambulacrum 1, near the peristome. It is about 0-9 mm
long and is not in life position. It is straight, flattened and ornamented by rows of thorns. It
tapers proximally into a short neck and ends in a slightly expanded base.

COMPARISON with other species. The specimen, which was first mentioned by George & Vincent
(1977), was initially identified as Sa/enia cf. pellati on the basis of Cotteau’s figures (1860, 1892).
However, closer examination reveals some differences, the chief of which is the presence of three
suranal plates in S. trisuranalis and only one in S. pellati. The ornament on the ocular plates of
S. trisuranalis is like that on the genital plates, but on S. pe//ati it is different — the ribbing is
coarser on the ocular plates, and these are perforate. The apical disc covers about the same area
in both species. The peristome appears larger in S. pellati where it is about the same diameter as
the apical disc, whereas in S. trisuranalis it is smaller than the apical disc.

Ambulacra are similar to those of S. pel/ati in width, size and numbers of primary tubercles.
In S. pellati pore pairs open *. .. at the base of a small granular swelling .. .’, while in S. trisur-
analis they are surrounded by a deep wall. Interambulacra are different from those of S. pellati in
that S. triswranalis has more primary tubercles, although these are similar to that species in
structure and numbers of crenulations. S. garciae Cotteau (1890) differs slightly in detail, al-
though the type specimen was incomplete, so that the apical disc was not described. The inter-
ambulacra of S. garciae have many more granules around the primary tubercles and the ambu-
lacra are longer, narrower and have more tubercles. The pore pairs of S. garciae appear to have
little ornament around them.

Discussion

The Salenioids are divided in the Treatise of Invertebrate Paleontology (Fell & Pawson 1966)
into two families — the Acrosaleniidae and the Saleniidae. The Saleniidae are in turn divided
into the Saleniinae and the Hyposaleniinae. In the light of Hennigian methodology (Hennig
1966, 1969) it is interesting to consider the phylogenetic relationships of these and other groups.

All known Salenioids are distinguished from other echinoids, except primitive irregular forms,
by having one or more suranal plates tessellated into a large apical disc. Many echinoids have an
obvious suranal plate in the imago, but usually this later becomes lost amongst the other peri-
proctal plates (see, for example, Gordon 1926, on Psammechinus miliaris). The presence of suranal
plates tessellated into the apical disc is an advanced character with respect to other echinoids and
could be a synapomorphy of Salenioids and irregular echinoids. It suggests that Salenioids may
be the sister group of the irregular echinoids (see Jesionek-Szymanska 1968). Its retention in the
Salenioids is a paedomorphic character, but its tessellation into the apical disc is not paedomor-
phic.

119

SALENIA TRISURANALIS

“sprloluges oy} Jo sdiysuoneyer
a[qissod MoUs 0} WeIseIq § ‘SI

pesoepnquiesazul sad

sa}ejd yo suuiNjo> OM}

‘sajauaqgn}y ayejnuerd

so1jsiaa}oeseys

prou1yseanea ds1ydiowodeyne 1494}0
yO ul6i10 pue sazejd jesoejnquiesazul

punodwod ‘uoljipuod ayery2uesqo}oa

Snipessajul 10149}S0d
0} Buipuay yo01d1ied

1499} pelae

\o
2sip jeoside oyu ep le
sayejd jeueins jo uolyeyjasseay
geo
uoljyisod o\3*
4olsaysod ul paxty yooudued
‘ayejd jeueins auo Ajuo
uoljisod je1}xep—o41a}sod tayesojysaduit sejoseqny
ul paxly y204diied oo
4oro™
ov?
PN)
pedojanapas sayeid yo yc a
49° \o¥8
jeuesns ajdiyinw 3? ‘ qo
wo! eh
\
ee)
93
SD
3
ave?
a\°

120 D. N. LEWIS & R. P. S. JEFFERIES

The Acrosaleniidae probably include the most primitive members of the Salenioids. The reasons
for this belief are three. They are geologically early compared with other Salenioids, they have
perforate, crenulate primary tubercles, like those of other primitive Euechinoids and Miocidaris
and they have a tendency, not expressed in all forms, for the periproct to be posterior in position
(toward genital plate 5). These characters are likely to be primitive, by an outgroup comparison
with irregular echinoids. The difficulty with the comparison is that the position of the periproct
is not fixed in genera of the Acrosaleniidae.

However, the tessellation of more than one suranal plate into the apical disc in the Acrosalenii-
dae is likely to be an advanced condition with respect to many non-Salenioid echinoids, on the
grounds of geological age and correlation with the three characters listed above. Salenioids
outside the Acrosaleniidae have only one suranal plate tessellated into the apical disc, with the
single exception of S. trisuranalis.

The Hyposaleniinae are more advanced than the Acrosaleniidae in three respects. The primary
tubercles are imperforate, the periproct is fixed in a posterior position, towards genital plate 5,
and only one suranal plate is tessellated into the apical disc. The last two characters are likely to
be advanced with respect to the acrosaleniid condition on the grounds of geological age. Within
the Hyposaleniinae the genus Poropeltaris is strikingly characterized by the advanced feature of
smooth primary interambulacral tubercles.

The Saleniinae are more advanced than the Hyposaleniinae in having the periproct fixed in a
posterodextral position, i.e. near to ocular plate I.

The likely phylogenetic relationships within the Saleniids and some related groups are as
shown in Fig. 5. If this figure is correct then the Acrosaleniidae and Hyposaleniinae are groups
different in nature from the Saleniinae. In Hennigian terms they are paraphyletic, in that they
exclude forms descended from members of the group. Indeed, the Acrosaleniidae and the Hypo-
saleniinae would be parts of the Hennigian stem-group of the Saleniinae — they would be inter-
mediate categories, Zwischenkategorien in the terminology of Hennig (1969).

As regards the multiple suranals of S. trisuranalis, two possibilities need to be considered. The
first is that the condition is homologous with the similar state of the Acrosaleniidae. This would
imply that development of imperforate tubercles and the fixation of the periproct in a postero-
dextral position had happened twice. It would also imply that Salenioids with more than one
suranal plate should exist beyond the Upper Cretaceous. Despite an extensive geological record
up to the Upper Cretaceous, Salenioids with more than one suranal plate have not been observed
in post-Cretaceous rocks, with the exception of S. trisuranalis.

The second possibility is that the multiple suranal plates of S. trisuranalis are a convergence
with Acrosalenia, aping the primitive condition. This seems more likely than the first possibility.
Indeed, the condition in S. trisuranalis may even prove to be an individual variation. Nevertheless,
the rarity of the species has prompted us to describe it, in the hope that this note will encourage
the search for further specimens.

Acknowledgements

Grateful thanks are expressed to Mr William George for the presentation to the Museum of this
stimulating specimen, and to Mr T. W. Parmenter for taking the photographs.

References

Cotteau, G. H. 1860. Echinides nouveaux ou peu connus. Revue Mag. Zool., Paris, (2) 12 : 212-224, pls
123%

1889-94. Terrains Tertiaires, Eocéne, Echinides, 2. 789 pp., pls 201-384. Paléont. fr., Paris (1,

Animaux Invertébrés). ;

1890-91. Echinides Eocénes de la Province d’Alicante. Mém. Soc. géol. Fr., Paris, (3) 5 : 1-107,

pls 22-37.

SALENIA TRISURANALIS 121

Duncan, P. M. & Sladen, W. P. 1887. On some points in the morphology and classification of the Saleniidae,
Agassiz. Ann. Mag. nat. Hist., London, (5) 19 : 117-136.

Fell, H. B. & Pawson, D. L. 1966. Echinodermata. Jn Moore, R. C. (ed.), Treatise on Invertebrate Paleon-
tology, U : U211—U640. Kansas.

George, W. & Vincent, S. 1977. Report of field meeting to Walton-on-the-Naze and Wrabness, Essex,
2.10.1976 with notes on the London Clay of Walton. Tertiary Res., West Wickham, 1 (3) : 83-90.

Gordon, I. 1926. The development of the calcareous test of Echinus miliaris. Phil. Trans. R. Soc., London,
B 214 : 259-312.

Hennig, W. 1966. Phylogenetic Systematics. 263 pp. Urbana, IIl.

— 1969. Die Stammesgeschichte der Insekten. 436 pp. Frankfurt a.M.

Jesionek-Szymanska, W. 1968. Irregular echinoids — an insufficiently known group. Lethaia, Oslo, 1 (1) :
50-62.

Lovén, S. 1874. Etudes sur les Echinoidées. K. svenska Vetensk Akad. Handl., Uppsala & Stockholm,
11 : 1-191, 53 pls.

a % —

z A
= = ‘
. ty So ==
me
7 =~
7 : i
. = f
4
;

f e —
-
be:
-_ 4
;

: f 1
: ;
< 3

Tertiary and Cretaceous brachiopods from Seymour,
Cockburn and James Ross Islands, Antarctica

E. F. Owen

Department of Palaeontology, British Museum (Natural History), Cromwell Road, London
SW7 S5BD

Synopsis

The systematics of the brachiopod fauna described by Buckman (1910) from the Lower Tertiary of Sey-
mour and Cockburn Islands is revised. New species Notosaria seymourensis, Terebratulina buckmani,
Terebratella crofti and Liothyrella anderssoni are described, as is the dallinoid ‘Laqueus’ cockburnensis,
broadly assigned to that genus. Present views on the stratigraphical position and latest age estimates for
the Lower Tertiary beds of Seymour Island are briefly discussed.

Descriptions of new genera and species Protegulorhynchia meridionalis (Hemithyrididae) and Rossithyris
humpensis (Laqueidae), from the Upper Cretaceous, Campanian of James Ross Island, are given for the
first time.

Introduction

Andersson (1906) recognized two distinct lithologies within the Tertiary sequence on Seymour
Island. One of these he described as a calcareous sandstone with tuff fragments, located chiefly
in the Cross Valley area, while the remaining deposits were said to be friable sandstones with
conglomerate intercalations. Plant remains were abundant in the calcareous sandstones and
numerous invertebrate fossils, mainly brachiopods and molluscs, occurred in the more friable
beds. The term Seymour Island Series was applied to the entire sequence.

Adie (1958 : 9) pointed out the existence of a major unconformity between the Tertiary and
Cretaceous deposits which occur south-west from Cross Valley and also crop out at the most
northerly point, Cape Wiman. Since this work, very few stratigraphical observations have been
made on the island, but in 1974-75 an expedition, conducted by members of the Institute of
Polar Studies at Ohio University, led to a publication by Elliot & Trautman (1979) who revised
the Tertiary deposits and proposed the following formations.

a. Cross Valley Formation, for the sequence of non-marine sandstone and plant-bearing fine sand-
stone which crops out in Cross Valley and is regarded as of Palaeocene age.

b. La Meseta Formation, for the unconsolidated marine sands exposed on the flanks and to the
north of the mesa situated on the north-eastern part of the island. The suggested age of the beds
was late Eocene to early Oligocene, a view expressed also by Zinsmeister (1977) on the basis of
faunal and floral evidence.

In the same paper, Zinsmeister (1977) points out that Wilkens’ (1911) correlation of the
‘Seymour Island Series’ with the Patagonian Formation, though correct in chronological position,
was wrongly dated as Oligocene—Miocene and should be regarded as older. He quoted the work
of Comacho (1974; Comacho & Fernandez 1955), recording several species of Venericardia
planicosta group bivalves from the lower part of the Patagonian Formation in central Patagonia.
The presence of this cosmopolitan bivalve suggested to Comacho that the lower two-thirds of the
formation was of Eocene age and that only the uppermost marine unit was of Oligocene age.
Thus, according to Zinsmeister, the beds of the Seymour Island Series can be assigned an Eocene
to Early Oligocene age with perhaps some of the Lower and Middle Eocene missing.

Other opinions vary with those of Cranwell (1959), who regarded the same deposits as Palaeo-
cene, on the evidence of plant microfossils, and Simpson (1971), who assigned a late Eocene to

Bull. Br. Mus. nat. Hist. (Geol.) 33 (2) : 123-145 Issued 31 January 1980
123

124 E. F. OWEN

\ Ele

8 So oN
Si VEGA ISLAND >

P
= So 6
z SC LY
: 3

cyHUMPS 1. 64°S

: ae

COCKBURN
uaa
JAMES ROSS é 5
ISLAND f SG
Sf
f
oN SEYMOUR
oe ISLAND
iY “f
e Be L/N a J
ct as ai : pee
Oy re a J
D GO ae a_a /
ar ye y) pa Hi
if
7 af | SNOW RIEL 7
ISLAND A 64 30S
‘ ee:
ve 4
km
(0) 10 20 30 40 50

° a
58 W. S7W

Fig. 1 Sketch map showing the relative positions of Seymour, Cockburn and
James Ross Islands.

possibly early Oligocene age from the evidence of fossil penguin bones. Zinsmeister points out
that Cranwell’s data apply to the lower 110 m of the Seymour Island Series and that the upper
420 m contains a molluscan fauna which suggests an early Oligocene age.

Buckman (1910) interpreted the age of the beds on Seymour Island as Pliocene—Miocene on
the evidence of a collection made by the Swedish explorer Dr J. Gunnar Andersson (1901-03)
from one locality which Andersson called Loc. 11, and compared the brachiopods to Patagonian
species. Those specimens from two further localities, Locs 12 and 13 on Cockburn Island, were
ascribed a Miocene—Oligocene and a Pleistocene age, respectively, and compared to Australasian

and South American species.

The brachiopods described here are from two collections, one made in 1947 by the late W. N.
Croft from Seymour and Cockburn Islands, with additional Cretaceous material from James

ANTARCTIC BRACHIOPODS 125

50
516

*

N

i@)
D
oO
(ep)
)
D492 2 MESETA
> a Dv
c
c
m
wa
D
D

PENGUIN BAY

(0) km 2
| 1 i

Fig. 2 Outline map of the northern sector of Seymour Island, showing the stations from which

brachiopods have been collected. Prefix D indicates W. N. Croft’s localities. Plain numerals are
IPS localities.

Ross Island, and the other from a more recent visit to Seymour Island by D. H. Elliot and
T. A. Trautman in 1974-75.
Unless otherwise stated, all material cited is preserved in the fossil brachiopod collections of

the Department of Palaeontology, British Museum (Natural History); registered numbers are
prefixed BB.

IPS refers to the Institute of Polar Studies, Ohio State University, U.S.A.

Using simple age and lithological terms, Croft divided the formations on Cockburn Island into
the following groups.

1. Intrusive rocks of uncertain age

2. Pecten Conglomerate of Pliocene—Pleistocene age
3. Ross Island Formation

4. Tertiary beds with a brachiopod fauna

5. Cretaceous beds with ammonites

Croft stated in his original unpublished report (1947 : 3) that Andersson’s locs 12 and 13
were visited and their positions corrected to points further north than those given on Ander-
son’s sketch map, and on the west side of the island. It is clear from his notes that Croft was
somewhat disorientated, since Andersson’s sketch (1906 : 6) shows the localities situated on the
south-eastern part of the island. Furthermore, Andersson (1906 : 41, fig. 3) gives a section
through the beds on the ‘East side of Cockburn Island at locality 12’.

Both the brachiopod fauna from Seymour Island and the fauna from the Tertiary beds of
Cockburn Island are almost identical in constitution to those described by Buckman (1910),

COCKBURN
ISLAND

0 km
_

Fig. 3. Sketch map of Cockburn Island showing W. N. Croft localities.

which were based on material collected by Drs O. Nodenskjéld and J. Gunnar Andersson during
the Swedish Antarctic Expedition of 1902-03. The present paper is, therefore, very largely a
revision of Buckman’s work, bringing the systematic and taxonomic aspects up to date.

Systematic descriptions
Superfamily LINGULACEA Menke, 1828
Family LINGULIDAE Menke, 1828
Genus LINGULA Bruguiére, 1797

Lingula antarctica S. S. Buckman
Fig. 10

1910 Lingula antarctica S. S. Buckman : 9; pl. 1, fig. 7.

Buckman described this species as an elongate, parallel-sided shell with rather flat valves, having
growth-lines forming an oblong pattern throughout its life history. He figured a rather poorly
preserved and incomplete specimen which he felt resembled Glottidia palmeri Davidson, but
without any of the characteristic median ridges within the dorsal valve.

The specimen figured here, Fig. 10, is just 30 mm in length, about 5mm longer than Buck-
man’s figured specimen. It was collected by members of the expedition of 1974-75 mounted by
the IPS (see p. 123) and comes from their loc. 16, approximately | km north of the eastern
headland of Penguin Bay, Seymour Island. It is preserved in a highly glauconitic sandstone of
Lower Tertiary age.

Buckman’s specimen, which was collected from a point further eastward and nearer the coast,
is from Loc. 11 of Nordenskjéld & Andersson, 1902-03, and mentioned by Buckman (1910 : 3)
as having abundant Bouchardia specimens and “. . . crowded with examples of a small Cerithium’,
recently reidentified by Mr R. J. Cleevely, British Museum (Natural History), as Turritella sp.

As Buckman pointed out (1910 : 34), the presence of a species of Lingula in beds of Tertiary
age within the Antarctic is of interest since it extends the records of that genus to localities
further south than Patagonia and the Australasian continent. With the additional material
collected by the late W. N. Croft in 1947 and the single specimen figured here, from the collections

ANTARCTIC BRACHIOPODS 127

of the IPS, not only is its presence confirmed, but the new record lends weight to the argument
for considerably increased climatic temperatures during Lower Tertiary times in this area, as living
species of Lingula today prefer warmer climates and are seldom, if ever, found in other than
tropical seas.

The presence of a high percentage of glauconite within the Tertiary beds of Seymour Island
also suggests a somewhat shallow, high energy sea area, perhaps shelf or inshore conditions.

Loca ity. IPS loc. 16 and loc. D.515 of W. N. Croft, said to be the fossil penguin locality and
equivalent to loc. 11 of Andersson (1906).

Horizon. Lower Tertiary.

MATERIAL. The figured specimen (IPS) and ten fragments, BB.76752-9.

Superfamily RHYNCHONELLACEA Gray, 1848
Family HEMITHYRIDIDAE Rzhonsnitskaya, 1956
Genus HEMITHIRIS dOrbigny, 1847

Hemithiris antarctica S. S. Buckman
Figs 5a—c

1910 MHemithyris antarctica S. S. Buckman : 13; pl. 1, figs 8, 9.

The specimen figured here is one of three almost complete examples collected from the Glauco-
nitic Bank on Cockburn Island. It departs from the original description and figured example of
Buckman (1910 : pl. 1, figs 8, 9) in its broader trigonal general outline and in having very faint
radial striae. The difference in outline can be explained in terms of variability within the species
which is manifest in the few specimens available here, but Buckman referred to the absence of
radial striae from the surface of H. antarctica and used this as a point of distinction between his
species and H. psittacea (Gmelin) and H. woodwardi (Adams). Close examination under a low-
power microscope of the specimens available shows them to have very faint radial striae on the
shell surface of both valves. The produced beak, broad sulcus and large foramen appear to be
characters common to all species of Hemithiris.

Loca.ity. W. N. Croft’s locs D.491.4 and D.526, north-eastern section of Cockburn Island.
Horizon. Lower Tertiary. Glauconitic Bank and Pecten Conglomerate.

MATERIAL. The figured specimen, two almost complete specimens and a small fragment from the
Glauconitic Bank and two well-preserved specimens and a single pedicle valve from the Pecten
Conglomerate, BB.76592, BB.76749-51, BB.76601-3.

Genus NOTOSARIA Cooper, 1959

Notosaria seymourensis sp. nov.
Fig. 4

DESCRIPTION. A single crushed rhynchonellide 18 mm long, 12 mm wide, broadly oval in general
outline with surface ornament of 25 coarse, rounded, radiating costae interrupted by marked
concentric growth-lines becoming more lamellose towards the shell margins. A short, suberect
beak is truncated by a large, subcircular foramen guarded by slightly extended deltidial plates.
A striated apical plate is just visible posteriorly. [The internal characters of this specimen were
impossible to define owing to the crushing of the shell.]

Name. From Seymour Island.

Locatity. W. N. Croft’s loc. D.492.1, south-west of Cross Valley, Seymour Island. Croft’s
notes (1947 : 8) record this locality as yielding traces of plant remains in hard bands and a
rhynchonellide which was collected from the top of knoll in the sandy clay beds.

128 E. F. OWEN
HOLotyPe. The specimen BB.76590 figured here, Fig. 4.
HORIZON. Lower Tertiary.

REMARKS. In his original description of the genus Notosaria, Cooper (1959 : 48) distinguished
it from Tegulorhynchia Chapman & Crespin by its different costellae and fine growth-lines. The
present specimen has comparatively coarse costae and rather marked concentric growth-lines.
The nature of the growth-lines, though somewhat more marked than in the type species N.
nigricans (G. B. Sowerby), appear to conform to the pattern illustrated by Cooper (1959 : pl. 6,
figs 8, 9) and show the characteristic lamellose thickening at the margins.

The presence of a specimen of Notosaria within the Tertiary deposits of Antarctica extends the
record southwards. The only other species so far recorded from the Tertiary of the southern
hemisphere are NV. nigricans (G. B. Sowerby) and N. sublaevis (Thomson) from the Miocene of
New Zealand.

Genus PLICIRHYNCHIA Allan, 1947
Plicirhynchia sp.
1910 Hemithyris plicigera Ihering; Buckman : 12; pl. 1, fig. 10.

A single damaged pedicle valve and two crushed, but complete, specimens show generic charac-
ters of Plicirhynchia. The single valve is subtriangular in general outline, with a produced umbo
and sharp beak. The shell surface of all three specimens is posteriorly striate with anterior costae
developing marginal plicae. A marked growth-line occurs at about half the length of the shell and
another nearer the margins, giving a somewhat lamellose appearance to the ornament.

The specimen figured by Buckman is not instantly recognizable as Rhynchonella plicigera
Ihering to which he assigned it, but, in view of the poor original figure of the species by Ihering
(1897 : 270, fig. 7) and the fact that Buckman’s specimen was equally poor, it is not easy to draw
any conclusion from the comparison. Cooper (1959 : 52) had similar doubts as to their direct
specific relationship. When more material is available, it may be possible to compare the two
forms more effectively.

Loca.ity. W. N. Croft’s loc. 491.4, Cockburn Island.
Horizon. Lower Tertiary. Glauconitic Bank.

MATERIAL. One pedicle valve and two crushed specimens referable to this genus, BB.76746,
BB.76747, BB.76748.

Genus TEGULORHYNCHIA Chapman & Crespin, 1923

Tegulorhynchia imbricata (S. S. Buckman)
Figs |la-c

1910 Hemithyris imbricata S. 8S. Buckman : 11; pl. 1, fig. 12.

EMENDED DESCRIPTION. Tegulorhynchia 12-9 mm long, 17 mm wide and 6:9 mm thick. Trans-
versely oval in general outline. Umbo short, beak sharp; foramen small, circular. Beak-ridges
distinct, interarea short. Brachial valve with a low incipient median fold. Pedicle valve with a
broad shallow anterior sulcus. Anterior commissure arcuate; linguiform extension long. The
shell surface consists of fine, rounded, radiating costellae interrupted by numerous concentric
growth-lines, giving a lamellose or imbricate appearance. [Internal characters unknown owing to
lack of suitable material.]

Loca.ity. W. N. Croft’s loc. D.491.4, Cockburn Island.
Horizon. Lower Tertiary. Glauconitic Bank.

MATERIAL. The specimen BB.76593 figured here, Fig. lla-c, and one single pedicle valve,
BB.76745, from the same horizon and locality.

ANTARCTIC BRACHIOPODS 129

REMARKS. The genus Tegulorhynchia is the subject of a forthcoming revision of the Tertiary
Rhynchonellidae by Daphne Lee of Otago University, New Zealand. In her work she includes
the species T. squamosa (Hutton) from the Oligocene of Broken River, Castle Hill Basin, New
Zealand, and 7. imbricata (Buckman). She concludes that the two species are probably synony-
mous. Topotype specimens of Tegulorhynchia squamosa (Hutton), kindly donated by Dr Lee,
have been compared with specimens collected from Cockburn Island and identified as T. imbricata.
Although the shell ornament is similar, in all other respects the two species have little in common.
The general outline of 7. squamosa is transversely triangular to subpentagonal, whereas that of
T. imbricata remains transversely oval. Furthermore, 7. squamosa appears to be more robust
than T. imbricata, with a marked posterior inflation of the brachial valve. The umbo is slightly
more produced in 7. squamosa and consequently gives rise to a more extensive interarea bounded
by well-marked beak-ridges.

Buckman’s citation of a rhynchonellide specimen under the name of Hemithyris squamosa
Hutton (Buckman 1910 : 10) and the accompanying figures (pl. 1, figs 13a, b) does not justify
its assignment to Tegulorhynchia squamosa (Hutton) as originally described (Hutton 1873 : 37).
It is probable that this specimen is merely a rather poorly preserved example of 7. imbricata,
but Buckman’s figure does not show the typical ornament.

Genus PROTEGULORHYNCHIA nov.
TYPE SPECIES. Protegulorhynchia meridionalis sp. nov.

DIAGNOSIS. Subcircular to broadly oval costellate rhynchonellide. Fold low; anterior com-
missure arcuate. Linguiform extension long. Umbo small, slightly produced. Beak sharp, sub-
erect. Beak-ridges distinct. Foramen small.

Name. Allied to Tegulorhynchia.

Protegulorhynchia meridionalis gen. et sp. nov.
Figs l5a-c, 16

DESCRIPTION. Small rhynchonellide, 18 mm long, 15 mm wide and 5-6 mm thick. Ornament
of 32-35 fine, rounded costellae, with 8-10 on low dorsal fold and a corresponding number in
shallow ventral sulcus in the anterior part of the pedicle valve. Marked concentric growth-lines
become more lamellose anteriorly and thicken at the shell margins. [Internal characters unknown
owing to lack of material.]

Name. ‘Southern’.
HototyPe. The specimen BB.76770 figured here as Figs 15a-c.

Locatity. W. N. Croft’s loc. D.533. On the saddle between the two peaks on Humps Islet,
63°59’ S, 57°25’ W, James Ross Island.

Horizon. Upper Cretaceous. Lower Campanian.
MATERIAL. The holotype, BB.76770 and two other fragmentary specimens, BB.76771, BB.76772.

REMARKS. The arrangement of costellae and the lamellose nature of the concentric growth-lines
are suggestive of a relationship with Tegulorhynchia from Tertiary and Recent localities. Similar
shell ornament, however, occurs in Rhynchonella crenifera which was described by Stoliczka
(1872-73) from the Arrialoor Groop (Campanian) of the Upper Cretaceous of Mulloor, southern
India (pl. 1, figs 9 & 10 only). Protegulorhynchia meridionalis differs from this species in its more
circular general outline and shorter or less produced umbo. Nevertheless, it is felt that the two
forms are closely related and Rhynchonella crenifera Stoliczka is therefore here assigned, albeit
tentatively, to the genus Protogulorhynchia, largely on similarity of general morphology and
ornament.

130 E. F. OWEN
Family CANCELLOTHYRIDIDAE Thomson, 1926
Subfamily CANCELLOTHYRIDINAE Thomson, 1926
Genus TEREBRATULINA dOrbigny, 1847
Terebratulina buckmani sp. nov.
Fig. 9
1910 Terebratulina lenticularis Tate; Buckman : 28, pl. 3, fig. 4.

DescrIPTION. Terebratulina 24:9 mm long, 20 mm wide and approximately 18 mm thick. Broadly
oval in general outline, greatest width just anterior to mid-line. Slightly produced umbo truncated
by large circular, submesothyridid, attrite foramen. Deltidial plates obscured by slightly inflated
posterior part of dorsal umbo. Beak-ridges distinct; interarea moderately well defined, fairly
extensive. A faint sulcus is just discernible in the anterior part of the ventral valve. The anterior
commissure is incipiently uniplicate.

Name. For S. S. Buckman.

Loca.ity. W. N. Croft’s loc. D.491.5, Cockburn Island.
HOoLotyPe. The specimen BB.76594 figured here, Fig. 9.
Horizon. Lower Tertiary. Glauconitic Bank.

REMARKS. Buckman figured a uniformly oval specimen (1910 : pl. 3, fig. 4) from the Glauco-
nitic Bank of Cockburn Island with a maximum width about midway between the umbo and

Fig. 4 Notosaria seymourensis sp. noy. Holotype, BB.76590, Lower Tertiary, SW of Cross Valley,
Seymour Island. 1-5.

Figs 5a—c Hemithiris antarctica S. S. Buckman. BB.76592, Lower Tertiary, Glauconitic Bank,
northeastern section, Cockburn Island. x 1-5.

Figs 6a—c Liothyrella lecta (Guppy). BB.76591, Lower Tertiary, northeastern section, Cockburn
Island. x 1-5.

Figs 7a-c ‘Terebratella sp. BB.76784, Lower Campanian, Lachman Crags South, James Ross
Island. x 1-5.

Fig. 8 ‘Laqueus’ cockburnensis sp. nov. Holotype, BB.76595, Lower Tertiary, Glauconitic Bank,
Cockburn Island. x 1-5.

Fig. 9 Terebratulina buckmani sp. nov. Holotype, BB.76594, Lower Tertiary, Glauconitic Bank,
Cockburn Island. x 1-5.

Fig. 10 Lingula antarctica S. S. Buckman. Lower Tertiary of IPS loc. 16, Seymour Island. x 1-5.

Figs lla-c Tegulorhynchia imbricata (S. S. Buckman). BB.76593, Lower Tertiary, Glauconitic
Bank, Cockburn Island. x 1°5.

Figs 12a-c Magella australis (S. S. Buckman). BB.76615, Lower Tertiary, Glauconitic Bank,
Cockburn Island. x 1-5.

Figs 13a-c ‘Terebratula’ sp. BB.76788, Upper Cretaceous, Lower Campanian, Lachman Crags
South, James Ross Island. 1°5.

Figs 14a-c Terebratulina sp. BB.76789, Lower Campanian, Humps Islet, James Ross Island.
i195).

Figs 15a—c, 16 Protegulorhynchia meridionalis gen. et sp. nov. Figs 15a—c, Holotype, BB.76770.
Fig. 16, BB.76771. Both Lower Campanian, Humps Islet. 1-5.

a—Dorsal view. b—Lateral view. c—Anterior view.

ANTARCTIC BRACHIOPODS 131

132 E. F. OWEN

anterior commissure. In assigning the specimen to Tate’s Terebratulina lenticularis, which was
originally described from the Miocene of Aldinga, South Australia, Buckman pointed out certain
discrepancies between the two forms which amounted to a difference in overall outline, beak
features and sulcation of the ventral valve. He considered that the more circular outline and
considerably smaller dimensions of Tate’s specimens were because of their younger individual
age. Comparison of the specimen figured here as Terebratulina buckmani sp. noy. has been made
with a series of specimens in the British Museum (Natural History), which were collected and
identified by R. Tate and sent to T. Davidson. The largest of nine individuals was only 8-5 mm
in length. All the specimens showed less extensive interareas, and the shell ornament of fine
imbricate costellae was given to more frequent bifurcation than in our Antarctic specimen.

It must be admitted that, even with careful reconstruction, the present crushed specimen
could never assume the uniform outline shown in Buckman’s drawing (1910: pl. 3, fig. 4). It
is possible that such differences in outline would be within the normal limits of variability of
T. buckmani and, since neither form can be positively assigned to any other described species of
Terebratulina within the Tertiary, it is felt that a new taxon is justifiable.

Family TEREBRATELLIDAE King, 1850

Subfamily BOUCHARDIINAE Allan, 1940
Genus BOUCHARDIA Davidson, 1850
Bouchardia antarctica S. S. Buckman
Figs 19a—c, 20a—b, 2la—c, 22a—b, 23, 24, 25, 26a—b
1910 Bouchardia antarctica S. S. Buckman : 16; pl. 1, figs 2, 3.
1910 Bouchardia ovalis S. S. Buckman : 16; pl. 1, fig. 1.
1910 Bouchardia angusta S. 8S. Buckman : 16; pl. 1, fig. 4; pl. 3, fig. 2.

1910 Bouchardia elliptica 8S. S. Buckman : 17; pl. 1, fig. 5.
1910 Bouchardia attenuata 8. S. Buckman : 17; pl. 1, fig. 6.

EMENDED DESCRIPTION. Bouchardia approximately 23:8 mm long, 15-2 mm wide and 9:9 mm
thick. Elongate-oval in general outlme. Test thick, smooth, punctate. The umbo is short and
massive with an obtuse beak dominated by a moderate to large, circular, epithyridid foramen.
Extensive interarea bounded by sharp beak-ridges. The slightly concave symphytium is well
exposed.

Eight to ten marked concentric growth-lines feature prominently on the almost equally convex
valves. A strong carination of the pedicle valve persists anteriorly giving rise to an acute carinate
anterior commissure.

Internal structures. The umbonal cavity of the pedicle valve is reduced in size by the lateral
thickening of the shell and the swollen bases of the elongate hinge-teeth. A low septal ridge
extends the length of the valve.

Within the brachial valve the cardinalia are thickened and fused. The area normally occupied
by the hinge-trough in other Terebratellidae is filled with the thickened crural bases to form a
hinge-platform. The premagadiniform brachial loop is supported by a strong, high, median
septum.

Loca.itTigs. W. N. Croft’s locs D.498, D.504, D.506, D.507, D.515, D.519 and those of the IPS
T-30, 22, 4.49; all from Seymour Island.

Horizon. Lower Tertiary.
MATERIAL. 122 specimens, BB.76596-98, BB.76620-743, and 77 specimens in the IPS.

REMARKS. Buckman described five species of Bouchardia from the Younger Beds of the Tertiary
deposits on Seymour Island. He distinguished them chiefly on grounds of overall general outline,
some appearing more oval or elliptical and others with less acute carination of the pedicle valve.

ANTARCTIC BRACHIOPODS 133

Fig. 17 Scatter diagram showing relationship of length to width in specimens of Bouchardia
antarctica S. S. Buckman; same material as in Fig. 18. @-—IPS loc. T-30. A — IPS locs 4, 22, 45.
O-W.N. Croft locs.

All the specimens examined by him were collected from localities within the same geological
horizons specified at that time. There is little to be said in favour of such narrow taxonomic
splitting as Buckman effected, and it is more likely that, within the entire assemblage occurring
in the Tertiary rocks on Seymour Island, no more than two actual species exist. Variants assigned
to the species Bouchardia ovalis, B. elliptica, B. attenuata and B. angusta can be recognized in
random samples from the same horizon and localities as those of the dominant species B. antarc-
tica. The variation in form seems to be affected principally by the breadth of the valves or the
extent of the hinge-line, those individuals with narrower or less extensive hinge-lines appearing
somewhat more oval in outline.

Samples containing a mixture of specimens previously identified as B. antarctica, B. ovalis,
B. elliptica, B. attenuata and B. angusta, together with samples of both large and small unidenti-
fied specimens collected from Seymour Island by members of the IPS expedition of 1974~75,
have been measured. Taking straightforward parameters of length, width and thickness of the
valves, the measurements have been plotted and are figured here as scatter diagrams, Figs 17 and
18.

134 E. F. OWEN

Fig. 18 Scatter diagram showing relationship of length to thickness in specimens of Bouchardia
antarctica §. 8. Buckman; same material and symbols as in Fig. 17.

The scatter diagrams clearly mdicate a single species for the five taxa erected by Buckman
(1910) and for the species represented by specimens collected by the IPS. These forms are here
assigned to Bouchardia antarctica S. S. Buckman.

Ihering (1897 : 268) described Bouchardia zitteli from the Tertiary of Patagonia. He figured
a specimen (1897 : fig. 6) which, although similar in general outline to B. antarctica Buckman,
differs in having fewer and more pronounced concentric growth-lines and also in its faint carina-
tion of the anterior part of the dorsal valve. B. zitte/i was again cited by Ihering (1903 : 334)
and a specimen figured (1903 : pl. 3, fig. 9), together with a specimen of Bouchardia patagonica
(1903 : pl. 3, fig. 10) which he had described from the Palaeocene of St Jorge Gulf, Pico, Sala-
manea, Chile. An error in the numbering of the specimens figured on the plate has caused con-
fusion over the identification of these species. The numbers have in fact been transposed. Levy
(1964 : 213) drew attention to this error when describing specimens of B. patagonica from the
type locality.

In the same publication Levy (1964) also described and figured examples of Bouchardia zitteli
Ihering (unfortunately misspelled as B. zitte//i) from the Lower Tertiary of Santa Cruz, Argen-
tina. She figured four examples (1964 : pl. I, figs 3, 4) showing a fair range of variation in outline

ANTARCTIC BRACHIOPODS 113)5)

from broadly elongate-oval to subpentagonal in form. On the same plate (pl. |, figs 5a—-d) she
figured examples of B. transplatina from the Miocene of the Buenos Aires area which she com-
pared to B. antarctica Buckman, pointing out the difference in the carination of the ventral valve
which, in B. transplatina \hering, appears to be more acute than in Buckman’s species. She
figured a damaged specimen (figs Sa—c) with a strong carination of the pedicle valve and a corres-
ponding sulcus in the anterior part of the brachial or dorsal valve.

Thomson (1918 : 260) had considered B. angusta Buckman and B. transplatina Ihering to
be indistinguishable. As previously stated, B. angusta and B. antarctica are merely variants within
the same species. It may be proved, at some future date, that B. antarctica Buckman and B.
transplatina Ihering are synonyms. No examples of B. transplatina were available for compara-
tive study when the present publication was being prepared and, therefore, no confirmation of
Thomson’s ideas was possible.

Subfamily TEREBRATELLINAE King, 1850
Genus TEREBRATELLA d Orbigny, 1847
Terebratella crofti sp. nov.

Figs 27a—c

DESCRIPTION. Medium-sized biconvex pentagonal terebratellide approximately 27-7 mm long,
24-9 mm wide and 13 mm thick. Dorsal valve with moderate umbonal convexity. Median sulcus
originating from about the middle of the dorsal valve and broadening anteriorly. A median
septum can be seen extending from the posterior umbo to just over half the length of the dorsal
valve. Concentric growth-lines appear irregularly spaced but are well defined.

The ventral valve is acutely carinate with a broad hinge-line and massive umbo truncated by a
large circular, mesothyridid foramen. Beak-ridges are distinct and define an extensive, smooth
interarea. Anterior commissure sulcate. Faint radiating costellae are just visible on the shell
surface, becoming stronger anteriorly.

NAME. For W. N. Croft.
Locatity. IPS coll. loc. 4, north Seymour Island.
Hototyer. The specimen figured here, Figs 27a—c. IPS, Ohio State University, U.S.A.

Horizon. Lower Tertiary.

REMARKS. The figured specimen bears a superficial resemblance to a specimen described by Buck-
man (1910: pl. 1, fig. 17a-d) as Magasella antarctica from the Glauconitic Bank of Cockburn
Island. It differs from that species in its greater dimensions, more acutely pentagonal general
outline, wider hinge-line and less acutely inflated dorsal umbo. It is somewhat similar, however,
to the specimen described and figured by Ortmann (1902 : 78; pl. 14, fig. 2a, b) as Terebratella
gigantea from the Cape Fairweather Beds, Cape Fairweather, Patagonia. Ortmann’s specimen
is considerably larger and has a less pentangulate outline.

Levy (1961 : 87; pl. 3, fig. la-c) described and figured a similar form from the Lower Tertiary
of Patagonia as Pachymagas gigantea (Ortmann), following the assignment of that species by
Thomson (1927: 286) to the genus Pachymagas. Levy’s specimen, however, does not appear to
have many of the characters associated with Pachymagas, such as hinge teeth with swollen bases,
a produced umbo and an extensive interarea. Instead it resembles many specimens hitherto
assigned to Magellania venosa (Solander). Nevertheless, it does bear a resemblance to the speci-
men described and figured here as Terebratella crofti sp. nov.

‘Terebratella’ sp.
Figs 7a-c

DESCRIPTION. Elongate-oval terebratellide 19mm long, 14:3 mm wide and 18-1 mm thick.
Shell biconvex, ornamented by 25-30 strong, rounded radiating costae, given to frequent bifurca-
tion and interrupted by marked, step-like concentric growth-lines. The short, massive umbo of

136 E. F. OWEN

the pedicle valve is truncated by a large, circular foramen. The beak-ridges are indistinct and the
interarea is poorly defined. The anterior commissure is incipiently uniplicate.

Loca.ity. W. N. Croft’s loc. D.422.2, Lachman Crags South, James Ross Island.
Horizon. Upper Cretaceous. Campanian.
MATERIAL. The single specimen BB.76784 figured here, Fig. 7a-c.

REMARKS. Although tentatively assigned to ‘Terebratella’ sp., the specimen is of a distinctly
Magellania aspect. It does not have a median fold in the brachial valve, nor does it develop a
sulcus in the pedicle valve, two features which distinguish it from other costate species assigned
to Magellania. A faint septum, extending to about half the length of the shell, is just visible in the
brachial valve. Owing to the rather poor state of preservation of the specimen, there is no clear
indication of the development of dental lamellae in the pedicle umbo.

The specimen was found in beds of supposedly Campanian age and from a locality recorded
by Spath (1953 : 58) as both C.41370 and C.41377, from which the ammonite Gunnarites kalika
(Stoliczka) was also recorded.

Genus MAGELLANIA Bayle, 1880

Magellania antarctica (S. S. Buckman)
Figs 30a-—c, 31

1910 Pachymagas antarcticus S. 8. Buckman : 21; pl. 2, figs 5-7.

EMENDED DESCRIPTION. Elongate-oval Magellania, approximately 54mm long, 37 mm wide
and 26 mm thick. Umbo massive, suberect, truncated by large, subcircular, mesothyridid fora-
men; beak-ridges attrite, fairly prominent, interarea extensive. Large conjunct deltidial plates
form broad, subquadrate symphytium. Shell surface smooth with numerous fine concentric
growth-lines. Anterior commissure incipiently sulcate.

Pedicle valve more acutely convex than the brachial valve and with a degree of carination
posteriorly. A low median fold is developed at about midway between the pedicle umbo and the
anterior commissure.

Loca.ity. W. N. Croft’s locs D.491.4, Glauconitic Bank, Cockburn Island, and D.519.3, 45 m
SW of headland of Penguin Bay, Seymour Island.

Horizon. Lower Tertiary.

MATERIAL. Two damaged specimens, one almost complete, but with damaged anterior, BB.76599,
and one single pedicle valve, BB.76744, showing massive, peg-like hinge teeth.

REMARKS. Although unlike other Tertiary species of Magellania previously described, M. antarc-
tica is nevertheless comparable in size and general outline to Magellania patagonica (Sowerby)

Figs 19-26 Bouchardia antarctica S.S. Buckman. All x 1-5. Figs 19a—c, BB.76596, Lower Tertiary,
Seymour Island. Figs 20a—b, 22a—b, IPS coll. 4.49, Figs 20a—b showing the atypical wide hinge-
line, somewhat similar to B. zitteli Ihering. Figs 21a—c, IPS coll. T-30, showing slightly more
parallel sides to the valves. Fig. 23, interior of brachial valve from IPS coll. T-30, showing the
cardinalia and brachidium. Figs 24, stereo pair showing brachial valve with cardinalia and brachidium
as internal moulds; BB.76597, Lower Tertiary of Seymour Island. Figs 25, stereo pair of a brachial
valve showing cardinalia and brachidium; the hinge-line is more quadrate in outline than in the
previous specimen. Lower Tertiary, Seymour Island. Figs 26a—b, a large specimen from IPS
T-30, cleaned of matrix to show the massive brachial platform and strong median septum support-
ing the thickened brachial loop.

Figs 27a-c Terebratella crofti sp. nov. Holotype, IPS coll. loc. 4, Seymour Island. 1-5.

Fig. 28 Rossithyris humpensis gen. et sp. nov. Holotype, BB.76773, Upper Cretaceous, Lower Cam-
panian, Humps Islet. 1-5.

a—Dorsal view. b—Lateral view. c—Anterior view.

ANTARCTIC BRACHIOPODS 137

20a 22a

138 E. F. OWEN

from the Miocene of Patagonia. It differs from this species, however, in its less acute inflation
of the brachial umbo and in the incipient sulcation of the anterior commissure.

Buckman (1910) figured three specimens under the name of Pachymagas antarcticus. One of
these figures (1910 : pl. 2, fig. 3) is said to be greatly restored. There was enough of the specimen
left to remind Buckman of the species described by Ihering (1903) as Pachymagas venter from
the Lower Patagonian of Patagonia. Buckman also pointed out that there was a similarity in
form between his species and that of Magellania venosa as figured by Davidson (1886-88 : pl.
14, fig. 2). It is very largely on the strength of this remark and not so much on the rather poor
reconstruction of Buckman’s figured specimen (1910 : pl. 2, fig. Sa-c) that the specimens des-
cribed and figured here as Magellania antarctica (Buckman) are assigned to that genus and
species.

Genus MAGELLA Thomson, 1915

Magella australis (S. S. Buckman)
Figs 12a—c

1910 Magasella australis S. S. Buckman : 19; pl. 1, figs 14-16; pl. 3, fig. 3a—b.

EMENDED DESCRIPTION. Subquadrate to oval biconvex Magella, 9:7 mm long, 9 mm wide and
5-1 mm thick. Test thin, smooth, with irregularly-spaced but well-defined concentric growth-
lines. Pedicle umbo massive with suberect beak dominated by large incomplete submesothyridid
foramen, guarded by short discrete deltidial plates.

Brachial valve with slightly inflated umbo. A shallow sulcus develops anteriorly and gives rise
to a sulcate anterior commissure.

Internal characters. Long, high median septum in the brachial valve supports a strong magelliform
brachial loop.

Loca.ity. W. N. Croft’s locs D. 491.4 and D.526, Cockburn Island.
Horizon. Lower Tertiary. Glauconitic Bank and Pecten Conglomerate.

MATERIAL. Eight specimens from the Glauconitic Bank, BB.76615-22, and two from the Pecten
Conglomerate BB.76623-4.

REMARKS. In his original description of the species Buckman (1910 : 19) refers to the anterior
commissure as being *.. . folded to an M-Shape.’ In fact, the specimen he figured (1910 : pl. 1,
fig. 14c) is shown with the ventral valve uppermost which, by present-day standards, would be
upside down. Thus, his description of M-shaped anterior commissure should be interpreted as
w-shaped. Even so, the term does not fit the specimen figured by Buckman, nor can any speci-
men recently examined for the present study be described as having such a commissure.

The eight specimens examined here show a range from near rectimarginate to having a faintly
sulcate anterior commissure, the sulcation developing late in the development of the shell and,
therefore, being more apparent in the more mature specimens.

Thomson (1915 : 396), in his description of the genus Magella, suggested that Magasella
gouldi, which Dall (1870) had described from recent seas of Japan, was based on the young
stages of another terebratellide, possibly a species of Magellania, and, on this account, referred
Buckman’s species Magasella australis to Magella.

Family LAQUEIDAE Thomson, 1927
Subfamily LAQUEINAE Thomson, 1927
Genus ROSSITHYRIS nov.
Type SPECIES. Rossithyris humpensis sp. nov.

DIAGNosIs. Small, oval, evenly biconvex laqueinid. Umbo short, foramen small, permesothyridid.

ANTARCTIC BRACHIOPODS 139

Test smooth, evenly punctate. Commissure plane. Dental lamellae short. Brachial loop laqueini-
form.

NAME. From James Ross Island.

Rossithyris humpensis sp. nov.
Fig. 28

DESCRIPTION. The shell is approximately 11-6mm long, 9:°8mm wide and 5mm thick. Al-
though evenly biconvex, there is a slight inflation of the umbonal region of each valve. A well-
marked median septum extends just over half the length of the brachial valve. Numerous very
faint concentric growth-lines appear at irregular intervals on the shell surface.

Internal characters. Pedicle valve with short subparallel dental lamellae supporting thick, in-
wardly-directed, peg-like hinge-teeth.

Brachial valve with a comparatively narrow, v -shaped, steep-sided hinge-trough with well-
developed inner and outer socket ridges broadening anteriorly and supported by a strong, high
median septum. Elongate triangular hinge-plates give rise to the descending branches of the
brachial loop which quickly develop inwardly-curving crural processes.

The broad transverse band of the ascending branches, which forms the simple laqueiniform
hood, appears, in transverse serial sections, at an earlier stage in Rossithyris than in either the
Recent Laqueus or Waconella from the Cretaceous, and develops anteriorly. The descending
branches connect with the median septum by means of long, thin, lateral connecting bands,
Fig. 29. Similar bands are seen in transverse serial sections of Laqueus californicus (Koch) and
Waconella wacoensis (Roemer) (Owen 1970 : 76-77).

NAME. From Humps Islet.

Locatity. W. N. Croft’s loc. D.533.2 and recorded by Spath (1953 : 60) as ‘on the saddle between
the two peaks on Humps Islet, lat. 63°59’S; long. 57°25’W.’

Ho.otyPe. The specimen BB.76773 figured here as Fig. 28.

Horizon. Upper Cretaceous, probably Lower Campanian. The fauna is associated with Gun-
narites antarcticus (Weller) and Tetragonites sp., both ammonites previously described from
localities of Campanian age within the Antarctic (Spath 1953 : 29, Howarth 1958 : 10).

MATERIAL. The holotype, BB.76773 and four fragmentary specimens, BB.76774-7.

REMARKS. Although only five specimens of this species have been collected so far, they represent
a form hitherto undescribed and closely related to the Recent genus Laqueus. Rossithyris humpensis
differs from that genus in that it is considerably smaller and in minor differences of brachial
loop development, having a less complex but fundamentally similar expansion of the transverse
band of the ascending branches. In Laqueus this band develops a more expansive hood with lateral
lacunae. The attachment of the descending branches to the median septum occurs at an earlier
stage than in any previously-described member of the Laqueinae and, in this respect, Rossithyris
is similar to a species described below as “Laqueus’ cockburnensis.

Genus LAQUEUS Dall, 1870

‘Laqueus’ cockburnensis sp. nov.
Fig. 8
DESCRIPTION. Brachial valve 22:4 mm long, 23-1 mm wide. Subpentagonal to elongate-oval in
outline. Test thin, smooth with numerous faint concentric growth lines on external surface. A
faint median sulcus is developed late in the anterior portion of the brachial valve. No cardinal
process seen. Subquadrate hinge-trough, with well-developed inner and outer socket ridges, is
supported by a strong, high median septum which extends to about one-third the length of the
valve. The descending branches of the brachial loop are given off from the anterior part of the
hinge-trough and quickly develop long, inwardly curving crural processes. A thin straight band

140 E. F. OWEN

M=M™

Ey nee

"A \ "a N

Rs Oe ee
mews os

Fig. 29 Fifteen transverse serial sections through the umbo of a specimen of Rossithyris humpensis
gen. et sp. nov., Lower Campanian, Humps Islet. 4.

ANTARCTIC BRACHIOPODS 141

of attachment between the descending branches of the brachial loop and the median septum
develops at an earlier stage than is found in the Recent species Laqueus californicus (Koch). It
is noted, however, that in specimens of Laqueus vancouveriensis (Davidson) from the waters of
Naha Bay, south-eastern Alaska, an earlier attachment of loop to septum occurs and that this
stage is accompanied by a less expansive or simpler development of the transverse band or
laqueiniform hood in the ascending branches.

NAME. From Cockburn Island.

Loca.ity. W. N. Croft loc. D.491, Cockburn Island.
Ho.oryPe. The specimen BB.76595 figured here, Fig. 8.
Horizon. Lower Tertiary. Glauconitic Bank.

MATERIAL. Two brachial valves and three fragmentary pedicles valves, BB.76595, BB.76785,
BB.76786, BB.76787.

REMARKS. As no complete brachial loop has been preserved with any of the specimens available,
it is not possible to assign the species described here to any genus beyond a broad sensu lato
determination. It is hoped that more material will become available for further study at some
future date.

The fact that there are certain morphological features which all members of the Laqueinae
have in common does not imply that there is any direct relationship between ‘Laqueus’ cock-
burnensis and Rossithyris humpensis gen. et sp. nov., described here (p. 139) from the Upper
Cretaceous, Campanian of Humps Islet, Ross Island. It is interesting to note, however, that in
both species the attachment of descending branches of the loop to the median septum occurs
earlier than in other members of the Laqueinae.

Superfamily TEREBRATULACEA Gray, 1840
Family TEREBRATULIDAE Gray, 1840

Genus LIOTHYRELLA Thomson, 1916

Liothyrella lecta (Guppy)
Fig. 6a—c
1866 Terebratula lecta Guppy : 296; pl. 19, fig. 3.

1910 Terebratula lecta Guppy; S. S. Buckman : pl. 2, figs 1, 2; pl. 3, fig. 1.
1910 Terebratula vitreoides Tate; S. S. Buckman : pl. 2, fig. 3.

Examples of this species have been adequately described and figured by Buckman (1910) and no
emendation of the original description is required.

The specimen figured here, Fig. 6a—c, is an elongate-oval terebratulide 33 mm, long, 26-3 mm
wide and 12 mm in thickness. The maximum width occurs at a little over half the length of the
shell anteriorly. The fine concentric growth-lines on the smooth surface of the shell are crossed by
numerous faint radiating grooves. This character was included by Guppy (1866) in his original
description of the species and was noted by Buckman when recording and describing the species
from the Antarctic. It is thus distinct from Terebratula vitrioides Tenison-Woods, 1878, from the
Miocene of Tasmania, which it closely resembles, and differs also from that species in its less
robust appearance and less inflated brachial valve. The anterior margin of L. /ecta is incipiently
uniplicate, whereas that of Terebratula vitreoides is seen to develop a very faint biplication in
some young and adult forms.

Loca.ities. W. N. Croft’s locs D.491.4, D.491.5, D.491.6 and D.525.3, Cockburn Island.
Horizon. Lower Tertiary. Glauconitic Bank.

MATERIAL. Eighteen specimens and shell fragments, BB.7659!, BB.76604—14, BB.76778-83.

33a

Figs 30a—c, 31 Magellania antarctica (S. S. Buckman). All 1-5. Figs 30a—c, BB.76599, Lower
Tertiary, Glauconitic Bank, Cockburn Island. Fig. 31, single pedicle valve BB.76744, Lower
Tertiary of Seymour Island.

Figs 32a—c, 33a—e_ Liothyrella anderssoni sp. nov. Lower Tertiary, IPS coll. loc. 45, Seymour Island.
<1. Figs 32a—c, Holotype.

a—Dorsal view. b—Lateral view. c—Anterior view.

ANTARCTIC BRACHIOPODS 143

Liothyrella anderssoni sp. nov.
Figs 32a—c, 33a—c

DESCRIPTION. Elongate-oval Liothyrella, averaging 48-2 mm long, 33-1 mm wide and c. 24 mm
thick. Test smooth with numerous irregularly-spaced, well-defined concentric growth-lines.
Massive, slightly produced, pedicle umbo dominated by a large subcircular mesothyridid labiate
foramen. Delthyrium subquadrate, deltidial plates conjunct. Short, slender hinge-teeth, well
developed. Anterior commissure rectimarginate.

Name. For J. G. Andersson.

Loca.ity. IPS coll. loc. 45, c. 2 km NE of Cross Valley, Seymour Island.

HoLotyPe. The specimen figured here, Figs 32a—c. IPS, Ohio State University, U.S.A.
Horizon. Lower Tertiary.

MATERIAL. Three fragmentary pedicle valves.

REMARKS. This is the first record of a species of Liothyrella from the Tertiary of Seymour Island.
The species occurs in the top beds of the La Meseta Formation (unit IV of Elliot et al. 1975)
and cannot be directly related to any known species of Liothyrella previously described. It is a
robust biconvex form with a massive umbo and labiate foramen, which somewhat resembles
L. oamarutica (Boehm, 1904) from the Miocene of Kakanui, North Olap, New Zealand, in its
general morphology. It differs from that species, however, in its more elongate-oval outline or
narrower development of the anterior part of the shell. The marked pedicle collar seen in L.
oamarutica is absent in L. anderssoni, but the arrangement of the concentric growth-lines on the
shell surface and the degree of convexity of the pedicle valve is reminiscent of that species.

In addition to the above systematic descriptions, a biplicate terebratulide of Campanian age is
figured here, Fig. 13a—c. The locality is given as Lachman Crags South, James Ross Island, given
in Croft’s notes as D.422.2. Approximately 50 specimens said to be from the same locality can
be definitely assigned to the genus Bouchardia. The occurrence of these specimens in beds of
presumably Upper Cretaceous age leaves an element of doubt regarding the accuracy of the
locality citation.

Without knowledge of the internal structures of this form, it is impossible to give any more
than an arbitrary suggestion as to its generic status. It has external morphological features
similar to those of a Concinnithyris sp. from the Middle to Upper Cretaceous of the northern
hemisphere.

Conclusions

Buckman (1910 : 34) drew up a table dividing the Tertiary faunas of both Seymour Island and
Cockburn Island into three main groups, as follows.

a. The Pleistocene, represented by the Pecten Conglomerate of Cockburn Island and containing
a distinct fauna of Hemithiris antarctica, Magasella australis and Magellania fontainei.

b. Pliocene—Miocene of Seymour Island in which his Bouchardia antarctica and the varieties
B. ovalis, B. angusta, B. elliptica and B. attenuata were dominant and Lingula antarctica was
comparatively rare.

c. Miocene—Oligocene, represented by the Glauconitic Bank of Cockburn Island, containing
Hemithiris australis, H. imbricata, H. plicigera, H. squamosa, Magasella antarctica, Pachymagas
antarcticus, Terebratula trinitatensis, T. bulbosa, T. lecta, T. vitrioides, Terebratulina lenticularis
and T. oamarutica.

As far as can be ascertained, there is no palaeontological evidence for the age determination
of these deposits except by comparison with similar brachiopod faunas from Tertiary deposits
in South America and Australasia. Buckman’s descriptions and records of the species, however,
were based upon poorly-preserved specimens, some of which had been greatly restored and re-
constructed.

144 E. F. OWEN

The fauna described here from similar localities on Seymour Island varies little in constitution
from that described in Buckman’s original work except for the addition of Notosaria seymourensis
sp. nov., a first record for that rhynchonellide genus from as far south as Antarctica, Terebratella
crofti sp. nov., Magellania antarctica, previously described and recorded only from the Glauco-
nitic Bank, Cockburn Island as Pachymagas antarcticus, and Liothyrella anderssoni sp. nov., a
new record for the genus from Seymour Island. By far the most important of these records is
that of Notosaria since no species assigned to that genus has been recorded outside the Miocene
or Recent seas.

There is no additional evidence, therefore, either to confirm Buckman’s claims to the age of
the deposits or to deny them. Zinsmeister (1977) records a new species of the gastropod genus
Struthioptera from the Seymour Island Series and considered this, together with a reference to a
member of the Venericardia planicosta group, sufficient evidence for a Late Eocene age for
these beds, supporting Cranwell’s (1959) data for the slightly earlier age of Palaeocene for the
basal 80 m of the same deposits.

The recommendation by Elliot & Trautman (1979) that the term Seymour Island Series
should be replaced by Seymour Island Group containing two major formations, La Meseta
Formation and the Cross Valley Formation is accepted here, but in their description of the beds
to the north of the island, little mention is made of the existence of brachiopod faunas contained
within the 450 m type secion. A broad reference to invertebrate fossils and an occasional
Lingula appear in the text but the positions within the column of examples of Bouchardia antarc-
tica, which was collected extensively, are not indicated. If, at some future date, account is taken
of the relative positions of the varieties of Bouchardia within the column, together with data on
their size differences, perhaps some useful information regarding the stratigraphical significance
of the brachiopod faunas can be expected.

Acknowledgements

I would like to record my sincere thanks to Dr Michael Thomson, of the British Antarctic
Survey, Cambridge, and Dr D. H. Elliot, of the IPS, for their helpful co-operation; and also to
Dr C. H. C. Brunton, Dept of Palaeontology, British Museum (Natural History), for reading the
typescript and offering sound advice. I would also like to thank Mr Adrian Rissoné of the same
department for his work with the statistical analysis.

References

Adie, R. J. 1958. Geological investigations in the Falkland Islands Dependencies since 1940. Polar Rec.,
Cambridge, 9 (58) : 3-17.

Andersson, J. G. 1906. On the Geology of Grahamland. Bull. geol. Instn Univ. Upsala 7 : 19-71, 6 pls.

Boehm, G. 1904. Uber tertiare Brachiopoden von Oamaru, Siidinsel Neu-Seeland. Z. dt. geol. Ges.,
Berlin, 56 (4) : 146-150, 1 pl.

Buckman, S. S. 1910. Antarctic Fossil Brachiopoda. Wiss. Ergebn. schwed. Stidpolarexped., Stockholm,
3 (7) : 1-40, 3 pls.

Comacho, H. H. 1974. Biostratigrafia de las formaciones marinas del Eoceno y Oligoceno de La Pata-
gonia. An. Acad. nac. Cienc. exact. fis. nat. B. Aires 26 : 39-57.

— & Fernandez, J. A. 1955. La trangression en la costa Atlantica entre Comodoro Rivadavia y el curso
inferior de Rio Chabut. Revta Asoc. geol. argent., Buenos Aires, 11 (1) : 23-45.

Cooper, G. A. 1959. Genera of Tertiary and Recent rhynchonelloid brachiopods. Smithson. misc. Collns,
Washington, 139 (5) : 1-90, 22 pls.

Cranwell, L. M. 1959. Fossil pollen from Seymour Island, Antarctica. Nature, Lond. 184 : 1782-1785.

Croft, W. N. (1947). Notes on the geology of Cockburn and Seymour Islands. 17 pp. Falkland Islands
Dependencies Survey (internal report, unpublished).

Dall, W. H. 1870. A revision of the Terebratulidae and Lingulidae. Am. J. Conch., Philadelphia, 6 : 88-168,
3 pls.

Davidson, T. 1886-88. A Monograph of Recent Brachiopoda. Trans. Linn. Soc. Lond. (2) Zool. 4 : 1-248,
30 pls.

ANTARCTIC BRACHIOPODS 145

Elliot, D. H., Rinaldi, C., Zinsmeister, E. J., Trautman, T. A., Bryant, W. A. & del Valle, R. 1975. Geological
investigations on Seymour Island. Antarctic JI U.S., Washington, 10 (4) : 182-186.

—— & Trautman, T. A. 1979. Lower Tertiary strata on Seymour Island. Proc. 37th biol. Collog.(Corvallis,
Ore., 1976). (In press).

Guppy, R. J. L. 1866. Tertiary Brachiopoda from Trinidad. Q. J/ geol. Soc. Lond. 22 : 295-297, 1 pl.

Howarth, M. K. 1958. Upper Jurassic and Cretaceous ammonite faunas of Alexander Land and Graham
Land. Scient. Rep. Falkld Isl. Depend. Sury., London, 21 : 1-16, 5 pls.

Hutton, F. W. 1873. Catalogue of the Tertiary Mollusca and Echinodermata of New Zealand in the collection
of the Colonial Museum. xvi+48 pp. Wellington.

Ihering, H. von 1897. Molluscos dos terrenos terciarios de Patagonia. Rev. Mus. paul., Sao Paulo, 2 : 217-
382, 7 pls.

—— 1903. Les Brachiopodes Tertiares de Patagonia. An. Mus. nac. B. Aires (3) 2 : 321-348, 1 pl.

Levy, R. 1961. Sobre algunos Terebratellidae de Patagonia (Argentina). Ameghiniana, Buenos Aires, 2
(5) : 79-88, 2 pls.

— 1964. Acerca de los Géneros Bouchardiella y Bouchardia (Braquidpodos) en el Terciario de Patagonia
(Argentina). Ameghiniana, Buenos Aires, 3 (7) : 212-218, 2 pls.

Owen, E. F. 1970. A revision of the brachiopod subfamily Kingeninae Elliott. Bull. Br. Mus. nat. Hist.,
London, (Geol.) 19 : 27-83, 14 pls, 17 figs.

Ortmann, A. E. 1902. Tertiary Invertebrates. Rep. Princeton Univ. Exped. Patagonia 4, Palaeontology:
48-332, 27 pls.

Simpson, G. G. 1971. Review of fossil penguins from Seymour Island. Phil. Trans. R. Soc., London, B5
(178) : 357-387.

Spath, L. F. 1953. The Upper Cretaceous cephalopod fauna of Graham Land. Scient. Rep. Falkld Isl.
Depend. Sury., London, 3 : 1-60, 13 pls.

Stoliczka, F. 1872-73. The Cretaceous fauna of southern India. IV. The Brachiopoda. Mem. geol. Surv.
India Palaeont. indica, Calcutta, 4 (7) : 1-32, 7 pls.

Tenison-Woods, J. E. 1878. On the Tertiary deposits of Australia. J. Proc. R. Soc. N.S.W., Sydney, 11 :
65-82, 1 pl.

Thomson, J. A. 1915. Brachiopod genera: the position of shells with Magaselliform loops, and of shells
with Bouchardiform beak characters. Trans. Proc. N.Z. Inst., Wellington, 47 : 392-403.

— 1916. Additions to the knowledge of the Recent & Tertiary Brachiopoda of New Zealand and Aus-
tralia. Trans. Proc. N.Z. Inst., Wellington, 48 : 41-47, 1 pl.

— 1918. The genus Bouchardia and the age of Younger Beds of Seymour Island. Geo/. Mag., London,
(6) 5 : 258-263.

— 1927. Brachiopod morphology and genera (Recent and Tertiary). N.Z. Board of Sci. & Art. Manual 7 :
V+ 338 pp., 2 pls.

Wilkens, O. 1911. Die mollusken der antarktischen Tertiar formation. Wiss. Ergebn. schwed. Siidpolarex-
ped., Stockholm, 3 : 1-62.

Zinsmeister, W. J. 1977. Note on a new occurrence of the Southern Hemisphere Aporrhaid gastropod
Struthioptera Finlay & Marwick on Seymour Island. J. Paleont., Menasha, 51 (2) : 399-404.

Revision of the rugose coral Diphyphyllum concinnum
Lonsdale, 1845 and historical remarks on
Murchison’s Russian coral collection.

B. R. Rosen and R. F. Wise

Department of Palaeontology, British Museum (Natural History), Cromwell Road, London
SW7 5BD

Synopsis
The type specimens of Diphyphyllum concinnum Lonsdale, 1845, which is also the type species of the im-
portant Carboniferous rugose coral genus Diphyphyllum, had been thought lost, but one has recently
been found again, and is here redescribed. Comparison with recently erected neotypes suggests that they
are only doubtfully conspecific. Remarks are given on the history of Murchison’s Russian coral collec-
tions, which some authors had thought lost, and a list is given of those of his corals held in the British
Museum (Natural History) collections.

Introduction

The objects of this note are to redescribe an important type specimen that has been found again,
to correct the belief held by some other coral workers that Murchison’s Russian corals described
by Lonsdale are lost, and to give some historical background to these corals.

The type specimens of Diphyphyllum concinnum Lonsdale, 1845, collected by Sir Roderick
Murchison from the Lower Carboniferous of the Urals, Russia (Murchison ef a/. 1845), are
important because of the status of D. concinnum as the type species of the widespread Carbon-
iferous rugose coral genus Diphyphyllum. Moreover, this genus has long been of additional
problematic interest because there has always been uncertainty about its relation to the equally
important genus Lithostrotion, with which it would seem to intergrade. In this respect the genus
has therefore been the type example of the ‘diphymorph’ trend or condition (Smith & Lang 1930).
The type specimens have been apparently missing for about half a century, but one was found
again by one of us (R. F. W.) in 1972. Rediscovery provides an opportunity to give a new des-
cription and figures, rendered desirable by the need for a good transverse section and detailed
confirmation of the description given originally by Lonsdale. We have done so here on the
recommendation of Professor Dorothy Hill (University of Queensland, Australia). Redescription
is now made additionally interesting by the recent designation of a neotype of D. concinnum
from the type locality by Ivanovski1 & Shurygina (1975). Dr J. R. Nudds (Trinity College,
Dublin) is currently preparing a detailed study of the genus and its relationship to other litho-
strotionids.

The above Russian authors have stated (1975 : 14) that Murchison’s Russian corals are
lost, an impression which is presumably shared by some at least of other coral workers. This is
only partly correct. We have found several other type specimens very recently as a result of check-
ing carefully through material in our care, stimulated by this statement by Ivanovskii & Shurygina.
Others have never been lost at all. Several well-established taxonomic names were first used by
Lonsdale for corals described by him in Murchison’s collections, and a list of relevant material
held by the British Museum (Natural History) is now desirable. A brief history of Murchison’s
coral collection is also appropriate. We are currently preparing redescriptions of other corals
from Murchison’s Russian travels held by the Museum.

All specimen numbers used here are register numbers of the Department of Palaeontology,
British Museum (Natural History). In the remainder of this paper, the three institutions concerned

Bull. Br. Mus. nat. Hist. (Geol.) 33 (2) : 147-155 Issued 31 January 1980
147

148 B. R. ROSEN & R. F. WISE

will be referred to by their initials, MPG (Museum of Practical Geology), GSL (Geological
Society of London) and the BM(NH).

It is a pleasure to acknowledge help from Dr Pierre Semenoff-Tian-Chansky (Institut de
Paléontologie, Paris) on the coral terminology used by Ivanovskii & Shurygina, and also from
Dr John R. Nudds (Trinity College, Dublin); Mr R. J. Cleevely, BM(NH), has helped us on the
historical matters. Dr G. F. Elliott, BM(NH), Dr Semenoff-Tian-Chansky and Dr J. Fedorowski
(A. Mickiewicz’s University, Poznan) kindly read the manuscript and offered helpful criticisms.

Historical note on Murchison’s Russian corals

Murchison’s Russian corals came to the BM(NH) by at least two routes. Some were presented
by him to the MPG (which was, in effect, part of the Geological Survey and is now, as the Geo-
logical Museum, part of the Institute of Geological Sciences) of which Murchison was for a long
time the Director. Others appear to have been presented to the collections of the GSL or other-
wise found their way into these collections. In this connection Mr R. J. Cleevely, BM(NH), tells
us that he knows of no Murchison donation of Russian material to the GSL, and has drawn our
attention to the fact that Lonsdale was the Society’s Curator and Librarian during the period
1829-42. All non-British material from both these large and important collections was later
transferred to the BM(NH), the largest part of the material coming from these Institutions in
1880 and 1911 respectively. Woodward (1904 : 314) lists Murchison and Lonsdale in his entry
for the MPG, and gives this 1880 date for the transfer to the BM(NH); see also p 231. Also on
p. 314 there is an entry for Murchison, but the information given applies to material other than
the present Urals corals. There was also a smaller transfer from the MPG to the BM(NH) in
1878 (R. J. Cleevely, personal communication).

The history of the transfer of the Geological Society collections is not summarized anywhere,
and because of their historic interest and importance, this is now given here. In 1911 the GSL
decided to relinquish its collections (Watts 1911a; 19116 : Ix-Ixii; GSL 1911a). British material
was to be offered to the MPG, and ‘foreign and colonial’ material to the Trustees of the BM(NH)
(GSL 19115). Both institutions duly accepted and the collections were transferred (Watts 1912a;
19126: Ixix—Ixxi). Teall (1913) reported on the transfer of the British fossil material and Wood-
ward (1913) on the foreign fossil material. As Woodward explained, it was the intention not to
incorporate any of the material received by the BM(NH) until ‘the whole have been thoroughly
curated, registered and studied’. This task was for some reason never completed, and the type of
D. concinnum was found together with other material from the GSL, still unregistered. At various
times, foreign specimens have been transferred from the Institute of Geological Sciences to the
BM(NH). They were probably unintentionally sent to the MPG in the 1911 transfer, or were
perhaps even left over from the 1880 transfer. Such reasons would account in part for the apparent
loss of some of Murchison’s Russian corals.

List of Lonsdale’s corals from Murchison’s Russian travels

The species below are listed in Lonsdale’s original order using his names and quoting his specimen
details, and giving his page and figure references. However, the species names themseives are
given according to modern procedure. Species represented in the BM(NH) collections are shown
by the listing of their specimen numbers; redescriptions are in preparation. Species not repre-
sented are preceded by an asterisk (*).

Although referred to as corals by Lonsdale, the following species discussed by him are now
regarded as belonging to other groups; their details are not included here. Stromatopora concen-
trica, Monticularia sternbergii, Stenopora spinigera, S. crassa, Fenestella infundibuliformis, F.
retiformis ?, F. veneris ?, F. martis ?.

Of the 30 coral species (excluding the above list) described by Lonsdale, the BM(NH) holds
types, or figured or described specimens, of 19 (28 specimens). Four out of nine Silurian species

DIPHYPHYLLUM AND MURCHISON’S RUSSIAN CORALS 149

are represented and 15 out of 19 Carboniferous species. The remaining corals unrepresented are:
Silurian/Devonian (one), Devonian (one), Devonian/Carboniferous (one), Permian (one). Of
the ten species founded by Lonsdale, eight are represented. Three of these are type species of
genera founded by Lonsdale in this same work and still recognized today. Two of these are
represented.

1. *Syringopora parallela Fischer. Lonsdale 1845 : 591-592. ‘Perimishel, south of Kaluga; Vitegra,
Ilinsk on the Tchussovaya: Carboniferous limestone. Odoyef near Lichvin: Upper Devonian.’

2. Syringopora distans (Fischer). Lonsdale 1845 : 592-593. ‘Ilinsk, on the river Tchussovaya, west of
the Ural Mountains. Carboniferous limestone.’ Specimens: R35936—7 (described only).

3. *Catenipora labyrinthica Goldfuss. Lonsdale 1845 : 593. ‘Isle of Dago, Upper Silurian. Top of Lower
Silurian, at Naissi, in Lithuania.’

4. Chaetetes radians Fischer. Lonsdale 1845 : 595-596; pl. A, figs 9, 9a. ‘Kaluga, and Vitegra, Boro-
vitchi near Valdai, Carboniferous limestone.’ Specimen: 26356 (figs 9, ? 9a), Kaluga.

5. *Chaetetes dilatatus Fischer. Lonsdale 1845 : 596. ‘Borovitchi; Miatchkova. Carboniferous lime-
stone.’

6. *Chaetetes petropolitanus (Pander). Lonsdale 1845 : 596-597; pl. A, figs 10, 10a. ‘Nikolsk to Petro-
pavlosk [sic], on the river Volkof; banks of the Siass river; ravines of Pulkovka and Popovka, south of
St. Petersburgh; plateau of Czarskoe-celo; sea cliffs from Narva to Reval. Lower Silurian.’

7. Lithodendron costatum sp. nov. Lonsdale 1845 : 598-599, text-figs a, b. ‘Perimishel, south of
Kaluga. Carboniferous limestone.’ Specimen: R33575 (text-figs a, b).

8. Lithodendron annulatum sp. nov. Lonsdale 1845 : 599, pl. A, figs 5, Sa. ‘River Issetz, east of Ekaterin-
burg; Ilinsk, on the Tchussovaya. Carboniferous limestone.’ Specimens: R35827 (figs 5, ? Sa), R35828-9
(described only).

9. *Lithodendron concameratum sp.nov. Lonsdale 1845 : 599-600. ‘River Oceter, Government of
Tula. Carboniferous limestone.’

10. Lithodendron fasciculatum J. Phillips. Lonsdale 1845 : 600. ‘River Tchussovaya, west flank of Ural
mountains. Carboniferous limestone.’ Specimen: R33574 (described only).

11. Cladocora ? sarmentosa sp.nov. Lonsdale 1845 : 600-601. ‘Kamensk, east of Ekaterinburg.
Carboniferous limestone.’ Specimens: R49819—20 (described only).

12. Columnaria sulcata Goldfuss. Lonsdale 1845 : 601-602; pl. A, figs 1, la—lc. ‘Habsal, near Reval.
Lower Silurian.’ Specimen: R33606 (figs 1, 1a—Ic).

13. Lithostrotion emarciatum (Fischer). Lonsdale 1845 : 603-605, text-figs a-f on p. 603. ‘Borovitchi,
near Valdai. Carboniferous limestone.’ Specimens: R17177 (text-figs a-c), R17178 (text-fig. f), R17180
(text-fig. d); specimen for text-fig. e not known.

14. Lithostrotion mammillare (Fischer). Lonsdale 1845 : 606-607, text-figs a, b on p. 606. ‘Priksha
(Valdai), Government of Novgorod. Carboniferous limestone.’ Specimen: R17176 (text-figs a, b).

15. Lithostrotion astroides sp. nov. Lonsdale 1845 : 607-608, text-figs a-c on pp. 607-608. ‘Pinega
(sixty versts west); Carboniferous limestone. Tchussovaya banks, above Ust-Koiva; Carboniferous lime-
stone.’ Specimen: R17179 (text-figs a, b, c), Pinega.

16. Lithostrotion floriforme Fleming. Lonsdale 1845 : 608-610, text-figs a-c on p. 609. ‘Borovitchi, near
Valdai. Carboniferous limestone.’ Specimen: R49833 (text-figs a—c).

17. *Favosites alveolaris Goldfuss. Lonsdale 1845 : 610. ‘Isle of Dago, Petropavlosk [sic] and Volshanka
River (North Ural). Upper Silurian.’

18. *Favosites polymorpha Goldfuss. Lonsdale 1845 : 610-611. ‘Katchukof, on the Upper Belaia and
Uziansk Zavod, in the South Ural Chain; Silurian. East of Alatau, South Ural; Devonian ? [sic]. Banks
of the lake of Petropavlofsk sixty versts north-west from the Works of Bogoslofsk, North Ural; Upper
Silurian ? [sic]’.

19. Michelinia concinna sp. nov. Lonsdale 1845 : 611-612; pl. A, figs 3, 3a. ‘East of Ust Koiva, on the
Tchussovaya. Carboniferous limestone.’ Specimen: R33605 (figs 3, ? 3a).

20. Cyathophyllum turbinatum Goldfuss. Lonsdale 1845 : 612-613. ‘Petropavlofsk, N. Ural.
Upper Silurian.’ Specimen: R36125 (described only).

21. *Cyathophyllum (Tryplasma) aequabilis (subgen. et sp. nov.). Lonsdale 1845 : 613-614; pl. A,
figs 7, 7a. ‘The river Kakva; East side of the North Ural Mountains; Silurian. Petropavlofsk, northern-
most Russo-Uralian mines. Silurian or Devonian ? [sic]’.

22. Strombodes sp. Lonsdale 1845 : 614-615; pl. A, fig. 13. ‘Ussa River, junction with the Volga near
Samara. Upper Carbon[iferous] limestone.’ Specimen: R49741 (fig. 13).

23. Cystiphyllum impunctum sp. nov. Lonsdale 1845 : 615. ‘Margin of the lake of Petropavlofsk, sixty
versts north-west from the works of Bogoslofsk. Silurian.’ Specimens: R36144—5 (described only).

150 B. R. ROSEN & R. F. WISE

24. *Caninia sp. Lonsdale 1845 : 616-617. ‘East of Usolie, on the Volga above Samara. Carboniferous
limestone.’

25. Caninia ibicina ? [sic] (Fischer). Lonsdale 1845 : 617-619; pl. A, figs 6, 6a—6d. ‘Velikovo, between
Vladimir and Kovrof. Upper Carboniferous limestone.’ Specimens: R49743 (fig. 6), R49744 (fig. 6a),
R49745 (fig. 6c); specimens for figs 6b and 6d not known.

26. Stylastraea inconferta (gen. et sp. nov.). Lonsdale 1845 : 621-622; pl. A, figs 2, 2a—2c. ‘Kossatchi-
Datchi, south of Miask, eastern side of the Ural Chain. Carboniferous limeste.’ Specimen: R17562
(figs 2, 2a—2c).

27. Diphyphyllum concinnum (gen. et sp. nov.). Lonsdale 1845 : 624-625; pl. A, figs 4, 4a—4c. ‘Hill of
Tchirief, Kamensk, on the river Issetz, eastern side of the Ural Chain. Carboniferous limestone.’ Specimen:
R49740 (figs 4a, 4b, ? 4c; not fig. 4). See below.

28. Porites pyriformis Ehrenberg. Lonsdale 1845: 625-626. ‘Isle of Dago; Petropavlofsk; Gothland;
Malmoe Isle, in Christiania Bay; Upper Silurian.’ Specimen: R31192 (described only), Isle of Dago.

29. *Aulopora conglomerata ? [sic] Goldfuss. Lonsdale 1845 : 626. ‘Isle of Dago. Upper Silurian.’

30. *Anthophyllum ? incrustans [sic] (Phillips). Lonsdale 1845 : 631. ‘Ust-Vaga, i.e. débouchtire of the
Vaga into the Dwina. Permian.’

Systematic description
Class ANTHOZOA
Order RUGOSA

Family LITHOSTROTIONIDAE
Genus DIPHYPHYLLUM Lonsdale, 1845

TYPE SPECIES. Diphyphyllum concinnum Lonsdale, 1845, by original designation of Lonsdale
(1845 : 624); Carboniferous, Urals.
Diphyphyllum concinnum Lonsdale, 1845
(Figs 1-3)
1845 Diphyphyllum concinnum Lonsdale : 624; pl. A, figs 4, 4a—c.
non 1876 Diphyphyllum concinnum Lonsdale; Thomson & Nicholson : 123; pl. 8, figs 1, la [= D. smithi
Hill].

non 1883 Diphyphyllum concinnum Lonsdale; Thomson : 384; pl. 8, fig. 2 [= D. lateseptatum McCoy].
non 1887 Diphyphyllum concinnum Lonsdale; Thomson : 35; pl. 4, fig. 1 [= D. lateseptatum McCoy].

21975 Diphyphyllum concinnum Lonsdale; [neotypes] Ivanovskii:& Shurygina : 17; pl. 2, figs la, b.

MATERIAL. One of Lonsdale’s syntypes of D. concinnum, corresponding to his pl. A, figs 4a, 4b
(possibly also 4c, but not his fig. 4). R49470.

Loca.ity. Lonsdale gives the locality as ‘Hill of Tchirief [=Chiriev], Kamensk, on the river
Issetz [=Iset’], eastern side of the Ural Chain’. Kamensk corresponds to the modern name,
Kamensk-Ural’skiy, 56°25’ N, 61°54’ E (U.S. Army 1970). Carboniferous.

HISTORY OF THE TyPE SPECIMENS. Thomson (1887 : 33) appears to have been the last author
to have worked with the type specimens of D. concinnum, though he did not publish explicit
comparisons of his own material with the types. Blake (1902 : 28) listed the D. concinnum
types, and this represents the most recent previous citation of Lonsdale’s material. Unlike most
of the other coral specimens from Murchison’s Urals collection, listed in the previous section
and still extant, Blake’s citation shows that it definitely came from the GSL collections rather
than the MPG collections. It was found in the BM(NH) in a small collection of unincorporated,
unregistered GSL material representing several groups including corals, and including several
other type specimens (not of Lonsdale). This collection might have arrived at a later date than the
main 1911 transfer (see p. 148), or for some reason was overlooked in the original programme of
incorporation which followed this transfer and was then put on one side. By the time Smith
(1928) wrote his Nemistium paper, the specimen was thought to be lost (1928 : 114). Later
authors confirmed this, including Smith & Lang (1930), who perforce recommended that the
genus Diphyphyllum should be based on another B-form, D. /ateseptatum (1930 : 180), which

DIPH YPH YLLUM AND MURCHISON’S RUSSIAN CORALS 151

Figs 1-3 Diphyphyllum concinnum Lonsdale. Type specimen, R49470. Fig. 1, transverse section,
x16. Fig. 2, longitudinal section, x4. Fig. 3, detail of longitudinal section, 8.

152 B. R. ROSEN & R. F. WISE

they believed to be closely related to D. concinnum. (Hill revised Smith’s original designation
of D. lateseptatum as an a-form in her account of the genus (1940 : 181, 184)). The type material
of D. concinnum had therefore been missing since 1928 at least, and possibly since 1911. Lonsdale
actually based his description on several specimens, of which the one corresponding to his pl.
A, fig. 4 is still missing. His fig. 4c might be the present specimen, but there is no way of knowing
this.

DESCRIPTION. The specimen consists of a single corallite just under 9 mm in width, showing
axial division into two corallites. Specimen length 24 mm. The apparently longitudinal section
is in fact tangential, though quite close to the axis. Of the original specimen, only the section
corresponding to the minor segment is to hand, and full axial details are therefore missing.
Preservation is mostly good.

In the incomplete transverse section, eleven major septa are present, alternating with minors.
Majors in the tabularium taper very slightly towards the axis, and they also thin marginally
in the dissepimentarium to about half their tabularial thickness. Majors reach at least two-thirds
of the distance to the inferred axis; minors are between one-third and a half of the length of the
majors and project only very slightly beyond the dissepimentarium. Majors slightly zigzag in
the tabularium and all septa slightly more so in the dissepimentarium, where there is also a
slight tendency for minors to become marginally indistinct and the dissepiments to appear
inosculating. The wall is lost. Dissepiments concentric to angular marginally, with innermost
dissepiments more regular and smaller in the inner area of the dissepimentarium. The axial region
is lost by original section cutting but a few tabular plates are visible in the most axial region.

In longitudinal section, the dissepimentarium is about two-fifths of the corallite radius. There
are two marginal rows of larger, nearly horizontal dissepiments within which is an inner row of
one or two very much smaller vertical dissepiments. Tabulae are in inner and outer series. Inner
tabulae conspicuous, about 10 per cm in the corallite below division; mostly flat, or very slightly
arched, outermost parts of each downturned steeply to rest upon the one below. Outer tabulae
not conspicuous, mostly horizontal or sloping gently either inwards or outwards. No axial
structures can be seen in this section. The septa are clearly seen as amplexoid spines on upper
surfaces of tabulae.

Division is axial, ‘parricidal’. In longitudinal section, at the point of corallite division, one axial
tabula is strongly arched upward, and tabulae below it show an upward trend of increasingly
arched form prior to corallite division. The new corallite wall rests directly on the summit of
the most arched tabula. The full dissepimentarium structure is established within the new walls
almost at the outset.

INFERRED CHARACTERS. Because of the importance of this specimen, we felt it worthwhile to
attempt a reconstruction of the transverse section in order to infer some of the missing informa-
tion. By extrapolating the septa axially, we found that they mostly converge at a point, which we
could then use to generate a complete circular corallite circumference, making a best fit with the
fragment outline by eye. This reconstructed diameter is found to be a little more than 9 mm.
Using the septal pattern of the existing fragment (Fig. 1) as a template, we found the number of
septa we could fit into the reconstructed outline was 52 (majors and minors). For this we assumed
a fully radial septal arrangement without strongly bilateral or tetrameral features, as this is both
suggested by the existing septa and approximates to the general appearance of many species of
Diphyphyllum and Lithostrotion. On this reconstruction the plane of the existing tangential
longitudinal section is found to be 1:25 mm off centre, 1.e. it lies about two-thirds the way along
the reconstructed radius from the margin. A second reconstruction of the transverse corallite
section, made purely geometrically, gives a slightly larger diameter of 10 mm, over 60 major
and minor septa, and a position of the plane of the tangential section rather further from the
axis, at about five-eighths of the radial distance from the margin. On the basis of both reconstruc-
tions the septal number would seem therefore to be between 50 and 65 majors and minors.

The extent to which the septa extend towards the axis cannot be directly reconstructed, but
those major septa which lie at right-angles to the plane of the longitudinal cut (Fig. 1) are seen

DIPHYPHYLLUM AND MURCHISON’S RUSSIAN CORALS 153

to stop just short of it. The major septa would therefore have a length of at least two-thirds of
the radial distance from the margin. With more material to hand this would probably prove to
be variable.

Discussion. The absence of axial details is of course most unfortunate in a type specimen of
Diphyphyllum and for this reason we have refrained from designating this specimen as the lecto-
type. Perhaps the remaining type material will be found in due course. Lonsdale (1845) placed
great emphasis on the general absence of a columella in his own type descriptions of the present
genus and species, whilst he also explained carefully how an intermittent partial columella spine
was sometimes present on occasional tabulae in both his type material and numerous similar
Bristol specimens. We conclude that a fully reconstructed transverse corallite section of the pre-
sent specimen would belong with Smith’s (1928) Diphyphyllum B-group. It would, however,
differ from his actual figured example of the B-form (1928 : 115) in having slightly longer septa,
a different dissepimentartum and a narrower inner tabularium (than the transverse figure)
with rather flatter tabulae. Similar differences separate D. concinnum from the B-form D. latesep-
tatum, which Smith & Lang (1930, see p. 150) proposed as a type species for Diphyphyllum
because the D. concinnum types were lost at the time. Differences between D. concinnum and
D. lateseptatum are summarized in Table 1. We especially draw attention to the aulos-like struc-
ture, horseshoe-like dissepiments, the shorter and nearly equal minor and major septal length,
and the broad inner tabulae in D. /ateseptatum. The two species are close. however, especially
with regard to septal number, transient columella and tabular form.

Table 1 Comparison between D. Jateseptatum McCoy (in Smith & Lang 1930; Hill 1940) and D.
concinnum

Diphyphyllum lateseptatum McCoy Diphyphyllum concinnum Lonsdale

corallite diameter 3-8 mm 9 mm (7-9-5 mm in Lonsdale)

length of major septa c. + corallite diameter c. 4 corallite diameter

length of minor septa c. + major septa (4 to equal in Hill) c. 4-4 major septa

width of dissepimentarium narrow: c. + wide: c. 2

character of dissepiments presence of globose, almost horseshoe-like series not present

horseshoe-like series

aulos-like structure present not developed

outer tabulae concave flat, but may slope either outwards
or inwards

inner tabulae c. + corallite diameter c. 4 corallite diameter

In view of Dr J. R. Nudd’s forthcoming revision of the whole family, we have not made an
exhaustive survey of other relevant material, but the most important comparison that must be
made here is that with Ivanovskii & Shurygina’s D. concinnum neotypes. Their vertical sections
show a similar dissepimentarium but with a less frequent development of the small vertical
innermost dissepiments and a narrower inner tabularium. The outer tabularium differs in being
clearly trough-shaped and rather wider. Their transverse sections show a distinct but variable
axial feature made up of septal lamellae, the steep sides of the inner tabulae, and even a few
elements that look like tabellae. One of their corallites exhibits an axial plate. The major septa,
though rarely continuous within the tabularium, frequently converge in the axial region. The
authors do not give a full description, but point out that major septa ‘sometimes unite in the
centre forming an intermittent axial column’, a feature which they relate to the development of
amplexoid septa. Their figures however suggest that ‘sometimes’ should really be ‘often’. While
their figures, like Lonsdale’s type, are in Smith’s B-group (above), they are actually closer to
Smith’s (1928) Nemistium than they are to D. concinnum. Smith regarded Nemistium as having
B-characteristics and therefore probably being a development of B-stock. The D. concinnum
neotypes, however, are not conspecific with N. edmondsi Smith 1928.

154 B. R. ROSEN & R. F. WISE

In the absence of axial details in the present specimen these Nemistium-like features of the
neotypes should be contrasted with Lonsdale’s type description, which is completely unambiguous:
‘The indications of an axis were very faint, being confined to the occasional appearance of a
single line in the centre of the area, or to a few instances of conical irregularities in the dia-
phragms.’ The appearance of such irregularities he relates to the axial division. While Lonsdale’s
observation obliges the concepts of both the species and genus to embrace an intermittent axial
feature, there is as yet no evidence that his species would show quite the development of axial
structures seen in Ivanovskii & Shurygina’s figures. Even if the differing degree of axial discon-
tinuity is disregarded, the other axial features mentioned above are clearly excluded by Lonsdale’s
own description.

Perhaps examination of further material will show that the range of variation in Lonsdale’s
specimens overlaps more convincingly with that in the neotypes or that complete intergradation
exists between them. We tried to test this by considering the forms placed in D. concinnum
Lonsdale by Thomson, who strongly advocated the proper recognition of the genus. Thomson
(1887) was evidently the last person to publish coral descriptions based on his having seen the
type specimens before they were lost (see p. 150). We follow Hill (1940 : 184), however, in
placing Thomson’s 1883 and 1887 D. concinnum in the synonymy of D. lateseptatum McCoy
(a different B-group form, as explained by Hill, 1940 : 181). And we believe Thomson & Nichol-
son’s (1876) D. concinnum is a good a-form, and belongs with D. smithi Hill (1940 : 181).

Pending further studies, therefore, we would maintain that Ivanovskii & Shurygina have
provided insufficient evidence that their neotypes are the same species as Lonsdale’s type, though
we recognize the interest and value in presenting topotypic material. It should be noted that our
conclusion is based on Lonsdale’s own type descriptions rather than on any new details observed
here in the type specimen.

References

Blake, J. F. 1902. List of the types and figured specimens recognized by C. D. Sherborn, F.G.S., in the col-
lection of the Geological Society of London. 100+ xxxii pp. London.

Geological Society of London 191 1a. [Special General Meeting, January 25th, 1911.] Proc. geol. Soc. Lond.
(Feb. 1911). In: Q. JI geol. Soc. Lond. 67 : vii.

19115. [Special General Meeting, June 14th, 1911.] Proc. geol. Soc. Lond. (Aug. 1911). In: Q. JI geol.
Soc. Lond. 67 : cii-ciii.

Hill, D. 1940. A monograph of the Carboniferous rugose corals of Scotland, 3 : 115-204, pls 6-11.
Palaeontogr. Soc. (Monogr.), London.

Ivanoyskii, A. B. & Shurygina, M. V. 1975. Reviziya rugoz Urala (V. Lonsdaeyl, 1845; F. N. Chernyshev,
1885, 1893; E. D. Soshkina, 1937; T. V. Nikolaeva, 1849). Trudy Inst. Geol. Geofiz. sib. Otd., Novo-
sibirsk, 218 : 1-44, pls 1-20. [In Russian.]

Lonsdale, W. 1845. Description of some characteristic Palaeozoic corals of Russia. Appendix A in
Murchison, de Verneuil & von Keyserling 1845: 591-634, pl. A.

Murchison, R. I., de Verneuil, E. & von Keyserling, A. 1845. The geology of Russia and the Ural Mountains,
1 (Geology). xxiv+700 pp., 12 pls, 5 col. sections, 2 maps, text-figs. London.

Smith, S. 1928. The Carboniferous coral Nemistium edmondsi gen. et sp. n. Ann. Mag. nat. Hist., London,
(10) 1 : 112-120, pl. 5, text-figs.

—— & Lang, W. D. 1930. Descriptions of the type-specimens of some Carboniferous corals of the genera
‘Diphyphyllum’, ‘Stylastraea’, Aulophyllum, and Chaetetes. Ann. Mag. nat. Hist., London, (10) 5 : 177-
194, pls 7-8.

Teall, J. J. H. 1913. Fossils. In Strahan, A., Annual General Meeting, February 21st, 1913. Report of the
Council for 1912. Proc. geol. Soc. Lond. (March 1913). In: Q. JI geol. Soc. Lond. 69 : x.

Thomson, J. 1883. On the development and generic relation of the corals of the Carboniferous system of
Scotland. Proc. phil. Soc. Glasg. 14 : 296-520, pls 1-14.

— 1887. On the occurrence of species of the genus Diphyphyllum, in the Lower Carboniferous strata of
Scotland, with a description of some new species and notices of varieties. Q. J/ geol. Soc. Lond. 43 :
33-38, pls 4-5.

—— & Nicholson, H. A. 1876. Contributions to the study of the chief generic types of the Palaeozoic corals
[part]. Ann. Mag. nat. Hist., London, (4) 17 : 123-128, pl. 8.

DIPH YPHYLLUM AND MURCHISON’S RUSSIAN CORALS 155

United States Army 1970. Gazeteer No. 42 — U.S.S.R. (2nd ed.) 7 vols. Geographic Names Division, U.S.
Army Topographic Command, Washington.

Watts, W. W. 1911a. Annual General Meeting, February 17th, 1911. Report of the Council for 1910.
Proc. geol. Soc. Lond. (May 1911). In: Q. JI geol. Soc. Lond. 67 : ix—xi.

— 19116. The Anniversary Address of the President. Proc. geol. Soc. Lond. (May 1911). In: Q. JI geol.
Soc. Lond. 67 : lii—xciii.

— 1912a. Annual General Meeting, February 16th, 1912. Report of the Council for 1911. Proc. geol.
Soc. Lond. (Feb. 1912). In: Q. JI geol. Soc. Lond. 68 : vii—ix.

— 1912b. The Anniversary Address of the President. Proc. geol. Soc. Lond. (June 1912). In: Q. JI geol.
Soc. Lond. 68 : x\vili-ci.

Woodward, A. S. 1904. The Department of Geology. In: The history of the Collections contained in The
Natural History Departments of the British Museum, 1 : 197-340. London, Trustees of the British
Museum.

— 1913. Fossils. Jn Strahan, A., Annual General Meeting, February 21st, 1913. Report of the Council
for 1912. Proc. geol. Soc. Lond. (March 1913). In: Q. JI geol. Soc. Lond. 69 : viii—ix.

Neuroptera (Insecta) in amber from the Lower
Cretaceous of Lebanon

P. E. S. Whalley

Department of Entomology, British Museum (Natural History), Cromwell Road, London
SW7 5BD

Synopsis

Glaesoconis fadiacra sp. nov. (Coniopterygidae), Banoberotha enigmatica gen. et sp. nov. (Berothidae)
and Paraberotha acra gen. et sp. noy. (Berothidae) are described from Lebanese amber.

Material and techniques

In recent years the Lebanese amber has aroused considerable interest (Schlee & Dietrich 1970,
Schliiter 1976) as the oldest known amber with insect inclusions. Recently the British Museum
(Natural History) received some Lebanese amber from Professor Aftim Acra of Beirut, who has
been processing and studying the amber for some years.

The material has been variously dated as Aptian or Neocomian, with ages ranging from 100
to 130 m.y. (Acra et al. 1972, Schliiter 1976).

A preliminary examination of the insect inclusions has been made, but the plant fragments
and other inclusions have not been studied. Schlee (1970) points out that the amber is difficult
to handle; it is very fragile and often fragmented with many cracks developed inside the pieces.
These cracks affect the clarity of the amber and its strength, and various clearing and polishing
techniques have been developed to make the inclusions clearer.

The amber varies in colour from pale yellow to dark red, and in a number of examples there is
evidence of some flow of the amber after the insects were trapped. In these cases parts of single
insects may be separated by several millimetres or more. It is not possible to determine whether
the flowing of the amber took place as the insects were trapped, resulting in the head, thorax
and abdomen being separated in the same lump of amber, or whether the amber melted and
flowed long after the insects were trapped. The flow within one piece of amber does not always
occur in the same plane and the pieces of a single insect may be separated in three dimensions.

The amber can be examined dry or by immersion in liquids. Some organic solvents (e.g.
toluene, xylene) rapidly attack the amber and usually 70% alcohol or glycerin was used. Alcohol
spreads through the cracks, giving better visibility, but can lead to further fractures; it will also
slightly attack the amber. Polishing the surface was done with a jewellery burnishing paste which
gave satisfactory results. For greater resolution some of the more fragile pieces were embedded
in plastic and then mechanically polished. In some cases minute contraction of the embedding
plastic can occur as it hardens. Even slight pressure resulted in the collapse of some of the amber
along the fault-cracks, distorting the block.

Other examination methods included the standard microscopical preparation techniques. The
amber, after briefly soaking in xylene, was embedded in Canada balsam. This produced mixed
results, generally proving satisfactory but occasionally the balsam was found partially to dissolve
the amber, causing slight fragmentation of the insect specimen. A similar result was obtained by
embedding in Euparal, but here the Euparal tended to clear the amber considerably, improving
visibility of the specimen. The fragmentation does not destroy the fossil but results in the chitin
being apparently stretched slightly in all directions by the solution effect of the mountant. This
breaks the insect up into a series of small blocks which together retain the original shape of the
fossil. The result is similar to a newspaper picture; at low magnification the picture is clear but
at high magnification the individual particles can be seen. Some details become clearer in this
manner but the specimens should always be studied fully before embedding. In this way any
minor loss of detail may be more than compensated for by the improved definition of other parts.

Bull. Br. Mus. nat. Hist. (Geol.) 33 (2) : 157-164 Issued 31 January 1980
157

158 P. E. S. WHALLEY

Specimens did not always fragment when embedded in Euparal and embedding can result in
dramatically improved visibility of the specimen.

Small pieces embedded in ‘Trylon’ resin have been cut into serial sections on a Lastac wire
saw using a 0-008-inch (0:2 mm) diamond-impregnated wire. The results have been promising
but care is needed in handling the sections owing to the fragility of the amber.

Preliminary infra-red spectroscopic examination has shown a spectrum clearly distinct from
that of Baltic amber and copals.

Systematics

From the amber collected by Professor Acra and at present being studied at the British Museum
(Natural History), twelve orders of insects and three of Arachnida (Pseudoscorpions, Mites,
Spiders) have been recognized. By far the most abundant insects in the amber are the Nemato-
ceran Diptera, with the Hemiptera-Homoptera next, a high proportion of which are Aleyrodidae.
Specimens of the following Orders have been found in the sample from Professor Acra: Thy-
sanura, Collembola, Orthoptera, Dictyoptera, Psocoptera, Hemiptera, Thysanoptera, Neurop-
tera, Lepidoptera, Hymenoptera, Coleoptera and Diptera.

An account of the Lepidoptera in Lebanese amber has been published (Whalley 1977, 1978);
in the present paper the Neuroptera are described.

Six specimens of neuropterous insects have been identified in the sample of Lebanese amber.
These are two Coniopterygidae, three Berothidae and one fragment, probably referable to th-
Myrmeleonidae. No Neuroptera were recorded from Lebanese amber by Schliiter (1976), but he
found specimens in resins of Cretaceous age from France (Schliiter 1975).

Family MYRMELEONIDAE ?

One of the specimens is only fragments of a wing (Figs 1-2), possibly a Myrmeleonid, but there
is insufficient evidence to be certain of this determination. The group is known from the Triassic
onwards (Crowson et al. 1967).

Family CONIOPTERYGIDAE

This family, including the fossils, was monographed by Meinander (1972); later he described
other fossil species (Meinander 1975). One of the species he described is from Karatau (U.S.S.R.)
and is generally regarded as Upper Jurassic; the other is from Siberian amber of Cretaceous age
(Coniacian—Santonian). From this Siberian amber, younger than the Lebanese (Schliter 1976),
Meinander described a new genus and species of Coniopterygidae, Glaesoconis cretica (Aleuro-
pteryginae), placing it in the tribe Fontenelleini. The new specimen from the Lebanese amber is
congeneric with Meinander’s species.

Glaesoconis fadiacra sp. nov.

DESCRIPTION. Front of head damaged. Flagellum with 19 segments, each slightly longer than
broad. Maxillary palp five-segmented, apical segment large and broader than the others. Labial
palps with enlarged terminal segment, number of labial segments unclear. Forewing about
1-7 mm (Fig. 3). Two crossveins in basal costal area. Sc, meeting R, basally of radial crossvein.
Radial crossvein arises from fork of R,+ Rs. M, and M, fork, running separately to wing margin.
Hindwing, partly obscured venation apparently similar to forewing. Apical r-m crossvein on
branch M, of M,+M,. The abdomen is obscured but some evidence of the plicaturae on ab-
dominal segments. [The genitalia are not visible in the specimen. ]

Types. Holotype and Paratype: Lebanese amber, Lower Cretaceous. Coll. Professor Aftim
Acra.

Discussion. There are three main characters which separate this species from G. cretica. The latter
has more antennal segments (24-27). G. fadiacra has longer branches to Rz +R, and R,+R,;,

LEBANESE AMBER NEUROPTERA 159

2

Figs 1-2 Fragments of forewing,? Myrmeleonid.

with their junction more basad, and the terminal segment of the maxillary palp is broader than
the other segments.

With only two specimens available the validity of the intraspecific variation is uncertain. The
presence of a separate M, on the wing margin, which is similar to G. cretica, is considered by
Meinander to be a primitive character. In all Recent genera there are only two branches to the
median vein, but three in G. fadiacra and G. cretica.

The difference between these fossil coniopterygids and Recent species is relatively small and
all the main features of the family were evidently well established by the Lower Cretaceous. The
presence of complex morphological features shows that the coniopterygid type of organization
must have arisen well before the Lower Cretaceous, and the presence of a coniopterygid in

Fig. 3 Forewing, Glaesoconis fadiacra sp. nov. Coniopterygidae. See Fig. 7.

160 P. E. S. WHALLEY

Upper Jurassic rocks (although with much less detail visible compared with the amber specimens)
suggests an even earlier date for the separation of the Coniopterygidae from their sister group.

Family BEROTHIDAE

The characters given below suggest that both the adult specimens of the species described here
are in the Osmyloidea (sensu Crowson et al. 1967). Riek (1970) divides the Osmyloidea, separating
the Mantispoidea to include the Mantispidae, Sisyridae and Berothidae. One of the amber
lacewings has characters which occur in both the Berothidae and the Sisyridae. Comparison of
Banoberotha enigmatica sp. nov. with neuropterous fossils described from the Permian, Triassic
and Jurassic does not provide any more information on its systematic position. Certainly Per-
moberotha, as illustrated by Carpenter (1954), is quite distinct.

The presence of unbranched costal veinlets in the fossil B. enigmatica is the most striking
difference from Recent Berothidae. The hindwings are very similar, particularly with Cu, parallel
to the hindwing margin. The anastomoses near the base of the hindwing in Recent Berothidae
figured by Tjeder (1959) cannot be seen in the fossil. There are, however, sufficient characters
for the fossil to be placed as a new genus and species in the Berothidae.

Gonarcus

Figs 4-5 Banoberotha enigmatica gen. et sp. nov. Berothidae. Fig. 4, forewing.
Fig. 5, apex of abdomen and gonarcus. See Fig. 8.

Genus BANOBEROTAA nov.
TyPE SPECIES. Banoberotha enigmatica sp. nov.

GENERIC DIAGNOSIS. Prothorax elongate. Forelegs cursorial. Forewing with costal veinlets
unbranched. Rs and Sc fused near apex. Radial single with subparallel branches. MP bifurcate.
Hindwing with Cu, parallel to hmdwing margin.

REMARKS. The separation of this genus from Recent berothid genera is based on the costal
veinlets and the differences in cross venation (Fig. 4). While the branching or otherwise of the
costal veinlets is variable within one species of Recent berothids, there are usually some veinlets
that are forked. In Banoberotha all are unbranched. The genus is placed tentatively in the Nosy-
binae (MacLeod & Adams 1967).

Banoberotha enigmatica sp. nov.

DESCRIPTION. Male: Wing 6mm, wingspan about 12-5 mm. Head damaged, eyes prominent.
Mouthparts not clear. Antennae with circlet of hairs on each bead-like segment. Legs broken
but very hairy. Prothorax at least twice as long as broad. Forewing (Fig. 4) with many macro-
trichia on veins. No recurrent humeral vein. Trichosors all round wing margin. Costal margin
broader near base. Radial and subcostal veins join near apex of wing. R, single with four

LEBANESE AMBER NEUROPTERA 161

roughly subparallel branches. MA from radial (sensu Martynoy 1928), MP forked. Cubital
branched. Anal veins not clear. Hindwing frenulum group of bristles. Cubital vein almost
parallel to hindwing margin. Basal area indistinct. Abdomen damaged, posterior segments as in
Fig. 5. This shows the gonarcus (?) forced out; its shape is distorted by fine air bubbles around
it and the figure represents the best assessment of the outline of the structure.

Hotorype. Male, Lebanese amber, Lower Cretaceous. Coll. Professor Aftim Acra.

DISCUSSION. This small fossil lacewing has been compared with examples of all the Recent families
and with figures of all the fossil families currently regarded as neuropterous. The fossil shows
clearly the specialization reached by the Lower Cretaceous Neuroptera. The frenulum, which can
be clearly seen in the fossil, has not previously been seen in fossil specimens earlier than the
Tertiary, because of problems of preservation. The whole insect, except for certain venation
differences, has the appearance of a Recent berothid. The abdomen is not clear enough for de-
tailed examination but its apical shape is indicated (Fig. 5), together with what may be the
gonarcus. The latter has fine spines at the apex and may well have the arms as indicated in the
figure, although the shape is distorted by the air present.

Banoberotha possesses characters of both berothids and sisyrids; the simple costal veinlets are
typical of Sisyridae, but all the other characters, e.g. the branched median vein of the forewing
and Cu, in the hindwing, are typical of the Berothidae. The elongate prothorax occursin berothids
but not in sisyrids, where the pronotum is broader than long. This elongate prothorax is similar
to many hemerobiids, but the fossil lacks the recurrent humeral vein and the arrangement of the
radial veins differs from that family.

Amongst the Lebanese amber material collected by Professor Acra is the larger part of a neurop-
terous larva (Figs 9-10). While some details of the abdomen are obscured, the mouthparts quite
clearly suggest a berothid larva.

DescriPTION. Including the long mandibles, the larva is 6-8 mm long, very hairy, probably
broader in the middle and with a relatively elongate prothorax. The mandibles are long and
thin, while the maxillae are shorter but broader at the base, giving a distinctly triangular look to
the mouthparts. Lateral ocelli are clear; each side has five ocelli, arranged in a half-circle of four
plus one ocellus ventrally placed to this group. The antennae end in a fine seta, but the number of
segments is not clear. The frons is rounded with four long hairs near the anterior margin. There
is a prominent epicranial suture, and the labial palps are 4- or 5-segmented. The single apical
tarsal segment preserved ends without a claw, but has a trumpet-shaped empodium.

Discussion. The larvae of only a few Recent species of Berothidae are known, but the most
important diagnostic feature is the broad bases to the maxillae (Riek 1970), which show clearly
in the fossil. No evidence for the subfamily position of this larva can be deduced since adult
lacewings from two distinct Berothid subfamilies have been found in the same amber.

The final specimen from the Lebanese amber, although damaged, shows most of the characters
of the Berothidae, but especially the raptorial forelegs characteristic of the endemic African
subfamily, Rhachiberothinae. It is distinct from the two Recent genera but still within the sub-
family Rhachiberothinae, which evidently has a long fossil history.

Genus PARABEROTHA nov.

TYPE SPECIES. Paraberotha acra sp. nov.

GENERIC DIAGNOSIS. Prothorax not elongate, forelegs raptorial. Forewing with Sc and R joining
near apex. Trichosors on wing margin. Subcostal veinlets branched in basal third.

REMARKS. From Rhachiberotha Tjeder the new genus is separated by the shorter pronotum and
the fusion of Sc and R near the apex. It is similar to that genus in the more medial position of the

162 P. E. S. WHALLEY

antennal sockets. In the Recent genus Mucroberotha Tjeder these are placed higher up, level with
the upper part of the eye. The genus Paraberotha is placed in the Rhachiberothinae; species in this
subfamily are currently known only from Zimbabwe-Rhodesia.

Paraberotha acra sp. nov.

DESCRIPTION. Wing 4mm. Head damaged, mouth parts missing, eyes prominent. Antennae
with circlets of hairs on each segment, long, with bead-like, slightly elongate, segments. Pronotum
broader than long. Tip of abdomen and most of legs missing. Forelegs raptorial (Fig. 12), with
two rows of teeth on femur interlocking with spines on tibia and tarsi. Teeth 21-23 in each row
on femur, 10-15 long spines on tibia. Probably 5 tarsal segments, 4 visible. First segment longer
than rest with long spines, segments 2-3 enlarged apically with prominent spine. Wings partially
obscured. Sc and R fused near apex, separate towards base. Costal veinlets bifurcate in basal third
of forewing. No recurrent humeral vein. Hindwings partially obscured but Cu elongate, parallel

to hind margin.

4 1 VSIA a
Ps tees —

Fig. 6 Forewing, Paraberotha acra gen. et sp. nov. Berothidae. See Figs 11-12.

HototyPe. Male (?), Lebanese amber, Lower Cretaceous. Coll. Professor Aftim Acra.

Discussion. Comparison of Paraberotha acra with the Recent species Rhachiberotha signifera
Tjeder shows many interesting parallels and some differences. The forewing venation is broadly
similar, but the Recent species has more subcostal veinlets bifurcating and Sc and R not fusing
below the apex. The teeth of the femur are very similar in Recent and fossil species, especially in
their construction and insertions, and differ only in the arrangement, being alternately long and
short in the fossil while rather irregular in length, though in similar rows, in the Recent species.
The arrangement of teeth on the femur differs amongst the few Recent species in the Rhachi-
berothinae.

The overall similarity of Recent and fossil species is striking. Probably one of the most funda-
mental differences is in the shape of the pronotum. In most Recent species (Mantidae, Mantis-
pidae, Rhachiberothinae), where the forelegs are modified for grasping and holding the prey,
the pronotum is elongate; in the fossil Paraberotha it is quite clearly not elongated.

Tjeder (1959) regarded the Rhachiberothinae as a possible connecting link between the Manti-
spidae and Berothidae. MacLeod & Adams (1967) review this relationship in more detail and,
while not agreeing with the reasons given by Tjeder, agree with his conclusions and this possible

relationship.

Apart from Lebanese amber, fossil Berothids are known only from Baltic amber (Eocene/
Oligocene); these are probably in the Berothinae (MacLeod & Adams 1967). Schliiter (1978)
describes Retinoberotha stuermeri as a new genus and species from the Cenomanian of north-west

France.
Little is known of the biology of Recent Berothidae. For example, while the forelegs of Recent

Rhachiberothinae are typically raptorial and are almost certainly used to catch prey, this has
not actually been observed, and the modifications of the legs might even be used to assist in
courtship, although I believe this unlikely.

LEBANESE AMBER NEUROPTERA 163

$

Fig. 7 Glaesoconis fadiacra sp. nov. See Fig. 3.
Fig. 8 Banoberotha enigmatica gen. et sp. nov. See Figs 4—5.
Figs 9-10 Berothid larva. Fig. 10, enlargement of mouth parts.

Figs 11-12 Paraberotha acra gen. et sp. nov. Fig. 11, head and raptorial foreleg (arrowed) on left.
Fig. 12, enlargement of foreleg. See Fig. 6.

When considering the faunal and floral changes since the Lower Cretaceous, the morphological

stability of the Nosybinae and Rhachiberothinae is apparent. It is possible that this very stability
also accounts for their apparent rarity in Recent faunas and their extremely limited distribution;

while the Berothinae are widespread throughout the world, although not abundant, the Recent
Rhachiberothinae and Nosybinae are rare and restricted to central and southern Africa.

164 P. E. S. WHALLEY

Acknowledgements

I am grateful to Professor Aftim Acra, Beirut, for permission to study this material. I appreciate
the advice of my colleague, Dr P. Barnard, who has made useful comments on the problems
involved with these fossils. Advice and assistance with embedding, cutting and polishing from
the BM(NH) Departments of Mineralogy and Palaeontology is gratefully acknowledged.

References
References marked * are not cited in the text, but are included as contributing to the bibliography of the subject.

Acra, A., Milki, R. & Acra, F. 1972. The occurrence of amber in Lebanon. Abstr. Réun. scient. Ass.
Libanaise Avymt Sci., Paris, 4e : 76-77 (UNESCO).

*Carpenter, F. M. 1940. A revision of the Nearctic Hemerobiidae, Berothidae, Sisyridae, Polystoechotidae
and Dilaridae (Neuroptera). Proc. Am. Acad. Arts Sci., Boston, 74 : 193-280.

— 1954. In Brues, C. T., Melander, A. L. & Carpenter, F. M., Classification of Insects. Bull. Mus.
comp. Zool. Hary., Cambridge, Mass., 108 : 1-917.

Crowson, R. A., Rolfe, W. D. I., Smart, J., Waterston, C. D., Willey, E. C. & Wootton, R. J. 1967. Arthro-
poda. /n Harland, W. B. er a (eds), ine Fossil Record : 499-534. Spec. publs geol. Soc. Lond., 2.
*Kriiger, L. 1922. Berothidae. Beitrage zu einer Monographie der Neuropteren-Familie der Berothiden.

Stettin. ent. Ztg 83 : 49-88.
1923. Neuroptera succinica baltica. Stettin. ent. Ztg 84 : 68-92.
*MacLeod, E. G. 1967. Studies on the systematics of the Berothidae, 1. The genus Sphaeroberotha Navas
with a critique of the taxonomic characters used in the Berothinae (Neuroptera). Psyche, Camb. 74 :
342-352.
1970. The Neuroptera of the Baltic Amber: Ascalaphidae, Nymphidae and Psychopsidae. Psyche,
Camb. 77 : 147-180.
& Adams, A. 1967. A review of the taxonomy and morphology of the Berothidae with the description
of a new subfamily from Chile. Psyche, Camb. 74 : 237-265.

Martynov, A. 1928. Permian fossil insects of North-East Europe. Trudy geol. Muz., Leningrad, 4 : 1-118,
19 pls.

Meinander, M. 1972. A revision of the family Coniopterygidae (Neuroptera). Acta zool. fenn., Helsinki,
136 : 1-357.
—— 1975. Fossil Coniopterygidae (Neuroptera). Notul. ent., Helsinki, 55 : 53-57.
Riek, E. F. 1970. Neuroptera. /n The insects of Australia : 472-494. Melbourne (C.S.I.R.O.).
Schlee, D. 1970. Insektenfossilien aus der unteren Kreide, 1. Verwandtschaftsforschung an fossilen und
rezenten Aleyrodina (Insecta, Hemiptera). Stuttg. Beitr. Naturk. 213 : 1-72.

& Dietrich, H. G. 1970. Insektenfiihrender Bernstein aus der unteren Kreide der Libanon. Neues Jb.
Geol. Paldont. Mh., Stuttgart, 1970 (1): 40-50.

Schliiter, T. 1975. Nachweis verschiedner Insecta-Ordines in einem mittelkretazischen Harz Nord-
westfrankreichs. Ent. germ., Berlin, 1 : 151-161.

— 1976. Die fossile Harz der gegenwartige Erforschungsstand. Naturw. Rdsch. Stutt. 29 : 350-354.

—— 1978. Zur Systematik und Palokologie harzkonservierter Arthropoda einem Taphoz6nose aus dem
Cenomanian von NW-Frankreich. Berl. geowiss. Abh. A 9 : 1-150.

Tjeder, B. 1969. Neuroptera planipennia, the Lacewings of Southern Africa. 2. Family Berothidae.

S. Afr. anim. Life, Stockholm, 6 : 256-314.
Whalley, P. 1977. Lower Cretaceous Lepidoptera. Nature, Lond. 266 : 526.
— — 1978. New taxa of fossil and Recent Micropterigidae with a discussion of their evolution and a
comment on the evolution of the Lepidoptera (Insecta). Ann. Transy. Mus., Pretoria, 31 : 71-86.

>

4
be
5
‘
,
i
®

Titles to be published in Volume 33

An account of the Ordovician rocks of the Shelve Inlier in West
Salop and part of north Powys. By the late W. F. Whittard, F.R.S.
(Compiled by W. T. Dean)

Miscellanea

The Caradoc faunal associations of the area between Bala and
Dinas Mawddwy, north Wales. By M. G. Lockley

Fossil insects from the Bembridge Marls, Palaeogene of the Isle of
Wight, southern England. By E. A. Jarzembowski

The Yorkshire Jurassic fern Phlebopteris braunii (Goeppert) and its
reference to Matonia R. Br. By T. M. Harris

The entire Geology series is now available as a series

Printed by Henry Ling Ltd, Dorchester

British Museum (Natural History)

The Yorkshire Jurassic Flora by T. M. Harris

Hard bound in grey cloth covers, size 30-8 x 24-5 cm

Vol. I. Thallophyta - Pteridophyta
1961, v+212pp, 71 text figures £9.00

Vol. Il. | Caytoniales, Cycadales and Pteridosperms
1964, viii+ 19lpp, 7 plates, 67 text figures £10.50

Vol. III. Bennettitales
1969, vit+ 186pp, 7 plates, 69 text figures £12.00

Vol. IV. Ginkgoales and Czekanowskiales

Contributions by W. Millington & J. Miller
1974, viii+150pp, 8 plates, 41 text figures £19.50

Vol. V._— Coniferales
1979, vit 166pp, 7 plates, 68 text figures £22.50

Complete in five volumes

SG6Ml.o.

Bulletin of the N
peeuish Museum (Natural History)

4

: Caradoc faunal associations of the

_ area between Bala and Dinas Mawddwy,
q morth Wales

M. G. Lockley

Geology series. Vol 33.No3 27 March 1980

The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four
scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology, and
an Historical series.

Papers in the Bulletin are primarily the results of research carried out on the unique and
ever-growing collections of the Museum, both by the scientific staff of the Museum and by
specialists from elsewhere who make use of the Museum’s resources. Many of the papers are
works of reference that will remain indispensable for years to come.

Parts are published at irregular intervals as they become ready, each is complete in itself,
available separately, and individually priced. Volumes contain about 300 pages and are not
necessarily completed within one calendar year. Subscriptions may be placed for one or more
series. Subscriptions vary according to the contents of the Volume and are based on a forecast
list of titles. As each Volume nears completion, subscribers are informed of the cost of the
next Volume and invited to renew their subscriptions. Orders and enquiries should be sent to:

Publications Sales,
British Museum (Natural History),
Cromwell Road,
London SW7 5BD,
England.

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

© Trustees of the British Museum (Natural History), 1980

ISSN 0007-1471 Geology series

Vol 33 No 3 pp 165-235

British Museum (Natural History)
Cromwell Road
London SW7 5BD Issued 27 March 1980

The Caradoc faunal associations of the area between
Bala and Dinas Mawddwy, north Wales

M. G. Lockley .
Department of Geology, Lilybank Gardens, University of Glasgow G12 8QQ, Scotland

Contents

Synopsis : : : : ; : : : : : ‘ , . 166
Introduction . : : : : ; ; ; , ; ; ; . 166
Stratigraphy . : : : : : é . 167
Aims, sampling procedures and data analysis : 5 : : 3 ‘ . 169
Data synthesis and presentation : : : : : : P F ele
The faunal succession : : : : : : ; : 3 : 5 7s}
Faunal associations . : : 5 3 . 3 : 2 : : . 188
Diversity patterns . : 5 : : 4 : : ! ‘ : . 197
Taxonomic descriptions . ; : : : ‘ : ; ; . 202
Phylum Brachiopoda Dumeril 3 : : : 5 ‘ 3 4 . 202
Class Inarticulata Huxley . ; 2 : : x : ; : . 202
Superfamily Lingulacea Menke é : : : 3 : : . 202
? Lingulella sp. . : : : : : 7 ; 202
Lingulella cf. ovata (M’ Coy) ; é : : : : : . 203
Palaeoglossa cf. attenuata (Sowerby) . : ‘ 5 ; : . 203
? Pseudolingula sp. : : ; ‘ é : : : : . 203
Paterula sp. ; : : : : : . 203
Paracraniops cf. paeehen Williams , : 5 é : . . 203
Paracraniops glaber sp. nov. ' 5 : , : F é . 204
Paracraniops cf. glaber Lockley . : : 5: 5 : i . 205
Superfamily Craniacea Menke : : ; : : F 3 . 205
Orthisocrania sp. : : 3 ; : , ; : : . 205
Class Articulata Huxley 5 ; : ; ; : : 2 . 205
Superfamily Orthacea Woodward 2 : 5 : é ; : . 206
Nicolella cf. actoniae obesa Williams : : : 5 : . 206
Rhactorthis crassa Williams . 5 ; 4 : : ; : . 209
Superfamily Enteletacea Waagen . ; : : : ‘ 3 . 209
Howellites cf. ultimus Bancroft. : ; ; ; : : . 209
Howellites cf. antiquior (M’Coy) . : : ‘ ; ; : . 209
Onniella ostentata Williams A , : q j : : . 209
Bancroftina sp. . : : : ; : 3 : : : . 209
Salopia sp. ; ' F 5 : ; ‘ 3 . 210
Superfamily Tripleciacea Schuchert 5 ; 5 ; : F : . 210
Triplesia maccoyana Davidson : : é ; : : : . 210
Superfamily Plectambonitacea Jones : ; : ‘ : 5 . 210
Palaeostrophomena canalis sp. nov. ; ; : : ; : . 210
Anisopleurella cf. multiseptata Williams . : s : : é . 211
Sericoidea abdita Williams, emend. : P : 3 : . , Alli
Sericoidea abdita complicata subsp. nov. . 3 5 : : 5 wl2
Bimuria dyfiensis sp. nov... 4 : : : : : 5 PAS
Superfamily Porambonitacea Davidson ; ! ; ’ 5 ; . 216
Parastrophinella brenchleyi sp. nov. : : : : : : . 216
Superfamily Atrypacea Gill . : : : é 5 : , 5 lls)
Protozyga musculosa sp. nov. } ‘ ; : : ; s 5 ZA
Superfamily Dayiacea Waagen : : : : P : : 5 AS
Cyclospira aff. bisulcata (Emmons) : ; 3 5 : : . 219
Superfamily unknown . ‘ ' 3 : ; : : ; . 220
? Spiriferide, gen. indet. : : é : : ; ; 220

Bull. Br. Mus. nat. Hist. (Geol.) 33 (3): 165-235 Issued 27 March 1980

165

166 M. G. LOCKLEY

Phylum Arthropoda. ; : ; : : ‘ : ‘ ; . 220

Class Trilobita Walch ; : : : ; : : . 220

Family Phillipsinellidae Whittington. 7 , 3 : ; ‘ . 220

Phillipsinella sp. : : : : : : ; . 220

Family Raphiophoridae Angelin : ‘ : ; : : ; nee:

Lonchodomas sp. : j : ‘ : : ? 3 5 in,

Family Cheiruridae Salter 3 : : : : : ‘ ; 5 222?)

Sphaerocoryphe sp... : ‘ ‘ ; Z F ; : 5 2p

Family Proetidae Salter . ; ; : : ; ; ; ; » 22

Cyphoproetus sp. : ‘ : : ; : : ; ; 5 | pep

Phylum Mollusca ; : : : : ‘ ; ; : : 5

Class Bivalvia Linné . ; ; ' : : : i eee,

Family Praenuculidae McAlester , : : : : ; ; 5 ee

? Cardiolaria sp. : : ; ; : : 5 2733

Family Malletiidae Adams & Adams : : ; 5 5 : eres

Nuculites sp. : ‘ : : : : ; ; 5 3

Family Modiomorphidae Miller : : ; , : : : = 223

Modiolopsis sp. . : : 5 . , 5 : : 5 2

Family Cyrtodontidae Ulrich ; 5 : : : 3 ‘ ; 5

Cyrtodonta sp. . : : : 3 : : , ; a 2723)

Family Grammysiidae Miller : ; ‘ : F ; ‘ : . 224

Cuneamya sp. . : : : : : : : . : . 224

Class Gastropoda Cuvier . ; i : ; : 4 : . 224

Class Monoplacophora Wenz in Knight : : F 5 : 2 . 226

Family Archinacellidae Knight : : : j ; F ‘ 5226

? Archinacella sp. : ‘ : : é : ; ‘ : 5 PNG

Class Cephalopoda Cuvier F ; ; ; : : 3 : . 226

Phylum Echinodermata : 2 ; . 3 : ; : : 26

Class Crinoidea Miller : : ; : 5 : . 226

Family Archaeocrinidae Moore & Lauden ° : : : : . 226

Balacrinus basalis (M’Coy) . : : : é 3 : , . 226

Class Stelleroidea Lamarck : : 3 : : : ; : =» 226

Family Stenasteridae Schuchert ‘ : ‘ j : : : > 226

Stenaster obtusus (Forbes) . ; : : : ; : : . 226

Faunal communities and associations F ; : F ‘ 5 s g Pi

Conclusions . : : ; : : j j ; : ; 3 a 278)

Acknowledgements . : ; ; : 5 5 ; : : : . 230

References : 3 é ; : ; j ‘ ‘ ' : : . 230

Index ; é : ‘ : : : , ; ; : : + 233
Synopsis

A systematic scheme of sampling in key sections is used to provide a quantitative outline of the distribu-
tion of the total fauna in the greater part of the structurally less-disturbed, fossiliferous portion of the
Lower Bala Group (and some equivalent deposits) south of Bala Lake. Sampling, data collecting and
analysis techniques are discussed and derived quantitative parameters are used to describe six inter-
related, brachiopod-dominated associations.

Representatives of five brachiopod, four trilobite and six molluscan genera are recorded in the
group for the first time. Four new brachiopod species and a subspecies are described. These are Para-
craniops glaber sp. nov., Palaeostrophomena canalis sp. noy., Bimuria dyfiensis sp. nov., Protozyga musculosa
sp. nov. and Sericoidea abdita complicata subsp. nov.; Parastrophinella brenchleyi sp. nov. is also described
from contemporary deposits in the NE Berwyns.

The composition of faunal associations from the Lower Bala Group is compared with the structure
and characteristics of named contemporary associations in adjacent areas.

Introduction

The researches presented here are linked to a reappraisal of the geology of the Llanuwchllyn to
Llanymawddwy area presented elsewhere (Lockley 1980). The present paper is principally

CARADOC FAUNAL ASSOCIATIONS 167

concerned with the quantitative description of faunal assemblages and associations found in the
fossiliferous upper part of the lower to middle Caradoc succession between Bala and Dinas
Mawddwy. To this end a series of suitable sections were chosen for detailed examination, at
more or less regular intervals along a 20 km portion of the strike belt shown in Fig. | (inset).
Laterally spaced sample sites provide a framework in which to examine the relationships between
faunal distribution patterns and vertical and lateral facies changes.

At each of the twelve localities shown in Fig. 1 a series of from 3 to 84 samples were collected
at regular, measured vertical intervals. Horizons from which samples were derived are informally
referred to — in conjunction with sample numbers — as ‘beds’; e.g. “bed H1’ (Fig. 4) refers to the
horizon from which the material of sample H1 was derived.

Where exposure permitted, including key localities such as type sections, beds were extensively
sampled; at remoter, less well-exposed sites fewer samples were recovered. In all, 250 collections,
totalling 1-5 metric tons of rock, were made from the twelve named sections and subjected to
thorough analysis employing the methods described below. All identifiable material was assigned.
to its respective taxon and counted; in all some 25 000 individual specimens were examined, in
addition to numerous fragmentary remains. The resultant tabulated data are presented here as
the basis from which further extrapolation and inference is derived. A limited quantity of bio-
metric data is presented in conjunction with the taxonomic descriptions; additional information
is given in my unpublished thesis (Lockley 1977). Regional cleavage caused the deformation of a
large proportion of the material, prevented some accurate measurement and in some cases even
identification.

In the course of mapping the Llanuwchllyn to Llanymawddwy area (Lockley 1980) a few
important specimens were recovered from various localities other than those shown in Fig. 1;
these, together with new material from the main sections, are described in the taxonomic part of
this paper. Sample numbers not prefixed with a consonant (or consonants) correspond to map
locality numbers on the author’s field maps (1 : 10 000) and generally refer to isolated outcrops or
poorly exposed parts of minor sections; these numbers fall between 1 and 1162, and in certain
cases have a suffix (e.g. A or B).

Full taxonomic descriptions are confined to the new Brachiopoda from this region and to a
single new form from the Berwyns. Other figured material, including representatives of the
Trilobita and Mollusca, is distinguished either by being previously unknown or undescribed in the
Caradoc succession of this area or in having a particular stratigraphical significance (e.g. a
considerably extended stratigraphical range). Numbers BB92200-BB92264 are used for the
Brachiopoda, It.14294-It.14309 for the Trilobita, PL 4440-PL 4451 for the Bivalvia and PG 5022,
C 81324, E 53698 and E 67750 for four other miscellaneous specimens; all other material is
housed in the National Museum of Wales under accession number NMW 79.5G.

Conclusions are confined to correlative comparisons between the faunal assemblages and
associations described here and the known contemporary faunas and named communities
elsewhere. Brief comments on the definition of contemporary faunal associations (and com-
munities) are also included where appropriate.

Stratigraphy

The stratigraphy of the Lower Bala Group was outlined by Bassett et al. (1966), who recognized
four formations within the group in the Bala area (Fig. 1). Recognition, and indeed definition, of
these formations is considerably facilitated by the occurrence of distinctive ash members within
the succession. In contrast, however, the monotony of the entirely argillaceous Hengae Group
described by Pugh (1923, 1928) in the Corris and Dinas Mawddwy areas is broken only by the
black graptolitic shales of the Nod Glas Formation which represent the uppermost 20 m of the
group.

Although Pugh (1929) attempted detailed correlations between the Bala and Dinas successions
in the intervening Llanuwchllyn to Llanymawddwy area, his efforts have recently been shown
to be at best imprecise and at worst quite inaccurate (Lockley 1977, 1980). Stratigraphical
revisions of the succession in the Llanuwchllyn to Llanymawddwy area have not only been

168 M. G. LOCKLEY

LLANUWCHLLYN LLANYMAWDDWY DINAS MAWDDWY

PONT Y CEUNANT FORMATION
ASH MEMBER

COWARCH
PHOSPHATE BED (P)
see inset

aie
Ul
|
|
|
i}

FRONDDERW ASH MEMBER et ee cb iiscatie a hana
ee : ; " "LLAETHNANT .
ey ee Ee oe oe a FORMATION — ,

GLYN GOWER FORMATION - 2 eal coe
Let idly A a CEISWYN

|
. : 1 .
SS 2 oS eS agi: aw MUDSTONE
CEFN GWYN ASH MEMBER

FORMATION
|

BAL A - LLANUWCHLLYN-LLANYMAWDDWY - DINAS
CORRIS
—2SHALE

~Gs)_O)

DYFI MUD Si

GROUP

uN
!
it
|
|
]
|
|
|
|
|
|
]
|
|
|
i}
1
|
|

NANT. HIR
FORMATION

' LLANYMAWDDWY

4
QIceEN GWYN ASH

LLAETHNAN
FORMATION Jz!
HENGAE

LOWER BALA GROUP

INANT HIR|GLYNGOWER|
FM MATION

DINAS MAWODWY

EISWYN MUDSTONE FORMATION

Cc

JOERFEL Ls 2
ARAN VOLCANIC GROUP

1
|
|
|
|
|
|
|
|

x ARAN VOLCANIC GROUP x

Fig. 1 Scale representation of sampled sections (solid vertical bars I-XIJ) in the fossiliferous upper
part of the Lower Bala and Hengae Groups between Bala and Dinas Mawddwy. I, Gelli-grin type
section. II, Maes-Meillion section. III, Craig y Gath. IV, Lledwyn Bach. V, Ty nant. VI, Beudy
Isaf. VII, Nant Tan y Bwlch. VIII, Afon Twrch. IX, Craig Ty nant section at Rhiw March. X,
Pistyll Gwyn. XI, Y Ceunant. XII, Aber Cowarch. Insets (left, centre and right) respectively show
regional setting, local stratigraphy and local geographical and geological setting.

effected by the discovery of outcrops representing distal extensions of the Frondderw Ash but
have also been considerably facilitated by detailed examination of the faunal succession in
sections throughout the area. The stratigraphical revisions (Lockley 1980) are summarized in Fig. |
(inset).

The Llaethnant Formation is proposed as a name for a thick group of alternating mudstones
and siltstones which, in addition to representing a different facies from the stratigraphically
thinner Glyn Gower Formation, is not necessarily the chronostratigraphical equivalent of that
unit.

The base of the Allt Ddu Formation, defined, following Bassett et a/. (1966) as the base of the
Frondderw Ash member, is accurately located as far south as Rhiw March in the Dyfi Valley.
Similarly the base of the Gelli-grin Formation is accurately located as far south as this locality.
The greater part of the Nod Glas Formation (between Aber Cowarch and Rhiw March) is
shown to be equivalent to the upper part of the Gelli-grin Formation and to include the distal
portions of the Cymerig Limestone member which becomes discontinuous towards the south.

CARADOC FAUNAL ASSOCIATIONS 169

Distinctive lateral changes in the Nod Glas facies in the area in question have resulted in the
subdivision of the formation into two members which both represent facies of a diachronous
nature. The Dyfi Mudstone is dominated by a distinctive Sericoidea-dominated fauna and the
Corris Shale is characterized by being generally unfossiliferous except for local graptolitic
assemblages.

Aims, sampling procedures and data analysis

The primary aim of this study has been to name and describe quantitatively the Caradoc faunal
associations and assemblages found between Bala and Dinas Mawddwy, and to compare these
with contemporary associations outlined by Williams (1973) and described by Pickerill &
Brenchley (1979). Thorough sampling provides a census of the fossil faunas, whether representa-
tive of various biocoenoses or thanatocoenoses, which furnishes data from which recurrent
combinations of taxa may be noted. Where similar combinations of taxa show chronological
(stratigraphical) and geographical persistence in like facies (which lack current-produced sedi-
mentary structures) they are referred to as ‘associations’ and considered to resemble disturbed
neighbourhood assemblages (sensu Scott 1974) or in situ communities. Such ‘associations’
predominate in the area considered here. Where clearly transported shelly deposits are noted
they are categorized as ‘assemblages’.

Collecting techniques

Bulk samples of rock (mean weight 6 kg) were collected at vertical intervals varying from 0:5 m
to >3 m, depending on the extent of the section and the faunal content. It was found to become
increasingly impracticable to collect large, closely-spaced samples as the succession got progres-
sively less fossiliferous, as in some parts at lower horizons of the Caradoc of this area, so discretion
in the choice of sample interval was exercised. However, in general it was necessary to collect a
larger sample in poorly fossiliferous rocks to recover a representative number of specimens.
In the six most thoroughly sampled sections (i.e. numbers I, II, III, VII, VIII and IX of Fig. 1)
a mean sample interval of 2-75 m was employed for a total of over 500 m of section. Locally, as
in the perfectly exposed fossiliferous upper 135 m of the Rhiw March section (IX), a smaller
mean interval was employed, in this case 1-75 m.

To eliminate preferential collecting biases all samples were ‘broken up’ on a rock crushing
machine in such a way as to ensure that all the fossils or fossil fragments were retained for identifi-
cation purposes. All rock fragments were reduced to a size of 1-2 cm® before being discarded if
unfossiliferous. With the crushing machine (a converted fly press) it was easy to reduce fossili-
ferous rocks to a series of chips only fractionally larger than the individual fossils themselves
which, in this case, generally approximated to the above chip size. All extracted identifiable
specimens were then examined under a binocular microscope and counted, totals for each taxon
being considered to represent random samples of residual fossil species populations.

Faunal densities

Essentially only two types of faunal density may be calculated, density per unit volume of rock
or density per unit area of bedding plane. Neither method is entirely adequate for the analysis
of more than a limited variety of fossiliferous lithologies. Volumetric densities are more suited
for the expression of faunal densities in homogeneous (or isotropic), poorly-bedded strata,
whereas areal densities better describe the density of fossils on bedding planes in strata where the
rock fabric is essentially anisotropic. Both methods are size dependent; i.e. the mean size of the
fossils controls the mean number of specimens per unit area. The calculation of areal and
volumetric densities produces more accurate results when applied to relatively low density
assemblages and associations; it becomes more difficult to derive accurate measurements from
high-density deposits such as shell beds. The practical problems of accurate counting increase
with the density of specimens in the rock.

170 M. G. LOCKLEY

Not only are contemporary sections of varying lithology likely to require different methods of
density measurement, but within the same section it may be necessary to apply alternative
methods of density evaluation at successive horizons. In the present study it was primarily
volumetric densities which were calculated, although a few areal density measurements were
obtained from well-defined bedding planes (Fig. 2). The volumetric densities calculated for the
two sections shown in Fig. 2 are a useful guide to the faunal content (i.e. density of fossils) in
the sequence at successive horizons. For example, the ‘faunal depletion’ zone described by
Bassett et al. (1966 : 236) is graphically illustrated and shows a marked contrast to the fossili-
ferous horizons of the overlying Gelli-grin Formation, which include the dense accumulation of
small Sericoidea valves indicated.

CRAIG Y GATH RHIW MARCH
at ee eee pester chee
NOD GLAS ren ConcENTRATION OF
SERICQIDEA VALVES
GELLI-GRIN FORMATION
=}
al} = | 7/
S39 Z 5 a 5 ,
ALLT DDU FORMATION Fig. 2 Density of fossils per kg in the Craig y
Gath and Rhiw March sections. Box (right)
contains counts of specimens per 10cm? on
bedding planes in the latter section.
Faunal )
Depletion = 7
Zone
1
! eae
0 100 0 100
! Specimens per kg
ie
i)

The use of volumetric density measurements in this study was preferred not only because of the
disposition of fossils in the rock (i.e. generally scattered throughout rather than concentrated on
easily sampled bedding planes) but also because of the strong regional cleavage inclined at a high
angle to bedding, which almost invariably has the effect of breaking up bedding planes to the
extent that a complete sample of more than 100 cm? is hard to obtain; the mean area of the four
bedding surface samples shown in Fig. 2 is only 25 cm®. The single most significant effect of
cleavage on sampling procedure in the area under study is that rock splits so as to produce a
sample representing a vertical range of strata extending at least 15 cm above and below any given
horizon. In effect the vertical range of a sample approximates to 0-3 m.

Volumetric and areal density measurements may be compared and shown to have consistent
relationships. Theoretical considerations and empirical observations support the validity of
relating the two methods of evaluation. It is easily shown that 1 kg of rock (400 cm? at a density

CARADOC FAUNAL ASSOCIATIONS 171

of 2:5 g/cm?), if broken into cubic chips between 2 cm® and | cm, will expose a surface area of
between 600 cm? and 1200 cm? of which one third of the area (200-400 cm?) will represent surface
area in the horizontal (bedding) plane. Empirical observations by the author and Dr J. M. Hurst
(personal communication 1977) have consistently shown that, when measuring the surface area
of exposed bedding plane during the ‘breaking up’ of weighed samples, about 400 cm? of fresh
surface is exposed for each kilogram of processed rock. When processing homogeneous, poorly-
bedded lithologies it is generally found that the rock can be reduced to a smaller chip size thereby
exposing up to double the surface area. For example, Hurst & Hewitt (1977 : 154) equated
2500 cm? with a 3-5 kg sample (i.e. > 700 cm? per kg).

Although faunal density is a useful parameter which aids in the description of fossiliferous
successions, values should be used with caution; for comparative purposes, only like or similar
facies should be compared. It should also be noted that fossil remains at any given horizon not
only represent the remains of the ‘standing crop’ or the living (biological) populations which
inhabited that surface but also represent dead assemblages representing contemporary and
earlier generations. However, since palaeoecologists can rarely distinguish between these cate-
gories effectively, density estimates refer simply to fossils in the rock and are not in any way
precisely indicative of original population structures.

Identification and counting

All the fossils extracted from the rock by the ‘crushing’ process were identified and counted. The
counting procedures outlined below are aimed at assessing the number of individual organisms
in any sample. Different counting methods are required for the various fossil groups under
consideration and produce varying degrees of accuracy. However, consistent methods are used
for counting all representatives of any given group.

The Brachiopoda represent the most diverse and abundantly represented phylum encountered
in this study. The number of individuals (N) per sample was estimated using the formula N=
A+4I+P (if P>B) or A+431+B (if B>P), where A, P, B and I represent the number of arti-
culated pairs of valves, pedicle valves, brachial valves and indeterminate valves, respectively.
The same method (with right and left being substituted for pedicle and brachial) was used to
count the Bivalvia. This method has been used by other workers including Hurst (1975) and
Watkins (1979). The numbers of individual gastropods, cephalopods, macheridians, tentaculitids
and graptolites were counted singly.

Estimates of individual numbers in the five above-mentioned groups are far more accurate than
for the other groups considered here. In assessing numbers of bryozoa it was assumed that a
single complete colony could be regarded as an individual ‘unit’. Since fragmentary remains do
not generally outnumber complete or relatively complete specimens they too were regarded as
each representative of a complete individual colony. Bryozoans were therefore also counted on a
one to one basis; a significant overestimate of the abundance of colonies is considered no more
probable than the likelihood of biased quantification in assessment of the numbers of other
groups also to some extent represented by fragmentary remains.

The number of individuals represented by a collection of arthropod fossils is problematical
since individuals may shed their exoskeleton many times during ontogeny (ecdysis). Ostracods
almost invariably moult seven times before their ontogeny culminates in the eighth, maturation
moult. Similar ontogenetic patterns are recorded in modern and fossil ostracods (Anderson
1964). In contrast, however, existing evidence on the ontogeny of various trilobite groups suggests
considerable variation in the number of moults produced, for example between representatives
of the Agnostina (Hunt 1967) and the Olenellinae (Raw 1927) or the Olenidae (Palmer 1957,
Cisne 1973). Estimates of the number of instars produced during ontogeny range from 9 in the
case of the agnostids to about 29 for the olenellids and olenids, with the majority considered
representative of the adult (holaspid) stage. Variations in trilobite and ostracod moult patterns
are reviewed in greater detail elsewhere (Lockley 1977).

A consideration of arthropod ecdysis favours the conclusion that a given number of trilobite
or ostracod exoskeletons in any fossil assemblage is likely to represent fewer individual organisms.

172 M. G. LOCKLEY

For example, an ostracod could theoretically produce up to eight pairs of valves (albeit of
differing sizes) whilst a trilobite could produce many exoskeletal moults, with up to 50% repre-
senting adult instars showing little or no significant size differences. With our present incomplete
understanding of trilobile ecdysis, in addition to known variation amongst groups and complicat-
ing factors such as sexual dimorphism and fragmentation associated with ecdysis, any correction
factor used to avoid overestimation of numbers is highly arbitrary. Nevertheless, various authors
have estimated numbers of trilobites either without using any correction factor (e.g. Bayer 1967,
Hurst 1975) or by dividing a total number of exoskeleton remains by a correction factor such as
10 (e.g. Pickerill 1974). This later example echoes the suggestion of Harrington et al. (1959 : 111)
that less than 10% of trilobite remains are likely to belong to dead individuals. Either method
tends to produce estimates approaching theoretical extreme values.

In this study correction factors of 8 and 4 were used for the Ostracoda and Trilobita respec-
tively. Since all ostracod valves were counted without making a distinction between left and right,
it was assumed that, ignoring the first instar (the egg), an individual could be represented by
between 0 and 16 valves. A correction factor of 8 was therefore chosen as a mean estimate between
theoretical minima and maxima. Similarly, it was assumed that if none of the trilobites en-
countered in this study were represented by more than eight holaspid instars then a mean correc-
tion factor would approximate to four. Corrected counts were therefore derived by estimating
one individual for every four pygidia or cephala depending on which fragment was most numerous.
Throughout the study fragments of pygidia or cephala for all species were generally found to be
complete and representative of the holaspid stage. It could therefore be argued that the absence
of a complete series of instars would call for a relatively low correction factor (i.e. less than 10)
to avoid excessive underestimation. However, since the correction factors are acknowledged as
arbitrary it is emphasized that the use of a correction constant does not obscure the original data.

In this study most of the crinoid remains recovered were fragmentary; it was therefore impos-
sible to assess a representative number of individuals. Counts made of all ossicles and stem
fragments are presented in the data tables where the presence of crinoid material is otherwise
indicated simply by the addition of one to the total of individuals per sample.

Data synthesis and presentation

Having assessed the number of individuals in each taxon, totals for each sample were calculated;
these totals accompany counts for each taxon and are presented in a series of tables which each
represent one of the sampled sections.

The taxonomic level of classification employed in these tables, although variable, is essentially
specific. In all but a few cases the generic names of Brachiopoda and Trilobita refer unequivocally
to members of single-species populations as defined by Williams (1963) and Whittington (1962-
68). Categories which do not necessarily represent only a single species population are as follows:
Inarticulata, dalmanellid indet., Stroph. indet. and Trilobita indet. The remaining fauna is
classified into the following series of generic or suprageneric groups: Mollusca or Bivalvia,
Gastropoda (or Sinuites and Cyclonema), Cephalopoda, Macheridia, Monoplacophora, Tenta-
culites, Ostracoda (or Tallinnella and Primitia), Bryozoa indet. (or ramose bryozoa, ‘Prasopora’
and cateniform bryozoa) and Crinoidea. These classifications generally represent one morpho-
type or morphospecies, although in certain cases categories such as Mollusca are used either to
group minimal numbers of representatives of different molluscan taxa or to indicate uncertainty
about the taxonomic affinities of particular specimens.

Where known species or genera are grouped within a broader classification (e.g. the Bivalvia,
in Fig. 8, p. 179, are represented by Modiolopsis, Cuneamya, Cyrtodonta and others), full details
are given at the appropriate point in the text.

All tabulated figures refer to original counts of specimens except in the case of the trilobites
where numbers represent 25% of the maximum number of either pygidia or cephala following the
rationalizations given above. Although ostracod numbers are assessed by a similar arbitrary
method the original counts of valves are given (in brackets). Similarly, for Crinoidea counts of
ossicles (in brackets) and stems [in square brackets] are also presented.

CARADOC FAUNAL ASSOCIATIONS 17/3)

The totals of individuals per sample are tabulated and used as a basis for estimating the relative
abundance (°%) of the taxa at each horizon. Graphic representations of these distributions are
presented in Figs 5 (p. 177), 11 (p. 182) and 16 (p. 189) where the percentage of taxa from all
samples with over 20 individuals is plotted; those with less than 50 individuals are indicated by
a dot to the left of the columns. A sample with at least 50 individuals is considered to reflect the
composition (%) of the fauna at a given horizon adequately (Watkins 1975), whilst samples with
20-50 individuals, which characterize many less fossiliferous horizons, are invariably found
to give consistent percentage values when compared with larger (>50) samples. Although
arbitrarily chosen, these minimum sample size figures ensure at least a consistent and minimum
level of statistical constraint on data used for further extrapolation.

The size (weight) of each sample is indicated on most tables and can be used to assess faunal
densities.

The total number of taxa in any sample is used as an estimate of the faunal diversity (or species
richness) at any horizon. The diversity values presented here cannot necessarily be calculated
from the tables since, in certain cases, the suprageneric taxonomic categories represent two or
more taxa. In addition to the gross species richness (taxa per sample) presented here corrected,
size-standardized diversity graphs are also given using the Margalef (1958) method and the
Sanders (1968) rarefaction technique.

Trace fossils

Trace fossils in the Bala to Dinas Caradoc succession are neither common nor varied; for this
reason they are not described in detail. However, a few examples are noted (Lockley 1977, 1980);
Pickerill (1977) has outlined contemporary Caradoc trace faunas from the Berwyn region.

The faunal succession

The sections (numbered I-XII according to geographical location in Fig. 1, p. 168) are described
here in a different sequence based on relative stratigraphical position. The description of these
twelve sections is prefaced by the presentation of some additional information on the faunal
composition of the Derfel Limestone, the basal member of the Lower Bala Group.

The Derfel Limestone

A large (10 kg) sample from the fossiliferous shelly shales representative of the Derfel Limestone
at the type locality (SH 850395) was subjected to routine processing and analysis, to compare the
composition of the fauna with that found in the younger Gelli-grin Formation. Since Williams
(1973) referred faunas from both stratigraphical levels to the Nicolella association, after having
described the older fauna as a Nicolella—Kullervo—Palaeostrophomena association (Williams in
Whittington & Williams 1955), both faunas were similarly analysed to establish points of com-
parison. The sample yielded specimens representative of at least 28 taxa, which are listed below in
ranked order of abundance, with corrected numbers of individuals in brackets. Ramose bryo-
zoans (35), Dolerorthis (28), Platystrophia (15), Anisopleurella* (13), Oxoplecia (10), Nicolella (9),
Salopia (9), Plectambonitacea indet. including Sericoidea and three other* listed genera (9),
dendroid graptolite (9), Onniella (6), Leptestiina* (4), prasoporid (4). Howellites, Kullervo,
Eoplectodonta*, Leptaena, Deacybele ? sp., Broeggerolithus (all 2).? Lingulella, Palaeostrophomena,
Cyrtonotella, Brachiopoda indet., Platylichas, Ostracoda, Macheridia, Crinoidea, Cystoidea,
spicules (all 1). A total of 173 individuals is estimated from the above list.

Cyphoproetus is recorded in the Ordovician of Wales for the first time; similarly the occurrence
of Lingulella is the first record of a representative of the Inarticulata in this member (Lockley
1980). The relative abundance of the brachiopods Dolerorthis and Nicolella compare with
abundances noted in parts of the Gelli-grin Formation, whilst the occurrence of Platystrophia,
Anisopleurella, Oxoplecia, Salopia, Onniella, Leptestiina, Sericoidea, Eoplectodonta and Palaeo-
strophomena at both horizons is also noteworthy.

174 M. G. LOCKLEY
AN

o
@
|
ac
w
[e)

LAMINATED
SILTSTONE BEDS

MUDSTONE

— Frondderw
—H24 Ash

Peas -H9 Wee
-H42 _48
-H41 hg
nly --H40 “He
-H39 =
sae
-H2
0
[L -H38 —-H1
Hale tee een a Ran A AN nV4O) ~~
an an an—H38
H44 259
951
VAN
Eager
ee
-20m peers eSHiss
—H34
+10m

Shes
Lo is
— H31 952 — MAP LOCALITES.

Fig. 3 Sample points in the Afon Twrch section with inset map showing outcrop of sampled beds
associated with shell beds and the Frondderw Ash in the upper reaches of Afon Twrch.

CARADOC FAUNAL ASSOCIATIONS 175

The Nant Hir Mudstones, Glyn Gower Siltstones and equivalent deposits

Although Bassett et al. (1966 : 229-230) listed faunas from various horizons in the Nant Hir
Formation, no shelly fauna is known from the Nant Hir or the equivalent part of the Ceiswyn
Mudstone at any locality south of northing 315. Similarly a large part of the overlying Glyn
Gower Formation and equivalent beds to the south are only sparsely fossiliferous, with a low
diversity fauna typified by the forms listed by Bassett et a/. (1966 : 231). For this reason ambitious
sampling schemes were not applied to sections through this part of the succession. However,
near the headwaters of Afon Twrch an important series of fossiliferous beds is exposed in a
section which contains a newly-discovered outcrop of the Frondderw Ash Member. The fauna,
although dominated by Heterorthis and Sowerbyella, also contains elements previously con-
sidered as representative of the Nicolella association.

The Afon Twrch section and equivalent beds

The horizons sampled in this section are representative of the upper part of the Llaethnant
Siltstone Formation (equivalent to the upper part of the Glyn Gower Formation) and the lower
part of the Allt-Ddu Formation. Figs 3, 4 and 5 show respectively the field location and sample
grid, faunal distribution and relative abundance patterns relating to this part of the succession.
A series of 44 samples (total weight 232 kg) was collected from riverside exposures between map
locs 950 (grid ref. 9133 2330) and 958 (9101 2305). However, in this section faulting (Lockley
1977, 1980) has caused duplication of part of the faunal succession (Figs 3, 5). Outcrops of
Heterorthis-dominated shell beds are found in the repeated parts of the sequence and, since both
outcrops are sampled at close vertical intervals, samples representative of beds thought to be
precisely equivalent are bracketed together in Fig. 4.

There is a dual significance in the distribution of taxa recorded in the Twrch section. Firstly,
the occurrence of Heterorthis (in abundance) at an horizon some 40 m below the Frondderw
Member contrasts with the occurrence, elsewhere to the north, of Heterorthis assemblages in the
Lower part of the Allt-Ddu Formation (see Bassett et al. 1966 : 234 and Fig. 6, p. 178).

A second noteworthy aspect of the distributions shown here is the occurrence of certain genera
hitherto thought to be confined to the Gelli-grin Formation and the Derfel Limestone Member.
These include Orthisocrania, Nicolella, Onniella, Salopia and Chasmops. Dolerorthis is also
recorded; although known from the Upper Allt Ddu Formation it is otherwise confined to the
Derfel Limestone and the Gelli-grin Formation. All these forms are characteristic of the Nicolella
association and although, with the exception of Onniella, they are rare at this horizon, their
occurrence can be considered indicative of the sustained establishment of this type of association,
in this general area, through the Lower Caradoc. With the exception of Dolerorthis and Chasmops,
representatives of the above-mentioned genera are figured in the taxonomic section.

When traced southwards to a gully on the north side of the Dyfi Valley (map and sample
loc. 202, grid ref. 896217) beds equivalent to the Frondderw Member and the underlying shell
beds are known (Lockley 1977, 1980). Fig. 5, which outlines the faunal distributions recorded at
horizons in this part of the succession, indicates a significant lateral change in the composition
of the shell bed beneath the Frondderw Ash. There is no evidence for the presence of Heterorthis
and only the occurrence of Onniella, Bicuspina and the association of a minor ash are comparable
with characteristics of the Twrch section. Since these shell beds are associated with thin parallel
laminated, storm-generated siltstone sheets they are considered to represent transported material.

It is of particular interest to note that Sowerbyella and Heterorthis are almost entirely mutually
exclusive though the two forms occur in abundance in beds separated by only a fraction of a metre
(Fig. 5). J. M. Hurst (personal communication, 1978) has noted some degree of segregation
between these two genera in the Alternata Limestone of Shropshire. It is possible that such
patterns represent a differential response to the effects of transportation. Both forms have
atrophied pedicles and would have therefore been relatively susceptible to disturbance by currents.

The Beudy Isaf and Ty-nant sections
Uniike the relatively remote Twrch and Dyfi sections, the Beudy Isaf and Ty-nant sections are

176

M. G. LOCKLEY

9

v

s 9
2 a a] oO} 5 =| a] @
o = yw =| =| ==}
£ 4 = o|
=} gate sg qi) os 4 > Al S| gl) 2
rv) z ao = S| Ss} S| = oo oof Of cl cc
ss a aw s a ol OS
iS, o BS 3 g 2 8 ss] Sl J
ag o S| 6 oO s] a} 5] S| 8 ao g 8
& al als = cx] zi wl a =|) 3 te

9
$ nN
z g 48
=| g Sy SG
° os Tye
e 48 = g
a] 3 5
a OS as en = v
an@ > Bore 55
ee oe
oom aren acne ‘=
agomO OU ae re Vv

1 (35) —
(234)[ 5]

|
o
3

AIG 52 SE Sta ops 2578 5 oa we tipo pes epe ae

3 1 (6) 1 (20)(3]

2 (2) hk CO) =

— 1 5 — — 1 (90)(4]

a ee 9) IS

pee ee ee — rE eu) US)

ee, Ea ee Tet ie SDC
en ee 1 (3) —
Oe UP) es
= 9 Je SS] Hae Se SS |] SS SSS SS 1 (20) —
ee So = FF HLA

—-=—---—-=-= 1 (27) —

as CA al hb) a) (2) —

1 6 — 3 (25)1 (1) (1)

ee Seen 1 4 7 2 (17) 1 (238)[5)

3°95 & 1 (5) 1 (79)—

1912 13 1 (7) 1 (47) —

4 16 10 1 (6) 1 (27) —

— 5 10 1 (10) 1 (31) —

— — 13127— 61—-—-— 1— 12—2 4 2 1 (4) 1 (40) —
—— 4 167 10 -_ = — 2, 2 5 6 FW (1) 1001) —
204" 106463) — —F 22] 8) 1) 1 US)=
— $) (3 IS I= = 1 1 AS ey
ASN = SI aS 1 SS SS YD ED) 1 (A) =
S91 35 GOS i 2 2h Sh Sree ae
SS] OUP Saye ee SS 1 PS Haeae AP 1) 1 Mia
Pe Sy Ta te a a
6.62, 4. =) 1024 ee 2 nS) ately
eh Se eth eS ee eee Bee Ect)
3G 1 6 — 9 226 28
ee ee) — ee | ee
=ss=Woscre Moa} sa fsesaafas =a 1 WH
Sa ea i Wile tee Me 16h Ut CG NCa ONT
127965 S 112 45——— 1 1 3 1 — — 6 — 1 (10) 1 (73)[2)
SSeS SS. ee eS SS eS) ESOT)
=, Sa 2 i 2 TG)
33. — Po} oe] = = & 24) OT
16 oes a eee Meee Ses Ce) a =
2- 3-—- —--—- —- —- |= | —- — —- —- -—- = 4 — — 1 (5) (4)
Je Ge eae esoacsa Vea ese 2 2 =«1 (1) 1 (32)(3)
Na SaaS Sea" AS) SS SU I

Fig. 4 Fauna from the Afon Twrch section.

CARADOC FAUNAL ASSOCIATIONS

oe a

(lege ee ei
AINA
1 I NARTICULATA
2 DINORTHIS
3) DALMANELLIDAE INDET. NE
4 HOWELLITES
5 ONNIELLA a ll ial ip Eee |
6 REUSCHELLA
7 HETERORTHIS
8 BICUSPINA
9 SOWERBYELLA
10 BSROEGGEROLITHUS
11 BRONGNIARTELLA
12 RAMOSE BRYOZOA
13 PRASOPORID
14 OSTRACODA

FRONDDERW ASH peee:

MMT at

= settee ver MINOR ASH
r LP ’ 202-|an]| | > > b> p p HORIZON

Leave
so
Ye
ae

| | GAP L0m
>>

WEL IL iteL Ltt

G ly tay bet

Fig. 5 Sample points in Afon Twrch section (NE) and the lower part of the Dyfi Valley section
(SW) below lowest upper Allt Ddu sample (R086). Dotted lines indicated inferred correlations.

accessible by road. Fig. 6B shows the composition of faunas collected from six stratigraphically
sequential horizons in the small stream gully east of Beudy Isaf (9105 2495). The section covers
some 20 m of beds, passing upward from the gully to fossiliferous roadside exposures (9115 2505)
containing numerous Heterorthis specimens. This section represents a part of the lower to middle
Allt Ddu Mudstone (Fig. 1, p. 168) and is not therefore contemporary with the deposit found in
the upper reaches of Afon Twrch. Sowerbyella and Heterorthis are again found to be mutually
exclusive.

The Ty-nant section (Fig. 6A) is represented by a series of widely-spaced samples recovered
from the Nant Bwlch-y-pawl valley east of Ty-nant (9050 2625). The Frondderw Ash exposed

178 M. G. LOCKLEY

ps
Ky] ho SOs So Dalmanellidae
Reuschella
Heterorthis
= 1 1 | Bicuspina
Sowerbyella
ggerolithus
Crinoidea

= = = | Broe
1 4 — | nuculid

1 4 4 | Gastropoda

nm =» = 1 | Bryozoa

Aya Sn Tf

HM} § § 4 | — | Paracranio

Fig. 6 Fauna from the Ty Nant section (A) and
the Beudy Isaf section (B).

under the bridge in the farmyard succeeds mudstones with Sowerbyella shell beds and is overlain
by massive siltstones with a sparse fauna. Outcrops higher in the succession at streamside
locs 400 m and 700m east of the farm yield a Heterorthis-dominated fauna similar to that
recorded in the upper part of the Beudy Isaf section. Loc. 35 (9167 2665) yielded a typical middle
to upper Allt Ddu fauna characterized by the presence of graptolites and a nuculid bivalve
(Fig. 96, p. 225). Immediately to the north of this locality and higher in the succession on the
ridge known as Pen-y-Cefn-Coch two outcrops of upper Gelli-grin beds were discovered. These
yielded a fauna dominated by bryozoa, Leptestiina, Dolerorthis and Skenidioides and containing
characteristic elements like Estoniops, Nicolella and Rhactorthis (Fig. 7A). Outcrops at loc. 615
(9150 2685) consist mainly of well-bedded limestones typical of the Cymerig Member, whereas at
loc. 615A (9170 2700) a coarse tuffaceous, calcarenite lithology is indicative of beds immediately
above the limestone member. The recognition of precise stratigraphical relationships is hampered
by poor exposure and local faulting (Lockley 1977, 1980).

lymene

Flexica.
neni,

Paracraniops
Dolerorthis
Skenidioides
pina
lectodonta
testiina
jtaena
isotelinid
iartella
Broeggerolithus
“entaculite
ramose bryozoa
prasoporid

Ostracoda
Crinoidea

Lep:

\ | Sericoidea
= | Platylichas

— | Estoniops

~ | Lep

1(10)(2)
1(110)(3)

- # | Nicolella

1 | Rhactorthis
Howellites

= \© | dalmanellid, indet-

1 — | Reuschella

1 4 | Bicus:

1 4 | =9P

n©

1 1 | Bro

1 4 | Cephalopoda

1 — | Conularida

nw

555B 366 aes ji o> = = 4(5)
555A 188 tL Seles Aktien) em 1(8)(2)
trcesresssss+(Stratigraphical gap of 35 m):-'-: . : : Pytko 0x4

Im 2a iS 3) af SS - 1(30)

Fig. 7 Fauna from Pen y Cefn Coch (A) and Lledwyn Bach (B). (For Conularida read Macheridia.)

The Allt Ddu Formation at Craig y Gath and Rhiw March

Figs 8 and 9 show the distribution of faunas in the upper part of the Allt Ddu Formation at
Craig y Gath and Rhiw March respectively and Fig. 11 shows percentage abundances.

Bassett et al. (1966 : 235) stated that the Allt Ddu succession ‘is best seen at Craig y Gath
(915306) and pointed out that, in addition to considerable repetition caused by faulting at the
type locality, the junctions with neither the underlying nor with the overlying member are seen
in this area. For this reason Craig y Gath is regarded here as the alternative type section.

CARADOC FAUNAL ASSOCIATIONS 179

9 oo 5 8
vw = o| =| =z 0 oo

my 3 Z| a = °
ad 3 w 2 3] 42 eel cele SG 3 So 5 og &
| al a %@ wv =| > | Of v oo) ie 2
ce] 4) S) 5 =! re So ale Sls sl ble Rog oR SSS 8 °
| sc} US =} Go 2 Bo GS] o] el ow -S oS 5G a oo 2 r)
SUC Tel ie SM oS) ey ES Sil es NS Se ee 2 A i 8 ° a
So} Oo oe cs] 3 So} Se} a Oo eB! Uw gl SI CS $u 3 a Eces& < oO
o} S| =} S| oO | 0 S| rc) Ci OES ESlcling, 1S) a Ooh aon iM = sa
al cl wi of zi al wal wi wi = Oo al oaloal- © ao & 2 yo o rc)

1 (76) —
1 (263)—
1 (16)(1)

1) =
1 (34) —
(22) —
aps] 7| 35)1 -——6 —~—2—7 —~—112——-—-—— (5) —

(24)(1)

~u Fr Funny

Fig. 8 Fauna from the Craig y Gath section. (For Conularida read Macheridia.)

The sample grid (covering 150 m of the succession) and a locality map are shown in Fig. 10.
The two uppermost samples, recovered from the base of the Gelli-grin Formation, yield a
characteristic fauna dominated by Onniella and Eoplectodonta. However, the distribution of
upper Allt Ddu faunas, both here and at Rhiw March, is characterized by an association of
mainly long-ranging forms, all of which maintain a relatively consistent pattern of relative
abundance throughout the succession. The fauna is dominated by Howellites, Paracraniops,
Macrocoelia and Broeggerolithus.

180 M. G. LOCKLEY

“ o Guo
| Z|» g 85
C] zi} 0 M4 eo
oo 6c] = Sl c¢ o ° uv
a og 2 gy = =] 2 c c 6
a ¢ el He PEE EEE 5| vse = be o
a5 = ec e o & = = tl oo £| ac. 75 2 e
rH £] > vl] of ec] E] oe} > FH ce A ZS a Os iy D3 3
£aqs= 4 a ae} vy of Sl ay el = 2 = sf ovgQonte aS
3 C) 4 5 el oo el ae D2 . ST Sn pw es °
rr) -| 4 5} oy Ye} S| St =] =<] cg) 2 eo 0 C) oc F0Fo oe = ec
a] S ¢ zs) El gy] o oY o} © Oo] = = > @ ce >» = =
So0 ¢| 2 3) ol & oo] So} o Ss ee - © Oo Foo” .
™ al~o [o) al ol al x! Sl sl el lal Y ole @Oumgaooc ala 0 & a)
10
7—-—----- 2---e--1tedd ——-—-—-— —— 1 — 3 (25)1 (2) —
2 ee ee ec —- 1 —-—-—- — — — — — — 2 (16)—- —
5
Pomme e J SoS So SoS to 1 = SS aS SS SS SS BIS
ee ee SS Se — 819)050=
ts)

a Te OS SIS Ss So — — — — 15 — 26 — — $ (31) 1 (16)f1)
Sse OM) Pa 0S ene eH = 1 Wd Ao Saseogod dd 2 a= i Wd (Oi
oe ee 2——1 24 2 — — 4 (@29)1 62)13)
—d & a eee ms th II)

CBR ype AE oprah 5 Sth hy Dee art “Ny a eens MFI? (Coes

Se 2S SSeS Sy SSH 1 OY Oe

10 a 4] 1 2 1 (3) 1 (14)(3)

Se ae ee ee 124) =. Ca eS Eee ‘l (6b) S (0b) =

—— 33—- — — — — — — 11—1—2— — 1 2—— 8 — — — 1 (1) 1 (25)—

rs, MOEN IG) eet fey a i eee ef es a (
bone, enn | ew ey ON Bee ees 4) oe 1 AGRO SST
es) ee ee es A RG: a PS a) 2 ee an
ee) ae ee ae Es aoe ee = {te 1 (7)14)
a a has i ee ee 302 SSS ioe
67s Se ae Se a is sissy st
2h 39 = S52 oS = SS (8G. = 2 zene
a ee Ge eee et) ee ee SS Ose sd 0 Se

SBasaSSeIGSI9 SSeSafots acl) Ae) — 10) 5p ca

SAMetross FOS 1 I aoe 123) 40 2 7-38: 1 —, 3 ee A) costar
Sipe! shad eats alin a ee es ee ee ee ee ey Se OT CO),
Sj 62S 2] 9 10 S=2eI1=%4 1S) 258 FSS tt ww!) Ge

Vik Se Sd Sh eS re 1 S39 Sat Sk = 7 ti) 1 (Calg
eae A Al ST

S

et 1 te] 1 S24
Ao tot Sow Pow a a) Wd Gey
2

25 2 a ai a) (6 = 2 — = — i6' 10: 1 1 yeaa
a. Np ER Ea ae Doe. See ae eee

Fig. 9 Fauna from the Rhiw March section (Allt Ddu beds). (For Bellerophon read Sinuites, and for
Conularida read Macheridia.)

CARADOC FAUNAL ASSOCIATIONS 181

MAP BASED ON

0.S. Air photo no. 262
\ Film no.72 410
sc. :7500
FRONDDERW ASH —_ i AOze
AD X=

i CRAIG Y GATH

204m

MUDSTONE

Silty bands
& bioturbation

:| SILTSTONE

Fig. 10 Map of Craig y Gath exposures showing sample points and their stratigraphical relationships.

182 M. G. LOCKLEY

Paracr, Din. Howellites , B Reu, Bic, Sowerb, Ser, Mac.Lep, R. ge B'62 Bic. B!, B82 Ostracoda.6.

Os ® 22 ®owosF | “eGo t a (7)

}
. ret rf aS prtae

:

=
NJ
N

CRAIG Y GATH 0 100 % 10m
fran 5 |

ETE TE WU Te Te yp a

-

f
RHIW MARCH

Fig. 11 Percentage distribution of taxa in the upper part of the Allt Ddu Formation at Craig y Gath
and Rhiw March. Abbreviations refer to Paracraniops (Paracr), Dinorthis (Din), Dalmanella (D),
Onniella (QO), Reuschella (Reu), Bicuspina (Bic), Sowerbyella (Sowerb), Sericoidea (Ser), Macro-
coelia (Mac), Leptaena (Lep), Rostricellula (R), Broeggerolithus (Broeg), Brongniartella (B),
Sinuites (B*), Gastropoda (G?*), Bivalvia (Bi), Macheridia (C), ramose Bryozoa (B'), prasoporid
Bryozoa (B*), Graptoloidea (G). Lithostratigraphical boundary shown by dotted line.

CARADOC FAUNAL ASSOCIATIONS 183

Paracraniops (P. glaber sp. nov., Figs 33-36, p. 207) is an important element throughout the
Allt Ddu and is even known from upper Glyn Gower and Llaethnant horizons. This form was
previously unrecorded by workers in this area (e.g. Bassett et a/. 1966), who ignored it because of
uncertainty about its taxonomic affinity (A. Williams, personal communication, 1976). However,
in the Caradoc successions of the Berwyn Hills to the east, Pickerill & Brenchley (1979) have
described Paracraniops as an important element in the Howellites community. As demonstrated
below their observations allow useful comparisons to be made between the composition of
related faunal associations.

According to Bassett et al. (1966) and Whittington (1968) Flexicalymene was unknown in the
upper part of the Allt Ddu Formation, but it has now been found at several horizons (Figs 8, 9).

Differences in the upper Allt Ddu faunal succession between Craig y Gath and Rhiw March are
only pronounced with respect to particular taxa. For example, Dinorthis, which is confined to the
uppermost part of the Craig y Gath succession, is unknown at Rhiw March, and Sericoidea,
which is commonly found in the Rhiw March succession, is unknown at Craig y Gath and at
other Allt Ddu localities north of Ty-nant.

The assemblage zones outlined by Bassett et al. (1966 : 236) are only entirely valid for the
area they mapped (1966 : pl. 2). Although the lower, middle and upper Allt-Ddu assemblages
(represented respectively by the Heterorthis faunule, a zone of faunal depletion and a zone
characterized by ‘new’ stocks) can be recognized, where exposed, as far south as the Ty-nant
area — with the possible exception of the upper zone — it is impossible to recognize these divisions
within the formation in the Rhiw March section. For example, the zone of faunal depletion
noted at Craig y Gath is not characteristic of equivalent horizons in the Rhiw March section
(Fig. 2, p. 170). With respect to this zone it is of interest that these argillaceous beds, both at
Craig y Gath and other localities (e.g. loc. 35, Fig. 6), are characterized by the presence of grap-
tolites and small nuculoid bivalves.

Characteristic ‘bursts’ of Sowerbyella noted by Bancroft (1945) and Bassett et al. (1966 : 236)
are apparently widespread in space but, in the upper part of the formation, restricted in time to
only a few horizons. A Sowerbyella shell carpet or ‘burst’ occurs at about 8 m below the base of
the Gelli-grin Formation at Craig y Gath, and at about 16m below the base at Rhiw March
(samples ADU and R18). The likelihood of these two horizons, at localities 7 km apart, being
contemporaneous is strongly supported by the occurrence of an unusually high concentration of
archaeogastropods at an horizon exactly 14m below the Sowerbyella ‘burst’ in both sections
(samples ADO and R13). It is suggested that the widespread occurrence of abundant gastropod
and Sowerbyella specimens at particular horizons could be indicative of a large successful spat-
fall or conditions otherwise conducive to the temporary proliferation of such specific groups.

The uppermost 30 m of the Allt Ddu Formation at Rhiw March are apparently equivalent to
only 22 m of beds at Craig y Gath, where erosion before the deposition of the Gelli-grin could
have removed some 8 m of uppermost Allt Ddu deposits. Since respective lithologies for these
uppermost beds are well-bedded silty mudstones and fine-grained argillaceous mudstones it
is probable that reworking took place in the north while continuous deposition prevailed in the
south.

Sampled sections through the Gelli-grin Formation

Dominant faunal elements. Although only 60 m in thickness, the Gelli-grin Formation represents
the most lithologically varied, fossiliferous, and faunally diverse, stratigraphical unit in the Lower
Bala Group; for this reason it has been examined here in particular detail. The formation crops
out between Pont y Ceunant (SH 944346, about | km north of the type locality west of Gelli-grin
farm) and the Rhiw March section in the Dyfi Valley. To the south of this latter locality the
formation passes laterally into argillaceous beds representative of the upper part of the Ceiswyn
Mudstone and the overlying Nod Glas Formation. The distribution of faunas in five sections
through the formation is shown in Figs 7B, p. 178, and 12-15. The chosen sections are at the
following localities: west of Gelli-grin farm, 944340 (Fig. 12), Maes-Meillion, 925305 (Fig. 13,
p. 185), Lledwyn Bach, 912279 (Fig. 7B), Nant Tan y Bwich, 914240 (Fig. 14, p. 186) and the

184 M. G. LOCKLEY

: be} < 4% 9 a
; < 9 zg yc v ; 3 ¢ 3 ¥ > o
v0 w| 4 S Bl = -| 5 c) S| os U
Ee gd aaa2 3 a Jags 4 qgds Badass adye z= gi. esitag g
=| = | tT] © e@ 2 9) eo} | r) 0) fv coy | 7S > -=- ob D3 c
Pate as =| 3} S| 9 ads a BI b>} Sag z i] 9 2 a a]
SS s1 gs? ZT aT esas segs ges es PPssssege &
Ae EEE Ree EEE EEEELEELEE EEE Sezesiz
witsHdadadadsaddd S| el ol SI a Is] 8 5 wl ol al = YS ae 3 moor acer rr)
10
SSaMaodol ven tose = =e) soe ee aed eS Ht J e=== + 112 — — 1 (8)[1)
m
1- 380-11 1-21--—--—-3 -—-538—--—- 43 —1)/2 —~-4—-----=- =-=—-=-—= 8 4 — — 1 (22)(1)
°
—-—- 686 2=—==-==+-=5+-=-==- 2—-—-=---|/--------=- ---=-=-=-=-—-=-= 1 —(1)
Gap
Brn |= coigtis7a|—- — 4 8 —37 2 — — 4 4011 1 — — — —18 — — —~— — ich — th) Kid oe) 4 ——— 6 2 46 18 — — 1 (23)—
x—x—x
Gap ]
PER eyes SICH oo eS ea) esate ee oe Oy ee ES SO es SS ee Po=} sah YW i GH) 4D deal
x-x-
PSS eS hrAi = = = = = Jes aN Hossa sae 231 —--— 1 — 2/4 —_—=—=-=-+= tS SS] SS 5 = 13/1 (2) 1
GG 23 | (2) (5) 01)
| |
———|— 66if ||175|— = = — — Sern = Sa Sekhar il a Hes — ee Se a 3:25) 1 04) —
———
P= J— G6le | 66)— — 1 — — 162 — 4% — 7 462 = — 1 = 2 — — i = lee 2aesotds Sess ss qt 71 (ES) (Gb)
SESS SH! Ze vie = GS ae Bae ae Se SH Yee SSeS = i oy = ee a CS
22252 | GENER = Gl = fe = WS OS) BAotae= SS SS) —------- 1 — 2-1 —-—-— 4 1 —— 1 (M1) —
x—x-x |- 661b]223|1 1 1 = i) = 15 8 42 1 2333 — 53 1) 89 tod ==" 1 12 Poon 28 FC) WOH=
aS
x= x= |
x *x**x|_ GG6la|—| Pont-y—Ceunant no fauna H — —
Xx K Xx
xx x xx
— 661z| 1 = = = = = == = = 1 = Ses Sess =
|= Ceiba oe ee — = = a = 1 1 SSS s5Ss5=2255
TCA FE ae cS i A ——15 ——— 1 — 1 (1) 1 63)[6)
GELLI-GRIN OG LIMESTONE
%Xi| TUFFACEOUS SILTSTONE and SANDSTONE MUO STONE
ext] TUFFACECUS MUOSTONE = SILTSTONE
‘ a LAMINATED and HOMOGENEOUS
OOLITIC TUFFACEOUS SANDSTONE itya] SANDSTONE |

Fig. 12 Fauna from the Gelli-grin (type) section. (For Bellerophon read Sinuites, and for Conularida
read Macheridia.)

cliffs north of Rhiw March, 899219 (Fig. 15, p. 187). All sections, with the exception of the
poorly-exposed outcrops at Lledwyn Bach, have been sampled thoroughly throughout.

In the Tan y Bwlch and Rhiw March sections the uppermost 20 m of beds not only belong to the
Gelli-grin Formation, but also represent the northern part of the outcrop of the Nod Glas
Formation now known to be equivalent to this upper part of the Gelli-grin Formation (Lockley
1977, 1980). The relative abundance of faunas in the four main Gelli-grin sections is shown in
Jenga Iles yoy, Ite

At the type locality a series of 15 samples was collected. Here the underlying uppermost
Allt Ddu beds, where fossiliferous, contain a fauna dominated by Dinorthis, Howellites and
bivalves. The succeeding coarse Pont y Ceunant Ash is unfossiliferous at this locality but is in turn
overlain by highly fossiliferous, tuffaceous mudstones containing a diverse brachiopod-dominated
fauna. Various elements of these lowermost Gelli-grin faunas, in particular Eoplectodonta, are
found so closely packed that in a few instances they show primary growth distortions; such a
phenomenon indicates an in situ association. These fossiliferous beds are characterized by the
following rapidly changing succession of local associations; an association dominated by
Onniella, Eoplectodonta, Bicuspina and Reuschella is succeeded firstly by a Nicolella-dominated
fauna, then by a Skenidioides-dominated fauna. These horizons are in turn overlain by beds
containing Dalmanella, Leptestiina and Howellites. A species of the latter, H. antiquior (M’Coy),
occurs in particular abundance in the middle part of the Gelli-grin Formation and is associated
with Rhactorthis and an abundance of the trilobites Kloucekia, Broeggerolithus and, to a lesser
extent, Flexicalymene.

Although the middle part of the formation is not entirely exposed, examination of other
contemporary sections has revealed that the Howellites-dominated fauna persists into sub-
Cymerig beds. The change in facies associated with the onset of Cymerig Limestone deposition

CARADOC FAUNAL ASSOCIATIONS 185

Anisopleurella
Sericoidea
Leptaena
Macrocoella
Stroph. indet,
Kloucekia
Estoniops
ps
atylichas
exicalymene
ongniartella
lobita Ind.
isotelinid
Cephalopoda
Mollusca
tentaculites
prasoporid
ramose bry
Tatlinnella
Crinoidea
Cystoidea

S kenidiodes
F
Bi
Tr

Cremnorthis
Daimanelia
Platystrophia
Reuschella

Rhactorthis

Dolerorthis

Dinorthis

i F
Lingulida
Nicolella

Chasmo

1 (2) (65)_

1 (29)(32) 1
1 (4)

1 (8) 1 (8)(2)—

1 (6) 1 (2) — —

1 (1) 1 0)(6) —

1 (11) 1 (63)— —

219 14 — 58 - 1 - — 1 (4) 1 16)(2] —

31 — < 7 - 5 (42) 1 (100)(4) —

1 (3) 1 (6) 1) —

=SsS5=1 >=

MAES-—MEILLION

Fig. 13 Fauna from the Maes-Meillion section.

is accompanied by a corresponding change in the fauna. Cymerig and supra-Cymerig beds
contain a fauna dominated by Dolerorthis, Nicolella, Eoplectodonta and Skenidioides. Other
characteristic elements include Cremnorthis, Chasmops and Estoniops; the latter form is only
known from horizons above the limestone member.

At Maes Meillion the faunal succession is remarkably similar to that recorded at Gelli-grin
(Figs 12, 13). Notable differences include the composition of the uppermost Allt Ddu fauna
(dominated by Bicuspina, Reuschella and gastropods) and the occurrence of a fauna in the upper
part of the Pont-y-Ceunant Ash Member which is thinner and finer here than to the north.
The faunal succession throughout the formation mirrors the pattern observed in the Gelli-grin
section; all the dominant genera named above are found at equivalent horizons in this section
4 km to the south.

Three samples from newly-discovered Gelli-grin exposures on Lledwyn Bach have yielded a
fauna characteristic of the lower to upper middle part of the formation (i.e. from 5-10 m to
40-45 m above the base). The lowest, poorly-exposed beds contain an association characterized
by Eoplectodonta, Dalmanella, Howellites and Leptestiina and are succeeded, after a 35 m gap,
by larger outcrops yielding (from two sampled horizons | m apart) a fauna dominated by
Howellites and trilobites (Fig. 7B, p. 178).

Analysis of the Gelli-grin sequences in the Tan y Bwich and Rhiw March sections reveals
patterns of faunal succession in the lower and middle part of the formation which compare
closely with those noted in the sections to the north (Figs 14-16). However, the faunal succession
in the upper part of the formation differs from that recorded to the north and corresponds to a
distinctive lateral facies change occurring south of Pen y Cefn Coch. At Tan y Bwlch the Gelli-grin
Formation is only about 45 m in thickness and is mainly argillaceous throughout.

The lower part of the formation at Tan y Bwlch is dominated by Onniella and Eoplectodonta,

= 5 3
£ g e| 2 3 8
2 = os oY c 3 2] S| g 9 wee S| 3 os a Fo
2 Ja 20 2 2 rc) Sg = 3 925 2
Sgt trast ada F{addagssse § Ce 9
eer er ee cEeELce Se Peake =
tay rye ay 345 3 2328 ss shes qs g a Ses) Bubes, 75
28333 sq7 ss eee TFs sasysase Stssss «
f£ziadaarwdadnt | S| al al Sl zl di xl alclal ¥ ai eoooofre 5
eauees| LIMESTONE (RHIWLAS)
waa) is) Wise bd ene a hk Ae meh eet RE Eh ee (ood
Ea) |e an Se Meat eS oe ahi aan im. Tem LIMESTONE NODULES
Paes Se> Sess (CYMERIG.
Wh. HS == So GS eS ae = (eS peg tag ap 1?esoti@ IO Was
\SLSPLILET ET: +4) — 10-3) — 17 — = 2 = wo--—-24-—--—-—--—-— —= = = 2 — Zi
bedding A 2 DARK GRAPHITIC
‘sol— 5 — 8 2 —— —90— — —— wHI8 —— —— 451-1 —- =—=—3 2 3 —— 1 (IH —
sea Spstecs 27 ee pn ee, Sk a ne (4) a
GREY MUOSTONE
=== = = — (os SSS fea Gg 7 Sa a =——-—=— 1 31%) 1 01)
------- we- - —- -- -—--=-— 2-4-—-—-2 ~~~ 2 — 141 wii
SSS Se Se Yojse sea eases jas =s=) ------- Es
TUFFACEOUS
SHELLY MUDSTONE
H=0 1 6GH8BeeM i= 1 Se seoe= = Joo ee = 36 6 1 (19)
13962 1 — 2% 3 —14 157 4 9a = 1h = 4 — — 4 § 23 1 (9) 1 (30)(2)
Boley 103|)1] 5 — (6) — —"31) (5) 1) —— 98) 9) 3) a nares)
55 |(—91S)— 4) —) 9) —1 (921 9) 8) i) —) 9) 2) 1 —--—1— 714) 1 aot)
1—--—-s—-—7—-—4 01 —-—10311—-—1—24—14 1— 1 3101 a 1 45)—
=== GAP OF 8Ometres
oe T SAMPLED
—"—"—~ ONLY PARTIAL EXPOSJRE
- 4
|e Bere wiz6] 1 —- — —~— 68 — 7 — — —-— 4 —— 9 9 —~— —~— =» — 4 ---31-—1H—
eared GAPOF 40m
-- = 5-«3 [1 20] Se ee 2a fe eg gee er a d= 1 SSssasf 71 @=
|S

Fig. 14 Fauna from the Nant Tan y Bwlch section. (For Conularida read Macheridia.)

with Skenidioides and Nicolella also representing important elements. The middle part of the
formation is dominated by Howellites and, to a lesser extent, Kloucekia, Broeggerolithus and the
Ostracoda. However, in the uppermost part of the formation (i.e. the Dyfi Mudstone Member)
the fauna is characterized mainly by Sericoidea and rarer forms like Cyclospira.

In the Rhiw March section the Gelli-grin Formation has been sampled more thoroughly than
at any other locality. The pattern of faunal distribution is similar to that noted in the Tan y
Bwlch section. The lower beds are dominated by Onniella and Bicuspina, with Eoplectodonta
being less abundant than at contemporary horizons to the north. The Onniel/la-dominated. beds
pass up into strata in which Nicolella and Skenidioides are important elements and are in turn
succeeded by mudstones of the middle part of the formation which are dominated by Howellites,
Kloucekia, Broeggerolithus and the Ostracoda. The upper (Dyfi Mudstone) part of the formation
is dominated by Sericoidea and Onniella in association with less abundant forms including
Skenidioides, Cyclospira, Eoplectodonta, Nicolella, various inarticulates and macheridians.
The uppermost metre of the Rhiw March succession consist of soft, coal-black graptolitic shale
containing a monospecific assemblage of Climacograptus minimus (Carruthers); specimens were
not counted for inclusion in Fig. 15.

Non-dominant elements in the Gelli-grin Formation. Throughout its area of outcrop the lowermost
beds of the formation contain a number of brachiopod genera which are either unique to this
part of the Lower Bala Group or only otherwise known from the Derfel Limestone. These genera
include ? Pseudolingula, Platystrophia, Anisopleurella, Oxoplecia, Palaeostrophomena and Bimuria.
Of these Bimuria was previously unknown in Wales and Palaeostrophomena and Anisopleurella
were hitherto unknown in the Gelli-grin Formation. Salopia is also characteristic of the lower
part of the formation but, in addition to being known from the Derfel Limestone, is now also
recorded from horizons beneath the Frondderw Ash in Afon Twrch. A few specimens assigned
to Kjaerina (Hedstroemina) have been recovered from sample R28; this genus was previously only
known from the Glyn Gower ‘unit’ of the Lower Bala Group (Williams 1963 : 460). The distribu-
tion of seven of the eight above-mentioned genera is shown in Figs 12-15; the eighth genus,

187

CARADOC FAUNAL ASSOCIATIONS

M(t Ge

—@p1 ms
Cel(s) 4
— (9) | (@)
—@)b E

— @s)1

— (6) 4 (ed Zz

— (08) + (9) +
(yz) 4 (e) 6
Cel) 4 (sd 4
— (6) b (A)
“(zd CA
—(Z)b Uz

et Cy

— (bh) 4 4) s

= =A TS

SS SY)

sl eater (15 A

ail (ZEEE:

—(y) eae

ee ap
— Gb) t (9) s

— (ht di

ay at can (OE),9;
oe Sees

Depioujsa
eyeuul)) DL

—---t9%---—t - z ze == [Yeh CY OE
SS 8 Oh ge SS eS a Oa aS eh OL Ege tke a Ge
ae a ak Lea) st t i eh ie Pa YS a i
Si SS SS § SS SSS SSS CO STI) Geen a arta ta ae a a
Si dS = © ‘ = CSS ARS = fe 976 € Lae C3 = 8S 1 = 2 bo t
TS SOT Oe = SS SS Sh SS Sk SS SiS SS eS he SS SS Sie
2) S Sale eee CUD, mae tig ings Op eee eee a Vit E eake ie bom Calne REC © eh = we - e =
(CHL OG. SG = Fania as) I IS =F ht SS z & Saath eS
= = Gy PSUS pats iat ie reel oa See O = t = 3 = =

Cb PP es Sz hence ig Pi FE Sai SET SS eee es ee i ee ian eee afl ate me ie
= = 3 = P= = Stk Se ge ae a Sa eee eh 2 tt al 2 ci oe ee) eka
> ih Le eee Spee i at eM Seen a Nm a tea Caer? Ag mgt pees te AEE, ACT ae mes
= Si Ggo= iy t= aS (> fp SaaS SS SS SS SS SS e == ait
2S bey Ea 2) Sam Ge Saat - ee ag gt ate a Dan) Ile a dT
a aaa 6 I ae = 9 ae Carseat f) = 8 SiS EGS ieee
Fe Se 2 Ca) a ert (| = == pee = wl SSS SS SS eS
= SS fe P fi YS Sp i CS i oo
CC i i t = eee Sis ee aes) ote ad HRS SS SO Se fr
SSSe hoe e eos e= SSI US GS SS SS CSS SSeS = (YS SS
SS SS = Be Sh eh a a 8 a a Ra eh
Ee IS Sa i = fl Sh ee DOSOCR SS) SR a5 meme ae a
SPAS ML, «BP Be = SS bot SS SS 6S Ss SS 1962S. >i Sis
Se Spee es NEL (gee = So ee OS Sa. Soa ee ae ae C1 0 pened aka Rese ae Sis eae ‘

}1)N20}u 24
“3q @s0uDy
DIA]DAIG
Dplav}nuod

AT

Dweuoj>

pul "Dyqoy lat

SDWOPpoyruC7

Boe
Piuljayos!

jos |

mya

Buc

Ded

ay sD1UI

jauaw)

Diyeanow

DBAzojoId

‘DaydsoysA5,

darysoe0)} 0g

%

DulseD
Di) 202032 04
Dounwig

yepur ydouys

s

Deploaise

TRIOpO}D0 1403
yda7,

DyeAgqsa MO:
Duns
Duidsn 21g
Dyewsney

D Seyiemon

ou

Fayyloseyol
Di109'N
OsDI0d

016

“yepul ‘wiop

st 0O€

Se 07

= —
~ | euoyspran—
soa 2D)5n) —

| ne use|
ers

Fauna from the Rhiw March section (Gelli-grin/Nod Glas beds). (For Bellerophon read

Fig. 15

Sinuites, and for Conularida read Macheridia.)

188 M. G. LOCKLEY

Oxoplecia (not recorded in the samples from which the data tables were compiled) is now known
to occur at horizons GGlce and GGle (Fig. 12) and TB9 and TB12 (Fig. 13), following the
examination of additional material.

The middle part of the formation generally contains fewer brachiopods and more trilobites
than the lower beds. With the possible exception of Rhactorthis, less dominant elements show no
significant restriction to the middle part of the formation. Conversely the distribution of Para-
craniops macellus Williams suggests that this form has an affinity with the faunal associations of
the lower and upper parts of the formation (Fig. 15).

The upper part of the formation has yielded several taxa which were hitherto unrecorded in
the Upper Bala Group. These forms, which include Paterula, Palaeoglossa, Protozyga, Phillip-
sinella, Lonchodomas and Sphaerocoryphe, were all recovered from the Cymerig Limestone
member or associated beds in the Tan y Bwlch and Rhiw March sections (Figs 14-15). The
discovery of Protozyga at these localities represents the first record of this genus in Wales. The
trilobites Phillipsinella, Lonchodomas and Sphaerocoryphe are all known from the Upper Bala
Group of this area but were previously unknown at these earlier horizons. Cyclospira is an
important element of the Cymerig fauna at these two localities; although recorded by Williams
(1963) the material recovered in this study has facilitated a more thorough appraisal of the
specific affinities and distribution of this form than was hitherto possible. Full taxonomic
descriptions are given below, p. 219.

The Pistyll Gwyn, Y Ceunant and Aber-Cowarch sections

Fig. 17 contains faunal data derived from the study of the three above-named sections. These
sections, the southernmost in the area under study, cover the uppermost 30 m of the Caradoc
succession which, at all localities, comprises the Nod Glas Formation and a part of the underlying
Ceiswyn Mudstone. The faunal succession is similar at each locality.

The grey mudstones underlying the Dyfi Mudstone are dominated by Howellites and, to a
lesser extent, the trilobites K/oucekia and Broeggerolithus. These are succeeded by the rather more
pyritous, grey Dyfi Mudstone which is dominated by the small brachiopod Sericoidea. The
member is also characterized by the variable occurrence of Kloucekia, Broeggerolithus, Cyclospira
and the Macheridia. At Pistyll Gwyn these mudstones contain a locally-developed phosphatic
limestone which contains Sericoidea, Broeggerolithus, Nuculites and Sinuites. The discontinuous
Cymerig Limestone Member, consisting of variably fossiliferous, crystalline nodules measuring
about 30 cm x 15 cm, occurs at an horizon in the upper part of the Dyfi Mudstone.

Above the Cymerig Member the Dyfi Mudstone grades rapidly up into the sparsely fossili-
ferous or entirely unfossiliferous, dark grey Corris shale.

Faunal associations

Introduction

The data presented in Figs 4-17 represent as thorough a quantitative description of the faunal
content of the succession as the sampling scheme allows. Since only a cursory glance at these
data indicates that samples from like facies consistently contain recurrent combinations of taxa
in similar proportions (whilst samples from other facies contain different combinations and
proportions of mainly different taxa) it must be concluded that the associations are largely

Fig. 16 Percentage distribution of taxa in named sections through the Gelli-grin Formation.
Abbreviations refer to Nicolella (Nic), Dolerorthis (Dol), Rhactorthis (R), Platystrophia (P),
Skenidioides (Sk), Cremnorthis (C), Dalmanella (Dal), Reuschella (Re), Bicuspina (Bi), Leptestiina
(Ls), Sowerbyella (S), Eoplectodonta (Eopl), Bimuria (B), Strophomenacea/Macrocoelia (S),
Leptaena (La), Palaeostrophomena (P), Cyclospira (Cy), Kloucekia (K\), Broeggerolithus (Br),
Gastropoda (G), Macheridia (Con), ramose Bryozoa (B‘), prasoporid Bryozoa (B?), Ostracoda
(Ost). Dashed lines represent boundaries between faunal associations.

RHIW MARCH

190 M. G. LOCKLEY

ASHGILL———+ CI
MUDSTONES | ~~ .

-

Liz -YC7

Za RHIWLAS- LST

- YC6
—— — -9-PS3_ ___cymERIG----— §G- ves -----
“yYc4
=PG2
e- PG 1b ~ MES
\Apg= PGla 3
set a eS -Yo2
- PG
“YC!
5m
0

(Howellites)
Reuschella
Kloucekia
ggerolithus

Trilobita, indet.
Ostracoda
Bivalvia

phon
Other Gastropoda
Conularida
Crinoidea
Graptoloidea
Miscellanea?

Dalmanellidae, indet.
Bellero

Inarticulata

Sericoidea

Cyclospira

Broe

1(3)

1(3) 13)

BEEBEBEEEE

a8

Fig. 17 Fauna from the Pistyll Gwyn section (PG), the Y Ceunant section (YC) and the Aber
Cowarch section (AB). (For Bellerophon read Sinuites, and for Conularida read Macheridia.)

CARADOC FAUNAL ASSOCIATIONS 191

facies-related and that the sampling procedures consistently provide an adequate census method.
This being the case it is necessary to proceed by outlining distinctive associations and assemblages
and testing that at least in the general sense they differ significantly from each other in terms of
their overall composition.

Although the associations named here (see also Lockley 1978) were picked out initially by
‘simple inspection’ of the data (cf. Watkins 1975 : 48) samples considered representative of given
associations were compared with each other and with those from other associations using the
Similarity Index or Index of Affinity (Murray & Wright 1974 :3; Rogers 1976 : 504-506).
Details are given below. Furthermore, all numerical data are presented here (Figs 4-17) in such
a way as to be readily available for subjection to further quantitative analysis.

The associations and assemblages defined here (in descending order of their stratigraphical
occurrence) are as follows.

5. The Onniella—Sericoidea and the Sericoidea Associations from the Nod Glas Formation
respectively north and south of Llanymawddwy

4. The Howellites—Kloucekia Association from the middle part of the Gelli-grin Formation

3. The Nicolella—Onniella Association from the Lower part of the Gelli-grin Formation (with
variant earlier and later phases from the Derfel Limestone and upper part of the Gelli-grin
Formation respectively)

2. The Howellites—Paracraniops Association from the upper part of the Allt Ddu Formation

1. The Heterorthis Assemblage from respective upper and lower parts of the Llaethnant and
Allt Ddu Formations.

The term ‘assemblage’ is used to distinguish clearly transported faunas from those considered
to be either in situ or of the ‘disturbed neighbourhood’ type (sensu Scott 1974: 321), which
are referred to here as ‘associations’. The term ‘phase’ is used informally to indicate the different
time intervals represented by the repetitive stratigraphical occurrence of the same (albeit varying)
faunal association.

Although multivariate analysis has not been used to cluster like samples and to define associa-
tions, the use of the Similarity Index or Index of Affinity (IA) to test affinities between the
majority of the larger representative samples (90 in all) serves the same purpose and clearly
indicates that the associations named herein are relatively homogeneous in internal composition
whilst being quite distinct from each other. Over 220 representative IA values consistently
indicate that intra-association IA values are high whilst inter-association IA values are very low.
For example, respective mean IA values for the Howellites—Paracraniops and the Howellites—
Kloucekia Associations (with range in brackets) and number of IA values considered [in square
brackets] are 61-4°% (40-91 %) [16] and 72-8 % (37-98 %) [54]. Similarly, respective values for the
Nicolella—Onniella Association in the lower part of the Gelli-grin Formation and the Onniella—
Sericoidea Association are 57-4% (17-87 %) [59] and 48-8 % (7-83 %) [36], whilst the phase of the
Nicolella—Onniella Association in the upper part of the formation exhibits lower values, i.e.
40-8 % (20-63%) [10]. Although mean IA values are below the 80% level considered to indicate
‘identical assemblages’ (Murray & Wright 1974 : 3), individual IA values above 80% are recorded
for each of the associations except the phase of the latter one mentioned here which in any case
is described from only a small number of samples. Mean JA values show a marked contrast to
the low inter-association values calculated in order to compare the Howellites-Kloucekia Associa-
tion with both the Howellites-Paracraniops Association and the Nicolella—Onniella Association
(lower Gelli-grin phase). Respective mean IA values (with range) and number of IA values used,
as before, are 7:8 % (3:2-13:2%) [4] and 9-9 % (2:5-13-1%) [4]. An outline of these recognizable
associations is presented in quantitative terms using the parameters of persistence of occurrence
and mean relative abundance. Fig. 18 outlines the stratigraphical distribution of associations
identified in the fossiliferous upper part of the Caradoc succession between Bala and Dinas
Mawddwy, and Fig. 19 outlines their composition.

Calculations of the relative abundance of taxa from each sample (Figs 5, p. 177, 11, p. 182,
16, p. 189) permitted the construction of a series of tables (one for each sample) in which faunal
elements were ranked in order of abundance (Lockley 1977). The numerically dominant taxa

192 M. G. LOCKLEY

which make up 80% of the fauna may be regarded as the Trophic Nucleus; Neyman (1967)
proposed this quantitative definition of the nucleus originally defined by Turpaeva (1948). The
dominant faunas listed in Fig. 19 therefore represent the Trophic Nucleus of their respective
named associations either at named localities or for the association as a whole.

Although some palaeontologists (e.g. Titus & Cameron 1976) have named associations or
communities after their rarer component species, such methods do not conform with the more
popular tendency of naming associations after their dominant component taxa. The classic work
of Petersen (1924), summarizing his studies of marine animal communities in Danish waters,
included an outline of four categories of component taxa — Ist, 2nd and 3rd order characterizing
species and associated animals. Thorson (1957 : 477) subsequently proposed that these respective
characterizing species be quantitatively defined as representing at least 5, 5, 10 and 2% of the
total living weight (biomass) in at least 50, 50, 70 and 25% of samples from any given community.

Sane GRIN NANT- aia
Sanne: ee Soci Palen ne PI ACY

ee ONRIELLA SERICOIDEA oun a,
“SL, ASSOCIATION ERICOIDEA! “AS aka LAS
= Az

MARGINAL _H¢ HOWELLITES

Assoc ier oo

20m
PARACRANIOPS ASSOCIATION

FAUNAL DEPLETION
ZONE

A

Fig. 18 Stratigraphical distribution of named faunal associations in the upper part of the Lower
Bala Group (drawn to scale).

The only differentiation made between Ist and 2nd order characterizing species was that Ist
order species occur ‘practically everywhere’ in a specific community whereas 2nd order species
occur ‘only in certain parts’ of such specific communities. Biomass estimates cannot be derived
from fossil associations without at least making numerous assumptions and repetitive measure-
ments for the calculation of the mean size of each population. Nevertheless, in the absence of
other evidence, relative abundance can be used as it is in this study as an alternative estimate of an
association’s composition. If this is done with the aim of identifying characterizing species it
becomes clear that only the dominant species fall within this category.

Named Associations from the Caradoc Series of the Bala to Dinas Mawddwy area

Associations in the Allt Ddu Formation

(1) The Howellites-Paracraniops Association, characteristic of the upper part of the Allt Ddu
Formation, contains no genera, other than these two and Broeggerolithus, which are representative
of biospecies which could be considered strictly analogous to Ist or 2nd order characterizing
species (Fig. 19). Although several other genera listed here occur in at least 50°% of samples, in

193
TAXA/GENERA OCCURRENCE (%) OCCURRENCE (%) ABUNDANCE
Howellites 12/12 100 % B 23/24 95.8% 37-42%
Paracraniops 12/12 100 % 17/24 70.8% 7.13%
Bryozoa 12/12 100 % 22/24 91.7% 13.35%6
Broeggerolithus 12/12 100 % 24/24 100 % 8.67%
Macrocoelia 10/12 83.3% 8/24 33.3% 1.51%
Brongniartella 8/12 66.7% 20/24 83.3% 2.03%
Ostracoda 5/12 41.7% 19/24 79.2% 4.42%
Bivalvia 5/12 41.7% 8/24 33.3% 0.75%6
Gastropoda 5/12 41.7% 16/24 66.7% 3.39%
Reuschella 4/12 33.3% 5/24 20.8% 1.42%
Dinorthis 3/12 0/24 = -
Leptaena 3/12 10/24 41.7% 1.73%
Sowerbyella 2/12 1/24 4.7% 0.84%
Bicuspina 2/12 14/24 58.3% 4.49%
Sericoidea 0/12 12/24 50.0% 3.48%
TOTAL 90.63%

Onniella 26/26 27.52%
Bryozoa 26/26 17.11%
Skenidioides 23/26 6.08%
Bicuspina 22/26 7.11%
Eoplectodonta 18/26 10.59%
Reuschella 17/26 2.93%

Nicolella 15/26 4.26%
Dolerorthis 1/26 0.32%

Howellites
Kloucekia
Broeggerolithus
Sericoidea

Skenidioides

Onniella
Dalmanellidae
Sericoidea
Bryozoa
Skenidioides
Conularida
Cyclospira
Ostracoda
Kloucekia

Eoplectodonta

Broeggerolithus
Nicolella

Graptoloidea

Fig. 19 A & B. Dominant and persistent taxa/genera in the Howellites—Paracraniops association
in the upper part of Allt Ddu Formation at Craig y Gath (A) and Rhiw March (B); inthe Nicolella—
Onniella association from the (C) lower and (D) upper parts of the Gelli-grin Formation; in the
Howellites—Kloucekia association from the middle part of the Gelli-grin Formation (E) and the
marginal Howellites association from the uppermost Ceiswyn Mudstone Formation (F); in the
Onniella—Sericoidea association (for Conularida read Macheridia) (G) and the Sericoidea associa-
tion (H) of the Dyfi Mudstone.

194 M. G. LOCKLEY

most cases their mean relative abundance is less than 5%. With the probable loss of soft-bodied
representatives of the fauna in the process of fossilization, all relative abundance figures would
tend to be proportional overestimates. It is therefore probable that even fewer species can be
confidently considered analogous to characterizing species (sensu Thorson 1957).

Although it is only possible confidently to use samples with at least 20 individuals for the
purposes of calculating relative mean percentages of taxa comprising associations, the calculation
of mean diversity is derived from the consideration of all samples in order to avoid an over-
estimate of diversity. The Howellites—Paracraniops Association has a mean diversity of 9-68 but
can locally be found to yield samples with a diversity of up to 17. The association is named
after Howellites ultimus Bancroft and Paracraniops glaber sp. nov. (p. 204) which both occur
persistently and relatively abundantly; respective values for the two sections studied are 95-8-
100°% occurrence (with mean relative abundance of between 37-4 and 53-5 %) for H. ultima and
70-8-100°% occurrence (with a mean relative abundance of between 7-1 and 12°) for P. glaber.
Broeggerolithus (cf. B. soudleyensis Bancroft) is of equal importance, occurring in all samples
with a mean relative abundance of between 4 and 8-7°%. Bryozoa are also an important element
of the fauna with two morphospecies (a prasoporid and a ramose bryozoan) representing 6—-13-4%
of the fauna in almost all samples (92-100%); see Fig. 19, p. 193. Amongst the other more
important ‘associated species’ are Macrocoelia prolata Williams and Reuschella horderleyensis
undulata Williams from the Craig y Gath section and Bicuspina spiriferoides (M’Coy), Sericoidea
sp. and Brongniartella cf. minor (Salter) from the Rhiw March section, all of which occur in at
least 25°% of samples with a mean relative abundance of between 2 and 4:7%. The Gastropoda
(including Sinuites sp. and Cyclonema sp.) are an important element of the association at the
latter locality.

Although there is variation in the composition of the association between the two localities of
Craig y Gath and Rhiw March its extent is insufficient to suggest the existence of more than one
association at these horizons.

Associations in the Gelli-grin Formation

The Gelli-grin Formation is characterized by three stratigraphically successive associations
(Figs 16, 18), which are here described in sequence. These are the Nicolella—Onniella Association
(phase 2) and the Howellites—Kloucekia Association respectively from the lower and middle parts
of the formation between Gelli-grin and Rhiw March, and also the Nicollela—Onniella Association
(phase 3) of the upper part of the formation between Gelli-grin and Pen y Cefn Coch. This
latter association is related to the association from the lower part of the formation and to the
Onniella—Sericoidea Association of the Dyfi Mudstone.

(ii) The Nicolella—Onniella Association (phase 2). From the lower part of the Gelli-grin Formation
this is a high diversity association (mean 17-5) named after Nicolella actoniae obesa Williams and
Onniella ostentata Williams. The latter form is particularly abundant (Fig. 19) and is associated
with large numbers of Eoplectodonta rhombica (M’Coy). This association is to some extent
similar in composition to the Derfel Limestone association (phase | of the Nicolella—Onniella
Association), which includes Dolerorthis tenuicostata Williams, Nicolella humilis Williams,
Onniella aff. avelinei Bancroft and Eoplectodonta lenis Williams amongst the most common
Brachiopoda. The Derfel Limestone association is also of a high diversity, yielding up to 28
species from a single large sample; such values compare with maximum values of 25 for samples
from the lower part of the Gelli-grin Formation. The association from the upper part of the
Gelli-grin Formation (phase 3; mean diversity 13:7 and maximum 19) is also referred to here as
the Nicolella-Onniella Association although it is again different in composition from the associa-
tion in the lower part of the formation (Fig. 19). N. actoniae obesa Williams and E. rhombica
(M’Coy) are common at horizons in both the lower and upper parts of the formation; however,
Dolerorthis duftonensis prolixa Williams, which is uncommon from the lower part of the forma-
tion, occurs abundantly in the uppermost parts. O. ostentata Williams is not recognized from the
upper part of the formation; the Dalmanellidae are poorly represented and assigned to Onniella

CARADOC FAUNAL ASSOCIATIONS 195

sp., Dalmanella sp. or ina few cases Bancroftina sp.; Figs 7A, 12 and 13 do not distinguish between
these three genera.

The composition of the Nicolella—Onniella Association (phase 2) from the lower part of the
Gelli-grin Formation is calculated from the analysis of samples GGIb-GGle, GG2a—GG2c,
ADZ, ADY, TB9-TB13 and R24—-R35. The summary of these data, shown in Fig. 19, indicates
that, at this level, Onniella and Eoplectodonta are analogous to ‘characterizing species’ with
Reuschella and Nicolella representing ‘associated’ forms. Although Skenidioides, Bicuspina and
Reuschella are important constituents of this association, conspecific forms also constitute a
significant element in other associations; they cannot therefore be considered strictly analogous
to Ist or 2nd order characterizing species in the Nicolella—Onniella Association.

There is some lateral variation in the composition of the lower Gelli-grin faunal association;
this corresponds to a decrease in sediment coarseness towards the SSW. The most notable
changes are a relative increase in the abundance of Bicuspina and Leptaena in this direction
and a corresponding decline in the abundance of Eoplectodonta and Nicolella (Figs 12, 13).
Rare occurrence of Sowerbyella in the Rhiw March section are unique to the formation and
indicative of the modification of this association towards the SSW.

(iii) The Howellites—Kloucekia Association. The middle part of the Gelli-grin Formation is
mainly argillaceous; representative samples from beds GGIf-GGlgl, GG2d-GG2g, TB14-
TB17 and R36—R44 reveal a faunal association dominated by Howellites antiquior (M’Coy),
Kloucekia apiculata (M’Coy) and Broeggerolithus nicholsoni (Reed). The association has a mean
diversity of 12-7. Maximum diversity values in excess of 20 have been recorded from samples
GGlgl, GG2d and GG2e; however, these are atypical in being recovered from horizons closely
associated with overlying and underlying strata containing higher diversity associations. The
other 18 samples, particularly those from Tan y Bwich and Rhiw March, exhibit lower diversities;
minimum values do not exceed 4.

Although Kloucekia and Broeggerolithus are considered to each represent less than 4% of the
total fauna (Fig. 19) in this association (when using the arbitrary correction factor of 4 as a
compensation for ecdysis), it is clear that a correction factor of say 2 or 3 would indicate a mean
relative abundance analogous with that of ‘characterizing’ species. Sericoidea, whilst being
analogous to an ‘associated’ species, is known to be mainly restricted to the two southern sections
(Figs 12, 13). The occurrence of Skenidioides in most samples from the Gelli-grin associations is
noted.

(iv) The Nicolella—Onniella Association (phase 3). The composition of this in the upper part of the
Gelli-grin Formation. differs from that noted for the related phases of the association elsewhere
in the succession; for example, it is only at this level that Dolerorthis duftonensis prolixa Williams
can be considered analogous to a characterizing species (Fig. 19). Differences in composition are
thought to be directly related to the distinctive calcareous facies associated with the Cymerig
Limestone Member north of Pen y Cefn Coch. Caradoc strata above the Cymerig member
in this region consist predominantly of coarse, shelly calcarenites containing tuffaceous material
and ? ferruginous oolites. At certain horizons (GG2k and sample 615) vertical crinoid stems and
bryozoan fronds are noted; these are considered indicative of the rapid deposition of coarse
material in a shallow-water environment. Bulk samples are only readily obtained from a few
horizons in the Cymerig and higher beds; in general the limestones and associated rocks are
unyielding enough to present practical sampling problems. Of the 9 samples collected in this
study only 7 yielded sufficient specimens to permit the estimation of relative abundances (Fig. 19).
Owing to the small number of relatively variably composed samples, it is considered that the only
clearly-defined characteristics which distinguish this facies fauna from related associations
elsewhere in the area is the relative dominance of Dolerorthis and Nicolella and the small numbers
of dalmanellids.

(v) The Onniella—Sericoidea Association. This facies fauna is found in association with the
argillaceous Dyfi Mudstone Member between Tan-y-Bwlch and Rhiw March (Fig. 18). It is

196 M. G. LOCKLEY

laterally equivalent to the coarse, calcareous beds containing the continuous Cymerig Member
north of Pen y Cefn Coch and, although containing the discontinuous ‘distal’ parts of the
Cymerig Member south of this locality, otherwise represents the most marked lateral facies and
faunal change observed in this study.

The association is dominated by Sericoidea abdita complicata subsp. nov. (p. 212) and Onniella
sp., which between them represent over 75 °% of the total fauna and are the only forms analogous
to ‘characterizing ’species. Poor preservation (associated with relatively intense cleavage in this
argillaceous facies) militates against sound statistical assessment of the Onniella specimens for a
specific determination. Less distorted specimens from the Cymerig Member display only external
features.

Cyclospira, Skenidioides and Eoplectodonta, together with the ostracod Tallinnella, can all be
considered analogous to ‘associated animals’. The association has a mean diversity of 10-9 with
maximum values (up to 17) representative of the Limestone Member and associated more
calcareous shales. Broeggerolithus and Flexicalymene are particularly abundant at certain horizons
(e.g. R49) in the Rhiw March section.

Similarities between the fauna from this association and the Nicolella—Onniella Association to
the north help substantiate correlations between quite different facies, respectively of argillaceous
and coarse calcareous sediment (Fig. 18). It is of interest to note that Nicolella, Dolerorthis,
Eoplectodonta, Leptestiina, Reuschella, Rhactorthis, Leptaena, Flexicalymene, Cyclospira, Onniella
and Sericoidea are amongst the genera common to both associations; the latter three dominate
in the Onniella—Sericoidea Association whilst the other forms, common to the north, are rarer
here to the south. Protozyga, Paterula, Phillipsinella, Lonchodomas and Sphaerocoryphe are all
unique to this association being currently unknown elsewhere in the area.

(vi) The Sericoidea Association, although related to the Onniella—Sericoidea Association, is
considerably less diverse (mean diversity 4:25, maximum 8). The association characterizes the
Dyfi Mudstone Member between Pistyll Gwyn and Aber Cowarch (Fig. 18). Representative
samples (PGla, PGlb, PG2, YC3, YC4, YCS5, AB4 and ABS) indicate an association entirely
dominated by Sericoidea with a few poorly-preserved macheridians and dalmanellids showing
distribution patterns analogous to those of ‘associated animals’. Broeggerolithus, Kloucekia and
unidentifiable graptolites also occur. Cyclospira and a spired gastropod are known from the
Cymerig member whilst additional molluscan material is known from the locally-developed
underlying limestone beds at Pistyll Gwyn.

(vii) The Marginal Howellites Association. Samples PG, YC1-2, ABO0S5-—09 and ABI-2 (Fig. 17,
p. 190) have yielded a low diversity (mean 3-25) Howellites-dominated fauna. Although preserva-
tion of most of the material is poor it has been possible to identify H. antiquior (M’Coy) as the
dominant species (83:7°%). Sericoidea is also ‘characteristic’ of this association; however, its
abundance is almost one order of magnitude less than in the succeeding Sericoidea Association.
Kloucekia and Broeggerolithus both occur relatively abundantly in beds underlying those associ-
ated with the first major influx of Sericoidea (i.e. samples PG, YC2 and AB3; Fig. 17). Elsewhere
throughout these successions their distribution is variable.

The Marginal Howellites Association is clearly related to the Howellites—Kloucekia Association
typical of the middle part of the Gelli-grin Formation both in terms of faunal composition and
stratigraphical relationships. However, since the Cowarch Phosphate Bed and associated shell
beds may be contemporaneous with the Pont y Ceunant Ash (Lockley 1980), the Marginal
Howellites Association could also have been established in the Abercowarch area in Lower Gelli-
grin times and subsequently migrated diachronously northwards in middle Gelli-grin times.

(viii) The Graptolitic Association. The uppermost part of the succession in the Tan y Bwlch to
Aber Cowarch sections (Figs 13-17) consists of dark grey or black graptolitic or virtually
unfossiliferous shales referred to as the Corris Shale Member (Lockley 1977, 1980). These beds
are lithostratigraphically equivalent to the graptolitic Nod Glas seen at the type locality Nant y
Nod. Since these dark pyritous shales have yielded only one or two small shelly fossils (samples

CARADOC FAUNAL ASSOCIATIONS 197

YC6, AB6 and AB7), from lower beds, it is convenient to refer to the fauna of this member as a
graptolitic association.

(ix) The Heterorthis Assemblage. The shell beds dominated by Heterorthis and Sowerbyella
beneath the Frondderw Ash in the Llaethnant Siltstone in Afon Twrch are considered separately
from the younger named associations described above. The distinctive shell beds are concentrated
in only about 15 m of strata (samples HI—H19 and equivalents, see Figs 3-5, pp. 174-177), and
are considered to represent transported assemblages. These richly fossiliferous beds are underlain
and overlain by mudstones and siltstones which are generally sparsely fossiliferous (Fig. 3);
mean percentage values from these small samples would be of little value. Even within the shell
bed sequence there are sudden changes in the composition of the faunas at successive horizons;
samples H2, H4-H6 and H16—H19 are dominated by Sowerbyella, sample H7 is dominated by
Reuschella and the Dalmanellidae whilst samples H8—H15 are dominated by Heterorthis. Mean
Index of Affinity (IA) values for samples H8—H15 and H16—H19 respectively are 84:3 and 88:2;
these contrast markedly with the IA value of 5-0 derived from a comparison of samples H15 and
H16 and serve to demonstrate just how pronounced the sudden changes in faunal composition
are. Similarly, the mean diversity of taxa (species) in the shell bed sample HI-—H19 is 11-25
(maximum value 17) whereas the mean diversity for overlying samples H20—H29 is 4:5 (maximum
9).

The composition of faunas above and below the shell beds is essentially similar to that observed
in less fossiliferous parts of the Allt Ddu, Glyn Gower and Llaethnant Formations and so could
be considered a poorly-developed expression of the Howellites—Paracraniops Association described
from the upper part of the Allt Ddu Formation.

In contrast, however, the shell beds containing Heterorthis cf. retrorsistria (M’Coy) and Sower-
byella sericea permixta Williams, whilst being reminiscent of the Heterorthis ‘faunule’ (sensu
Bassett et al. 1966 : 237), contain Onniella ostentata Williams, Nicolella sp., Orthisocrania sp.,
Salopia sp., Dolerorthis sp. and Chasmops sp., all of which are considered characteristic of the
Nicolella—Onniella Association. They are also characterized by significant numbers of Dinorthis
and Reuschella and various cephalopod and gastropod species.

Although the relationship between this Heterorthis-Sowerbyella Assemblage and its rare
elements is unknown, it is reasonable to assume that a Nicolella—Onniella type of association was
established locally in pre-Frondderw times.

Diversity patterns

There are various methods by which relative diversity values D may be calculated. The two
methods chosen here are the Margalef method which uses the formula D=(S-1)/log N (Margalef
1958) and the Sanders (1968) rarefaction technique which is similarly designed to calculate
diversity values for a standard sample size (in this case 50 individuals). Both methods have been
applied to palaeoecological studies; for example, Ziegler et al. (1968) used the Margalef index
in the modified form D=S/(log N), and various authors including Calef & Hancock (1974),
Antia (1977), Fiirsich (1977) and Watkins (1979) have used the Sanders rarefaction techniques.

MacArthur (1965 : 511) stated that the simplest measure of species diversity was a count of
number of species. Similar simple definitions of species diversity or species richness given in
numerous standard texts, e.g. Krebs (1978 : 374) and Valentine (1973 : 288), usually refer to
community diversity but may equally well be applied to the diversity of individual samples.

Since a simple count of species diversity is a prerequisite for the calculation of relative diversity
values in a series of samples, the original, size-dependent diversity values are plotted alongside
standardized values for comparative purposes (Figs 20, 21).

The two standardizing methods applied here have the effect of smoothing out patterns of
excessive fluctuation which are related to sample size; they also serve to emphasize real patterns of
variation like the pronounced fluctuations observed in the Rhiw March succession (Fig. 20).
Here a moderately diverse association in the lower part of the sampled Allt Ddu succession is
succeeded by less fossiliferous strata, with a less diverse fauna, in the uppermost part of the

198 M. G. LOCKLEY

Io ae et

:

|
ALLT DDU
FM

WATE EE dd de eet tin

ates ae
RHIW MARCH

Fig. 20 Faunal diversity patterns in parts of the Allt Ddu (AD) and Gelli-grin Formations (including
the Cymerig (C) and Pont y ceunant (Pyc) Members) at three named localities. A represents
species per sample; B represents diversity values standardized to a sample size of 50 individuals
using the Sanders (1968, 1969) rarefaction technique; C represents diversity values derived from
the Margalef (1958) index.

199

“suOT}eIDOSse [eUNeJ 0} diysuOTje[oI sy} pue suIs}}ed AjISIOAIP [Pune] TZ “BI

ajduios sad
salzads jo saqunNy =)

Peg ayoydsoyg Yy21DMOD eee
spag auojsawi| UAMS |JAtsig commer

@D D
UENCE

speq uaiing sapnjour WY

uolyDIDOSSW I141;0¥dDID

0% 0 0 N 50}

ay)

auojsawy BisewAd|

WOL

uolj;DUIOY

UD 11199 40 as0g

(A

Hittite iid

Heditit titi
ty titit
ii

yi
|

SOq1][2MOH / ate

CARADOC FAUNAL ASSOCIATIONS

= jpulBiow = ad TCEIEL SS
! x = / SSIES) : -
'S ee ee ee a eel es ee , 3 =} -- Ce i — " _ Poco

é eco] POS =|eee i a= = E go ee ey -|Ps00 -PTRPFIN

| sgl ~* 77 = =Jemo| 7 uorssazans /” Sn as a
“G, YW ~-="2n - -=lass = Ye 20poiD> jo doy a | ee —

OO ES 1 Sip == a

laatiea! Ss late (srr (Sisal lan oe alien
Ol (0) Ol (0) Ol (0) Oz 0 SZ 0 74 0

200 M. G. LOCKLEY

formation; these beds are in turn succeeded by those of the lower part of Gelli-grin Formation
which contain a high-diversity association related to a less argillaceous, calcareous facies. The
middle part of the formation, representing a return to more argillaceous deposition, is character-
ized by a low-diversity association which contrasts with the associations of higher diversity of
both the preceding and succeeding beds.

The use of the Sanders rarefaction technique generally involves adherence to a particular,
chosen sample size. In this case, the use of diversity values calculated for a sample size of 50
avoids the necessity of calculating diversity values for smaller samples, although this may be
done if desired. When used in isolation no such restraints affect the use of the Margalef diversity
index which produces size-dependent results; however, size-related variation in diversity values is
not excessive in this case where values for N vary only within a single order of magnitude. In the
present study Margalef diversity values were calculated only for samples with at least 20 indi-
viduals, except in the southernmost three sections where a few samples associated with the
Graptolitic Association are characterized by minimal numbers of specimens. The estimation of the
Sanders and the Margalef indices has the primary effect of reducing the numerical value of the
original whole-number count of species per sample to non-integral numbers of little more than
50% of the value of original counts; the latter method results in greater reductions (Fig. 20).

Although the diversity indices used here do not take into account species equitability (or
evenness of distribution of species within samples or associations) it is to be expected that low-
diversity associations would exhibit less even distributions than high-diversity associations
(Pielou 1969 : 233; Krebs 1978 : 456). Such relationships are intuitively evident in the case of the
associations discussed here. Although detailed discussions of species equitability are outside the
present scope, following the above positive diversity/equitability correlation we may say,
for example, that the Sericoidea-dominated association and the two dominated by Howellites
have less equitable species distributions than the diverse Nicolella—Onniella Association.

Since the Sanders method is limited to a particular sample size and produces values which fall
between those calculated by the other two methods used here (Fig. 20), it was not used in the
compilation of Fig. 21. Here, the two most different diversity indices are plotted for all the main
thoroughly-sampled sections (other than the Afon Twrch shell bed section) and shown in con-
junction with the stratigraphical distribution of named faunal associations. In addition to the
parallel patterns resulting from the application of differing methods to the same data, the most
striking features of the stratigraphical variation in diversity trends are as follows. Firstly, it is
apparent that there is a continual fluctuation in the diversity of sequentially collected samples;
this has been noted by Watkins (1979) and can be attributed to a number of possible factors
including patchiness of faunal distribution within associations, actual original fluctuations of
diversity with time and differences in post-mortem disturbance at successive horizons or variations,
including variations in fossilization processes, in the physical environment. The second distinctive
pattern pertains to the relationships between faunal associations and diversity trends; Fig. 21
serves as a comprehensive graphic description of inter- and intra-association diversity trends.
The relatively high diversity values noted for the Nicolella—Onniella Association from the lower
part of Gelli-grin Formation contrast with lower values noted for both the underlying and over-
lying associations and indicate that the variation in trends is directly related to original differences
in the composition of these associations. Therefore, although short-term fluctuations cannot
be unequivocally explained, longer-term variations can be accounted for to some degree.

Relationships between density and diversity
In lower parts of the succession it was found difficult to derive a statistically-valid sample (e.g.
50 individuals) from sparsely fossiliferous rock. For this reason tests were devised to establish
whether low diversity values were an actual phenomenon or the product of small numbers of
individuals in a sample (i.e. low density). Since high- and low-density samples often show similar
patterns of taxonomic composition significant increases in diversity would probably not result
from corresponding increases in sample size.

Two large collections, H29 (10 kg) and H26 (8 kg), from the sparsely fossiliferous lower Allt-
Ddu Mudstones of the Afon Twrch section were chosen for testing. Each sample was divided into

CARADOC FAUNAL ASSOCIATIONS 201

2 kg ‘sub-samples’ of rock and the total fauna from each was broken out and identified separately.
A consistent number of taxa was derived from each of these sub-samples and the cumulative
number of taxa was found to increase only gradually as the data from each sub-sample were
pooled (Fig. 22). A x? test revealed that there is no significant difference between the taxonomic
diversity in a small (2 kg) sample and a larger (8-10 kg) one; for H29 98%>P>95°% and for
H26 95%>P>90%. This implies a positive relationship between low density and low diversity.

Number of species

Cumulative no. of spp.

Sample DIVERSITY DENSITY
Number Specimens/kg.

[ 8 27
R 26
R 25
R 24

Number of Individuals

Fig. 22 A, x? data for sample H29 with graph showing cumulative number of species and individuals
derived from five 2 kg subsamples comprising a 10 kg sample. B, x? data for sample H26, divided
into four 2 kg subsamples. C, x? data for test of density—diversity relationships in a part of the
Rhiw March section. IFB: interformational boundary.

Further evidence of such a relationship results from the analysis of a series of samples from
contrasting facies (Fig. 22). Eight samples collected through 16 m of strata and across a facies
boundary all show a positive correlation between density and diversity (77, 50%>P>20.%).
Conditions favourable for increased diversity therefore favour increased numbers of individuals;
conversely, factors limiting diversity tend also to limit numbers.

Such positive density—diversity correlations as those noted here are by no means universally
evident; many authors have sought to demonstrate an inverse relationship between the two
parameters. For example, Calef & Hancock (1974 : 779) and Hancock et al. (1974 : 151) referred
to a negative correlation between density and diversity in their respective analyses of Silurian
benthic communities.

Valentine (1972 : 195) stated that there was no single, well-verified theory of diversity regula-
tion. Many environmental factors, including temperature, temperature stability, depth, salinity,
current activity and substrate composition affect the diversity and density of benthic associations.
The present author, like many workers, avoids considering any single factor in isolation and
tends to favour unifying theories such as the ‘stability-time hypothesis’ (Sanders 1969).

202 M. G. LOCKLEY
Taxonomic descriptions

The main aim of this section is to describe and figure genera and species hitherto unknown in
the Lower Bala Group. Amongst these are representatives of the trilobite genera Cyphoproetus,
Lonchodomas, Phillipsinella and Sphaerocoryphe. Species of the brachiopod genera Bimuria and
Protozyga are recorded in Wales for the first time. Excluding Cyphoproetus, the above-mentioned
genera are from the Gelli-grin Formation (at Rhiw March) and current evidence suggests
that their distribution elsewhere is very restricted. Representatives of the brachiopod families
Obolidae and Paterulidae, together with species of the genera Orthisocrania, Paracraniops,
Palaeostrophomena, Anisopleurella, Sericoidea and Cyclospira, are described and figured. Although
species belonging to some of these taxa are already known from the Lower Bala Group, those
described here are either new species, species new to the area or well-preserved specimens of
previously poorly-represented species. New representatives of the Lingulacea recovered during
the present study are given only a generic classification because they cannot be assigned with
confidence to any more specific taxonomic grouping owingto their lack of internal morphological
features. This precludes lengthy discussion of diagnostic features, so only brief descriptions
accompany certain plates.

Dr P. J. Brenchley has recovered several specimens of the brachiopod genus Parastrophinella
from the Caradoc Bryn Beds of the NE Berwyns. Since this genus was hitherto unknown in the
Caradoc of north Wales, and has only been recorded in the Caradoc of Salop and south Wales
(Williams 1974, Addison 1974), the Berwyn material is best described in conjunction with new
material from the Lower Bala Group.

Representatives of the Mollusca (6 genera) and the Echinodermata (2 species) are also figured.

Measurements for most of the material are included; however, values in brackets refer to
variably distorted specimens and should therefore be regarded only as approximate estimates of
original size.

Further information on those brachiopod species (and related forms) not fully described
here may be found in Williams (1963) and Hurst (1979); similarly for information on the Trilobita
see Whittington (1962-68).

Phylum BRACHIOPODA Dumeril, 1806
Class INARTICULATA Huxley, 1869
Order LINGULIDA Waagen, 1885
Superfamily LINGULACKEA Menke, 1828
Family OBOLIDAE King, 1846
Subfamily LINGULELLINAE Schuchert, 1893

Genus LINGULELLA Salter, 1866

? Lingulella sp.
Fig. 24

DESCRIPTION. Suboval, convex specimen of exfoliated Lingulella, with pedicle valve 80% as wide
as long and an acute beak (<90°); ornamented with concentrically-arranged growth lines
numbering at least four per mm at 3 mm anterior of the umbo, with fine fila between growth
lines.

MATERIAL AND LOCALITY. Complementary internal and external parts (BB92200a, b) of exfoliated
pedicle valve (length 8 mm, width 6:5 mm, depth at umbo 0-7 mm) from fossiliferous shales at
Pont Aber Derfel (SH 850395), the type locality for the Derfel Limestone.

DiscussiON. This is the first record of a genus belonging to the Inarticulata in this oldest member
of the Lower Bala Group.

CARADOC FAUNAL ASSOCIATIONS 203
Lingulella cf. ovata (M’Coy, 1846)
Fig. 25

DESCRIPTION. Suboval, convex specimen of a pedicle valve 70% as wide as long and with an acute
beak (< 90°). Ornamented with fine concentric growth lines seen only at anterior commissure
(spacing 0:3 mm).

MATERIAL AND LOCALITY. Complementary internal and external parts (BB92201a, b) of exfoliated
pedicle valve from bed R36 in the Gelli-grin Formation at Rhiw March.

Genus PALAEOGLOSSA Cockerell, 1911 (emend. Williams 1974)
Palaeoglossa cf. attenuata (J. de C. Sowerby, 1839)
Figs 26, 27

DESCRIPTION. Subtriangular convex pedicle valve 65% as wide as long with an acute beak
(60° + 5°), ornamented with strong concentric growth lines (2 per mm) in addition to fine fila.

MATERIAL AND LOCALITY. Measurements in mm.
length width depth

External mould of conjoined valves (BB92202) from bed GGld . 14 9 1-5
Internal and external part of an exfoliated pedicle valve
(BB92203a, b) from bed R52 . 4 4 z : ; : 10 6.5 =

Discussion. P. cf. attenuata is recorded here in the Lower Bala Group for the first time.

Subfamily GLOSSELLINAE Cooper, 1956
Genus PSEUDOLINGULA Mickwitz, 1909 (emend. Williams 1974)
? Pseudolingula sp.
Fig. 28

DESCRIPTION. Parallel-sided, convex glossellinid 42% as wide as long with an acute beak (< 90°)
and ornamented with (poorly preserved) growth lines.

MATERIAL AND LOCALITY. An external mould of a brachial valve (BB92204), length 24 mm,
width 10 mm, was recovered from bed GG1b in the Gelli-grin Formation.

Discussion. Pseudolingula is recorded here in the Lower Bala Group for the first time.

Family PATERULIDAE Cooper, 1956
Genus PATERULA Barrande, 1879
Paterula sp.
Figs 29a, b, 30
DESCRIPTION. Smooth oval Pateru/a with rounded posterior and anterior margins. 76-84% as
wide as long, beak poorly defined. Well-developed marginal limbus.

MATERIAL AND LOCALITY. Single specimens of the external and internal moulds of convex
pedicle (?) valves (BB92205, BB92206) respectively from beds R52 and R53 in the Cymerig
Limestone at Rhiw March.

Discussion. Paterula is recorded here in the Lower Bala Group for the first time.

Family CRANIOPSIDAE Williams, 1963

Genus PARACRANIOPS Williams, 1963

Paracraniops cf. macellus Williams, 1963
Figs 32a, b

DESCRIPTION. Elongately oval and subequivalve Paracraniops, with flattened posterior margin

204 M. G. LOCKLEY

and strong concentric ornamentation consisting of 7-8 irregularly-spaced growth lines per mm
on the anterior part of valve.

MATERIAL AND LOCALITY. Five well-preserved external moulds were obtained from beds GG2c
(BB92208), R29 (BB92263), R46 (BB92264) and R47 (two specimens) in the Gelli-grin Formation.

DiscussION. Williams (1963 : 347-348) discussed the problems of differentiating between pedicle
and brachial valves belonging to this genus and revised his earlier diagnosis (1962 : 88-89) of
which valve was which, concluding that the pedicle valve bears the shield-shaped muscle scars
while the brachial valve bears posterior adductor and oblique scars extending anterolaterally
for about a quarter of the valve length. Mitchell (1977 : 22-23) apparently ignored Williams’
later proposals, preferring to adhere to his earlier definitions. However, the more recently
proposed orientation is adhered to here and it is noted that P. glaber sp. noy. has distinctive
posterior adductor (?) scars (Figs 33-36).

When describing P. macellus! (Williams 1963 : 348; pl. 1), insufficient material was available
to illustrate the external morphology adequately and the reader was referred to descriptions of
P. pararius Williams (1962), emend. Williams (1963) for additional information. Material obtained
during this study shows the close resemblance between the exterior morphology of P. cf. macellus
and P. pararius.

Paracraniops glaber sp. nov.
Figs 33-36

DiaGnosis. Large, externally smooth, oval, dorsibiconvex (?) Paracraniops with well-developed
limbus, slightly convex to asymmetrically conical dorsal (?) valve and slightly convex ventral (?)
valve.

NAME. ‘Smooth’.

DESCRIPTION. Oval Paracraniops 66-84 % as wide as long in populations with smaller and larger
mean size respectively. Dorsal (?) valve with pair of variably-developed faint posterior adductor
scars (?) arising anterior to the posterior margin (10% of valve length) and extending for 20-30 %
of valve length, only known in large specimens (Figs 33, 34). External ornament smooth except for
well-developed marginal limbus. (Rare development of concentric growth lines seen on some
internal moulds.)

TYPE MATERIAL. Measurements in mm.

length width
Holotype, internal mould of brachial (?) valve, BB92209 : F SPS) 4:8
Paratype, internal mould of pedicle (?) valve, BB92210 ; ; 5:0 4:2
Paratype, internal mould of brachial (?) valve, BB92211 : : (4:5) (4-1)
Paratype, internal mould of brachial (?) valve, BB92212 : : 3-4 3-0

TYPE HORIZON AND LOCALITIES. Fig. 23 gives biometric parameters of specimens of P. glaber
sp. nov. from sample loc. 55 near Fedw Farm (9089 2975) and from the Rhiw March section
(Fig. 9, p. 180). The former locality, on the south side of Nant Rhyd Wen (opposite the cliffs of
Craig y Gath), is the ‘alternative’ type locality for the Allt Ddu Formation (p. 178). Fig. 45A,
p- 208, gives statistics of two small populations of P. glaber from beds R093 and R096 in the
Rhiw March section.

Discussion. The morphology of most of the specimens belonging to this common Allt Ddu
species is so featureless that previous studies have disregarded them owing to uncertainty about
their taxonomic affinity (Williams, personal communication 1976). However, a fortuitous dis-
covery of a small population of well-preserved, large Paracraniops specimens in a partially
decalcified calcareous nodule has revealed that a few of these normally featureless valves have
internal muscle scars preserved. The internal morphology of the dorsal (?) valve closely resembles

‘Called by Williams P. macella; however, a ruling of the I.C.Z.N. (1974, Bull. zool. Nom., London, 31 : 81-83)
is that generic names in -ops shall be regarded as masculine in all cases.

CARADOC FAUNAL ASSOCIATIONS 205

that of P. pararius Williams, but the lack of external ornament on P. glaber distinguishes it from
other related forms at least at the specific level. P. glaber appeared much earlier in the Lower
Bala Group than the related form P. macellus Williams, which is only known from the Gelli-grin
Formation.

The mean length/width ratio of P. glaber varies allometrically with growth. Specimens from loc.
55 apparently grew to a relatively large size, increasing in width relative to length.

LENGTH a
A 1 C
5mm Width

Fig. 23 Biometric parameters (length and width) of Paracraniops glaber sp. nov. A, specimens from
sample 55 including two showing dorsal (?) muscle scars. B and C respectively represent specimens
from samples R096 and R093; corresponding size frequency histograms are also shown.

Paracraniops cf. glaber Lockley, herein
Fig. 37
MATERIAL AND LOCALITY. A single slightly distorted internal and external mould (BB92213a, b)
of a specimen apparently conspecific with P. glaber recovered from bed H6 (9001 2305) is figured.

The dorsal (?) posterior adductor scars and well-developed internal growth lines are seen.
(Length of specimen 4:6 mm, width 3-1 mm.)

Suborder CRANIIDINA Waagen, 1885
Superfamily CRANIACEA Menke, 1828
Family CRANIIDAE Menke, 1828

Genus ORTHISOCRANIA Rowell, 1965
Orthisocrania sp.
Figs 3la, b
DESCRIPTION. Subcircular, convex brachial valve with elliptical anterior adductor scars.

MATERIAL AND LOCALITY. Single internal mould of deformed brachial valve (BB92207) from
bed H8 in Afon Twrch (9001 2305); see Fig. 4, p. 176.

Discussion. This is the first record of Orthisocrania from beds below the Gelli-grin Formation.
This Soudleyan occurrence, the earliest known in Britain, was predicted by Wright (1970 : 102).
Class ARTICULATA Huxley, 1869

INTRODUCTION. Systematic descriptions of members of this class are given for representatives of
the Orders Strophomenida, Pentamerida and Spiriferida only. Although representatives of seven
species in the Order Orthida are figured, they are all considered to be conspecific with those

206 M. G. LOCKLEY

described by Williams (1963); they are included in order to illustrate well-preserved morphological
features and, in the case of Nicolella cf. actoniae obesa, Onniella ostentata and Salopia sp.,
represent early occurrences of species previously known only from the Gelli-grin Formation.

Order ORTHIDA Schuchert & Cooper, 1932
Superfamily ORTHACEA Woodward, 1852
Family ORTHIDAE Woodward, 1852
Subfamily PRODUCTORTHINAE Schuchert & Cooper, 1931

Genus NICOLELLA Reed, 1917

Nicolella cf. actoniae obesa Williams, 1963
Fig. 38

MATERIAL, LOCALITY AND DISCUSSION. A single undistorted pedicle valve external mould
(BB92215), length 6:0 mm, width 8-5 mm, depth 1-2 mm, was recovered from bed H6; two
distorted brachial valves were also found (beds H6 and H40, BB92214). Only the pedicle valve
specimen is figured. This is the earliest record of this species in the Lower Bala Group.

Fig. 24 ?Lingulella sp.(p. 202). Derfel Limestone, Nant Aber Derfel. BB92200a. Internal part of
exfoliated pedicle valve, 4:5.

Fig. 25 Lingulella cf. ovata (M’Coy) (p. 203). Gelli-grin Formation, Rhiw March, Llanymawddwy.
BB92201a. Internal part of exfoliated pedicle valve, x 4.

Figs 26-27 Palaeoglossa cf. attenuata (J. de C. Sowerby) (p. 203). Fig. 26, Gelli-grin Formation,
near Gelli-grin Farm, Bala. BB92202. Articulated valves, x3. Fig. 27, Cymerig Limestone
Member, Rhiw March, Llanymawddwy. BB92203a. Internal part of exfoliated pedicle valve, x 3.

Fig. 28 ? Pseudolingula sp. (p. 203). Gelli-grin Formation, near Gelli-grin Farm, Bala. BB92204.
External mould of a brachial valve, x 2.

Figs 29-30 Paterula sp. (p. 203). Cymerig Limestone Member, Rhiw March, Llanymawddwy.
Figs 29a, b, BB92206. External mould of a pedicle (?) valve and latex cast of the same, both x7°5.
Fig. 30, BB92205. Internal mould of a pedicle valve, x 6.

Figs 3la,b Orthisocrania sp. (p. 205). Llaethnant Formation, Afon Twrch, near Bwlch y Groes.
BB92207. Internal mould of a brachial valve, x4; latex cast of same, x 5.

Figs 32a, b Paracraniops cf. macellus Williams (p. 203). Gelli-grin Formation, near Gelli-grin Farm,
Bala. BB92208. Latex cast (a) of external mould of valve (b), both x9.

Figs 33-36 Paracraniops glaber sp. nov. (p. 204). Allt Ddu Formation; all internal moulds. Figs
33-35, near Fedw Farm, Llangower. Fig. 33, Holotype, BB92209. Brachial (?) valve, x 6. Fig. 34,
Paratype, BB92211. Brachial (?) valve, x6. Fig. 35, Paratype, BB92210. Pedicle (?) valve, x4.
Fig. 36, Rhiw March, Llanymawddwy. Paratype, BB92212. Brachial (?) valve, x6.

Fig. 37 Paracraniops cf. glaber Lockley (p. 205). Llaethnant Formation, Afon Twrch, near Bwlch
y Groes. BB92213a. Internal mould of a brachial (?) valve, x 6.

Fig. 38 Nicolella cf. actoniae obesa Williams (above). Llaethnant Formation, Afon Twrch, near
Bwlch y Groes. BB92215. External mould of a pedicle valve, x 3.

Fig. 39 Rhactorthis crassa Williams (p. 209). Nod Glas Formation, Rhiw March, Llanymawddwy.
BB92216a. Internal mould of a brachial valve, x 3.

Fig. 40 Howellites cf. ultimus Bancroft (p. 209). Allt Ddu Formation, Rhiw March, Llanymawddwy.
BB92217. External moulds of articulated valves, x 6.

Fig. 41 Howellites cf. antiquior (M’Coy) (p. 209). Ceiswyn Formation, Aber Cowarch, Dinas
Mawddwy. BB92218a. Internal mould of a brachial valve, x 4.

Figs 42-43 Onniella ostentata Williams (p. 209). Llaethnant Formation, Afon Twrch, near Bwich
y Groes. Fig. 42, BB92219. Internal mould of a pedicle valve, x3. Fig. 43, BB92220a. Internal
mould of a brachial valve, x 3.

Fig. 44 Salopiasp.(p. 210). Llaethnant Formation, Afon Twrch, Bwlch y Groes. BB92222. Internal
mould of a brachial valve, x 3.

CARADOC FAUNAL ASSOCIATIONS 207

*(SuOWIWIg) vynIjnS1q “ye Daidsoja4D

«

> ‘Aou ‘dsqns vivajdwos vjipqv vapiodisas

a

(L10°0)LO°T
vrvl"o
(VS0°0)¢S°T

C
(800°0)LLT°T

9678°0
(S80°0)67S°T

H 3) a
(OT0°0)TO*T (L20°0)90°T (600°0)2T°T
69£6°0 6628°0 1¢68°0
(290°0)0S°T (vE0°0)SE°T (860°0)6S°T
(190°0)SL°0 ©(SL0°0)8L°0 (940°0)TO"T
0S£6°0 zvz8"0 T168°0
(LSE°T)T9°V }~=— (eve )E6°< ~ (TOS*T)vO°S
(TOE"0)6T°2 + (G6E°0)9z°% += (69£°0) 08°
iS GT GT 62
isa]
g cS u 26 u sb uy
Oo
fe)
=]
c) “SUOT}O9S YOIeW MTYY pue yoTmg A uey 94} UT
s speq pe yetoosse pues suojseuTy Staz9swAD 944

wory (Suoumq) e}eOTNSTq ‘jJye BaATdSOTOAD
JO SOATVA (gq) TeTyovrq pue (Vy) eToTped jo
(4) UAPTM pues (T) 4ySUeT FO SOTIST}EIS 5

*sorTdures owes worly SeATeA [eTYyoerq jo
SOTIST181S (H-dA) ‘S®ID PON 94d UT
suoztzoy Inoy woary ‘dsqns ejeo0TTduwoo.

*aou
e1Ipqe Boeprootszag JO SaATeA aTotped jo
(4) WaPTM pue (T) YIBUeT Fo soTsTivIs g

“yoIeW MTYY 7% 9UOISPNW NPC IITV eu

ut (d) €608 pue (V¥) 960H SUOZTI0Y wor
*aou ‘ds seqe,Ts sdotueioeieq I0y (M) 4UZPTM

unuTxeW pue (T) Y4SUeT FO SoT4STIeIg Vv

208

‘q ‘Aou ‘ds daqgnjs sdoluvsovavg ‘WY :10J SONSNRIS Sp “Sly

0)

(900°0)660°T
L6£8°0
(680°0)L00°T

€

V

(800°0)6£2°T (210°0)S16°0

z0v8"0
(0L0°0)L8S°T

OLvL*O

( "IA ) re)

3
i

(z70°0)TS*T(m* “ST *zea) mM: BT

(L70°0)t8°0
G6¢L°0

(290°0)9£6°0
9678°0

(Lee*T)2lL*v §=(LST°2)S°6*y) = (LEB*T)ebr-r

(€l2°0)8¢°2 (Ovv°0)TE9°2 (89h°O)9LV-2
oe 8v 6S

co ua

(vL0°0)698°0
Tve8"0

lv 4 26 u

(212 °0)¢2TS°1(9L00°0)¢evo°T
92¢L°0 v796°0
(620°0)LOv°T

(£10°0)¢8¢°T

(v60°0)Lv2°T
(180°0)¢ve°T

Togl°0
(£0S°0)VT°V

L296°0
(6TE°T)S9°¢
(S0Z°0)TO"v }~=6(£Sv°T)O0°r
L

€

(970°0)VL6°0
8S¢8°0
(998°T)90°S
(£ve°0)OL*2

vs
ev

(69%°0)2v0°T
8£8°0
(8S0°0)798°0
(¥S0°0)29e°T
(£T0°0)¢0L°0
Lv¢8°0
(6S¢°0)vv-z
(82L°0)8z29°¢
VT

a

(vv0°0)16°0(1* St" aea)T° “BT

62vL°0

(S06°0)19°v

(682°0)S°2
8¢

lv a

(0L°0)000°T
109°0
(€90°0)009°0
(£90°0)L00°T
(2£0°0)999°0
966S°0
(2¢2°0)z288°T
(22S°0)L28°2
TI

V

I

(“°z2ea) umm

(1 °2eA)um-T

N

£4 TT B00T

( 0 -IeA) ”

3

(m?37 zea) M*9T
(1281 -zea) 1787

(e°rea)e

a

(a°zea) umm" A

(T° rea) um’ T

N

CARADOC FAUNAL ASSOCIATIONS 209

Family PLECTORTHIDAE Schuchert & Le Vene, 1929
Genus RHACTORTHIS Williams, 1963

Rhactorthis crassa Williams, 1963
Fig. 39

MATERIAL, LOCALITY AND DISCUSSION. Well-preserved, complementary internal and external
moulds of a brachial valve (BB92216a, b) from bed R49 in the Rhiw March section; its length
is 5-1 mm, width 7-0 mm.

Superfamily ENTELETACEA Waagen, 1884
Family DALMANELLIDAE Schuchert, 1913

Genus HOWELLITES Bancroft, 1945

Howellites cf. ultimus Bancroft, 1945 (emend. Williams 1963)
Fig. 40

MATERIAL, LOCALITY AND DISCUSSION. A group of at least seven articulated Howellites valves
(BB92217) from bed R089 show attachment to an elongate strand of unknown material. The
illustration of this specimen supplements descriptions of Howellites ‘life assemblages’ from the
Berwyn area (Brenchley 1966).

Howellites cf. antiquior (M’Coy, 1852)
Fig. 41

MATERIAL, LOCALITY AND DISCUSSION. A well-preserved internal and external mould of a brachial
valve (BB92218a, b), length 5-6 mm, width 7:0 mm, showing characteristic muscle scars, was
recovered from bed ABO7, associated with the Cowarch Phosphate Bed at Aber Cowarch.

Genus ONNIELLA Bancroft, 1928

Onniella ostentata Williams, 1963
Figs 42, 43

FIGURED MATERIAL. Representative internal moulds of a pedicle (BB92219) and a brachial valve
(BB92220a) are shown.

HORIZON, LOCALITY AND DISCUSSION. Several populations of O. ostentata were recovered from
beds H4-H7 and H38-H40 in Afon Twrch. The H4 population was compared with a population
from bed GGlb at Gelli-grin (type locality); no significant differences could be observed in the
three allometrically controlled morphological features that were tested, the outline of the pedicle
valve (P>10%), the outline of the brachial valve (P>10%) and the shape of the cardinalia
(0%>P>5%). (For statistics see Lockley 1977 : fig. 12.2). For a full description of this species
see Williams (1963 : 405); it is discussed here to draw attention to its early appearance in pre-
Frondderw beds.

Genus BANCROFTINA Sinclair, 1946

Bancroftina sp.
Figs 46a, b

MATERIAL, LOCALITY AND DISCUSSION. A well-preserved internal mould of a Bancroftina brachial
valve (BB92221, length 13 mm, width 18 mm; cardinalia length 2:5 mm, width 5:5 mm) was
recovered from bed GGlIh above the Cymerig Limestone. This form is poorly represented in the
Lower Bala Group.

210 M. G. LOCKLEY

Family LINOPORELLIDAE Schuchert & Cooper, 1931
Genus SALOPIA Williams, 1955
Salopia sp.
Fig. 44
MATERIAL, LOCALITY AND DISCUSSION. This internal mould of a brachial valve (BB92222), from

bed H6 below the Frondderw Ash, is one of several found in association with species which
make an early appearance in the Lower Bala Group at about this horizon.

Suborder TRIPLESIIDINA Moore, 1952
Superfamily TRIPLECIACEA Schuchert, 1913
Family TRIPLECIIDAE Schuchert, 1913

Genus TRIPLESIA Hall, 1959
Triplesia maccoyana Davidson, 1869, emended

1852 Hemithyris depressa (J. de C. Sowerby); M’Coy in Sedgwick & M’Coy: 201.

1869 Triplesia ? Maccoyana Davidson: 199; pl. 24, fig. 29.

1978 Triplesia? maccoyana Davidson; Cocks: 86, 187.

DESCRIPTION. Small dorsibiconvex, globular, plicate Triplesia with pedicle valve averaging 96%
as long as wide in 5 specimens (range 80-108 %) and 22 °% as deep as long in 4 specimens (range
15-27%); dorsal valve averaging 87 % as long as wide in 13 specimens and 45 % as deep as long in
12 specimens; ventral sulcus and corresponding dorsal fold arising between the 2 and 3 mm
growth stages and averaging 62% as wide (wavelength) as valve (range 56-67 °%) and 30% as high
(amplitude) as wavelength (range 21-40°%) at the commissure of 4 smaller valves between 3-6
and 5:0 mm in width, deepening to respective average width : wavelength and amplitude : wave-
length ratios of 59% (range 55-63°%) and 44% (range 39-47%) at the commissure of 4 larger
valves between 5:5 and 7:2 mm in width; ventral umbo pointed, overhanging incurved dorsal
umbo and with short narrow curved apsacline interarea divided by elongate pedicle groove;
surface smooth except for fine concentric growth lines numbering about 10 per mm beyond the
3 mm growth stage. Interior of both valves unknown.

MATERIAL AND LOCALITY. Lectotype (selected Cocks 1978: 86) articulated valves (SMA.42436),
length 5-6 mm, width 7:0 mm, and 19 other paralectotypes (SMA.42437-48) from a limestone
lens in the Allt Ddu Mudstones, Bryn Bedwog Quarry, near Bala, Gwynedd (grid. ref. SH
931329).

Discussion. The species is known only from M’Coy’s original material and its importance as an
element of the Lower Bala Group fauna is difficult to evaluate. It is, however, the first Triplesia
species recorded in the Ordovician of north Wales; Williams (1974) recorded a contemporary
Triplesia sp. from the Soudleyan of Salop. There is also an earlier species, Triplesia edgelliana
(Davidson), from the Upper Llanvirn of the Llandeilo area (Lockley & Williams, in press);
T. maccoyana compares most closely with this, differing only in respect of its significantly deeper
dorsal valve (0:05< P<0-02). Relevant statistics for the dorsal valve length (I), width (w) and
depth (th) are as follows: (n = 13) | mm (var. 1) 5-14 (1-273), W mm (var. w) 5-91 (2-185), r =
0-9077, a (var. a) 1-3103 (0:0275); (n = 12) 1 mm (var. 1) 5:32 (0-938), th (var. th) 2:42 (0-552),
r = 0:8066, a (var. a) 0-7675 (0-0206).

Order STROPHOMENIDA Opik, 1934
Superfamily PLECTAMBONITACEA Jones, 1928
Family LEPTESTIIDAE Opik, 1933
Subfamily LEPTESTIINAE Opik, 1933

Genus PALAEOSTROPHOMENA Holtedahl, 1916

Palaeostrophomena canalis sp. nov.
Figs 47a, b, 48-52

DIAGNOsIS. Subquadrate, gently biconvex to biplanate Palaeostrophomena with pedicle valve

CARADOC FAUNAL ASSOCIATIONS ANI

characterized by well-developed mantle canal system arising anteromedially and restriction of
genital markings to posterolateral part of valve.

NAmE. With reference to the mantle canal system.

DESCRIPTION. Pedicle valve 85-90% as long as wide with interior characterized by elongate,
narrow, divergent diductor scars flanking narrow, small anterior median adductor scars. Muscula-
ture otherwise poorly defined and intricately related to well-defined mantle canal system.

Interarea long apsacline, dental lamellae absent, but divergent false dental plates present.
Narrow delthyrium open medially but covered laterally by pseudodeltidium. Up to 12 mantle
canals branching radially into at least 30 tributary canals at the commissure. At least seven
concentric rugae (wavelength 0-2 mm) developed posterolaterally at an acute angle to the hinge.
Omament very fine, unequally parvicostellate; at least seven ribs arise in umbonal area defining
sectors bisected by ribs arising 3-4 mm from umbo, and these in turn define sectors bisected by
ribs arising 5-6 mm from umbo. Brachial valve unknown.

TYPE MATERIAL. Measurements in mm.

length width
Holotype, internal mould of pedicle valve, BB92223 . ; ; (15) (16-5)
Paratype, internal mould of pedicle valve, BB92224 . ‘ : (13) (18)
Paratype, internal mould of pedicle valve, BB92225 ; . (12) (12)
Paratype, internal mould of pedicle valve, BB92226 . . : (9) (12)
Paratype, internal mould of pedicle valve, BB92227 : , (9) (15)
Paratype, external mould of pedicle valve, BB92228 P : (9) (11)

TYPE HORIZON AND LOCALITIES. Holotype from sample GG1b in the Gelli-grin type section;
paratypes from samples R27 and R28 in the Rhiw March section.

DIscussION. This species is quite distinct from P. magnifica Williams, which is known from the
Derfel Limestone. The lack of well-developed genital markings and the closely-spaced mantle
canals are its most distinctive features. Cooper (1956 : 703) noted the complete absence of well-
preserved brachial valves in a large population of P. angulata and attributed this to their fragility.
Although similar absence of brachial valves of P. canalis is noted here, explanations are not
proposed.

Family SOWERBYELLIDAE Opik, 1930
Subfamily SOWERBYELLINAE Opik, 1930

Genus ANJISOPLEURELLA Cooper, 1956
Anisopleurella cf. multiseptata (Williams, 1955)
Fig. 53

Discussion. A well-preserved specimen of a pedicle valve internal mould (BB92229) from bed
GG2b represents the first record of Anisopleurella in the Gelli-grin Formation. The large, paired
diductor scars are well displayed and the fragmentary remains of external moulds which were
also recovered reveal the essentially smooth exterior of the shell which is ornamented only by
widely-spaced primary costae. The specimen compares closely with A. multiseptata Williams
from the Derfel Limestone (Williams in Whittington & Williams 1955 : 416).

Subfamily AEGIROMENINAE Havlicek, 1961
Genus SERICOIDEA Lindstrém, 1953
Sericoidea abdita Williams, 1955, emend. herein

DIAGNOsIS. Sericoidea with variable arrangement of paired or single strong lateral septules and a
well-defined median septum all extending into the anterior half of the valve; dorsal muscle
platform bilobed.

212 M. G. LOCKLEY

Sericoidea abdita complicata subsp. nov.
Figs 54-59

DiaGnosis. Differs only from the nominate subspecies in the variable development of septules
and the presence of dorsal muscle scars.

NAME. ‘Complicated’.

DESCRIPTION. Semicircular, planoconvex Sericoidea with mean length/width ratio of 54-4%
and 52:7% for pedicle (N=199) and brachial (N=91) valves respectively. Ornamentation
consisting of fine, differentially thickened, radial costae and costellae with a mean of 12-4 per mm
(N=56) found anteromedially. Dorsal septules arranged in arcs over two-thirds of the length
of the brachial valve anterior of the umbo. Septules are generally but not invariably arranged in
arcuate, single or double rows, or both, on either side of the median septum. The mean number of
septules per row (as counted for the longest row) is 5-8 (N=25). Dorsal adductor scar bilobed,
extending anteriorly for less than half the length of the valve and laterally for over one-third
the width of the valve.

TYPE MATERIAL. Measurements in mm.

length width

Holotype, internal mould of brachial valve, BB92231 Fs ae ; 5:0 2:8
Paratype, internal mould of brachial valve, BB92230 . : : - -

Paratype, exterior of pedicle valve, BB92232 6 é : : 5:3 3°5
Paratype, exterior of pedicle valve, BB92233 : : : . 4-6 3-0
Paratype, internal mould of brachial valve, BB92234 . : : - -

Paratype, internal mould of brachial valve, BB92235 . : : 5:0 B32)
Paratype, exterior of pedicle valve, BB92236 ; 3 : 5 7-0 4-0

TYPE HORIZON AND LOCALITY. Nod Glas Formation, Rhiw March: holotype from bed R47
and paratypes from beds R52 and RS3.

Discussion. The genus Sericoidea is abundantly represented in the Nod Glas. The largest popula-
tions were recovered from beds R47, R48, R52 and R53; full statistical details are given in Fig.
45B. Estimates of a, the growth ratio for samples affected by allometry, have been derived from
the data in Fig. 45B and tests show that none of the populations differs significantly from any
other (P>0-1). Furthermore, they do not differ significantly from S. restricta from the Sularp
Shale and from S. aff. restricta from the Craighead Limestone; nor do they differ significantly
from S. aff. abdita from the Balclatchie Mudstones either at Laggan Burn or Byne Hill (P>0-1,
for all tests). Even when using growth ratios not corrected for allometry, no significant difference

Figs 46a,b Bancroftina sp.(p. 209). Gelli-grin Formation, Gelli-grin Farm, Bala. BB92221. Internal
mould and latex cast of a brachial valve, both x 2°5.

Figs 47-52 Palaeostrophomena canalis sp. nov. (p. 210). Gelli-grin Formation. Figs 47a, b, Gelli-grin
Farm, Bala. Holotype, BB92223. Internal mould and latex cast of a pedicle valve, x 2°5. Figs 48-52,
Rhiw March, near Llanymawddwy. Figs 48-50, Paratypes, BB92226, BB92225, BB92227. Internal
moulds of pedicle valves, all «2. Fig. 51, Paratype, BB92224. Internal mould of a pedicle valve,

«2-5. Fig. 52, Paratype, BB92228. External mould of a pedicle valve, x 2°5.

Fig. 53. Anisopleurella cf. multiseptata (Williams) (p. 211). Gelli-grin Formation, Maes-Meillion,
near Llangower. BB92229. Internal mould of a pedicle valve, x 2-5.

Figs 54-59 Sericoidea abdita complicata subsp. nov. (above). Nod Glas Formation, Rhiw March,
Llanymawddwy. Fig. 54, Holotype, BB92231. Internal mould of a brachial valve, x 7:5. Figs 55—57,
Paratypes, BB92230, BB92235, BB92234. Internal moulds of brachial valves, x12, x6 and 7-5
respectively. Figs 58-59, Paratypes, BB92233, BB92232. Exteriors of pedicle valves, both x 6.

Figs 60-62 Bimuria dyfiensis sp. nov. (p. 215). Gelli-grin Formation, Rhiw March, near Llany-
mawddwy. Figs 60-61, internal moulds of pedicle valves, 2-5. Fig. 60, Holotype, BB92237.
Fig. 61, Paratype, BB92239a. Figs 62a, b, Paratype, BB92238a. Posterior and lateral views of a
pedicle valve, both x 2-5. See also Figs 64-65.

CARADOC FAUNAL ASSOCIATIONS 213

214 M. G. LOCKLEY

XESS EE VESTED

costellae per mm

papal Tt te
SU ES) A
EC EEED MN ECVECS) CUES) CSIECS) EC TC)
Pases} pal [eahopapofel | | [se
Pesef [ [pe] polafal | ho | ow
EE ND)

9-72] 091-0 +0] 790

|
'
C =
Fig. 63 A, frequency of counts of costellae per mm occurring at the anteromedian margins of
Sericoidea abdita complicata subsp. nov. B, pooled data for costellae frequency in S. aff. abdita
from Laggan Burn and Byne Hill and in S. abdita complicata subsp. nov. C, the distribution of
various types of lophophore platforms, with from 1 to 10 septules either side of the median
septum (MS) in S. aff. abdita from Laggan Burn (number of individuals in brackets) and S. abdita
complicata subsp. nov. from the Nod Glas (number of individuals not in brackets).

can be detected between S. abdita complicata and Chonetoidea radiatula (Barrande) from the
Ashgill of Pomeroy (P>0-1). This evidence highlights the similarity in the shape of these related
aegiromeninids.

Williams (1963 : 188-190) measured the frequency of costellae per mm at the anteromedian
margins of Sericoidea and found that S. aff. abdita from the Balclatchie Mudstones was signific-
antly different from S. aff. restricta from the Craighead Limestone. In this study similar measure-
ments on the frequency of costellae at the anteromedian margin in S. abdita complicata were
compared with measurements obtained for S. aff. abdita, using a y? test with cell groupings
identical to those used by Williams (1963 : 189). No significant difference was detected between
the two populations (0:2>P>0-1); the data are given in Fig. 63.

However, a comparison between the different types of septule arrangement in S. aff. abdita
and S. abdita complicata reveals that the latter subspecies was significantly different (7? test,
P<0-001) in having longer rows of single and paired septules of comparable size; see Fig. 63.

CARADOC FAUNAL ASSOCIATIONS 215

Since there is a high degree of variability in the disposition of septules in the S. abdita complicata
population the differences are considered subspecific.

Although attempts have been made by Williams (1962) and the author to define Sericoidea
populations quantitatively, a certain element of qualitative judgement attaches to whether smaller
or larger septules are considered important as representatives of well-defined or less well-defined
rows. According to Williams (1962 : table 46) septule rows never contain more than three septules
(either in S. restricta or in S. aff. abdita). Similarly, S. abdita was diagnosed as having ‘up to
three pairs of strong lateral septules’ (Williams in Whittington & Williams 1955 : 418). However,
whilst acknowledging that palaeontologists might differ as to what constitutes a septule rather
than a tubercle (the specimens in Williams 1974 : pl. 24, figs 8, 9 and Williams 1962 : pl. 18, fig. 8
might for example be considered to display at least four septules in each lateral row) the differences
between S. cf. abdita, S. aff. abdita and S. abdita complicata are considered significant because
all measurements and counts of morphological features were estimated in the same way.

Family BIMURIDAE Cooper, 1956
Genus BIMURIA Ulrich & Cooper, 1942

Bimuria dyfiensis sp. nov.
Figs 60, 61, 62a, b, 64a, b, 65

D1AGNosIs. Concavo-convex, small, slightly sulcate Bimuria with pedicle valve umbo strongly
incurved and overlapping dorsal interarea, teeth simple.

Name. From the Dyfi river.

DESCRIPTION. The pedicle valve interior morphology is well defined. Variably developed, narrow,
divergent diductor scars extend anterolaterally for approximately three-quarters of the length of
the valve, enclosing less well defined, radial adductor and mantle canal impressions postero-
medially. Simple teeth developed laterally for about one-third of the width of the valve but short
anteromedially. Slight development of pedicle valve sulcus evident from broad indentation of
anterior commissure.

Brachial valve interior unknown; exterior essentially smooth in the only known specimen.
Ventral exterior poorly known, essentially smooth, comae absent or very indistinct, resembling
faint concentric growth lines where present.

TYPE MATERIAL, HORIZON AND LOCALITY. Holotype, internal mould of pedicle valve (BB92237)
from bed R28 (length 12 mm, width 16 mm). Paratypes, internal and external moulds of pedicle
valves (BB92238a, b-BB9224 1a, b) and internal moulds of pedicle valves (BB92242 and BB92243)
and a brachial valve (BB92244), all from bed R28.

The above specimens, all from bed R28 in the Gelli-grin Formation at Rhiw March, are
distorted so that measurements are inaccurate; but the mean size of this population was clearly
small, only two specimens being slightly wider than 12 mm.

DISCUSSION. The morphological features described above suggest that this species cannot be
assigned to any of the Scoto-Irish species (B. cf. buttsi Cooper, B. youngiana Davidson, B.
youngiana recta Williams and B. cf. youngiana recta). The Welsh specimens apparently belonged
to a population with a smaller mean size than the two former Scoto-Irish species populations
(and the latter subspecies population). More significantly, however, the lack of any comae or
other distinctive concentric ornamentation must be regarded as an important morphological
difference. This lack of ornamentation is not simply the result of the small size of the Welsh
specimens; Williams (1962 : 174-175) reported comae originating at 4-6 mm from the ventral
umbo in all three of the Scottish species he described. Although the length of the Welsh specimens
is not determined precisely, if one can assume, following Williams (1962) and Mitchell (1977),
that Bimuria is at least two-thirds to three-quarters as long as wide, then all the Welsh specimens
must be 6 mm or more in length. B. dyfiensis represents the first known occurrence of the genus
Bimuria in Wales; since it is morphologically distinct, in respect of its essentially smooth external
ornament, from all other known species in Britain, it is given specific recognition.

216 M. G. LOCKLEY

Order PENTAMERIDA Schuchert & Cooper, 1931
Suborder SYNTROPHIIDINA Ulrich & Cooper, 1936
Superfamily PORAMBONITACEA Davidson, 1853
Family PARASTROPHINIDAE Ulrich & Cooper, 1938

Genus PARASTROPHINELLA Schuchert & Cooper, 1931
Parastrophinella brenchleyi sp. nov.
Figs 66-69
DiaGnosis. Large, subpentagonal, biconvex Parastrophinella with vestigial fold and sulcus and
commonly 12 costae, with a wavelength of 1-:0-1:5 mm, ornamenting anteromedian part of the
Shell.

Name. For Dr P. J. Brenchley.

DescrIPTION. Unequally biconvex, rostrate Parastrophinella with a subpentagonal outline. 80%
as long as wide with pedicle valve about 30% as deep as long. Transverse profile convex with
steep lateral slopes; longitudinal profile unevenly convex to anteriorly geniculate in larger
specimens. Faintly and sporadically developed low dorsal fold with at least 4 costae and shallow
ventral sulcus with 3 costae. Twelve variably developed ventral costae are characteristically
angular and well developed anteromedially (being 1-0-1-5 mm in width and amplitude) but are
rounded and indistinct posterolaterally. Variably developed concentric corrugations up to 4 per
mm anteromedially and anterolaterally are faint or absent in posteromedian part of shell.

Ventral interior with spondylium which is sessile posteriorly but supported anteriorly by
sporadically developed medium septum extending forward for an average of 49 % of the length of
three pedicle valves. Muscle scars very faint or absent. Dorsal interior unknown.

TYPE MATERIAL. Measurements in mm.

length width
Holotype. Internal mould of a pedicle valve, NMW 77.11G.24 (10:0) (15-0)
Paratype. Internal mould of a pedicle valve, NMW 77.11G.25 : (7:0) (9-0)
Paratype. Internal mould of a pedicle valve, NMW 77.11G.26 ; (10-5) (10-5)
Paratype. Internal mould of a pedicle valve, NMW 77.11G.27 F (9-0) (11-0)

Paratype. Internal mould of a pedicle valve, NMW 77.11G.28 : - -
(Measurements given here indicate respective mean length and width values of 9-12 and 11-37 mm)

TYPE HORIZON AND LOCALITY. Bryn Beds (Lower Longvillian) exposed in small overgrown quarry
55 m east of Pandy Quarry (SJ 202363).

Figs 64-65 Bimuria dyfiensis sp. nov. (p. 215). Gelli-grin Formation, Rhiw March, near Llany-
mawddwy. Figs 64a, b, Paratype, BB92241a. Internal mould of a pedicle valve, x 4; corresponding
latex cast, 2-5. Fig. 65, Paratype, BB92240a. Internal mould of a pedicle valve, x4. See also
Figs 60-62.

Figs 66-69 Parastrophinella brenchleyi sp. nov. (above). Bryn Beds, Pandy Quarry, near Glyn
Ceriog. Fig. 66, Holotype, NMW.77.11G.24. Internal mould of a pedicle valve, «4. Figs 67-69,
Paratypes, NMW.77.11G.26, NMW.77.11G.25 and NMW.77.11G.27 respectively. Internal
moulds of pedicle valves, all » 4.

Figs 70-76 Protozyga musculosa sp. nov. (p. 218). Nod Glas Formation; all internal moulds.
Figs 70, 74-76, Rhiw March, near Llanymawddwy. Fig. 70, Paratype, BB92249. Pedicle valve,
x8. Fig. 74, Paratype, BB92246. Pedicle valve, x 16. Fig. 75, Paratype, BB92252. Brachial valve,

<10. Fig. 76, Paratype, BB92247. Pedicle valve, x 16. Figs 71-73, Nant Tan y Bwlch, near Bwlch
y Groes. Fig. 71, Paratype, BB92253. Pedicle valve, x 9. Fig. 72, Holotype, BB92245. Pedicle valve,
x12. Fig. 73, Paratype, BB92254. Pedicle valve, x9.

Figs 77a, b Cyclospira aff. bisulcata (Emmons) (p. 219). Cymerig Limestone Member, Rhiw March,
near Llanymawddwy. BB92255. Ventral and dorsal views of the exterior of an articulated specimen.
both x6. See also Figs 78-82.

CARADOC FAUNAL ASSOCIATIONS PAT)

218 M. G. LOCKLEY

Discussion. This material, collected by Dr P. J. Brenchley from a single locality, has hitherto
been described only briefly and informally (Brenchley 1966 : 242; figs 161-162). Parastrophinella
is known from Scotland (Williams 1962), from Wales (MacGregor 1961, Addison 1974) and from
the Welsh Borderland (Williams 1974). The Scottish species P. youngi (Reed) from the Craighead
Limestone (Caradoc) differs from P. brenchleyi in being characteristically tumid with poorly-
developed costae primarily associated with plication. Similarly, P. parva MacGregor (1961 : 197)
from the Llandeilo rocks of the Berwyns is small, tumid and lacks well-developed costae; Addison
(1974 : 47) also recorded this species in rocks of early Caradoc age near Narbeth(Pembrokeshire).
However, the other smaller species, P. costata MacGregor (1961 : 199) from the Llandeilo of the
Berwyns, resembles P. brenchleyi in outline, length/width ratio and spondylial arrangement, but
differs in being considerably smaller (mean size) and having a larger number of costae (13-22).
P. musculosa Williams (1974 : 151) from the Spy Wood Grit, Salop, resembles P. brenchleyi
sp. nov. in ‘commonly’ having 11-13 costae (mean c. 12) and exhibiting a poorly-developed fold
and sulcus and a similar spondylial arrangement; however, it differs in being smaller than P.
brenchleyi (55 °% of mean width) and relatively but not significantly wider (25>P> 10). A mean
width/length ratio of 0-93 for 4 brachial valves of P. musculosa implies a smaller value for the
corresponding, relatively longer pedicle valves (say < 0-90); this differs considerably from the ratio
of 1-25 for 4 pedicle valves of P. brenchleyi. Although allometric growth might account for a
relative increase in width with size, even in the absence of more material the differences in size,
shape and ornament between P. musculosa and P. brenchleyi are sufficient to merit the systematic
recognition of the latter species. Williams (1974 : 152) also described Parastrophinella sp. from the
Hagley Volcanics; although resembling P. brenchleyi in its sub-pentagonal outline and large size it
differs from this form in having about 20 well-developed costae.

Order SPIRIFERIDA Waagen, 1883
Superfamily ATRYPACEA Gill, 1871
Family ATRYPACEA Gill, 1871

Genus PROTOZYGA Hall & Clarke, 1893

Protozyga musculosa sp. nov.
Figs 70-76

DIAGNOsIS. Small subcircular to subspherical Protozyga with well-developed dental plates and
pedicle valve musculature.

Name. With reference to the well-developed musculature.

DESCRIPTION. Small subcircular to subeliptical or subpentagonal ventribiconvex Protozyga as
wide as long, with long subparallel to slightly divergent dental plates extending anteromedially
for between half and two-thirds the length of the pedicle valve and enclosing equally long diductor
muscle scars best developed in larger specimens. External features poorly known; ornamentation
smooth but nature of anterior arc of commissure unknown.

Brachial valve interior with posterolateral socket plates and long median septum arising at a
point anterior to the hinge line and extending anteriorly for most of the length of the valve.

Statistics for three measured pedicle valve paratypes; mean length (2:73 mm), mean width
(2:73), r (0:9736), a (1:077) and b (—0-:1415).

TYPE MATERIAL, HORIZON AND LOCALITIES. Holotype, internal mould of pedicle valve, BB92245
(length 3 mm, width 3 mm). From bed TB19 in the Nod Glas at Nant Tan y Bwlch. Paratypes,
internal moulds of pedicle valves, BB92246-BB92251 and BB92253-BB92254, and a single
brachial valve, BB92252. From beds TB19 and R54 in the Nod Glas at Rhiw March.

DiscussION. This occurrence is the first record of Protozyga in Wales; related forms including
P. diversa (Reed), P. rotunda Cooper and P. perplexa Williams from Girvan and P. cf. perplexa
and P. cf. diversa from Pomeroy are unlike the Welsh species both in the ventral arrangement of
dental plates and diductor muscle scars and in the posterior origin of the brachial valve median
septum.

CARADOC FAUNAL ASSOCIATIONS 219

Superfamily DAYIACEA Waagen, 1883
Family DAYIIDAE Waagen, 1883
Subfamily CYCLOSPIRINAE Schuchert, 1913

Genus CYCLOSPIRA Hall & Clarke, 1893

Cyclospira aff. bisulcata (Emmons, 1842)
Figs 77-82

DiaGnosis. Unequally biconvex, subpentagonal Cyc/ospira with rounded anterior margin.

DESCRIPTION. Unequally biconvex Cyclospira with pedicle valve nine-tenths as wide as long and
brachial valve about as wide as long. Pedicle valve strongly convex in median part of valve with
slight development of anteromedian sulcus and flat or concave posterolateral flanks or ‘wings’
subparallel to commissural plane.

External ornament mainly smooth but with fine concentric growth lines developed anteriorly
and fine radial striations emphasized by slight exfoliation of shells. Pedicle interior characterized
by well-developed ‘shoelifter process’ partly anterior to ventral muscle field.

Brachial valve sulcate with slight development of median plication dividing anterior part of
sulcus. Interior characterized by median septum originating just anterior of medially cleft
hinge line and extending anteriorly for at least three-quarters of the valve length. Thin socket
plates present.

MATERIAL. Measurements in mm.

——_P V——~ —— BV———~.

length width length width

Complete articulated specimen, BB92255_ . : : 5:0 S25) 4:5 5:5
Complete articulated specimen, BB92256 . : ; 40 3-5 3:9 3°5
External of pedicle valve, BB92257__.. : ¢ 5-6 5:0 - =
Latex cast of articulated specimen, BB92258 : é 4-6 4-4 4:2 4-4
Complete articulated specimen, BB92259 . ; : 4-7 4-6 41 4-6
Internal and external mould of pedicle eS

BB92260a, b ; é @:0) @) - =
Internal mould of brachial valve, BB92261_ : - (3-0)

Statistics of length and width for 12 pedicle and 7 sail! valves are given in Fig. 45C.

HORIZON AND LOCALITIES. From bed R53 in the Cymerig Limestone at Rhiw March and bed
TB20 in the Tan y Bwlch section.

DISCUSSION. The material recovered in this study affords an excellent opportunity to elaborate on
the description of the Cyclospira sp. which Williams (1963 : 469) described as ‘reminiscent’ of
C. bisulcata (Emmons). C. bisulcata is one of six species described by Cooper (1956); of these
only a form related to C. ? Jonga (Cooper) has previously been described from British successions
(Williams 1962 : 251; pl. 23). C. carrickensis (Reed) and C. nana (Davidson) have also been
described from Ordovician rocks by Williams (1962) and Mitchell (1977) respectively. The
specimens from north Wales are unlike C. ? /onga, which has an almost triangular outline with
incipient plication of the anterior commissure (Cooper 1956 : pl. 142.1). They are also unlike
C. nana in having a longer median septum and thinner socket plates. They differ from C. carrick-
ensis in not having an emarginate anterior commissure (Williams 1962 : 250) but are otherwise
similar in exterior morphology.

The present C. aff. bisulcata bears a general resemblance to both C. quadrata Cooper and
C. preciosa Cooper, but these two American species lack any median plication in the dorsal sulcus
and show little or no ventral sulcation. C. parva Cooper and particularly C. sulcata Cooper are
considerably more elongate than the present form. The species described by Cooper are hard to
differentiate on external morphology; internal features are not illustrated, though C. parva, C.
quadrata and C. ? longa are all known to have a dorsal median septum. The species based on
British specimens (including C. aff. bisulcata) are probably better known in their internal morph-
ology. The designation of this species as C. aff. bisulcata supports the suggestion of Williams
(1963) that the Welsh Cyclospiridae belong to a group closely related to C. bisulcata.

220 M. G. LOCKLEY

Superfamily unknown

? Spiriferide, gen. indet.
Fig. 83

DESCRIPTION AND DISCUSSION. Pedicle valve about as wide as long, with well-defined ‘shoelifter
process’, which differs from C. aff. bisculcata in having paired raised scars in median part of valve
between flanks of anterior part of process. External ornament of (unfigured) counterpart smooth.
Posterior lateral flanks of valves show more ‘wing-like’ extension than in C. aff. bisulcata, but
the possibility that this specimen is closely related to that form cannot be ruled out.

MATERIAL AND LOCALITY. Pedicle valve internal and external moulds (BB92262a, b), from loc.
615a in beds above the Cymerig Limestone on Pen y Cefn Coch.

Other phyla

In addition to the brachiopod taxa described above, representatives of other phyla, notably the
Arthropoda (Trilobita) and Mollusca (Bivalvia), are also recorded in the Lower Bala Group for
the first time. These include four trilobite genera and five bivalve genera which are figured below
but not accompanied by full systematic descriptions. Brief discussion on the significance of these
newly recorded genera is included where appropriate.

Phylum ARTHROPODA Siebold & Stannius, 1845
Class TRILOBITA Walch, 1771
Family PHILLIPSINELLIDAE Whittington, 1950

Genus PHILLIPSINELLA Novak, 1886

Phillipsinella sp.
Figs 84-87

MATERIAL AND LOCALITY. Three pygidia (It.14294-6) and four cephalic (glabellar) fragments
(It.14297-300) from bed R53 in the upper part of the Cymerig Limestone at Rhiw March.

Discussion. These specimens, from beds of ‘presumed’ Upper Longvillian age, are the oldest
known Phillipsinella specimens in Britain (Ingham, personal communication 1977); they are the
first record of this genus in the Caradoc Series of north Wales. The phylogeny of the genus is
discussed in Bruton (1976).

Figs 78-82 Cyclospira aff. bisulcata (Emmons) (p. 219). Figs 78-81, Cymerig Limestone Member,
Rhiw March, near Llanymawddwy. Fig. 78, BB92258. Posterodorsal view of latex cast of an
articulated specimen, <7°5. Fig. 79, BB92257. Exterior of a pedicle valve, x5. Fig. 80, BB92261.
Internal mould of a brachial valve, 8. Figs 81a, b, BB92256. Ventral and dorsal views of the
exterior of a broken articulated specimen, both x5. Fig. 82, Nod Glas Formation, Nant Tan y
Bwlch, near Bwlch y Groes. BB92260. Internal mould of a pedicle valve, x 5. See also Figs 77a, b.

Fig. 83 ? Spiriferide, gen. indet. (above). Gelli-grin Formation, Pen y Cefn Coch. BB92262a.
Internal mould of a pedicle valve, x5.

Figs 84-87 Phillipsinella sp. (above). Cymerig Limestone, Rhiw March, near Llanymawddwy.
Figs 84a, b, It.14297. Oblique left-lateral and left-lateral views of a cranidium, x16. Fig. 85,
It.14298. Oblique left-lateral view of a cranidium, x16. Fig. 86, It.14299. Dorsal view of a
small cephalon, x 15. Fig. 87, It.14294. Dorsal view of a pygidium, x 16.

Figs 88-89 Lonchodomas sp. (p. 222). Cymerig Limestone Member, Rhiw March, near Llany-
mawddwy. Fig. 88, It.14301. Dorsal view of plasticine cast of cranidium, x3. Fig. 89, It.14302.
Dorsal view of cranidium, x 4.

Figs 90-91 Sphaerocoryphe sp. (p. 222). Cymerig Limestone Member, Rhiw March, near Llany-
mawddwy. Dorsal views of cranidia, both x6. Fig. 90, It.14307. Fig. 91, It.14308.

Fig. 92 Cyphoproetus sp. (p. 222). Derfel Limestone, Nant Aber Derfel. It.14309. Dorsal view of
cranidium, x15.

CARADOC FAUNAL ASSOCIATIONS 221

222: M. G. LOCKLEY

Family RAPHIOPHORIDAE Angelin, 1854

Genus LONCHODOMAS Angelin, 1854
Lonchodomas sp.
Figs 88-89

MATERIAL AND LOCALITIES. Five cephalic specimens (It.14301—5) from bed R52 at Rhiw March
and a single cephalic specimen (It.14306) from bed R49 are the first record of this genus in the
Lower Bala Group.

Family CHEIRURIDAE Salter, 1864
Genus SPHAEROCORYPHE Angelin, 1854

Sphaerocoryphe sp.
Figs 90-91

MATERIAL AND LOCALITY. Two spherical, pustulose glabellae (It.14307-8) from bed R52 in the
Cymerig Limestone at Rhiw March are the first record of this genus in the Lower Bala Group.

Family PROETIDAE Salter, 1864
Genus CYPHOPROETUS Kegel, 1927

Cyphoproetus sp.
Fig. 92

MATERIAL AND LOCALITY. The internal and external mould of a cranidium (It.14309) from the type
Derfel Limestone at Nant Aber Derfel (SH 850395) is the first record of this genus in the Ordo-
vician of Wales.

Phylum MOLLUSCA
Class BIVALVIA Linné, 1758

Introduction

Recent studies by Brenchley (1966), Pickerill (1974), Hurst & Hewitt (1977), Hurst (personal
communication 1978) and Pickerill & Brenchley (1979) show that the Mollusca, and in particular
the Bivalvia, are, after the Articulata (Brachiopoda) and Trilobita, one of the most important
classes of organism found in the shelly facies of the Caradoc Series in Wales and the Welsh
Borderland. But their importance has not been emphasized, and in Wales, largely because of
poor preservation and indistinctive, mainly external morphology, the class has received little
attention. Since the above authors record the occurrence and distribution of various bivalve
genera in Wales and the Welsh Borderland, distinctive forms recovered from the Lower Bala
Group are noted here.

Although Ordovician Bivalvia from eastern North America (Bretsky 1970, Pojeta 1971) are
generally more common in contemporary Middle to Upper Ordovician deposits of the Appa-
lachians (Bretsky, personal communication 1977), and the overall distribution of Anglo-Welsh
Caradoc Bivalvia has yet to be outlined in any detail, the occurrence of congeneric taxa in both
areas implies reasonable prospects for future correlations and comparisons.

Figured material

The forms figured here are assigned to genera but not described in any detail. In general the
material from the Bala area is less abundant and well-preserved than that from the Berwyns and
Salop. Any attempt at full systematic description of the Bala material would prove unrewarding
without detailed consideration of other, more abundant material in other parts of the Anglo-
Welsh region, but full locality details are given for the material discussed.

CARADOC FAUNAL ASSOCIATIONS 223

Order NUCULOIDA Dall, 1889
Family PRAENUCULIDAE McAlester, 1969

Genus CARDIOLARIA Munier-Chalmas, 1876

? Cardiolaria sp.
Figs 93-95

MATERIAL AND LOCALITIES. Internal moulds of nuculoid bivalves from beds AD A (PL 4441-3)
and loc. 35 (PL 4440) in the middle part of the Allt Ddu Formation are provisionally assigned to
? Cardiolaria sp. The four numbered specimens have respective height and length measurements
(mm) as follows: 2:5 and 4-0; — and 3-5; 3-0 and 5:0; 3-5 and 5-5 (means 3-0 and 4-5). Dentition
is observed in the first two specimens.

Family MALLETIIDAE Adams & Adams, 1858
Genus NUCULITES Conrad, 1841

Nuculites sp.
Fig. 96

MATERIAL AND LOCALITY. A single phosphatized internal mould of an articulated specimen
(PL 4444), height 5-0 mm, length 8-0 mm, was recovered by Mr P. Magor from the locally-
developed limestone beds in the lower part of the Nod Glas at Pistyll Gwyn. The posterior part
of the hinge exhibits about 12 poorly-preserved, fine teeth each about 0:2 mm in width.

Discussion. The above two nuculoids, though uncommon in most of the succession, are the only
sessile benthos recovered from parts of the ‘faunal depletion zone’ in the middle part of the
Allt Ddu Formation. They are considered to be infaunal deposit feeders. In the absence of
better-preserved material Dr N. J. Morris, who has helped the author with the identification of
bivalve taxa, has suggested that forms resembling the genera Praenucula and Palaeosolen are also
present amongst the unfigured material collected from the middle part of the Allt Ddu Formation.

Order MODIOMORPHOIDA Newell, 1965
Family MODIOMORPHIDAE Miller, 1877

Genus MODIOLOPSIS Hall, 1847
Modiolopsis sp.
Figs 97, 98
MATERIAL, LOCALITIES AND DISCUSSION. Specimens (PL 4445 and PL 4446a, b) from beds at
loc. 928 (8990 2234) and from bed AD 3J, respectively, in the upper part of the Allt Ddu Formation
are assigned to the genus Modiolopsis. The Modiomorphidae are usually considered to have been
byssally attached, semi-infaunal or epifaunal suspension feeders.

Order ARCOIDA Stoliczka, 1871
Family CYRTODONTIDAE Ulrich, 1894

Genus CYRTODONTA Billings, 1858
Cyrtodonta sp.
Figs 99-101

MATERIAL, LOCALITIES AND DISCUSSION. Specimens from beds GG 1X (PL 4447), AD 0 (PL 4448)
and H13 (PL 4449) are assigned to this genus. The former, an articulated specimen, was found in
presumed life position with the umbones pointing downwards. Another specimen, PL 4450 from
bed CYG 5 (see Fig. 8, p. 179), is assigned to ? Cyrtodonta sp.

nN
i)
BSS

M. G. LOCKLEY

Order PHOLADOMYOIDA Newell, 1965
Family GRAMMYSIIDAE Miller, 1877

Genus CUNEAMYA Hall & Whitfield, 1875

Cuneamya sp.
Figs 102a—c

MATERIAL, LOCALITY AND DISCUSSION. A single, well-preserved specimen (PL 4451) of an articul-
ated Cuneamya was recovered from bed AD 3H in the Allt Ddu Formation. This genus is
commonly regarded as having had a burrowing, infaunal mode of life.

Discussion of the Bivalvia

Brenchley (1966), Pickerill (1974) and Pickerill & Brenchley (1979) record Modiolopsis, Gonio-
phorina, Arca, Psilonychia and Vlasta in their Howellites Community, Byssodesma, Psilonychia
and a pteriacean in the Dinorthis Community and Colpomya in the Dalmanella Community.
Pickerill (1974) also referred to Ambonychia, Ctenodonta sp. and M. modiolaris in the Berwyn
area. Most of the bivalves in the Caradoc Series of north and central Wales are confined to the
Soudleyan Stage.

Dr J. M. Hurst has collected a large number of bivalves from the Caradoc Series of Shropshire,
many of which are congeneric with, or otherwise closely related to, bivalves from Wales. The
author, who is currently examining some of this material, has collected Cyrtodonta from the lower
part of the Chatwall Flags (Soudleyan) and has also examined faunas rich in nuculoid bivalves
from the Acton Scott Beds. In the light of forthcoming publications on the Shropshire faunas,
however, further comment on the known and recently-discovered bivalvia is considered pre-
mature.

Class GASTROPODA Cuvier, 1797

Elles (1922) recorded the gastropod genera Cyclonema, Bellerophon (Sinuites), Lophospira and
Murchisonia in the Caradoc Series at Bala. In the present study representatives of the three
former genera have been recorded in 55 of the 196 samples corresponding to the main sections
(Figs 2, 8, 12-15). Dr J. S. Peel confirmed the identifications and noted that two distinct species of

Figs 93-95 ? Cardiolaria sp. (p. 223). Allt Ddu Formation, south side of Pen y Cefn Coch. Figs
93a, b, PL 4440. Left-lateral and anterior views of internal mould of an articulated specimen,
both x7-:5. Fig. 94, PL 4442. Internal mould of a right valve, x6. Fig. 95, PL 4443. Internal
mould of a right valve, x9.

Fig. 96 Nuculites sp. (p. 223). Limestone beds beneath the Cymerig Member in the Nod Glas
Formation at Pistyll Gwyn, near Llanymawddwy. PL 4444. Right-lateral view of internal mould
of an articulated specimen, x5.

Figs 97-98 Modiolopsis sp. (p. 223). Allt Ddu Formation. Fig. 97, Craig y Gath, near Llangower.
PL 4445. Right-lateral view of external mould of an articulated specimen, x 1-5. Fig. 98, north of
Graig Ty nant, near Llanymawddwy. PL 4446a. External view of a right valve, x 1-5.

Figs 99-101 Cyrtodonta sp. (p. 223). Figs 99-100, Allt Ddu Formation. Fig. 99, west of Gelli-grin
Farm, Bala. PL 4448. Left-lateral view of an articulated specimen, x2. Fig. 100, Craig y Gath,
near Llangower. PL 4447. External mould of a left valve, x 4:5. Fig. 101, Llaethnant Formation,
Afon Twrch, near Bwlch y Groes. PL 4449. External mould of a left valve, x 4-5.

Figs 102a-c Cuneamya sp. (above). Allt Ddu Formation, Craig y Gath, near Llangower. PL 4451.
Dorsal, anterior and left lateral views of the external mould of an articulated specimen, all > 2.

Fig. 103 ? Archinacella sp. (p. 226). Nod Glas Formation, Rhiw March, Llanymawddwy. PG 5022.
Left-lateral view of an internal mould of a complete specimen, x 6.

Fig. 104 Balacrinus basalis (M’Coy) (p. 226). Allt Ddu Formation, Rhiw March, Llanymawddwy.
E 67750. External mould of cup remains and attached arm and pinnule remains, x 2.

Fig. 105 Stenaster obtusus (Forbes) (p. 226). Allt Ddu Formation, Ty nant Farm, south of
Llanuwchllyn. E 53698. Internal mould of a complete specimen, x 4.

225

CARADOC FAUNAL ASSOCIATIONS

226 M. G. LOCKLEY

Cyclonema are found, in the Allt Ddu and Gelli-grin formations respectively. Material from this
study is not figured here.

Pickerill & Brenchley (1979) record Cyclonema, Cyrtolites, ? Seelya, Sinuites and Bucanopsis
from the Howellites Community, and Lophospira and Murchisonia in the Dinorthis and Dalmanella
Communities respectively. Pickerill (1974) also referred to Bucania sp. and Kokenospira in the
Lower Cwm Rhiwarth Siltstones.

Class MONOPLACOPHORA Wenz in Knight, 1952
Family ARCHINACELLIDAE Knight, 1956

Genus ARCHINACELLA Ulrich & Scofield, 1897

? Archinacella sp.
Fig. 103

A single asymmetrically conical shell mould (PG 5022) from bed R45 at the base of the Nod Glas
Formation resembles Archinacella and is provisionally assigned to this genus.

Class CEPHALOPODA Cuvier, 1797

MATERIAL AND DISCUSSION. Although representatives of the Cephalopoda are recorded in 33 of
the 196 samples corresponding to the main sections, most material 1s poorly preserved. However,
a single well-preserved specimen (C 81324) from bed H40 (Fig. 4, p. 176) is assigned to ? Ortho-
ceras and deposited with the other material figured here.

Phylum ECHINODERMATA
Class CRINOIDEA Miller, 1821
Subclass CAMERATA Wachsmuth & Springer, 1885
Family ARCHAEOCRINIDAE Moore & Laudon, 1943

Genus BALACRINUS Ramsbottom, 1961

Balacrinus basalis (M’Coy)
Fig. 104

MATERIAL, LOCALITY AND DISCUSSION. A well-preserved external mould (E 67750) of cup showing
characteristic plate pattern, arm and pinnule remains was recovered from sample R14 in the Allt
Ddu Formation at Rhiw March. Despite the occurrence of numerous fragmentary crinoid
remains in the majority of samples (see Figs 4, 6-9, 12-15, 17) this specimen is the only complete
one recovered during this study.

Class STELLEROIDEA Lamarck, 1816
Subclass OPHIUROIDEA Gray, 1840
Order STENURIDA Spencer, 1951
Family STENASTERIDAE Schuchert, 1914

Genus STENASTER Billings, 1858

Stenaster obtusus (Forbes)
Fig. 105

MATERIAL, LOCALITY AND DISCUSSION. A complete internal mould (E 53698) was recovered from
beds immediately above the Frondderw Ash in the stream at Ty Nant (SH 906262); this is the
same horizon from which sample 1022B was obtained (Fig. 6A, p. 178). The specimen is from a
similar stratigraphical horizon (associated with the Frondderw Ash) to the Moel y Garnedd
locality from which Salter recovered a specimen (Spencer 1927: pl. 23). Elles (1922: 138)
recorded another ophiuroid Protaster salteri (Salter, ex Forbes) from immediately above the Cefn
Gwyn Ash (Cefn Gwyn) (Spencer 1934 : pl. 31).

CARADOC FAUNAL ASSOCIATIONS Di
Faunal communities and associations

Unlike the Silurian faunas categorized by Zeigler (1965), Ziegler et al. (1968) and others into
benthic ‘communities’, British Ordovician faunas have not, until recently, been classified in any
such way, but elsewhere, e.g. in eastern North America, several Ordovician communities have
been named by Bretsky (1969) and others.

Williams (1973 : 242-243) concluded that the Lower and Middle Caradoc faunas of Wales
and the Welsh Borderland contained four facies-related ‘faunal associations’ (or “communities’),
named after Dinorthis, Nicolella, Onniella and Howellites, which are respectively characterized by
4,5, 2 and | other named brachipod genera. He also named a fifth association, the “Bicuspina set’,
which is characteristic of the Upper Llandeilo and Lower Caradoc succession of the Shelve area
(Williams 1976 : 39); the quantitative composition of this set (Williams 1974 : tables 7-11)
allows it to be distinguished, by cluster analysis, from two other pre-Caradoc, inarticulate-
dominated Shelve sets. The Bicuspina set is readily compared with the Dinorthis Association as
they contain elements in common, in particular Da/manella, Heierorthis and Bicuspina itself.

Pickerill (1974, 1977 : figs 3, 4) and Pickerill & Brenchley (1979) followed Williams (1973)
by naming a Dinorthis, a Nicolella, an Onniella, a Howellites and a Dalmanella ‘community’ from
the South Berwyns; the Dinorthis community is considered to consist of component Dinorthis
and Macrocoelia ‘sub-communities’, which are essentially analogous to ‘populations’ referred
to by Bretsky (1970). Unlike Williams they used quantitative measurements of ubiquity and
average abundance to define four of these five communities; they referred only briefly to the
fifth, Onniella community with associated genera. As the author, independently, chose to define
the associations named here in a similar fashion to the quantitative method of Pickerill &
Brenchley, an excellent opportunity is afforded for comparisons between the compositions of
related associations in adjacent areas. Like Williams (1973, 1976) and Pickerill & Brenchley (1979)
I use dominant genera in naming the associations described here; this permits the following direct
comparisons.

The Howellites community of the South Berwyns is essentially similar to the contemporary
Howellites—Paracraniops Association described here; both occur in a mudstone or silty mudstone
facies and have Howellites, Paracraniops, Macrocoelia, Reuschella, Bicuspina, Leptaena, Sowerby-
ella, Broeggerolithus and Brongniartella amongst the more persistent (ubiquitous) and abundant
elements. The main difference between the two associations is the relative importance of Sower-
byella in the South Berwyns. This is partly a reflection of a difference in the range of strata
considered in the two areas; if the fauna of pre-middle Allt Ddu beds bearing Sowerbyella and
Heterorthis were considered in conjunction with the upper Allt Ddu association the composi-
tional resemblance of the associations in these two areas would be even more striking.

The composition of the Howellites-Kloucekia Association (and the closely related marginal
Howellites association, Fig. 19, p. 193) is quite distinct, in detail, from the other Howellites-
dominated associations referred to here. The brachiopods Paracraniops, Macrocoelia, Reuschella,
Bicuspina, Leptaena and Sowerbyella and the trilobite Brongniartella are rare, absent or, like
Howellites and Broeggerolithus, represented in this association by species distinct from those found
in the older association(s) elsewhere. The association is further characterized by the relative
persistence of Skenidioides and Kloucekia.

Although Howellites and Broeggerolithus are characteristic of all associations incorporating
the name Hovwellites, it is clear that the Allt Ddu and Gelli-grin associations of the area south of
Bala are quite distinct and should not be considered jointly as Williams (1973) has done.

The Dinorthis Association, sensu Williams, although recognized by Pickerill & Brenchley in
the South Berwyns, has not been identified in the Lower Bala Group south of Llanuwchllyn.
Current evidence suggests that it is recognizable in parts of the Soudleyan succession in the Bala
area (sensu Bassett et al. 1966) but not in equivalent beds to the south. For example, Dinorthis
is a dominant element in sample GGIX (Fig. 12, p. 184) but is rare in equivalent beds at Craig y
Gath (Fig. 8, p. 179) and unknown in the equivalent Rhiw March section (Fig. 9, p. 180). Williams
(1973) states that the association is characterized by Bicuspina, Dalmanella, Heterorthis and
Leptaena; this is confirmed by Pickerill & Brenchley (1979) who recognize these genera together

228 M. G. LOCKLEY

with Howellites, Reuschella, Sowerbyella, Paracraniops, Macrocoelia, Broeggerolithus and Brong-
niartella amongst the dominant and persistent elements of the Dinorthis sub-community in the
South Berwyns. It is significant that these latter seven genera are all of equivalent importance in
the South Berwyn Howellites community and the Howellites—Paracraniops Association. However,
in the Dinorthis sub-community (sensu Pickerill & Brenchley 1979) of south Salop, Bicuspina
and Reuschella are not recorded whereas Harknessella and Salopia are considered important
elements. At various specified horizons in the Salop, south Berwyns, Breidden Hills and Snow-
donia successions, these authors consider the community to be represented by the Macrocoelia
sub-community which, apart from being dominated by this genus, is essentially similar in
composition to the Dinorthis sub-community. Similarly, their Dalmanella community, which
is not recognized in the Bala area, is composed essentially of elements which characterize their
Howellites and Dinorthis communities; its only very distinctive feature is the relative importance
of Dalmanella.

The Kullervo—Nicolella—Palaeostrophomena Association (Williams in Whittington & Williams
1955), the Nicolella association/community sensu Williams (1973) and of Pickerill & Brenchley
(1979) and the phases of the Nicolella—Onniella Association (defined here) are clearly closely
related. The six brachiopod genera, including Nicolella, referred to by Williams as components of
the association are not those which are most characteristic of the Welsh associations. Pickerill &
Brenchley (1979) and the present study have shown, for their respective associations, that the
most dominant and persistent elements are Nicolella, Dolerorthis, Skenidioides, Eoplectodonta,
Leptestiina, Onniella, Platystrophia, Bicuspina and Reuschella. Of these Platystrophia is partic-
ularly abundant in the South Berwyns, but not at Bala, whilst Onniella, Eoplectodonta, Reuschella
and Bicuspina are considerably more important in this latter area.

In the area considered here the importance of Onniella and Eoplectodonta in the Nicolella—
Onniella Association cannot be overlooked. Not only are both forms considerably more abundant,
at certain horizons, than any other brachiopod genera but they also represent a ‘dalmanellid—
plectambonitacean combination’ noted in numerous Ordovician faunal associations in Britain
and elsewhere. The so-called Onniella association/community referred to by Williams/Pickerill &
Brenchley, although not quantitatively defined, appears to differ from the Onniella associations
of this paper.

The Onniella—Sericoidea Association quantitatively outlmed here is clearly related to the
Onniella association/community referred to by Williams/Pickerill & Brenchley. Both the Pen y
Garnedd Shale and Dyfi Mudstone associations contain Onniella, Sericoidea and Paterula. This
particular dalmanellid—plectambonitacean association is typical of argillaceous parts of the
Caradoc Series elsewhere in Britain. For example, Hurst (personal communication 1978) reports
an Onniella—Sericoidea type of association in the Onny Shales. Similarly, Dean (1959) described
Marshbrookian to Pusgillian faunas characterized by Onniella and Chonetoidea (and Sericoidea,
A. D. Wright, personal communication 1975) from the Cross Fell area.

The Sericoidea Association of the Pistyll Gwyn, Y Ceunant and Aber Cowarch sections (Figs
17 & 19, pp. 190-193) is simply a diluted (low diversity, low equitability) marginal variety of the
Onniella—Sericoidea Association. Between Bala and Dinas Mawddwy there is a pronounced lateral
variation in the faunal associations (and facies) of the uppermost 20 m of the Caradoc succession.
The WNicolella—Onniella Association (occupying coarse, calcareous clastic substrates) passes
laterally into the Onniella—Sericoidea and Sericoidea Associations (calcareous mudstones), then
into a graptolitic association (in black shales); see Fig. 18, p. 192.

Contemporary Scoto-Irish faunas described by Williams et a/. (1962) and Mitchell (1977) have
yet to be categorized into named associations. These faunas are fundamentally different from
those of ‘Anglo-Welsh’ affinity; this is a reflection of the mid-Ordovician separation of northern
and southern Britain by the Proto-Atlantic ocean (Smith et a/. 1973). The discovery of Bimuria
and Protozyga in Wales and the recognition of wider distributions for Cyclospira, Palaeostro-
phomena and Anisopleurella are evidence of more mixing of Scoto-Irish with Anglo-Welsh
faunas than was hitherto supposed.

Despite the fact that Whittington & Williams (1955 : 398) and Williams (1962 : table 2; 1969 :
131-135) have shown that faunal associations with Scoto-American (or Scoto-Appalachian)

CARADOC FAUNAL ASSOCIATIONS 229

affinities show little similarities to those known from the Anglo-Welsh region, recent studies by
Bayer (1967), Fox (1968), Bretsky (1969, 1970), Bretsky & Bretsky (1975, 1976), Walker &
Laporte (1970), Walker (1972), Walker & Alberstadt (1975), Walker & Parker (1976) and others
have led to the naming of a number of associations or communities from the Mid- to Upper
Ordovician of the Appalachians which show striking parallels with those named from the Anglo-
Welsh region.

Among these ‘American communities’ are several named after familiar dalmanellid—plectam-
bonitacean combinations including the Onniella—Sowerbyella community (Bretsky 1969), the
Eoplectodonta (Thaerodonta)—Onniella community (Bayer 1967) of the mid-continent upper
Ordovician and the Resserella—Sowerbyella assemblage (Fox 1968); similarly, data given by Titus
& Cameron (1976 : 1216-1217) reveal that Paucicrura and Sowerbyella represent about two-
thirds (mean relative abundance) of the total fauna in four out of five communities named after
rarer elements.

Other communities named after or dominated by genera common to the Anglo-Welsh succes-
sions are as follows. The mid-Appalachian Rafinesquina, Sowerbyella, Lophospira Association and
the Onniella—nuculoid bivalve Association named by Bretsky (1969) from siltstones and mudstones
respectively, as component associations (or ‘populations’ ) in the Sowerbyella—Onniellacommunity,
are reminiscent of parts of the Berwyn and Salopian faunal successions respectively. Similarly,
the succession dominated by Onniella and Cryptolithus outlined by Bretsky & Bretsky (1975 :
228) from the upper Ordovician of Quebec is shown subsequently by these authors (1976 :
table 3) to be dominated at different horizons by combinations of genera including Nuculites with
Cryptolithus and Nuculites with Onniella (cf. Bretsky 1969). Such combinations are again re-
miniscent of parts of the Anglo-Welsh Caradoc successions.

The upper Ordovician Platystrophia—Leptaena assemblage (Fox 1968) is comparable to the
Nicolella community (sensu Pickerill & Brenchley 1979) which, in addition to occurring in a
similar calcareous facies, contains these two former genera amongst its most important elements.

Walker & Laporte (1970) and Walker (1972) referred to Strophomena and Dalmanella respec-
tively as ‘major’ taxa in inferred low intertidal and subtidal carbonate facies of the middle
Ordovician of New York. Similarly, Walker & Alberstadt (1975) and Walker & Parker (1976)
have named a Rostricellula—~Strophomena community from the middle Ordovician of the southern
Appalachians. All these genera represent important or dominant elements in parts of broadly
contemporary Anglo-Welsh successions.

Although Williams (1969 : 133-137) referred to Sowerbyella, Sericoidea, Macrocoelia, Rafines-
quina, Strophomena, Platystrophia, Dalmanella, Howellites and Onniella as widespread or pandemic
genera in Caradoc to Ashgill times, there has hitherto been little comment on their importance as
‘community dominants’ in parts of both the American and north European provinces. In the
Anglo-Welsh and Appalachian regions the importance of dalmanellid—plectambonitacean
dominated associations cannot be overemphasized; combinations like Dalmanella with Sower-
byella (in the Upper Llanvirn—Llandeilo of south Wales) and the other Caradoc to Ashgill
combinations mentioned above occupy a wide variety of facies including limestones, tuffs,
sandstones, siltstones and fine mudstones.

Conclusions

The present study suggests that detailed local examination of (Caradoc) faunal successions can
lead to the identification of distinct associations dominated by different species of the same, often
ubiquitous, genera. Since associations are commonly named after such dominant forms, a
paradoxical situation arises, where on the one hand local interassociation variation is easily
demonstrated, yet on the wider provincial scale considerable uniformity appears to be the rule.
Such a paradox is the result of a dual influence. Firstly, the use of generic rather than specific
terminology in the naming of communities, although favouring brevity, masks important
differences. Secondly, the widespread occurrence of various species of ubiquitous genera strongly
suggests that “parallel communities’ (sensu Thorson 1957, 1966) characterized Ordovician marine
benthic environments. For these reasons it is emphasized that the generic names employed in the

230 M. G. LOCKLEY

classification of associations should refer unequivocally to particular species at particular
horizons, as stated or implied here in all cases. This approach permits named communities to be
specifically identified and distinguished from other local or distant, more or less contemporary
parallel communities dominated by ubiquitous congeneric forms.

Acknowledgements

Amongst the many geologists to whom I wish to extend my thanks are Professor P. W. Bretsky, Professor
H. B. Whittington and in particular Professor A. Williams; also Drs L. R. M. Cocks, G. E. Farrow,
R. A. Hewitt, J. M. Hurst, J. K. Ingham, P. Magor, N. J. Morris, J. S. Peel, I. Strachan, G. B. Slater,
R. M. Watkins and C. Wilcox. I also wish to thank Mr & Mrs Davies, Mr & Mrs Edwards, Mr & Mrs
Lewis, Mr D. Roberts and my wife Mary for assistance during my periods of fieldwork. A N.E.R.C.
grant is also gratefully acknowledged.

References

Addison, R. 1974. The biostratigraphy of the Llandeilo Facies of South Wales. Ph.D. Thesis, Univ. of
Belfast (unpubl.).

Anderson, F. W. 1964. The law of ostracod growth. Palaeontology, London, 7: 85-105.

Antia, D. D. J. 1977. A comparison of diversity and trophic nuclei of live and dead molluscan faunas
from the Essex Chenier Plain, England. Pa/eobiol., Menlo Park, 3 : 404-414.

Bancroft, B. B. 1945. The brachiopod zonal indicies of the Stages Costonian—Onnian in Britain. J.
Paleont., Menasha, 19 : 181—252, pls 22-38.

Bassett, D. A., Whittington, H. B. & Williams, A. 1966. The stratigraphy of the Bala district, Merioneth-
shire. Q. JI geol. Soc. Lond., 122 : 219-271.

Bayer, T. N. 1967. Repetitive benthonic community in the Maquoketa Formation (Ordovician) of
Minnesota. J. Paleont., Tulsa, 41 : 417-422.

Brenchley, P. J. 1966. The Caradoc rocks of the north and west Berwyns. Ph.D. Thesis, Univ. of Liverpool
(unpubl.).

Bretsky, P. W. 1969. Central Appalachian Late Ordovician communities. Bull. geol. Soc. Am., Boulder,
Col., 80 : 193-212.

1970. Upper Ordovician ecology of the Central Appalachians. Bull. Peabody Mus. nat. Hist., New

Haven, 34 : 1-150.

& Bretsky, S. S. 1975. Succession and repetition of Late Ordovician fossil assemblages from the
Nicolet River Valley, Quebec. Paleobiol., Menlo Park, 1 : 225-237.

—— —— 1976. The maintenance of evolutionary equilibrium in late Ordovician benthic marine
invertebrate faunas. Lethaia, Oslo, 9 : 223-233.

Bruton, D. L. 1976. The trilobite genus Phillipsinella from the Ordovician of Scandinavia and Great
Britain. Palaeontology, London, 19 : 699-718.

Calef, C. E. & Hancock, N. J. 1974. Wenlock and Ludlow marine communities. Palaeontology, London,
17 : 779-810.

Cocks, L. R. M. 1978. A review of British Lower Palaeozoic brachiopods, including a synoptic revision of
Davidson’s Monograph. Palaeontogr. Soc. (Monogr.), London. 256 pp.

Cooper, G. A. 1956. Chazyan and related brachiopods. Smithson. misc. Collns, Washington, 127 : 1-1024,
1025-1245, pls 1-269.

Cisne, J. L. 1973. Life history of an Ordovician trilobite Triarthrus eatoni. Ecology, Brooklyn, 54: 135-142.

Davidson, T. 1869. A Monograph of the British Fossil Brachiopoda. Part VII. The Silurian Brachiopoda.
3: 169-248, pls 23-37. Palaeontogr. Soc. (Monogr.), London.

Dean, W. T. 1959. The stratigraphy of the Caradoc Series in the Cross Fell Inlier. Proc. Yorks. geol. Soc.,
Hull, 32 : 185-228.

Elles, G. L. 1922. The Bala country: its structure and rock succession. Q. J/ geol. Soc. Lond., 78 : 132-175.

Fox, W. T. 1968. Quantitative palaeoecologic analysis of fossil communities from the Richmond Group.
J. Geol., Chicago, 76: 613-641.

Fiirsich, F. T. 1977. Corallian (upper Jurassic) marine benthic associations from England and Normandy.
Palaeontology, London, 20 : 337-385.

Harrington, H. J. et al. 1959. Arthropoda I. Jn Moore, R. C. (ed.), Treatise on Invertebrate Paleontology O.
xix + 560 pp., 415 figs. Lawrence, Kans.

CARADOC FAUNAL ASSOCIATIONS 231

Hancock, N. J., Hurst, J. M. & Fiirsich, F. T. 1974. The depths inhabited by Silurian brachiopod com-
munities. J. geol. Soc. Lond., 130 : 151-156.

Hunt, A. S. 1967. Growth variation and instar development of an agnostid trilobite. J. Paleont., Tulsa,
41 : 203-208.

Hurst, J. M. 1975. Some observations on the brachiopods and the level-bottom community ecology of
Gotland. Geol. Fér. Stockh. Forh., 97 : 250-264.

1979. The stratigraphy and brachiopods of the upper part of the type Caradoc of South Salop.

Bull. Br. Mus. nat. Hist., London, (Geol.) 32 (4): 183-304.

& Hewitt, R. H. 1977. On tubular Problematica from the type Caradoc (Ordovician) of England.
Neues Jb. Geol. Paldont. Abh., Stuttgart, 153 : 2, 147-169.

Krebs, C. J. 1978. Ecology. The experimental analysis of distribution and abundance. (2nd edn.) London,
694 pp.

Lockley, M. G. 1977. The geology of the Llanuwchllyn to Llanymawddwy area, . . . (&c.). Ph.D. Thesis,
Univ. of Birmingham (unpubl.).

— 1978. The application of ecological theory to palaeoecological studies with particular reference to
equilibrium theory and the Ordovician system. Lethaia, Oslo, 11 : 281-291.

— 1980. The geology of the Llanuwchllyn to Llanymawddwy area, south Gwynedd, North Wales.
21 msp, 9 figs. Geol. J., Liverpool, 15 (1) (In press.)

—— & Williams, A. (1980). Lower Ordovician Brachiopoda from mid and south-west Wales. Bull. Br.
Mus. nat. Hist., London, (Geol.) 34 (4) (In press.)

MacArthur, R. H. 1965. Patterns of species diversity. Biol. Rev., Cambridge, 40 : 510-533.

MacGregor, A. R. 1961. Upper Llandeilo brachiopods from the Berwyn Hills, north Wales. Palaeontology,
London, 4 : 177-209.

Margalef, R. 1958. Information theory in ecology. Gen. Systs, Ann Arbor, Mich., 3 : 36-71.

Mitchell, W. I. 1977. The Ordovician Brachiopoda from Pomeroy, Co. Tyrone. Palaeontogr. Soc.
(Monogr.), London. 138 pp., 28 pls.

Murray, J. W. & Wright, C. A. 1974. Palaeogene Foraminiferida and Palaeoecology, Hampshire and
Paris Basins and the English Channel. Spec. Pap. Palaeont., London, 10 : 1-129.

Neyman, A. A. 1967. Limits to the application of the trophic group concept in benthic studies. Oceanology,
Washington, 7 : 149-155 [transl. from Russian of Okeanologiya, Moscow, 7 (2) : 195-202].

Palmer, A. R. 1957. Ontogenetic development of two Ollenellid trilobites. J. Paleont., Tulsa, 31 : 105-128.

Petersen, C. G. J. 1924. A brief survey of the animal communities in Danish waters... . Am. J. Sci.,
New Haven, (5) 7 : 343-354.

Pickerill, P. K. 1974. Geology of the south Berwyn Hills (North Wales) . . . (&c.). Ph.D. Thesis, Univ. of
Liverpool (unpubl.).

— 1977. Trace fossils from the upper Ordovician (Caradoc) of the Berwyn Hills, central Wales.
Geol. J., Liverpool, 12 : 1-16.

—— & Brenchley, P. J. 1979. Caradoc marine benthic communities of the south Berwyn Hills, north
Wales. Palaeontology, London, 22 : 229-264.

Pielou, E. C. 1969. An Introduction to Mathematical Ecology. 286 pp. London.

Pojeta, J., jr 1971. Review of Ordovician Pelecypods. Prof. Pap. U.S. geol. Surv., Washington, 695.
46 pp., 20 pls.

Pugh, W. J. 1923. The geology of the district around Corris and Aberllefenni. Q. J/ geol. Soc. Lond.,
79 : 50-51.

— 1928. The geology of the district around Dinas Mawddwy (Merioneth). Q. J/ geol. Soc. Lond., 84 :
345-381.

—— 1929. The geology of the district between Llanymawddwy and Llanuwchllyn (Merioneth). Q. J/
geol. Soc. Lond., 85 : 242-306.

Raw, F. 1927. The ontogenetics of trilobites, and their significance. Am. J. Sci., New Haven, (5) 14:
7-35, 131-149, 240; 25 figs.

Rogers, M. J. 1976. An evaluation of an index of affinity for comparing assemblages, in particular of
Foraminifera. Palaeontology, London, 19 : 503-515.

Sanders, H. L. 1968. Marine benthic diversity: a comparative study. Am. Nat., Lancaster, Pa., 102 : 243-
282.

— 1969. Benthic marine diversity and the stability-time hypothesis (contrib. 2353, Woods Hole
Oceanogr. Inst.). Jn Diversity and stability in ecological systems. Brookhaven Symp. Biol., Upton,
INDY. 22)2 71-81.

Scott, R. W. 1974. Bay and shoreface benthic communities in the Lower Cretaceous. Lethaia, Oslo, 7 :
315-330.

232 M. G. LOCKLEY

Sedgwick, A. & M’Coy, F. 1851-55. A synopsis of the classification of the British Palaeozoic Rocks, witha
systematic description of the British Palaeozoic Fossils ...(&c.) xcvili+661 pp., 27 pls. London &
Cambridge.

Smith, A. G., Briden, J. C. & Drewry, G. E. 1973. Phanerozoic world maps. Jn Hughes, N. F. (ed.),
Organisms and continents through time. Spec. Pap. Palaeont., London, 12 : 1-42.

Spencer, M. A. 1913-65. British Palaeozoic Asterozoa. Parts I-X, with supplement and index by H. G.
Owen. Palaeontogr. Soc. (Monogr.), London. 583 pp., 37 pls.

Thorson, G. 1957. Bottom communities (sublittoral or shallow shelf). Jn Hedgpeth, J. W. (ed.), Treatise
on marine ecology and paleoecology, 1. Mem. geol. Soc. Am., Washington, 67 (1) : 461-534.

—— 1966. Some factors influencing the recruitment and establishment of marine benthic communities.
Neth. J. Sea Res., Den Helder, 3 : 267-293.

Titus, R. & Cameron, B. 1976. Fossil communities in the Lower Trenton Group (Middle Ordovician)
of central and northwestern New York State. J. Paleont., Tulsa, 50 : 1209-1225.

Turpaeva, E. T. 1948. [The feeding of some benthic invertebrates of the Barents Sea.] Zool. Zh., Moscow,
27 (6) : 503-512. (In Russian.)

Valentine, J. W.,1972. Conceptual models of ecosystem evolution. Jn Schopf, T. J. M. (ed.), Models in
Paleobiology : 192-215. San Francisco.

—— 1973. Evolutionary paleoecology of the marine biosphere. 511 pp. New Jersey.

Walker, K. R. 1972. Community ecology of the Middle Ordovician Black River Group of New York
State. Bull. geol. Soc. Am., Boulder, Col., 83 : 2499-2524.

& Alberstadt, L. P. 1975. Ecological succession as an aspect of structure in fossil communities.

Paleobiol., Menlo Park, 1 : 238-257.

& Laporte, L. F. 1970. Congruent fossil communities from the Ordovician and Devonian of New
York. J. Paleont., Tulsa, 44 : 928-944.

—— & Parker, W. C. 1976. Population structure of a pioneer and a late stage species in an Ordovician
ecological succession. Paleobiol., Menlo Park, 2 : 191-201.

Watkins, R. M. 1975. British Ludlow palaeoecology and its bearing on the Silurian marine ecosystem.
D.Phil. Thesis, Univ. of Oxford (unpubl.).

— 1979. Benthic community organization in the Ludlow Series of the Welsh Borderland. Bull. Br.
Mus. nat. Hist., London, (Geol.) 31 (3) : 175-280.

Whittington, H.B. 1962-68. The Ordovician tribolites of the Bala area, Merioneth. Parts I-IV.
Palaeontogr. Soc. (Monogr.), London. 138 pp., 32 pls.

& Williams, A. 1955. The fauna of the Derfel Limestone of the Arenig district, North Wales.
Phil. Trans. R. Soc., London, B 238 : 397-427.

Williams, A. 1962. The Barr and Lower Ardmillan Series (Caradoc) of the Girvan district, south-west
Ayrshire, with decriptions of the Brachiopoda. Mem. geol. Soc. Lond., 3 : 1-267.

—— 1963. The Caradocian brachiopod faunas of the Bala district, Merionethshire. Bull. Br. Mus. nat.
Hist., London, (Geol.) 8 : 327-471.

—— 1969. Ordovician faunal provinces with reference to brachiopod distribution. Jn Wood, A. (ed.),
The Pre-Cambrian and Lower Palaeozoic rocks of Wales : 117-154. Cardiff.

— 1973. Distribution of brachiopod assemblages in relation to Ordovician palaeogeography. Jn
Hughes, N. F. (ed.), Organisms and continents through time. Spec. Pap. Palaeont., London, 12 : 241-
269.

— 1974. Ordovician Brachiopoda from the Shelve district, Shropshire. Bull. Br. Mus. nat. Hist.,
London, (Geol.) Suppl. 11 : 1-163, pls 1-28.

—— 1976. Plate tectonics and biofacies evolution as factors in Ordovician correlation. Jn Bassett, M. G.
(ed.), The Ordovician System : 29-66. Cardiff (Proc. palaeont. Ass. Symp., Birmingham 1974).

et al. 1972. A correlation of Ordovician rocks in the British Isles. Spec. Rep. geol. Soc. Lond.,
3 : 1-74.

Wright, A. D. 1970. The stratigraphic distribution of the inarticulate brachiopod Orthisocrania divaricata
(M’Coy) in the British Isles. Geol. Mag., Cambridge, 107 (2): 97-103.

Ziegler, A. M. 1965. Silurian marine communities and their environmental significance. Nature, Lond.,
207 : 270-272.

——, Cocks, L. R. M. & Bambach, R. K. 1968. The composition and structure of Lower Silurian marine
communities. Lethaia, Oslo, 1 : 1-27.

CARADOC FAUNAL ASSOCIATIONS P33)
Index

New taxonomic names and the page numbers of the principal references are in bold type. An asterisk

(*) denotes a figure.

Aber Cowarch 168, 188, 190, 196, 209, 228
Aegiromeninae 211-15
Afon Twrch 174-7, 186, 197, 200, 209
Agnostina 171
Allt-Ddu Formation 168, 175, 177-80, 182-4,
191-3, 197-8, 200, 204, 210, 223-4, 226-7
Ambonychia 224
Anisopleurella 173, 184-6, 202, 211, 228
cf. multiseptata 211, 213*
Arca 224
Archaeocrinidae 226
? Archinacella sp. 225*, 226
Archinacellidae 226
Arcoida 223
Arthropoda 220
Atrypacea 218

Bala 167-9, 173, 191-2, 210, 222, 224, 227-8
Lake 166; see Lower, Upper Bala Group
Balacrinus 226
basalis 225*, 226
Balclatchie Mudstones 214, 216
Bancroftina 195, 209, 213*
Bellerophon 180, 184, 187, 190, 224
Berwyn 166-7, 173, 183, 202, 209, 222, 224, 227,
229
Beufy Isaf 175, 177-8
Bicuspina 175-80, 182, 184-7, 189, 193, 195,
227-8
spiriferoides 194
Bimuria 186-7, 189, 202, 215, 228
ef. buttsi 215
dyfiensis 166, 213*, 215, 217*
youngiana 215
recta 215
Bimuridae 215
Bivalvia 167, 172, 176, 179, 180, 182, 184-5, 187,
190, 193, 220, 222
Brachiopoda 167, 172-3, 194, 202, 222
Broeggerolithus 173, 176-80, 182, 184-90, 192,
194-6, 227-8
nicholsoni 195
soudleyensis 194
Brongniartella 176-80, 182, 184-7, 193, 227-8
cf. minor 194
Bryn Beds 202, 216
Bryozoa 172, 178, 180, 187, 190, 193
Bucania 226
Bucanopsis 226
Byne Hill 212, 214
Byssodesma 224

Camerata 226
Caradoc 167, 169, 173, 183, 188, 191-2, 195, 199,
202, 218, 220, 222, 224, 227-9

Cardiolaria 223, 225*
Ceiswyn Mudstone 175, 183, 188, 193
Cephalopoda 172, 176, 178-80, 184—6, 226
Chasmops 175, 184—5, 187, 197
Cheiruridae 222
Chonetoidea 228
radiatula 214
Climacograptus minimus 186
Colpomya 224
Corris 167
Shale 169, 188, 196
Craig y Gath 168, 170, 178-9, 181-3, 193-4, 204,
DOs
Craniacea 205-6
Craniidae 205-6
Craniidina 205-6
Craniopsidae 203-5
Cremnorthis 184-5, 187, 189
Crinoidea 172-3, 176, 177-80, 184-7, 190, 226
Cryptolithus 229
Ctenodonta 224
Cuneamya 172, 224, 225*
Cyclonema 172, 180, 184, 187, 194, 224, 226
Cyclospira 178, 185-90, 193, 196, 202, 219, 228
aff. bisulcata 208, 217*, 219, 220, 221*
carrickensis 219
longa 219
nana 219
preciosa 219
quadrata 219
sulcata 219
Cyclospirinae 219
Cymerig Limestone 168, 178, 184-5, 188, 195-6,
198-9, 203, 209, 219-20, 222
Cyphoproetus 173, 202, 221*, 222
Cyrtodonta 172, 223, 224, 225*
Cyrtodontidae 223
Cyrtolites 226
Cyrtonotella 173
Cystoidea 173, 185

Dalmanella 178, 180, 182, 184-7, 189, 195, 224,
226-9

Dalmanellidae 178

indet. 177, 180, 184-7, 190, 193-4, 197, 209

Dayiacea 219

Dayiidae 219

Deacybele? sp. 173

Derfel Limestone 173, 175, 186, 191, 194, 202, 222

Dinas Caradoc 173

Dinas Mawddwy 167-9, 191-2, 228

Dinorthis 176-7, 179, 182-5, 187, 193, 224, 226-8

Dolerorthis 173, 175-6, 178, 184-6, 189, 193,
196-7, 228

234 M. G. LOCKLEY

duftonensis prolixa 194-5
tenuicostata 194

Dyfi Mudstone 169, 186, 188, 193-6, 228
Valley 168, 175, 177, 183

Echinodermata 202, 226
Enteletacea 209
Eoplectodonta 173, 178-9, 184-7, 189, 193, 195-6,
228-9
lenis 194
rhombica 194
Estoniops 178, 184-5

Flexicalymene 178, 180, 183-7, 196
Frondderw Ash 168, 174-5, 186, 197, 209-10, 226

Gastropoda 172, 176, 178-9, 182, 185-7, 189-90,
193-4, 224

Gelli-grin Formation 168, 170, 173, 175, 178-9,
183-8, 191, 193-6, 198-200, 202-6, 209, 211,
215

Glossellinae 203

Glyn Gower Formation 168, 175, 183, 186, 197

Goniophorina 224

Grammysiidae 224

Graptoloidea 178-9, 182, 190, 193

Hagley Volcanics 218

Harknessella 228

Hedstroemina 186

Hemithyris depressa 210

Hengae Group 167-8

Heterorthis 175-8, 183, 191, 197, 227
cf. retrorsistria 197

Howellites 173, 176-80, 183-9, 190-4, 196-7,

199-200, 209, 224, 226-8

antiquior 184, 195-6, 207*, 209
ultimus 194, 207*, 209

isotelinid 178-80, 184-7

Kiaeromena 180

Kjaerina 187
(Hedstroemina) 186

Kloucekia 178, 184-91, 193-6, 199, 227
apiculata 195

Kokenospira 226

Kullervo 173, 228

Laggan Burn 212, 214
Leptaena 173, 176, 178-80, 182, 184-7, 189, 193,
195-6, 227, 229
Leptestiidae 210
Leptestiina 173, 178, 184-7, 189, 196, 228
Leptestiinae 210
Lingulacea 202
Lingulella 173, 202, 207*
cf. ovata 203, 207*
Lingulellinae 202

Lingulida 184-5, 202

Linoporellidae 210

Llaethnant Formation 168, 175, 183, 191, 197

Llandeilo 210

Llanuwchllyn 166-7, 227

Llanymawddwy 166-7, 191

Lledwyn Bach 178, 183-5

Lonchodomas 187-8, 196, 202, 221*, 222

Lophospira 224, 226, 229

Lower Bala Group 166-8, 173, 183, 186, 192,
202-3, 205-6, 209-10, 220, 222, 227

Macheridia 172-3, 178-80, 182, 184, 186-90, 193
Macrocoelia 176, 179-80, 182, 184-7, 189, 193,
227-9

prolata 194
Maes-Meillion 168, 183, 185, 198
Malletiidae 223
Modiolopsis 172, 223, 224, 225*
Modiomorphoida 223
Mollusca 172, 185, 202, 220, 222
Monoplacophora 172, 187, 226
Murchisonia 224, 226

Nant Aber Derfel 222
Nant Bwlch-y-pawl 177
Nant Hir 175
Nant Rhyd Wen 204
Nant Tan y Bwlch 183, 186, 218
Nant y Nod 196
Nicolella 173, 175-6, 178, 184-7, 189, 191, 193-7,
199-200, 206, 227-8
cf. actoniae obesa 194, 206, 207*
humilis 194
Nod Glas Formation 167-9, 183-4, 187-8, 191,
196, 212, 214, 218, 223, 226
Nuculoida 223
Nuculites 188, 223, 225*, 229

Obolidae 202
Olenellinae 171
Olenidae 171
Onniella 173, 175-80, 182, 184-7, 191, 193-6, 199-
200, 209, 227-9

aff. avelinei 194

ostentata 194, 197, 206, 207*, 209
Ophiuroidea 226
Ordovician 173, 210, 222, 227-9
Orthacea 206-9
Orthida 205, 206-10
Orthidae 206
Orthisocrania 175, 197, 202, 205-6, 207*
Ostracoda 172-3, 177-9, 186, 189-90, 193
Oxoplecia 173, 186, 188

Palaeoglossa 187-8, 203
cf. attenuata 203, 207*
Palaeosolen 223

CARADOC FAUNAL ASSOCIATIONS 235

Palaeostrophomena 173, 184, 186-7, 189, 202, 210,
228
angulata 211
canalis 166, 210-11, 213*
magnifica 211
Paracraniops 178-80, 182-3, 187, 191-4, 197, 199,
202-3
glaber 166, 183, 194, 204-5, 207*, 208, 227-8
macellus 188, 203-4, 205, 207*
pararius 205
Parastrophinella 202
brenchleyi 166, 216-8, 217*
costata 218
musculosa 218
parva 218
youngi 218
Parastrophinidae 216
Paterula 187-8, 196, 203, 207*, 228
Paterulidae 202-3
Paucicrura 229
Pentamerida 205, 216-8
Pen-y-Cefn-Coch 178, 185, 194-6, 220
Phillipsinella 188, 202, 220, 221*
Phillipsinellidae 220
Pholadomyoida 224
Pistyll Gwyn 188, 190, 196, 199, 223, 228
Platylichas 173, 178, 184-7
Platystrophia 173, 178, 184-7, 189, 228-9
Plectambonitacea 210
Plectorthidae 209
Pont y Ceunant 183-4, 196, 198
Porambonitacea 216
Praenucula 223
Praenuculidae 223
Prasopora 180, 187
prasoporid bryozoa 175-9, 182, 184-6, 189
Primaspis 185
Primitia 172, 187
Productorthinae 206
Proetidae 222
Protaster salteri 226
Protozyga 186-8, 196, 202, 218, 228
diversa 218
musculosa 166, 217*, 218
perplexa 218
rotunda 218
Pseudolingula 186, 203, 207*
Psilonychia 224

Rafinesquina 229
ramose bryozoa 175-80, 182, 184-7, 189
Raphiophoridae 222
Resserella 229
Reuschella 176-80, 182, 184-7, 189-90, 193,
195-7, 227-8
horderleyensis undulata 194

Rhactorthis 178, 184-9, 196, 209
crassa 207*, 209
Rhiw March 169-70, 178-80, 182-8, 193-8,
201-4, 209, 211-12, 215, 218-20, 222, 226-7
Rostricellula 179-80, 182, 229

Salopia 173, 175-6, 184-7, 197, 206, 207*, 210,
228
Seelya 226
Sericoidea 169-70, 173, 176, 178, 180, 182-8,
190-1, 193-5, 199-200, 202, 211-5, 228-9
abdita complicata 166, 196, 208, 212-5, 213*
restricta 212
Sinuites 172, 180, 182, 184, 187-8, 190, 194, 224,
226
Skenidioides 178-9, 184-7, 189, 193, 195-6, 227-8
Soudleyan 205, 210, 224, 227
Sowerbyella 175-80, 182-3, 186-7, 189, 193, 195,
197, 227-9
sericea permixta 197
Sowerbyellidae 211-15
Sowerbyellinae 211
Sphaerocoryphe 188, 196, 202, 221*, 222
Spiriferida 205, 218
Spiriferide, gen. indet. 220, 221*
Stelleroidea 226
Stenaster obtusus 225*, 226*
Stenasteridae 226
Stenurida 226
Strophomena 229
Strophomenida 205, 210-11
Syntrophiidina 216

Tallinnella 172, 176, 180, 184-7, 196
Tan y Bwlch 184-5, 188, 195-6, 219
Tentaculites 172, 184-5, 187
Trilobita 167, 172, 180, 184-5, 187, 190, 202, 220,
222

Tripleciacea 210
Triplesia 210

edgelliana 210

maccoyana 210
Triplesiidae 210
Triplesiidina 210
Ty-nant 175, 177-8, 183, 226

Upper Bala Group 188
Upper Llanvirn 210

Vlasta 224

Y Ceunant 188, 190, 228

Accepted for publication 19 June 1979

Sy had eee a

British Museum (Natural History)
Handbooks

" eed primarily at schools and colleges, and written by Museum scientists, these profusely
‘illustrated paperbacks are invaluable for courses and as background reading for students.
_ Especially recommended for students of Palaeontology are:

al
=e

si

e

British Caenozoic Fossils (Tertiary and Quaternary)
5th edition, 1975, vi+ 132 pp, including 44 line plates,
coloured map, 8vo. £1.00

British Mesozoic Fossils
‘Sth edition, 1975, vi+ 207 pp, including 73 line plates,
coloured map, 8vo. £1.50

British Palaeozoic Fossils
4th edition, 1975, vi+203 pp, including 69 line plates,
coloured map, 8vo. £1.00

Titles to be published in Volume 33

An account of the Ordovician rocks of the Shelve Inlier in West
Salop and part of north Powys. By the late W. F. Whittard, F.R.S.
(Compiled by W. T. Dean)

Miscellanea

The Caradoc faunal associations of the area between Bala and f
Dinas Mawddwy, north Wales. By M. G. Lockley . Me

Fossil insects from the Bembridge Marls, Palaeogene of the Isle of
Wight, southern England. By E. A. Jarzembowski

The Yorkshire Jurassic fern Phlebopteris braunii (Goeppert) and its
reference to Matonia R. Br. By T. M. Harris y

The entire Geology series is now available as a series

Printed by Henry Ling Ltd, Dorchester

A yt het

Bulletin of the
British Museum (Natural History)

Fossil insects from the Bembridge Marls,
Palaeogene of the Isle of Wight, Southern
England

EAJ arzembowski

Geology series Vol33No4 27 March 1980

The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four
scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology,
and an Historical series.

Papers in the Bulletin are primarily the results of research carried out on the unique and
ever-growing collections of the Museum, both by the scientific staff of the Museum and by
specialists from elsewhere who make use of the Museum’s resources. Many of the pepee ar
works of reference that will remain indispensable for years to come. ‘

Parts are published at irregular intervals as they become ready, each is complete in itself, ve
available separately, and individually priced. Volumes contain about 300 pages and are no
necessarily completed within one calendar year. Subscriptions may be placed for one or m
series. Subscriptions vary according to the contents of the Volume and are based on a fe :
list of titles. As each Volume nears completion, subscribers are informed of the cost of the —
next Volume and invited to renew their subscriptions. Orders and enquiries should be sent to:

Publications Sales,
British Museum (Natural History),
Cromwell Road,
London SW7 S5BD,
England.

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

© Trustees of the British Museum (Natural History), 1980

ISSN 0007-1471 ” Geology serial

British Museum (Natural History)
Cromwell Road
London SW7 5BD Issued 27

/
Fossil insects from the Bembridge Marls, Palaeogene
of the Isle of Wight, southern England

E. A. Jarzembowski

Department of Emtomplocy. British Museum (Natural History), Cromwell Road, London
SW7 S5BD

Contents
Synopsis ; : : ; : : ; ; : : : ‘ . 238
Introduction . 5 : : ; : ; : 5 : ; A 5 62238)
Previous studies : 5 : : ; : . 240
Notes on lithology, inclusions and method of study : : : : : . 240
Field sections . : f : ; : : : : : ; : . 242
Correlation. é : : : : : : 5 : : . 243
Systematic descriptions ‘ : ‘ : ; : 5 : 4 : . 244
Order Isoptera_. : : . : : : : : : . 244
Family Mastotermitidae : ‘ : é : : : : : . 244
Genus Mastotermes Froggatt . 5 , : 3 : : : . 244
Mastotermes anglicus von Rosen . ; : 3 : ‘ : . 245
Family Kalotermitidae é ; : : - : : 3 : . 249
Genus Kalotermes Hagen ‘ : : 5 ; : : . 249
Kalotermes disruptus (Cockerell) : : é : : ; : . 249
Family Rhinotermitidae . 5 : ; : ; ; : o Zyl
Genus Reticulitermes Holmgren : ; i A ' ‘ ; 5 225i
Reticulitermes sp. : ; f 5 : é ‘ : ; > 25
Family Termitidae . : : F : P , : 5 : 5 255)
Species A . : 5 : : ; ; : ; é ; » 258
Family uncertain : : : : z : : : : 258
Order Plecoptera . " : : : ? : 5 : : : « 253)
Family Leuctridae_ . : : ‘ : 5 , , : 3 5 253
Genus Leuctra Stephens . , : : : : 5 5 : 5 38
Leuctra priscula (Cockerell) . : : : ; : ' : S23
Order Neuroptera A A : i : : : : ; 5 5 2S)
Family Mantispidae . : : : ; ; : ; ; , 5 2S)
Genus Promantispa nov. ; 5 ; : : ; . 2 E255
Promantispa relicta (Cockerell) ; é : : ‘ é ‘ o PSs)
Family Hemerobiidae : : . : ; ; ; : ‘ 5 AS/
Genus Hemerobius Linné 5 : ‘ 5 : f : : a 25)
Hemerobius tinctus sp. nov. . : : ; : 5 : : a 9 5)//
Genus Neuronema McLachlan : : : : : : : o 2SL)
? Neuronemasp.A . : . : : : : j : a ay)
? Neuronema sp. B : : : ; : ; : ; , 2 259)
Genus indet. 4 a. 5 ; : : ; : : : . 260
Species A . ‘ : 3 : : ; : ‘ : . . 260
Species B . : ; ; : : ‘ : F : ‘ . 260
Species C . : ; : : ; ; : j ; : . 260
Family Chrysopidae . : : ; ‘ ; é ; : ; . 260
Species A . - : : ; : : : : : ; . 260
Family Sisyridae é ; : : ‘ : 3 : ; : . 261
Species A . : - ; : : : ; : : . 261
Family uncertain : : : : : : : : : 5 Ay
Order Mecoptera . : : : : : 2 ; ’ : : 5 AW
Family Bittacidae : : : : : : : : é : . 262

Bull. Br. Mus. nat. Hist. (Geol.) 33 (4) : 237-293 Issued 27 March 1980

Nm
ies)
(oo)

E. A. JARZEMBOWSKI

Genus Bittacus Latreille . : F ; : 2 d ; ‘ . 262
Bittacus veternus (Cockerell) , : ; : : : : - XS?
Bittacus sp. A. : ; : ? i : : ‘ : . 263
Order Lepidoptera ; ; ; : , : . : : ‘ . 263
Family Micropterigidae : : ; : 7 , , j . 263
Genus Micropterix Hiibner . ; ; 3 , ; F . 263
Micropterix anglica sp. nov. ; ; : ‘ : ; : . 263
Family Hepialidae . ‘ : ; ; ; ; : : j . 265
Genus Prohepialus Piton ; 2 ; j : s : : . 265
Prohepialus sp. . : ; : : : : : : : . 265
Family Tineidae : : 3 : ; : : : ; a AGT
Genus Paratriaxomasia nov. . : 3 ‘ F : j a Asy/
Paratriaxomasia solentensis sp. nov. : : 3 : j é . 267
Genus indet. 3 : F ; : ; ; : : ; . 269
Species A . : ; : : i 5 : : ; . 269
Superfamily Gelechioidea : : : : F : 5 ; F . 269
Species A . : : ; 5 : : : : : . 269
Family Gopronionphidce : : 3 : ; : F : ‘ 5 ZI)
Genus Copromorpha Meyrick . ; ‘ 3 : . : ; 5 27K)
Copromorpha fossilis sp. nov. ; F 2 : : : F . 270
Microlepidoptera, Family uncertain . ; ; : : ‘ ; 270
Species A-L 3 : : : : 5 : : : : 270-275
? Family Cossidae . 2 : : : ; : : : ; a 4S

Genus Gurnetia Cockerell ; : j ‘ : : : ‘ TS

Gurnetia durranti Cockerell . : : 5 ; : ; 5 ~ TS
Family Pyralidae : : ‘ : : : ; : : : ©)

Genus Pyralites Heer. : : : F ; ; ‘ P . 276

Pyralites preecei sp. nov. j F : ‘ : : : : . 276
Family Geometridae . : : F : F : ‘ : . 278

Genus Geometridites Clark et al : : : : : : ; a) ZS

Geometridites larentiiformis sp. nov. : ; P E ; F 2 218
Family Nymphalidae : j : 5 : : ; ; : =~ 219

Genus Nymphalites Scudder . : : ; 5 q : : 5 2i9)

Nymphalites zeuneri sp. nov. : : : ; : : : . 279
Family Lycaenidae . : ; : i F : : : 3 . 282
Genus Lithopsyche Butler ‘ : : ; 5 : ; ; 5 ey
Lithopsyche antiqua Butler . F : P ; ; ‘ : 5288
? Superfamily Papilionoidea : : ; : ; ; ; : . 284
Genus indet. . : j ; , ‘ : 3 : : : . 284
Conclusions . : : ; : ; j ‘ : : : : . 284
Acknowledgements . F 5 : : : j , : ; : . 286
References ; : : : , j ; : : ; ; ; . 286
Index . : ; ; : ; ; : ; ; : A : . 291
Synopsis

The palaeontology, sedimentology and stratigraphy of the Insect Bed (Bembridge Marls) are discussed
and previous work summarized. Five insect orders are revised, notes on the biology of living relatives
are included and the palaeoecology considered. Isoptera: Mastotermes anglicus von Rosen is redescribed,
Sisyra (?) disrupta Cockerell is transferred from the Neuroptera to Kalotermes s. lat. (Kalotermitidae);
a species of Reticulitermes (Rhinotermitidae) and a termitid are described but not named. The Isoptera
suggest a warm subtropicaltropical climate. Plecoptera: Nemoura priscula Cockerell is transferred to
the genus Leuctra (Leuctridae). Megaloptera: the record of Raphidia is shown to be erroneous. Neurop-
tera: Mantispidae: Mantispa relicta Cockerell is transferred to Promantispa gen. nov. Hemerobiidae:
Hemerobius tinctus sp. nov., ? Neuronema spp. A and B, and 3 unnamed species are described. Chry-
sopidae: a species of Chrysopinae (s. lat.) is described but not named. The affinities of an extinct sisyrid
are discussed. Mecoptera: Panorpa veterna Cockerell (Panorpidae) is transferred to the genus Bittacus
(Bittacidae); an unnamed species of Bittacus is described. Lepidoptera: Micropterix anglica sp. nov.

PALAEOGENE INSECTS 239

(Micropterigidae), Paratriaxomasia solentensis gen. et sp. nov. (Tineidae), Copromorpha fossilis sp. nov.
(Copromorphidae), Pyralites preecei sp. nov. (Pyralidae), Nymphalites zeuneri sp. nov. (Nymphalidae),
Prohepialus sp. (Hepialidae) and Geometridites larentiiformis sp.nov. (Geometridae) are described.
Lithopsyche antiqua Butler is transferred to the Lycaenidae and the placement of Gurnetia durranti
Cockerell in the Cossidae is questioned. A tineid, a tentative gelechioid, 12 other species of microlepid-
optera and a possible papilionoid are described but not named.

Introduction

The Bembridge Marls are a predominantly argillaceous formation, 21-36 m thick, preserved only
in the northerly half of the Isle of Wight in southern England (Fig. 1). Near the base of the
formation is a thin clay bed, generally less than a metre thick, with concretions and tabular bands
of fine-grained argillaceous limestone and hard marl. The latter, on the NW coast near Cowes,
have been noted for their insect remains for just over a century and constitute the ‘Insect Lime-
stone’ of authors (Daley 1973a). However, the calcareous and argillaceous developments together

1 West Cowes East Cowes
ae 2

Sticelett Ledge
Thorness Bay
Saltmead Ledge

Newtown

Priory Bay
Node’s Point

St. Helens
Cliff End

A

Howgate Bay

Whitecliff Bay
< are
é \ one > Km Y,

Fig. 1 Outcrop map of the Bembridge Marls.

form a distinct unit (/bid.) and the name Insect Bed (Bembridge Marls) used by Curry (1958)
for a single limestone is here applied collectively. This horizon has yielded the only sizeable
insect fauna in the British Tertiaries above the Lower Eocene, some fifteen orders being repre-
sented. The fauna also includes molluscs, an ostracod, an anostracous crustacean, an isopod,
arachnids (spiders) and avian remains (feathers, Fig. 61, p. 268), the last three of which are
scarce. Although named after the insects, the bed is probably better known in the palaeonto-
logical literature for its flora due to the work of Reid & Chandler (1926). However, apart from
the leaves of a reed, well-preserved plant macrofossils are rare. Insect fragments are fairly
frequent, and richer pockets are sometimes encountered (Jarzembowski 1976). The most extensive
collections from this bed are now preserved in the British Museum (Natural History). These were
largely collected by E. J. A’Court Smith (1814-1900), collecting between West Cowes and the
Newtown River from about 1859 (Smith 1874). Apparently he discovered the insects following a
comment by Forbes (1856 : 58-59) on the insect-bearing nature of some similar limestones in
the Purbeck Beds (late Jurassic-early Cretaceous) on the mainland (Hughes 1922 : 67; Jackson
1933). Much of A’Court Smith’s material subsequently went to R. W. Hooley and the Rev. P. B.
Brodie, and the history of the collections and their acquisition by the Museum are given in Reid &
Chandler (1926) and Crane & Getty (1975). Notable smaller collections from the Insect Bed
include one at the Museum of Isle of Wight Geology, which was mainly the work of
G. W. Colenutt, and another consisting of some of Brodie’s material in the Lacoe Collection
at the United States National Museum; I have myself collected from the bed since 1966.

The fossil insect collections at the British Museum (Natural History) contain over 3900 pieces
of ‘Insect Limestone’ of which a large proportion bear two or more specimens. However, a

240 E. A. JARZEMBOWSKI

number of these are indeterminate and some species, especially in the Formicidae, are repre-
sented by numerous examples. The present paper is confined to the Isoptera, Plecoptera, Neurop-
tera, Mecoptera and Lepidoptera, which although rare or uncommon have yielded a varied
range of taxa; several other orders are being studied. All available material is documented, and
in addition to systematic and taxonomic considerations, the ecological and biological require-
ments of Recent relatives are briefly discussed.

Previous studies

A’Court Smith (1874) appears to be the first published record of the occurrence of arthropod
remains in the Insect Bed near Cowes. Woodward (1878, 1879) published a list of identifications
made by Frederick Smith of the Department of Zoology, British Museum (Natural History),
based on a sample of A’Court Smith’s collection. In the 1879 paper Woodward described an
anostracan, Branchipodites yectensis, and an isopod, Eosphaeroma fiuviatile, from the same
horizon; Martini (1972) showed the latter to be a synonym of the continental species Eosphaeroma
margarum (Desmarest 1822). McCook (1888a, b) gave two identical descriptions of a mygalo-
morph spider, Eoatypus woodwardii, and Jones & Sherborn (1889) described an ostracod Pota-
mocypris brodiei which Haskins (1968) compared with Cypridopsis, another freshwater genus.

The first description of an insect, Lithopsyche antiqua Butler 1889, was accompanied by a
colour plate. This was followed by von Rosen (1913) on fossil Mastotermes, and in 1915, T. D. A.
Cockerell described 33 new insect species from the Lacoe Collection. He subsequently made
selective studies of the main collection at the BM(NH): Cockerell (1917a, 1921a, b, c, 1922);
Cockerell & Andrews (1916); Cockerell & Haines (1921). The Formicidae were studied by
Donisthorpe (1920), the Culicidae by Edwards (1923) and Orthoptera by Zeuner (1939). These
constitute the main works on the insect fauna and, although the flora was monographed by
Reid & Chandler (1926), there is no comprehensive work on the arthropods. Chandler (1964)
reviewed the Bembridge Flora, including the microfloral work of Machin (née Pallot) published
later in 1971. Although the junction of the Bembridge Marls and Bembridge Limestone is
unconformable (Daley & Edwards 1971), the palaeobotanists considered the Bembridge Beds to
contain a single flora, and it is not always clear which taxa occur in the Insect Bed. Daley (1969)
studied the palaeoenvironment of the Bembridge Marls, and (1971, 1972a) was specifically
concerned with sedimentary deformational structures in the ‘Insect Limestone’. Jarzembowski
(1976) first recorded insects in this bed on the east side of the island near St Helens, 18 km from
Cowes.

Notes on lithology, inclusions and method of study

The insect-bearing lithology is usually a fine-grained argillaceous limestone with a conchoidal
fracture and blue-grey colouration when fresh. It weathers externally producing a distinctive
brownish yellow rind, and on broken surfaces the weathered zone has a sharp junction with the
unweathered material. Older pebbles, especially near West Cowes, may only have a small region
of ‘blue’ near the centre, and the weathered zone occasionally shows concentric colour banding.
At outcrop it occurs in dark blue-grey clay as oval concretions, up to 20 cm thick, or as thinner
tabular bands, and can be of a more marly constitution; thin, laterally continuous shell bands
delimit the bed (Figs 2-4, p. 242). The calcareous developments tend to occur in courses within
the clay more or less parallel to the bedding, which is of local stratigraphic value. There is a
definite arching of the upper shell bed over large concretions, but in some places the undulations
are less easy to explain.

Silty laminae, 1-4 mm thick and 3-90 mm apart, are usually evident in the limestone, marked
by finely-comminuted carbonaceous detritus. They appear to have no vertical regularity and
erosional bases are occasionally evident. The laminae may contain ostracod valves, fragments of
drifted vegetation —the only common recognizable ones being of Typha /atissima Braun — and
insect wings and body fragments; groups of shells belonging to the freshwater genus Ga/ba some-

PALAEOGENE INSECTS 241

times occur. The freshwater genera Hippeutis and Viviparus also occur but are infrequent. Plant
fossils are usually accompanied by a brown carbonaceous residue, and small black fragments of
vegetable material are occasionally present. One of the latter has been examined and identified as
natural fusain by Professor T. M. Harris. The limestone has yielded some small grains of resin:
one of these (In.17436) was analysed by infrared spectroscopy, but the spectra obtained were too
weak for comparison with those of other Tertiary resins.

The laminae occasionally yield concentrates of articulated insect bodies, often associated with
abundant, complete examples of Branchipodites vectensis, but such rich developments appear
to be very local. Salt pseudomorphs, comminuted shell débris, seeds, spiders and portions of bird
feathers are rare: the last are the only vertebrate fossils found. Insect bodies and wings are
normally preserved in good relief, although substantial compression is evident in some of the
more argillaceous laminae; well-preserved insects as well as fragments occur in the interbedded
limestone, but there is never any concentration of material. Chitin is absent only from weakly
sclerotized areas and the colour pattern may be well preserved: comparison with related forms in
the living fauna indicates that some colours are close to those in life (pp. 259, 262). The body
cavities are often infilled with a pink calcareous material which obscures structures, but external
moulds in the enclosing matrix show very fine detail (Fig. 59, p. 267) and the ‘Insect Limestone’
has been justifiably referred to as ‘opaque amber’.

The insects are predominantly imagines of small to medium size. The preservation of relief
and soft-bodied, immature stages (Fig. 62, p. 268) as well as the fragile Branchipus-like crustacean,
together with the arching of the upper shell bed, indicate early lithification of the matrix, but the
occasional penetration of burrows from above, and development of the small-scale sedimentary
deformation described by Daley (1971, 1972a) and evident in the limestone at all the main
localities, suggest that it was not immediate. Pantin (1958) provides evidence for the length of
time involved in the formation of diagenetic calcareous concretions. It is likely that the limestone
was formed within several thousand years of deposition of the sediment, and perhaps less judged
by the preservation of the soft-bodied arthropods. I have not observed any insect macrofossils in
the clays adjacent to fossiliferous concretions although a similar lamination is evident on fresh
surfaces.

The remarks so far are based on the north-western localities. Whilst the lithological comments
apply equally well on the east coast, there are some differences in the biota. St Helens and Priory
Bay have yielded five orders of insects but only a few more complete specimens and the occasional
ostracod and plant fragment. The author has only collected here since late 1975 and it is likely
that more will be found. However, sustained search near Bembridge Foreland has yielded only a
single insect specimen (see p. 243). Whilst the calcareous part of the Insect Bed is well developed
and can be traced south into Whitecliff Bay, no fossils were found except for two fragments of
lymnaean shell near the middle of Howgate Bay. The fossiliferous localities for insects are there-

fore confined to the northern part of the outcrop area of the Bembridge Marls (Fig. 1). Insect
remains do occur at other levels in the formation, but very few specimens have been found
to date. The A’Court Smith collection includes examples in a clay-ironstone matrix from the
north-west coast, and I have seen one or two fragments in the sideritic ironstone concretions
from the upper part of the Bembridge Marls littering the shore near Whitecliff Bay Point (Daley
1973).

Where the limestone is exposed in the lower part of a cliff or slump and is accessible to marine
erosion, it is usually well represented amongst the blocks and pebbles of the upper part of the
shore, and splitting of this débris can be more rewarding. A hand-lens is necessary for examining
fractured surfaces and small forms are more readily observed in oblique light. Specimens are
rarely exposed properly owing to the curved fracture of the rock and both halves may show
unique features. All the necessary preparation for this study was done by the author under a
microscope with an electric engraver (vibrotool) or a tungsten needle. Routine examination of
rock specimens under the microscope (x 16) also revealed a number of forms previously un-
noticed. The examination of chitinous material was often facilitated by immersing the specimen
in liquid to heighten contrast; as the matrix may contain salt, especially the softer limestone and

242 E. A. JARZEMBOWSKI

hard marl, aqueous solutions were avoided and xylene was used for short examinations and
paraffin for longer periods. Dry examination was also aided by using oblique lighting from several
directions and referring to modern specimens of related forms.

Field sections

Natural exposures of the Insect Bed are confined to coastal sections and there appear to be no
inland records from artificial excavations. The sections were examined between summer 1976
and spring 1977.

North-west coast. The ‘Insect Limestone’ is developed here only to the east of the Newtown River
(Fig. 1, p. 239). The Insect Bed is represented by a thin band of clay just west of the river behind
Hamstead Ledge and appears to be absent at Cliff End (Daley 1973a). Across the river, the
strata descend gently on the eastern limb of the Porchfield anticline and the bed is seen at 433927},
a short distance before the western end of Saltmead Ledge. It is exposed in the lower part of the
cliff for some 500 m as far east as 438928 and fossiliferous blocks occur on the shore. The bed is
0:65-0:85 m thick and four calcareous courses are recognizable (Fig. 2). The uppermost includes
some large concretions and may account for the thinness of the course below. At 440930 two
blocks of limestone were seen apparently in situ in the upper part of the foreshore. Pebbles of
‘Insect Limestone’ die out on the beach a short distance east of Saltmead Ledge (=Thorness
Point/Ledge).

2:

BW IV

I

[ly
|

|

|

{I
eer OL

.

GUR III i

GUR I-11
0:'17m above Bembridge Limestone

Figs 2-4 Sections of the basal part of the Bembridge Marls. Fig. 2, near Saltmead Ledge. Fig. 3,
immediately south of Gurnard Ledge. Fig. 4, near St Helens. (Black: limestone. Oblique shading:
hard marl. Dashed lines: laminated marly clay. Bivalve: Corbiculidae. Small gastropod: Turebia.
Large gastropod: Potamides. The bed numbers to the right of Figs 2-3 are after Daley 1973a).

The bed is next seen on the east side of Thorness (=Thorney) Bay, where a limestone course
outcrops on the upper part of the foreshore behind Sticelett Ledge (461941). The bed gradually
rises from the shore north towards Gurnard Ledge (= Gurnard Point) on the northern limb of the
Thorness syncline and is about 2 m above cliff base behind the latter (463946). It is best exposed
in the low cliff for some 140m south of the ledge and is 0:65-0:75 m thick. Two calcareous
courses are developed: a lower more or less continuous tabular band of limestone (Insect Bed of
Curry 1958) and an upper discontinuous band of concretions, often less well cemented (Fig. 3).
The clay yields sporadic examples of worn Potamides and corbiculid valves, rare examples of the
latter also being found in the calcareous part. Between Gurnard Ledge and Gurnard (=Gurnet)
Bay there is a kilometre cf tumbled cliff line in the Bembridge Beds but there are no further

! National Grid Reference, 100 km square 40 (SZ), as are the subsequent references.

PALAEOGENE INSECTS 243

sections in the Insect Bed on the northern coast of the island before St Helens. Some character-
istic limestone was however found at several points amongst the slips 500-800 m NE of Gurnard
Ledge, rising towards Gurnard from approximately 6 to 11 m above cliff base. Pieces of fossili-
ferous limestone occur amongst the shore débris as far as West Cowes (490966).

A’Court Smith worked along the coast from West Cowes to the Newtown River (White 1921)
although the specimens are generally labelled “Gurnet (or Gurnard) Bay’. The most important
collecting site was on the east side of Thorness Bay where the limestone outcropped on the shore
(Brodie 1878, Reid 1889), which suggests that it was at Sticelett Ledge, and the latter has recently
produced some relatively well-preserved material (Jarzembowski 1976). G. W. Colenutt refers
to the site as Sticelett in correspondence (ibid.) but an accompanying sketch map by the same
author is less helpful.

East Coast. Commencing in the north, a section of the Insect Bed is found NE of St Helen’s
Church (Fig. 4), 200-300 m south of Node’s Point (638900). Although numerous blocks of
limestone occur in the SW corner of the adjacent Priory Bay, no sections were observed amongst
the slips. At the Point the bed is 9 m above the base of the cliff, and 0-7 m at the southern end of
the section as the strata dip towards the axis of the Bembridge syncline. It is 0-60—0-85 m thick
with a single course of limestone concretions. The lower shell bed is only well-developed in the
southern part of the section.

Continuing south, the Insect Bed is next seen near Bembridge Foreland, 2:7 km SE of St
Helens, where a limestone course outcrops on the upper part of the foreshore amongst the
beach cover and can be traced into Howgate Bay. At 654872 the limestone is 15 cm thick and
yielded a single Oecophylla wing. Further south the bed is seen in the lower part of the cliff near
Whitecliff Bay Point, gradually ascending to the cliff top in Whitecliff Bay: the limestone becomes
replaced by a more marly course, and there are no insect remains.

Correlation

Edwards (1966) gives a summary of the stratigraphy of the younger Palaeogene strata of the
Hampshire Basin and an extensive bibliography is found in Edwards (1971). The Bembridge Marls
are near to the Eocene—Oligocene boundary, but different workers have drawn it variously above
and below the formation. Edwards (1966) followed Curry (1958) who regarded the Bembridge
Marls as of Oligocene, Lattorfian age, but as can be seen from Table 1, there is still no general
agreement on the position of the boundary. The Insect Bed (Bembridge Marls) can therefore be
considered as of late Eocene or early Oligocene age, with an absolute dating of about 35 million
years BP (cf. Cavelier & Pomerol 1977).

Table 1 The age of the Bembridge Marls according to various authors, 1966-76

Oligocene : Lattorfian . j : : . Edwards 1966

Oligocene : Lattorfian . ; ; . Curry 1966

Middle Oligocene : early Rupelian ; : . Rey 1967

Lower Oligocene . : : 3 : . Castel 1968

Eocene: Ludian . F : 3 : . Cavelier 1968

Oligocene : Tongrian . : : . Denizot 1968

Upper Eocene or Lower Oligocene : ; . Curry, Gulinck & Pomerol 1969
Oligocene : Lattorfian . : , - . Daley 1969

Oligocene : Tongrian . : : : . Haskins 1971

Late Eocene-early Oligocene : : me Neen 972

Lower Oligocene . F : 2 . Martini 1972

Late Eocene or early Oligocene : f . Bosma & Schmidt-Kittler 1972
Late Eocene or early Oligocene ‘ 3 . Bosma 1974

Early Oligocene . ; : : . Stinton 1975

Upper Eocene : late Priabonian : : . Cavelier & Pomerol 1976

244 E. A. JARZEMBOWSKI

Systematic descriptions

The family and higher classification follows the papers of Gay, Riek and Common (all 1970)
with the following departures: the Leuctrinae are regarded as a family following modern usage,
a superfamily classification of the Neuroptera is not adopted because of lack of general agreement,
and the classification of the Lepidoptera is also based on Karsholt & Nielsen (1976).

Cockerell (19176 : 12) mentioned that he had described a species of Raphidia (order Megalop-
tera) from the Insect Bed; in fact (1917a : 373) he actually described a species of Riphidia (order
Diptera : Tipulidae) which is a mis-spelling of Rhipidia Meigen 1818.

A supplementary list of the material examined in the course of this study is deposited in the Palae-
ontology Library of the British Museum (Natural History). Except where otherwise indicated,
the material is from the Insect Bed (Bembridge Marls) on the NW coast of the Isle of Wight,
between West Cowes and the Newtown River. The figures prefixed ‘I.’ or ‘In.’ are registration
numbers of the Department of Palaeontology, British Museum (Natural History) (BM(NH))
and the material is currently housed in the Department of Entomology at the same institution.
The major collections are abbreviated thus:

Sis E. J. A°Court Smith
lst R. W. Hooley
18} 2 P. B. Brodie

MIWG: Museum of Isle of Wight Geology
Other collections are named in full.
Wing veins are referred to by standard abbreviations; damaged or faint veins or margins are
indicated by dashed lines, folds by dashed-and-dotted lines and extrapolations by dotted lines.

Order ISOPTERA
Family MASTOTERMITIDAE

Genus MASTOTERMES Froggatt, 1896

TYPE SPECIES. Mastotermes darwiniensis Froggatt 1896 by monotypy; Recent, Australia.

See ge . = 1 mt ie
ewe ie “— 4 - .

a . hath:

farnee —

Fig. 5 Mastotermes anglicus von Rosen. In.24571. (Scale line = 2 mm)

PALAEOGENE INSECTS 245

at : Sitti me

re
eo

Figs 6-7 Mastotermes anglicus von Rosen. Fig. 6, lectotype (hindwing), I.15037. Fig. 7, paralecto-
type (forewing, 1.15033. (Scale line=2 mm)

Mastotermes anglicus von Rosen
Figs 5-7, 15, 20

1879 Phryganea F. Smith in Woodward : 344.
1913 Mastotermes anglicus von Rosen : 322; pl. 27, figs 5-8.
1965 = Mastotermes batheri von Rosen 1913; Emerson : 29.

EMENDED DESCRIPTION. Forewing: elongate, some incomplete specimens suggesting that a greater
elongation is possible than in the specimen figured by von Rosen (1913 : pl. 27, fig. 7). Sc short
and single. R, forking distad or basad of humeral suture, the outer branch terminating on the
anterior margin at 0-2-0-4 of the forewing length (as measured from the suture). R, and R,
elongate and separate at suture, R, sometimes forked. Stem of Rs terminating near wing apex
with a more or less marked anterior curve in its distal part. Rs with about four branches to the
anterior margin, the outer ones usually subdivided. M and Cu separate at suture, the former closer
to Rs than CuA with one to three major branches distally and perhaps several short branches
towards base. In the left forewing of In.17143 and right forewing of 1.8659, M is briefly linked
with Rs near base. Rs and CuA are closer in some specimens than in von Rosen’s figure with M
nearly median. CuA occupies about half of the wing area, dividing some ten times, the branches
frequently forking once and occasionally several times. Humeral suture arcuate with maximum
apical extension near M, and the basal part of the forewing is usually downcurved anterior to Rs.
A brown pigmentation is often preserved on the scale, parts of M and on the more anterior veins.
Forewings are preserved in 24 specimens and probably in two others.

Hindwing: elongate, rounded apically; anterior and posterior margins nearly parallel, apex
nearer the latter. Sc single, elongate, on average about five times as long as in the forewing;
some faint branches to the anterior margin may be present. R, long terminating at 0-6 of the wing
length from base, with none to four subparallel branches. R,,3 has a common stem with Rs and

246 E. A. JARZEMBOWSKI

lla 11b

Figs 8-13 Kalotermes disruptus (Cockerell) comb. nov. Fig. 8, In.64534 (forewing). See also Fig. 17.
Fig. 9, In.24619 (? hindwing). Fig. 10, In.24591 + 1n.24600 (hindwing). Fig. lla, 1.9845 (right
forewing base). Fig. 11b, 1.9845 (left forewing, anterior part of scale). Fig. 12, 1.9576 (hindwing
base). Fig. 13, 1.8644 (holotype). See also Fig. 19.

Fig. 14 Termitidae species A, In.24603. See also Fig. 35.

Fig. 15 Mastotermes anglicus von Rosen, In.24610 (hindwing base).

(Scale line = 2 mm. Figs 8-14 same scale, minor reticulation and membranous irregularities omitted).

the appearance of being the first branch of the latter. Stem of Rs more or less oblique towards
apex with two to four unforked major branches to the anterior margin. Stem of M nearer Rs
than CuA, united with Rs at base, and producing two to three major branches to the apical part
of the posterior margin. Some very short branches are developed towards the base of M. CuA
dividing ten to eleven times, the branches becoming progressively shorter towards base. CuP
single with a broad terminal fork. 1A parallel to anal fold with numerous short branches directed
obliquely towards the latter. Anal lobe unknown, probably because it is folded under or broken
away. The stem of M and the more anterior veins may show traces of brown pigmentation. As in
the forewing, there is a well-developed reticulation. In addition to the lectotype, the hindwings
are seen in six specimens and probably in two others.

Body: about 15 mm long and all parts pigmented brown as in Recent Mastotermes. Head with
a well-defined Y-shaped suture. Antennae about 5:0 mm long, segments transversely oval near
base, circular near middle and elongate distally. Pronotum widest anteriorly with sides strongly

PALAEOGENE INSECTS 247

downcurved (Fig. 20, p. 249). In.17357 has a longitudinal groove on the underside of the abdomen
near the lateral margins as in Recent Mastotermes. In the same specimen, sternite 7 is not enlarged,
indicating that it isa male. Body remains are preserved in 17 specimens.

LectoryPe. 1.15037 (S), designated herein. Fig. 6.

OTHER MATERIAL. There are 18 paralectotypes and 42 other specimens (including one from St
Helens), plus three possible specimens.

DIMENSIONS. Lectotype: wing length (apex—approximate position of ‘suture’) 25-8 mm, width
8-38 mm.

Other material. Forewing: length from humeral suture to apex (single specimen) 20-8 mm,
maximum width 7-8-8:8 mm, length of Sc 1-0-2:°8 mm, maximum length of scale 2:8—3-2 mm.
Hindwing: length from wing base to apex 27-2-29:1 mm, maximum width 8-5-8-8 mm. Body:
head, length 3-2-4-0 mm, maximum width 2-8-3-2 mm; pronotum, length 1-9-2-0 mm, maximum
width 2:8-3:8 mm.

REMARKS. Emerson (1965 : 30) states that he examined the ‘holotype’ and ‘paratypes’ of M.
anglicus and that the species was described from three specimens. Examination of the fossil
insect collection at the BM(NH) showed that three specimens of this species, labelled ‘holotyp2’
(1.15037) and ‘paratype’ (1.15033 and 1.8989), had been singled out and in addition there were a
further 18 specimens each labelled ‘paratype’. Von Rosen (1913 : 321) stated that he had examined
21 specimens of M. anglicus, and the former three correspond to the figures of his pl. 27; a foot-
note (1913 : 335) says that these are ‘nach den Typen’, but no holotype is designated in the paper.
In a reprint from von Rosen in the library of the Department of Palaeontology, BM(NH), the
registration numbers of the 21 specimens are hand written in the margin; I.15037 is indicated as
‘type’ and the remainder as ‘paratypes’. Von Rosen’s material is therefore a syntypic series and
1.15037 is here designated as lectotype. However, only 18 paralectotypes are recognized, as there
are two pairs of part and counterpart; a further 42 specimens from the Insect Bed are also referred
to this species. Eight specimens from the A’Court Smith collection, purchased in 1883 (I.15034-6,
1.15038—42) are labelled ‘Phryganea’, with or without a roman numeral in the same unidentified
hand. Woodward (1879) recorded eight specimens of the trichopterous genus Phryganea in the
A’Court Smith collection, on the basis of which a predator and prey association with Branchi-
podites vectensis was inferred by Tasch (1969). This misidentification was, however, realized not
long afterwards, for on I.8804 Brodie had written ‘Phryganea’ which he then crossed out and
replaced by ‘termite’; the specimen was purchased in 1898, a year after his death. The lectotype,
a single hindwing, is adequately figured by von Rosen (1913 : pl. 27, fig. 8) except that Sc probably
continues apically for twice the figured length but is intertwined with the fine anterior branches
from R,.

Both the wing and body remains indicate that M. anglicus is close to the living M. darwiniensis.
The wing venation of modern primitive termites is prone to individual variation and this was
also true of anglicus. Emerson (1965) was probably aware of the difficulty of separating the two
species when he stated ‘Mastotermes anglicus seems to have a greater proportional distance
between the costal border and the radial sector than does M. darwiniensis’ (1965 : 30). The distance
from the stem of Rs to the anterior margin divided by the wing width, measured near the mid-
point of the wing, gave values of 0-14—0-18 and 0-15—0-18 for the fore- and hindwings respectively
of the fossil species. This compares very well with the living species (cf. Gay 1970 : fig. 15.2;
von Rosen 1913 : pl. 26, figs 1-2; Tillyard 1931 : pl. 21, figs 1-2). M. anglicus appears to differ
from M. darwiniensis in that more branches of CuA originate directly from the main stem in the
forewing: in the latter there are five-six main branches. The soldier caste is generally more useful
for separating species in modern termites but only the reproductives are so far known from the
Isle of Wight.

Present day Mastotermes is naturally confined to tropical northern Australia, although it has
been introduced to New Guinea, its southern limit being the Tropic of Capricorn. The termite
avoids high-rainfall areas and is absent from rainforest and areas where such forest has been

248 E. A. JARZEMBOWSKI

cleared. It is very destructive of timber, although some trees including Ficus spp. are hardly
affected; this genus is known from the Insect Bed (Reid & Chandler 1926; personal examination
of matrix). The termite normally nests in or under the boles of trees, or in logs or stumps, and
produces galleries in the soil. Alates are released in early November and have been recorded
through the summer monsoon season (Gay & Calaby 1970 : 395-396).

“

&.
SS ‘
% SS ad a

~ Lae Sn, Soe i em <i f nt aes)
16 fee es Ne - f e <a
tts St agg : é
Fig. 16 Kalotermes disruptus (Cockerell) comb. nov. 1.9845. (Scale line = 2 mm)

— ry +.

Fig. 17 Kalotermes disruptus (Cockerell) comb. nov. In.64534, forewing. See also Fig. 8.
(Scale line=2 mm)

E = # ‘ * AEs Ez:
Fig. 18 1In.64535, incomplete counterpart of a termite wing from the Insect Bed to show preser-
vational similarity to Fig. 19.
Fig. 19 Kalotermes disruptus (Cockerell) comb. noy. Holotype, posterior margin uppermost. See
also Fig. 13. (Scale line = 2 mm)

PALAEOGENE INSECTS 249
Sc+R

CuA CuP =

Fig. 20 Mastotermes anglicus von Rosen, In.43415, head and pronotum. Fig. 20a, dorsal view.

Fig. 20b, lateral view.
Fig. 21 Reticulitermes sp., In.64535 (forewing). Venation and main reticulation only. See also

Fig. 26.
Fig. 22 Leuctra priscula (Cockerell) comb. nov. Forewing venation. Holotype, In.17498.

(Scale line = 2 mm)

Family KALOTERMITIDAE
Genus KALOTERMES Hagen, 1853

TYPE SPECIES. Termes flavicolle Fabricius 1793, by subsequent designation; Recent, western
Palaearctic.

Kalotermes disruptus (Cockerell) comb. nov.
Figs 8-13, 16-19

1917a_ Sisyra(?) disrupta Cockerell : 381; pl. 31, fig. 13.

EMENDED DESCRIPTION. The holotype consists of an apical fragment of a wing with parts of Cu, M
and possibly Rs preserved. The following account is based on a series of specimens considered to
be conspecific with the type.

The body is known from a single dorsal impression (1.9845). The head is missing except for
traces of the basal region which indicate that it was inclined downwards at burial. Pronotum
gently tapered anteriorly and with sides slightly curved ventrally; meso- and metanota subequal.
Abdomen with nine visible segments, the ninth curved anteroposteriorly with blunt posterior
angle. A single leg is preserved: tibia with remains of an inner terminal spur; tarsus with elongate
distal segment and short basal segments.

Wings elongate, gradually widening towards the apex with maximum width at 0:3 of the length
from the latter; apex rounded. R, single and close to the anterior margin. Costal area (stem of
Rs to anterior margin) moderately wide and frequently downcurved especially near base as in
Recent termites, and consequently difficult to examine. Rs with 3-6 oblique, single branches to
the anterior margin. Cu occupies just over half of the wing area, branching 8-10 times. Branches
of Cu running obliquely to the posterior margin and less inclined than those of Rs : they may
fork once or occasionally twice at a variable distance from base. M nearly parallel to the stems of
Rs and Cu, and slightly closer to the former. M may have a longish apical fork or two to three
short terminal branches and may unite with Rs for a short distance apically. The terminations of
Rs, M and Cu are often difficult to identify in the variable pattern of short veins or reticulations
near the wing apex.

250 E. A. JARZEMBOWSKI

The wing membrane is finely tuberculate (in dorsal aspect) posterior to Rs; the sculpture may
be well-defined (Fig. 17), or only a general irregularity of the surface may be discernible. A few
crossvein-like reticulations are also developed. R,, much of Rs and the basal portions of M and
Cu often show traces of a dark brown pigmentation apparently corresponding to the more
sclerotized areas in life; body light brown.

Fore- and hindwings are not readily identifiable from incomplete specimens but the former are
represented by 1.9845 and In.64534, and the latter by I.9576, In.24591 +In.24600 and In.24621.
The humeral suture in the forewing is gently curved towards base posterior to R,; the forewing
scale has a strongly curved anterior margin and overlaps the base of the hindwing scale (1.9845).
The hindwing suture is shorter and more arcuate, Rs and M separate distad of the suture, A
present.

Ho.otyPE. 1.8644 (B). Figs 13, 19.
OTHER MATERIAL. 14 specimens and possibly one other.

DIMENSIONS. Wings: length (distance from humeral suture—apex) c. 6-0-8-1 mm, maximum width
2-0-2:4 mm.

Body (1.9845 only): length of thorax 2 mm, length of pronotum 0-7 mm, maximum width of
pronotum |-3 mm, maximum width of meso- and metathorax | mm; length of abdomen 2:6 mm;
length of tibia 0-6 mm.

REMARKS. Cockerell described this species from a single wing fragment, and on the basis of a
comparison with Recent Sisyra vicaria (Walker) it was placed in the Neuroptera (Hemerobiidae:
Sisyrinae). He was, however, doubtful about its generic placement (Cockerell 1917a : 381-382).

Comparison of the fragment with S. vicaria and Recent Sisyra spp., and with fossil insects
from the Insect Bed in the BM(NH), indicates that the species should be transferred to the
Isoptera, because

i. The vein pattern does not agree with Sisyridae (cf. Parfin & Gurney 1956);

i. Neither veins nor wing margin show any trace of macrotrichia or trichosors which are
present in sisyrids (/bid.); preservational failure is improbable when compared with other Insect
Bed Neuroptera;

ill. The veins are relatively thick and lacking in pigmentation, unlike the Insect Bed Neurop-
tera;

iv. S. vicaria and the Insect Bed Neuroptera may show a faint, linear fold in the membrane
between two veins, but this is different from the membranous irregularities in Cockerell’s
specimen;

vy. The fragment compares best with the negative impressions of wings belonging to a hitherto
unrecognized termite in which the venation is well developed near the apex.

The type and new material are referred to the family Kalotermitidae because the pronotum is
flatly arched and not narrower than the head, the anterior wing scale overlaps the hindwing scale,
Sc is short and sclerotized in the forewing and absent in the hindwing, R, is sclerotized and single,
Rs is sclerotized and parallel to the anterior margin with a number of branches to the latter, Cu
is unsclerotized with numerous branches to the posterior margin and an anal vein is present in the
hindwing only (Krishna 1961 : 315).

The wing sculpture is similar-to Recent G/yptotermes Froggatt, Rugitermes Holmgren and
Kalotermes, and of these three, the venation comes closest to Kalotermes (Krishna 1961). However,
the relatively small number of branches of Rs in the forewing (Fig. 8) is more like Recent Incisi-
termes Krishna than Kalotermes (Krishna 1961 : 334, 354). Krishna erected the former genus for
a segregate of Kalotermes in which one of the diagnostic characters is the absence of tuberculation
of the wing membrane. Emerson (1969 : 43, 49) accepted the new genus but rejected the distinction
with regards to wing sculpture. Examination of the alates of K. flavicollis (Fabricius), K. rufinotum
Hill, /. minor (Hagen) and /. tabogae (Snyder) suggests to me that sculpture does not always
allow a definite separation of the two genera (Figs 23-24). In Kalotermes rhenanus Hagen from
the Oligocene of Rott, Germany, Rs has 5-8 branches in the forewing and 3-4 in the hindwing

PALAEOGENE INSECTS 251

(Emerson 1969 : 34), which compares well with the Bembridge Marls species (Figs 8-10). How-
ever, there appears to be no indication of wing sculpture in K. rhenanus. The costal area of the
forewing is too poorly known in disruptus for detailed comparison with the Recent genera.
Emerson (1969 : 29) expresses some uncertainty about distinguishing alates of the two genera
in fossils where the dentition is unknown, and the Bembridge Marls species is therefore referred
to Kalotermes in the wider sense.

The wings of kalotermitids are especially subject to variation (cf. Fuller 1919 : pl. 8, pl. 9 figs
124-134), and no two examples in Kalotermes disruptus (Cockerell) comb. nov. are likely to be
exactly similar in the finer details of venation and reticulation. Recent kalotermitids are xylo-
phagous, predominantly dry-wood termites, digestion being made possible by an abundant
intestinal fauna of flagellate Protozoa. Incisitermes and Kalotermes are both hardwood termites,
the former occurring in the New World and across the western Pacific and the latter with a
cosmopolitan distribution (Emerson 1969 : 45; fig. 11). Incisitermes is warm temperate (sub-
tropical)-tropical, but Kalotermes is mainly temperate (ibid : 45, 49), although preferring the
warmer part of that zone (ibid : fig. 7). The Kalotermitidae are represented in Europe at the
present time by a single species restricted to southern parts (Harris 1970 : 297).

Fig. 23 Kalotermes flavicollis. Recent; forewing. Oblique top lighting.
Fig. 24 Incisitermes tabogae. Recent; forewing. Oblique top lighting, same scale as Fig. 23. (Scale
line = 2 mm)

Family RHINOTERMITIDAE
Subfamily HETEROTERMITINAE

Genus RETICULITERMES Holmgren, 1913

Type species. Termes flavipes Kollar 1837 by subsequent designation; Recent, Holarctic.

Reticulitermes sp.
Figs 21, 26

DESCRIPTION. The species is only known from wing material. Wings elongate, forewing with a
length/width ratio of 4-3-3:1. Humeral suture nearly straight in forewing (In.17119, In.64535).
Sc+R prominent forming the anterior margin. Rs prominent, close to Sc+ R, terminating near

sy) E. A. JARZEMBOWSKI
m-
& : =I,

25

Fig. 25 Reticulitermes tibialis. Recent; forewing. Transmitted light. (Scale line=2 mm)

taal cc ete ail ae sa ge

Fig. 26 Reticulitermes sp., In.64535. See also Fig. 21. (Scale line =2 mm)

the wing apex. Cu occupies nearly half the wing width, branching 8-10 times, some branches
dividing again distally. M single, closer to Cu than Rs and parallel to the former, with a posterior
curve apically.

Membrane reticulate posterior to Rs; a few incipient, transverse ridges may be discernible
near the apical extremities of Rs and Sc+R. The reticulation is best developed in the distal part of
the Rs space, consisting of oblique transverse ridges tending to join medially to form an irregular
longitudinal ridge. Wing colouration : Sc+ R and Rs brown along much of their length; Cu and
M only pigmented near the suture.

In.24623 and In.25047 are relatively small wings and may represent a separate species.

In.24821 is probably a hindwing as Rs and M appear to meet just distad of the humeral suture
and the wing is comparatively narrow near base, but the specimen is too poorly preserved to be
referred to this species with certainty.

MATERIAL. Seven specimens and possibly one other.

DIMENSIONS. Forewing length 6:0-6-6 mm, width 1-5—2:0 mm. Length of humeral suture (fore-
wing) 0:5 mm. (In.24623 length 4-5 mm, width 1-3 mm. In.25047 length 5 mm, width 1-4 mm).

REMARKS. The presence of reticulation, a comparatively reduced venation with Rs simple and
parallel to the anterior margin, and general habitus place the wings in the Rhinotermitidae :
Heterotermitinae. A well-defined, irregular transverse reticulation is like Reticulitermes Holmgren
(Fig. 25) and unlike Heterotermes Froggatt (Banks & Snyder 1920 : 42; Hill 1942 : 121; Weesner
1970 : 495).

Recent Reticulitermes spp. are xylophagous insects tunnelling in pieces of damp wood which
are frequently buried; adjacent pieces are linked by subterranean galleries (Noirot 1970 : 77).
The Pyrsonymphidae, a family of intestinal Protozoa, are confined to Reticulitermes. The genus is
mainly Holarctic, but there are a few oriental species (Roonwal 1970 : 334). It is essentially
temperate; it normally occurs to latitude 46°N in Europe, a little further north than Kalotermes
(Harris 1970 : 299) and occurs as far south as the borders of Vietnam where it inhabits mountain-
ous regions (Harris 1968).

PALAEOGENE INSECTS 253
Family TERMITIDAE

Species A
Figs 14, 35

DESCRIPTION. The material consists of a single wing broken off at the humeral suture. Wing
narrow basally, apex rounded; humeral suture short, straight. Sc+R prominent, gently curved.
Rs prominent, close and parallel to Sc+ R for most of its course terminating just before the wing
apex. Rs with a short, forked apical branch, but this is weak and not pigmented like the main
stem. M separates from R a short distance beyond the suture; Cu relatively short, terminating
near the mid-length of the wing and branching five times. Sc+R, Rs and base of Cu are brown-
coloured. There is some faint reticulation of the wing membrane especially near the apex.

MATERIAL. In.24603 (H) (Figs 14, 35) and counterpart In.24596 (H).
DIMENSIONS. Maximum length 5:5 mm, width 1-6 mm. M divides 2:0 mm from base.

REMARKS. The simple venation, M well separated from Rs and close to Cu, and reduced reticula-
tion are indicative of the higher termites (Termitidae). This is the largest family of Isoptera at
the present day and is essentially tropical although some genera range into warm temperate
regions (Harris 1970 : 309). Termitids are humivorous or phytophagous, the latter including
wood feeders; nesting habits are varied (Noirot 1970).

The wing venation of the fossil approaches that of extant Apicotermes occultus Silvestri
(Apicotermitinae); the reticulation suggests some affinity with the Macrotermitinae, e.g. Micro-
termes spp. The systematic position of the fossil within the family is uncertain and it may well
represent an extinct genus.

Family Uncertain

In.24417 (H). Basal half of a wing, 9 mm long, with only the posterior part preserved, showing
part of Cu and several branches. Membrane slightly tuberculate.

1.8648 (B), 1.10204 (B), In.24414 (H), In.24612 (H) and counterpart In.25358 (H), In.24616 (H)
and In.24618 (H) are indeterminate wing fragments belonging to the Kalotermitidae or Rhino-
termitidae.

Order PLECOPTERA
Suborder FILIPALPIA
Family LEUCTRIDAE

Genus LEUCTRA Stephens, 1836

TyPE SPECIES. Phryganea fusca Linné 1758 by subsequent designation; Recent, Palaearctic.

Leuctra priscula (Cockerell) comb. nov.
Fig. 22

1922 Nemoura (s. lat.) priscula Cockerell : 34; fig. 2.

EMENDED DESCRIPTION. Single specimen consisting of an incomplete forewing with the outer
half and most of the anterior area preserved. Estimated original length 7 mm. Sc reaches anterior
margin 2:75 mm from wing apex. Rs, M and Cu two-branched. Crossvein r—m originating just
basad of the fork of Rs, m-cu from Mj ,, just distad of the fork of M. Four Cu crossveins
preserved, the only crossveins apical to the cord. Sc, and i.r nearly continuous and perpendicular
to anterior margin.

Ho otype. In.17498 (S). Fig. 22.

REMARKS. The venation is plecopteran of the Leuctra type (Needham & Claassen 1925 : pl. 2,
fig. 21). It is distinguished from Nemoura Latreille by the absence of an oblique crossvein in the
apical marginal space (ibid. : 196). The fossil is referred to Leuctra Stephens s. /at., the material

Fig. 27 Sisyridae species A, forewing. In.17287.

Fig. 28 Sisyra fuscata (Fabricius). Recent; forewing.

Figs 29-30 Promantispa relicta (Cockerell) gen. and comb. nov. Fig. 29, holotype, In.24326.
Fig. 30, In.24597 (forewing).

Fig. 31 Hemerobius tinctus n. sp. Holotype, 1.9160.

Fig. 32 Hemerobiidae species B. The wings are in fact at a slight angle to the body but are drawn
as if coplanar. In.17173.

Fig. 33 Hemerobiidae species A. In 43477.

Fig. 34 Hemerobiidae species C. In.17445.

(Scale line =2 mm)

PALAEOGENE INSECTS DSS

being too fragmentary for certain reference to any of the subgeneric or generic segregates of
Leuctra currently recognized by entomologists.

Leuctrid adults are found amongst stones and foliage near freshwater streams, rivers and lakes
which are the habitat of the aquatic nymphs. Both adults and young are herbivorous, the former
feeding on lichens and small algae (Claassen 1931 :97; Hynes 1941 : 529, 534, 547-548). The
family is mainly Holarctic and the common name of needleflies stems from the adult habit of
rolling the wings around the body in repose.

Order NEUROPTERA
Family MANTISPIDAE

Genus PROMANTISPA nov.

DiaGnosis. A Palaeogene genus of Mantispinae differing from Mantispa Illiger in the possession
of an apically extended pterostigma.

Type Species. Mantispa relicta Cockerell 1921b.

Promantispa relicta (Cockerell) comb. nov.
Figs 29, 30

1921b Mantispa relicta Cockerell : 477-478; fig. 46.

EMENDED DESCRIPTION. The original description is based on a single wing fragment 7 mm long,
preserving the distal portions of the branches of M and five branches of Rs, plus part of a gradate
series of crossveins (Fig. 29). The veins are coloured brown and macrotrichial bases are visible
on the veins and wing margin.

The discovery of an unrecognized left forewing (In.24597+ In.24631) requires a redescription
of the species. During fossilization, the wing was broadly folded about a longitudinal axis and
Fig. 30 is based on a combination of three camera lucida drawings from slightly different positions.
Original length estimated at 15 mm. Costal area with basal expansion as usual in mantispid
forewings, with two well-separated crossveins preserved. Sc curved posteriorly in its apical
portion to form the posterior boundary of the pterostigma. The latter is pale brown, very elongate
and blunt apically. Sc deflected towards R, just distad of 2nd r,-1s indicating the position of a
short pterostigmal crossvein at 0-4 of the length of the pterostigma from base. R, on prominent
wing fold, close to Sc basally, and producing three oblique crossveins to the anterior margin
beyond the pterostigma. Rs undulose, originating 10:0 mm from wing apex, with six branches
to the outer posterolateral margin and three equidistant crossveins to R, partitioning the branches
1, 2, 2, 1. M dividing nearly opposite R with a short r—m crossvein a small distance distad of the
fork; an oblique m-—cu just basad of the fork is continuous with an incompletely preserved
cubital crossvein. Branches of Rs, M and CuA, end-twigged and linked by one gradate series of
crossveins; branches of Rs and MA undulose. CuA, short and apparently single.

35

Fig. 35 Termitidae species A, In.24603. See also Fig. 14. (Scale line =2 mm)

E. A. JARZEMBOWSKI

Sar ae
fk Pee
&.¢ Lhe een ee emt
~ prnmiaaia - x
-
= ba ee sence
ae tt # mt
c eran ¥

*

ee ee * : 36.

Fig. 36 Bittacus veternus (Cockerell) comb. nov. Holotype, In.24330, ventral half. See also Fig. 44.
(Scale line = 2 mm)

37 ya

Fig. 37 Mantispa tenella Erichson. Recent. (Scale line =2 mm)

os

by . mM
Si oF £
- ~ be % a
4 ; ee é
. SA Rg
\,
5 \ Ms

og . 38
2 *
> ¥

as

e Wiis ace
ing. In.17464. (Scale line =2 mm)

Fig. 38 Prohepialus sp., w

PALAEOGENE INSECTS 257

Veins dark brown; macrotrichial bases visible on veins, pterostigma and wing margin; macro-
trichial hairs preserved along the margin near the apex. Trichosors absent.

Ho vorype. In.24326 (H.15381). Fig. 29.
OTHER MATERIAL. In.24597 (Fig. 30) and incomplete counterpart In.24631 (H).
DIMENSIONS. Forewing: maximum length 11:5 mm, width 3:5 mm.

REMARKS. The Mantispa-like venation and absence of trichosors place the fossil in the
Mantispidae : Mantispinae (Tjeder 1959 : 275). However, the pterostigma in Mantispa is
relatively short (Handschin 1959 : 198) and usually has an obliquely truncate apical margin
just beyond the pterostigmal crossvein. The pterostigma in the fossil bears some resemblance to
Ditaxis biseriata (Westwood), but the venation of the latter is quite different. Comparison of the
fossil with species representing all the known genera of Mantispidae (about 38) indicates that a
new genus is required. Promantispa relicta (Cockerell) gen. and comb. nov. appears to be the
only known fossil representative of the family.

Recent mantispids (mantis flies) are found in all major regions including Europe, but they no
longer occur in Britain. The adults are found on vegetation where they are predacious on other
insects, the raptorial forelegs (Fig. 37) having a striking convergence with those of mantids
(praying mantises) and berothids (Rachiberothinae). The older larvae of Mantispinae enter the
egg sacs of spiders where they feed, hypermetamorphose and pupate (McKeown & Mincham
1948; Tjeder 1959 : 260, 273); the Order Araneida is represented in the Insect Bed.

Family HEMEROBIIDAE
Genus HEMEROBIUS Linné, 1758

TYPE SPECIES. Hemerobius humulinus Linné 1758, by subsequent designation; Recent, Holarctic.

Hemerobius tinctus sp. nov.
Fig. 31

DiaGnosis. A Palaeogene species of Hemerobius differing from H. humulinus in that the inner
branch of Rs is divided near base and there is an sc-r crossvein near the base of the middle
branch of Rs.

DESCRIPTION. The species is known from a single forewing and is smaller than Hemerobius
humulinus (Killington 1936: pl. 13) with an estimated forewing span of 11 mm. Costal area
narrow, humeral vein recurved with four single branches to the anterior margin. Sc terminating on
anterior margin and branching about 17 times. Rs arising from R on three branches, inner
branch dividing close to base and outer branch separating near mid-point of wing; the latter
branch with three main forks. Sc linked to stem of R by two crossveins immediately basad of the
origins of the middle and inner branches of Rs. No basal r—m nor basal crossvein between the
branches of Rs. MA and MP separating near the base of M, forking distally. CuA branching
four times distally, CuP with a single fork. Two gradate series of crossveins present, outer series
extending from R, to 1A. Veins with macrotrichia and usually end-twigged marginally. Trichosors
present. Colour pattern consisting of a light brown pigmentation of the membrane especially in
the outer part of the wing and tending to form transverse bands.

Ho orypee. 1.9160 (B). Fig. 31.
DIMENSIONS. Forewing length 5-0 mm, maximum width 2:2 mm.

REMARKS. Rs arising on more than two stems, Sc terminating on the costal margin and general
form of the wing place it in the family Hemerobiidae (Riek 1970). The fossil is close to the extant
Hemerobius humulinus (Killington 1936 : fig. 11; pl. 13, fig. 3) but comparison with this and other
Hemerobius spp. in the BM(NH) indicates that the basal fork of the inner branch of Rs and the

45

Fig. 39 ? Neuronema species A. In.17404.

Fig. 40 ? Neuronema species B. 1n.64536.

Fig.41 Nothochrysa capitata (Fabricius). Recent; hindwing tip. The crossveins in the subcostal space
are not prominent and less likely to be preserved if fossilized (cf. Fig. 42).

Fig. 42 Chrysopidae species A. 1.8643.

Fig. 43 Bittacus sinensis Walker. Recent; forewing tip.

Fig.44 Bittacus veternus (Cockerell) comb. nov. Holotype, In.24330. See also Fig. 36.

Fig. 45 Bittacus chlorostigma McLachlan. Recent; outer part of forewing.

Fig. 46 Bittacus species A. 1n.64537.

(Scale line=2 mm; Figs 39-41 are to same scale.)

PALAEOGENE INSECTS 259

crossvein near the base of the middle branch of Rs is unusual for the genus and may form the
basis of a supraspecific grouping. The colouration of H. tinctus sp. nov. resembles Recent H.
atrifrons McLachlan.

Recent hemerobiids (brown lacewings) are terrestrial insects and both adults and larvae prey
on small soft-bodied arthropods. They are commonly arboreal but also occur on low herbage.
The genus Hemerobius has a cosmopolitan distribution and can be economically important in
the biological control of aphids (Balduf 1939 : 251-253, 256-257, 259; Tjeder 1961 : 296).

Genus NEURONEMA McLachlan, 1869

TyPe SPECIES. Hemerobius decisus Walker 1860, by original designation; Recent, Hindustan.

? Neuronema sp. A
Fig. 39

DESCRIPTION. The species is known from a single impression of an incomplete left forewing,
original length estimated at about 12 mm. Costal area very broad near base, narrowing apically
and with a distinctive longitudinal line as in some Recent hemerobiids. Costal crossveins numer-
ous, usually forking once and often twice in their distal halves and becoming increasingly oblique
towards apex. The first seven crossveins are evenly truncate at base, indicating the course of a
strongly recurved humeral vein, although the vein itself is difficult to discern. Sc and stem of R
are marked by strong folds in their proximal portions, flattening out distally and curving towards
the apex. Subcostal area with a slight basal expansion. Five branches of Rs can be distinguished
and the spacing indicates that this is close to the actual complement. Four to five branches of Ry
are discernible. The median area is bounded by a faint fold on its outer side and a prominent
fold on its inner side. M and Cu indicated by some vague traces of venation. Colouration : dark
brown mottles in costal area and traces of an oblique brown band across the wing; similar browns
are typical in the wing colouration of Recent hemerobiids.

MATERIAL. In.17404 (S). Fig. 39.
DIMENSIONS. Forewing length 10:5 mm, maximum width 4:8 mm.

REMARKS. Rs arising on several branches from R is indicative of the Hemerobiidae (Fraser 1959 :
18; Riek 1970 : 482). The wide costal area, recurved humeral vein, Rs with four to six or seven
branches, pectinate branching of R, and form of the costal crossveins place the fossil close to
Recent Drepanacra and Neuronema (Nakahara 1960 : 61-62; pls 15-16), especially the latter
oriental genus (e.g. Neuronema irrorata Kimmins).

Drepanacra Tillyard is an Australian genus and known to feed on Neomyzaphis Theobald
(Balduf 1939).

? Neuronema sp. B
Fig. 40

DESCRIPTION. The fragmentary forewing is of a similar size and colouration to Species A. The
costal area differs in that there are a few longitudinal crossveins towards base. A basal crossvein
between Sc and R is present as well as two more apical crossveins. Rs arises on four branches,
R, with three well-spaced forks. R, with basal crossvein to R stem, R, with three crossveins to
R, stem. There are also several crossveins in the apical portions of R,-R,. M divides near base.

MATERIAL. 1In.64536 (E. A. Jarzembowski, J. Cooper and S. F. Morris). Fig. 40.

REMARKS. Species B is close to Species A but differs principally in the presence of the longitudinal
crossveins in the costal area and R, appears to have fewer branches.

260 E. A. JARZEMBOWSKI

Genus indet.

Species A
Fig. 33

DESCRIPTION. The species is known from two fragments of the apico-anterior region of the wings.
Forewing: three branches of Rs preserved, the outer branch apparently with only two main forks.
Inner gradate series represented by two separate crossveins, the anterior of which is well distad
of the above-mentioned forks. Trichosors present. The colouration consists of brown mottles.

MATERIAL. In.43477 (H). Fig. 33.

REMARKS. The fragmentary specimen may well belong to Hemerobius but is too incomplete for
its generic affinities to be ascertained. The species differs from H. tinctus sp. nov. in its larger size
(the distance from the origin of the outer branch of Rs to the apical end of Sc is about 3-3 mm
compared with 1-7 mm in H. tinctus sp. nov.) and position and incompleteness of the inner
gradate series.

Species B
Bip 32

DESCRIPTION. The species is known from a single body impression and forewing fragments. Head
wide, with globular eyes and moniliform antennae. The abdominal sclerites are well preserved
with a well-defined membranous area between the tergites and sternites, the modification of
tergites 8-10 being indicative of a female (cf. Killington 1936 : fig. 13). Forewing: costal area
moderately narrow, Rs apparently arising on more than one stem; outer branch of Rs with two
main forks and two crossveins to R, a little distad of the forks. Trichosors present. The mem-
brane is pigmented light brown in the pterostigmal area.

MATERIAL. In.17173 (S). Fig. 32.
DIMENSIONS. Length of body 4:0 mm.

REMARKS. The generic position of the species is uncertain, like Species A, but it is readily separated
from the latter by the wing colour-pattern.

Species C
Fig. 34

DESCRIPTION. Dorsal impression of a fragment of a right forewing from near the base of the wing,
maximum length 3-0 mm, width 1-5 mm. Sc on prominent upfold with the bases of eight oblique
costal crossveins preserved. Stem of R slightly deflected posteriorly before converging on Sc
basally; a transverse groove possibly represents a basal crossvein. Rs represented by four branches
with fine parallel folds in the wing membrane between them. Some brownish pigmentation of the
veins and membrane.

MATERIAL. In.17445 (B). Fig. 34.

REMARKS. Rs arising on several branches, the basal expansion of the subcostal area and general
appearance of the fragment would place it in the Hemerobiidae (cf. corresponding area in
Fig. 39, p. 258). The generic affinities of the fragment are indeterminate but size comparison with
? Neuronema sp. A indicates a species with probable wing length of about or a little less than
two-thirds of the latter.

Family CHRYSOPIDAE
Subfamily CHRYSOPINAE

Species A
Fig. 42

DESCRIPTION. Dorsal impression of an apical fragment of a right hindwing, maximum width

PALAEOGENE INSECTS 261

3-1 mm, length 5 mm. Estimated original length 13-16 mm. Wing membranous; apex pointed,
relatively sharp. Macrotrichial bases visible on wing margin and veins, especially numerous along
C. Sc marked by prominent fold; Sc and R, curved towards apex distally. Rs zigzagged, with
eight branches preserved, the basal four showing end-twigging. Rs linked to R, by a number of
crossveins; branches of Rs joined by two series of gradate crossveins.

MATERIAL. 1.8643 (B). Fig. 42.

REMARKS. R, distinct from Rs but linked by numerous crossveins, branches of Rs arising from a
single (zigzagged) stem and the absence of trichosors are characteristic of Chrysopidae (Fraser
1959 : 15; Riek 1970 : 482). The reduced venation excludes the fossil from the Apochrysinae
(cf. Kimmins 1952) and places it in the Chrysopinae in the wider sense (Riek 1970). The specimen
is too fragmentary for more detailed determination. A Recent wing is illustrated in Fig. 41 for
general comparison.

Recent chrysopids (green lacewings, etc.) are found in both warm and temperate regions of
the world (Tjeder 1966 : 243). The larvae and adults (Chrysopinae s. /at.) are predominantly
arboreal although they also occur on low vegetation, where the larvae and many of the adults
prey on small, soft-bodied arthropods, mainly insects (Balduf 1939 : 271-272, 291; Tjeder
1966 : 228).

Family SISYRIDAE

Species A
Fig. 27

DESCRIPTION. The species is known from a single left forewing, estimated original length 5-0 mm.
The wing has been folded longitudinally in two upon itself, but the two halves are readily dis-
tinguished by (a) the general orientation of the veins and (b) the veins of the anterior portion being
convex and ridge-like and the posterior ones concave since they are seen from the underside.
Costal area moderately wide, macrotrichial bases visible along the anterior margin and on some
crossveins. Sc curved towards R, apically, costal crossveins numerous, unforked. R forked near
wing base, R, parallel to Sc basally but not preserved apically. Rs branching four times with R,
apparently divided. Short inner gradate series of crossveins present and traces of an outer series.
MA and MP divided apically, branches forked. CuA with seven short distal branches, the latter
simple except for one which appears to be forked. CuA linked by prominent crossveins to CuP
basally and MP apically. Colouration: veins and pterostigmal area light brown. The membrane
shows traces of a similar pigmentation in the middle and outer parts of the wing.

MATERIAL. In.17287 (S). Fig. 27.

REMARKS. The wing form and venation of the fossil are closest to the extant genus Sisyra Bur-
meister (Fig. 28); Sc curving to terminate on R, distally is typically sisyrid (Nakahara 1958). The
inner gradate series resembles that of Sisyra panama Parfin & Gurney and the fork of R, 1s
similar to S. vicaria (Walker). Comparison with Recent species in the BM(NH) shows that the
fossil differs from Sisyra in that the costal crossveins and branches of CuA are more numerous
and the apical r,-rs crossvein appears to be immediately below rather than beyond the termination
of Sc. The branching of CuA is more like the Nearctic sisyrid Climacia areolaris (Hagen) and
the costal crossveins resemble Neurorthus spp. The Neurorthinae have been removed from the
Sisyridae in recent years and elevated to family rank (Zwick 1967). The larvae of both families live
in fresh water except those of Austroneurorthus Nakahara which probably inhabit damp places;
adults are found near the larval habitat. The Sisyridae have a cosmopolitan distribution and the
larvae are parasitic on fresh-water sponges (Porifera : Spongillidae) (Parfin & Gurney 1956;
Zwick 1967; Riek 1970). Both neurorthids and sisyrids are known in the Baltic amber and a
number of specimens await description (Macleod 1970). The systematic position of the amber
‘sisyrid’ Rophalis relicta Hagen is unsettled (Parfin & Gurney 1956; Nakahara 1958).

262 E. A. JARZEMBOWSKI
Family Uncertain

In.20557 (B). Fragment probably of the costal area of a wing, preserving a straight portion of the
wing margin and distal parts of 22 crossveins. Length 5mm, width | mm. Crossveins slightly
inclined to the margin, evenly spaced, and single except for one with a terminal fork. Macro-
trichial bases dense on wing margin and also evident along the veins, no trichosors. Membrane
pale brown, veins and margin dark brown.

Order MECOPTERA
Suborder EUMECOPTERA
Family BITTACIDAE

Genus BITTACUS Latreille, 1805

TYPE SPECIES. Panorpa italica Miller 1766, by subsequent designation; Recent, western Palaearctic.

Bittacus veternus (Cockerell) comb. nov.
Figs 36, 44

19216 Panorpa veterna Cockerell : 478; fig. 47.

EMENDED DESCRIPTION. The species is known from the apical part of a right wing, original length
estimated at 12-14 mm. Wing membranous, slightly folded about R;. Macrotrichia visible along
part of anterior and posterolateral margins. R, represented by the curved apical end; Rs four-
branched, R,,; forking 4mm from wing apex, R,,3 opposite the end of R,. R;_; linked to
each other by two transverse crossveins. M four-branched, M, ,. forking a little nearer the apex
than R,,;; M, and M, converging but fork not preserved. First r-m oblique, originating close
to the base of R;; Ist 1.m. also oblique commencing near the end of the r—m but 1s incompletely
preserved. R; and M, _, are also linked to each other by two transverse crossveins with one to M,.
CuA represented by a small apical fragment. Wing colouration: veins brown; dark brown areas
around the end of R, (pterostigma) and the transverse R crossveins; paler brown around the
transverse M crossveins. Wing tip also dark. The pigmentation is similar in colour to Recent
Mecoptera.

/

HovotyPe. In.24330, with counterpart (H). Figs 36, 44.
DIMENSIONS. Maximum width 3 mm, length 5 mm.

REMARKS. The reduced venation excludes the fossil from Meropeidae and Notiothaumidae
(Grassé 1951 : 107) and places it in the Eumecoptera. Cockerell (19215) stated that it resembles
the living Panorpa germanica L. (Panorpidae) except in details of colour pattern. I have examined
a series of specimens of this European species in the museum collection, and in addition the fossil
differs in that R, is short and does not fork and that Ist r-m is more or less continuous with Ist
i.m. which is directed towards the anastomosis of M3 ,4. In P. germanica, and Panorpa Linné in
general, R, is usually forked (Carpenter 1931 : 213) although there are exceptions. Comparison
of P. germanica with other panorpid species, e.g. in Esben-Petersen (1921), suggests that Ist i.m.
is not normally present close to Ist r-m and fork M, ,, is more basal in Panorpidae.

The venation of the fossil in fact agrees best with that of the Bittacidae (Esben-Petersen 1921 :
115-162), and I have compared the specimen with published accounts of the type species of the 12
currently recognized bittacid genera, as well as some additional species in the larger genera
(Bittacus, Harpobittacus Gerstaecker). The shape and venation compares well with Austrobittacus
Riek, Bittacus (e.g. B. sinensis Walker, Fig. 43), Kalobittacus Esben-Petersen, and possibly
T ytthobittacus Smithers. The absence of pterostigmal crossveins in the fossil could be preserva-
tional failure, as these may be faint in Recent specimens, and are sometimes obscure (Esben-
Petersen 1921] : 127). The size corresponds to more than one of these genera: in Bittacus a Baltic
amber species, B. minutus Carpenter, has wings of 3 x 11-5-13 mm. The Bembridge Marls species
approaches the living B. marginatus Mikayé in the pigmentation of the pterostigma, apex and
more apical crossveins, and is provisionally referred to the genus Bittacus Latreille.

PALAEOGENE INSECTS 263

Recent bittacids (hanging flies) are found on vegetation in sheltered places and are predacious
on other insects. The larvae are ground-dwelling scavengers, pupating beneath the surface
(Grassé 1951; Byers 1971). The family is temperate-tropical (Carpenter 1931 : 206) with a world-
wide distribution except for the northern parts of Europe, Asia and America (Esben-Petersen
1921 : 116). It is no longer found in Britain (Fraser 1959) although represented on the continent
(Grassé 1951).

Bittacus sp. A
Fig. 46

DESCRIPTION. The species is known from a single ventral impression of an incomplete left wing
without counterpart, of which pieces have flaked away and are not preserved. Original length
estimated at c. 15 mm. The anterior and posterior margins are sufficiently well preserved to
indicate a narrow elongate wing. The longitudinal veins are distinct, but some of the crossveins
cannot be discerned although their positions are indicated by the undulations in R and M. R,,
R,; and CuA are on prominent upfolds. Sc 1s close to anterior margin in the basal part of the wing,
continuing beyond the fork of Rs. R, curved posteriorly at its distal end bounding the pterostigma
on its posterior side. No definite traces of pterostigmal crossveins. Rs forking just basad of M,
the former four-branched and the latter with only three branches preserved. R,;—-M, interlinked
by well-spaced crossveins; Ist 1.m. oblique and well distad of the termination of CuP on the
posterior margin. Wing colouration: veins brown, pterostigma dark brown; a slightly lighter
brown area between R, and Rs immediately beneath and distad of pterostigma.

MATERIAL. In.64537 (H). Fig. 46.
DIMENSIONS. Maximum length 6.5 mm, width | mm.

REMARKS. The elongate form and well-developed pterostigma are typically bittacid (cf. Setty
1940 : 282). The venation is close to that of Bittacus veternus (Cockerell) comb. nov. but it differs
from the latter in that the pigmentation is confined to the pterostigma and adjacent membrane.
The wing details agree with Recent Bittacus (Fig. 45), but in the absence of further information
on this species it can only be referred to this genus in the wider sense.

Order LEPIDOPTERA
Suborder ZEUGLOPTERA
Family MICROPTERIGIDAE

Genus MICROPTERIX Hiibner, 1825

TYPE SPECIES. Tinea podevinella Hiibner 1813, by subsequent designation; Recent, western
Palaearctic.

Micropterix anglica sp. nov.
Fig. 49

DiaGnosis. A Palaeogene species of Micropterix with Sc forking distally beyond the origin of
R, and R, unbranched in the forewing.

DESCRIPTION. The single known specimen consists of the greater part of a forewing, original
length estimated at about 4 mm. Sc forked, R, simple. R, and R, originating from the outer half
of the areole, chorda faintly preserved. R, and R; single, and like M,, converging on the apical
end of the areole; forks of M, ,, and CuA, ,, approximately beneath the basal end of the areole.
CuP single. 1A represented by a short distal portion. Crossvein r—m faint but no trace of sc—r
nor m-cu. Colouration: veins mostly dark brown with intervening membrane light brown.

Hovortype. In.17411 (S). Fig. 49.

DIMENSIONS. Forewing: preserved length 2-4 mm, width 1-1 mm; length of Sc, 0-93 mm.

264 E. A. JARZEMBOWSKI

51

Fig. 47 Prohepialus sp. Composite drawing of wing In.64528 (i-i11).

Fig. 48 Hepialus fusconebulosa (De Geer). Recent, Britain. Outer part of forewing.
Fig. 49 Micropterix anglica sp. nov. Holotype, In.17411.

Fig. 50 Microlepidoptera family uncertain sp. A, 1.9492. (b= bristle).

Fig. 51 Copromorpha gypsota Meyrick. Recent, Fiji. Wing venation.

Fig. 52. Copromorpha fossilis sp. noy. Holotype, 1n.25766.

(Scale line=2 mm. Figs 51—S2 are to same scale.)

REMARKS. The homoneurous type of venation and vein details agree with the extant genus
Micropterix. The absence of crossveins sc-r and m-cu is probably due to preservational failure,
as they may be weakly developed in Recent micropterigids. The fossil differs from *‘Micropterix’
proavitella Rebel (Baltic amber) in that Sc forks more apically and R, is unbranched (cf. Rebel
1935 : fig. 17). It approaches the Recent British species Micropterix [= Eriocephala] calthella
(Linné) but differs principally in that Sc, is shorter and appears to originate more apically of the
origin of Ry compared with the latter (cf. Meyrick 1928 : figure on p. 873).

PALAEOGENE INSECTS 265

Recent micropterigids have a widespread distribution but Micropterix is confined to the
Palaearctic Region. This primitive family includes the earliest known Lepidoptera, from the
Lower Cretaceous of Lebanon (Whalley 1977). They are small, metallic-coloured, largely diurnal
moths with functional mandibles and no proboscis, feeding on pollen grains; the larvae are
considered to feed on bryophytes but some may be detritus feeders (Common 1970). Adult
Micropterix visit blossoms including Carex L. and Ranunculus L., both genera being represented
in the Insect Bed.

Suborder EXOPORIA
Superfamily HEPIALOIDEA
Family HEPIALIDAE

Genus PROHEPIALUS Piton, 1940

TYPE SPECIES. Prohepialus incertus Piton 1940, by monotypy; Palaeocene, Menat (Puy-de-D6me).

Prohepialus sp.
Figs 38, 47, 59

1976 hepialid, Jarzembowski : 13.
1977 hepialid, Robinson : 108; figs 5-6.

DESCRIPTION. The material consists of four incomplete wings, estimated wing length 20-45 mm.
Sc marked by prominent fold, close to R, distally and like R, continuing well beyond the end of
the discal cell. R, ,3 closer to Ry than R, at base. Base of Ry well separated from origin of r—m.
Discal cell widening towards apex, M bifurcate in cell. The distance from posterior end i.m. to the
anterior end r—m is 0-6 of width of the discal cell across apical angles. R, ,, forking beyond the
cell at c. 1-3 of maximum width of the latter; M,;+r—m and CuA,+m-cu form outwardly-
directed angles from the apical end of the cell. Crossvein i.m. oblique, originating from the fork of
M, 42 or a little basad of the fork. No colour pattern is preserved, but the wing surface has a
brownish tinge in some of the specimens because of the presence of scales. In.64528(i) shows
convex in the part indicating that this is the dorsal impression of the wing with scales from the
same surface; the counterparts therefore show the ventral scales. Adpressed marginal scales can
be seen in In.64528(iii) and In.64538+ counterpart, the latter also with long scales on both

surfaces.

In.64538 (Fig. 59) and counterpart on reverse of In.64528(11); In.64539 (H).
Dimensions. Width of discal cell across apical angles 9-8-c. 4-4 mm (In.64538-9).

REMARKS. The general habitus of the wings is indicative of the Hepialoidea, and M forked in the
discal cell, m—cu (M, auctt.) oblique, and comparatively large size place them in the Hepialidae,
commonly known as swift moths (Common 1970 : 787, 789). Relationships with Recent genera
are uncertain, but the Hepialus-like venation (cf. Fig. 48) agrees with the Palaeogene genus
Prohepialus Piton. The latter genus is based on a single specimen of which the venational details
are incompletely known, rendering specific comparison impossible.

The hindwings of Hepialus spp. are slightly smaller than the forewings and this would increase
the total variation in wing size. In.64538 and In.64528 could be the fore- and hindwings of the
same individual as they come from the same piece of rock, both show traces of pigmented scales,
and the difference in size of the discal cell (apical width differs by c. 33°) is within the range of
some Recent hepialid specimens. Examination of both surfaces of the forewing of the Recent
British Hepialus fusconebulosa (De Geer) (Fig. 48) showed a similar shape difference in the
scales as in In.64528, and the presence of long scales in the hindwing as in In.64538 + counterpart.
The presence of broad longitudinal ridges as well as more numerous, finer ridges is a feature of
hepialid scales that was noted by early workers (e.g. Kellogg 1894 : 80-82); both types of

266 E. A. JARZEMBOWSKI

longitudinal ridge and other ultrastructural details are preserved in the scale shown in Fig. 59.
The family has a cosmopolitan distribution but is most diverse in the southern hemisphere.

The larvae are phytophagous, tunnelling in wood or soil; the adults are short-lived with mouth-

parts more or less reduced and functionless (Bourgogne 1951 : 372; Common 1970 : 787).

55

4

Fig. 53 Paratriaxomasia solentensis gen. et sp. nov. Holotype, 1.9166.

Fig. 54 Triaxomasia caprimulgella (Stainton). Recent, Europe. Wing venation, after Zagulyaev
(1964).

Fig. 55 Tineidae species A. 1.9614.

Fig. 56 Gelechioidea species A. 1.9042.

Figs 57-58 Microlepidoptera family uncertain. Fig. 57, species C. In.25512+1n.25252. Fig. 58,
species K. In.25219.

(Scale line=2 mm)

PALAEOGENE INSECTS 267

Suborder DITRYSIA
Superfamily TINEOIDEA
Family TINEIDAE

Genus PARATRIAXOMASIA nov.

DiaGnosis. A Palaeogene genus of Tineidae approaching the extant genus Triaxomasia Zagulyaev
but differing in that the forewing is slightly broader and the hindwing lacks a marginal indentation
near the termination of Sc+R, and is slightly narrower towards base.

TYPE SPECIES. Paratriaxomasia solentensis gen. et sp. nov.
Paratriaxomasia solentensis gen. et sp. nov.
ley, 33)
DIAGNOsIS. Species of Paratriaxomasia with a forewing span of 8-0 mm.

DESCRIPTION. The species is known from a single ventral impression of the body and wings.
Head rounded anteriorly. Forewing approximately three times as long as broad, attenuated
apically. Sc close to anterior margin-terminating opposite the origin of R,. R, originating a short
distance from wing base at about 0:25 of the length of the discal cell. Areole elongate, 0-4 of the

| 5 oom A - _
TBSP i \ 3 i aie 4

Fig. 59 Prohepialus sp., 1n.64538 (scale, x 10 000; Ir longitudinal ridge, cr cross rib, w window).

Fig. 60 Gurnetia durranti Cockerell, holotype. I1n.24324. See also Fig. 66. (Scale line=2 mm)

268 E. A. JARZEMBOWSKI

Fig. 61 Fragment of small bird feather, occurring on slab In.17194. See p. 239. (Scale line = 2 mm)

ae

Fig. 62 Insect larvae inside cavity of a plant stem. In.64544. See p. 241. (Scale line =2 mm)

cell length; R3,-R;, M,-M; and CuA,—-CuA, originating from the apical part of the cell, the
origin of R, basad of that of CuA,. R; appears to terminate immediately anterior to the wing
apex. Hindwing lanceolate, 4:5 times longer than broad and gradually tapering towards the apex.
Sc+R, well separated from the anterior margin, terminating at midlength of wing opposite the
apical end of the discal cell. M, towards apex, converging on M, basally and continuing as a
fold into the discal cell. Ms, originating close to M, and CuA,. CuP prominent, anal vein present.
Stem of CuA well separated from the posterior margin.

HovoryPe. I.9166 (B). Fig. 53.
Dimensions. Length of body 3-6 mm, forewing 3-6 mm and hindwing 3-3 mm.

REMARKS. The wings resemble those of the extant monotypic genus Triaxomasia (Fig. 54),
from which they differ in that the forewing is slightly broader, there is no faint trace of M in the
discal cell and the chorda and 2A appear to be stronger veins; the hindwing is narrower basally,
and there is no shallow indentation on the anterior margin near the distal end of Sc+ R, as in

PALAEOGENE INSECTS 269

Triaxomera Zagulyaev. The forewing span of the European Triaxomasia caprimulgella (Stainton)
(Fig. 54) is 10:5—11:5 mm, and like other Nemapogoninae, it is a forest dweller and the larvae
are fungus feeders. The caterpillars of this species inhabit fungus-infested rotten wood including
Fagus L., Ulmus L. and Quercus L. (Zagulyaev 1964 : 86, 133, 155-160).

Genus indet.

Species A
ene, BS)

DESCRIPTION. The species is known from a single specimen with thorax, basal part of abdomen
and a single forewing preserved. Meso- and metathorax distinct. Forewing elongate, about three
times as long as broad, apex bluntly angular; membrane with distinct transverse lineation.
Discal cell elongate, 0-6 of wing length. Sc close to anterior margin, terminating near middle of
wing. R, originating at a quarter of wing length from base, terminating a short distance beyond
the apical end of the discal cell. Branches of R, M and CuA originating relatively close to each
other, R; towards apex; origin of R, a little more basad in position than that of CuA, and
marking an anterior deflection in the stem of R. Discocellular crossveins indistinct. Chorda
apparently absent, but preservational failure in the discocellulars raises the possibility that a
weak chorda was present in life. 1A+2A with a short basal loop and well separated from the
posterior margin.

MATERIAL. 1.9614 (B). Fig. 55.

DIMENSIONS. Length of thorax 1-8 mm. Length of forewing 5:8 mm, maximum width of forewing
2:0 mm.

REMARKS. The forewing bears some resemblance to that of the incurvariine Adela reamurella
(Linné) but differs notably in that the apex is more rounded and there is a wider area between
the apical end of the discal cell and the outer margin (cf. Meyrick 1928 : figure on p. 843). It is
similar in size and form to Tinea pellionella (Linné) (cf. Forbes 1923 : fig. 73) although in the
latter the forewing is more elongate (about four times as long as broad) and a chorda is present.
The venation is also close to that of Nemapogon granella (L.) in which the chorda 1s faint (Forbes
1923 : fig. 74, Tinea granella (L.)) and the fossil apparently belongs in the Tineidae, a widespread
present-day family.

Superfamily GELECHIOIDEA

Species A
Fig. 56

DESCRIPTION. The single specimen consists of a ventral impression of the body, incomplete
hindwings and part of the right forewing. The last shows portions of Sc and R, the latter with two
well-spaced branches. Hindwing: moderately wide, Sc+ R, and Rs sinuous and preserved in the
left wing only; CuA preserved in both wings, CuP prominent, anal area broad, angular.

MATERIAL. [.9042 (B). Fig. 56.
DIMENSIONS. Length of body 4:5 mm.

REMARKS. Although the fossil is rather fragmentary, the hindwing form and venation resemble
some extant gelechiids including Agnippe sp. (Brues, Melander & Carpenter 1954: fig. 421),
Parachronistis albiceps (Zeller) (Spuler 1913 : fig. 123) and Polyhymno_ luteostrigella Chambers
(Forbes 1923 : 256). It is tentatively referred to the Gelechioidea. The superfamily has a cosmo-
politan distribution; the larvae are usually phytophagous.

270 E. A. JARZEMBOWSKI

Superfamily COPROMORPHOIDEA
Family COPROMORPHIDAE

Genus COPROMORPHA Meyrick, 1886

TYPE SPECIES. Copromorpha gypsota Meyrick 1886, by monotypy; Recent, Australian Region
(Pacific).

Copromorpha fossilis sp. nov.
Fig. 52

DiaGnosis. A Palaeogene species differing from the living C. gypsota in the smaller size of the
wings, the slightly more basad termination of the branches of R in the forewing, and M, closer
to Rs in the hindwing.

DESCRIPTION. The single specimen consists of the dorsal impression of a decapitated body and
the remains of the right wing pair.

Body: posterior end of mesothorax rounded, metascutum distinct. Abdomen completely
flattened except for two lateral cavities in the basal segment; seven annuli are discernible and the
remains of the apical scale cover. The abdomen Is comparatively large even allowing for compres-
sion and, together with the pointed posterior, suggests that the holotype is a female.

Forewing: anterior margin rounded basally, straightening distad of the termination of Sc and
curving strongly past R,. Sc terminating near the mid-length of the wing; R, originating at
0-25 and terminating at 0-65 of the forewing length (from base); R; and R, terminating at 0°85
and 0-94 of the same length. Base of R, well separated from that of R,. CuA, 1A and 2A repre-
sented by short basal portions close to the CuP fold; anal veins well separated from the margin.

Hindwing: apex rounded with an angle of 85°. Anterior margin straight, outer margin curving
gently towards base away from the apex. Sc+R, elongate continuing apically well beyond the
end of the discal cell. Rs directed towards apex. Bases of M,—M, linked by oblique crossveins.
M, equidistant between M, and Mg, M, continuing as a distinct fold in the discal cell. Base of M3
touching CuA, at 0-4 of its length from the separation of CuA,.

HovortyPe. In.25766 (H). Fig. 52.
DIMENSIONS. Length of forewing 8-8 mm.

REMARKS. The wing form and venation come closest to that of extant moths of the family
Copromorphidae and the type genus itself (Fig. 51). The genitalia of Copromorphoids show
affinities with those of the Gelechioidea and the venation of the fossil shows some resemblance
to the Oecophoridae. Although the bases of R, and R; are not preserved, their courses indicate
that they do not form a distinct fork as is characteristic of the Oecophoridae. The apical angle is
close to 90°, in agreement with the Copromorphidae.

The family has a southern distribution, only Aegidomorpha Meyrick entering the Palaearctic
in China; Copromorpha occurs in Africa and from India through to Australia. Copromorphid
larvae are tunnellers in leaf veins, twigs and fruits (Common 1970 : 827).

Microlepidoptera, Family uncertain

Species A
Fig. 50

DESCRIPTION. The species is known from the dorsal impression of a single specimen with the
left pair of wings and anterior part of the body preserved. Estimated wing span 6-0 mm. Body
pale brown; base of abdomen, meso- and metathorax distinct but dorsoventrally compressed.
Head rounded anteriorly. Wings elongate, apices pointed, with some scale traces.

Forewing: broader and slightly longer than hindwing. Anal area broad, CuP marked by a
prominent fold terminating on the posterior margin at 0-6 of the wing length from base; Sc close

PALAEOGENE INSECTS 271

to anterior margin and terminating at 0-4 of the same length. Distal part of stem R + M preserved
close to CuA; two branches of R well defined as in Fig. 50 and apical venation suggested by some
faint traces. CuA and A single and close to CuP fold.

Hindwing: Sc + R, thickened basally, long and close to anterior margin: a detached bristle may
represent the frenulum. Rs+M forking near mid-length of wing; Cu single.

MATERIAL. 1.9492 (B). Fig. SO.

DIMENSIONS. Forewing length 2:7 mm; hindwing length 2:3 mm.

REMARKS. The venation approaches that of the modern leaf-miners Heliozela Herrich Schaffer
(Common 1970 : fig. 36.15B) and Nepticula Heyden (N. terminella Braun 9, Forbes 1923 : fig. 56),
and Species A probably belongs in either the Incurvariina or Nannolepidoptera.

Species B

DESCRIPTION. The species is known from a single specimen consisting of the ventral impression of
the body plus forewings. Estimated wing expanse 6:0 mm. Body elongate, head well rounded
anteriorly. Forewing lanceolate, with some traces of scale cover. Sc close to but well separated
from anterior margin, probably not continuing beyond mid-length of wing. Anal area broad,
CuP fold prominent. Stem of R+M close to CuA; R producing three branches to the anterior
margin in outer half of wing, some faint suggestions of venation posterior to last branch. Vein A
single, and like the stem of CuA, close to CuP fold.

MATERIAL. In.64540 (H).
DIMENSIONS. Body length 2:8 mm, forewing length 2-7 mm.

REMARKS. The forewing venation comes close to that of Heliozela but also resembles the gele-
chioid Dyselachista sericiella (Haworth) [=saltatricella (Fischer von Réslerstamm)] (cf. Spuler
1913) and the lyonetiid Lewcoptera laburnella (Stainton) (cf. Meyrick 1928); the exact affinities
are uncertain. The latter two species are a little larger than Species B.

Species C
Fig. 57

DESCRIPTION. A very small moth with an estimated wing expanse of about 3-8 mm. Head rounded
anteriorly. Mesothorax oval, scutellum short and pointed posteriorly; metascutum laterally
lobate (In.17142).

Forewing: lanceolate, CuP fold prominent extending for 0-6 of wing length, CuA and A close
to fold. Sc close to anterior margin, extending approximately to mid-length of wing. R represented
by a distal fork. Anal area broad.

Hindwing: \inear-lanceolate, with some well-preserved scales (In.25512+ 25252). Sc+R, well
separated from the anterior margin and elongate like CuA; Rs+ M apparently with three branches
in outer part of wing.

MATERIAL. In.25512 (H) and counterpart In.25252 (H), both Fig. 57; In.17142 (S).

Dimensions. Length of body 1:9 mm, forewing 1:7 mm.

REMARKS. Although rather small, the venation shows some resemblance to the extant lyonetiids

Leucoptera Hibner and Bedellia Stainton; the species probably belongs in the Tineoidea.
Species D

DESCRIPTION. The single specimen consists of the abdomen and incomplete fore- and hindwings.
The abdomen tapers to a point posteriorly and the specimen is probably 9. The forewing is

7/2 E. A. JARZEMBOWSKI

represented by a median fragment and suggests an elongate wing with a discal cell; two oblique
branches of R are preserved. A hindwing fragment shows the basal portions of M, and of both
branches of CuA; Ms originates very close to CuA, and is linked by a short crossvein.

MATERIAL. In.17392 (S).
DIMENSIONS. Length of abdomen 2:7 mm.

REMARKS. The general appearance of this fragmentary fossil is suggestive of the Ditrysia.

Species E

DESCRIPTION. A single dorsal impression of a body and basal half of a forewing. Head rounded
anteriorly, thorax typically lepidopterous. Mesoscutellum 0-3 of the length of the mesothorax
and rounded posteriorly with traces of the postnotum, suture with mesoscutum gently curved
anteriorly; metascutum short with a narrow central portion between the meso- and meta-
scutellum. Abdomen with some traces of segmentation.

Forewing: area between the stem of R and anterior margin moderately wide, Sc well separated
from the margin. Base of R, closer to R, than Sc but well separated from the former. Discal cell
moderately narrow. CuA, CuP fold and 1A+ 2A close, the last apparently with a basal loop.

MATERIAL. In.25251 (H).
DIMENSIONS. Length of body 3:5 mm, abdomen 2:1 mm.

REMARKS. The affinities of this moth are uncertain but the general appearance suggests that it
belongs to one of the ‘tineoid’ superfamilies.

Species F

DESCRIPTION. A single dorsal impression of the anterior part of the body and the remains of the
left wing pair. Head rounded anteriorly. Scutoscutellar suture of mesothorax curved anteriorly,
scutellum small and truncate posteriorly. Metathorax elongate with a straight transverse suture
with the mesothorax. Abdomen represented by two basal segments.

The wings consist of an incomplete hindwing overlying part of the forewing in which 1|A+2A
and the basal loop can be discerned. The venation of the hindwing is more complete and suggests
an elongate wing: Sc well separated from the anterior margin, CuA, and CuA, slightly divergent.
M apparently arising on two branches, M, equidistant between M, ,, and CuA,. M, ,, relatively
close to Rs, M, continuing into the discal cell. Discocellular crossveins present. Fine transverse
lines and some scales are visible on the wing membrane.

MATERIAL. 1.9783 (B).
Dimensions. Length of thorax 1-5 mm.

REMARKS. The affinities of this moth are uncertain: the hindwing venation shows some resemb-
lance to Adela ridingsella Clemens (Incurvariidae) and Haplotinea insectella (Fabricius) (Tineidae).

Species G
Fig. 63

DESCRIPTION. A single dorsal mould of the anterior part of the body plus forewings. Antennae
typically lepidopterous with pedicel larger than the flagellar segments and an expanded scape.
Mesoscutellum pointed posteriorly, and markedly convex (in dorsal view) like the mesoscutum.

Forewings moderately broad and directed posteriorly as in rest position. Estimated original
wing length 10-0 mm. Discal cell elongate with traces of M. R, originating at 0-35 from the base
of the cell, R.—R, from the posterior part of the areole, R, closer to Ry. M,j-M; and CuA,-CuA,

67 A ete 3 68

Figs 63-65 Microlepidoptera family uncertain. Fig. 63, species G. 1.8917. Fig. 64, species H. 1.8809.
Fig. 65, species L. In.24506 with details of head and right forewing from counterpart In.64543.

Fig. 66 Gurnetia durranti Cockerell, holotype. In.24324. See also Fig. 60.

Fig. 67 Acrolophus cf. cossoides Felder & Rogenhofer. Recent, Brazil. Outer part of forewing.

Fig. 68 Culama sp. Recent, Australia (Meyrick coll. BM(NH) 1938-290). Outer part of forewing.

(Scale line=2 mm. Figs 66-68 same scale.)

274 E. A. JARZEMBOWSKI

originating close together from the apical end of the discal cell, the origin of CuA, less basad
than that of Ry. Fine transverse lines present on wing membrane.

MATERIAL. 1.8917 (B). Fig. 63.
DIMENSIONS. Length of discal cell in forewing 5-6 mm.

REMARKS. The form of the thorax and primitive venation are very close to that of the tineid genus
Acrolophus Poéy (cf. Forbes 1923 : fig. 19) but also to the incurvariine Adela Latreille (cf. Spuler
1913 : fig. 223).

Species H
Fig. 64

DESCRIPTION. A single ventral impression of the body with wing remains. Head rounded anteriorly.
Estimated forewing span 5-6 mm.

Forewing: discal cell elongate and moderately narrow. Sc close to anterior margin and appar-
ently terminating in basal half of wing. R, short, originating at 0:3 of the length of the discal cell
from base. R, and R; originating from the apical end of the cell, bases well separated. Only a
single branch of CuA is preserved, the base opposite that of Rs. CuP fold slightly closer to CuA
than A.

Hindwing: elongate, wing tip attenuate. Sc+R, close to anterior margin. Rs + M forking near
mid-point of wing, M forking a short distance apically; Rs to the outer part of the anterior
margin, M,—M, to the outer part of the posterior margin, M, and M, with common stem. CuA
forking opposite Rs+M, branches short.

MATERIAL. I.8809 (B). Fig. 64.
DIMENSIONS. Length of discal cell in forewing c. 2:0 mm.

REMARKS. The fossil’s venation shows some resemblance to that of the gracillariid Ca/optilia
alchimiella (Scopoli) (cf. Meyrick 1928 : figure on p. 788) and some small gelechioids, and it
apparently belongs amongst the ‘tineoid’ superfamilies.

Species I

DESCRIPTION. A single specimen, the body being represented by the abdomen, parts of the thorax
and a single leg, all of which are crushed. Estimated forewing span 11-12 mm. Hindwing repre-
sented by a median fragment indicating a broad wing, with part of CuA and Ms preserved.
Forewing elongate, over 3-5 times as long as broad, apex unknown. Discal cell relatively narrow,
oblique. R, originating at 0-45 of the length of the discal cell from base, a short distance basad
of the termination of Sc. R,-R;, M,-M, and CuA,—-CuA, apparently all present and originating
from the apical part of the cell, base of R, nearly opposite that of CuA,. 1A+2A with basal loop.
Forewing covered with numerous well-preserved scales.

MATERIAL. In.64541 (E. A. Jarzembowski & Tertiary Research Group, London). Sticelett Ledge.
Dimensions. Length of discal cell in forewing 3:2 mm, of abdomen 2:8 mm.

REMARKS. The venation resembles extant species of Gelechia Hiibner, Plutella Schrank (cf.
Forbes 1923) and Stegasta Meyrick; the fossil apparently belongs in one of the ‘tineoid’ super-
families.

Species J

DESCRIPTION. A single fragmentary specimen consisting of the thorax and incomplete fore- and
hindwings. Forewing elongate, stem of A and CuA close to prominent CuP fold. R with three—
four branches preserved in outer half of wing. Sc apparently terminating on anterior margin a
little distad of the origin of R,. Hindwing venation too poorly preserved for comment.

PALAEOGENE INSECTS DiS)
MATERIAL. [n.25157 (H).

DIMENSIONS. Length of forewing 3-4 mm.

Species K
Fig. 58

DESCRIPTION. A single ventral impression of the head, thorax and basal part of abdomen with
remains of fore- and hindwings. Body elongate, wing membrane with a faint transverse lineation.

Forewing: discal cell elongate with traces of M. Venation complete, R.—-CuA, originating from
the apical end of the discal cell. Base of R, well separated from that of R,, Sc terminating on
anterior margin a short distance beyond the base of R,.

Hindwing: an incomplete left wing is preserved, anteriorly overlapping the posterior part of
the corresponding forewing. Sc+ R, and Rs represented by two short distal fragments. M forking
in the discal cell. M, linked to CuA, by a short crossvein.

MATERIAL. In.25219 (H). Fig. 58.
Dimensions. Length of discal cell in forewing 3-1 mm.

REMARKS. The general habitus is indicative of the ‘tineoid’ superfamilies and the fossil resembles
the extant yponomeutid Plutella annulatella (Curtis) (Forbes 1923 : fig. 200), but its exact
systematic position is unknown.

Species L
Fig. 65

DESCRIPTION. A single specimen with body, both forewings and fragments of left hindwing.
Mesothorax oval with apparent traces of the reduced prothorax anteriorly (In.24506); meta-
scutum short and laterally expanded. Wing membrane with transverse lineation.

Forewing: broad, discal cell elongate with traces of M. R with four branches. R, originating
at 0-5 of the length of the discal cell from base. R, closer to Rj than to R, but well separated from
the former. Sc well separated from the anterior margin terminating on the latter just beyond the
base of Ry. Ry ..;, M,_3 and CuA, _, originating closely from the apical end of the discal cell,
CuA, opposite R;. Fork of 1A+2A well separated from posterior margin of wing.

Hindwing: two fragments with vein traces probably of Rs, M, CuA and A.

MATERIAL. In.24506 (H) and counterpart In.64543. Fig. 65.
DIMENSIONS. Length of body 5-8 mm, length of discal cell in forewing 3-2 mm.

REMARKS. The venation resembles that of the extant psychid Lypusa maurella (Denis & Schiffer-
miiller) although there is some trace of a chorda in the forewing of the latter (cf. Spuler 1913 :
fig. 206). The fossil probably belongs in the Tineoidea.

? Superfamily COSSOIDEA
? Family COSSIDAE
Genus GURNETIA Cockerell, 1921
TYPE SPECIES. Gurnetia durranti Cockerell 1921, by monotypy.

Gurnetia durranti Cockerell, 1921
Figs. 60, 66

19216 Gurnetia durranti Cockerell : 473; fig. 38.
HovotyPe. In.24324 (H). Figs 60, 66.

DIMENSIONS. Forewing (incomplete) : maximum length 9-4 mm, width 6-5 mm.

276 E. A. JARZEMBOWSKI

REMARKS. The genus and species is known from a single incomplete forewing with associated
fragments probably of the hindwing. Estimated original length of forewing 15 mm. The transverse
lineation on the membrane (Fig. 60) readily excludes it from the trichopteran family Limnephilidae
(cf. Cockerell 19216). Cockerell and Durrant (Cockerell 19216) placed Gurnetia in the Cossidae
and comparison with Recent material shows that the venation is close to that of the Australian
genus Culama Walker (Fig. 68). The primitive type of venation shown by Gurnetia is also develop-
ed in certain microlepidoptera. Cockerell commented on the resemblance to the incurvariine
Nemophora Hoffmannsegg, but the forewing of this is only about 6-8 mm long. CuA, originating
close to CuA, near the apical end of the discal cell readily excludes the fossil from the Tortricidae.
However, the fossil’s venation does approach certain genera in the ‘tineoid’ superfamilies, a
possibility not discussed by Cockerell and Durrant (ibid.). Thus it resembles the yponomeutid
Orthotaelia sparganella (Thunberg) which has a maximum forewing length of 13 mm, although
the posterior part of the wing is narrower in the latter. It also resembles the extant New World
genus Acrolophus (Fig. 67), in which wing spans of 30-40 mm are often attained. The placement of
Gurnetia in the Cossoidea is therefore uncertain although it undoubtedly belongs to the lower
Ditrysia.

Superfamily PYRALOIDEA
Family PYRALIDAE
Subfamily PYRALINAE

Genus PYRALITES Heer, 1856

TYPE SPECIES. Pyralites obscurus Heer 1856, by monotypy; Oligocene, Aix-en-Provence.

Pyralites preecei sp. nov.
Fig. 69

DESCRIPTION. Single specimen consisting of decapitated body in dorsal aspect and remains of
both fore- and hindwings. Mesothorax oval, scutellum rhomboidal, 0-4 of mesothoracic length.
Left tegula distinct. Metascutum typically lepidopterous with two lateral lobes. Abdomen
compressed but with some pink calcareous infill showing a number of fine tubular structures.
The segmentation of the abdomen is obscure except for the more sclerotized basal segment.

The left wings are partly overlapping, but details of the hindwing venation can be seen beneath
the forewing where the posterior part of the latter has flaked away. No trace of colour except
some pale brownish pigmentation of the veins; scaly areas may be locally discernible on the
membrane.

Forewing: discal cell long and narrow with traces of a median fold. Sc close to anterior margin
continuing beyond the end of the discal cell. R producing three branches in outer half of cell,
the second branch originating a little nearer the third than R,. The branches of R and Sc run
closely and very obliquely towards the anterior margin but their terminations are not preserved.
M,_3 curving towards outer margin, M, forming an angular fork with the third branch of R,
and M, originating near to Ms. CuA, and CuA, diverging from the cell less obliquely than Sc
and R branches, CuA, separating about midway between R, and the second branch of R. The
venation suggests a relatively narrow forewing with original length estimated at 15 mm.

Hindwing: anterior margin obscure, discal cell much shorter and broader than in forewing.
Sc+R, prominent, close to Rs for a short distance in distal part of its course. Rs indistinct

Fig. 69 Pyralites preecei sp. nov. Holotype, 1.8640. Body and left wing pair.

Fig. 70 Chloroclysta truncata (Hufnagel). Recent, Britain. Outer part of forewing.

Fig. 71 Geometridites larentiiformis sp. nov. Holotype, 1.8866 and 1.8935 (composite drawing).

Fig. 72} Nymphalites zeuneri sp. nov. Holotype. Fig. 72a, part 1.10384; Fig. 72b, counterpart
1.10384a. See also Figs 75-76.

Fig. 73 ? Papilionoid, 1n.64545a.

(Scale line=2 mm. Figs 69-71 same scale; Figs 72-73 same scale.)

PALAEOGENE INSECTS ET

278 E. A. JARZEMBOWSKI

inwards of the fork with M,. Bases of Mz, M3 and CuA, closely spaced, apparently linked by
crossveins. Only the ends of mde are preserved, indicating a strong asymmetric basal arch.

The right wings are preserved at an angle of c. 15° to the plane of the body, inclined downward
and slightly overlapping the abdomen on the same side. They show little useful detail, although a
transverse lineation is well preserved.

Ho.oryPe. 1.8640 (B). Fig. 69.
Dimensions. Length of mesothorax 3-0 mm, of tegula 1:5 mm.

REMARKS. The general habitus, M absent from the discal cell, M, close to M, basally in both wings
and hindwing with Sc+R, close to Rs beyond the discal cell are indicative of the Pyraloidea :
Pyralidae (Common 1970 : 833, 836). The venation approaches the Recent genera Paractenia
Ragonot, Bostra Walker, Diloxia Hampson and Tyndis Ragonot (cf. Hampson 1896 : figs 101,
104, 105, 111) in the Pyralinae (and Pyralini of Whalley 1961). The fossil agrees with Pyralites
obscurus Heer from Aix, southern France (mid-Oligocene, Denizot 1956 : 5) in having a rounded
mesothorax, but differs from the latter in the larger size of the mesothorax and in having a
distinct mesothoracic scutoscutellar suture (cf. Heer 1856 : 30; pl. 2, fig. 6). However, a re-
examination of P. obscurus may require a separate genus for P. preecei sp. nov.

The adults of Recent pyralids are usually nocturnal insects, pyraline larvae generally feeding
on dry or decomposing vegetable material (Bourgogne 1951 : 399-400). The subfamily has a
widespread distribution.

Superfamily GEOMETROIDEA
Family GEOMETRIDAE

Genus GEOMETRIDITES Clark et al., 1971

TYPE SPECIES. Geometridites repens Kernbach 1967, by original designation of Clark et al. 1971;
Pliocene, Willershausen.

Geometridites larentiiformis sp. nov.
Fig. 71

DiaGnosis. A fossil species of Geometridae with the forewing venation resembling that of the
extant larentiine Chloroclysta truncata (Hufnagel), but differing in that R,_; are not curved
apically.

DESCRIPTION. Fragment of a forewing, estimated original length 15 mm. Sc represented by a
short median portion. Two elongate areoles present, the inner smaller than the outer; R,_; from
apical extremity of latter, the separation of R, a little more basad. R,_, with common stem, base
of R, equidistant between that of R, and areole. M, gently curved and continuous with stem of R.
CuA, and !A+2A represented by two short sections, strongly divergent and indicating a broad
wing.

HovoryPe. I.8866 (B) and counterpart 1.8935 (B). Fig. 71.

Dimensions. Length of inner areole 1-5 mm, of outer areole 2:1 mm; fork of Rs ,, 2-4 mm from
latter.

REMARKS. Sc separate from R, two areoles anterior to the apical end of the discal cell, R; with
areolar origin, and general form of the venation place the wing in the Geometridae in the Ster-
rhinae or Larentiinae (Common 1970 : 846, 848). The venation is especially close to some Recent
Larentiinae (cf. Fig. 70) but the fossil is too incomplete for certain subfamily placing; its clear
geometrid affinities render it referable to the broad fossil genus Geometridites Clark et al.

Both Recent subfamilies have a cosmopolitan distribution, the Larentiinae with a temperate
preference (Bourgogne 1951). Geometer larvae are found on foliage, usually pupating in the
ground or amongst débris (Common 1970 : 846).

PALAEOGENE INSECTS 279

Superfamily PAPILIONOIDEA
Family NYMPHALIDAE

Genus NYMPHALITES Scudder, 1889

TyPE SPECIES. Nymphalites obscurum Scudder 1889, by original designation; Oligocene, Florissant.

Nymphalites zeuneri sp. nov.
Figs 72, 75-6

1878 Lithosia (sp.); Smith in Woodward : 88.
1879 Lithosia (sp.); Smith in Woodward : 344.
1894a Butterfly; Scudder in Brodie : 168.
18945 Butterfly; Scudder in Brodie : 70.

1907 (? Lithosia); Handlirsch : 923.

1961 cf. Euthalia; Zeuner : 310.

DiAGnosis. Medium-sized Palaeogene species of Nymphalidae close to the Recent species
Neurosigma siva Westwood, Abrota mirus Fabricius and Cymothoe theobene Doubleday &
Hewitson in wing form and venation, but the forewing differs in that M, is less close to M,
basally and the hindwing is differentiated by the short humeral vein towards the humeral angle
and relatively narrow anterior area of the wing.

DESCRIPTION. The body is represented by some black chitinous fragments of the thorax near the
cleaved edge of the rock, the remainder apparently having broken away during collection.

The remains of both pairs of wings are preserved on the opposite sides of I.10384. The two pairs
diverge from the body at an angle of about 15° resembling the upright rest position of the wings
in butterflies. However, the hindwings are here overlapped on the outside by the forewings,
indicating that they are folded beneath the body as may happen in dead Recent specimens. The
two wings on either side are separated from each other at many points by a fine parting of
limestone, although this has flaked away in places on I.10384a exposing the forewing. The follow-
ing account is based on the left pair (Figs 72, 76) as only the bases of the right pair are exposed
and these are similar to the left.

Forewing: moderately wide, outer part missing. Discal cell closed, elongate and widest near
origin of R, with maximum length to width ratio 3-4 : 1. Cells C and Sc moderately narrow.
R, arising 0-7 of the length of the discal cell from base. R, closer to R3_; than R, but well
separated from the former. R,_; and M, arising from the anterior apical angle of the discal cell;
stem of R;_; 0-4 of the cell length and only the initial portion of the first fork preserved. Cu trifid,
CuA, and CuA, with relatively divergent courses and separating at 0-8 and 0-5 respectively of the
length of the discal cell from base. M, closer to M, than M, at 0-4 of the span between the bases
of M, and M3. Branches of M gently curved towards outer margin. Discocellular crossvein fine
but clear : ude absent, mdc curved basally and Idec nearly straight. |A+2A single.

Hindwing: moderately elongate, margins missing except for the basal part of the anterior
margin. Discal cell apparently open and broader than in forewing: about 0-66 of the length of the
forewing cell. Cell Sc+ R, wide. Humeral area expanded, angular. Humeral vein indicated by a
straight groove just over | mm long arising from the fork of Sc+R and directed towards the
humeral angle. Rs slightly sinuous. M, closer to M, than M3. Cu trifid, CuA, and CuA, divergent
arising in the outer half of the discal cell. Inwards of 1A+2A on I.10384a are some traces of wing
membrane which apparently include fragments of the anal area, but these are difficult to separate
from the right wing pair which is very close at this point. The hindwing shows faint impressed
traces of the forewing venation.

Colouration: both wings show a light brown pigmentation which is better preserved in the
hindwing. The ventral counterpart of the hindwing shows the underside colour pattern (Fig. 72a).
This includes two dark brown lines in the outer part of the discal cell, another traversing cells C
and Sc+R, and an irregular postdiscal band. The last consists of broad crescents with dark

E. A. JARZEMBOWSKI

N

(wu ¢ = oul] ayes)

“OL ‘Q ‘RZL “SBI{ OS[R 29g “PEEOT'] ‘[BUONUSAUOD UONeIUATIO ‘j9adse [eI]UDA ur adAjofoxy “Au “ds t4aunaz sajypydud N JOSBUIM Sf “Sq
“LL “814 Osye 9ag “yed Woy yoa0dse [BSIOp UI xe10Y} aed

SUIMPUIY 142]

‘

Qpl “814 ‘J4ed19}UN09 puv jied WoJ sjrejap uO paseq SUIMIIOJ Io] “ep, “BIJ ‘adAjOOP -

Vot+V1

~19]UN09 WOIJ UsUIOpgR pue peay : Apo ‘dp “314 “(P866I J) Wed Wo s[rejap owos YUM (69EOT I) Weds9}UNODd UO SUIMAIO] JO [RAOUIAI Jaye

Jayng vnbijun ayodsdoynzT Pl ‘sly

281

PALAEOGENE INSECTS

(wu ¢ =9Ul] 9]89$)
"2 ‘q ‘epl ‘SBI OSTe 995 “PR66I'T ‘(Q4ed) edMojoH ‘JoyINg vAbUUD ayadsdoyuy LL “3th
"CL ‘q ‘BZL SSI OSTR 99S “PEO I ‘G4ied) adAjopoxy “AOU “ds Jounaz sajyoydudn OL *3ty

LL

282 E. A. JARZEMBOWSKI

brown margins and pale centres; it is interrupted in cell M, and offset basally in cell Rs. The
discal line in cells C and Sc+ R, shows traces of a pale area near Sc on its outer side.

Ho vorypee. 1.10384 (B) and counterpart I.10384a (B). Figs 72a, b, 75, 76.

DIMENSIONS. Forewing: maximum width of discal cell (opposite the base of R,) 4:0 mm, length
13-5 mm.
Hindwing: base of M, to base of CuA,, 4-4 mm; base of discal cell to base of My, 9:2 mm.

REMARKS. The general habitus of the fossil is indicative of the Nymphalidae. Zeuner (1961)
regarded it as close to the Recent genus Euthalia Hiibner, but his proposed description (/bid.)
was never published. However, a note in the fossil insect collection at the BM(NH) indicates
that he regarded it as nearest to the allied Tanaecia pelea Fabricius (=T. pulasara Moore). This
species differs principally from the fossil in that the forewing lacks crossvein Ide and the hindwing
has an apically curved humeral vein which originates from Sc+R, distad of the fork of Sc+R
(Schatz & Rober 1885-92 : pl. 26). The absence of Idec in the forewing is also characteristic of
Euthalia Hiibner (Bingham 1905 : 267), the venation of which 1s very similar to Tanaecia Butler
(Schatz & Rober 1885-92 : pl. 26). It is possible that Zeuner’s comparison with T. pelea was
based on the large crescentic markings of the postdiscal band in the hindwing of this Malayan
species, but the band is more irregular and of a different form in the fossil.

Zeuner (1961 : 310) stated that this is the fossil which Handlirsch recorded as Lithosia Fabricius,
a living genus of arctiid moth, based on an old label. Handlirsch (1907 : 923), however, queries
the identification of Lithosia in the Insect Bed and gives Woodward (1879) as the source of the
record. Radcliffe-Grote (1901) also doubted the identification. The determinations in Woodward’s
paper (by F. Smith) were based on material received from A’Court Smith, and although the
specimen is from Brodie’s collection it is quite likely that it subsequently went to the latter, for
Brodie (1878) mentions that he obtained material from A’Court Smith. Specimen 1.10384a is
actually labelled Lithosia in Brodie’s hand, but on labels in the same hand accompanying I.10384
the latter is recognized as a butterfly hindwing accompanied by Scudder’s name. Brodie (1894a, b)
published the determination made by the latter.

The venation of the fossil does not resemble Lithosia, but comes close to that of the living
nymphalid species Neurosigma siva Westwood, Abrota mirus Fabricius and Cymothoe theobene
Doubleday & Hewitson (Schatz & ROber 1885-92 : pls 24-26). It differs principally in that in
these Recent species the forewing appears to have an incipient second anal and M, is much closer
to M, basally; the hindwing has a longer humeral vein which is curved apically and the anterior
part of the wing is wider. The venation of the fossil also resembles that of the living Aeropetes
[= Meneris] tulbaghia (Linné), but the forewing of the latter differs in that R, is closer to Ry,
mde is not curved basally, and in the hindwing the humeral vein has a slight basal curve, Mg is
more central between M, and Mg, and lIdc is present (Schatz & Rober 1885-92 : pl. 35).

The colour pattern of the fossil does not agree with specimens of the above Recent genera in
the collection of the BM(NH) nor with any of the butterflies illustrated in Seitz (1909-39). Of
the known Palaeogene genera of Nymphalidae the fossil can only be placed as a separate species
in the broadly-based genus Nymphalites Scudder.

Personal observations on the decay of the wings of /nachis io (Linné) 1n water indicate that the
darker browns remain when the brighter colours have faded. Considering that the wings of
other insects in the Insect Bed may preserve a dark brown colouration essentially similar to
Recent relatives, e.g. the hemerobiids described herein, it seems likely that the dark brown
markings in Nymphalites zeuneri are original, with a light centre in the postdiscal band.

Recent Nymphalidae are a large, cosmopolitan family feeding on a variety of angiosperms.

Family LYCAENIDAE

Genus LITHOPSYCHE Butler, 1889
(non Lithopsyche Scudder, 1889, =Lithodryas Cockerell, 1909)

1889 Lithopsyche Butler : 294.

PALAEOGENE INSECTS 283
1974 Calospilites van Schepdael : 15, 18.
TyPeE spEciES. Lithopsyche antiqua Butler 1889 : 294; pl. 21, fig. 3; by monotypy.

Lithopsyche antiqua Butler 1889
Figs 74, 77

DESCRIPTION. Body: the head is preserved as a rough cast on the counterpart and shows few
details; labial palpi elongate. An external mould of the dorsal surface of the thorax is well pre-
served in the part; mesothorax elongate, scutoscutellar suture distinct, arcuate; metathorax
short, expanded laterally. Abdomen dorsoventrally compressed with scale traces especially at
the posterior end.

The following account of the wings is based on the left pair which are well separated from
each other by matrix. The right pair are superimposed and their anterior parts are missing; what
can be discerned of the venation is similar to the left pair.

Forewing: original span estimated at 60 mm. Costal area moderately wide, Sc represented only
by basal portion. Three divergent branches of R are preserved apically. Discal cell elongate,
about 0:2 as wide as long. M, _, nearly parallel, with a median M, and faint trace of Ide. I[A+2A
with traces of a basal fork.

Hindwing: broad, cells C and Sc+ R, expanded. Humeral angle rounded with margin thickened
on its inner side. Humeral vein curved apically, originating very close to the separation of Sc+R.
Discal cell short, 0-55 of length of forewing cell. M similar to forewing, CuA, separating near the
posterior end of Idc. 1A+2A represented by a distal portion and well separated from CuAg.
Anal area largely obscured by the abdomen.

Ho orype. 1.19984 (S) and counterpart I.10369 (B). Figs 74a, b, c, 77.
DIMENSIONS. Length of discal cell 15 mm (forewing) and 8-2 mm (hindwing).

REMARKS. Scudder (1883 : 280) used the generic name Lithopsyche for a Palaeogene nymphalid
from Florissant, but this was a nomen nudum and a full description was not published and a type
species designated until 1889. In the same year Butler independently described the present genus of
geometrid moth from the Insect Bed using the same name, and thus created a homonym. Butler’s
paper is dated 21st May but was not published until October (Duncan 1937). Scudder’s paper
has no month of publication although the advertisement with which it is bound is dated Ist
October. The original wrapper of the part in which Butler’s paper appeared is preserved in the
Zoology Library, BM(NH), bound with the journal; the date of issue is given as Ist October.
Butler's name therefore would seem to have priority and the replacement name inadvertently
offered by van Schepdael to be superfluous and invalid.

Butler saw only the part (1.19984), of which he published a colour lithograph (1889 : pl. 21,
fig. 3). Under normal lighting neither half (Fig. 77) shows evidence of the colour banding in
Butler’s figure, although browns are well preserved in other Insect Bed Insecta. Examination
under ultraviolet and infrared lighting only emphasizes some vague traces of pigmentation in
the forewing (Fig. 74a). The left hindwing, now exposed on the counterpart, shows some well-
preserved brown mottles (Fig. 74b).

Butler compared the fossil with 13 species of extant Geometridae, and specimens of these
have been examined. Their only resemblance is that they show much dark pigment in their wing
colouration: the venation is quite different. The wing form and venation of Lithopsyche Butler 1s,
however, close to that of the extant Riodininae species Metacharis ptolomaeus (Fabricius),
Mesene phareus (Cramer), Anteros formosus (Cramer), Theope publius Felder, Uraneis hyalina
(Bates) and Poly stichtis emylius (Cramer) (Schatz & Rober 1885-92 : pls 43-45; Stichel 1910-11 :
pls 8, 13, 15, 22). It differs principally in that there is no evidence of a fork in the outermost
branch of R; however, the latter is incomplete in the fossil and a terminal fork may have been
developed. The fork is absent in some extant Lycaeninae, e.g. Neo/l)caena de Nicéville.

Recent Lycaenidae are a large, cosmopolitan family feeding on a variety of angiosperms; the
larvae are frequently myrmecophilous.

284 E. A. JARZEMBOWSKI
? Superfamily PAPILIONOIDEA

Genus indet.
Fig. 73

DESCRIPTION. The single specimen consists of a decapitated body with wing remains. The thorax
is uncrushed but few details are discernible due to mineralization. Abdomen elongate, flattened
dorsoventrally except near base and apex; the lateral interior of the basal segment contains
traces of cream-coloured calcified muscle, and the apical segments are covered with a dense coat
of piliform scales.

Forewing: (In.6454Sa, left. Fig. 73). Cells C and Sc moderately narrow. Discal cell elongate,
slightly constricted apically and apparently open. Sc terminating on the costal margin a little
distad of the end of the discal cell. Branches of M only slightly divergent and gently inclined
relative to the long axis of the cell; M, nearly equidistant between M, and Ms. R, originating
at 0-7 of the length of the discal cell from base, R, and Rg; ,, originating close together where
the cell begins to constrict; R, parallel to M,. CuA, leaves the cell opposite the base of R, and
CuA, a little basad of base R,. The posterior part of the forewing is obscured by the corres-
ponding hindwing.

Little further can be said concerning the venation except that the veins immediately to the
right of the abdomen (In.64545a) are probably CuA and an anterior anal of the opposite hind-
wing. There is no indication of colour pattern on the wings although scale traces may be dis-
cerned at higher magnification.

MATERIAL. In.64545a, b (H). Fig. 73.

Dimensions. Abdomen: length 13 mm. Forewing: estimated original length 20-25 mm; length of
Rz,, stem, 3-4 mm.

REMARKS. The size and venation place the fossil amongst the higher ditrysian macrolepidoptera.
The forewing venation bears a general resemblance to some Lycaenidae (cf. Bingham 1907 :
figs 72-73), including Curetis siva Evans and Ogyris genoveva Hewitson, although there are
differences in detail. The close proximity of the two outer branches of R resembles certain
Nymphalidae, e.g. Apatura iris (Linné), although the venation differs in other respects such as in
the position of M,. The fossil probably belongs to the Papilionoidea, though in the absence of
further material the placing Is uncertain.

The abdominal muscles probably underwent mineral replacement like some seeds from the
Insect Bed, where the endospermal cells are replaced by a ‘cream-coloured granular substance’
(Reid & Chandler 1926 : 7).

Conclusions

During late Eocene and early Oligocene times southern Hampshire and the northern part of the
Isle of Wight are considered to have formed part of a wet coastal lowland subject to marine
inundation; gradual crustal subsidence resulted in the accumulation of tens of metres of sediment
The Bembridge Marls represent one such marine transgression with a much longer regressive
period (Daley 1973a). Unique conditions near the base of this deposit resulted in the formation of
fine-grained limestone in a unit sandwiched between brackish-water clays (/bid.), thus preserving
the fragile water-logged remains of various small and medium-sized Insecta; the associated biota
has already been discussed (p. 239). Insect orders represented are Hymenoptera, Diptera,
Coleoptera, Neuroptera, Mecoptera, Lepidoptera, Trichoptera, Hemiptera, Thysanoptera,
Psocoptera, Isoptera, Dictyoptera, Orthoptera, Plecoptera, Odonata and possibly Dermaptera.
Of these the first three constitute over 70° of the insects in the collections examined, the first
two being especially common: more than 120 species of Diptera and Hymenoptera have been
described. Although taxonomically diverse, the insects considered in the present paper represent
about 1:5°, of the total number of insects in the collections examined. The A’Court Smith

oO

PALAEOGENE INSECTS 285

collection, much of which went to Brodie and to Hooley (p. 239), is biased towards better-
preserved material, for Brodie states ‘Mr Smith observes that owing to the perverse fracture he
has lost a very considerable number of specimens . . .’ (Brodie 1878 :9) The A’Court Smith,
Hooley and Brodie collections at the British Museum (Natural History) contain some 9000
insects. A bulk sample collected in 1975 yielded nearly 700 insect remains. Of the orders studied,
the numbers of specimens in the main collections compared with the sample (the latter in
parenthesis) are: Isoptera 92 (2), Plecoptera | (0), Neuroptera 10 (1), Mecoptera 2 (0), Lepidoptera
24 (1). The proportional representation of these orders is therefore broadly similar except in the
Isoptera and here the most common species, Mastotermes anglicus von Rosen, is a relatively
large insect and collector’s bias is to be expected.

The Lepidoptera proved the most taxonomically diverse in this study, with 24 species ranging
from pollen feeders to non-feeders amongst the adults and both surface feeders and miners
amongst the larvae. The Bembridge fauna includes the most primitive Lepidoptera (Micropteri-
gidae) through to the most advanced (butterflies); adult Micropterix occurs with its food plants
(Ranunculus, Carex) in the same bed. Of the orders considered here the Lepidoptera are the most
diverse in the British and world fauna at the present day with a total of over 165 000 known
species. They are usually thought to be very rare as fossils but that is probably partly because the
small species are overlooked; however, these specialized terrestrial insects cannot be expected to
form a significant part of an aquatic death assemblage. Further work on some of the microlepi-
doptera from the Insect Bed, such as scanning electron microscope examination for aculae, may
help to clarify their affinities. The Neuroptera (Sisyridae excepted) and Mecoptera are represented
by families in which both adult and immature stages are terrestrial and, like the Lepidoptera,
their infrequency in an aquatic sediment is not surprising. The bionomics of Recent hemerobiids
and chrysopids (Neuroptera) are very similar and both are active predators of plant-feeding
insects such as aphids and coccids (Balduf 1939 : 250, 292). Aphids are present in the Insect Bed
and since the chrysopids and hemerobiids are close to extant forms, it is quite likely that they had
a similar ecological role in Bembridge Marls time. The Mecoptera are represented in the Insect
Bed only by Bittacidae and there is no evidence of Panorpidae or Boreidae in British Palaeogene
strata, the only two families of Mecoptera in Britain at the present day. Carpenter (1954), from
a study of Baltic amber, concluded that bittacids were more diverse during the Palaeogene than
today, although the virtual absence of resin in the British Tertiaries and its abundance in con-
temporaneous strata in the Baltic area (Larsson 1978) is somewhat puzzling.

Aquatic insects in the orders studied are represented by the Neuroptera (Sisyridae) and
Plecoptera, but these freshwater groups are only known from two incomplete wings of the
terrestrial imagines which suggests a limited development of local freshwater habitats; my
observations on the Insect Bed Trichoptera tend to support this as the order is relatively rare
and taxonomically restricted.

The occurrence of four families of termites indicates a warmer climate than today in the
Hampshire Basin during Bembridge Marls time. At the present day their normal poleward limit
in the Palaearctic is latitude 45°N (Harris 1970 : 295). Kalotermitids, rhinotermitids and termi-
tids overlap south of about latitude 35°N in the eastern Palaearctic (/bid.). In the southern
hemisphere the poleward limit of Mastotermes is at the Tropic of Capricorn: the northern tropic
approximates to the equatorial limit of Reticulitermes. The evidence from the termites therefore
suggests that, in temperature terms, the palaeoclimate was close to the warm temperate (sub-
tropical)-tropical boundary in the sense of Miller 1961. This is compatible with the evidence from
the other groups studied. Daley (19726) has suggested that the Eocene climate of Britain was such
as to allow some overlap of modern tropical and temperate biotas and this would facilitate the
coexistence of Reticulitermes and Mastotermes. The Insect Bed Isoptera are represented by
alates which only emerge in the open for their ephemeral swarming-flights. The bodies are
preserved in many specimens of M. anglicus, often with visible appendages, suggesting a fairly
local provenance. The occurrence of Ficus in the Insect Bed is of ecological interest as this is one
of the woody plants which is resistant to attack by extant Mastotermes; the occurrence of this
termite also suggests that precipitation was less than that of modern rainforest. The Insect Bed
termites are represented by wood-feeding forms with one possible exception (p. 253). Modern

286 E. A. JARZEMBOWSKI

Termitidae include many non-lignivorous species and the single detached wing suggests derivation
from a more distant habitat, perhaps in more open country.

The British insect fauna in Bembridge Marls time included a number of groups not represented
here at the present day. In addition to the Isoptera, other notable absentees include the Bittacidae,
Mantispidae and Copromorphidae; the distribution of these and the other taxa is discussed above.

The Insecta are a very diverse group at the present day and were undoubtedly already so in
more recent geological times. Many of the characters used by modern entomologists for generic
and specific classification are often missing or poorly preserved in fossils. Whilst this does not
preclude the possibility of a satisfactory fossil taxonomy, the task of relating Tertiary fossils
precisely to their extant relatives is difficult: further collecting and comparative study will
probably help to clarify relationships. Meanwhile the fossils provide useful data on Tertiary
distribution, ecology and comparative morphology of many modern groups.

Acknowledgements

The author is indebted to the following in various ways. The Lepidoptera sections and Dr P. E.S.
Whalley of the Department of Entomology, BM(NH), for general help with Lepidoptera and the
latter also for guidance on fossil nomenclature; Dr P. C. Barnard of the same department for
general advice on Neuroptera and reading the manuscript; Mr R. M. C. Williams and the
Termite Research Unit of the Centre for Overseas Pest Research for reading the manuscript
and general help, respectively; Dr A. Insole of the Museum of Isle of Wight Geology for access
to collections and loan of material; Mr G. Jones of the Department of Mineralogy, BM(NH), for
spectral analysis of resin; the Electron Microscopy Unit, BM(NH), for use of facilities; Drs R. H.
Bate and C. R. Hill of the Department of Palaeontology, BM(NH), for access respectively to
fossil ostracod and fossil plant collections; Messrs J. Cooper and E. F. Owen of the same depart-
ment for reading the manuscript; and Messrs P. V. York, P. J. Green and T. W. Parmenter of the
Photographic Unit, BM(NH), especially the former, for photographic work. Last but not least
my thanks are due to Norman and the late Mary Preece of West Cowes, Isle of Wight, for help with
fieldwork.

References
Authors of Recent genera and type species are not included.

A’Court Smith, E. J. See Smith.

Balduf, W. V. 1939. The bionomics of entomophagous insects, 2. 384 pp. (1974 reprint) Hampton, Middlesex.

Banks, N. & Snyder, T. E. 1920. A revision of the Nearctic termites with notes on biology and geographic
distribution. Bull. U.S. natn. Mus., Washington, 108. vili+ 228 pp., 35 pls.

Bingham, C. T. 1905-07. Butterflies. Pt 1. Fauna Br. India, London, (Ent.). xxii+501 pp., 10 pls. Pt 2.
loc. cit. Vili+ 480 pp.

Bosma, A. A. 1974. Rodent biostratigraphy of the Eocene—Oligocene transitional strata of the Isle of
Wight. Utrecht micropaleont. Bull. spec. Publs, 1 : 1-128, pls 1-7.

—— & Schmidt-Kittler, N. 1972. Ectropomys exiguus n. gen., n. sp., member of the Otinomyinae n. sub-
fam. (Theridomyidae, Rodentia), from Palaeogene deposits of the Isle of Wight (England) and southern
Gerinany. Proc. K. ned. Akad. Wet., Amsterdam, B 75 : 181-192, pl. 1.

Bourgogne, J. 1951. Ordre des Lépidopteres. /n: Grassé, P.-P. (ed.), Traité de Zoologie 10 (1) : 174-448,
pls 1-3. Paris.

Brodie, P. B. 1878. On the discovery of a large and varied series of fossil insects and other associated
fossils in the Eocene (Tertiary) strata of the Isle of Wight. Proc. Warwick. Nat. Archaeol. Fld Club,
Warwick, 1878 : 1-12.

—— |1894a. Further remarks on the Tertiary (Eocene) insects from the Isle of Wight and on others
from the Lias and Coal Measures. Geo/. Mag., London, (2) 1 : 167-169.

—— 1894b. Notes on the Eocene Tertiary insects of the Isle of Wight. Proc. Warwick. Nat. Archaeol.
Fld Club, Warwick, 1894 : 67-70.

PALAEOGENE INSECTS 287

Brues, C. T., Melander, A. L. & Carpenter, F. M. 1954. Classification of insects. Bull. Mus. comp. Zool.
Hary., Cambridge, Mass., 108. v+917 pp.

Butler, A. G. 1889. Description of a new genus of fossil moths belonging to the geometrid family Eusche-
midae. Proc. zool. Soc. Lond., 59 : 292-297, pl. 31.

Byers, G. W. 1971. Ecological distribution and structural adaptation in the classification of Mecoptera.
Int. Congr. Ent., Moscow (1968), 1 : 486 (only).

Carpenter, F. M. 1931. Revision of the Nearctic Mecoptera. Bull. Mus. comp. Zool. Hary., Cambridge,
Mass., 72 (6) : 205-277, 8 pls.

— 1954. The Baltic amber Mecoptera. Psyche, Camb., 61 (1) : 31-40.

Castel, M. 1968. Zones de Charophytes pour l’Oligocéne d’Europe occidentale. C. r. somm. Séanc. Soc.
géol. Fr., Paris, 4 gull aileal28)5

Cavelier, C. 1968. L’Eocéne supérieur et la base de l’Oligocéne en Europe Occidentale. Mém. Bur. Rech.
géol. miniér., Paris, 58 : 507-527.

—— & Pomerol, C. 1976. Les rapportes entre le Bartonien et le Priabonien. Incidence sur la position de
la limite Eocene moyen—Eocéne supérieur. C. r. somm. Séanc. Soc. géol. Fr., Paris, 18 : 49-51.

—— —— 1977. Proposition d’une échelle stratigraphique standard pour le Paléogéne. News/. Stratigr.,
Leiden, 6 : 56-65.

Chandler, M. E. J. 1964. The Lower Tertiary floras of southern England. IV. xii+ 151 pp., 4 pls. London.

Claassen, P. W. 1931. Plecoptera nymphs of America (North of Mexico). 199 pp., 36 pls. Springfield,
Illinois.

Clark, J., Cole, R., Fawcett, S., Green, M., Howcroft, J., Niedbala, S., Rawlins, K., Theobald, O. &
Tobias, M. 1971. Insecta. Zoo/. Rec., London, 104 (13). xv+ 704 pp.

Cockerell, T. D. A. 1909. A catalogue of the generic names based on American insects and arachnids
from the Tertiary rocks with indications of the type species. Bull. Am. Mus. nat. Hist., New York,
26 : 77-86.

—— 1915. British fossil insects. Proc. U.S. natn. Mus., Washington, 49 : 469-499, pls 60-65.

—— 1917a. New Tertiary insects. Proc. U.S. natn. Mus., Washington, 52 : 373-384, pl. 31.

— 1917b. Fossil insects. Ann. ent. Soc. Am., Washington, 10 : 1-22.

—— 192la. Fossil arthropods in the British Museum, V : Oligocene Hymenoptera from the Isle of
Wight. Ann. Mag. nat. Hist., London, (9) 7 : 1-25.

— 1921b. Fossil arthropods in the British Museum, VI : Oligocene insects from Gurnet Bay, Isle of
Wight. Ann. Mag. nat. Hist., London, (9) 7 : 453-480.

— 1921c. Fossil arthropods in the British Museum, VII. Ann. Mag. nat. Hist., London, (9) 8 : 541-545.

1922. Fossil arthropods in the British Museum, VIII : Homoptera from Gurnet Bay, Isle of Wight.

Ann. Mag. nat. Hist., London, (9) 10 : 157-161.

& Andrews, H. 1916. Dragon-flies from the English Oligocene. Proc. biol. Soc. Wash., 29 : 89-92,
(oll, 2s

—— & Haines, F. H. 1921. Fossil Tipulidae from the Oligocene of the Isle of Wight. Entomologist,
London, 54 : 81-84, 109-112.

Common, I. F. B. 1970. Lepidoptera. Jn: The insects of Australia : 765-866. Melbourne (C.S.1I.R.O.).

Crane, M. D. & Getty, T. A. 1975. An historical account of the palaeontological collections formed by
R.W. Hooley (1865-1923). News/. geol. Curators Grp, Keele, 4 : 170-179.

Curry, D. 1958. Palaeogene : England, Wales and Scotland. Lexique strat. int., Paris, 1 (3aXI1) : 1-82.

—— 1966. Problems of correlation in the Anglo—Paris—Belgian Basin. Proc. Geol. Ass., London, 77 :
437-467.

——, Gulinck, M. & Pomerol, C. 1969. Le Paléocéne et l’Eocéne dans les Bassins de Paris, de Belgique et
d’Angleterre. Mém. Bur. Rech. géol. miniér., Paris, 69 : 361-369.

Daley, B. (1969). A palaeoenvironmental study of the Bembridge Marls (Oligocene) of the Isle of Wight,
Hampshire. Thesis (unpublished), University of Reading.

1971. Diapiric and other deformational structures in an Oligocene argillaceous limestone. Sedim.

Geol., Amsterdam, 6 : 29-51.

1972a. Slumping and micro-jointing in an Oligocene lagoonal limestone. Sedim. Geol., Amsterdam,

7 : 35-46.

19726. Some problems concerning the early Tertiary climate of southern England. Palaeogeogr.

Palaeoclimat. Palaeoecol., Amsterdam, 11 : 177-190.

1973a. The palaeoenvironment of the Bembridge Marls (Oligocene) of the Isle of Wight, Hampshire.

Proc. Geol. Ass., London, 84 : 83-93.

19736. Fluvio-lacustrine cyclothems from the Oligocene of Hampshire. Geo/. Mag., London, 110 :

235-242, pls 1-2.

288 E. A. JARZEMBOWSKI

—— & Edwards, N. 1971. Palaeogene warping in the Isle of Wight. Geol. Mag., London, 108 : 399-405.

Denizot, G. 1956. Tertiare. Lexique Strat. Int., Paris, 1 (4aVI1) : 1-217, 14 maps.

—— 1968. Bartonien, Ludien et Tongrien. Mém. Bur. Rech. géol. miniér., Paris, 58 : 533-552.

Donisthorpe, H. St J. K. 1920. British Oligocene ants. Ann. Mag. nat. Hist., London, (9) 6 : 81-93, pl. 5.

Duncan, F. M. 1937. On the dates of publication of the society’s ‘Proceedings’ 1859-1926. Proc. zool.
Soc. Lond., 107 A : 71-84.

Durrant, J. H. See Cockerell, 19216.

Edwards, F. W. 1923. Oligocene mosquitoes in the British Museum, with a summary of our present
knowledge concerning fossil Culicidae. Q. J/ geol. Soc. Lond., 79 : 139-155, pl. 7.

Edwards, N. (1966). A study of the Headon Beds, Osborne Beds and Bembridge Limestone (Upper Bartonian
and Lattorfian), in the Isle of Wight and Hampshire, England. Thesis (unpublished), University of
Reading.

—— 1971. Stratigraphy and correlation of the Headon, Osborne, Bembridge and Hamstead Beds
(Palaeogene), Hampshire Basin —a bibliography (1814-1970). J. Soc. Biblphy nat. Hist., London, 6 :
50-60.

Emerson, A. E. 1965. A review of Mastotermitidae (Isoptera), including a new fossil genus from Brazil.
Am. Mus. Novit., New York, 2236 : 1-46.

—— 1969. A revision of the Tertiary fossil species of the Kalotermitidae (Isoptera). Am. Mus. Novit.,
New York, 2359 : 1-57.

Esben-Petersen, P. 1921. Mecoptera. Monographic revision. Collections Zoologiques de Baron Edm. de
Selys Longchamps, 5 (2). 172 pp., 2 pls. Brussels.

Forbes, E. 1856. On the Tertiary Fluvio-marine Formation of the Isle of Wight. Mem. geol. Sury. U.K.,
London. xvili+ 162 pp., 10 pls.

Forbes, W. T. M. 1923. The Lepidoptera of New York and neighboring states. Primitive forms — Bombyces.
729 pp. Ithaca, New York.

Fraser, F. C. 1959. Mecoptera, Megaloptera, Neuroptera. Handbk Ident. Br. Insects, London, 1 (12-13).
40 pp.

Fuller, C. 1919. The wing venation and respiratory system of certain South African termites. Ann. Natal
Mus., Pietermaritzburg, 4 (1) : 19-102, 9 pls.

Gay, F. J. 1970. Isoptera. In: The insects of Australia : 275-293. Melbourne (C.S.1.R.O.).

—— & Calaby, J. H. 1970. Termites of the Australian Region. Jn: Krishna, K. & Weesner, F. M.
(eds), The Biology of Termites 2 : 393-448. New York & London.

Grassé, P.-P. 1951. Supre-ordre des Mécoptéroides. Jn: Grassé, P.-P. (ed.), Traité de Zoologie 10 (1) :
71-124. Paris.

Hampson, G. F. 1896. Moths, 4. Fauna Br. India, London, (Ent.). xxvi+ 594 pp.

Handlirsch, A. 1906-08. Die fossilen Insekten und die Phylogenie der rezenten Formen. 2 vols. ix + 1430 pp.,
51 pls. Leipzig.

Handschin, E. 1959. Beitrage zu einer Revision der Mantispiden (Neuroptera). 1. Mantispiden des Musée
Royal du Congo Belge, Tervuren. Revue Zool. Bot. afr., Brussels, 59 : 185-227.

Harris, W. V. 1968. Isoptera from Vietnam, Cambodia and Thailand. Opusc. ent., Lund, 33 : 143-154.

—— 1970. Termites of the Palearctic Region. Jn: Krishna, K. & Weesner, F. M. (eds), The Biology of
Termites 2 : 295-313. New York & London.

Haskins, C. W. 1968. Tertiary Ostracoda from the Isle of Wight and Barton, Hampshire, England.
Part II. Revue Micropaléont., Paris, 11 : 3-12, 2 pls.

—— 1971. The stratigraphical distribution and palaeoecological significance of the Ostracoda from the
Lower Tertiary Beds of the Hampshire Basin, England. Bull. Cent. Rech. Pau, 5 (Supplément) : 545—
Sy:

Heer, O. 1856. Ueber die fossilen Insekten von Aix in der Provence. Vjschr. naturf. Ges. Ziirich, 1 : 1-40,
pls 1-2.

Hill, G. F. 1942. Termites (Isoptera) from the Australian Region. 479 pp., 24 pls. Melbourne.

Hughes, J. C. 1922. The geological story of the Isle of Wight. viii+ 115 pp., 4 pls, 1 map. London.

Hynes, H. B. N. 1941. The taxonomy and ecology of the nymphs of British Plecoptera with notes on the
adults and eggs. Trans. R. ent. Soc. Lond., 91 (10) : 459-557.

Jackson, J. F. 1933. An outline of the history of geological research in the Isle of Wight. Proc. Isle Wight
nat. Hist. archaeol. Soc., Newport, 2 : 211-220.

Jarzembowski, E. A. 1976. Report of Easter field meeting: the Lower Tertiaries of the Isle of Wight.
Tertiary Res., London, 1 ;: 11-16.

Jones, T. R. & Sherborn, C. D. 1889. A supplementary monograph of the Tertiary Entomostraca of
England. Palaeontogr. Soc. (Monogr.), London. 55 pp., 3 pls.

PALAEOGENE INSECTS 289

Karsholt, O. & Nielsen, E. S. 1976. Catalogue of the Lepidoptera of Denmark. 128 pp. Klampenborg.

Keen, M. C. 1972. The Sannoisian and some other Upper Palaeogene Ostracoda from north-west Europe.
Palaeontology, London, 15 : 267-325, pls 45—S6.

Kellogg, V. L. 1894. The taxonomic value of the scales of Lepidoptera. Kans. Univ. Q., Lawrence, 3 (1) :
45-89, pls 9-10.

Kernbach, K. 1967. Uber die bisher im Pliozin von Willershausen gefundenen Schmetterlings- und
Raupenreste. Ber. naturhist. Ges. Hannover, 3 : 103-108.

Killington, F. J. 1936. A monograph of the British Neuroptera, 1. xix+ 269 pp., 15 pls. London.

Kimmins, D. E. 1952. A revision of the genera of the Apochrysinae (family Chrysopidae). Ann. Mag.
nat. Hist., London, (12) 5 : 929-944.

Krishna, K. 1961. A generic revision and phylogenetic study of the family Kalotermitidae (Isoptera).
Bull. Am. Mus. nat. Hist., New York, 122 : 303-408.

Larsson, S. G. 1978. Baltic amber —a palaeobiological study. Entomonogr., Klampenborg, 1 : 1-192,
12 pls.

McCook, H. C. 1888a. A new fossil spider, Eoatypus woodwardii. Proc. Acad. nat. Sci. Philad., 1888 :
200-202.

—— 1888». A new fossil spider (Eoatypus woodwardii). Ann. Mag. nat. Hist., London, (6) 2 : 366-369.

Machin, J. 1971. Plant microfossils from the Tertiary deposits of the Isle of Wight. New Phytol., London
& Cambridge, 70 : 851-872, 2 pls.

McKeown, K. C. & Mincham, V. H. 1948. The biology of an Australian mantispid (Mantispa vittata
Guerin). Aust. Zool., Sydney, 11 (3) : 207-224, pls 14-15.

MacLeod, E. G. 1970. The Neuroptera of the Baltic amber, 1. Ascalaphidae, Nymphidae, and Psycho-
psidae. Psyche, Camb., 77 : 147-180.

Martini, E. 1972. Die Gattung Eosphaeroma (Isopoda) im europdischen Alttertiar. Senckenberg. leth.,
Frankfurt a.M., 53 : 65-79, pls 1-2.

Meyrick, E. 1928. Revised handbook of British Lepidoptera. vit 914 pp. Hampton, Middlesex.

Miller, A. A. 1961. Climatology (9th ed.). xii+ 320 pp. London.

Nakahara, W. 1958. The Neurorthinae, a new subfamily of the Sisyridae (Neuroptera). Mushi, Fukuoka,
32 : 19-32, pls 7-9.

—— 1960. Systematic studies on the Hemerobiidae (Neuroptera). Mushi, Fukuoka, 34 : 1-69, 16 pls.

Needham, J. G. & Claassen, P. W. 1925. A monograph of the Plecoptera or stoneflies of America north of
Mexico. 397 pp., 50 pls. La Fayette, Indiana.

Noirot, C. 1970. The nests of termites. Jn: Krishna, K. & Weesner, F. M. (eds), The Biology of Termites
2 : 73-125. New York & London.

Pallot, J. See Machin, J.

Pantin, H. M. 1958. Rate of formation of a diagenetic calcareous concretion. J. sedim. Petrol., Menasha,
28 : 366-371.

Parfin, S. I. & Gurney, A. B. 1956. The spongilla-flies, with special reference to those of the western
hemisphere (Sisyridae, Neuroptera). Proc. U.S. natn. Mus., Washington, 105 : 421-529, 3 pls.

Piton, L.-E. 1940. Paléontologie du Gisement Eocéne de Menat (Puy-de-D6éme). vii + 306 pp., 26 pls. Paris.

Radcliffe-Grote, A. 1901. Fossile Schmetterlinge und der Schmetterlingsfliigel. Verh. zool.-bot. Ges.
Wien, 51 : 655-661.

Rebel, H. 1935. Bernstein Lepidopteren. Dt. ent. Z. Iris, Dresden, 49 : 162-186.

Reid, C. 1889. Jn: Reid, C. & Strahan, A., Geology of the Isle of Wight. xiv +349 pp., 5 pls, 4 appendices.
London, Mem. geol. Sury. U.K.

Reid, E. M. & Chandler, M. E. J. 1926. Catalogue of Cainozoic Plants in the Department of Geology.
1 : The Bembridge Flora. viii+ 206 pp., 12 pls. London.

Rey, R. 1967. Essais de corrélations entre bassins Oligocenes de l’Europe occidentale a |’aide des gastéro-
podes continentaux. Bull. Soc. géol. miner. Bretagne, Rennes, 1966 : 1-15.

Riek, E. F. 1970. Plecoptera, Neuroptera, Mecoptera. Jn: The insects of Australia : 314-322, 472-494,
636-646. Melbourne (C.S.I.R.O.).

Robinson, G. S. 1977. A taxonomic revision of the genus Callipielus Butler (Lepidoptera : Hepialidae).
Bull. Br. Mus. nat. Hist., London, (Ent.) 35 : 101-121, 6 pls.

Roonwal, M. L. 1970. Termites of the Oriental Region. /n: Krishna, K. & Weesner, F. M. (eds), The
Biology of Termites 2 : 315-391. New York & London.

Rosen, K. von 1913. Die fossilen Termiten : eine kurze Zusammenfassung der bis jetzt bekannten Funde.
Int. Congr. Ent., Oxford (1912), 2 : 318-335, pls 26-31.

Schatz, E. & Rober, J. 1885-92. Die Familien und Gattungen der Tagfalter. Jn: Staudinger, O. & Schatz,
E., Exotische Schmetterlinge 2. ii+ 284 pp., 50 pls. Furth (Bayern).

290 E. A. JARZEMBOWSKI

Schepdael, J. van 1974. Macrolépidopteres fossiles du domaine paléarctique. Naturalistes belg., Brussels,
SS (0) 3 Sasi7e

Scudder, S. H. 1883. The tertiary lake basin at Florissant, Colo., between South and Hayden Parks.
Rep. U.S. geol. geogr. Sury. Territ., Washington, 12 (1) : 271-293, 1 map.

1889. The fossil butterflies of Florissant. Rep. U.S. geol. Surv., Washington, 8 (1) : 433-475, pls 52-53.

—— 1894. See Brodie 1894a, b.

Seitz, A. 1909-39. Les Macrolépidoptéres du Globe. (1) 1, viii+ 380 pp., 89 pls (1911); Supplément, vii+
404 pp., 16 pls (1939). (2) 5 (planches), ix pp., 194 pls (1926). (2) 13 (planches), 79 pls (1909). (3) 9
(planches), xii pp., 175 pls (1929). Stuttgart.

Setty, L. R. 1940. Biology and morphology of some North American Bittacidae (Order Mecoptera).
Am. Midl. Nat., Notre Dame, Ind., 23 (2) : 257-353.

Smith, E. J. A’Court 1874. Discovery of the remains of plants and insects. Nature, Lond., 11 : 88 (only).

Smith, F. See Woodward, H.

Spuler, A. 1913. Die sogenannten Kleinschmetterlinge Europas. cxxvii+ 402 pp., 22 pls. Stuttgart.

Stichel, H. 1910-11. Family Riodinidae. Genera Insect., Brussels, 112 : 1-452, pls 1-27.

Stinton, F. C. 1975. Fish otoliths from the English Eocene. Pt 1. Palaeontogr. Soc. (Monogr.), London.
56 pp., 3 pls.

Tasch, P. 1969. Branchiopoda. Jn: Moore, R.C. (ed.), Treatise on Invertebrate Paleontology, R : 128-191.
Kansas.

Tillyard, R. J. 1931. The wing venation of the order Isoptera, 1. Introduction and the family Mastoter-
mitidae. Proc. Linn. Soc. N.S.W., Sydney, 56 : 371-390, pl. 21.

Tjeder, B. 1959-66. Neuroptera—Planipennia. The Lacewings of southern Africa: 2, Family Berothidae;
4, Family Hemerobiidae; 5, Family Chrysopidae. S. Afr. anim. Life, Stockholm, 6 : 256-314, 8 : 296—
408, 12 : 228-534.

van Schepdael, J. See Schepdael.

von Rosen, K. See Rosen.

Weesner, F. M. 1970. Termites of the Nearctic Region. Jn: Krishna, K. & Weesner, F. M. (eds), The
Biology of Termites 2 : 477-525. New York & London.

Whalley, P. E. S. 1961. A change of status and a redefinition of the subfamily Endotrichinae (Lep.
Pyralidae) with the description of a new genus. Ann. Mag. nat. Hist., London, (13) 8 : 733-736.

— 1977. Lower Cretaceous Lepidoptera. Nature, Lond., 266 : 526 (only).

White, H. J. O. 1921. A short account of the geology of the Isle of Wight. Mem. geol. Surv. U.K., London.
vili+201 pp.

Woodward, H. 1878. On the occurrence of Branchipus (or Chirocephalus) in a fossil state, associated with
Archaeoniscus and with numerous insect remains in the Eocene freshwater limestone of Gurnet Bay,
Isle of Wight. Geol. Mag., London, (2) 5 : 88-89.

— 1879. On the occurrence of Branchipus (or Chirocephalus) in a fossil state, associated with
Eosphaeroma and with numerous insect remains, in the Eocene freshwater (Bembridge) Limestone of
Gurnet Bay, Isle of Wight. Q. J/ geol. Soc. Lond., 35 : 342-350, 1 pl.

Zagulyaeyv, A. K. 1964. Lepidoptera, IV (2). Tineidae, 2 : Nemapogoninae. Fauna SSSR, Moscow &
Leningrad, (Ent.) 86 : 1-416, pls 1-2. [In Russian; transl. (1968) vii+436 pp., Jerusalem].

Zeuner, F. E. 1939. Fossil Orthoptera Ensifera. xiii+ 321 pp., 80 pls. London.

1961. Notes on the evolution of the Rhopalocera (Lep.). Int. Congr. Ent., Vienna (1960), 1 : 310-313.

Zwick, P. 1967. Beschreibung der aquatischen Larve von Neurorthus fallax (Rambur) und Errichtung der
neuen Planipennierfamilie Neurorthidae fam. nov. Gewdss. Abwdss, Dusseldorf, 44—45 : 65-86.

PALAEOGENE INSECTS 291

Index

New taxonomic names and the page numbers of the principal references are in bold type. An asterisk (*)

denotes a figure.

Abrota mirus 279, 282
Acrolophus 274, 276

cf. cossoides 273*
Adela 274

reamurella 269

ridingsella 272
Aegidomorpha 270
Aeropetes tulbaghia 282
Agnippe sp. 269
Anostraca 240-1, 247
Anteros formosus 283
Apatura iris 284
Apicotermes occultus 253
Araneida 257
Austrobittacus 262
Austroneurorthus 261

Bedellia 271
Bembridge Beds 240, 242
Foreland 241, 243
Marls 239, 239*, 240-1, 242*, 243, 284
syncline 243
bird feather 241, 268*
Bittacidae 262
Bittacus 262-3
chlorostigma 258*
marginatus 262
minutus 262
sinensis 258*, 262
veternus 238, 256*, 258*, 262-3
sp. A 258*, 263
Bostra 278
Branchipodites vectensis 240-1, 247
Branchipus 241; see Branchipodites
British Museum (Natural History) 239-40, 244,
247, 250, 262
Brodie, Rev. P. B. 239, 247, 285
butterfly 279

Caloptilia alchimiella 274
Calospilites 283
Carex 265, 285
Chloroclysta truncata 271*, 278
Chrysopidae, Chrysopinae 260
sp. A 258*, 260
Chrysopinae 261
Cliff End 242
Climacia areolaris 261
Colenutt, G. W. 239, 243
Copromorpha 270
fossilis 239, 264*, 270
gypsota 264*, 270
Copromorphoidea, Copromorphidae 270
Corbiculidae 242

? Cossidae 275

? Cossoidea 275

Cowes 237-40

Culama sp. 273*, 276
Culicidae 240

Curetis siva 284

Cymothoe theobene 279, 282
Cypridopsis 240

Diloxia 278

Ditaxis biseriata 257

Drepanacra 259

Dyselachista saltatricella 27\
sericiella 271

Eoatypus woodwardii 240

Eosphaeroma fluviatile 240
margarum 240

Euthalia 279, 282

Fagus 269

fauna, Insect Bed (Bembridge Marls) 239-241
feather, bird 241, 268*

Ficus 248, 285

flora, Insect Bed (Bembridge Marls) 239-241
Formicidae 240

fossil insects (generally) 239-41, 284-6

fusain 241

Galba 240
Gelechia 274
Gelechioidea sp. A 266*, 269
Geometridae 278
Geometridites 278
larentiiformis 239, 277*, 278
repens 278
Geometroidea 278
Glyptotermes 250
Gurnard 243
Bay 242-3
Ledge 242-3
Gurnetia 275, 276
durranti 239, 267*, 273*, 275

Hamstead Ledge 242
Haplotinea insectella 272
Harpobittacus 262
Harris, T. M. 241
Heliozela 271
Hemerobiidae 250

spp. A-C 254*, 260

292 E. A. JARZEMBOWSKI

Hemerobius 257, 259-60 Microlepidoptera sp. A 264*, 270
atrifrons 259 sp. B 271
decisus 259 sp. C 266*, 271
humulinus 257 sp. D 271
tinctus 238, 254*, 257, 259-60 sp. E 272
Hepialidae, Hepialoidea 265 sp. F 272
Hepialus 265 sp. G 272, 273*
fusconebulosa 264*, 265 sp. H 273*, 274
Heterotermes 252 sp. I 274
Hippeutis 241 sp. J 274
Hooley, R. W. 239, 285 sp. K 266*, 275
Howgate Bay 241, 243 sp. L 273*, 275
Micropterigidae 263
Waachisioo Micropterix 263-5, 285

anglica 238, 263, 264*

Incisitermes 250-1 calthella 264

minor 250 ee Be 3
tabogae 250, 251* Regu papel 264
Insect Bed (Bembridge Marls) 239-44, 248, 285 LAM ES

: 62x muscle replacement 284
Dee ee 239-42 Museum of Isle of Wight Geology 239
Isopoda 240 mygalomorph spider 240

Nemapogon granella 269
Nemophora 276
Nemoura 253

priscula 238, 253
Neolycaena 283
Neomyzaphis 259
Nepticula 271

terminella 271
Neuronema 259

Kalobittacus 262
Kalotermes 250-2
disruptus 238, 246*, 248*, 249, 251
flavicollis 250, 251*
rhenanus 250-1
rufinotum 250
Kalotermitidae 249

Lacoe Collection 239-40 irrorata 259

Leucoptera 271 ? Neuronema sp. A 238, 258*, 259, 260
laburnella 271 sp. B 238, 258*, 259

Leuctra 253, 255 Neurorthinae 261
priscula 238, 249*, 253 Neurorthus 261

Leuctridae 253 Neurosigma siva 279, 282

Leuctrinae 244 Newtown River 239, 242-4

Lithodryas 282 Node’s Point 243

Lithopsyche 282, 283 Nothochrysa capitata 258*
antiqua 239-40, 280*, 281*, 283 Nymphalidae 279, 282

Lithosia 279, 282 Nymphalites 279, 282

localities 239* obscurum 279

Lycaenidae, Lycaeninae 283 zeuneri 239, 277*, 279, 280*, 281*, 282

Lypusa maurella 275
Oecophoridae 270

Mantispa 255, 257 Oecophylla 243

relicta 238, 255 Ogyris genoveva 284

tenella 256* Orthoptera 240
Mantispidae 255 Orthotaelia sparganella 276
Mantispinae 255, 257 Ostracoda 240
Mastotermes 238, 240, 244, 246-7, 285

anglicus 244*, 245*, 245, 246*, 247, 249*, 285 Panorpa 262

batheri 245 germanica 262

darwiniensis 244, 247 italica 262
Mastotermitidae 244 veterna 238, 262
material studied 244 Panorpidae 262
Mesene phareus 283 ? Papilionoidea sp. 277*, 284

Metacharis ptolomaeus 283 Parachronistis albiceps 269

PALAEOGENE INSECTS 293

Paractenia 278 Sisyra 250, 261
Paratriaxomasia solentensis 239, 266*, 267 fuscata 254*
Phryganea 245, 247 panama 261
fusca 253 vicaria 250, 261
Plutella 274 Sisyra (?) disrupta 238, 249
annulatella 275 Sisyridae 250, 261
Polyhymno luteostrigella 269 sp. A 254*, 261
Polystichtis emylius 283 Sisyrinae, see Sisyridae
Porchfield anticline 242 Smith, E. J. A°Court 239, 241, 243, 247, 284-5
Porifera 261 Smith, Frederick 240
Potamides 242 Spongillidae 261
Potamocypris brodiei 240 Stegasta 274
Priory Bay 241, 243 Sticelett Ledge 242-3
Prohepialus 265
incertus 265 Tanaecia 282
sp. 239, 256*, 264*, 265, 267* pelea 282
Promantispa 255 pulasara 282
relicta 238, 254*, 255, 257 Tarebia 242
Protozoa 251 Termes flavicolle 249
Purbeck Beds 239 flavipes 251
Pyralidae, Pyralinae 276 Termitidae sp. A 246, 253, 255*
Pyralites 276 Theope publius 283
obscurus 276, 278 Thorness Bay 242-3
preecei 239, 276, 277* Thorness syncline 242
Pyrsonymphidae 252 Tinea pellionella 269
podevinella 263
Tineidae 267
Quercus 269 sp. A 266*, 269

Triaxomasia 267-8
caprimulgella 266*, 269

Triaxomera 269

Tyndis 278

Typha latissima 240

Tytthobittacus 262

Rachiberothinae 257
Ranunculus 265, 285
Raphidia 238; see Rhipidia
resin 241
Reticulitermes tibialis 252*
sp. 238, 249*, 251, 252*, 252, 285
Rhinotermitidae 251
Rhipidia 244
Riphidia, see Rhipidia
Rophalis relicta 261
Rugitermes 250

Ulmus 269
United States National Museum 239
Uraneis hyalina 283

Viviparus 241

West Cowes 239-40, 243-4
St Helens 240-1, 243, 247 Whitecliff Bay 241, 243
Saltmead Ledge 242 Point 241, 243

Accepted for publication 19 July 1978

Monographs on the Palaeobotany of Southern
England, published by the British Museum
(Natural History)

Catalogue of the Caenozoic Plants in the Department of Geology.

Vol. I. The Bembridge Flora. E. M. Reid and M. E. J. Chandler with a section
on the Charophyta by J. Groves
1926, viii+206 pp, 12 plates, 20 text figures, 8vo, £7-15

Lower Tertiary Floras of Southern England

Vol. I. Palaeocene Floras, London Clay Flora (supplement).
M. E. J. Chandler
1961, v+ 354 pp, 54 text figures, 4to

Atlas of 34 plates, 1961, 4to, £18-00 the set

Vol. II. Flora of the Pipe-Clay Series of Dorset. M. E. J. Chandler
1962, v+ 176 pp, 29 plates, 25 text figures, 4to, £12-50

Vol. III. Flora of the Bournemouth Beds; The Boscombe and the Highcliffe
Sands. M. E. J. Chandler, 1963, v+169 pp, 25 plates, 33 text figures, 4to, £15-00

Vol. IV. A Summary and Survey of Findings in the Light of Recent Botanical
Observations. M. E. J. Chandler
1964, xii+ 150 pp, 4 plates, 6 text figures, 4to, £12-00

Titles to be published in Volume 33

An account of the Ordovician rocks of the Shelve Inlier in West

Salop and part of north Powys. By the late W. F. Whittard, F.R.S.

(Compiled by W. T. Dean)
Miscellanea

The Caradoc faunal associations of the area between Bala and
Dinas Mawddwy, north Wales. By M. G. Lockley

Fossil insects from the Bembridge Marls, Palaeogene of the Isle of
Wight, southern England. By E. A. Jarzembowski

The Yorkshire Jurassic fern Phlebopteris braunii (Goeppert) and its
reference to Matonia R. Br. By T. M. Harris

The entire Geology series is now available as a series

Printed by Henry Ling Ltd, Dorchester

‘
ee Se ee ee ee

.
;

Bulletin of the | |
British Museum (Natural History)

The Yorkshire Jurassic fern Phlebopteris
_braunii (Goeppert) and its reference to
_ Matonia R. Br.

-T. M. Harris, F.R.S.

"Geology series Vol 33 No5 27 March 1980

The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four
scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology,
and an Historical series.

Papers in the Bulletin are primarily the results of research carried out on the unique and
ever-growing collections of the Museum, both by the scientific staff of the Museum and by
specialists from elsewhere who make use of the Museum’s resources. Many of the papers are
works of reference that will remain indispensable for years to come.

Parts are published at irregular intervals as they become ready, each is complete in itself,
available separately, and individually priced. Volumes contain about 300 pages and are not
necessarily completed within one calendar year. Subscriptions may be placed for one or more
series. Subscriptions vary according to the contents of the Volume and are based on a forecast
list of titles. As each Volume nears completion, subscribers are informed of the cost of the
next Volume and invited to renew their subscriptions. Orders and enquiries should be sent to:

Publications Sales,
British Museum (Natural History),
Cromwell Road,
London SW7 5BD,
England.

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

© Trustees of the British Museum (Natural History), 1980

This number completes Volume 33

ISSN 0007-1471 Geology series
Vol 33 No 5 pp 295-311
British Museum (Natural History)
Cromwell Road
London SW7 5BD Issued 27 March 1980

The Yorkshire Jurassic fern Phlebopteris braunii | -
(Goeppert) and its reference to Matonia R. Br. \*,

T. M. Harris, F.R.S.

Professor Emeritus in the University of Reading, Department of Geology.

Contents
Synopsis , ; : ; ; 3 : : : ; ; ; > 295
Introduction . : f : : ; j : : ‘ : ‘ e295
Material 3 E ‘ ; : : : : 3 : ; : . 296
Description . ; 5 : : 2 : ; ’ t d : . 296
Discussion : ; : : : é : : é , : i . 300
The indusium . ‘ 3 : ; : F : , : - 300
Contour of fertile pinnule 3 : 3 : : , : . 302
Identity of Phlebopteris muensteri with P. Brauinil : : : : : . 303
Identification of Yorkshire specimens . : : : ; 3 5 . 304
Reference of Phlebopteris braunii to Matonia : : : : : . 304
Comparison of M. braunii with Yorkshire Matoniaceae : : : . 305
Distribution of M. braunii : ; ‘ F : : : : . 306
Field key to the Yorkshire Matoniaceae_ . j : ; 5 ‘ . 307
Notes on fossil matoniaceous leaves described up to 1841 F ‘ : : eS Oi
Summary ; 5 ; , : : : ; ‘ ; ‘ ‘ . 309
Acknowledgements . : : é : : ; F : ‘ : . 310
References , ' : ; A : : : : ; : : . 310

Synopsis

Recently-collected material of the matoniaceous fern Phlebopteris braunii (Goeppert), first described and
best known from the Lower Lias of Germany, is described from the Middle Jurassic (Bajocian) of
Yorkshire. The Yorkshire leaves are shown to have indusiate sori like the living Matonia pectinata R.
Br., whereas the Liassic P. braunii was firmly believed to have naked sori. Search has however revealed
an indusium in Liassic specimens also, and reasons why it had been missed are suggested. Phlebopteris
muensteri (Schenk 1867) is regarded as a synonym of P. braunii and the combined species is renamed
Matonia braunii (Goeppert) n. comb.

The present status of the Yorkshire Jurassic Matoniaceae is considered. A few specimens of M. braunii
had been mistakenly identified with other species, blurring their definition. In addition, reasons are
given for conserving the specific name braunii Goeppert 1841 against names of certain Yorkshire specimens
described earlier. The early descriptions of fossil Matoniaceae are also reviewed in general, with notes on
the type status of some of the specimens.

Introduction

Fronds agreeing with the well-known Lower Liassic fern Phlebopteris braunii (Goeppert) (P.
muensteri (Schenk)) were collected from the Bajocian of Yorkshire in 1972-73. When suitably
prepared the new specimens proved to have indusiate sori like the living Matonia whereas P.
braunii was believed to have naked sori, a position which seemed, at first, confusing. I then
realized that certain Yorkshire specimens which I had identified with Matonidium goeppertii

Bull. Br. Mus. nat. Hist. (Geol.) 33 (5) : 295-311 Issued 27 March 1980
295

296 T. M. HARRIS

Schenk because they possessed indusia agreed better with the new material. Later still I found
that specimens collected from the classic Gristhorpe Bed and described early in the nineteenth
century under various names were also similar; see notes on p. 301. Finally I found that
Phlebopteris braunii itself, from the original localities in the Lias of Bavaria and from East
Greenland, also has indusiate sori.

When exposed by splitting the rock, most of the new fronds show their upper surface, and the
fertile pinnules only reveal their sori satisfactorily when prepared as transfers. However, a number
of Lower Liassic specimens from Bavaria and from Greenland show their lower surface as
collected, and since their indusia are in most cases missing the sporangia are seen clearly. There
thus seemed no point to early workers, including myself, in preparing transfers of such specimens.
Certain Greenland specimens did show sori covered by indusia but when I worked on them 50
years ago I dismissed them as unsatisfactory and kept them merely as duplicates for exchange.
Apparently Bavarian collectors did the same. Happily the duplicates were sent to the British
Museum (Natural History) and I re-examined them recently, when ready to perceive their nature.

Harris (1931) gave reasons for identifying Phlebopteris muensteri (Schenk) with P. braunii
(Goeppert), a conclusion to which I adhere. The discovery of the indusium, however, distinguishes
the combined species from the genus Phlebopteris, and it is here placed in Matonia, as Matonia
braunii (Goeppert) n. comb.

Material

Most of the new specimens were collected from the Hillhouse Nab Plant Bed, Farndale, at about
54°23'3''’N, 0°59’0’’W. This bed was first found in about 1950 and consists of a 2 m thick and
rather extensive layer of plant-bearing rocks at the base of the deltaic beds. Like many other
Yorkshire plant beds it includes lenticles providing abundant specimens of a particular species
and one such lenticle, about 2 m broad and 10 cm thick, gave the specimens described here. This
lenticle is the only rich source of P. braunii in Yorkshire though material had earlier been collected
at Saltwick, also at the base of the deltaics, and from the Gristhorpe Bed at a higher level, but only
as occasional specimens.

The Hillhouse Nab matrix is a fragile silty shale and rarely gave blocks more than 20 cm wide.
Together with overlap of the crowded fronds this limited the size of the specimens as collected,
though some of the fronds may well have been complete when deposited. The matrix is one in
which the coaly substance of fern and other leaves with delicate cuticles has become crossed by
fine cracks and is apt to crumble away as dust. This is the cause of gaps in drawings of the veins
(Figs 5, 6, 15). In most leaves the lamina is opaque but in some layers it had undergone oxidative
decay at or following deposition, and the veins show as dark strands on a brown lamina. Other-
wise the lateral veins, if seen at all, are marked only by their elongated epidermal cells though the
midribs form a small ridge above and a larger ridge below. Most fertile pinnules merely show their
sori as obscure bulges on the upper surface and when transferred prove denuded, showing the
placenta as a boss. But the best leaves have sori which retain both their sporangia and indusia
(though these are apt to be damaged when a transfer is made). Nearly all sporangia were found to
be empty when macerated but one was full of ripe spores.

The Gristhorpe and Saltwick specimens are as well preserved as Lower Lias ones and compared
with the Hillhouse Nab material the lamina is less completely flattened and the coaly substance
more coherent.

Description

Frond form. All the new specimens are somewhat damaged and not one shows a perfect frond
centre spread out in a symmetrical fan as in a carefully pressed leaf of Matonia pectinata. The
specimen in Fig. | is the left half of a frond centre, together with a small, probably extraneous
axis on the right. The pinna rachises increase in width, left to right, from 1:2 to 1-6 mm and they
are borne at the end of a bare ‘arm’. Some herbarium specimens of M. pectinata in the

MATONIA BRAUNII (GOEPPERT) 297

YU

Figs 1-6 Matonia braunii (Goeppert) comb. nov. Hillhouse Nab, Farndale. Fig. 1. Half of leaf centre,
drawn from part and counterpart, V.59743, x 1. The separate axis on right is probably extraneous.
Fig. 2. Half of leaf centre, drawn from part and counterpart, V.59744, « 1. Top of rachis twisted and
folded. Fig. 3. Apex of fertile pinna, V.59748, x 1. Fig. 4. Apex of sterile pinna, V.59744, x 1. Fig. 5.
Venation of small sterile pinnule, shown on imprint of under surface, V.59746, 6. Vein meshes are
numerous in this specimen. Fig. 6. Venation of lowest pinnule from pinna in Fig. 8, p. 298. At x note
the branch vein which ends blindly at a point where a placenta might be expected. V.59745, x 5.

BM(NH) show ‘arms’ bare below and could provide such a fragment. The frond centre in Fig. 2
shows the right half only; the left ‘arm’ may be folded beneath or missing and the rachis has been
folded in compression. Eleven pinnae have rachises about 1:0 mm wide but the two on the right
are narrower. None of the specimens contributes to settling the controversial architecture of the
frond centre discussed by Hirmer (1927), Hirmer & Hoerhammer (1936), Lundblad (1950) and
Appert (1973) as well as others cited by them.

No Hillhouse block gives the full length of a pinna, though from the taper of the rachis I estimate
it about 20 cm long. The pinna rachis grades into a wing of lamina, making it hard to measure
its width accurately. Over most of a pinna the pinnules remain uniform but at the base they
become shorter and broader and finally merge as a continuous wing (Fig. 1). At the apex they
become narrower and are borne at a smaller angle, and if fertile they have few sori (Fig. 3). If
sterile they look like a Cladophlebis with very oblique veins (Fig. 4).

At Hillhouse Nab, where no large leaves are known, the larger pinnae have pinnules 18-25 mm

298 T. M. HARRIS

< 3-5 mm, the smaller pinnae having pinnules 11 x 2:5 mm. In the Gristhorpe Bed pinnules
larger than Hillhouse ones occur, along with small ones.

Adjacent pinnules of all specimens are connected by a web of lamina 0-5-1-0 mm wide though
this web has often suffered in preservation. The pinnule margins are in contact in some fronds
but separated by gaps up to 3mm wide in others. In fertile fronds and most sterile ones the
pinnule margins of the longer pinnules (Figs 7, 9) are parallel to near the apex but in a few sterile
fronds the pinnules taper from a broad base (Figs 6, 8); this is the typical “‘braunii’ form as
distinct from the ‘muensteri’ form. Intermediate forms are frequent. The classic Bavarian material
named Laccopteris braunii (Goeppert 1841) included a remarkable series of small leaves, L.
germinans, which are generally accepted as belonging to juvenile plants, but the Yorkshire
collections include no such specimen.

Sterile Hillhouse pinnules are flat, apart from the raised midrib. In transfers their margins are
dark and possibly fibrous. The fertile pinnules are nearly flat but the margins may be slightly
depressed. Above each sorus there is a bulge, broad in some pinnae, small in others and never very
pronounced. It may be so low that the presence of a sorus is uncertain until a transfer is made.
In the Gristhorpe matrix the sterile lamina is slightly convex between midrib and margins and
more strongly convex in fertile ones.

Figs 7-9 Matonia braunii (Goeppert) comb. nov. Hillhouse Nab, Farndale. All x 1. Fig. 7. Pinna with
short, densely fertile pinnules, V.59746. Fig. 8. Sterile pinna with broad-based pinnules, venation shown
in Fig. 6, p. 297. Part, V.59745. Fig. 9. Pinna with longer and sparsely fertile pinnules, V.59747.

Venation. In sterile pinnules and in sterile parts of fertile pinnules the veins characteristically fork
twice and their branches run obliquely, meeting the margin at 35°. Very occasional veins of large
pinnules anastomose, the union being near the midrib or further out and thus irregular (Fig. 5).
Many pinnules and all the smaller ones have no anastomosis at all. In the smallest pinnules and
in the apical parts of large ones the veins fork only once. In the basal flange along the rachis there
is a forward-running branch vein which joins a backward one from the next pinnule, and from the
combined vein further branches run to the sinus, though often this part is poorly preserved. All
vein branches end in the margin.

In fertile regions of the pinnules each lateral vein has a forward branch ending at the sorus and
an outward branch which forks, the branchlet nearest the rachis meeting the margin nearly at a
right angle (Fig. 18, p. 302). There are no anastomoses except in the basal web between pinnules.

No vein or rachis shows any hairs or scales, either directly or in transfer.

Sori. Although the upper surface of the lamina is the one exposed in all the Yorkshire specimens,
the coaly substance has sometimes crumbled away to reveal sporangia, though I did not recognize

MATONIA BRAUNII (GOEPPERT) 299

Fig. 10 Diagrams comparing the venation of sterile and partly fertile pinnules of Matonia braunii. Note
that both are often relatively longer. A higher proportion of sterile veins may anastomose. Fertile
pinnules are often more fully fertile, relatively narrower and have narrower webs along the rachis.

the indusium. In transfers of the more perfect fertile specimens the state of the sori varies, even
between adjacent sori. In its most perfect state, however, the sorus is entirely covered by the large
convex indusium (Fig. 12). Often this is partly broken in preparation and then the outer parts of
sporangia are visible; sometimes only the middle of the indusium remains (Fig. 11). More often
both indusium and sporangia have been lost, apparently before the preparation was made, and
nothing remains but the projecting placenta (Harris 1961: fig. 38D, named Matonidium); I
consider that this loss occurred before preparation because I believe I can distinguish between the
bright surface of coal fractured in preparation and the dull surface of coal preserved in contact
with mud. Not one Hillhouse specimen showed a sorus of ripe but undischarged sporangia
exposed by the loss of the industum, though sori in this state have been figured from Bavaria and
from Greenland.

Many Hillhouse sporangia were macerated and proved empty but one from an otherwise empty
sorus gave a coherent mass of spores (Fig. 14). The maceration failed to separate them but I

12

—— ——————— ee )
eee ye Ee Pon

Figs 11-12 Matonia braunii (Goeppert) comb. noy. Hillhouse Nab, Farndale. Transfers of fertile
pinnules, x 20. Fig. 11. V. 59750. The sori have crumbled in preparation, the upper one having lost
its indusium and the lower ones all but some of the middle portion. Fig. 12. V.59751. The indusia are
intact apart from cracks caused by shrinking. The small sori to the right are abortive and at the top left
there is a denuded placenta.

300 T. M. HARRIS

Figs 13-14 Matonia braunii (Goeppert) comb.
nov. Hillhouse Nab, Farndale. V.59752a.
Fig. 13. Spore at the edge of the mass shown
in Fig. 14, x 800. The shaded part is obscured
by overlapping spores. Fig. 14. Spore mass
isolated from a sporangium in situ, x 80.
Some of the spores are outlined more definitely
than is certain.

estimate their number at about 100. The spores have smooth brown walls about 1:5 ym thick
and the one most clearly seen (Fig. 13) is figured. They are about 50 ym wide. As the spores
are coherent and smooth-walled it is possible that they died before maturing and the fact that this
was the only sporangium of the sorus which yielded spores also implies arrested development.
Harris (1961) figured a similar spore from a Gristhorpe specimen and a smaller one from a
Saltwick specimen, both as Matonidium.

Discussion

The indusium. I determined the new specimens in the field as Ph/ebopteris braunii because they
agreed in every character visible with a lens. But when I made transfers and found that the sori

Figs 15-16 Matonia braunii (Goeppert) comb. nov. Hillhouse Nab, Farndale. Fig. 15. Partly fertile

pinna in transfer, V.59749, x 3. The placentae are small and possibly abortive. Fig. 16. Longest
available pinna fragment, sori obscure. V.59748, x 1.

MATONIA BRAUNII (GOEPPERT) 301

had indusia I decided this was impossible. Like others I was convinced that P. braunii had naked
sori, this being the character that distinguishes Phlebopteris from Matonia. There seemed a grave
practical difficulty. There were two groups of well-preserved specimens, P. braunii in the Lower
Lias of western Europe and East Greenland in which certain specimens, though a minority,
showed naked sori, and the new Yorkshire group in which certain specimens showed indusia
when transferred. But most specimens were indistinguishable and this applied also to nearly all
those figured previously except the ones showing details of sori. It seemed that the groups were
generically distinct but could seldom be distinguished and I tried to escape from the difficulty with
elaborate ae

Fig. 17 Two So eeuies of Matonia Bani V.21660, Seneee Sound, East Gesatand: x 10. Shrinkage
after collection caused cracks to form in the indusium between the sporangia and around the placenta.
The sporangia contain ripe spores but their annulus is not visible.

Later, when I saw Liassic specimens of P. braunii in the BM(NH) the difficulty vanished. P.
braunii also has an indusium but is apt to lose it. I suggest that collectors selected against specimens
with sori concealed by indusia, as I evidently did in Greenland.

For example, specimen 15043 from Bayreuth, Bavaria, labelled Laccopteris muensteri, shows
the under-surface of many fertile pinnules, some with intact sori, others denuded. In the intact
sori the sporangia form vague bulges but do not show their annulus cells and I conclude that this
is because they are covered by an indusium. The evidence is not fully convincing because it is
possible that this specimen had been coated with varnish which obliterated them, although I see
no sign of any such varnish.

Equally, specimen V.21660, collected by me from the Lower Lias of East Greenland and sent
by the Copenhagen Museum in exchange, has sori with an essentially smooth surface though the
coaly substance has shrunk since being collected and there is a circular crack around the placenta.
There are radial cracks between sporangia but nothing at all of their annulus cells is to be seen
(Fig. 17).

There is thus no difference between the indusiate sori of the Yorkshire specimens and the Lower
Liassic ones. There is, however, a difference in size between the Yorkshire and the Liassic leaves,
recognizable in the pinnules. Those from Hillhcuse Nab range from about 12 mm to 25 mm in
length and one pinna from the Gristhorpe bed has pinnules 35 mm long. Mature and fertile
Liassic leaves also range from about 12 mm upwards, but many specimens are 50 mm or more

302 T. M. HARRIS

long; the largest seen (Bavarian) was almost 100 mm x 5mm. AsI can find no other difference
I suppose that the Yorkshire environment, particularly at Hillhouse Nab, was unfavourable to the
production of very large leaves. The fact that scarcely any unopened sporangia are preserved there
suggests that the air may have been dry.

Contour of fertile pinnule. Hirmer & Hoerhammer (1936) illustrated the variation in the surface
contour of the fertile pinnule (in specimens named P. muensteri), and similar variation has been
seen in the Greenland and Yorkshire material. Their pl. 3, figs 3, 3a show a nearly flat upper
surface and their pl. 5, fig. 1 a nearly flat lower surface. On the other hand their pl. 3, fig. 6, of the
upper surface of another pinnule, shows high mounds above the sori and depressions near the
midrib and margins. Their pl. 4, fig. 6 of the under surface is the reverse; the midrib and margins
are strongly raised and they are connected by transverse ridges which mark out hollows where the
sori were situated. No Yorkshire specimen is as strongly contoured as this, though some approach
it.

18

20

Figs 18-20 Matonia braunii (Goeppert) comb. nov. Gristhorpe Bed, V.31982. Fig. 18. Part of a pinnule
of Fig. 20, x 10. The imprint is seen below and to the right, the coaly substance of the Jamina above
and to the left. Fig. 19. Diagrams representing the contour of the imprints of the pinna of Figs 18 and
20 and an inverted pinna on the same block. Horizontal scale of lower section = 10, of upper x 13,
vertical scale increased four times to compensate for matrix compression. Fig. 20. Pinna x 1. The sori
are shown as dots representing the placentae, but some in addition show the indusium or the sporangia.
A transfer of the counterpart, V.31982a, was figured by Harris (1961: figs 38A-D) as Matonidium.

The absolute difference in elevation differs in different matrices; in the Yorkshire specimens it is
lower for example in the Hillhouse siltstone than in the Gristhorpe claystone but in each the range
is of the same kind. The differences can be explained on Walton’s (1936) theory of the changes
which occur when a plant organ is compressed in a less compressible matrix, the final form being
determined by the surface which originally faced downwards. The original shape of a fertile M.
braunii pinnule can thus be deduced from the shape of its upper and lower surfaces. Original
differences in different pinnules may have occurred also, though it is unnecessary to invoke them
to explain what is seen in the fossils.

Specimen V.31982 from Gristhorpe is helpful in working out pinnule contour because it shows
three pinnae, two in normal orientation and one inverted, and in all three some of the coal remains
but most is missing and the imprint is exposed. By comparison with other blocks I infer that this
specimen has the same orientation as when it was part of the Gristhorpe Bed. The imprints and
remaining coal are illustrated (Figs 18-20) as seen under very oblique lighting. In describing the
imprints I reverse the words ‘high’ and ‘low’ as though describing the fossil that made the imprint.

MATONIA BRAUNII (GOEPPERT) 303

The upper surface (shown by the imprint of the inverted pinnules, Fig. 20 and upper part of 19)
has low relief, high regions being at most 100 nm above low ones. The midrib forms a narrow
ridge bounded by shallow grooves in the lamina and the margins are depressed. The sori form
scarcely perceptible bulges. The lateral veins are sometimes visible as very low ridges. Leaf
substance remaining as coal, the original under surface, has just the same relief though the midrib
is broader and the placentae and occasionally indusia can be seen (but are nearly flat).

The imprint of the lower surface (shown by the normally orientated pinnules, Fig. 18 and lower
part of 19) has relief of up to 200 pm. The broad midrib and the margins are prominent and the
sori form conspicuous hollows usually with a central boss, the placenta, sometimes still with the
indusium or part of it. One sorus shows the imprint of a full set of sporangia but no indusium.
Again the upper surface is mainly just the same but the midrib is narrow and the strong soral
bulges often have a dimple above the placenta. The lateral veins may be fairly conspicuous as
ridges on the lower surface but the branch to the sorus is usually concealed in the soral pit.

Walton’s (1936) theory includes the effects of compression of the matrix. This causes a general
and proportionate lowering of relief. Certain pinnules are distorted and tilted, and from their
apparent narrowing and the elevation of the raised side I estimate that this sample of
the Gristhorpe clay was compressed to about one quarter. I use this figure in reconstructing the
original shape of the pinnule (Fig. 19), but unfortunately no laterally compressed pinnule is
available to show its original thickness. I have merely made the lamina thick enough to allow for
the difference between the deep soral pit below and the low mound above.

The matrix of V.31982 is too coarse to show the imprint of epidermal cells like those illustrated
by Appert (1973). He comments (1973: 28) on the difference between his figure of collodion
imprints and the figures of Schenk (1867) which apparently represent macerated cuticles. My
own efforts to prepare a cuticle from Greenland and from Yorkshire ferns failed. I suggest that
Schenk figured a fragment of gymnosperm cuticle which adhered to his specimen.

Identity of Phlebopteris muensteri with P. braunii. It is convenient to consider this aspect of the
taxonomy before identifying the Yorkshire specimens with both. Hirmer (1927) united the two
taxa under Laccopteris elegans Presl, a species he later placed—with no good reason—in the
Marattiales. Hirmer & Hoerhammer (1936), however, separated them and set out the differences
in detail. They have been followed in this by most authors, though Krausel (1958) expressed
doubt. On the other hand some have firmly united them under Phlebopteris muensteri Schenk,
chiefly Harris (1931), Reymanowna (1963), Weber (1968) and Appert (1973). (Goeppert’s name
braunii has priority and possibly these authors mistakenly used muensteri because Schenk’s fertile
specimens were better).

The evidence for identity is simple; the two forms occur together in the field and they intergrade
morphologically. Thus they are recorded from the same localities in Bavaria (though I know no
details). In Greenland the material was only locally common but was found in five localities and
in each the two forms are closely associated (Harris 1931). The two forms also occur together
in Yorkshire, particularly at Hillhouse Nab and in the Gristhorpe Bed.

Leaves of the braunii form have in fact the appearance of shade leaves. They are always sterile
and have a broad, flat lamina showing the veins very clearly. Their pinnules may be triangular
and always taper from the base. Leaves of the typical muensteri form on the other hand may be
sterile or fertile. Their pinnae are long, relatively narrow and parallel-sided to near the apex. The
substance of the lamina is thick and the veins are hard to see until a transfer is prepared, when the
branching of the lateral veins is seen to be of braunii form in sterile pinnules but different in fertile
ones. The pinnule margins are often bent downwards in fertile leaves.

Hirmer & Hoerhammer stated additional differences regarded by them as of specific rank but
they do not even apply to all the specimens they themselves figured. The most impressive evidence
of identity of the two forms is probably from partly fertile specimens, where braunii venation
is seen in the sterile portions but muensteri where there are sori; see for example Harris (1926:
text-fig. 6 A, E as ‘Laccopteris groenlandica’ ; 1961: fig. 37E as Matonidium goeppertii) and Fig. 15,
p. 300 here.

304 T. M. HARRIS

Identification of Yorkshire specimens. There is no doubt that all the Hillhouse Nab specimens
belong to a single species. They are numerous and intergrade, the mean is the most frequent form
and the more complete specimens show much of the range of form in the one frond. No similar
species is associated with them.

The Gristhorpe Bed specimens are few, though even there a number of fragments may occur
on one block. A difficulty is that Phlebopteris polypodioides also occurs there and is more frequent,
but it is usually easy to distinguish. The specimens accepted as M. braunii are like those from
Hillhouse Nab but are better preserved and some are larger; one pinna has pinnules 35 mm x
3-5 mm (the largest at Hillhouse Nab was 25mm x 3:5 mm); if Lindley & Hutton’s figure
(1831-33: pl. 60) accurately represents M. braunii they may be at least 5 mm wide.

At Saltwick a few fragments only were obtained, from a place where it was difficult to collect.

As all of these Yorkshire specimens agree in fine details I identify them with one another
confidently.

Reference of Phlebopteris braunii to Matonia. The genera of Mesozoic Matoniaceae were formerly
distinguished on vague and undefined characters, for instance Matonidium Schenk on its resem-
blance in aspect to Matonia pectinata. We owe clear distinctions to Hirmer & Hoerhammer
(1936). The genera dealt with here differ in a single organ, the indusium. Phlebopteris had no
indusium at all, Matonidium had a placenta which expanded slightly and covered the bases of the
sporangia and living Matonia has the sorus entirely covered by the indusium. Since P. braunii
has a large indusium it must be removed from Phlebopteris and does not fit Matonidium, so the
only question is whether it should be placed in the living genus Matonia.

Only late in this study did I realize that Matonia pectinata, like P. braunii, may also show naked
sori through loss of the indusium. Seward (1899) figured such a sorus in detail whilst a photograph
in Hirmer & Hoerhammer (1936: text-fig. A) shows a herbarium specimen with some sori
indusiate and others naked.

I have examined herbarium specimens of M. pectinata in the BM (NH) and other herbaria.
Fronds with small sori which I presume are immature retain all their indusia but those with full-
sized sori nearly all show some sporangia exposed by the loss of the indusium. There is nothing to
show whether the indusium dropped off while the frond was alive, as apparently happened in the
specimen figured by Seward (1899), or only after it had been dried for the herbarium. But no
frond shows all its sori exposed as do certain Lower Liassic ones of M. braunii. The excellent
Liassic specimens of M. braunii showing all their sporangia exposed but still full of ripe spores
must be counted as biological failures since they died without releasing their spores, a fate which
happens to modern ferns living in continuously damp air. In this respect the Middle Jurassic
Hillhouse Nab specimens are more normal since almost every sorus has discharged all its spores.

I can find no account of the behaviour of Matonia sori as they ripen. However, I fortunately
consulted Mr A. C. Jermy (Department of Botany, BM (NH)) shortly before his recent visit to
Sarawak and he gave me the following interesting information about Matonia by letter.

I was able to find it on the mountain ridges of Gunong Mulu. It was growing at about a 5000 ft
[1500 m] altitude in those areas which had been bared by land-slips and also partly in glades in the
Lithospermum-—Quercus woodland growing on the crest of the ridges. I looked at many fronds to see if
the indusium was falling to expose the sporangium, but in no case did I find this was so. For the most
part I should add that the complete sorus was attached, although on many fronds the slightest touch
would remove them in toto, leaving the characteristic papilla-like receptacle.

I then collected a number of mature fronds and exposed them to the sun. I found that the indusium on
these shrivelled or partially folded back thus exposing the very short-stalked sporangia which very soon
dehisced.

I also found Phanerosorus sarmentosus on this trip. This is a fascinating fern, hanging in festoons against
limestone cliffs, with fronds up to 20 ft [6 m] or more. It behaved in the same way as Matonia except
here there were occasional sori which had lost their indusia.

The sori of the leaves described must have been fully ripe, riper than on most of the herbarium
fronds I have seen. There is thus a difference of behaviour between our fossil and the living
Matoniaceae, but perhaps a slight approach in Phanerosorus. Clearly in M. braunii, under some

MATONIA BRAUNII (GOEPPERT) 305

conditions in the Liassic at least, the ripe sori may have lost all their indusia before any sporangia
dehisced. The fronds were lost from the plant, transported by water and buried with intact
sporangia but I cannot tell what interval there may have been between loss of the indusia and of
the fronds. I can only imagine that continuous damp weather led to the failure of the sporangia
to dehisce, but clearly the difference between M. pectinata and M. braunii is real. This occurrence
of intact sporangia in the fossil Matoniaceae of the Lias is not unique, for in the Yorkshire
Oolite many leaves of Todites williamsonii and Klukia exilis have all their sporangia intact, but
containing apparently ripe spores.

M. braunii has much larger pinnules than M. pectinata and different venation; in particular its
sorus is supplied by a single vein, not several. However, similar differences have been accepted
in the fossil genus Phlebopteris (the sorus of P. dunkeri being supplied just as in M. pectinata) and
I treat these differences between the living and fossil Matonia species as specific rather than generic.
It is hard to assess the importance to attach to the different behaviour of the indusia of M.
braunii, which is partly demonstrable but partly hypothesis, but again I treat it as less than generic.
There is a major difference in habitat for it is impossible that fronds should be transported from
mountain tops to deltaic pools and buried intact or as large pieces in considerable numbers.
Clearly the present observations on the indusium of M. braunii call for fresh study of the sori of
other fossil Matoniaceae. This study would be best made by transfer preparations of fronds with
slightly immature sori facing downwards into the rock. No such study has been made but since in
P. polypodioides the exposed sorus shows nothing suggesting a central scar, I think there was
probably no indusium in this species.

I designate the species as Matonia braunii (Goeppert) n. comb. As lectotype I designate the
specimen named Laccopteris braunii by Goeppert (1841: pl. 5, fig. 1), refigured by Schenk (1867:
pl. 23, fig. 12) and by Hirmer & Hoerhammer (1936: pl. 2, fig. 1).

Matonia mesozoica Appert 1973, based on abundant, well-preserved material from the Upper
Jurassic of Madagascar, is distinguished from M. braunii by its short, broad-based and triangular
pinnules up to 9mm x 3mm. The sterile veins branch twice as in M. braunii. Anastomoses,
apart from one in the basal web, are rare.

Comparison of M. braunii with Yorkshire Matoniaceae. The recognition of Matonia braunii called
for reassessment of the Yorkshire Matoniaceae in general, and in particular those with indusia,
which had been referred to Matonidium on this one character. There are changes in three species
but these are slight and most of the specimens remain undisturbed: the species are Matonidium
goepperti, Phlebopteris woodwardii and P. polypodioides.

The following comparison consists of short statements of the main diagnostic characters of
these three species, drawn up as far as possible in the same manner, preceded by a statement of
M. braunii.

Matonia braunii (Goeppert, 1841), Yorkshire material. Pinnules 12-35 mm x 2-5-3-5 mm, rarely
5 mm, basal connecting web up to | mm broad. Substance of lamina dense, sori sunken on lower
surface but above forming ill-defined bulges. In basal web a vein anastomosing, other veins in
sterile pinnules typically forking twice, branches reaching margin at 40° and at a concentration
of 22 per cm, anastomosing only occasionally and then irregularly. In fertile parts of pinnules,
lateral vein giving a forward branch to sorus and an outward branch which forks, branchlet
nearest rachis almost perpendicular to margin. Indusium present, large but often lost and seldom
visible without transfer. Sori at least 1 mm apart.

Matonidium goeppertii (Ettingshausen) Schenk, 1871. Pinnules usually less than 10 mm long and
2 mm wide, web along rachis narrow and with no anastomosing vein. Substance of lamina dense.
Upper surface in sterile pinnule slightly convex, in fertile pinnule more convex and sori sometimes
forming hemispherical bulges in contact with one another; bulges extending from midrib to
margin. Veins once forked in most sterile pinnules, branches reaching margin at about 70° and at
up to 50 per cm. In fertile pinnules veins forked, forward branch ending in a sorus, outward

306 T. M. HARRIS

branch sometimes forked again near margin and proximal branchlet perpendicular to margin
(as in M. braunii), but successive sori at less than | mm apart.

In Yorkshire M. goeppertii is known from the Lower Deltaic only, where it is locally common,
but elsewhere it occurs in younger rocks, particularly the Lower Cretaceous. Nearly all the
Y orkshire leaves identified as Matonidium goeppertii in the BM (NH) agree fully with that species
and form a homogeneous group including several excellent specimens. But a few fragments of
M. braunii were determined as M. goeppertii and figured by Harris (1961) because they showed
points of interest; they are from the Gristhorpe Bed (Middle Deltaic) and Saltwick (1961:
figs 37A-C, E; figs 38A—G). There is now no detailed figure of the sorus of M. goeppertii from
Yorkshire, nor has its spore been figured from anywhere except at a low magnification. The sori
when exposed in transfer look just like the figures of M. goeppertii by Hirmer & Hoerhammer
(1936), but as happens in M. braunii some of them have lost their sporangia and the indusium too,
leaving merely a boss representing the placenta.

Phlebopteris polypodioides Brongniart, 1836. Pinnules commonly 40mm x 5mm, sometimes
larger, basal web often 2mm broad. Substance of lamina delicate and translucent. Sori not
sunken and not projecting above. In basal web an anastomosing vein; veins otherwise essentially
similar in sterile and fertile pinnules, laterals at once dividing into forward and backward branches
which anastomose to form regular and somewhat rounded broad arches near midrib; sorus
situated on primary arch. Branch veins arising from arch and reaching margin at an angle of
about 60° and at a concentration of 20 per cm, occasionally branching further and anastomosing.
Indusium probably absent.

The specimens now redetermined as M. braunii are a few collected from the Gristhorpe Bed in
about 1830 and described under various names (see pp. 307-8). They were referred to Phlebopteris
polypodioides by Seward (1900) and by Harris (1961) but did not fit well.

It should be noted that broad basal arches are not exclusive to P. polypodioides, for they occur
also in Piazopteris lorchi Appert (1973) where the pinnules, though much shorter, have very similar
venation.

Phlebopteris woodwardii Leckenby, 1864. (This species is almost always represented by fusainized
(charred) pinnule fragments). Pinnules often 5mm broad, sometimes broader; full length un-
known but probably relatively long. Substance of lamina originally dense, preserved as un-
compacted charcoal showing individual cells. (Pinna rachis and basal web scarcely known).
Veins (not known in basal web) forking and anastomosing to form primary arches which are
higher than broad, branch veins arising from arch, forking and sometimes anastomosing, reaching
margin at about 80° and a concentration of 50 per cm. Sorus situated in middle of primary vein
arch, borne on a forward branch vein. Sorus strongly raised and forming a sharply bounded
mound on upper surface, deeply sunken below. (Existence of indusium uncertain and details of
sporangia poorly known).

The recognition of Matonia braunii in Yorkshire has made it necessary to reconsider the
determination of certain unusual fragments determined as P. woodwardii which seemed to give
additional information about the sorus. These fragments, from Roseberry Topping and from
Hawsker, Jack Ass Trod are preserved as coaly compressions and showed Matonidium-like
sporangia and a small indusium. I still think them rightly determined but the information they
gave needs confirmation.

Apart from these compressed fragments, P. woodwardii is represented by widespread fusainized
pinnule fragments, many of them fertile but all having entirely denuded placentae. They are
occasionally abundant, along with fusainized wood. Hirmer & Hoerhammer (1936) removed the
species from Phlebopteris and placed it in Nathorstia of the Marattiaceae, but I do not think that
a sorus or synangium of this family would drop off from its placenta, even if scorched by fire.
No other Yorkshire leaf is mainly known as fusain.

It is to be noted that this species and P. dunkeri, also included by Hirmer & Hoerhammer in
Nathorstia, are known only as fragments, giving no evidence of the form of the whole leaf.

Distribution of M. braunii. Hirmer & Hoerhammer (1936) reduced a huge number of fossil species

MATONIA BRAUNII (GOEPPERT) 307

to six and for two of them, Phlebopteris braunii and P. muensteri, they gave references to specimens
spread over most of the northern hemisphere and ranging in age from the Trias to the Cretaceous.
They wrote with such confidence that a reader may think the determination of a matoniaceous
fossil from its figure is a clear-cut procedure to well-informed workers, as these two clearly were.
But unfortunately it is not. As evidence for this I mention their treatment of the earliest named
figure, Phillips’ (1829) Pecopteris caespitosa (1936: 7). It heads their list for Phlebopteris braunii
and also their list for Matonidium goeppertii, and as redrawn by Seward (1900) it is included in
Phlebopteris polypodioides. This taxonomic slip arises from real difficulty and had they given cf.
to each of these taxa, with cross references, I would have thought their treatment perfect.

Their long lists of citations under P. braunii and P. muensteri are valuable references to
specimens of similar aspect. These specimens and later ones can be graded according to the
published evidence. The most convincing is an assemblage of specimens showing matoniaceous
leaf form with similar venation in sterile and fertile pinnules and no discordant feature. Just three
regions produce such material: Bavaria (Franconia) and East Greenland, both in the lowest
part of the Lias, and Yorkshire in the Bajocian. If we accept separate pinnae of suitable character
we can add Bornholm, Middle or Upper Lias (Moller 1902), and Madagascar, Upper Jurassic
(Appert 1973). If we omit the fertile venation (which is often very obscure) thereis Poland, ?Upper
Lias (Reymanowna 1963), and other localities in the Lower Lias of Bavaria (Krausel 1958, Weber
1968). These specimens are excellent in other respects. Sikstel’ (1960) figures similar pinnae under
various names from central Asia (Upper Trias), but details are obscure in the published figures
as are those from the Lias of Romania (Semaka 1956). Hirmer & Hoerhammer’s citations of
unfigured specimens form in my view a category without evidence.

Finally there are some figured specimens cited by Hirmer & Hoerhammer which fail to show
the main diagnostic features and in my judgement also show a discrepant character. These include
all specimens cited from the Cretaceous and from the Trias older than the Rhaetic, and also
those from Asia and Africa.

Field key to the Yorkshire Matoniaceae

This key may fail with exceptional forms though normal forms are likely to occur with them.

1 Specimens (pinnule fragments) preserved as fibrous fusain . . : .Phlebopteris woodwardii
— Specimens preserved as a continuous film of coal. : 3 F ‘ F : : 2
2 Pinnules 10 mm x 2 mm or smaller ; ; : ‘ : F . Matonidium goeppertii
— Pinnules larger than 10 mm x 2 mm ; : : : ; : :
3 Pinnules 5 mm or more broad z : ‘ : : : ; : ; : : 4
— Pinnules less than 4 mm broad : : : : 3 , ; k : é 5 5
4 Veins forming broad arches along midrib 5 : § Phlebopteris polypodioides
— No arches along midrib, veins forming many small meshes : ; : Phlebopteris dunkeri
5 Lamina thick; in sterile pinnule vein branches at a small pat in fertile pinnule the branches

nearly perpendicular . ; ; : , . Matonia braunii
— Lamina thin, basal vein arches present : : j a small form of Phicbopteris polypodioides

Notes on fossil matoniaceous leaves described up to 1841

All the specimens cited below are from the Gristhorpe Bed apart from a few whose origin is
mentioned.

1828 ‘Polypodium’ Murray: 313; pl. 5, fig. 2. Fertile. Specimen not seen but figure looks like P.
polypodioides.

1829 Pecopteris caespitosa Phillips: pl. 8, fig. 10. Specimen in Yorkshire Museum. No description
and name therefore not valid. Refigured Phillips 1875: pl. 8, fig. 10 and with descriptive notes on p. 207

308 T. M. HARRIS

together with lign. 20, another specimen (from Haiburn). Original specimen redrawn by Seward (1900:
fig. 8) as Phlebopteris polypodioides; this determination accepted by Harris (1961). Hirmer & Hoerhammer
(1936) determined Phillips’ drawing as Phlebopteris braunii and also as Matonidium goeppertii (1936: 18).
Specimen in Phillips 1875: lign. 20 determined by Harris (1961) as M. goeppertii. Specimen of Phillips
1875: pl. 8, fig. 10 has its largest pinnules 17 mm x 3 mm, margins slightly depressed, web along pinna
rachis narrow, substance of lamina thick and opaque, low bulges at 1 mm intervals perhaps representing
denuded sori beneath; venation not observed (fine details damaged by varnish). Certainly not Matonidium
but agrees well with Matonia braunii. 1 now think it unlikely to be P. polypodioides because substance is
dense and pinnules are small. A few other Yorkshire Museum specimens determinable with confidence as
P. polypodioides are labelled P. caespitosa.

1829 Pecopteris crenifolia Phillips: pl. 8, fig. 11. Drawing shows basal region of specimen in Yorkshire
Museum. No description and name therefore not valid. Refigured by Phillips (1875), with description as
Phlebopteris crenifolia and P. propinqua Lindley & Hutton included as a synonym. Specimen well preserved
and normal P. polypodioides, pinnule margins mostly entire. The broad sori possibly give an illusion of
lobing, but scarcely as in the figure. In places there are real bulges in the margin up to 0-5 mm high and
these may be opposite sori or between them. At several points, however, the substance at the margins has
flaked off or is overlapped by matrix and this gives some appearance of lobing. In the Yorkshire Museum
there are also a few other specimens of P. polypodioides \abelled crenifolia Phillips.

1832 or 1833 Pecopteris polypodioides Brongniart: pl. 83, fig. 1. Name on plate but description not
published until 1836. Sterile pinna showing venation, specimen said to be in Paris.

1833 Pecopteris polypodioides Lindley & Hutton: 167; pl. 60 (nomen nudum of Brongniart 1828 mis-
applied). Specimen in Scarborough Museum but not seen by me; venation in figure agreeing with fertile
Matonia braunii; accuracy of drawing confirmed by Phillips (1875: 202). Figure regarded by Seward
(1900) and Harris (1961) as badly drawn P. polypodioides. Accepted as P. braunii by Hirmer &
Hoerhammer (1936). Now accepted by me also as M. braunii because I know that leaves with venation as
in the figure do occur in the Gristhorpe Bed. This confusing name has never come into use, see below.

1834 Pecopteris propinqua Lindley & Hutton: 101; pl. 119. A description and drawing by W.
Williamson jr of a specimen collected by W. Williamson sr. Lindley & Hutton never saw the specimen
and I have not recognized it in a museum. The figure strongly resembles that of P. crenifolia of Phillips,
i.e. P. polypodioides. Williamson, though, compares it with P. polypodioides of Lindley & Hutton, i.e.
Matonia braunii. The name propinqua never came into general use, see p. 309.

1835 Phlebopteris contigua Lindley & Hutton: 177; pl. 144. Specimen in good condition in
Scarborough Museum. This is the first appearance in print of the genus Phlebopteris but it was not defined
and therefore not valid. The text shows the authors had seen Brongniart’s plate of P. polypodioides but
supposed that their specimen, in fact typical P. polypodioides, was distinct. This name for Phlebopteris
polypodioides never came into general use and although strictly valid I hope for this reason it will never
be revived, see p. 309.

1836 Phlebopteris Brongniart: 371. Diagnosis and discussion of genus, first species described as P.
polypodioides. Species now placed in Dictyophyllum were included in the genus.

1836 Phlebopteris polypodioides Brongniart: 372 (diagnosis, corresponding to figure, pl. 83, fig. 1, la
already cited; see 1832 or 1833 above). Generally accepted as the holotype, though description is predated
by the scarcely used and less well defined names P. contigua and P. propinqua of Lindley & Hutton.

1836 Phlebopteris propinqua Brongniart: 373; pl. 132, fig. 1; pl. 133, fig. 2. Same figures as P. crenifolia
Phillips and P. propinqua Lindley & Hutton. P. crenifolia Phillips cited as synonym.

1836 Phlebopteris schouwii Brongniart: 374; pl. 132, figs 4, 4a, 5, 6. Sterile and fertile fragments from
Lias of Bornholm. Specimens considered by Moller (1902: pl. 46) as possibly P. woodwardii but otherwise
indeterminable; accepted by some authors as P. polypodioides.

1836 Polypodites crenifolius Goeppert: 343, for Pecopteris crenifolia Phillips and Pecopteris propinqua
Lindley & Hutton.

1836 Polypodites lindleyi Goeppert: 342; pl. 38, figs 1, 2, for Pecopteris polypodioides Lindley &
Hutton. Name /indleyi applied to a few specimens of P. polypodioides in the Yorkshire Museum, but later
dropped.

1838 Laccopteris elegans Sternberg: 115; pl. 32, figs 1-3, 8a-c. A block showing sterile and fertile
pinnule fragments from the Lower Lias of Bavaria. Specimen re-examined by Jung & Knobloch (1972)
who remained uncertain about its nature; see p. 309.

1841 Laccopteris braunii Goeppert, Lief 1-2 (2): 7; pl. 5, figs 1-7. Lower Lias, Bavaria. The specimen
in fig. 1 is considered here to be the Type.

1841 Laccopteris germinans Goeppert, Lief 1-2 (2): 9; pl. 6, figs 1-12. Lower Lias, Bavaria. (Widely
regarded as leaves of young plants of L. braunii).

MATONIA BRAUNII (GOEPPERT) 309

Summary

This work leaves the validity of three important specific names uncertain.

Is Laccopteris braunii Goeppert, 1841 identical with Laccopteris elegans Sternberg, 1838?
Many have thought so, for example Hirmer (1927), but later he changed his mind and with
Hoerhammer (1936) firmly separated Laccopteris elegans as a marattiaceous fern with round sori.
The matter can only be settled by a fresh study of Bavarian material. If for example specimens
matching the original figures were discovered which have unmistakable synangia I would be
convinced. At present I am not but leave the name braunii standing.

There are still earlier descriptions of braunii under other names. Of these, Phillips’ (1829)
Pecopteris caespitosa (see p. 307) is not valid because there is no description. The name Pecopteris
polypodioides of Lindley & Hutton, 1833, raises a difficult problem. The matter is complicated by
the almost simultaneous publication of the relevant works of Brongniart (1828-36) and Lindley
& Hutton (1831-37) in numerous parts. Both include descriptions of plants from the Gristhorpe
Bed. Another complication is caused by Brongniart’s publication of many nomina nuda and also
of named figures without descriptions, both being completed in later parts. Lindley & Hutton
deeply respected Brongniart and co-operated with him and treated his nomina nuda as valid names,
even when unsure how they applied.

The first matoniaceous fern of Brongniart, 1828, is his nomen nudum Pecopteris polypodioides.
The first description in valid form is of a fertile pinna showing the specific character of braunii
by Lindley & Hutton (1833: pl. 60). The description is good but unfortunately they misapplied
Brongniart’s nomen nudum to their specimen. The words ‘Pecopteris polypodioides’ head their
text but just beneath in the synonymy they give ‘?P. polypodioides Brongniart p. 57’. In the text
they say: ‘It is evident that our fossil is referable to Adolphe Brongniart’s genus Pecopteris;
but as the figures illustrative of that genus are not yet published, we have no means of knowing to
what species; we conjecture only that it must be his P. pol ypodioides, from the aptness of the name,
and from its having been procured by him from the Lower Oolite’. Soon after they wrote this,
Brongniart’s (1832 or 1833) named figure of P. polypodioides was published and their error was
plain. They had a new species. Can we accept Lindley & Hutton’s expressed doubt as making the
application of the name polypodioides to their specimen infirm and therefore not valid in nomen-
clature ? It would be greatly in the interests of palaeobotany to do so and this is the view I take.
So polypodioides L. & H. is not after all the first valid name of braunii. But some may disagree.

I turn to P. polypodioides of Brongniart. Its earliest named figure, Pecopteris crenulata Phillips,
1829, has no description. After Lindley & Hutton had seen Brongniart’s (1832 or 1833) figure they
received Williamson’s drawing which they published (1834: pl. 119) under the name Pecopteris
propinqua. They recognized that the veins were as in P. pol ypodioides of Brongniart, but considered
that the lobed margins of the pinnules distinguished it. I suppose that the specimen is like Phillips’
drawing of P. crenulata, a fertile pinna where the figure gives an exaggerated impression of lobes
and that it is merely a somewhat inaccurate drawing of polypodioides Brongniart, but asI have not
seen the specimen I cannot be as sure as I am about Phillips’ very similar figure. The next validly
published name is Pecopteris contigua Lindley & Hutton, 1835: pl. 144. The figure and the
specimen itself show the venation of a typical specimen of P. polypodioides but Lindley & Hutton
thought that its crowded pinnules distinguished it. All later authors have considered it well
within the form range of P. polypodioides and have dropped the name contigua, though I can see
no way in which this can be done in accordance with the Code.

Thus as I see it, the name braunii (or elegans) is saved, though by a narrow margin, from
becoming ‘polypodioides L. & H.’. But the valid name for Brongniart’s species is either propinqua
L. & H. or contigua L. & H.

I have never flouted the Botanical Code but may have breached it inadvertently. But now I
continue to use the name polypodioides. To change it wastes time and causes frustration. Palaeo-
botanists will be aware that there is nothing basic in taxonomy that makes this change of name
necessary. There are only the provisions of the Code and rulings of Botanical Congresses against
the conservation of any specific name, though they did agree to conserve generic names. Had these
fossils been animals there would have been little difficulty. The zoologists (including palae-

310 T. M. HARRIS

ontologists) can and do conserve old specific names. They could have surmounted the trouble
caused by polypodioides L. & H. in that way, but I suppose they could, as an alternative, have
deemed that Brongniart’s drawing was published slightly earlier (as it may have been) and accepted
it as a valid characterization of the species, cutting out the Lindley & Hutton figure and name and
making propinqua and contigua synonyms.

I state the facts as I see them and leave the name Phlebopteris polypodioides standing and merely
express the hope that before anyone changes it, the Botanical Code will be altered in a way which
will make this name regular.

These nomenclatorial difficulties began through ordinary taxonomic errors. I think they were at
first perpetuated through the early palaeobotanists having in their minds principles other than
that of priority of publication, or at least they did not give that overwhelming weight. To some,
admiration for Brongniart’s masterly work was a reason for preferring his names to ones
published a year or two earlier. Much later, Seward in his revisions of Mesozoic floras was un-
willing to put nomenclatorial priority over the advantages of letting a well-understood name
continue to stand; that is on his own judgement he conserved specific names. These names still
stand.

Acknowledgements

I take this opportunity to thank the staff of the Scarborough Museum and especially of the
Yorkshire Museum where I spent several days working on the classic material in their charge;
also Dr C. R. Hill of the Department of Palaeontology, British Museum (Natural History),
London.

References

Appert, O. 1973. Die Pteridophyten aus dem Oberen Jura des Manamana in Siidwest-Madagaskar.
Schweiz. palaont. Abh., Basel, 94 : 1-62, 90 pls.

Brongniart, A. 1828. Prodréme d’une Histoire des Végétaux fossiles. viii + 223 pp. Paris.

1828-38. Histoire des végétaux fossiles, ou recherches botaniques et géologiques sur les végétaux
renfermés dans les divers couches du globe. 1. xii + 488 pp, 171 pls. Paris. Parts cited (dates taken from
Zeiller 1903: 306) Liv. 7: 265-288, pls 83-97, 1832 or 1833; Liv. 11: 369-416, pls 135-146, 1836.

Goeppert, H. R. 1836. Die fossilen Farrnkréute (Systema filicum fossilium). Nova Acta physico-med.,
Halle a.S., 17 (Suppl.). xxxiii + 488 pp., 44 pls.

1841-46. Die Gattungen der fossilen Pflanzen vergleichen mit denen der Jetzwelt und durch Abbildungen
erlautert. 151 pp., 51 pls. Bonn.

Harris, T. M. 1926. The Rhaetic Flora of Scoresby Sound, East Greenland. Meddr Gronland, Copenhagen,
68 : 45-148, pls 1-13.

—— 1931. The Fossi] Flora of Scoresby Sound, East Greenland. 1. Cryptogams (exclusive of Lyco-
podiales). Meddr Gronland, Copenhagen, 85 (11) : 1-102, pls 1-18.

— 1961. The Yorkshire Jurassic Flora, 1. Thallophyta—Pteridophyta. ix + 212 pp. London, Brit. Mus.
(Nat. Hist.).

Hirmer, M. 1927. Handbuch der Paldobotanik. xvi + 708 pp., 817 figs. Munich & Berlin.

& Hoerhammer, L. 1936. Morphologie, Systematik und geographische Verbreitungen der fossilen
und rezenten Matoniaceen. Palaeontographica, Stuttgart, B 81 : 1-70, pls 1-10.

Jung, W. & Knobloch, E. 1972. Die ‘STERNBERG-Originale’ der Bayerischen Staatssammlung fir
Palaontologie und historische Geologie zu Miinchen. Mitt. bayer. St. Paldont. Hist. Geol., Munich,
12 : 105-111.

Krausel, R. 1958. Die Juraflora von Sassendorf bei Bamberg. Senckenberg. leth., Frankfurt a.M., 39, 1-2:
67-102, pls 3-8.

Leckenby, J. 1864. On the sandstones and shales of the Oolites of Scarborough, with descriptions of some
new species of fossil plants. Q. J/ geol. Soc. Lond. 20 : 74-82, pls 8-11.

Lindley, J. & Hutton, W. 1831-37. The fossil flora of Great Britain; or, figures and descriptions of the
vegetable remains found in a fossil state in this country. 1 (1831-33), li + 223 pp., pls 1-79. 2 (1833-35),
xxvil + 208 pp., pls 80-156. 3 (1835-37), 208 pp., pls 157-230. London.

MATONIA BRAUNII (GOEPPERT) 311

Lundblad, A. B. 1950. Studies in the Rhaeto-Liassic floras of Sweden. 1. Pteridophyta, Pteridospermae
and Cycadophyta from the mining district of N.W. Scania. K. svenska Vetensk Akad. Handl., Stockholm,
(4) 8 : 1-82, pls 1-13.

Moller, H. 1902. Bidrag till Bornholms fossila flora. Pteridofyter. Acta Uniy. lund. 38 2 (5) : 1-63, pls 1-6.

Murray, P. 1828. Account of a Deposite of Fossil Plants, discovered in the Coal Formation of the Third
Secondary Limestone, near Scarborough. Edinb. new phil. J. 5 : 311-317, pl. 5.

Phillips, J. 1829. I/lustrations of the Geology of Yorkshire, or a description of the strata and organic remains
of the Yorkshire coast (i.e. Part 1, 1st edn). xvi + 192 pp., 9 + 14 pls. York.

— 1875. Illustrations of the Geology of Yorkshire; or, a description of the strata and organic remains.
Part I, The Yorkshire Coast. (3rd edn, ed. Etheridge, R.) xii + 354 pp., 28 pls, map. London.

Reymanowna, M. 1963. 1. Review of investigations on Polish Jurassic floras. 2. The Jurassic flora from
Grojec near Cracow in Poland, Part I. Acta palaeobot. Cracov. 4 (2) : 1-48, pls 1-9.

Schenk, A. 1865-67. Die fossile Flora der Grenzschichten des Keupers und Lias Frankens. xxiv + 232 pp.,
Atlas 45 pls. Wiesbaden.

— 1871. Beitraige zur Flora der Vorwelt, IV. Die Flora der nordwestdeutschen Wealdenformation.
Palaeontographica, Stuttgart, 19 : 203-262, pls 22-43.

Semaka, A. 1956. Contributii Ja Flora Liasica de la Vulcan-Codlea. Nota II. Buletin sti. Acad. Republ.
pop. Rom. (ser. Geol. Geogr.), Bucharest, 1 (1-2) : 107-121, figs 1-30.

Seward, A. C. 1899. On the structure and affinities of Matonia pectinata R. Br. with notes on the geological
history of the Matonineae. Phil. Trans. R. Soc., London, B 191 : 171-209, pls 17-20.

— 1900. The Jurassic Flora. I, The Yorkshire Coast. Catalogue of the Mesozoic Plants in the Department
of Geology, British Museum (Natural History). 3, xii + 341 pp., 21 pls. London.

Sikstel’, T. A. 1960. [Stratigraphy of the continental deposits of the upper Permian and Triassic of
Central Asia.] Trudy tashkent. gosud. Univ. (n.s.) 176: 1-147, pls 1-19 (numbered as pages). [In
Russian].

Sternberg, C. von 1820-38. Versuch einer geognostisch-botanischen Darstellung der Flora der Vorwelt.
200 pp., 70 pls col. Leipzig & Prague.

Walton, J. 1936. On the factors which influence the external form of fossil plants; with descriptions of the
foliage of some species of the Palaeozoic Equisetalean genus Annularia Sternberg. Phil. Trans. R. Soc.,
London, B 226 : 219-237, pls 31, 32.

Weber, R. 1968. Die fossile Flora der Rhat-Lias-Ubergangsschichten von Bayreuth (Oberfranken) unter
besonderer Beriicksichtigung der Coenologie. Erlanger geol. Abh. 72 : 1-73, pls 1-17.

Zeiller, R. 1903. Flore fossile des Gites de Charbon du Tonkin. viii + 328 pp., 56 pls. Etud. Gites minér.
Fr., Paris.

Accepted for publication 6 November 1978

ee bd
a) a te |
Pilar lle } i tgs

‘ ; ive y Bard]

= a Fi co]
é \ y i a
| ; ue
f et ; ; - , , i H ‘aA oh ‘ ye
i. iy :
: : ne — . * ¥ Sh oi ot

7 ; re "|
ue a
7 . -s it iy » 7
{ " , : ’ on : 7 fix vo OB eee
j ;
2 ie mf ?
: i 7 i
z , : e I
he 7
7 SS } c iF ' tp Te
- e (ea At, 4 hl
— —— i
PEPER D> Ciyin | eg
~ 1 ; =
7 v*;
7 : :
» = le
a \_ = >
- y . Pay nt ae
P oe \
t “< “
L .—
» — i im
Ss : a
- Ti rv)
1 Zz
_ 7 e
i t ;
f *@
as

-

British Museum (Natural History)

ee ee

The Yorkshire Jurassic Flora by T. M. Harris
Hard bound in grey cloth covers, size 30:8 x 24-5 cm

Vol. I. Thallophyta — Pteridophyta
1961, v+212pp, 71 text figures £9.00

Vol. II. Caytoniales, Cycadales and Pteridosperms
1964, viiit+ 19lpp, 7 plates, 67 text figures £10.50

Vol. II. Bennettitales 5
1969, vit+ 186pp, 7 plates, 69 text figures £12.00

Vol. IV. Ginkgoales and Czekanowskiales
Contributions by W. Millington & J. Miller
1974, viii+ 150pp, 8 plates, 41 text figures £19.50

Vol. V. _Coniferales
1979, vit+ 166pp, 7 plates, 68 text figures £22.50

.
7
4
j
i

Complete in five volumes

The series, begun in 1961 (and still in print in its entirety), provides a definitive
statement on this extremely important group of fossil plants, which will remain a
standard work for many years. The main aim throughout has been the accurate
description of the fossils, their evolution and the light they throw on the origins
of modern plants. Not least is their contribution to the stratigraphical
palaeontology of the Jurassic, now assuming greater commercial importance in the
search for oil.

Titles to be published in Volume 33

An account of the Ordovician rocks of the Shelve Inlier in West
Salop and part of north Powys. By the late W. F. Whittard, F.R.S.
(Compiled by W. T. Dean)

Miscellanea

The Caradoc faunal associations of the area between Bala and
Dinas Mawddwy, north Wales. By M. G. Lockley

Fossil insects from the Bembridge Marls, Palaeogene of the Isle of
Wight, southern England. By E. A. Jarzembowski

The Yorkshire Jurassic fern Phlebopteris braunii (Goeppert) and its
reference to Matonia R. Br. By T. M. Harris

The entire Geology series is now available as a series

Printed by Henry Ling Ltd, Dorchester

'
|
|
4

Se ae a ee

say A ; =

330e% B40°%°
B25e°° Tass

GEOLOGICAL MAP OF THE
SHELVE INLIER, WEST SALOP
AND PART OF NORTH POWYS

PONTESFORD >

oO

305°%°

By the late W. F. Whittard, FRS,
compiled by W. T. Dean

Upper

"
Pir D__ Hoastow

ebteriey
To accompany Bull. Br. Mus. nat. Hist. (Geol.) 33 (1), 1979 ... i

© Trustees of the British Museum (Natural History) 1979

Marton
Pool

MARTONG

300°% =

“ GEOLOGICAL
BOUNDARY DIP

Yi 4 VERTICAL
pe) AUER STRATA

WHITTERY STAPELEY
| SHALE SHALE
J] MEMBER MEMBER
Boke WHITTERY STAPELEY 2
VOLCANIC VOLCANIC S
MEMBER MEMBER Be
HAGLE Fe
v¢
SHALE >6 HOPE 2
MEMBER Se MEMBER
oa Ww
HAGLEY zz zi
bal'S
VOLCANIC rye) MVTTON a
295°00 MEMBER MEMBER 2)
ALDRESS STIPERSTONES
MEMBER MEMBER
ROAD
[=] SPY WOOD SHINETON
MEMBER
GENERALIZED J sna
BOUNDARY OF
CURLER NAR Damn] RORRINGTON | =
DEPOSITS "| MEMBER LONGMYNDIAN |
o
A
22 =
ER 3
RIV! 8 COAL Be MEADOWTOWN| >> 3S
oR MEASURES {Swi MEMBER eye} URICONIAN w
SS ok we &
= 3 as
Q & eECN 95 PIcRITE | W
= a MEMBER =u > 3
& n 7)
A & ; DOLERITE >>
METRES sy PONTESFORD | OZ
g 83 WESTON
saute ete anoesie
5 <
ap5e00 330%" nang eS oe MEMBER

rae

whee

"oh

7

Bi
sg
e

Ati

wonton

Nees
# bay

em

7
oe

Shes
C4

i
het
ig,

2
fe Ee ete
f aie ete!
BERR aeneeete,

a ae span
SSS

: hain

3h}
fH Ph
id

Stok!