- W

_ (/> ■ — u> — cn

aVH9Jl LIBRARIES SMITHSONIAN INSTITUTION NOIiniliSNI NVINOSHimS SBIHVyan LIBRA
z: , w ^ ^ z. \ (n z

^ >Sk ? 5 V 1 < /<22^ - -

o X o ^ fM;-WA o X o

2 5 -X W 2 E

5, XjVA$V^^/ 1^ > '^ S >

c/> '>■• 2 w 2 W 2 t/5

ITUTION NOIifliliSNI NVINOSHiilNS S3IWVHan LIBRARIES SMITHSONIAN INSTITUTION NOIiOJ

to 5 \ ^ = w> ~ <n

^ ^ ^ w w ^^'5oT;5x 2;

;3 ;: h z W§m, h

w'W^^

O ■ ■ W _ o “ • o

-I 2 ' _l 2 _) 2

Hvaan libraries smithsonian institution NoiiniiiSNi nvinoshiiins saiavaan libr/>

2 f~ 2 (" 2

m 2 rrs ^ rn 'll m

^ (/> — t/> _ (/)\5 ^

ttution NoiiniiisNi NviNOSHims saiMvaan libraries smithsonian institution Noiin

CO 2 W 2 .., . CO 2 ,V,,

> "W*- S > '>?r” s '- > Xj2iL!55>' S >

' 2 «0 2 00 *2 </) 2

avaan libraries smithsonian institution NoiinmsNi NviNosHims saiavaa n_LiBR/

CO 2 ^ ^ c> — (/)

U) A '~ ‘ ‘ ' /!5?^S^2nx CO

O qc^ _ NtOiii.‘it*X O " ''<iyu522 o

2 _J Z _J 2 ‘ _l 2

‘ITUTION NOIiniliSNI NVINOSHIIINS SHIBVaail LIBRARIES SMITHSONIAN INSTITUTION NOliO

2 ^ H ^ ■" ^ !Z ^

^ Z /AovoV O m xfSvoX o

33
-

rn <' V?^' ^ XjvAiq^ no XivwiiiZ n m ’'

CO ■■■ _ CO ± CO “

avaan libraries smithsonian institution NoiiniiiSNi nvinoshiiins saiavaan libr/

2 > w ^ 2 > CO 2 ^ ^ 5 ■■•

2 H 2 ^ 2 -C 2:

i g 1

2 <0 ■■*‘" 2 CO 2 CO

ITUTION Noii iliSNi NviNosHii^vs S3iavaan libraries smithsonian institution NOIin.

CO 5 \ ^ ^ ^ ^ ^

LiJ /v rn ^ y*

5 ® 5 'g

2 _l 2 _J 2

avaan libraries smithsonian institution NoiiniiiSNi nvinoshiiins saijavaan LiBR<f

2 r- 2 r- _ 2

- fT, ^ xcoii^ rn

CO — CO _ coX— ^

riTUTlON NOIiniliSNI NVINOSHillNS S3iaVaail libraries smithsonian INSTITUTl0N,„N0Iin

... CO 2 CO 2 ..- w 2 W

" .y /SSfC’m 3 .Vi./ A s 3 ^ 3 /.-y

c/) ^ (/> E CO _

SMITHSONIAN INSTITUTION NOliniliSN! NVINOSHilWS S3iyVHan LIBRARIES SMITHSONIAN INS

^ Z > t/) z c/> z

2 ^ <

o ± d I o

^ O >^V ■ X

t :z,/^//-'^' t

CO

__ __ ^ . >

</) •'* Z C/> Z CO '^' Z in

NviNOSHilws S3 1 a vy an libraries Smithsonian institution NoiiniiiSNi nvinoshiiws s3

^ X CO — CO 5 \

^ m y^Q A/7Nw CU >45^33^. #« LJ ✓<d)'CIT(/^x rn txi

O — O ^xI^osvA}j/ _

-4 Z _j z _

SMITHSONIAN INSTITUTION NOIlfUliSNI NVINOSHilWS S3iaVMan LIBRARIES SMITHSONIAN INS

z ^ ^

o ~ 2 o

m '<ayo(^ w m 2 m xi^jvoo^ m

^ . CO V. ~ CO ^ i(/)

NVmOSHillNS S3 1 ava an LIB R ARIES, SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHilWS S3

i" X (O Z V- to 2- to

z "/J ^ ^ ^ ^ ^

X . ./yj^ O .-S^Vx Z /PA*”'S«VC^\ o X o

>. iw i ^ i' I

^ ^ C/> 2?

SMITHSONlAN_INSTITUTION NOIlDlllSNI NVINOSHIIIAJS S3iavaan LIBRARIES SMITHSONIAN IN

^ ~ CO X to ::^

5 D

CO

o --iiXDx/ _ o ''' IS X^Omsgx o Xo<v~D(^ " O

NVINOSHillNS S3 I aVB a IT LI BRAR I ES^ SMITHS0NIAN~^INST1TUTI0N ^ N0liniliSNI~'NVIN0SHillMS^S3

>2; r- z r- X r~vX

* —

m Xivosv^ ^ m XjNost^ ^ xjvAi,>^ rn "S-vW

.. CO Z to 2 CO _

SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHIUNS S3iavaan LIBRARIES SMITHSONIAN IN!
^ z » to z to z , to X

^ t i , w t z /^^^ t i

5 ' > 2 ,

CO ••'“ Z CO Z CO Z Jn

NviNOSHims S3 lava an libraries Smithsonian institution NoiiniiiSNi nvinoshiiws S3

to — to — to — •, to

< (ig'"'3lsl 3 IW^i 5 ,^^5 0^'"'S4 =5 5

O N^ostC^X __ O X^AJOSW^^ XjVAiV;^?^ O

-j X _j 2 — I Z _

SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHIIWS S3iavaan LIBRARIES SMITHSONIAN IN'
” “ c X ^ ^ z >“

,x'^vA<VA*'\. aSSsxsI >''^vj^n*7Nv

DtJ
:o
>

rn w Jo m xo^Dcjx ^ m

^ — c/>*. ? t/> — C/J

NviNOSHiiws S3iavaan libraries Smithsonian institution NoiiniiisNi nvinoshiiins S3

2. . C/> ^ ... CO “

, < ^

v'X

-H

CO

&x.

VOLUME 9 • PART 3

Palaeontology

OCTOBER 1966

PUBLISHED BY THE
PALAEONTOLOGICAL ASSOCIATION
LONDON

Price £3

THE PALAEONTOLOGICAL ASSOCIATION

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

Institute membership £7. Oj. (U.S. $20.00)

Ordinary membership £3. 3j. (U.S. S9.50)

Student membership £2. 2s. (U.S. $6.50)

There is no admission fee. Student members will be regarded as persons receiving
full-time instruction at educational institutions recognized by the Council; on first
applying for membership, they should obtain an application form from the Secretary
or the Treasurer. Subscriptions are due each January, and should be sent to the
Treasurer, Dr. C. Downie, Department of Geology, The University, Mappin Street,
Sheffield 1, England.

Palaeontology is devoted to the publication of papers (preferably illustrated) on all
aspects of palaeontology and stratigraphical palaeontology. Four parts are published
each year and are sent free to all members of the Association. Members who join for
1966 will receive Volume 9, Parts 1 to 4.

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

Vol. 1 (for 1957-8) in 4 parts at £3 or U.S. $9.00 per part.

Vol. 2 (for 1959) in 2 parts at £3 or U.S. $9.00 per part.

Vol. 3 (for 1960) in 4 parts at £3 or U.S. $9.00 per part.

Vol. 4 (for 1961) in 4 parts at £3 or U.S. $9.00 per part.

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

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

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

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

A complete set. Volumes 1-8, consists of 30 parts and costs £90 or U.S. $270.

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

© The Palaeontological Association, 1966

PREDATION AND SHELL DAMAGE IN A
VISEAN BRACHIOPOD FAUNA

by HOWARD BRUNTON

Abstract. A description is given of a variety of borings and signs of predation which occurred either during the
life or after the death of some Visean brachiopods. These structures, thought to have been produced by sponges,
bryozoans, gastropods, and fish are of interest as indicators of environment and of the type of death assemblage
met within these rocks.

A RICH silicified brachiopod fauna obtained from Upper Visean rocks in Co. Fer-
managh, Northern Ireland, by the process of acid etching shows signs of predation or of
having provided a habitat for organisms that required a hard surface upon or within
which to live. The etched limestones were from a predominantly argillaceous series
immediately underlying the widespread and prominent Dartry Limestone group.
Within this fauna it is possible to distinguish damage indicating activity upon the living
shells, and that reflecting settlement of organisms after death, in which only the shelly
material was aflFected. It is assumed that any organism that lived wholly on the inner sur-
faces of valves, or which covered areas such as articulatory processes, or extended from
the inner to the outer surfaces via the posterior margins must have colonized the valve
after the death of the brachiopod.

Damage to living shells. One of the commonest signs of damage to brachiopod shells con-
sists of neatly bored circular holes, usually penetrating the full thickness of the valve
(PI. 60, fig. 1). These holes vary from OT mm. to just over 1 mm. in diameter and occur
in as many as 50% of the collected pedicle valves of Productina margaritacea. However,
the proportion is usually about 30% of pedicle valves and 15%, or less, of brachial valves,
e.g. in Rugosochonetes cf. celticus about 35% of the pedicle valves and 7% of the brachial
valves are bored. This higher occurrence of bored pedicle valves is probably a reflection
of their accessibility to predators on or within the substratum, and of their immobility
as compared to the free brachial valve. A surprising thing is that large valves may have
three or four penetrating holes. It can only be assumed that the predators were only able
to feed from the tissue close to the hole and bored in several places to expose more of the
soft parts. The holes are commonly on the postero-median region of pedicle valves
within about 6 or 7 mm. of the ventral beak, but not invariably so, and there seems to be
no area in which they are particularly concentrated (text-fig. 1). Two borings out of
forty-six into the pedicle valves of P. margaritacea were unsuccessful in that they do not
penetrate the complete shell thickness. However, it is possible that these are the rare in-
stances when predation was not successful in killing the brachiopod. This might have
been possible if the boring took place just behind the shell edge and if the predator was
unable to remove much of the internal mantle lobes. If this were so, the mantle could
probably have repaired itself and subsequent deposition of secondary shell would have
sealed the hole internally. From other evidence of damage to shells, it does seem that the

[Palaeontology, Vol. 9, Part 3, 1966 pp. 355-9, pi. 60.)

C 4179 B b

356

PALAEONTOLOGY, VOLUME 9

mantle was capable of repair allowing normal anterior growth to proceed and only
leaving a temporary scar at the surface of the shell.

It is most probable that the predators involved were carnivorous gastropods living
within the surface layer of the substratum. Modern gastropods, such as Nucella, bore
similar holes, of about 1 mm. diameter, into lamellibranch shells. It is commonly noted
that one valve of the lamellibranch is bored in preference to the other and this probably
resulted from a preferred orientation of the shells, allowing greater accessibility to one
valve. A comparable preference was shown by the Carboniferous predators for the
pedicle valve of brachiopods. Fretter and Graham (1962) have pointed out that modern
Natica only bores into shells when it is buried in sand. It is likely that the pedicle

TEXT-FIG. 1. Diagram of the pedicle valve exterior of Productina margaritacea (Phillips)
showing the positions and approximate sizes of 44 borings through the valves.

valves of the Fermanagh brachiopods were partially buried in a soft substratum and that
gastropods, occupying a similar niche to Natica, bored these areas of accessible shell.
A boring of particular interest, seen on a German Permian strophalosiid, is one in which
the boring did not penetrate the shell and which has a small boss within the circular
cavity. Natica is known to bore upon its prey by rasping away the shell at the edges of
the hole, so leaving a central boss similar to that seen on the Permian shell. Within the
etched faunas there is a paucity of gastropod remains, and only a few specimens of
Platyceras, Euomphahis, Bellerophon, and Subulitidae are to be found in residues from
near Derrygonnelly, Co. Fermanagh.

Modern boring naticids and muricacids have been traced back to the Triassic and
Cretaceous periods respectively, and no Palaeozoic genera are known to have been
borers. However, it is likely that this style of feeding was employed by gastropods, and
Fenton and Fenton (1931) have discussed such borings found in American Palaeozoic
brachiopods. They distinguish borings with bevelled edges, such as described in Tertiary
lamellibranchs by Hayasaka (1933), from those with parallel sides, as described here.
Fenton and Fenton suggested that Platyceras may have been related to naticids, and
therefore a borer, but this concept is no longer commonly accepted.

On the external surfaces of some shells are slightly twisting open canal-like borings.
In some specimens these extend right through the shell substance, in others they appear
as simple pits and holes. In some specimens of Eomarginifera these borings more or less

H. BRUNTON: PREDATION IN A VISEAN BRACHIOPOD FAUNA 357

follow the radial ribbing and measure between 0-8 and 1 -5 mm. wide (PL 60, fig. 2). Their
alignment with the ribs may result from the shell substance being thinner in these regions
because of the supposed underlying mantle canals, which may have provided nutrients
to the predator. There is no clear evidence as to whether these borings were made
during the brachiopod’s life or after its death. But as the borings invariably seem to enter
the shell from its external surfaces, and tend to be concentrated around the ventral
umbo, it is more likely that they were made while the shell was alive. Had the borings
been made purely as a protective habitat it is probable that the thicker parts of the shell,
between the ribs, would have been excavated. The borings are quite unlike the neat holes
described above or the ramifications of bryozoans described below, and are thought to
have been made by sponges.

Distortion of the normal shell growth is not uncommonly seen, either locally confined
to a small area, or more extensive and with scar shell material dwindling anteriorly into
normal growth. Sarycheva (1949) described forms of damage to productoid brachiopods
and distinguished several morphological types based upon the extent and position of the
damage. She suggested that much of the damage resulted from mechanical action, such
as waves and moving stones, but that the jaw action of predators (p. 288) may have
caused the more sharply localized damage. Sarycheva also described the irregular in-
ternal morphology, such as muscle scars, seen on some specimens. I believe that preda-
tion and irregular growth resulting from the close proximity of a foreign object account
for most of the distorted shells seen in the Fermanagh fauna.

Irregularities in growth commonly developed from an indentation in the shell, as if
locally crushed, or as a cleft extending to the valve margin. In the former instance the
point of crushing may be confined to one valve or involve both valves, as can be seen in
a specimen of Delepinea destinezi (PI. 60, figs. 6, 7) from limestones of Lower Yisean age
bordering Lough Erne. In this example a series of points have scarred and crushed part
of the dorsal valve, producing a similarly shaped series of protuberances from the sur-
face of the pedicle valve. The damage must have been considerable and broken right
through the shell from the dorsal surface. Secondary shell occurs anterior to the damaged
areas, so that the shell must have survived and continued its growth. In the scar area, and
beyond, the costellation is irregular and remains distorted for a distance of about 20 mm.
It seems likely that death would have resulted had this damage been infiicted to the
visceral region, but that the mantle epithelium of the brachial cavity was capable of re-
formation and continued growth. Growth-lines radiate antero-medianly from the
lateral edges of the scars, showing the way in which growth must have continued sealing
around the front of the damaged areas, and indicating that the mantle edge was probably
damaged. A crest was formed where these two growing edges reunited in front of the
broken shell. Signs of the mantle edge having been damaged are to be seen in some
specimens of Krotovia spimdosa and Eomarginifera in which a continuous groove de-
veloped up to the anterior margin (PI. 60, fig. 5). Here shell substance was only de-
posited from the sides and normal growth was not resumed. One specimen of Eomargini-
fera (PL 60, fig. 4) appears to have had a portion of its pedicle valve removed without
having irrevocably damaged the mantle. Here shell growth was continued level with the
inner surface of the existing shell, but was not only secondary shell, as the external orna-
ment was soon redeveloped and the shell levels became equal towards the anterior
margin. It seems, therefore, that even after having had part of its shell ripped away, the

358

PALAEONTOLOGY, VOLUME 9

mantle and mantle edge were still capable of shell deposition. The continuation or re-
development of the normal surface ornamentation is indicative of normal primary shell
formation, at the mantle edge, while featureless scar material was probably the product
of the mantle surface along with the normal secondary shell substance.

So far as is known, the only possible predators large enough to inflict damage on the
scale described above would have been fishes, probably members of the cladodont
sharks. The deep impression left in the brachial valve of Delepinea, described above,
matches the shape that is typical for a cladodont tooth. The impression lacks the mark of
the large central cusp because of the deposition of secondary shell in that region, but the
smaller marginal cusps have all left their marks. The teeth of these vertebrates are to be
found occasionally in nearby sediments.

Damage to Dead Shells. One of the commonest forms of damage to shells is that of in-
festation, most likely by ctenostomatous bryozoans, belonging to the subphylum
Ectoprocta. These stoloniferous, zooid-bearing organisms usually appear on the inner
surfaces of the valves as thin crack-like tubes of about 0-04 mm. diameter. Older regions of
the colony increase in size to about OT mm. diameter and are covered by minute pores,
but a gradation between the two sizes is rare. The stolons may be on the valve surface or
more or less sunk into the shell. Rarely they disappear completely into the shell and
may penetrate to the outer surface. Branching is common and the stolons may cross one
another. Colonization of the valves commonly took place in a more or less radial fashion
from one or more centres of growth, which became enlarged and commonly deeply sunk
into the shell. The surfaces of the stolons of these older regions are closely covered by
minute pores, giving a granular appearance, as if the zooids were crowded over the entire
surface rather than being arranged in single rows, as along the thin newly formed stolons.

Of the described ctenostomatous genera (Bassler 1953), the nearest in morphology is
Vinella Ulrich 1890, although the Fermanagh ramifications have more than a ‘single
row of small pores’ (1953, G35) on many of the stolons. Further study of these organ-
isms may reveal that they should not be referred to Vinella and that the pores themselves
housed polypides rather than being the scars from where the zooecia have broken.

EXPLANATION OF PLATE 60

All specimens are silicified and collected from Co. Fermanagh.

Fig. 1. Krotovia sp., showing six holes bored through the pedicle valve. BB52878, X4T.

Fig. 2. Eowargiuifera sp., showing borings on the visceral region of the pedicle valve. BB52893, x4 1.

Fig. 3. Plicatifera plicatilis (J. de C. Sowerby), damaged shell showing irregularly developed plication.
BB52924, x3-8.

Fig. 4. Eoinarginifera sp., damaged and bored pedicle valve. BB52925, X 3-7.

Fig. 5. Krotovia spinulosa (J. Sowerby), damaged pedicle valve. BB52923, x4T.

Figs. 6, 7. Delepinea destinezi (Vaughan), fragment of shell showing the damage believed to have been
caused by a cladodont shark; viewed ventrally and dorsally. BB52927, X 1-2.

Fig. 8. Schizophoria sp., fragment of pedicle valve showing worm tubes that developed over the dental
plates. BB54634, x4T.

Fig. 9. Portion of the ventral interior of Schizophoria showing ctenostome infestation. BB52931, X 10-2.

Fig. 10. Eoinarginifera sp., portion of the upper, posteriorly directed, portion of the pedicle valve trail
showing ctenostome infestation. BB52930, X 8 -6.

Fig. 11. Schizophoria sp., portion of the ventral interior showing a young ctenostome colony. BB52932,
Xl7.

Palaeontology, Vol. 9

PLATE 60

BRUNTON, Predation and damage in Visean brachiopods

H. BRUNTON: PREDATION IN A VISEAN BRACHIOPOD FAUNA

359

As colonization of the shells always started on their inner surfaces and because the
stolons extended over articulatory surfaces, it is clear that infestation took place after
the death of the brachiopod. It is possible, therefore, to distinguish some of the shells
that died and lay upon the sea floor prior to burial. An investigation into the sizes of
shells and valves colonized within each species shows that the ctenostomes are commonly
restricted to the larger specimens of the Fermanagh faunas. Juvenile and immature
specimens of most species are rarely colonized and this suggests that only adult shells
were being added to the death assemblage. Two species, a Schizophoria and Rugoso-
chonetes, do provide young valves, down to about 3 mm. long, which are colonized
by the bryozoans. These must have had a higher infant mortality rate than other species
in the fauna and this may indicate that the habitat was not the optimum for these species.

The presence of these colonizing epifaunas can be taken as evidence of a period in
which the skeletal parts lay unburied upon the sea floor. Thus, it cannot be assumed that
such a community was killed by the sudden advent of sediment and consequent burial.

Valves in which the stoloniferous Ctenostomata are developed commonly have super-
ficial irregularly winding and slightly coiled tubes of OT to 0-25 mm. diameter. These
show no external ornament or structure, but appear to incorporate fine sediment and
are probably the remains of tubiferous marine worms similar to the polychaetes Serpula
and Spirobis. It is quite clear that these worm tubes were late in development as they are
adherent to the free inner surfaces of shells, and commonly cover bryozoan stolons. Like
the stolons, they may partially fill the sockets or cover the teeth of Schizophoria or
Rhipidomella (PI. 60, fig. 8). The tubes were probably calcareous, with a little adherent
fine sediment, and the silicification has bonded the two together.

It may be significant that brachiopods seen in the older black bituminous and pyritic
limestones of Lower Yisean age, or those from the overlying reefal limestones of
Knockmore, miles south-south-west of Bunnahone, do not appear to have been sub-
jected to attack by boring organisms other than gastropods.

It is clear that the brachiopods of the communities studied were not only prey to car-
nivorous animals, but provided a habitat for other organisms after their death, when
their shells were added to the skeletal debris on the sea floor. Such debris was in turn of
value to larval brachiopods of subsequent generations when they settled to the sea floor
and became attached by their pedicles, cementation, or clasping spines.

REFERENCES

BASSLER, R. s. 1953. Treatise on Invertebrate Paleontology. Pt. G, Bryozoa, i-xiii, G1-G253. Lawrence,

Kansas.

FENTON, c. L. and FENTON, M. A. 1931. Some snail borings of Paleozoic age. Am. Midi. Not. 12, 522-8.
FRETTER, V. and GRAHAM, A. 1962. British Prosobranch Molluscs, their functional anatomy and ecology.

Ray Society, London.

HAYASAKA, I. 1933. Fossil occurrcncc of pelecypod shells bored by certain gastropods. Mem. Fac. Sci.

Taihoku imp. Univ. 65-70, pi. 1-4.

SARYCHEVA, T. G. 1949. The study of damage to Carboniferous productoid shells. Trudv Paleont. Inst.

280-92, pi. 1, 2. [In Russian.]

HOWARD BRUNTON

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

London, S.W. 7

Manuscript received 14 April 1965

FOSSIL WOOD OF ANACARDIACEAE FROM
THE BRITISH EOCENE

by DONALD W. BRETT

Abstract. Silicified wood from the sands of the Woolwich Series (Landenian) at Herne Bay, Kent, is described
as Edenoxylon aemidum sp. nov. (Anacardiaceae). The specimen closely resembles wood described from the
Lower Eocene Green River Formation of Wyoming, U.S.A., as Edenoxylon parviareolatiim Kruse. The character
of the wood is essentially that of a tropical type with no strongly marked seasonal increments in its radial growth.

The fossil wood described in this paper is closely similar to the wood of Edenoxylon
parviareolatiim Kruse (1954) which is founded on small stems and roots from the Lower
Eocene of Wyoming, U.S.A. The following diagnoses are based on the holotype (stem)
of E. parviareolatiim Kruse, which I have examined, and the new British material which
is named E. aemulwn sp. nov.

Family anacardiaceae
Organ-genus edenoxylon Kruse

Diagnosis. Fossil secondary wood, or stems or roots with some secondary wood. Rays
of two kinds: narrow 1-2 seriate, and wider fusiform rays containing ducts. Vessels
solitary and in radial multiples (2-9), about 50 per sq. mm; diameter 20-140 p', inter-
vascular pitting typically alternate but opposite in places, minute (about 5 p) ; perfora-
tions simple, more or less oblique. Wood parenchyma scanty paratracheal. Pitting be-
tween vessels and ray cells and between vessels and parenchyma irregularly elliptical,
about twice as long as broad, or sub-circular. Remainder of wood tissue narrow fibres,
mostly septate.

Type species. Edenoxylon parviareolatiim Kruse.

Description of type specimen. Stem about 3 cm. wide with medulla, secondary xylem, and phloem.
Vessels about 50 per sq. mm; 20-70 p diameter. Narrow rays mostly uniseriate, rarely biseriate in
part; 2-20 cells high; about 20 per mm. Secondary phloem containing vertical ducts mainly in concen-
tric rows, fibres, and stone cells. (PI. 61, figs. 1, 2.)

Holotype. B-3280 in Palaeobotanical Collection, University of Cincinnati, Ohio, U.S.A.

Horizon. Lower Eocene of Eden Valley, Wyoming, U.S.A.

Edenoxylon aemidum sp. nov.

Plate 61, figs. 3-7

Diagnosis. Secondary xylem. Vessels about 44 per sq. mm; 30-140 p diameter. Narrow
rays almost all uniseriate nearer centre of axis; further out almost all biseriate and 5-9
cells high, or higher when including uniseriate parts. Rays average about 8 per mm.

Holotype. V44297 in Department of Palaeontology, British Museum (Nat. Hist.), London.

Horizon. Lower Eocene, Landenian; Herne Bay, Kent, England.

(Palaeontology, Vol. 9, Part 3, 1966, pp. 360-4, pi. 61.]

D. W. BRETT; FOSSIL WOOD OF ANACARDIACEAE

361

Description. The material consists of several fragments taken from a large piece of silici-
fied wood lying on the beach at Bishopstone, Herne Bay, Kent. Borings through the
wood made by a species of Teredo are infilled with a greyish sand quite unlike that of
the existing beach but matched in the lower beds of the Woolwich Series which overlie
the Thanet Sands in the cliflTs at Herne Bay. It is deduced from this that the fossil derives
from the Woolwich beds.

The wood is diffuse porous. There is an indistinct and irregular periodicity evident in
the dimensions of the fibres and vessel chains; and some broader rings have a definite
late-wood with smaller and fewer vessels. In transverse section the vessels are solitary
and in radial multiples and chains of 2-6 (exceptionally 8); solitary vessels and groups
together average 44 per sq. mm. ; tangential diameter averages 95 p but the range is wide,
30-140 p. Perforations are simple, more or less oblique; intervascular pitting is alternate,
about 5 p. Pits to the ray cells are irregularly elliptical, about twice as long as broad,
commonly obliquely inclined or horizontal, or sub-circular and little wider than the
intervascular pitting. Pitting between vessels and wood parenchyma is similar to the ray
pitting but more commonly horizontal and elliptical (i.e. scalariform). Tyloses fill all the
vessels. The fibres are narrow, mostly less than 15 p wide, thin walled, septate and with
dark contents, in strict radial rows forming a uniform tissue. Pits not seen. Wood
parenchyma is scanty paratracheal. The rays are of two types, low narrow rays and
larger fusiform rays containing radial ducts, together about 8 per mm. (range 6-12).
Nearer the centre of the axis the narrow rays are almost uniseriate, 3-16 cells high, and
the fusiform rays are only a little higher than the taller of the narrow rays and 3-4 cells
wide at the duct. In the more mature wood the narrow rays are almost all biseriate, 5-9
cells high, the rarer uniseriate rays being only 3-4 cells high. Where the biseriate rays
include a uniseriate portion they may reach a height of 14 cells. These narrow rays are
markedly heterogeneous in cellular composition with a marginal row of square or up-
right cells and procumbent cells forming the body of the ray. Upright cells frequently
occur in the biseriate portions however and included uniseriate portions are of upright
cells. The fusiform rays containing ducts are up to 1 mm. high and 0-25 mm. across the
widest part of the duct. Occasionally two ducts occur in the ray. The duct is surrounded
by 2-3 layers of thin walled procumbent cells much narrower than those of the rest of
the ray and with more oblique tangential walls. The cellular composition is otherwise
similar to that of the smaller rays. Most of the cells in all types of rays are richly pitted
and have dark contents and many show evidence of having contained crystals.

Comparison. The material on which the new species is based differs very little from the
wood of E. parviareolatum. The vessels of the new species are a little wider and fewer per
sq. mm. ; the narrow rays are more commonly biseriate and on the whole not so high and
are more widely spaced.

Such small variations in vessel size and number, and ray width and height are com-
monly found between different samples of a single living species in which case they may
be the result of differences in the age or size of a tree, conditions of growth, relation to
branches, and so on. The scarcity of biseriate rays in the younger wood of the new
material accentuates the similarity, since the type specimen is obviously a young stem or
branch while the new material consists for the most part of more mature wood.

The possibility that all the material belonged to a single natural species of Eocene

362

PALAEONTOLOGY, VOLUME 9

times cannot be excluded. Nevertheless I have felt justified in giving the English specimen
a new name, more particularly because the type species includes roots and stems and
secondary phloem is present in both. The specific epithet has been chosen to stress the
similarities between the new species and the type species.

AFFINITIES OF THE GENUS EDENOXYLON

Kruse (1954) placed Edenoxylon in the Anacardiaceae ‘without too much confidence’
since it could not be identified with any modern genus of the family. My own findings
confirm this.

Two families of dicotyledons, the Anacardiaceae and Burseraceae, have wood structure
similar to that of Edenoxylon. The similarity between the wood of these families is well
known and has been put forward as evidence of their close relationship; the families had
earlier been united in the Terebinthaceae even before the wood anatomy was well known.
More recent studies of wood anatomy have suggested that the Anacardiaceae and Burser-
aceae should be considered along with the Meliaceae, Sapindaceae, Rutaceae, and
Simarubaceae as a single phyletic complex (Heimsch 1942, Webber 1941).

In the Table below are set out some of the more important and variable characters of
the wood of the five tribes of the Anacardiaceae and of the Burseraceae and Edenoxylon.

Mangiferae

Spondiae

Rhoideae

Semecarpeae

Dobineae

Burseraceae

Edenoxylon

Vessel multiples

+

chains in
few spp.

+

chains

rare

+

-T

seldom

+

chains in
few spp.

+

chains

Banded parenchyma

+

—

rare

extended

aliform

—

Scanty paratracheal parenchyma

few spp.

+

+

—

+

+

+

Radial ducts

+

+

+

1 genus

—

+

+

Septate fibres

1 genus

+

+

—

+

+

-1-

Opposite inter-vascular pitting

~

1 sp.

some

genera

—

+

—

small

areas

Unilaterally compound pitting

—

—

—

—

+

—

(Data chiefly from Heimsch 1942, Metcalfe and Chalk 1950, Webber 1941.)
+ commonly present in tribe or family; — not reported in tribe or family.
These signs denote presence or absence in Edenoxylon.

It will be seen that considering distribution of wood parenchyma and the occurrence of
septate fibres and radial canals, Edenoxylon has most in common with the Spondiae and
Rhoideae of the Anacardiaceae and with the Burseraceae. The Burseraceae however are
usually readily distinguished by the unilaterally compound pits communicating between
vessels and ray cells. Furthermore the presence of small areas of opposite intervascular

EXPLANATION OF PLATE 61

Figs. 1 , 2. £■. parviareolatum. 1 , Transverse section of secondary wood of holotype ; X 20. 2, Longitudinal
section of same, mostly tangential, more or less radial to the right; X 20.

Figs. 3-7. E. aemuhim. 3, Tangential section showing a large fusiform ray with duct; some narrower
rays and vessels filled with tyloses are also shown; x70. 4, Transverse section showing vessel
multiples and numerous narrow rays, X 35. 5, The same at higher magnification showing 2- and 3-
seriate rays and the uniform tissue, probably mainly septate fibres, between the large vessels; X 120.

6, Intervascular pitting on vessel wall; the arrangement is seen to be irregularly alternate; X 500.

7, Radial longitudinal section; x70.

Palaeontology, Vol. 9

PLATE 61

BRETT, Wood of Eocene Anacardiaceae

D. W. BRETT: FOSSIL WOOD OF ANACARDI ACEAE

363

pitting in Edenoxylon is an additional feature linking the fossil genus with the Rhoideae
and Spondiae.

The name 'parviareolatiim ’ was given to the type because the small intervascular pits
(4-5 /x) present in this species ‘are not characteristic of the family as a whole’ (Kruse
1954). In the Rhoideae, however, small pitting is common and the larger intervascular
pitting which is characteristic of the Mangiferae is rarely found. Intervascular pitting of
about the same order of size as that of Edenoxylon is to be found in some species of Rhus
(R. toxicodendron 6 p, R. thiinhergii 5 p). and very small pitting is reported in Eaguetia
and Trichoscypha {WitXc2L\k and Chalk 1950), all three genera belonging to the Rhoideae.

DISCUSSION

Several fossil woods have been ascribed to the Anacardiaceae, and five organ-genera
named. The first to be described, the genus Rlioidiuni Unger (1850), has been used by
authors for fossil wood resembling the wood of Rhus spp. The genus Anacardioxylon
Felix (1882, diagnosis 1894) was erected for fossil wood resembling wood of Spondias.
Neither Rhoidium spp. nor Anacardioxylon spp. have radial ducts, with the exception of
Anacardioxylon nwllii Krausel which was subsequently called Sumatroxylon by den
Berger because of doubt about its affinities (Edwards 1931). Radial ducts occur
commonly throughout the family, in both Rhus and Spondias, and in the other three
genera of fossil woods.

Gluloxylon Chowdhury (1934, diagnosis 1936) has been characterized so as to include
wood of a type common to the living genera Ghtta and Melanorrhoea. In addition to
Edenoxylon another new organ genus, Schinoxylon, was described by Kruse (1954), and
this is said to differ from the living Schinus only in the pattern of the vessels in transverse
section.

Although not recorded previously from the lower beds of the English Tertiary (the
wood described above has been noted in Brett 1960 and Chandler 1964) the Anacardi-
aceae are well represented among the fruits and seeds of the London Clay and the succeed-
ing Lower Bagshot and Bournemouth Beds. Reid and Chandler (1933) described eleven
fossil species. Only three living genera were recognized. Most of the remaining speci-
mens were assigned to four new genera presumed to be extinct, and it seemed to these
authors that some of the features characterizing the fossil genera have been recombined
in the living genera (Reid and Chandler 1933, p. 299). All the fossils were, however,
assigned to Spondiae. In addition to three species assigned to the Spondiae, the Lower
Bagshot and Bournemouth Beds yielded four or five species of Rhus (Chandler 1962,
1963, 1964). Identifications of pollen grains from the London Clay (Sein 1961) and the
Oligocene of the Isle of Wight (Pallot 1961) have also confirmed the presence in these
fioras of Spondiae and Rhoideae.

Of the few plants so far known from the Thanet Sands and Woolwich and Reading
Beds several suggest a lowland tropical environment whilst others present a more tem-
perate aspect. The situation is in fact similar to that of the London Clay flora and it is
becoming clear that the vegetation of the Landenian probably differed very little, if at
all, from that of the succeeding Ypresian (Chandler 1961, 1964: a full discussion of
recent work is given in the latter). The character of the anacardiaceous wood Edenoxylon
is essentially that of a tropical type with no strongly marked seasonal increments in its
radial growth. In Wyoming Edenoxylon is found among other fossil woods of tropical or

364

PALAEONTOLOGY, VOLUME 9

subtropical affinities including palm stems. Palm stems are also known from the Lan-
denian of England, several having been found by collectors but none described.

The overall aspect of the vegetation of the Green River Formation, from which the
Wyoming woods were derived, is warm temperate or subtropical. Like most large fossil
floras the various remains (leaves, pollen, wood) probably include elements of a more
extreme nature but, unlike the Lower Eocene floras of southern England, the Green
River flora from Wyoming does not include any large assemblage of lowland rain forest
types. Considering the present altitude of the Green River fossiliferous beds (5,000-
10,000 ft.) it is not improbable that the area was already at some considerable height
above sea level in Eocene times, although it has been concluded that the altitude could
not have been above 3,000 ft. (Brown 1934).

Acknowledgements. I am indebted to the following for their help: the Director of the Royal Botanic
Gardens, Kew, and Dr. C. R. Metcalfe, for access to the Kew collection of specimen slides and for
permission to work in the Jodrell Laboratory; Dr. W. L. Stern, Curator of the Samuel James Record
Memorial Collection, School of Forestry, Yale University for specimens of wood of Anacardiaceae;
the Curator of the Palaeobotanical Collection of the University of Cincinnati, Ohio, for the loan of the
type specimen of Edenoxylon.

REFERENCES

BRETT, D. w. 1960. Fossil oak wood from the British Eocene. Palaeontology, 3, 86-92.

BROWN, R. w. 1934. The recognizable species of the Green River flora. Prof. pap. U.S. geol. Snrv. 185,
45-78.

CHANDLER, M. E. j. 1961. The Lower Tertiary floras of Southern England, 1. Palaeocene floras. London
Clay flora (Supplement). Brit. Mus. (Nat. Hist.), London.

■ 1962. The Lower Tertiary floras of Southern England, 2. Flora of the Pipe-Clay Series of Dorset

(Lower Bagshot). Brit. Mus. (Nat. Hist.), London.

1963. The Lower Tertiary floras of Southern England, 3. Flora of the Bournemouth Beds, the

Boscombe, and the Highclitf Sands. Brit. Mus. (Nat. Hist.), London.

1964. The Lower Tertiary floras of Southern England, 4. A summary and survey of findings in the

light of recent botanical observations. Brit. Mus. (Nat. Hist.), London.

CHOWDHLJRY, K. A. 1934. A fossil dicotyledonous wood from Assam. Curr. Sci. 3, 255-6.

1936. A fossil dicotyledonous wood from Assam. Ann. Bot. 50, 501-10.

1952. Some more fossil woods of Glutoxylon from southeast Asia. Ann. Bot. n.s. 16, 373-7.

EDWARDS, w. N. 1931. EossiUiim catalogus, 2: Plantae, pars 17, Dicotyledones (Ligna). Berlin.

EELix, j. 1882. Studien iiber fossile Holzer. Inaiig. Diss., Leipzig.

1894. Untersuchungen iiber fossile Holzer. Z. dt. geol. Ges. 46, 79-100.

HEiMSCH, c. 1942. Comparative anatomy of the secondary xylem in ‘Gruinales’ and ‘Terebinthales’ of
Wettstein with reference to their taxonomic grouping. Lilloa, 8, 83-198.

KRDSE, H. o. 1954. Some Eocene dicotyledonous woods from Eden Valley, Wyoming. Ohio J. Sci. 54,
243-68.

METCALFE, c. R. and CHALK, L. 1950. Anatomy of the Dicotyledons. Oxford.

BALLOT, J. M. 1961. Plant microfossils from the Oligocene of the Isle of Wight. Ph.D. Thesis, University
of London.

REID, E. and CHANDLER, M. E. J. 1933. The London Clay flora. Brit. Mus. (Nat. Hist.), London.

SEiN, MA KHiN. 1961. Eossil spores of the London Clay. Ph.D. Thesis, University of London.

UNGER, F. 1850. Genera et species plantarum fossilium. Vindobonae.

WEBBER, I. E. 1941. Systematic anatomy of the woods of the Burseraceae. Lilloa, 6, 441-65.

D. W. BRETT
Dept, of Botany,
The University,
Glasgow, W. 2

Manuscript received 18 May 1965

PARACHONETES, A NEW LOWER AND MIDDLE
DEVONIAN BRACHIOPOD GENUS

by J. G. JOHNSON

Abstract. Parachonetes is proposed as a new chonetid genus with Chonetes luacrostriata Walcott from Nevada
as its type species. It is thought to have been derived from the protochonetid genus Ecceiitiicosta. Barrande's
species Chonetes veriienili is assigned to Parachonetes as are two named species from south-eastern Australia.
Parachonetes is absent in eastern North America, but is represented in Nevada, the Canadian Arctic, Novaya
Zemlya, the Ural Mountains, Central Asia, south-eastern Asia, Czechoslovakia, and in south-eastern Australia.

In the course of studies elucidating the Lower Devonian brachiopods of central Nevada
(Johnson, in prep.) Walcott’s species 'Chonetes' niacrostriata was studied through the
preparation of internal moulds and found to be unassignable to any chonetid genus pre-
viously described. Further work on some excellent material of Barrande’s species
‘ Chonetes' verneuili from the Schary Collection at the Museum of Comparative Zoology,
Harvard, showed that that species is congeneric. At the time these studies were in progress
Talent (1963) published descriptions and illustrations of chonetids from the Lower
Devonian of Victoria, Australia, which proved also to belong to the new genus. In addi-
tion silicified chonetids sent by Dr. Talent from the Receptacidites Limestone of New
South Wales are illustrated herein and show the cardinalia very well.

SYSTEMATIC PALAEONTOLOGY

Superfamily chonetacea Bronn 1862
[notn. trails/. Shrock and Twenhofel 1953 (c.y Chonetidae Bronn 1862)]

Family chonetidae Bronn 1862
Subfamily devonochonetinae Muir-Wood 1962

Discussion. The two principal Silurian chonetid genera, Protochonetes and Stropho-
chonetes, are much alike internally as a recent investigation of the interior of Stropho-
chonetes has shown (Johnson, in prep.) and it seems likely that they shared a common
ancestor. Strophochonetes, however, was assigned to a separate subfamily by Muir-
Wood (1962, p. 40). Parachonetes gen. nov. is here assigned to the Devonochonetinae
because of its close relation to Eccentricosta which must have been derived from
Protochonetes, all of which suggest that a more critical appraisal needs to be made into
the phylogenetic relations of Protochonetes and Devonochonetes. If the latter genus was
indeed derived from the former, Parachonetes is a member of the Devonochonetinae
although antipodal to the type genus Devonochonetes.

Genus parachonetes gen. nov.

Type species. Chonetes niacrostriata Walcott 1884, p. 126.

Diagnosis. Protochonetids with low rounded costae originating along the hinge-line of
some specimens. Brachial valve cardinalia characterized by a pair of posteriorly conjunct

[Palaeontology, Vol. 9, Part 3, 1966, pp. 365-70, pis. 62-63.]

366

PALAEONTOLOGY, VOLUME 9

cardinal process lobes that develop an externally quadrilobate myophore and a more
or less deep alveolus. Dorsal median septum present except on small specimens.

Discussion. Parachonetes is erected to include a group of chonetid species that were
widespread geographically during most of Early Devonian and early Middle Devonian
time. Medium and larger size shells are among the largest Lower Devonian chonetids
and are deeply concavo-convex. In addition to the type species which is presently known
only from Nevada, the genus includes Chonetes verneui/i Barrande, which is widely re-
ported in the Bohemian-Uralian faunal province in Czechoslovakia (Barrande 1879),
Central Asia (Nikiforova 1937), the Ural Mountains (Rzhonsnitskaya 1960, p. 128), and
as far north as Novaya Zemlya (Tscherkessowa 1960, p. 178). Topotype material of
‘ Chonetes' venieuili was studied by the writer who was able to prepare internal moulds of
both valves to confirm the generic assignment. These specimens are illustrated in figs. 4-8
on Plate 63. Parachonetes verneuiJi is a large species with much more regular ribbing than
P. macrostriatus and in the brachial valve the alveolus of P. verneuili is shallower and
broader. The writer has seen fragmentary remains of chonetids like P. verneuili from the
upper Lower Devonian Stuart Bay Lormation on Bathurst Island in the Canadian
Arctic.

Chonetes kwangsiensis Wang (1956, p. 385, pi. 1, figs. Dl-4) and Strophomena
carnica Gortani (1915, p. 122, pi. 1, fig. 3) are species that resemble Parachonetes ex-
ternally; however, their internal structures are unknown. The species Plectambonites
yenlacensis Mansuy (1916, p. 49, pi. 7) probably is a Parachonetes. It has the typical
shape and ribbing, a pair of lateral septa and a median septum in the brachial valve, and
the posterior impression of the cardinal process is multilobate (Mansuy 1916, pi. 7, fig.
36). Patte (1926, p. 62) suggested that several of the Indochinese chonetoids given differ-
ent names by Mansuy, including P. yenlacensis, are no more than junior synonyms of
Chonetes zeili Mansuy (1908, p. 36). It seems highly probable that this is true, but the
figured specimens of Mansuy (1908, pi. 7, fig. 13) are insufficient for firm assignment to
Parachonetes.

In addition to the North American and Eurasian occurrences of Parachonetes, the
genus is represented by Chonetes baragwanathi Gill and by Chonetes? suavis Talent, both
of which have recently been described and illustrated by Talent (1963) from the Lower
Devonian of Victoria, Australia. P. suavis may prove to be synonymous with P. barag-
wanathi since the principal distinction of the former is its split dorsal septum, a feature
that is evanescent in P. macrostriatus. There remains some uncertainty regarding generic
affinities of Chonetes? foedus Talent (1963, pi. 37, 38).

Etienne Patte (1926, pp. 62, 63) first called attention to the resemblance between
P. verneuili, P. macrostriatus, and several strongly concavo-convex coarse-ribbed species
of Mansuy from Indochina. His astute evaluation, based wholly on considerations of
external morphology, are borne out by the present investigation of the interior structures.
Patte suggested further (1926, p. 63) that the Indochinese species ‘C.’ zeili may be only
a variety of P. macrostriatus.

Parachonetes macrostriatus and P. verneuili appear not to be threatened by senior
synonyms, nor is there any basis to merge them together, and they may be usefully em-
ployed to split Parachonetes into two species groups, (1) the macrostriatus group with
low, irregular, and commonly bifurcating costae and (2) the verneuili group, with even.

J. G. JOHNSON; PARACHONETES, A DEVONIAN BRACHIOPOD GENUS 367

more elevated costae that rarely bifurcate. On this basis group 1 comprises species from
Nevada, south-east Asia, and Australia while group 2 is restricted to Bohemia, the
Carnic Alps?, Central Asia, the Ural Mountains, Novaya Zemlya, and the Canadian
Arctic.

Comparison. Eccentricosta is smaller and less strongly concavo-convex than Parachonetes
and on this basis Eccentricosta and small Parachonetes are indistinguishable. Internally
the cardinal process lobes of Eceentrieosta are relatively smaller and more blade-like
than the stout curved lobes of Parachonetes and the latter bears a well-defined alveolus.
In addition, the elaborate posterior quadrilobation of the cardinal process of Para-
chonetes is not developed in Eceentrieosta (Berdan 1963, text-fig. Ic; Bowen, in press,
pi. 2, fig. 9) nor is there a dorsal median septum in Eccentricosta although this is a well-
developed structure in Parachonetes.

Paraehonetes macrostriatus (Walcott)

Plate 62, figs. 1-17; Plate 63, figs. 1-3

1884 Chonetes macrostriata Walcott, p. 126, pi. 2, fig. 13 (?); pi. 13, figs. 14, 14c/, \Ab, 14c.

1940 Clwnetes macrostriata Merriam, p. 55, pi. 6, fig. 4.

1944 Clwnetes macrostriatus Cooper, p. 345, pi. 134, fig. 16.

1962 Longispina macrostriatus Muir-Wood, p. 47.

Diseussion. Muir-Wood (1962, p. 47) incorrectly assigned this species to Longispina, but
she evidently had not seen interiors of either valve since they do not show any marked
resemblance to the interiors of Longispina (cf. Muir-Wood 1962, p. 46, figs. 7 a, b).

Description. Exterior. The shells are transversely shield-shaped with the cardinal angles
approximately equal to right angles in small shells. Large shells are commonly auriculate.
The valves are concavo-convex in lateral profile, becoming strongly concavo-convex in
large specimens. The hinge-line is long and straight and is the place of maximum width.
Very small spine bases are commonly developed along the hinge-line, but spines are sel-
dom seen. On one small shell measuring 8 mm. along the hinge-line there are six short
tubular spines on either side diverging laterally from the hinge-line at varying angles
from approximately 30 to 50 degrees. The spines are hollow and are filled with matrix
like that of the enclosing rock. The ventral beak is small, inconspicuous, and strongly
incurved. The interarea of the pedicle valve is orthocline to anacline, commonly ortho-
cline laterally, but anacline near the middle. The interarea of the brachial valve is hyper-
cline. The delthyrium is triangular and very broad at its base enclosing an angle of
greater than 100 degrees; however, the teeth fill a large portion of the lateral part of the
delthyrium at the hinge-line leaving an opening that is semicircular rather than triangular.

The ornament consists of irregular low radial costae emanating from the beak, and on
large specimens also along the lateral portions of the hinge-line. Costae commonly and
irregularly increase in number by bifurcation and there may be 5 to about 8 costae in a
space of 5 mm., 10 mm. anterior to the beak.

Interior of pedicle valve. The hinge teeth are small and triangular in cross section and
attach directly to the interior of the shell at the inner edges of the delthyrium. There is
a low median septum developed slightly anterior to the apex of the beak and it may con-
tinue anteriorly, dividing the vascular trunks. The median septum may be slightly

368

PALAEONTOLOGY, VOLUME 9

thickened or swollen at its posterior end. The diductor muscle scars are broad and
flabellate, lightly impressed, and they blend with the interior of the shell imperceptibly
along their anterior margin. The adductor scars are elongate and well defined. They are
commonly fairly well impressed posteriorly and blend with the interior of the shell an-
teriorly. Laterally they may be deeply impressed or may be bounded by short, slightly
arcuate, muscle bounding ridges. The vascular trunks are variably impressed anterior to
the adductor scars. The internal surface of the valves is commonly pustulose due to the
internal projection of the pseudopunctae and it is variably crenulated by the impress of
the costae.

Interior of brachial valve. Broadly divergent sockets are widely set apart along the
posterior edge of the shell and are bounded medially by low rounded socket plates. The
socket plates join with a pair of posteriorly conjunct plates that form the cardinal pro-
cess. The area between the eardinal process lobes is not only enclosed, but is depressed
below the level of the surrounding shell material to form an approximately circular,
broad, deep alveolus. On small shells the two cardinal process lobes are still partially
separated medially at their posterior ends, but in larger specimens shell material is de-
posited around the plates, forming a more or less complete U-shaped strueture, on the

EXPLANATION OF PLATE 62

Figs. 1-17. Parachoiietes macrostriatiis (Walcott). " Spirifer' pinyouensis zone (Emsian), central Nevada.
1, Ventral view of the lectotype (x2), USNM 13809, Lone Mountain. Illustrated by Walcott (1884,
pi. 13, fig. 146). 2, Ventral view of a paratype (former cotype) (x 1), USNM 13809A, Lone Mountain.
Illustrated by Walcott (1884, pi. 13, fig. 14). 3, Ventral view (x3), USNM 140435, locality 10769,
northern Roberts Mts. 4, Ventral view (X 2), USNM 140436, locality 10709, Sulphur Spring Range.

5, Dorsal view of internal mould (x 2), USNM 140437, locality 10762, northern Simpson Park Mts.

6, 7, Posterior view of middle part of hinge line (X 5) and dorsal view (x 1), USNM 140438, Lone
Mountain. 8, Ventral view (xl), USNM 140439, locality 10788, Roberts Creek Mtn. 9, Rubber
impression of internal mould of a small brachial valve ( X 5), impression of USNM 140440, locality
10729, McColley Canyon, Sulphur Spring Range. 10, 1 1, Internal mould of brachial valve and rubber
impression (x3), USNM 140441, locality 10729, McColley Canyon, Sulphur Spring Range. 12, 13,
Internal mould of brachial valve and rubber impression (x2), USNM 140442, locality 10729, Mc-
Colley Canyon, Sulphur Spring Range. 14, Posterior view of internal mould of pedicle valve (x 2),
USNM 140443, locality 10705, northern Roberts Mts. 15, Internal mould of pedicle valve (x2),
USNM 140444, locality 10729, McColley Canyon, Sulphur Spring Range. 16, 17, Posterior and
ventral views of internal mould of pedicle valve (x 1-5), USNM 140445, locality 10729, McColley
Canyon, Sulphur Spring Range.

EXPLANATION OF PLATE 63

Figs. 1-3. Parachonetes macrostriatiis (Walcott). Posterior, ventral, and anterior views of pedicle valve
(xl), USNM 140446, locality 10705, " Spirifer' pinyonensis zone (Emsian), northern Roberts
Mountains, central Nevada.

Figs. 4-8. Parachonetes verneiiili (Barrande). Upper Koneprusy Limestone (lower Emsian), Czecho-
slovakia. 4, Posterior view of internal mould of pedicle valve (x 2), USNM 140447. 5, 6, Posterior view
(x2i and ventral view (x 1) of internal mold of pedicle valve, Schary Collection, Museum Com-
parative Zoology, Flarvard, 9441. 7, 8. Dorsal views (x 1-5 and X 3) of internal mould, Schary
Collection, Museum Comparative Zoology, Harvard, 9440.

Figs. 9-14. Parachonetes cf. P. macrostriatiis (Walcott). Receptaciilites Limestone, Taemas, N.S.W.,
Australia. 9, Dorsal view (x2), Geol. Surv. Victoria 61055. 10, Interior of brachial valve (x6),
Geol. Surv. Victoria 61056. 11, Interior of pedicle valve (x3), Geol. Surv. Victoria 61054. 12, 13,
Interior and posterior views of brachial valve (x6), Geol. Surv. Victoria 61058. 14, Interior of
brachial valve (x3), Geol. Surv. Victoria 61057.

Palaeontology, Vol. 9

PLATE 62

JOHNSON, Devonian Parachonetes gen. nov.

Palaeontology, Vol. 9

PLATE 63

JOHNSON, Devonian Pamclionetes gen. nov.

J. G. JOHNSON: P ARACHONETES, A DEVONIAN BRACHIOPOD GENUS 369

posterior face of which four lobes of the cardinal process are developed. A broad flat
median septum originates at the anterior edge of the alveolus and becomes narrow and
strong anteriorly. The flat spatulate posterior end of the median septum tends to be split
on some specimens and joins the inner edges of the cardinal process lobes. Adjacent to
the point of origin of the median septum there is a pair of lateral septa that enclose the
adductor scars. The lateral septa are short and diverge only slightly from one another.
The shell is commonly thickened at the anterior edges of the socket plates forming
irregularly triangular umbonal platforms. The interior of the valve is made pustulose by
the pseudopunctae and may be more or less crenulate by the impression of the costae.

Comparison. Parachonetes macrostriatus from Nevada is very much like specimens from
the Receptaculites Limestone in external form and ribbing. Internally as well the two
bear considerable resemblance to one another. However, consistent differences may be
seen in the thickness of the shell of the two species. The shell material of the Australian
specimens is generally thinner than that of P. macrostriatus from Nevada so that the
ribbing corrugates the interior surface rather strongly (PI. 63, figs. 1 1 and 14). Specimens
of the same size from Nevada have much thicker shells with the ribbing little impressed,
or not at all, internally (PI. 62, figs. 10-17), consequently their internal structures are
generally a bit more well defined or strongly impressed so that the ventral vascular
trunks and dorsal internal papillae are well displayed (PI. 62, figs. 15-17, fig. 11).

Parachonetes macrostriatus differs from P. verneuili in having more irregular, more
commonly bifurcating radial costae. In addition the ribbing of P. verneuili is relatively
strongly impressed on the interior of the valves (PI. 63, figs. 6, 7).

Occurrence. Parachonetes macrostriatus first appears in Nevada in the Acrospirifer
kobehana zone (Emsian; see Johnson, 1962; 1965, fig. 4) on the north side of the Roberts
Mountains. It is present in the ' Spirifer' pinyonensis zone in the Sulphur Spring Range,
in the Roberts Mountains, at Lone Mountain, and in the northern Simpson Park Range.

Cooper and others (1942, p. 1773) reported the presence of the “'Chonetes macro-
striata' fauna in the Muddy Mountains of southern Nevada. In order to investigate this
occurrence the collections in question were borrowed from the U.S. National Museum.
The material consists of two collections made by Dr. C. R. Longwell in 1926 and 1927.
Both of these bear specimens of Parachonetes macrostriatus in a fauna that is indicative
of the " Spirifer' pinyonensis zone, but the collection localities are from north of Las
Vegas rather than from the Muddy Mountains and one of the two has a fairly detailed
locality slip indicating that the collection was made near the east flank of the Spotted
Range.

Figured specimens. The locality numbers cited in the plate legends are those of the U.S. National
Museum. The cotypes, USNM 13809, come from Walcott’s locality 555, Lone Mountain. Both speci-
mens are refigured and the writer designates the specimen in figure 1 on plate 62 as the holotype.
Walcott (1884, p. 127) lists several localities with P. macrostriatus, but Lone Mountain is not among
them. However, the labels with Walcott's two best preserved illustrated specimens (pi. 13, figs. 14,
146) examined on loan from the U.S. National Museum give Lone Mountain as their source locality
and there is no doubt that the discrepancy is due to an error of omission in the text. The illustrated
hypotype specimens include USNM nos. 140435-46.

Acknowledgements. The writer wishes to thank Dr. G. A. Cooper of the U.S. National Museum for
the loan of Walcott 's types of ‘ Chonetes ’ macrostriata and for photographs of two specimens ; Professor
H. B. Whittington of the Museum of Comparative Zoology, Harvard, for loan of topotype specimens

370

PALAEONTOLOGY, VOLUME 9

of Parachonetes venieiiili from the Schary Collection; Dr. D. J. McLaren of the Geological Survey of
Canada for the opportunity to examine collections from the Stuart Bay Formation of Bathurst Island ;
Professor A. J. Boucot of Pasadena for access to his collection of brachiopods from the upper Kone-
prus Limestone of the Barrandian. The writer is especially indebted to Dr. J. A. Talent of the Geological
Survey of Victoria, Australia, for loan and permission to illustrate silicified specimens from the Recep-
taculites Limestone Taemas, N.S.W. Work on this project was supported by a grant from the National
Science Foundation, number GP 2290.

REFERENCES

BARRANDE, J. 1879. Svsteme silurien dii centre de la Bolieme, 5, Brachiopodes. 1-226, 153 pis., Prague,
Paris.

BERDAN, J. M. 1963. Ecceiitricosta, a new upper Silurian brachiopod genus. J. Paleont. 37, 254-6.

BOWEN, z. p. (in press). Brachiopoda of the Keyset Limestone (Siluro-Devonian) of Maryland and
adjacent areas. Mem. geol. Soc. Am.

COOPER, G. A. 1944. Phylum Brachiopoda in shimer, h. w. and shrock, r. r.. Index Fossils of North
America. 277-365, pi. 105-43.

and others 1942. Correlation of the Devonian sedimentary formations of North America. Bull.

geol. Soc. Amer. 53, 1729-94, 1 pi.

GORTANi, M. 1915. Fossili eodevonici della base del Capolago (Seekopfssockel). Palaeontogr. Ital. 21,
118-87, 3 pis.

JOHNSON, J. G. 1962. Brachiopod faunas of the Nevada formation (Devonian) in central Nevada.
J. Paleont. 36, 165-9.

1965. Lower Devonian stratigraphy and correlation, northern Simpson Park Mountains, Nevada.

Bull. Canadian Petrol, geol. 13, 365-81.

(in prep.) Great Basin Lower Devonian Brachiopoda. Manuscript, 1-578, pi. 1-75.

MANSUY, H. 1908. Contribution a la carte geologique de ITndo-Chine (Paleontologie). Indochina,
Serv. Mines. Hanoi-Haiphong, 1-73, pi. 1-18.

1916. Etude complementaire des faunes paleozoiques et triasiques dans FEst du Tonkin. Indo-
china, Serv. geol., Mem. 5 (4), 39-71, pi. 6-8.

MERRiAM, c. w. 1940. Devonian stratigraphy and paleontology of the Roberts Mountains region,
Nevada. Spec. Pap. geol. Soc. Am. 25, 1-114, 16 pis.

MUIR-WOOD, H. M. 1962. On the morphology and classification of the brachiopod suborder Chonetoidea.
British Museum (Nat. Hist.), 1-132, 16 pis.

NIKIFOROVA, o. I. 1937. Upper Silurian Brachiopoda of the central Asiatic part of the USSR. Central
Geol., Mon. Pal. Prosp. Inst. USSR, 35, 1-93, 14 pis.

PATTE, ETIENNE. 1926. Etudes paleontologiques relatives a la geologie de Test du Tonkin (Paleozoique
et Trias). Indochina, Bull. Serv. geol. 15 (1), 1-240, pi. 1-12.

RZHONSNiTSKAYA, M. A. 1960. Die Korrelatioii der karbonatischen Sedimente des Unter- und Mittel-
Devons in der Sowjetunion und in Westeuropa. Prager Arbeitstagiing iiber die Stratigraphie des
Silurs und des Devons, 123-35.

TALENT, J. A. 1963. The Devonian of the Mitchell and Wentworth Rivers. Mem. Victoria geol. Siirv. 24,
1-118, 78 pis.

TSCHERKESSOWA, s. w. 1960. Stratigraphie des Obersilurs und Devons des westlichen Sektors der
Sowjetischen Arktis. Prager Arbeitstagiing iiber die Stratigraphie des Silurs und des Devons, 175-85.

WALCOTT, c. D. 1884. Paleontology of the Eureka District. Mon. U.S. geol. Surv. 8, 1-298, 24 pis.

WANG, Y. 1956. Some new brachiopods from the Yiikiang formation of southern Kwangsi province.
Scientia Sinica, 5 (2), 373-88, 3 pis.

J. G. JOHNSON

Division of Geological Sciences,
California Institute of Technology,
Pasadena,

California, U.S. A.

Manuscript received 28 May 1965

MORPHOLOGY AND STRATIGRAPHIC RANGE
OF THE PHYLLOCARID CRUSTACEAN
CARYOCARIS FROM ALASKA AND THE
GREAT BASIN

by MICHAEL CHURKIN, JR.

Abstract. Exceptionally well-preserved specimens of Caryocaris from Alaska and the Great Basin show a
peculiar shell enrolment; several other important skeletal details are described for the first time. In the light of
this new material, Ruedemann’s types that are the basis for his reconstructions of Caryocaris are re-examined.
Finally, a review of the stratigraphic range of Caryocaris leads to the conclusion that none of its reported Silurian
species are definitely Caryocaris. The range of the genus should therefore be restricted for the present to the
Ordovician and probably to the lower half of the system.

Since Salter (1863) named Caryocaris, a pod-shaped crustacean from the Skiddaw
Slates of the Lake District, England, it has been reported from Ordovician graptolitic
shales in North America (Gurley 1896; Ruedemann 1921), South America (Bulman
1931), Australia (Chapman 1903), and Scandinavia (Stormer 1937), and new species of
Caryocaris have been described from Silurian graptolitic shales in south-eastern Alaska
(Ruedemann 1934) and Oklahoma (Ruedemann 1935). The flattened and crushed speci-
mens hitherto described are difficult to interpret and have resulted in a number of
different restorations.

The purpose of this paper is to describe exceptionally well-preserved specimens of
Caryocaris from Alaska and Nevada that show for the first time a peculiar shell enroll-
ment and several morphological details. In the light of this new material, Ruedemann’s
types that have served as the basis for his classical reconstructions are re-examined, and,
finally, the stratigraphic range of Caryocaris is reviewed and greatly reduced.

Acknowledgements. This study is the Tesult of the discovery of exceptionally well-preserved Caryocaris
in mapping the Charley River (1 : 250,000) and Eagle D-1 ( 1 : 63,360) quadrangles in east-central Alaska
(Brabb and Churkin 1964, 1965). I am indebted to Earl E. Brabb (U.S. Geological Survey) for his
interest and numerous useful suggestions; to W. D. Ian Rolfe (Hunterian Museum, Glasgow) for
critically reading the manuscript; to J. Wyatt Durham (University of California, Berkeley) for sug-
gesting shell enrollment to explain the spirally coiled carapaces; to Copeland MacClintock (Peabody
Museum of Natural History) for ideas concerning shell wall microstructure; to Robert A. Gulbrandsen
(U.S. Geological Survey) for interpreting the X-ray patterns of Caryocaris shell material; to Charles W.
Merriam (U.S. Geological Survey) for kindly furnishing his collections of Caryocaris from Nevada;
to A. R. Palmer (U.S. Geological Survey) for arranging loans of type specimens from the U.S. National
Museum; and to Donald W. Eisher and Clinton E. Kilfoyle (New York State Museum) for loaning
type specimens in their care. Publication is authorized by the Director, U.S. Geological Survey.

CARYOCARIS IN EAST-CENTRAL ALASKA

Caryocaris occurs in several graptolitic shale horizons of Ordovician age in the Road
River Formation (Ordovician and Silurian) of east-central Alaska and is especially

[Palaeontology, Vol. 9, Part 3, 1966, pp. 371-80, pis. 64, 65.]

C 4179 C C

372

PALAEONTOLOGY, VOLUME 9

abundant in chert at the base of the formation (Churkin and Brabb 1965). The best pre-
served specimens described below were obtained from loc. Ml 001 -CO (see Appendix)
near the Alaska-Canada border. Here abundant specimens of Caryocaris are preserved
with the graptolites Isograptus forcipiformis, Isograptiis caduceus cf. var. maximo-
divergens, and Didymograptus cf. D. nitidus in a greyish-black chert 5 to 8 feet strati-
graphically above unnamed limestone containing Late Cambrian trilobites. Both
Caryocaris and the graptolites are preserved in relief and are easily distinguished by their
phosphate bloom, apparently a white phosphatic- weathering product (PI. 64, figs. 1, 2).
The graptolites associated with the Caryocaris indicate a latest Arenig or earliest
Llanvirn age equivalent to graptolite Zone 5 of Didymograptus hirundo or Zone 6 of
Didymograptus bifidus of Elies and Wood (1901-18). Conodonts found in chert directly
below the Caryocaris locality were identified by John W. Huddle, of the U.S. Geo-
logical Survey, and suggest late Arenig. Thus, all of the Tremadoc and most of the
Arenig seem to be absent in the less than 5-foot chert interval between the conodont
locality, the lowest fauna in the Road River Formation, and the highest trilobite collec-
tion from the underlying Cambrian limestone. At loc. 63ACn 1533, about 3 miles east of
loc. MlOOl-CO, Caryocaris is again very abundant in the basal Road River Formation,
where it occurs in thin beds of chert pebble conglomerate and chert grit interstratified
with and cemented by dark-grey chert. The poorly sorted angular grains, cross-laminae,
and scour-and-fill structures suggest that the Caryocaris material, in places reaching
coquinoid proportions, was laid down in a high energy environment as part of a basal
deposit unconformably overstepping Cambrian limestone.

MORPHOLOGICAL FEATURES OF CARYOCARIS IN
EAST-CENTRAL ALASKA

The Alaskan specimens of Caryocaris are unusual in that they have in various degrees
their third dimension preserved in chert instead of appearing as completely flattened
films in shale. Beeause the Alaskan specimens are not completely flattened, they reveal
a peeuliar spiral coiling of the carapace wall here described for the first time. In addition,
other important skeletal features hitherto unknown or inadequately figured are de-
scribed below, and lastly the microstructure and eomposition of the carapace wall are
discussed. The terminology of the exoskeletal elements of Caryocaris has not been
standardized, and therefore the nomenclature used in the following description is given
in text-fig. 1.

Carapace enrolment. Spirally coiled carapaces of Caryocaris appear in long view as
cylindrically-shaped body walls tapering to blunt points at each end (PI. 64, figs. 3, 4).
All the larger specimens were partially broken, but several are still nearly 20 mm. long.
Their corresponding diameters are about 2 mm. ; and width to length ratio is about 1 ; 10
as contrasted with about 1 : 4 in associated flattened specimens.

In a number of specimens transverse breaks show in cross-section the internal spiral
coiling of the carapace wall (PI. 64, figs. 5, 6). The fragile carapace walls, nearly black in
colour, are supported by microcrystalline white quartz deposited between each whorl. In
thin sections random profiles through coiled carapaces reveal a number of differently
appearing patterns of the shell wall (PI. 64, fig. 7) of which the shapes of the longitudinal
(PI. 64, fig. 8) and nearly transverse profiles (PI. 65, fig. 1) are highly charaeteristic.

M. CHURKIN, JR.: PHYLLOCARID CRUSTACEAN CARYOCARIS

'ill,

Apparently each coiled piece is a single valve that is the result of separation, perhaps by
moulting, along the dorsal hinge line followed by in-curling of the valve wall starting
from the ventral margin.

Spined posterior margin. Ruedemann (1921) originally described several features of
Caryocaris from graptolitic shale here referred to the Road River Formation at the
Alaska-Yukon boundary, but the coiled and three-dimensional specimens preserved in
chert from the base of the formation apparently were not available to him.

‘The Alaska-Yukon material’, Ruedemann wrote, ‘which is not obscured by an im-
perfect cleavage leaves no doubt that the posterior margin of the carapace was indeed
furnished with a fine comb of uniform bristles or teeth. ’ As Ruedemann pointed out, the

RIGHT

TEXT-FIG. 1. Orientation and terminology of Caryocaris used in this paper.

posterior margins of the Alaska-Yukon species certainly have a row of ‘bristles or
teeth’ (here referred to as spines and spinules to divorce these structures of unproved
function from a genetic connotation) and are not a product of structural deformation.
These spines, however, are not of equal size as Ruedemann claimed. The spine at each end
of the posterior margin is several times larger than the 13 intervening spinules (PI. 65,
fig. 3). In addition, the ventral margin in its posterior portion, has a row of 5 and possibly
more posteriorly directed spinules (PI. 64, fig. 3 and PI. 65, fig. 3) which have not been
previously recorded anywhere and are rarely preserved even in the best Alaskan speci-
mens.

Carapace horn. The Alaskan Caryocaris certainly has a horn-like projection at the
anterior end of its dorsal margin (PI. 65, fig. 2) but it is simply a prolongation of the
carapace wall instead of a ‘distinct plate’ or true rostrum as Ruedemann (1921, p. 96,
figs. 45, 51 ; p. 99) indicated.

Segmentation and appendages. Although the Alaskan collections consist of several
hundred specimens, none of the carapaces has either an abdomen or telson attached as
Ruedemann (1921) showed in figs. 46, 48, and 51. Several fragments that resemble
‘abdomina’ and ‘telsons’, were found but these segmented structures, though they are
well preserved, are completely detached from any carapace and their biological affinities
to Caryocaris can only be conjectured. An examination of the original specimens in the
New York State Museum from which Ruedemann’s figs. 46, 48, and 49 were drawn has

374 PALAEONTOLOGY, VOLUME 9

also failed to show any justification for the segmentation or appendages shown in his
drawings.

Ventral margin. The ventral margins of the Alaskan Caryocaris have a thickened border
that is more or less well preserved. In flattened specimens this border is frequently im-
pressed in various positions on the opposite valve wall, thus suggesting that the carapace
walls were in these cases enrolled before flattening.

Carapace surface ornamentation and wall microstructure. The surface of the Caryocaris
exoskeleton has the iridescent lustre common to other phosphatic shells and appears
smooth except in flattened specimens from shale, where it is generally finely wrinkled.
Under the microscope, however, the exterior of the carapace is seen to be distinctly
corrugated into very fine furrows and ridges that form a slightly sinuous pattern.

The microstructure of the carapace wall in thin section consists of three distinct layers
(PI. 64, fig. 7; PI. 65, fig. 1). The outer and inner layers range in colour from pale
yellowish-orange to shades of yellowish-brown and are separated from each other by a
much thinner middle layer that is dark brown. Frequently the edges or centre of this
middle layer of the carapace wall serve as a plane along which the inner layer separates
from the outer layer. A very faint fibrous structure (possibly representing in part very
fine canals) normal to the carapace wall can be seen in the inner and outer layers, but the
middle layer instead seems to have a faint lamination or parting, especially along its
centre, that parallels the carapace layering. Several specimens seem to have a very
narrow hollow space where two carapace walls come to a point. Presumably this is the
union of two valves along the hinge line.

In thin sections the shells have the faint birefringence of microcrystalline apatite that
probably represents very nearly the original chitino-phosphatic material of the exo-
skeleton. The mineralogical composition of Caryocaris shell material separated out
from Coll. 63ACn 1533 gave the X-ray pattern of carbonate-fluorapatite.

Involute phosphatic shells similar to those described above but only one-third normal

EXPLANATION OF PLATE 64

Fig. 1. Crushed carapaces of Caryocaris ciirvilata associated with Isograptus caduceus cf. var. maximo-
divergens. The phosphatic fossils weather greyish- white and stand out clearly against the greyish-black
chert matrix; USNM 147441, loc. MlOOl-CO, Road River Formation, east-central Alaska. X 2.

Fig. 2. Right valve view of Caryocaris curvilata associated in the same chert fragment with Isograptus
forcipiformis\ USNM 147442, loc. MlOOl-CO, Road River Formation, east-central Alaska. X2.

Fig. 3. Enrolled valve of Caryocaris (lower centre) next to two partially flattened valves. The valve in
the upper left corner shows the spined posterior margin characteristic of Caryocaris curvilata and
the rarely preserved fine spinules along the postero-ventral border; USNM 147443, loc. MlOOl-CO,
Road River Formation, east-central Alaska. X 3. Photographed with ammonium chloride coating.

Fig. 4. Enrolled Caryocaris reflecting light from its phosphatic carapace that has an iridescent lustre;
USNM 147444, loc. M1006-CO, Road River Formation, east-central Alaska. x2.

Figs. 5, 6. Enrolled Caryocaris showing transverse views of the spiral coiling of the very thin and dark
carapace wall. The delicate spiral structure is supported by white microcrystalline quartz filling the
space between each whorl; 5, USNM 147445; 6, USNM 147446; loc. MlOOl-CO, Road River Forma-
tion, east-central Alaska. X 6.

Figs. 7, 8. Photomicrographs of a thin section through coiled carapaces of Caryocaris; 7, random cross-
section across a partially crushed specimen showing a complicated pattern of carapace wall folding
and the characteristic three-layered microstructure of the carapace wall, Xll4; 8, longitudinal
profile, X45; USNM 147447, loc. 63ACn 1533, Road River Formation, east-central Alaska.

Palaeontology, Vol. 9

PLATE 64

CHURKIN, Phyllocarid crustacean Caryocaris

M. CHURKIN, JR.: PHYLLOCARID CRUSTACEAN CARYOCARIS 375

size (PI. 65, figs. 4-6) were found in chert grit layers of the basal Road River Formation
near the Yukon River (loc. 63ABa 3181B). Fig. 6 shows that the shell wall has the same
three-layered microstructure as in Caryocaris. In the heart-shaped cross-sections there is
a very narrow but definite break in the shell microstructure where the shell wall comes to
a point, and this suture probably represents the dorsal hinge line. The shell walls coiled
into a double spiral can then be interpreted as two connected valves curled in along their
ventral margins. In some cross-sections the incurled valves on either side of the supposed
dorsal hinge line have shell walls of variable and unequal thickness but this is largely due
to splitting of the layered walls, as in the feather edge of the outer whorl in fig. 6, rather
than because of any original difference between the two valves.

Mertie (1930, pi. 9) illustrated involute shells from cherty grit in the same area. Edwin
Kirk {in Mertie 1930, p. 83) wrote that ‘the zoological affinities of these curious fossils
are doubtful’ and assigned them to the Pteropoda ‘mainly by a process of elimination’.
These involute fossils reported by Mertie appear in thin section identical to those in
Plate 65, figs. 4-6, and in both cases are similar in outline and have the same shell wall
microstructure as those of Caryocaris described above.

MORPHOLOGICAL FEATURES OF CARYOCARIS FROM NEVADA

AND IDAHO

Gurley (1896) described Caryocaris wrightii Salter and named a new species, C.
cnrvilatns, from Summit, Nevada (this locality lies in what is known today as Garden
Pass). In a large collection of Caryocaris, made by C. W. Merriam from the same grap-
tolitic shale at Summit, Nevada, I found several specimens of C. curvilata with a well-
developed carapace horn not originally recognized by Gurley. In these specimens the
posterior border, where preserved, is lined with a row of spinules. The larger spines at
each end of the posterior border, although generally incompletely preserved in the
Summit specimens, were recognized and figured by Gurley (pi. 5, fig. 3) but were not
mentioned by Ruedemann (1921) in his subsequent description of C. curvilata from the
Alaska-Yukon boundary. The type species of Caryocaris may lack the posterior spines
and carapace horn so characteristic of the western North American species discussed in
this paper, and as W. D. Ian Rolfe pointed out {in Theokritoff 1964, p. 183), Caryocaris
curvilata and other species of Caryocaris in America may belong to Rhinopterocaris
Chapman, a Lower Ordovician genus from Australia that has been synonymized with
Caryocaris but perhaps is generically distinct.

The major differences in shell width as well as the ‘complicated foldings and refold-
ings of the marginal filaments’ reported by Gurley (1896, p. 87) along the ventral margin
are probably a result of carapace enrollment before flattening as noted above in the
Alaskan material. Now that specimens of Caryocaris curvilata from its type locality are
known to have a carapace horn and a distinctively spined posterior margin, it may be
concluded that the better-preserved Alaskan specimens are more certainly nonspecific
with C. curvilata from Summit, Nevada. The Alaskan specimens provide details not
preserved in the type material from Nevada.

Several tridentate fossil parts resembling caudal appendages of various phyllocarids
(Jones and Woodward 1888-99, pi. 10, figs. 10, 11; Ruedemann 1934, pi. 22, fig. 7) occur
with Caryocaris curvilata at the Summit locality but in every case are detached from any

376

PALAEONTOLOGY, VOLUME 9

carapace. In this and in all collections of Caryocaris studied no single appendage was
found attached to a carapace, nor was there evidence of body segmentation. Ruedemann
(1921, p. 96, hg. 51) in a restoration of the Caryocaris ciirvilata from the Alaska-Yukon
boundary connected a segmented ‘abdomen’ to a spined posterior margin and then
attached a tridentate ‘telson’ to the end of the ‘abdomen’. Later Ruedemann (1934,
p. 92, pi. 22, figs. 8, 9) illustrated a specimen of C. ciirvilata from Trail Creek, Idaho, that
he implied showed ‘both the abdomen and the telson in place’ for the first time. A re-
examination of Ruedemann’s plesiotypes indicated that his figs. 8 and 9 are actually
a highly inferential reconstruction and that none of the specimens in the USNM type
collection no. 90858 from Trail Creek loc. 1367 has a telson-like structure attached to the
posterior end of a carapace as Ruedemann’s figures show. According to W. D. Ian
Rolfe, of the Hunterian Museum, Glasgow (written communication. May 1965), the
original specimen from which Ruedemann’s figs. 8 and 9 were made lacks the anterior
part of the carapace including any ‘articulated rostrum’, and has a telson near the pos-
terior end of the carapace, but definitely detached from it. Unfortunately, this specimen
has been recently damaged and the posterior end of the carapace has been chipped off
and lost, further obscuring any connexion between the carapace and the associated
telson. Another and more completely preserved specimen (text-fig. 2a) on the same shale
slab as the type for Ruedemann’s figs. 8 and 9, has a tridentate structure that is probably
part of another specimen adjoining its dorsal margin instead of being attached to the
posterior end, and also lacks any definite articulated rostrum or segmented abdominal
parts as indicated by Ruedemann.

In shales Caryocaris is generally preserved with either the right or left valve showing.
Rarely are both valves found connected. Ruedemann (1934, pi. 22, fig. 5) illustrated
such a specimen from Trail Creek, Idaho, and noticed ‘prominent spines on the inner
point of the posterior margins on both sides of the dorsal hinge’. He failed to mention,
however, a correspondingly large spine at the opposite (ventral) end of the posterior
margin of the left valve (see text-fig. 2b, a camera lucida drawing of Ruedemann’s
figured specimen). In the same collection from Trail Creek a specimen of C. ciirvilata

EXPLANATION OF PLATE 65

Fig. 1. Photomicrograph of nearly transverse cross-section through coiled valve of Caryocaris', USNM
147448, loc. 63ACn 1533, Road River Formation, east-central Alaska. x86.

Fig. 2. Right valve view of small Caryocaris with part of its carapace horn still intact; USNM 147449,
loc. Ml 001 -CO, Road River Formation, east-central Alaska. X 5. Photographed with ammonium
chloride coating.

Fig. 3. Close-up of the posterior portion of Caryocaris ciirvilata shown on PI. 64, fig. 2. The two spines
at each end of the posterior margin are separated from each other by 1 3 intervening spinules. The
row of 5 posteriorly directed spinules along the postero-ventral margin is also preserved; USNM
147442, loc. MlOOl-CO, Road River Formation, east-central Alaska. X 10. Photographed with am-
monium chloride coating.

Figs. 4-6. Photomicrographs of cross-sections of very small involute shells resembling Caryocaris. In
the heart-shaped cross-sections there is a break in the shell wall where it comes to a point. 4, X 35;
5, x70; 6, X 114. All in thin section USNM 147450, loc. 63ABa 3181B, Road River Formation,
east-central Alaska.

Figs. 7-9. Caryocaris ciirvilata from shale at Ninemile Canyon, east-central Nevada; 7, left valve view
of flattened specimen, USNM 147451, x 3; 8, anterior portion of carapace with the entire carapace
horn intact, USNM 147452, x3; 9, spined posterior margin, USNM 147453, X 10.

Palaeontology, Vol. 9

PLATE 65

CHURKIN, Phyllocarid crustacean Caryocaris

M. CHURKIN, JR.; PHYLLOCARID CRUSTACEAN CARYOCARIS 377

shows both pairs of prominent spines apparently at the posterior tips of each valve (text-
fig. 2c).

In Ninemile Canyon at the north end of the Antelope Range, east-central Nevada,
abundant and exceptionally well-preserved specimens of Caryocaris occur in a grey
shale devoid of graptolites (Nolan etal. 1956; Merriam 1963). Numerous specimens have
carapace horns and spined posterior margins particularly well preserved in a collection
made by C. W. Merriam from the Ninemile locality (PI. 65, figs. 7-9). In addition, the
very rarely preserved spinules along the posterior portion of the ventral margin are pre-
served on several specimens. An exceptionally well-preserved posterior margin shows
for the first time that the flanks of the major posterior spines, and the flanks of the
intervening spinules themselves bear a uniform set of microscopic spinules parallel to the
larger spines (PI. 65, fig. 9, and text-fig. 2d). Another specimen has along the antero-
ventral portion of the valve a serrated margin (see text-fig. 1 ) probably originally present
but not preserved in other specimens.

STRATIGRAPHIC RANGE OF CARYOCARIS IN ALASKA AND THE
WESTERN UNITED STATES

Although Caryocaris has been reported from Ordovician and Silurian shales in
western North America it seems more likely from the evidence given below that Caryo-
caris is restricted to the Ordovician and probably only to the lower half of the Ordo-
vician.

Ordovician. The oldest record of Caryocaris in western North America may be from the
grey shale provisionally included as the uppermost division of the Windfall Formation
of Late Cambrian age (Nolan et al. 1956, p. 21) or as the basal member of the overlying
Goodwin Limestone of Early Ordovician age (Merriam 1963, pp. 14-18, 21). This 150-
foot Caryocaris-hQdiXmg shale at Ninemile Canyon, Nevada, is overlain, however, by
limestone bearing the Early Ordovician Kainella trilobite fauna according to Nolan el
al., and thus may be itself Early Ordovician. Elsewhere in the Great Basin and Alaska,
Caryocaris is associated with graptolites that generally permit an accurate age assign-
ment. Caryocaris was found in the following graptolite zones of the Road River Forma-
tion in Alaska (Churkin and Brabb 1965); uppermost Arenig Zone 5 of Didynwgraptus
hirimdo or lower Llanvirn Zone 6 of Didynwgraptus bifidus (locality MlOOl-CO); upper
Llanvirn Zone 7 of Didynwgraptus nmrchisoni (locality M 1003-CO) ; and lowest Caradoc
Zone 9 of Nemagraptus gracilis (locality M1019-CO). At Summit, Nevada, Caryocaris
curvilata is associated with graptolites of the Llandeilo Zone 8 of Glyptograptus teretius-
cidus. The best specimens of Caryocaris curvilata in the graptolitic shale sequence at
Trail Creek, Idaho, occur in Zone 4 of Didynwgraptus extensus.

Silurian. The range of Caryocaris was extended from the ‘lowest Ordovician to the
Silurian’ by Ruedemann (1934) when he described the new species Caryocaris silurica
(p. 95, 96, pi. 22, figs. 15-20) from Silurian graptolitic shale at Klakas Inlet, Prince of
Wales Island, south-eastern Alaska. The types of C. silurica in the U.S. National Mu-
seum were re-examined and are considered as not identifiable to genus or even to phy-
lum with certainty because of poor preservation. The original illustrations of C. silurica
given in pi. 22, figs. 15 and 16 are highly interpretive, and the general shape, ‘eye’, and

TEXT-FIG. 2

M. CHURKIN, JR.: PHYLLOCARID CRUSTACEAN CARYOCARIS 379

apparent segmentation of Ruedemann’s types seem to reflect more the pattern of ex-
foliation of fossil material than the true outline and structures of the original fossils.
Text-figs. 2e,f, prepared from the original specimens, show only a general agreement in
broad outline with Ruedemann’s illustrations, but I could not positively identify any of
the internal segmentation or the ‘ eye ’ . In the field season of 1 964 1 made large collections
of Silurian graptolitic shale from Klakas Inlet, the type locality for C. sihirica, but no
phyllocarids were found.

Besides C. silurica the only other Silurian record of Caryocaris was made by Ruede-
mann (1935), when he named two new species, C. magnus and C. okJahomensis, from the
Henryhouse Shale of Oklahoma. An examination of Ruedemann’s type specimens indi-
cates that C. magnus is simply a pod-shaped depression in the rock and probably not a
fossil. C. okJahomensis, on the other hand, seems to have a branched structure (text-fig.
2g) instead of the segmented abdomen that Ruedemann’s figs. 2 and 3 indicate. Thus the
general outline of C. okJahomensis is unlike any phyllocarid, and its identification as to
phylum is uncertain. It may even be a plant.

None of the three species of Caryocaris reported from the Silurian can definitely be
identified as Caryocaris, and the stratigraphic range of the genus should therefore be
restricted to the Ordovician.

APPENDIX

Geographic position of Caryocaris collections in the basal part of the Road River

Formation of east-central Alaska

Collection numbers are those of the U.S. Geological Survey.

MlOOl-CO. From greyish-black chert at the base of the Road River Formation and 5 to 8 feet strati-
graphically above limestone containing Late Cambrian trilobites. On the ridge crest in NE. f , SW. f.
Sec. 28, T. 2 N., R. 33 E., Eagle D-1 quad. (1 : 63,360), lat. 64“ 58-3' N., long. 141° 2-4' W.

63ABa 3181B. From pale yellowish-brown and greyish-black laminated chert about 6 feet above the
base of the Road River formation. South flank of BM2814 (Crow) at elevation 2600', NE. f, SW. I,
Sec. 26, T. 2 N., R. 32 E., Eagle D-1 quad. (1 : 63,360), lat. 64° 58-3' N., long. 141° 10-9' W.

EXPLANATION OF TEXT-FIG. 2

a-c, Caryocaris ciirvilata from Lower Ordovician graptolitic shale near Trail Creek, Idaho (USGS
Coll. 1367-SD): a, one of the most complete specimens in collection. Ruedemann’s plesiotype (1934,
pi. 22, figs. 8, 9) in the U.S. National Museum (USNM 90858a), that according to Pmedemann has
‘both the abdomen and the telson in place’ lies on the same slab of shale; x 5. b, Ruedemann’s fig-
ured specimen (1934, pi. 22. fig. 5; USNM 90858b), showing the spined posterior portions of both
valves still connected along the dorsal hinge line; x 5. c, nearly complete right valve and posterior
portion of left valve showing both pairs of prominent spines at the posterior tips of each valve;
USNM 90858c, X 5.

d. Posterior margin of Caryocaris ciirvilata showing that the flanks of the major posterior spines and
the flank of the intervening spinules themselves bear a uniform set of microscopic spinules, X 14;
USNM 147454; Ninemile Canyon, east-central Nevada.

e, f. Type specimens of Caryocaris silurica Ruedemann from Silurian graptolitic shale at Klakas Inlet,
Prince of Wales Island, south-eastern Alaska (USGS loc. 1037); e, Ruedemann’s figured specimen
(1934, pi. 22, fig. 15), USNM 90857a, x5; /, ‘small specimen’ (Ruedemann 1934, pi. 22, fig. 16),
USNM 90857b, X 5.

g, Caryocaris oklahomeiisis Ruedemann (1935, fig. 2), holotype, USNM 114534, x5. Henryhouse
Shale, Silurian, Oklahoma.

All camera lucida drawings.

380

PALAEONTOLOGY, VOLUME 9

63ACn 1533. From chert grit interbedded with greyish-black laminated chert and shale. In Yukon
Territory, Canada, about 0-4 mile north-east of McCann Hill (International boundary monument
105), lat. 64° 55-5' N., long. 140° 59-5' W.

63ACn 1713. From small isolated outcrop of dark-grey, thin-bedded chert and shale. North side of
Hard Luck Creek valley SW. }, SW. i. Sec. 6, T. 4 N., R. 32 E., Charley River A-1 quad. (1 ; 63,360),
lat. 65° 11-8' N., long. 141° 19-5' W.

REFERENCES

BRABB, E. E. and CHURKIN, M. Jr. 1964. Preliminary geologic map of the Charley River quadrangle,
east-central Alaska. U.S. Geol. Surv. open-file map, scale 1 ; 250,000.

1965. Preliminary geologic map of the Eagle D-1 quadrangle, east-central Alaska. U.S. Geol.

Surv. open-file map, scale 1 : 63,360.

BULMAN, o. M. B. 1931. South American graptolites, with special reference to the Nordenskiold collec-
tion. Ark. Zool. 22a, no. 3, 1-111.

CHAPMAN, E. 1903. New or little-known Victorian fossils in the National Museum, Melbourne. Pt. 1,
Some Paleozoic species. Proc. R. Soc. Viet. 15, n.s., 104-22.

CHURKIN, M. jr. and brabb, e. e. 1965. Ordovician, Silurian, and Devonian biostratigraphy of east-
central Alaska. Bull. Am. H55. Petrol. Geol. 49, 172-85.

ELLES, G. L. and WOOD, E. M. R. 1901-18. A monograph of British graptolites. Palaeontogr. Soc.
[Monogr. ]

GURLEY, r. r. 1896. North American graptolites; new species and vertical range. /. Geol. 4, 63-102,
291-311.

JONES, T. and woodward, h. 1888-99. A monograph of the British Palaeozoic Phyllopoda (Phyllo-
carida, Packard). Palaeontogr. Soc. [Monogr.]

merriam, c. w. 1963. Paleozoic rocks of Antelope Valley, Eureka and Nye counties, Nevada. Prof. pap.
U.S. geol. Surv. 423.

mertie, j. b. jr. 1930. Geology of the Eagle-Circle district, Alaska. Bull. U.S. geol. Surv. 816,

NOLAN, T. B., MERRIAM, C.W., and WILLIAMS, J. s. 1956. The stratigraphic section in the vicinity of Eureka,
Nevada. Prof. pap. U.S. geol. Surv. 276.

RUEDEMANN, R. 1921. Paleontologic contributions from the New York State Museum. Note on Caryo-
caris Salter. Bull. N.Y. St. Mus. 227-8, 95-100.

1934. Paleozoic plankton of North America. Mem. geol. Soc. Am. 2.

1935. Silurian phyllocarid crustaceans from Oklahoma. J. Paleont. 9, 447-8.

SALTER, J. w. 1863. Note on the Skiddaw Slate fossils. Q. J. geol. Soc. Loud. 19, 135-40.

ST0RMER, L. 1937. Planktonic crustaceans from the Lower Didymograptus Shale (3b) of Oslo. Norsk
geol. Tidsskr. 15, 267-78.

THEOKRiTOFF, G. 1964. Taconic stratigraphy in northern Washington County, New York. Bull. geol.
Soc. Am. 75, 171-90.

MICHAEL CHURKIN, Jr.

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

Manuscript received 9 June 1965

DEVONIAN SCHIZOPHORIID BRACHIOPODS
FROM WESTERN EUROPE

by YVONNE P. POCOCK

Abstract. Six species and one subspecies of the genus Schizophoria are described from the Devonian of western
Europe, with their stratigraphical ranges and postulated phylogeny. All taxa are shown to be externally and
internally distinct. Neotypes are proposed for Schizophoria providvaria (Maurer) and S. strigosa (Sowerby).

Specimens of Schizophoria have been studied in this work from the Lower to Upper
Devonian of the southern border of the Dinant basin, the Upper Devonian of Bou-
lonnais; the Lower Devonian of the Rheinischen Schiefergebirge, and the Middle
Devonian of the Eifel. The stratigraphical succession of the Dinant basin is given by
Maillieux (1922, 1941), together with stratigraphical correlations with the Rheinischen
Schiefergebirge (1922). The succession of the Boulonnais inlier is presented by Pruvost
(1924), and that of the Eifel by Struve (1963).

The stratigraphical distribution of the species studied from these regions is shown on
text-fig. 1.

Abbreviations. In the following descriptions, relevant museum collections listed are as follows: BC —
Bedford College, University of London; BM — British Museum (Natural History); GSM — Geological
Survey Museum (London); GMUS — Geology Museum, University of Saskatchewan; HMUG —
Hunterian Museum, University of Glasgow; IRSN — Institut royal des sciences naturelles de Belgique;
MNB — Museum fiir Naturkunde, Berlin ( Haupt-Sammlung) ; SMF — Senckenberg Museum, Frankfurt.

In each text-fig. of serial sections, the numbers represent distances in millimetres
measured anteriorly from the umbones. Muscle field patterns and vascular markings are
also illustrated on figures of internal moulds, since moulds, rather than discrete valves,
are the common form of preservation.

SYSTEMATIC DESCRIPTIONS

Suborder dalmanelloidea Moore 1952
Eamily schizophoriidae Schuchert and Ee Vene 1929
Subfamily schizophoriinae Schuchert and Le Vene 1929
Genus schizophoria King 1850

Type species. Concbyliolitinis Anomites resupinatus Martin 1809.

Outline transversely rectangular to elliptical, quadrate to rounded, ventribiconvex to
biconvex to dorsibiconvex, the dorsal valve generally deeper in adult forms. Ventral
valve convex umbonally, flattening laterally, depressed medially. Dorsal valve evenly
convex longitudinally, or greatest convexity umbonally, flattening laterally and an-
teriorly. Hinge-line straight, submegathyrid; cardinal angles rounded. Beaks small,
pointed, incurved; brachial beak more incurved. Interareas curved to beaks, pedicle
interarea higher; delthyrium and notothyrium open. Anterior commissure varying from

[Palaeontology, Vol. 9, Part 3, 1966, pp. 381-412, pi. 66.]

382

PALAEONTOLOGY, VOLUME 9

rectimarginate to uniplicate, unisulcate, sulciplicate, biplicate. Dorsibiconvexity and
height of anterior plication increase with age. Ventral sulcus frequently developed;
dorsal fold occasionally developed adjacent to anterior commissure. Shell costellate,
rugate, punctate. Costellae separated by narrower, more angular striae, costellae in-
creasing by bifurcation and intercalation. Growth rugae variably developed with age.

q:

S. ANTIOUA SOLLE

SPROVULVARIA (MAURER)

FAMENNIAN

FRASNiAN

CIVETIAN

EIFELIAN

S.PYCMAEA STRUVE

s.striatula(schlotheim)

S.SCHNURI BLANKENHEIMENSIS STRUVE

s.stricosaCsowerby)

s.vulvariaCquenstedt)

BELGIUM

and

GERMANY

BELGIUM

and

FRANCE

GERMANY

PYCMAEA

holotype

BLANKENHEIMEN5IS
! holorype

I GERMANY
I BLANKENHEIMENSIS

BELGIUM

and

GERMANY

holotype
I GERMANY

SIECENIAN

BELGIUM

and

GERMANY

neotype

BELGIUM

and

GERMANY

holotype

BELGIUM

ond

GERMANY

neotype

GERMANY

I PROBABLE CONNECTING RANGES

TEXT-FIG. 1. Stratigraphical range of species of Schizophoria from Belgium, NE. France, and Germany.

and between species, concentrated anteriorly and laterally. Puncta subrounded, con-
centrated along striae on shell surface, concentrated along costellae in lower shell
layers, evenly distributed in inner shell layers.

Ventral muscle field parallel-sided or flabellate, bounded posteriorly by dental lamellae
supporting compound teeth, laterally and anteriorly by ridge-like extensions of lamellae, j

Ridges decreasing in height anteriorly, reflexed, uniting with end of median septum
generally to form anterior re-entrant. Diductor muscle field longitudinally divided by i
median septum, originating near apex of delthyrial cavity. Adductor muscles attached to , I

Y. P. POCOCK: DEVONIAN SCHIZOPHORIID BRACHIOPODS 383

median septum. Two subparallel vascula media originating from anterior of muscle
field. Genital markings developed laterally and postero-laterally. Shell partially filling
delthyrial cavity, decreasing in thickness and disappearing anteriorly.

Dorsal valve with simple or compound serrated myophore. Compound form con-
sisting of central ridge bounded by two to four shorter, narrower ridges (one or two
either side). Myophore bounded by divergent or curved brachiophore plates supporting
stubby brachiopores. Shell partially filling notothyrial cavity, decreasing in thickness and
disappearing anteriorly. Ventral teeth articulating with dorsal sockets. Sockets oval in
transverse section, bounded internally by smaller, shallower accessory sockets, and ex-
ternally by larger, irregularly shaped accessory cavities. Dorsal muscle field generally
one-third to one-half valve length, quadripartite, bounded posteriorly by brachiophores,
brachiophore plates, laterally and anteriorly by accessory ridges. Ridges decreasing in
height anteriorly, reflexed to form anterior re-entrant, uniting with median septum.
Median septum originating at base of notothyrial cavity. Minor septum frequently
dividing each half of adductor muscle field into pyriform anterior scar and digitate or
tripartite posterior scar. Four subparallel vascula media originating from anterior of
muscle field. Two vascula myaria occasionally developed laterally to vascula media from
ends of minor septa. Genital markings developed laterally and postero-laterally.

Schizophoria antiqua Solle
Plate 66, figs. \a, b\ text-figs. 2-5

1907 Orthis (Schizophoria) striatiila Schlotheim; Walther, p. 279, pi. 13, fig. 9.

1910 Orthis striatula Schlotheim; Assman, p. 161, pi. 9, figs. 1, 2.

1916 Orthis striatula Schlotheim; Victor, p. 452, pi. 18, fig. 10.

1936 Schizophoria antiqua Solle, p. 208, figs. 14, 15.

Type. The holotype. Nr. XVII 533«, and specimen Nr. XVII 533Z> are deposited in the Senckenberg
Museum, Frankfurt.

Diagnosis. Medium to small, rectangular to elliptical, rugate shell, dorsibiconvex in
adult form. Ventral muscle field strongly incised, flabellate, with broad, rounded median
septum. Dorsal muscle field longitudinally elliptical, bounded posteriorly by curved
brachiophore plates.

Description. Shell medium to small, ventribiconvex to dorsibiconvex, rectangular to
elliptical, with greatest shell width at mid-length. Ventral sulcus ill-defined, originating
near anterior border. Low, broad, subrounded anterior uniplication. Costellae coarse, 4
to 5 in 1 mm. at 10 mm. from beaks. Prominent growth rugae.

Teeth compound, supported by anteriorly divergent ventrally subparallel to divergent
dental lamellae (text-fig. 2, sections 1.8-3. 8). Ventral muscle field (text-fig. 3a) one-third
to one-half valve length, flabellate, strongly incised. Deep, broad, rounded anterior re-
entrant. Median septum generally prominent, rounded, broadening and increasing in
height, and becoming flat-topped anteriorly (text-fig. 2, sections 1.1-9. 5). Two slightly
divergent vascula media. Genital markings developed postero-laterally (text-fig. 3a).

Myophore small, simple or rudimentarily compound, with central ridge bordered by
two lateral ridges. Stubby brachiophores fused to strong, curved brachiophore plates
(text-fig. 2, sections 1.4-3. 8). Dental sockets oval, articulating with ventral teeth
(text-fig. 2, sections 3.4, 3.5). Dorsal muscle field (text-fig. 3b) moderately incised.

384

PALAEONTOLOGY, VOLUME 9

3 8

TEXT-FIG. 2. Schizopboria antiqiia Solle. Transverse serial sections (IRSN IG6154).

Y. P. POCOCK: DEVONIAN SCHIZOPHORIID BRACHIOPODS 385

longitudinally elliptical, with greatest width anteriorly, one-third to one-half valve length.
Accessory ridges continuous with brachiophore plates, smoothly reflexed anteriorly.
Median septum angular, increasing in height and broadening very slightly anteriorly
(text-fig. 2, sections 0.8-5. 5). Anterior adductor muscle scar pyriform; posterior muscle
scar possibly digitate. Four weakly divergent vascula media, and pair of divergent vas-
cula myaria (text-fig. 2>b). Genital markings developed postero-laterally.

Dimensions. External dimensions and muscle field dimensions are plotted on text-fig. 4.

Q VENTRAL MUSCLE FIELD b DORSAL MUSCLE FIELD

TEXT-FIG. 3. Schizophoria antiqiia Solle. Ventral and dorsal muscle fields.

Remarks. Although Solle described Schizophoria antiqua from the Lower Devonian
(Emsian) of Germany, the bulk of the material examined is deposited in the Institut
royal des sciences naturelles de Belgique, and was collected from the Frasnian of the
Dinant Basin. Two additional specimens from the Museum fiir Naturkunde, Berlin,
collected from the Gedinnian of the Taunus region, extends the range of the species. This
medium to small, tumid, rugate form of Schizophoria is distinct from other Devonian
species (text-fig. 5).

Schizophoria antiqua closely resembles S. woodi Bond of the Carboniferous in outline,
tumidity, prominent growth rugae, curved brachiophore plates, and flabellate pedicle
muscle field. There is a closer resemblance in size with the smaller form of S. woodi from
the Treak Cliff, Cracoe, and Craven areas of reef limestone. But S. antiqua is more
coarsely costellate, lacks spine bases, and the brachial muscle field is more elliptical in
outline.

Youthful forms of S. antiqua resemble S. connivens (Phillips) of the Carboniferous, in
rectangular outline and coarse costellae, but internally there are distinct differences. The
flabellate ventral muscle field, and broad median septum, contrast with the less flabellate,
elliptical form and narrower septum of S. connivens. The elliptical dorsal muscle field,
curved brachiophore plates, and six pallial sinus trunks of S. antiqua contrast with the
elliptical to rounded muscle field, divergent brachiophore plates, and four pallial sinus
trunks of Y. connivens.

386

PALAEONTOLOGY, VOLUME 9

Material. Belgium: Frasnian, Assise de Frasnes (F2)-F2d (IRSN IG5911, 6154, 8439, 9179), F2h
(IRSN IG4591, 6418), F2i (IRSN IG8701), Dinant basin. Germany: Gedinnian, Hobracher Schichten
(MNB), Taunus; Lower Emsian (MNB), Villmar. Middle Devonian (MNB), Boppard; Lower Fras-
nian (BC Bl-9), Paffrather Syncline, near Cologne; Upper Middle Devonian (SMF), Villmar.

S.ANTIOUA SOLLE

WIDTH

8

7 h

6

5

J ■*

o

z

3

2

. © holotype

X venfrol muscle field
• dorsol muscle field

1 > I 11 1 I )

WIDTH

01 234 5678 mm.

WIDTH

TEXT-FIG. 4. Dimensions of Schizophoria antiqiia Solle.

Schizophoria provulvaria Maurer

Plate 66, figs. 2, 3, 10; text-figs. 6-9

1864-5 Ortliis hipparionix Vanuxem (?); Davidson, p. 90, pi. 17, figs. 9, 10? non 8, 11.

1886 Orthis provulvaria Maurer, p. 21.

1890 Orthis personata Zeiler; Kayser, pi. 12, fig. 3.

1893 Orthis provulvaria Maurer; Maurer, p. 7, pi. 3, figs. 1-4.

1904 Orthis {Schizophoria) provulvaria Maurer; Drevermann, p. 267, pi. 30, figs. 29? 30; pi. 31,
figs. 11-19 (11? 16? 18?).

1936 Schizophoria provulvaria Maurer; Termier and Termier, p. 1126, pi. 3, figs. 3, 4: 1950,
pi. 71, figs. 10, 11?; pi. 72, figs. 12? 13?

1938 Schizophoria provulvaria (Maurer); Shirley, p. 465, pi. 4, figs. 10-13.

1942 Schizophoria provulvaria (Maurer); Gill, p. 36, pi. 6, fig. 1.

C 4179

TEXT-FIG. 5. Comparisons of Devonian species of Schizophoria.

I

388

PALAEONTOLOGY, VOLUME 9

TEXT-FIG. 6. Schizophoria provulvaria (Maurer). Transverse serial sections of plaster internal moulds.
A— ventral valve (HMUG L5345/2), b— dorsal valve (HMUG L5341/2).

Il

Y. P. POCOCK: DEVONIAN SCHIZOPHORIID BRACHIOPODS

389

Type. Maurer’s specimens (1886, 1893) cannot be traced. These were collected from the Lower
Devonian of Seifen, Germany. A neotype has been selected, BM B42943, a dorsal internal mould. This
was also collected from the Lower Devonian of Seifen, Dierdorf.

Diagnosis. Shell large, elliptical. Ventral muscle field broad, flabellate, strongly incised,
with broad, rounded median septum. Brachiophores-brachiophore plates thick, sub-
subparallel to divergent. Short peripheral follicular markings.

Description. Shell mould large, dorsibiconvex, elliptical in outline, with greatest width
at or slightly anterior to midlength. Anterior commissure rounded uniplicate.

a VENTRAL MUSCLE EIELD b DORSAL MUSCLE FIELD

BMB42935 Internal moulds BM B42943

BMB42943 > — BMB42936

(neotypej

TEXT-FIG. 7. Schizophoria provulvaria (Maurer). Ventral and dorsal muscle fields.

Ventral muscle field (text-fig. la) one-half to two-thirds valve length, broad, flabellate,
strongly incised. Subrounded anterior re-entrant. Median septum narrow, rounded,
rapidly broadening and increasing in height anteriorly (text-fig. 6a, sections 6.2-15.4).
Two parallel vascula media originating in anterior re-entrant of muscle field. Genital
markings developed laterally and postero-laterally (text-fig. la).

Prominent cardinal process; broad, compound myophore, narrow shaft (text-fig. 8).
Stubby brachiophores fused to strong, thick, subparallel to divergent brachiophore
plates (text-fig. 6b, sections 3. 6-5. 6). Dorsal muscle field (text-fig. lb) moderately incised,
transversely rectangular to elliptical, one-half valve length. Low accessory ridges
smoothly reflexed anteriorly to form shallow, sub-rounded anterior re-entrant. Median
septum low, broad, rounded, narrowing anteriorly (text-fig. 6b, sections 3.8-7. 1). An-
terior adductor muscle scar pyriform ; posterior muscle scar more incised, digitate, with
slightly longer inner lobe (text-fig. lb). Two vascula media, each bifurcating (text-fig.
lb). Short follicular markings developed peripherally.

Dimensions. Dimensions of available muscle fields are plotted on text-fig. 9.

Remarks. Apart from a few doubtful specimens assigned to S', provulvaria preserved as
fragmentary external moulds and illustrating a coarsely costellate shell, all specimens are
preserved as fragmentary internal moulds. Serial sections of plaster internal moulds

390

PALAEONTOLOGY, VOLUME 9

DORSAL VALVE

plosticine cast

25

X

X

X X
X

X

X

X

• X

X

X

X

© neotype

S. PROVULVARIA (MAURER)

X ventral muscle field
• dorsol muscle field

I I I I 1 I I I I 1 I I I 1 I iJ

5 lO 15 20 mm.

WIDTH

TEXT-FIG. 9. Dimensions of muscle fields of Schizophoria provulvaria (Maurer).

EXPLANATION OF PLATE 66

Fig. 1 . Schizophoria antiqua Solle. 1 a, b. Ventral and dorsal views of internal mould, MNB ( X 1 ^).

Figs. 2, 3, 10. Schizophoria provulvaria (Maurer). 2, Internal mould of dorsal valve, neotype, BM
B42936 (x 1). 3, 10, Internal moulds of dorsal valves, MNB (x 1).

Fig. 4. Schizophoria pygmaea Struve. 4u, b. Dorsal and anterior views, BC B55 (x2).

Fig. 5. Schizophoria schniiri blankenheimensis. 5a-c, Dorsal, ventral, and lateral views, BC B68 (x 1).

Figs. 6, 7. Schizophoria striatula (Schlotheim). 6, Ventral view, BC B90 (xl). 7, Lateral view, BC
BIOS (xl).

Figs. 8, 9, 11, 12. Schizophoria strigosa (Sowerby). 8, Internal mould of dorsal valve, neotype, MNB
B102.1 (xl). 9, 11, Internal moulds of dorsal valves, MNB (xl). 12, Internal mould of dorsal
valve, IRSN IG8219 (xl).

Fig. 13. Schizophoria vulvaria (Quenstedt). 13o, b, Ventral and dorsal views of internal mould, BM
B62947 (X 1).

Palaeontology, Vol. 9

PLATE 66

V’

POCOCK, Devonian schizophoriid brachiopods

^ A.

Y. P. POCOCK: DEVONIAN SCHIZOPHORIID BRACHIOPODS 391

(Stanley, 1964) show the general appearance of internal structures. Only discrete valves
were available, showing no ventral-dorsal valve relationship (text-fig. 6).

S. provulvaria (Maurer) superficially resembles S. strigosa (Sowerby). Comparisons
are listed under S. strigosa (see text-fig. 5). S. provulvaria is distinguished externally from
S. vulvaria (Quenstedt) by its wider outline. Internally, the flabellate ventral muscle
field, broad median septum, and deep anterior re-entrant, contrast with the longer,
lanceolate to weakly flabellate muscle field, narrow septum, and shallow, or lack of re-
entrant of S. vulvaria. The moderately incised, rectangular dorsal muscle field, thick
brachiophores and parallel-to-divergent brachiophore plates, and bipartite posterior
adductor muscle scars of S. provulvaria contrast with the more quadrate, strongly in-
cised muscle field, thinner brachiophores and divergent brachiophore plates, and
commonly tripartite posterior muscle scar of S. vulvaria.

Davidson’s use of Or this hipparionix (1864-5 p. 90) for his specimens resembling
S. provulvaria is invalid. He did state that his large internal moulds resembled Or this
hipparionyx of American authors, but could not be certain as to their identification. The
genus Hipparionyx was established by Vanuxem in 1842, and is synonymous with the
genus Streptorhynchus King. In 1853, Schnur, working in the Eifel, discovered specimens
with a similar flabellate ventral muscle field, which he considered belonged to the genus
Orthis, and changed Vanuxen’s nomenclature to Orthis hipparionyx. But Schnur’s speci-
mens are orthotetid brachiopods. Davidson (1864-5) presumably recognized the
flabellate ventral muscle field of his specimens and listed them in synonymy with
Schnur’s Orthis hipparionyx.

Material. Belgium, Dinant Basin: Siegenien, Gres d’Anor, Sg 2 (IRSN IG 12533); Grauwacke de
Saint-Michel, Sg 3 (IRSN IG5382, 5746, 8219); Grauwacke Inferieur de Laroche, Sg 3III (IRSN IG9382).
Quartzophyllades de Saint-Vith, Sg 5III (IRSN IG8633); Emsien Inferieur, Grauwacke de Pesche, Em
la (IRSN IG8791); Gres de Mormont, Em Ig (IRSN IG8390). Germany: Lower Devonian (BM
B24563, B24565, B42935, 6, B42942, 3, B42945, 6, B49,920, HMUG L5341/4), Seifen, Dierdorf.
Siegener Schichten (HMUG L5345/1, 2, SME), Seifen. Siegener Schichten, Rauhflaserschichten
(MNB), Seifen. South-west England: Lower Devonian (GSM 49692), New Drive above Hope’s Nose,
Torquay.

Schizophoria pygmaea Struve
Plate 66, figs. 4 a, b; text-figs. 10, 12, 15
1 963fl Schizophoria pygmaea Struve, p. 251, pi. 39, 40.

(See Struve for synonymy.)

Schizophoria schnur i blankenheimensis Struve 1965

Plate 66, figs. 5a~c; text-figs. 11, 13-15

1853 Orthis striatida d'Orbigny; Schnur, pi. 38, fig. 1 e-gl

1942 Schizophoria excisa (Quenstedt); Spriesterbach, p. 182, pi. 5, figs. 9-14.

Types. Struve (1963o) deposited the holotype, SME 17298, and paratypes of Schizophoria pygmaea
in the Senckenberg Museum, Erankfurt. The holotype of Schizophoria schnuri blankenheimensis,
SME 19559, and paratypes, are also deposited in the Senckenberg Museum.

Diagnosis. Shell small {S. pygmaea) to large (,5. schnuri blankenheimensis) rectangular to
elliptical, weakly dorsibiconvex, with prominent ventral sulcus. Ventral muscle field

392 PALAEONTOLOGY, VOLUME 9

flabellate, strongly incised, with broad, rounded median septum. Strong, curved
brachiophore plates.

Description. Shell small to large, weakly dorsibiconvex, thin, rectangular to elliptical,
with greatest width at or anterior to mid-length. Ventral sulcus well defined, originating

lO

TEXT-FIG. 10. Schizophoria pygmaea Struve. Transverse serial sections (BC B58).

below umbo, flaring and deepening anteriorly. Gentle dorsal fold developed in older
specimens. Anterior commissure rounded uniplicate. Costellae fine, 5 to 6 in 1 mm. at
10 mm. from beaks. Spine bases developed at anterior ends of scattered costellae.
Growth rugae developed on older specimens.

Teeth prominent, compound, supported by anteriorly and ventrally divergent dental
lamellae (text-figs. 10, sections O.T-1.75; 11, sections 2.2-5.0). Articulation supple-
mented by interlocking ends of brachiophores and dental lamellae (text-fig. 10, sections

Y. P. POCOCK: DEVONIAN SCHIZOPHORIID BRACHIOPODS

393

TEXT-FIG. 11. Schizophoria schmiri hlankenheimensis. Transverse serial sections (BC B70).

394

PALAEONTOLOGY, VOLUME 9

1.4-.1.6). Ventral muscle field (text-figs. 12, 13) one-third to one-half valve length,
flabellate, strongly incised. Deep, subrounded anterior re-entrant. Median septum
rounded, broadening and increasing in height, and becoming flat-topped anteriorly
(text-figs. 10, sections 0.4-6.0; 11, sections 1.1-13.0). Two vascula media; genital mark-
ings developed postero-laterally (text-fig. 13).

a VENTRAL MUSCLE FIELD b DORSAL MUSCLE FIELD

TEXT-FIG. 12. Schizophoria pygmaea Struve. Ventral and dorsal muscle fields.

a VENTRAL MUSCLE FIELD b DORSAL MUSCLE FIELD

Myophore prominent, compound, average width 2-3 mm., with central ridge bordered
by four lateral ridges, all finely serrated (text-fig. 14). Stubby brachiophores fused to
strong curved brachiophore plates (text-figs. 10, sections 0.4-1. 5; 11, sections 1.1-5. 3).
Brachiophore plates thickened posteriorly by shell filling notothyrial cavity (text-figs. 10,
sections 0.5-1. 5; 11, sections 2.2-4.2). Deep, oval dental sockets articulating with ven-
tral teeth (text-figs. 10, sections 1.0-1. 3; 11, sections 2. 8-3. 8). Dorsal muscle field (text-
figs. 12, 13) moderately incised, rectangular to rounded. Accessory ridges smoothly
reflexed anteriorly to form deep subrounded re-entrant. Median septum rounded,

1

Y. P. POCOCK: DEVONIAN SCHIZOPHORIID BRACHIOPODS 395

decreasing in width and becoming sharp-crested anteriorly (text-figs. 10, sections 0.4-5.2;
11, sections 1. 1-7.0). Anterior adductor muscle scar pyriform, posterior muscle scar
digitate, both parts of similar length (text-figs. 12, 13). Two parallel vascula media, both
bifurcating. Genital markings developed laterally and postero-laterally (text-fig. 13).

I mm

TEXT-FIG. 14. Schizophoria schtmri blankenheimensis, Myophore structure.

Dimensions. External dimensions and dimensions of muscle fields are plotted on text-fig. 15.

Remarks. Struve (1963u, p. 25.1) described and illustrated a small, relatively thin form of
Schizophoria, S. pygmaea, from the Eifelian Hundsdell, Bildstock, and Flesten Horizons
of the Eifel region. Spriesterbach (1942, p. 182) previously described and illustrated
a large, relatively thin form of Schizophoria from the Middle Devonian of the Blanken-
heim region, also in the Eifel, which he listed in synonymy with S. excisa (Quenstedt).
But S. excisa (Quenstedt), (Quenstedt 1868-71, p. 561, pi. 55, figs. 138-46) is a separate,
distinct form and is synonymous with another species, S. striatula (Schlotheim 1813,
p. 8, pi. 1, fig. 6; 1820, p. 254, pi. 5, fig. 4). Struve recognized the similarity between S.
pygmaea and the specimens illustrated by Spriesterbach (1942 pi. 5, figs. 9-14), since he
listed the specimen of figure 14 in synonymy with his species. But he made no reference
at that time to the larger specimens illustrated on this plate. Figure 14 is probably a more
youthful, smaller specimen of the form illustrated by Spriesterbach.

396

PALAEONTOLOGY, VOLUME 9

Struve (1965 p. 204) has since described these large specimens under a new subspecies,
S. schnuri blankenheimensis. His new species, S. schnuri, is here considered to be a large
form of S. striatula, and S. schnuri blankenheimensis is considered to be closely related to
S. pygmaea Struve. S. pygmaea appears to be a dwarf form of S. schnuri blanken-
heimensis. Size is the only distinction. The largest (adult) specimens of S. pygmaea are

S. PYGMAEA STRUVE

S. SCHNURI BLANKENHEIMENSIS STRUVE

40
35
30
25
20
X 15

I—

o

X

^ 10
5

hoIo^ype

.pi' .

30
25
20 1
x'^ E

t—

Q_ *

Q 10 -

5 r

► S. PYGMAEA

iS. SCHNURI BLANKENHEIMENSIS

hoIob/pe-<p‘

X * , x»

tt ** JL *

* V *

• S- PYGMAEA

* 5. SCHNURI BLANKENHEIMENSIS

I I I 1 1 1 1 1 1 1 I I 1 1 ' I ■ 1 1 ■ ■ ■ ■ I . . I ' I . . . ■ I . . . . I . . . . I

5 10 15

20 25

WIDTH

30 35 40 45 50mm

■ I ■ ■ ■ ■ > ■ ■ ■ ■ 1 • » ■ > I < ■ ■ > I ■ < ■ » I ■ ■ ■ ■ I ■

ro 15 20

WIDTH

25 30 35 40 45 50mm

30 r
25
ilO

. S. PYGMAEA

*S. SCHNURI BLANKENHEIMENSIS
I ..-■I .... I , I .... I .... I .... I I I

S- PYGMAEA
• venTrol muscle field
X dorsol muscle held
S SCHNURI BLANKENHEIMENSIS
o ventral muscle fidd
tdorsol muscle field

-1— 1 1 I I » 1 ■ ■

10 15 20 25 30 35 40 45 50mm

WIDTH

\ \ I I ■ 1

lO I5mm

TEXT-FIG. 15. Dimensions of Schizophoria pygmaea Struve, and S. schnuri blankenheimensis.

comparable with some of the smallest (youthful) specimens of S. schnuri blanken-
heimensis (text-fig. 15). The two forms are similar externally and internally (cf. text-
figs. 10, 11 and 12, 13).

The dwarf species, S. pygmaea, occurs in the Hundsdell, Bildstock, and Flesten Hori-
zons, where other brachiopods are smaller than normal. This dwarf form is preceded
and succeeded stratigraphically by the larger form of S. sehnuri blankenheimefisis.
Struve (1965) makes no reference to the similarity between S. pygmaea and S. schnuri
blankenheimensis.

Schizophoria schnuri blankenheimensis superficially resembles S. provulvaria (Maurer)
in the flabellate ventral muscle field and strong median septum, and the dorsal muscle
field and four parallel vascula media. However, externally, S. schnuri blankenheimensis is

Y, P. POCOCK: DEVONIAN SCHIZOPHORIID BRACHIOPODS 397

generally less convex, and has a well-defined ventral sulcus. Specimen BC B64, from the
lowermost Middle Devonian (Wolfenbach Horizon) is more convex, but when sec-
tioned, illustrated the characteristic internal structures of S. schnuri blankenheimensis.
S. pygmaea resembles S. provulvaria in the same manner, but is much smaller in size.

Schizophoria schnuri blankenheimensis is distinguished from S. striatula (Schlotheim)
by its more rectangular outline and weaker convexity. Schizophoria striatula is generally
more quadrate to rounded, with greatest shell width situated anteriorly. Internally there
are muscle field differences (text-figs. 13, 17). The strongly fiabellate ventral muscle field,
deep re-entrant, broad median septum, and widely separated, slightly divergent vascula
media of S. schnuri blankenheimensis contrast with the less fiabellate to elongate
elliptical muscle field, narrower re-entrant and median septum, and closely spaced,
parallel vascula media of S. striatula from the Middle Devonian. In the dorsal valve, S'.
striatula has a longer inner portion to the posterior adductor muscle scar, divergent
pallial sinus trunks, and shorter genital markings. The divergent brachiophore plates also
contrast with the curved plates of S. schnuri blankenheimensis.

Material. Germany, Eifel region: Eifelian, S. pygmaea — Ahrdorf Beds, Bildstock Horizon, MTB
Dollendorf (BC BIO). Same stratigraphical level, MTB Dollendorf (BC Bll). Ahrdorf Beds, Flesten
Horizon, MTB Dollendorf (BC B12, 13). Same stratigraphical level, MTB Dollendorf (BC B14, 15).
Same stratigraphical level, Ahrdorf Syncline, MTB Dollendorf (BC B 16-59). Ahrdorf Beds, MTB
Mechernich (BC B60-63). Nohn-Ahrdorf Beds, Hundsdell-Bildstock Horizon, Sotenicher Syncline,
MTB Mechernich (SMF 17267). Schwirzheim Horizon, Gerolstein (SMF 17278). Ahrdorf Beds,
Bildstock Horizon, Hillesheimer Syncline, MTB Dollendorf (SMF 17286). S. schnuri blankenheimensis
— Lauch Beds, Wolfenbach Horizon, MTB Dollendorf (BC B64). Junkerberg Beds, Blankenheim Rail-
way Cutting (BC B65-67, 69-71). Same stratigraphical level, MTB Blankenheim (BC B72-82). Middle
Devonian, Gerolstein (BM B42948). Upper Junkerberg Beds, Blankenheim Railway Cutting (GMUS
Eu DE 77(5). Lower and Upper Middle Devonian, Blankenheim (MNB). Middle Devonian, Gerol-
stein (MNB).

Schizophoria striatula (Schlotheim)

Plate 66, figs. 6-7; text-figs. 16-20

1777 Terebratiilae miniitissime striatae; Schroter, p. 390, pi. 4, fig. 24.

1813 Anomia terebratulites striatulus Schlotheim; Leonhard, p. 8, pi. 1, fig. 6.

1820 Terebratulites striatulus-, Schlotheim, p. 254, pi. 15, fig. 4.

1841 Orthis resupinata-. Phillips, pi. 27, fig. 115.

1842-4 Orthis striatula-, De Koninck, p. 224, pi. 13^ fig. 6; non. pi. 13, fig. 11.

1850- 6 Orthis striatula-, Sandberger and Sandberger, p. 355, pi. 34, fig. 4.

1851- 5 Orthis striatula, Schlotheim; Davidson, pi. 7, figs. 128-33.

1853 Spirifer striatulus Schlotheim sp.; Geinitz, p. 61, pi. 15, figs. 10-12.

1853 Orthis striatula d’Orbigny; Schnur, p. 215, pi. 38, fig. \a-d, h-i; e-gl
1860 Orthis striatula Schlotheim; Griinewaldt, p. 87, pi. 2, fig. 6.

1864-5 Orthis striatula Schlotheim; Davidson, p. 87, pi. 17, figs. 4-7.

1868-71 Orthis excisa-, Quenstedt, p. 561, pi. 55, figs. 138-45.

1908 Orthis (Sch) striatula (Schlotheim); Cowper-Reed, p. 79, pi. 13, figs. 19-24.

1922 Orthis (Schl) striatula (Schlotheim); Cowper-Reed, p. 34, pi. 6, figs. 12, 13.

1930 Orthis (Schizophoria) resupinata var. striatula (Schloth.); Paeckelmann, p. 158, pi. 9,
figs. 3-10.

1959 Schizophoria striatula (Schlotheim); Biernat, p. 54, pis. 7-9; pi. 10, fig. 3.
non 1907 Orthis (Schizophoria) striatula Schlotheim; Walther, p. 279, pi. 13, fig. 9.
non 1916 Orthis striatula Schl; Victor, p. 452, pi. 18, fig. 10.

non 1932 Schizophoria aff. striatula (Schlotheim); Schuchert and Cooper, pi. 23, figs. 22-25.

398

PALAEONTOLOGY, VOLUME 9

TEXT-FIG. 16. Schizophoria striatula (Schlotheim). Transverse serial sections (BC B94).

Y. P. POCOCK: DEVONIAN SCHIZOPHORIID BRACHIOPODS

399

Type. Schlotheim’s holotype is deposited in the Museum fiir Naturkunde, Berlin.

Diagnosis. Shell medium to large, elliptical to quadrate, with prominent anterior dorsal
fold in older specimens. Ventral muscle field oval to flabellate. Strong divergent brachio-
phore plates.

Description. Shell medium to large, dorsibiconvex, quadrate to elliptical, with greatest
shell width generally anterior to mid-length. Ventral sulcus originating half way along
valve, broadening and deepening anteriorly. Dorsal fold developed in older specimens.
Anterior commissure rounded uniplicate. Costellae moderately coarse, 4 to 5 in 1 mm.
at 10 mm. from beaks. Growth rugae prominent on older specimens.

a VENTRAL MUSCLE FIELD b DORSAL MUSCLE FIELD

delthyrial cavity
dental lamello

loterol ridge

oddoctor-diductor
muscle field

genitol morkings

posterior

adductor muscle field

medion septum

onterior

odductor muscle field

vasculo media

BC B 127

j— notothynol cavity

brachiophorc

brachiophore plate
genital markings
minor septum
occessory ridge

anterior re-entront

internal moulds

TEXT-FIG. 17. Schizophoria striatula (Schlotheim). Ventral and dorsal muscle fields.

Teeth prominent, compound, supported by anteriorly divergent, ventrally parallel to
divergent dental lamellae (text-fig. 16, sections 1.7-3. 2). Articulation supplemented by
interlocking ends of brachiophores and dental lamellae. Shell partially filling delthyrial
cavity (text-fig. 16, sections 0.9-3. 4). Ventral muscle field (text-fig. \la) one-third to one-
half valve length, elliptical to flabellate, strongly incised. Rounded anterior re-entrant,
or re-entrant absent. Median septum rounded, broadening and increasing in height an-
teriorly (text-fig. 16, sections 0.4-9. 0). Two parallel vascula media, divergent anteriorly.
Genital markings developed postero-laterally (text-fig. \la).

Myophore prominent, compound, average width 3 mm., with central ridge generally
bordered by two lateral ridges, all coarsely serrated (text-fig. 18). Stubby brachiophores
fused to strong divergent brachiophore plates (text-fig. 16, sections 1.1-3. 4). Deep dental
sockets articulating with ventral teeth (text-fig. 16, sections 1.9-2. 2). Dorsal muscle field
incised, quadrate to rounded, one-third to one-half valve length. Accessory ridges
smoothly reflexed anteriorly to form moderately deep, rounded re-entrant. Median
septum angular to subrounded, broadening and increasing in height, then narrowing
anteriorly (text-fig. 16, sections 0. 9-7.0). Anterior adductor muscle scar pyriform, with
acute apex; posterior muscle scar digitate, with longer inner lobe (text-fig. 176). Two di-
vergent vascula media, each bifurcating. Two narrower vascula myaria with lateral

400

PALAEONTOLOGY, VOLUME 9

markings apparently developed, one either side main trunks. Genital markings developed
postero-laterally (text-fig. 11b).

Dimensions. External dimensions and dimensions of muscle fields are plotted on text-fig. 19.
Remarks. Variation in the form of the ventral muscle field is shown on text-fig. 20.

MYOPHORE STRUCTURE

Schizophoria striatula (Schlotheim) is a long-ranging species (Eifelian-Frasnian), and
shows little variation in morphology, except in size. Although specimens from the Middle
and Upper Devonian have been differentiated on text-fig. 19, they have comparable
dimensions. But many Upper Devonian specimens lack the characteristic dorsal an-
terior fold of the Middle Devonian forms, and have their greatest shell width at the mid-
length. Internally, the ventral muscle field of many Upper Devonian specimens is more
flabellate, with a broader septum. However, specimen BC B131 collected from the
Middle Devonian, although representing a minority, has a flabellate muscle field, and

Y. P. POCOCK: DEVONIAN SCHIZOPHORIID BRACHIOPODS

401

specimen GSM 34/20, an Upper Devonian form, is an exception, with an elliptical
muscle field (text-fig. 20).

Specimens of Schizophoria striatiila from the Geisdorf Horizon (Eifelian) of the Eifel
region are much larger. Other members of fauna at this level are also larger. Struve (1965,

S. STRIATULA (sCHLOTHEIM)

20

15 -

sA'.* .

• U. Devonlon
% M. Devonian

30 -

25 -

20

10

I I ■■ I I I I I I I . I I I I I 1 t I I I I I I I I 1 I ! 1 1 I I I 5

15 20 25 30 35 40 45mm

WIDTH

,«f 4

• U- Devonian
I M. Devonion

I ■ I ■ I ' I I I I I ■ ' '

I I I I ' 1 I ' I I I I I ' I I I 1 t I I I I I 1 I

lO 15 20 25 30 35 40 45mm

LENGTH

30

25

20

. O

- X

, , * i . • \ * X

• >* t

U. Devonlon
M. Devonion

I 1 1 1 1 1 1 ' 1 1 ' I ' 1 1 ' ' 1 1 1 1 1 1 1 1 , 1 1 1 1 , 1 1 , 1

15
14
13
12
n
10 ■
9
8

7 -

S b

i;

3 -

2

I

• ventral muscle field

X dorsal muscle field
-L- I I I I I I I I I I I I I

O I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 r
WIDTH

lO 15 20 25 30 35 40mm

WIDTH

TEXT-FIG. 19. Dimensions of Schizophoria striatiila (Schlotheim).

p. 202) described a new species, Schizophoria schnuri, from the Middle Devonian of
the Eifel, which also appears to be a large form of S. striatula. At the same time, he
established the subspecies S. schnuri schnuri, subexcisa,junkerbergiana, and biscissa. There
are no illustrations of internal structures, and externally the subspecies could represent
variation within S. striatula.

Schizophoria striatula superficially resembles S. resupinata (Martin) of the Carboni-
ferous in general outline and muscle fields. Early workers frequently considered them as
one species. But S. striatula is generally smaller, more quadrate, lacks a dorsal sulcus,
has a higher anterior plication, and lacks spine bases. Schizophoria resupinata is fre-
quently larger, rectangular to elliptical, with a rectimarginate to uniplicate to unisulcate

402

PALAEONTOLOGY, VOLUME 9

to sulciplicate anterior commissure, and is frequently covered in spine bases. The
elliptical to weakly flabellate ventral muscle field of S. striatula superficially resembles
that of S. resupinata, and the dorsal muscle fields are comparable.

Material. France, Boulonnais; Lower Frasnian, NE. end Carriere Parisienne (BC B83-89); Devonian,
Ferques (BM B19213, B82765, B82778); Devonian (BM B26,209). Belgium, Dinant Basin: Couvinien
Superieur (CO 2)— CO 2a (IRSN IG49 1 6, 6887, 8663), CO 2c ( IRSN IG4761 , 49 1 6, 5 1 27, 9694) ; Frasnien
Moyen (F2)— F2a (IRSN IG3031, 5911, 8254, 11.349), F2b (IRSN IG3349), F2e (IRSN IG4591),
F2i (IRSN IG3031, 2731, 4761, 5408). Germany, Eifel region: Eifelian-Lower Nohn beds, Weilersbach
Horizon, Hillesheimer Syncline, MTB Dollendorf (BC B90-92); Lower Nohn Beds, low Schleit

vanalion in ventral muscle field

TEXT-FIG. 20. Schizophoria striatula (Schlotheim). Variation in ventral muscle field.

Horizon, MTB Dollendorf (BC B93, 94); Upper Junkerberg Beds, Geisdorf Horizon, MTB Gerol-
stein (BC B95); same stratigraphical level, Priim Syncline (BC B96-98); same stratigraphical level,
MTB Mechernich (BC B99). Upper Junkerberg Beds to Upper Ereilingen Beds, MTB Gerolstein (BC
BlOO-12); Lower Freilingen Beds, MTB Mechernich (BC B1 13-17); same stratigraphical level, MTB
Miinstereifel (BC B1 18-28). Freilingen Beds, MTB Dollendorf (BC B129); Freilingen Beds, Eilenberg
Horizon, MTB Dollendorf (BC B130, 131); Givetian, Loogh Beds, Rech Horizon, Hillesheimer Syn-
cline (BC B132, 133); Middle Devonian, Gerolstein (BM B39562, 3); Middle Devonian (BM B62946,
B86023); Mitteldevon, ostiolatiis Horizon, Geisdorf (SMF); Mitteldevon, Gerolstein (SME). South-
west England: Upper Devonian (probably Petherwin Beds, Gatehouse Quarry), Petherwin (GSM
34/20).

Schizophoria strigosa (Sowerby)

Plate 66, figs. 8, 9, 11, 12; text-figs. 21, 22

1842 Orthis? strigosa, Sowerby, p. 409, pi. 38, fig. 7.

1886 Orthis ■, Maurer, p. 18.

1887 Orthis strigosa Sowerby; Bedard, p. 88, pi. 4, figs. 15, 16.

1893 Orthis occulta Maurer, p. 9, pi. 3, figs. 5-9.

Y. P. POCOCK: DEVONIAN SCHIZOPHORIID BRACHIOPODS

403

non 1871 Orthis strigosa Quenstedt, pi. 56, figs. 55, 56.
non 1890 Orthis personata Zeiler; Kayser, pi. 2, figs. 3-6.

Type. Sowerby’s specimen (1842) cannot be traced. This was collected from Devonian rocks of the
Dill synclinorium. A neotype, a dorsal internal mould, has been selected from the Museum fiir Natur-
kunde, Berlin. This was collected from the Lower Devonian, Rauhflaserschichten, at Seifen.

Diagnosis. Shell medium to large, quadrate to elliptical. Ventral muscle field moderately
long, flabellate, strongly incised, with narrow to broad, rounded median septum.
Moderately thin brachiophores and brachiophore plates. Long peripheral follicular
markings.

a VENTRAL MUSCLE FIELD b DORSAL MUSCLE FIELD

I cm

TEXT-FIG. 21. Schizophoria strigosa (Sowerby). Ventral and dorsal muscle fields.

Description. All specimens examined are preserved as internal moulds. Internal mould
medium to large, dorsibinconvex, quadrate to elliptical, with greatest width at mid-
length. Anterior commissure rounded uniplicate.

Ventral muscle field (text-fig. 21a) one-half to two-thirds valve length, flabellate,
strongly incised. Shallow anterior re-entrant, or re-entrant absent (text-fig. 21a).
Median septum varying in width, rounded, broadening, and increasing in height
anteriorly.

Cardinal process differentiated into oval myophore supported by narrower shaft.
Myophore compound, with up to twelve lateral ridges. Dorsal muscle field (text-fig.
2\b) incised, quadrate to rounded. Accessory ridges smoothly reflexed anteriorly to form
shallow, sub-rounded anterior re-entrant. Median septum narrow, subangular, narrow-
ing and decreasing in height anteriorly. Anterior adductor muscle scar pyriform ; pos-
terior adductor muscle scar bipartite, with longer inner lobe. Long follicular markings
developed peripherally.

Dimensions. Dimensions of available muscle fields are plotted on text-fig. 22.

Remarks. Sowerby (1842, pi. 38, fig. 7) illustrated a fragmentary ventral internal mould,
under Orthisl strigosa, from the Silurian of Haiger Siilbach (Dillenburg) in the German
Rhineland. More recent work has shown that Devonian rocks outcrop in the Dill

E e

C 4179

404

PALAEONTOLOGY, VOLUME 9

synclinorium, and not Silurian rocks as previously supposed. Other German and Belgian
material of this species is Lower Devonian in age. Bedard (1887, pi. 4, fig. 17) illustrated
a similar ventral internal mould under Orthis strigosa Sowerby, from the Dinant Basin,
Belgium.

Detailed and accurately localized collections from the Lower Devonian of Belgium
(deposited in the Institut royal des sciences naturelles) include forms closely resembling
Sowerby’s, Bedard’s, and Maurer’s illustrations. These have been listed under S.
strigosa (Sowerby). The dorsal muscle fields of these specimens (text-fig. 2lb) closely

X

X . •

X

XX • ‘

X ^ ^®neorype

X • * .

S. STRIGOSA (SOWERBY)

X ventral muscle field
• dorsal muscle field

I . I . . I . ■ ■ . I

O 3 lO 15 20 mm.

WIDTH

TEXT-FIG. 22. Dimensions of muscle fields of Schizophoria strigosa (Sowerby).

resemble Maurer’s illustrations. But the ventral muscle field is often less fiabellate, and
the median septum narrower, as shown by Sowerby.

The ventral field occasionally resembles that of Schizophoria vulvaria (Quenstedt) in
outline and narrow median septum. These variations have previously been illustrated by
authors with specimens under S. vulvaria (eg. Oehlert 1887, pi. 5, figs. 1, 5) and S.
provulvaria (Drevermann 1904, pi. 30, fig. 20). These could possibly belong to S. strigosa.
The specimens of S. vulvaria illustrated by Oehlert (1887) have also been listed by
Maillieux (1936, p. 53) under S. provulvaria, indicating further the presence of specimens
with close similarities with both S. provulvaria and S. vulvaria.

The specimens illustrated by Drevermann (1904, pi. 31, figs. 16-18) under S. pro-
vulvaria have long follicular markings, characteristic of S. strigosa. Those of S. pro-
vulvaria are shorter.

Schizophoria strigosa appears in the Siegenian, and ranges into the lower Emsian,
where it is succeeded by S. vulvaria.

Maurer (1893, p. 10, pi. 3, figs. 5-9) described and illustrated another species, Schizo-
phoria occulta, which is here considered synonymous with S. strigosa. The ventral
muscle fields and follicular markings are similar. Maurer also illustrated a dorsal

Y. P. POCOCK: DEVONIAN SCHIZOPHORIID BRACHIOPODS 405

internal mould, not shown by Bedard or Sowerby. Maurer recognized S. occulta (i.e. S.
strigosa) as distinct from the contemporaneous species S. provulvaria. The ventral
muscle field of S. occulta is strongly incised, flabellate, with a rounded median septum,
while that of S. provulvaria is much more strongly incised, protuberant in profile, and
often with a broader median septum.

There are distinct differences in the dorsal muscle fields of S. occulta (S. strigosa) and
S. provulvaria. The more slender brachiophores and brachiophore plates, and longer
inner lobe of the digitate posterior adductor muscle scar, contrast with the strong
brachiophores and brachiophore plates and more equal lobes of S. provulvaria.

Material. Belgium, Dinant Basin; Siegenien, Grauwacke de Petigny, Sg 3b (Hersdorferschichten),
(IRSN IG8254); Grauwacke de Petigny, Sg4 (IRSN IG8190); Grauwacke de Saint-Michel, Sg 3 (IRSN
IG8219); Emsien Inferieur, Grauwacke de Pesche, Em la (IRSN IG8791). Germany: Siegener Schich-
ten (MNB.B102.1 — neotype), Rauhflaserschichten (MNB), Seifen, Lower Coblenzian (SMF).

Schizophoria vulvaria (Quenstedt) 1867

Plate 66, figs. 13 a, 6; text-figs. 23-25

1655 Hysterolitlios; Worm, p. 83, text-fig. on same page.

1719 Hysterolithus', Wolfart, pi. 3, figs. 3, 5, non 4.

1763-4 Hysterolites vulva marina\ Baumer, p. 327, fig. 28.

1768 Hysterolites-, Walch, p. 90, pi. B4, figs. 5, 6.

1820 Hysterolites viilvarius Schlotheim, p. 247, pi. 29, fig. 26? non 2a, 3.

1853 Orth. Beaumont i de Verneuil; Schnur, p. 215, pi. 37, fig. 9.

1867 Hysterolitlies viilvarius Quenstedt, p. 577, pi. 49, fig. 2.

1868-71 Hysterolithus viilvarius-, Quenstedt, p. 565, pi. 56, figs. 2-6.

1882 Hysterolitlies viilvarius-, Quenstedt, p. 737, fig. 252.

1885 Hysterolitlies viilvarius; Quenstedt, pi. 57, fig. 13.

1887 Orthis (Hysterolitlies) viilvarius Schlotheim sp. ; Qehlert, p. 53, pi. 5, figs. 1-9.

1889 Orthis hysterita Gmelin; Kayser, p. 53, figs. 1, 7-9.

1893 Orthis vulvaria Quenstedt; Maurer, pi. 4, figs. 1, 2.
non 1753 Hysterolithus; Tessin, p. 90, pi. 5, fig. 2.
non 1850 Orthis Beaumonti de Verneuil, p. 180, pi. 4, fig. 8.
non 1938 Schizophoria vulvaria (Schlotheim); Compte, p. 13, pi. 1, figs. 2, 3.

Type. Schlotheim ’s specimens (1820, pi. 29, figs. 2a, 3, and possibly 2b) were probably spiriferids (see
"Remarks’’). Quenstedt (1867, 1868-71, 1882, 1885) is apparently the first author to describe and illus-
trate Schizophoria vulvaria proper. These specimens are deposited in the Geological Museum, Tubingen.

Diagnosis. Shell large, quadrate to rectangular. Ventral muscle field long, lanceolate to
weakly flabellate, strongly incised, divided by narrow, subrounded median septum.
Strong divergent brachiophore plates. Each posterior adductor muscle scar tripartite or
quadripartite.

Description. Internal mould large, dorsibiconvex, quadrate to rectangular, with greatest
width at or slightly anterior to mid-length. Anterior commissure rounded uniplicate.

Ventral muscle field (text-fig. 24a) long, one-half to two-thirds valve length, lanceolate
to weakly fiabellate, strongly incised. Dental lamellae ventrally convergent anteriorly di-
vergent. Shallow, subangular re-entrant, or re-entrant absent (text-fig. 24a). Median
septum narrow, subrounded, broadening slightly anteriorly, first increasing, then de-
creasing in height (text-fig. 23, sections 3.5-16.2). Two sub-parallel vascula media, with
lateral branches, originating from diductor muscle field or anterior re-entrant. Genital

406

PALAEONTOLOGY, VOLUME 9

TEXT-FIG. 23. Schizophoria vulvaria (Quenstedt). Transverse serial sections of plaster internal mould

(BM B23179).

Y. P. POCOCK: DEVONIAN SCHIZOPHORIID BRACHIOPODS

407

markings arranged concentrically, developed laterally and postero-laterally (text-fig.
24a).

Cardinal process prominent, broad myophore supported by narrower shaft (text-fig.
24c). Myophore compound, with up to six lateral ridges. Stubby brachiophores fused to
strong, divergent brachiophore plates (text-fig. 23, sections 2.2-4. 1). Dental sockets deep,
oval.

a VENTRAL MUSCLE FIELD b DORSAL MUSCLE FIELD

C DORSAL MUSCLE FIELD ENLARGED

TEXT-FIG. 24. Schizophoria vuharia (Quenstedt). Ventral and dorsal muscle fields.

Dorsal muscle field (text-fig. 24 b, c) quadrate to rounded, length and width approxi-
mately equal. Accessory ridges smoothly reflexed to form shallow, subrounded anterior
re-entrant. Median septum low, broad, rounded, decreasing in height and narrowing
anteriorly (text-fig. 23, sections 2. 2-9. 6). Anterior adductor muscle scar pyriform; pos-
terior muscle scar generally tripartite or quadripartite, with longer inner lobe, shorter
middle lobe(s), short outer lobe (text-fig. 24 b, c). Two main divergent vascula media,
each bifurcating, branching peripherally; two lateral vascula myaria occasionally de-
veloped from anterior adductor scars. Genital markings developed laterally and postero-
laterally (text-fig. 24b). Short peripheral follicular markings (text-fig. 24a).

Dimensions. External dimensions and dimensions of available muscle fields are plotted on text-fig. 25.

408

PALAEONTOLOGY, VOLUME 9

Remarks. All the specimens examined are preserved as internal moulds.

Comparisons with Schizophoria sirigosa (Sowerby), S. provulvaria (Maurer), and S.
striatula (Schlotheim) are made under these species. Schizophoria strigosa and S. pro-
vulvaria range from the Sigenian to Lower Emsian. Schizophoria vulvaria appears higher
in the Emsian, ranging into the Lower Eifelian, when it is succeeded by S. striatula.

Maillieux (1932, p. 24) presented a long synonymy of S. vulvaria, and discussed the
naming of the species, with reference to authors of the seventeenth century. The name

S. VULVARIA COUENSTEDT)

O 1-
Z Q-

35 p
30 -
25 -
20
15
O lO
5

25

20

lO

5 lO 15 20 25 30 35 40 mm

WIDTH

V it

A

* ventrol muscle field

• dorsol muscle field

I I I I I I I I I 1 I 1 I 1 I I I I

O 5 10

WIDTH

TEXT-FIG. 25. Dimensions of Schizophoria vulvaria (Quenstedt).

15

20 mm.

S. vulvaria is synonymous with Hysterolites. Gmelin (1790, p. 3345) very briefly de-
scribed a form under Auomia hysterita, which has been included under S. vulvaria by
later authors. However, this description is not specific.

Schlotheim (1820, p. 247, pi. 29, figs. 2, 3), the stated author of the species by Quen-
stedt and later authors, described and illustrated some specimens grouped under
Hysterolites vulvar ins, which have the transverse outline and long hinge-line charac-
teristic of a spiriferid. His figure 2b has a lanceolate ventral muscle field characteristic of
S. vulvaria, but the hinge-line is curved, obscuring its length. The mould outline and
muscle form of figure 2b also probably represents a spiriferid.

Later authors have apparently misinterpreted Schlotheim’s work as actually repre-
senting the schizophoriid species S. vulvaria, mainly on the basis of the ventral muscle
field in his figure 2b.

Quenstedt (1868-71 p. 565) stated that Schlotheim (1820, pi. 29, figs. 2a, 3) incorrectly
described and figured specimens under S. vulvaria, and that only figure 2b could possibly
represent the ventral valve of S. vulvaria. However, Schlotheim makes no reference to
the genus Schizophoria (then Orthis) in his description, and was probably describing a

Y. P. POCOCK: DEVONIAN SCHIZOPHORIID BRACHIOPODS 409

new spiriferid, since Hysterolithes is an old group term for spiriferids. He described the
specimens under Hysteroliten.

The ventral muscle field of Schlotheim (1820, pi. 29, fig. 2b) has confused later authors,
who incorrectly recognized Schlotheim as author of the schizophoriid species S. vulvaria.
Quenstedt (1867, 1868-71, 1882, 1885) was the first author to describe and illustrate S.
vulvaria as such. Quenstedt is hereby listed as the author of S. vulvaria.

De Verneuil (1850, p. 180, pi. 4, figs. Sa-d) described and illustrated a new species
Orthis Beaumonti from the Devonian of northern Spain, which resembles S. vulvaria in
shell outline, and elongate form of the ventral muscle field. But the ventral valve pallial
sinus and genital markings of O. Beaumonti are radially arranged, in contrast to the
concentric arrangement in S. vulvaria. In the dorsal valve of O. Beaumonti, the anterior
adductor muscle scars are very small, the posterior scars apparently non-digitate, and
only two parallel vascula media originate from the anterior of the muscle field. In S.
vulvaria, the anterior muscle scar is larger, the posterior scar tripartite or quadripartite,
and four to six trunks diverge from the muscle field.

Compte (1938 p. 13, pi. 1, figs. 2, 3) described and illustrated specimens from the
Lower and Middle Devonian of northern Spain under 5*. vulvaria, with Orthis Beaumonti
de Verneuil listed in synonymy. Compte’s specimens resemble those of Verneuil, and
were collected from the Upper Siegenian, Emsian, and Lower Eifelian stages. Schizo-
phoria vulvaria from Belgium and Germany is restricted to the Emsian and Lower
Eifelian stages.

Although resembling and probably related to S. vulvaria, Orthis Beaumonti has not
been listed in synonymy with S. vulvaria, but is here considered a separate form, based
on its dorsal muscle field and pallial sinus markings.

Material. Belgium, Dinant Basin; Lower Devonian (BM B 15708). Emsien Superieur, Grauwacke de
Hierges, Em 3 (IRSN IG4916, 5391, 5746, 5910, 5911, 8254, 8284, 8573); Couvinien Inferieur, Assise
de Bure (CO l)-CO la (IRSN IG12409), CO lb (IRSN IG5746); Germany, Lower Devonian (BM
B19002) Lahnstein; (BM B23179, B34290, B49920, B39450) Eifel; (BM B39435) Coblenz; (BM
B86626, 7), mouth of River Lahn; Ober Coblenzian (HMUG L2031) Daleiden; Coblenzian (HMUG
L5344) Grimbach; Ober Ems (SME) Priim.

PHYLOGENY

The postulated phylogeny of Devonian species of Schizophoria studied, and possible
links with the Carboniferous, is shown on text-fig. 26. This chart is based solely on material
examined, and could conceivably represent only a part of the true picture of descent.

The relationship of species has been based externally on outline, and internally on
muscle field patterns and form of the brachiophore plates, correlated with stratigraphical
occurrence. External and internal features are illustrated on text-fig. 26.

The species appear to have been derived from two, perhaps three, root stocks.
Schizophoria antiqua (Gedinnian-Erasnian) is the first species to appear in Europe.
Morphologically, S. antiqua bears little relationship to other Devonian species, but re-
sembles more closely S. woodi from the Carboniferous. Schizophoria antiqua is repre-
sented as a separate line of development.

Other Devonian species could have been derived from the S. provulvaria (Siegenian-
Emsian) and S. strigosa (Siegenian-Emsian) stocks. Schizophoria provulvaria appears to
have given rise to S. pygmaea and S. schnuri blankenheimensis (Eifelian). There are close
resemblances in external outline and muscle field patterns. Although S. pygmaea is

S CONNIVENS S RESUPINATA-CARBONIFEROUS CARBONIFEROUS-S.WOODI

TEXT-FIG. 26. Postulated phylogeny of Schizophoria from the Devonian of western Europe.

Y. P. POCOCK: DEVONIAN SCHIZOPHORIID BRACHIOPODS

411

much smaller than S. provulvaria, S. pygmaea is a dwarf form of S. schnuri blanken-
heimensis. The subspecies blankenheimensis and S. provulvaria are comparable in size.

Schizophoria strigosa is replaced by S. vulvaria in the Emsian. The two species have
comparable outline and muscle field patterns.

Schizophoria striatula (Eifelian-Frasnian) shows both external and internal affinities
with S. vulvaria, and is considered to have been derived from this line of development.

Acknowledgements. The author thanks Dr. C. H. Holland of Bedford College for his advice on
presentation, and to all museums for allowing access to relevant collections. Thanks are extended to
Mr. J. W. Keith (Bedford College) for photographic assistance. This research work formed part of
a thesis submitted for a Ph.D. degree, University of London.

REFERENCES

ASSMAN, p. 1910. Die Fauna der Erbsloch-Grauwacke bei Densberg im Kellerwald. Jb. preufi. geo!.
Landesanst. 31, 136-72.

BAUMER, J. w. 1763-4. Naturgeschichte des Mineralreichs. 1-520, Gotha.

BECLARD, F. 1887. Les fossiles Coblenziens de Saint-Michel pres de Saint-Hubert. Bull. geol. Soc. Belg.

I, 60-96.

BiERNAT, G. 1959. Middle Devonian Orthoidea of the Holy Cross Mountains and their ontogeny.
Palaeont. polon. 10, 53-63.

COMPTE, p. 1938. Brachiopodes Devoniens des gisements de Ferrones (Asturies) et de Sabero (Leon).
Ann. Paleont. 27, 39-88.

cowPER-REED, F. R. 1908. Devonian fauna of the northern Shan States. Palaeont. indica. New Ser. 2,
1-183.

1922. Devonian fossils from Chitral and the Pamirs. Ibid. 6, 1-134.

DAVIDSON, T. 1851-5. British Fossil Brachiopoda. v. 1, Tertiary, Cretaceous, Oolitic, and Liassic species,
with a General Introduction. Palaeontogr. Soc. [Monogr.]

1864-5. A monograph of the British fossil Brachiopoda, pt. 6, The Devonian Brachiopoda.

Ibid.

DREVERMANN, F. 1904. Die Fauna der Siegener Schichten von Seifen unweit Dierdorf (Westerwald).
Palaeontographica, 50, 229-87.

GEiNiTZ, H. B. 1853. Die Versteinemngen der Grauwackenfonnation in Sachsen and den angrenzenden
Ldnder-Abtheilungen, 2, 1-95.

GILL, E. D. 1942. On the thickness and age of the type Yeringian strata, Lilydale, Victoria. Proc. roy.
Soc. Viet. New Ser. 54, 21-52.

GRUNEWALDT, M. 1860. Beitriige zur Kenntnis der sedimentiiren Gebirgsformationen. Mem. Acad,
imp. Sci. St.-Petersb. 7me ser. 2, 1-144.

KAYSER, E. 1889. Die Fauna des Hauptquarzits und der Zorger Schiefer des Unterharzes. Abli. preufi.
geol. Landesanst . 1, 1-139.

■ 1890. Beitrage zur Kenntnis der Fauna der Siegenschen Grauwacke. Jb. preufi. geol. Landesanst.

II, 95-107.

KONiNCK, L. E. DE 1842-4. Description des animaux fossiles qui se trouvent dans le terrain carbonifere
de Belgique, iv+650 pp., Liege.

LEONHARD, c. c. 1813. Tasclieubucli fiir die gesammte Mineralogie, mit Hinsicht auf die neuesten
Entdeckungen, 1 (2), 612 pp., Frankfurt am Main.

MAiLLiEUX, E. 1922. The Geology of Belgium 11. The Palaeozoic Formations of the southern part of the
Dinant Basin. Proc. geol. Ass. 33, 9-19.

1932 La faune de I’Assise de Winenne (Emsien moyen) sur les bordures meridionale et orien-
tate du Bassin de Dinant. Mem. Inst. Sci. nat. Belg. 52, 1-102.

1936 La faune et Page des quartzophyllades Siegenien de Longlier. Ibid. 73, 1-140.

1941. Les Brachiopodes de I’Emsien de I’Ardenne. Ibid. 96, 1-74.

MAURER, F. 1 886. Die Fauna des rechtsrheinischen Unterdevon. 32 pp. Darmstadt.

■ 1893. Paliiontologische Studien im Gebiet des rheinischen Devon. Neues Jb. Min. Geol. Paldont.

1, 1-14.

412 PALAEONTOLOGY, VOLUME 9

OEHLERT, D. p. 1887. Etudc SUE quclqucs fossiles devoniens de I’ouest de la France. Ann. Sci. geol.
Angers, 19, 1-80.

PAECKELMANN, w. 1930. Die Brachiopoden des deutschen Unterkarbons. 1. Ab/i. prenfi. geol. Landes-
anst. 122, 143-326.

PHILLIPS, J. 1841. Figures and descriptions of the Palaeozoic Fossils of Cornwall, Devon, and West
Somerset, xii + 231 pp., London.

PRUVOST, p. 1924. A synopsis of the geology of the Boulonnais, including a correlation of the Mesozoic
rocks with those of England. Proc. geol. 35, 29-67.

QUENSTEDT, F. A. 1868-71. Petrefaktenkioule Deiitschlands: Abt. 1, Band 2. Die Brachiopoden. iv+748
pp., Leipzig.

• 1871. Atlas zu den Brachiopoden. 37-61 pis., Leipzig.

■ 1882-5. Handbiich der Petrefaktenkioule. viii+1239 pp., Tubingen. 3rd ed. (1852 — 1st ed. ;

1867— 2nd ed.).

SANDBERGER, G. and SANDBERGER, F. 1850-6. Die Versteinerungeii des rheinischen Schichtensystems in
Nassau, xiv+564 pp., Wiesbaden.

SCHLOTHEiM, E. E. 1820. Die Petrefaktenkioule aiif ihrem jetzigen Standpiinkte diirch die Beschreibiing
seiner Sammhing versteinerter iind fossiler Uberreste des Thier- and Pflanzenreichs der Vorwelt.
Ixii 1437 pp., Gotha.

SCHNUR, j. 1853. Zusammenstellung und Beschreibung sammtlicher im Uebergangsgebirge der Eifel
vorkommenden Brachiopoden. Palaeontographica, 3, 169-254.

SCHROTER, J. s. 1777. Abluindlioigen der iiber verschiedene Gegenstdnde der Naturgeschichte, 2, 335-504.
SCHUCHERT, c. and COOPER, G. A. 1932. Brachiopod Genera of the suborders Orthoidea and Pentam-
eroidea. Mem. Peabody Mas. 4(1), xii + 270 pp.. New Haven.

SHIRLEY, J. 1938. The fauna of the Baton River Beds (Devonian), New Zealand. Quart. Jl. geol. Soc.
Loud. 94, 459-506.

SOLLE, E. 1936. Revision der Fauna des Koblenzquarzits am Rhein und Mosel. Senckenbergiana, 31,
154-215.

sowERBY, J. 1842. Description of Silurian fossils from the Rhenish Provinces. Trans, geol. Soc. Lond.

6, 303-410.

SPRiESTERBACH, J. 1942. Lenneschiefer (Stratigraphic, Fazies und Fauna). Abh. Reichsamte Bodenf.
203, 1-219.

STANLEY, J. w. 1964. Serial sectioning of steinkerns. Palaeontology, 7, 105-7.

STRUVE, Vv'. 1963. Das Korallenmeer der Eifel vor 3C0 Millionen Jahren. Natur. u. Mus. 93, 237-76.

• 1963u. Beitrage zur Kenntnis devonischer Brachiopoden, 1: Schizophoria pygmaea n. sp.

Senckenbergiana, 44, 251-63.

1965. Beitrage zur Kenntnis devonischer Brachiopoden, II: Schizophoria striatula und Schizo-
phoria excisa in ihrer urspriinglichen Bedeutung. Ibid. 46, 193-215.

TERMiER, H. 1936. Etudcs gcologiqucs sur le Maroc Central et le Moyen Atlas Septentrional. Tome 3,
Paleontologie, Petrographic. Notes Serv. Min. Maroc, 33, 1087-421.

and TERMIER, G. 1950. Invertebres de I’Ere Primaire, fasc. 2, Bryozoaires et Brachiopodes. Paleont,

Maroc, 2, 1-253.

TESSIN, c. G. 1753. Museum Tessinianiim, opera ilhistrissimi Comitis. 123 pp. Holmiae.

VERNEUiL, E. de 1850. Note sur les fossiles devoniens du district de Sabero (Leon). Bull. Soc. geol. Fr.

7, 155-86.

viETOR, w. 1916. Der Koblenzquarzit. Seine Fauna, Stellung und linksrheinische Verbreitung. Jb.

preufi. geol. Landesanst. 31, 317-476.

WALCH, E. i. 1768. Die Naturgeschichte der Versteinerungeii, 2, 90-95.

WALTHER, K. 1907. Beitrage zur Geologic und Palaontologie des alteren Palaozoicums in Ostthiiringen.

Neues Jb. Min. Geol. Paldont. 24, 221-324.

WOLEART, p. 1719. Flistoriae Natiiralis Hassiae Inferioris pars prima. 52 pp., Cassel.

WORM, o. 1655. Museum Wormianum sen Historia serum rariorum, 389 pp., Lugduni Batavorum.

Y. P. POCOCK

Department of Biology and Geology,
Northern Polytechnic,

London, N. 7

Manuscript received 26 May 1965

THE MICROMETRIC FORMULA AND THE
CLASSIFICATION OF FENESTRATE CRYPTOSTOMES

by R. TAVENER-SMITH

Abstract. A critical assessment of the use of the micrometric formula in classifying fenestrate cryptostomes
shows that although the device may be of some use as an aid to description and as a means of indexing species,
it is an ineffective basis for structural comparisons. The method employed in making such comparisons is also
unsound. Nevertheless, taxonomic conclusions are commonly drawn from them, and a result of this is an
unreasonable increase in the number of recognized species. It is suggested that the micrometric formula should
be discarded for comparative purposes and its place taken by one of the orthodox biometrical tests of signifi-
cance. Such tests afford a simple and objective way of comparing sets of data. Samples for comparison should
consist of groups of specimen means: owing to the colonial nature of the organisms, data from a single colony
are inadequate for the purpose. A procedure for comparing samples is outlined, and an example given.

Since colonies of fenestrate bryozoa were first examined in detail it has been evident
that the pattern of structural elements in a zoarium offers the basis for a numerical means
of discriminating between species. M‘Coy (1844) and his contemporaries incorporated
measurements of these features in systematic descriptions with a view to their use in com-
parison, and at a later date Shrubsole (1881, p. 189) presented similar data in tabular
form with the same end in view. With the passage of time a desire for greater refinement
led to the inclusion of increasing numbers of measurements in descriptions, particularly
by Russian authors. These became so numerous that Nekhoroshev (1926) introduced the
practice of extracting those that seemed most critical and presenting them separately in
the form now known as the ‘meshwork formula’ (Condra and Elias 1944, pp. 56-57), or
the ‘micrometric formula’ (Miller 1961, p. 224). These figures were intended to convey
the essential structural characteristics of the forms described. They were based on
measurements of four kinds: the number of branches in 10 mm., measured perpendicular
to the axis of growth ; the number of fenestrules in the same distance, measured along the
branch length; the number of zooecial apertures in a single row in 5 mm., and the num-
ber of carinal nodes, also in 5 mm. The formula stated the frequency or (more often)
the observed range of each feature.

This procedure was made known to western workers by Condra and Elias in the paper
mentioned above, in which they also proposed a standard method for making the re-
quired measurements. These authors strongly advocated the use of the formula in de-
scriptive work and also used it as a basis for taxonomic comparisons. Since that time it
has been accepted into general use and become a principal means of discriminating be-
tween the numerous species of fenestrate cryptostomes.

For descriptive purposes the micrometric formula has much in its favour. It is readily
obtained even from small specimens, and provides a convenient shorthand expression
of important structural characteristics of colonies. Because these formulae are now
available in the literature for almost all adequately described species they also provide
a useful basis for indexing (Miller 1961, p. 224), a valuable asset in a genus like Fenes-
tella with more than 500 named species. As a basis for taxonomic discrimination, how-
ever, the formula has less to commend it and the purpose of this paper is to examine its

[Palaeontology, Vol. 9, Part 3, 1966, pp. 413-25.)

414

PALAEONTOLOGY, VOLUME 9

function in this respect and to suggest improvements. Before doing so it is helpful to
consider eertain structural characteristics of fenestrate colonies, and also the nature of
the information that the formula contains.

Structural variation in fenestrate colonies. The tendency towards structural variation
within a species is a widely recognised characteristie of fenestrate eryptostomes and one
that has been commented on by many authors (e.g. Foerste 1887, p. 84; Condra and

Apertures per 5mm.

Branches per 10mm.

Fenestrules per 10mm.

TEXT-FIG. 1. a. Graph of measurements made on a single colony (a homeotype of
Fenestella hemispherica M‘Coy, Sedgwick Museum specimen E 17841); b. Distribu-
tion of the modes of 55 colonies of Ptilofenestella carrickensis Tavener-Smith;
c. Bivariate distribution of modal values from the same colonies, showing the
location of the specimens listed in Table 2.

Elias 1944, p. 56). This variation takes two main forms: that within a colony, and that
between colonies. If a number of readings from a colony are represented graphically they
form an approximately normal distribution (text-fig. \a). Each variable is distributed in
this way, occurring between certain observable limits, and it is therefore possible to con-
struct a micrometric formula for a single zoarium, or even a fragment. It is important to
remember, however, that a zoarium is a clone, that is, an association of asexually pro-
duced individuals that are genetically alike, and the range of variation within it will not,
therefore, be of direct use in classification.

The central value, on the other hand, has much greater importance as it is from many
points of view the most representative measurement for the colony concerned. This value
is determined by the interaction of two sets of factors : genetic and ecological. A colony
originates by sexual reproduction and therefore has its own particular genetic consti-
tution that distinguishes it from all other colonies. Ecological considerations involve
the relationship between a colony and its environment, and will vary with the situation.
Within a colony disparities in the biochemical control of growth, and differences of
micro-environment cause the dispersion of the data into a normal distribution, of which
the central value reflects the interaction of the two basic factors.

Where a series of data shows an approximately normal distribution, either the mean,
median or mode may, according to circumstance, be used to measure their central

R. TAVENER-SMITH: CLASSIFICATION OF FENESTRATE CRYPTOSTOMES 415

tendency. Although the mode is usually the obvious choice for this purpose, it has dis-
advantages in small samples, for in them it is very subject to random fluctuation.
Difflculty may also arise becauses uch samples occasionally show more than one mode
(e.g. three of the graphs in text-fig. 2a). The median and arithmetic mean are therefore
often of greater practical value and, of these, the mean is the more useful statistic in

A. B.

Number of branches in 10mm. Number of fenestrules in 10mm.

TEXT-FIG. 2. Graphs of measurements from colonies of Ptilofenestella carrickensis to
illustrate the range of morphological variation. Each graph incorporates 15 readings,
and the order of arrangements is the same in both diagrams.

comparative work. The mean of a series of readings made on a colony is, for these
reasons, usually the single measurement best suited to represent it.

If sets of readings from a number of conspecific colonies are plotted graphically, a
series of overlapping distributions result (text-fig. 2). In such a series the ranges of indi-
vidual colonies may differ appreciably from one another, and may even (as the diagram
shows) be mutually exclusive. It is apparent from this that in a morphologically variable
group such as the fenestrate cryptostomes the range of variation shown by a single
colony may bear little relation to that of the species to which it belongs.

The lack of a direct relationship between intra- and inter-colonial variation can also be
demonstrated by means of the technique of analysis of variance, to which the data are
readily adapted. Four such analyses were made, one for each variate of the micrometric
formula, the data being derived from specimens of Ptilofenestella carrickensis Tavener-
Smith 1965. Between ten and fifteen measurements from each of 10 colonies were used,
and in each case it was found that a significant difference {P < 0-05) existed between the

416

PALAEONTOLOGY, VOLUME 9

variance estimates. There is therefore a recognizable difference in pattern between intra-
colonial and inter-colonial variation, and this being so, the range of a variate within
a colony cannot justifiably be used as the basis for taxonomic comparisons between
colonies.

When the central values of a number of colonies of the same species are assembled into
a histogram it is seen that they also have an approximately normal distribution. In text-
fig. \b 0. group of modes is used to illustrate this point: the corresponding means would
show a similar pattern. Samples of this kind are likely to be taxonomically useful because
they are based on the most representative measurements of a group of colonies and

Table 1

Fenestrule

width.

Inter-ap.

space.

Internodal

space.

Branch

width.

Apertural

diameter

Fenestrule

length.

♦0-6886

♦ 0-4217

♦0-3898

-0-4836

♦0-5226

Fenestrule

width.

♦0-3899

♦0-3274

-0-4471

♦0-7169

Inter- ap.
space.

♦0-3014

-0-3214

♦0-3605

Internodal

space.

-0-2640

♦0-1119

Branch

width.

♦0-3760

Ptilofenestella carrickensis: coefficients of correlation between pairs of structural
features. The continuous variables used here include those that correspond most
closely with the micrometric formula, namely: fenestrule width and length, inter-
apertural space, and internodal space. In all cases n=55.

(according to the number of specimens measured) will provide a more or less reliable
indication of the range of variation in the species concerned. Distributions of this type
are used as the basis for the comparative technique outlined later in this paper.

Another characteristic of fenestrate cryptostomes is the significant, though weak,
correlation that exists between different structural elements in a colony (Table 1). This is
evident from a consideration of micrometric formulae, in which a high count for the
number of branches in 10 mm, is often accompanied by high numbers of fenestrules,
apertures, and nodes (e.g. Fenestella bicellulata Etheridge: 24-27/27-28//27-29/29-31).
The converse is generally true when the branch count is low (e.g. F. oblongata Koenig:
9-1 5/4-7//1 4— 19/4-7). Although it is technically more correct to use a multivariate
approach where sets of data are correlated, the correlations are here so weak that little is
lost by using simpler univariate methods in comparing samples. Comparisons based on
the micrometric formula are in this respect quite well adapted to the situation, for they
function on this principle. Although each formula embodies the frequency or observed
range of four variates, these are dealt with separately, and the comparisons are quite
independent of one another.

Construction of the micrometric formula. There is no reason to doubt that the geo-
metrical arrangement of structural elements in a fenestrate colony is of taxonomic value.

R. TAVENER-SMITH: CLASSIFICATION OF FENESTRATE CRYPTOSTOMES 417

and the spatial distribution of branches, dissepiments, zooecial apertures, and carinal
nodes were considered by Nekhoroshev (1926) to be the most important variates in-
volved. These features have come, by usage, to be the ones on which most reliance is
placed in discriminating between species (Condra and Elias 1944, p. 54). They are there-
fore weighted for taxonomic purposes, as compared with others such as the width of
branches or dissepiments, and the diameter of zooecial apertures, which are not in-
cluded in the formula. Nevertheless, all these have been recognized at one time or
another to have potential diagnostic value fe.g. Nekhoroshev 1932, p. 302; Miller 1962,
p. 120). Their relative neglect is probably due to the tendency for secondary calcification
to alter dimensions with increasing age, thus apparently nullifying the usefulness of these
features in classification. It seems likely, however, that this objection is not insuperable.
Certainly, the restriction of taxonomic consideration to the variates of the micrometric
formula is in itself a disadvantage, for it is generally agreed that the best classification is
that based on all relevant morphological data.

A standard method for measuring variates of the micrometric formula is described by
Condra and Elias (1944, pp. 54-55). They recommend the use of the so-called space-unit
count, which means that it is the space between selected structural features that is
counted, and not the features themselves. Thus, the total per standard distance (5 or 10
mm.) is not the actual number in that distance, but one less than this. It is worth notic-
ing that because the basis of each count is the linear distance between adjacent features,
the variates are essentially continuous and not discontinuous, as first appearances
suggest. It is therefore permissible and advantageous to use the mean rather than the
mode as the central value of distributions relating to them.

The method of presenting structural data in the orthodox micrometric formula is ex-
tremely rudimentary. Only the observed range of the measurements is given for each
feature, and sometimes this is abbreviated to a single figure, implying that there was no
variation in the sample examined. No supporting data of any kind relating to the pattern
of the distribution are given. Nor is it stated how many readings were made, or whether
all were taken from a single specimen or from several. The work of Perry and his asso-
ciates is, in this respect, an exception to the general rule. Utgaard and Perry (1960) give
formulae supported by histograms showing the distribution of the variables, and Malone
and Perry ( 1965) state a mean and standard deviation for each variate and the number of
measurements made. They do not say, however, how many zoarial fragments were
examined, or how many readings were taken from each. Both facts are relevant if com-
parisons are to be taxonomically valuable.

Effectiveness of the formula in taxonomic discrimination. To illustrate the use of the
micrometric formula in structural comparisons an example is necessary, and a typical
case occurs in recent work by Burckle (1960, p. 1083). This author measured some new
material in order to ascertain whether it was conspecific with Fenestella rectangularis
Ulrich, to which there was a superficial resemblance. Having derived a formula, he
compared it with that of Ulrich’s species in the usual way, namely by placing the two
side by side and examining the ranges of each variable separately. The formulae were as

follows: p rectangularis Ulrich: 20-26/171-24//23-25/14-17

Burckle’s material: 23-27/ 16-24//20-24/ 16-20.

He concluded that, while the ranges for the first three variables were close enough to be

418

PALAEONTOLOGY, VOLUME 9

considered identical, those for carinal nodes indicated a clear difference between the
samples. This he took to be of taxonomic as well as numerical significance, and accord-
ingly founded a new species which he named F. tooelensis.

It is necessary to decide whether comparisons of this pattern are, in general, acceptable
as a basis for taxonomic discrimination. The reasons given below suggest that they are
not.

1 . Micrometric formulae are usually stated in terms of the observed range of variates.
This is generally acknowledged to be a poor basis for comparison, as the range of a
sample is directly related to its size (Simpson 1941). Unless samples are uniform in this
respect their ranges will be expected to differ, even if they are drawn from the same popu-
lation, and such differences need have no taxonomic significance. If, in the example

Table 2

Branches In 10mm.

Fenestrules in 10mm.

Apertures in 5mm.

Nodes in 5mm.

A.

20 - 22

12-13

21 - 24

27 - 33

B.

13 - 17

8-12

17-19

21 - 27

C.

20 - 22

11 - 12

18 - 21

25 - 28

D.

14 - 18

9-10

17 - 21

30 - 38

Micrometric formulae measured on four selected specimens of Ptilofenestella carrickensis.

quoted above, each formula was based on only one specimen it would be quite un-
justifiable to assume that the difference in node counts had taxonomic significance. But
if, on the other hand, each was derived by measurement of twenty specimens, the case
for a difference between them would be a strong one. As the number is not stated there is
no way of deciding where the truth hes.

Formulae giving single figures for variates are even less useful than those that show
the observed range. A single figure provides no idea of variation, and all that can
reasonably be done is to treat it as the mean of the distribution it represents. Even then it
is of no value for comparative purposes without information as to the number of speci-
mens examined and measurements made. In view of the structural variability of fenes-
trate zoaria it seems likely that formulae giving this kind of information were measured
on single small fragments.

2. Many micrometric formulae in the literature appear to have been measured on one
specimen only, often a holotype (Condra and Elias 1944, p. 107; Elias and Condra
1957, p. 77; Koenig 1958, p. 135; Burckle 1960, p. 1087; Miller 1961, p. 231, and many
others). Such formulae can only record intra-zoarial variation and this offers no founda-
tion for inter-zoarial comparison because it is possible for the range of a variate to differ
markedly in two specimens that are conspecific (text-fig. 2). A notable numerical dis-
crepancy between observed ranges may have no taxonomic significance at all if the
ranges were measured on single specimens. Table 2 shows the formulae of four speci-
mens of PlilofenesteUa carrickensis Tavener-Smith which occur at the extremes of a
bivariate distribution of the number of branches and fenestrules per 10 mm. The graph
is shown in text-fig. Ic. Because a weak positive correlation exists between the variables
of the formula, extreme ranges for apertures and nodes occur in the same specimens.
Discrepancies between the ranges of these formulae are enough to suggest separation

R. TAVENER-SMITH; CLASSIEICATION OE FENESTRATE CRYPTOSTOMES 419

into 2, if not 4 species of the kind that Burckle recognized. Yet all 4 colonies are, in fact,
conspecific. Comparisons involving formulae derived from single specimens are, it
seems, unreliable for taxonomic purposes.

3. In comparisons of formulae such as that outlined earlier, the method is visual and
subjective, each case being decided solely by the personal judgement of the author con-
cerned. While such a procedure may be satisfactory if formulae happen to be identical or
when they differ widely, in the great majority of intermediate cases there is much room
for error. Results are most likely to be unreliable where the overlap between ranges is
appreciable but not complete, and the dependability of the test will therefore be least
exactly where it needs to be greatest. Because there is no objective way to decide whether
an observed difference between the ranges of samples has taxonomic significance, uni-
formity of treatment can hardly be expected. A difference that is enough to justify the
erection of a new species in the opinion of one author may seem insufficient for the pur-
pose to another, and the classification suffers accordingly.

It is evident that there are serious deficiencies in the present method of comparing
samples of structural data from fenestrate colonies. One result of this is an unreasonable
increase in the number of recognized species. An improved procedure is needed, and two
alternative courses are open: either the micrometric formula may be retained in a re-
vised and expanded form, or it may be discarded in favour of a new approach. For
reasons already stated, the formula can only provide a reasonable basis for comparison
if the ranges given are those of groups of specimen means, and if the number in each
group is known. Approximate statistics for the distribution can then be calculated, and
a rougn but objective comparison of samples made (Simpson 1941, pp. 788 and 793).
However, the micrometric formula is founded on the observed range and even in its most
acceptable form permits only a crude comparison of the kind mentioned above. It seems
better, from all points of view, to discard the formula for comparative purposes, though
it could be retained in descriptions as a means of indexing species. In effecting numerical
comparisons it is the distribution of variables, not their ranges, that is important, and
this is best defined in terms of orthodox statistics such as the mean and variance. If
these are available samples can be compared by utilizing one of the significance tests
common in modern biometric usage. Before outlining the kind of procedure that might
be followed, it is advisable to consider the morphological features that could be com-
pared, and the way in which they might be measured.

Dimensions to be compared. The fundamental purpose of the method outlined by Condra
and Elias (1944, pp. 54-57) for measuring variable features on fenestrate bryozoans is to
establish the average dimensions that characterize a particular fragment or series of
fragments. Their procedure for doing this is clumsy and fails to provide an effective basis
for comparison. Counting the number of branches, fenestrules and so on, per unit dis-
tance is merely an indirect and ineffective way of assessing the average distance between
these structures. It is more satisfactory as well as simpler to make the necessary measure-
ments directly, and to derive from them statistics that can be quickly and objectively
compared by means of a significance test.

For comparisons to be valid it is essential that measurements should be made in a
standardized manner, and text-fig. 3 suggests how this could be done. The variables of
the micrometric formula are adequately represented by the basic measurements

F f

C 4179

420

PALAEONTOLOGY, VOLUME 9

concerned, namely fenestrule width and length, together with inter-apertural and inter-
nodal distance. Fenestrule width is measured through the mid-point of a fenestrule,
from the centre of the branch on one side to the centre of that on the other. Fenestrule
length is the distance, along the mid-line of the fenestrule, between the centres of adjacent
dissepiments. Inter-apertural distance and internodal distance are measured between
the centres of pairs of the appropriate structures that are situated in the same row.

To these variates may be added others, such as branch width, dissepiment width, and
apertural diameter. It is true that these are subject to secondary calcification, and their
dimensions may be influenced by this factor. Nevertheless, if samples consist of groups
of specimen means there is no reason why it should vitiate comparisons. Random

TEXT-FIG. 3. Method of making measurements: a, fenestrule
length; b, fenestrule width; c, inter-apertural space; d,
internodal space; e, branch width; /, dissepiment width;
g, apertural diameter.

samples of fenestrate colonies should contain sufficiently similar numbers of both young
and old colonies to cause the influence of secondary accretion on measurements to
balance out between samples. Its importance is therefore much less than would be the
case in comparing single colonies.

Branch width is recorded at right angles to the branch axis and away from points of
bifurcation and branch-dissepiment junctions. The width of dissepiments is measured
midway along their length, where the structure is narrowest. In the case of zooecial aper-
tures it is the internal diameter that is measured: for ovoid or pyriform apertures a
longer and shorter dimension could be given.

Elias and Condra (1957, pp. 70-72) attached great taxonomic value to the number of
zooecial apertures per fenestrule, and largely based their classification upon this. Many
workers would not agree in according prime importance to this feature, but it is one that
can sometimes be used with advantage in comparative work. An examination of the
literature reveals, however, that there is a discrepancy in the method employed to make
the necessary measurements. Some workers (e.g. Shulga-Nesterenko, 1951) count the
actual number of apertures along one side of a fenestrule, while others use the space-
count method of Condra and Elias and record one less than the actual number. As the
numbers are always small such a discrepancy is likely to have unfortunate results, and it

R. TAVENER-SMITH; CLASSIFICATION OF FENESTRATE CRYPTOSTOMES 421

is important to standardize procedure. The second method of counting is recommended,
as it accords with that used in measuring other variables, and in practice the number of
apertures per fenestrule for a colony or fragment may be derived from the figures for
mean fenestrule length and mean inter-apertural space.

Measurements of these features and any others that are desired can be made in the
usual way with a microscope eyepiece micrometer. More accurate results are obtained
by using a screw micrometer with travelling cross-wire, or better still, a traversing stage
with a screw micrometer. Measurements should, wherever possible, be taken from
mature parts of zoaria where structural variation is likely to be at a minimum (Miller
1961, pp. 222-3), rather than from the proximal region where growth is often irregular.

Method of comparison. In order to permit a realistic comparison to be made, samples
must contain information from a number of specimens, not only one. This is because
data from a single colony or fragment reflect intra-zoarial variation alone and this, for
reasons already given, has little value in taxonomy. Also, only the central value of a series
of measurements from a colony is taxonomically useful, and significance tests are not
competent to discriminate between single values, but only between groups. The mean
is the most convenient central value for most purposes, and a sample for comparison
should therefore consist of a number of specimen means. The larger the number in the
sample, the more accurately will it reflect the range of variation in the population from
which it was drawn. Small samples of only a few specimens can also be used however.

The use of a recognized significance test ensures that an objective comparison of
samples is made, and the one best suited to present requirements is the well-known /-test
described in standard statistical texts (e.g. Fisher 1948, p. 122). When this method is used
to compare samples, / is essentially the ratio of the difference between means to the
standard error of the difference. It may be written :

t = — TTY — — for tu-^iu—2 degrees of freedom,

sf{llfh+lin2)

where Xi and .x.^ are the means of two samples, and /q, n.^ are the numbers of specimens in
those samples. Also;

2

1

IKw

(A'a-.w)

21

)•

Having calculated t and knowing the number of degrees of freedom, the value of P
(the probability that the difference between the means is due to the chances of sampling
only) may be read from appropriate tables. If P is greater than 0-05 it is probable that
the results are due to chance, and they are described as ‘not significant’. If less than 0-01,
then not once in 100 times could such a result have arisen by chance, and it may be con-
sidered significant. If the value of P lies between the 0-01 and 0-05 levels the result is
probably significant. A suggested sequence for comparing samples of structural data by
this method is outlined below. The procedure would have to be repeated for each variate.

(1) A series of readings should be made on each specimen for the variate under con-
sideration. The number of measurements would depend on the size and state of preser-
vation of the material: between 10 and 20 would be satisfactory. From these calculate
the specimen mean.

(2) Repeat for each specimen, and then assemble the specimen means into a separate

422

PALAEONTOLOGY, VOLUME 9

distribution. It is this group, consisting only of specimen means, that constitutes the
sample for comparison. Compute the sample mean and variance: these, together with
the number of specimens in the sample, are the basic statistics used in comparison.

(3) Any two samples of this kind may then be objectively compared by using the /-test,
as outlined above.

To illustrate the working of the method an actual example is quoted below. It con-
cerns three superficially similar forms of Fenestella for each of which a number of speci-
mens was available. It was desired to test these numerically in order to find whether the
three groups could have been drawn from the same parent population. For this purpose
each was represented by a sample of 25 specimens, referred to here as samples A, B, and
C respectively. A number of variates were measured in each sample and comparisons
made between them. The data for fenestrule width (corresponding to the number of
branches per 10 mm. in the micrometric formula) were as follows:

Sample A Sample B Sample C
x: 0-573 mm. 0-598 mm. 0-432 mm.

2 (a— .v)“: 0-069 mm. 0-155 mm. 0-104 mm.

On testing samples A and B it was found that:

X1—X2 = 0-025 mm.

52 = ^ (0-069+0-155) = 0-0047.

So

Then

V = 0-068.

0-025 /p5x25\
^ ^ (H)68 V \ 50 /

1-299.

From the tables it is seen that for n = 48 this value of l indicates a probability of more
than 0- 1 . The result is therefore not significant, and the samples could very well have been
drawn from the same population. On comparing samples + and C, however, it is found
that the value of t is much larger, being, in fact, 8-29. For the same number of degrees of
freedom this represents a probability level of less than 0-001, and it is very unlikely that
these two samples could belong to the same population. Comparison of forms B and C
yields a similar result, the value of t this time being 7-41 (i.e. P < 0-001). From these tests
it appears that while samples + and B cannot be differentiated from one another, both
are significantly different from sample C. Further tests on other variates gave confirm-
atory results and it was concluded that two distinct species were represented, + and B
belonging to one, and C to the other. Additional confirmation of quite a different kind
appeared later when it was found that specimens of groups + and B had triangular
zooecial base shapes, while those of group C were hemi-hexagonal. It is worth mention-
ing that micrometric formulae based on the three samples showed much overlap in their
ranges and gave no indication of the result that emerged quite clearly from the above
tests. The formulae were as follows:

Group +: 7-12/3-6//1 1-15/2-6.

Group B\ 8-13/3-8//12-16/3-7.

Group C: 10-14/5-9//13-17/4-9.

R. TAVENER-SMITH: CLASSIFICATION OF FENESTRATE CRYPTOSTOMES 423

The recognition of significant differences between samples in all tests of a series leaves
little doubt that the groups concerned belong to different species. Difficulty would arise,
however, if it was found that the level of significance was exceeded in only, say, three out
of six cases. Would it then be reasonable to differentiate the samples at specific level? In
earlier work new species have sometimes been erected on the basis of a single quantitative
difference between samples, as in the case of Burckle’s species. More frequently two such
differences are cited, and occasionally more. In discussing the application of biometrical
methods to the classification of Caradocian brachiopods Williams (1962, p. 79) suggested
that a significant difference in one feature, particularly if it could arise phenotypically,
might serve to indicate the presence of separate sub-species, but that two or more such
differences are needed to justify separation at the specific level. Such a scheme is, of
course, arbitrary and the need to make exceptions to it might arise from time to time.
Nevertheless, it seems to be in general accordance with established practice in the classi-
fication of fenestrate bryozoa, and its adoption would promote uniformity of treatment.

SUMMARY

The foregoing arguments and suggestions can be summarized as follows:

(a) The micrometric formula is of use as a descriptive aid because it conveys an
immediate impression of the general characteristics of a fenestrate colony. It is also a
convenient basis for indexing the numerous species of Feuestella, and as such will no
doubt continue to be used.

{h) The formula presents information in the form of observed ranges of measurements
unsupported by other data, and this precludes the use of conventional numerical tech-
niques in comparing samples. Instead, simple visual methods are relied on to determine
whether two formulae relate to the same species, and such tests are strongly subjective.
In addition, it is probable that many formulae were measured on single specimens and
are therefore unreliable as a basis for taxonomic comparisons. In view of these dis-
abilities it is recommended that the micrometric formula should be discarded in compara-
tive work; its continued use can only lead to further confusion.

(c) If adequate data are available, reliable and objective comparisons can be made be-
tween sets of measurements by utilizing one of the significance tests commonly used in
biometrics. A method for doing this is outlined, based on the t-test. Techniques of this
kind are only competent to discriminate between groups of data, and not between
single measurements. Samples for comparison should therefore consist of a representa-
tive measurement from each of a number of colonies, and the arithmetic mean is best
suited to this purpose. If provision is to be made for testing new material against estab-
lished species, systematic descriptions must include certain essential statistics for each
variate. These are the sample mean and variance, and the number of specimens in the
samples. An indication of the number of measurements made on each specimen would
also be of assistance, though not essential.

It may be objected that, although the comparative technique suggested here is in
theory superior to the micrometric formula, it suffers from an important practical dis-
advantage, namely, that numbers of specimens are not usually available for comparison,
but only one or two. To this criticism there is only one reply: unless adequate samples

424

PALAEONTOLOGY, VOLUME 9

are available, attempts to make numerical comparisons of any kind are futile and the re-
sults misleading. If there is insufficient material on which to base such a comparison,
there is no point in making one, and it is much better not to do so. Half a dozen speci-
mens constitute a sample of about the minimum permissible size: less than that would
yield results of doubtful value.

Although this paper is exclusively concerned with the numerical comparison of sets of
structural data, it is not suggested that the classification of fenestrate cryptostomes
should rest on these alone. Other factors, not so amenable to mathematical treatment,
must also be considered, for example, the shape of the zooecial chamber. The prime
purpose of the present paper is to draw attention to the shortcomings of the micro-
metric formula as a means of structural comparison, and to stress the need for a
better comparative technique if taxonomic conclusions are to depend on the results
of such comparisons. Finally, it seems possible that significance tests, used along the
lines indicated above, might be a means of discriminating between other kinds of
colonial organisms besides bryozoa, provided that samples of data are constituted as
here suggested.

Acknowledgements. The author is indebted to Professor Alwyn Williams and to Messrs. C. D. Kemp
and D. H. McNally for useful discussions on statistical aspects of the work. Professor Williams also
kindly undertook to read through the manuscript prior to publication. Thanks are due to the Curator
of the Sedgwick Museum for the loan of a specimen from which measurements were taken.

REFERENCES

BURCKLE, L. H. 1960. Some Mississippian fenestrate bryozoa from Central Utah. J. Pcdeont. 34, 1077.
CONDRA, G. E. and ELIAS, M. K. 1944. Study and revision of Archimedes (Hall). Spec. Pap. geol. Soc.
Am. 53, pp. viiiH 243.

ELIAS, M. K. and CONDRA, G. E. 1957. Fenestella from the Permian of West Texas. Mem. geol. Soc. Am.
79, pp. ix+ 1 58.

FISHER, R. A. 1948. Statistical Methods for Research Workers. Edinburgh: 10th ed., xii+351 pp.
FOERSTE, A. F. 1887. Flint Ridge bryozoa. Bull. sci. Labs. Denison Univ. 2, 71.

KOENIG, J. w. 1958. Fenestrate bryozoa in the Chouteau Group of Central Missouri. J. Pcdeont. 32,
126.

MALONE, p. G. and PERRY, T. G. 1965. Fenestellid bryozoans from oncolites in Sappington Sandstone
of Southwestern Montana. Ibid. 39, 41.

m'coy, f. 1844. Synopsis of the characters of the Carboniferous Limestone fossils of Ireland. Dublin:
pp. viii + 207.

MILLER, T. G. 1961. Type specimens of the genus Fenestella from the Lower Carboniferous of Great
Britain. Palaeontology, 4, 221.

1962. North American species of Fenestella from the Carboniferous of Great Britain and Ireland.

J. Paleont. 36, 120.

NEKHOROSHEV, B. 1926. Lower Carboniferous bryozoa from the Kuznetsk basin. Bull. Com. geol.
Leningrad, 43, 1237.

1932. Microscopical method of exploration of Palaeozoic Bryozoa of the fam. Fenestellidae.

Izv. Vses. geol. razved. Ob"d. 51, 279. [Russian with English summary.]

SHRUBSOLE, G. w. 1881. Further notes on the Carboniferous Fenestellidae. Quart. J. geol. Soc. Lond.
37, 178.

SHULGA-NESTERENKO, M. I. 1951. Kamennougolnye fenestellidy ruskoii platformy. Trudy Paleont. inst.
32, pp. 156. [In Russian.]

SIMPSON, G. G. 1941. Range as a zoological character. Am. J. Sci. 239, 785.

TAVENER-SMiTH, R. 1965. A ncw fenestrate bryozoan from the Lower Carboniferous of County Fer-
managh. Palaeontology, 8, 478.

R. TAVENER-SMITH: CLASSIFICATION OF FENESTRATE CRYPTOSTOMES 425

UTGAARD, j. and PERRY, T. G. 1960. Fenestrate bryozoans from the Glen Dean Limestone (Middle
Chester) of Southern Indiana. Bull. Indiana Dep. Conserv. geol. Surv. 19, pp. 32.

WILLIAMS, A. 1962. The stratigraphy and brachiopod faunas of the Barr and Lower Ardmillan Series
(Caradoc) of the Girvan District of SW. Ayrshire. Mem. geol. Soc. Loud., 3.

R. TAVENER-SMITH
Department of Geology,
The Queen's University,

Manuscript received 15 April 1965 Belfast

A NEW PRODUCTID BRACHIOPOD FROM
THE UPPER VISEAN OF SCOTLAND

by K. A. G. SHIELLS

Abstract. Promarginifera Ueantensis gen. et sp. nov. a productid brachiopod from the Upper Visean of Ayr-
shire, Scotland is described. It occurs in the Dockra Limestone 65 ft. below the Visean-Namurian junction.
Morphologically the form has affinities with both the marginiferids and dictyoclostids but on the evidence of a
marginiferid cardinal process and discontinuous marginal ridges it is classified as a marginiferid of the subfamily
Costispiniferinae.

Specimens of an unrecorded productid have been collected from the Upper Visean of
Ayrshire, Scotland. The form was first found at Trearne quarry (NS 37155340) 1-4 miles
N. 108° E. of Beith, and has subsequently been obtained from the neighbouring Hessil-
head quarry (NS 37755315) 0-4 miles N. 288° E. of Hessilhead Farm. Both of these
quarries expose the Dockra Limestone, here 20-25 ft. thick, and up to 10 ft. of the over-
lying argillaceous succession. The base of the Dockra Limestone locally occurs 10 ft.
above the base of the Lower Limestone Group of Scotland and is approximately 65 ft.
below the shale underlying the Top Hosie Limestone which is considered on the evidence
of Cravenoceras scoticimi within the limestone and Cravenoceras sp. indet., within the
shale to approximate to the Visean-Namurian boundary (Currie 1954, pp. 534, 535).
The Lower Limestone Group which is 80 ft. thick in the Beith district consists of a 95%
calcareous-argillaceous succession; the massive section of the Dockra Limestone thins
from 35 ft. near Dairy to 13 ft. near Lugton, being replaced in a west to east direction by
shales and alternations of shales and limestone bands (Richey, Anderson, and Mac-
Gregor 1930, pp. 141-8, fig. 12).

The Dockra Limestone is in part lenticular with the development of a shell-bank facies
of cream-coloured limestones within dark, thinly bedded, argillaceous limestones. Text-
fig. 1 illustrates a measured section on the east end of the working north face of Trearne
quarry which was accessible in May 1964. Cream-brown lenticular bedded calcarenites
overlie a shale parting which in turn rests on grey argillaceous calcilutite forming the
uniformly bedded upper portion of the main calcareous unit. The lenses of calcarenite
are highly fossiliferous, the organic content of some is 90% composed of productids and
crinoid remains, the brachiopod valves forming a shell-supported fabric with the inter-
stices infilled with crinoid debris and matrix. These shell-bank accumulations were de-
posited in a moderately turbulent environment. The productids in the lenses include
species of the genera Productm, Krotovia, Buxtonia, Aiitiquatonia, Eomarginifera, and
Proniarginifera gen. nov. ; in some of the lenses Promarginifera forms 95% of the produc-
tid fauna.

The material for this study was collected from Trearne quarry. The specimens are pre-
served in three different ways; (i) calcareous shells in a calcareous matrix — the normal
unreplaced condition, (ii) siliceous shells in a siliceous matrix — complete silicification of
the rock either along layers parallel to the bedding or in irregularly shaped masses of

[Palaeontology, Vol. 9, Part 3, 1966, pp. 426-47, pi. 67.]

K. A. G. SHIELLS; A NEW PRODUCTID BRACHIOPOD

427

scattered distribution, (iii) siliceous shells in a decomposed ochreous matrix — the most
useful, but relatively rare material. The visceral cavities of all the partially or completely
silicified shells and 90% of the calcareous shells are inhlled with a drusy growth
of transparent quartz. Incomplete silicification and the growth of the quartz crystals
within the visceral cavity have been responsible for the imperfection of internal detail ob-
served in most specimens. Furthermore the infilling is stronger than the shell substance

Calcareous shale with a marine fauna
passing upwards into blue-black
unfossiliferous shale.

Lenses and lenticular beds of cream -brown
limestone with calcareous shale intercalations;
highly fossiliferous with a varied
brachiopod-molluscan-bryozoan macrofauna
in the shale and a brachiopod-crinoid
macrofauna predominant in the limestone.
Partial silicification of the limestone.

Colcareous shale with a brachiopod-
molluscan-bryozoan macrofauna.

Grey argillaceous limestone with a relatively
sparse but varied macrofauno. The
thicknesses of the beds are constant.

shale

limestone

H

silicified limestone

TEXT-FIG. 1. Vertical section through the upper beds of the Dockra Limestone at

Trearne Quarry, Ayrshire.

and its removal from the visceral cavity difficult. It is evident from the nature of the de-
position of the Promarginifera-hearing lenses that the brachiopods are not found in their
growth positions but have undergone varying degrees of reworking. This movement
damaged the shells, especially stripping them of their spines and breaking the ears and
trail. In a collection of approximately 500 specimens only 119 were sufficiently well
preserved for statistical analysis. Of the latter, 62 specimens show the exterior surface
of the pedicle valve, 7 of these also show the pedicle valve interior, and 37 others also
show the brachial valve exterior. The remaining 57 specimens are fragmental brachial
valves showing both interior and exterior surfaces. The collection on which the
statistical analysis is based is deposited in the Hunterian Museum in the University
of Glasgow. All linear measurements are given in millimetres, angular measurements
in degrees.

428

PALAEONTOLOGY, VOLUME 9

SYSTEMATIC DESCRIPTION

Suborder productoidea Maillieux 1940
Superfamily productacea Waagen 1883
Family marginiferidae Stehli 1954
Subfamily costispiniferinae Muir-Wood and Cooper 1960
Genus promarginifera gen. nov.

Plate 67, figs. 1-14

Derivation of name. Latin, — early; rnarginifera — margin bearer.

Type species. Promarginifera trearnensis sp. nov.

Diagnosis. Shell subquadrate, costate and rugate, with reticulate umbo. Spinose pedicle
valve, spines arranged (i) in row along ear, (ii) in row down flank, (hi) scattered over
valve on costae. Ears large and enrolled, demarcated from flank by sharp fold. No
geniculation or thickening of trail of brachial valve. Non-crenulate marginal ridges ex-
tending laterally from cardinal process parallel to hinge-line. Dendritic, elevated adduc-
tor scars, variable cardinal process. Pedicle valve with flabellate diductor scars and
adductor platform.

Description. Shell small, subquadrate, width exceeds straight length. Pedicle valve con-
vex, flanks approximately normal to hinge-line at the posterior before curving anteriorly
on to a broad venter; venter with or without sulcus, trail curved; flanks moderately
steep and umbo incurved forming a uniformly convex visceral disc; ears large and en-
rolled, demarcated from flanks by sharp fold. Brachial valve with posterior of visceral
disc plane or slightly concave, becoming rounded and concave anteriorly to form trail,
no geniculation; trail is a single structure and is not thickened by additional laminae.

Both valves costate and rugate, umbonal region reticulate. Pedicle valve with erect or
suberect spines, arranged, (i) in row near hinge margin, (ii) in curved row down flanks,
and (iii) scattered over the visceral disc, venter, and flanks on supracostal positions. A few
scattered spines may be present on the brachial valve.

Interior of pedicle valve with flabellate diductor scars which are longitudinally ridged
and separated from each other by an elevated platform bearing two ovate adductor
scars; the ears are demarcated from the visceral cavity by a sharp flexure.

Interior of the brachial valve with trilobate marginiferid cardinal process which may
be highly variable in its attitude with respect to the posterior part of the valve and in the
degree of separation of the lateral lobes; posteriorly the median septum is low and
rounded or ill defined, anteriorly it extends into a blade-like structure; away from the
median septum the marginal ridges follow close to the hinge-line and then swing an-
teriorly along the inner margin of the ears to finally die out on the flanks ; brachial ridges
are narrow, low, sometimes obscure, given off at anterior-lateral margin of adductor
scars at a low angle to the hinge-line direction; each adductor has a double dendritic
surface within the attachment area, the anterior scar being upraised along its anterior
border; rows or scattered endospines occur anteriorly.

Remarks. Promarginifera combines both dictyoclostid and marginiferid characters. The
reticulate umbo, flabellate diductor scars, adductor platform, and dendritic adductor

K. A. G. SHIELLS; A NEW PRODUCTID BRACHIOPOD 429

scars are features usually associated with dictyoclostids. On the other hand the cardinal
process is of a type strictly comparable to that found in the marginiferids and the
possession of the marginal ridges is an essential character of that group. Furthermore,
both the size and shape of the shell and the large, enrolled, well-demarcated ears are like
those found in many marginiferids. The striking arrangement of spines on the pedicle
valve of Promarginifera is not completely diagnostic since it is similar to that of several
other genera including dictyoclostids such as Anliquatonia and Costiferina, marginiferids
such as Costispinifera and Liostella, and the alleged buxtoniid ProtonieJIa. An assess-
ment of the taxonomic placing of Promarginifera suggests, therefore, that it is a margini-
ferid with dictyoclostid similarities. Since the marginiferids almost certainly evolved
from the dictyoclostids in Lower Carboniferous times the balance or morphological
attributes found in Promarginifera (a high Lower Carboniferous form) may indicate
phylogenetic affinity with a dictyoclostid stock.

According to Muir-Wood and Cooper (1960, p. 205) two of the four subfamilies of the
Marginiferidae, the Marginiferinae and the Costispiniferinae, are separated on internal
features, especially in the degree of completeness of the marginal ridges. The condition
in which the marginal ridges fuse anteriorly to form a continuous rim around the an-
terior of the visceral disc characterizes the Marginiferinae. If the marginal ridges are dis-
continuous and die out anteriomedianly the form belongs to the Costispiniferinae. The
remaining two subfamilies of the Marginiferidae have special features such as the tubi-
form trail and row of endospines in the brachial valve of the Retariinae and the supple-
mented brachial trail of the Probolioniinae.

Promarginifera has certain external resemblances to the genera Marginifera and
Hystriciilina, both members of the subfamily Marginiferinae. The general shape and size
of the shells, the dimensions of the costae, and to a lesser extent, the distribution of
spines, are similar in all three genera. Another marginiferinid, Eomarginifera, is found
associated with Promarginifera but can be easily distinguished externally by its six
symmetrical spine bases. In internal detail, however, Promarginifera has discontinuous
marginal ridges and is therefore distinct from the marginiferinids; it is clearly a member
of the Costispiniferinae and is the first British genus to be ascribed to that group.
Promarginifera shares some external features with other members of the Costispini-
ferinae, notably with Costispinifera, Liostella, and EUiotella. Promarginifera can be dis-
tinguished from Costispinifera by its less numerous spines, non-crenulate marginal
ridges, non-geniculate but more costate brachial valve, and shorter endospines. Ex-
ternally the pedicle valve of Promarginifera differs from that of Liostella in having a re-
ticulate umbo and a row of spines on the ears close to the hinge-line. Internal distinc-
tions include the more elevated and smooth marginal ridges of Promarginifera and
differences in detail of the cardinal process, median septum, and adductor scars.
EUiotella is more coarsely costate than Promarginifera and does not possess rugae on the
pedicle valve. The orientation of the two spine rows near the hinge of EUiotella is
different from that of Promarginifera. Internally, the marginal ridges of both EUiotella
and Costispinifera diverge from the hinge immediately after leaving the cardinal process;
they cannot, therefore, buttress the articulating surface as do those of Promarginifera
(see description of P. trearnensis). The genus Protoniella simulates Promarginifera ex-
ternally, particularly in the distribution of the spine bases; internally, however, they can
be readily distinguished by the buxtoniid cardinal process and antron of the former.

430 PALAEONTOLOGY, VOLUME 9

Promarginifera trearuensis sp. nov.

Plate 67, figs. 1-14

Diagnosis. As for genus.

Type specimens. The type series consists of a holotype and 118 paratypes, the total collection on which
the statistical analysis is based.

Specifications of the holotype: Hunterian Museum number L6202; Plate 67, figures 1, 2, 3:

Hinge width 19-5 mm.

Curved length of pedicle valve 29 0 mm.

Straight length 13-4 mm.

Thickness of visceral cavity (inclusive of shell) 6 0 mm.

Umbonal angle 90°

Extension of umbo posterior to hinge-line 0-7 mm.

Width of ears 8 5 mm.

Length of ears 3-6 mm.

Height of ears 0-7 mm.

Number of rugae 4

Number of costae 24

Nonsulcate; width of flat zone on venter 2 0 mm.

Number of spines on pedicle valve exterior 99

The figured paratypes (PI. 67, figs. 4-14) are given Hunterian Museum numbers L6203-11, and the
remaining paratypes are given Hunterian Museum numbers L6212/1-107.

Description. The shells are small (Muir-Wood and Cooper 1960, p. 18) with a sample
mean hinge width (= maximum width) of 17-3 ±1-0 mm. The size of the species is more
fully described by the statistics of Table 1 (all confidence limits are quoted at the 0-01
level of probability unless otherwise stated). The shell outline is subquadrate (PI. 67, figs.
1, 3, 5); the outer surface of the pedicle valve is strongly convex and the outer surface of
the brachial valve is strongly concave; neither valve is geniculated (text-fig. 2). The cur-
vature of the posterior to anterior profile of the pedicle valve in the median plane of the
shell (text-fig. 3, method after Prentice 1956) is a compound of three logarithmic spirals
which from umbo to trail have spiral angles of 37°, 44°, and 81° respectively, the initial
spiral being subject to secondary modification and is not precisely defined. In all

EXPLANATION OF PLATE 67
Promarginifera trearuensis gen. et sp. nov.

Figs. 1-3. Holotype; Hunterian Museum L6202. 1, ventral view; 2, dorsal view; 3, lateral view. X3.
Fig. 4. Paratype; H.M. L6209. Interior of brachial valve showing cardinal process of type 1. x2-5.
Figs. 5-6. Paratype; H.M. L6204. 5, posterior view showing reticulate umbo; 6, ventral view showing
costation of type 2. X 3.

Fig. 7. Paratype; H.M. L621 1. Interior of brachial valve showing cardinal process of type 5. X 2-5.
Fig. 8. Paratype; H.M. L6207. Lateral view showing enrolment of ear. x3.

Fig. 9. Paratype; H.M. L6208. Interior of pedicle valve showing diductor scars and adductor platform.

X4.

Fig. 10. Paratype; H.M. L6210. Interior of brachial valve showing cardinal process of type 6. x2-5.
Fig. 11. Paratype; H.M. L6203. Posterior view. x2-5.

Fig. 12. Paratype; H.M. L6206. Antero-ventral view showing costation of type 6. X 3.

Figs. 13-14. Paratype; H.M. L6205. 7, posterior view showing internal mould of diductor scars and
adductor platform; 8, antero-ventral view showing costation of type 3. x2-5.

Palaeontology, Vol. 9

PLATE 67

SHIELLS, Carboniferous Promarginifem gen. nov.

K. A. G. SHIELLS: A NEW PRODUCTID BRACHIOPOD

431

TEXT-FIG. 2. A median section of the shell of P. trearnemis
showing the profile of the valves, the thickening of the
pedicle valve in the region of the adductor platform, and
a large cardinal process.

mm.

T ;xT-FiG. 3. A plot of the straight length vectors which describe the growth spirals of the curved length
o f the pedicle valve. The principal changes in slope of the nine profiles plotted on (a) represent the
change in spiral of the curves x, y, and z on {b). This is shown on the summation diagram (c) and the

angles a, b, and c are the spiral angles.

432

PALAEONTOLOGY, VOLUME 9

specimens the hinge-line is straight and the hinge width is the maximum width of the
shell ; neither valve possesses an interarea.

GROWTH OF VALVES

The relative growth of the two valves and the increase in the thickness of the visceral
cavity of P. trearnensis follow a pattern for strongly concavo-convex productids which
is discussed elsewhere (Shiells 1965, pp. 878-80). Since the curvature of the valves is con-
stant (within 5% angular variation), measurements of any individual of the sample

mm,

-6

-4

-2 th

5 10 15 20 25 30 35 mm.

pv

TEXT-FIG. 4. A plot on logarithmic coordinates of uniform growth increments
to the curved length of the pedicle valve (/iv) against (i) the corresponding
increments to the curved length of the brachial valve (bv) shown as dots, and
(ii) the corresponding increments to the thickness of the visceral cavity (th)
shown as open circles.

It is observed that;

1 . For unit increase in the growth of the curved length of the pedicle valve, the corresponding consecu-
tive growth increments to the curved length of the brachial valve (u) systematically diminish until
the pedicle valve has reached a length of approximately 22-5 mm., (b) progressively increase as the
pedicle valve grows from approximately 22-5 to 32-5 mm., and (c) are equal in length for the remain-
der of growth (Table 2, text-fig. 4). From the first reliable measurements, i.e., when the curved length
of the pedicle valve equals 5 mm., and the curved length of the brachial valve equals 3-8 mm., to the
growth stage in which the curved length of the pedicle valve equals 1 1 mm., and the curved length of
the brachial valve equals 7-5 mm., there is a simple allometric growth relationship between the valves.
The allometric growth constant is 0-87 (text-fig. 5a). Subsequent growth of the two valves, however,
deviates from simple allometry. As the two valves increase in curved length to approximately 22
mm. and 10-5 mm. for the pedicle and brachial respectively, there is a systematic decrease in the
value of the allometric coefficient from 0-87 to 0-2. From this point until the growth stage in which

K. A. G. SHIELLS: A NEW PRODUCTID BRACHIOPOD

433

the curved length of the pedicle valve is 32-5 mm. and the curved length of the brachial valve is
16-5 mm., the values of the allometric coefficient rapidly increase to 1-98. Due to the parallelism of
the valves further values for the allometric coefficient decrease until growth ceases.

2. For unit increase of the curved length of the pedicle valve as above, the corresponding increments to
the thickness of the visceral cavity (a) are constant at first and then progressively increase until the
pedicle valve has attained a length of approximately 27-5 mm., and (b) successively decrease as
the pedicle valve increases from approximately 27-5 mm. to 32-5 mm. (Table 2, text-fig. 4). When the
growth of the pedicle valve reaches approximately 32-5 mm., the brachial valve has become parallel
to the pedicle valve. Since no specimen of P. trean:eusis shows any thickening of the existing brachial
valve or secretion of a new trail to the brachial valve, it follows that no further increase in the thick-
ness of the visceral cavity can take place once this condition has been achieved. Allowing for a total
maximum variation in the thickness of the shell substance of the valves, the sulcation, and costation,
the estimate of the maximum thickness of the visceral cavity inclusive of the shell is 8± 1-5 mm. As
shown on text-figure 5b an isometric growth relationship (coefficient of allometry equal to unity)
exists between the thickness of the visceral cavity and the curved length of the pedicle valve, until the
latter reaches 17-5 mm. For further increase in the curved length of the pedicle valve, subsequent
coefficients of allometry successively rise to a maximum value of 1 -7, corresponding to a curved length
of the pedicle valve of 27-5 mm. Finally, the values of the coefficient of allometry diminish in turn
to 0-95 when the maximum thickness of the visceral cavity is attained.

3. For unit increase in the growth of the curved length of the brachial valve, the corresponding
consecutive growth increments to the thickness of the visceral cavity (a) maintain a fairly constant
size at first until the curved length of the brachial valve is 9-9-5 mm., {h) progressively increase as the
curved length of the brachial valve is extended to 13-13-5 mm., and (c) successively diminish until
the parallelism of the valves prevents further enlargement of the visceral cavity (Table 3, text-fig. 6).
The relative growth of the thickness of the visceral cavity and the curved length of the brachial valve
is shown on text-fig. 5r; the deviations from standard allometry are complex.

4. As shown on Table 4 the correlation of the hinge width with the curved length of the pedicle valve
and the straight length is significant at a high level. However, such a correlation is subject to two inter-
pretations. Firstly, in the light of the complex growth relationships between the curved lengths of
the two valves it might be expected that an equally involved relationship would exist between the
length of a valve and the hinge width, the correlation of the total data simply averaging the effect of
the sample distribution about the true growth curve of the two characters. The growth of the two
valves is controlled, however, by the common origin and common line of termination of any growth
stage. In a comparison of the length with the width of the shell there is no similar direct functional
dependence although an indirect cause is presumed. In the second place, therefore, it is possible that
the growth of the hinge width matched that of either of the valves largely to substantiate the correla-
tion. Unfortunately the grcvvth laminae on the external surface of the shell proved insufficiently
distinct to test these possibilities.

The slope of the flanks is inclined 65-80° depending on the completeness of the trail
(PI. 67, figs. 6, 11, 14). The shape of the venter is slightly convex, approximately flat, or
sulcate (Table 1, PI. 67, figs. 1, 5, 9, 12, 13). Of 44 specimens 18 (41%) possessed a sulcus
and 26 (59%) were nonsulcate. A gradation exists between (i) nonsulcate forms, (ii)
forms with a closed sulcate depression restricted to the anterior part of the visceral disc
and the posterior end of the trail, and (iii) forms with an open sulcus beginning on the
anterior part of the visceral disc and extending to the anterior extremity of the trail. The
sulci of the open type are deepest on the anterior part of the visceral disc and the pos-
terior end of the trail as demonstrated by the data of Table 5. Examination of the degree
of correlation between the maximum width and maximum depth of the sulcus on 18
specimens just failed (by 0-001) to be significant at the 0-05 level of probability (Table 4).
No significant correlation occurs between the height of the visceral cavity and the inward
folding of the venter as possibly anticipated if the sulcation represented a simple struc-
tural modification of the test to strengthen an increasingly arched internal lumen (Table 4).

434

PALAEONTOLOGY, VOLUME 9

The umbo shows moderate expansion with a mean umbonal angle of 99-5 ±4-8^
(Table 1). In all specimens the umbo projects posterior to the hinge (Table 1, PL 67, figs.
3, 6, 11, 14). Inflation of the umbonal cavity is expressed by corresponding increments
(correlated at the 0-05 level of probability) to the umbonal angle and the distance to
which the umbo extends posterior to the hinge (Table 4). Such inflation accompanies
a general increase in gibbosity of the visceral cavity as shown by the significant correla-
tion at the 0-05 level of probability between the height of the visceral cavity (plus the
shell thickness of each valve) and the extension of the umbo posterior to the hinge. How-
ever, the very low (not significant) correlation between the hinge width and umbonal
angle (indeed a low negative value was obtained — Table 4) suggests that an overall
volumetric increase in the capacity of the shell interspace is not necessarily a function of
gibbosity.

Prominent ears are an intrinsic feature of P. trearnensis (PI. 67, figs. 1, 3, 8, 9). In
shape they are triangular and enrolled, approximating either to a cylindrical sector or to
a conical sector increasing in width and curvature towards the postero-lateral extremi-
ties. The mean dimensions of the structures are given on Table 1. Although the correla-
tion between the width of the ear and the hinge width is significant at the 0-01 level and
the correlation between the ear width and ear length is significant at the 0-05 level, the
height of the ear, a measure of its curvature or degree of enrolment, varies independently
(Table 4). The steeply inclined enrolled anterior slopes of the ears oppose the steeply in-
clined umbonal lateral surfaces to form a conspicuous groove, which either demarcates
the slant side of the conical sectors or traces obliquely round the more cylindrical types of
ear surface, increasing in depth towards the lateral margin of the shell.

EXTERIOR OF PEDICLE VALVE

The exterior surface of the pedicle valve is conspicuously costate (PI. 67, figs. 1, 5, 12,
13), the mean number of costae being 26±2 at the anterior margin (Table 1). Alternate
costae and sulci diverge from an umbonal focus, each costa and sulcus proportionately
increasing in width towards the anterior. The spread of the fan of costae is supplemented
on the central, anterior, and antero-lateral regions of the visceral disc through additional
costae being inserted or produced by bifurcation. On reaching the anterior of the visceral
disc or the posterior part of the trail, expansion ceases and the costae and intercostal
sulci continue as a set of parallel ridges and grooves down the length of the venter and
antero-lateral surfaces of the trail to the valve extremity. Rarely, coalescence of costae
takes place on the trail. Three properties of the costation (text-fig. 7), (i) the maximum
widths of the costae attained on any specimen, (ii) the regularity of costation, and (hi)
the shape or combinations of shape of the costae in transverse section, have been used to
arbitrarily subdivide the continuously variable pattern for descriptive purposes. The six
groups into which the costation may be subdivided are given in Table 6. The costation
of 50 specimens has the following distribution; group 1 — 10 specimens (20%), group 2 — 9
specimens (18%), group 3 — 16 specimens (32%), group 4 — 8 specimens (16%), group 5 — 6
specimens (12%), and group 6 — 1 specimen (2%).

The exterior surface of the pedicle valve is rugate (PI. 67, figs. 6, 11); the mean number
of rugae is 7Jzl (Table 1). The trace of the rugae and interrugal sulci approximates to
a series of semi-ellipses about an umbonal origin. The majority of the rugae traverse the
visceral disc but 0-2 of the inner rugae and 1-3 of the outermost rugae die out on the

K. A. G. SHIELLS: A NEW PRODUCTID BRACHIOPOD

435

S 20 ' 25 30 ' 35 mm

CURVED LENGTH OF PEDICLE VALVE

TEXT-FIG. 5. Deviations from simple allometry: (a) Coefficient of allometry for the relative growth of
the length of the brachial valve and the length of the pedicle valve plotted against the length of the
pedicle valve, (b) Coefficient of allometry for the relative growth of the thickness of the visceral cavity
and the length of the pedicle valve plotted against the length of pedicle valve, and (c) Coefficient of
allometry for the relative growth of the length of the brachial valve and the thickness of the visceral
cavity plotted against the length of the brachial valve.

mm.

I

4 6 8 lO 12 14 lb mm,

bv

TEXT-FIG. 6. A plot on logarithmic coordinates of uniform growth increments to the curved length
of the brachial valve (/tv) against the corresponding increments to the thickness of the visceral cavity (r/i).

Gg

G 4179

436

PALAEONTOLOGY, VOLUME 9

umbonal slopes. As the distance from the umbo increases there is a proportionate in-
crease in the dimensions of the rugae. Furthermore, each of the outer rugae changes
shape from an approximately equidimensional structure at the base of the umbonal
slopes to one in which the posterior to anterior width of the ruga exceeds the height by
a maximum factor of ten. Eight per cent, of the rugae bifurcate across the visceral disc.
Measurements from a single specimen to demonstrate the progressive change in size of
the rugae are given on Table 7. The rugae and costae lie normal to one another over the
mm.

(

TEXT-FIG. 7. Variation of costation on the pedicle valve. The costation is arbitrarily ;

grouped into six categories on the basis of regularity, size, and shape (b). Plots of I

examples of these groupings are shown on (a).

shell surface of the posterior region of the visceral disc, and the incidence of costae,
intercostal sulci, rugae, and interrugal sulci produces a reticulated pattern. Close to the
umbo the internodal intercepts on a costa are approximately twice the internodal inter-
cepts on a ruga. Flowever, as the rate of expansion of the rugation pattern exceeds that
of the costation, the reticulation becomes progressively elongate in an anterior direction
so that the internodal intercepts on a costa approaching the central part of the visceral
disc are four times the internodal intercepts on a ruga in the same region. Measurements
from a single specimen to show this relationship are given in Table 8. No significant
association occurs between the type of costation and the number of rugae (x^ contin-
gency only shows correlation at the 0-95 level of probability). The limitation of the rugae -
to the posterior surface of the visceral disc suggests that such corrugation may structu-
rally reinforce the tightly convex umbonal slopes; however, no significant correlation
(r = —0-244) occurs between the frequency of the rugae and the size of the visceral
cavity. j

The pedicle valve is highly spinose with 50-100 erect, or suberect, halteroid spines pro- 1
jecting from its surface (text-fig. 8; PI. 67, figs. 1, 2, 11, 13). The spines are situated on [

K. A. G. SHIELLS: A NEW PRODUCTID BRACHIOPOD

437

radii of umbonal origin and the spine bases progressively increase in diameter away from
the umbo. The two spine rows on each side of the umbo which are closest to the hinge-
line are especially distinct. They are characterized by a more regular spacing of the spine
bases and, for an equivalent distance from the umbo, the spine bases are slightly greater
in diameter than those on the remainder of the valve (Table 9). The posterior row is
situated on the posterior enrolled slope of the ear and is subparallel to the hinge-line
(maximum divergence from the hinge-line 15°). The anterior row is situated down the

a

I CM

TEXT-FIG. 8. Spinosity of the pedicle valve (ventral and lateral views), (a) Plot of spine bases on the holo-
type showing their distribution on 26 radii of umbonal origin (a-z): spine rows b to v are supracostal;
spine rows a and z occur on the ventral and posterior ear slopes, (b) Reconstruction of spines; positions
of the spine bases as for holotype, length and curvature of spines based on collective information.

flank and initially diverges from the posterior row at an angle of 25-30°. All of the spines
are supracostal except those on the ears and those on the posterior part of the row
down the flanks.

The exterior surface of the pedicle valve shows concentric growth lamellae; the fre-
quency of the lamellae is of the order of 7-9 per mm.

INTERIOR OF PEDICLE VALVE

The interior of the pedicle valve is strongly concave and is for the greater part an exact
reflection of the convex exterior. The internal ridges and furrows which radiate from the
umbo coincide with the intercostal sulci and costae of the exterior, simulating along
the anterior margin of any growth stage a series of truly concentric folds through which

438

PALAEONTOLOGY, VOLUME 9

the thickness of the shell remains constant. In the postero-central region, however, the in-
terior of the pedicle valve is considerably modified by the attachment areas of the muscles
(PI. 67, figs. 9, 14). On either side of the mid-line of the valve two large (length 12-1 5 mm.,
maximum width 5-6 mm.), flabellate diductor scars diverge away from the umbo; each
is longitudinally ridged. Posteriorly the diductor scars meet, but at a distance of approxi-
mately 5 mm. from the hinge-line they become separated by an upraised platform 1-5-2
mm. high, 5-6 mm. long, and 2-2-25 mm. wide. This is the adductor platform and it
bears two ovate adductor scars near its posterior end, each 1 mm. wide and 3 mm. long.
At a distance of approximately 1 mm. from the umbo the posterior margin of the
pedicle valve becomes internally thickened; on occasion the internal thickening forms
a distinct ridge up to 0-5 mm. high. Laterally this thickening or ridge bends towards the
anterior and passes into the sharp flexure which abruptly demarcates the visceral cavity
from the ears.

The interior surface of the pedicle valve is roughened by the anteriorly projecting cal-
cicular rods of the taleolae. The density of thetaleolae (i.e. the number per unit length
of the pedicle valve measured in thin section) ranges from 0-15 per mm. in various parts
of the valve, with a mean density of 1 1 per mm. for the anterior region of the visceral
disc and trail. Most of the taleolae are curved, forming a low angle with the plane of the
shell surface at their termination within the shell substance, but projecting from the in-
terior of the valve at an angle of approximately 50-70° to the valve surface.

EXTERNAL SURFACE OF BRACHIAL VALVE

The concave external surface of the brachial valve is costate and weakly rugate (PI.
67, fig. 3). The costation is the exact reflection of that of the exterior surface of the pedicle
valve, the costae of the latter being represented by the intercostal sulci of the brachial
valve exterior. The postero-central portion of the valve is featureless except for a small
protuberance situated on the hinge-line or the umbo (text-fig. 9). Up to six ill-defined
rugae may occur, at least half the number being incomplete. The ears are demarcated
from the rest of the shell by sharp folds which increase in amplitude towards the margin
of the flanks and are complementary to the grooves demarcating the ears from the visceral
disc of the pedicle valve. On two specimens several broken spines have been observed pro-
jecting at high angles to the anterior part of the visceral disc. The exterior surface of the
brachial valve is otherwise smooth except for growth lamellae which are irregularly
spaced but of the order of 7-9 per mm., as on the pedicle valve.

INTERNAL STRUCTURE OF BRACHIAL VALVE

Internally the bilateral symmetry of the brachial valve (PI. 67, figs. 4, 7, 10) is reflected
on either side of a slender median septum which extends 70-90% of the valve length from
the posterior margin. The median septum has a mean length of 8-4±0-5 mm. (Table 1),
and consists of two structural elements. The anterior 55% of the structure has a triangu-
lar blade-like form, 0-2-0-3 mm. wide, which arises from a position level with, or an-
terior to, the mid-point of the adductor scars. Towards the anterior the blade increases
in height attaining a maximum elevation at, or within 0-5 mm. of, its anterior termina-
tion; the mean value for the maximum height of the median septum is 1-33 ±0-28 mm.
(Table 1). The posterior section of the median septum is less well defined, comprising
a low structure, semicircular in cross-section, with a maximum width of approximately

K. A. G. SHIELLS: A NEW PRODUCTID BRACHIOPOD

439

1 mm. and a maximum height of approximately 0-4 mm. Anteriorly it is contiguous with
the blade-like portion of the septum; posteriorly it expands and fuses with the inflated
ends of the marginal ridges to form a reinforced attachment with the cardinal process
(PI. 67, figs. 4, 10). In those specimens with a large cardinal process the posterior section
of the median septum is weakly developed with an elevation of usually no more than
0-1 mm. The contour of the brachial valve is modified immediately anterior to the car-
dinal process by a dorsally concave and ventrally convex fold. The amplitude of this fold
is slight with a maximum value of approximately T5 mm.; in shape it is elliptical with
the long axis (3-6 mm.) orientated in a posterior to anterior direction. When the posterior
section of the median septum is reduced in size the amplitude of the fold is generally
large (1-T5 mm.), the corrugation of the valve appearing, therefore, to compensate in
part for the rigidity of a robust shaft. The value for the coefficient of correlation (r -=
0-458) of the length and height of the median septum is not significant, and similarly
there is no simple relationship between the size of the median septum and the size of the
visceral cavity.

TYPES OF CARDINAL PROCESS AND MUSCLE SCARS

The cardinal process is a highly variable structure assuming a general costispiniferid
form (Muir-Wood and Cooper 1960, pp. 28-29). It is trilobed, usually with a small
median lobe, and is sessile; the overall dimensions are given in Table 1. Different types
may be distinguished according to the degree of separation of the lateral lobes, and the
inclination of the process with regard to the surface of the brachial valve (text-fis. 9;
PI. 67, figs. 4, 7, 10).

Type 1 (text-fig. 9-1 ; PI. 67, fig. 4) is characterized by a horizontal posterior to an-
terior axis of the lateral lobes so that the process is in effect a lobed termination to the
median septum. At the posterior extremity all three lobes are distinct, the small median
lobe being little more than a V-shaped inflexion between the semiconical lateral lobes.
The axis of the median lobe is disposed at 50° to the valve surface, and the anterior end
continues into a distinct protuberance on the dorsal surface. The attachment area is di-
rected postero-dorsally and the surface is roughened by the accentuation of growth
lamellae. In ventral aspect the lateral lobes converge towards a mid-line before becoming
fused anteriorly with the median septum. The cardinal process is buttressed by the
swollen ends of the marginal ridges.

The cardinal process of type 2 (text-fig. 9-2) rises ventrally from the plane of the pos-
terior part of the brachial valve at a steep angle (60°). When viewed from the posterior,
two inflated lateral lobes are separated by a narrow median lobe which has a consider-
able extension in a postero-dorsal direction so that it makes an angle of only 30° with
the brachial valve surface. The flexuous surface to which the diductor muscles were
attached is roughened by irregular growth lamellae. Anterior to the median lobe the
hinge-line bears a conspicuous protuberance which extends dorsally for approximately
0-3 mm. at right angles to the valve surface. In ventral aspect the lateral lobes converge
and join the expanded end of the median septum.

The third type of cardinal process (text-fig. 9-3) has a similar upturned attitude in re-
lation to the posterior part of the brachial valve to that of type 2. It differs from type 2,
however, in the erect position of the median lobe and the anterior extension of the attach-
ment surfaces of the lateral lobes, two features that enable the sinuous profile of the

440

PALAEONTOLOGY, VOLUME 9

complete attachment surface to be clearly seen in the ventral aspect of the valve. The
attachment surface is almost normal to the plane of the valve and is made asperous by
uneven growth lamellae. Dorsally the cardinal process tapers into a distinct protuberance
which has a relief of 0-2-0-3 mm., at right angles to the exterior surface of the valve. The
process is supported by the marginal ridges which rise ventrally and fuse with the
lateral lobes.

ventral lateral posterior dorsal

TEXT-FIG. 9. Variation in the form of the cardinal process of P. trearnensis.

The strong median lobe of type 4 (text-fig. 9-4) distinguishes it from the other types of
cardinal process. The general attitude of the process is similar to that of types 2 and 3 but
the median lobe extends further postero-dorsally and is much incurved, forming a lip
which overhangs the attachment surface. The ventral surface of the median lobe is con-
cave and bears a small pit. The axis of the median lobe curves slightly but is approxi-
mately at an angle of 50° to the plane of the posterior part of the brachial valve.
Dorsally the tip of the median lobe forms a small protuberance. The attachment surfaces
of the lateral lobes are almost closed by the enlargement of the median lobe ; the whole
of the attachment surface is roughened by uneven growth lamellae. In ventral aspect the
cardinal process is supported by a swelling of the marginal ridge-median septum inter-
angle.

The cardinal process of type 5 (text-fig. 9-5; PI. 67, fig. 7) is distinguished both by its
size and attitude in relation to the brachial valve. The process is bulbous and its axis has

K. A. G. SHIELLS: A NEW PRODUCTID BRACHIOPOD 441

an antero-ventral to postero-dorsal incline, making an acute intersection with the
brachial valve at 40°. The lateral lobes are much inflated and extended in an antero-ven-
tral direction. In text-fig. 9-5 and Plate 67, fig. 7, the cardinal process has lateral lobes
which are fused in a unified structure ; the same type of cardinal process also occurs in
which the lateral lobes are quite distinct anteriorly (text-fig. 9-5a). The median lobe is
small and its dorsal extremity is posterior to the hinge-line and to a large protuberance
0-3-0-4 mm. in height which extends dorsally from the hinge-line at the mid-point of the
valve. The involute attachment area is roughened in the usual way by growth lamellae.
The cardinal process is supported by the thickened ends of the marginal ridges and the
median septum.

The cardinal process of type 6 (text-fig. 9-6; PI. 67, fig. 10) is similar to type 5 except
that the incline of the process is increased so that the lateral lobes are directed dorsally,
at right angles to the plane of the posterior part of the brachial valve. The median lobe
has a conspicuous postero-dorsal extension, its axis being 40° to the surface of the
valve. A protuberance extends dorsally, anterior to the median lobe. The attachment
surface is roughened by growth lamellae and the process is supported by a swelling of
the marginal ridges as they unite across the median septum.

The distribution of the six types of cardinal process in a sample of 38 specimens is:
type 1 — 7 specimens (18%), type 2 — 4 specimens (10-5%), type 3 — 6 specimens (16%), type
4 — 6 specimens (16%), type 5 and 5a — 7 specimens (18%), type 7 — 8 specimens (2T5%).

The marginal ridges are asymmetrical elevations on the brachial valve which close the
visceral cavity posteriorly and postero-laterally (PI. 67, figs. 7, 10). From the cardinal
process the ridges extend within 7° of the hinge-line a mean distance of 6-1 ±0-5 mm.
(Table 1), before swinging anteriorly through an angle of 110-145° and continuing a mean
distance of 4-5-6-5 mm. to their termination on the flanks. At the mid-point of the valve
they join the median septum and frequently swell to support the cardinal process (text-
fig. 9). The section of the marginal ridges which extends from a quarter to a half of the
distance along the hinge-line from the median septum to the lateral extremity of the ears,
represents the main articulating surface of the two valves. It is only in these two short
lengths on either side of the cardinal process that the marginal ridges are sulficiently
close to the posterior edge of the brachial valve for them to combine with the shell sub-
stance of the main part of the valve in producing a reinforced pivot against the strain of
a crude leverage of the valves imposed by the contraction of the diductor muscles.

The adductor scars are two subtriangular elevated regions, the long sides of which are
parallel to, and OT-0-3 mm. laterally separated from, the anterior blade of the median
septum. The overall dimensions of the adductor scars and their location on the valve sur-
face are given in Table 1. Each adductor scar has two main fields, each of which bears
dendritic markings (text-fig. lOn). The anterior margin of the anterior field is steeply
elevated above the valve surface (text-fig. 10/?).

BRACHIAL RIDGES AND ENDOSPINES

The brachial ridges originate at the antero-lateral region of the adductor scars (PI. 67,
figs. 7, 10) and extend antero-laterally in a loop circumscribing a comparatively smooth,
ovate area of the valve surface. The loop is incomplete at its anterior end (PI. 67, fig. 7).
The dimensions of the loop are given in Table 1 ; the positions of the measurements are
shown in text-fig. 11.

442

PALAEONTOLOGY, VOLUME 9

I mm.

TFXT-FiG. 10. (a) Reconstruction of adductor scars on either
side of the median septum, each scar having two fields of
attachment with dendritic relief, (b) Profile of adductor
scar (section AB on (a)) showing steeply elevated anterior
margin of the anterior field.

width A

TEXT-FIG. 1 1 . Diagram to show width A, width B, and angle of offset of the brachial

loop (see Table 1).

It is estimated that 80-150 endospines extend anteriorly from the interior surface of
the brachial valve. They are situated in a zone on the strongly convex anterior and an-
tero-lateral slopes between the relatively flat postero-central disc and the trail. The
majority are supracostal although others occur in sulci; they range in size up to 2 mm.
and are inclined to the valve surface at approximately 45°. The taleolae of the brachial
valve are variously distributed but attain a mean density of 1 1 per mm. at the anterior of
the visceral disc and trail. As with the taleolae of the pedicle valve they are curved and
anteriorly directed.

K. A. G. SHIELLS: A NEW PRODUCTID BRACHIOPOD

443

APPENDIX OF TAXONOMIC DATA
Table 1

Linear measurements in millimetres, angular measurements in degrees

Univariate Statistics of Selected Characters

Character

N

.V

5

OR

V

/i (P-0-05)

0-01

Hinge width of pedicle valve 45

17-3

2-49

13-22

14-38

16-6-18-0

16-3-18-3

Curved (or surface) length

of pedicle valve

19

32-9

502

25-41

15-13

30-5-35-4

29-6-36-3

Straight length
Thickness (or height) of

19

13-6

1-42

11-4-17-2

10-43

12-9-14-3

12-7-14-5

visceral cavity inclusive
of shell

29

5-4

1-82

40-9-5

24-70

5 -0-5 -9

4-8-6- 1

Width of sulcus

18

5-9

L39

3-8

23-73

5 -2-6- 5

4-9-6-S

Depth of sulcus

Width of fiat zone on venter

18

0-87

0-356

0-5-1 -5

40-69

0-70-1-05

0-63-1-12

of non-sulcate forms

26

4-5

1-98

1-8

45-47

3-5-5-1

3-3-5-4

Umbonal angle
Extension of umbo posterior

32

99-5

9-81

85-135

9-86

95-9-103-0

94-7-104-1

to hinge

32

0-75

0-245

0-3-1-3

32-65

0-66-0-83

0-63-0-87

Ear width

27

7-32

1-12

5-10

15-20

6-47-8-17

6-17-8-47

Ear length

27

303

0-606

2-4

19-99

2-79-3-27

2-70-3-36

Ear height

27

0-85

0-351

0-2-1-7

41-32

0-71-0-99

0-66-1-04

Number of costae

34

26-0

4-42

18-39

16-99

25-27

24-28

Number of rugae

36

70

2-14

3-13

30-51

6-8

6-8

Length of median septum
Height of median septum at

22

8-4

0-91

7-0-10-5

9-91

8-02-8-83

7-87-8-97

anterior end

16

1-33

0-38

0-7-1 -9

28-74

1-13-1-53

1-05-1-61

Length of cardinal process

33

0-94

0-34

0-6-1 -9

36-5

0-86-1-02

0-83-1-05

Width of cardinal process
Length of marginal ridge

33

117

0-26

0-6-1 -7

22-4

1-08-1-27

1-05-1-30

along posterior margin
Height of marginal ridge

1 1

61

0-54

5-3-7-0

8-89

5 .74-6-46

5-56-6-62

along posterior margin
Width of marginal ridge

28

0-25

0-064

0-2-0-4

25-53

0-22-0-27

0-21-0-28

along posterior margin

28

0-32

0-121

0-2-0-6

37-12

0-28-0-37

0-26-0-39

Length of adductor scar

19

2-43

0-40

1-7-3-3

16-56

2-23-2-62

2-16-2-69

Width of adductor scar
Distance of posterior end of

20

1-61

0-29

1 -3-2-5

17-91

1-48-1-75

1-43-1-80

adductor scar from hinge-
line

18

L88

0-28

1 -3-2-3

15-15

1-75-2-03

1-69-2-08

Distance of adductor scar

from median septum

17

017

0-069

0-10-3

40-21

0-13-0-21

0-12-0-22

Width A of brachial loop

5

4-28

0-415

40-5-0

9-69

3-76-4-79

3-43-5-13

Width B of brachial loop
Angle of offset of brachial

5

3-40

0-447

2-8-40

13-15

2-16-4-64

1-34-5-46

ridges

5

104

8

95-110

7-9

94-1 14

87-121

Width of brachial ridge
Distance of posterior end of

5

0-27

0-097

0-15-0-40

36-10

0-15-0-39

0-07-0-47

brachial ridge from hinge-
line

5

3-50

0-436

3 0-4-0

12-45

2-96-4-04

2-60-4-38

444

PALAEONTOLOGY, VOLUME 9

Table 2

Measurements to compare the relative increase in the length of the Brachial Valve and
the thickness of the Visceral Cavity for unit increases in the length of the

Pedicle Valve

Increment to

Increment to

Number of

Curved length

curved length

Curved length

thickness of

Thickness of

increments

of p.v.

of b.v.

of b.v.

visceral cavity

visceral cavity

1

2-5

208

208

—

—

2

5 0

1-75

3-83

—

0-83

3

7-5

1-58

5-41

0-42

1-25

4

100

1-54

6-95

0-42

1-67

5

12-5

1-33

8-28

0-42

2-09

6

15 0

0-92

9-20

0-42

2-51

7

17-5

0-75

9-95

0-42

2-93

8

20 0

0-34

10-29

0-59

3-43

9

22-5

0-29

10-58

0-66

4-09

10

250

0-67

11-25

0-75

4-84

11

27-5

0-91

12-16

0-83

5-67

12

300

1-92

14-08

0-66

6-33

13

32-5

2-42

16-50

0-50

6-83

14

3 5 0

2-50

19-00

15

37-5

2-50

21-50

Table 3

Measurements to compare the relative increase in the thickness of the Visceral Cavity
for unit increase in the length of the Brachial Valve

Increment

Number of

Curved

to thickness of

Thickness of

increments

length of b.v.

visceral cavity

visceral cavity

1

0-8

2

1-7

3

2-5

0-58

4

3-3

0-08

0-66

5

4-2

0-21

0-87

6

5-0

0-21

1-08

7

5-8

0-21

1-29

8

6-7

0-21

1-50

9

7-5

0-21

1-71

10

8-3

0-21

1-92

11

9-2

0-21

2-13

12

10-0

0-25

2-38

13

10-8

0-29

2-67

14

11-7

0-41

3-08

15

12-5

0-92

4-00

16

13-3

1-17

5-17

17

14-2

0-83

6-00

18

15-0

0-17

6-17

K. A. G. SHIELLS: A NEW PRODUCTID BRACHIOPOD

445

Table 4

Bivariate statistics of Selected pairs of Characters

Coefficient of
relative dispersion

Coefficient of

about reduced

Level of

Characters

N

correlation

major axis

significance

Hinge width X curved length

19

0-681

11-58

0-01

Hinge width X straight length

19

0-536

10-83

0-05

Sulcation : width X depth

18

0-467

25-01

not significant

Thickness of visceral cavity X sulcation
Umbonal angle X extension of umbo

10

-0-141

—

not significant

posterior to hinge

Extension of umbo posterior to hinge

32

0-351

11-23

0-05

X thickness of visceral cavity

23

0-500

26-01

0-05

Umbonal angle X hinge width

29

-0-285

—

not significant

Hinge width X ear width

27

0-729

11-58

0-01

Ear width X ear length

27

0-428

17-97

0-05

Ear width x ear height

27

0-095

29-67

not significant

Ear length x ear height

27

0-181

51-46

not significant

TABLE 5

Measurements showing the changing proportions of the Ventral Sulcus

Distance from

Width of

Depth of

Position on valve surface

umbo

sulcus

sulcus

Posterior of visceral disc

0-0-85

nonsulcate

nonsulcate

Mid-region of visceral disc

0-85-1-05

3-0

0-25

1-05-1-25

4-0

0-50

1-25-1-75

5-0

0-75

Anterior of visceral disc

1-75-2-25

5-0

1-00

Trail

2-25-2-60

5-0

0-75

2-60-3-35

5-0

0-50

Table 6

Variation in Costation of the Pedicle Valve

A — REGULAR COSTATION

Width

Nature of

(costae approximately
equal in width)

Pine < 1 -2 mm.

Coarse > 1-2 mm.

cross-section

Group 1

Group 4

Rounded

Group 6

Scalloped

B — IRREGULAR COSTATION

Range of mean widths

Nature of
cross-section

No. of Pine > Coarse
Mean width < 1 -0 mm.

No. of Coarse > Pine
Mean width > 1 -0 mm.

Group 2

Group 3

Rounded

Group 5

Rounded &
scalloped

446

PALAEONTOLOGY, VOLUME 9

Table 7

Measurements showing the progressive change in size of successive Rugae on the

Pedicle Valve of a single specimen

Trace of viscera!

Max. distance

Max. posterior

Height of ruga

No.

Total

disc in plane

from ruga im-

to anterior width

at position of

of ruga

length of ruga

of ruga

mediately posterior

of ruga

maximum width

1

2x4

11

—

—

—

2

12

12

0-8

0-4

005

3

14

14

0-8

0-5

005

4

16

16

0-7

0-4

0075

5

18

18

0-9

0-3

01

6

22

22

1-2

0-6

01

7

25

25

1-5

0-8

0-2

8

30

30

20

0-8

0-3

9

32

32

20

10

0 1

10

2x 10

33

—

—

—

Table 8

Measurements showing the progressive Elongation of the Reticulation on the Pedicle

Valve of a single specimen

Ratio of intercept on ruga

Distance from umbo

Intercept on ruga

Intercept on costa

to intercept on costa

5

0-4

0-8

1 : 2

9

0-45

1-6

1 : 3-5

12

0-5

2-0

1 :4

Table 9

Distribution of Spines on the Pedicle Valve of the holotype of Promarginifera

trearnensis gen. et sp. nov.

Distance
of largest

Spine spine from Spine bases numbered from anterior to posterior

umbo Distance between adjacent spines in each row — d.

1

Ear row

9-5

0-7

Flank row

14

0-6

Other

1

24

0-4

rows

2

22

0-3

3

24

0-4

4

25

0-4

d

2

d

3

d

4

20

0-4

1-4

0-35

10

0-2

0-9

0-55

1-2

0-4

0-8

0-3

4-5

0-3

4-5

0-25

2-3

0-2

2-5

0-3

4-2

0-2

20

0-2

2-2

0-2

1-9

0-3

1-5

0-25

2-3

0-35

2-3

0-2

1-8

0-2

d 5 d 6 d 1

10 01 — —

0-8 0-3 M 0-2 0-7 015

40

01

K. A. G. SHIELLS; A NEW PRODUCTID BRACHIOPOD

447

Acknowledgements. I thank Professor T. Neville George, Dr. J. D. Lawson, and Dr. W. D. I. Rolfe for
reading the manuscript; to them and to Dr. J. K. Ingham I am indebted for many helpful discussions
and advice. The work was carried out in the Department of Geology in the University of Glasgow and
I gratefully acknowledge the assistance of the technical staff during various stages of the work.

REFERENCES

CURRIE, E. D. 1954. Scottish Carboniferous Goniatites. Trans, roy. Soc. Edin. 62, 11, 527-602.
MUiR-wooD, H. and coorer, g. a. 1960. Morphology, Classification and Life Habits of the Productoidea
(Brachiopoda). Mem. geol. Soc. Am. 81, 1-447.

PRENTICE, J. E. 1956. Gigantoprodiictiis edelburgensis (Phillips) and related species. Proc. Yorks,
geol. Soc. 30, 229-58.

RICHEY, J. E., ANDERSON, E. M., and MACGREGOR, A. G. 1930. The Geology of North Ayrshire. Mem.
geol. Snrv. Scotl.

SHIELLS, K. A. G. 1965. The growth of a productid shell and its implication on a method of statistical
correlation. Nature, 205, 4974, 878-80.

DR. K. A. G. SHIELLS

Department of Geology
The University
Glasgow, W. 2

Manuscript received 7 December 1964

DESCRIPTION OF DIMORPHISM IN
STRIATOPORA FLEXUOSA HALL

by WILLIAM A. OLIVER, JR.

Abstract. Striatopora Hall comprises branching favositoid tabulate corals with thick, dilated walls near the
outer surface of the colonies. Study of serial sections of the type species, S.flexuosa Hall, shows that individuals
within colonies commonly originated in one of two positions, either near the axis of the colony or near the
boundary between inner thin-walled and outer thick-walled zones. Corallites originating in the two positions
are morphologically distinct and were produced alternately by corallites of the first type. Corallites of the second
type did not reproduce asexually. Previous descriptions of S. flexuosa have been based on exteriors of colonies
or on material other than the type specimens. Redescription of the type specimens based on thin section studies
provides a better basis for understanding the genus.

Other named genera of morphologically comparable corals can be differentiated on the basis of wall micro-
structure or other features, or are possible subjective synonyms of Striatopora.

The tabulate coral genus Striatopora Hall has been widely accepted since its first des-
cription in 1851, and specimens ranging in age from Silurian to Permian, from most con-
tinents, have been referred to it. Most workers have loosely conceived of the genus as
a branching favositoid with a certain amount of wall dilation in individual corallites,
especially near the surface of the colony. This concept was based on Hall’s description
and illustrations of the exterior of the Silurian type species, S. flexuosa. In recent years
the accepted concept has been narrowed, good redescriptions of the type species have
been published, and additional genera have been erected to include forms once referred
to Striatopora sensu lato. Although the type species is now comparatively well known,
much of the available descriptive information is based on specimens other than the syn-
types, which have not previously been sectioned. The purpose of this paper is to describe
the developmental pattern within colonies, to redescribe the type specimens so that there
will be a satisfactory basis for understanding the genus, and to outline the problems
connected with this and other morphologically similar genera.

Acknowledgements. The redescription was undertaken at the suggestion of Professor Dorothy Hill,
Brisbane, Australia, who pointed out the need for a fuller understanding of Striatopora flexuosa. I am
indebted to Professor Hill for her comments on an early draft of the description and generic discussion,
many of which have been incorporated in the text. The manuscript also profited from review by Dr.
Richard S. Boardman, U.S. National Museum, who suggested that the included analysis of colony
growth be undertaken. The syntype collection is housed in the American Museum of Natural History
(AMNH) and was loaned to me by Dr. Roger L. Batten. Topotype collections in the U.S. National
Museum (USNM) were made available by Dr. Boardman. Photographs are by Robert H. McKinney
(exteriors) and David H. Massie (thin sections). Publication is authorized by the Director, U.S.
Geological Survey.

PATTERN OE GROWTH

After initial study of Hall’s original specimens (see below), two topotype fragments of
S.flexuosa were prepared by making two series of 8 and 10 thin sections, transverse to
the branch axis at 1 mm. intervals. In these and other specimens, offsets (new corallites;

[Palaeontology, Vol. 9, Part 3, 1966, pp. 448-54, pis. 68-71.]

W. A. OLIVER, JR.: DIMORPHISM IN STRIATOPORA FLEXUOSA HALL 449

‘buds’) were formed only in two positions, at the branch axis and near the boundary
between the inner (thin-walled) and outer (thick-walled) zones.

Offsets originating at the branch axis (a-oflFsets in the following discussion) appear on
the axial side of the thin-walled part of the protocorallite (a on PI. 69, fig. 1, and PI. 70,
fig. 3). A polygonal bulge in the parental wall precedes the development of a new wall
separating parent and a-olTset (several examples can be seen in PI. 71, figs. 1-6). Initially,
a-offsets have thin walls with minimal light-coloured lamellar deposits.

Offsets appearing at the boundary between the inner and outer zones (b-oflFsets) also
appear on the axial side of the protocorallite but seemingly form on the thick lamellar
part of the parental wall (b on PL 69, fig. 1, and PI. 70, fig. 3).

a- and b-corallites are morphologically distinct. Measured to the nearest millimetre, 14
a-corallites range in length from 5 to more than 9 mm., 6 mm. being the most common
observed length ; 1 5 b-corallites range from 2 to 4 mm. in length with 10 individuals being
3 mm. long. In addition, wall dilation is limited to the distal one-third to one-half of each
a-corallite, whereas b-corallite walls are dilated along their entire length.

Within the studied material the pattern of colony development seems to be consis-
tent. Each a-corallite gives rise to two offsets (‘buds’). An a-offset is formed as the parent
separates from the axis; a b-oflfset is formed as the parent becomes thick-walled. Both
types are labelled in longitudinal sections illustrated on Plates 69 and 70. b-corallites do
not produce offsets. The result of obtaining two offsets from each a-corallite and none
from b-corallites is the a-b alternation which is so clearly shown in the illustrations. The
lack of alternation at the top of the figured longitudinal sections is only apparent. Presum-
ably each of the adjacent a-corallites produced b-corallites above the limits of the thin
section.

If the described pattern of development were invariable, a- and b-offsets should
appear in equal numbers. Counts were made in the two serial series with these results: in
specimen USNM 146505 (PI. 71, figs. 1-6), I identified 10 a-offsets and 10 b-offsets in the
5-mm. length illustrated ; two other offsets are of uncertain origin. In the second specimen
I found 8 a-, and 10 b-offsets in 5 mm. Although based on a small sample, these data do
support the suggested alternation of a- and b-types. There are exceptions, however. On
Plate 71, corallite 7 gives rise to a-oflfset 23 in fig. 4; 2 mm. higher in the branch (fig. 6),
the same corallite gives rise to two more offsets which are completely separated from
corallite 7 in a thin section taken 1 mm. higher. The upper offsets are probably b-types,
although this is not certain from the section ; neither is included in the previous offset
count because no new wall was formed within the 5-mm. portion on which the count was
based. Perhaps the ‘extra’ corallite was needed to fill space in the colony.

The dimorphism represented by a- and b-corallites was probably significant in the
development of the living colony, a-polyps went through a longer development period.
This may have been required for the attainment of full maturity, and a-polyps may have
reproduced sexually as well as asexually. b-polyps apparently skipped early growth
stages. They may have attained normal maturity for some functions, such as feeding or
defence, but not for others, such as reproduction. Alternate production of a- and b-
polyps permitted faster colony growth, which may have given the species certain advan-
tages in food and/or oxygen intake or in some other way. Presumably both types of
polyps competed for space with surrounding individuals but in somewhat different
micro-environments.

450

PALAEONTOLOGY, VOLUME 9

The formation of a- and b-offsets is in some ways similar to lateral and peripheral off-
setting, respectively, in rugose corals.

SYSTEMATIC DESCRIPTION
Genus striatopora Hall

Striatopora Hall 1851, p. 400; 1852, p. 156; Wells 1944, pp. 259-60 [part]; Hill and Stumm
1956, pp. 464 [part]; Lafuste 1959, pp. 85-87.

Type species. By monotypy, S.flexiiosa Hall 1852, p. 156, pi. 40b, figs. 1 a-e. Middle Silurian, Rochester
Shale, Lockport, New York.

Diagnosis. Ramose favositoid coralla with cylindrical or slightly compressed branches.
Corallites gently curve away from axial region, opening obliquely to surface on small
branches, perpendicularly on large branches. Corallite walls thin axially, strongly di-
lated distally, distinctly lamellar. Corallites polygonal in cross-section, but lumen round
because of dilation. Mural pores common. Septal spines project into lumen, expressed as
septal ridges in calice. Tabulae complete.

Discussion. Several genera of ramose cerioid tabulates with greater or lesser dilation of
distal corallite walls have been described, and it is not at all clear how they interrelate.
Wells (1944) discussed the morphologic series Favosites^Thamnopora^Striatopora—^
Tracin’pora as one of increasing wall dilation with complete gradation in this character,
but subsequent work indicates that the phylogenetic relationships are more complex than
this would suggest. The genera were originally described without adequate analysis of
similarities and differences and without knowledge of the morphology of even the type
species of earlier genera. Some of these genera are briefly discussed below with em-
phasis on apparent differences from Striatopora.

Thanmopora Steininger (1833; see Lecompte 1939, pp. 102-4) has cylindrical or com-
pressed branches; corallite walls are moderately dilated, the dilation increasing distally;
septal spines are weak or lacking. Lecompte (1936, pp. 14-16), working with the Middle
Devonian type material, and Lafuste (1958) have described the wall structure as fibrous,
the fibres arranged more or less normal to the wall surface (radial-fibrous). Similar wall

EXPLANATION OF PLATE 68

Figs. 1-7, Striatopora flexuosa Hall. 1-4, Lectotype corallum, AMNH 1685:1, the original of Hall
1852, pi. 40b, fig. \a. 1, Exterior of complete specimen ( x 1 ), thin sections were taken from the distal
end of the right-hand branch. 2, Enlargement of central portion of corallum ( x 2). 3-4, Longitudinal
and transverse thin sections ( X 20), note the relatively thin walls in this ‘ young ’ portion of the colony.
5, Paralectotype, AMNH 1685:2; exterior (x2); the original of Hall 1852, pi. 40b, fig. \b. 6, Para-
lectotype, AMNH 1685:6; exterior (x 2); not illustrated by Hall; see also pi. 70, fig. 6. 7, Paralecto-
type, AMNH 1685:3; exterior (x2); note that septal ridges are especially well preserved on this
specimen; original of Hall 1852, pi. 40b, fig. Ic.

EXPLANATION OF PLATE 69

Figs. 1-5, Striatopora flexuosa Hall. Paralectotype, AMNH 1685:4; not illustrated by Hall; see also
PI. 70. 1-2, Longitudinal thin section (x 10) and detail of same (x 50), note mural pores, septal spines
and lamellar wall-structure; a- and b- corallites are so labelled. 3-4, Transverse thin section ( x 10)
and detail of same ( X 50), note septal spines and lamellar wall-structure. 5, Another transverse thin
section (x 10), showing mural pores and excessive wall dilation.

Palaeontology , Vol. 9

PLATE 68

OLIVER, Striatopura flexuusa Hall

Palaeontology, Vol. 9

PLATE 69

OLIVER, Striatopora flexuosa Hall

W. A. OLIVER, JR.: DIMORPHISM IN STRIATOPORA FLEXUOSA HALL 451

Structure is known in many Devonian species Favosites" (Swann 1947, Lafuste 1962)
and Tbamnopora is considered by many workers to be closely allied to that genus.
Lafuste, however, has restudied the wall structure of Favosites gotlaudica Lamarck, the
Silurian type species, and describes it as lamellar (concentric) and ‘apparently identical
to that which has been observed in Striatopora flexiiosa Hall . . . ’ (Lafuste 1962, p. 106).
Lafuste suggests that Favosites s.s. should include only forms with lamellar wall struc-
ture.

The type species of Trachypora Milne-Edwards and Haime (1851, pp. 158, 305, T.
davidsoni, pp. 305-6, pi. 17, figs. 7, la. Upper Devonian) has widely spaced calice pits
separated by a great deal of stereoplasm (possibly a symbiotic stromatoporoid, accord-
ing to Lecompte, 1939, pp. 147-8), the outer surface of which is covered by vermiform
marks. As far as 1 know, the type species has never been sectioned. Subsequent workers
have based the genus on the excessive amounts of stereoplasm and the general external
appearance. Hollard and Lafuste (1961, pp. 71-76) described the microstructure of
Trachypora based on the American Middle Devonian T. limbata (Eaton). According to
them, the heavily dilated walls of this species are lamellar and similar in structure to
those of Favosites gotlaudica and Striatopora.

The type species of Cladopora Hall (1851, p. 400; 1852, p. 137; C. seriata Hall, 1852,
pp. 137-8, pi. 38, fig. 1 a-m) has recently been redescribed on the basis of thin sections
(Stumm 1960; Oliver 1963). The species is of small diameter; corallites intersect the sur-
face at a low angle, walls are only slightly thickened, mural pores are rare, and tabulae
are lacking; no septa have been observed; the microstructure is lamellar.

Parastriatopora Sokolov (1949; 19556, p. 32, based on F. rhizoides, p. 32, pi. 50, figs.
2-4) is diagnosed as follows: ‘Fasciculate and finger-like coralla, formed by prismatic
corallites, radially diverging from the axis of the corallum and opening normal to its sur-
face. The stereoplasm zone is sharply demarcated at the periphery’ (Sokolov 19556, p. 32,
translated by George Rabchevsky). The species description and subsequent usage of the
genus by Russian workers emphasize the sharp separation of an inner thin-walled zone
from an outer zone of heavily dilated walls. Illustrations make clear that dilation is
strong, and that the two zones are sharply and smoothly separated. Sokolov also noted
that septal spines are weakly developed (p. 32). Microstructure of Parastriatopora
rhizoides was described and diagramed by Chudinova (1958) and Sokolov (1962).

Chudinova (1958, following in part Sokolov, 1955n)and Sokolov (1962) described three
types of microstructure within their family Thamnoporidae or Pachyporidae (Chudinova
1958, pp. 30-31, figs. 8-10; pp. 36-37, 45, 53, 67-68; Sokolov 1962, p. 206, fig. 1 a-c\
pp. 228-9):

1. ‘Fibrous microstructure’ has ‘fibres’ arranged parallel to the corallite walls (i.e.
concentrically) without layering. This characterizes the Parastriatoporinae.

2. ‘Concentric microstructure’ has short ‘ fibres ’ arranged perpendicularly to the wall
(i.e. radially) ; the fibres are thin, short, hard to discern, and in distinct concentric layers.
This characterizes the Striatoporinae.

3. ‘ Radially-fibrous microstructure’ has ‘fibres’ arranged perpendicularly to the wall
(i.e. radially); the fibres are coarse, long and tend to fuse and obliterate the boundaries
of the concentric layers. This characterizes the Thamnoporinae.

As indicated in previous paragraphs, the separation of a radial structure (3) from a

C 4179 H h

452

PALAEONTOLOGY, VOLUME 9

concentric one (1 and 2) has been observed by several workers. The recognition of two
types of concentric structure seems doubtful, as this will be very much a matter of preser-
vation. This opinion is strengthened by present observations which indicate that the
structure of typieal Striatopora is closer to the Chudinova-Sokolov type 1 (their Para-
striatoporinae) than to their type 2 (their Striatoporinae).

Chudinova separates the Parastriatoporinae from the Striatoporinae on the following
charaeters in addition to microstructure : the Parastriatoporinae have a sharp break be-
tween an inner zone of little wall dilation and the outer dilated zone, they have more
widely distributed mural pores, and their corallites open at right angles to the outer sur-
face of the branch ; the Striatoporinae show gradual increase in wall dilation with growth,
fewer rows of mural pores, and corallites opening at an angle to the surface. These
features are a questionable basis for even generic separation and seem quite inadequate
for subfamily definition.

If wall structure can be shown to be primary, the following grouping of the discussed
genera would be suggested by the available data: 1. radial-fibrous group, Thanmopora
and many ' Favosites’ ; 2. concentric group, Favosites s.s., Striatopora (and Parastriato-
pora), Cladopora, and Trachypora.

The problem of phylogenetic relationships is beyond the scope of the present paper
which primarily is to redescribe the species and type specimens on whieh the genus
Striatopora is based. It is clear that much remains to be done with all genera. The diffi-
culties that result from long-distance interpretation of genera and type species are not to
be easily overcome.

Striatopora flexuosa Hall
Plates 68-71

Striatopora flexuosa Hall 1851, p. 400 (nomen nudum); 1852, p. 156, pi. 40b, figs. 1 a-e\ Wells
1944, p. 260, pi. 40, figs. 1-2; Hill and Stumm 1956, figs. 352, 5a-b\ Lafuste 1959, pp. 85-87,
text-figs. 1, 2, pi. 1, figs. 1-3.

External features. Coralla are ramose with cylindrical or compressed branches varying
in diameter from 3 to 5 mm. The lectotype corallum bifurcates at intervals of from 1 1 to
22 mm. ; bifurcation is common in other specimens, but none is complete from one

EXPLANATION OF PLATE 70

Figs. 1-6, Striatopora flexuosa Hall. 1-2, Paralectotype, AMNH 1685:4, not illustrated by Hall; see
also PI. 69; thin section tangential to colony branch showing transverse sections of corallites in
mature region (xlO); detail (x50). 3-4, Paralectotype, AMNH 1685:5, not illustrated by Hall;
longitudinal and transverse thin sections (x 10). 5, Topotype, USNM 146506; detail of longitudinal
thin section (x 50) showing tunnel-like mural pores. 6, Paralectotype, AMNH 1685:6; longitudinal
thin section (x 10) of specimen illustrated on PI. 68, fig. 6 (section was taken from the portion of the
branch below the crack).

EXPLANATION OF PLATE 71

Figs. 1-6, Striatopora flexuosa Hall. Topotype, USNM 146505; serial transverse thin sections (x 15)
taken at right angles to the branch axis at 1 mm. intervals; figure 1 is the lowest section, 6 is the
highest. Corallites are numbered for recognition in successive sections but numbers are omitted
where their inclusion would cover important detail. Corallite origin is as follows: a-corallites, 13, 14,
15, 16, 17, 23, 24, 27, 28, 29; b-corallites, 8, 9, 10, 12, 18, 20, 21, 22, 25, 26; corallites of uncertain
origin, ll(b?), 19(a?).

Palaeontology, Vol. 9

PLATE 70

OLIVER, Striatopora flexuosa Hall

Palaeontology, Val. 9

PLATE 71

iPtSME

OLIVER, Striatopora flexuosa Hall

W. A. OLIVER, JR.: DIMORPHISM IN STRIATOPORA FLEXUOSA HALL 453

bifurcation to the next. Above bifurcations, branch diameters are commonly sub-equal,
although unequal in some cases. The largest known specimen is the lectotype (PI. 68,
fig. 1) with a spread of 85 mm. between the ends of the two main branches.

Corallite boundaries are clearly defined. Calices are broad and shallow, normally
deepest near the proximal margin. Calice walls slope gently from margin to pit; in
slender corallum fragments, walls tend to be concave, rising to sharp ridges which
separate corallites ; in more robust fragments of coralla, outer parts of calice walls are
gently convex so that the marginal ridges are subdued.

Where well preserved, calices are marked by low, broad septal ridges radiating from
the axial pit ; these are best developed on more robust fragments of coralla. In the few
calices where septa can be counted with any degree of certainty, they are 12 in number.

The number of corallites was counted on the outer surface of five branch fragments
with counts ranging from 2-9 to 4-5 corallites per mm. of branch length. This agrees with
counts of 4-4 and 3-6 new buds per mm. in two serially sectioned branches. The variation
between specimens indicates differences in rates of budding relative to rates of skeletal
accretion and probably represents varying environmental conditions.

Internal features. Corallites originating near the axis of the corallum (a-corallites) extend
upward and outward, initially at very small angles to the axis ; distally, these corallites
turn away from the axis and intersect the surfaee at angles of 60 to 80 degrees,
b-corallites originate in the zone where the outward bending of earlier corallites takes
place; these expand more rapidly and are initially at a high angle to the branch axis.
Corallite walls consist of a very thin dark granular layer (epitheca of some workers)
which separates corallites, and a light lamellar layer which is very thin near coralla axes,
but which thickens with growth and is very thick peripherally. Mural pores are common
at all stages of corallite growth; in thickened walls they appear as tunnels with lengths
several times their diameters. In places, a low ridge surrounds the pore on one side of the
wall (e.g. PI. 70, fig. 5).

Tabulae are thin, complete, and irregularly spaeed, composed of a dark, basal tissue
overlain by varying thieknesses of light-coloured stereoplasm in lamellae continuous
with those of the wall. They are commonly not at right angles to the axis or walls of the
corallite but are inclined away from the branch axis.

Septa are irregular in their appearance and occurrence. They are rare in thin-walled
parts of corallites, common in thick- walled parts. They appear to be low, broad ridges
or spines projecting into the lumen, formed by the inward bending of the wall laminae,
whieh are continuous around and through the septa.

Mierostructure of the dark wall layer is obscure. The light tissue is distinctly lamellar
parallel to the inner surface of the corallite wall.

Remarks. As the description indicates, two more or less distinct zones can be recognized
in seetions through or at right angles to the corallum axes. An inner zone is characterized
by corallites having small total diameters, thin walls, occurring at low angles to branch
axes, and being polygonal in internal cross-section. The outer zone has corallites having
large total diameters, and thick walls ; these corallites occur at high angles to branch axes
and are circular in internal cross-section. Transverse sections of mature parts of corallites
are seen in sections tangential to the branches (PI. 70, figs. 1, 2).

Slender corallum fragments (d'ameters 3-4 mm.) apparently represent portions of

454

PALAEONTOLOGY, VOLUME 9

colonies with minimal development of outer zone characters. The portion of the lecto-
type that was removed for thin sectioning was from a slender part of the right-hand
branch (PI. 68, fig. 1) near a growing tip and shows the characters of corallites in an
early stage of ‘maturity’ (PI. 68, figs. 3, 4). Although not sectioned, the oldest part of the
holotype specimen is more robust, with larger corallites and thicker walls. The paraleeto-
type illustrated on Plate 69 and Plate 70, figs. 1 and 2, is robust and shows full develop-
ment of ‘mature’ characters. Note that the lectotype thin sections are illustrated at x20
whereas the others are at x 10. Presumably corallite growth was continuous, and branch
robustness and ‘maturity’ are a function of corallite age and environmental conditions.

Material. Lectotype, AMNH 1685:1; paralectotypes illustrated by Hall, 1685:2, :3; other paralecto-
types, 1685:4-:8; 12 topotype fragments, USNM 1 14214; illustrated topotypes from the same collection
have been renumbered USNM 146505 and 146506.

REFERENCES

CHUDINOVA, 1. 1. 1958. [Devonian thamnoporids of Southern Siberia]: Trudy Palaeout. iiist. 73, 146 pp.,
34 pi.

HALL, J. 1851. New genera of fossil corals from the report of James Hall, on the Paleontology of New
York: Am. J. Sci. [2], 11, 398-401.

• 1 852. Containing descriptions of the organic remains of the lower middle division of the New York

system. Paleont. New York, 2. 362 pp., 84 pis.

HILL, D. and STUMM, E. c. 1956. Tabulata, in Moore, R. C. Treatise on Invertebrate Paleontology, Pt. F.
Coelenterata. Kansas.

HOLLARD, H. and LAFUSTE, J. 1961. Tracltypora liinbata (Eaton) 1832 (Coelenterata, Tabulata) dans le
Devonien du sud Marocain. Notes Serv. geol. Maroc. 20, 71-79.

LAFUSTE, J. 1958. Sur la microstructure des parois calicinales chez Thanmopora Steininger. C. r. Acad.
Sci. Paris, 246, 3658-60.

1959. Sur la microstructure du genre Striatopora Hall, 1851. Annls Hebert Hang, 9, 85-88.

1962. Note preliminaire sur la microstructure de la muraille chez Favosites Lamarck (Coelenterata,

Tabulata). C. r. Seanc. Soc. geol. Fr. 1962, 105-6.

LECOMPTE, M. 1936. Revision des Tabules devoniens decrits par Goldfuss. Mem. Mns. r. Hist. nat.
Belg. 55, 112 pp., 14 pi.

1939. Les Tabules du Devonien moyen et superieur du bord sud du bassin de Dinant. Mem. Mus.

r. Hist. nat. Belg. 90, 111 pp., 23 pi.

MiLNE-EDWARDS, H. and FFAiME, J. 1851. Monographic des Polypiers fossiles des terrains paleozoiques.
Arclis. Mns. Hist, nat., Paris, 5, 502 pp., 20 pi.

OLIVER, w. A., JR. 1963. Redescription of three species of corals from the Lockport Dolomite in New
York. Prof. Pap. U.S. geol. Snrv. 414-G, 9 pp., 5 pi.

SOKOLOV, B. s. 1949. Tabulata and Heliolitida of the Silurian of the USSR, in Atlas of the guide forms
of the fossil faunas of the USSR, vol. II. [Not seen.]

1955n. [Tabulates of the Paleozoic of the European part of the USSR, Introduction.] Trudy

VNIGRI, 85, 527 pp., 90 pi.

19556. [Subclass Tabulata in Field atlas of Ordovician and Silurian faunas of the Siberian

Platform, ed. by O. I. Nikiforova]: VSEGEI, Moscow, pp. 25-34, illustr.

1962. Subclass Tabulata, in Orlov, Yu. A., Osnovi paleontologii, 2.

STUMM, E. c. 1960. The type species of the Paleozoic tabulate coral genera Cladopora and Coenites.
Contr. Mns. Paleont. Univ. Mich. 15, 133-8, 1 pi.

SWANN, D. H. 1947. The Eavosites alpenensis lineage in the Middle Devonian Traverse Group of Michi-
gan. Ibid. 6, 235-318, 17 pi.

WELLS, J. w. 1944. New tabulate corals from the Pennsylvanian of Texas. J. Paleont. 18, 259-62, pi.
40-41.

Manuscript received 14 July 1965

WILLIAM A. OLIVER, JR.

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

A PHYLLOCERATID AMMONITE FROM THE
SPEETON CLAY (LOWER CRETACEOUS)
OF YORKSHIRE

by P. F. RAWSON

Abstract. Hypophylloceras cf. perlobatiim (Sayn) is described and illustrated. It is the first phylloceratid
ammonite to be recorded from the Neocomian of Britain.

In Europe during the Lower Cretaceous period, the phylloceratid ammonites were
essentially restricted to the Tethyan faunal province, and few specimens have been
recorded from more northerly areas. In Britain, Spath (1923, pp. 15-20) described a few
specimens of Phylloceras subalpinwn (d’Orb.) and Hypophylloceras seresi tense (Per-
vinquiere) from the Gault but no phylloceratids have been recorded from the Lower
Greensand (Casey 1960, p. xxxv) or from earlier Cretaceous beds. Outside Britain few
phylloceratids have been described from the Neocomian Boreal province. Von Koenen
(1902, p. 39) recorded one specimen of Phyllopachyceras aff. winkleri (Kilian) from the
Lower Hauterivian Radiatus Zone of North Germany, and Donovan (1953, p. 100) de-
scribed a single specimen of Phylloeeras sp. from the Middle Valanginian of Traill
Island, East Greenland.

Recent collecting at the type locality of the Speeton Clay in Filey Bay, Yorkshire, has
yielded a specimen of Hypophylloceras cf. perlobatwn (Sayn) from bed C8 of the Lower
Hauterivian. Its occurrence in this bed is of particular interest as several other genera of
Tethyan affinity found at Speeton are first recorded at this horizon, including Lytoceras
(C. W. Wright collection), Eodesmoeeras, Spificliseiis, and rare crioceratids. The crio-
ceratids become much more common in the overlying bed Cl.

The specimen is deposited in the author’s collection in the Department of Geology,
the University of Hull, catalogue number HU.C/Rn.460.

SYSTEMATIC DESCRIPTION

Family phylloceratid ae Zittel 1884
Subfamily phylloceratinae Zittel 1884
Genus hypophylloceras Salfeld 1924
Hypophylloceras cf. perlobatiim (Sayn)

Plate 72, figs. 1-3; text-fig. \a

Cf. 1901 Phylloceras serum Oppel var. perlobata Sayn (p. 7, fig. 3, pi. 1, figs. 6-8).

Description. The specimen is a wholly septate pyritized nucleus, with fragments of the
original shell preserved. The test is compressed and involute, with a steep umbilical wall.
The whorl flanks are flat, even near the umbilical edge, and the venter rounded. The shell
is covered with fine, dense, and slightly sinuous striae, which cannot be seen on the in-
ternal mould. The suture (fig. \a) is complex; the first lateral lobe is asymmetrical and

[Palaeontology, Vol. 9, Part 3, 1966, pp. 455-7, pi. 72.]

456 PALAEONTOLOGY, VOLUME 9

much larger than the ventral and second lateral lobes, and the saddle endings are
tetraphyllic.

Dimensions: Maximum diameter: 31-5 mm.; at diameter of 28-2 mm., whorl height 16-9 mm. (59-9%),
thickness 9-5 mm. (33-7%), width of umbilicus 2 0 mm. (71%).

Remarks and A ffinities. The classification of the Phylloceratid ammonities at generic and
subgeneric level fluctuates considerably from author to author, but most recent authori-
ties (for example Arkell 1957, and Wiedmann 1963) place the compressed, finely ribbed

TEXT-FIG. 1. a, Suture-line of Hypopliylloceras ci. perlobatuin, HU.C/Rn.460.
b. Suture-line of Hypophylloceras perlobatiiin, reproduced from Sayn (1901,
p. 8, fig. 3). Both figures X 6.

species with tetraphyllic saddle endings and asymmetrical lobes in Hypophylloceras. It is
this more complex suture-line which distinguishes H. cf. perlobalian from Phylloceras
serum and allied species which resemble H. cf. perlobatum in general form and sculpture.

Few species of Hypophylloceras have been described from the Valanginian and
Hauterivian of the Tethyan province, so that it is very difficult to make a close compari-
son with other forms.

EXPLANATION OF PLATE 72 (contliiucd on p. 460)

Figs. 1-3. Hypophylloceras cf. perlobatum (Sayn). Lateral and ventral views. HU.C/Rn.460, x2.

Palaeontology, Vol. 9

PLATE 72

RAWSON, Hypophylloceras from the Speeton Clay
JENKINS, Cheiloceras from New South Wales

P. F. RAWSON; A PH YLLOCERATID AMMONITE

457

The Valaiiginian specimens of H. perlobatwn figured by Sayn are closely similar to the
Speeton form. Sayn’s drawing of the suture is reproduced in text-fig. lb. It can be seen
that the terminal folioles of the second lateral saddle are only slightly subdivided, while
those of the present specimen are more strongly tetraphyllic. The minor differences
between the two suture-lines could be accounted for by specific variation.

Phylloceras spatlii Collignon (1949, p. 63) from the Hauterivian of Madagascar, com-
pares closely in the degree of flattening of the whorl flanks, but the suture line is in-
sufficiently known for closer comparison to be made.

From the Barremian of Algeria, Busnardo (in Busnardo and David 1957, p. 86) has
described HypophyUoceras barremense, which is slightly more inflated than H. cf.
perlobatwn. The sutures of the two forms are very similar, but the terminal folioles of
H. barramense are more elongate. H. betieri (Busnardo) (op. cit., p. 84) is a form with
marked flattening of the flanks, but the terminal folioles are even more elongate: this is
a feature typical also of the more numerous Aptian, Albian, and Upper Cretaceous
species of HypophyUoceras, such as H. seresitense (Pervinquiere), with which H. cf. per-
lobatwn could otherwise be compared.

Straiigrapliical horizon: Lower Hauterivian, Radiatus Zone, Sulcosa Subzone. Bed C8, 2 ft. below the
top and approximately 6 ft. above the base, in Middle Cliff, Speeton.

Acknowledgements. This paper forms part of a larger study of British Valanginian and Hauterivian
ammonites being carried out during the tenure of an S.R.C. Research Studentship. I thank Professor
D. T. Donovan and Mr. C. W. Wright for helpful discussion and advice. Dr. J. W. Neale and
Mr. C. W. Wright for reading the manuscript, and Mr. B. Nettleton for assistance with photography.

REFERENCES

ARKELL, w. J. 1957. Treatise on Invertebrate Paleontology, Part L, R. C. Moore (ed.), pp. Ll-490.
New York.

BUSNARDO, R. and DAVID, L. 1957. Contribution a I’etude des faunes d’Ammonoides de Medjez Sfa
(Est Constantinois). Pnbis Serv. Carte geol. Alger. 13, 67-123, 3 pi.

CASEY, R. 1960. The Ammonoidea of the Lower Greensand. Palaeontogr. Soc. [Monogr.], Pt. I.
COLLIGNON, M. 1949. Fauiic neocomienne des couches a Crioceres de Belohasifaka (Cercle de Sitam-
piky) (Madagascar). Annis geol. Serv. Mines Madagascar, fasc. XV, 55-85, 6 pi.

DONOVAN, D. T. 1953. The Jurassic and Cretaceous stratigraphy and palaeontology of Traill O, East
Greenland. Medd. Gronland, Bd. Ill, Nr. 4. 150 pp., 25 pi.

KOENEN, A. von. 1902. Die Ammoniten der norddeutschen Neocomien. Abh. preass. geol. Landesanst,
N.E. Heft 24, 451 pp.. Atlas with 51 pi.

SAYN, G. 1901. Les Ammonites pyriteuses des marnes Valanginiennes du Sud-Est de la France. Mem.

Soc. geol. Fr. Tome IX, mem. 23, 1-27, pi. 7 and 8.

SPATH, L. F. 1923. Ammonoidea of the Gault. Palaeontogr. Soc. [Monogr.], Pt. I.
wiEDMANN, J. 1963. Unterkreide-Ammoniten von Mallorca. 2. Lieferung; Phylloceratina. Abh. math.-
natnrw. Kl. Akad. Wiss. Mainz, Jahrgang 1963, Nr. 4, 151-264, pi. 11-21.

P. F. RAWSON

Department of Geology,
The University,

Hull,

Manuscript received 16 August 1965 Yorkshire

THE UPPER DEVONIAN INDEX AMMONOID
CHEILOCERAS FROM NEW SOUTH WALES

by T. B. H. JENKINS

Abstract. CheUoceras aciitum (Munster) from a presumed equivalent of the Baldwin Formation (Upper
Devonian) near Tamworth, N.S.W. is described as the first occurrence of this index genus in eastern Australia.

Ammonoid cephalopods of Devonian age are very rare in eastern Australia. Teichert’s
(1948) description of goniatites from the Buchan district of Victoria was the first authen-
tic report of Devonian ammonoids from eastern Australia, previous records being
probably erroneous (Teichert 1943). Subsequently, Pickett (1960) has described a new
species of clymenid and a doubtful ?Platyclymenia from different horizons in the Upper
Devonian of northern New South Wales. Taxonomic work by Erben (1960) has in-
directly contributed to a revised view of the correlation of Teichert’s Buchan goniatites
(Philip and Pedder 1964), but there have been no further records of Devonian
ammonoid discoveries in eastern Australia. The present note is thus only the third
such published record. A fourth occurrence is mentioned herein and will shortly be
documented.

The goniatites here recorded were found near Keepit Dam (see text-fig. 1), a recently
completed structure on the Manilla (or Namoi) River between the towns of Manilla,
Tamworth, and Gunnedah. A reasonably complete statement of the regional geology of
the surrounding country has been achieved by the work of several authors summarized
by Voisey and Williams (1964) and White (1964, 1965). The present writer has mapped
formations in detail through a north-south belt which includes Keepit Dam.

The dam lies on the western limb of the long submeridionally trending syncline which,
by reversals of plunge, forms the closed Werrie and Belvue Basins. Dips are consistently
40° to 50° at N. 30° to 55° E. within the area of text-fig. 1, and only minor faulting occurs.
The succession is broken by a disconformity at the base of the Carboniferous strata, with
the lowest Carboniferous formation overstepping northwards on to lower beds within
the Upper Devonian. Evidence from fossil ammonoids points to the absence hereabouts
of the upper half of the Upper Devonian.

There is also a possibility of an erosional break lower in the Upper Devonian, at the
base of what has been called the Keepit Conglomerate, but there is no local evidence for
an angular unconformity at this horizon such as has been claimed for adjacent areas
(White 1964, 1965).

Below the Keepit Conglomerate the Upper Devonian succession near the dam con-
sists chiefly of volcanic detritus. It contains a thick unit, over 2,000 ft., of alternating
shales and tuff's which yield marine fossils (including CheUoceras) at the level indicated
in text-fig. 2.

Described specimens are in the fossil collection of The University of Sydney Geo-
logical Department and are referred to by USGD catalogue numbers.

[Palaeontology, Vol. 9, Part 3, 1966, pp. 458-63, pi. 72.)

T. B. H. JENKINS; UPPER DEVONIAN INDEX AMMONOID CHEILOCERAS 459

SYSTEMATIC DESCRIPTION

Superfamily cheilocerataceae Freeh 1897
Family cheiloceratidae Freeh 1897
Genus cheiloceras Freeh 1897

Type species (by subsequent designation of Wedekind 1918, p. 144): Goniatites siibpartitus Munster
1839, p. 18 (loc. cit. fide Freeh 1902, p. 69).

Remarks. The mode of division of the dorsal lobe of the suture has been mainly used to
separate from Cheiloceras Freeh 1897 the subgenera Torleyoceras Wedekind 1918 (?=

TEXT-FIG. 1. Map of Water Cress Gully locality, showing tracks.

Staffites Wedekind 1918) and Dyscheiloceras Schmidt 1921 and the genus Paratorleyo-
ceras Bogoslovsky 1957, thus leaving Cheiloceras (Cheiloceras) to receive the species
having an undivided, flat, or gently bowed dorsal lobe of the suture.

In the southern hemisphere Cheiloceras is known only from Western Australia
(Fitzroy Basin) but is now known from all the northern continents having been recently
reported from North America (House 1962, House and Pedder 1963).

Cheiloceras (Cheiloceras) aculiim (Munster)

Plate 72, figs. 4-6; text-fig. 3 a-e

1839 Goniatites acutiis Miinster, p. 1 10, pi. 16, fig. 1 1 (fide Freeh 1902, p. 71).

\S52 Goniatites retrorsiis yar. aentnsG. andF. Sandberger, p. 108, pi. 10, fig. 10;pl. 10<7,figs. 1,2.

1902 Cheiloceras aentnni Minister emend. Sandberger; Freeh, p. 71, pi. (3)4, fig. 6.

1918 Cheiloceras acntuni Sandberger; Wedekind, p. 146, fig. 46, i.

1918 Cheiloceras acutnm Freeh; Wedekind, pi. 18, fig. 7.

460

PALAEONTOLOGY, VOLUME 9

TEXT-FIG. 2. Stratigraphic column for the area near Keepit Dam, showing ammonoid horizons.

Material. One specimen, an uncrushed internal mould of a segment of the outer whorl, lacking the
umbilicus but preserving the greater part of several sutures (external and internal) and a part of the body
chamber.

Description. Shell form thickly lenticular with whorls deeply overlapping. Whorl sides
are flatly curved and meet at an acute angle along the venter. No trace of an umbilical
shoulder is to be seen. The dorsum is more broadly rounded, the flanks of the impressed
area converging to a wider but still noticeable angulation. Whorl sides and the adjacent
flanks of the impressed area converge very gradually towards the axis of coiling; extra-
polation of the partial cross-section (text-fig. 3r/) suggests a closed umbilicus.

EXPLANATION OF PLATE 72 (sCC alSO p. 456)

Figs. 4-6. Clieiloceras (Cheiloceras) acutimi (Munster) from an equivalent of the Baldwin Formation
in Water Cress Gully, L9 miles north-west of Keepit Dam, near Tamworth, N.S.W. Ventral,
lateral, and dorsal views respectively of USGD6870, x2.

T. B. H. JENKINS; UPPER DEVONIAN INDEX AMMONOID CHEILOCERAS 461

Maximum dimension of the specimen is 46 mm., measured as an incomplete chord of
the venter’s spiral curve and ending adorally within the body chamber. Diameter of the
entire shell therefore exceeded 46 mm. by an amount depending mainly on the unknown
length of the body chamber. Maximum whorl width of the preserved segment is 15-1
mm.; the reconstruction (text-fig. ?>d) shows whorl width as 16-5 mm. at a diameter of 47
mm.

TEXT-FIG. 3. a-e, Clieiloceras (C/ieiloceras) acutum (Munster), based on specimen USGD6870. a-c,
dorsal, lateral, and ventral views; d, cross-section, reconstructed from preserved portion shown by
oblique ruling; e, suture./, g, based on specimens USGD 6871 and 6872 respectively, cross-sections of
two indeterminable external moulds. All X 1. All specimens from the same bed in an equivalent of the

Baldwin Formation.

Sutures have six lobes, of which the pair on the umbilicus is partly inferred, the umbili-
cus being missing (text-fig. 3e). The ventral lobe and flanking saddles are narrow ele-
ments, in contrast to the broad, rounded character of other preserved sutural elements.
External sutures are spirally confluent where the broad, slightly asymmetrical U-shaped
lateral lobe passes into a wide, evenly rounded dorso-lateral saddle. Internal sutures are
similarly confluent where the flatly rounded dorsal lobe passes into the adjacent saddle.

Remarks. In shell form and external suture Clieiloceras acutum (Munster) closely re-
sembles the coeval Tornoceras acutum Freeh so that complete internal moulds of the two
species are homoeomorphic. Distinctions have been recognized on form of growth-lines
and on internal sutures (e.g. Wedekind 1908, p. 585; 1918, p. 101 ). The internal suture of
Tornoceras acutum Freeh shows a deep, narrow dorsal lobe flanked by broader saddles
(Wedekind 1908, pi. 39, fig. 4; Schindewolf 1923, p. 508). The broad, flatly rounded dorsal
lobe of the present specimen is typical of Clieiloceras sensu stricto and resembles that of
the internal suture of C. circumflexitm Sandberger which was cited by Wedekind (1908,
p. 585) as similar to that of C. acutum.

Fetter (1959, p. 195) has pointed out that Tornoceras iowaense Miller 1938 may belong

462 PALAEONTOLOGY, VOLUME 9

to Cheiloceras acutum since neither its internal suture nor its growth-lines have been
recorded.

The specimen here described is the largest heretofore figured as, or authentically
assigned to, Cheiloceras acutum. Lange (1929, p. 36) records a specimen 90 mm. in
diameter but indicates doubt on generic allocation. The venter of the described specimen
is somewhat more acute than that of other figured specimens of C. acutum but since whorl
shape changes with growth to a more acute form of venter, this constitutes no objection
to the specific allocation (cf. Sandberger 1852, pi. 10a, fig. 2).

Cheiloceras acutum seems to be previously recorded only from Germany. Wedekind
noted acutum as a rare species of the lower portion of the Cheiloceras Stufe (= Ila).
Lange (1929) named ' ?Cheiloceras acutum' as a zonal fossil for the lowest of three sub-
divisions of the Cheiloceras Stufe.

Other material. Two indeterminable external moulds of coiled cephalopods were found
in the bed of tuff which yielded C. (C.) acutum. The cross-section of one of these ex-
ternal moulds (text-fig. 3a) may be compatible with an immature stage of C. (C.)
acutum. The other mould is a slightly asymmetrical oxycone (text-fig. 3/) with an um-
bilical structure somewhat similar to that figured by Clarke (1899, p. 112) for Tornoceras
imiangulare (Conrad), i.e. the umbilical lip is produced laterally giving a small spiral
prominence on the axis of coiling.

Horizon and locality. The specimens were found in Water Cress Gully, 1-9 miles north-west of Keepit
Dam, which is on the Manilla (= Namoi) River, between Tamworth and Gunnedah in northern New
South Wales. The fossil locality is in shaly volcanic tuff, a presumed equivalent of the Baldwin Forma-
tion. According to local inhabitants it is close to the spot whence Mitchell (1921, 1924) described
Merista plebeia Sowerbyand other brachiopods. The horizonissome 3,000ft. below another ammonoid
bed with Platyclynienia and Gennclyinenia, indicating Platydynienia Stufe and strongly suggesting an
expanded stratigraphic succession. The bed with Cheiloceras acutum is some 1,700 ft. above the highest
known local occurrence of the conodont Paimatoiepis.

Acknowledgements. The writer thanks Dr. B. F. Glenister, State University, Iowa, U.S.A. for helpful
correspondence and acknowledges financial assistance from the University of Sydney Research Grant.

REFERENCES

BOGOSLOVSKY, B. I. 1957. New genera of Devonian ammonoids. Material)’ k osnovam paleontologii.
Pal. Inst., Acad. Sci. USSR, 1, 45-48. [In Russian.]

CLARKE, j. M. 1899. The Naples fauna (fauna with Manticoceras intumescens) in western New York.
16th Ann. Rep. N.Y. State Geol. 29-161, pi. 1-9.

ERBEN, H. K. 1960. Primitive Ammonoidea aus dem Unterdevon Frankreichs und Deutschlands.

Neues Jb. Geol. Paldont., Abh. 110, 1-128, pi. 1-6.

FRECH, F. 1897. Lethea geognostica, TheH 1\ Lethea palaeozoica. Lief. 1, Stuttgart.

1902. Uber devonische Ammoneen. Beitr. Paldont. Geol. Ost.-Ung. 14, 27-111, pi. 2-5.

HOUSE, M. R. 1962. Observations on the ammonoid succession of the North American Devonian.
J. Paleont. 36, 247-84, pi. 43-48.

and REDDER, A. E. H. 1963. Devonian goniatites and stratigraphical correlations in Western

Canada. Palaeontology, 6, 491-539, pi. 70-77.

LANGE, w. 1929. Zur Kenntnis des Oberdevons am Enkeberg und bei Balve (Sauerland). Abh. Preufi.
geol. Landesanst., n.f. 119, 1-132, pi. 1-3.

MITCHELL, J. 1921. Some new brachiopods from the Middle Palaeozoic rocks of New South Wales.
Proc. Linn. Soc. N.S.IV. 45, 543-51, pi. 31.

■ 1924. A further reference to the occurrence of Merista plebeia Sowerby in New South Wales.

Ibid. 49, 499-503, pi. 53.

T. B. H. JENKINS: UPPER DEVONIAN INDEX AMMONOID CHEILOCERAS 463

MUNSTER, G. Graf zu. 1839. Naclitrag zii den Goiiiatiteii des Fichtelgebirges. Beitrage zur Petre-
factenkiinde. Heft 1, 16-31, pi. 3. Bayreuth.

FETTER, G. 1959. Gouiatites devoniennes du Sahara. Mem. Carte geol. Alger. Paleont. 2, 1-313, pi. 1-26.

PHILIP, G. M. and PEDDER, A. E. H. 1964. A re-assessment of the age of the Middle Devonian of south-
eastern Australia. Nature, Lond., 202, 1323-4.

PICKETT, j. w. 1960. A clymeniid from the Wocklumeria zone of New South Wales. Palaeontology, 3,
237-41, pi. 41.

SANDBERGER, G. and SANDBERGER, F. 1850-56. Systeinatisclw Beschreibiing nnd Abbildung der Ver-
steinernngen des rheinischen Schichtensystems in Nassau. 1-564, pi. 1-39 (Atlas). Wiesbaden.

SCHINDEWOLF, o. H. 1923. Beitriige zur Kenntnis des Paliiozoicums in Oberfranken, Ostthtiringen und
dem Sachsischen Vogtlande. 1, Stratigraphie und Ammoneenfauna des Oberdevons von Hof. a. S.
Neues Jb. Min. Geol. Paldont. 49, 250-357, 393-509, pi. 14-18.

SCHMIDT, H. 1921. Das Oberdevon-Culm-Gebiet von Warstein i. W. und Belecke. Jb. Preufi. geol.
Landesanst. 41, 254-339, pi. 12.

TEiCHERT, C. 1943. The Devonian of Western Australia. A preliminary review. Am. Jour. Sci. 239,
148-53.

1948. Middle Devonian goniatites from the Buchan district, Victoria. J. Paleont. 22, 60-67, pi. 16.

voiSEY, A. H. and williams, k. l. 1964. The geology of the Carroll-Keepit-Rangari area of New South
Wales. /. Proc. t oy. Soc. N.S. W. 97, 65-72.

WEDEKIND, R. 1908. Die Cephalopodenfauna des Hoheren Oberdevon am Enkeberge. Neues Jb. Min.
Geol. Paldont. 26, 565-634, pi. 39-45.

1918. Die Genera der Palaeoammonoidea (Goniatiten), (Mit Ausschluss der Mimoceratidae,

Glyphioceratidae und Prolecanitidae). Palaeontographica, 62, 85-184, pi. 14-22.

WHITE, A. H. 1964. Geological map of New England I : 100,000, Attunga sheet (No. 321), with marginal
text. The University of New England, Armidale, New South Wales, Australia.

1965. The stratigraphy and structure of the Upper Palaeozoic sediments of the Somerton-Attunga

district, N.S.W. Proc. Linn. Soc. N.S.W. 89, 203-17.

T. B. H. JENKINS

Department of Geology and Geophysics,
The University of Sydney,

Sydney, N.S.W.,

Manuscript received 3 August 1965 Australia

SCHIZOCHROAL EYES AND VISION IN SOME
PHACOPID TRILOBITES

by E. N. K. CLARKSON

Abstract. Some aspects of vision in nine species of phacopid trilobites are described. Methods of study, which
primarily consist of the investigation of the morphology of the visual surface and the extent and nature of the
visual field, are identical with those of previous work.

Divergent interpretations of intraspecific variation in the lens number of Phacops sp. are discussed and sexual
dimorphism is advanced as a tenable hypothesis. Phacopids generally have a fixed pattern of lens distribution,
though the individual lens formation and arrangement varies in different species.

Individuals within a species have visual fields of similar extent and type, regardless of dimorphism. In different
species, however, there is wide variation in visual type, largely as a result of distinct modes of curvature of the
visual surface, affecting the angular bearings of the lens-axes.

Three principal visual types are distinguished, presumably reflecting adaptations to distinct ecological con-
ditions. In the first, the visual field has a relatively wide vertical range and the eye is slightly astigmatic; in the
second there is much greater astigmatism and a narrower visual field; the third type combines features of the
other two.

Individual phacopid trilobites are sometimes found in an excellent state of preserva-
tion, showing no distortion and retaining their original convexity. Such specimens are of
great value for functional as well as for morphologieal studies, since it is often possible to
interpret exoskeletal structures or organs in terms of adaptive morphology. The method-
ology of such functional interpretation of fossil structure is fully discussed by Rudwick
(1964).

Recently the author studied the eyes and some aspects of vision in a small group of
well-preserved Silurian trilobites (Acastinae) (Clarkson 1966u). Some other phacopids,
of ages ranging from Ordovieian to Devonian, have also been studied and the results of
these investigations are presented here. The chief limitation to this work has been the
difficulty in obtaining enough perfectly preserved material ; only a few adult specimens
of each species were available. The methods of approach were the same as in my pre-
vious work beeause the material consisted of museum specimens which could not be
sectioned. These methods include detailed deseriptions of the external visual surface, the
size and distribution of the lenses, the angular extent of the visual field, and the nature of
vision as deduced from the manner in which the axes of the individual lenses are distri-
buted within the visual field.

This information gives some conception of trilobite visual processes but needs to be
supplemented by further work, and in partieular by the study of the internal structure of
the eyes of well-preserved specimens. The only significant research on this subject
hitherto was undertaken by Lindstrom (1901), who obtained little information about the
sublensar structure of schizochroal eyes.

Techniques of study are identical with those described in my former work. The same
terminology has been retained throughout, apart from the use of the term ‘genal field’
proposed by Shaw and Ormiston (1964) which is substituted for ‘eye platform’ as used
by Harrington et al. in the ‘Treatise’.

[Palaeontology, Vol. 9, Part 3, 1966, pp. 464-87, pis. 73-75.]

E. N. K. CLARKSON; EYES AND VISION IN PHACOPID TRILOBITES 465

The material examined includes the following species:

1. Superfamily Phacopacea. Phacops rana (Green), Devonian of New York State;
Phacops fecwuhis Barrande, Silurian, Bohemia; Phacops hreviceps Barrande, Devonian,
Bohemia; Phacops latifrons (Bronn), Devonian, Germany (briefly referred to); Phacops
hatracheus Whidborne, Devonian, S.W. England; Phacops boecki Hawle and Corda,
Devonian, Bohemia.

2. Superfamily Dalnianitacea. Dahnanites vulgaris (Salter) and Dahnanites caudatus
(Brunnich), Silurian, Dudley, England; Chasmops (Eichwald), Ordovician, Estonia.

The eyes of some of these species are quite well known and have been described by
systematic palaeontologists. The existing accounts of their morphology are referred to
throughout but since a clear understanding of the function of any organ depends so
greatly upon detailed knowledge of its structure, it has been necessary to supplement
these short accounts by fuller descriptions.

The eye and its position upon the cheek are of great importance in phacopid taxonomy,
and certain matters of taxonomic interest came to light in the course of this study. An
example of this is the variation in lens number between adults of an apparently single
species, which at least in the case of the genus Phacops led to divergent interpretations of
systematic relationships. But although much phacopid taxonomy is rather confused,
any attempt at systematic revision of the species in question would be beyond the scope
of the present work. Elence existing systematic categories are used, even though some of
these are known to be unsatisfactory. Such taxonomic problems as were encountered
are, however, noted and described in full.

Owing to the relative scarcity of perfect specimens it has been possible to analyse only
one or two adults of each species, and comparatively few other individuals were norm-
ally available for comparison. It has been already established that the number of
lenses per eye and their distribution may vary considerably within a species. These
factors are related to the size of the specimen and in some cases to sexual dimorphism
(Clarkson 1966a). The problem thus arises as to how far the specimens examined are
representative of the species as a whole.

This problem, as shown below, is rather complex, and in the absence of full growth
stages cannot fully be eliminated. I have endeavoured to minimize it as far as possible by
the following procedures : (a) by selecting only large sized specimens showing the most
obviously adult features, (b) by comparison with other material and previously published
descriptions, particularly in the case of the Bohemian material exhaustively described by
Barrande, and (c) by distinguishing dimorphs where the supply of specimens was ade-
quate.

It is worth noting that in spite of the variation in lens number, etc. within a species,
the angular range of vision described for single specimens appears to be fairly constant
for the species as a whole, whatever the stage of development. Increase in lens number
with age does not lead to an increase in the angular range of vision. In Acaste downingiae
(Murchison) and in Phacops fecimdus Barrande, enough specimens were available to
demonstrate that the surface curvature of the eye is established at an early stage in
post-larval development, and remains much the same though the visual surface grows
while new lenses are added. This leads to a larger number of lens axes covering the
same area.

A short discussion of the variation in the eyes of Phacops is given in the next section.

466

PALAEONTOLOGY, VOLUME 9

FUNCTIONAL DESCRIPTIONS OF EYES
Superfamily phacopacea

Variation in the eyes o/Phacops 5.5. The eyes of various species of Phacops s.s. have been
studied in somewhat greater detail by previous authors than have those of other Phaco-
pina. Among the best known of all are those of Phacops rana (Green) where the varia-
tion present has led to divergent taxonomic and developmental interpretations. This
confusion is illustrated by the following history of research upon the eyes of P. rana.

1 . A very detailed account of the morphology and development of the eyes of P. rana
was given by Clarke (1889). From observations of the number, and manner of develop-
ment of the lenses in specimens of various sizes, he concluded that lens number increased
from youth to maturity following a constant pattern of development, and then declined
again from maturity to senescence. This decline, as he stated, could have taken place
only by the sclerosis of the lenses, or by their absorption by the palpebral lobe.

2. Beckmann (1951) worked principally upon the development of the eyes of P. cf.
breviceps and P. schlotheiini from Germany, but referred also to P. rana. In the German
species he found an increase in lens number from youth to maturity but no geronto-
logical decline. He gave strong evidence that two forms of P. rana had been confused by
Clarke, one with a larger adult number of smaller lenses than the other. The author’s
studies of acastid eyes (Clarkson \966a) gave a general agreement with Beckmann’s
conclusions; two variant eye forms were distinguished and tentatively attributed to
sexual dimorphism.

3. Systematic revision of the species P. rana by Stewart (1927) and Stumm (1953) led
to the erection of several subspecies. The principal diagnostic features on which these
subspecies were distinguished included the overall structure of the eye and the number
and arrangement of the lenses, though neither author referred to Clarke’s work. Typical
eyes of all these subspecies were hgured by Stumm.

A similar variation occurs in several species of Phacops related to P. rana. Barrande’s
(1852) observations on Bohemian phacopids were not mentioned by Clarke, but in many
respects they are quite comparable. Barrande’s fullest descriptions (1852, p. 514) were of
the eyes of the common Bohemian species P. fecundus. With his anatomical studies he
included a table showing the number and distribution in files of the lenses in the eyes of
twelve specimens in various stages of development. In these the maximum number of
lenses per eye ranged from 64 to 1 36. Barrande noted that the number of lenses depended
upon the size of the specimens, but was also partially controlled by the particular con-
formation of the eyes of each individual. None of the largest specimens possessed the
maximum number of lenses, but they did exhibit a higher degree of scleral inflation and
a relatively wider degree of spacing of the lenses. He also figured the eyes of three speci-
mens of P.fecundiis major, reproduced here (text-fig. 1), which if considered as growth
stages would indicate an exactly comparable situation to that described by Clarke, i.e. an
apparent decrease of the lens number with senescence. Very large specimens of P. breviceps
and P. latifrons again have larger, fewer, and more widely spaced lenses than smaller
specimens.

The nature of eye-variation in P. rana and in all these other species is therefore a
matter of some uncertainty, and has been variously interpreted by these diflerent authors

E. N. K. CLARKSON: EYES AND VISION IN PHACOPID TRILOBITES 467

firstly as the result of a peculiar developmental pattern ; secondly as ecological or sexual
dimorphism ; and thirdly as merely natural subspeciation.

At present the state of taxonomic confusion of the genus Phacops renders it difficult to
distinguish the most probable of these alternatives. It is quite likely that more than one
factor is involved. Abandonment of Clarke’s hypothesis is, however, strongly advocated
by Beckmann’s work, and this conclusion is substantiated by the extreme variability of
P. rana as shown by its number of possible subspecies. The similarity between the differ-
ent eye- variants of P. rana and those of related species would indicate, nevertheless, that
natural subspeciation could be complicated by sexual dimorphism occurring in all these

a b c

TEXT-FIG. 1 . Phacops fecwidus major Barrande. Parts of the visual surface of the eyes of three
specimens increasing in size from a to c. The last may possibly be interpreted as a different
eye-variant from a and b, hence this series may reflect a sexually dimorphic condition. (Re-
drawn from Barrande 1852.) (a x8, ^ x8, c x6.)

related species. The possibility of such dimorphism has not always been taken account of
by systematists.

It has been necessary to defer complete studies of eye development and variation
within these species until such taxonomic revisions have been made which would provide
a sounder basis for such studies. The purpose of the present study is to show something
of the range of visual types within phacopid trilobites. Thus the course adopted here has
been to describe and analyse functionally, as individual visual organs, one or two of the
representative eye forms of each species or subspecies in question.

Phacops rana (Green 1832)

1832 Calymene bufo var. rana Green, p. 42,

1888 Phacops rana (Green); Hall and Clarke, pp. 19-26, pi. 7, figs. 1-1 1 ; pi. 8, figs. 1-8; pi. 8u,
figs. 21-33.

1889 Phacops rana (Green); Clarke, pp. 253-70, pi. 21.

1940 Phacops rana (Green); Delo, p. 22, pi. 1, figs. 1, 2.

All available material was referred to the two subspecies P. rana nhlleri Stewart and
P. rana crassituberculata Stumm. Stumm remarks that the only characters whereby the
subspecies may be distinguished are the structure of the eyes and their number of lenses.
It is therefore of particular interest to compare the functional characteristics of their
eyes.

468

PALAEONTOLOGY, VOLUME 9

Phacops rana miJIeri Stewart 1927
Plate 73, figs. 10-11 ; text-fig. 2d-f

1927 Phacops rana var. milleri Stewart, pp. 58-60, pi. 5, figs. 14-17.

1940 Phacops rana var. milleri Stewart; Delo, p. 23, pi. 1, fig. 3.

1953 Phacops rana milleri Stewart; Stumm, pp. 137-8, pi. 9, figs. 1-4, pi. 10, figs. 1-10.

A single adult specimen, SM H 6033, from the Hamilton Group, Devonian, Western
New York State, was analysed. Both eyes are preserved, the left is slightly warped, the
right is in perfect condition apart from the very bottom of the central region, which is
slightly dislodged from its true position. The external morphology of the specimen
accords with P. ram milleri but there are slightly fewer lenses than in P. rana rana
which has 104-124. Nevertheless the comparative diameters and the degree of spacing of
the lenses is identical with Stewart’s type specimens.

Eye morphology (left eye SM H 6033). Dimensions of eye: L. (max.) 7-5 mm. (min.)
5-8 mm. ; W. (max.) 4-8 mm. (min.) 4-0 mm. ; H. 4-8 mm., where cephalic length (sag.) is
14 mm. and breadth 26 mm.

Eye large and prominent, occupying much of the upper part of the librigena, about
half the length and two-thirds the total height of the cephalon but not rising quite as
high as the crown of the glabella. The anterior and posterior edges lie approximately in
the same exsagittal plane. The eye, which is in the form of a lunate segment of a cone,
stands out laterally from the librigena, its base being bounded by a deep groove. In plan
the visual surface is rather flattened anteriorly, but very sharply reflexed behind ; in pro-
file, the lower part of visual surface again is flattened but the curvature increases up-
wards, becoming very strong just below the facial suture. Palpebral lobe broad (/r),
coarsely tubercular, slightly inflated, rising gently from the narrow rim above the facial
suture, and declining to the distinct palpebral furrow. Palpebral area lying below the
level of lobe, open posteriorly and continuous with surface of librigena.

Lenses disposed in 17 files, which diverge ventrally at a moderate angle, as follows;

466 767 676 655 442* 2* 2; Max. 7; Total 84. (* denotes irregularity.)

(The lower ascending diagonal row is complete apart from the bottom lenses of rows 15
and 16. If this row were unbroken, as would probably be the case in a more normal eye,
there would be 86 lenses.) Lenses rather widely spaced between files, but within each file
the lowermost lenses are almost contiguous, the spacing increasing upwards so that the
top lenses are quite far apart. Lens diameters: Maximum 0-5 mm. (in the upper lenses of

EXPLANATION OF PLATE 73

Figs. 1-4. Phacops rana crassituberculata Stumm. SM H 7259. Silica Formation, Devonian, Lucas
Co., Ohio. Right and left eyes in lateral and dorsal aspects. X 5.

Figs. 5, 6. Phacops iatifrons (Bronn). SM H 4840. Devonian, Rhineland. Right eye in dorsal and
lateral aspects, x 5.

Figs. 7-9. Phacops fecitndiis Barrande. SM A 49374. Stage £3, Silurian, Dlauha Hora, Bohemia.
7, Central region of the left eye, showing irregular distribution of the lenses. X 25. 8, 9, Left eye
in lateral and dorsal aspects. X 5.

Figs. 10, 11. Phacops rana milleri Stewart. SM H 6033. Hamilton Group, Devonian, New York State.
Right eye in lateral and dorsal aspects, x 5.

Palaeontology, Vol. 9

PLATE 73

CLARKSON, Schizochroal eyes in phacopid trilobites

E. N. K. CLARKSON: EYES AND VISION IN PHACOPID TRILOBITES 469

the first and last few files), average 0-4 mm. (in the central files), minimum 0-2 mm. (in
the centre of the upper horizontal row). Apart from in the latter region and in the first
and last files, the lens diameters are fairly constant.

Interlensar sclera only slightly inflated, generally giving the appearance of indistinct
intersecting hexagons but flattened above the upper horizontal row, so that the upper

TEXT-FIG. 2. a-c. Phacops rana crassitiiberciilata Stumm. SM H 7259. a, b, Stereograms showing visual
fields of left and right eyes respectively, c. Projected visual surface of left eye (x 5). This eye has one
more lens (file 9) than the right eye and the bottom lens of file 8 is slightly displaced, d-f. Phacops rana
milleri Stewart. SM H 6033. i/, Stereogram of right eye. e. Contoured stereogram showing coverage of
different regions of the visual field by lenses of various sizes./, Visual surface of right eye ( X 5).

rim of the inflated part of the sclera appears undulose, with each lens of the upper row
lying in a trough. Each lens is situated in a shallow crater-like depression, above which
the crown of the lens just projects, but in the posterior few files these depressions become
shallower so that the bases of the lenses are almost flush with the surface of the sclera.
All the lenses of the upper horizontal row are set at an angle to the surface of the sclera
so that their axes have a lower latitudinal bearing than would be expected from the de-
clination of the surface of the eye.

470

PALAEONTOLOGY, VOLUME 9

Character of visual field. The maximum range of vision is 16-184° longitudinally andO-
32° latitudinally. The upper visual limit is latitudinal, but the elevation of the lens-axes
of the lower ascending diagonal row increases posteriorly, hence there is a postero-ven-
tral lacuna in the potential visual field. The coverage of different regions of the visual
field by lenses of various sizes is shown in text-fig. 2e.

The strong differentiation in the curvature of certain regions of the visual surface in
plan and profile results in a high concentration of lens-axes in the lower region of the
visual field between 45° and 80° long., which is covered by files 3-10. Here the longi-
tudinal axial angles average 6°, but increase anteriorly to 14°, and posteriorly where the
curvature is extreme, to 20°. Latitudinal axial angles about 3° in the region of maximum
concentration, increasing upwards to 14°; in other parts of the visual field averaging 10°.
The axes of the lenses of the upper horizontal row are latitudinal, those of the lower rows
are sublatitudinal.

Whereas in each of the anterior and central files, the lens-axes all have about the same
longitudinal bearing, in the posterior files there is a marked change in axial direction
from top to bottom, and this arrangement of lens-axes enables a considerable angular
range to be covered with a maximum economy of lenses.

Phacops rana erassitubereidata Stumm 1953

Plate 73, figs. 1-4; text-fig. 2a~c

1953 Phacops rana erassitubereidata Stumm, pp. 136-7, pi. 9, figs. 5-13, pi. 10, figs. 19-21.

The available material was an enrolled and complete topotype, SM H 7259, of very
large size with both eyes perfectly preserved.

Eye morphology (SM H 7259). Dimensions of eyes: L. (max.) 11 mm. (min.) 8 mm.; W.
(max.) 5-5 mm. (min.) 4-5 mm.; H. 6 mm., where cephalic length (sag.) is 22 mm., and
breadth 38 mm.

The chief morphological differences between the eyes of this subspecies and P. rana
milleri consist only of the relative size of the eye, the number and arrangement of
lenses, and the degree of inflation of the sclera. The lenses in SM H 7259 are disposed as
follows in 18 files:

Right eye: 345 565 655 555 544 323; Max. 6; Total 80
Left eye: 345 565 656 555 544 323; Max. 6; Total 81

The files diverge ventrally as in P. rana mdleri, but the lenses are much more widely
and evenly spaced, and none are contiguous. Variation in lens-size in different parts of
the visual surface resembles that in P. rana milleri. Lens diameters : Max. (posterior two
files) 0-7 mm., average (central and anterior files) 0-6 mm., Min. (centre of upper
horizontal row) 0-25 mm.

Interlensar sclera more highly inflated than that of P. rana milleri, hence all the lenses
are more deeply sunken, and the depressions in which the posterior lenses lie are more
profound, though less deep than those in the central region.

Charaeter of visual field. In both eyes, the visual range is almost identical in extent with
that of P. rana milleri in spite of the difference in arrangement of lenses (15-180°, long.,
0-34° lat.), and there is a similar concentration of lens-axes anterior to the polar

E. N. K. CLARKSON; EYES AND VISION IN PHACOPID TRILOBITES 471

meridian. There is less extreme variation in latitudinal axial angles owing to the shorter
dorso-ventral files; here they range from 4 to 12°. The difference in longitudinal bearing
between the upper and lower lenses of the posterior files is rather higher than in H 6033,
hence the rows of points indicating the axes of the lenses of these files on the stereogram
appear more slanted.

Phacops fecimdus Barrande 1846
Plate 73, figs. 7-9, Plate 74, figs. 3, 6; text-fig. 3 a-d

1 846 Phacops fecimdus Barrande, p. 46.

\%51 Phacops fecimdus Barrande; Barrande, pp. 514-18, pi. 21, figs. 1-27, pi. 22, figs. 32, 33.

1872 Phacops fecimdus Barrande, and vars. comimmis and degeiier Barrande, pp. 24-25, pi. 13,
figs. 1-14.

Variation in the eye of P. fecimdus as described by Barrande, has already been noted
(p. 466). In subsequent accounts of recorded occurrences of P. fecimdus and varieties
(Erben 1952, p. 328-30, with synonomies), no further details as to the morphology of the
eyes have been given. Only one adult eye-variant has been examined. Several specimens,
SM A 49374, 49375, from Stage Eg (Barrande 1852), Silurian, Dlauha Hora, Bohemia,
and others were examined. The former is the best preserved, and shows an interesting
irregularity in the disposition of the lenses.

Eye morphology (left eye SM A 49374). Dimensions of eye: L. (max.) 4-3 mm. (min.) 4-0
mm.; W. (max.) 2-8 mm. (min.) 2-0 mm.; El. 2-0 mm., where cephalic length (sag.) is 12
mm., and breadth 20 mm.

Eye of moderate size, not very prominent, situated near the anterior angle of the
librigena, and occupying the central third of the total cephalic length. The anterior edge
almost touches the axial furrow; the posterior edge lies further from the sagittal line and
is separated from the axial furrow by a broad inflated palpebral area. In this specimen
the latter edge falls far short of the posterior marginal furrow, but Barrande regarded
this character as variable. In profile the base of the eye lies about half way between the
plane of the genal angle and the glabellar crown ; the palpebral lobe lies well below the
latter.

Curvature of visual surface moderate and fairly uniform, in plan slightly increasing
posteriorly. Lenses set on a thickened inflated pad, which is bounded ventrally by a deep
groove, and which projects laterally outwards from the librigena, but does not reach as
far as the plane of the cephalic border. Outer tuberculate rim of palpebral lobe greatly
thickened, projecting outwards above facial suture, and separated adaxially from the
slightly inflated centre of the palpebral lobe by a smooth deep groove. Palpebral area
strongly inflated, slightly higher than lobe, opening posteriorly to the outwardly declined
librigena.

Lenses disposed in 17 files, regularly arranged apart from in the hiatus which lies be-
tween the 12th and 16th dorso-ventral files.

456 666 666 664* 655 43; Max. 6; Total 90. (* denotes irregularity.)

The first three files converge ventrally, the rest diverge at a moderate angle. Spacing of
lens centres within files is almost uniform. The lens pattern is complicated by the great
variation in lens size, which decreases ventrally in the outer files. Lenses of central files
separated by more than a lensar radius; elsewhere they may be almost contiguous.

472

PALAEONTOLOGY, VOLUME 9

Largest lenses 0-25 mm., average 0-2 mm., smallest 0-1 mm. Very small lenses occur at
the bottom of the diagonal rows as well as in the centre of the upper horizontal row. The
former are immature. Accessory upper horizontal row present, arising in file 11, but dis-
rupted by hiatus in files 12-16, where the upper lenses are displaced anteriorly from their
true position.

Interlensar sclera hardly inflated, finely granular, with larger tubercles in places,
forming irregular zigzag rows between the dorso-ventral files.

Text-fig. 2)d represents for comparison the visual surface of a slightly larger specimen
SM A 49375, in which the lenses (maximum diameter 0-35 mm.) are disposed regularly
in 19 files as follows:

Character of visual field. In SM A 49374 the maximum visual range is 27-150° longi-
tudinally, and —3 to 36° latitudinally. A large postero-ventral lacuna is present. The
lens-axes are dispersed more uniformly than in P. rana, except for the region of the de-
scribed hiatus. Axial separations in both directions range between 4 and 7°, increasing
slightly posteriorly and ventrally. The visual field of SM A 49375 is very similar in ex-
tent, 25-153° long., and —4 to 39° lat. Owing to the larger number of lens-axes cover-
ing a visual field of similar extent to SM A 49374, latitudinal separations are slightly
less and are more uniform.

c

TEXT-FIG. 3. P/iacops fecinulus Barrande. a, b. SM A 49374. Visual surface of left
eye (x 5) and stereogram. The hiatus breaking up of the regularity of the lenses,
and the effects of this hiatus upon the dispersal of lens-axes are visible, c, d. SM
A 49375. Visual surface of right eye (x 5) and stereogram.

346 676 767 667 665 543 2; Max. 7; Total 102.

Phacops breviceps Barrande 1 846

Plate 74, figs. 1, 2, 4, 7; text-fig. 4a-e

1846 Phacops breviceps Barrande, p. 71.

1852 Phacops breviceps Barrande; Barrande, pp. 518-19, pi. 22, figs. 24-31.
1951 Phacops cf. breviceps Barrande; Beckmann, pp. 126-41, pi. 10, 16 figs.

E. N. K. CLARKSON; EYES AND VISION IN PHACOPID TRILOBITES 473

Barrande (1852) recorded 18-22 files per eye, each file containing 5-8 lenses (normally
7), the overall number of lenses ranging from 83-128 according to the age of the speci-
men. He noted that in the larger specimens the lenses were much more widely spaced and
the degree of scleral inflation was greater.

Beckmann (1951) did not distinguish any notable variation in the number and arrange-
ment of the lenses in P. cf. breviceps, but in the species defined by Barrande variation
similar to that of P.fecimdiis major (text-fig. 1) occurs in different specimens found in
the same beds. The two principal adult eye-variants found here are for convenience desig-
nated eye-variant A and B. Nine cephala, all from Koneprusy Limestone, Stage Fo
(Barrande 1852), Devonian, Koneprus, Bohemia, were examined. A specimen of
average size, H 8442 (eye-variant A), and the largest specimen, SM H 8440 (eye- variant
B) are described here.

Eye-Variant A (right eye SM H 8442). Dimensions of eye: L. (max.) 6-5 mm. (min.) 5-0
mm.; W. (max.) 3-5 mm. (min.) 2-0 mm.; H. 3-0 mm., where cephalic length is 10 mm.
and breadth 22 mm.

Eye medium-sized, quite prominent, visual surface extending from the axial furrow
just anterior to the median transverse line, almost to the posterior marginal furrow.
Both edges lie approximately in the same exsagittal plane, but owing to the strong
rearward convergence of the axial furrows, the posterior edge is separated from the
glabella. In profile, the eye occupies the central third of the height of the cephalon, hence
the palpebral lobe falls short of the glabellar crown. Plan curvature of visual surface
moderate, increasing behind transverse line; in profile the lower part of the eye is seen to
be flattened; the degree of flexure increases upwards, becoming extreme just below the
facial suture. Eye surface projects outwards above the shallow groove at its base, reach-
ing laterally almost to the plane of the antero-lateral cephalic border. Palpebral lobe
sparsely tubercular, rising gently from thickened rim above the facial suture to the
curved palpebral furrow, beyond which the smooth palpebral area shelves adaxially to
the axial furrow. Lenses disposed as follows in 18 dorso-ventral files:

456 566 666 565 544 332; Max. 6; Total 87.

All files diverge ventrally at a high angle, which is less in the first three files. Lenses
quite closely packed, showing little variation in spacing within each file, almost contigu-
ous in outer files. Size distribution as in P. rana, largest lenses 0-45 mm., average 0-35-
0-4 mm., smallest 0T5 mm. In the ascending diagonal rows there is a rearward increase
in diameter, and in the descending diagonals a forward increase.

Interlensar sclera somewhat inflated and smooth.

Eye-Variant B (left eye, SM H 8440). Dimensions of eye: L. (max.) 10 mm. (min.) 7-0
mm., W. (max.) 6-0 mm. (min.) 4-5 mm., H. 4-5 mm., where cephalic length (sag.) is 18
mm. and breadth 38 mm.

The position and overall appearance of the eye is not dissimilar to that of SM H 8442
(eye-variant A), but the palpebral lobe shows a higher degree of inflation and is more
coarsely tuberculate, and its outer edge where it contacts the facial suture is thickened.
As seen directly in lateral aspect this ridge is slightly downwarped. The chief morpho-
logical differences between the two eye-variants are exhibited by the visual surface. In
this specimen the last two files have been broken olf. Impressions of the inner surfaces of

474

PALAEONTOLOGY, VOLUME 9

the lenses have, however, been left on the underlying matrix, and have been used for
spatial and angular measurements. Lenses disposed as follows in 19 files which diverge
ventrally as in eye-variant A :

345 455 555 454 544 443 1 ; Max. 5; Total 79.

Lenses quite widely spaced, the periphery of each generally lies at a distance of more than
a lensar radius from that of its neighbours, but in the first and last few files, where their
diameter is particularly large, they are almost contiguous.

TEXT-FIG. 4. Phacops breviceps Barrande. a, b, SM H 8442 (eye-variant A). Stereogram
and visual surface of right eye (X 5). c-e, SM H 8440 (eye-variant B). c-d. Stereogram
and visual surface of left eye ( X 5). e, Contoured stereogram showing latitudinal separa-
tions of lens-axes. Longitudinal separations average 7-10°.

Largest lenses (files 17, 18, 19), 0-55 mm. diam., average 0-4 mm., smallest 0T5 mm.,
those of the first file (1) increasing ventrally from 0-45 mm. to 0-525 mm.

Interlensar sclera inflated so that each lens lies in a shallow cup, but the top of each
lens projects above the level of this sclera.

Characters of visual field. The species Phacops rana, P. fecimdus, P. latifrons (PI. 73, figs.
5-6), and P. breviceps all exhibit a fairly similar pattern of lens number and distribution.
The visual fields of the first three are likewise similar, having a wide latitudinal range,
(c. 35°), and a strong postero-ventral lacuna. The visual field of P. breviceps however,
differs radically from these, having a comparatively narrow latitudinal range and no

E. N. K. CLARKSON: EYES AND VISION IN PHACOPID TRILOBITES 475

lacuna. In both eye-variants the total range is about 15-175° longitudinally, and 12-32°
latitudinally.

The pattern of lens-axis dispersal is peculiar. The profile surface curvature of the eye
in the lower part of the visual field is slight, and axial separations are only 2° or so, but
owing to increasing upward curvature, they may reach 10°. The plan curvature is much
higher and axial separations vary between 7 and 10°. There are thus partial visual
strips, strongly evident in the lower and central parts of the visual field, but which be-
come less distinct towards the top. In eye-variant B there are fewer lenses per file, and
the strips are not quite so apparent. Anteriorly the strips are longitudinal, but in the
central and posterior regions they run diagonally across the visual field.

Phacops batracheus Whidborne 1889

Plate 74, figs. 5, 8; text-fig. 5 a, b

1889 Phacops batracheus Whidborne, pp. 2-4, pi. 1, figs. 2-7.

Whidborne (1889, p. 2) commented in his full description of this species that the lenses
of the eyes were ‘very large and convex, in 18 perpendicular rows of from four to six
lenses each, between 80 and 90 in all’. He figured the cephalon and left eye of SM H 4072
(1889, pi. 1, fig. 5), which is described here.

All Whidborne’s syntypes from the Middle Devonian, at Lummaton, Devon
(SM H 4067-74), and many other specimens were available. There is apparently only
one eye-variant present, typified by Whidborne’s figured specimen, which is perfectly
preserved and undistorted, though the surfaces of a few lenses are slightly damaged.

Eye morphology (right eye SM H 4072). Dimensions of eye; L. (max.) 4-5 mm. (min.)
3-5 mm.; W. (max.) 2 mm. (min.) 1-5 mm.; H. 1-75 mm., where the length (sag.) of the
cephalon is 9-5 mm., and breadth 20 mm.

Eye large, prominent, lunate, lying in the posterior half of the cephalon, occupying
half its total length (sag.), extending from just anterior to the median transverse line to
within 0-5 mm. of the posterior marginal furrow. The posterior edge is further from the
sagittal line. In profile the eye occupies less than a third of the total height of the cepha-
lon. The base of the visual surface is situated midway between the plane of the genal
angle and the glabellar crown. Since the vertical extension of the eye is small as compared
with its longitudinal span, the palpebral lobe lies well below the level of the glabella. The
visual surface projects laterally almost to the antero-lateral cephalic border.

Plan curvature moderate, increasing posteriorly; profile curvature similar increasing
ventrally. Palpebral lobe sparsely tuberculate, rising adaxially above narrow outer rim,
becoming horizontal near axial furrow. Palpebral area continuous with lobe; there is no
palpebral furrow.

Lenses strongly convex, disposed in 1 8 files ;

355 565 656 554 444 332; Max. 6; Total 80.

The first three files are parallel, the others diverge ventrally at a rather high angle. In-
dividual lenses quite closely packed, almost contiguous in outer files, elsewhere there is
about half a lensar radius between them. Within the files each lens centre is situated at
a uniform distance from its neighbours. Lens diameters: largest (files 17 and 18) 0-4 mm.

476

PALAEONTOLOGY, VOLUME 9

diam. ; average 0-375 mm., minimum 0-15 mm., in the centre of the upper horizontal
row, and at the lower ends of the longest diagonals. There is no accessory upper hori-
zontal row.

Interlensar sclera only slightly inflated, each lens projects conspicuously from the
visual surface. The lenses of the upper horizontal row lie in shallow crater-like de-
pressions, each of which has a prominent scleral rim. The lower edges of the rims are
more strongly pronounced, and the lenses set obliquely to the surface of the sclera, and

TEXT-FIG. 5. Phacops batracbeus Whidborne. SM H 4072. a, b. Visual
surface of right eye (X5) and stereogram.

are thus directed at a lower angle to the horizontal than if they had been normal to the
surface (cf. P. rana).

Characters of visual field. The maximum range of vision is 20-156° longitudinally, 3-32°
latitudinally. The upper visual limit is latitudinal and there is a deep postero-ventral
lacuna present. The lens-axes are dispersed regularly and uniformly without any ten-
dency to cluster or form visual strips. Both the latitudinal and longitudinal separations
are about the same, averaging 7°. The separations in the lower part of the visual field are
slightly greater.

Phacops boecki Hawle and Corda 1847

Plate 74, figs. 9, 10; text-fig. 6 a, b

1 847 Phacops boeki Hawle and Corda, p. 1 07.

1852 Phacops boecki Corda; Barrande, p. 513, pi. 20, figs. 30-32.

The fullest previous description is that of Barrande (1852) who recorded 18-21 files
per eye, with 5-8 lenses as a maximum depending on the size of the specimen. The total
number of lenses ranged from 66-134.

EXPLANATION OF PLATE 74

Figs. 1, 2, 4, 7. Phacops breviceps Barrande. Koheprusy Limestone, Stage Fj, Devonian, Koheprus,
Bohemia. 1, 2, SM H 8440 (eye-variant B). Left eye in lateral and dorsal aspects. X 5. 4, 7, SM H 8442
(eye-variant A). Right eye in lateral and dorsal aspects. X 5.

Figs. 3, 6. Phacops fecundus Barrande. SM H 49375. Stage E,, Silurian, Dlauha Hora, Bohemia. Right
eye in lateral and dorsal aspects, x 5.

Figs. 5, 8. Phacops batracheus Whidborne. SM H 4072. M. Devonian, Lummaton, Devon. Right eye
in lateral and dorsal aspects. X 5.

Figs. 9, 10. Phacops boecki Hawle and Corda. SM H 8439. Dvorce Limestone Stage Gi (lower part),
Devonian, Dvorce, Bohemia. Left eye in lateral and dorsal aspects. X 5.

Palaeontology Vol. 9

PLATE 74

CLARKSON, Schizochroal eyes in phacopid trilobites

E. N. K. CLARKSON: EYES AND VISION IN PHACOPID TRILOBITES 477

Only one well-preserved specimen was available, so only the one eye-variant has been
noted. SM H 8439 is a very large adult from Dvorce Limestone, Stage (lower part)
(Barrande 1852), Devonian, Dvorce, Bohemia. Both eyes are well preserved, though the
surfaces of some of the lenses are slightly roughened by weathering.

Eye morphology (left eye SM H 8439). Dimensions of eye : L. (max.) 7 mm. (min.) 5 mm. ;
W. (max.) 4 mm. (min.) 3 mm. ; H. 4 mm., where cephalic length is 13 mm., and breadth
23 mm.

Eye very large, prominent, occupying almost half length (sag.) of cephalon, visual sur-
face extending from just anterior to median transverse plane of cephalon almost to the
posterior marginal furrow. Both edges lie in approximately the same exsagittal plane. In
profile, the eye base is set low on the cephalon, the palpebral lobe standing not quite as

TEXT-FIG. 6. Phacops boecki Hawle and Corda. SM H 8439. a, b. Visual surface
of left eye ( x 5) and stereogram.

high as the glabellar crown. The visual surface forms a conic segment, with parallel
upper and lower borders, whose height/base ratio, if projected to form a cone would be
5:1. Eye surface bounded ventrally by an inwardly shelving flat strip, projecting laterally
so as to overhang cephalic border. Plan curvature uniform, moderate ; profile curvature
very low ventrally, increasing dorsally. Palpebral lobe smooth, bounded laterally by
thickened rim, somewhat inflated, shelving adaxially to indistinct palpebral furrow.
Palpebral area flat, open posteriorly.

Lenses, as far as can be ascertained, strongly convex, disposed in 21 files:

678 788 888 888 878 776 654; Max. 8; Total 150.

All files diverge ventrally at a very low angle.

Lenses rather widely spaced, generally separated by distance of a lensar radius and re-
markably constant in size throughout the visual surface, peripheral lenses generally
slightly smaller. Largest lenses 0-35 mm., diam. smallest 0-2 mm.

Interlensar sclera only slightly inflated between lenses of central files.

Character of visual field. The maximum range of vision is 25-172° longitudinally, 5-32°
latitudinally. Both the upper and lower limits are almost entirely latitudinal. In spite of
the proximity of the posterior edge of the eye to the rear border of the cephalon, there is
only a slight upward deflection of the lower limit in the hindermost part of the visual
field, posterior to 140° long, and covered by files 18-21.

Owing to the extremely low profile curvature of the visual surface near the base of the

478

PALAEONTOLOGY, VOLUME 9

eye, the axes of the lower lenses of the dorso-ventral hies are concentrated in visual strips,
in which axial separations are as low as 1°; but the upward increase in curvature results
in separations of up to 12° near the top of the visual held and the strips fade out dor-
sally. The visual strips are quite regularly spaced, and the angles separating them vary
between about 7 and 10° of longitude.

Superfamily dalmanitacea
Dahnauiles vulgaris (Salter 1849)

Plate 75, figs. 1-4; text-fig. 1 a-e

1822 Asaphus caudatus Brongniart, p. 22, pi. 2, figs. Aa-d.

1845 Dalmamua caudatus (Brunnich); Emmerich, p. 38.

1849 Phacops caudatus (Brunnich) var. vulgaris Salter, pp. 1-6, pi. I, figs. 1-12, 15.

1864 Phacops (Odoutochile) caudatus (Brunnich) var. vulgaris Salter; Salter, pp. 49-53, pi. 3,
figs. 4-17. (= Dalmanites caudatus (Brunnich) of later writers.)

1935 Dalmanites vulgaris (Salter); Delo, pp. 424-6, pi. 48, figs. 8, 9.

Salter’s descriptions of 1849 and 1864 include good illustrations of the eye. He re-
corded about 240 lenses per eye, and 8-10 per hie. He also noted that in some specimens
the eyes were larger than in others and referred this difference to sex, concluding that
the male had larger eyes.

Many specimens from the Wenlock Limestone have been examined, but although
some variation is apparent within the species there is as yet no conclusive evidence of
sexual dimorphism.

In the best-preserved specimen examined, SM A 28635, both eyes are well preserved,
though part of the left eye is chipped and two patches of polyzoa have encrusted the
upper rim of the right eye, obscuring a few lenses. The lenses retain their full convexity.

Eye morphology (right eye SM A 28635). Dimensions of eye; L. (max.) 5 mm. (min.)
3 mm.; W. (max.) 4 mm. (min.) 2-5 mm.; H. 3 mm., where cephalic length (sag.) is 17
mm., and breadth 30 mm.

Eye large, situated posteriorly on the cephalon, standing high above the glabella, and
forming a truncated cone whose height/base ratio, if projected to form a cone would be
4:1. The anterior edge is separated from the axial furrow, and the posterior edge lies
further from the sagittal line. The eye base occupies the summit of the librigena, and is
surrounded by a moat-like depression, with an external concentric ridge, which limits
the genal field. Plan curvature very strong and fairly uniform, but in profile the visual
surface is almost flat and inclined at about 80° to the horizontal. In the ‘natural atti-
tude’ (text-fig. 7 b, e). the upper rim is oblique to the horizontal, and the eye appears to
slope forwards and downwards, though the ‘horizontal’ lens-rows actually lie in a
horizontal plane.

The narrow {tr) palpebral lobe is almost uniformly curved in plan and consists of
a narrow strip with a shallow concentric groove running parallel with the facial suture.
It plunges steeply adaxially to the palpebral furrow, which lies a short distance below
the upper rim of the eye. The tuberculate palpebral area declines with increasing steep-
ness towards the axial furrow.

E. N. K. CLARKSON: EYES AND VISION IN PHACOPID TRILOBITES 479

There are 34 dorso-ventral files in which the lenses are disposed as follows :

467 899 109*10* 10*11*10* 11*10*11* 10*11*10 11 10 10 10 109 988 788 765
2; Max. 1 1 ; Total 294 ( * In the regions where the eye surface is covered by
polyzoa, the number of lenses in the other eye have been counted.)

TEXT-FIG. 7. Dahnanites vulgaris (Salter). SM A 28635. a, c, Stereogram and visual surface of right eye
(X5). The blank circles and the corresponding enclosed areas on the stereogram represent lenses
covered by polyzoa. The equivalent lenses and lens-axes of the left eye are substituted here, b. Left eye
in ‘natural orientation’ showing a horizontal row (hr) and one of the dorso-ventral files (dvf), which
in this attitude appears vertical, d, Limits of binocular vision shown by the bearings of the upper
(hvu) and lower (bvl) lenses of the first and last files, e. Profile view of the cephalon showing limits of
binocular vision with respect to the upper surface of the head. Apparent angular bearings are shown
here (apparent latitude = tan 6 cos (f> where 6 is the latitudinal and <j> is the longitudinal bearing of

the lens-axis in question).

Apart from the first and last three dorso-ventral files, which are parallel, the files
diverge ventrally at a very high angle (the highest observed so far in any Phacopina).
Within the files the lenses are closely packed together. In the lower part of the visual sur-
faee the lenses of adjacent files are quite widely spaeed owing to the high downward
divergence of the files so that there is about the distance of a lensar radius between them ;
but in the median and upper parts of the eye they are almost contiguous. The distance
between the lenses within the files gradually inereases upwards. The largest lenses

480

PALAEONTOLOGY, VOLUME 9

(diam. 0-3 mm.) occur in the central region of the visual surface. On the periphery of this
region they are smaller and all round the borders of the visual surface lenses of diameter
less than 0-2 mm. are present. Hence the lowest two or three aseending and descending
diagonal rows eontain small lenses.

Minute lenses are found in the centre of the upper horizontal row and in an ineipient
aecessory upper horizontal row which is present at the top of files 26, 28, and 30.

Interlensar selera is indistinguishable in the region of the large closely paeked lenses,
but is visible at the bottom, of the eye, and is tubereulate. Each lens here is surrounded
by an irregular polygon of fine points.

Character of visual field. The maximum range of vision is 265° longitudinally, and 12-34°
latitudinally. The visual fields of the two eyes overlap anteriorly and posteriorly and, as
in all dalmanitids, the lenses covering the overlapping parts are small. The axes of the
top and bottom lenses of the first files have slightly different longitudinal bearings (40
and 35°) and meet just in front of the glabella and the anterior margin of the cephalon
respectively. Lens-axes of the last file likewise have different bearings (—35 and 45°) and
meet at some distance behind the oeeipital ring clearing it by a fair angle (text-fig. 7 d, e).
The upper limit is largely latitudinal, as is the lower over the anterior region, but pos-
teriorly there is a shallow laeuna.

As in A caste domungiae the strong differential curvature of the visual surfaee results
in the arrangement of the lens-axes of the files into very distinet visual strips. Within the
strips axial separations are usually less than 2° and never more than 4°. Between strips
axial angles are about 6-7° though axial separations twice this figure occur between files
19 and 20, just posterior to the polar meridian. A unique feature of this eye is that all
the visual strips slant the same direetion and the top lens-axis in each file has about the
same angular bearing as the bottom lens-axis of the preceding file. This condition may
be contrasted with that in A. downingiae, where the strips follow a radial pattern, and
that in Chasmops odini where they are curved. It may be considered as an adaptation
whereby the whole longitudinal field of view is covered, with maximum economy of
lenses.

In the overlapping parts of the visual field the separation of the visual strips is rather
more variable.

Dalmauites eaudatus (Briinnieh 1781)

Plate 75, figs. 5, 6; text-fig. 8 a-j
1781 Trilobus eaudatus Briinnieh, p. 392.

1839 Asaphus longicaudatus Murchison, p. 656, pi. 14, figs. 11-14.

EXPLANATION OF PLATE 75

Figs. 1-4. Dalmauites vulgaris (Salter). SM A 28635. Wenlock Limestone, Silurian, Dudley, England.
1 , 2, 4, Right eye in lateral, dorsal, and dorso-lateral aspects. X 7-5. 3, Lenses near the lower rim of the
visual surface. X 20.

Figs. 5, 6. Daimanites eaudatus (Briinnieh). SM A 28644. Wenlock Limestone, Silurian, Dudley,
England. Left eye in lateral and dorsal aspects. X 5.

Figs. 7-1 1. Chasmops odini (Eichwald). SM A 53427. C2 Kukruse shale or C3 Itfer shale, Caradocian,
Estonia. 7, Right eye, lenses near the lower rim of the visual surface. X 25. 8-1 1 , Right eye in lateral,
frontal, dorso-lateral, and dorsal aspects. X 5.

Palaeontology, Vol. 9

PLATE 75

-.-•A

m.

CLARKSON, Schizochroal eyes in phacopid trilobites

E. N. K. CLARKSON; EYES AND VISION IN PHACOPID TRILOBITES 481

1845 Dalmaiuiia longicaudota (Murchison); Emmerich, p. 40.

1849 Phacops caudatus (Briinnich) var. longicaiidatiis (Murchison); Salter, pp. 1-6 (pars), pi. 1,
figs. 13-14.

1864 Phacops (Odontochile) longicaiidatus Murchison; Salter, pp. 55-56, pi. 3, figs. 19-28.
( = Dalinanites longicaiidatus of later authors.)

1935 Dalinanites caudatus (Briinnich); Delo, pp. 424-6, pi. 48, figs. 6, 7.

The fullest previous descriptions of this Wenlock Limestone species are those of
Salter (1849, 1864). He made only brief reference to the eye.

Many specimens have been examined but only one eye-variant has been distinguished.
SM A 28644 is well preserved, though the lensar surfaces have in some cases been
planed off. Both eyes are identical.

Eye morphology (left eye SM A 28644). Dimensions of eye: L. (max.) 5 mm. (min.) 3-5
mm.; W. (max.) 3-5 mm. (min.) 3 mm.; H. 3 mm., where cephalic length is 13 mm. and
width 20 mm.

The form and situation of the eye resembles that of D. vulgaris; the major differences
lie in the degree of curvature, the character of the palpebral lobe, and the number and
arrangement of the lenses. The visual surface is very strongly and regularly curved in
plan, whereas in profile the degree of curvature of the anterior part of the visual surface
is considerable but lessens posteriorly so that in the region of the last few files the eye
is almost flat. The palpebral lobe is extremely narrow (tr), and plunges very steeply to the
U-shaped palpebral furrow, which lies well below the upper rim of the eye.

Lenses disposed as follows in 32 files;

457 889 899 999 989 899 999 988 888 776 52; Max. 9; Total 249.

The outer files converge ventrally, but the central ones diverge at quite a high angle.
Within files lens spacing increases ventrally. In the upper regions they are so closely
packed as to be hexagonal, elsewhere they are rounded. The largest lenses (0-25 mm.)
occur in a wide crescent running through the upper anterior, lower central, and upper
posterior parts of the visual surface. Outside this crescent lie smaller lenses (0T75 mm.).
In SM A 28644 the central lenses of the upper horizontal row have failed to develop.
Three minute lenses at the top of files 25, 27, and 29 represent an incipient accessory
upper horizontal row.

The lenses are so closely packed that no interlensar sclera can be seen.

Character of visual field. The maximum range of vision is 254° longitudinally, 3-37° lati-
tudinally. There is an overlap of the visual fields of the two eyes comparable with that of
D. vulgaris (text-fig. Se,f), though there is a greater difference (15°) between the longi-
tudinal bearings of the top and bottom lenses of file 1, which seems to be associated with
the relatively longer distance between the front of the glabella and the anterior point.

Lens-axes of the dorso-ventral files form visual strips which are separated by angles of
5-8°. Some of these strips are longitudinal, others, particularly those associated with the
more posterior files, are divergent. Although the strips are quite regularly spaced they
are quite unlike those of D. caudatus because of the strong differentiation in profile cur-
vature of the eye from front to rear. On the contoured stereogram (text-fig. 86) this
change in curvature is shown by a band of closely spaced lens-axes which traverses the
visual field diagonally in the anterior part of the visual field, and lies almost horizontally

482

PALAEONTOLOGY, VOLUME 9

posteriorly. Within this band latitudinal axial separations average 2-3°, outside it they
may be as high as 10°, and normally approximate the longitudinal separations.

The elevation of the lens-axes within the horizontal rows declines gradually back-
wards, but as the three small lenses of the accessory upper horizontal rows have eleva-
tions of 37°, the upper visual limit is unbroken.

TEXT-FIG. 8. Dalmanites caiidatus (Briinnich). SM A 28644. a, d. Stereogram and visual surface of left
eye ( X 5). b. Contoured stereogram showing latitudinal separations of lens-axes. The diagonal band
containing the lowest separations is distinct. Longitudinal separations average 6-8°. c. Left eye in
natural orientation, e, f. Limits of binocular vision, notation as in text-fig. 7.

Chasmops odini (Eichwald 1 840)

Plate 75, figs. 7-11 ; text-fig. 9 a,b
1840 Calymeiie odini Eichwald, p. 62.

1851 Chasmops odini (Eich. sp.); M‘Coy, p. 164, pi. Ig, figs. 22, 23.

1881 Chasmops odini (Eichwald); Schmidt, p. 101, tab. 2, figs. 1-13.

1937 Chasmops odini (Eichwald); Opik, p. 77, tab. 9, figs. 5, 6, text-figs. 23-25.

E. N. K. CLARKSON: EYES AND VISION IN PHACOPID TRILOBITES 483

In Eichwald’s original description of this species little reference was made to the eye,
but Schmidt gave a very detailed account, noting about 240 lenses in 20-22 files, and
about 12 lenses per file. His excellent figures (1881, figs. 2, 8, 9) showed the almost flat
visual surface. No new information was recorded by Opik.

The material studied was a single cephalon, SM A 53427, from the Caradocian of
Estonia (€3 Kukruse shale or C3 Itfer shale (after Opik)). It is preserved in full relief and
both eyes are present. The upper surface of the left eye is broken, but the right is un-
damaged apart from the posterior upper rim and a few of the topmost lenses which have
been broken off.

Eye morphology (right eye SM A 53427). Dimensions of eye; L. (max.) 5-5 mm. (min.)
5-0 mm. ; W. (max.) 4-5 mm. (min.) 4-0 mm. ; H. 5-5 mm., where cephalic length (sag.) is
16 mm. and breadth 30 mm.

Eye very prominent and set high on the cephalon forming a high truncated cone, which
if projected to form a cone would have a height/base ratio of 10; 1. The anterior edge is
situated almost exactly half way between the anterior border and the occipital ring, on
a level with glabellar furrow (3p), and the posterior edge lies some way (2 mm.) anterior
to the posterior marginal furrow. Both the anterior and posterior edges of the eye lie
almost in the same exsagittal plane, the former being separated from the axial furrow by
the highly inflated fixigena. The visual surface reaches almost to the antero-lateral
border.

A deep lunate depression surrounds the base of the eye, external to which lies the
narrow genal field from which the librigena plunges vertically. In plan the visual surface
is strongly curved and parabolic about the transverse line but in profile it is almost flat
with only the slightest curvature. The prominent outer rim of the palpebral lobe
surrounds the whole of the upper edge of the visual surface and projects outwards over
it, becoming thicker where it plunges downwards subparallel with the anterior and pos-
terior dorso-ventral file. Adaxially to the rim the inner face of the palpebral lobe slopes
to meet the palpebral furrow, which in its central part is very sharply reflexed so as to
be almost V-shaped (the point of the V being directed outwards). At the base of the eye
the palpebral furrow joins with the deep groove surrounding the bottom of the visual
surface, so that the whole eye is encircled by one continuous furrow. The palpebral area
is inflated and open to front and rear.

There are 22 dorso-ventral files, with the following lens distribution ;

5 7 10 11 12 13 14 14 14 14 14 14 14 13 12 12 11 10 975 4; Max. 14; Total 239.

In all other phacopids which have been investigated the dorso-ventral files are straight,
but here only the central files are straight and the outer files are curved outwards, the de-
gree of flexure increasing in opposite directions away from the centre. (In text-fig. 9 the
curvature of these files is shown as accurately as possible, but in representing the distance
apart of the outer lenses there is some inevitable distortion.) Thus the files diverge from
the top downwards at a moderate angle from each other and just below the median
plane they being to converge again. All the lenses are quite closely packed, but there is
a gradual upward increase in the distance between the lenses in the dorso-ventral files,
hence the lowest lenses are elliptical and almost contiguous, whereas the smaller upper

C 4179 K k

484

PALAEONTOLOGY, VOLUME 9

lenses are rounded and are nearly their own diameter apart. In certain files elliptical
lenses occur near the top of the eye but generally they are round. Between the lenses of ad-
jacent dorso-ventral files the separation is generally greater than the vertical distance
between lenses within a file.

The lenses of the lower central and outer parts of the visual surface are the largest
(0-3 mm.) and are of relatively uniform diameter. In the central files the lenses become
progressively smaller upwards from the median region, decreasing to 0-15 mm. diam.
Those of the upper horizontal row are of about the same diameter, and only increase
slightly in size away from the centre. The interlensar sclera is flat but furnished with
small round tubercles in the interstices of the lenses which form irregular poly-
gons surrounding each lens. At the base of the visual surface these tubercles are very
numerous and completely cover the lower rim.

TEXT-FIG. 9. Chasmops odiiii (Eichwald). SM A 53427. a, b. Stereogram and visual

surface of right eye ( X 5).

Characters of visiiaffield. The maximum range of vision is 180° longitudinally, 5-17°
latitudinally. Both visual limits are latitudinal, except at their posterior extremities,
where the elevation of the lenses of the last few files increases greatly.

The pattern of lens-axis dispersal, as shown on the stereogram, is one of the most re-
markable observed in any phacopid. The axes of the numerous lenses in the dorso-ventral
files are accommodated within the visual field in very distinct, widely separated, and
peculiarly curved visual strips. The curved appearance of these strips obviously relates
to the curving files themselves; both are bowed outwards in opposite directions away
from the central file, whose equivalent strip has a bearing of about 105° long. Within
strips the latitudinal axial separations of the lens-axes may be as low as though to-
wards the extremities it increases to 2° or 3°. Longitudinal separations between strips
may be as high as 15°. Apart from those in the last few files, where there is a sudden in-
crease in elevation, all the lens axes of the horizontal rows have approximately constant
bearings.

SUMMARY

This paper, together with my previous one (Clarkson, 1966u) describes the relation-
ship between eye-form and the spatial and angular characteristics of the lenses, in a
number of typical schizochroal trilobite eyes. Full physiological interpretations of this
and other information relating to trilobite vision are to be made in a later paper but a
brief concluding summary of observations to date can be given.

E. N. K. CLARKSON: EYES AND VISION IN PHACOPID TRILOBITES 485
External morphology of eyes

(i) Lenses upon the visual surface are normally arranged in a regular pattern of
hexagonal close packing. Breakdown in regularity occurs in a few cases.

(ii) During postlarval ontogeny there is an increase in lens number until full maturity
is reached. The adult lens-number within species is approximately constant, though in
some species there are apparently two distinct adult eye-variants, which may reflect
a sexually dimorphic condition. In the present state of taxonomic confusion it is not yet
possible to elucidate this matter further.

(iii) Within the visual surface of individual specimens of the Phacopina there is nor-
mally considerable variation in lens-size. Small lenses are always found in the upper
central region, and the newly emplaced lenses at the lower ends of the dorso-ventral
files are likewise small. The pattern is distinct in all the species investigated except
Chasmops odini.

(iv) Lens size, number, and spacing vary greatly within the Phacopina as a whole.
Eyes with very large lenses, e.g. Phacops rana crassitubercidata, normally only have a
limited lens number. In some eyes increased lens number has been achieved by the close
packing of the lenses so that they become hexagonal. Characteristic patterns of the
arrangement of lenses upon the visual surface relate to systematic categories and pre-
sumably reflect evolutionary relationships.

(v) Variation in lens-size within the visual surface of individuals is normally indepen-
dent of differential surface curvature.

(vi) The lenses are not always set normal to the visual surface but owing to their large
size their axial directions may readily be inferred from the external morphology of the eyes.

Characters of visual fields. Three main patterns of the arrangement and dispersal of lens-
axes are found within Phacopina, largely as a result of distinct modes of curvature of the
visual surface. In species with two adult eye-variants the visual fields of both may be of
similar extent, although the lens-axes may have a slightly different arrangement. These
patterns are as follows:

(i) Visual fields of wide (c. 30-40°) latitudinal extent, normally associated with eyes of
100 lenses or less, in which the plan curvature is rarely more than twice the profile
curvature. The upper visual limit is normally latitudinal, though a deep triangular
postero-ventral lacuna is commonly present which truncates the equatorial lower limit.
This device apparently prevents occlusion of vision in this region by the high posterior
cephalic border. Axial separations are high though some of the lens-axes may be clus-
tered together in the lower anterior part of the visual field. Examples are Phacops rana
milleri, P. rana crassitubercidata, P. Jatifrons, P. fecundus, P. batracheus, P. musheni
(see Clarkson, 1966/?), Acastoides const ricta.

(ii) Visual fields of narrow (10-20°) latitudinal extent usually relating to eyes with up-
wards of 100 lenses, curving very much more in plan than in profile. Both visual limits
are generally latitudinal ; if there is a lacuna it is shallow. The lower visual limit lies at 1 0°
or more above the equator. Such eyes are highly astigmatic, as the lens-axes are dis-
persed in visual strips which are present from a very early stage in postlarval develop-
ment. These strips cross the visual field from top to bottom and may be disposed either
radially as in Acaste downingiae, diagonally in Dalmanites vulgaris, or may themselves be
curved, e.g. Chasmops odini.

486

PALAEONTOLOGY, VOLUME 9

(iii) Visual fields which combine many of the features of (i) and (ii). The number of
lenses in the eye is variable. Both visual limits are latitudinal, the lower normally lying
a few degrees above the equator (12° in P. breviceps). Incomplete visual strips are
developed which run in a band normally parallel with the visual limits (apart from
Dalminites caudatus) but which do not extend the full latitudinal width of the visual field.
In P. breviceps and P. boecki this band of partial visual strips lies at the base of the visual
field; in Acaste dowmingiae macrops it is above this level; and in D. caudatus it runs
diagonally across the anterior part of the visual field and posteriorly runs parallel to and
just above the lower limit.

Acknowledgements. The author's thanks are due to Prof. O. M. B. Bulman for research facilities in the
Sedgwick Museum, Cambridge, and to Mr. A. G. Brighton for the loan of specimens in his charge.
Dr. M. J. S. Rudwick offered stimulating supervision throughout the work, and Mrs. C. M. Clarkson
gave helpful advice, and assisted with the preparation of the manuscript.

Much of the original research was carried out while the author was in receipt of a D.S.I.R. Research
Studentship.

REFERENCES

BARRANDE, J. 1846. Notlcc preliminaire sur le Systeme silnrien et les Trilobites de Bohenie. 1-97, pi. 1.
Leipzig.

1852. Systeme sUurien dit Centre de la Boheme. partie; Crnstaces; Trilobites. 1, 1-935,

pi. 1-51. Prague and Paris.

1872. Systeme silnrien du Centre de la Boheme ; Supplement an vol. I. Trilobites, Crnstaces diverses

et Poissons. 1-647, pi. 1-37. Prague and Paris.

BECKMANN, H. 1951. Zur Ontogenie der Sehflache groBaugiger Phacopiden. Paldont. Z. 24, 126-41,
pi. 10.

BRONGNiART, A. and DESMAREST, A. G. 1822. Histoire natnrelle des Crnstaces fossiles, sons les rapports
zoologiqnes et geologiqnes. 1-154, pi. 1-11. Paris.

BRUNNiCH, M. T. 1781. Beskrivelse over Trilobiten. K. danske Vidensk. Selsk. Skr. n.s. 1, 1-384.
CLARKE, J. M. 1 889. The structure and development of the visual area in the trilobite Phacops rana
Green. J. Morph. 2, 253-70, pi. 21.

CLARKSON, E. N. K. 1966u. Schizochroal eyes and vision of some Silurian Acastid Trilobites. Palae-
ontology, 9, 1-29, pi. 1-3.

\966b. The life attitude of the Silurian trilobite Phacops mnsheni Salter 1864. Scot. J. Geol. 2,

76-83, pi. 1.

DELO, D. M. 1935. The genotype of Dalmanites. J. Palaeont. 9, 424-6, pi. 48.

1940. Phacopid trilobites of North America. Spec. Pap. geol. Soc. Am. 29, 1-135, pi. 1-13.

EiCHWALD, c. E. VON. 1840. Sur le systeme silnrien de TEstonie. 1-122. St. Petersbourg.

EMMERICH, H. E. 1845. Uebcr die Trilobiten. Nenes Jb. Miner. Geol. Paldont. (Bd. 1845), 18-52, pi. 1.
ERBEN, H. K. 1952. Trilobiten aus dem alteren Hercyn (Unterdevon) des Unterharzes. Nenes Jb. Geol.
Paldont. Abh. 94, 150-362, pi. 17-24, 64 figs.

GREEN, J. 1832. A monograph of the Trilobites of North America. 1-93, pi. 1-4. Philadelphia.

HALL, J. and CLARKE, J. M. 1888. Trilobites and other Crustacea. Paleontology of New York, 7, 1-236,
pi. 1-36. Albany.

HAWLE, I. and CORDA, A. J. c. 1847. Prodrom einer Monographie der bohmischen Trilobiten. K. Bdhm.
Ges. Wiss. Abh. 5, 129-212, pi. 1-7.

LiNDSTROM, G. 1901. Researches on the visual organs of the trilobites. K. Svensk. Vetensk. Akad. Handl.
34, 1-86, pi. 1-6.

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

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

MURCHISON, R. I. 1839. The Silurian System. 1-768, pi. 1-40. London.

OPiK, A. A. 1937. Trilobiten aus Estland. Acta Comment. Univ. tartu. A. Math. — Phys.—-Med. 32,
1-163, pi. 1-26, figs. 1-42.

E. N. K. CLARKSON: EYES AND VISION IN PHACOPID TRILOBITES 487

RUDWiCK, M. J. s. 1964. The inference of function from structure in fossils. Br. J. Phil. Sci. 15, 27-40.
SALTER, J. w. 1849. Figures and descriptions illustrative of British Organic Remains. Decade II. Mem.
geol. Surv. U.K.

1864-8.3. A monograph of British Trilobites. Palaeontogr. Soc. [Monogr.], 1-224, pi. 1-30.

SCHMIDT, F. 1881. Revision der ostbaltischen silurischen Trilobiten nebst geognostiche Ubersicht des
ostbaltischen Silurgebiets. Abt. 1 Phacopiden, Cheiruriden, und Encrinuriden. Mem. Acad. imp.
sci. St.-Petersboiirg, 30, 1-237, pi. 1-16.

SHAW, F. c. and ormiston, a. r. 1964. The eye socle of trilobites. J. Pcdeont. 38, 1001-2, fig. 1.
STEWART, G. A. 1927. Fauna of the Silica Shale of Lucas County. Bull. geol. Surv. Ohio, 32, 1-76, pi. 1-5.
STUMM, E. c. 1953. Trilobites of the Devonian Traverse Group of Michigan. Contr. Mus. Paleont.
Univ. Mich. 10, 101-57, pi. 1-12, 1 map.

WHiDBORNE, G. F. 1889-92. A monograph of the Devonian Fauna of the South of England. Vol. 1.
Palaeontogr. Soc. [Monogr.], 1-344, pi. 1-31.

E. N. K. CLARKSON

Grant Institute of Geology,
University of Edinburgh,

Manuscript received 12 August 1965 Edinburgh, 9

TWO CRISTATE MEGASPORES FROM THE
EOWER CARBONIFEROUS OF SCOTLAND

by K. L. ALVIN

Abstract. A new megaspore described under the name Triletes pannosus sp. nov. resembles in its crest of ana-
stomosing hairs T. suhpalaeocristatiis Alvin from the same locality. A new specimen of this latter species
has yielded further details of its structure. Comparison is made between the two species.

Some fifteen specimens of a new megaspore were obtained from samples of shale col-
lected at Oxroad Bay, East Lothian. The shale contains abundant seed megaspore
membranes, microspores, and fragmentary carbonized plant remains. It was collected
and macerated in the hope that it might yield whole specimens of a megaspore described
recently from petrified material from the same locality under the name Triletes sub-
palaeocristatus Alvin (1965). This latter spore was so closely associated with the cone of
the lycopod Oxroadia gracilis Alvin that it was almost certainly borne by this plant. It
was described only from serial peel sections and was of interest because of its unusual
morphology in possessing a crest of hairs along the lips of the trilete ridge. Oddly enough
the new megaspore also possesses a crest, but appears genuinely distinct.

A new specimen of T. suhpalaeocristatus has been found in the original block con-
taining Oxroadia. This has been macerated out from the matrix and has yielded some
new information about this species for which an emended diagnosis is now given.

SYSTEMATIC DESCRIPTIONS
Genus triletes Bennie and Kidston ex Zerndt
Triletes pannosus sp. nov.

Plate 76, figs. 1-6; text-fig. I a-d

Diagnosis. Megaspore spheroidal; mean equatorial diameter 1,360 p. (range for nine
spores measured 1,180-1,580 p.); mean height (excluding crest) 1,260 ft (range for ten
spores measured 1 ,100-1, 440 ft). Wall about 30-50 ft thick. Arms of trilete ridge extending
about three-quarters the radius of the spore. Wall (except for contact areas) bearing
numerous spines about 200 ft long, sometimes up to 400 ft and up to 200 ft wide at the
base. Spines simple or forked near the tip, straight or commonly curved; base expanded
and buttressed; buttresses often forming separate ‘legs’. Papillae and simple hairs up to
100 ft long also present among the spines. Papillae, 10-20 ft in diameter, usually present
on contact areas. Lips of trilete ridge bearing a prominent crest up to 660 ft high at the
apex (mean height for twelve spores 525 ft), diminishing to about half this at the ends
of the trilete arms. Crest resembling a series of anastomosing hairs, but sometimes form-
ing a more or less continuous membrane with free teeth distally; hairs or teeth of very
varied thickness, generally 5-20 ft.

[Palaeontology, Vol. 9, Part 3, 1966, pp. 488-91, pi. 76.)

K. L. ALVIN: TWO CRISTATE MEGASPORES 489

Locality and horizon. Oxroad Bay, East Lothian, Scotland. Cementstone Group (Upper Tournaisian),
Calciferous Sandstone Series, Lower Carboniferous.

Holotype. British Museum (Natural History), Palaeontology Department, V52016a. The specific name
is derived from the ragged appearance of the crest.

Discussion. The new spore most closely resembles Triletes echinoides Chaloner (1954),
a species described originally from the Beaver Bend Limestone of the Mississippian of
Indiana, but one which has since been recorded from a number of localities including

TEXT-FIG. I . A-D, Triletcs paiwosus sp. nov. E, Triletes snbpalaeocristatiis Alvin. A, Specimen showing the
general features. The crest, part of which is shown in Plate 76, fig. 6, shows an especially high degree of
fusion among the hairs so as to form a more or less continuous membrane. Slide V52016e. b, c.
Spines, short hairs, and papillae in profile. Slide V52016f, g. d. Part of the spore wall in surface view
showing a single buttressed spine and papillae. Slide V52016e. e. Part of the wall in surface view.

Slide V51513Z.

Britain where it has been found in the sporophyll known as Lepidostrobopliylliini fini-
briatiim (Kidston) (Allen 1951; Lacey 1962). Since the publication of the original
description, the range in the equatorial diameter for the species has been extended to
L2-4'0mm., though the two original specimens were 2-66 mm. and 2-00 mm. The
wide range virtually covers that for the new Scottish spore. The spines of T. echinoides
are up to about 1 mm. long (Winslow 1959), whereas the maximum length in T.pannosits
is only less than half this. The chief difference between the two species lies in the crest:
in T. echinoides this has the form of a continuous fluted membrane with no indication
of hairs or teeth or even of perforations.

In the general character of the ornamentation of the distal face, the new spore
resembles a number of other Carboniferous spores, especially perhaps T. crassiacideatiis
(Zerndt). However, this spore differs in the quite different form of the apical prominence.

The trilete crest of T. pannosus, in resembling a series of anastomosing hairs arising
from the lips of the ridge, may be compared with that of T. subpalaeocristatus Alvin

490

PALAEONTOLOGY, VOLUME 9

(1965). Indeed, when the new spore was discovered it was thought that it might be
identical with this species which had been based on specimens from a block of petrified
plant material from the same locality. A new specimen of this spore has now been
macerated from the original block and the following emended diagnosis incorporates
the new information that this has yielded.

Triletes subpalaeocristatus Alvin 1965
Plate 76, figs. 7, 8; text-fig. 1e

Emended diagnosis. Megaspore nearly spherical ; mean diameter 1 ,450 /u, (range for four
spores 1,250-1,750 /x); height (excluding crest) only about 50 /x less than equatorial
diameter. Wall about 35-45 [j. thick. Arms of trilete ridge extending the whole radius
of the spore. Distal face and contact areas bearing scattered papillae and short hairs up
to about 50 IX long. No distinct arcuate ridge, but contact areas delimited distally by a
narrow zone of relatively dense papillae and hairs. Lips of trilete ridge bearing a crest
of branched anastomosing hairs (3-5-) 6-5 (-13-0) jx thick; maximum height of crest at
apex up to 405 ix, height diminishing to about half this at the ends of the trilete arms.

Discussion. This species, on the evidence of the new specimen, is rather less like T.
palaeocrislatus Chaloner than was originally believed. It differs in possessing an
ornamentation of short hairs and papillae scattered over the surface and in the de-
limitation of the contact areas by arcuate zones of denser ornamentation. The new
specimen, in being somewhat larger than the original specimens, extends the size range
so that this approaches more closely that for T. palaeocrislatus.

GENERAL DISCUSSION

The similarity in certain characters, most notably the presence of an apical crest of
hairs, between T. subpalaeocristatus and T. pannosus is remarkable, especially as these
are the only known megaspores from Oxroad Bay. Since the two species are based on
material preserved in a different manner, the question immediately arises as to whether
the difference between them might not be due either to the modes of preservation or
even to the techniques used in preparing clean specimens. T. pannosus (PI. 76, figs.
1-6; text-fig. 1a-d), coming as it does from the shale, is always compressed, but although
some specimens are split open, the preservation is excellent. These spores were obtained
by first breaking down shale samples in nitric acid and potassium chlorate, neutralizing
and picking out the spores, and then freeing them from adhering mineral particles in
hydrofluoric acid. The specimen of T. subpalaeocristatus obtained free of matrix (PI.
76, figs. 7, 8; text-fig. 1e) was etched from the block slowly in acetic acid and sub-
sequently freed from siliceous material in hydrofluoric acid and from pyrite in nitric acid.

EXPLANATION OF PLATE 76

Figs. 1-6. Triletes pannosus sp. nov. 1, Holotype; Slide V52016a, x35. 2, Slide V52016b, X35.
3, Spines and papillae in profile; Slide V52016c, X 80. 4-6, Portions of the crest of three different
specimens showing variation in form; Slides V52016c, d, e, x80.

Figs. 7, 8. Triletes subpalaeocristatus Alvin. 7, New specimen isolated from the nodule; Slide V51513z,
X 35. 8, Portion of the crest of the same spore, X 80.

Palaeontology, Vol. 9

PLATE 76

ALVIN, Lower Carboniferous megaspores

K. L. ALVIN: TWO CRISTATE MEGASPORES

491

It was carefully examined at every stage and the only structural damage that was
observed was that some hairs protruding through patches of pyrite on the surface came
away when this was dissolved at the final stage.

The most obvious difference between the two species is the presence in T. parmosus
of large buttressed spines and their apparently complete absence in T. subpalaeocristatus.
It has been noticed that the spines are frequently broken (PI. 76, fig. 3, middle) and also
that touching them with a needle often causes them to break away cleanly from the
surface. Abrasion of the spore prior to preservation could conceivably have removed
all the spines in the spores from the nodule. However, if abrasion sufficient to have re-
moved all the spines had occurred, the short delicate hairs scattered over the wall and
the hairs of the trilete crest would probably not have survived as they have. Moreover,
all the known specimens of T. subpalaeocristatus, including the new one, were lying in
close proximity to the cone of Oxroadia which probably bore them.

Disregarding the spiny character, there are other differences between these two kinds
of spores which seem quite sufficient to indicate that they are genuinely distinct species.
These are: (1) the greater length of the arms of the trilete ridge in T. subpalaeocristatus',
(2) the finer hairs of the crest and the lesser degree of fusion among them; (3) the
arcuate zones of papillae and hairs of which there is no indication in T. panuosus',
(4) the presence of some short hairs on the contact areas, particularly near the tips of
the ridge which again are entirely lacking in the species from the shale where in fact the
lips are very sharply delimited from the contact areas.

The coincidence at the same locality of these two spores sharing the very unusual
character of a trilete crest of anastomosing hairs is remarkable. The fact that one is
apparently confined to the shale bands and the other is known only from a nodule may
be of no significance. On the other hand, it may indicate an ecological difference between
the two plants, possibly related lycopods, that they represent. Nothing is yet known of
the composition of the microfossil flora of the shales, and of the nodule flora, only a
small number of macrofossil species have been described.

Acknowledgement. My thanks are due to Mr. A. Horne for photographic assistance.

REFERENCES

ALLEN, K. c. 1961. LepidostrobopbyUuni fimbriatinn (Kidston 1883) from the Drybrook Sandstone
(Lower Carboniferous). GeoL Mag. 98, 225-9.

ALVIN, K. L. 1965. A new fertile lycopod from the Lower Carboniferous of Scotland. Palaeontology,
8, 281-93.

CHALONER, w. G. 1954. Mississippian megaspores from Michigan and adjacent states. Contr. Mas.
Paleont. Univ. Mich. 12 (3), 23-35.

LACEY, w. s. 1962. Welsh Lower Carboniferous plants. I. The flora of the Lower Brown Limestone in
the Vale of Clwyd, North Wales. Palaeontographica, B 111, 125-60.

WINSLOW, M. R. 1959. Upper Mississippian and Pennsylvanian megaspores and other plant microfossils
from Illinois. Bull. III. geol. Surv. 86, 1-135.

K. L. ALVIN

Department of Botany,
Imperial College,
London, S.W.7

Manuscript received 13 August 1965

UPPER CRETACEOUS FORAMINIFERA FROM
THE BALLYDEENLEA CHALK, COUNTY KERRY,

IRELAND

by F. T. BARR

Abstract. Twenty-three species of Upper Cretaceous foraminifera have been recovered from the newly dis-
covered (Walsh I960) Ballydeenlea Chalk of County Kerry, Ireland. These are the first Upper Cretaceous
foraminifera reported from the Republic of Ireland and the westernmost Upper Cretaceous fauna found in
Europe. This faunal assemblage shows that the Ballydeenlea Chalk can be correlated with part of the Upper
Chalk (Senonian) of England and Northern Ireland. The presence of BoUvinoides decorata (Jones) in this fauna
indicates further that the Ballydeenlea Chalk is, at least in part. Upper Campanian in age and equivalent to the
Belenmitella mucronata zone of the British Isles and Western Europe.

In 1960, while systematically mapping the Upper Palaeozoic rocks of the Killarney
District, County Kerry, Dr. P. T. Walsh discovered a small pocket of hard white chalk
within the typical dark Palaeozoic shale and limestone of the area. This chalk was
exposed in a disused quarry in Ballydeenlea townland, six miles north of Killarney, and
had been quarried for perhaps over a century until 1930. The exposure may have been
known to geologists in the past, but until Walsh’s (1960) initial investigation, there is no
record in the literature that any geological significance had ever been attributed to this
exposure. Walsh collected material from this quarry and soon both megafossils and
microfossils suggestive of an Upper Cretaceous age were recognized. The megafossils
consisted chiefly of fragments of Inoceramus sp. and fragments of unidentifiable
echinoids. Foraminifera were observed in thin section and although specific identifica-
tions could not be made, generic determinations suggested that the Ballydeenlea Chalk
was Upper Cretaceous in age. This was the first recognition of Cretaceous rocks in the
Republic of Ireland.

The Ballydeenlea Chalk is a pocket-like mass surrounded by Palaeozoic rocks. The
minimum dimensions of this deposit are approximately 130 ft. by 300 ft. and the thick-
ness is about 100 ft. The Ballydeenlea Chalk shows no bedding and consists chiefly of a
hard white limestone. Flint nodules typical of the Upper Cretaceous Chalk of England
and Northern Ireland are common. Angular fragments of brecciated dark shale are also
abundant throughout most of the deposit. These fragments vary in size from a few
millimetres to pieces several feet across. This shale has so far been found to be unfossili-
ferous.

Walsh (1966) gives a detailed description of the Ballydeenlea Chalk and several other
deposits in the Killarney district which may possibly be related. He concludes that the
Ballydeenlea Chalk represents a karst collapse deposit. Prior to Upper Cretaceous times,
a large cavern formed in the Carboniferous limestone. During the late Cretaceous, an
arm of the sea transgressed over this area, probably from the west, depositing a marine
sequence of white chalk over the Killarney district. Soon afterwards, the roof of the
limestone cavern collapsed and the cavern was filled by the overlying Upper Cretaceous
Chalk along with fragments of dark Palaeozoic shales from the roof. The Ballydeenlea

[Palaeontology, Vol. 9, Part 3, 1966, pp. 492-510, pis. 77-79.]

F. T. BARR; UPPER CRETACEOUS EORAMINIEERA

493

Chalk was thus protected from later erosion, whereas the great bulk of Upper Cretaceous
sediments deposited over this part of Ireland was removed by Cenozoic erosion.

The preservation of this small deposit of Upper Cretaceous Chalk is indeed fortunate.

TEXT-FIG. 1. Map showing location of the Ballydeenlea Chalk in Co. Kerry, south-western Ireland.

Its presence has considerably changed some of the fundamental ideas about the palaeo-
geography of Ireland during the Upper Cretaceous. The purposes of this present study
are; (1) to date the Ballydeenlea Chalk more exactly, which should add to the precision
of palaeogeographic reconstruction; (2) to show that earlier suggestions (Walsh 1960,
p. 113) that the Ballydeenlea Chalk was equivalent to the Middle Chalk (Turonian) are
incorrect; and (3) to describe and document the first Upper Cretaceous foraminifera
found in the Republic of Ireland, and which also constitute the westernmost Upper
Cretaceous fauna in Europe.

494

PALAEONTOLOGY, VOLUME 9

Acknowledgements. I thank Dr. P. T. Walsh for supplying the material from the Ballydeenlea Chalk on
which this study is based, and discussing various problems related to this deposit; Miss Elizabeth R.
Hill, Mr. Wilf Austin, and Mr. John Smith for help with the manuscript; and my wife, Melza, for
preparing the illustrations of foraminifera.

Deposition of types. All specimens illustrated in this paper and additional material from the Ballydeen-
lea Chalk are deposited in the British Museum (Natural History).

PALAEONTOLOGY AND AGE OF THE BALLYDEENLEA CHALK

The Ballydeenlea Chalk is predominantly a hard white limestone which was found
difficult to disaggregate. First attempts to obtain foraminifera in a free form from this
well-indurated rock were unsuccessful. Consequently, foraminifera were first observed
only in thin section. Thin sections of the Ballydeenlea Chalk characteristically exhibit
fairly common foraminifera in addition to abundant small fossil fragments. On the basis
of these foraminifera, Dr. F. T. Banner and Dr. W. H. Blow (British Petroleum Co.)
expressed the opinion that this Irish chalk was equivalent to the Middle Chalk (Turon-
ian) (Walsh 1960, p. 113).

I subsequently experimented with various techniques to disaggregate this tough
limestone and obtain specimens of foraminifera in a free state, and met with partial
success using the following method;

1 . The limestone was crushed down to fine pebble size.

2. This material was then heated in a dilute solution of hydrogen peroxide for several
hours.

3. The residue was then washed through a 200-mesh sieve.

4. The remaining fraction was then placed in an ultrasonic vibrator for 30 to 45
minutes.

This is rough treatment for delicate foraminifera and no doubt many specimens were
destroyed or damaged. Nevertheless, by using large amounts of material and repeating
this process many times, a large, although generally poorly preserved, fauna was finally
obtained. In addition to the foraminifera recovered, molluscan and echinoid fragments
were common, siliceous sponge spicules were abundant, fish teeth were rare, and a single
ostracod was found. The following foraminifera are recorded from the Ballydeenlea
Chalk:

Ammodisciis cretacea (Reuss)

Glomospira cf. gordialis (Jones and Parker)

Textidaria sp.

Trochamminoides sp.

Haplophragmoides sp.

Ataxophragmium variablis (d’Orbigny)

Arenobiilimina cf. sphaerica Marie
Lagena acuticosta Reuss
Nodosaria limbata (?) d’Orbigny
Dentalina cf. communis (d’Orbigny)

Lenticidina sp.

Praebulimina obtiisa (d’Orbigny)

Bolivinitella eleyi (Cushman)

Bolivinoides decorata (Jones)

F. T. BARR: UPPER CRETACEOUS FORAMINIEERA

495

Heterohelix globulosa (Reuss)

Gyroidinoides iimbilicata (d’Orbigny)

Eponides cf. concinna Brotzen
Globigerinelloides aspera (Ehrenberg)

Riigoglobigerina sp.

Gaveliiiella lontiana (d’Orbigny)

Gavelinella thalmamu (Brotzen)

Gaveliiiella sp.

Cibicides beaumontiana (d’Orbigny)

It is not certain how well this list of foraminifera represents the complete fauna. By
analogy with faunas usually obtained from the Upper Cretaceous Chalk of southern
England, one would suspect that this list represents only a fraction of the complete
fauna. The methods used to disaggregate the limestone probably introduced a consider-
able bias in the fauna obtained.

Nevertheless, the foraminifera recovered from the Ballydeenlea Chalk can be used to
date this deposit precisely. It must be pointed out that it is possible that the Ballydeenlea
Chalk may consist of a mixture of sediments of several different ages. However, all the
species recovered can occur together within a faunal assemblage of a single age. Many
of the species range in the British Isles and Western Europe throughout most of the Sen-
onian (see text-fig. 2); other species have a stratigraphic range of Middle Senonian to
Maestrichtian. The bulk of the fauna, therefore, indicates that the Ballydeenlea Chalk
is a time equivalent to part of the Upper Chalk (Senonian) of England, not the Middle
Chalk (Turonian) as was previously suggested. Furthermore, about twelve specimens of
Boliviiioides decorata (Jones) were recovered. This species is restricted to the Belenmitella
mucronata Zone (Upper Campanian) of England, Northern Ireland, and Western Europe
in general (Barr 1966), and its presence therefore indicates that the Ballydeenlea Chalk,
at least in part, is Upper Campanian in age and a time equivalent to the B. mucronata
Zone of England and Northern Ireland.

SYSTEMATIC PALAEONTOLOGY

Family ammodiscidae Reuss 1862
Subfamily ammodiscinae Reuss 1862
Genus ammodiscus Reuss 1862

Ammodiscus cretacea Reuss
Plate 77, fig. 2

OpercuUna cretacea Reuss 1845, p. 35, pi. 13, figs. 64, 65.

Cormispira cretacea Reuss; Reuss 1860, p. 177, pi. 1, fig. 1.

Ammodiscus gaulticiis Berthelin; Brotzen 1936, p. 31, pi. 1, figs. 3 a, b.

Ammodiscus cretacea (Reuss); Marie 1941, p. 18, pi. 1, figs. 5-6.

Ammodiscus cretaceus (Reuss); Cushman 1946, pp. 17, 18, pi. 1, fig. 35 (see synonymy).
Ammodiscus cretaceus (Reuss); Hagn 1953, pp. 4, 5, fig. 3.

Involutiiia cretacea (Reuss); Belford 1960, pp. 22, 23, pi. 6, fig. 1.

Occurrence. Rare distorted specimens of A. cretacea were recovered from the Bally-
deenlea Chalk. This species, although usually rare, has a wide geographic distribution
and ranges throughout much of the Upper Cretaceous.

496

PALAEONTOLOGY, VOLUME 9

THXT-FiG. 2. Chart showing the stratigraphic ranges in the British Isles and Western Europe of some of
the Foraminifera recovered from the Ballydeenlea Chalk. Based chiefly on Barr (1961, 1962, 1966).

F. T. BARR: UPPER CRETACEOUS EORAMINIEERA

497

Genus glomospira Rzehak 1885
Glonwspira cf. gorclialis (Jones and Parker)

Trochammina squamata var. gordialis Jones and Parker 1 860, p. 304.

Trochanvnina squamata var. gordialis Jones and Parker; Parker and Jones 1865, p. 408, pi. 15,
fig. 32.

Glomospira gordialis Jones and Parker; Cushman 1946, pp. 18, 19, pi. 1, figs. 38-40 (see
synonymy).

Occurrence. A single specimen was recovered from the Ballydeenlea Chalk; it appears
to be identical to specimens illustrated by Cushman (1946, p. 18, pi. 1, figs. 38-40).
Cushman (op. cit., p. 19) recorded G. gordialis from the Austin Chalk, Taylor Marl,
and Velasco Shale of the Gulf Coastal Area of the U.S.A. and Mexico.

Family ataxophragmiidae Schwager 1877
Subfamily ataxophragmiinae Schwager 1877
Genus ataxophragmium Reuss

Ataxophragmium variahlis (d’Orbigny)

Bidimina variablis d'Orbigny 1840, p. 40, pi. 4, figs. 9-1 1.

Ataxogyroidina variablis (d’Orbigny); Barnard and Banner 1953, pp. 205, 206, pi. 9, figs. 6 a, b;
text-figs. 7 a-l (see synonymy).

Remarks. There is considerable variation in the general test shape, configuration of
apertural face, and in shape of aperture. Barnard and Banner (1953, text-figs. 1 a-I)
illustrated much of this variation.

Occurrence. Rare specimens of A. variablis were recovered from the Ballydeenlea Chalk.
This species is abundant in the Senonian Chalk of southern England, the Paris Basin,
and other parts of Western Europe.

Subfamily globotextulariinae Cushman 1927t7
Genus arenobulimina Cushman 1927<7

Arenobuliniina cf. sphaerica Marie
Arenobulimina sphaerica Marie 1941, p. 49, pi. 4, figs. 36a-e.

Remarks. Three specimens of a small Arenobulimina recovered from the Ballydeenlea
Chalk closely resemble A. sphaerica Marie in most respects. The poor preservation of
these specimens, however, does not allow a positive identification.

A. sphaerica was originally described from the B. mucronata Zone of the Paris Basin.

Family lituolidae de Blainville 1825
Subfamily haplophragmoidinae Maync 1952
Genus trochamminoides Cushman 1910

Trochamminoicles sp.

Plate 77, figs. I a, h

Description. Test planispiral, loosely coiled, evolute, biumbilicate, often distorted; peri-
phery in side view circular; 6-9 indistinct chambers in final whorl, uniformly and rapidly

498 PALAEONTOLOGY, VOLUME 9

increasing in size; sutures indistinct, weakly depressed, slightly curved, radial; wall
finely arenaceous.

Remarks. This species is similar to Trochamminoides velascoensis Cushman in some
respects, but can easily be distinguished by its larger size, greater and rapidly increased
thickness, and less compressed test.

Occurrence. This species is rare in the Ballydeenlea Chalk. Specimens of an unnamed
species of Trochamminoides have been observed in the Campanian of the Sacramento
Valley, California, and appear to be conspecific with this Irish form.

Family nodosariidae Ehrenberg 1838
Subfamily nodosariinae Ehrenberg 1838
Genus nodosaria Eamarck 1812

Nodosaria limbata (?) d’Orbigny

Nodosaria limbata d’Orbigny 1840, p. 12, pi. 1, fig. 1.

Nodosaria limbata d’Orbigny; Franke 1928, p. 42, pi. 3, figs. 27, 28.

Nodosaria limbata d’Orbigny; Cushman and Jarvis 1932, p. 32, pi. 10, fig. 5.

Nodosaria conciima Reuss; Cushman and Jarvis 1932 (non Reuss), pp. 31, 32, pi. 10, fig. 4.
Nodosaria limbata d’Orbigny; Cushman 1946, p. 74, pi. 27, figs. 1, 2 (see synonymy).

Dentalina concinna (Reuss); Hagn 1953 (non Reuss), pp. 43, 44, pi. 4, fig. 18.

Nodosaria limbata d’Orbigny; Said and Kenawy 1956, p. 133, pi. 2, fig. 32.

Dentalina catemda catemda Reuss; Pozaryska 1957 (non Reuss), pp. 76, 77, pi. 9, fig. 8.
Nodosaria limbata d’Orbigny; Belford 1960, pp. 38, 39, pi. 11, figs. 7-9.

Occurrence. Rare broken specimens of Nodosaria were recovered from the Ballydeenlea
Chalk. These fragments appear to be identical to the specimens of N. limbata in the
d’Orbigny Collection in Paris; however, because of the incompleteness of the only
specimens recovered, there is still some uncertainty as to this identification. I have also
found specimens of N. limbata in the B. mucronata Zone (Upper Campanian) of southern
England and in the B. lanceolata Zone (Lower Maestrichtian) at Sidestrand, Norfolk.
D’Orbigny (1840, p. 12) originally described this species from the B. mucronata Zone at
Meudon in the Paris Basin.

Genus lagena ’Walker and Jacob 1798
Lagena acuticosta Reuss
Plate 77, figs. 7, 8

Lagena acuticosta Reuss 1862, p. 305, pi. 1, fig. 4.

Lagena acuticosta Reuss; Egger 1899, p. 106, pi. 5, fig. 12.

EXPLANATION OF PLATE 77
All specimens are from the Ballydeenlea Chalk, X 125.

Fig. 1. Trochamminoides sp. la, side view; \b, end view; P45707.

Fig. 2. Ammodiscus cretacea Reuss, side view; P45710.

Fig. 3. Praebidimina obtusa (d’Orbigny); P45702.

Figs. 4, 5. Bolivinoides decorata (Jones). 4, side view, final chamber broken; P45693. 5a, side view;
5b, end view; P45692.

Fig. 6. Bolivinitella eleyi (Cushman). 6a, side view; 66, edge view; P45698.

Figs. 7, 8. Lagena acuticosta Reuss. 7, P45699. 8, P45700.

Palaeontology, Vol. 9

PLATE 77

BARR, Foraminifera from Ballydeenlea Chalk

F. T. BARR: UPPER CRETACEOUS FORAMINIFERA

499

Lagena Isabella (d’Orbigny); Franke 1925 (non d’Orbigny), p. 60, pi. 4, fig. 40.

Lagena Isabella (d’Orbigny); Franke 1928, p. 87, pi. 8, fig. 1.

Lagena acuticosta Reuss; Cushman 1931o, p. 308, pi. 35, fig. 12.

Lagena acuticosta Reuss; Cushman 1932, p. 337, pi. 50, fig. 13.

Lagena Isabella (d’Orbigny) ; Brotzen 1 936 {non d’Orbigny), pp. 1 1 1 , 1 1 2, pi. 7, fig. 5 ; text-fig. 37.
Lagena cayeiixi Marie 1941, pp. 74, 75, pi. 9, figs. 82, 83.

Lagena amphora Reuss; Schijfsma 1946, p. 54, pi. 2, fig. 17.

Lagena acuticosta Reuss; Cushman 1946, p. 94, pi. 39, figs. 14, 15.

Lagena acuticosta Reuss; Visser 1951, pp. 234, 235, pi. 2, fig. 1.

Lagena acuticosta Reuss; Frizzell 1954, p. 102, pi. 14, figs. 1, 2.

Lagena cf. cayeuxi Marie; McGugan 1957, p. 339, pi. 31, fig. 22.

Lagena acuticosta Reuss; Martin 1964, p. 61, pi. 5, figs, la, b.

Occurrence. Six specimens of L. acuticosta Reuss were found in the Ballydeenlea Chalk.
This species has a long stratigraphical range and great geographical distribution; it
ranges from the M. cor-testudinarium Zone (Coniacian) to the top of the B. mucronata
Zone (Upper Campanian) on the Isle of Wight. I have also found this species in the
Lower Maestrichtian of Norfolk, the Lower Maestrichtian of southern Limberg, and
the Upper Maestrichtian of Stevns Klint, Denmark. Cushman (1946, p. 94) reported
that L. acuticosta had a wide distribution in the Gulf Coastal area of the U.S.A.,
occurring in strata of Austin to Navarro age.

Family turrilinidae Cushman 1927u
Subfamily turrilininae Cushman 1927a
Genus praebulimina Hofker 1953

Praebulimina obtusa (d’Orbigny)

Plate 77, fig. 3

Bulimina obtusa d’Orbigny 1840, p. 39, pi. 4, figs. 5, 6.

Bulimina laevis Beissel 1891 (part), p. 66, pi. 12, figs. 42, 43 (non figs. 39-41).

Bulimina elegans d’Orbigny; Chapman 1892 (non d’Orbigny), p. 516 (list), pi. 15, fig. 9.
Bulimina elegans d’Orbigny; Heron-Alien and Earland 1910, p. 409, pi. 6, fig. 11.

Bulimina ovulum Reuss; Franke 1925, p. 25, pi. 2, fig. 17.

Bulimina ovulum Reuss; Franke 1928, p. 157, pi. 14, fig. 14.

Buliminella carseyae Plummer; Dain 1934 (non Plummer), p. 37, pi. 4, fig. 38.

BuUminella laevis (Beissel); Cushman and Parker 1936, p. 6, pi. 2, figs. la-c.

Buliminella obtusa (d’Orbigny); Brotzen 1936, p. 131, pi. 8, figs, la, b.

Buliminella obtusa (d’Orbigny) forma typica; Marie 1941, p. 198, pi. 30, figs. 291 a~f.
Buliminella obtusa (d’Orbigny) inflata Marie 1941, p. 199, pi. 30, figs, \91a-e.

Buliminella obtusa (d’Orbigny) laevis (Beissel); Marie 1941, p. 199, pi. 30, figs. 293 a-e; pi. 31,
figs. 19Aa-e.

Buliminella obtusa (d’Orbigny); Schijfsma 1946, p. 80, pi. 4, fig. 9.

Description. Test free, tapering, approximately U times longer than wide, consisting of
4 to 5 whorls each containing 3 to 4 weakly inflated chambers rapidly and uniformly
increasing in size; spiral sutures depressed; chamber sutures curved, flush or weakly
depressed; aperture slit-like or loop-shaped opening along inner margin of final cham-
ber, often also extending up into chamber face bordered by thin narrow lip; tooth plate
simple, narrow, attached to inner side of aperture and extending into penultimate
aperture; wall calcareous, smooth, finely perforate.

Remarks. P. obtusa occurs commonly throughout the Senonian chalk of southern
England. In the middle and upper parts of the B. mucronata Zone, it becomes

C 4179 L 1

500

PALAEONTOLOGY, VOLUME 9

progressively larger, more inflated and more ovate, and eventually evolves into a dis-
tinctively larger form which is indicative of the latest Campanian and Maestrichtian
in Western Europe.

Occurrence. D’Orbigny (1840, pp. 39, 40) originally described this species from the
Upper Cretaceous localities at Meudon and Saint-Germain in the Paris Basin. Marie
(1941, pp. 197-9) reported three varieties of this species from the B. mucronata Zone of
the Paris Basin, and Schijfsma (1946, p. 80) records P. obtusa from the Campanian of
southern Limberg, Netherlands. F. obtusa is common in the Ballydeenlea Chalk.

Family bolivinitidae Cushman 1927a
Genus bolivinoides Cushman 1927a

Bolivinoides decorata (Jones)

Plate 77, figs. 4, 5

Bolivinoides decorata (Jones); Barr 1966, pp. 220-43, pi. 34, figs. 2-6, 12; pi. 35, figs. 6-9;
pi. 36, figs. 1-5. Full synonymy.

Remarks. B. decorata (Jones), a species common to the B. mucronata Zone of southern
England and Northern Ireland, has recently been described in detail (Barr 1966). A
lectotype was established from a series of syntypic specimens in the Joseph Wright
Collection, which is located in the Queen’s University, Belfast (op. cit. pp. 231-4,
pi. 36, figs. 1 a, b). B. decorata was shown to have exceptional value as an index fossil of
the Upper Campanian.

The direct ancestor of B. decorata appears to be B. hiltermanni Barr. The latter gave
rise to B. decorata at the beginning of the Upper Campanian (base of the B. mucronata
Zone). The A. quadratus Zone-5, mucronata Zone (Lower Campanian-Upper Camp-
anian) contact in southern England can be located with some accuracy at the strati-
graphic position where B. hiltermanni is replaced by B. decorata (s.s.). B. decorata differs
from its ancestor, B. hiltermanni, by having a larger, more flaring test with a small
length/breadth ratio (usually 1-5 to 1-8 compared with 1-9 to 2-3 which is most usual for
B. hiltermanni). The surface sculpture is more distinct on B. decorata and there are more
lobes on the final chambers, most often 4 or 5, whereas B. hiltermanni usually possesses
3 on its latest chambers.

Bolivinoides miliaris Hiltermann and Koch is similar to B. decorata (Jones) in many
respects and appears to have been derived from B. decorata during the late B. mucronata
Zone. B. decorata is distinguished from B. miliaris by having a slightly greater length/
breadth ratio and a more elongate, less rhomboidal lateral outline. The maximum
breadth of B. decorata is closer to the apertural end, whereas the maximum breadth of
B. miliaris is nearer the mid-point of the test. The surface lobes of B. miliaris are less
regular than those of B. decorata and there is more of a tendency for them to fuse with
the lobes on the earlier chambers.

Edgell (1954, pp. 71, 72) described a new subspecies which he named B. decorata
(Jones) australis. This subspecies is a form transitional between B. decorata (Jones) and
B. gigantea Hiltermann and Koch. Barr (1966) re-examined type specimens of B. decorata
australis and concluded that this form was closer morphologically to B. gigantea and
was quite distinct from B. decorata (s.s.). B. australis appears to have evolved from

F. T. BARR: UPPER CRETACEOUS FORAMINIFERA

501

B. decorata (s.s.) at the beginning of the Maestrichtian. It can be distinguished from
B. decorata by ; (1) having a smaller length/breadth ratio ; (2) having more surface lobes on
the final chambers (5 to 7 on the final chamber compared with 4 or 5 for B. decorata s.s.) ;
(3) possibly being more commonly ornamented near its initial end; and (4) sometimes
having a more rhomboidal outline.

Occurrence. Approximately twelve specimens of Bolivinoides decorata were recovered
from the Ballydeenlea Chalk. However, most of these specimens were broken or dam-
aged to varying degrees.

B. decorata occurs in abundance throughout the B. mucronata Zone in southern
England and appears to be restricted to this zone (Barr 1966). The type locality for this
species is at Keady Hill, County Derry, Northern Ireland, which is probably in the
upper part of the B. mucronata Zone.

B. decorata also occurs in the Ballycastle Pellet Chalk of Northern Ireland (McGugan
1957, p. 339, pi. 32, figs. 10-15; Barr 1966). However, it appears that this formation was
derived by Tertiary erosion of an Upper Cretaceous land surface (Charlesworth 1963,
p. 369) and that the contained foraminifera were also derived from the nearby Upper
Cretaceous Chalk.

Genus bolivinitella Marie 1941

In 1854, Ehrenberg proposed the generic name Loxostomum, but unfortunately
neither gave a description of this new genus nor designated a type species. However, he
listed seven new Upper Cretaceous species which he considered to belong to this genus :
Loxostomum aculeatum, L. anglicum, L. curvatum, L. rostratum, L. subrostratum,
L. tumens, and L. vorax. None of these species was described and the only illustrations
were of specimens mounted in balsam and viewed by transmitted light. Consequently,
there have been few attempts to use Ehrenberg’s species of Loxostomum and most of
them must be considered nomina dubia.

Cushman (19276) subsequently fixed the type species of the genus Loxostomum as
L. subrostratum Ehrenberg. In various editions of Cushman’s classification (e.g. 1955,
p. 269, pi. 27, figs. 30-32), he described Loxostomum as follows : ‘Test in early stages simi-
lar to Bolivina, adult tending to become uniserial ; aperture terminal. — Cretaceous to
Recent.’ Cushman included Bolivina plaita Carsey in the genus Loxostomum. Subsequent
authors have considered this species typical of the genus and have based their concept
of Loxostomum on this and similar forms.

Marie (1941, pp. 189, 190) erected the genus Bolivinitella and designated the well-
known species Bolivinita eleyi Cushman as the type species. Marie (op. cit., p. 191, pi.
29, figs. 282 a-c) illustrated a typical specimen of B. eleyi from the B. mucronata Zone
(Upper Campanian) at Montereau, and listed the occurrence of this species from
numerous other localities in the Paris Basin in the B. mucronata Zone, including
Meudon (one of the two localities from which Ehrenberg (1854) originally recovered
his specimens of Loxostomum subrostratum).

Eollowing suggestions made by Hofker (1952), Loeblich and Tappan (1962, pp. 110,
111) placed Bolivinitella Marie and Loxostomum Ehrenberg in synonymy. They sup-
pressed Bolivinitella as a junior synonym of Loxostomum and erected a new genus,
Corypliostoma, in which they placed many species previously referred to Loxostomum
(e.g. Bolivina plaita Carsey). Loeblich and Tappan pointed out that Marie (1941)

502

PALAEONTOLOGY, VOLUME 9

recorded Bolivinitella forma typica from Meudon, a locality from which Ehrenberg
originally described Loxostomum subrostralum. They concluded: ‘the specimens illus-
trated as B. eleyi by Marie (1941, pi. 29, figs. 282 a-c) are typical of L. subrostratum and
the two “species” are not only congeneric, but almost certainly conspecific.’

At best the suppression of BoJivimtella as a junior synonym of Loxostomum is
premature. Ehrenberg never described L. subrostratum and the only illustration is a
drawing of a specimen viewed by transmitted light. Consequently, L. subrostratum is
not well enough known to be recognized with any assurance. Furthermore, I have
collected material from the Upper Cretaceous Chalk at Meudon with M. Pierre Marie
which indeed contains numerous specimens of Bolivinitella eleyi \ however, there are a
number of other biserial forms present (e.g. Bolivina plaita) which might conceivably
belong to the true Loxostomum subrostratum. Therefore, at least until Ehrenberg’s
original specimens can be re-examined, it would seem best to continue the usage of
Bolivinitella as a valid genus.

Bolivinitella eleyi (Cushman)

Plate 77, figs. €>a, b

Textularia obsoleta Reuss; Eley 1859 (non Reuss), p. 202, pi. 8, fig. 11c.

Bolivina obsoleta Eley; Jones 1872, p. 124 (list).

Bolivinita eleyi Cushman 1927a, p. 91, pi. 12, figs. \ \a, b.

Bolivinita eleyi Cushman; Cushman 1931Z>, p. 39, pi. 5, figs. 8a, b.

Bolivinita eleyi Cushman; Cushman 1932, p. 338, pi. 51, figs, la, b.

Bolivina quaclrilatera (Schwager); Macfadyen 1932, pp. 492, 493, pi. 35, figs. 21 a, b.

Bolivinita quadrilatera (Schwager); Dain 1934, pp. 34, 35, pi. 3, fig. 35.

Bolivinita eleyi Cushman; Brotzen 1936, p. 122, pi. 9, figs. 5a, b, text-fig. 41.

Bolivinitella eleyi (Cushman); Marie 1941, p. 190, pi. 29, figs. 282 a, b.

Bolivinita eleyi Cushman; Cushman 1946, p. 114, pi. 48, figs. 18-20.

Bolivinitella eleyi Cushman; Schijfsma 1946, pp. 72, 73, pi. 6, fig. 10.

Bolivinita eleyi Cushman; Hagn 1953, p. 76, pi. 6, fig. 24.

Bolivinitella eleyi (Cushman); Erizzell 1954, p. 112, pi. 16, figs. 23a, b.

Bolivinitella eleyi (Cushman); McGugan 1957, p. 340.

Bolivinitella eleyi (Cushman); Montanaro Gallitelli 1957, p. 150, pi. 34, figs. 14-17.
Bolivinitella eleyi (Cushman); Belford 1960, p. 62, pi. 15, figs. 20, 21.

Bolivinitella eleyi (Cushman); Akimets 1961, pp. 192, 193, pi. 19, figs. 9a, b.

Bolivinita eleyi Cushman; Kaptarenko-Chernousova et al. 1963, p. 113, pi. 19, figs, la, b.
Bolivinitella eleyi (Cushman); Graham and Church 1963, p. 51, pi. 5, figs. 25a, b.

Description. Test elongate, compressed, sometimes twisted, rectangular in cross-section,
with up to 16 chambers arranged biserially, almost becoming uniserial in final stage;
sides nearly parallel or gradually tapering; lateral sides flat or slightly concave, chambers
reniform, arched, overlapping; sutures distinct, limbate, flush or very slightly raised,
fusing at edges, forming a thin longitudinal keel along each edge ; wall calcareous, very
finely perforate, surface smooth; aperture a thin terminal slit; approximate length
0-56 mm.; breadth 0-20 mm.; thickness 0-09 mm.

Occurrence. B. eleyi is fairly common in the Ballydeenlea Chalk. It is common in
southern England, where it ranges from the base of the M. cor-anguinum Zone to the
top of the B. mucronata Zone. B. eleyi also occurs in lowermost Maestrichtian strata

F. T. BARR: UPPER CRETACEOUS FORAMINIFERA 503

of Western Europe and North Africa. The holotype is from the upper Brownstone
Marl of Arkansas.

Family heterohelicidae Cushman 1927a
Subfamily heterohelicinae Cushman 1927a
Genus heterohelix Ehrenberg 1843

Heterohelix globulosa (Ehrenberg)

Plate 78, figs. 5, 6

Textularia globulosa Ehrenburg 1840, p. 135, pi. 4, figs. 2, 4, 5, 7, 8.

Textilaria globulosa Ehrenburg; Ehrenburg 1854, pi. 21, fig. 87.

Textularia globulosa Ehrenberg; Eley 1859, pp. 194, 202, pi. 2, fig. 9; pi. 9, fig. 9.

Textilaria globifera Reuss 1860, p. 232, pi. 13, figs. 7, 8.

Textularia decurrens Chapman 1892, p. 515, pi. 15, fig. 6.

Giimbelina globulosa (Ehrenberg); Egger 1899, p. 32, pi. 14, fig. 43.

Textularia globifera Reuss; Franke 1925, p. 11, pi. 1, fig. 13.

Textularia globulosa Ehrenberg; Franke 1928, p. 134, pi. 12, fig. 11.

Pseudotextularia globulosa (Ehrenberg); Macfadyen 1932, pi. 35, fig. 22.

Giimbelina globulosa (Ehrenberg); Morrow 1934, p. 194, pi. 29, figs. 18 a, b.

Giimbelina globifera (Reuss); Eoetterle 1937, p. 34, pi. 5, fig. 3.

Giimbelina globulosa (Ehrenberg); Cushman 1946, pp. 105, 106, pi. 45, figs. 9-15 (see synonymy).
Giimbelina globulosa (Ehrenberg); Williams-Mitchell 1948, p. 99, pi. 9, fig. 2.

Giimbelina globulosa (Ehrenberg); Eoeblich 1951, p. 108, pi. 12, figs. 4, 5.

Giimbelina globulosa (Ehrenberg); Hagn 1953, p. 73, pi. 6, figs. 16, 17.

Giimbelina globulosa (Ehrenberg); Frizzell 1954, p. 109, pi. 15, figs. 24-27.

Heterohelix globulosa (Ehrenberg); Montanaro Gallitelli 1957, pi. 31, figs. 12-15.

Giimbelina globulosa (Ehrenberg); Belford 1960, p. 59, pi. 15, figs. 10, 11.

Heterohelix globulosa (Ehrenberg); Said and Kerdany 1961, p. 331, pi. 2, fig. 1.

Heterohelix globulosa (Ehrenberg); Graham and Church 1963, pp. 61, 62, pi. 7, figs. \\a,b.
Heterohelix globulosa (Ehrenberg); Barr and Cordey 1964, pp. 306, 307, pi. 49, fig. 4.
Heterohelix globulosa (Ehrenberg); Takayanagi 1965, pp. 195, 196, pi. 20, figs. \a, b (see
synonymy).

Remarks. H. globulosa is distinguished by its tapering test, lobulate periphery, and
smooth globular chambers. Specimens of H. globulosa are fairly common in the Bally-
deenlea Chalk, but most are somewhat smaller than typical specimens of this species.

Occurrence. H. globulosa occurs in abundance throughout the Senonian of southern
England. Williams-Mitchell (1948, pi. 10) recorded the range of this species in the Chalk
of England as I. labiatus Zone (Lower Turonian) to the B. mucronata Zone. Textularia
decurrens Chapman (= H. globulosa) was originally described from the phosphatic
chalk at Taplow, Buckinghamshire, which is referred to the M. testudinarius Zone
(Santonian) (Barr and Cordey 1964, pp. 306, 307).

Family planomalinidae Bolli, Eoeblich, and Tappan 1957
Genus globigerinelloides Cushman and ten Dam 1948

Globigerinelloides aspera (Ehrenberg)

Plate 78, figs. 4 a, b

Rotalia aspera Ehrenberg 1854, p. 24, pi. 27, figs. 57, 58; pi. 28, fig. 42; pi. 31, fig. 44.
Phanerostomum asperum Ehrenberg 1854, p. 23, pi. 30, figs. 26 a-b.

504

PALAEONTOLOGY, VOLUME 9

Rotalia aspera Ehrenberg; Beissel 1891, p. 73, pi. 14, figs. 1-6.

Globigerina aequilateralis {no?i Brady); Chapman 1892, p. 517, pi. 15, fig. 14.

Globigerina aequilateralis (non Brady); Heron-Alien and Earland 1910, p. 424, pi. 8, figs.

11, 12.

Globigerinella aspera (Ehrenberg); Brotzen 1936, p. 170, pi. 13, fig. 2.

Globigerinella aspera (Ehrenberg); Schijfsma 1946, pp. 94-96, pi. 6, fig. 8.

Globigerinella aspera (Ehrenberg); Bandy 1951, p. 508, pi. 75, fig. 3.

Globigerinella aspera (Ehrenberg); Belford 1960, p. 91, pi. 25, figs. 4-6.

Planomalina aspera (Ehrenberg); Barr 1962, pp. 561, 563, pi. 69, figs. A a, b.

Planomalina (Globigerinelloides) aspera aspera (Ehrenberg); van Hinte 1963, p. 97, pi. 12,
figs, la, 3.

"Globigerinella' aspera (Ehrenberg); Graham and Church 1963, pp. 64, 65, pi. 7, figs. \1 a-c.
Globigerinelloides aspera (Ehrenberg); Barr and Cordey 1964, p. 309.

Globigerinelloides asper (Ehrenberg); Takayanagi 1965, pp. 201, 202, pi. 20, figs. 9 a-c.

Occurrence. Approximately twelve specimens of G. aspera were recovered from the
Ballydeenlea Chalk, and additional specimens were observed in thin section. G. aspera
appears to be the most common planktonic species found in this chalk; it has a wide
geographic distribution, occurring in most areas of the world where Senonian planktonic
faunas are found. Barr (1962) recorded G. aspera in abundance from the Isle of Wight,
ranging from the lower M. cor-anguinum Zone to the upper B. mucronata Zone. It is
also found in the lower and middle Maestrichtian of continental Europe. Barr and
Berggren (1965) recorded specimens of G. aspera from the Palaeocene Thanet Formation
of eastern Kent. These specimens, along with other Upper Cretaceous species found in
the Thanet Formation, were undoubtedly derived from the nearby Campanian Chalk.

Family globotruncanidae Brotzen 1942
Genus rugoglobigerina Bronnimann 1952

Rugoglobigerina sp.

Plate 78, figs. 2, 3

Remarks. Several small specimens of Rugoglobigerina were recovered from the Bally-
deenlea Chalk. The chambers of these forms were coarsely hispid and arranged in
a very low trochospiral. These specimens could not be specifically identified, but are
generally similar to several species that occur in the Campanian chalk of western
Europe.

EXPLANATION OF PLATE 78
All specimens are from the Ballydeenlea Chalk, X 175.

Fig. 1. Gavelinella thalmanni (Brotzen). la, dorsal view, juvenile specimen; \b, end view; Ic, ventral
view; P45703.

Figs. 2, 3. Rugoglobigerina sp. 2a, dorsal view; lb, end view; 2c, ventral view; P45709. 3a, dorsal view;

lb, ventral view, showing final chamber broken; P45708.

Fig. 4. Globigerinelloides aspera (Ehrenberg). 4a, lateral view; 4b, end view; P45701.

Figs. 5, 6. Heterohelix globulosa (Ehrenberg). 5, lateral view, initial end missing; P45704. 6, lateral
view of exceptionally small specimen; P45705.

Palaeontology, Vol. 9

PLATE 78

BARR, Foraminifera from Ballydeenlea Chalk

V

F. T. BARR; UPPER CRETACEOUS FORAMINIFERA

505

Family eponididae Hofker 1951
Genus eponides de Montfort 1808

Eponides cf. concinna Brotzen
Plate 79, figs. 2a-c

Epionides concinna Brotzen 1936, p. 167, pi. 12, figs. Aa-c.

Description. Test trochoid, consisting of about 3 whorls, biconvex; dorsal side weakly
convex, nearly evolute, early whorls somewhat obscured by thin smooth mass of hyaline
calcite; ventral side convex, involute, umbilicate; periphery in side view circular, weakly
lobulate, with slight keel; chambers weakly inflated, ventrally overlapping, 9 to 12 in
final whorl, uniformly and gradually increasing in size; ventral sutures distinct, de-
pressed, dorsal sutures indistinct on early whorls, distinct on final whorl, limbate, flush,
slightly curved, oblique; aperture narrow slit at base of final chamber, extending from
umbilicus to periphery, bordered by thin, narrow lip; umbilicus small, shallow, often
partially filled by low umbilical plug; wall calcareous, finely perforate, surface smooth.

Remarks. The morphology of the umbilical area of specimens referred to Eponides cf.
concinna varies considerably. The umbilicus varies from a shallow simple form through
ones containing a slightly raised plug to umbilici almost completely filled by a well-
developed umbilical plug. The original illustration of the holotype of E. concinna shows
a small shallow umbilicus with no umbilical plug. Although the specimens in the present
study are similar in most respects to the type specimen from Sweden, there is still some
uncertainty as to their relationship. On the other hand, the Ballydeenlea specimens
appear to be identical to specimens from the Upper Cretaceous Chalk of southern
England which range from the base of the M. cor-anguinum Zone to the upper part of
the B. mucronata Zone.

Family osangulariidae Loeblich and Tappan 1964
Genus gyroidinoides Brotzen 1942

Gyroidinoides wnbilicata (d’Orbigny)

Plate 79, figs. 3 a-c

Rotalina umbilicata d’Orbigny 1840, pi. 3, figs. 4-6.

Rotalina nitida Reuss; Beissel 1891, p. 71, pi. 14, figs. 14-16.

Rotalia soldanii umbilicata (d’Orbigny); Franke 1927, p. 692.

Gyroidina umbilicata (d’Orbigny); Cushman 19316, p. 43, pi. 6, fig. 3.

Gyroidina umbilicata (d’Orbigny); Macfadyen 1932, p. 489 (list), pi. 35, figs. 21 a, b.

Gyroidina nitida (Reuss); Brotzen 1936, pp. 157-9, pi. 11, figs. 3n-c; text-fig. 58.

Gyroidina umbilicata (d’Orbigny); Marie 1941, pp. 219, 220, pi. 34, figs. 318a-c.

Gyroidina nitida (Reuss); Schijfsma 1946, pp. 85, 86, pi. 5, figs. 1 a~c.

Gyroidina nitida (Reuss); Williams-Mitchell 1948, p. 97, pi. 8, figs. 6 a-c.

Gyroidina umbilicata (d’Orbigny); McGugan 1957, p. 342, pi. 33, fig. 4.

Description. Test trochoid, planoconvex, composed of 3 to 3| whorls; dorsal side flat,
evolute, ventral side strongly convex, involute, umbilieate, periphery in side view broadly
rounded; chambers weakly inflated, 7 to 8 chambers in final whorl, uniformly and
gradually increasing in size; ventral sutures straight, radial, flush; early dorsal sutures on
final whorl depressed, slightly curved, nearly radial, dorsal spiral suture in early whorls

506

PALAEONTOLOGY, VOLUME 9

flush, indistinct, in final whorl distinct, depressed; aperture narrow slit at base of final
chamber, extending from ventral umbilicus on to dorsal side, bordered by thin narrow
lip ; umbilicus narrow, shallow, sometimes partially covered by apertural lip ; wall cal-
careous, finely perforate, surface smooth.

Occurrence. G. umbilieata is common in the Ballydeenlea Chalk. This species has a long
stratigraphic range in Western Europe. In southern England, G. umbilieata ranges from
the Upper Turonian to the top of the Campanian. It is also abundant in the Lower
Maestrichtian chalk at Sidestrand, Norfolk.

Family anomalinidae Cushman 1927a
Subfamily anomalininae Cushman 1927a
Genus gavelinella Brotzen 1942

Gavelinella lorneiana (d’Orbigny)

Plate 79, figs. 1 a-c

Rosalina lorneiana d’Orbigny 1840, p. 36, pi. 3, figs. 20-22.

Anonialina (Rosalina) lorneiana (d’Orbigny); Egger 1909, p. 45, pi. 4, figs. 10-12.

Anomalina clementiana (d’Orbigny); Cushman 19316, p. 46, pi. 6, figs. 10 a-c.

Discorbis lorneiana (d’Orbigny); Marie 1941, pp. 214-16, pi. 33, figs. 314a-c.

Anomalina clementiana (d’Orbigny); Cushman 1946 (part), p. 155, pi. 63, figs. 13 a-c (non
figs. 12 a-c).

Gavelinella lorneiana (d’Orbigny); Loeblich and Tappan 1964, p. 759, figs. 621 (6a-c).

Occurrence. Rare specimens of G. lorneiana were recovered from the Ballydeenlea
Chalk. I have found this species in abundance at Culver Clilf, Isle of Wight, where it
ranges from the M. cor-testudinarium Zone (Coniacian) to the lower part of the B.
nmeronata Zone. D’Orbigny (1840, p. 37) originally described this speeies from the Upper
Campanian localities of St. Germain and Meudon in the Paris Basin, where he recorded
it in abundance, and from Sens (Lower Campanian) and England, where he found only
rare specimens. Marie (1941, p. 215) also reported G. lorneiana from various localities
in the B. nmeronata Zone (Upper Campanian) of the Paris Basin.

Gavelinella thalmanni (Brotzen)

Plate 78, figs. 1 a-c

Cibicides thalmanni Brotzen 1936, pp. 190, 191, pi. 14, figs. 1 a-c.

Description. Test flatly trochoid, nearly planispiral, consisting of about 2| tightly coiled
whorls; dorsal side weakly convex, semi-evolute, umbilieate; ventral side flat, involute.

EXPLANATION OF PLATE 79
All specimens are from the Ballydeenlea Chalk.

Fig. 1. Gavelinella lorneiana (d’Orbigny), X95. la, dorsal view; 16, end view; Ic, ventral view,
umbilicus obscured by matrix; P45694.

Fig. 2. Eponides cf. concinna Brotzen, X 125. 2a, dorsal view; lb, end view; 2c, ventral view; P45697.
Fig. 3. Gyroidinoides umbilieata (d’Orbigny), X 125. 3a, dorsal view; 36, end view; 3c, ventral view;
P45695.

Fig. 4. Cibicides beaumontiana (d’Orbigny), X 125. 4a, ventral view; 46, end view; 4c, dorsal view,
final chamber broken, umbilical area obscured by matrix; P45696

Palaeontology, Vol. 9

PLATE 79

BARR, Foraminifera from Ballydeenlea Chalk

r

F. T. BARR: UPPER CRETACEOUS EORAMINIEERA

507

weakly umbilicate; periphery in side view circular, in end view rounded; chambers
uninflated, 11-12 in final whorl, uniformly and gradually increasing in size; dorsal
sutures indistinct, weakly depresssed, gently curved, nearly radial, bordered by low, broad
septal ridges which become less raised toward periphery; ventral sutures weakly de-
pressed, usually obscured by low, very broad, gently curved, nearly radial septal ridges,
composed of opaque, white, shell material which merges into thickened carinal band ;
septal ridges most prominent in early chambers of final whorl becoming less raised in
final chambers ; primary aperture low arch at base of final chamber with slit extending to
ventral umbilicus, bordered by thin narrow lip, umbilical portion of apertures of suc-
cessive chambers appear to remain open as relict apertures; dorsal umbilicus broad,
fairly deep; ventral umbilicus narrow, shallow; wall calcareous, coarsely perforate.

Remarks. Specimens of G. thahnawn from the Ballydeenlea Chalk are very similar to
Brotzen’s (1936, pi. 14, figs. 1 a-c) figures of the holotype, although his illustrations do
not show the early whorls dorsally, which are usually apparent on the Irish specimens.
G. thahnanni is easily distinguished from G. pertusa (Marsson), G. costata Brotzen,
G. bullata Brotzen, and most other species of Gavelinella by having a dorsal umbilicus.

Occurrence. G. thahnanni is common in the Ballydeenlea Chalk. I have also found this
species in the Chalk of southern England, ranging from the M. cor-testudinarium Zone
to the upper part of the B. nmcronata Zone. The holotype of G. thahnanni is from the
Senonian strata of southern Sweden.

Family cibicididae Cushman 1927n
Subfamily cibicidinae Cushman 1927a

Genus cibicides de Montfort 1808

Cibicides beaumontiana (d’Orbigny)

Plate 79, figs. 4 a-c

Tnnicatidina beaumontiana d’Orbigny 1840, p. 35, pi. 3, figs. 17-19.

Truncatulina lobatula (d’Orbigny); Marsson 1878, p. 167, pi. 5, figs. 38 o-/.

Tnincatiilina convexa Reuss; Egger 1899, p. 149, pi. 18, figs. 25-27.

Truncatulina lobatula (Walker and Jacob); Franke 1925 {non Walker and Jacob), p. 83, pi. 7,
fig. 7.

(?) Cibicides excavata Brotzen 1936, p. 189, pi. 13, figs. 7, 8.

Cibicides ribbingi Brotzen 1936, pp. 186-9, pi. 13, figs. 5, 6, text-figs. 67, 68.

Cibicides beaumontiana (d’Orbigny); Marie 1941, pp. 249, 250, pi. 37, figs. 352-4.

Cibicides excavata Brotzen; Schijfsma 1946, pp. 100, 101, pi. 6, figs. 1 a-c.

Cibicides ribbingi Brotzen; Williams-Mitchell 1948, p. 103, pi. 9, figs. 5fl, b.

Cibicides beaumontianus (d’Orbigny); McGugan 1957, p. 344, pi. 32, figs. 14 a-c.

Remarks. Cibicides beaumontiana exhibits considerable variation in shape of test. The
dorsal surface may be slightly concave, flat, or convex. The dorsal side of attached forms
reflects the shape of the surface to which the foraminifer had attached itself. Non-
attached forms exhibit more inflated, slightly convex dorsal surfaces.

Brotzen (1936, pp. 186-90, pi. 13, figs. 5-8) described two new species, Cibicides
ribbingi and C. excavata, from the Senonian chalk of southern Sweden. It is possible that
both of these forms are variants of C. beaumontiana. C. ribbingi is the more compressed
form which was attached to a flat or convex surface, while C. excavata is a more inflated

508

PALAEONTOLOGY, VOLUME 9

form which was only partially attached, or not attached at all. Both Brotzen’s species,
although superficially quite dilferent, appear to be within the range of variation of
C. beaumontiana.

Occurrence. C. beaumontiana is common in the Senonian chalk of southern England,
where it ranges from the M. testudinarius Zone (Santonian) to the top of the B. muero-
nata Zone. It also occurs in abundance in the B. lanceolata Zone (Lower Maestrichtian)
of Sidestrand, Norfolk. D’Orbigny originally described C. beaumontiana from the Chalk
{B. mucronata Zone) of Meudon, and from unspecified localities in England.

REFERENCES

AKiMETS, V. s. 1961. Stratigrafiya i foraminifery verkhnemolovykh otlozhenii Belorussii. Paleontologiya
i Stratigrafiya BSSR, Izdat. Akad. Nauk beloruss. SSR. 1-245, pi. 1-19.

BANDY, o. L. 1951. Upper Cretaceous Foraminifera from the Carlsbad area, San Diego Co., Calif.
J. Paleont. 25, 488-513, pi. 72-75.

BARNARD, T. and BANNER, F. T. 1953. Arenaceous Foraminifera from the Upper Cretaceous of England.
Quart. J. geol. Soc. Loud. 109, 173-216, pi. 7-9.

BARR, F. T. 1961. Senonian Foraminifera from the Isle of Wight. Unpubl. Ph.D. thesis, Univ. London,
pp. 1-543, pi. 1-36.

1962. Upper Cretaceous planktonic Foraminifera from the Isle of Wight, England. Palaeontology,

4, 552-80, pi. 69-72.

1966. The foraminiferal genus Bolivinoides from the Upper Cretaceous of the British Isles. Ibid.

9, 220-43, pi. 34-38.

and BERGGREN, w. A. 1965. Planktonic foraminifera from the Thanet Formation (Paleocene) of

Kent, England. Stockholm Contr. Geol. 13, no. 2, 9-26.

and GORDEY, w. g. 1964. Some Upper Cretaceous Foraminifera from the Chapman Collection

(1892). J. Paleont. 38, 306-10, pi. 49.

BEissEL, I. 1891. Die Foraminiferen der Aachener Kreide. Abh. preuss. geol. Landesanst., pt. 3, 1-78,
pi. 1-16.

BELFORD, D. J. 1960. Upper Cretaceous Foraminifera from the Toolonga calcilutite and Gingin chalk.
Western Australia. Bull. Bur. Miner. Resour. Geol. Geophys. Aust. 57, 1-198, pi. 1-35.

BROTZEN, F. 1936. Foraminiferen aus dem schwedischen untersten Senon von Eriksdal in Schonen.

Sver. geol. Unders. Afh. ser. C, no. 396, 1-206, pi. 1-8.

CHAPMAN, F. 1892. Microzoa from the Phosphatic Chalk of Taplow. Quart. J. geol. Soc. Lond. 48,
514-8, pi. 15.

CHARLESWORTH, J. K. 1963. Historical Geology of Ireland. Edinburgh and London.

CUSHMAN, J. A. 1927a. An outline of a re-classification of the Foraminifera. Contr. Cushman Lab.
foramin. Res. 3 (1), 1-105, pi. 1-21.

\911b. Notes on foraminifera in the collection of Ehrenberg. J. Wash. Acad. Sci. 17, 487-91.

1931a. The Foraminifera of the Saratoga Chalk. J. Paleont. 5, 297-315, pi. 34-36.

19316. Cretaceous Foraminifera from Antigua, B.W.I. Contr. Cushman Lab. foramin. Res. 1 (2),

33-46, pi. 5, 6.

1932. The Foraminifera of the Annona Chalk. J. Paleont. 6, 330-45, pi. 50, 51.

1946. Upper Cretaceous Foraminifera of the Gulf Coastal Region of the United States and ad-
jacent areas. Prof. Pap. U.S. geol. Surv. 206, 1-241, pi. 1-66.

1955. Foraminifera, their classification and economic use. 4th ed., Cambridge.

and JARVIS, p. w. 1932. Upper Cretaceous Foraminifera from Trinidad. Proc. U.S. natn. Mus.

80, art. 14, 1-60, pi. 1-16.

• and PARKER, F. L. 1936. Notes on some Cretaceous species of Buliminella and Neobulimina.

Contr. Cushman Lab. foramin. Res. 12 (1), 5-10, pi. 2.

DAiN, L. G. 1934. Foraminifera of the Upper Jurassic and Cretaceous beds of the Djaksy-bai Oil Field
(Temir region). Trans. Geol. Oil Inst., Leningrad, ser. A, no. 43, 1-63, pi. 1-5.

F. T. BARR: UPPER CRETACEOUS FORAMINIFERA 509

EDGELL, H. s. 1954. The stratigraphical value of Bolivinoides in the Upper Cretaceous of north-west
Australia. Contr. Cushman Fdn foramin. Res. 5 (2), 68-76, pi. 13, 14.

EGGER, J. G. 1899. Foraminiferen und Ostracoden aus den Kreidemergeln der oberbayerischen Alpen.
Sber. bayer. Akad. Wiss. 21, 1-230, pi. 1-27.

■ 1909. Foraminiferen der Seewener Kreideschichten. Ibid. 39, art. 11, 1-52, pi. 1-6.

EHRENBERG, c. G. 1838 (1840). Ubcr die Bildung der Kreidefelsen und des Kreidemergels durch
unsichtbare Organismen. Abb. dt. Akad. Wiss. Berl.

1854. Mikrogeologie. Leipzig.

ELEY, H. 1859. Geology in the garden; or, the fossils in the flint pebbles. London (Bell and Daldy).
FRANKE, A. 1925. Die Foraminiferen der pommerschen Kreide. Abh. geol.-palaeont. Inst. Greifswald,
6, 1-96, pi. 1-8.

1927. Die Foraminiferen der Aachener Kreide Erganzungen und Berichtigung zu dem gleich-

namigen Buche mit Atlas von Ignaz Beissel, herausgegeben von E. Holzapfel 1891. Jb. preuss. geol.
Landesanst. BergAkad. 48, 667-98.

1928. Die Foraminiferen der Oberen Kreide Nord und Mitteldeutschlands. Abh. preuss. geol.

Landesanst., N.s., no. Ill, 1-207, pi. 1-18.

FRIZZELL, D. L. 1954. Handbook of Cretaceous Foraminifera of Texas. Rep. Invest. Bur. econ. Geol.
Univ. Tex., no. 22, 1-234, pi. 1-21.

GRAHAM, J. J. and CHURCH, c. c. 1963. Campanian Foraminifera from the Stanford University Campus,
California. Stanford Univ. Pubis, Geol. Sci. 8(1), 1-90, pi. 1-8.

HAGN, H. 1953. Die Foraminiferen der Pinswanger Schichten (Unteres Obercampan). Ein Beitrag zur
Mikropalaontologie der helvetischen Oberkreide Sudbayerns. Palaeontographica, 104, 1-118,
pi. 1-8.

HERON-ALLEN, E. and EARLAND, A. 1910. On the Recent and fossil Foraminifera of the shore-sands at
Selsey Bill, Sussex. Part V. The Cretaceous Foraminifera. Jl R. microsc. Soc. 401-26, pi. 6-11.
HiNTE, J. E. van. 1963. Zur Stratigraphic und Mikropalaontologie der Oberkreide und des Eozans
des Krappfeldes (Karnten). Jb. geol. Bundesanst., Wien, 8, 1-147, pi. 1-22.

HOFKER, J. 1952. The Foraminifera of the Siboga Expedition. Part 3. Siboga Exped, 1-513.

JONES, T. R. 1872. Notes on Eley’s Foraminifera from the English Chalk. Geol. Mag. 9, 123-6.

• and PARKER, w. K. 1860. On the rhizopodal fauna of the Mediterranean, compared with that

of the Italian and some other Tertiary deposits. Quart. Jl geol. Soc. Lond. 16, 292-307.

KAPTARENKO-CHERNOUSOVA, O. K., GOLYAK, L. M., ZERNETSKII, B. F., KRAEVA, E. Ya, and LIPNIK, E. S.

1963. Atlas kharakternikh foraminifer yury, mela i paleogena platformennoi chasti Ukrainy. Trudy
Inst. geol. Nauk, Kiev, ser. strat. paleont. 45, 1-200, pi. 1-47.

LOEBLiCH, A. R., JR., 1951. Coiling in the Heterohelicidae. Contr. Cushman Fdn foramin. Res. 2 (3),
106-11, pi. 12.

and TAPPAN, H. 1962. Six new generic names in the Mycetozoida (Trichiidae) and Foraminiferida

(Fischerinidae, Buliminidae, Caucasinidae, and Pleurostomcllidae), and a redescription of Loxosto-
mum (Loxostomidae, new family). Proc. biol. Soc. Wash. 75, 107-14.

— — 1964. Treatise on Invertebrate Paleontology. Part C. Protista 2 (Sarcodina, chiefly ‘Thecamoebians’
and Foraminiferida). Ed. r. c. moore, 2 vols. Univ. of Kansas Press and Geol. Soc. Am.
LOETTERLE, G. J. 1937. The micropaleontology of the Niobrara Formation in Kansas, Nebraska and
South Dakota. Bull. geol. Surv. Neb., ser. 2, no. 12, 1-73, pi. 1-11.

MACFADYEN, w. A. 1932. Foraminifera from some late Pliocene and glacial deposits of East Anglia.
Geol. Mag. 69, 481-96, pi. 34, 35.

MARIE, p. 1941. Les Foraminiferes de la Craie a Belemnitella mucronata du Bassin de Paris. Mem. Mus.
natn. Hist, nat., Paris, N.s. 12, 1-296, pi. 1-37.

MARSSON, T. F. 1878. Die Foraminiferen der weissen Schreibkreide der Insel Riigen. Mitt, naturw. Ver.
Neu-vorpomm. 10 Jg., 115-96, pi. 1-5.

MARTIN, L. 1964. Upper Cretaceous and Lower Tertiary Foraminifera from Fresno County, California.
Jb. geol. Bundsesanst., Wien, 9, 1-128, pi. 1-16.

MCGUGAN, A. 1957. Upper Cretaceous Foraminifera from Northern Ireland. J. Paleont. 31, 329-48,
pi. 31-35.

MONTANARO GALLiTELLi, E. 1957. A revision of the foraminiferal family Heterohelicidae. Bull. U.S.
natn. Mus. 215, 133-54, pi. 31-34.

510 PALAEONTOLOGY, VOLUME 9

MORROW, A. L. 1934. Foraminifera and Ostracoda from the Upper Cretaceous of Kansas. /. Paleont.
8, 186-205, pi. 29-31.

d’ORBiGNY, A. 1840. Memoire sur les Foraminiferes de la Craie Blanche du Bassin de Paris. Mem. Soc.
geol. Fr. 4 (1), 1-51, pi. 1-4.

PARKER, w. K. and JONES, T. R. 1865. On some Foraminifera from the North Atlantic and Arctic
Oceans, including Davis Straits and Baffin’s Bay. P/iil. Trans. R. Soc. 155, 325-441, pi. 13-19.

POZARYSKA, K. 1957. Lagcnidae du Cretace Superieur de Pologne. Palaeont. pol. 8, 1-190, pi. 1-27,
text-pl. 1-6.

REUSS, A. E. 1845, 1846. Die Versteinenmgen der bohmischen Kreideformation. Abt. 1. (1845) pt. 1,
25-40, 55-88, pi. 8, 12, 13. (1846) pt. 2, 106-10, pi. 24. Stuttgart.

1860. Die Foraminiferen der westphalischen Kreideformation. Sber. baver. Akad. Wiss. Wien,

40, 147-238.

1862. Palaontologische Beitrage (a) Die Foraminiferen des Kreidetuffs von Maastricht, (b) Die

Foraminiferen des Schreibkreide von Riigen. (c) Die Foraminiferen des senonischen Grunsandes
von New Jersey. Ibid. 44, 303-43, pi. 1-8.

SAID, R. and KENAWY, A. 1956. Upper Cretaceous and Lower Tertiary Foraminifera from northern
Sinai, Egypt. Micropaleontology, 2, 105-73, pi. 1-7.

and KERDANY, M. T. 1961. The geology and micropaleontology of the Farafra Oasis, Egypt. Ibid.

7, 317-36, pi. 1, 2.

SCHIJFSMA, E. 1946. The Foraminifera from the Hervian (Campanian) of southern Limberg. Meded.
geol. Sticht (c), sect. 5, no. 7, 1-174, pi. 1-10.

TAKAYANAGi, Y. 1965. The Upper Cretaceous planktonic Foraminifera from the Putah Creek sub-
surface section along the Yolo-Solano County Line, California. Sci. Rep. Tolioku Univ., ser. 2,
36 (2), 161-237, pi. 20-29.

vissER, A. M. 1951. Monograph of the Foraminifera of the type-locality of the Maestrichtian (South-
Limburg, Netherlands). Leid. geol. Meded. 16, 197-360, pi. 1-15.

WALSH, p. T. 1960. Cretaceous Chalk in the Killarney District, Eire. Proc. Geol. Soc. No. 1581, 112-13.

1966. Cretaceous Outliers in south-west Ireland and their implications for Cretaceous palaeo-

geography. Quart. J. geol. Soc. Lond. 122, 63-84, pi. 7.

wiLLiAMS-MiTCHELL, E. 1948. The zonal value of Foraminifera in the Chalk of England. Proc. Geol.
Ass. 59, 91-112, pi. 8-10.

F. T. BARR

Oasis Oil Company of Libya, Inc.,
Oasis Oil Building,

P.O. Box 395,

Tripoli, Libya

Manuscript received 20 September 1965

PODOCNEMIS SOMALIENSIS, A NEW
PLEURODIRAN TURTLE FROM THE
MIDDLE EOCENE OF SOMALIA

by C. A. WALKER

Abstract. A new pleurodiran turtle, Podocnemis somaliensis, is described from the Middle Eocene of Somalia.

Between 1928 and 1930 the fragments of a fairly complete turtle shell were collected
by W. A. Macfadyen and J. A. Hunt from the Middle Eocene of Somalia (then British
Somaliland), along with numerous nautiloids. The specimen was presented to the
Sedgwick Museum, Cambridge, and was then sent to the British Museum (Natural
History) for repair and description, but it has remained undescribed until now.

The geology of the area concerned was dealt with in detail by Macfadyen (1933,
1952) and, to some extent, by Haas and Miller in their description of the nautiloids
(1952).

Family pelomedusidae
Subfamily pelomedusinae
Genus podocnemis Wagler 1830

Podocnemis somaliensis sp. nov.

Plate 80

Diagnosis. Nuchal large and broad; its lateral sutures not straight but rather sigmoid;
notch in anterior border probably absent (or, if present, very small). Carapace almost
as broad as long, widest part just behind bridge. Six neurals, behind which 6th-8th pairs
of pleurals meet in mid-line. Anterior lobe of plastron wide, with blunt anterior edge;
entoplastron broader than long; xiphiplastral symphysis longest.

Holotype. The unique specimen, Sedgwick Museum C 54.276: an almost complete carapace and plas-
tron.

Occurrence. Lower half of the Nautilus Beds, Lower Daban Series, Middle Eocene
(Lutetian); Bijo Gora River, near Las Daban (10° 22' N., 45° 14' E.), some 25 km.
ESE. of Berbera, Somalia.

State of preservation. The shell has suffered some crushing. The outline of the carapace
is distorted a little at both ends. The plastron has been damaged to a greater degree,
and many of the sutures have been obscured, especially on the outer surface where
plaster has been used to fill in eroded areas of bone. There are no shield furrows visible.

The nuchal lacks a portion from the leading edge backwards. All the neurals are
present, but the suture between the first and second cannot be distinguished. All the
pleurals, except the seventh and eighth pairs and the first right, are distorted to some
extent.

[Palaeontology, Vol. 9, Part 3, 1966, pp. 511-16, pi. 80.]

512

PALAEONTOLOGY, VOLUME 9

The plastron has suffered slight distortion in such a way that the sagittal suture no
longer forms a straight line, its front part being bent to the right. The sutures of the ento-
plastron are obscured on the left side by plaster; the sagittal suture and the posterior
suture of the right epiplastron have been covered by plaster on the outer surface but

TEXT-FIG. 1. Podocnemis somaliensis sp. nov., reconstruction of the carapace in
dorsal view, Sedg. Mus. C 54.276. X f.

remain visible in places on the visceral side. On the left side the mesoplastron is almost
entirely absent and large sections of the hyoplastron and hypoplastron are also missing.
The mesoplastron can be seen on the right, but there has been distortion and its inner

EXPLANATION OF PLATE 80

Podocnemis somaliensis sp. nov., holotype, Sedgwick Museum C 54.276. Middle Eocene of Somalia.
Fig. 1 . Dorsal view of carapace. (Nuchal, first neural, and suprapygal are foreshortened.)

Fig. 2. Right lateral view of whole shell.

Fig. 3. Visceral view of plastron.

Fig. 4. Ventral view of plastron, with carapace in position.

Magnification § natural size.

Palaeontology, Vol. 9

PLATE 80

WALKER, Eocene pleurodiran turtle

C. A. WALKER: PODOCNEMIS SOMALIENSIS

513

edge is imperfeet. The xiphiplastra are broken off in front of the notch and their lateral
margins are chipped.

Of the two sets of buttresses the right is more perfect than the left. Short stumps of
the ilia are fused to the carapace and of the pubes and ischia to the plastron.

TEXT-FIG. 2. P. somaliensis sp. nov., reconstruction of the plastron in ventral view,
Sedg. Mus. C 54.276. X f .

Description. Carapace. The carapace is
almost as wide as long, with its greatest
width just behind the bridge. The nuchal is
relatively large, being broader than long,
and its lateral sutures are slightly sigmoid
rather than straight (text-fig. 3). The first
neural, which butts on to the nuchal, forms
an elongated hexagon widest in the middle;
neurals 2-5 are also hexagonal and almost
equal in size, but the greatest width of each
is near its anterior end; neural 6 is smaller,
pentagonal, and wedged between pleurals 5
and 6. The suprapygalis roughly pentagonal

TEXT-FIG. 3. P. somaliensis sp. nov., dorsal view
of the nuchal. X T

514

PALAEONTOLOGY, VOLUME 9

with wide angles and is considerably broader than long. The pygal, although incomplete,
does not seem to be enlarged. There are 8 pairs of pleurals, the first being the largest
and the others becoming successively smaller towards the suprapygal ; pleural 2 is con-
cave anteriorly, 3 is biconcave, while 4-8 are convex in front and concave behind. The
axillary and inguinal buttresses are located under the first and fifth pleural plates, the
axillary being the larger (as is usual in Podocnemis). The visceral surface is not well
preserved; the iliac scars occupy a position on either side of the mid-line under the
seventh and eighth pleurals. Of the 1 1 pairs of peripherals nos. 4-7 are connected with
the bridge.

Plastron. The anterior lobe is wide with a straight anterior edge. The suture separating
the two epiplastra in the mid-line is short; from its posterior end the margin of the epi-
plastron runs obliquely outwards and backwards, separating it from the entoplastron,
and then outwards and forwards, separating it from the hyoplastron. The entoplastron
appears short and wide on the outer surface, but on the inner surface its outline is different
(text-figs. 4, 5). Only one mesoplastron is preserved; it is badly crushed and the sutures
are difficult to see, but it is wider than long. The sutures between the hyoplastra and
hypoplastra are straight, but those between the hypoplastra and xiphiplastra are slightly
curved, concave towards the rear. The xiphiplastra form the longest part of the sagittal
suture but are incomplete posteriorly, thus making it impossible to give any true indica-
tion as to the size and shape of the notch. The pubic and ischial scars are well represented,
with the ischia probably meeting in the mid-line, although some breakage has occurred
at this point.

Measurements. (E) = estimated, (R) =

right side, (V)

= visceral, (O) = outer.

Carapace

Plastron

Length

450 mm. (E)

Sagittal length

377 mm. (E)

Maximum width

424 mm.

Sagittal length of

Maximum

Width

Anterior

Width

anterior lobe

68 mm.

Length

Sagittal length of bridge
Sagittal length of

177 mm.

Nuchal

76-5

120

74 mm.

posterior lobe

132 mm. (E)

Neurals

1.

56-5

29-5

15 mm.

Length of epiplastral 1

1

( 18 mm. (V)

2.

41-5

33-5

17 mm.

symphysis j

1

[ 19-5 mm. (O)

3.

41-5

35-5

21-5 mm.

Length of entoplastron

62 mm.

4.

38

35

18 mm.

Length of hyoplastral 1 1 98

(R)

93 mm. (EV)

5.

38

35

20 mm.

symphysis j (

94 mm. (O)

6.

25-5

27

17 mm.

Length of hypoplastral ) j 94

(R)

90 mm. (EV)

Suprapygal

48

64

—

symphysis j \

90 mm.

Pygal

50 (E)

34-5

65 mm. (E)

Length of xiphiplastral 1 j 103

(R)

1 1 1 mm. (EV)

Inner

Outer

symphysis j (

106 mm. (EO)

length

length

Length of mesoplastron
Width immediately anterior

64 mm. (RO)

Peripherals

1.

27

55 mm.

to bridge

254 mm.

(R)

2.

41

49 mm.

Width immediately posterior

3.

44

57-5 mm (E)

to bridge

225 mm.

4.

51-5

55 mm. (E)

Width of entoplastron

87 mm. (V)

5.

45-5

51-5 mm.

98 mm. (O)

6.

47

58-5 mm.

Width of mesoplastron

68 mm. (EO)

7.

59

73 mm. (E)

C. A. WALKER; PODOCNEMIS SOMALIENSIS

515

Comparisons. The presence of small, laterally placed mesoplastra and of only six neurals
indicates that the species should be placed in the sub-family Pelomedusinae (see Zangerl
1948, pp. 48 et seq.). Reference to Podocnemis is suggested by the exact pattern of bones
in the carapace (with the nuchal and 1st neural in direct contact, not separated by the
1st pair of pleurals) and by the absence of a central fontanelle between hyo- and hypo-
plastra.

Podocnemis somaliensis is the largest species of the genus yet discovered in Africa.
Although no shield furrows are visible in this specimen, it may be differentiated from

TEXT-FIGS. 4, 5. P. somaliensis sp. nov., ventral and visceral views of the entoplastron. x4.

Other contemporary African species of the genus by the size, shape, and proportions of
the nuchal, the carapace as a whole and the entoplastron. These species include only
P. antiqua Andrews 1903 and P. stromeri Reinach 1903, both from the Middle Eocene
of Egypt (see also Dacque, 1912).

P. antiqua resembles the new species in the shape of the carapace, but it is much smaller
and differs considerably in its anatomy. The nuchal contrasts with that of P. somaliens,s
in having straight lateral sutures and, behind it, a pair of foramina on either side of the
1st neural which are certainly absent in the new form. The suprapygal differs also in
being much broader than long.

P. stromeri and its variety major both have oval shells, the proportions of which are
like those of the Recent P. madagascariensis Grandidier but unlike those of P. somali-
ensis, in particular their nuchals and entoplastra are of different proportions. The xiphi-
plastral symphyses are not the longest in the mid-line, as they are in the Somalian species.

Remarks. It would be unwise to draw any conclusions as to the relationships of P.
somaliensis with other species of Podocnemis until more material becomes available; it
would be especially useful to have knowledge of its shield furrows and to compare it
first hand with other African forms. However, the specimen is obviously specifically
distinct and sufficiently well preserved to merit designation as the holotype of a new
species.

Acknowledgements. I thank Dr. A. J. Charig for criticizing the manuscript; the Sedgwick Museum,
Cambridge, for the further loan of the specimen; and Mr. A. Tanner who took the photographs.

M m

C 4179

516

PALAEONTOLOGY, VOLUME 9

REFERENCES

ANDREWS, c. w. 1903. On some pleurodiran chelonians from the Eocene of the Fayum, Egypt. Ann.
Mag. nat. Hist. (7), 11, 115-22, pi. 7-8.

1906. A descriptive catalogue of the Tertiary Vertebrata of the Fayum, Egypt. London.

DACQUE, E. 1912. Die fossilen Schildkroten Aegyptens. Geol. paldont. Abh. (n.s.) 10, 4, 275-337, pi. 1-2.
HAAS, o. and miller, a. k. 1952. Eocene nautiloids of British Somaliland. Bull. Am. Mus. nat. Hist.
99, 5, 314-54, pi. 21-31.

MACFADYEN, w. A. 1933. The geology and palaeontology of British Somaliland : Part I. The geology of
British Somaliland. Government of the Somaliland Protectorate.

1952. Note on the geology of the Daban area and the localities of the described nautiloids. Bull.

Am. Mus. nat. Hist. 99, 5, 347-9.

REiNACH, A. VON. 1903. Scliildkrotenreste aus dem agyptischen Tertiar. Abh. Senckenb. Naturforsch.
Ges. 29, 1-64, pi. 1-17.

ZANGERL, R. 1948. The vertebrate fauna of the Selma formation of Alabama: Part I. Introduction.
Part II. The pleurodiran turtles. Fieldiana, Geol. Mem. 3, 1 and 2, 1-56, pi. 1-4.

C. A. WALKER

Department of Palaeontology,
British Museum (Natural History),

Manuscript received 29 October 1965 London, S.W. 7

SOME LATE SILURIAN BRYOZOA FROM THE
CANADIAN ARCTIC ISLANDS

by THOMAS E. BOLTON

Abstract. One new species of trepostome Diplotrypa frankUni, and two species of cyclostome Fistulipora(?)
mutabilis Hennig and Cyclotrypa sihirica Hennig originally described from the Liidlovian rocks of Gotland are
characteristic of the Early Ludlovian shelly faunas scattered throughout the Canadian Arctic Islands.

Bryozoa form only a small part of the Silurian shelly faunas collected from various
islands in the Canadian Arctic Archipelago. However, some species are rather widespread
and through their similarity with precisely dated fauna on the Island of Gotland assist in
the regional correlation of the associated rock units.

In North America, Llandoverian and Wenlockian bryozoan faunas have been detailed
by Bassler (1906, 1928) and Perry and Hattin ( 1960) but, with the exception of the Bryozoa
of the Tonoloway Formation of West Virginia (Bassler 1923), Ludlovian Bryozoa
are little known. In contrast, the Silurian Bryozoa of the Soviet Republic of Tuva have
been studied in great detail by Astrova ( 1959, 1965), and recently the Ludlovian Bryozoa
of Great Britain have been discussed by Owen (1960, 1962). Certain elements of both
these faunas, especially among the Cyclostomata, are closely related to the Canadian
Arctic bryozoan faunas. A still closer relationship is evident, however, between the
Canadian species and forms described by Hennig (1905, 1906, 1908) from the Island of
Gotland. Several species or closely related forms characteristic of the Early and Middle
Ludlovian Hemse and Hamra Groups (Hede 1960) have been identified in the ‘Atrypella
schei faunas’ of the Read Bay Formation on Somerset, Cornwallis, and south-western
Devon Islands and of the Douro Formation on north-western Devon Island (text-fig. 1 ).
The most abundant form is the trepostome Diplotrypa frankUni sp. nov., mainly from
Early Ludlovian rocks, but in addition identified in Late Silurian or Early Devonian
strata in east-central Ellesmere Island. Associated with this species are the cyclostomes
Fistulipora(?) mutabilis Hennig and Cyclotrypa sihirica Hennig, the cryptostome FenesteUa
sp., and poorly preserved trepostomes of the genus Eridotrypa. Among the latter is a
specimen (GSC No. 20430, Read Bay Eormation, Cornwallis Island) with zooecia
0-25-0-3 mm. in diameter and rare diaphragms that is closer to Eridotrypa wnbonensis
Owen (1962, p. 203) from the Lower Bringewood Beds of the Welsh borderland than
to E. ramea Hennig (1908, p. 38) with its more numerous diaphragms found within the
Hemse Group.

SYSTEMATIC DESCRIPTIONS

Order trepostomata Ulrich 1882
Genus diplotrypa Nicholson 1879

Type species. Favosites petropolitanus Pander 1830

[Palaeontology, Vol. 9, Part 3, 1966, pp. 517-22, pis. 81-82.]

518

PALAEONTOLOGY, VOLUME 9

Diplotrypa franklini sp. nov.

Plate 81, figs. 1-6; Plate 82, fig. 4

1958 Mesotrypa sp. cf. M. suprasilurica Hennig; Thorsteinsson, pp. 49, 50, 52, 67.
1963 M. suprasilurica Hennig; Fortier et aL, pp. 132, 205.

1963 M. cf. M. suprasilurica Hennig; ibid., p. 240.

1965 M. sp. cf. M. suprasilurica Hennig; Bolton, p. 12.

TEXT-FIG. 1. Late Silurian bryozoan localities Canadian Arctic Archipelago. 1. Cape Admiral M’Clin-
tock, Somerset Island. 2. Resolute Bay area, Cornwallis Island. 3. Goodsir Creek, Cornwallis Island.
4. Radstock Bay, Devon Island. 5. Colin Archer Peninsula, Devon Island. 6, Darling Peninsula,

Ellesmere Island.

EXPLANATION OF PLATE 81

Figs. 1-6. Diplotrypa fraukliui sp. nov. 1. Longitudinal section multilaminate colony, paratype GSC
20425, X 23. 2. Tangential section showing large mesopores, paratype GSC 20423, c. x 32. 3, 5, 6.
Tangential and longitudinal sections, holotype GSC 20421, c. X 32 and x 100. 4. Tangential section,
paratype GSC 20422, c. x 32.

Palaeontology, Vol. 9

PLATE 81

BOLTON, Silurian Bryozoa from Canada

r

•)

9

'WR

„■ . T I

T. E. BOLTON; LATE SILURIAN BRYOZOA

519

Description. Zoaria vary from small, cylindrical to subhemispherical to shallow conical
mounds, some concentric layered; largest is 45 mm. wide and 30 mm. high; suggestion
of a few low small monticules.

In tangential section, the zooecia are polygonal to subrounded, the openings slightly
more oval where the zooecial walls are thicker (paratype 20425), 0-20-0-22 mm. in
diameter, with a minimum of 0-15 mm. and a maximum of 0-28 mm. Generally eight to
ten zooecia are present in 2 mm. Single rows of subtriangular to rectangular mesopores
completely isolate, form a half circle, or are lacking from between zooecial openings,
the arrangement varying within one complete colony horizontally and vertically, as well
as between layers within a multilaminate colony (paratype 20425). Diameter of meso-
pores normally 0-1 to 0-15 mm., but some of the more rectangular mesopores extend up
to 0-3 mm. (PI. 81, fig. 2). Walls between mesopores and zooecial openings clear,
amalgamate (?), thin (PI. 81, fig. 6). Rare megacanthopore-like structures are confined
to thickened walls at zooecial junctions, localized in paratype 20422, larger and more
numerous in paratype 20426; absent from most colonies.

In longitudinal sections of multilaminate colonies, the normally erect zooecia and
mesopores both may be inclined for a very short distance along the basal laminae. Thin
diaphragms are present throughout the length of the zooecial tubes, horizontal to slightly
oblique, but abruptly bending up at the boundary of the zooecial wall, 0-2 to 0-3 mm.
apart (four to six per 1 mm.) with a minimum of 0-1 5 mm. and a maximum of 0-45 mm.
(paratype 20423). Diaphragms much closer in mesopores, OT mm. apart normally (ten
to eleven per 1 mm.). Mesopores are continuous throughout zoaria, walls often curved
inward at diaphragms. Walls clear, skeletal microstructures not preserved, of equal
thickness throughout length of tubes.

Discussion. This new species is assigned to Diplotrypa rather than to Mesotrypa because
of the absence or rarity of acanthopores and cystiphragms, despite amalgamate (?) walls
characteristic of the latter genus. Diplotrypa franklini differs from D. nwnniiforniis (HaW),
from the Rochester and Osgood Formations (Niagaran) of New York and Indiana
States, in that the latter discoidal species has more closely spaced, more inclined dia-
phragms and normally six to seven zooecia in 2 mm. ( Bassler 1 906, p. 27 ; Perry and Hattin
1960, p. 707). D. wa/keri Bassler (1906, p. 47) from the Rochester Formation of Ontario
and New York State and D. neglectoformis Astrova ( 1959, p. 26) from the Wenlockian
both have larger zooecia, discontinuous mesopores, and fewer diaphragms.

Mesotrypa suprasihirica Pfennig (1908, p. 30, text-figs. 35-37), principally from the
Late Llandoverian (Hede 1921, p. 31) but also recorded from the Early Ludlovian of
Gotland, with which this new species was originally compared, is identical in zoarial
form. It difters in that the species has more distinct acanthopores, some with rounded
lumen often projecting into the hollow of the zooecial tubes and more oblique or funnel-
shaped diaphragms. "Prasopora' [= Diplotrypa?] gotlandica Hennig(1908, p. 28) from
the Early Ludlovian Hemse Group of Gotland has smaller zoaria and some cystiphragm-
like and closer diaphragms. Astrova (1959, p. 29) has suggested that some of the species
assigned to Mesotrypa that lack cystiphragms and acanthopores might be included in
her Wenlockian genus Mesotrypella. This genus characteristically has branched colonies,
with rounded or subrounded zooecial openings, oblique or flexed diaphragms, and very
undulating walls of constant thickness throughout the entire colony, whereas Diplotrypa

520

PALAEONTOLOGY, VOLUME 9

has massive colonies and straight or slightly flexed walls. Specimens of D. franklini
lacking acanthopores nevertheless have many features in common with Mesotrypella
alasheusis Astrova.

Distribution and types. D. franklini is most abundant in the Lower Ludlovian Member A of the Read
Bay Formation, 459 to 731 ft. (holotype 20421; paratypes 20422, 20423) and at 1,365 ft. below the
top on Goodsir Creek on the central-east coast of Cornwallis Island. Additional specimens have been
identified 22 ft. above the base of Upper Ludlovian Member C of the Read Bay Formation at the same
locality and in Member A near Resolute Bay, central-south coast of Cornwallis Island; upper beds of
the Read Bay Formation, Cape Admiral M’Clintock, north coast of Somerset Island (paratype 20424)
and the west side of Radstock Bay, south-western Devon Island (paratype 20425); lower beds of the
Douro Formation, Colin Archer Peninsula, north-western Devon Island; and 3,380 to 3,400 ft. above
base of undifferentiated Allen Bay-Read Bay Formation (Late Silurian or Early Devonian) on Darling
Peninsula, east-central Ellesmere Island (paratype 20426).

Order cyclostomata Busk 1852
Genus fistulipora M‘Coy 1850

Type species. Fistulipora minor M‘Coy 1850

Fistulipora (?) mutabilis Hennig 1908
Plate 82, figs. 2, 5, 7, 8

1908 Fistulipora mutabilis Hennig, p. 19, pi. 2, figs. 1-7; pi. 7, figs. 3, 4; text-figs. 21-23.

1958 F. sp. cf. mutabilis Hennig; Thorsteinsson, pp. 50, 59.

Description. Zoaria thin small expansions, largest over 13 mm. long and 26 mm. thick.

In tangential section, the zooecial openings are oval, 0-2-0-25 mm. in diameter ranging
between 0-18 to 0-3 mm., four to five zooecia in 2 mm. Lunaria are poorly developed,
broadly arched to slightly indenting zooecial cavities, lacking or obscure on most zooecia.
Vesicles subpolygonal or polygonal, variable in size and number, at least one between
adjacent zooecia.

In longitudinal section, zooecia are recumbent near the base of a colony, erect with
thin walls throughout the remainder of the zoarium. Diaphragms are rare, thin, horizon-
tal. Zooecia are separated by single polygonal tubes, 0-1 mm. in diameter with closely
spaced horizontal diaphragms, by interlocking tubes of varying diameter and horizon-
tal or convex diaphragms, or by vesicles; the latter structures are particularly well
developed in early stages of growth (PI. 82, fig. 7). Thickness of these compound zones
ranges between 0-4 and 0-7 mm.

Discussion. Similar species include F. strawi Owen, but with fewer intervening vesicles,
and certain forms of F. promiscua Perry and Hattin with barely discernible lunaria. As

EXPLANATION OF PLATE 82

Figs. 1, 3. Cyclotrypa silurica Hennig. Longitudinal and tangential sections, GSC 20429, x23.

Figs. 2, 5, 7, 8. Fistulipora (?) mutabilis Hennig. 2, 7. Tangential and longitudinal sections, GSC 20427,
X 25. 5, 8. Longitudinal section showing complete colony and enlargement of upper right corner,
GSC 20428, X 10 and x23.

Fig. 4. Diplotrvpa franklini sp. nov. Tangential section showing a few acanthopores, paratype GSC
20426, x23.’

Fig. 6. Cheilotrypa ostiolata [HaW). Longitudinal section to show irregular axial canal, GSC 20431, X 10.

Palaeontology, Vol. 9

PLATE 82

BOLTON, Silurian Bryozoa from Canada

O’

;rJ»:

/

;■ ;f*'Vif?

-■ ;
" ""

( '

'tfi- '

' j; i --V' , 'bswJM

.

:': 1,

T. E. BOLTON: LATE SILURIAN BRYOZOA

521

many zooecia of F. miitabilis apparently lack lunaria, or where developed the lunaria are
inconspicuous, this species might equally be assigned to Cyclotrypa Ulrich 1896.
Specimen No. 20428 (PI. 82, figs. 5, 8) with its dense stereom (?) and pseudo-ramose habit
is strikingly similar in both tangential and longitudinal sections to Cheihtrypa? opina-
bilis Astrova (1965, p. 156, pi. 13, figs. 2a, c) from the Lower Ludlovian. The irregular
central filling is 3-4 to 4-6 mm. thick in the Canadian specimen and between 0T8 and
2-8 mm. thick in the Russian specimen. In neither specimen is a true irregular axial canal
or adjacent non-tabulated vertical zooecia characteristic of Cheilotrypa (C. ostiolata
(Hall) from the Rochester Formation at De Cew Falls, Southern Ontario, PI. 82, fig. 6)
evident in the central filling; accordingly the Canadian form is interpreted as an encrust-
ing form of Fistulipora mutabilis.

Distribution and types. Lower Ludlovian Member A of the Read Bay Formation, 618 and 698 ft. below
the top (GSC Nos. 20427, 20428), Goodsir Creek on central-east coast, and upper beds of the Read
Bay Formation near Resolute Bay, central-south coast of Cornwallis Island. The syntypes were from
the Early Ludlovian Hemse Group (Hede 1921, p. 57), but the species was recognized by Hennig
throughout the Late Llandoverian to Middle Ludlovian strata of Gotland.

Genus cyclotrypa Ulrich 1896
Type species. Fistulipora communis Ulrich 1890.

Cyclotrypa sihtrica Hennig 1 908
Plate 82, figs. 1, 3

1908 Cyclotrypa silurica Hennig, p. 24, pi. 1, fig. 19; pi. 4, figs, 8, 9; text-figs. 26-29.

1958 C. sp. cf. C. silurica Hennig; Thorsteinsson, pp. 49, 50.

Description. Zoarium cylindrical, branching, up to 12 mm. in diameter.

In tangential section, zooecia are thin walled, oval, 0-2 mm. in diameter with a
maximum of 0-28 mm., five zooecia in 2 mm. Lunarium undeveloped. Adjacent zooecia
nowhere in contact, separated by at least one polygonal vesicle; interzooecial distance
ranges from OT mm. to more than 0-4 mm.

In longitudinal seetion, zooeeia open obliquely to the zoarial surface, thin walled,
with a few thin, horizontal diaphragms in both axial and peripheral regions up to
0-8 mm. apart. The interzooecial tissue normally is vesicular; in single tubes the
cystiphragm-like diaphragms are OT to 0-3 mm. apart, closest in the peripheral regions.

Discussion. The branching form and cystiphragm (?) structure of C. silurica allies it with
the Ludlovian species Fistuliranms sinensis Astrova (1965, p. 153, pi. 12, figs. 2a~d;
pi. 13, figs. 1 a, b), but it lacks the lunaria and heavy concentration of ‘cystiphragms’ in
the peripheral zone characteristic of that genus.

Distribution and types. This species was identified from collections 331 (GSC No. 20429) and 618 ft.
below the top of the Lower Ludlovian Member A of the Read Bay Formation, Goodsir Creek, central-
east coast of Cornwallis Island. The syntypes were from the Middle Ludlovian Hamra Group (Hede
1921, p. 76) but the species also was recognized by Hennig in the Late Llandoverian, Wenlockian,
and Early Ludlovian strata of Gotland.

522

PALAEONTOLOGY, VOLUME 9

REFERENCES

ASTROVA, G. G. 1959. Silurian Bryozoa from central and western Tuva Republic. Trans. Pal. Inst. Acad.
Sci. USSR, 79, 3-72, pi. 1-12.

1965. Morphology, evolution history and classification of Ordovician and Silurian Bryozoa.

Ibid. 106, 3-336, pi. 1-84

BASSLER, R. s. 1906. The Bryozoan fauna of the Rochester shale. Bull. U.S. geol. Surv. 292, 1-65,
pi. 1-31.

1923. Systematic Paleontology of Silurian deposits. Bryozoa. Maryland geol. Surv. Silurian,

405-12, pi. 13, 14.

1928. Bryozoa. In Twenhofel, W. H., Geology of Anticosti Island. Mem. geol. Surv. Can. 154,

143-68, pi. 5-14.

BOLTON, T. E. 1965. Trilobites from the Upper Silurian rocks of the Canadian Arctic Archipelago:
Encrinurus (Frammia) and Hemiarges. Contr. Can. Pal., Bull. geol. Surv. Can. 134.

FORTIER, Y. o. et cd. 1963. Geology of the north-central part of the Arctic Archipelago, Northwest
Territories (Operation Franklin). Mem. geol. Surv. Can. 320, 1-671.

HEDE, j. E. 1921. Gotlands silurstratigrafi. Sverig. geol. Unders. 305 (C) (Arsb. 14, 1920, No. 7).

1960. The Silurian of Gotland. Guide to Excursion Cl 7. Intern, geol. Congress. XXI Session,

44-87.

HENNiG, A. 1905. Gotlands Silur-Bryozoer. 1. Arkiv.fdr Zoologi, 2, 10, 1-37, pi. 1, 2.

1906. Gotlands Silur-Bryozoer. 2. Ibid. 3, 10, 1-62, pi. 1-7.

1908. Gotlands Silur-Bryozoer. 3. Ibid. 4, 21, 1-64, pi. 1-7.

OWEN, D. E. 1960. Upper Silurian Bryozoa from central Wales. Palaeontology, 3, 69-74, pi. 16.

1962. Ludlovian Bryozoa from the Ludlow district. Ibid. 5, 195-212, pi. 28-32.

PERRY, T. G. and HATTiN, D. E. 1960. Osgood (Niagaran) Bryozoans from the type area. J. Paleont. 34,
695-710, pi. 85-90.

THORSTEiNSSON, R. 1958. Comwallis and Little Cornwallis Islands District of Franklin, Northwest
Territories. Mem. geol. Surv. Can. 294, 1-134.

THOMAS E. BOLTON

Geological Survey of Canada,
Ottawa,

Manuscript received 8 November 1965 Ontario, Canada

THE PALAEONTOLOGICAL ASSOCIATION

COUNCIL 1966-7

President

Professor T. S. Westoll, The University, Newcastle upon Tyne
Vice-Presidents

Dr. W. S. McKerrow, University Museum, Oxford
Professor F. H. T. Rhodes, University College, Swansea

Treasurer

Dr. C. Downie, Department of Geology, The University, Mappin Street, Sheffield 1

Secretary

Professor C. H. Holland, Department of Geology, Trinity College, Dublin, 2 Ireland

Assistant Secretary

Dr. J. M. Hancock, Department of Geology, King’s College, London, W.C.2

Editors

Mr. N. F. Hughes, Sedgwick Museum, Cambridge
Dr. Gwyn Thomas, Department of Geology, Imperial College of Science, London, S.W. 7
Dr. I. Strachan, Department of Geology, The University, Birmingham, 15
Dr. M. R. House, University Museum, Oxford

Other members of Council

Professor P. M. Butler, Royal Holloway College, Surrey
Mr. M. A. Calver, Geological Survey Office, Leeds
Dr. G. Y. Craig, Grant Institute of Geology, Edinburgh
Miss Grace Dunlop, Bedford College, London
Dr. T. D. Ford, The University, Leicester
Dr. B. M. Funnell, Sedgwick Museum, Cambridge
Dr. R. P. S. Jefferies, British Museum (Natural History), London
[Dr. G. A. L. Johnson, The University, Durham City
Dr. F. A. Middlemiss, Queen Mary College, London
Dr. W. D. I. Rolfe, Hunterian Museum, Glasgow
Professor Scott Simpson, The University, Exeter
Dr. A. H. Smout, British Petroleum Company, Sunbury-on-Thames
Dr. L. B. H. Tarlo, The University, Reading

Overseas Representatives

Australia-. Professor Dorothy Hill, Department of Geology, University of Queensland, Brisbane
Canada-. Dr. D. J. McLaren, Geological Survey of Canada, Department of Mines and Technical
Surveys, Ottawa

India-. Professor M. R. Sahni, 98 The Mall, Lucknow (U.P.), India

New Zealand-. Dr. C. A. Fleming, New Zealand Geological Survey, P.O. Box 368, Lower Hutt
West Indies and Central America: Mr. John B. Saunders, Geological Laboratory, Texaco Trinidad,
Inc., Pointe-i-Pierre, Trinidad, West Indies

Western U.S.A. : Professor J. Wyatt Durham, Department of Paleontology, University of California,
Berkeley 4, Calif.

PALAEONTOLOGY

VOLUME 9 • PART 3

CONTENTS

Predation and shell damage in a Visean brachiopod fauna. By h. brunton 355

Fossil wood of Anacardiaceae from the British Eocene. By d. w. brett 360

Parachonetes, a new Lower and Middle Devonian brachiopod genus. By J. g.
JOHNSON 365

Morphology and stratigraphic range of the phyllocarid crustacean Caryocaris
from Alaska and the Great Basin. By m. churkin, jr. 371

Devonian schizophoriid brachiopods from Western Europe. By yvonne p.
pococK 381

The micrometric formula and the classification of fenestrate cryptostomes. By
R. TAVENER-SMITH 413

A new productid brachiopod from the Upper Visean of Scotland. By k. a. g.
smELLS 426

Description of dimorphism in Striatopora jiexuosa Hall. By w. A. Oliver, jr. 448

A phylloceratid ammonite from the Speeton Clay (Lower Cretaceous) of
Yorkshire. By p. F. rawson 455

The Upper Devonian index ammonoid Cheiloceras from New South Wales.

By T. B. H. JENKINS 458

Schizochroal eyes and vision in some phacopid trilobites. By E. N. K. clarkson 464

Two cristrate megaspores from the Lower Carboniferous of Scotland. By k. l.

ALVIN 488

Upper Cretaceous Foraminifera from the Ballydeenlea Chalk, County Kerry,
Ireland. By f. t. barr 492

Podocnemis somaliensis, a new pleurodiran turtle from the Middle Eocene of
Somalia. By c. A. walker 511

Some late Silurian Bryozoa from the Canadian Arctic Islands. By t. e. bolton 517

PRINTED IN GREAT BRITAIN AT THE UNIVERSITY PRESS, OXFORD
BY VIVIAN RIDLER, PRINTER TO THE UNIVERSITY

-s- h i

VOLUME 9 • PART 4

Palaeontology

DECEMBER 1966

PUBLISHED BY THE
PALAEONTOLOGICAL ASSOCIATION
LONDON

Price £3

THE PALAEONTOLOGICAL ASSOCIATION

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

Institute membership .... £7 (U.S. $20.00)

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

Student membership . . . £2. 2s. (U.S. $6.50)

There is no admission fee. Student members will be regarded as persons receiving
full-time instruction at educational institutions recognized by the Council; on first
applying for membership, they should obtain an application form from the Secretary
or the Treasurer. Subscriptions are due each January, and should be sent to the
Treasurer, Dr. C. Downie, Department of Geology, The University, Mappin Street,
SheflBeld 1, England.

Palaeontology is devoted to the publication of papers (preferably illustrated) on all
aspects of palaeontology and stratigraphical palaeontology. Four parts are published
each year and are sent free to all members of the Association. Members who join for
1966 wiU receive Volume 9, Parts 1 to 4.

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

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

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

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

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

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

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

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

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

Vol. 9 (for 1966) in 4 parts at £3 or U.S. $9.00 per part.

A complete set. Volumes 1-9, consists of 34 parts and costs £88 or U.S. $264.

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

© The Palaeontological Association, 1966

THE SILURIAN BRACHIOPOD
EOCOELIA HEMISPHAERICA
(J. DE C. SOWERBY) AND RELATED SPECIES

by A. M. ZIEGLER

Abstract. The atrypacean Eocoelia occurs widely in the Upper Llandovery and is now known from the lower
Wenlock. It shows evolutionary trends in the progressive suppression of ribs and in the modification and
strengthening of the articulatory mechanism. These trends are used to define four successive species, Eocoelia
hemisphaerica (J. de C. Sowerby) of C1-C2 age, E. intermedia (Hall) of C3-C4 age, E. curtisi sp. nov. of C5 age,
and E. sulcata (Prouty) of Cs-Wenlock age. The taxa Eocoelia hemisphaerica sefinensis (Williams) and Eocoelia
qiiebecensis Amos and Boucot are relegated to the synonymy of Eocoelia hemisphaerica. Because of the
abundance of Eocoelia at some localities, the term Eocoelia Community seems appropriate; this community
probably occurred in a near-shore environment. Brachiopod lineages, such as the Eocoelia lineage, provide
a reliable basis for fine-scale correlation.

The small costate atrypacean Eocoelia is a characteristic element of Upper Llandovery
and low Wenlock shelly deposits in Great Britain, eastern North America, and Siberia.
It is also known from Norway and South America (Harrington 1950). Differences,
considered to be evolutionary in nature, have been noticed in Eocoelia by Williams
(1951) and Amos and Boucot (1963), and make the fossil an important one for correla-
tion purposes. Material for the present investigation was collected from the southern
part of the Welsh Borderland (Gloucestershire, Herefordshire, and Worcestershire)
where Eocoelia occurs in profusion and can be obtained from many stratigraphic
levels ranging in age from (early Upper Llandovery) to early Wenlock.

The type species of Eocoelia, Atrypa hemisphaerica, was originally described by
J. de C. Sowerby in Murchison’s The Silurian System (1839, p. 637, pi. 20, fig. 7).
Unfortunately the locality of the figured specimen was not specifically stated. There is
a specimen in the Geological Survey Museum labelled ‘Sil Syst PI. 20, f. 7.’; it resembles
the figure only in being a block containing many specimens. However, the block may
be accepted tentatively as the type specimen with the theory that the figure is an artist’s
interpretation of this block. The label on the block is Ankerdine Hill (an area in
Worcestershire at the southern end of the Abberley Hills), one of the two localities
mentioned by Sowerby for his species. Davidson (1866, pi. XI 11, fig. 23) was probably
wrong in stating the locality of Sowerby ’s specimen to be May Hill, a locality that
was not mentioned in the first edition of The Silurian System (1839).

Davidson further confounded the issue by figuring specimens from several localities
and inadvertently giving greater prominence to specimens which, on close examination,
are quite distinct from the type. Thus, he showed enlargements of specimens from
Penwhapple Glen, Ayrshire (1866, pi. XIII, figs. 29a, 30u), which, in contrast with the
type, do not have discrete dental lamellae and which have relatively weak ribbing.
It has been assumed by subsequent workers, namely Williams (1951, pp. 113-14), and
Amos and Boucot (1963, pp. 445-7), that specimens like Davidson’s Penwhapple Glen
specimens are identical with the type; in each case these workers have erected new names

[Palaeontology, Vol 9, Part 4, 1966, pp. 523-43, pis. 83, 84.]

C 4397 N n

524

PALAEONTOLOGY, VOLUME 9

to describe species which, in reality, are identical with the type of Eocoelia hemi-
sphaerica. These names, Coelospira hemisphaerica sefinensis Williams and Eocoelia
quebecensis Amos and Boucot, must be relegated to the synonymy of E. hemisphaerica.
However, the important fact is that Williams, and Amos and Boucot, showed that the
form with strong ribs and dental lamellae is definitely earlier in date than the other, and
Williams, working in the type area of Llandovery, established this date as of the
Upper Llandovery.

A second species of Eocoelia, conspecific with the type of Hall’s (1860, p. 147)
E. intermedia, occurs in C4 beds at Llandovery. Specimens of E. intermedia are similar
to E. hemisphaerica, but possess relatively weaker ribs and display intermediate stages in
the disappearance of the umbonal chambers of the pedicle valve, with the consequent
disappearance of the dental lamellae as distinct entities.

The complete disappearance of the umbonal chambers as well as a further reduction
in rib strength is seen in a third species, here named Eocoelia curtisi. Davidson’s
Penwhapple Glen specimens mentioned above are of this type, which occurs abundantly
in the basal Damery beds of the Tortworth Inlier.

The highest Llandovery and lowest Wenlock beds of Tortworth, May Hill, and the
Malvern Hills contain an Eocoelia with mere vestiges of ribs; the ribs occur only as
faint traces in the early growth stages of the shell, that is, in the beak region. This form
occurs in North America and has been described as Eocoelia sulcata (Prouty) (Prouty
and Swartz 1923, p. 466).

Four species of Eocoelia, E. hemisphaerica, E. intermedia, E. curtisi, and E. sulcata
may thus be recognized. All four occur on both sides of the Atlantic and their strati-
graphic succession may be demonstrated in localities as widely separated as Maryland,
Nova Scotia, and the Welsh Borderland.

In the present investigation bulk collections were made from forty-five localities in
the southern part of the Welsh Borderland and thirty-two of these contained specimens
of Eocoelia. Numbers of Eocoelia range up to 212 in a collection, and the fourteen
largest collections have been treated statistically (Appendix; for a description of other
localities see Ziegler 1963). Collections from Llandovery and Ireland have also been
examined, as well as eighty-one collections from a variety of regions in eastern North
America. Such dimensions as length, width, distance between ribs at a set distance
from the umbo, and height of ribs at the same distance from the umbo, were measured
(text-fig. 1); also, the number of ribs was noted, as was the presence or absence of such
features as the umbonal cavities and the cardinal process.

Acknowledgements. I would like to thank Dr. W. S. McKerrow who supervised the research, which
was part of a D.Phil. thesis at Oxford University; Mr. J. M. Edmonds (University Museum, Oxford),
who housed the thesis collections; Dr. A. J. Boucot (California Institute of Technology), who made
available for study his extensive collections of North American and European material; Dr. M. L. K.
Curtis (City Museum, Bristol), and Dr. L. R. M. Cocks (British Museum (Natural History)), who
conducted the writer to localities at Tortworth, Gloucestershire, and Norbury, Shropshire, respectively.
The research was supported in part by a grant from the Burdett-Coutts Eund, Oxford University.

ASSEMBLAGE ANALYSIS AND INTERPRETATION

The extent to which a fossil sample is representative of the original population is
a problem which should be considered by any taxonomist who deals with closely related

A. M. ZIEGLER: SILURIAN BRACHIOPOD EOCOELIA H EM I SP H AERIC A 525

species. Differences between samples cannot be used for taxonomic discrimination
unless it can be shown that the samples were unaffected by current activity capable of
selectively removing portions of the assemblage. Recently many techniques have been
suggested for detecting preburial alteration of fossil assemblages (Fagerstrom 1964), and
some of these are applicable to the Eocoelia assemblages preserved in the Upper Llan-
dovery sandstones and shales of the Welsh Borderland.

Measured ■{

Derived-^

w = width
I = length

ljj = distance from umbo to point of deflection
6X= distance across 6 ribs at 3.5 mm from umbo
hf = rib " height"at 35mm from umbo

6e = ongle enclosed by the 6 central ribs

2tan3e = ^-|^ —

3.5mm

= rib "width" at 3.5mm, from umbo
Wr = 3.5mm. x 2tan(-^)

TEXT-FIG. 1. Brachial valve of Eocoelia showing the parameters measured.

It should be noted that the fossils of the Upper Llandovery rocks are typically
confined to the bottom of some of the beds, and not evenly distributed throughout the
rock either vertically through several beds or horizontally along one bed, so the activity
of currents in concentrating the shells may be inferred. Furthermore, most of the valves
are disarticulated; of a total of 704 brachial valves collected, only 6 remain attached to
pedicle valves. This consideration is not very significant since Eocoelia did not possess
completely locking valves, and so these would become disarticulated easily once the soft
parts had disappeared. Of more concern is the fact that the ratio of opposing valves
may be as high as 1 ; 1 -5. On the other hand, less than 1 % of all the valves collected show
signs of wear or abrasion and what little breakage exists may be largely attributed to
post-burial compaction of the sediment.

The character of the size-frequency distributions has been used by some workers to
distinguish sorted assemblages. Boucot (1953) theorized that undisturbed assemblages
should be highly skewed to the right due to high infant mortality, while current-sorted

526

PALAEONTOLOGY, VOLUME 9

assemblages would have normal distributions. But Rudwick (1962) and Craig and
Hallam (1963) showed that an undisturbed assemblage might very well have a modal
peak in the large size range, as well as one in the very small range; this is because larvae,

Collection Collection Brachial valves

Width (mms.)

TEXT-FIG. 2. Size-frequency distributions. For localities see Appendix.

once established, may have very low mortality rates until mature. The Eocoelia
assemblages (text-fig. 2) have modal peaks in the larger size ranges ; there is no peak in
the small range and it must be concluded that, owing to their very small size, the victims
of high infant mortality were removed or destroyed by currents.

Size-frequency distributions may be used to better advantage in another way. Since
the brachial valve of Eocoelia is flat, or nearly flat, and the pedicle valve quite convex.

A. M. ZIEGLER: SILURIAN BRACHIOPOD EOCOELIA HEMISPHAERICA 527

and since it is known that sphericity is an important hydrodynamic factor (Menard
and Boucot 1951, p. 150), it is evident that the mean sizes of the valves would dilfer
significantly if carried any distance by a current. The size-frequency distributions of the
brachial and pedicle valves of sample 13 are compared in text-fig. 2. Although there are
more brachial than pedicle valves, the shapes of the distributions correspond very closely,
an indication that current sorting was at a minimum.

TABLE 1. The relation of the Lip and Deflection to size in three collections

LIP

Width

Collection

no. 13

Collection no. 11

Collection no. 15

(mm.)

Present Absent

Present

Absent

Present

Absent

-4-5

4

2

1

4-6-S-5

3

17

1

5

0

S-6-6-5

6

25

3

3

4

6-6-V-5

15

15

3

10

5

7-6-8-5

12

22

8

9

18

8-6-9-S

2

3

3

1

10

9-6-10-5

1

1

4

10-6-

8

DEFLECTION

-4-5

5

2

1

4-6-5-5

17

5

2

5

0

5-6-6-5

27

7

3

4

4

6-6-7-5

23

5

10

6

1

4

7-6-8-5

31

4

9

8

7

11

8-6-9-5

5

1

4

1

8

3

9-6-10-5

1

2

5

10-6-

3

2

The examination of maturity indices may give useful information as to the relative
ages represented in the samples. Two such indices, the Tip’ and ‘deflection’ (text -fig. 1)
are present, at least in the two earlier species, Eocoelia hemisphaerica and E. intermedia.
The relationship of lip and deflection to size in several samples is shown in table 1 ;
the specimens with these structures tend to be larger, but the dilference in size is by no
means great and, in fact, in only the lip distribution of sample 13 is the association
significant (x^ = 8-87, p < 0-003). If, as seems likely, the lip is the final stage in the
ontogeny of the animal, then several conclusions follow; (1) the wide size range ex-
hibited by specimens with lips indicates that some individuals reached maturity at a
considerably smaller size than others; (2) the large number of specimens with lips shows
that many specimens actually lived to maturity; and (3) the fact that the specimens
without lips are nearly as large as the ones with lips suggests that the former had nearly
attained full maturity.

Unlike the lip, the deflection was formed some time before the animal finished growing.
The distance from the umbo at which the deflection occurred {Ifl may be compared with
the overall length of the valve (/) in the ratio Iflx 100. Text-fig. 3 shows the distribution
of this statistic for sample 13. The distribution has a sharp peak showing that most of
the deflections occur at about 85% of the length of the valve. More important, there
is no apparent correlation between / and %/xlOO, i.e. smaller specimens have the

528

PALAEONTOLOGY, VOLUME 9

deflection in the same relative position as larger specimens. As with the case of the
lip, this suggests that the specimens were mostly mature at the time of their death.

In conclusion, most of the fossil beds studied are thought to be the result of some event,
such as a storm, which stirred up the bottom, concentrating the shells lying on the sea
floor, most of which were dead and disarticulated, and burying them with a layer of
sediment. There is no clear indication that current sorting seriously affected the samples,
except that the victims of high infant mortality appear to be unrepresented.

I/I X 100

TEXT-FIG. 3. Histogram showing the percentage of the length at
which the deflection occurs (/ =- length = distance from umbo
to point of deflection). Measurements were made on the brachial
valves of Collection 13.

SYSTEMATIC DESCRIPTIONS

Superfamily atrypacea
Family leptocoeliidae Boucot and Gill 1956
Genus eocoelia Nikiforova 1961

Type species. Atrvpa hemisphaerica J. de C. Sowerby 1839, in Murchison, The Silurian System, p. 637,
pi. 20, fig. 7.

Diagnosis. The valves of Eocoelia are plano-convex and effectively corrugated by
prominent radial ribs. The plane of bilateral symmetry is emphasized by a strong median
rib in the pedicle valve and a corresponding trough in the brachial valve; internally, this
trough is a positive feature and supports a low, rounded median ridge. The articulation
mechanism is developed so that, in addition to the normal tooth-socket relationship,
the crural plates fit into well-defined grooves, or fossettes, on the internal surfaces of the
teeth (text-fig. 4).

Comparison. Leplocoelia and Australocoelia, the two other genera of the Leptocoeliidae,
always possess well-developed cardinal processes, whereas this structure, if developed
at all in Eocoelia, is usually no more than a low inconspicuous ridge. Also, Leptocoelia
and Australocoelia possess a sharply defined fold and sulcus involving two ribs of the
brachial valve and one of the pedicle valve; the valves of Eocoelia are usually gently

A. M. ZIEGLER: SILURIAN BRACHIOPOD EOCOELIA HEMISPHAERICA 529

warped into a fold and sulcus, but this structure is not distinct enough to be defined in
terms of number of ribs involved.

Discussion. Shape. The brachial valve is usually flat, but may be gently concave or
gently convex. The pedicle valve is always convex and in some cases approaches a
hemispherical shape. The width is slightly greater than the length. The beak of the pedicle
valve is small and overhangs the brachial valve by approximately 0-2 mm. or 0-3 mm. in
specimens about 1 cm. long. Young shells tend to be equidimensional but become
slightly wider with respect to their length in their older growth stages.

TABLE 2. Statistics of the allometry eurves for the variates length (/) and ^vidth (vv)

For a discussion of the parameters a, Ug. b, and r see Kermack and Haldane 1950, p. 40,

or Imbrie 1956, p. 228

Species of
Eocoelia

Collection

no.

N

a

h

r

II

X

E. sulcata

5

10

0-961

0-060

0-978

0-964

90-57

E. curtisi

9

8

1-113

0-053

0-719

0-977

89-71

3

17

0-924

0-094

0-965

0-850

83-29

E. intermedia

15

16

0-817

0-042

1-218

0-972

85-29

6

21

0-927

0-053

0-916

0-941

79-57

E. hemisphaeriea

7

55

0-936

0-038

0-948

0-922

83-71

11

27

1-060

0-050

0-740

0-931

83-14

13

45

0-917

0-041

0-991

0-926

84-29

14

11

0-851

0-100

1-159

0-890

86-71

The posterior margins form an obtuse angle at the umbo which varies from 165° to
145° as measured on the brachial valve. The delthyrium is small (from 0-2 to 0-4 mm.
wide in adult specimens) and is uncomplicated by deltidial plates. The cardinal ex-
tremities are rounded. Interareas are not developed in Eocoelia.

Eoeoelia possesses structures caused by a change in the direction of growth of the
anterior margin of the shell. In one case, anterior growth ceased completely and the
valves grew directly toward one another to form a ‘lip’ around the margin of each valve.
This structure varies in height up to 0-32 mm., values of OT mm. and 0-2 mm. being
common (text-fig. 1). In another case, an inward ‘deflection’ occurs in the direction of
growth of each valve (text-fig. 1). Abundant examples of the lip and deflection are to
be found in the early populations of Eocoelia, that is, those that are grouped with
E. heinisphaerica and E. intermedia, but the structures are rarely developed in the later
populations, that is, E. curtisi and E. sulcata. The disappearance of these structures was
gradual and is thought to represent an evolutionary trend.

There is usually a weak sulcus in the brachial valve of Eocoelia and a corresponding
fold in the pedicle valve.

Ribs. Eocoelia possesses unbranching ribs which should be considered both from the
standpoint of strength and number. The early species, E. hemisphaeriea, has strong,
rather angular ribs whereas the last species, E. sulcata, has ribs which are present only
as vestiges and can be seen just in the area around the umbo, that is, in the young growth
stages. The species which are intermediate stratigraphically have ribs of intermediate
strength (PI. 83).

530

PALAEONTOLOGY, VOLUME 9

The modal rib number varies considerably in Eocoelia', however, this feature does not
appear to behave consistently with time (Tables 3, 4). The early species, E. hemisphae-
rica, has a moderate number of ribs with populations showing modes of 14 and 16 ribs.
E. intermedia tends to have more ribs, the modes being 16 and 18. E. curtisi and E.
sideato have about 12 and 10 respectively, though the decrease in these cases is due as
much to the progressive fading of ribs in the postero-lateral regions of the shell as to an
increase in rib spacing.

TABLE 3. Rib Number

Species of

Collection

Eocoelia

no.

N

7-8

9-10

11-12

13-14

15-16

17-18

E. sulcata

8

5

2

2

1

2

3

1

1

1

4

5

2

1

1

1

5

21

3

7

9

2

E. curtisi

9

7

2

4

1

10

8

3

2

2

1

3

25

1

17

6

1

E. intermedia

15

18

1

12

4

1

6

15

1

8

5

E. hemisphaerica

7

81

2

37

36

6

11

8

2

5

1

13

79

9

52

16

2

14

31

9

19

3

12

11

2

7

2

*

8

7

1

* Material from Williams’s Cj locality in the Llandovery District (1951, p. 114).

Articulation. Like most atrypids, the early species E. hemisphaerica possesses, in the
brachial valve, a pair of blade-like hinge plates medianly bounding the sockets, and, in
the pedicle valve, a pair of dental lamellae supporting teeth which are connected to the
posterior margin of the shell by planareas. Lateral to the dental lamellae are umbonal
chambers. However, in the later populations, these are not present and the teeth and
dental lamellae are fused into one large tooth-like mass. The intermediate stage of the
disappearance of these chambers is the characteristic feature of E. intermedia and they
are never present in later populations, that is, E. curtisi and E. sulcata (PI. 84).

The inner surfaces of the teeth of Eocoelia each have a groove, or fossette, into which
fit the laterally directed crural plates of the brachial valve. Thus Eocoelia has an inter-
locking articulation mechanism which consists of the normal tooth-socket relationship

EXPLANATION OF PLATE 83

All specimens are rubber casts of external moulds of brachial valves, x 4. For localities see Appendix.
Figs. 1-6. Eocoelia sulcata. 1, 2, Collection 1, USNM 140431, 140432; 3, 4. Collection 8, OUM
C6457a, C6454a; 5, 6, Collection 2, USNM 140433, 140434.

Figs. 7-8. E. curtisi. Collection 3, OUM C3198Z), C3225a.

Figs. 9-12. E. intermedia. 9, 10, Collection 15, OUM C8050Z>, C8066(>; Collection 6, OUM C1644<7,
C1669fl.

Figs 13-16. E. hemisphaerica. 13, 14, Collection 7, OUM C1095(), C1012n; Collection 13, OUM
C6857Z?, C6810/5.

Palaeontology, Vol. 9

PLATE 83

ZIEGLER, Silurian atrypacean brachiopod Eocoelia

A. M. ZIEGLER: SILURIAN BRACHIOPOD EOCOELIA HEMISPHAERICA 531

as well as the crural plate-fossette relationship. The fossettes, which are present in all
species of EocoeUa, are more deeply impressed in the later species.

Lophophore support. Presumably the lophophore was attached to the anterior portions

TABLE 4. Statistics of 69, where 69 = angle enclosed by six ribs

Lower

Upper

Species of

Collection

confidence

confidence Standard Observed

Eocoelia

no.

N

limit

Mean limit

deviation

range

E. sulcata

8

3

44-67

60-67 70-67

6-44

56-68

2

4

54-44

61-50 68-56

4-44

58-68

4

3

40-12

60-00 79-88

8-00

52-68

5

12

58-16

61-33 64-50

4-50

50-66

E. curtisi

9

11

56-15

58-18 60-21

3-03

52-64

10

12

57-68

60-50 63-32

4-44

56-70

3

33

57-70

59-27 60-84

4-44

50-68

E. intermedia

15

47

55-55

56-72 57-89

3-94

48-66

6

42

51-34

52-67 54-00

4-30

46-60

E. hemisphaerica

7

114

55-85

56-88 57-91

5-56

40-70

11

50

57-61

59-08 60-55

5-19

48-70

13

94

57-88

59-17 60-46

6-29

46-74

14

29

57-74

60-48 63-22

5-85

52-76

12

14

58-65

61-57 64-49

5-06

52-78

TABLE 5.

Statistics of the

size {width)-frequeney distributions

Collection

no.

N

Number of
fossil beds

Mean Observed Description Oj

width (mm.) range (mm.) distribution

8

6

1

6-02

4-7-7-2

. —

2

9

many

6-23

5 -6-7 -3

—

4

7

2 or 3

7-03

5-9-8-0

■ — ■

5

23

1

5-71

2-8-7-4

skew, left

9

19

1

5-98

2-8-9-0

unimodal

10

15

several

6-68

3 -0-8 -6

skew, left

3

45

2

6-16

3 -4-8 -8

bimodal

15

56

1

8-63

3-9-12-6

skew, right

6

47

several

8-29

5-5-13-4

skew, right

7

129

many

8-87

4-2-13-8

skew, left

11

56

1

7-06

4-2-9-0

skew, left

13

141

2

6-72

3-7-10-2

bimodal

14

38

2 or 3

7-16

3-3-10-6

irregular

12

19

1

6-74

4-5-9-4

unimodal

of the hinge plates. The existence of spires has been investigated by Nikiforova (Niki-
forova and Andreeva 1961, p. 253), who sectioned many specimens without finding
spires, although they were present in other atrypids in the same deposit. As has been
mentioned, the crural plates served in the role of articulation as well as for lophophore
support. They are rather blade-like and curved laterally. The bases of the crural plates
bound the notothyrial cavity and diverge anteriorly and ventrally.

Muscle attachment. In the brachial valve the adductor scars are faintly impressed
immediately anterior to the crural plates and on either side of a low, rounded median

532

PALAEONTOLOGY, VOLUME 9

TEXT-FIG. 4. Changes in the posterior end of the pedicle valve. All specimens are internal moulds and
all are x 4, with the exception of the line drawings, which are x 8. All the specimens are in the Oxford
University Museum collections. For localities see Appendix. Top row, Collection 2, Eocoelia sulcata;
1, C3734fl. Second row. Collection 3, E. curtisi; 2, C3240/b; 3, C3241a (a diagram of the latter is shown
in the bottom right of the figure). Third row. Collection 15, E. intermedia; 4, C8169fl; 5, C8167a;
6, C8168rt. Fourth row. Collection 6, E. intermedia; 7, C\l\9a; 8, Cl 702^; 9, C\699b. Bottom row.
Collection 13, E. hemisphaerica; 10, C6937b; and ) 1, C7004a (a diagram of the former is shown in the

bottom left of the figure).

A. M. ZIEGLER: SILURIAN BRACHIOPOD EOCOELIA HEMISPHAERICA 533

ridge. The cardinal process (PI. 84, fig. 5) is rarely distinct and not developed in all
specimens; it is usually not more than a low ridge running along the notothyrial cavity.

In the pedicle valve, the paired diductor scars are oval with their long axes parallel to
that of the valve and are sometimes impressed to such a depth that a distinct and very
narrow ridge, or myophragm, is left between them (PI. 84, fig. 4). Small elongate
adductor muscle scars are often seen postero-medianly to the diductor scars.

Size. Eocoelia is typically small, shell widths usually being under 10 mm.; the largest
recorded width is 13-4 mm. The size varies considerably from collection to collection
(Table 5).

Species

Atrypa hemisphaerica J. de C. Sowerby, 1839, in Murchison, The Silurian System, p. 637, pi. 20, fig. 7.
Leptocoelia intermedia Hall, 1860, p. 147, fig. 5.

Eocoelia curtisi sp. nov.

Coelospira sulcata Prouty 1923, p. 466, pi. 27, figs. 6-8.

Species questionably assigned

Atrypina{7) paraguayensis Harrington 1950, p. 62, pi. 1, figs. 9, 10, 13-16.

Eocoelia hemisphaerica (J. de C. Sowerby)

Plate 83, figs. 13-16; Plate 84, figs. 5-11

Atrypa hemisphaerica J. de C. Sowerby, in Murchison 1839, p. 637, pi. 20, fig. 7.

Atrypal hemisphaerica J. de C. Sowerby; Davidson 1866, p. 136, pi. 13, fig. 23.

Coelospira hemisphaerica sefinensis Williams 1951, p. 113, pi. 5, figs. 19, 20.

Eocoelia quebecensis Amos and Boucot 1963, p. 447, pi. 62, figs. 1-10.

Definition. Included in this species are all populations of Eocoelia which consist entirely
of individuals possessing umbonal cavities lateral to the dental lamellae in the pedicle
valve.

Discussion. The species is strongly ribbed; the index of rib strength, mean hjw,x 100
(where h, = height of rib and m,. = width of rib), for the populations identified with
this species is about 40% (text-fig. 5, table 6). The individuals commonly have from
13 to 16 ribs, but may have as few as 1 1 or as many as 18 ribs. It should be mentioned
that E. hemisphaerica sefinensis Williams, here synonymized with E. hemisphaerica, was
originally said to have 12 ribs but larger collections now available clearly demonstrate
a modal rib number of 14, the same as E. hemisphaerica', in the present work, rib counts
were made only on external moulds of complete brachial valves, and failure by Williams
to adhere to these standards would account for his low values.

A particularly distinctive feature of populations of E. hemisphaerica is the presence in
many individuals of the ‘lip’ and ‘deflection’. The lip usually occurs in about 30% of
the specimens; the occurrence of the deflection is more variable and may be observed in
as many as 87-5% of the individuals or as few as 15%, depending on the population
(Table 7). A lip and a deflection may occur on the same specimen and occasionally two
deflections are present on a single specimen. Both the lip and deflection are probably
indications of maturity in Eocoelia', the lip would have been the final portion of the
shell to grow, and the deflection usually occurs at a late growth stage of the animal,
at about 85% of the shell length as measured from the umbo (text-fig. 2).

Holotype. The specimen preserved in the Geological Survey Museum (Geol. Soc. Coll. GSd 4829)
is thought to be the holotype, although the artist’s drawing in The Silurian System (pi. 20, fig. 7)

534

PALAEONTOLOGY, VOLUME 9

by no means confirms this. The type locality of the specimen is Ankerdine Hill, seven miles west of
Worcester. The species is very abundant in this small inlier of Upper Llandovery rocks, but E. ciirtisi
also occurs on the west side of the hill between the Knightwick Sanatorium and the River Teme.

Col lection
No.

Stratigraphic

Successions

13

14

1 1

12

^rfuoUy

smooth

E3-

confidence interval

pwj

/77eg/T4:f*| ■! I
observed range

1 Tortworth Inlier

2. Shropshire ^

J- May Hill Inlier

4. Old Storridge Common

5 Ankerdme HU I

{E3-

4 5

0 10% 20% 30% 40% 50% 60%

hf/ Wr X too at 3 5mm. from umbo, where
h|. = heightof rib, Wf= width of rib

TEXT-FIG. 5. The range of variation in rib strength. The collections have been
arranged in chronological order by palaeontological and stratigraphical evi-
dence. The latter is summarized at the right of the figure. For localities see

Appendix.

Occurrence. In the southern part of the Welsh Borderland, E. hemisphcierica occurs in
the Cowleigh Park Beds of Ankerdine Hill and the Old Storridge Common area, both
to the north of the Malvern Hills, and in the Huntley Hill Beds of the May Hill Inlier. In
Shropshire the species occurs in the Pentamerus beds at Ticklerton, near Wenlock Edge^
and at Hillend, near the Longmynd. There is a small inlier of Upper Llandovery sandstone
immediately south of Presteigne, Radnorshire, where the species occurs in abundance.

A. M. ZIEGLER: SILURIAN BRACHIOPOD EOCOELIA HEMISPHAERICA 535

At Llandovery E. hemisphaerica occurs in the Sefin shales, where it has been identified
as E. hemisphaerica sefiuensis Williams (1951, p. 113); the species also occurs just to the
south of this area in the bottom ten feet of the Upper Llandovery beds of the Sawdde
River section.

TABLE 6. Statistics of the rib proportions hJw,.X 100, where h^ is the height of a rib at
3-5 mm. from the umbo and w^ is the width of the rib at 3-5 mm. from the umbo

Lower

Upper

Species of

Collection

confidence

confidence

Standard

Observed

Eocoelia

no.

N

limit

Mean

limit

deviation

range

E. sulcata

8

3

000

7-77

17-55

3-94

4-12

2

3

2-01

9-87

17-73

3-16

7-13

3

10

11-53

13-71

15-89

3-05

10-21

E. ciirtisi

9

6

12-98

18-84

24-70

5-58

12-28

10

6

16-23

20-85

25-47

4-40

14-25

3

15

20-23

22-37

24-51

3-87

14-28

E. intermedia

15

20

24-59

27-35

30-11

5-90

15-41

6

13

28-14

33-82

39-50

9-40

20-50

E. hemisphaerica

7

25

37-74

41-72

45-70

9-65

21-55

11

3

40-19

48-40

56-61

3-30

46-52

13

39

39-39

41-97

44-55

7-94

27-62

14

12

35-69

40-48

45-27

7-55

27-53

12

5

39-11

45-56

52-01

5-20

40-54

TABLE 7. Occurrence

of the Lip and Deflection

Species of Collection

Eocoelia no.

N

Lip

present

(%)

N

Deflection

present

(%)

E. sulcata 2

10

0

1

0

5

21

0

24

0

E. curtisi 9

2~>

0

21

0

10

21

0

19

0

3

55

1-8

57

0

E. intermedia 1 5

61

3-3

56

39-3

6

44

11-4

46

2-2

E. hemisphaerica 7

42

28-8

40

87-5

11

60

30-0

71

52-1

13

139

30-9

145

77-9

14

38

36-8

47

14-9

12

11

27-3

24

16-7

The species is known from the Pointe-aux-Trembles Formation of the Lake Temis-
couata-Lake Touladi region of eastern Quebec where it has previously been identified
as E. quebecensis Amos and Boucot (1963, p. 448). At Arisaig, Nova Scotia, E. hemi-
sphaerica occurs in the lowest shell beds of the Ross Brook Formation at a point on the
shore about 1000 ft. north-east of the mouth of Arisaig Brook.

The Atrypina{l) paraguayensis of Harrington (1950, p. 62) from beds of probable
Lower Llandovery age (Wolfart 1961, p. 53) cannot at present be distinguished from
E. hemisphaerica.

536

PALAEONTOLOGY, VOLUME 9

Stratigraphic range. At the type area of Llandovery, E. hemisphaerica occurs in the Sefin
shales, an impersistent stratigraphic unit, which overlies beds of typical lithology
and which was grouped with the Q beds by Williams (1951, p. 130). The Sefin shales
contain graptolites of the zone of Monograptus sedgwicki (Jones 1949, p. 62) which are
of early Upper Llandovery age.

The Paraguay specimens occur in beds which have been correlated with the Lower
Llandovery on the basis of graptolite and trilobite evidence (Wolfart 1961, p. 53); they
are the earliest known of the genus Eocoelia. No other Lower or Middle Llandovery
occurrences are known.

The upper limit of E. hemisphaerica must be prior to the C4 beds of Llandovery
which contain E. intermedia.

Eocoelia intermedia (Hall)

Plate 83, figs. 9-12; Plate 84, figs. 1-2
Leptocoelia intermedia Hall 1860, p. 147, fig. 5.

Definition. This species includes all populations that contain individuals both with and
without umbonal chambers lateral to the dental lamellae of the pedicle valve; the
umbonal chambers when present are typically reduced in size by comparison with the
earlier species, E. hemisphaerica.

Discussion. The species is distinctly ribbed, with the two populations measured showing
values for the mean hjw.x 100 30%. The values measured are significantly different

from the larger populations of E. hemisphaerica measured, which are about 40%. The
populations identified as E. intermedia show considerable variation in rib number ; sample
no. 6 has an unusually high mode of 16 ribs with some specimens showing as many as
20 ribs, while sample no. 15 has somewhat fewer. The ribs of E. intermedia show a
slight tendency to fade out and disappear in the postero-lateral parts of the shell.

The lip and deflection are developed in only a few specimens.

Holotype. The type specimen of Hall’s Leptocoelia intermedia, now preserved in the American Museum
of Natural History, no. 1634, is similar to the Welsh Borderland collections, nos. 6 and 15, in rib
strength and in the very reduced umbonal cavities (PI. 84, figs. 1-2). Hall’s specimen came from Arisaig,
Nova Scotia, where Eocoelia occurs at many horizons in the Ross Brook Formation. All four species

EXPLANATION OF PLATE 84

For localities of Figs. 5-17 see Appendix.

Figs. 1-2. Eocoeliaintermedia(Wa\\\^()t),y). 144,fig.5). Holotype, Arisaig, Nova Scotia. AMNH 16341:0,
pedicle valve; I, posterior view of internal mould; 2, internal mould.

Figs. 3-4. E. sulcata, Arisaig, Nova Scotia. AMNH 16341 : 1 pedicle valve; 3, posterior view of internal
mould; 4, internal mould.

Figs. 5-11. E. hemisphaerica. Collection 13. 5, OUM C6849n, internal mould of brachial valve; note
cardinal process. 6-8, OUM C6857a, brachial valve; 6, internal mould; 7, rubber cast of internal
mould; 8, oblique view of rubber cast of internal mould. 9-11, OUM C6966a, pedicle valve; 9,
internal mould; 10, rubber cast of internal mould; 1 1, oblique view of rubber cast of internal mould.

Figs. 12-17. E. cnrtisi. Collection 3. 12-14, OUM C3189^>, brachial valve; 12, internal mould; 13,
rubber cast of internal mould; 14, oblique view of rubber cast of internal mould. 15-17, Holotype,
OUM C3241n, pedicle valve; 15, internal mould; 16, rubber cast of internal mould; 17, oblique view
of rubber cast of internal mould.

Palaeontology , Vol. 9

PLATE 84

ZIEGLER, Silurian atrypacean brachiopod Eocoelia

A. M. ZIEGLER: SILURIAN BRACHIOPOD EOCOELIA HEMISPHAERICA 537

of Eocoelia occur in this sequence; included with Hall's type specimen in the American Museum of
Natural History are some specimens with extremely weak ribs which probably came from a higher
horizon than the type (PI. 84, figs. 3-4).

Occurrence. In the Welsh Borderland E. intermedia occurs in the Huntley Hill Beds of
May Hill, as does E. hemisphaerica, and in the Pentamerus Beds of Norbury, Shropshire.
At Llandovery the species occurs in beds mapped by Jones (1925) as C4 in an old quarry
just west of the road from the village of Myddfai to the farm Gorllwyn-fawr, at the
point adjacent to the farm Gorllwyn-fach.

Stratigraphic range. As stated, at Llandovery, the species occurs in beds mapped as C4
(Jones 1925). The upper and lower ranges cannot be determined rigorously because
sufficient specimens of Eocoelia have not yet been found in C3 or C5 at Llandovery.
Nor can the superposition of E. intermedia on E. hemisphaerica be rigorously demon-
strated; however, mapping evidence at May Hill and Llandovery suggests this relation-
ship.

Eocoelia curtisi sp. nov.

Plate 83, figs. 7-8; Plate 84, figs. 12-17

Atrypal hemisphaerica J. de C. Sowerby; Davidson 1866, p. 136, pi. 13, figs. 24u, b, c.

Definition. This species is defined to include populations which show no sign of umbonal
chambers or dental lamellae in the pedicle valve, but which do have distinct ribs.

Discussion. The teeth of this species are broadly based on the floor of the pedicle valve
instead of being supported by dental lamellae as is the case with E. hemisphaerica
and some specimens of E. intermedia. Fossettes (grooves on the median sides of the
teeth) are more deeply incised in E. curtisi than in earlier species. The fossettes are a
type of socket into which fit the laterally directed hinge plates of the brachial valve.

The ribs of E. curtisi are weaker than in earlier species, the value of mean h,lw,x 100
being about 20% for the populations measured. The differences with earlier popula-
tions are statistically significant in some cases. Rib number varies about a mode of 12;
this decrease with respect to earlier populations is due both to a wider spacing of ribs
and to the progressive disappearance of ribs in the postero-lateral regions of the shell.
The lip and deflection are virtually absent in E. curtisi.

Holotype. Oxford University Museum specimen no. C3241 from collection 2, Charfield Green,
Tortworth Inlier (PI. 84, figs. 15, 16, 17).

Occurrence. The species is particularly abundant in the Damery Beds of the Tortworth
Inlier and may be collected at Charfield Green and Damery Quarry, localities that were
frequented by the early workers. At May Hill a few specimens, probably E. curtisi, have
been found in the Yartleton Beds. E. curtisi occurs in the basal Wyche Beds both at
Gullet Quarry, Malvern Hills district, and at the Gunwich Mill site. Old Storridge
Common. A specimen, probably of this type, has been discovered in C5 beds at Llan-
dovery, 500 yards north of the farm Llwyn-Meredith in a stream section. In Ireland
E. curtisi is abundant in the lowest beds of the Annelid Grit, north-west Co. Galway.

In North America the species has been identified from Arisaig, Nova Scotia, where it
occurs at the mouth of Arisaig Brook.

538

PALAEONTOLOGY, VOLUME 9

Stratigraphic range. E. curtisi probably does not range higher or lower than the C5 beds
of Llandovery. E. intermedia occurs in C4 at Llandovery, and in the southern part of the
Welsh Borderland E. sulcata occurs in the top part of the Llandovery sequences, that is,
in beds that are probably equivalent to the unfossiliferous Cg shales of the type area.

Eocoelia sulcata (Prouty 1923)

Plate 83, figs. 1-6; Plate 84, figs. 3-4

Coelospira sulcata Prouty, in Prouty and Swartz 1923, p. 466, pi. 27, figs. 6-8.

Eocoelia sulcata (Prouty), Boucot et al. (in press).

Definition. Populations identified with this species have vestigial ribs, that is, ribs that
are present only in the young growth stages and are not apparent at the anterior com-
missure of the average-sized specimen.

Discussion. The index of rib strength (mean lijw^x 100) is very low, the values being
about 10%, as compared with those of E. curtisi, which are about 20%.

E. sulcata is similar to E. curtisi as far as the structures of the beak and of the pedicle
valve are concerned; that is, umbonal chambers and dental lamellae are absent.

Holotype. Prouty ’s specimens from the Rochester Formation and the upper part of the Rose Hill
Formation of Maryland are the types of this species. It may eventually prove necessary to designate
new types as Prouty ’s specimens cannot be located at present (G. A. Cooper, personal communication).

Occurrence. At Tortworth E. sulcata occurs in the Tortworth Beds, that is, the top of
the Llandovery section, and in beds mapped as Wenlock (Curtis 1955, p. 4). The species
occurs in the upper part of the Yartleton Beds of May Hill and in the upper part of the
Wyche Beds near Old Storridge Common. In Norway, the species is abundant in Stage
^a-b at Gjettum Station, Baerum, Oslo-Asker District. Apart from the type localities
in Maryland, the species also occurs in North America in the Long Reach Formation
of New Brunswick (Boucot et al., in press).

Stratigraphic Range. The superposition of beds containing E. sulcata on E. curtisi may
be demonstrated at Tortworth and Old Storridge Common. Since E. curtisi is of C5 age,
E. sulcata probably originated about Cg. Eocoelia may have become extinct about the
middle of the Wenlock as there is no record of it in the upper Wenlock or Ludlow.
Amos and Boucot would derive the Devonian brachiopod Leptocoelia from Eocoelia
(1963, p. 443), but the later species of Eocoelia show trends in the loss of ribs that make
this connexion unlikely. Leptocoelia may have been derived from one of the early
species of Eocoelia, which are similar in several respects, but there is no direct stratigraphic
evidence for this.

PALAEOECOLOGY

Eocoelia is a widespread element in Upper Llandovery and low Wenlock faunas, but
it is only really abundant at certain localities. At these localities Eocoelia dominates the
assemblage to such an extent that the expression Eocoelia Community is appropriate
(Ziegler 1965). The Eocoelia Community occurs in the Cowleigh Park Beds of Ankerdine
Hill and Old Storridge Common where it is of C1-C2 age, in the Huntley Hill Beds of

A. M. ZIEGLER: SILURIAN BRACHIOPOD EOCOELIA HEMISPHAERICA 539

May Hill where it is of C1-C3 age, and in the Damery Beds of Tortworth where it is of
C5 age. Each of these stratigraphic units is a basal unit deposited by the transgressing
Silurian sea and in each of these areas the Eocoelia Community is succeeded by a
community dominated by Pentameroides sp. At Presteigne the Eocoelia Community is
well developed in beds of C1-C2 age and is succeeded by a Community dominated by
Pentamerus sp. The Eocoelia Community occurs in both the Conglomerate Series of
Marloes Bay, Pembrokeshire, and the Annelid Grit of Co. Galway, Ireland; in each of
these places it is associated with basal beds deposited by a transgressing sea.

Owing to its association with transgressive seas, and often with relatively coarse clastic
rocks, the Eocoelia Community is thought to represent a coastal environment. It was
not the only coastal community, however, as the basal Cowleigh Park Beds of the
Malvern District contain a very restricted fauna dominated by Lingula pseudoparallela
Stubblefield, ‘ Camarotoecliia' decemplicata (J. de C. Sowerby), some bivalves and
gastropods, and little else.

INTERPRETATIONS OF THE TRENDS AND VARIATIONS

The most easily understood trend in Eocoelia is the gradual strengthening of the
articulatory mechanism. Individuals with deeper fossettes, stronger hinge plates, and
larger, more broadly based teeth would have had a selective advantage, particularly in
the relatively rough coastal environment inferred for this brachiopod. Trends toward
increased articulation have been noticed in other brachiopods; for instance, several
groups of stropheodontids independently display increases in the number of denticles
along their hinge lines with time (Williams 19536).

The reason behind the rather striking decline in rib strength is less obvious. The ribs
may have originally helped in the protection of the animal by their interlocking effect.
If this were so, then the development of the articulatory mechanism might well have
made the ribs redundant.

Other features of Eocoelia disappear during the time of the development of the
articulatory mechanism; both lip and deflection are unknown in the later species. The
formations of these structures, seen particularly well in populations of E. hemisphaerica,
involved changes in the direction of growth at the anterior commissure with consequent
demands for rapid reorientation of the teeth in the sockets. With time, the deepening of
the fossettes to produce more of an interlocking articulation may have made this
reorientation difficult and eventually impossible.

There is considerable variation between collections in both the mean size and relative
numbers of Eocoelia. Moreover, the two factors seem to be related in a positive way,
that is, when Eocoelia occurs in large numbers, as in the Eocoelia Community, the
individuals tend to be large; but when it occurs sporadically the individuals are always
small. The fact that size and numbers are related suggests that they were originally
controlled by the general favourability of the environment for Eocoelia.

Finally, the reasons behind the variation in the spacing of the ribs are not at present
understood. It cannot yet be established whether or not variation exists between popula-
tions at one horizon. At best, it can be said that the rib spacing appears to vary in a
random way in a given stratigraphic section.

o o

c 4397

540

PALAEONTOLOGY, VOLUME 9

UPPER LLANDOVERY CORRELATION

The detailed correlation of strata should be based on organisms whose evolution is
well understood. The mere linking of various species does not definitely establish an
evolving lineage. It must be shown that all gradations from one species to the next

Eocoelia Stricklandiids

WENLOCK

U

P

P

Eoc

su/

^

oetia Costisti

cata Urata

' \

'ickiandia

typica

'

E

^ C

Eoco

cur

^

- 7

Costisti
*isi lirati

\

ncktandia
7 alpha

>

L

L

A

/

Eocc

infer

\ )

Stric

tens

letia

media

\

kian dia
ultima

N

^ C

0 ^2-3

V

E

y

/

r -7

/

\

R

^ c,

Eoco

hemiS)

eiia

onoenca

lens pi

ktandia

'ogressa

>

)

TEXT-FIG. 6. Tentative correlation of the evolving lineages.

exist; this is manifestly the case with Eocoelia, where the sampling is such that the prob-
lem is more in dividing up a completely intergrading set of populations than in linking
together distinct species. An important check on an evolutionary trend is to detect its
occurrence in widely separated areas ; all the species of Eocoelia are known from North
America and Britain and their sequence can be established on both sides of the Atlantic.
Finally, evolutionary trends in unrelated brachiopods may be used as a check on each
other; trends in the Strieklandia-Costistricklandia lineage (St. Joseph 1935, p. 421;
Williams 1951, p. 98; Boucot and Ehlers 1963, p. 49) are useful in this respect.

The correlation of the various evolving lineages with the lithologic zones mapped by

A. M. ZIEGLER: SILURIAN BRACHIOPOD EOCOELIA HEMISPHAERICA 541

Jones (1925, 1949) at Llandovery may be established from occurrences of the species
both in the type area and in surrounding areas (text-fig. 6). At Llandovery Eocoelia
hemisphaerica and Stricklandia lens progressa Williams occur in beds mapped as Ci.
The C2-C3 beds are largely unfossiliferous and only one brachial valve of Eocoelia, not
enough for specific identification, has been found. In C4 beds Eocoelia intermedia and
Stricklandia lens ultima Williams occur together. Both lineages are represented, though
somewhat unsatisfactorily, in C5 beds; one pedicle valve of Eocoelia, probably E.
curtisi, has been found, and stricklandiids occur, but are insufficiently complete to tell
whether weak ribs are present. However, in an adjacent area to the south, the Sawdde
River section near Llangadock, beds mapped as C4-C5 (Williams 1953n) contain abun-
dant Eocoelia curtisi and Costistricklandia lirata alpha (St. Joseph). The Cg beds of
Llandovery are unfossiliferous and are thought to be equivalent to beds in the Welsh
Borderland, such as at May Hill or Old Storridge Common, that contain Eocoelia
sulcata and Costistricklandia lirata typica (St. Joseph); these beds occur just below the
base of the Woolhope Limestone, the basal horizon of the Wenlock.

APPENDIX

Description of Localities

Collection

Field Grid

Species Oj

no.

no. reference

Exposure

Location

Horizon

Eocoelia

TORTWORTH INLIER

I

T-W-B ST/

hill-side

Little Daniel’s Wood,

Wenlock

E. sulcata

6947 9376

exposure

750 yds. N. 78° W. of Old
Court farm, Tortworth

Series

2

T-D-A ST/

loose blocks

About 150 yds. N. 60° E.

Base of

E. sulcata

6962 9390

in stream

of SW. corner of Daniel 's

Tortworth

bed

Wood near Tortworth

Beds

3

T-M-A ST/

stream

270 yds. N. 10° W. of Hill-

Lower

E. curtisi

7268 9212

section

house farm, Charfield

Damery Beds

MAY HILL INLIER

4

M-S-C SO/

stream

380 yds. N. 71° W. of Hay

Upper

E. sulcata

6936 2271

section

Farm on N. side of May

Yartleton

Hill

Beds

5

M-S-B SO/

stream

430 yds. N. 81° W. of Hay

Upper

E. sulcata

6932 2270

section

Farm on N. side of May

Yartleton

Hill

Beds

6

M-H-A SO/

track-side

Just W. of old quarry

Huntley

E. intermedia

7047 1811

exposure

on E. side of Nottswood
Hill, 760 yds. S. 62° W.
of Hinders

Hill Beds

7

M-N-A SO/

loose blocks

Newent Wood, 400 yds.

Huntley

E. hemisphaerica

7014 2104

from old

N. 45° W. of Folly Farm

Hill Beds

quarries

on SE. side of May Hill

OLD STORRIDGE COMMON

8

H-M-C SO/

stream

W. bank of Leigh Brook,

Upper

E. sulcata

7405 5167

section

210 yds. N. 41° E. of Mouse-

Wyche

hole Bridge (SW. of Alfrick
Pound)

Beds

9

H-L-C SO/

sunk track

Coneygore Coppice, 690

Wyche Beds

E. curtisi

7464 5108

exposure

yds. N 30 ’ W. of The Beck

542

PALAEONTOLOGY, VOLUME 9

Collection Field Grid

Species of

no. no. reference

Exposure

Location

Horizon

Eocoelia

10 H-M-B SO/

stream

W. bank of Leigh Brook,

Basal

E. curtisi

7430 5152

section

440 yds. E. of Mousehole
Bridge (SW. of Alfrick
Pound)

Wyche Beds

11 H-F-C SO/

stream

Small stream 950 yds.

Top of Cow-

E. hemisphaerica

7444 5124

section

N. 36° W. of The Beck

leigh Park
Beds

12 H-L-A SO/

track

Coneygore Coppice,

Cowleigh

E. hemisphaerica

7467 5115

exposure

740 yds. N. 25° W. of The
Beck

Park Beds

ANKERDINE HILL

13 A-Q-A SO/

disused

Near summit of Ankerdine

Cowleigh

E. hemisphaerica

7363 5630

quarry

Hill 335 yds. N. 45° E. of
church at Knightsford
Bridge

Park Beds

14 A-H-A SO/

track

Hay Wood, 200 yds.

Cowleigh

E. hemisphaerica

7376 5696

exposure

N. 45° E. of Knightwick
Sanatorium on NW. side
of Ankerdine Hill

Park Beds

NORBURY, SHROPSHIRE

15 S-N-A SO/

disused

N. of the Linley to Nor-

Pentamerus

E. intermedia

3586 9286

quarry

bury road, 280 yds. ESE.
of Linley Cottage

Beds

REFERENCES

AMOS, A. and boucot, a. j. 1963. A revision of the brachiopod family Leptocoeliidae. Palaeontology, 6,
440-57.

BOUCOT, A. J. 1953. Life and death assemblages among fossils. Am. J. Sci. 251, 25-40.

and EHLERS, G. M. 1963. Two new genera of stricklandid brachiopods. Contr. Mus. Faleont.,

Univ. Mich., E. P. Wright Mem. Vol., 47-66.

JOHNSON, J. G., HARPER, c. w., and WALMSLEY, V. G. Silurian brachiopods and gastropods of

Southern New Brunswick. Bull. geol. Surv. Can. (in press).

CRAIG, G. Y. and HALLAM, A. 1963. Size-frequency and growth-ring analyses of Mytilus edulis and
Cardium edide, and their palaeoecological significance. Palaeontology, 6, 731-50.

CURTIS, M. L. K. 1955. Lower Palaeozoic, 3-7, in Chap. I, Geology, Bristol and its adjoining counties,
ed. C. M. Machines and W. F. Whittard, Brit. Assn. Adv. Sci. Bristol.

DAVIDSON, T. 1866-71. British Fossil Brachiopoda, 3, Palaeontogr. Soc. [Monogr.l
FAGERSTROM, J. A. 1964. Fossil Communities in paleoecology ; their recognition and significance.
Bull. geol. Soc. Am. 75, 1197-1216.

HALL, J. 1860. Descriptions of new species of fossils from the Silurian rocks of Nova Scotia. Canadian
Nat. Geol. and Proc. nat. Hist. Soc. Montreal, 5, 144-59.

HARRINGTON, H. J. 1950. Geologia del Paraguay Oriental. Contrnes. dent. Fac. dent, exact, fls. nat.
Univ. B. Aires, Ser. E., Geol. 1, 1-82.

IMBRIE, J. 1956. Biometrical methods in the study of invertebrate fossils. Bull. Am. Mus. nat. Hist.
108, 213-51.

JONES, o. T. 1925. Geology of the Llandovery District, Part 1 : The Southern Area. Q. Jl. geol. Soc.
Fond. 81, 344-88.

1949. Geology of the Llandovery District, Part 2: The Northern Area. Ibid. 105, 43-64.

KERMACK, K. A. and HALDANE, J. B. s. 1950. Organic correlation and allometry. Biometrika, 37, 30^1.
MENARD, H. w. and BOUCOT, A. J. 1951. Experiments on the movement of shells by water. A?n. J. Sci.
249, 131-51.

MURCHISON, R. I. 1839. The Silurian System. London.

NIKIFOROVA, o. I. and ANDREEVA, o. N. 1961. Ordovician and Silurian stratigraphy of the Siberian
Platform and its palaeontological basis (Brachiopoda). Trudy VSEGEI, 56.

A. M. ZIEGLER: SILURIAN BRACHIOPOD EOCOELIA HEMISPHAERICA 543

PROUTY, w. F. and swartz, c. k. 1923. Brachiopoda. Maryland geol. Surv., Silurian vol., 412-67.
RUDWiCK, M. j. s. 1962. Notes on the ecology of brachiopods in New Zealand. Trans. R. Soc. N. Z.
25, 325-35.

ST. JOSEPH, J. K. s. 1935. A critical examination of Stricklandia {= Stricklandinia) lirata (J. de C. Sowerby)
1839 forma typica. Geol. Mag. 72, 401-22.

WILLIAMS, A. 1951. Llandovery brachiopods from Wales with special reference to the Llandovery
District. Q. Jl. geol. Soc. Loud. 107, 85-136.

1953fl. The Geology of the Llandeilo District, Carmarthenshire. Ibid. 108, 177-207.

1953Z>. North American and European stropheodontids : their morphology and systematics.

Me/n. geol. Soc. Am. 56.

WOLFART, R. 1961. Stratigraphie und fauna des alteren Palaozoikums (Silur, Devon) in Paraguay.
Geol. Jb. 78, 29-102.

ZIEGLER, A. M. 1963. The stratigraphical palaeontology of the Upper Llandovery rocks in the southern
part of the Welsh Borderland. Unpublished D.Phil. thesis, Oxford.

1965. Silurian marine communities and their environmental significance. Nature, Loud. 207,

270-2.

A. M. ZIEGLER

Department of Geophysical Sciences,
University of Chicago,
Chicago,

Manuscript received 24 June 1965 Illinois 60637, U.S.A.

SPONGOPHYLLIDAE FROM THE DEVONIAN
GARRA FORMATION, NEW SOUTH WALES

by D. L. STRUSZ

Abstract. The structure of rugosans belonging to the family Spongophyllidae Edwards and Haime 1873 (here
taken to include the Family Ptenophyllidae Wedekind 1923) is discussed in detail. Representatives of the Family
occurring in the Emsian (or possibly early Eifelian) Garra Formation are revised or described. The Australian
species ascribed to Acanthophyllum are reviewed, PseudochonophyUum Soshkina is fully revised, and its type (and
only) species, P. pseiidohelianthoides (Sherzer) is shown to occur only in Czechoslovakia and eastern Australia.
Grypophyllum aggregatiim Hill is placed in Lyrielasma Hill. Australophylliim Stumm is discussed and its species
are reviewed. Several of the eastern Australian Devonian faunas, previously thought to range from Coblenzian
to late Eifelian, are considered to be probably Emsian.

New species described are: Acanthophyllum {Acanthophyllum) aeneae, A. (Neostiingophyllum) implicatum,
A. {N.) turni, A. {Grypophyllum) jenkinsi, Lyrielasma! micrum, Australophylliim bilaterale.

In this paper are described those species of corals, occurring in the Emsian? Garra
Formation of New South Wales, which comprise the acanthophyllid and spongophyllid
groups. For reasons outlined below, these are for the present at least combined into the
Family Spongophyllidae.

The Garra Formation has been described in a previous paper on the disphyllids and
phacellophyllids (Strusz \965b), and in a purely stratigraphic Note (Strusz 1965a).
Further references, both geological and palaeontological, may be found therein. An
additional reference of some interest in the correlation of this formation is Philip and
Pedder (1964).

Abbreviations. The following abbreviations (fully explained in Strusz 1965Z), p. 522) are used herein :

Dc Corallite diameter
Dt Tabularium diameter
R Corallite radius

Ts Tabularial spacing (astraeoid, etc., coralla)
n Number of septa
Li Length of major septa
L2 Length of minor septa

The repositories of type and other specimens are indicated by the following prefixes to their catalogue
numbers :

AM Thin section numbers, Australian Museum, Sydney, N.S.W.

AM F Fossil numbers, Australian Museum

GSQ Geological Survey of Queensland

GSV Geological Survey of Victoria

NM National Museum, Melbourne, Victoria

SU University of Sydney Palaeontological Collection; Sydney, N.S.W.

UCT F University College of Townsville Palaeontological Collection; Townsville, Qld.

UQF University of Queensland Palaeontological Collection; Brisbane, Qld.

[Palaeontology, Vol. 9, Part 4, 1966, pp. 544-98, pis. 85-96.]

D. L. STRUSZ: SPONGOPHYLLID AE FROM DEVONIAN GARRA FORMATION 545

SYSTEMATIC PALAEONTOLOGY

Family spongophyllidae Dybowski 1873, emend. Pedder 1964

Remarks. Following Pedder (1964, p. 436), and a tentative suggestion by Birenheide
(in litt.), I place the Family Ptenophyllidae Wedekind 1923 in synonymy with this
family. The thus enlarged family has the following characters;

Septa. The septal structure is rather similar to that of the Disphyllidae. The septa
consist of fine trabeculae, in most cases arranged in single radial series, although in some
genera there may be more than one series in outer parts of the septa. The trabeculae are
arranged either broadly parallel, and inclined upwards and inwards, or in half-fans
within the dissepimentarium, and usually again in
the tabularium. This arrangement has been well
described by Birenheide (1961).

The trabeculae are monacanthine; in cross-
section they are round to square in attenuate por-
tions of a septum, becoming oval to rectangular
(the long axis perpendicular to the plane of the
septum) in dilated portions. Within a trabecula the
fibres diverge from the axis, strongly in a tangential
direction (i.e. longitudinal, and perpendicular to the
plane of the septum), but only weakly in a radial
direction (text-fig. 1). In transverse section a septum
gives the appearance of a median dark line, from
which the fibres diverge at a large angle, apparently
without clear trabecular organization. As shown
by Birenheide (1961, p. 85) the median dark line
represents that region in which the fibres are per- text-fig. 1. Generalized septal micro-
pendicular to the section. In tangential longitudinal *e Family Spongophyllidae,

section the fibres are seen to diverge strongly from f^om thin sections of Eifel species of
the mid-line of the septum. Only in radial longi- Accmthophyllum in the collection of the
tudinal section may the trabecular arrangement be University of Queensland, a, tangential
seen, and even then often not very clearly. In such longitudinal section along the axis of one
sections there is also seen a parallel banding perpen- i radial longitudinal section

... ,• ,1 . ■ along the median plane of the septum;

dicular to the direction of the trabeculae. This is boundaries between trabeculae
apparently formed partly by periodic variation in emphasized; b", appearance in thin
either optical orientation or composition, and partly section; c, transverse section, with tra-
by periodic insertion of additional trabeculae as the becular boundaries as for b', b".
trabecular fan expands.

In cases where parts of the septa consist of more than a single series of trabeculae, the
median dark line is normally absent. The individual trabeculae are round in section, and
do not greatly diverge from the direction of the septal plane. Moreover, there is generally no
regular arrangement into radial series in such cases. This condition is known peripherally
in Dohmop/iyllum, and over much of the dissepimentarium in Pseudoehonophyllum.

There are no carinae (ridges on the side of a septum, formed by the lateral extension
of one, or a series of successive, trabeculae at all stages of growth, and so growing in the
same direction as the trabeculae). Septal modifications may be of several forms;

546

PALAEONTOLOGY, VOLUME 9

TEXT-FIG. la. Map showing location of the Garra Formation in New South Wales, and generalized
geological map showing positions of fossil localities mentioned in the text.

D. L. STRUSZ: SPONGOPHYLLIDAE FROM DEVONIAN GARRA FORMATION 547

TEXT-FIG. 2b. Southern extension of locality and geological map, text-fig. 2a.

548

PALAEONTOLOGY, VOLUME 9

(a) Lateral denticles. These may be formed by individual fibres continuing a little
beyond the face of the septum, or by the temporary lateral divergence of part of a
trabecula from its upward direction of growth.

(b) Septal flanges. These are parallel to the growing edge of septum, and are formed
by the lateral divergence in unison of all the trabeculae in a septum. In Acanthophyllwn,
for example, they slope downwards towards the axis. Their greatest development is
frequently within the tabularium.

(c) Axial processes. These are usually difficult to interpret because of their complexity,
but appear to be formed either by the irregular extension of trabeculae beyond the axial
edge of the septum, or by irregularities in the course of the flanges in the axial region.
The latter would appear to be the most common.

{d) Naotic plates. These are much as in Craterophyllum ; they occur in Pseudochono-
phyl/uni, some Dohmophyllwn, and occasionally in other genera.

(p) Finally, in some instances the septa peripherally break up into parallel strands,
each consisting of a single series of trabeculae. These strands are usually anastomosing,
and in some cases may be modified and extended flanges.

D:M I T:M I D-M

tr

abed e

TEXT-FIG. 3. Diagrammatic representation of calical forms in the Spongophyllidae, shown in longi-
tudinal section, with German terms as used by Birenheide (1961, 1962) and the equivalent terms used
herein, a, ‘Trichterkelch’: inversely conical calice (characteristic of Acanthophyllum (Neostringo-
phylluni), A. (Grypophyllum))\ b, ‘Krempenkelch’: bell-shaped calice (characteristic of A. {Acantho-
pbyllum)); c, ‘Flachkelch’: saucer-shaped calice; d, ‘Wulstkelch’: everted calice; e, ‘ sattelrandiger
Wulstkelch’: everted calice with flared rim. D;M — dissepimentarium ; T:M — tabularium. Note the
correlation between arrangement of horizontal elements, attitude of trabeculae, and shape of calice.

Tabulae. The tabulae may be flat to gently concave, and mostly complete, as in
Spongophyllum, or gently to deeply concave with a median depression, in some cases
strongly infundibuliform, as in Acanthophyllwn and Xystriphyllum. In the second case,
the tabulae are generally predominantly incomplete to vesicular.

Dissepiments. Characteristically the dissepimentarium is wide, and consists of
numerous small, globose to elongate dissepiments. Lonsdaleoid dissepiments are
common. In Spongophyllum the dissepimentarium is comparatively narrow, but the
presence of well-developed lonsdaleoid dissepiments is characteristic, and serves to link
the genus with the group around Acanthophyllum with wide dissepimentaria.

Genus acanthophyllum Dybowski 1873

1873 Acanthophyllum Dybowski, p. 339 {fide Lang, Smith, and Thomas 1940, p. 13).

1961 Acanthophyllum Dybowski 1873; Birenheide, p. 80 ; contains an extensive synonymy to 1 958.

1949 Pseudochonophyllum Soshkina, p. 119 {non Soshkina 1937).

1962 Acanthophyllum Dybowski 1873; Philip, p. 184 (partim).

1964 Grypophyllum Wedekind 1922; Pedder, p. 439 {partim).

D. L. STRUSZ; SPONGOPH YLLID AE FROM DEVONIAN GARRA FORMATION 549

Type species. CyathophyUum heterophylliim Edwards and Haime 1851, p. 367, pi. X, fig. 1 ; subsequent
designation Schliiter, 1889, p. 296 (for description see Birenheide 1961, pp. 81, 89).

Diagnosis. Solitary, predominantly subcylindrical corals with more or less frequent
growth swellings. Calice inversely conical or bell-shaped. Septa trabecular, thin or
dilated, generally of only one radial series of trabeculae; lamellar thickening may occur
in some or all growth stages; flanges, lateral denticulae, and axial processes common.
Septa radial, or directed to a generally short median plane, frequently marked by
elongation of the cardinal septum. Tabulae incomplete to vesicular, forming concave
floors with a median depression. Lonsdaleoid dissepiments may be well developed.

Remarks. I follow Birenheide (1961) in subdividing the genus into three subgenera,
distinguished on calical shape, degree of septal dilatation, and to a lesser extent degree of
development of lonsdaleoid dissepiments. All three appear to be present in the Garra
Formation. Acanthophyllum mansfieldense of Philip (1962) is a species of A. {Neostringo-
phyllwn), and is therefore the earliest representative of that subgenus yet known.
Similarly, the earliest A. {Acanthophyllum) would seem to be that described by Philip
(1962) from the Gedinnian or Siegenian of Victoria. The Garra record of A. {Grypo-
phyllwn) would seem to be the earliest yet known for that subgenus.

The Russian records of this group are a little difficult to interpret. Most of the
specimens placed by Soshkina (1949, p. 119; 1952, p. 98) in Pseudochonophyllum are
Acanthophyllum. Bulvanker (1958) described several species under the names
Acanthophyllum, Neostringophyllum, Grypophyllum, and Stenophyllum, but also
included species better placed in other genera of this family. Similarly, species of
Acanthophyllum are placed by Soshkina and Dobrolyubova (in Orlov 1962) in Pteno-
phyllum, Acanthophyllum, Stenophyllum, Neostringophyllum, and Grypophyllum.

Subgenus acanthophyllum Dybowski 1873

Type species. See above.

Diagnosis. Subgenus of Acanthophyllum with bell-shaped calice. With thin peripheral
stereozone, few lonsdaleoid dissepiments. Septa generally weakly to strongly fusiform,
frequently further thickened during ontogeny by lamellar sclerenchyme. Trabeculae
arranged in half-fans in dissepimentarium.

Remarks. According to Birenheide (1961, p. 128), A. {Acanthophyllum) characterizes the
Eifelian of the Eifel. In Australia species are known as early as the Gedinnian or
Siegenian, and if recent interpretations are correct (Philip 1960; Philip and Pedder
1964), are not certainly known above the lower Eifelian. Of the five species so far
assigned to Acanthophyllum s.k, four are herein considered to be A. {Acanthophyllum).
These fall into two groups, one of large corallites and the other of small ones. The small
species are A. {A.) aequiseptatum Hill 1940, and A. {A.) asper Hill 1940, both from the
Murrumbidgee limestones (see pi. 85, figs. 1, 2). Both have Dc = circa 2 cm.,
n = circa 60. The holotypes differ in the manner of dilatation of their septa (strong,
axially thickened, and convoluted, major more than minor, in A. (A.) asper; mode-
rate, fusiform, uniform, in A. {A.) aequiseptatum), and in Dt {circa ^Dc in A. {A.)
asper, circa jDc in A. {A.) aequiseptatum). However, I have collected several corallites
from Clear Hill, type locality of A. {A.) aequiseptatum, which suggest that these

550

PALAEONTOLOGY, VOLUME 9

characters are rather variable, and that in fact the two species may be synonymous.
This will require further study outside the scope of this paper.

The large Australian species are A. (A.) clermontense (Etheridge fil. 1911) from Cler-
mont, Queensland, and A (A.) mansfieldense (Dun 1898) from Mansfield, Victoria. The
former is solitary to weakly fasciculate, the latter trochoid. Both have Dc over 2-5 cm., the
former reaching 6 cm. In both, Dt = circa JDc. In A. (A.) clermontense, n = 80 (max.),
in A. (A.) mansfieldense n = 50. Their most important distinction is in their septal
dilatation: in the former the septa are fusiform, and axially dilated and contorted; in
the latter the dilatation is in the form of concentric series of wedges increasing in size
peripherally, while the septa are radial, and not greatly dilated or contorted axially.

The new species described below belongs with this second group. It is easily dis-
tinguished from the previous species by the greater width of its oval tabularium, by the
strongly bilateral arrangement of its septa, and by its deeply infundibuliform tabulae.

Acanthophyllum {Acanthophyllum) aeneae sp. nov.

Plate 85, figs. 5a, b

Derivation of name. Aeneas, a Graeco-Roman mythological hero.

Holotype. SU 11291 (PI. 85, figs. 5a, b); loc. Be-10.

Diagnosis. Large A. {Acanthophyllum) with wide, oval tabularium of deeply infundibuli-
form incomplete tabulae; septa unequal, directed towards long axis of tabularium,
moderately dilated; minor septa long, very unequal.

Description. Solitary, trochoid, or ceratoid. The external characteristics are unknown,
as all free corallites collected were worn. The epitheca is apparently thin and lined with
a narrow fibrous stereozone. As deduced from longitudinal sections, the calice is the
bell-shaped one characteristic of the subgenus (see text-fig. 3), with a fairly wide sloping
rim and a wide, funnel-shaped axial pit. The greatest known diameter is 3-6 cm.+
(estimated circa 4-5 cm.), but most corallites are about 2 cm. across.

Dimensions. In mm. of representative specimens.

No.

loc.

Dc

Dt

DtlDc

n

SU 11291*

Be-10

est. 45

15x22

ca. 0-3-0-5

ca. 74

SU 11296

18-5-t-

7x7-5

7

ca. 60

SU 14121

Ct-6

est. 38

6-5

ca. 0-2

ca. 60

*Holotype; measured Dc = 36 mm.

Adult corallites have some 60-70 septa. The major septa are long and unequal,
extending to the axis where they are arranged about a long median plane. This plane is
further delineated by an elongate septum (cardinal?) at one end, which is flanked by
long minor septa. Opposite may be a rather short major septum. The septa are fusiform
and moderately to strongly dilated ; the dilated septa at most are equal to the interseptal
loculi. The septa are rough-sided, but not markedly flanged in the dissepimentarium;
thick flanges are prominent in the tabularium, where they are gently axially inclined. In
the zone of maximum septal dilatation there may be intermittent reinforcement by
lamellar sclerenchyme, which spreads over the surfaces of adjacent dissepiments (here
strongly geniculate towards the periphery), so that an incomplete zigzag stereozone

D. L. STRUSZ: SPONGOPHYLLIDAE FROM DEVONIAN GARRA FORMATION 551

may form. In large corallites there may be two or three concentric stereozones so formed.
The dilatation is somewhat variable within even the same transverse section ; however,
the major and minor septa are generally equally dilated. The minor septa vary in length,
generally ending at the inner margin of the dissepimentarium, but at times somewhat
withdrawn or extending a little way into the tabularium.

The septal structure is of fine trabeculae, approximately parallel, and diverging from
the median plane of the septum. They are arranged in a broad half-fan within the dis-
sepimentarium, the maximum inclination at the margin of the tabularium being about
50° from the vertical. The arrangement within the tabularium is less clear, but seems to
consist of a much less divergent half-fan, the trabeculae being nearly vertical near the
dissepimentarium, becoming moderately axially inclined (about 20° from the vertical ?)
near the axis. The flanges are trabecular and are perpendicular to the direction of growth
of the trabeculae. Even in longitudinal section they can be readily distinguished from the
lamellar sclerenchyme reinforcing the septal dilatation.

The epitheca has been worn from all available specimens, so the relative width of the
tabularium is uncertain. It does appear to be rather wide (see table of dimensions).
In all specimens the tabularium is markedly oval, the long axis being in the presumed
counter-cardinal plane. The tabular floors are funnel-shaped, with sides inclined axially
at about 30-40° from the horizontal ; the margins are upturned, at times sharply, while
there is a strong axial depression about JDt across, and as deep. The tabulae are close,
elongate, varying from nearly complete to incomplete (particularly marginally and
within the axial depression).

The dissepimentarium consists of numerous small rather globose to elongate vesicles,
which increase in inclination but decrease in size towards the axis. There is a definite
region towards the tabularium where the rate of increase in inclination is temporarily
increased. This change in slope is, however, not quite as abrupt as in the German species
of Accmthophyllum s.s. In transverse section the dissepiments are strongly geniculate,
the apices directed peripherally, over the greater part of the dissepimentarium. Lons-
daleoid dissepiments do not appear to be developed, but it is possible that complete
specimens may show a narrow peripheral series.

Comparison. A. (A.) aeneae differs from the two other large Australian species — A. (A.)
mansfieJdense (Dun 1898) and A. (A.) clermontense (Etheridge fil. 1911) — in its wide tabu-
larium and in the extreme bilaterality of the septa and tabularium. A. (A.) clermontense
has similar septal dilatation, except that the major are thicker, but the tabular floors are
not as deeply infundibuliform; the species may be weakly colonial. A. (A.) mansfieJdense
has fewer septa which are dilated in a distinctive manner — concentric series of wedges —
and somewhat withdrawn from the axis ; the calix is considerably shallower.

The closest German species is the type A. (A.) heterophyllum (Edwards and Haime
1851); this has rather more septa which are generally markedly convolute axially, and
not nearly so strongly bilaterally arranged ; moreover, the minor septa are thinner than
the major. The tabulae are smaller, rather more globose, and not so steeply inclined.
Similarly, the dissepiments are more globose, and are more elongate axially, whereas
in A. (A.) aeneae they are more elongate peripherally. The manner of reinforcement of
the septal dilatation by sclerenchyme differs somewhat from the German species.

Known localities. Be-10, Ct-6, Ct-53.

552

PALAEONTOLOGY, VOLUME 9

Acanthophyllum (Acanthophyllum) clermontense (Etheridge fil. 1911)

191 1 Cyathophyllum{l) clennontensis Etheridge, p. 5, pi. B, figs. 1, 2, pi. D, fig. 3. Douglas Ck.,
Clermont, Queensland; ‘Upper Couvinian’.

19396 Acanthophyllum clermontense Etheridge; Hill, p. 57, pi. IV, figs. 1, 2.
non 1962 Acanthophyllum clermontense (Etheridge); Philip, p. 185, pi. 27, figs. 1, 2. Tyers R.,
Victoria; Gedinnian.

Diagtwsis. Large, solitary or weakly fasciculate A. (Acanthophyllum) with narrow tabu-
larium; septa moderately dilated, fusiform, frequently with second zone of fusiform
dilatation just outside tabularium, generally axially contorted; tabulae incomplete,
forming bowl-shaped floors with wide axial deepening.

Remarks. This of all the Australian species most resembles the German species as de-
scribed by Birenheide (1961); from them it differs particularly in being weakly fasciculate,
but also in its narrow tabularium (Dt/Dc = circa J). The specimen described and figured
by Philip (1962) from Victoria has distinctly domed tabular floors and apparently an
inversely conical calice; it is probably a species of Dohmophyllum Wedekind 1923 (see
Birenheide 1963), although somewhat older than described species of the genus.

A. (A.) sp. cf. clermontense (Etheridge fil. 1911)

Plate 85, figs. 3, 4, 6, Plate 86, fig. 1 ; text-fig. 4.

Material. Several specimens (9 sectioned) from Iocs. Cr-100, P-43.

Description. Most of the specimens appear to be solitary, ceratoid to cylindrical, but one
at least (SU 13230, loc. Cr-100) is clearly fasciculate, consisting of a small radiating
corallum of 3 or 4 corallites, increase being peripheral. The epitheca is thin, marked by
narrow septal grooves. Repeated sharp rejuvenescence rims often occur. The calice has
a wide, sharp-rimmed, gently sloping border, and a rather shallow, bowl-shaped axial
pit. The septa form prominent ridges on the floor of the calice. The largest known coral-
lite is about 3 cm. in diameter, but for most corallites Dc = 15-20 mm.

In adults n = 50-58. The major septa are long, extending unequally to the axis.
Their axial structure and arrangement is variable: in some they are thin, with thin
flanges, and are irregularly curved as in 4. (A.) clermontense; in others, the arrangement
is similar, but the septa are rather thickened, and heavily flanged. This leads on to those
specimens in which the septa are thick and strongly flanged axially, and are bilaterally
directed to a short median plane, rather than rotated as in typical A. (A.) clermontense.

Dimensions. In mm. of representative specimens.

No.

loc.

Dc

Dt

DtlDc

n

Group a: septa not greatly dilated or flanged:

SU 13227

Cr-100

est. 22

5-5

ca. 0-25

54

SU 13230*

?9

4-8

2-8

0-58

38

>5

99

est. 20

6-7

ca. 0-34

58

Group b: septa strongly dilated and flanged:

SU 17237

P-43

est. 24

7-5

ca. 0-31

54

SU 17258

99

est. 23

70

ca. 0-30

52

SU 17261

99

est. 29-5

7-0

ca. 0-24

52?

* colonial.

D. L. STRUSZ: SPONGOPHYLLIDAE FROM DEVONIAN GARRA FORMATION 553

Similarly, the septal dilatation is highly variable, even in the one transverse section.
Generally the minor septa are as thick as the major and terminate at the margin of the
tabularium. In some specimens the septa are fusiform in transverse section, and only
moderately dilated, with broad wedge-shaped bases. Other specimens show this type,
and as well strongly dilated septa tapering inwards from the thick bases, with perhaps a
second zone of thickening corresponding to the zone of fusiform dilatation of the
thinner septa. The other extreme (group b in the above table) is of extremely dilated
septa, in contact over the outer half to two-thirds of the dissepimentarium, and then
tapering inwards, perhaps with a narrow zone of slight fusiform dilatation just outside
the tabularium. Occasionally in these specimens some of the septa are peripherally
attenuate, but rapidly dilate inwards, being in contact a short distance in from the
epitheca. The septa in some sections bear fine lateral denticulae, and in most cases are

10-6 . . . group "a"

-OF • ' .....

p + group b

5 10 20 ^ 30mm

TEXT-FIG. 4. A. (Acanthopliyllum) sp. cf. clermontense. Scatter diagram for Dt/Dc:Dc.

flanged; the inclination of the flanges increases inwards to the margin of the tabularium
and then decreases to nearly horizontal. The flanges are generally weak but may become
strong in the tabularium.

The septa consist of thin trabeculae arranged in broad half-fans in the dissepimen-
tarium : near the periphery they are directed upwards and inwards at a small angle from
the vertical. This angle increases steadily inwards until near the tabularium, where there
is a sudden fairly marked increase to about 30-40°; within the tabularium they are
directed more nearly vertically as the axis is approached.

In juveniles the tabularium is wide, Dt = c/>caO-5Dc; however, it increases in width
more slowly than the corallite, so that in adults it is narrow {circa 0-3-0-4Dc) — see
text-fig. 4. The tabulae are generally incomplete, axially inclined, and only moderately
convex. The tabular floors are bowl-shaped, with upturned margins, and a wide,
shallow axial depression.

The dissepimentarium in adults is wide, consisting of numerous series of small dis-
sepiments. In the outer zone these are gently to moderately axially inclined and rather
globose. In a narrow zone around the tabularium they are steeply axially inclined and
tend to be elongate. Lonsdaleoid dissepiments are almost entirely absent; in only one
or two sections have I seen a few small ones.

Remarks. These specimens differ from the Clermont ones in their smaller size, and in
their far more variable septal structure, dilatation, and axial arrangement. Were it not
for the gradation shown by some specimens I would be inclined to place those with
axially dilated and flanged septa directed to a short axial plane (group b in the table of

554

PALAEONTOLOGY, VOLUME 9

dimensions) into a distinct species. Until the extent of variation in the Clermont species
is known fully, and until considerably more Garra material is available, it is possible
only to compare them.

Known localities. Cr-100 (9 or 10 specimens), P-43 (5 or 6 specimens), WC-9. Possibly also Cr-77.

Subgenus neostringophyllum Wedekind 1922

Type species. By monotypy, N. ultimum Wedekind 1922, p. 16, figs. 17, 18. Type locality: Dusseltal,
Bergisches Land; upper Givetian.

Diagnosis. Subgenus of Acanthophyllum with inversely conical calice. Epitheca thick;
septa predominantly strongly dilated, composed of trabeculae directed upwards and
inwards, approximately parallel. Lonsdaleoid dissepiments absent or poorly developed.
Tabulae mostly incomplete, vesicular.

Remarks. None of the named Australian species can be certainly placed herein. How-
ever, several specimens, from a number of basins, previously referred to Acanthophyllum
aequiseptatum Hill 1940, are actually representatives of A. {N.) implicatum sp. nov.
If this subgeneric assignment is accurate — and there are no marked morphological
differences from the German species as figured by Wedekind (1922), Walther (1928), and
Birenheide (1961) — then this is a very early appearance of the subgenus, for in Germany
and in England (Webby 1964, p. 12) it is characteristic of the Givetian. The Australian
occurrences are all Lower or basal Middle Devonian. Homomorphy is possible, but for
the present I consider it best to use the one subgeneric name, as in both cases the species
are probably derived from A. {Acanthophyllum), and so are closely related.

Acanthophyllum {Neostringophyllum) implicatum sp. nov.

Plate 86, figs. 2-8 ; text-figs. 5, 6

19406 Acanthophyllum sp. Hill, p. 253, pi. 9, fig. 6. Wee Jasper, N.S.W.; Emsian or Eifelian.
1940c Acanthophyllum spp. Hill, p. 152, pi. 2, fig. 2; ?p. 153, pi. 2, figs. 3a-c. Silverwood,
S. Qld. ; Tower Couvinian’ (Emsian?).

1940 Acanthophyllum sp. Hill and Jones, p. 179, pi. 2, figs. 1, 2. Molong District, N.S.W.;
Emsian?

19426 Acanthophyllum sp. Hill, p. 14, pi. 1, figs, la, 6. Mt. Etna, Qld.; Emsian?

?19426 Acanthophyllum sp. or Lyrielasma sp. Hill, p. 14, unfigured. Mt. Etna, Qld.

?1942cl Acanthophyllum cf. mansfieldense (Dun); Hill, pp. 182, 183, pi. 5, fig. 1. Wellington,
N.S.W. ; Emsian?

1942c/ Acanthophyllum aequiseptatum Hill; Hill, p. 183, pi. 6, figs. \a, 6 {non Hill 19406, p. 251).
Wellington, N.S.W.

1 950 Acanthophyllum sp. Hill, p. 1 39, pi. 5, fig. 2. Buchan, Vic. ; Emsian ? (or early Eifelian ?).
1962 Acanthophyllum aequiseptatum Hill; Philip, p. 184, pi. 26, figs. 2, 3. Tyers R., Vic.;
Gedinnian (or Siegenian?).

Derivation of name. Latin implicatus = entangled, confused; referring to the septal meshwork at the
base of the calice.

Holotype. SU 17219 (PI. 86, figs, la, 6). Type locality: P-40.

Diagnosis. Small A. {Neostringophyllum) with deep, inversely conical calice whose base
is filled with an extensive meshwork of septal ends and flanges; septa moderately to
extremely dilated, flaring to wedge-shaped peripherally, tapering inwards, without

D. L. STRUSZ; SPONGOPHYLLID AE FROM DEVONIAN GARRA FORMATION 555

secondary lamellar deposits; tabularium rather wide, consisting of moderately to deeply
concave incomplete tabulae, generally with a small axial deepening.

Description. Solitary; corallites may form loose colonies of adherent individuals,
frequently invested by a stromatoporoid. Corallites attach themselves by lateral ex-
pansions of the dissepimentarium (PL 86, figs. 2, 8). They are ceratoid to cylindrical,
with irregular although not pronounced growth swellings and contractions; occasional
rejuvenescence rims occur. The calice is deeply inversely conical, with a sharp rim,
steep sides, and a concave base. The septa extend as low ridges down the sides, but at the
base form a deep meshwork of lamellae and flanges, whose upper surface is gently
concave, and up to 10 mm. above the topmost tabula (PI. 86, figs. 4, 7). For normal
adults, Dc = 10-20 mm.; the maximum known is 22-5 mm.

Dimensions. In mm.

No.

loc. Dc

Dt

DtjDc

n

SU 17219

P-43 2-5

1-4

0-56

30

99

„ 9-6

ca. 0-5

ca. 0-52

48

99

„ 12-4

60

0-48

56

99

,, ca. 13-5

6-0

ca. 0-44

56

99

14-0

ca. 0-6

ca. 0-43

50

Measured from random sections of corallites in type specimen.

There are up to 66 septa, the usual number lying between 48 and 62. Generally they
are equally dilated and straight in the dissepimentarium, to which the minor septa are
confined; within the tabularium the major septa may be straight, but are more usually
curved. They are frequently arranged in four ill-defined groups, apparently symmetrical
about the counter-cardinal plane. This plane is frequently marked by elongation of the
cardinal septum, which is often then flanked by elongate minor septa. Septal dilatation
is quite variable in extent, at times even within the one corallite. It is generally moderate
to strong, tapering inwards; the septal bases are generally flared or wedge-shaped,
forming a stereozone which is usually about 0-5-1 mm. across, but may rarely be up
to 3 mm. wide. In a few cases the septal bases do not contact at all, when there is between
them a fibrous stereozone about 0-3 mm. thick. In some sections the septal dilatation is
slightly fusiform or lanceolate. As the septal dilatation decreases inwards, so there is a
steady increase in the degree of lateral flanging; these flanges are trabecular outgrowths,
parallel to the calical surface, and perpendicular to the direction of the trabeculae. They
are steeply axially inclined in the dissepimentarium, the inclination decreasing in the
tabularium, such that they are best seen in slightly oblique longitudinal sections. The
combination of rotated or convoluted septal ends and flanges frequently presents a
highly complex picture in transverse sections of the tabularium.

The septa are composed of thin trabeculae directed upwards and inwards at about
30° to the horizontal in the dissepimentarium. This inclination changes rapidly as the
septa enter the tabularium, so that they are nearly vertical. Variation in the length of
the component fibres produces fine lateral denticulae on the septal sides in many
specimens.

The tabularium in adults is about 0-4-O-5 Dc. It consists of slightly convex incomplete
tabulae, generally increasing a little in size axially; these are arranged in gently to deeply

pp

C 4397

556

PALAEONTOLOGY, VOLUME 9

concave tabular floors, with upturned margins, and generally with a narrow median
depression of variable depth.

The dissepimentarium consists of up to 9 series of small, globose to rather elongate,
highly inclined dissepiments. The inclination may be uniform, or may increase gradually
towards the axis ; there is, however, never the marked change in inclination from moderate
to almost vertical near the tabularium which characterizes species of T. (Accmthophylliim).
Only in relatively few cases is the increase in inclination sufflcient for the dissepiments
at the inner margin of the dissepimentarium to be vertical. Lonsdaleoid dissepiments
are not developed.

Variation. This species shows considerable variation, particularly in the degree and
manner of septal dilatation — features which unfortunately do not lend themselves to
statistical analysis. In some specimens the septa are uniformly thin, and change little in
thickness inwards, beyond the characteristic peripheral wedge-like or flaring dilatation.
At the other extreme are those corallites in which the septa are in contact peripherally
for up to 3 mm., then taper inwards, but remain thick even in the tabularium. As
well, dilatation varies from the usual inwards taper to moderately lanceolate (maxi-
mum dilatation in outer third of septum, decreasing to either side) or fusiform
(maximum dilatation at mid-length), again always with a peripheral expansion.

Further variation occurs in the degree of inclination of the dissepiments. This varies
from about 30° to about 45° from the vertical. The concavity of the tabular floors also
varies within wide limits.

Measurements of Dc, Dt, and n were made for 77 individuals. As this was done from,
thin sections, the majority of which were consciously made of adult corallites, simple
frequency plots are of slight value. Graphs were plotted for Dt .Dc (text-fig. 5a) and
for n:Dc (text-fig. 6). The former is the most significant: there is relatively little scatter,,
and straight-line axes show a definite change in slope at Dc = 10 mm. This is taken to
represent the boundary between juvenile and adult corallites. For the juveniles, Dt =
0-5Dc; for adults, Dt = 0-33Dc+l-67 approximately. There is greater scatter in the
plot of n : Dc, but here also there seems to be a distinct break in slope at Dc = circa 10 mm.

A plot of Dt/Dc:Dc shows a fair degree of scatter (text-fig. 5b). Dt/Dc changes
gradually from about 0-5-0-6 for juveniles to about 0-4-0-5 for most adults; very large
corallites may have a tabularium as narrow as 0-33 Dc.

EXPLANATION OF PLATE 85

All figures X 2.

Fig. 1. A. {Acanthophyllum) aequiseptatum Hill 1940. la, b, transverse and longitudinal sections,
holotype AM F 9577, Bluff Limestone, Clear Hill, Taeinas, N.S.W. Photographs by courtesy of the
Australian Museum, Sydney.

Fig. 2. A. {A.) asper Hill 1940. 2a, b transverse and longitudinal sections, holotype UQ F 4270, Wee
Jasper, near Taemas, N.S.W. Photographs supplied by Mr. J. Jell, University of Queensland.

Figs. 3, 4. A. (A.) sp. cf. dennontense (Etheridge fil. 1911), group a (see pp. 552-3). 3a, b, transverse
and longitudinal sections, SU 13227, loc. Cr-100. 4a, longitudinal section of a small colony, 4b,.
transverse section of the corallite on the right in 4a; SU 13230, loc. Cr-100.

Fig. 5. A. {A.) aeneae sp. nov. 5a, b, transverse and longitudinal sections, holotype SU 11291, loc.
Be-10.

Fig. 6. A. (A.) sp. cf. clermontense, group b (see pp. 552-3). 6a, b, transverse and longitudinal sections,
SU 17237, loc. P-43.

Palaeontology, Vol. 9

PLATE 85

STRUSZ, Devonian Spongophyllidae from New South Wales

D. L. STRUSZ: SPONGOPHYLLIDAE FROM DEVONIAN GARRA FORMATION 557

TEXT-FIG. 5. A. {Neostriiigophyllniu) impUcatum sp. nov. Scatter diagrams and limiting polygons show-
ing variation of {a) Dt;Dc, and (b) Dt/Dc:Dc.

oJ-

0

5

10

15

20

□ c (mm) — ►

25

TEXT-HG. 6. A. {Neostringophyllum) iwplicatum sp. nov. Scatter diagram and limiting polygon for

n:Dc.

558

PALAEONTOLOGY, VOLUME 9

Remarks. The strongest objection to placing this species in A. {Neostringophyllum) is
that in the Eifelian succession the earliest known species, A. {N.) concavum (Walther
1928), appears at the Eifelian-Givetian boundary (see Birenheide 1961, p. 128).
However, there appears to be no significant morphological dilTerence between the
German and Australian species; I feel it is unwise at this juncture to erect an endemic
Australian genus solely on the basis of apparent age.

Range. This species is now known from a number of localities in eastern Australia.
It is likely that most of these are Coblenzian in age. Hill (1940Z?) suggested an age about
the boundary of the Lower and Middle Devonian for the Bluff Limestone fauna of the
Taemas region. Philip and Pedder (1964) have forcefully suggested a Siegenian or
perhaps early Emsian age for this succession. Similarly, they consider the Buchan
sequence to be mainly Siegenian, as against Hill’s (1950) late Emsian or early Couvinian.
The two limestones have very similar faunas, including A. (A.) aequiseptatum and
A. (N.) implicatwn.

The Silverwood fauna. Hill (1940c) compared most closely within Australia with that
of the Nemingha Limestone of Tamworth: the latter she (1942c) considered to be
probably late Lower Devonian, while the former she placed in the lower Couvinian.
I consider them both to be Coblenzian; A. (N) implicatum occurs at Silverwood, but is
not yet known from Tamworth. It is also known from the Mt. Etna Limestone of
Queensland, which is quite close to the Garra faunally, and which Hill (19426) con-
sidered to be (at least in part) Coblenzian in age.

The oldest fauna containing A. (N.) implicatum is apparently the Tyers R. fauna of
Victoria, which Philip (1960, 1962) considers to be Gedinnian or possibly early
Siegenian.

A. {N.) implicatum would therefore seem to be characteristic of the Lower Devonian
faunas of eastern Australia.

Known Garra localities. A. (N.) implicatum is widespread in the Garra Formation, second perhaps only
to the digonophyllids. It has been collected from the following localities: Br-3, Br-10, Ca-1, Cr-1,
Cr-2, Cr-5, Cr-86, Cr-89, Cr-94, Cr-100 (very common), Cr-103, Cr-106?, Cr-1 11, Ct-16?, Ct-55?,
E-5, E-37?, Ge-3 (very common), Ge-7b (common), Ge-12?, Gn-7?, MM-1, MM-5b, MM-6,
MM-7, MM-11, P-22 (rather common), P-23, P-25, P-26, P-38 (rather common), P-40 (type
locality), P-43 (abundant), WC-4, WC-5; BRV177, BRV965?

EXPLANATION OF PLATE 86

Fig. 1. A. (Acanthophyllum) sp. cf. clermontense (Etheridge 191 1), group b (see pp. 552-3). Longitudinal
section through calice, SU 17258, loc. P-43; x2.

Figs. 2-8. A. (Neostringophyllum) implicatum sp. nov.

2a, b, transverse to longitudinal sections of a group of corallites invested by a stromatoporoid;
holotype SU 17219, loc. P-40; X2. 3, oblique transverse section, SU 15296, loc. E-5; x2. 4a,
transverse section, 4b, longitudinal section showing septal mesh in calice; SU 7297, loc. Ge-3; X2.
Figured Hill ]942d. 5a, b, transverse and longitudinal sections, SU 18222, loc. P-43; x2.

6a, b, transverse and longitudinal sections of a specimen with septa thickened in the tabularium,
and with nearly fiat tabular floors; SU 12177, loc. Ca-1 ; X 2.

la, transverse section low in calice, showing septal mesh, lb, longitudinal section; UCT F 1983,
loc. P-43 ; X 2.

8, longitudinal section of a juvenile corallite adherent to the surface of a stromatoporoid colony;
SU 13222, loc. Cr-100; x4.

Palaeontology, Vol. 9

PLATE 86

6b

7b

STRUSZ, Devonian Spongophyllidae from New South Wales

D. L. STRUSZ; SPONGOPH YLLIDAE FROM DEVONIAN GARRA FORMATION 559

Accmthophyllum {NeostringophyJhim) turni sp. nov.

Plate 87, figs. 1,2; text-fig. 7

Derivation of name. Turnus, commander of the Latins resisting Aeneas’ settlement in Italy.

Holotype. SU 14247 (pi. 87, figs, la, b). Type locality: Ct-44.

Diagnosis. Small A. (Neostringophyllum) of irregular growth shape, with frequent sharp
rejuvenescence rims; septa dilated to contact at periphery, and intermittently as inverse
cones of septal tissue and sclerenchyme; the degree of dilatation increases distally.
Tabularium fairly narrow, of concave incomplete tabulae, often axially deepened.
Dissepiments steeply inclined, elongate.

Description. The solitary corallite is generally curved, ceratoid to trochoid, with marked
growth irregularities; it is commonly oval to irregular in cross-section. Rejuvenescence
is generally frequent and strong, forming sharp rims which are often close, so that the
contracted intervening spaces are deep but almost slit-like (see text-fig. 7). The calice is
deep, inversely conical, with a sharp rim and a narrow concave apex; the septa form only
low ridges on its sides and floor. The epitheca, apart from transverse growth irregularities,
is marked by narrow septal grooves.

Adult corallites are about 15-20 mm. in diameter.

Dimensions. In mm. of representative corallites.

No.

Dc

Dt

DtjDc

n

SU

14247*

8-6

4-8

0-56

38

est. 15

7-0

0-47

50

SU

14248

16-4

5-9

0-36

66

SU

14250

15-5

6-5

0-42

52

*Holotype; all specimens from loc. Ct-44. Mean values of Dc are given for oval corallites.

In adult corallites there are usually 50-64 septa. These extend inwards from the
epitheca (0-03 mm. thick), unbroken by lonsdaleoid dissepiments; the major septa
generally almost reach the axis, where they are usually directed to a prominent counter-
cardinal plane. There may be slight to moderate rotation of the septal ends. The counter
and cardinal septa are elongate, but do not meet. The shortest major septa flank these,
and the septal length gradually increases away from these, being again greatest trans-
verse to the counter-cardinal plane. The minor septa tend to be irregular in length,
although on the whole they extend inwards for about R/2. Those flanking the counter
(or cardinal ?) septum may be somewhat lengthened. The septa are dilated to contact in
a peripheral stereozone about 1-T5 mm. wide, whose inner surface is reinforced by
lamellar sclerenchyme. In this narrow region of reinforcement the septa themselves
attenuate rapidly inwards, so that inside the stereozone they are thin or only weakly
dilated, in the latter case being irregular to fusiform. In most corallites inversely
conical sheets of sclerenchyme extend inwards from the stereozone, coating septa and
dissepiments. These appear to correspond in position to prominent and very close
rejuvenescence rims, seemingly representing tissue deposited on the surface of the calice
immediately before constriction, and to the peripheral stereozone lining the epitheca of
the immediately subsequent expansion (see text-fig. 7).

The septa are generally somewhat wavy and bear moderately developed flanges.

560

PALAEONTOLOGY, VOLUME 9

which increase in frequency in the tabularium. Axial processes are weakly developed or
absent, as are lateral denticulae.

The septal microstructure is that typical for the subgenus : approximately parallel fine
monacanthine trabeculae directed upwards and inwards at 20-30° from the horizontal
within the dissepimentarium, the inclination becoming much steeper in the tabularium.

TEXT-FIG. 7. A. {Neostringophylliim) tiinii sp. nov. Longitudinal
section of holotype SU 14247, X 4, showing sharp rejuvenescence
(r) above deep, slit-like re-entrants, and the associated inverse
cones of septal dilatation. Drawn from photograph.

EXPLANATION OF PLATE 87

All figures x 2.

Figs. I, 2. A. {Neostringophylliim) tiirni sp. nov. In, transverse and oblique sections, \b, longitudinal
section showing rejuvenescence (cf. text-fig. 7); holotype SU 14247, loc. Ct-44. 2a, b, transverse and
longitudinal sections, SU 14248, loc. Ct-44.

Figs. 3, 4. A. (GrypophyUiim) jenkinsi sp. nov. 3a, transverse section at base of calice, 3b, transverse
celluloid peel near proximal end, 3c, longitudinal section showing unflanged septa; holotype SU
16247, loc. Ge-2. Aa, transverse section showing septal thickening in tabularium, Ab, longitudinal
section; UCT F 1947, loc. Ge-2.

Fig. 5. Lyrielasma sp. cf. siibcaespitoswn (Chapman 1925). Transverse section, SU 19238, loc.
St/2565.

Figs. 6-8. Pseudochonophylliim pseiidohelianthoides (Sherzer 1892). 6a, b, off-centre longitudinal, and
transverse sections, holotype, Rominger Collection, Museum of Paleontology, University of
Michigan; Koneprusy, Bohemia. Photographs supplied by Prof. E. C. Stumm. 7, transverse section
showing well-developed naotic plates (lower right), SU 13241, loc. Cr-100. 8n, b, longitudinal and
transverse sections, UQ F 3412, Silverwood, Qld.; Acantliophylliiin sp. cf. maiisfieldeuse of Hill
1940c, pi. II, figs. In, b. Photographs by courtesy of Prof. D. Hill.

Palaeontology, Vol. 9

PLATE 87

STRUSZ, Devonian Spongophyllidae from New South Wales

D. L. STRUSZ: SPONGOPH YLLID AE FROM DEVONIAN GARRA FORMATION 561

The tabularium decreases from about |Dc in juveniles to about 0-4Dc in adults.
The tabular floors are moderately to strongly concave, with upturned margins and an
axial deepening which is apparently a little eccentric. The tabulae are incomplete,
elongate, and only slightly domed. They are thin, or lightly to moderately coated with
sclerenchyme in the conical zones of dilatation.

The dissepimentarium consists of up to ten series of small, fairly globose to elongate,
highly inclined dissepiments. These are generally uniform in size and inclination, but
may become vertical at the margin of the tabularium; this is generally not vertically
persistent in a single corallite.

Variation. Data are too few for rigorous treatment; plots (not figured) were made for
Dt: Dc, Dt/Dc:Dc, and n:Dc. The first two appeared to be straight-line plots, the first
suggesting Dt = 04Dc-}-0-4. The plot for n : Dc was apparently curved, the rate of
insertion of septa decreasing steadily with age of the corallite.

Comparison. On the basis of available material, this species is inseparable in size,
number of septa, and relative width of tabularium from A. {N.) implicatum sp. nov.,
and from A. {Grypophyllum) jenkinsi sp. nov. (described below). Examination of the
data plots suggests that further material may show that Dt/Dc is slightly less than for
A. {N.) implicatum, from which it is distinguished by its prominent rejuvenescence, its
septal dilatation, and its lack of an intricate septal mesh work axially. A. (G.) jenkinsi
lacks the wide stereozone and sharp rejuvenescence; moreover, it may have some
lonsdaleoid dissepiments distally. Both of these species lack the periodic conical
dilatation characteristic of A. {N.) tiirni. In this feature it differs also from all of the
European species.

Remarks. This species is placed in A. (NeostringophyJlnm) because of its peripherally
strongly thickened septa and its complete lack of lonsdaleoid dissepiments.

Known localities. Ct-44, Ct-45.

Subgenus grypophyllum Wedekind 1922

Type species. G. deuckmanni Wedekind 1922, p. 13, figs. 13, 14. Type locality: Bergisch-Gladbach,
Bergisches Land; Biichel Beds, Givetian.

Diagnosis. Subgenus of Acanthophyllnm with inversely conical calice, usually with
fairly rounded margin. Septa predominantly thin, composed of fine trabeculae directed
upwards and inwards, roughly parallel. Lonsdaleoid dissepiments, particularly inter-
rupting minor septa only, common; dissepiments inclined uniformly inwards. Tabulae
incomplete, vesicular.

Remarks. Hill (19406, p. 267) interpreted Grypophyllum as fasciculate, and has described
several species from Australia on this basis. Of these, ‘ G.' aggregatum Hill 1940 and ‘ G.'
lophophylloides Hill 1942u I tentatively place in Lyrielasma Hill 1939. The other species
from north Queensland assigned by Hill (1942<7) to Grypophyllum have tabulae which
suggest that they may not even be spongophyllids.

Pedder (1964, p. 439) followed Hill in interpreting the genus as fasciculate, and
described a Canadian species which seems to be allied to ‘G.’ aggregatum Hill 1940,

562

PALAEONTOLOGY, VOLUME 9

both of which differ from typical LyrieJasma in their rather thin peripheral stereozone.
For the present, however, I would leave them in that genus.

AcanthophyUum {Grypophyllum) jenkinsi sp. nov.

Plate 87, figs. 3, 4

Derivation of name. For Dr. T. B. H. Jenkins, of the University of Sydney.

Holotype. SU 16247 (pi. 87, figs. 3 a-c). Type locality: Ge-2.

Diagnosis. A. {Grypophyllum) up to 25 mm. across, with thin epitheca; minor septa well
developed; cardinal and counter septa elongate; septal processes absent; lonsdaleoid
dissepiments occur only in adult stage.

Description. Solitary, cylindrical or ceratoid, with some growth irregularities. The
epitheca has distinct interseptal ridges and faint growth lines. The calice is inversely
conical, rather deep, with a concave base. Septa form low ridges down its sides, and
somewhat higher ridges across its base. The greatest known Dc = 23 mm. ; it is usually
about 15-20 mm.

Dimensions. In mm.

No.

Loc. Dc

Dt

DtjDc

n

SU 16247*

Ge-2 8-5

4-0

0-47

48

5?

„ 10-9

4-3

0-40

50

„ 16-0

5-0

0-31

56

„ 16-5

5-2

0-32

56

UCT F 1947

„ 13-9

8-0

0-58

52

*Holotype: serial sections over a length of 40 mm.

In adults, n = 50-60. The septa are long, fusiform, and only slightly dilated. Their
peripheral ends, however, expand rapidly as broad, blunt wedges, to form a dentate
stereozone. The septal bases are slightly concave externally (in transverse section), so
giving rise to the interseptal ridges of the epitheca. The septa are smooth, or bear fine
lateral denticulae; they are gently waved on a small scale, appearing straight at first
glance. Weak, discontinuous flanges occur near and in the tabularium. There are no
axial processes. In juveniles the septa are complete, but in adults some peripheral
discontinuities generally develop. The minor septa end just inside the tabularium.
The major septa extend unequally to a fairly short median plane, which is marked by
elongate counter and cardinal septa. The counter (?) septum is flanked by elongate
minor septa.

The septa consist of fine almost parallel trabeculae directed upwards and inwards at
about 20° from the horizontal. Their arrangement within the tabularium is unknown.

The tabularium is rather narrow — about 0-3Dc in adults, somewhat greater in
juveniles. The tabular floors are concave and often moderately deepened axially. The
tabulae are incomplete, elongate, and weakly convex.

The dissepimentarium consists of up to 9 series of globose to rather elongate dissepi-
ments, steeply axially inclined. They are fairly uniform in size and increase in inclina-
tion only slightly towards the tabularium. Lonsdaleoid dissepiments, only rarely
interrupting the major as well as the minor septa, are poorly developed in adult stages.

D. L. STRUSZ: SPONGOPHYLLID AE FROM DEVONIAN GARRA FORMATION 563

Comparison. A. (G.) jenkinsi sp. nov. does not closely resemble any previously described
Australian species. A. (N.) impUcatum sp. nov. is at first glance rather similar, and their
data plots coincide, but it may be distinguished by the different septal dilatation (taper-
ing axially, and rather greater), and the complete lack of lonsdaleoid dissepiments;
also it has a generally thicker epitheca and stereozone, and a wider tabularium. Finally,
the septal mesh at the base of its calice is highly distinctive of A. {N.) impHcatiim.

A. (G.) jenkinsi is rather close to A. (G.) denckmanni (Wedekind 1922) sensu Engel
and Schouppe (1958) in size, number of septa, septal dilatation and arrangement, and
structure and relative width of tabularium. However, the German species has more
frequent lonsdaleoid dissepiments, so that in transverse section the minor septa are
usually highly discontinuous; also it has a rather thicker and less dentate peripheral
septal stereozone. In all other described European species the minor septa are much
shorter and the tabularium wider.

Known localities. Ge-2 (type); Be-7, Cr-36b, ?Ct-36.

Genus pseudochonophyllum Soshkina 1937

1937 Psendochonopliyllnm Soshkina, pp. 59-60, 96 (partiinl)] remaining synonymy as for
species.

T}’pe species. Chonophyllinn pseiidohelianihoides Sherzer 1892, p. 275, pi. 8, fig. 6; designated Soshkina
1937, p. 59. Type locality: Koneprusy, Bohemia; Lower Devonian.

Diagnosis. Trochoid to cylindrical rugosan with an everted calice. Septa trabecular, the
trabeculae arranged in half-fans in the wide dissepimentarium. Septa strongly dilated in
dissepimentarium, generally to contact, and peripherally partly replaced by more or less
well-developed naotic plates. Septa flanged in narrow tabularium. Tabularium of
incomplete tabulae arranged in concave floors, frequently depressed axially.

Discussion. Pocta (in Barrande 1 902, pp. 1 23-4), describing topotypic material, considered
Sherzer ’s species to have characteristics sufficiently removed from those of Chonophydlimt
Edwards and Haime 1850 for it to be a distinct genus, but went no further. Soshkina
(1937) erected the new genus Pseudochonophyllum with Sherzer ’s species as type; in the
English summary (p. 96) she referred to Sherzer’s original description as being a
sufficient diagnosis. In the Russian text (pp. 59-60) she gives the following (my transla-
tion) :

Coral solitary, now and then giving off buds. External form subcylindrical, frequently turbinate or
mushroom-shaped. Calice small, goblet-shaped, with very wide top and weak basal boss. Septal
apparatus constructed of stout lamellar septa of 2 orders. These are peripherally strongly thickened
with stereoplasm, and towards the axis become very thin and sometimes curved; subsequently some of
them again are a little thickened at the immediate centre. Each vertical lamellar septum consists of
curved roof-like platelets, superimposed one on another. Interseptal apparatus in the peripheral zone
consists of numerous lamellae in the form of small vesicles. In the central zone are developed numerous
very thin, strongly vesicular incomplete and confused tabulae.

Distinguished from the genus Chonophyllinn by its pleomorphic structure and the development of
2 orders of septa. At the same time, they are similar in: 1) external form, 2) form of calice, 3) roof-
shaped axial boss of complicated incomplete tabulae.

Since this paper, Soshkina has several times referred Russian species to the type
species which in fact differ markedly from it, and are closer to Acanthophylhim. This
has been recently recognized by Bulvanker (1958, p. 95), who considers Soshkina’s

.564

PALAEONTOLOGY, VOLUME 9

species to be Acanthophyllum, but apparently without explicitly committing herself on
the Bohemian type species. She does, however, place Pseudochonophyllum in the synonymy
of Acanthophyllum. She also points out the further confusion arising from Soshkina
{1952, p. 98) naming as type species P. pseudohelianthoides Soshkina non Sherzer.
Spasskiy (1960, p. 43) places Soshkina ’s species in synonymy with Acanthophyllum
heterophyllum (Edwards and Haime 1851), but Birenheide (1963, pp. 407,409) considers
it to be a species of Dohmophyllum Wedekind 1923. I agree with this, and also with
Birenheide ’s specific exclusion of Sherzer ’s species.

Professor E. C. Stumm has kindly sent me photographs of sections he has recently
had made of the holotype of P. pseudohelianthoides. From this, and from the descrip-
tions and figures given by Sherzer (1892) and Pocta (in Barrande 1902) of material from
the type locality, 1 consider Pseudochonophyllum to be closely related to Acanthophyllum
and Dohmophyllum. It differs from the former essentially in having a reflexed calice, in
having septa composed of more than one radial series of trabeculae, and in the naotic
modification of the peripheral parts of the septa. Dohmophyllum has flat to domed
tabular floors, and septa whose greatest dilatation is near the tabularium; however, the
multitrabecular structure of the peripheral parts of the septa in Dohmophyllum (Biren-
heide, 1963, p. 372: text-fig. 3) is identical with the structure of the dilated portions of the
septa in Pseudochonophyllum. The dissepimentarium and the arrangement of the septa
in Pseudochonophyllum are typically acanthophylloid.

Of the Russian material, only that figured by Soshkina (1937, pi. 18, figs. 1-4) from
the Coblenzian of the Urals probably belongs to the genus.

1892
1902
71937
1940c
non 1949
non 1950
non 1952
non 1955
nonl 1955
1962

Pseudochonophyllum pseudohelianthoides (Sherzer 1892)

Plate 87, figs. 6-8, Plate 88, figs. 1-3

Chonophylliim pseudohelianthoides Sherzer, p. 275, pi. 8, fig. 6. Koneprusy, Bohemia;
Lower Devonian.

Chonophyllum pseudohelianthoides Sherzer; Pocta, p. 123, pis. 47, 48, 109 (figs. 3-6),
113 (figs. 21, 22), in Barrande. Same locality.

Pseudochonophyllum pseudohelianthoides Scherzer {sic)\ Soshkina, p. 60, pi. 18, figs. 1-4.
Urals; Coblenzian.

Acanthophyllum sp. cf. mansfieldense (Dun); Hill, p. 152, pi. 2, figs, la, b. Silverwood,
Qld.; Couvinian?

Pseudochonophyllum pseudohelianthoides (Scherzer); Soshkina, p. 119, pi. 27, figs. 1, 2,
pi. 28, tigs. 2-5, pi. 29, figs. 2-5. Urals; D^-DJ.

Chonophyllum pseudohelianthoides Sherzer; Termier and Termier, p. 99, pi. 39, figs. 23,
24. Morocco, N. Africa; Emsian.

Pseudochonophyllum pseudohelianthoides Soshkina non Sherzer; Soshkina, p. 98, pi. 38.

Urals and Armenia ; Coblenzian and Eifelian. (The specimen figured is Acanthophyllum.)
Pseudochonophyllum pseudohelianthoides Scherzer (s/c); Krayevskaya, p. 218, pi. 42,
figs. 4a, b, in Khalfin. W. Siberia; Coblenzian to Eifelian.

Pseudochonophyllum pseudohelianthoides Soshkina; Spasskiy, p. 314 {fide Bulvanker
1958, p. 95).

Pseudochonophyllum pseudohelianthoides Sherzer; Soshkina, p. 309, in Orlov. Bohemian
material only (unfigured).

Holotype. The specimen figured by Sherzer (1892, pi. 8, fig. 6) from the Lower Devonian (E2) Kone-
prusy Limestone of Bohemia, now in the Rominger Collection of the Museum of Palaeontology,
University of Michigan. Sections have recently been made; they are figured herein (PI. 87, figs. 6a, b).

D. L. STRUSZ: SPONGOPH YLLID AE FROM DEVONIAN GARRA FORMATION 565

Diagnosis. Solitary, up to 5 cm. in diameter, with strongly reflexed calice and numerous
septa which may split into parallel strands peripherally; dissepiments small, gently
domed, near-horizontal except in narrow zone around tabularium, where vertical.
Description. The corallite is turbinate or trochoid, becoming irregularly cylindrical, with
narrow rejuvenescence rims, and growth swellings and contractions. Sherzer recorded
a maximum Dc = 5 cm. ; this is also known from the Wellington specimens, which
generally have Dc = 3-3-5 cm. The calice is weakly to strongly everted, sometimes with
a flared outer margin (the ‘ sattelrandiger Wulstkelch’ of Birenheide 1963, p. 371, text-
fig. 1). The axial pit is narrow and shallow and may have a low axial boss formed of a
meshwork of septa. The septa form rounded ridges on the surface of the calice. The thin
epitheca may be faintly marked by shallow interseptal grooves : unfortunately in most of
the Wellington specimens it has been weathered away.

As noted also by Soshkina (1937, p. 59), and figured by Pocta (pi. 47, fig. 21, in
Barrande 1902), occasionally a single peripheral calical bud may occur. I have seen
such small buds on 2 of the 25 or more specimens collected from the Wellington
district.

Dimensions. Of representative corallites in mm.

No.

loc.

Dc

Dt

DtjDc

n

SU 13180

Cr-94

32*

4-5

0-14

62

13246

Cr-100

45*

4-5

010

50

13248

55

34*

5-0

0-15

64

55

50*

6-0

0-12

68

13252

55

47

5-5

0-12

70

14104

Cr-1 1 1

34

4-2

0-12

66

* estimated value.

In the Wellington specimens, n = 52-68; Sherzer recorded 72. The major septa
extend equally towards the axis, leaving an axial space up to 2 or 3 mm., into which
septal flanges and axial processes may extend; the minor septa end just inside the tabu-
larium. In the tabularium the septa are thin, straight, radial, and bear strong flanges;
there may be some slight axial dilatation. The flanges appear in longitudinal section to
be arranged in a regular zigzag pattern, and extend inwards and downwards at a low
angle. In the dissepimentarium the septa are equally dilated; this increases rapidly
outward from the tabularium. Within 2 or 3 mm. the septa are generally wider than the
loculi, and over most of the outer half of the dissepimentarium they are in contact.
An unusual feature of a number of specimens in transverse section is the apparent fusion
of 2 or 3 septa to become one, as they are traced inwards from the periphery (pi. 88,
fig. 3). The dilatation appears to be periodically interrupted by intervals of growth during
which the septa were comparatively thin; in these intervals the septa are horizontally
corrugated. This is best seen on the surfaces of weathered corallites.

In the zone of extreme dilatation the septa are frequently modified. In some instances
a few septa may split into two or three parallel or anastomosing strands for a greater or
lesser distance, as may occur in Dohmophyllum (Birenheide 1963, pi. 59, fig. 56). More
often, the septa are replaced by naotic plates, supported by isolated trabeculae. These
plates are small, very gently domed, radially elongate, and parallel in inclination to the
dissepiments. Unlike Craterophyilum Foerste, the plates generally do not extend from

566

PALAEONTOLOGY, VOLUME 9

one side of the septum to the other in a regular manner; they overlap in a manner more
like that of dissepiments, and may occasionally extend across two or three septa (see
Sherzer 1892).

The septal trabeculae are up to 0-2 mm. thick; they diverge from a median plane in
the septum, although that plane is not expressed morphologically; they are radially
arranged in half-fans. A few millimetres outside the tabularium they are vertical;,
outside this, they are directed upwards and slightly outwards, until at the periphery
they may again become vertical. Towards the tabularium they are directed axially up-
wards at an angle of about 30° from the vertical. The trabeculae in the tabularium are
too fine to show whether a second fan develops there as in Acanthophyllum s.s.

The tabularium is narrow: Dt = circa 5 mm., and Dt = 0T2Dc approximately, being
generally between OTl and 0T4Dc. The tabular floors are gently sagging, with slightly
to strongly upturned margins; there may be a slight median depression. The tabulae are
incomplete, thin, and flat to slightly convex.

The very wide dissepimentarium consists of small gently to strongly domed dis-
sepiments. Over most of the dissepimentarium they are horizontal to slightly inclined
outwardly, following the calical shape. In a narrow zone around the tabularium they
are vertical and elongate. The dissepiments may be coated with thin fibrous extensions
of the septal dilatation.

Remarks. From an examination of the photographs kindly supplied by Professor
Stumm, and of the figures of Sherzer and Pocta, there would seem to be very little
difference between the Garra and Koneprusy material. The naotic plates in the Bohemian
specimens apparently show a greater tendency to extend across two or three septa, and
the dilatation of the septa does not show the periodic reduction to nearly the same
extent. The only possible argument against specific identity is that, as the Russian
material is excluded from the species, there is no record of it between Bohemia and
Eastern Australia.

The Silverwood specimen described by Hill (1940c) as Acanthophyllum sp. cf. mans-
fieldense has all the features of the present species — extreme septal dilatation consisting
of thick trabeculae, peripheral naotic modification, reflexed calice — and I consider it
to be specifically identical with the present species. It is figured herein for comparison
(pi. 87, fig. 8rt, b).

Range. The Koneprusy Limestone is considered to be time-transgressive, extending
through the whole of the Lower Devonian (see, e.g. Havlicek 1959). Hill (1940c) con-
sidered the Silverwood fauna to be probably Lower Couvinian, and equivalent to the
Nemingha fauna of Tamworth. I consider both these, and the Garra Formation, to be
more likely Emsian in age (see Strusz 1965); this would be supported by correlation
with Koneprusy.

EXPLANATION OF PLATE 88

Figs. 1-3. PseudochonophyUum pseiidohelianthoides (Sherzer 1892). 1, transverse section showing
septa at various degrees of dilatation; SU 13248, loc. Cr-100; x2. 2, oblique longitudinal section
showing acanthophylloid tabularium, and septal flanges; SU 13249, loc. Cr-100; x2. 3, transverse
section showing details of septal structure; ‘fused’ septa are bracketed peripherally (see p. 565);
SU 13248. loc. Cr-100; x4.

Palaeontology , Vol. 9

PLATE 88

STRUSZ, Devonian Spongophyllidae from New South Wales

0

}

K

D. L. STRUSZ: SPONGOPHYLLID AE FROM DEVONIAN GARRA FORMATION 567

Known Gana localities. All specimens have been collected from three localities in a biostrome cropping
•out in the generally massive limestones on the west bank of the Bell River, opposite the Wellington
Caves Reserve: Cr-94, Cr-100 (common), Cr-111.

Genus lyrielasma Hill 1939

1925 Cyatlwphyllinn ipartini) Chapma.n, p. 112.

1939a Lyrielasma Hill, p. 243.

Type species. Cyathophyllum subcaespitosum Chapman 1925, p. 112, pi. 13, figs. 15, \6a, b. Cave Hill
Lilydale, Vic:; Siegenian? — Philip 1960, p. 151.

Diagnosis. As AcanlhophylJum, but fasciculate to subcerioid; calice inversely conical;
with wide peripheral stereozone.

Discussion. Corallites of Lyrielasma are narrow, cylindrical, with an inversely conical
calice (the ‘trichterkelch’ of Wedekind — see Birenheide 1961). There is usually a
relatively wide peripheral stereozone formed by the dilatation of the septal bases; this
is widest in the type species, and narrowest in ' Grypophyllum' aggregatum Hill 1940.
In some cases this stereozone is partly broken up by lonsdaleoid dissepiments. The
characters of the dissepimentarium and tabularium are much as in Acanthophyllwn
{Grypophylhmi) .

Grypophyl/um aggregatum Hill 1940Z), from the Eifelian? of the Murrumbidgee
River, is a phaceloid species, and so is here removed from Grypophyllum, which Biren-
heide (1961, p. 114) has shown to be solitary, to Lyrielasma (see pp. 570-1).

Lyrielasma curvatum Hill 1942n (p. 238), from the Givetian Fanning River Limestone,
differs from the type species in having flat to domed tabular floors. It is more akin to
Grypophyllum compactum Hill 1942 (p. 255) and G. sp. Hill 1942 (p. 255), from the
same formation, none of these species being either Lyrielasma or Grypophyllum.

Lyrielasma^. micrum sp. nov.

Plate 89, figs. 1, 2; text-fig. 8
Derivation of name. Greek puKpos = small.

Holotype. SU 18200 (PI. 89, fig. la, b). Type locality: P-43.

Diagnosis. Lyrielasma about 5-6 mm. in diameter with relatively wide peripheral
stereozone and zigzag septa, and lacking lonsdaleoid dissepiments; calice deeply
conical.

Description. The form of the corallum is unknown, as all specimens are of fragmentary
cylindrical corallites; however, the holotype shows evidence of a dendroid growth form,
with lateral increase. The epitheca is thin, longitudinally and transversely wrinkled, but
without regular septal grooves. Narrow, sharp rejuvenescence rims occur (holotype,
and SU 18195). Longitudinal sections suggest a very deep calice, inversely conical in
shape, with a sharp or slightly rounded rim and a narrow concave base. Maximum
Dc = 7 mm., normal adults being 4-6 mm. across.

568 PALAEONTOLOGY, VOLUME 9

Dimensions. In mm. of representative corallites.

No.

loc.

Dc

Dt

DtjDc

n

Ua

SU 16299

Gg-2

4-9

2-7

0-55

32

0-3 R

SU 18195

P-43

7-0

3-7

0-53

38

0-4-0-7 R

SU 18200*

59

1-2

0-6

0-50

12

—

>5

99

4-5

2-4

0-53

36

0-4 R

5-0

2-1

0-42

—

—

* Holotype

Lining the thin epitheca is a relatively wide stereozone composed of dilated septal
bases, which in some instances apparently merge into lamellar sclerenchyme parallel to
the epitheca. This last may, however, be an effect of recrystallization. The stereozone is
0-5-T0 mm. wide. In adults, n = 30-40; the major septa extend unequally to the axis.

TEXT-FIG. 8. Lyrielasmal micrwn sp. nov. Scatter diagram for
Dt : Dc; the estimated approximate mean, Dt = 0-5Dc, is added.

where they are curved and arranged about a median plane. One major septum (K?) may
be elongate, extending across the axis. The minor septa vary in length from 0-3R to
0-7R, being generally about 0-5R. Inside the stereozone the septa are equally and
moderately dilated, irregularly wavy, and bear zigzag flanges which are inclined steeply
down towards the axis.

The trabeculae are very narrow and are approximately parallel throughout their
length. The structure of the trabeculae is much as described by Birenheide (1961) for
AcanthophyUum. They are directed axially upwards at a very small angle above the
horizontal.

Dt = 0-5Dc approximately (see text-fig. 8), ranging between 0-39-0-56. The tabular

EXPLANATION OF PLATE 89

Figs. 1, 2. Lyrielasmal micriim sp. nov. \a, b, transverse and longitudinal sections, holotype SU 18200,
loc. P-43 ; X 4. 2a, transverse section showing wide stereozone, thick septa, 2b, longitudinal section ;
SU 18195, loc. P-43; x4.

Figs. 3-6. Lyrielasma aggregation (Hill 1940). 2a-c. transverse and longitudinal sections, holotype
AM F 10132, Wee Jasper, near Taemas, N.S.W. ; figured Hill 1940Z?; X 2. Photographs by courtesy
of the Australian Museum, Sydney. 4, transverse section at base of calice, SU 19147, loc. St/1821 ;
x2. 5, random section through a fragmented colony, UCT F 1946, The Gap (loc. E-18?); x2.
6a, b, transverse and longitudinal sections, SU 14257, loc. Ct-50; x2.

Palaeontology, Vol. 9

PLATE 89

STRUSZ, Devonian Spongophyllidae from New South Wales

D. L. STRUSZ; SPONGOPHYLLID AE FROM DEVONIAN GARRA FORMATION 569

floors are deeply inversely conical, with a narrow concave base, but a median depression
has not been seen. The tabulae are incomplete, steeply inclined, elongate, and but little
inflated.

The dissepimentarium consists of 2-4 series of almost vertical fairly elongate to
moderately globose dissepiments, generally uniform in size.

Comparison. L? micrum differs from all other Australian species of Lyrielasma in its
small size and proportionately wide stereozone. The type species is closest in appearance,
differing essentially in size and number of septa.

Remarks. Because of the slight doubt whether this species is colonial or not, it is for the
present only tentatively assigned to Lyrielasma.

Known localities. Gg-2, P-43; BR^/177.

Lyrielasma subcaespitosum (Chapman 1925)

1962 Lyrielasma subcaespitosum (Chapman); Philip, p. 188, pi. 28, figs. 6-7. Cam Syn.
Holotype. NM 1731, 14065: Cave Hill, Lilydale. Figured Hill (1939fl), pi. XIV, figs. 1-3.

Diagnosis. Phaceloid Lyrielasma about 12 mm. in diameter, with long wavy to zigzag
or flanged septa seldom interrupted by lonsdaleoid dissepiments.

Remarks. L. subcaespitosum praeciirsor Philip 1962 (p. 1 19) differs from L. subcaespitosum
s.s. in being wider, with a wider stereozone, wide tabularium (Dt = -iDc), and frequent
lonsdaleoid dissepiments, interrupting both major and minor septa. The specimens
described below have features in common with both subspecies.

Lyrielasma sp. cf. subcaespitosum
Plate 87, fig. 5

Material. SU 12245 (loc. Cr-4)?; SU 19238 (PI. 87, fig. 5), 19241 (loc. St/2565).

Description. The material consists of fragmentary corallites; it is not known whether or
not they are from fasciculate coralla. The external characters are unknown.

Dimensions. In mm.

No.

Dc

Dt

DtjDc

n

SU 12245

est. 10

est. 4-5

ca. 0-45

60?

SU 19238

est. 9

5-5

ca. 0-6

46

The corallites are about 10 mm. in diameter. There is a peripheral stereozone about
0-7 mm. wide, formed of the dilated ends of septa. Inside this the septa suddenly reduce
in thickness, and maintain a fairly uniform moderate dilatation to the axis. Irregular
flanges occur in both dissepimentarium and tabularium.

The major septa extend unequally to the axis, being directed towards a median plane
which in SU 19238 is marked by one long and one short major septa, n = about 50-60.
The minor septa vary considerably in length between and ^R, being mostly about
R/3. Septal microstructure is not clear in the available material. Dt = circa 0-5 to 0-6 Dc.
The tabular form is not known. The rather narrow dissepimentarium is not well known;,
for the most part it is masked by the peripheral stereozone and the septal dilatation.

570 PALAEONTOLOGY, VOLUME 9

It apparently consists of small, highly inclined dissepiments. I have seen no lonsdaleoid
dissepiments.

Remarks. The available material differs from L. subcaespitosiim s.s. in its wider tabula-
rium, and possibly in having smaller dissepiments. From L. subcaespitosum praecursor
Philip 1962, which also has a wide tabularium, the specimens differ in lacking all sign of
lonsdaleoid dissepiments, and in having a narrower stereozone.

Lyrielasma aggregation (Hill 1940)

Plate 89, figs. 3-6, Plate 90, fig. 1 ; text-figs. 9, 10

19406 Grvpophvllimi aggregatum Hill, p. 268, pi. X, figs. Sa, b, pi. XI, figs. la-d. Wee Jasper,
N.S.W.; Eifelian.

\942d 2Grypophylliim taggregatum Hill; Hill, pi. VI, fig. 4. Wellington, N.S.W. ; ‘Early
Couvinian’ (probably Emsian).

Holotype. AM F 10132, Wee Jasper, Goodradigbee R., N.S.W., figured Hill (19406) pi. XI, figs. \a-d;
figured herein Plate 89, figs. 3 a-c.

Diagnosis. Lyrielasma with thin septa, narrow stereozone, wide tabularium, and absent
or rare lonsdaleoid dissepiments.

Description of Garra material. Available material is more or less fragmentary, but
UCT F 1946 shows evidence of a dendroid corallum in which increase is both sexual
(planulae settling on the sides of existing corallites) and by peripheral budding. The
latter apparently produces only one bud at a time and is non-parricidal. The epitheca is
smooth, or marked with faint longitudinal grooves which show little relationship to the
septa; it is thin. The calice is uncertain; longitudinal sections suggest a deeply inversely
conical calice with a narrow rounded rim. The maximum observed Dc = 11-8 mm.;
adults are normally about 5-9 mm. across.

Dimensions. In mm. of representative corallites.

No. loc.

Dc

Dt

DtjDc

n

stereozone

SU 14257 Ct-50

11-8

ca. 0-5

cfl.0-42

50

0-2-0-4 R

0-5 mm.

UCT F 1946 ‘The Gap’

2-9

1-7

0-59

26

01-0-2 R

01

95 99

4-3

3-6

0-84

30

01-0-3 R

0-3

99 99

6-5

3-2

0-49

38

0-3-0-4 R

0-6

99 99

8-8

4-2

0-48

38

0-3-0-5 R

0-4

SU 19147 St/1821

101

4-5

0-45

44

0-3-0-8 R

0-4

Maximum observed n = 50; for normal adults n = 34-44, mostly 38 or 40. The
major septa extend unequally towards a median plane, and generally leave a small
axial space. The minor septa are considerably more variable in length, even in the
same corallite, ranging between OT-0-8 R; they are usually 0-3-0-5 R. The septa are
equally dilated; their bases are flared to form a peripheral stereozone 0-5 mm. wide at

EXPLANATION OF PLATE 90

Fig. 1. Lyrielasma aggregatum (Hill 1940). Random section through a fragmented colony; note the
peripheral buds in the longitudinal section of a large corallite, lower left; UCT F 1946, The Gap
(loc. E-18 ?); X 2.

Figs. 2, 3. Spongophyllitm halysitoides halysitoides Etheridge fil. 1918. 2, transverse section, SU 16224,
loc. E-30; x4. 3a, 6, longitudinal sections, SU 16223, loc. E-30; x4.

Palaeontology, Vol. 9

PLATE 90

STRUSZ, Devonian Spongophyllidae from New South Wales

'}

4

T ,

■.tS''

I

Si

I

<

1

.. , 1

D. L. STRUSZ; SPONGOPHYLLIDAE FROM DEVONIAN GARRA FORMATION 571

most. Inside this they are tapered, generally only moderately to slightly dilated, and
frequently attenuate in the tabularium. They are zigzag, and occasionally bear weak
flanges; their inner ends may be curved. In SU 14257 the minor septa are contratingent,
and thus indicate the median plane: at one end they lean away from the major septum
(C ?), and they continue leaning in that sense until at the other end the one major septum
(K?) has two touching minor septa. Generally, however, the minor septa are far less
regular.

Dt = 0-5Dc approximately. The tabularium consists of elongate, highly inclined,
slightly convex incomplete tabulae, forming moderately to strongly inverse conical
tabular floors.

There are from one to four series of globose highly inclined dissepiments which tend
to become vertical and rather elongate near the tabularium. Lonsdaleoid dissepiments
are rare, and when present only intersect the minor septa. I have seen them in SU 5280
(collected E. M. Basnett 1940, from loc. MM-10) and in two corallites from loc. Ct-50.

Variation. Plots have been made for Dt : Dc, Dt/Dc: Dc, and n : Dc. The estimated curve
of best fit for the first approximates to Dt = 0-5Dc. The points fall slightly above this
line between Dc = 0 and 8-0 mm., and then slightly below it for greater Dc (text-fig. 9a).
This is shown more accurately by the plot Dt/Dc :Dc (text-fig. 9b)\ at Dc = 3 mm.,
Dt/Dc is about 0-55; at Dc = 8 mm., Dt/Dc = 0-50; and at Dc = 11 mm., Dt/Dc has
fallen to about 0-45.

The number of septa apparently increases rapidly at first (text-fig. 10), as at Dc = 2-9
mm. (the smallest measurement), n = 26. Beyond this, however, the relationship of
n:Dc is linear, approximating to n = 20+2-5Dc.

In no cases is the scatter of points excessive.

Comparison. L. aggregation differs from L. subcaespitosiun s.s. in a smaller average size,
a narrow stereozone, and thinner septa. From L. subcaespitosiun praecursor Philip 1962
it differs especially in the general lack of lonsdaleoid dissepiments, and also in size and
width of stereozone. From L. fioriforme Hill 1942 it differs in size, width of stereozone,
and in having tapered rather than fusiform septa; L. aggregation also has a wider
tabularium. L. lophophyUoides Hill 1942 is very similar; it is distinguished by having
four prominent, longer and thicker major septa, and a narrower tabularium.

Remarks. Hill (19406, p. 268) placed this species in GrypophyUwn because at that time
she considered that that genus included phaceloid species. However, she also drew
attention to the similarity of the species to Lyrielasma. As Birenheide (1961) has shown
that GrypophyUiim is a solitary subgenus of Acanthophylhon, and as ‘G. ’ aggregation
differs from other species assigned to Lyrielasma essentially only in having a narrow
stereozone, I have herein placed it in Lyrielasma.

Range. The species has been recorded from the Murrumbidgee limestones (Couvinian
according to Hill 19406, p. 249; Emsian according to Philip and Pedder 1964), and
from the Garra Formation (which I consider to be probably Emsian). It has also been
found at Douglas Creek, Clermont, Qld. (UQ F 36328), considered by Hill (19396) to be
probably upper Couvinian in age.

Known Garra localities. Ct-50, MM-10, MM-12; St/181, St/1125, St/1821; also collected by Dr.
J. Conolly from ‘The Gap’, probably loc. E-18.

Qq

C 4397

572

PALAEONTOLOGY, VOLUME 9

TEXT-FIG. 9. Lyrielasma aggregatiim. Scatter diagrams for (a) Dt:Dc, and (b) Dt/Dc; Dc, with data for
the holotype added (measured from published photographs); for comparison, the line Dt = 0-5Dc is

added.

TEXT-FIG. 10. Lyrielasma aggregation. Scatter diagram for n:Dc, with data for the holotype
added; the estimated approximate mean n = 20-l-2-5Dc is also added.

D. L. STRUSZ: SPONGOPH YLLID AE FROM DEVONIAN GARRA FORMATION 573

Genus spongophyllum Edwards and Haime 1851

1851 Spongophyllum Edwards and Haime, p. 425 (fide Lang, Smith, and Thomas 1940, p. 121).
1962 Spongophyllum Edwards and Haime 1851; Birenheide, p. 69 (partim). Contains an ex-
tensive synonymy, to which add:

1962 Neomphyma Soshkina 1937; Soshkina and Dobrolyubova, p. 335, in Orlov (partim).

Type species. By monotypy, S. sedgwicki Edwards and Haime 1851, p. 425. Neotype, selected Jones
(1929, p. 88), figured Birenheide (1962) pi. 9, fig. 8, pi. 10, fig. 10. Torquay, Devonshire; Middle?
Devonian.

Diagnosis. Phaceloid or, more usually, cerioid, with small eorallites; narrow tabularium
of flat or gently eoneave complete tabulae; dissepimentarium of only a few series of
globose lonsdaleoid dissepiments and smaller interseptal dissepiments; septa thin,
poorly developed, generally based on lonsdaleoid dissepiments.

Remarks. The genus has been well revised by Jones (1929), Prantl (1952), and Birenheide
(1962). Birenheide would include Australophyllum Stumm 1949. As proposed by Stumm,
there is certainly no difiference; however, the type species shows certain points which
I consider sufficient to justify retention of the genus. The tabulae are concave, with a
definite median depression such as typifies the acanthophylloid genera. Also the dis-
sepimentarium is rather wide for typical Spongophyllum.

Spongophyllum halysitoides halysitoides Etheridge fil. 1918
Plate 90, figs. 2, 3; text-figs. 11, 12

1918 Spongophyllum halysitoides Etheridge fil., p. 49, pi. VII. Tamworth, N.S.W. ; Coblenzian.
1932 Spongophyllum halysitoides Etheridge fil.; Jones, p. 56 (partim — not text-fig. 2). Tam-
worth.

1942c Spongophyllum halysitoides Etheridge; Hill, p. 161, pi. 11, figs. 5a, b. Tamworth.
non 1958 Spongophyllum halisitoides (sic) Etheridge; Bulvanker, p. 132, pi. LXIV, figs. \a, b, 2.
Kuznets Basin, U.S.S.R.; lower Eifelian.

71960 Spongophyllum halysitoides Etheridge; Spasskiy, p. 55, pi. XXIX, figs. 3, 4. Altai,
U.S.S.R.; lower Middle Devonian.

Holotype. AM F 16453 (AM 187), figured Etheridge, jr. 1918, pi. Vll, figs. 1-3. Type locality: road near
Beedle’s Farm, Moonbi, Co. Inglis, Tamworth, N.S.W. Nemingha Limestone, Coblenzian.

Diagnosis. Cerioid Spongophyllum with eorallites mostly 4-6 mm. in diameter, with
thick moniliform walls; septa much reduced or absent; tabularium narrow.

Remarks. The specimens from the Garra Formation agree closely with the descriptions
and figures of Etheridge, jr. (1918) and Jones (1932). The only differences lie in a slightly
stronger and more frequent development of septa, and in thicker walls. Some of the
Garra specimens have noticeably irregular eorallites.

The Garra material has been statistically compared with a specimen I have collected
from the Nemingha Limestone in a quarry just north of Attunga (near Tamworth), and
with SU 6241 from the same horizon at For. 256, Ph. Burdekin, near Tamworth.
The Tamworth and Garra specimens agree completely (text-figs. 11, 12). The range of
variation is greater than stated by Etheridge, the maximum diameter being nearly 9 mm. ;
most eorallites are about 3-5-7 mm. across. For both, Dt = 0-3Dc approximately, and
Dt/Dc seems to decrease only slightly with increasing Dc. A plot of n/Dc:Dc suggests
a slight change in rate of septal insertion with age.

TEXT-FIG. 1 1 . Spongophylluni halysitoides halysitoides. Scatter diagrams for (a) Dt ; Dc, and (b) Dt/Dc : Dc ;
data from specimens from the Garra Formation and the Nemingha Limestone (the type horizon).
The estimated approximate mean line Dt ^ 0-3Dc is shown, and also in (a) the range of values for
S. halysitoides minor (from the figures given by Hill).

60

r

30

• Garra Fm.

+ Nemingha Lst

figures given in
SOSHKINA 1952

Spongophyllum halysitoides

minor

5

Qc (mm ) ►

To

1

n

Dc

4

5

Dc ( mm )-

10

TEXT-FIG. 12. Spongophyllum halysitoides. Scatter diagrams for {a) n:Dc, and (b) n/Dc:Dc for .S'.
halysitoides halysitoides from the Garra Formation and the Nemingha Limestone. In (o) the range of
values given by Soshkina (1952), and for S. halysitoides minor by Hill (1940c) are also shown.

D. L. STRUSZ: SPONGOPHYLLID AE FROM DEVONIAN GARRA FORMATION 575

‘Varieties’ of S. halysitoides have been established by both Hill (1940c) and Soshkina
(1949), distinguished principally on corallite size. I consider that their geographic
separation, taken with the morphological differences, suffice for these to be subspecies.
Etheridge jr.’s original species is a distinct subspecies. S. halysitoides minor Hill 1940
is diminutive (Dc = 2-4 mm.), with a very narrow tabularium; from the figures given by
Hill, its range of variation for Dt ; Dc falls below the area of scatter for the nominate
subspecies. It is known from the Coblenzian? of Silverwood, Qld., and the Urals.
S. halysitoides media Soshkina 1949, from the Eifelian of the Urals, is close to, indeed
may be synonymous with, the nominate subspecies (Dc — 5-6 mm.), but has better
developed septa. The size range, and number of septa, given for the species by Soshkina
(1952) — probably based on the Russian material — falls well above the scatter area for
n:Dc for S. halysitoides halysitoides (text-fig. 12), suggesting that the Russian material
may differ in this respect also. S. halysitoides major Soshkina 1949, also from the
Eifelian of the Urals, is large (Dc = 8-9 mm.), with well-developed septa; it should
perhaps be considered a separate species.

The specimen figured by Bulvanker (1958, pi. 64) from the Eifelian of the Kuznets
Basin has very well-developed septa — in fact in most corallites they appear to be
continuous — as well as smaller dissepiments and distinctly acanthophylloid tabulae.
It is probably a species of either Australophyllum or Xystriphyllum.

Range. The species is known from the Coblenzian of Eastern Australia, and the Coblen-
zian and Eifelian of the U.S.S.R. The nominate subspecies is known only from the
Coblenzian.

Known Garra localities. Ct-29?, E-24, E-30, E-31 ; Bt/430, Bt/936.

Spongophyllum rosiforme Yoh 1937

1937 Spongophyllum rosiforme Yoh, p. 54, pi. VI, figs. \a-c. Kwangsi, China; middle Middle
Devonian.

Diagnosis. Phaceloid Spongophyllum with corallites about 9 mm. in diameter, buttressed
by lateral expansions; septa few, irregular, confined to tabularium; one peripheral
series of large geniculate dissepiments, supplemented axially by intermittent small,
vertical elongate dissepiments; tabularium narrow, of close, gently concave tabulae.

Spongophyllum sp. cf. rosiforme Yoh 1937

Plate 91, figs. \a, b\ text-fig. 13

Material. SU 17167, loc. P-1.

Description. The corallum is phaceloid, composed of cylindrical corallites with frequent
lateral processes. The epitheca is apparently smooth, with gentle growth swellings. The
calice is deep, inversely conical to bell-shaped. Dc = 6-8 mm.

The septa, consisting of discontinuous crests on the dissepimental surfaces, are
unequal and not divisible into two orders, n = 14-16 in general. Most are confined to
the tabularium but exceptionally they may extend outwards as far as the epitheca. The
longest septa often interfinger axially. The thin epitheca is lined by a lamellar stereozone
0-2 mm. thick, in which septal bases are not visible. Septal microstructure is unknown.

576 PALAEONTOLOGY, VOLUME 9

Dt = jDc or less. The tabulae are close, and gently but irregularly sagging, at times
with upturned edges.

The wide dissepimentarium consists of a peripheral series of large globose to
geniculate lonsdaleoid dissepiments, and an incomplete inner series of very elongate,

TEXT-FIG. 13. Spongophyllim sp. cf. rosifonne. Transverse and longitudinal sections of SU 17167, X 3;

drawn from photographs.

Steeply inclined dissepiments, some of them lonsdaleoid, which may be supplemented
by occasional small convex plates.

Comparison. This specimen differs from Yoh’s only in a slightly smaller diameter, and
14-16 instead of 20 septa. The differences could be of subspecific value, but as the pos-
sible variation in both cases is unknown, I consider it best for the present simply to
compare them.

. „ . EXPLANATION OF PLATE 91

All hgures X4.

Fig. 1. Spongophylliim sp. cf. rosifonne Yoh 1937. 1<7, b, transverse sections of individual corallites
(cf. text-fig. 13), SU 17167, loc. P-1.

Fig. 2. Xystriphyllum dunstani (Etheridge fil. 1911). 2a, b, longitudinal and transverse sections,
paralectotype AM F 9492-3, AM 733 A, B, Clermont, Qld. Photographs by courtesy of the
Australian Museum, Sydney.

Palaeontology, Vol. 9

PLATE 91

STRUSZ, Devonian Spongophyllidae from New South Wales

D. L. STRUSZ; SPONGOPH YLLID AE FROM DEVONIAN GARRA FORMATION 577

Genus xystriphyllum Hill 1939

19396 Xystriphyllum Hill, p. 62.

19406 Xystriphyllum Hill; Hill, p. 269.

1940c Xystriphyllum Hill; Hill, p. 163.

1942c Xystriphyllum Hill, 1939; Hill, p. 147.

\9A2d Xystriphyllum-, Hill, p. 183.

1949 Xystriphyllum Hill; Stumm, p. 33.

1949 Steuophyllum-, Soshkina, p. 127 (partim).

1950 Xystriphyllum Hill, 1939; Hill, p. 140.

1950 Xystriphyllum Hill, 1939; Wang, p. 218.

1952 Steuophyllum Wdkd. ; Soshkina, p. 83 (partim).

71955 Pseudospongophyllum Zhmaev; Krayevskaya, p. 213, in Khalfin (partim'l).

1955 Stenophyllum Amanshauser, em. Wedekind; Krayevskaya, p. 211, in Khalfin (partim).
71955 Spougophyllum M. Edwards et Haime; Krayevskaya, p. 214, in Khalfin (partim'l).

1956 Xystriphyllum Hill; Clauss, p. 16.

1956 Xystriphyllum Hill, 1939; Hill, p. F304, in Moore.

1958 Stenophyllum (Amanshauser) emend. Wedekind, 1925; Bulvanker, p. \A6 (partim).
non 1960 Yi’iP7p6_v/6//;; Hill; Crickmay, p. 11.

71961 Xystriphyllum Hill 1939; Fontaine, p. 158 (partim'l).

1962 Xystriphyllum Hill 1939; Philip, p. 188.

1962 Xystriphyllum Hill, 1939; Soshkina and Dobrolyubova, p. 335, in Orlov.

Type species. Cyathophyllum dunstani Etheridge fil. 191 1, p. 3, pi. A, figs. 1, 2 (see below).

Diagnosis. Cerioid; septa long, wavy, thin axially, dilated peripherally; tabulae close,
complete or incomplete, concave with axial depression; dissepiments globose, axially
inclined, becoming steep at margin of tabularium; minor septa occasionally cut by
lonsdaleoid dissepiments.

Remarks. Many of the Russian species assigned to Stenophyllum Wedekind {non
Verhoeff 1897) are XyslriphyUitnr, others are Acanthophyllwn s.l. I would place the
following in Xystriphyllum-. Stenophyllum ahum Soshkina 1949, p. 128 (Eifelian, Urals);
S. uralicum Soshkina 1949, p. 129 (Eifelian, Urals); S. gorskii Bulvanker in Krayevskaya,
1955, p. 211, in Khalfin (Coblenzian, Kuznets Basin, Central Salair; Eifelian, Kuznets
Basin — Bulvanker 1958); this probably is synonymous with X. dunstani, which see;
S. sibiricum Bulvanker in Krayevskaya 1955, p. 211, in Khalfin (Eifelian, Kuznets
Basin); S. devonicum Bulvanker in Krayevskaya 1955, p. 211, in Khalfin (Eifelian,
W. Siberia, Kuznets Basin, and Altai mineral field); S. salairicum Zhmaev in Krayev-
skaya 1955, p. 212, in Khalfin (Eifelian, Kuznets Basin); S. soshkinae Zhmaev in
Krayevskaya 1955, p. 212, in Khalfin (Eifelian, Kuznets Basin); S. taimyricum Kravtsov
1963, p. 40 (Emsian, central Taimyr).

The following may also belong in Xystriphyllum: Spongophyllum minimum Zhmaev
in Krayevskaya 1955, p. 215, in Khalfin (Lower Devonian, ^2d2t\x)-, Pseudospongophyllum
massivum Zhmaev in Krayevskaya 1955, p. 213, in Khalfin (Eifelian, Kuznets Basin);
Spongophyllum halysitoides Etheridge of Bulvanker 1958, p. 132 {non Etheridge jr. 1918)
(Lower Eifelian, Kuznets Basin).

Range. The genus is known from the Lower Devonian of America and Czechoslovakia,
the upper Lower to basal Middle Devonian of Australia, the Coblenzian to Eifelian of
the U.S.S.R., the Couvinian of Laos?, and the Frasnian of Menorca.

578 PALAEONTOLOGY, VOLUME 9

Xystriphyllwn dimstani (Etheridge fil. 1911)

Plate 91, fig. 2, Plate 92; text-figs. 14-17

1911 Cyctthophyllum dimstani Etheridge fil., p. 3, pi. A.
non 1935 ICyathophyllnm dimstani Etheridge; Allan, p. 6, pi. V, figs. 4, 5. Reefton, N. Zealand;
Eifelian?

1939fi Xystrip/iylliini dimstani (Etheridge); Hill, p. 62, pi. V, figs. 5-8. Clermont, Qld. ;
Eifelian?

1940c Xystripliyllmn dimstani (Etheridge); Hill, p. 163, pi. Ill, figs. Aa, b. Silverwood, Qld.;
Eifelian?

1942c/ Xystripliyllmn dimstani (Etheridge); Hill, pp. 183, 184, pi. VI, figs. 2a, b. Wellington,
N.S.W. ; Emsian?

?1955 StenopliyUmn gorskii Bulvanker in Krayevskaya, p. 211, pi. XXXVII, fig. 2, in Khalfin.

Kuznets Basin and Central Salair, U.S.S.R.; Coblenzian.

?1958 StenopliyUmn gorskii Bulvanker (partim)-, Bulvanker, p. 147, pi. LXXI, figs. \a, b, 2,.
pi. LXXll, figs, la, b. Kuznets Basin, U.S.S.R.; Eifelian.

Lectotype. GSQ Cl. 6; subsequent designation Hill (1939fi) p. 62, where figured plate V, figs. 5-8.
Type locality: Douglas Ck., Clermont, near Rockhampton, Qld. Eifelian? Paralectotype AM
F 9492-9493 (sections AM 733), figured Etheridge jr. (1911), pi. A, figs. 1, 2; figured herein plate 93,,
figs, la, b.

Diagnosis. Xystriphylhim with corallites of average diameter 6-8 mm., in which the
major septa interdigitate axially.

Description. I have collected a large amount of material from the Garra Formation,
and have compared it with the lectotype, and with other material in the collections of
the University of Queensland, the University of Sydney, and the Australian Museum.
The following description is based on all of this, but principally on the extensive
Garra material.

The corallum is massive and may reach a large size. Thus SU 17184 (loc. P-22) is
30 cm. high and over 30 cm. across, and is part of a non-radiating corallum. The
holotheca is unknown. Calices are deep, conical to slightly bell-shaped, with steep,
straight to moderately convex walls, narrow rims, and narrow concave bases. In-
dividual corallites are 4- to 8-sided, generally with straight or slightly crenulate walls.
The majority of adult corallites are 6-8 mm. in diameter (see text-fig. ISn). The corallite
epitheca is longitudinally ribbed, but shows only sparse and very faint transverse growth
marks. The wall structure comprises a thin dark median surface, representing the
epitheca of adjacent corallites (the ‘axial plate’ of Flower 1961). On either side of this
is a layer of calcite fibres growing perpendicularly outwards, forming a non-septal
stereozone. Quartz euhedra are common within these wall tissues in the Garra coralla.

Up to 44 septa have been counted, but the usual maximum for a corallum is 38-40.
Normal adult corallites contain 32-36 septa (see text-fig. 16). The septa, of two orders,
are thin in the tabularium, becoming gradually dilated outwards, until about 0-2-0-3 mm.
from the periphery, where there is a rapid wedge- wise dilatation. These prominent septal
bases merge with the fibrous wall tissue to form a narrow {circa 0- 1 mm.) composite stereo-
zone. The septal bases of adjacent corallites may be opposite or alternate. The major
septa are unequal, and approach or interdigitate at the axis; generally one extends
across the axis, or two opposite major septa may unite axially. The minor septa are

D. L. STRUSZ: SPONGOPHYLLIDAE FROM DEVONIAN GARRA FORMATION 579

TEXT-FIG. 14. Xystriphylliim dimstani, SU 13263, x6. («)-(<?) : transverse sections showing stages in
the growth of a juvenile corallite. (/) longitudinal section through a single bud; (1) inception of bud,
(2) appearance of first series of dissepiments, at a length of about 3 mm., (3) adult appearance attained,

at a length of about 9 mm.

580

PALAEONTOLOGY, VOLUME 9

generally |-| the length of the major, and extend some distance into the tabularium.
The septa are commonly sinuous but never flanged.

Dt = circa JDc. The tabulae are crowded, complete and incomplete, moderately to
deeply concave or often strongly funnel-shaped. Their margins are upturned.

There are 2-4 series of globose dissepiments, rather irregular in size. They are axially
inclined, the inclination being moderate peripherally, increasing inwards. Those at the
inner margin of the dissepimentarium are very steeply inclined or vertical and often
elongate.

Ontogeny. Budding and the development of corallites has been reconstructed from the
large number of sections made, although serial sections were not used (text-fig. 14).

Budding is peripheral. The bud forms in the corner of the parental calice with an
initial diameter of about 1-1-5 mm. The first skeletal element seen in transverse section
is a very short segment of wall tissue within the dissepimentarium (text-fig. 14a). The
initial septa are inserted immediately in a very irregular pattern — the typical fourfold
rugosan pattern apparently does not occur. The ends of the wall segment rapidly
expand to meet the walls of the parent corallites, and septal insertion continues at a
rapid rate (text-fig. \Ab-d). At a diameter of about 2 mm. the first series of dissepiments
appears; the corallite at this stage is about 3 mm. long (text-fig. 14/). Once the young
corallite is completely enclosed by wall tissue it expands by a lengthening of all its wall
segments, and so quickly assumes a polygonal shape, simulating an ‘inter-mural offset’
(text-fig. \Ae). The second series of dissepiments appears about 4-5 mm. above the
proximal extremity.

The above sequence suggests that in most, if not all, massive rugosan corals, the so-
called ‘inter-mural increase’ (see Hill, p. F248, in Moore 1956) is in fact peripheral
calical increase.

Variation (text-figs. 15-17). Dc, Dt, and n were measured for 6 coralla from the Garra
Formation (up to 100 measurements of each), as well as 37 measurements of sections
of 3 north Queensland coralla. It was found that frequency curves for all but Dc were
of little value. Dc was plotted with a class interval of 0-7 mm. ; the curves show consider-
able variation. Most are bimodal, with peaks at about 3-4 mm. and 5-7 mm. SU 5282
(loc. MM-12) and 13263 (loc. Cr-100) have skewed unimodal curves.

Bivariate plots of n:Dc for three coralla from loc. P-22 showed no significant
differences, but that for the Queensland coralla was inconclusive (clearly about 100
measurements are needed from one corallum for great accuracy). In both this set of
curves, and that for n/Dc:Dc, scatter diagrams show a marked change of slope in the
region of Dc = 3-5 mm., and n = 28. This is interpreted as indicating rapid septal
insertion in juvenile polyps, followed by slow insertion after the attainment of maturity,
at an average diameter of 3-5 mm.

Comparison. As noted by Hill (19396), X. dimstani is very similar to X. /nzYcAc/// (Etheridge
fil. 1892). The latter has proportionately more septa, which show less tendency to inter-

EXPLANATION OF PLATE 92

Xystriplivllum dimstani (Etheridge fil. 1911). Transverse and longitudinal sections, SU 13263, loc.
Cr-100; X4.

Palaeontology , Vol. 9

PLATE 92

STRUSZ, Devonian Spongophyllidae from New South Wales

D. L. STRUSZ: SPONGOPHYLLID AE FROM DEVONIAN GARRA FORMATION 581

digitate axially. Also, coralla of X. mitchelli from the Murrumbidgee Devonian may have
much larger corallites (up to Dc = 18 mm.). Plots of n:Dc and n/Dc:Dc show signifi-
cant differences from those for X. dimstani (text-figs. 16, 17). The plot for n/Dc:Dc has
the same form as that for X. dimstani — perhaps, therefore, a generic character — but
differs in that the change of slope occurs at Dc = circa 6 mm. A further morphological

TEXT-FIG. 15. XystriphyUum spp. Frequency polygons for Dc. («) five Garra coralla and coralla from
north Queensland localities (combined), X. dinislatu \ (b) coralla of X. iiisigne and X. mitchelli, including

* type localities.

distinction between the two species is in the nature of the peripheral septal dilatation.
In X. dimstani this is wedgewise and distinctly triangular in transverse section. In
X. mitchelli the dilated portions are longer and often ogival in section.

Scatter diagrams for X. insigne Hill 1940 are much closer to those for X. dimstani than
are those of X. mitchelli (text-figs. 15-17), but the two species are readily distinguishable
morphologically. X. insigne is smaller than X. dimstani, with fewer septa and propor-
tionately thick walls.

Remarks. Stenophylhmi gorskii Bulvanker in Krayevskaya 1955, from the Coblenzian
and Eifelian of Asian U.S.S.R., cannot be distinguished from the type sections of
X. dunstani. Despite the considerable geographic separation, I at least tentatively place

40-

h

20-

0

0

t"

40-

DC ( mm )

5

20-

10

• Clermont
+ Silverwood
o Thornb orough

Dc (mm )

ID

• XvstriDhvUum m it chell i
+ XystriphyU um insiqne

Dc ( mm )

o-J ^ 1 1

0 5 10 15 20

TEXT-FIG. 16. XystripbyUum spp. Variation of n;Dc. (a) scatter diagram for X. diinstani, SU 17183
(Garra Formation) with mean curve obtained from density-contouring of plot; {b) idem, north
Queensland X. dunstani; (c) scatter diagrams, X. insigne and X. rnitchelli; (d) comparison of mean
curves and limiting polygons for the three species.

D. L. STRUSZ: SPONGOPHYLLIDAE FROM DEVONIAN GARRA FORMATION 583

them in synonymy. This is to a certain extent supported by the presence in the same
region of Rhizophyllum enorme Etheridge fil. 1903 (Bulvanker 1958, p. 46), another
Garra species.

Range. Xystriphyllwn chmstani is now known from a number of localities in the Lower
and Middle Devonian of Eastern Australia. Hill (1943) considered the type locality —
Douglas Creek, Clermont, N. Qld. — to be Couvinian, and equivalent to part of the
Murrumbidgee succession. The presence of Lyrielasma aggregatum (see p. 570) supports

XvstriphvU um dunstani : SU 17183, loc. P- 22
XvstriphvUum insigne : all localities

Xvstriphvllum mitchell i : all Localities

0-

0

D c (mm)

10 15 20

TEXT-FIG. 17. Xystriphyllum spp. Scatter diagram and limiting polygons for n/Dc;Dc, X. chmstani
(SU 17183), X. insigne, and X. mitchelli, showing the close similarity of form for the three species.

this; I feel the type locality may be as old as Emsian, although the fauna is too small for
accuracy.

The Silverwood fauna agrees very closely with that of the Buchan-Taemas-Garra
faunal association, and is also probably Emsian, as pointed out by Hill (1943). In the
Garra Formation X. chmstani apparently is commonest in a biostrome high in the suc-
cession, and may be Emsian or basal Eifelian. Philip (1960, p. 149) reports X. chmstani
from the limestone at Toongabbie, Victoria, an horizon which he correlates with his
Gedinnian Cooper’s Creek Formation. Professor Hill’s collection contains specimens
from the Middle? Devonian of Thornborough, N. Qld., and the Eifelian? of the Broken
River, N. Qld. (locality BRS 48 — see Hill and Denmead 1960, fig. 21). SU 3193 is from
the Emsian or Eifelian of Wee Jasper, N.S.W.

X. dunstani would therefore seem to be characteristic of the Emsian and Eifelian of
Eastern Australia, possibly occurring as early as the Gedinnian. It apparently occurs in
rocks of a similar age in the Kuznets Basin and Central Salair, Asian U.S.S.R.

Known Garra localities. Cr-100 (common), Ct-11, Ct-13, Ct-15, Ct-22, MM-11, MM-12, P-22
(common), P-25; Bt/1838.

584

PALAEONTOLOGY, VOLUME 9

Xystriphyllum magnum Hill 1942

Plate 93, fig. 1

1942c Xystriphyllum magnum Hill, p. 147, pi. Ill, figs, la, b. Attunga, Tamworth, N.S.W. ;
Emsian or early Eifelian.

1942cf Xystriphyllum magnum Hill; Hill, p. 183, pi. VI, fig. 3. Wellington, N.S.W. ; Emsian?

Holotype. SU 7270. Type locality: pors. 88/115, ph. Burdekin, behind Sulcor Quarry, Attunga, N.S.W.
Sulcor Limestone.

Diagnosis. Xystriphyllum with large corallites, up to 28 mm. in diameter, and about
22 long septa of each order. (After Hill 1942c, p. 147.)

Remarks. No further specimens have been found. The only comment that may be added
to the description of Hill (1942c) is that the corallites in the type sections are up to 28 mm.
across, although Hill gives the maximum as 20 mm. The Garra specimen has a maximum
diameter of 20 mm.

Genus australophyllum Stumm 1949

71888 Endophylhmv, Foerste, p. 131.

71889 Lonsdaleial', Etheridge fil., p. 22.

1899 Spongophyllunr, Etheridge iW., p. 158.

1911 Spongophyllum; Etheridge fil., pp. 7-8 (partirn).

71925 Spongophyllunv, Cha.prr\an, pp. \ \2-\5 (partiml).

1932 Spongophyllum', Jones, p. 50 ff. (partirn).

1939/) Spongophyllum', Hill, pp. 60-61.

71940c Spongophyllum', Hih’, p. (partiml).

1940 Spongophylloidesl; Hill and Jones, p. 181.

1942c Spongophyllum', Hill, p. 161 (partirn).

1949 Australophyllum Stumm, p. 34.

1956 Australophyllum', Hill, p. F306, in Moore.

71960 Xystriphyllum', Crickmay, p. 11.

1961 Australophyllum', Lenz, p. 509.

71961 Xystriphyllum', Fontaine, p. 158.

1962 Spongophyllum', Birenheide, p. 69 (partirn).

1962 Neomphyma', Soshkina and Dobrolyubova, p. 335, in Orlov (partirn).

71962 Australophyllum', McLaren, p. 8, in McLaren, Norris, and McGregor.
non 1964 Australophyllum', McLaren, p. 13, in McLaren and Norris.

Type species. Spongophyllum cyathophylloides Etheridge fil. 1911, p. 7, pi. A, fig. 3, pi. C, figs. 1, 2.
Subsequent designation Stumm 1949, p. 34.

Diagnosis. Cerioid, with septa as in Acanthophyllum; tabulae close, concave with axial
depression; dissepimentarium wide, with well-developed lonsdaleoid dissepiments;
inclination of dissepiments increasing axially, becoming steep at margin of tabularium.

EXPLANATION OF PLATE 93

Fig. 1. Xystriphyllum magnum Hill 1942. Oblique transverse section, SU 5292, loc. MM-10, figured
Hill 1942c/; X 3. Photograph supplied by Dr. T. B. H. Jenkins, University of Sydney.

Fig. 2. Australophyllum^ spongophylloides (Foerste 1888). la, transverse section showing similarity to
typical Spongophyllum, lb, longitudinal section showing acanthophylloid tabulae as in Australo-
phyllum', UCT F 1621, Hume Limestone, Yass, N.S.W.; x 2.

Palaeontology, Vol. 9

PLATE 93

STRUSZ, Devonian Spongophyllidae from New South Wales

D. L. STRUSZ: SPONGOPHYLLID AE FROM DEVONIAN GARRA FORMATION 585

Discussion. Stumm (1949) considered the type species to be related to Hexagonaria,
differing from ‘typical Spongophyllum' in its wide dissepimentarium and carinate septa.
This interpretation is wrong on two counts: firstly, Stumm apparently based his concept
of Spongophyllum principally on the American species, which have very narrow dis-
sepimentaria. These Birenheide (1962, p. 81) has since referred to the new genus
SmithiphyJIum, which Pedder (1964, p. 436) suggests may be an Endophyllid. Australo-
phyUum is in fact quite close to true Spongophyllum, probably being derived from it.
Secondly, Stumm ’s interpretation of the type species is inadequate. " S. ' cyathophylloides
does not have septal carinae, such as occur in the Disphyllidae, but the flanges and
axial septal processes of the Spongophyllidae; moreover, the structure of both dis-
sepimentarium and tabularium, and the septal microstructure, differ from those of
Hexagonaria.

Birenheide (1962, pp. 50, 69, 72) places Australophyllum in synonymy with Spongo-
phyllum on the grounds that septal processes (the character stressed by Stumm) occur
also in true Spongophyllum. However, after examining the holotype and topotypes of
A. cyathophylloides, I consider that there are sufficient differences to warrant retention
of the genus. The dissepimentarium is better developed than in S. sedgwicki and other
European Spongophyllum. More significantly, the tabulae are of the ptenophylloid type
— concave, often incomplete, with a median depression — rather than the more regular
spongophylloid type, with no median depression. As pointed out by Hill (19396, p. 62),
the resemblance of the septal arrangement and tabulae to those of Xystriphyllum is
striking, the only significant difference being in the far greater development of lonsdaleoid
dissepiments.

I would refer the following Australian species to Australophyllum: Spongophyllum
eyathophylloides 'EXhQndgQfiX. 1911 — type, Clermont, N. Qld., Eifelian?; S. giganteum
Etheridge fil. 1899 (Hill 1942c, p. 161), Tamworth, N.S.W., Givetian; Spongophylloides?
thomasae Hill and Jones 1940, Cudal, N.S.W., Emsian? (see below).

In addition, I feel that the Silurian species Spongophyllum spongophyUoides (Foerste
1888) from Yass (see Hill 1940a) is sufficiently close to the Devonian species of
Australophyllum to be considered a probable ancestral, or near-ancestral, species. I here
assign it tentatively to Australophyllum (see PI. 93, figs. 2a, b).

Xystriphyllum hyperbolieum Crickmay 1960, placed in Australophyllum by McLaren
(in McLaren, Norris, and McGregor 1962), differs in having the flat tabulae of
Spongophyllum, although the septa and dissepimentarium are similar to those of
Australophyllum. Its affinities are uncertain. Probably allied to it is Australophyllum^
cf. A. 2 thomasae of McLaren (p. 13, in McLaren and Norris 1964), which differs from
true Australophyllum in having axially domed, peripherally depressed tabulae. The close
similarity of this form to the Australian species is superficial. The only remaining North
American record is that of Lenz (1961); I cannot be sure from his figures of the generic
affinities of this form.

Xystriphyllum laosense Fontaine 1961 is almost certainly Australophyllum, charac-
terized by considerable variation in the development of lonsdaleoid dissepiments; how-
ever, these do not apparently become as rare as in X. dunstani.

Range. Wenlockian?, Emsian to Givetian of Australia; Eifelian? of Laos; Eifelian
of Canada?

586

PALAEONTOLOGY, VOLUME 9
Australophylluni thomasae (Hill and Jones 1940)

Plate 94, figs. \a, b\ text -fig. 18

1940 Spoiigopliylloides(l) thomasae Hill and Jones, p. 181, pi. IV, figs. \a, b. Cudal, N.S.W.,
Emsian?

non 1964 Aiistralophylliinil cf. AA thomasae (Hill and Jones); McLaren, p. 13, pi. IV, figs, ha-c,
in McLaren and Norris. Horn Plateau, Canada, Givetian.

Holotype. SU6188(= NS 1296) (PI. 94, figs. 1 u, Z>). Type locality : pors. 2, 3, ph. Cudal, west of Orange,
N.S.W. (probably my loc. Cu-2). Garra Formation, Emsian?

Diagnosis. Large Aitstralophylhim with large lonsdaleoid dissepiments in several series;
tabulae deeply funnel-shaped; septa rarely developed in zone of lonsdaleoid dissepi-
ments; intercorallite walls lack septal bases.

Description. See Hill and Jones 1940, p. 181.

Remarks. The peripheral septal crests described by Hill and Jones are confined to a
narrow zone at the outer ends of the septa, and almost exclusively represent minor septa,
resting on an innermost series of ‘secondary’ lonsdaleoid dissepiments (see Birenheide
1962, pp. 46-48). In a wide zone outside this, large geniculate lonsdaleoid dissepiments
are developed, and in this zone there are no septal crests; a very few major septa extend
across the zone to reach the epitheca.

The one offset visible in the transverse section of the holotype resembles in every way
the peripheral calical buds described herein (pp. 71-72) for Xystriphyllum diinstani.

Known localities. Cu-2 (or nearby); Ct-14? (one poor specimen).

Australophylhirnl sp. cf. thomasae (Hill and Jones 1940)

Plate 94, figs. 2-5, Plate 95, figs. 1-3; text-fig. 18

Material. 28 specimens (12 sectioned) from loc. Ct-18; one each from Iocs. Ct-28, CAT/2193.

Description. The corallum form is a little uncertain. The specimens are from a pink
crinoidal calcarenite at the base of a bioherm; their appearance in the outcrop suggests
that they grew unattached in the sand. Many of the corallites appeared to be solitary,
conical, with sharp rejuvenescence. Others had flaring distal ends, giving rise to several
buds which then became a small cerioid corallum. Most of the material collected from
the massive limestone is fragmentary, and the few sections of colonial coralla do not
conclusively show whether they are truly cerioid, closely phaceloid, or even at times
astraeoid. In several cases critical parts of the sections show solution phenomena or
broken skeletal tissues. The calice is broadly bell-shaped to slightly everted, with a wide.

, EXPLANATION OF PLATE 94

All figures X 2.

Fig. 1. Aiistralophylliim thomasae (Hill and Jones 1940). lu, transverse section (note peripheral bud,
lower right), \b, longitudinal section, holotype SU 6188, loc. Cu-2. Photographs by courtesy of
Prof. D. Hill.

Figs. 2-5. Australophylluml sp. cf. thomasae. 2, transverse section of two corallites showing cerioid?
habit; SU 18114, loc. Ct-18. 3, transverse section showing strongly dilated septa; SU 14155, Ct-18.
4, longitudinal section, SU 14149, loc. Ct-18. 5, transverse section of three corallites showing
phacelo-cerioid habit; SU 20125, loc. Ct-18.

Palaeontology, Vol. 9

PLATE 94

STRUSZ, Devonian Spongophyllidae from New South Wales

D. L. STRUSZ: SPONGOPH YLLID AE FROM DEVONIAN GARRA FORMATION 587

flaring, sharp-edged rim, and a narrow inversely conical to funnel-shaped axial pit.
Diameters up to 46 mm. have been measured, adults apparently varying from about
16 mm. to 46 mm. (or more?).

There are generally 62-72 septa. The major septa extend well into the tabularium,
where they are bilaterally or quadrilaterally arranged about two opposite septa which
may meet or pass each other at the axis. There may or may not be a slight axial vortex.
The minor septa vary greatly in length. Most do not reach the inner margin of the
dissepimentarium, but an occasional one extends well into the tabularium. In the
tabularium the septa are thin and may be strongly waved. The degree of dilatation in
the dissepimentarium is highly variable. In some the septa remain only slightly dilated
throughout. Most corallites have strongly dilated septa, which in some cases are in
contact over much of the dissepimentarium. The dilatation may increase continuously
outwards or may take the form of a series of wedges on succesive dissepiments. There is
a peripheral stereozone up to 1 mm. wide formed by wedge-wise dilatation of the septal
bases.

The septa are generally highly discontinuous, the minor more than the major; the
degree of discontinuity increases outwards, and there may be a narrow peripheral zone
of lonsdaleoid dissepiments. The major septa are generally continuous near the
tabularium, and always within it, as are those few elongate minor septa. In extreme cases
the septa consist of short wedge-shaped crests on successive series of dissepiments. In the
tabularium the septa are zigzag flanged, frequently strongly; in those corallites where
the septa are relatively continuous and thin, these flanges may extend out into the
dissepimentarium.

The septa consist of monacanthine trabeculae about OT mm. thick. In longitudinal
sections these are seen to be arranged as in Acanthopliyllum: half-fans in the dis-
sepimentarium, again becoming nearly vertical in the tabularium. In thin parts of the
septa they are in one series, but in dilated parts all regularity of arrangement is lost.
In these parts the septa consist of numerous trabeculae set in lamellar sclerenchyme,
the lamellae being parallel to the attitudes of the dissepiments. The dilatation, both
trabecular and lamellar, may spread laterally as thin coatings on the dissepiments.

Dt is a little over OT Dc, decreasing gradually with age. The tabular floors are
moderately to strongly funnel-shaped, generally with upturned margins. The tabulae
are mostly incomplete; they are crowded, irregularly sagging to slightly convex, and
much broken up by the often wavy ends of the major septa.

The wide dissepimentarium consists of large elongate dissepiments. Peripherally,
these are slightly to strongly axially inclined and usually only slightly convex. Over the
outer half of the dissepimentarium they may be horizontal, even slightly peripherally
inclined, or more usually gently inclined axially; in this region they are generally moder-
ately convex. Towards the axis the dissepiments become increasingly steeply axially
inclined, and increasingly smaller, until at the inner margin of the dissepimentarium
they are vertical, and again fairly elongate.

Budding is apparently peripheral, calicinal.

Variation. Plots of Dt:Dc and Dt/Dc:Dc show only slight scatter, and suggest that
approximately Dt = OTDc+0-5. However, the data are insufficient for accuracy and
certainty. There are insufficient data for n : Dc to be of any use.

R r

C 4397

588

PALAEONTOLOGY, VOLUME 9

Comparison. In size and number of septa individual corallites of this form are very close
to those of A. thomasae. However, the latter has a wider tabularium (see text-fig. 18).
Other points of difference from the holotype of that species are several: the strongly
differentiated wide zone of lonsdaleoid dissepiments is lacking; the septa are far more
discontinuous, and the septal crests extend well outside the ends of the continuous
portions, if, indeed, the septa do not consist entirely of crests in the dissepimentarium; in
dilated portions, the septa consist of numerous trabeculae embedded in sclerenchyme.

TEXT-FIG. 1 8. Australophyllum thomasae, A ? sp. cf. thomasae. Scatter diagrams for Dt : Dc and Dt/Dc : Dc,
with estimated approximate mean curve Dt = OTDc+0-5 for sp. cf. thomasae.

and not arranged in a single series ; finally, the peripheral stereozone consists of wedge-
shaped septal bases (that of A. thomasae is of undifferentiated fibrous calcite). It is
likely that this form is a distinct species, but this cannot be ascertained with certainty
until {a) topotypic material is available of A. thomasae, and {b) the corallum form is
established.

Known localities. Ct-18 (= Bt/352), Ct-28?; CAT/2193. All the same horizon, a crinoidal biostrome.

EXPLANATION OF PLATE 95

All figures X 2.

Figs. 1-3. Australophyllnml sp. cf. thomasae (Hill and Jones 1940). 1, longitudinal sections of several
corallites, possibly forming a loosely dendroid corallum (note small peripheral calical bud at b)\
SU 20126, loc. Ct-18. 2, transverse section, SU 14162, loc. Ct-18. 3, transverse section of solitary
corallite showing wall structure; SU 14159, loc. Ct-18.

Fig. 4. Australophyllum bilaterale sp. nov. Transverse section, holotype SU 15158, loc. E-1.

Palaeontology, Vol. 9

PLATE 95

STRUSZ, Devonian Spongophyllidae from New South Wales

D. L. STRUSZ: SPONGOPHYLLIDAE FROM DEVONIAN GARRA FORMATION 589

Australophylliim bilaterale sp. nov.

Plate 95, fig. 4; Plate 96, fig. 1 ; text -figs. 19, 20

Derivation of name. Latin bis = i'NO^o\d-\- lateralis = lateral; refers to the marked bilateral septal
arrangement.

Holotype. SU 15158 (PI. 95, fig. 4; PI. 96, figs. 1«, b). Type locality: E-1.

Diagnosis. Large Australophylliim with septa typically arranged in four subequal
quadrants, having bilateral symmetry; minor septa reduced to septal crests; occasional
very large dissepiments appear in a wide lonsdaleoid dissepimentarium.

Description. The corallum is cerioid and large — the holotype was measured in the field
as about 50 x 25 cm., and 25 cm. high. The prismatic corallites are not greatly divergent.
The epitheca bears narrow septal grooves, and in longitudinal section is seen to be
finely wrinkled. The calice is deeply inversely conical, with a slightly rounded rim and a
concave base.

Adult corallites are generally 16-20 mm. in diameter, the maximum known being
22 mm. ; they are bounded by thick, straight, or slightly curved walls. The thin epitheca
is lined by a stereozone formed by strong wedge-shaped bases; these are generally
opposite, giving the wall a strongly beaded appearance in transverse section. There are
usually between 50 and 64 septa; these are attenuate inside the peripheral stereozone,
often strongly waved or zigzag, but are unflanged. The major septa are long, continuous
in the tabularium and inner dissepimentarium, but usually discontinuous peripherally.
The longest four septa divide the septal apparatus into four subequal quadrants,
arranged on either side of the plane containing two of the four septa (C and K?),
which reach, and may combine at, the axis. The other major septa are of variable
lengths, the shortest being midway between the longest; the shortest are about f R. The
minor septa are more variable in length, generally being about R. They are
generally discontinuous, particularly towards the periphery, where the individual septal
segments commonly only reach half-way to the succeeding dissepiment.

The septa are constructed of very fine trabeculae directed inwards and upwards from
the periphery at a fairly constant angle.

The tabularium is relatively narrow, Dt = 0-2 to 0-3 Dc. The tabular floors are
moderately concave, with upturned margins and an axial deepening. The thin tabulae
are crowded, complete and incomplete, gently concave to slightly convex.

There are up to four series of peripheral lonsdaleoid dissepiments, although generally
only the outermost of these interrupts the major septa. There are occasional very large
lonsdaleoid dissepiments which interrupt several major septa. Inside the zone of
lonsdaleoid dissepiments are several series of rather more elongate normal dissepiments,
whose inclination is steep, increasing to nearly vertical at the margin of the tabularium.

Budding is probably peripheral, but has not been seen in longitudinal section. In
transverse section the evidence is inconclusive (see text-fig. 20). In three sections there
are small trigonal buds at the intersection of three large corallites; these are about
2 mm. in diameter, with a single series of lonsdaleoid dissepiments and a few very short
septal segments on these. They lack a peripheral septal stereozone or prominent septal
bases: the epitheca is lined with fibrous sclerenchyme about 0-02 mm. thick. In two
other sections even smaller lenticular buds have been seen along the wall between two

590

PALAEONTOLOGY, VOLUME 9

corallites. These also lack a septal stereozone, and only one has any trace of internal
structure; one dissepiment?, and possibly very tiny peripheral septal traces. The evidence
suggests peripheral budding, with the corallite rapidly becoming prismatic, but does not
prove it.

TEXT-FIG. 19. Aiistralophyllum bilaterale sp. nov. Scatter diagrams and limiting polygons for (o) Dt:Dc
— reference lines Dt = 0-3Dc, Dt = 0-2Dc, added — (b) n:Dc, and (c) n/Dc:Dc.

„ ^ ^ EXPLANATION OF PLATE 96

All figures X 2.

Fig. 1. Australophylhmi bilaterale sp. nov. In, b, transverse and longitudinal sections, holotype SU
15158, loc. E-1.

Figs. 2-4. Taimyrophyllum expansum (Hill 1942). 2, longitudinal section, SU 14103, loc. Cr-111.
3, transverse section of a specimen with large Ts; holotype, SU 7290, loc. Cr-113. 4a, b, transverse
and longitudinal sections of a specimen with small Ts; SU 14102, loc. Cr-111.

Palaeontology, Vol. 9

PLATE 96

STRUSZ, Devonian Spongophyllidae from New South Wales

D. L. STRUSZ: SPONGOPH YLLID AE FROM DEVONIAN GARRA FORMATION 591

Variation. Only the holotype was well enough preserved for measurement. Reasonable
plots were obtained for this for Dt:Dc, n:Dc, and n/Dc:Dc. (text-fig. 19). The first
shows a slight change in slope at Dc = circa 12-13 mm. Dt changes gradually from
about 0-3Dc in juveniles to about 0-2Dc for adults.

In the plots involving n, the change at Dc = 12 mm. cannot be seen; in n:Dc there is,
however, apparently a change at about Dc = 8-9 mm. This does not show up very
strongly on the plot n/Dc : Dc, which shows a gradual change in the ratio with increasing
Dc.

TEXT-FIG. 20. AustralophyUinn bilaterale sp. nov. Transverse sections of holotype SU 15158, xlO,
showing stages in ontogeny. (<7) Dc = T4 mm., n ^ 4?; the smallest bud seen, forming between two
adults, so suggesting it may be peripheral calical rather than intermural, (b) Dc -^= 2 0 mm., n = 6;
a ‘corner’ bud; note the lack of septal bases peripherally, (c) Dc 2-6 mm., n =- 22 or 24?; in this
bud, the minor septa are apparently just being introduced as very low ridges on the corallite walls.

Comparison. A. bilaterale is intermediate in size between A. cyathophyiloides (the type
species), and A. giganteum and A. tlioniasae. A 7 spongophylloides, from the Silurian of
Yass, is almost as large, but has far fewer septa, and a very wide lonsdaleoid dissepimen-
tarium. A. bilaterale is distinguished from all of these by the marked bilateral symmetry
of its septal arrangement.

Known localities. Ct-17 (1 corallum), E-1 (1 corallum).

Genus taimyrophyllum Chernyshev 1941

1941 Tainnrophvllinn Chernyshev, pp. 12, 53.

1942c Eddastraea Hill, p. 147.

1942^/ Eddastraea Hill; Hill, p. 184.

1956 Eddastraea Hill; Hill, p. F306, in Moore.

1958 7c/hmTop/n7/im/ Chernyshev; Bulvanker, p. 154.

1964 Taimyrophyllum Chernyshev 1941 ; Pedder, p. 436.

Type species. Taimyrophyllum speciosum Chernyshev 1941, pp. 12, 53, text-fig. 1, pi. 1, figs. 1-3, pi. 2,
figs. 1-3, pi. 5, fig. 5; Tarei R., Taimyr, U.S.S.R.; Lower Devonian.

Diagnosis. See Pedder 1964, p. 436.

Remarks. The holotype has been refigured by Pedder (1964, pis. 63, 64).

592

PALAEONTOLOGY, VOLUME 9
Taimyr ophyllum expansum (Hill 1942)

Plate 96, figs. 2^

\9A2d Eddastraea expansa Hill, p. 184, pi. VI, figs. 5a, b.

Holotype. SU 7290 (PI. 96, fig. 3). Type locality: Cr-1 13.

Diagnosis. Taimyrophyllum with few, rather large corallites, with very wide dissepimen-
taria, and with dilated septa. (Modified after Hill 1942.)

Description. From an examination of the holotype, and additional material, the follow-
ing may be added to Hill’s description:

(1 ) Dt is generally about 3-4 mm., reaching 6 mm. only in the holotype. Similarly, n is
generally less than the 72 given by Hill; in SU 14102 it is 50-56.

(2) The tabulae are slightly to moderately sagging with upturned margins, and have
a very shallow median depression.

(3) The outer parts of the septa are frequently modified by lateral dissepiments.

(4) The amount of septal dilatation is quite variable, from slight to very great.

(5) Dc cannot generally be accurately measured. In the additional material, Ts is
highly variable, being generally 10-15 mm.

Comparison. All other species except T. carinatum Bulvanker 1958 have attenuate or only
weakly dilated septa. That species differs from T. expansum in being markedly aphroid
(see Bulvanker 1958, p. 155, pis. 76, 77). In other respects the two species are quite
similar.

Known localities. Cr-100, Cr-1 11, Cr-1 13.

FOSSIL LOCALITIES

Localities are listed in the same way as in Strusz (1965Z)); those already listed there do not appear

here. A full list of localities, together with detailed maps, may be found in Strusz (1963).

Parish of Bell, County of Ashburnam

Be-7: portion 175 (north end); grid reference 177.911 (Bathurst); Mandagery Ck., 150 yds.

south-south-west of road crossing (‘Crystal Springs’). Fossiliferous calcirudite.

Parish of Bryniedura, Co. Ashburnam

Br-3: portion 73 (south); grid ref. 173.907 (Bathurst); in gully just west of road, and 50 yds.

north of road junction. Calcarenite.

Br-10: junction, portion 81, 104, 105; grid ref. 174.899 (Bathurst); inroad. Interbeddedcalcilutite

and calcareous siltstone (‘rubbly limestone’).

Parish of Cardington, Co. Gordon

Ca-1 : portion 106 (extreme north-east corner, near portion 136); grid ref. 179.935 (Dubbo);

north-east of road. Thinly bedded fine calcarenite.

Parish of Curra, Co. Gordon

Cr-1: portion 103 (southeast); grid ref. 188.968 (Dubbo); about 120 yds. west of road, and

about 50 yds. north of portion 97. Pellet — oolite calcarenite.

Cr-2: portion 150 (south-west corner); grid ref. 188.970 (Dubbo); near small abandoned

quarry west of road. Pellet — oolite calcarenite.

D. L. STRUSZ: SPONGOPHYLLIDAE FROM DEVONIAN GARRA FORMATION 593

Cr-5: portion 171 (centre, north-east side); grid ref. 184.968 (Dubbo); in Curra Ck. Thickly

bedded calcarenite.

Cr-55: portion 172 (south-west); grid ref. 183.967 (Dubbo); south side of Wellington-Arthur-

ville road, between Curra Ck. road junction and Fingerpost, 218 yds. north-east of
junction of pors. 166, 167, 172, 173. 3-ft. band of rubbly calcarenite.

Cr-77: portion 99 (south-east); grid ref. 188.970 (Dubbo); 45 yds. west of portion 142, 120 yds.

north of portion 100. Calcarenite.

Cr-86: portion 44 (south); grid ref. 187.966 (Dubbo); 65 yds. west of portion 48, 110 yds.

north of southern boundary, portions 44, 84. Pelletal calcilutite.

Cr-89: portion 39; grid ref. 187.964 (Dubbo); 130 yds. south of portion 111, 250 yds. south-west

of junction, portions 111, 39, and the Bell River. Biostromal limestone (about 20 ft.
thick). Same horizon as Cr-94, -100, -111, -113.

Cr-103: portion 8; grid ref. 186.962 (Dubbo); 90 yds. north of portion 9, and 75 yds. west of

road. Black, foetid fossiliferous pellet calcarenite.

Parish of Catombal, Co. Gordon

Ct-6: boundary of portions 75, 124; grid ref. 176.946 (Dubbo); 110 yds. east of junction,

portions 38, 75, and 124. Ill-bedded fossiliferous calcirudite, possibly reef talus.

Ct-11 : portion 40 (north-west); grid ref. 176.944 (Dubbo); in tributary of Back Ck., 210 yds.

along creek east from portion 45 (= Bt/1838, measured section ‘BC, which see). Bio-
hermal limestone.

Ct-13: portion 40 (north); grid ref. 176.944 (Dubbo); in creek, 400 yds. east of portion 45, and

140 yds. south of portion 124. Grey biohermal limestone with large coral colonies.

Ct-14: portion 40 (north); grid ref. circa 176.944 (Dubbo); large area of limestone around head

of creek, about 500 to 600 yds. east of portion 45, and extending 180 yds. south from
portion 124. Grey biohermal limestone.

Ct-15: portion 40 (north-west); grid ref. 176.944 (Dubbo); in gully, 90 to 100 yds. east of por-

tion 45. Biohermal limestone.

Ct-16: portion 40 (north-west); grid ref. 176.944 (Dubbo); in gully, about 15 yds. east of por-

tion 45. Pink biohermal limestone, with numerous Favosites, thamnoporids.

Ct-17: portion 45 (north); grid ref. 176.944 (Dubbo); in gully, 150 yds. west of portion 40.

Bryozoan limestone (biohermal?).

Ct-22: portion 40 (east centre); grid ref. 176.944 (Dubbo); in gully, 160 yds. west of portion 57,

and 520 yds. south of portion 124. Crinoidal biohermal limestone.

Ct-29: portion 40 (south-east corner); grid ref. 176.943 (Dubbo); near junction, portions 40, 41,

and 57, and extending into portion 57 just north of that junction. Biohermal limestone.

Ct-36: portion 125 (south-east); grid ref. 176.941 (Dubbo); along north verge of road, extending

160 yds. north-west from portion 128. Thinly bedded fossiliferous calcarenite and lesser
siltstone.

Ct-44: portion 113 (south-west corner); grid ref. 176.942 (Dubbo); in tributary of Sawpit

Gully, 120 yds. east of portion 125, and 130 yds. north of portion 48. Olive-green
calcareous siltstone.

Ct-45 : portion 113; grid ref. 1 77.942 (Dubbo) ; in gully (as Ct-44), 200 yds. east of portion 1 25,

and 150 yds. north of portion 48. Slightly fossiliferous olive-green calcareous siltstone.

Ct-50: portion 46 (north); grid ref. 176.941 (Dubbo); in gully 90 yds. west of junction of

portions 31, 46, and 48, and 65 yds. east of portion 128. ‘Rubbly limestone.’

Ct-55: portion 58 (north centre); grid ref. 177.939 (Dubbo); south bank of Loombah Ck.,

90 yds. north of road, and 500 yds. west of junction of creek with portions 31, 58, and
77. Coarsely bedded calcarenite and minor calcirudite.

Parish of Cudal, Co. Ashburnam

Cu-2: portion 2; grid ref. circa 172.888 (Bathurst); large area of outcrop, just extending into

portion 61, and between 370 and 700 yds. west of westernmost bend in Boree Ck.
Calcarenite, calcilutite, and biostromal limestone.

594 PALAEONTOLOGY, VOLUME 9

Parish of Eurimbiila, Co. Gordon

E-1 : portion 127 (east); grid ref. 178.934 (Dubbo); in gully just west of road culvert, 500 yds.

south of junction with road of portions 1, 24, and 127. ‘Rubbly limestone.’

E-5: east junction, portions 98 and 160; grid ref. 178.932 (Dubbo); in gully about 170 yds.

due west of road. Biostromal limestone (circa 10 ft.) over ‘rubbly limestone.’

E-18: portion 67 (south-east); grid ref. circa 176.923 (Dubbo); north bank of gully, 100 to 180

yds. west of portion 52, and 200 yds. north of road. ‘Rubbly limestone.’

E-24: portion 43 (east); grid ref. 180.927 (Dubbo); 70 to 150 yds. west of portion 151, and

430 yds. north of portion 63. Massive calcarenite, with thin interbedded fossiliferous
shales.

E-30: portion 80; grid ref. 180.925 (Dubbo); in Blind Gully, 470 yds. east of portion 161, and

420 yds. north of portion 108. Biohermal limestone.

E-31 : portion 80; grid ref. 180.925 (Dubbo); in Blind Gully, 370 yds. east of portion 161, and

400 yds. north of portion 108. Biohermal limestone.

E-37: portion 51 (north); grid ref. 178.923 (Dubbo); 190 yds. east of road, and 35 yds. south of

portion 76. Calcirudite.

Parish of Geiirie, Co. Lincoln

Ge-2: grid ref. 177.990 (Dubbo); north-east bank of railway cutting between portions 197 and

198, about 300 yds. north-west of Geurie town boundary. Thin beds of fossiliferous
calcarenite in unfossiliferous pellet-oolite calcarenite.

Ge-7b: grid ref. 178.991 (Dubbo); south wall of abandoned quarry just north-east of town of

Geurie. Calcarenite.

Ge-12: boundary of portions 57, 59; grid ref. circa 179.988 (Dubbo); outcrop about 200 yds.

across. Calcarenite.

Parish of Gregra, Co. Ashburnain

Gg-2: portion 58 (north-east corner); grid ref. 179.897 (Bathurst); Walker’s Ck., just west of

road bridge. Thickly bedded calcirudite, with interbedded thin favositid biostromes.

Parish of Gundy, Co. Gordon

Gn-7: portion 131 ; grid ref. 183.972 (Dubbo); head of gully, 500 yds. north of portions. 130,

132, and 940 yds. east of portion 55. Thickly bedded calcarenite.

Parish of Mickety Midga, Co. Lincoln

MM-1 : grid ref. circa 181.987 (Dubbo); south side of Mitchell Highway, between portions 60 and

95, about 800 yds. north-west of turn-off to Combo railway siding (at junction of
portions 96 and 208). Pellet calcarenite.

MM-5b: portion 82 (south); grid ref. 182.983 (Dubbo); west of north-easterly road through

portion, and north of road to Ponto Falls. Thick layer of fossiliferous calcarenite within
unfossiliferous silty calcilutites and calcarenites.

MM-7: portion 78 (south); grid ref. 180.982 (Dubbo); about 50 yds. north of road, and 160 to

170 yds. east of portion 54. Calcarenite.

MM-1 1 : portion 206; grid ref. 181.984 (Dubbo); east side of road, 50 to 200 yds. south of junction,

portions 59, 206, and 207; just east of portion 59 (portion 206, near portion 59, of
Hill \9Ald). Calcarenite.

MM-1 2: portion 247 (north-west corner); grid ref. 180.986 (Dubbo); east of portion 39, and

south of road; large outcrop extending about 250 yds. east of portion boundary, and
about 500 yds. south of road. Calcarenite.

Parish of Ponto, Co. Gordon

P-1 : portion 14 (south-west corner); grid ref. 179.977 (Dubbo); hill to west of Mickety Trig.

station, 130 to 550 yds. north of road, and 350 to 600 yds. west of portion 118. Calciru-
dite.

D. L. STRUSZ: SPONGOPH YLLID AE FROM DEVONIAN GARRA FORMATION 595

P-22: portion 142 (east centre); grid ref. 182.976 (Dubbo); 90 to 110 yds. west of eastern por-

tion boundary, and about 750 yds. due north of road; just north of northern boundary,
Water Reserve 33680. Fossiliferous calcarenite and biostromal limestone.

P-23: portion 142 (north-west); grid ref. 182.976 (Dubbo); in gully, 360 yds. east of access

road to ‘Macquarie Park’, and 280 yds. south of portion 104. Calcarenite (6 in. beds).

P-25: portion 142; grid ref. 182.976 (Dubbo); just south of gully, 500 yds. due east of access

road, and about 460 yds. south of portion 104. Thick biostromal complex.

P-38: portion 104; grid ref. 182.977 (Dubbo); 70 yds. south of gully, about 450 yds. west of

access road, 350 yds. south of portion 141, and 420 yds. east of portion 62. Calcarenite.
P-40: portion 103; grid ref. 182.979 (Dubbo); small outcrop 12 feet east of fence (portion 79),

at gate near small wheat silos, ‘Macquarie Park’; about 330 yds. south of Macquarie
River. Partly dolomitized biostromal limestone.

Wellington Caves area: Pli. Wellington, Co. Gordon

WC-4: portion 208 (centre); grid ref. 188.963 (Dubbo); 95 yds. north-east of north-east corner,

portion 201, and 280 yds. east of portion 92. Pellet calcarenite, calcilutite.

WC-9: portion 46; grid ref. 189.965 (Dubbo); outcrop south of road to Wellington Caves, at

turn-off from Mitchell Flighway. Black calcarenite.

Measured section BR^ : portions 9, 10, and 80, ph. Curra; west bank of Bell River, south-west of
Wellington Caves. Section started at junction of three portions, and measured west along gully.

BR^/200: 200 ft. west of starting point. Black foetid pellet calcarenite.

BR®/214: 214 ft. west of starting point. Thin fossiliferous layer in black foetid pellet calcarenite.

Measured section Bt : portions 40 and 45, ph. Catombal; section measured eastward along tributary
of Back Ck., from fence at boundary of portions 38 and 45; grid ref. circa 175.944 (Dubbo).

Bt/430: grey, poorly bedded biohermal limestone with numerous dendroid bryozoa and laminar

Favosites, above pink crinoidal calcarenite; 430 ft. from portion 38.

Bt/936 : grey biohermal limestone with common digitate Favosites, stromatoporoids, Fasciphyllunt

sp. nov. ; 946 ft. from portion 38.

Bt/1838: portion 40; pale grey biohermal limestone with common stromatoporoids, Fasciphyllunt

sp. nov.; 1848 ft. from portion 38; = Ct-11.

Measured section CAT: portion 29, and boundary between portions 41, 43, ph. Catombal; section
measured east from grain silos to portion 41 (fence), then south along fence to corner, then east again.
The section is about 1 mile south of section Bt.

CAT/2193: grey biostromal limestone with Favosites, stromatoporoids, Teratopliyllum sp. nov.;

1 160 ft. along fence from corner. The same horizon as Iocs. Ct-18, -28.

Measured section St: portion 53, ph. Eurimbula; section started at first outcrop in gully near gate,
east of Cumnock-Molong road, and measured eastward; gully is a southern tributary of Spring Ck.

St/181 : grey calcarenite 181 ft. from start (i.e. near top of exposed section); near top of a 500-ft.

layer.

St/1125: black, foetid pellet calcareni4e, with crowded Favosites, stromatoporoids, 1,125 ft.

from start; 110-ft. bed, separated from that of St/181 by a 50-ft. bed of unfossiliferous
calcarenite and calcilutite.

St/1821 : top of 1 50-ft. bed of thinly bedded grey fossiliferous calcarenite; 1,821 ft. from start.

St/2565 : 18 in. bed of grey fossiliferous calcarenite in poorly fossiliferous black foetid calcarenite,

2,565 ft. from start, and near end of section.

Acknowledgements. Much of this work was carried out in the Department of Geology and Geophysics,
University of Sydney, with funds provided under a Sydney University Post-graduate Research
Studentship. I wish to thank Professor C. E. Marshall for allowing the use of the facilities of his depart-
ment. Further work has been done in the Department of Geology, University College of Townsville,
with the aid of research funds from the University of Queensland.

596

PALAEONTOLOGY, VOLUME 9

While responsibility for the contents rests solely with the author, this paper has profited from helpful
discussion with Professor D. Hill, University of Queensland, who critically read the manuscript, and
with Mr. J. S. Jell, also of the University of Queensland. Thanks are also due to Dr. T. B. H. Jenkins
and Mr. A. J. Wright, University of Sydney; Dr. G. M. Philip and Mr. A. E. H. Pedder, University of
New England; Dr. H. O. Fletcher, Australian Museum ; Professor E. C. Stumm, University of Michigan;
Dr. D. J. McLaren, Canadian Geological Survey; and Dr. R. Birenheide, Senckenberg Research
Institute. Mr. D. McNeill helped in the preparation of thin sections, while much of the photography
has been done by Mr. H. L. J. Lamont.

REFERENCES

ALLAN, R. s. 1935. The Fauna of the Reefton Beds (Devonian), New Zealand. Palaeont. Bull. N.Z., 14.

BARRANDE, J. 1902. Svsteme silitrieu du centre de la Bolieme. 8 (2). Anthozoaires et Alcyonaires (Pocta,
P.). Prague.

BIRENHEIDE, R. 1961. Die Acantliophyllum-ArtQn (Rugosa) aus dem Richtschnitt Schonecken-Dingdorf
und aus anderen Vorkommen in der Eifel. Senckenbergiana leth. 42, 77-146, pi. 1-7.

1962. Revision der koloniebildenden Spongophyllidae und Stringophyllidae aus dem Devon.

Ibid. 43, 41-99, pi. 7-13.

1963. Cvatiwphvlliim- und Dohmophylluiu- Arten (Rugosa) aus dem Mitteldevon der Eifel. Ibid.

44, 363-458, pi. 46-62.

BULV ANKER, E. z. 1958. Dcvonskic chetyrekhluchevye korally okrain Kuznetskogo basseina. Tnid.

VSEGEI, Leningrad (in Russian).

CHAPMAN, F. 1925. New or Little-known Fossils in the National Museum. Part 28 — Some Silurian
Rugose Corals. Proc. R. Soc. Viet. 37 (n.s.), 104-18, pi. 12-15.

CHERNYSHEV, B. B. 1941. Siluryiskic i Nizhnedevonskie korally Basseyna Reki Tarei (Yugo-Zapadnyi
Taymyr). Trudy Arkt. Inst. 158, 9-64, pi. 1-14 (in Russian).

CLAUss, K. A. 1956. liber Oberdevon-Korallen von Menorca. Neues Jb. Geol. Paldont., Abb., 103,
5-27, pi. 1-2.

CRICKMAY, c. H. 1960. The Older Devonian Faunas of the North-west Territories. 21 pp., 1 1 pi. Calgary,
Alberta.

DUN, w. s. 1898. Contributions to the Palaeontology of the Upper Silurian Rocks of Victoria, Based
on Specimens in the Collections of Mr. George Sweet, Part I. Proc. R. Soc. Viet. 10 (n.s.), 79-90,
pi. 3.

DYBOWSKi, w. N. 1873. Monographie der Zoantharia Sclerodermata rugosa aus der Silurformation
Estlands, Nord-Livlands und der Insel Gotland. Arch. Naturkde. Livl. Esthl. Kurl. 5 (ser. 3), 257-532,
pi. 1-5 (fide Birenheide 1961).

EDWARDS, H. MILNE, and HAiME, J. 1850. A Monograph of the British Fossil Corals. Part I. Introduc-
tion, etc. Palaeontogr. Soc. [Monogr.], lxxxv+71, pi. 1-11.

1851. Monographie des polypiers fossiles des terrains paleozoiques. Arch. Mus. Hist. nat.

Paris, 5, 1-502, 1-20.

ENGEL, G. and scHOUPPE, A. VON. 1958. Morphogenetisch-taxionomische Studien zu der devonischen
Korallengruppe StringophvUum, Neospongophyllum und Grypophyllum. Paldont. Z. 32, 67-114,
pi. 8-9.

ETHERIDGE, R., JR. 1889. On the Occurrence of a Coral, intermediate in its Structure, between the
Genera Lonsdaleia and Spongopbvllum, in the Upper (?) Palaeozoic Rocks of New South Wales.
Rec. geol. Surv. N.S.W. 1, 22-26,’ pi. 3.

1892. Descriptions of four Madreporaria Rugosa — Species of the Genera Phillipsastraea,

Heliophvllum, and Cvathophyllum — from the Palaeozoic Rocks of N. S. Wales. Ibid. 2, 165-74,
pi. ll-l’2.

1899. On the Corals of the Tamworth District, chiefly from the Moore Creek and Woolomol

Limestones. Ibid. 6, 151-82, pi. 16-38.

1903. An Unusually Large Form of Rhizophyllum, lately discovered in N. S. Wales. Ibid. 7,

232-3, pi. 47.

1911. The Lower Palaeozoic Corals of Chillagoe and Clermont, Part I. Publ. geol. Surv. Qld.

231, 1-8, pi. A-D.

D. L. STRUSZ: SPONGOPHYLLID AE FROM DEVONIAN GARRA FORMATION 597

1918. Two Remarkable Corals from the Devonian of New South Wales {Spongophyllum

halysitoides and Cohwmaria neminghensis). Rec. Aiist. Mus. 12, 49-51, pi. 7-9.

FLOWER, R. H. 1961. Montoya and Related Colonial Corals. Mem. N. Mexico lust. Min. Techuol.
7, 1-97, pi. 1-52.

FOERSTE, A. F. 1888. Notes on Palaeozoic Fossils. Bull. Sci. Lab. Uuiv. Denison, 3, 117-37, pi. 13.
FONTAINE, H. 1961. Les madreporaires paleozoiques du Viet-Nam, du Laos et du Cambodge. Arch,
geol. Viet-Nam, 5, 1-276, pis. 1-35.

HAVLiCEK, V. 1959. Spiriferidae v ceskem siluru a devonu (Brachiopoda). Rozpr. listr. Ust. geol. 25,
1-275, 1-28 (Czech, with English summary, 223-75).

HILL, D. 1939fl. The Devonian Rugose Corals of Lilydale and Loyola, Victoria. Rroc. R. Soc. Viet. 51
(N.S.), 219-56, pi. 13-16.

1939b. The Middle Devonian Rugose Corals of Queensland, I. Douglas Creek and Drummond

Creek, Clermont District. Proc. R. Soc. Qld. 50, 55-65, pi. 4-5.

1940(7. The Silurian Rugosa of the Yass-Bowning District, N.S.W. Proc. Linn. Soc. N.S.W.

65, 388-420, pi. 11-13.

1940^. The Lower Middle Devonian Rugose Corals of the Murrumbidgee and Goodradigbee

Rivers, N.S.W. J. Proc. R. Soc. N.S.W. 74, 247-76, pi. 9-11.

1940c. The Middle Devonian Rugose Corals of Queensland, II. The Silverwood-Lucky Valley

Area. Proc. R. Soc. Qld. 51, 150-68, pi. 2-3.

1942(7. The Middle Devonian Rugose Corals of Queensland, III. Burdekin Downs, Fanning R.,

and Reid Gap, North Queensland. Ibid. 53, 229-68, pi. 5-11.

1942(5. The Lower Devonian Rugose Corals from the Mt. Etna Limestone, Qld. Ibid. 54, 13-22,

pi. 1.

1942c. The Devonian Rugose Corals of the Tamworth District, N.S.W. /. Proc. R. Soc. N.S.W.

76, 142-64, pi. 2-4.

1942(7. Middle Palaeozoic Rugose Corals from the Wellington District, N.S.W. Ibid. 76, 182-9,

pi. 5-6.

1943. A Re-interpretation of the Australian Palaeozoic Record, based on a Study of the Rugose

Corals. Proc. R. Soc. Qld. 54 (1942), 53-66.

1950. Middle Devonian Corals from the Buchan District, Victoria. Proc. R. Soc. Viet. 62 (n.s.),

137-64, 5-9.

and DENMEAD, A. K. (eds.). 1960. The Geology of Queensland. J. geol. Soc. Aust. 7, 1-474, pi. 1-8.

and JONES, o. a. 1940. The Corals of the Garra Beds, Molong District, New South Wales. J. Proc.

R. Soc. N.S.W. 74, 175-208, pis. 2-8.

JONES, o. A. 1929. Qn the Coral Genera Endophyllum Edwards and Haime and Spongophyllum Edwards
and Haime. Geol. Mag. 66, 84-91, pi. 10.

1932. A Revision of the Australian Species of the Coral Genera Spongophyllum E. <& H. and

Endophyllum E. & H., with a Note on Aphrophyllum Smith. Proc. R. Soc. Qld. 44, 50-63, pi. 3-4.
XHALEIN, L. L. 1955. Atlas Rukovodyashchikh Eorm Iskopayemykh Eauny i Elory Zapadnoi Sibiri.
Moscow. (In Russian.)

KRAVTSOV, A. G. 1963. Rannedcvonskiye Chetrekhluchevye Korally s Reki Tarei (Tsentral'nyi
Taimyr). Nauch.-issled. Inst. Geol. Arkt., Gos. Kom. SSSR, lichen. Zap., Paleont. Biostrat. 3, 5-49,
pi. 1-15. (In Russian.)

LANG, w. D., SMITH, s., and THOMAS, H. D. 1940. Index of Palaeozoic Cored Genera. Brit. Mus. (Nat.
Hist.), London; vii + 231 pp.

LENZ, A. c. 1961. Devonian Rugose Corals of the Lower Mackenzie Valley, Northwest Territories.
Geology of the Arctic, 500-14, pi. 1-3. Toronto.

MCLAREN, D. J. and NORRIS, A. w. 1964. Fauna of the Devonian Horn Plateau Formation, District of
Mackenzie. Bull. geol. Surv. Canada, 114, 1-74, pi. 1-17.

and MCGREGOR, D. c. 1963. Illustrations of Canadian Fossils — Devonian of Western Canada.

Pap. geol. Surv. Canada, 62-64 (1962), 1-35, pi. 1-16.

MOORE, R. c. (ed.). 1956. Treatise on Invertebrate Paleontology, Part E — Coelenterata. xx + 498.
Geol. Soc. Amer.

ORLOV, Y. A. (ed.). 1962. Osnovy Paleontologii, 2. Gubki, Arkheotsiaty, Kishechnopolostuye, Chervi.
Akad. Nauk SSSR, Moscow; 1-485, 101 pis. (In Russian.)

598

PALAEONTOLOGY, VOLUME 9

PEDDER, A. E. H. 1964. Correlation of the Canadian Middle Devonian Hume and Nahanni Formations-
by Tetracorals. Palaeontology, 7, 430-51, pi. 62-73.

PHILIP, G. M. 1960. Victorian Siluro-Devonian Faunas and Correlations. Kept. XXI Inter. Geol. Cong..,
Copenhagen I960, 7, 143-57.

1962. The Palaeontology and Stratigraphy of the Siluro-Devonian Sediments of the Tyers Area,

Gippsland, Victoria. Proc. R. Soc. Viet. 75, 123-246, pi. 11-36.

■ and PEDDER, A. E. H. 1964. A Re-assessment of the Age of the Middle Devonian of South-eastern

Australia. Nature, Land. 202 (4939), 1 323-4.

PRANTL, E. 1 952. The Genera Endophylluin Edwards and Haime and Spongophyllinn Edwards and Haime
in the Silurian and Devonian of Bohemia. Sbornik listr. list. Geol., Palaeont. 18, 221-40, pi. 1-2.
SCHLUTER, c. A. J. 1889. Anthozoen des rheinischen Mittel-Devon. Abh. geol. Spezialk. prenfi.
Thiiring. Staat. 8 (4), 259-465, 1-16.

SHERZER, w. H. 1892. A Revision and Monograph of the Genus Chonophvllum. Bull. Geol. Soc. Am. 3,
253-82, 8.

sosHKiNA, E. D. 1937. Korally Verkhnego Silura i Nizhnego Devona Vostochnogo i Zapadnogo
Sklonov Urala. Trudy Paleozool. Inst., Akad. Nauk SSSR 6 (4), 5-155, pi. 1-21. (In Russian;
English summary p. 88.)

1949. Devonskie Korally Rugosa Urala. Ibid. 15, 1-160, pi. 1-68. (In Russian.)

• 1952. Opredelitel devonskikh chetyrekhluchevykh korallov. Trial. Paleont. Inst., Akad. Nauk

SSSR 39, 1-127, pi. 1-49. (In Russian).

SPASSKY, N. Y. 1955. Korally Rugosa i ikh znachenie dlya stratigrafii srednego devona zapadnogo
sklona Urala. Trudy VNIGRI (n.s.), 90 (fide Bulvanker 1958). (In Russian.)

1960. Paleontologicheskoe Obosnovanie Stratigrafii Paleozoya Rudnogo Altaya. Vypusk 3 —

Devonskie Chetyrekhlychevye Korally Rudnogo Altaya. Altaiskiy Gorno-inetall. Naucli.-Issled.
Inst. Akad. Nauk Kazakhskoi SSR 1-43, pi. 1-35; Moscow. (In Russian.)

STRUSZ, D. L. 1963. Studies in the Palaeontology, Petrography and Stratigraphy of the Garra Beds.
Unpublished Ph.D. thesis, Univ. Sydney.

1965fl. A Note on the Stratigraphy of the Devonian Garra Beds of N.S.W. J. Proc. R. Soc.

N.S.W. 98 (1).

1965Zi. Disphyllidae and Phacellophyllidae from the Devonian Garra Formation of New South

Wales. Palaeontology, 8, 518-71.

STUMM, E. c. 1949. Revision of the Families and Genera of the Devonian Tetracorals. Mem. geol. Soc.
Am. 40, i-viii-M-92, pi. 1-25.

TERMiER, G. and TERMiER, H. 1950. Paleontologic Marocaine, II. Invertebres de I’ere primaire.
Fascicule 1 : Foraminiferes, Spongiaires et Coelenteres. Notes Mem. Serv. geol. Maroc, 73, 1-220,
pi. 1-51.

WALTHER, c. 1928. Untersuchuugen fiber die Mitteldevon-Oberdevongrenze. Z. deiitsch. geol. Gesell.
80, 97-152.

WANG, H. c. 1950. A Revision of the Zoantharia Rugosa in the Light of their Minute Skeletal Structures.

Phil. Trans. R. Soc. London, ser. B, 234, 175-246, pi. 1-9.

WEBBY, B. D. 1964. Devonian Corals and Brachiopods from the Brendon Hills, West Somerset.
Palaeontology, 7, 1-22, pi. 1.

WEDEKIND, R. 1922. Zur Kenntnis der Stringophyllen des oberen Mitteldevon. S.B. Ges. Naturw.
Marburg (1921), 1, 1-16.

1923. Die Gliederung des Mitteldevons auf Grund von Korallen. Ibid. (1922), 24-35.

YOH, s. s. 1937. Die Korallenfauna des Mitteldevons aus der Provinz Kwangsi, Sfidchina. Palaeont ogra-
phica, 87, A, 45-76, pi. 4-9.

D. L. STRUSZ

Department of Geology,
University College of Townsville,
Townsville, Queensland

Manuscript received 25 September 1965

THE LATE PRECAMBRIAN FOSSILS FROM
EDIACARA, SOUTH AUSTRALIA

by MARTIN F. GLAESSNER and MARY WADE

Abstract. New genera and species are described from the richly fossiliferous Late Precambrian strata in which
a number of medusoid fossils were discovered by R. C. Sprigg in 1947 and from which various pennatulacean,
annelid, and other fossils have been described since. The new genera and their type species are: Mechisinites
asteroides (Sprigg), Mawsonites spriggi sp. nov., Coiiomedusites lobatus sp. nov., Lorenziniles ranis sp. nov.,
Kimberia qiiadrata sp. nov., Riigoconites emgmaticiis sp. nov., Arborea arborea (Glaessner), Ovatoscutum
coiicentriciim sp. nov., Praecambridiiim sigilliim sp. nov. In addition, the new species Cyclomedusa plana,
Rangea longa, R. grandis, Dickinsonia elongata, D. tenuis, and Spriggina‘1 ovata are described. The synonymies
of several genera and species are revised, and new morphological information on others, particularly
Tribrachidium heraldiciim Glaessner, is presented.

Since the first accounts of the discovery of new Late Precambrian fossils at Ediacara
were published (Glaessner 1958, \959a,b, 1960, 1961, 1962; Glaessner and Daily 1959),
much interest in this fauna has been expressed in the literature, in discussions, in the
form of requests for illustrations for publication in a great variety of books, and in
many other ways. Collecting and indexing of specimens from this locality, now number-
ing over 1,400, was continued and the authors have carried out jointly their study and
interpretation. A comprehensive publication on the accumulated material is in prepara-
tion. It will deal with its geological aspects, fossilization, systematics, biological,
evolutionary, and biostratigraphic interpretations. Nine new genera and thirteen new
species are described, the synonymy of others has been altered, and new material of
some known species has revealed additional important characters. Some of these new
data are now made available to those interested in the early history of metazoan evolution
and relationships, pending completion of more comprehensive investigations.

The description of this unusual collection of remains of soft-bodied animals preserved
in hard sandstone or quartzite requires a careful consideration of the circumstances of
their fossilization so as to discriminate, as far as possible, between results of post-mortem
alterations and original morphological characters. The general sedimentary environment
of the Ediacara fauna can be described as shallow marine and littoral. The position of the
shoreline at the time when deposition of the Pound Quartzite in the Adelaide Geosyncline
and on its western shelf margin was about to end is not known. Occasional emergence of
sandy shoals is proved by the occurrence of networks of drying cracks on some bedding
planes of the fossiliferous strata. The sediment is dominantly arenaceous; cross-
stratification indicates moving sand ripples, and ripple-marked bedding planes are
common. In this environment the fossils would not have been preserved but for one
important local feature. The sandstone, now partly altered to quartzite, is fiaggy, and the
sandy layers are separated by argillaceous laminae and thin lenses. They represent areas
of temporary quiescent conditions between the shifting current tracks, where fine par-
ticles could settle until they were covered again by sand waves. The animal remains
which came to rest on the muddy or silty flats and pool bottoms either made impressions
in the sediment or were retained there bodily until covered with sand by the next shifting

[Palaeontology, Vol. 9, Part 4, 1966, pp. 599-628, pis. 97-103.]

600

PALAEONTOLOGY, VOLUME 9

current. The sand, though partly coarse, must have had some of the qualities of a
foundry sand to be able to produce more or less perfect casts and moulds of the under-
lying sediment surface and of the animal moulds, tracks, and remains on it. The clay
lenses were subsequently highly compacted and altered and are now mostly only thin,
lenticular partings between the quartzite flags. Most of these partings can be opened
only by natural weathering. They reveal fossils mostly on the lower surfaces of the
quartzite flags. The upper portions of the sandy sedimentation units had apparently
only rarely the required composition and consistency to retain them.

This peculiar mode of fossilization, which explains the localization of fossils along a
few miles of outcrop of a stratigraphic horizon which has now been mapped regionally
over hundreds of miles, requires careful consideration of its implications for descriptive
work. Existing terminology must be applied with care to the five observed states of
preservation of these fossils. Most of the medusoid remains are preserved in the form of a
flattened convex cast (1) on the lower surface of a quartzite slab of the original depres-
sion in the underlying sediment. Sand-filled hollows or decayed parts of the organisms
are also preserved as casts but these are rare. The depression in the underlying stratum
which corresponds to the cast is its counterpart mould (2) which reproduces the surface
features of the cast as its mirror image. A superimposition of the features of the lower
surface of the organism (as deposited), moulded on the underlying stratum, and of those
of its upper surface, produces a composite mould (3) (McAlester 1962). Where a convex
feature on the underlying surface, possibly the dead organism itself, is impressed into the
lower surface of the overlying stratum, we speak of an impression or external mould (4).
The corresponding natural cast of this impression which is occasionally seen on a sandy
bed below is the counterpart cast (5). It will be noted in the systematic part that two of
the genera here described as medusoid are known as external moulds, as are all known
representatives of the genera Spriggina, Parvancorina, and Tribrachidium, though in the
last genus one external mould of the lower surface is known, in addition to those of the
upper surface of this and many other specimens. Uncertainties in the distinction between
the two sides of flat bodies which are caused by these modes of preservation add to the
difficulties of their morphological and functional analysis. The preferential preservation
as casts or external moulds in various taxonomic groups can be seen as a function of
their resistance to decay or of the toughness of their tissues. Fortunately, with very few
exceptions, each species is represented by many specimens, so that accidents of fossiliza-
tion can be recognized by comparison.

At the present stage of our inquiries we do not wish to introduce any suprageneric
classifications. The medusoid fossils cannot be generally classified into Scyphozoa and
Hydrozoa, though comparisons suggest that both taxa are represented. The two annelid
genera described clearly belong to two widely different families, but the level of their
separation will not be discussed here.

The geology of the Ediacara area has been the subject of very detailed studies by
geologists of the South Australian Mines Department who have carried out a geo-
chemical investigation of the metalliferous Lower Cambrian carbonate rocks at this
locality. In the course of this work (Nixon 1963, 1964) the fossiliferous strata of the
Pound Quartzite and the overlying Parachilna Formation (Dalgarno 1964) were
mapped in detail.

For reasons which have been stated (Glaessner 1960, 1962, 19636; H. and G. Termier

GLAESSNER AND WADE: LATE PRECAMBRIAN FOSSILS FROM EDIACARA 601

1960) the Ediacara fauna is considered as Late Precambrian, preceding stratigraphically
and chronologically the appearance of the first known Lower Cambrian faunas. South
Australian geologists have accepted this designation in their regional mapping. The
composition of the fossiliferous formation makes it unlikely that it will ever be dated
directly by radioactivity methods.

The fauna as known at present comprises the following 25 species. There are indica-
tions of the presence of additional species and of trace fossils which require further
study.

Phylum Coelenterata

A. Medusoid

Ediacaria flindersi Sprigg
Beltanella gilesi Sprigg
Medusinites osteroides (Sprigg)

Cydomediisa davidi Sprigg
C. radiata Sprigg

C. plana sp. nov.

Mawsonites spriggi gen. et sp. nov.

Conomedusites lobatiis gen. et sp. nov.

Lorenzinites ranis gen. et sp. nov.

Psendorhizostomites howchini Sprigg
Riigocoiiites enigmaticiis gen. et sp. nov.

Kimberia qiiadrata gen. et sp. nov.

Ovatosciitiim concentrician gen. et sp. nov.

B. Pennatulacean

Pangea longa sp. nov.

R. grandis sp. nov.

Pteridiniuin cf. simplex (Giirich)

Arborea arborea (Glaessner)

Phylum Annelida

Dickinsonia costata Sprigg

D. elongata sp. nov.

D. tennis sp. nov.

Spriggina floiindersi Glaessner

S. ? ovata sp. nov.

Phylum uncertain

Praecainbridiinn sigilhnn gen. et sp. nov.

Parvancorina minchami Glaessner
Tribracliidiiim heraldiciim Glaessner

Acknowledgements. Our work on the subject of this contribution was supported in 1961-3 by a generous
grant from the Nuffield Foundation. Current work is supported by an Australian Research Grant.
Material was made available by the South Australian Museum. Mr. R. C. Sprigg, the late Mr. John
Kimber, and Mr. H. W. Ziegler donated most valuable specimens from their collections. Information
on current geological work in the area was supplied by officers of the South Australian Mines Depart-
ment. Dr. B. McGowran and Mr. R. B. Major assisted in the field work. Many colleagues in the
Departments of Geology and of Zoology in the University of Adelaide and in many other University
Departments and research institutions in many countries supplied us with material, literature, and
information and made valuable suggestions and contributions to discussions. The hospitality given
to one of the authors (M. F. G.) in the Department of Geology and Palaeontology in the University of

602

PALAEONTOLOGY, VOLUME 9

Cambridge facilitated the completion of this paper. The text figures were re-drawn by Miss M. Boyce
(Adelaide). To all those who helped in this investigation we express our sincere thanks.

Depositories. Registered numbers prefixed P refer to specimens deposited in the South Australian
Museum. Those prefixed T or F refer to specimens in the Geology Department, University of Adelaide.

SYSTEMATIC DESCRIPTIONS
Phylum Coelenterata
A. Medusoid fossils

Most of the medusoid fossils from Ediacara are preserved as more or less flattened
convex casts on the lower surfaces of quartzite beds. In this preservation, a medusa
could show the shape of the exumbrellar surface and often also remains of any marginal
flange which, if present, could either spread around the entire periphery, or part of it,
or it could be hidden under the body. Such a cast could also show the subumbrellar side
on which one would expect oral or gastric structures, and a velum impression in hydroid
medusae. Actually, no clearly defined structures of this kind have been seen since Sprigg
(1947, 1949) interpreted in this manner some of the fossils described by him. The
distinction between preservation of the exumbrellar or subumbrellar surface is not
necessarily clear-cut, as a number of the medusoid fossils show signs of composite
moulding of parts of both surfaces. It is possible that in some of them remains of radial
grooving which do not reach the centre are to be considered as subumbrellar, while
most of the concentric grooving, particularly in the central part, was exumbrellar.
Gradual decay of the gelatinous bell during fossilization introduces further com-
plications. It is more important to recognize the complexities of structural interpretation
of ancient medusoid fossils than to attempt prematurely their placing in the existing
system. Further comparative studies are required but some distinctive genera and species
are here recorded.

Genus ediacaria Sprigg 1947

Type species. Ediacaria fUndersi Sprigg 1947.

Characters as for type species.

Ediacaria flindersi Sprigg 1947

Plate 99, fig. 6

1947 Ediacaria flindersi Sprigg, p. 215, pi. 5, figs. 1, ?2, text-fig. 3.

1949 Ediacaria flindersi Sprigg, p. 83, pi. 10, fig. 2, text-fig. 5.

1949 Madigania annulata Sprigg (parr/???), p. 93, pi. 17, figs. 1, 2 {non pi. 16, figs. 1, 2).

1949 Protodipleurosonia wardi Sprigg, p. 79, pi. 9, fig. 2, text-fig. 3e.

1956 Ediacaria flindersi Sprigg (partim), Protodiplenrosoma wardi Sprigg; Harrington and
Moore, in Moore, p. F74, fig. 60 (1); p. F79, fig. 64.

1959 Ediacaria flindersi Sprigg; Glaessner, in Glaessner and Daily, p. 378.

1962 Ediacaria Sprigg; Glaessner, p. 483.

Material and preservation. Over 40 specimens are definitely assigned to this species, while 30 small
specimens, 3 relatively uncompressed specimens, 5 composite moulds with rather strong radial furrows,
and a number of fragments may belong here.

Most specimens are rather flat composite moulds dominated by exumbrellar structures but some are
gently domed. Strong radial furrows on some specimens may indicate subumbrellar structures.

GLAESSNER AND WADE; LATE PRECAMBRIAN FOSSILS FROM EDIACARA 603

The species was gregarious; one slab measuring about 45x30 cm. bears parts of 9 specimens with
radii from 30-120 mm.

Holotype. Tl-2058.

Dimensions. This is the largest medusoid known from Ediacara. The largest complete specimen has a
mean radius of 120 mm. but a fragment has a radius of 210 mm. and another fragment indicates a
much larger specimen. The holotype has a mean radius of 50 mm.

Diagnosis. The surface (which is considered as exumbrellar) shows a central disc and an
outer ring, without any marginal flange. The disc is often slightly elevated above the
outer ring. The surface of the disc may be marked by a sharp annular furrow which
does not necessarily run parallel with its outer boundary. It could indicate the edge of
the gastric cavity. Some specimens show additional concentric grooves on the disc.
Several of them may be grouped close to the centre. Arcuate folds marked by grooves
parallel to the periphery may also occur on the outer ring. Radial furrows are mostly
confined to it but they are not present in all specimens. Crenulations of the periphery
are probably accidental. The radius of the central disc is from one-quarter to three-
quarters of the total radius.

Remarks. One-quarter of the holotype is poorly preserved and superimposed on a small
Cyclomedusa near the centre. This Cyclomedusa has caused distortion, straightening the
annular furrows near the centre of the Ediacaria. It may be the sole cause of some
of the straight furrows delimiting one side of the subtriangular structure which Sprigg
{1949, p. 85) interpreted as a manubrium. The other furrows delimiting this structure
are a radial and a concentric furrow, each in line with the normal furrows of the disc
but exceptionally deep.

Sprigg (1947, 1949) described ‘three pendant pouches extending radially from the
base of the manubrium’ in the position in which Harrington and Moore {in Moore
1956, p. F75) noted four ‘circular knobs’. One or two of these irregularities appear to
be due to distortion by the Cyclomedusa and the specimen does not support clearly
either of the quoted descriptions. The outer ring shows a number of partial concentric
folds. The general morphology of a more corrugated outer ring and a smoother, more
convex disc, both with a surface texture of fine, radial striae, indicates that this specimen
is an exumbrellar surface. None of the radial or concentric markings on the central disc
can be attributed with any probability to internal or subumbrellar structures. Many
radial furrows occur near the margin of the outer ring but less than half of them cross it.
Most of those that do, die out at the edge of the central disc but a few reach to, or just
across, a partly doubled annular depression on the disc which Sprigg (1947) regarded
as the outer edge of the ‘gastrovascular cavity’ or ‘stomach’. The radial depressions on
the outer ring of the holotype are more numerous than those in any other specimen
interpreted as an exumbrellar surface. They are deeper and less numerous than those on
the subumbrellar surfaces tentatively assigned to Ediacaria (Sprigg 1947, pi. 5, fig. 2,
and other specimens). There is very little branching or overlapping of the radial furrows
despite the striking arrangement of canals drawn in restorations (Sprigg 1947, text-
fig. 3a, b; 1949, text-fig. 5a-c).

The holotype of Protodipleurosoma wardi Sprigg (1949, pi. 9, fig. 2) is a composite
mould. All its exumbrellar characteristics are those of Ediacaria jiindersi. It difiers in
the relatively small size and elongate-oblate shape of the sharp furrow that may indicate

s s

C 4397

604

PALAEONTOLOGY, VOLUME 9

its gastric cavity (Sprigg 1949, p. 79). The enclosed portion is slightly elevated as if the
area had been filled with sand before compression. Groups of furrows radiate from
opposite points. They are deepest adjacent to the sharp furrow, and shallow away from
it. There is also a fine wrinkling across the surface with which the constriction across the
‘stomach’ shown by Sprigg (1949, text-fig. 3e) is associated. These furrows and wrinkling
are probably both compressional features due to the flattened infilling of sand; there are
no structures equivalent to them among the new specimens of medusoid fossils. Thus
the ‘radial canals’ (Sprigg) are not a convincing taxonomic character. The study of
' Prolodiplewosoma wardV shows that several partial concentric folds have been aligned
as a ring canal in Sprigg ’s text-fig. 3e, while the new material of Ediacaria shows that the
development of concentric furrows on the outer ring, varying in completeness and num-
ber, is common. The centre of the disc, where the area within the oblate furrow is least
raised, shows faintly a small, circular structure such as is seen on many specimens of
Ediacaria. The size of the infilled area is the only remaining distinction between Proto-
dipleurosoma and Ediacaria. This character cannot be considered as taxonomically
distinctive. Protodipleurosoma wardi Sprigg 1949 appears to be a composite mould of
Ediacaria Jiindersi Sprigg 1947, and the name is suppressed as a junior synonym. ,

Genus beltanella Sprigg 1947
Type species. Beltanella gilesi Sprigg 1947.

BeltaueUa gilesi Sprigg

1947 Beltanella gilesi Sprigg, p. 218, pi. 6, fig. 1, text-fig. 4a-c. j

1949 Beltanella gilesi Sprigg; Sprigg, p. 81, pi. 10, fig. 1, text-fig. 4d-f.

1956 Beltanella gilesi Sprigg; Harrington and Moore, in Moore, p. F70, fig. 56. ;

1959 Beltanella gilesi Sprigg; Glaessner, in Glaessner and Daily, p. 378.

1962 Beltanella Sprigg; Glaessner, p. 483, pi. 1, fig. 3.

Holotype. T3-2056.

Description. The holotype is a convex cast, highest at one side of the central disc. The type i
figure (Sprigg 1947 ; reproduced by Sprigg 1949, Harrington and Moore in Moore 1956) j
shows three bosses regularly spaced near the margins of the central disc; these are the |

clearest, but alternative lighting (Glaessner 1962) shows a fourth boss between and to the |

centre from two others. This distorts the concentric furrow which occurs at mid-radius ■'
of the central disc. In the possession of these bosses Beltanella gilesi remains unique
but its other characteristics are shared by several specimens which cannot be separated
from Ediacaria Jiindersi. I

Its outer flange appears narrower than the outer ring of Ediacaria', it is only 5-7 mm. f

wide over one-half of the circumference of Beltanella, but it widens out to a width of
13 mm. where the total radius is 56 mm. Even here, concentric folds show that it was not
stretched. The mean diameter of the central disc is just over 80 mm., so that the
proportions of this apparent ‘outer flange’ fall within those of the ‘outer ring’ of
Ediacaria. Similarly, the central disc is fairly smooth, bearing a concentric furrow (here
partly double) at mid-radius. The central structure is delimited by four annular furrows, *
the outer 12 mm. and the inner 4 mm. in mean diameter. It consists of two flat rings,
each 1 mm. across, and a narrower inner ring; probably the innermost portion was a

GLAESSNER AND WADE: LATE PRECAMBRIAN FOSSILS FROM EDIACARA 605

boss but it is now broken and a sharp-edged depression occupies the centre. The natural
features of this central structure can be closely matched among the available material of
Ediacaria. A structure that is not matched among the Ediacaria specimens is the double
radial furrow delimiting a low ridge, which crosses from the outer edge of the central
disc to the concentric furrow at mid-radius of the disc, and which is opposed by a low
ridge and indications of a double furrow on the opposite side of the central disc.
Sprigg (1947, 1949) reconstructed this as two of four paired radial canals at right angles,
but evidence of the other two is lacking and the significance of this character is uncertain.
A feature which has not been described is the faint radial furrowing extending towards
the centre from the concentric furrow at mid-radius {vide Glaessner 1962, pi. 1, fig. 3).
This is not present around the full circumference.

As the size, proportions, and main features of BeJtanella are within the range of similar
structures in Ediacaria, their distinctiveness hinges mainly on the presence in Beltanella
of the ‘ring’ of bosses which have been considered as possible gonads. The slab on which
the Beltanella occurs is exceptionally rich in medusoids; eight small specimens of Cyclo-
medusa (including two clumps of three each) and two other medusoids occur, besides
many vague markings. There are very few places on it where so large an individual as
the holotype of Beltanella would not over- or underlie some other specimen. The
bosses, as Sprigg noted from the first, have concentric furrows on them. A small group
of Cyclomediisa specimens superimposed on the holotype would satisfactorily explain
both this structure and the distribution of bosses (Glaessner 1962, pi. 1, fig. 3), which does
not fit the interpretation of eight regularly distributed gonads. On present evidence it is
not possible to prove conclusively whether these bosses are Cyclomedusa specimens or
parts of Beltanella. The discovery of one or more specimens showing such bosses on
slabs not rich in Cyclomedusa would be a strong indication of the distinctiveness of
Beltanella but in their continued absence the taxon must be regarded as doubtful and as
a possible synonym of Ediacaria.

Genus medusinites gen. nov.

Type species. Mediisina asleroides Sprigg 1949.

Characters of type species. This genus resembles Protolyella only in its smooth central
area, but the surrounding outer ring lacks the numerous, close-set, irregular, radial
grooves and ridges. Medusina' filamentis Sprigg does not belong here but is connected
with Pseudorhizostomites howchini Sprigg by transitional forms. Medusina Walcott
1898 is an objective synonym of Spatangopsis Torell 1870, a form unrelated to
Medusinites.

Medusinites asteroides (Sprigg) 1949

Plate 97, figs. 1-5

1949 Medusina asteroides Sprigg, p. 90, pi. 3, fig. 3, text-fig. 7c.

1949 1 Medusina mawsoni Sprigg, p. 89, pi. 13, fig. 4, text-fig. 7b.

1956 Protolvella asteroides (Sprigg), IP. mawsoni (Sprigg); Harrington and Moore, in Moore,
p. F155, fig. 127 (1), ?(2).

1959 Protolyella asteroides (Sprigg), ?F. mawsoni (Sprigg); Glaessner, in Glaessner and
Daily, p. 381.

606 PALAEONTOLOGY, VOLUME 9

Material and preservation. About 20 casts on the lower surfaces of quartzite slabs, many with counter-
part moulds.

Dimensions. Specimens with diameters from about 1 to 5 cm. can be identified. Smaller and slightly
larger specimens lacking distinctive characters may also belong here.

Diagnosis. Small circular bodies with a very narrow, generally distorted, marginal
flange surrounding a large outer ring separated from a central disc by a deep annular
furrow. The radius of the central disc is smaller than the width of the outer ring. Radial
furrows may be preserved on it. Faint concentric markings are more common on the
central disc than elsewhere.

Remarks. In closely similar medusae, even on the same rock-surface, the marginal
flange may be fully displayed, partly folded below the specimen (as in the holotype of
M. asteroides) or absent. Specimens without a flange and with a large central disc are
not clearly distinct from Medusina mawsoni Sprigg and it therefore seems on present
evidence that this form should be merged with M. asteroides. It is preferable to retain
the name of the form which shows most distinguishing characters, and suppress mawsoni
as a synonym.

Genus cyclomedusa Sprigg 1947

Type species. Cyclomedusa davidi Sprigg 1947.

1947 Cyclomedusa Sprigg, p. 220.

1949 Cyclomedusa Sprigg, p. 91.

1949 Madigania Sprigg, p. 93.

1949 Tateana Sprigg, p. 86.

1956 Cyclomedusa Sprigg; Harrington and Moore, in Moore, p. FI 53.

1956 Ediacaria Sprigg (partim)\ Harrington and Moore, ibid., p. F74, fig. 60 (4-5).

1956 Madigania Sprigg; Harrington and Moore, ibid., p. F154.

1958 Spriggia Southcott [= Madigania Sprigg 1949, non Whitley 1945], p. 59.

1959 Cyclomedusa Sprigg; Glaessner, in Glaessner and Daily, p. 378.

1959 Spriggia Southcott, Tateana Sprigg; Glaessner, ibid., p. 388.

Remarks. Cyclomedusa radiata Sprigg dilfers from the type species in its central boss-
and-ring structure and more widely spaced deep radial furrows, with fewer concentric
grooves. In several specimens of Cyclomedusa, however, the radial structures agree with
typical forms of C. radiata but extend further toward the centre, which is developed as
in C. davidi. C. gigantea Sprigg is represented only by its holotype, which is badly

EXPLANATION OF PLATE 97

All figures x 1 .

Figs. 1-5. Medusinites asteroides (Sprigg). 1-2, hypotypes, PI 3785-6, showing central disc, main
annular furrow, outer ring, and marginal flange. 3, holotype, T40-2021, showing also few radial
furrows. 4, holotype, T39, of ^Medusina mawsoni' Sprigg. 5, hypotype, P13783, with numerous
radial furrows on outer ring.

Figs. 6-7. Kimberia quadrata gen. et sp. nov. 6, holotype, PI 2734. 7, paratype, PI 2739. The deep
groove x-y is probably accidental rather than structural. Outlines marked where only faintly
recognizable on the specimens.

Fig. 8. Ovatoscutum concentricum gen. et sp. nov. Holotype, PI 3770. Arrow points to notch, centre,
and radial suture. It ends where a faint discontinuity in the rock surface sculpture may indicate the
edge of a smooth portion of the disc across the notch.

Fig. 9. Tribrachidium heraldician Glaessner. Hypotype, FI 7021. A distorted natural mould showing
impressions of bristles.

Palaeontology, Vol. 9

PLATE 97

GLAESSNER and WADE, Late Precambrian fossils from Ediacara, South Australia

GLAESSNER AND WADE: LATE PRECAMBRIAN FOSSILS FROM EDIACARA 607

weathered in the centre. Its specific distinctness from the type species is therefore
questionable. The presence of areas with radial furrows on several specimens formerly
placed in Spriggia annulata (Sprigg), including the holotype, makes its separation from
C. davidi impracticable because of intergrading.

Cyclomedusa plana sp. nov.

Plate 98, figs. 1-3

1959 ‘Unidentified medusoid fossils’, Glaessner, in Glaessner and Daily, pi. 45, fig. 2 (partiin).
Material and preservation. 8 specimens, casts on the lower surfaces of quartzite slabs.

Holotype. PI 3778. Paratypes. P13779, PI 3780.

Diagnosis. Adult distinguished by a broad, flat disc (maximum known radius 12 cm.)
surrounding a small, central, concentrically rugose cone rarely over 1 cm. in radius;
radial furrows very fine when present. The central cone is sometimes duplicated by
twinning.

Remarks. C. plana is distinguished from C. davidi in adult specimens by its smaller central
cone and large disc with few or without concentric rugosities, and by its finer radial
striations. The fineness of these, and the central cone, differentiate it from C. radiata.

Genus mawsonites gen. nov.

Type species. Mawsonites spriggi sp. nov.

Characters of type species. This genus is named after the late Sir Douglas Mawson,
geologist and explorer, who inspired and led his students and colleagues in the study of
the Flinders Ranges and the Ediacara area.

Mawsonites spriggi sp. nov.

Plate 99, figs. 1, 2

1958 ‘Unnamed jellyfish’, Sprigg, in Best, pi. 6, fig. b (figured only).

Material. Four casts on the lower surfaces of quartzite slabs. The holotype retains an elevation of
5 mm. from the rock surface in places. The paratype and other specimens are flattened.

Holotype. F17009. Paratypes. F17016, F17019.

Dimensions. Diameter of holotype 110-125 mm., of paratype F17016 89 mm. (maximum), of paratype
F17019 64-72 mm.

Diagnosis. Large, compressed, but in life presumably dome-shaped, becoming steeply
conical near centre. Central conical part smooth-walled, truncated, and compressed in
curved, overlapping folds in the holotype. The greater part of the surfaee is strongly
sculptured with arcs of prominent, large, irregular, bosses which increase in size out-
wards, and merge into the peripheral zone whieh is dominated by large, irregular,
radially elongate lobes separated by deep elefts. They form radial furrows on the outer
half of the disc. Periphery lobate. A circular area bearing the conical centre and the
enelosing one or two arcs of bosses was more eompressible than the remainder of the
bell and was enclosed by a shallow annular groove on the surface of the dome.

608

PALAEONTOLOGY, VOLUME 9

Remarks. The peculiar distribution of large bosses on the main body of the bell, and the
lobate margin, set this form apart from other fossil and living medusae. The bosses
must have been fairly solid to make such well-defined impressions in the underlying
sediment. In its possession of a truncate, easily compressed, possibly hollow central
cone with arcuate folds in its walls Mawsonites approaches Cyclomedusa davidi and
C. plana in the Ediacara fauna, and the Stauromedusae among modern forms.

Genus conomedusites gen. nov.

Type species. Conomedusites lobatus sp. nov.

Characters of type species.

Conomedusites lobatus sp. nov.

Plate 99, figs. 3, 4

Material and preservation. Three specimens, casts on the lower surfaces of quartzite slabs.

Holotype. PI 3789. Paratypes. PI 3788, PI 3790.

Dimensions. Diameter of holotype 21-22 mm., height 1-5 mm. Diameter of paratypes P13788 and
P13790 15-18 and 19 mm. respectively, height 3 and 1 mm. respectively. It should be noted that the
marginal flange is obscured in PI 3788.

Diagnosis. Small, convex body, composed of four almost equal lobes, delimited by sharply
defined grooves that deepen outwards, separating adjacent lobes at the periphery, or
above a broad, distorted, marginal flange with an entire margin. Further lobes may be
intercalated peripherally by bifurcation of the four apical grooves.

Remarks. Only the aboral side of C. lobatus is known. It is distinguished by its conical
shape from the exumbrellar view of the simpler specimens of Brooksella alternata
Walcott (1898, pi. 1, fig. 4; re-figured by Harrington and Moore, in Moore 1956,
fig. 11, 2a-c). There is no reason to assume that it was a related form as there is no trace
of subumbrellar lobes. A resemblance to the Ordovician Conchopeltis alternata Walcott,
a large, low, conical, tetramerous, presumably chitinous shell for which Moore and
Harrington (in Moore 1956, pp. F32, F57) erected a new Family and Suborder in the
Order Conulariida, could be fortuitous, for the marginal flanges preserved in two of the
three specimens are apparently entire. Conomedusites is the only medusoid fossil from
Ediacara with a clearly displayed tetrameral symmetry.

Genus lorenzinites gen. nov.

Type species. Lorenzinites rams sp. nov.

Characters of type species.

EXPLANATION OF PLATE 98

Figs. 1-3. Cyclomedusa plana sp. nov. l,paratype,P13779, xl-2. 2, holotype, PI 3778, Xl. 3, paratype,
P13780, part of specimen with twinned centre, xO-75. All specimens show the concentrically rugose
small central area. Strong radial grooves in 1 and 3 are probably accidental.

Fig. 4. Sprigginal ovata sp. nov. Holotype, P13754, x4.

Palaeontology, Vol. 9

PLATE 98

I

3

GLAESSNER and WADE, Late Precambrian fossils from Ediacara, South Australia

GLAESSNER AND WADE: LATE PRECAMBRIAN FOSSILS FROM EDIACARA 609

Lorenzinites rarus sp. nov.

Plate 100, fig. 1

Material and preservation. A single specimen, an external mould on the base of a quartzite bed.
Holotype. PI 3784.

Dimensions. A disc with a diameter of 5-6 mm., surrounded by eleven radiating lobes 2-3 mm. long.

Diagnosis. A small central disc from which lobes (eleven in the only known specimen)
radiate, broadened and flattened at the tips. The tips may bifurcate. The edge of the disc
describes a series of arcs from the base of each lobe to the base of its neighbouring lobes.

Remarks. At first glance this structure resembles the problematic medusoid genus
Lorenzinia but its lobes expand at the periphery of the fossil instead of being rounded
off, and may bifurcate. Only eleven lobes are known. In this it resembles Palaeose-
maeostoma Riiger, from the German Middle Jurassic, a genus classified as ?Trachy-
linida incertae sedis by Harrington and Moore {in Moore 1956, p. F76), but it lacks the
raised narrow collar around the centre of the disc.

Like Rugoconites enigmaticus sp. nov., but unlike the other medusoid fossils which are
casts, L. rants is preserved as an external mould. This specimen resembles the centre of
some specimens of R. enigmaticus but is much more regular. While its central disc is the
size of the central ring-ridge of some specimens of R. enigmaticus, its lobes are smaller
and more numerous than the innermost whorl of dichotomous radial ridges of this
species. It is not likely to be a juvenile of R. enigmaticus because of their differing
measurements and proportions, but it could be a related form.

Genus pseudorhizostomites Sprigg 1949

Type species. Pseudorhizostomites howchini Sprigg 1949. Plate 103, figs. 2-4.

1949 Pseudorhizostomites Sprigg, p. 87, pi. 12, figs. 1,3; text-fig. 6f-h.

1949 Pseudorhopilema Sprigg, p. 88, pi. 12, fig. 2; text-fig. 6e.

1949 Medusina Walcott (partim); Sprigg, p. 90, pi. 13, fig. 1; text-fig. 7d.

1956 Pseudorhopilema Sprigg, Pseudorhizostomites Sprigg; Harrington and Moore, in Moore,
pp. F51, 52, fig. 41 (1-3).

1956 Protolyella Torell (partim)', Harrington and Moore, ibid., p. F155, fig. 127 (7).

1959 Pseudorhizostomites Sprigg, Pseudorhopilema Sprigg; Glaessner, in Glaessner and Daily,
pp. 381-2.

Material. Over 80 specimens, most of which are impressions on the bases of quartzite slabs; and
occasional counterpart casts found on upper surfaces. The following descriptions are based on the
impressions.

Description. The specimens are extremely variable in appearance. Although the small
number available to Sprigg (1949) could be grouped in three apparent genera, such
grouping is no longer practicable for the many intermediate and new forms which are
now known.

The specimens, mostly under 7 cm. in diameter, consist of furrows which radiate
outwards from a centre of variable shape. As they radiate, they split and become shallow.

610

PALAEONTOLOGY, VOLUME 9

forming a zone of numerous very small furrows at the outer limits of many specimens.
Only one specimen possesses a definite outer edge, a sharp groove preserved around half
its circumference. Radial splitting and dendritic branching of the furrows are extremes
of variation (Sprigg 1949, pi. 12, figs. 1-3; pi. 13, fig. 1).

Some rock specimens have been split vertically across these fossils accidentally, and
others intentionally. In every example in which the lamina of quartzite immediately
above the specimen had appreciable thickness (9 specimens) there were indications of
post-depositional disturbance of sand grains above the centre, or above the deepest
furrows. Plate 103, figs. 2-4 shows a specimen which has been split along a linear central
furrow towards which all its other furrows converge. The main converging furrows can
be traced into vertical flutes on the sides of a smooth-walled passage through the over-
lying lamina of quartzite. This passage ends on the next bedding plane. Although com-
paction during diagenesis has forced its sides into contiguity, the individual sand grains
do not interlock. Another specimen with a linear central furrow shows a similar passage,
while three with point-centres show very narrow, funnel-shaped lines of disturbance
vertically above their centres, and more obscure disturbances are found in specimens
with ill-defined centres or circular central furrows. All seem to have resulted from the
escape of material from a decaying organism. The circular central furrows delimiting
one to several rounded bosses of quartzite (e.g. ' Medusina’’ filamentis Sprigg 1949,
pi. 13, fig. 1) seem to result from the subsidence of the overlying sand-lamina during the
escape of the organic matter, rather like the subsidence of the core of a ring-dyke.
Forms such as that figured here and by Sprigg (1949, pi. 12, fig. 1, the holotype of
Pseudorhizostomites howchini, and fig. 3b, Pseudorhizostomites sp.), which are more or
less radial structures, seem to have been shaped largely by viscous flow towards the
centre. These are the majority (over 40 specimens). The rather regular dichotomously
branched form Pseudorhopdema chapmani Sprigg (Sprigg 1949, pi. 12, fig. 2) occurs in
such numbers (about 20) that there is a possibility that some surface structures of the
organism, in addition to post-mortem processes, influenced their formation.

While Pseudorhizostomites howchini is thus primarily based on a peculiar state of
preservation of a decaying organism, it is unlikely to have been formed by the decay
of any of the other medusoid organisms here described, with the exception, perhaps, of
Rugocouites. It is likely to represent a morphologically distinctive though imperfectly
known taxon. Reliable specific distinctions cannot be made within it at this time.

Genus rugoconites gen. nov.

Type species. Rugoconites enigmaticiis sp. nov.

Characters of type species.

EXPLANATION OF PLATE 99

Figs. 1, 2. Mawsonites spriggi gen. et sp. nov. 1, holotype, FI 7009. 2, paratype, FI 701 6, X 1. The smaller
specimen is less strongly sculptured but the basic pattern is the same.

Figs. 3, 4. Conomedusites lobatus gen. et sp. nov. 3, holotype, PI 3789. 4, paratype, PI 3788, X 1.

Fig. 5. Tribrachidiiim beraldiciim Glaessner. Latex cast of holotype, PI 2898, X 1.

Fig. 6. Ediacaria flindersi Sprigg. Hypotype, FI 7039., showing smooth outer ring and flattened central
disc, marked only by an elliptical annular furrow, xO-5.

Palaeontology, Vol. 9

PLATE 99

GLAESSNER and WADE, Late Precambrian fossils from Ediacara, South Australia

GLAESSNER AND WADE; LATE PRECAMBRIAN FOSSILS FROM EDIACARA 611

Rugoconites enigmaticus sp. nov.

Plate 100, figs. 2, 3

Material and preservation. 16 impressions in the bases of quartzite flags. Large impressions are
relatively (and often actually) flatter than small specimens and their inner parts are obscured by
flattening. The following descriptions are based on latex casts of the impressions.

Holotype. PI 3781. Paratype. PI 3782.

Dimensions. The smallest specimen is 16 mm. in diameter; the largest, slightly oval, has an average
diameter of 64 mm.

Diagnosis. Periphery circular to oval. A few radial ridges diverge from a small central
elevation or polygonal ridge, and branch dichotomously one to three times. The last
dichotomy is, in each individual, a constant distance inside the periphery, and the
resultant ridges curve slightly to meet the periphery at a right angle.

Remarks. In its general form of dichotomous grooves in the rock surface R. enigmaticus
r QstmhlQS Pseudorhopilema chapmani Sprigg (1949). It differs from Sprigg’s specimen and
others since collected in usually having a definite periphery and in being much more
regular in its branching. The branches (‘ridges’, as here described, corresponding to the
‘grooves’ of P. chapmani Sprigg 1949) are much thicker in the new form, and do not
taper at their outer ends but are cut off sharply at, or just within, the periphery.

Genus kimberia gen. nov.

Type species. Kimberia quadrata sp. nov.

Characters of type species. The genus is named after Mr. John Kimber, student,,
teacher, and collector who lost his life during an expedition to Central Australia in 1964.

Kimberia quadrata sp. nov.

Plate 97, figs. 6, 7

1959 ‘Problematic fossil, possibly belonging to the Siphonophora’, Glaessner, in Glaessner
and Daily, p. 391, pi. 47, fig. 9.

Material. Four nearly complete and two fragmentary specimens are known. All are very flat casts on
the bases of quartzite slabs. Some portion of each specimen is broken off at the edge of the rock.

Holotype. PI 2734. Paratype. PI 2739.

Dimensions. Length of holotype 65 mm., width 36 mm.

Diagnosis. Ovate bodies, rounded at one end and with a smooth contour; internal
structures represented by several longitudinal, distinct zones of two kinds, coarsely
segmented or with fine, transverse, frill-like grooves bordering a smooth area.

Remarks. The bodies appear to have had the shape of a narrow bell which tapered a
little more towards one truncate and presumably open end than at the opposite, rounded
end. All are flattened and distorted in various ways, the truncate end of the bell being
noticeably less resistant to compression than the rounded end. All known specimens
have a frilled zone displayed centrally with segmented zones on either side. This sug-
gests that the bells were either flattened at the frilled zones (squarish in life) or that the
segmented zones supplied resilience so that the bodies preferentially lay on the area

612

PALAEONTOLOGY, VOLUME 9

between two segmented zones. In this position the segmented zones would overlie one
another and no specimen unequivocally shows how many there are, but there were more
than two and no more than four. This seems the most probable number, as three seg-
mented zones would not result in the frilled zone being displayed centrally. Even the two
fragmental specimens seem to have been preserved in this position.

Various interpretations of the shape of this peculiar organism in life are possible.
One line of argument leading to a tentative structural assessment is presented here.
There is an obvious resemblance between the segmented zones of this species and the
long gonads on four radial canals in some present-day medusae. These are principally
found among the Carybdeida, Trachymedusina, and Leptomedusae. The gonads of the
Carybdeida are paired, more lamellar, and less like tubes segmented by intermittent
transverse creases than the segmented zones of K. quadrata; they are positioned in the
corners of a more or less quadrate bell. The general shape of K. quadrata could have
approximated that of Carybdea marsupialis Peron and Lesueur (Mayer 1910), but the
segmented zones more closely resemble the gonads of, for example, adult Laodiceidae
(Leptomedusae), which run from the stomach almost to the margin of the bell. It is
difficult to relate the appearance of the frilled zone to these resemblances to structures
of present-day medusae. Too little is known to justify the allocation of K. quadrata to
one or other of the Leptomedusae, Trachymedusina, or Carybdeida, though the existence
of most of its characters in one or more of these groups indicates its medusoid affinities.

Genus ovatoscutum gen. nov.

Type species. Ovatoscutum concentricum sp. nov.

Characters of type species.

Ovatoscutum concentricum sp. nov.

Plate 97, figs. 8

Material. One external mould, flattened and showing some wrinkling but no fracturing. It is described
from a latex cast.

Holotype. PI 3770.

Dimensions. Length 6-2 cm., width 6T cm. 35-40 concentric ribs.

Diagnosis. A large, rounded shield sculptured with strong concentric corrugation, which
weakens adjacent to a triangular notch, where the outline of the shield is indistinct in the
only known specimen. The centre of the sculpture, a minute, smooth, oval, convex area,
is situated at a distance of about one-third of the length of the shield measured from a
line across the extreme ends of the notch. The contour of the ribbing changes from

EXPLANATION OF PLATE 100
Fig. 1. Lorenzinites rams gen. et sp. nov. Holotype, PI 3784, X L5.

Fig. 2, 3. Rugoconites enigniaticus gen. et sp. nov. 2, holotype, PI 3781. 3, paratype, PI 3782, one half
broken off. Broken line marks faint outline of disc. X 1.

Fig. 4. Rangea tonga sp. nov. Holotype, PI 3777, X 1.

Fig. 5. Rangea grandis sp. nov. Holotype, P12897, X 1.

Palaeontology, Vol. 9

PLATE 100

GLAESSNER and WADE, Late Precambrian fossils from Ediacara, South Australia

iFT-yx

GLAESSNER AND WADE: LATE PRECAMBRIAN FOSSILS FROM EDIACARA 613

distinctly elongate at a length of 2-2-5 cm. to almost equal in length and width, but it
becomes sinuous rather than circular. A suture-like zone about 1 mm. wide extends
through the centre from the tip of the notch towards the opposite margin.

Remarks. The shape, sculpture, and outline of the disc are somewhat distorted in the
anterior portion, apparently through draping over an irregularity of the underlying
surface. The ribs become wider and possibly more widely spaced towards the periphery.
In some places they appear to be asymmetrical in profile, sloping more steeply towards
the centre. The only radial lines are the suture-like zone where the profile of the ribs
appears to be flatter and more evenly convex, and the edges of the notch, where the
sculpture disappears along irregular and frayed-looking boundaries which are convex
towards the median line. The notch does not extend to the centre.

The fossil is strikingly similar in outline and sculpture to Plectodiscus corllandensis
Caster, from the Upper Devonian of New York State (Caster 1942). It is only half as
large as the holotype of this species and lacks the central cone described in much detail
by Caster. He does not mention any notch but the sculpture appears to be absent in a
triangular area between f and i (Caster 1942, pi. 1, fig. 1 ), corresponding in position to
the notch in Ovatoscutiim. The original substance of P. cortkmdensis was described as
Hough and flexible’. This would also fit O. concentricum. Caster described the grooves
between the ribs as ‘asymmetrical, the steeper slope being consistently peripherad’.

P. cortlandeusis was interpreted by Caster, as a result of detailed though ‘speculative’
interpretation, as a velellid float (pneumatophore). He examined other possibilities,
such as an affinity with brachiopod shells, but found only superficial resemblances.
Superficially, Ovatoscutiim may also be thought to resemble a brachiopod shell, but it
must have been quite flat, as the sculpture is not much distorted. The notch is unlike any
brachiopod structure. The possibility of the suture-like zone being a hinge of a bivalved
shell was also considered, but there is no discontinuity across this zone. Much of
Caster’s interpretation is concerned with proof of the existence of a ‘sail’ \n Plectodiscus.
No evidence of such a structure exists in Ovatoscutiim which, if correctly identified as a
pneumatophore of one of the Hydrozoa Chondrophorina, would have resembled
that of the living Porpita rather than that of Velel/a. Harrington and Moore (in Moore
1956, p. F149) gave the following diagnosis of the Family Porpitidae Brandt, 1835:
‘Corm highly vaulted or flattened disc-like, without crest or sails.’ Ordovician to Devon-
ian disc-like fossils assigned to this family have at least some radial sculptural elements.
In this respeet the new genus differs from all Porpitidae but this does not necessarily
aflect its tentative interpretation as a chondrophoran pneumatophore.

B. Pennatulacean Fossils

The interpretation of the genera Rangea, Pteridinium, and Chanda as representatives
of the Order Pennatulacea (Anthozoa Octocorallia) was discussed in previous pub-
lications to which reference is made in the following systematic descriptions. New
material confirms this assignment but does not yet permit a classification at family
group level. The family Pteridiniidae Richter 1955 was based on a genus which is still
incompletely known. It is not advisable to re-define it at this stage. The material from
Ediacara which was previously placed in the species Rangea arborea Glaessner can
now be separated into a species of Rangea (not including the holotype of arborea) and

614

PALAEONTOLOGY, VOLUME 9

another species. Its markedly different characters had been considered earlier, on the
evidence available from a limited number of specimens, as the result of post-mortem
alteration of Rangea. The description of the genera and species now known requires a
revision of descriptive terminology. For those genera which are close to Pennatulidae
the terms applied to living forms can be used, although no polyps are discernible. But
other genera deviate from living pennatulids and require new descriptive terms.

The body of all these fossils consists of a rhachis from which a leaf-like expansion
{frond) spreads with bilateral symmetry, as a plant leaf spreads from its mid-rib. The
term ‘leaf’ is not available, however, as in living pennatulids the numerous subordinate
individual ridges which spread from, or are wrapped around, the rhachis are known as
polyp leaves. They correspond to the primary branches which are delimited in fossils by
primary furrows extending laterally from the sides of the rhachis or from the mid-line.
In Pennatula the polyp leaves are deeply separated, so that there are no primary furrows,
while in Rangea they are free only near the periphery (text-fig. Ic). In the species arborea
they appear to have been rising above one surface of a foliate base (the main part of the
frond), so that the term ‘primary furrow’ here becomes ambiguous. It could refer to a
line on either side of the basal attachment of the polyp leaf or, since the polyp leaves
may be flattened against the foliate base in fossilization, it could also refer to what in
life was their free upper edge. The term basal trace is therefore used instead of primary
furrow for this structural type. In Pteridiniiim, however, the primary branches are
delimited by clear-cut primary furrows. In Rangea and Charnia the primary branches
are clearly divided into secondary branches which correspond to the anthosteles in
Pennatulacea. Such secondary branches and delimiting secondary furrows are also
seen on the ventral side of arborea, but only very rarely and faintly in Pteridiniwn.
The designation of the two sides of the foliate base as dorsal and ventral is conventional,
as in all Pennatulacea, the rhachis often appearing as a wide track dorsally and as a
narrow zigzag line between contiguous polyp leaves ventrally. The rhachis may extend
downwards (in the living position) into a stalk which in Recent forms may end in a
bulbous expansion. There are fossil specimens showing a long stalk with a pennatulacean
fossil (like Charnia Ford) at one end, but there are also similar stalks ending in a round
medusoid cast (like Charniodiscus Ford). No unquestionable fossil with a medusa-like
base and a pennatulacean upper end has yet been found.

Genus rangea Giirieh 1930
Type species. Rangea schneiderhoehni Giirieh 1930.

The relations between the genera Rangea, Charnia, and Pteridiniwn require further
study. A discussion of these general problems is beyond the scope of the present contri-
bution.

Rangea longa sp. nov.

Plate 100, fig. 4; text-fig. 1

1959a Rangea sp. nov. (part mi) Glaessner, pp. 1472-3.

1959 Rangea arborea (partim) Glaessner, in Glaessner and Daily, p. 383, pi. 45, fig. 1 (only).

1962 Charnia sp. b, Glaessner, pi. 1, fig. 5.

Material and preservation. Twelve specimens in good to fair preservation belong here. On one bedding
plane they are preserved on the upper surfaces of rock slabs (Glaessner, in Glaessner and Daily 1959,

GLAESSNER AND WADE: LATE PRECAMBRIAN EOSSILS EROM EDIACARA 615

pi. 45, fig. I) but elsewhere they occur on lower surfaces; this is most common in all groups of fossils at
Ediacara.

Holotype. PI 3777.

Dimensions. The holotype is 15 cm. long and about 6 cm. wide. The largest fragmentary specimen is
25 cm. long.

Diagnosis. Angle between axis and primary branches (polyp leaves) commonly acute,
but it could evidently be modified by bending. Polyp leaves made up of secondary
branches (anthosteles) of uniform appearance, elongated approximately at right angles
to the free (upper) edges of the polyp leaves. The secondary branches are separated by
furrows that are deeper at their upper ends. Twenty or more secondary branches in
larger leaves, fewer in smaller leaves in every example.

Description. Outline of frond variable, from
fusiform to very elongate. All specimens
lie flat (or almost flat) on dorsal or ventral
sides, indicating that their width must have
exceeded their thickness prior to compres-
sion. The margins are lobate, the outwardly
convex portions corresponding to the polyp
leaves, and the indentations to the furrows
between them. Normally these primary
furrows are straight, lying at angles from
25° to 50° to the axis, the angles being steeper
near the apex of the frond. The secondary
furrows form approximately right angles
with the upper edge of the polyp leaf. The
largest number of these depressions counted
is 17 but counts could not be made on
complete leaves. The secondary furrows are
most distinct near their upper ends, which
could have divided the upper margin of the
polyp leaves into separate lobes. In their
position and varying degree of distinctive-
ness they correspond to the anthosteles of
some modern pennatulids. In some speci-
mens some secondary branches are trans-
versely wrinkled. The width of the polyp
leaves in the mid-portion of a frond is from
about one-third to one-sixth of their length.
Approximately 1 : 5 is the most common
ratio.

TEXT-FIG. 1. Rangea iongasp. nov. Reconstruction
based on the holotype and three other specimens,
xO-5 (approx.), a, ventral view; b, dorsal view;
c, diagrammaticcross-section along transverse line,

Remarks. The most obvious differences
between the type species R. schneiderhoehni
and the Australian form are possibly due to preservation. They are the questionable
‘outer flange’ of the holotype of R. schneiderhoehni and the oblique angle which its

616

PALAEONTOLOGY, VOLUME 9

secondary furrows make with the primary furrows. The African material consists, as far
as we know, of only two specimens and is inadequate to elucidate which of its characters
are diagnostic. No other material of this species has been obtained to date. Rather
than doubtfully assign the better-known Australian material to the less well-known
African species, we have described it as a new species. The holotype and other specimens
from Ediacara of a species previously placed in Ranged can now be distinguished
generically from R. schneiderhoehni, on the basis of new material. Some others are
placed in Pteridinium. One unique specimen is considered as specifically distinct from
R. longa.

Rangea grandis sp. nov.

Plate 100, fig. 5

1959a Chaniia sp., Glaessner, p. 1472, text-fig. lb.

1959 Rangea? sp. (changed in addendum to Chaniia sp.), Glaessner, in Glaessner and Daily,
p. 397, pi. 46, fig. 2.

1961 Chaniia sp., Glaessner, p. 75, text-fig.

1962 Chaniia sp. a, Glaessner, pp. 484-5, pi. 1, fig. 4.

Material and preservation. One fragment of the mid-portion of a large frond preserved as a cast on the
lower surface of a rock slab.

Holotype. PI 2897.

Dimensions of holotype. Overall length of fragment approximately 160 mm., maximum measurable
width 75 mm. The longer dimension of the lowest secondary branches (anthosteles) on either side is
20 mm., decreasing fairly regularly to 14 mm. in the highest pair of polyp leaves.

Diagnosis. Frond large. Ventral zigzag furrow irregular. Primary branches (polyp
leaves) make an acute angle with the axis (30°-40° in the holotype). Secondary furrows
divide the polyp leaves into secondary branches (anthosteles) with bluntly rounded tips
but become shallow grooves across the lower one-half to two-thirds of the polyp leaves.
The secondary branches are larger and fewer than in other species; they narrow some-
what towards the outer margins of the frond, and 13 is the greatest number known in
one polyp leaf.

Remarks. R. grandis differs from R. longa and from R. schneiderhoehni in its much larger
secondary branches (anthosteles), which are about three times as wide as the largest
known in R. longa, and fewer in each primary branch. Diminishing width of the
anthosteles away from the axial furrow has not been observed in R. longa or R.
schneiderhoehni. Like R. longa and unlike R. schneiderhoehni, the secondary furrows
of R. grandis are roughly at right angles to its primary furrows.

Genus pteridinium Gurich 1930
Type species. Pteridinium simplex (Gurich) 1930.

Topotype material of Pteridinium simplex was described by Glaessner (1963u).

Pteridinium cf. simplex (Gurich)

cf. 1929 ‘Nr. r, Gurich, p. 85.

cf. 1930a ‘Nr. 1’ Gurich, pp. 671, 680, fig. 1.

cf. 19306 Pteridinium simplex Gurich, p. 637.

GLAESSNER AND WADE; LATE PRECAMBRIAN FOSSILS FROM EDIACARA 617

cf. 1933 Pteridinium simplex (Giirich); Giirich, p. 144, fig. 4«-c.

cf. 1955 Pteridinium simplex (Giirich); Richter, p. 246, pi. 1-6, pi. 7, fig. 11.

71959 Pteridinium sp., Glaessner, in Glaessner and Daily, p. 382, pi. 46, figs. 3, 4.

cf. 1963 Pteridinium simplex (Giirich); Glaessner, p. 113, pi. 1, pi. 2, fig. 1.

Material and preservation. Two small and obscure specimens (P12744a', b, described in Glaessner
1959fl), and four or five fragmentary larger specimens (collected 1964 by M. Wade). One of these
(F17010a-c) is preserved as a generally convex cast (a) showing two fragments of a frond, with their
median zones forming approximately a right angle. The margins of the larger fragment are obscured
but appear to converge away from the smaller one which has the margins curled down giving a poly-
gonal outline. There is a counterpart or natural mould (b) of this specimen, generally concave, showing
the features of A in mirror image. A small piece of rock (c) was removed from this counterpart block.
It shows a continuation of the larger fragment of a extending for 10-20 mm. along the median zone and
actually overlying the left-hand side of the smaller specimen by about 10 mm. vertically. The two
specimens could represent two individuals overlapping accidentally, or two fragments of the same
individual. The other three specimens are convex casts, without counterparts. Two show short frag-
ments of one side of the frond without the median zone. The lateral margin is preserved in one of them.
The third fragment shows a short portion of the median zone and part of one lateral margin. As in the
Nama quartzite from South-west Africa, the specimens are contorted and they are found on surfaces
fractured by weathering. Only very small areas can be freed by further mechanical preparation.

Dimensions. The larger fragment of FI 70 10 is about 80 mm. long, with a maximum width of about
55 mm. The smaller fragment is about 35 mm. long and up to 45 mm. wide. The other specimens are
48, 45-5, and 42 mm. The distances between primary furrows are 4-4-5 (FI 70 10), 5-4, 2-3-3-5, and 5 mm.

Occurrence. Remains of Pteridinium were first recorded from the fossiliferous flaggy
quartzites. Others have since been found in a massive quartzite on a low hill top about
3000 ft. north-north-east of the southernmost outcrop of these beds, some 60 ft. strati-
graphically below the main fossiliferous beds. At this locality the rock outcrop is affected
by later silcrete formation during weathering. Pteridinium is the only fossil found here.

Remarks. The new material from Ediacara obviously represents one species. The two
specimens described earlier from the richly fossiliferous beds can be included in it only
doubtfully, because of poor preservation. This material has been compared in detail
with P. simplex from the Kuibis Quartzite of South-west Africa, as represented by about
18 specimens described by Richter (including the neotype of which a cast is available to
us), 8 specimens lent by the Geological Survey of South Africa, and 8 specimens lent by the
Museum of South-west Africa. Thirty-four specimens of this species are now known.

The characters ofP. simplex and their variability have been comprehensively described
by Richter (1955), who gave careful attention to the influence of peculiarities of preserva-
tion on the appearance of these fossils. Little can be added to his observations. Some
additions, based on the South African Geological Survey material (Glaessner 1963)
can be summarized as follows: (1) The presence of secondary furrows between the
primary furrows, noted by Giirich but questioned by Richter, has been confirmed on
one specimen. (2) The interpretation of the course of the primary furrows as ‘forward
and outward’ as proposed by Richter has been questioned and a more transverse
course with distal convexity assumed. Admittedly, this is still hypothetical and the true
position of the proximal end has not been established beyond doubt in any specimen,
(3) A broad (5 mm.) median zone was found in one specimen (No. 1), instead of the
median narrow zigzag line, but here again there is no confirmation from other speci-
mens. The removal of a piece of quartzite at the lower end of this specimen (as figured on
pi. 2, fig. 1, Glaessner 1963) has revealed a part of a convex cast 70 mm. long along the

618

PALAEONTOLOGY, VOLUME 9

median zone. Its lateral zone is separated from that of the originally described concave
mould by up to 7 mm. of rock. This could be a ‘lower surface’ (as in Richter’s pi. 2,
fig. 3 and pi. 4, fig. 5c, x-x), but no less than four other interpretations of this specimen
are possible, based on varying assumptions of folding-over and subsequent loss by
erosion or lithification of half-specimens, as described in some of Richter’s specimens in
which duplication by apposition has occurred. Some of these interpretations would make
it doubtful whether the broad median zone now visible on the mould is an original
character of a single specimen. Two of the new specimens from Ediacara and five of the
eight specimens of P. simplex from the Museum of South-West Africa again show only a
narrow median zigzag line ; in the others this region is not preserved or not clearly visible.

The material from Ediacara, like that from South-West Africa (with the one possible
exception mentioned) shows identical aspects of Pteriditiiiim, consisting of a leaf-
shaped surface with a median zone developed as a narrow zigzag line from which
primary furrows extend laterally, with slightly staggered placing of their origins.
Measurements of the spacing of these furrows indicate a slightly greater width (5-0,
5-4 mm.) in two of the Ediacara specimens than in the African material, in which
4-7 mm. seems to be the widest and about 2 mm. the narrowest. No gradients of spacing
along the frond length have been observed and too few reliable widths of fronds to
which spacings could be related are observable to be significant for comparison.
Another possibly diagnostic character of the Ediacara material is a convergence of
primary furrows (and narrowing of primary branches) towards the margins of the fronds.
This, however, was observed only on 5-6 primary branches, while others give only a
general impression of greater curvature. The distinction between inner and outer lateral
zones is weaker than in some South-West African specimens but this is a variable
character.

The distinguishing characters of the few specimens of Pteridinium from Ediacara do
not appear to be sufficiently diagnostic from those of the much more abundant material
of P. simplex to establish a new species.

The interpretation of Pteridinium rests largely on its similarity with Pangea and
A rborea. Richter had referred both forms to the Gorgonacea, assuming that the frequently
observed apposition of specimens was indicative of the diagnostic branching of fronds
in this group. The large number of specimens of Pteridinium now known, without a single
branching one among them, disposes of his argument. Richter also concluded from the
apposition and from the apparently sediment-infilled specimens that Pteridinium had a
‘lower’ surface which was virtually identical with the upper surface and in which ridges
rising upwards corresponded in position to grooves, and depressions to the upwardly
convex primary branches. This remains questionable. Alternatively, the ‘lower’ surface
could have been featureless and perishable, leaving essentially an ‘upper’ lamina as the
only potential fossil to represent a former somewhat Renilla-MkQ organism.

Genus arborea gen. nov.

Type species. Rangea arborea Glaessner 1959.

Diagnosis. Dorsal track broad, ventral track narrow to fairly broad, zigzag-shaped;
rhachis extends downwards to form a stalk ending probably in a basal expansion.
The rhachis extends laterally into a thin, smooth, foliate base forming a frond with

GLAESSNER AND WADE: LATE PRECAMBRIAN FOSSILS FROM EDIACARA 619

entire margins. Lateral primary branches can be seen as ridges on both ventral and dorsal
sides of the foliate base, but they do not project as much on the dorsal side as they do
ventrally. The mainly ventral attachment of the
foliate base to the rhachis produces in less com-
pressed specimens triangular pocket-like hollows
between the rhachis and each of the primary
branches on the dorsal side. Flange-like polyp
leaves showing grooves (secondary branches),
apparently delimiting fused anthosteles, arise from
each basal trace. They are probably confined to the
ventral side and end before reaching the margin
or on it. Their preservation in varying positions
indicates that their free edges extended away from
the foliate base in life. Sharp, straight, linear
grooves on the rhachis (‘median field’) and
primary branches can be interpreted as impres-
sions of spicules.

Remarks. Abundant new material has led to a
reinterpretation of the pennatulacean fossils
from Ediacara. The forms with pronounced
spicular impressions, formerly considered as dis-
tinguished from typical forms of Rangea mainly
by their state of preservation, also possess flange-
like and flexible primary branches which do not
reach the leaf margins.

Arborea arborea (Glaessner) 1959

Plate 102, figs. 1,2; text-fig. 2

1959a i?a//grasp.nov. (/)a/7/m) Glaessner, pp. 1472-3.

1959 Rangea arborea Glaessner, in Glaessner and
Daily, p. 383, pi. 43, figs. 1-3, ?fig. 4; pi. 44,
figs. 1-3; pi. 45, fig. 2 (part), ?pl. 46, fig. 1.

Material. 35 specimens.

Dimensions. The incomplete holotype (PI 2891) measures
15 cm. along the axis. Its greatest measurable half-width
is 5-5 cm. The largest known specimen is about 60 cm.
long and over 10 cm. wide.

Diagnosis. Spicules in the axis lie approximately

longitudinally in the dorsal track, but zigzag from

side to side of the ventral track at the insertions

of primary branches. In the frond, spicules are

approximately parallel to the primary branches.

o I rru II i. u ■ j ■ TEXT-FIG. 2. Arborea arborea (Glaessner).

Remarks. The polyp leaves may be buried m a y., <• u w a a 'u

. y ^ Drawing of the holotype surrounded by

collapsed position, forming a ridge on the basal reconstruction based on other specimens.

trace, with all structure blotted out, or they may xO-5.

T t

4397

620

PALAEONTOLOGY, VOLUME 9

be folded flat against the foliate base, upwards (i.e. towards the apex), or downwards
(towards the stem). Composite moulds show the secondary branches (anthosteles) as
convex projections and the secondary grooves as furrows. They were soft enough to
collapse like medusae downwards into the sediment while those of Rangea Jonga (but
not R. grandis) usually stood up long enough to produce external moulds at the base of
the overlying sedimentary layer. This supports the supposition that they are often absent
because of comparatively rapid disintegration of the polyp leaves which also brings the
supporting spicules to the surface to form clear-cut impressions. As the polyp leaves of
Arborea show more clearly on the ventral than on the dorsal side of a foliate base,
this genus may be compared with Renilla, but in this living form there is no linear
arrangement of the polyps on the surface of the foliate base.

Phylum Annelida
Genus dickinsonia Sprigg 1947
Type species. Dickinsonia costata Sprigg 1947.

This genus, representing the Family Dickinsoniidae Flarrington and Moore 1956,
differs from Spinlher in the absence of the claw-like parapodia which this living genus
uses in its predatory life on sponges. Dickinsonia was a free-living organism, apparently
moving on the sediment or in the water above it by body undulations rather than by
parapodial crawling. Specimens of Dickinsonia are preserved as moulds or casts. There
are partly decayed specimens showing frayed edges and traces of longitudinal muscles,
and specimens with multiple impressions of the peripheral margin resulting from con-
traction after the first contact with the substratum.

Dickinsonia costata Sprigg
Plate 101, fig. 4

1947 Dickinsonia costata Sprigg, p. 221, pi. 7, fig. 2.

1947 Papilionata eyrei Sprigg, p. 223, pi. 8, fig. 2.

1949 Dickinsonia costata Sprigg, D. minima Sprigg, p. 95, pi. 18, fig. 2; pi. 19, figs. 1, 2; pi. 20,
figs. 1, 2; pi. 21, figs. 1-4; text-figs. 9, 10.

1955 Dickinsonia spriggi Harrington and Moore, p. 160.

1956 Dickinsonia costata Sprigg, D. minima Sprigg, D. spriggi Harrington and Moore, in

Moore, p. F24, figs. 13-16.

1956 Papilionata eyrei Sprigg; Harrington and Moore, ibid., p. F159.

19596 Dickinsonia costata Sprigg, D. minima Sprigg; Glaessner, p. 526, fig. 4.

1959 Dickinsonia costata Sprigg; Glaessner, p. 379.

EXPLANATION OF PLATE 101

Figs. 1-3. Pteridininm cf. simplex (Giirich). 1, 2, counterpart casts and moulds of specimen
F17010A-B above and possibly another specimen below. 3, same as 2 with part of cast (F17010c)
which was removed in 2 to reveal underlying surface, x 1 .

Fig. 4. Dickinsonia costata Sprigg. Hypotype, T53. Cast with faint segmental sculpture and with
infilled intestinal caeca on antero-median part. X 1.

Fig. 5. Trihrachidiitm heraldicum Glaessner. Holotype, PI 2898, X 3. As the holotype is an external
mould (impression), the direction of the curvature of the three arms is reversed from that in the
animal as represented by latex casts (PI. 99, fig. 5).

Palaeontology, Vol. 9

PLATE 101

GLAESSNER and WADE, Eate Precambrian fossils from Ediacara, South Australia

GLAESSNER AND WADE; LATE PRECAMBRIAN FOSSILS FROM EDIACARA 621

1959 Papilionata eyrei Sprigg; Glaessner, p. 380.

1961 Dickinsonia costata Sprigg; Glaessner, p. 74, text-fig.

1962 Dickiiisonia sp. Glaessner, pp. 484, 493, pi. 1, fig. 7.

Material. Approximately 250 specimens occur as impressions on the bases of beds; several have
counterpart casts on the tops of the underlying beds.

Dimensions. Generally the length exceeds the width but the largest specimen is about 17 cm. long
and 18-5 cm. wide, probably because of contraction. The smallest specimen is less than 1 cm. long.

Diagnosis. Broad, flat polychaet worms, growing to large size; with left and right
halves of anterior body-segment fused together pre-orally; with intestinal caeca that
branch dichotomously around pharynx ; pharynx situated at centres of first few segments ;
segmental furrows depressed dorsally and ventrally; dorsal lamellae rarely fossilized.

Remarks. The discovery of a specimen of D. costata showing some intestinal caeca has
established the position of these fossils among the annelids. The pattern of the caeca
has increased the previously noted known resemblances of Dickinsonia to Spinther, the
only other worm known with a similarly fused anterior body segment, which were
indicated by Glaessner (19596, 1961, 1962).

The specimen for which Harrington and Moore (1955) erected the species D. spriggi
is badly preserved and has had most of its edges broken off at the margins of the rock.
The lineation considered as the edge of the body by Harrington and Moore is a slight
step fault in the rock, and the impression is markedly shallower on the down-thrown
side than in the more axial portion of its body. This change of preservation occurs at
differing distances from the axis. The faintly preserved furrows are not discrete marginal
tentacles and this specimen does not provide evidence for the placing of Dickinsonia
in a coelenterate ‘Class Dipleurozoa’ as proposed by Harrington and Moore. It is
probably D. costata as Sprigg (1949) had thought. Numerous newly discovered speci-
mens have bridged the apparent gap separating D. costata from D. minima Sprigg,
which is accordingly placed in the synonymy of the type species.

Dickinsonia elongata sp. nov.

Plate 102, fig. 3

Material. One juvenile specimen with both anterior and posterior ends, one adult lacking anterior
end and right side, three distorted adults lacking anterior ends, and many fragments, some very large.

Holotype. PI 3767.

Dimensions. The holotype, a juvenile specimen, is just over 16 cm. long; it has approximately 100
segments. The adult specimen, which is partly preserved on a broken slab, measures 33-5 cm. with
250 segments but appears to have been buried in a limp and stretched condition. The three distorted
adults seem to have been at least 30-40 cm. long; one has 380 segments in about 31 cm., including the
diminishing segments of the posterior end. Fragments of greater width suggest larger specimens still;
one has 190 segments in a length of 19 cm., its half-width from periphery to median line tapering from
1 1 cm. to little more than 8 cm.

Diagnosis. Elongate, rounded anteriorly, tapering posteriorly. Only the anterior seg-
ments are appreciably shorter in the axial region than in the peripheral region. They are
followed by up to several hundred similar segments of constant length and width,
forming a long mid portion of the body. The proportion of width to length of these
segments is about 100:1 or more.

622

PALAEONTOLOGY, VOLUME 9

Dickinsonia tenuis sp. nov.

Plate 103, fig. 1

Material and preservation. 16 specimens of good to poor preservation and a number of fragments.
The segments are so short that the grain size of the sediment tends to blot them out. Some specimens
are preserved as casts on the bases of quartzite slabs, others as external moulds.

Holotype. P13769n.

Dimensions. Not many specimens are well enough preserved for accurate measurements. The species
is very variable in its proportions but the smaller forms are all relatively narrow (some being half as
wide as long, others relatively wider) while most of the larger forms are not much longer than wide.
The holotype is the largest specimen. It is almost complete but the posterior end is very faint. It is
19-21 cm. long and 15-16 cm. wide, the width probably being reduced a little by the position of the
body. It had approximately 380 segments.

Diagnosis. Rounded to elongate outline, very short segments, numbering up to several
hundred. They are three or more times as numerous as in specimens of D. costata or D.
elongata of the same length.

Remarks. While the overall shape of D. tenuis is similar to that of D. costata, its very
short segments clearly differentiate it from both D. costata and D. elongata. A con-
tracted specimen shows the same pattern of ridges and grooves as a contracted D.
costata. The preservation of several individuals as natural casts on the bases of quartzite
slabs may indicate that the tissues of D. tenuis were less resistant to decay than those
of D. costata and D. elongata, since in the Ediacara fossil beds natural casts on the
lower surfaces of beds are otherwise known only among the medusoids.

Genus spriggina Glaessner 1958

Type species. Spriggina floimdersi Glaessner 1958.

Studies of abundant new material of the species of Spriggina are in progress and will
be reported later. A new species tentatively assigned to this genus is recorded here.

Spriggina^. ovata sp. nov.

Plate 98, fig. 4

Material and preservation. Four external moulds. The fossils are here described from latex casts.

EXPLANATION OF PLATE 102

Figs. 1, 2. Arborea arborea (Glaessner). 1, Composite mould, hypotype, PI 3787, showing ventral
narrow zigzag line above, wide dorsal rhachis below, with some lateral spicule impressions to the
right and flattened primary branches (polyp leaves) ; secondary branches (anthosteles) on left side.
2, hypotype, FI 701 8, composite mould with dominantly ventral aspect, showing primary and
secondary branches, d marks lateral margins of rhachis on dorsal surface, preserved by composite
moulding. The polyp leaves appear to be mostly turned upwards on the left and downwards on the
right. X 1.

Fig. 3. Dickinsonia elongata sp. nov. Holotype, PI 3767. An impression (external mould) showing some
distortion of the axial furrow (a ridge in the mould), due probably to the intestine or other internal
organs. The surfaces of the segments are wrinkled by contraction of transverse muscles, with folds
continuous across anterior segments but interrupted at segmental boundaries of the smaller middle
and posterior segments. X 1 .

Fig. 4. Praecambridium sigillum gen. et sp. nov. Holotype, PI 3794, X5.

Palaeontology, Vol. 9

PLATE 102

GLAESSNER and WADE, Late Precambrian fossils from Ediacara, South Australia

GLAESSNER AND WADE: LATE PRECAMBRIAN FOSSILS FROM EDIACARA 623
Holotype. PI 3754.

Dimensions. In mm.

No.

Length along axis

Max. width

No. of body
segments

P13754

22

8-5

about 40

P13755

19

80

about 40

P13757

31 approx.

11

probably nearly 50

P13756

broken at 24

15

23 present

Diagnosis. Elongate-oval polychaet worms. Prostomium arcuate, narrower than the
broadest part of the body, which is a short distance behind it. The tapering postero-
lateral processes of the prostomium possibly bear setae. Segments up to 50 or more,
very short and broad, with parapodia apparently supported by bundles of long acicular
setae. An axial groove is bounded on each segment by a pair of small ridges which are
broadest and highest adaxially. On unflattened specimens the ridges terminate in a pair
of small bosses adjacent to the parapodia, at least on the larger segments. A pair of small
elongate ridges (such as teeth could form) occupy a near-axial position just behind the
prostomium.

On the best-preserved and on one flattened specimen, long setae are attached to the
segments, between the axis and the parapodia. These setae are generally oriented in
backward-curving groups and tend to obscure the segmentation.

Remarks. In its arcuate prostomium with postero-lateral processes S.l ovata resembles
S. Jioundersi, but its prostomium was not equally resilient and only its anterior edge is
clearly preserved. A pair of large teeth was probably present, while there is no evidence
of teeth in S. floimdersi. There were no long dorsal setae in S. fioundersi.

? Phylum uncertain
Genus praecambridium gen. nov.

Type species. Praecambridium sigillum sp. nov.

Characters of type species.

Praecambridium sigillum sp. nov.

Plate 102, fig. 4; text-fig. 3

Material and preservation. Seven moulds on the bases of quartzite slabs. Four specimens of good to
fair preservation are on one slab, and three on isolated slabs. Two further isolated specimens show
similar but less distinctive features. The species is here described from latex casts.

Holotype. PI 3794.

Diagnosis. Oval disc-shaped bodies less than 5 mm. in length and under 4 mm. wide.
One is slightly conical, rising 0T5 to 0-2 mm. above the bedding plane; the remainder
are flattened. The surfaces of the casts each bear three pairs of small raised lobes and an
axial lobe. These lobes are more or less confluent in the centre. The axial lobe of the
best-preserved specimen bears another pair of small lobes laterally. All the paired lobes
taper to pointed outer ends which are directed towards one end of the disc. This is here
considered the posterior end. The axial lobe is rounded at the opposite (anterior) end.

624

PALAEONTOLOGY, VOLUME 9

Remarks. The interpretation of these minute fossils is difficult. It seems reasonable to

assume that the lobes are muscle impressions in a thin and
probably rather soft shell. Like other Ediacara fossils some
of these forms were distorted but they are not known to
have broken. The fossils may be composite moulds of thin
shells with segmental muscles, but they do not conform
with the pattern of the Monoplacophora, as the muscle
impressions (if they are correctly interpreted as such) were
partly median. In this respect there is a resemblance with
the description of the larger Cambrian genus Carnbridium
Horny 1957, but no close relationship is suggested.

A more striking resemblance is found with the Lower
Cambrian genera Mobergello Hedstrdm 1923, and Disci-
nella Hall 1872. [The genus Barella Hedstrom 1930 is
invalid. It was based on Barrande’s ‘opercule isole H’,
from the middle Ordovician of Osek in Bohemia, which
was recognized by Zazvorka in 1930 as an operculum of
Hyolithes paxiUosus Novak.] The difficult question of the
status of these genera and their true nature is still undecided. It was reviewed recently
by Ahman and Martinsson (1965) who also gave the first clear illustrations of Mober-
ge/la holsli (Moberg). Their final conclusion is as follows:

As long as the number of muscle scars may be accepted as a distinctive character, this fossil should be
called Mobergella Iwlsti (Moberg). For the first time it has been found together with tubes with which
it may be associated as an operculum, but the material leaves us without an absolutely convincing
solution of the classical problem whether the disc has belonged to a sedentary hyolithellid or is the
remains of a free-living mollusc. Without turning a blind eye on the latter interpretation we may state
that the accumulated observations are very strongly in favour of the former one.

The new form has not been found together with Hyolithelliis-\ike tubes. It is dis-
tinguished by its elongate-oval outline and by the pattern of the presumed muscle
impressions, which show a concave curvature towards the posterior end of the disc,
while in the known genera they are straight or concave anteriorly. They are confluent
along an elongate median zone while in Mobergella and Discinella they merge at the
apex in the anterior third of the length of the shell. No such eccentric apex is known in
Praeeambridium. Whatever will be the final decision about the systematic position of
the problematic Cambrian genera, the new genus is morphologically distinct, differing
from them more than they differ from each other, yet similar enough to be considered
as probably related to them.

TEXT-FIG. 3. Praeeambridium
sigilliim gen. et sp. nov.
Holotype, PI 3794, x 10.

EXPLANATION OF PLATE 103

Fig. 1. Dickiitsonia tenuis sp. nov. Holotype, PI 3792, x 0-8. The specimen is a cast on the lower surface
of a bed. The anterior end is in the right upper corner. The body was deposited across a stem of a
large but poorly preserved Arborea which was partly sand-filled (s). Arrows mark its centre line,
pointing upward. Lateral branches begin to show on the rock about 5 cm. to the left and a bulbous
expansion about 10 cm. to the right from the margins of the picture.

Figs. 2-4. Pseudorhizostomites bowcluni Sprigg. 2, Hypotype, F17011, view from the lower surface of
the bed. 3, 4, views of the sides of the break seen across the specimen in 2, showing fluted surfaces
of the passage above the centre furrow. Arrows point to the upper surface of the bed. X 1.

Palaeontology, Vol. 9

PLATE 103

GLAESSNER and WADE, Late Precambrian fossils from Ediacara, South Australia

GLAESSNER AND WADE: LATE PRECAMBRIAN FOSSILS FROM EDIACARA 625

Genus parvancorina Glaessner 1958

Type species. Parvancorina ininchami Glaessner 1958.

Parvancorina minchami Glaessner 1958

1958 Parvancorina ntinchaini Glaessner, p. 187, pi. 1, fig. 4.

1959 Parvancorina minchami Glaessner, p. 528, text-fig. 3.

1959 Parvancorina minchami Glaessner, in Glaessner and Daily, p. 380, pi. 47, figs. 5, 6.

1960 Parvancorina minchami Glaessner, pp. 60, 62.

1961 Parvancorina minchami Glaessner, p. 75, text-fig.

1962 Parvancorina minchami Glaessner, p. 484, pi. 1, fig. 9.

Dimensions. The largest specimen, obliquely deformed, is 26 mm. long and 27 mm. wide. Less de-
formed specimens are 25, 23, and 22 mm. long and, respectively, 21,21, and 23 mm. wide. The smallest
specimen is 3 mm. long and less than 2 mm. wide.

Diagnosis. Small shield-shaped fossils with an elevated, anchor-shaped, antero-median
ridge bearing curved, linear postero-lateral appendages on its anterior arms and finer
transverse appendages on its median bar.

Remarks. About 60 specimens of this form are now available, representing a wide range
of sizes. Other specimens found recently have not shown any additional morphological
characters. They show the growth range and the post-mortem deformation of the
integument, which was pliable and possibly chitinous. It is not clear whether some
feeble ridges on the inverted anterior portion of one specimen are reflected dorsal or
independent ventral structures.

Genus tribrachidium Glaessner 1959
Type species. Tribrachidium heraldicum Glaessner 1959.

Tribrachidium heraldicum Glaessner 1959

Plate 97, fig. 9; Plate 99, fig. 5; Plate 101, fig. 5

1959 T. heraldicum Glaessner, in Glaessner and Daily, p. 389, pi. 47, figs. 7, 8.

1960 T. heraldicum Glaessner, pp. 60, 62.

1962 T. heraldicum Glaessner, p. 484, pi. 1, fig. 6.

Material and preservation. 40 specimens are available; 34 upper (oral) sides, one of which has a counter-
part showing the lower (aboral) side. The aboral side is an almost featureless disc and this explains the
failure to identify it when its casts or moulds occur alone. The remaining 5 specimens are composite
moulds dominated by the rather resistant marginal zone of the body, which forms an outer ring.

The material occurs as impressions (external moulds) on the lower surfaces of quartzite slabs, with
the exception of the one external mould of the aboral side which is separated from the overlying lamina
only by a thin, sericitic parting and the space contained between the two laminae. Elsewhere on the
same bedding plane the lower lamina remained sufficiently plastic during diagenesis to form a natural
cast of the impression of another specimen in the upper lamina. Another natural cast of a different
specimen of Tribrachidium is known from a rock with only a thin sericitic parting between the laminae.
Thus all except one of the counterparts are natural casts of the mould on the overlying lamina, or com-
posite moulds.

The species is described from latex casts of the external moulds.

Holotype. PI 2898.

626

PALAEONTOLOGY, VOLUME 9

Diagnosis. Disc-shaped fossils, slightly biconvex, with a steeply sloping margin; up to
about 4 cm. in diameter. The diameter of the smallest specimen on which the edge of the
disc can be seen, is 9 mm. The oral side has three raised brachia radiating from the
centre and curving in a clockwise direction to become parallel to the periphery, and
tapering to their extremities. The curve is more evenly rounded in small specimens and
more angular in large ones. A small, central, Y-shaped groove which may represent the
mouth is seen very rarely between the arms. Dome-shaped ‘bullae’ (Glaessner 1959 b)
occupy part of each interbrachial space, each being attached to the neighbouring
convexly curved side of an arm. The arms (brachia) and bullae are elevated above the
oral surface and have resisted flattening to some extent. The distal two-thirds of each
arm bears short stout tentacles on the outer side and tip. The number of tentacles is
generally greater in larger specimens, and some appear to be arranged in alternating
series. Fine, long, straight or gently curved bristles, all equal in width, extend from the
crest and the concave curve of each arm towards the periphery and can also be seen
extending from the surface of interbrachial spaces past the tip of an adjoining arm.
They are not preserved in all specimens.

Only a few, shallow, concentric grooves occur on the aboral side. A wide marginal
zone shows in composite moulds; though subject to some flattening, it does not transmit
structures of the oral side, which suggests considerable resilience.

Remarks. A distorted specimen, the largest (PI. 97, fig. 9), shows numerous bristles
extending radially from the crests of the arms, which are unusually sharp in this speci-
men. They obscure the surface of the arms. The greatest number of bristles that can be
counted as attached to one arm in this specimen is 40. The number of tentacles on the
three arms appears to differ slightly in some individuals and to differ strongly between
different individuals of the same size. In specimens over 2 cm. in diameter the number of
tentacles on one arm appears to vary between 16 and 30, though the exact number has
not been determined.

Glaessner (1959, 1960, 1962) has mentioned the superficial resemblance of Tribrachi-
diuni to certain Edrioasteroidea which show relics of tri-radial symmetry, but has
discounted suggestions of relationship between Tribrachidium and echinoderms because
of the complete absence of any traces of calcareous plates and of an ambulacral system.
Since then the resemblance has been strengthened by the discovery of specimens with a
central Y-shaped groove which could represent the mouth, and by the recognition of
bristle-like appendages on the surface. These are straight or gently curved, thin, long
and rather stiff but not brittle. In discussions with one of the present authors (M. F. G.),
Professor H. B. Fell (Museum of Comparative Zoology, Harvard University) argued
persuasively in favour of a relationship with echinoderms, considering the calcareous
skeleton as not essential for it and suggesting that the bristles might be tube feet. It is
difficult to see in the available material the precise manner of their attachment. They were
certainly not attached in regular rows to the crests of the arms, as the casual inspection of
an oddly preserved specimen (PI. 97, fig. 9) might suggest. They can be seen to cross the
arms from at least their inner margins, if not from the inter-brachial spaces. They could
have arisen in positions adjacent to the arms or, alternatively, all over the body like the
tube feet in Holothuria. They could have been respiratory organs, being stiff and
flexible, rather than prehensile and capable of being bent by muscles like true tube feet.

GLAESSNER AND WADE: LATE PRECAMBRIAN EOSSILS EROM EDIACARA 627

Thus they could have been the kind of ‘passive tube feet’ which Nichols (1962, p. 152)
suggests may have existed in Agelacrinidae. The basal surface of Tribrachidium was
probably a tough membrane like that which Nichols (p. 153) believed could have
existed in Edrioaster so that, as Bather had suggested earlier, it could have provided a
sucker for temporary attachment as well as a muscular base for limpet-like movement.
These new observations and suggestions provide a still tenuous line of argument con-
necting Tribrachidium with echinoderms, perhaps phylogenetically rather than taxono-
mically. The lophophore-like apparatus of tentacles will also have to be considered
in future functional and phylogenetic interpretations of Tribrachidium, but these are
beyond the scope of the present paper.

REFERENCES

AMMAN, E. and MARTiNSSON, A. 1965. Fossilifcrous Lower Cambrian at Aspelund on the Skaggenas
Peninsula. Geol. For. Stockli. Fork. 87, 139-51.

BEST, R. J. (ed.) 1958. Introducing South Australia. Australian and N.Z. Ass. Adv. Sci., Adelaide.

CASTER, K. E. 1942. Two siphonophores from the Paleozoic. Paleont. Americana, 3, 56-90, pi. 1, 2.

DALGARNO, c. R. 1964. Report on the Lower Cambrian stratigraphy of the Flinders Ranges, South
Australia. Trans. R. Soc. S. Aust. 88, 129-44.

GLAESSNER, M. F. 1958. Ncw fossils from the base of the Cambrian in South Australia. Ibid. 81,
185-8, pi. 1.

1959u. Precambrian Coelenterata from Australia, Africa and England. Nature, Loud. 183,

1472-3.

\959b. The oldest fossil faunas in South Australia. Geol. Rdsch. 47 (2), 522-31.

1960. Precambrian fossils from South Australia. Proc. 21st hit. geol. Congr., pt. 22, 59-64.

1961. Pre-Cambrian animals. Sclent. Am. 204, 72-78.

1962. Precambrian fossils. Biol. Rev. 37, 467-94, pi. 1.

1963(7. Zur Kenntnis der Nama-Fossilien Stidwest-Afrikas. Annin naturh. Mus. Wien, 66,

113-20, pi. 1-3.

19636. The base of the Cambrian. J. geol. Soc. Aust. 10, 223-42.

and DAILY, B. 1959. The geology and Late Precambrian fauna of the Ediacara fossil reserve.

Rec. S. Aust. Mus. 13, 369-401, pi. 42-47.

GURiCH, G. 1930(7. Die bislang iiltesten Spuren von Organismen in Siidafrika. hit. geol. Congr.,
S. Africa, 1929, 15, 670-80.

19306. Uber den Kuibisquarzit in Siidwestafrika. Z. deutscli. geol. Ges. 82, 637.

1933. Die Kuibis-Fossilien der Nama Formation von Sudwestafrika. Palaeont. Z. 15, 137-54.

HARRINGTON, H. J. and MOORE, R. c. 1955. Kansas Pennsylvanian and other jellyfishes. Bull. Kansas
geol. Sla v. 114 (5), 153-63, pi. 1, 2.

MAYER, A. G. 1910. The medusae of the world. Pubis Carnegie Instn, 109 (1-3), 1-735.

MCALESTER, A. L. 1962. Mode of preservation in early Paleozoic pelecypods and its morphologic and
ecologic significance. J. Paleont. 36, 69-73.

MOORE, R. c., (ed.) 1956. Treatise on invertebrate paleontology. Part F, Coelenterata. Geol. Soc. Am.

NICHOLS, D. 1962. Echinoderms. Hutchinson’s Univ. Libr., London.

NIXON, L. G. B. 1963. The Ediacara mineral field. Proc. Australas. hist. Min. Metall. 206, 93-112.

1964. Ediacara silver-lead-copper mineral field. Progress Report No. 1. Min. Rev., Adelaide,

116, 5-9.

RICHTER, R. 1955. Die iiltesten Fossilien Siid-Afrikas, Senckenberg. leth. 36, 243-89.

souTHCOTT, R. V. 1958. South Australian jellyfish. S. Aust. Nat. 32, 53-61.

SPRIGG, R. c. 1947. Early Cambrian ( ?) jellyfishes from the Flinders Ranges, South Australia. Trans. R.
Soc. S. Aust. 71, 212-24, pi. 5-8.

1 949. Early Cambrian ‘jellyfishes ’ of Ediacara, South Australia, and Mt. John, Kimberley District,

Western Australia. Ibid. 73, 72-99, pi. 9-21.

■628 PALAEONTOLOGY, VOLUME 9

TERMiER, H. and G., 1960. L’Ediacarien, premier etage paleontologique. Revue gen. Sci. pur. appl. 67,
79-87.

WALCOTT, c. D. 1898. Fossil medusae. Monogr. U.S. geol. Surv. 30, 1-201, pi. 1-47.

MARTIN F. GLAESSNER MARY WADE
Department of Geology,
University of Adelaide,

Manuscript received 26 August 1965 Adelaide, South Australia.

ADDENDUM

Two additional important specimens were collected at Ediacara in August 1966.
( 1 ) A complete external mould of a dickinsoniid worm, 77 mm. long, shows very fine trans-
verse grooving which is probably not segmental. Previously found specimens of this
form were distorted or badly preserved. Though its anterior end is similar to Dickinsonia,
the surface sculpture distinguishes this form specifically and possibly generically. (2) A
specimen attributed to Conomedusites is twice as large as the previous specimens. It
is a cast on the lower surface of a bed, with tentacles up to 14 mm. long around part
of its periphery but without a marginal flange. It resembles the Ordovician Conchopeltis
alternata Walcott 1876 (Moore and Harrington, in Moore 1956, p. F32, fig. 22, and
p. F57) but shows no radial striation.

DISTRIBUTION OF SPORE AND POLLEN
ASSEMBLAGES IN THE LOWER KITTANNING
COAL OF WESTERN PENNSYLVANIA

D. HABIB

Abstract. Palynological study of the Lower Kittanning coal seam of western Pennsylvania (Allegheny Series;
lower Westphalian D) has revealed an orderly vertical and geographical distribution of spore and pollen
assemblages.

Assemblages characterized by Densosporites and Piinctatisporites (P. obliquus) occur in the uppermost zones
where the seam is directly overlain by marine and restricted-marine faunal facies, respectively, of the Lower
Kittanning shale. The greatest vertical variation of assemblages occurs at these localities. Where the seam is in
direct contact with the overlying freshwater facies, however, it remains dominated throughout its thickness by
assemblages rich in Lycospora. Fewer assemblages occur at these localities. Because of the distribution of spore
and pollen assemblages in the seam and faunal facies in the overlying shale, a genetic relationship is suggested.

New taxa include two genera, Spackmanites and Paleospora, and thirty-three species.

The Lower Kittanning coal seam in western Pennsylvania is one of the more readily
identifiable and geographically extensive stratigraphic horizons of the Allegheny Series
(middle Pennsylvanian age). Because it is easily recognized, variations in the biological,
physical, and chemical properties of the seam and surrounding rocks have been investi-
gated. The known variation in these properties has prompted investigation of the seam
once again, this time to determine the vertical and geographic distribution of its spore
and pollen assemblages. It was assumed at the beginning of this study that if distinctive
spore and pollen assemblages at least partially represent in situ Lower Kittanning plant
communities, then some distributional pattern of their assemblage zones should be
discernible, in light of the known distributional pattern of other properties in the seam
and juxtaposed rocks.

Methods of study. Vertical columns of the Lower Kittanning seam were collected from 1 5
localities in the western part of the state (text-fig. 1). A one-inch increment of coal was
taken at three-inch levels in sequence in each column from the top downward. Where
the bottom-most sample did not coincide with this interval, it was taken as an additional
sample. The density of sampling sites was intended to cover a large area while keeping
the geographic interval small.

Maceration of the samples consisted of oxidation in nitric acid or Schulze’s solution,
and subsequent treatment in 8% potassium hydroxide. The spore residue was then
mounted on slides in glycerine jelly. Two hundred specimens were counted from each
sample (100 per slide). The entire slide was examined at x 100 and/or x450 magnifica-
tion, in order to determine the total number of species per sample.

Geological setting. The Kittanning formation comprises the middle third of the Allegheny
Series (Westphalian C-D). The formation is defined by three major coal horizons
which are separated by two thicker and lithologically more variable detrital units.
The lowermost coal seam, the Lower Kittanning, is overlain by the Lower Kittanning

] Palaeontology, Vol. 9, Part 4, 1966, pp. 629-66, pis. 104-9]

630

PALAEONTOLOGY, VOLUME 9

shale, a unit which separates the seam from the next overlying coal horizon by an
interval of from 30 to 50 ft.

On the basis of the distribution of fossil invertebrates in the Lower Kittanning shale,.
Williams (1959) differentiated facies depicting marine, restricted-marine, and fresh-
water environments of deposition. In a northwesterly direction across the sedimentary
strike of the marine embayment, the fresh-water facies of the Lower Kittanning shale^
represented by estherids, changes to restricted-marine {Lingula, Aviculopecten), and
marine {Mesolobus, other calcareous brachiopods, gastropods) facies in respective
order, before reverting through restricted-marine and fresh-water facies. Geochemical
evidence corroborates the delimitation of marine and fresh-water facies. Degens et aL
(1957,1958) distinguished these facies by relative abundance of trace element assemblages.
Greater concentrations of boron and rubidium were found in the marine shales, while
gallium was found to be more abundant in the fresh-water shales. Williams and Keith
(1963) subsequently showed a statistical relationship between the sulphur content of the
seam and the facies of the directly overlying rocks (Lower Kittanning shale), the content
being higher in the seam where it directly underlies ancient marine sediments. Weber
et a/. (1964) distinguished fresh-water siderite nodules in the Lower Kittanning shale
from marine nodules through distinct differences in their carbon isotope ratios.

Sampling Sites

Station 1. Section exposed in a limestone quarry, 7 miles west of New Castle, Pennsylvania. The
Vanport limestone is mined here and is exposed approximately 40-50 ft. below the Lower Kittanning
seam. Marine shale overlies the seam, characterized by Lingula and Mesolobus.

Station 2. Section exposed in coal strip pit, 1 mile east of West Pittsburg, Pennsylvania. The seam
is slightly thicker than that reported for the Lower Kittanning in this area.

Station 3. Section exposed in a road cut, on the east side of a toll bridge, 1 mile E. of Freedom,
Pennsylvania. The marine facies of the Lower Kittanning shale is represented here by Mesolobus.
Station 4. Section exposed in a coal strip pit, 5 miles north-west of Harlansburg, Pennsylvania. The
Lower Kittanning shale is characterized here by sandstone beds occurring 10 ft. above the seam.
Station 5. Section exposed in coal strip pit, 4 miles west of Slippery Rock, Pennsylvania. Plant remains
prevail in the nonmarine facies of the Lower Kittanning shale. Sandstone beds occur 6 ft. above the
seam.

Station 6. Section exposed in road cut 1 mile east of Nectarine, Pennsylvania. Thin irregularly bedded
sandstone, with plant remains, lies just above the seam.

Station 7. Section exposed in strip pit, 1 mile north-north-west of West Freedom, Pennsylvania. The
Vanport limestone is exposed approximately 40-50 ft. below the seam. The shale is characterized
by interlaminated light and dark layers (varves), with abundant plant fragments.

Station 8. Type locality of the seam. Seetion exposed in a railroad cut above west bank of Allegheny
River, outside of Kittanning, Pennsylvania. Mesolobus oeeurs in the shale a few inches above seam.
Station 9. Section exposed in strip pit 1 mile east of Turkey City, Pennsylvania. The Vanport limestone
is exposed approximately 25 ft. below the seam.

Station 10. Section exposed in strip pit, on a hill 2 miles west of U.S. Route 68, and 15 miles south of
Clarion, Pennsylvania. Mesolobus abounds in the upper portions of the shale, with Lingula just
above the seam. Shale is interlaminated in the lower part.

Station 11. Section exposed in strip pit 2 miles south of Limestone, Pennsylvania. The shale contains
Lingula.

Station 12. Section exposed in strip pit 1 mile west of Truittsburg, Pennsylvania. The shale contains
abundant Mesolobus a few inches above the seam. The seam varies within the pit from 64 to 38 ins.
in thickness, through a distanee of 100 ft.

Station 13. Section exposed in road cut on north side of creek, just north of Worthville, Pennsylvania.
Lingula occurs just above the seam, with Mesolobus predominant above.

D. HABIB: SPORE AND POLLEN ASSEMBLAGES

631

Station 14. Section exposed in strip pit 0-25 miles north-east of Hamilton, Pennsylvania. Avicidopecten
and Lingula are present just above the seam, with Mesolobus and other calcareous brachiopods
prevalent above.

Station 15. Section exposed in strip pit 1 mile south of Curwensville, Pennsylvania. Lingula and
Avicidopecten occur in the shale.

SYSTEMATIC DESCRIPTIONS

One hundred and forty species of spores and pollen, situated in 59 genera, were
recognized. Of these 33 are formally described as new, and 14 are given lettered specific
designations due to insufficient number of specimens. Two new genera, Spackmanites
and Paleospora, are proposed. All slides used in this study are stored in the Organic
Sediments Laboratory, The Pennsylvania State University. Position of specimens on the
slides is indicated by the scale settings of an E. Leitz ortholux microscope, serial no.
448497. All measurements are given with reference to the maximum diameter of speci-
mens. Wherever possible, holotypes were mounted as single grain wax mounts.

Anteturma sporonites (R. Pot.) Ibr. 1933
Genus reticulatasporites Leschik 1955

Reticulatasporites aletoreticuhis sp. nov.

Plate 104, fig. 1

Diagnosis. Alete spores; subcircular to circular in outline. Exine covered with concentric
reticulation pattern; lacunae largest in centre and become smaller and more lenticular
towards margin, as well as more low-lying; muri conversely become wider and more
low-lying towards margin, eventually grading into exine. No evidence of tetrad scar.
Exine approximately T5-2-0 p thick. Size range (twelve specimens) 53 to 66 p.

Holotype. Plate 104, fig. 1; 53x55 /x; slide LKC-2 (27-28) 1, coordinates 24-3 127-7.

Remarks. Spores of this species are characterized by the marginward diminution of its
reticulate pattern. The species was observed only at Station 2, where it is confined to
samples containing the Lycospora-Guthoerlisporites erectus assemblage. Reticulata-
sporites aletoreticuhis sp. nov. resembles the detached caps of specimens of Vestispora,
but could be distinguished in the Lower Kittanning seam by its larger size and orna-
mentational pattern.

Anteturma sporites H. Pot. 1893
Turma triletes Reinsch 1881
Subturma azonotriletes Luber 1935
Infraturma laevigati (Benn. and Kidst.) R. Pot. 1956
Genus leiotriletes Naumova ex Pot. and Kr. 1954
Leiotriletes sp. A
Plate 104, figs. 2, 3

Description. Trilete spores; triangular, with sharply rounded radial corners and concave
or slightly convex (due to folding) inter-radial margins. Exine one micron thick or
less; essentially levigate. Trilete mark distinct and slightly raised; it extends almost to
radial corners ; contact area defined by faintly sinuous (‘ wrinkled ’) curvaturae which turn

632 PALAEONTOLOGY, VOLUME 9

in inter-radially toward juncture of sutures. Two specimens, 41x41 /z, and 36x38 jix
(folded).

Remarks. Shape and thickness of spore coat relate the two specimens to Leiotriletes
ornatus Naum, and L. tumidus Butt, and Will. They are distinguished, however, by their
unique curvaturae. The specimens were observed only at Stations 5 and 7; both are
present in zones containing the Thymospora pseudothiesseni assemblage.

Genus calamospora S. W. and B. 1944
Calamospora multiplicata sp. nov.

Plate 104, figs. 6, 7

Diagnosis. Trilete spores; roundly elliptical in outline. Exine smooth, one micron thick
or less. From 3 to 9 curving, slender folds criss-cross exine in concentric and radial
directions; they are typically long and slender. Trilete mark short, extending less than
one-half radius; lips very thin and only slightly raised; contact area present, but only
slightly darker than remaining exine. Size range (twenty-five specimens) 43-60 p.

Holotype. Plate 104, fig. 6; 55 X 41 ju; slide LKC-10 (13-14) 2, 43-9 113-9.

Remarks. Calamospora multiplicata sp. nov. is present at every locality; it is most
common in the Thymospora pseudothiesseni and Lycospora-Guthoerlisporites erectus
assemblages. The species is distinguished from other species of Calamospora in the
Lower Kittanning seam by its size, roundly elliptical outline, and disposition of folds.

Calamospora elliptica sp. nov.

Plate 104, figs. 9, 10

Diagnosis. Trilete spores; elliptical to subelliptical-elongate in both polar and equatorial
views. Exine not over T5 thick; generally smooth, although the slightly darker contact
area may be slightly pitted; contact area not delimited by any specialized curvaturae.
Secondary folds always present, usually grouped near margin ; relatively long, slender.

EXPLANATION OF PLATE 104

All figures X 500.

Fig. 1. Reticidatasporites aletoreticidus sp. nov., holotype.

Figs. 2-3. Leiotriletes sp. A.

Figs. 4-5, 8. Calamospora pseudotriangulara sp. nov. 4, Holotype. 5, Proximal polar view with well-
developed contact area. 8, Thin dark folds attending the sutures prominent.

Figs. 6-7. Calamospora multiplicata sp. nov. 6, Holotype; slightly corroded but shows short trilete
mark and contact area. 7, Specimen with open trilete mark but which shows typical orientation of
folds.

Figs. 9-10. Calamospora elliptica sp. nov. 9, Holotype; thick lips border trilete mark. 10, Equatorial
view.

Figs. 1 1-14. Punctatisporites obliqiais Kos. ; note size variation, and nature of trilete mark.

Figs. 15, 18. Punctatisporites globulosiis sp. nov. 15, Holotype; lines of weakness extend longitudinally
from sutures on to distal surface. 18, Ruptured specimen, with triangular gap which extends to
distal hemisphere.

Figs. 16-17. Punctatisporites sphaerorigidus sp. nov. 16, Holotype; thick irregular fold on distal surface.
17, Specimen with T-shaped trilete mark.

Palaeontology, Vol. 9

PLATE 104

€

10

HABIB, Pennsylvanian miospores

D. HABIB; SPORE AND POLLEN ASSEMBLAGES

633

and smoothly tapering. Trilete mark distinct, extends about one-half radius bordered
by thick, raised lips. Size range (ten specimens) 80-100 jx.

Holotype. Plate 104, fig. 9; 85 x65 slide LKC-lOM-4, 37-7 128-5.

Remarks. The elliptical outline, long slender folds, and short thickly bordered trilete
mark distinguish this species.

Calamospora pseiidotriangulara sp. nov.

Plate 104, figs. 4-5, 8

Diagnosis. Trilete spores; subcircular to triangular in polar view, the outline appearing
triangular because of the large folds. Trilete mark distinct, extends from one-half to
two-thirds radius; thin but dark and distinct folds border it; contact area present,
barely discerned to very distinct. Large sharply tapering folds run peripherally and
often, because of their orientation, make the species appear triangular; largest peripheral
folds most often occur inter-radially. Exine less than one micron thick, levigate. Size
range (twenty specimens) 63-82 p.

Holotype. Plate 104, fig. 4; 75 X 63 /x; slide LKC-8 (36-37) 1, 30-2 127-1.

Remarks. The species is most common in the upper half of zones containing the
Lycospora-Guthoerlisporites erectus assemblage. It was not found at Station 13.
Calamospora pseudotriangu/ara sp. nov. is easily distinguished from other related
species by its size, triangular outline, thin exine, and long sutures bordered by thin,
distinct folds.

Genus punctatisporites Ibr. emend. Pot. and Kr. 1954
Pimctatisporites obJiquus Kosanke
Plate 104, figs. 11-14

Remarks. Because a very close relationship was observed between this species and
Laevigatosporites globosus Schem., they were counted together. Except for only a slight
difference in size ranges, the two species are very similar, and perhaps should be con-
specific. The true nature of the suture in either species is in doubt. Schemel (1951, p. 747)
noted the similarity between the species and admitted that if the nature of the suture
could not be determined, it would be doubtful to which species and genus the specimens
belong.

Punctatisporites sphaerorigidus sp. nov.

Plate 104, figs. 16, 17

Diagnosis. Trilete spores; circular to subcircular in outline, occasionally appearing
obliquely compressed, due to the presence of a thick fold. Exine generally smooth, but
may be irregularly pitted in localized areas; up to 5-6 p thick in polar view. Trilete
mark distinct and frequently open, extends about one-half radius or less; lips present
but only slightly raised; occasionally the trilete mark assumes a T-shape. Single large
compression fold, in some specimens accompanied by a smaller fold, usually present
restricted to distal hemisphere. Size range (twenty-five specimens) 75-98 p.

Holotype. Plate 104, fig. 16; 83 x 74 /x; slide LKC-2 (12-13) 1, 22-1 123-4.

634

PALAEONTOLOGY, VOLUME 9

Remarks. The species rarely exceeded one per cent in any sample. It was observed most
frequently in zones conidimmgihQ Punctatisporites obliquus and Thymospora pseudothies-
seni assemblages. Punctatisporites sphaerorigidus sp. nov. is distinguished by its size,
thick exine, and relatively short trilete mark. It differs from/*, obesus (Loose) Pot. & Kr.,
which also occurs in the seam, by its shorter trilete mark, smaller size, and relatively
thicker exine.

Punctatisporites globulosus sp. nov.

Plate 104, figs. 15, 18

Diagnosis. Trilete spores; elliptical to subcircular in outline. Exine covered over its
entire surface by dense, very closely spaced punctae; occasionally, probably due to
locally unornamented areas, a few low-lying (1 /x or less) rounded protuberances emerge
from the spore outline. Trilete mark distinct, and extends over one-half radius; lips
distinct, slightly raised; extending radially from the end of each suture is a thin line
along which the spores tend to rupture; ruptured specimens collapse usually to more
elliptical outline with opened trilete mark forming triangular outline and extending
well on to distal hemisphere. Size range (twenty-five specimens) 45-60 p.

Holotype. Plate 104, fig. 15; 56x53 p\ slide LKC-6 (15-16) 1, 20-6 118 0.

Remarks. The species was most frequently encountered in zones containing the
Punctatisporites obliquus assemblage, although it rarely exceeded two per cent.

Punctatisporites globulosus sp. nov. differs from other species of Punctatisporites by
its lines of weakness along which the specimens tend to rupture.

Punctatisporites ellipticus sp. nov.

Plate 105, fig. 1

Diagnosis. Trilete spores; distinctly elliptical in outline. Exine about one micron thick;
ornamented with minute but distinct, equidistant punctae which, as seen on the outline,
could also be considered coni. Trilete mark distinct; extends from one-half to two-
thirds radius; lips thin and slightly upturned. Size range (fifteen specimens) 55-65 p.

Holotype. Plate 105, fig. 1 ; 60x43 p; slide LKC-2 (12-13) 1, 33 113-6.

Remarks. This species is similar to P. globulosus sp. nov. but differs in its more elliptical
outline and trilete mark. It was encountered most frequently in Punctatisporites obliquus
assemblage zones.

Infraturma apiculati (Benn. and Kidst.) R. Pot. 1956
Subinfraturma granulati Dyb. and Jach. 1957
Genus granisporites Dyb. and Jach. 1957

Granisporites medius Dyb. and Jach.

Plate 105, figs. 2-3, 5

Remarks. This species is very similar to those ascribed to Granasporites Alpern,
the essential difference being the presence of trilete mark in Granisporites medius.
Occasionally, specimens were observed in tetrads (e.g. PI. 105, fig. 5).

0

SCALE in miles

20

“1 * 1

10 25

V

i

\

1

9

i

ib

i

\

' Y

\

I

I

4

/

t

I

r

i\

.1

■i \

I

\

1

VENANGO

COUNTY

CLARION

JEFFERSON

COUNTY

COUNTY

CLEARFIELD

LAWRENCE

COUNTY

COUNTY

ARMSTRONG

COUNTY

BEAVER

COUNTY

TEXT-RG. 1 . Geographic distribution of spore and pollen assemblages in the Lower Kittanning coal.
Faunal facies boundaries from Williams (1959) and Williams and Keith (1963).

SCALE in miles

Stole College

, . f ^'s

■- •t

., , ;.. • ; ' V VIM'* ■''■■•

' .1, ■ aW M'yf>

is^-' ^

t ' ' i f' \ * ' *. i ' ' - ^ ^ i '\

.1 .*

■OWt13V

YTMUO:>

•;.■ -ifhr

\ .... Q-fA

4^

. HO':?:-A.i::^

noiic-Dd"

YlidUOD

I X

■ >v a

. ; .. V^-,-.; !.■

o '.mm

■A:

\r

■Ni

.'5)r!!nncsfj!-.'i (5 ' v'

' ,|..rt;:^.A-iiOv:a’!y^ probaa};/\

..,^i) v; ;ut.('ud

ipo of e-i£i:V iutlirc
^„:i05*i>s coitepsc. iisua'By. trfK'htr
• ' ‘ .- L.r^;i>kf butUhg..:;^iai- exteodig%

'.,„■ ■ ■■ '"P"

,,--’'"-v:V-,Rfe ' ' '''va--v,; ■■ ■

■^ '

tc,.

i' - - ’' . - '

-YTl/iUCO

i fv F;viR';'.^ « - .,'S«e mkiop '‘'.hi#:- '

fIpioc!»1!i'^ p' ' /

‘ ’5. Y<^5?nV'-; ill'll ^ cU?ptic;'

■'■*^r; ■;•■•.■ ■■> >^PK-Ut.fh{*^M0 t'fMk • •■•

\ . , - ■ ,<I ,v..w..»i 'I «

■ ,; S956

to jtoliLidiY?aa'iarfqi«S'as?^’^ "p ■>'*•'*'■

■f' ■. iitcu. '

'■; ti ’’ti:>, , Gl >Si; >-'
■p’k. >a (Jrapki‘or{fo :oik...

,m. :'\: :

. 1^ Y

0 ’'’’Ik

I A’; .'■#yY • ' ' < . ' i-’^Vit'

D. HABIB; SPORE AND POLLEN ASSEMBLAGES

635

Subinfraturma verrucati Dyb. and Jach. 1957
Genus converrucosisporites Pot. and Kr. 1954
Convernicosisporites pseudovalvus sp. nov.

Plate 105, figs. 4, 7

Diagnosis. Trilete spores; triangular in polar view, with slightly convex to straight inter-
radial margins and subangular to subrounded radial corners. Exine covered with rounded
low-lying (approx. 2 /x high) verrucae which give the margin a roughened appearance;
verrucae all about the same size and evenly distributed, tending, however, to become
more numerous and slightly larger at radial corners; occasionally verrucae tend to align
themselves concentrically. Trilete mark distinct, extends almost completely to radial
corners; lips thin and only very slightly raised. Size range (five specimens) 38-50 p.

Holotype. (Plate 105, fig. 4; 38 X 38 p; slide LKC-8 (36-37), wax mount 33.

Remarks. Confined to the Lycospora-Guthoerlisporites erectus assemblage zones at
Stations 8 and 11. Converrucosisporites pseudovalvus sp. nov. is distinguished from
related species by the presence of larger and more numerous verrucae at the radial
corners.

Genus schopfites Kosanke 1950
Schopfites grossus sp. nov.

Plate 105, fig. 6

Diagnosis. Trilete spores; outline in polar view usually elliptical, but may be near
circular. Exine covered subequatorially and distally by short (to 3 p) but very wide
(up to 10 p) verrucate processes; it is free of this ornamentation only in a small area of
the proximal face, near the trilete mark; wall distinct, 3-6 p thick. Trilete mark extends
over one-half radius; usually closed but may be open, producing an irregular gap.
Size range (fifteen specimens) 84-116 p.

Holotype. Plate 105, fig. 6; 93 x 70 /x; slide LKC-10 (7-8) 2, 34-6 1 13-2.

Remarks. Although this species was rarely counted, it was observed more regularly in
zones of the Thymospora pseudothiesseni assemblage.

Genus verrucosisporites Ibr. emend. Smith et al. 1964
Verrucosisporites compactus sp. nov.

Plate 105, fig. 8

Diagnosis. Trilete spores; elliptical in both polar and equatorial views. Trilete mark
extends from between one-half to two-thirds radius, frequently torn open to form an
irregular gap; in some specimens, when not open, obscured by ornamentation; lips very
thin and only slightly turned up. Exine covered with relatively large (to 3 pin width) but
low, rounded verrucae with only slightly flattened apices; verrucae closely distributed
throughout surface, slightly higher than 1 p. Size range (nine specimens) 45-53 p.
Holotype. Plate 105, fig. 8; 48x 37 /x; slide LKC-12 (5-6) 1, 25-1 129-5.

Remarks. This species was never counted. It differs from the other species of Ver-
rucosisporites by its small size and relatively large verrucae.

C 4397

u u

636

PALAEONTOLOGY, VOLUME 9

Subinfraturma nodati Dyb. and Jach. 1957
Genus lophotriletes Naum, ex Pot. and Kr. 1954
Lophotriletes interruptus sp. nov.

Plate 105, fig. 12

Diagnosis. Trilete spores; triangular in polar view with round to semicircular radial
corners and gently to strongly concave inter-radial margins. Exine approximately one
micron thick, covered by relatively large (about 3 /a by 3 /a) pointed to rounded coni.
Trilete mark extends almost to the margin, usually closed and with thin lips. Size range
(ten specimens) 28-38 ij..

Holotype. Plate 105, fig. 12; 28 X 28 slide LKC-4 (18-19) 1, 26-2 116-3.

Remarks. This species was very rarely encountered.

Genus planisporites Knox emend. Pot. and Kr. 1954
Planisporites? sp. A

Plate 105, fig. 9

Description. Trilete spore; roundly triangular, with broadly rounded radial corners and
gently convex inter-radial margins. Exine slightly more than 1 p thick, covered over its
surface by coni 2-5-3 p high and wide at their base; the coni number 42 at the margin.
Trilete mark thin and very straight, extends almost to the margin; very thin but distinct
raised lips attend the sutures. Arcuate thickenings in exine present in proximal hemi-
sphere inter-radial and distinctly convex. Specimen 75x73 ft.

Remarks. The single specimen is very similar to P. granifer, differing only in the presence
of the arcuate thickenings.

EXPLANATION OF PLATE 105

All figures X 500.

Fig. 1. Punctatisporites ellipticus sp. nov., holotype.

Figs. 2-3, 5. Granisporites medius Dyb. and Jach. 2, trilete mark present, although indistinct. 3, Two
sutures torn, the third indistinct. 5, Tetrad.

Figs. 4, 7. Converrucosisporites pseiidovalvus sp. nov. 4, Holotype. 7, Verrucae tend to be concentrically
oriented.

Fig. 6. Schopfites grossus sp. nov., holotype.

Fig. 8. Verriicosisporites compactus sp. nov., holotype.

Fig. 9. Platusporitesl sp. A; note presence of arcuate ridges.

Figs. 10-11. Acanthotriletes dimorphus sp. nov. 10, Holotype; processes mostly blunted. 11, specimen
with mostly spinose ornamentation.

Fig. 12. Lophotriletes interruptus sp. nov., holotype.

Figs. 13-14. Acanthotriletes flexuus sp. nov. 13, holotype.

Fig. 15. Acanthotriletes tenuis sp. nov., holotype.

Figs. 16, 20. Knoxisporites sp. A. 16, Paired ridges distinct.

Figs. 17-18. Spackmanites ellipticus gen. et sp. nov. 17, holotype. 18, Rudimentary trilete mark can be
seen.

Fig. 19. Spackmanites facierugosus (Loose) comb. nov.

STATION

NUMBERS

TEXT-HG. 2. Presence of spore and pollen species in the Lower Kittanning coal.

Palaeontology, Vol. 9

PLATE 105

HABIB, Pennsylvanian miospores

D. HABIB: SPORE AND POLLEN ASSEMBLAGES

637

Genus acanthotriletes Naum, ex Pot. and Kr. 1954
Acanthotriletes dimorphus sp. nov.

Plate 105, figs. 10, 1 1

Diagrams. Trilete spores; roundly triangular, with broadly rounded radial corners and
straight to gently concave (most often) or convex inter-radial margins. Exine 1-2 /x
thick, covered with setaceous projections of essentially two types: (1) spine-like, broad-
based (to 2 jj.) projections which taper rapidly to a sharp point, (2) rod-like projections,
about 1-5 or 2 p wide, with rounded or flattened distal ends. Ornamentation usually
evenly distributed, with up to 25 projections from the outline; ratio of blunted to spinose
projections varies from 1:3 to 3:1. Trilete mark extends more than two-thirds radius;
frequently open with upturned lips, occasionally torn. Size range (twenty specimens)
48-65 p.

Holotype. Plate 105, fig. 10; 58 X 57 p; slide LKC-10 (8-9), wax mount 15.

Remarks. The species is irregularly distributed in the seam although it was found more
often in zones containing the Thymospora pseudothiesseni and Lycospora-Guthoerli-
sporites erectus assemblages.

Acanthotriletes fiexiius sp. nov.

Plate 105, figs. 13, 14

Description: Trilete spores; triangular in outline. Exine less than 1 p thick, covered with
evenly distributed closely packed sharp spines up to 3 |U. high; because of the thinness,
the exine is almost invariably torn and/or folded. Trilete mark distinct, extends over
two-thirds radius, it is attended by very dark thin folds. Size range (eight specimens)
22-32 p.

Holotype. Plate 105, fig. 13; 30x28 p; slide LKC-2 (8-9), 33-6 1 10-4.

Acanthotriletes tenuis sp. nov.

Plate 105, fig. 15

Diagnosis. Trilete spores; roundly triangular in polar view, with rounded radial corners
and straight to convex inter-radial margins. Exine about 1-2 p thick, sparsely covered
by slender, gently tapering spines, never exceeding a height twice their width, and up to
4 p high; about 15 to 20 spines are present at the outline, although in a few specimens
only 8 or 9; the spines tend to group inter-radially at the equator, although they are also
present at the radial corners. Trilete mark distinct, extends at least two-thirds radius;
lips very thin and ill-defined. Size range (fifteen specimens) 33-41 p.

Holotype. Plate 105, fig. 15; 33 x 33 p: slide LKC-12 (8-9) 1, 28 5 126-7.

Remarks. The species was observed from only three localities. It is distinguished mainly
by its very sparse, small, gently tapering spines.

Subinfraturma baculati Dyb. and Jach. 1957
Genus spackmanites gen. nov.

Type species. S. ellipticus sp. nov.

638

PALAEONTOLOGY, VOLUME 9

Diagnosis. Trilete spores; overall outline circular to elliptical. Spore body almost com-
pletely obscured by the presence of long, very closely packed rod-shaped or club-shaped
baculate processes, each of which is near-straight sided and approximately twice as long
as wide; no appreciable thickening of the bacula occurs at the bases; they are evenly
distributed over the entire surface of the exine, so much so that they are almost always
appressed to one another at the bases; in most cases the bacula are partially fused,
either at the base or at the distal ends; bacula usually rounded at the apices, or flat, with
indentations. Trilete mark usually short, or may even be represented onlyby a triangular
gap in the exine; usually can be seen only with careful focusing. Size range 30-70 yi.

Remarks. Spackmanites is characterized by closely spaced, long bacula which are
tightly appressed to each other and even partially fused, the effect of which is to obscure
the outline of the spore body. The genus most closely resembles Vernicosisporites and
Raistrickia; it differs from the former in the lack of verrucae, and from the latter in the
tight packing and partial fusing of the bacula.

Spackmanites ellipticus sp. nov.

Plate 105, figs. 17, 18

Diagnosis. Trilete spores; distinctly elliptical to oval in outline. Exine profusely covered
with rod-like bacula which in some cases fuse with each other near their bases; bacula
typically smoothly truncated or slightly indented at their distal terminations; many
bacula thicken distally, appearing club-shaped in profile. Trilete mark short, with
sutures of unequal lengths; frequently torn to form an irregular triangular gap. Size
range (twenty specimens) 50-67 y.-, exclusive of ornamentation, 34-49 y.

Holotype. Plate 105, fig. 17; 60x51 y (42x36 y)\ slide LKC-12 (63-64), wax mount 39.

Remarks. The truncated and slightly indented bacula, which are partially fused,
distinguish this species. It rarely exceeded one per cent in any zone, being found most
frequently in those containing the Lycospora-Guthoerlisporites erectus assemblage.

Spackmanites facierugosus (Loose) comb. nov.

Plate 105, fig. 19

1934 Reticulatisporites facierugosus Loose, p. 155, pi. 7, fig. 26.

1954 Verrucososporites facierugosus (Loose); Butterworth and Williams, p. 754, pi. 18, fig. 6,
text-fig. 1, 3.

Remarks. An excellent description of the species is given in Butterworth and Williams
(1954, p. 754). Its more circular outline, less densely packed bacula with rounded
apices, separates this species from S. ellipticus. Spackmanites facierugosus is usually
restricted to samples of the Lycospora-Guthoerlisporites erectus assemblage. It was
rarely counted.

Infraturma murornati Pot. and Kr. 1954
Genus knoxisporites Pot. and Kr. 1954

Knoxisporites sp. A
Plate 105, figs. 16, 20

Description. Trilete spores; outline distinctly polygonal. Exine characterized by thick

D. HABIB: SPORE AND POLLEN ASSEMBLAGES 639

ridges, approximately 3 p- wide, paired at polygonal corners but then separate, with
each ridge continuing around the peripheral area; exine otherwise about 2 jx thick,
levigate. Two specimens described, measure about 75 /x; similar to Knoxisporites
cinctus (Lub. and Waltz) Butt, and Will, but differ in size and opening out of
paired ridges. Specimens were recovered from the Lycospora-Giithoerlisporites erectiis
assemblage zone at Station 7.

Turma zonales (Benn. and Kidst.) R. Pot. 1956
Subturma auritotriletes Pot. and Kr. 1954
Infraturma auriculati (Schopf) Pot. and Kr. 1954
Genus triquitrites Wilson and Coe 1940

Triquilrites cheilus sp. nov.

Plate 106, figs. 1, 2

Diagnosis. Trilete spores; triangular in polar view, with gently to markedly concave
inter-radial margins and rounded, extending radial corners. Trilete mark distinct and
extends to just short of the corners; it is bordered by wide, thick, and darkened lips
(to 7 IX wide) which are raised. At the radial extremities round to flattened valvae are
present, which tend to grade in transitionally towards the polar area. Inter-radial
margin bordered by a relatively thin (to 2 p wide), sometimes indistinct, ridge which
thickens radially into the valvae. Exine levigate, with tendency to become irregular;
approximately 2 [x thick. Size range (fifteen specimens) 45-60 [x.

Holotype. Plate 106, fig. 1 ; 53 X 51 /x; slide LKC-6 (1-2) 1, 40-6 124-3.

Remarks. The species is ubiquitously distributed, but rarely exceeds one per cent in
any assemblage zone.

Triquitrites magnifieus sp. nov.

Plate 106, fig. 3

Diagnosis. Trilete spores; isodiametric triangular in polar view. Trilete mark distinct
and almost always open; extends almost to radial margins and attended by relatively
wide (to 5 p) lips that are slightly raised. Radial corners consist of flattened or stubby,
thick valvae. Ornamentation of rather long digitated, somewhat sinuous baculate
processes on an otherwise levigate exine. Projections, usually confined to areas of lips
and valvae, about 10-15 p long and near straight-sided. Size range (seven specimens)
58-78 p.

Holotype. Plate 106, fig. 3; 75x78 p\ slide LKC-13 (18-19) 1, 36-0 107-9.

Remarks. This rare species was never counted. Triquitrites magnifieus sp. nov. differs
from T. eheilus sp. nov. in the possession of baculate processes, and from other
species of Triquitrites in the concentration of processes at the lips and valvae.

Triquitrites sp. A
Plate 106, fig. 4

Description. Trilete spore; triangular in polar view, with convex inter-radial margins and
angular radial corners. Trilete mark extends almost to radial corners; lips well-defined.

640

PALAEONTOLOGY, VOLUME 9

slightly upturned. Radial corners thickened by rectangular to square-shaped valvae
which grade in towards the polar area. Ornamentation of scattered lobed cristae, radially
disposed; cristae low-lying, wider (to 5 /a) than high (to 3 /x), they are confined to distal
and equatorial areas. Specimen 75x68 /x.

Remarks. Triquitrites sp. A resembles T. crassus Kos. but has its ornamentation
restricted to the distal and equatorial areas of the exine.

Genus ahrensisporites Pot. and Kr. 1954
Ahrensisporites vagus sp. nov.

Plate 106, fig. 5

Diagnosis. Trilete spores; triangular in polar view, with rounded radial corners and
gently concave to convex inter-radial margins. Kyrtome best developed in inter-radial
area and can always be traced to radial corners where, however, it is distinguished
occasionally only with oil magnification; inner part raised inter-radially to form distinct
thickening. Trilete mark distinct, extends approximately three-fourths radius; usually
open, with thin lips. Exine smooth to minutely pitted. Size range (twenty specimens)
38-55 /X.

Holotype. Plate 106, fig. 5; 49x41 p; slide LKC-12 (8-9), wax mount LK-10.

Remarks. This species is characterized by the indistinct development of the kyrtome at
the radial corners. It was commonly encountered in samples of the Densosporites
oblatus and Punclatisporites obliqiais assemblages, and rare elsewhere.

Subturma zonotriletes Waltz 1935
Infraturma cingulati Pot. and Kr. 1954
Genus crassispora Bhardwaj 1957
Crassispora kosankei (Pot. and Kr.) Bhardwaj
Plate 106, figs. 9-10, 13

Remarks. Some representatives of Crassispora kosankei observed in this study displayed
a high degree of crassitudo development. The species was encountered in tetrads in
rare instances. It is present at every locality, becoming most abundant in the upper
portions of zones containing the Lycospora-Guthoerlisporites erectus assemblage.

Genus lycospora S. W. and B. 1944
Lycospora contacta sp. nov.

Plate 106, figs. 6, 8

Diagnosis. Trilete spores; roundly triangular to rounded in polar view and hemispherical
in equatorial view. Exine approximately 1 p thick, covered over its surface by very
smalt (probably less than 1 p) rounded grana. Equatorial crassitudinous thickening,
3-5 p wide, grades irregularly towards polar area; indents most markedly at radial
extension of trilete mark. Trilete mark distinct, straight, passes through indented

D. HABIB: SPORE AND POLLEN ASSEMBLAGES

641

crassitudinous thickening to margin; lips thin but distinct, slightly upturned. Size range
(twelve specimens) 30-35 /u,.

Holotype. Plate 106, fig. 6; 32x32 /x; slide LKC-12 (36-37) 1, 31-8 124-4.

Remarks. The species was very rarely counted, but was observed most frequently in
samples of the Thymospora pseudothiesseni and Lycospora-Guthoerlisporites erectus
assemblages. Lycospora contacia sp. nov. is distinguished by the irregular thickening at
the margin which extends towards the polar area at the radial corners.

Genus gravisporites Bhard. 1954
Gravisporites densus sp. nov.

Plate 106, fig. 11

Diagnosis. Trilete spores; subcircular to oval in polar view, but often triangular and
even pentagonal due to folding around the rigid trilete mark. Trilete mark distinct and
rigid because of thick, wide (to 5 /x), elevated ridges. Exine usually folds around trilete
mark; crassitudo distinct, to 4 /x wide, usually folded; exine 1-2 p. thick, irregularly
surfaced with grana of various sizes. Size range (nine specimens) 45-55 /x.

Holotype. Plate 106, fig. 11 ; 53x48 /tx; slide LKC-10 (30-31) 2, 39-6 122-6.

Remarks. The species was only very rarely encountered.

Genus rotaspora Schemel 1950
Rotaspora? perforata sp. nov.

Plate 106, fig. 7

Diagnosis. Trilete spores; roundly triangular in polar view; outline in some specimens
accentuated by protrusion of trilete sutures. Cingulum smooth, but characterized by
relatively large perforations aligned along its periphery, forming a peripheral rim.
Central body slightly more rounded than cingulum, but still retains triangular outline;
appears detached from cingulum, being connected solely by the sutures; ornamentation
minutely punctate or scabrate. Trilete mark prominent; raised and extends to outer
margin of cingulum. Size range (five specimens) 28-35 /x.

Holotype. Plate 106, fig. 7; 28x26 fx; slide LKC-7 (9-10) 1, 18-0 125-0.

Remarks. This species is distinguished by its distinctly perforated cingulum. It is only
provisionally assigned to Rotaspora on the basis of the presence of a rim ( ?) at the outer
margin of the cingulum. The actual occurrence of a rim is not firmly established, as it
may be the result solely of the marginal alignment of perforations. Were this to be the
case, the species would be better assigned to Lycospora.

Genus densosporites Berry emend. S. W. and B.

Densosporites oblatus sp. nov.

Plate 106, figs. 12, 14

Diagnosis. Trilete (?) spores; oblate triangular in polar view, the radial corners rounded

642

PALAEONTOLOGY, VOLUME 9

but always retaining some angularity. Cingulum oecupies from one-third to one-half
the area of the proximal surface of the central body, appears bizonate, becoming
appreciably darker at its inner margin; ornamented by radiating, prostrate spines up
to 5 ju,; spines taper to sharp point; they vary from one or two per specimen to over
fifty; in overmacerated specimens, many spines destroyed, leaving radial pits in cingulum.
Central body smooth to distinctly punctate; punctae, when they occur, restricted to
proximal surface; verrucae, up to 8 wide, on distal surface. No evidence of trilete
mark. Size range (twenty-five specimens) 57-70 fi.

Holotype. Plate 106, fig. 12; 68x 66 p; slide LKC-10 (2-3) 1, 16 113-3.

Remarks. This species is very similar to D. sphaerotriangularis Kos., differing only in
the presence of tapering prostrate spines and slightly larger size.

Densosporites spackmanii sp. nov.

Plate 106, figs. \5 a, b

Diagnosis. Trilete (?) spores; distinctly triangular in polar view, with a very irregular
outline. Cingulum heavily sculptured by irregular flaps which appear imbricate; it is
expressed on the outer margin by deeply cut crenulations, the outer portion of which
forms a mammoid sculpture; it occupies approximately 30% or more of proximal area
of central body. Central body more circular in outline; smooth to punctate on proximal
surface, distal surface covered with 15 to 25 large verrucae. Distinct trilete mark lacking,
although very faint trace can be seen on several specimens. Size range (nine specimens)
48-63 iJi.

Holotype. Plate 106, fig. 15; 58x61 slide LKC-10 (3-4) 3, 30-8 114-9.

Remarks. The heavily sculptured cingulum and distal verrucae distinguish this species.
Only nine specimens were found. Densosporites spackmanii sp. nov. closely resembles

EXPLANATION OF PLATE 106

All figures X 500.

Figs. 1-4. Triquitrites spp. 1-2. T. cheilus sp. nov. 1, Holotype. 3, T. magnificus sp. nov., holotype.
4, T. sp. A. Distal focus; distinct cristae.

Fig. 5. Ahrensisporites vagus sp. nov., holotype; ill-defined radial extensions of kyrtome.

Figs. 6, 8. Lycospora coutacta sp. nov. 6, holotype; polar indentation of crassitudinous margin at
radial corners well-expressed.

Fig. 7. Rotaspora? perforata sp. nov., holotype.

Figs. 9-10, 13. Crassispora kosankei (Pot. and Kr.) Bharad. Specimens possess very well-developed
crassitudinous thickenings. 9, Polar view; triangular gap. 10, Equatorial view; well-developed cras-
situdo. 13, Tetrad; area of contact in each specimen, upper specimen in focus.

Fig. 1 1 . Gravisporites densiis sp. nov. holotype.

Figs. 12, 14. Densosporites oblatus sp. nov. 12, holotype; prostrate radiating spines and distal verrucae.
14, Verrucae slightly smaller.

Fig. 15. Densosporites spackmanii sp. nov., holotype; a. Proximal focus, structured cingulum, and
proximally punctate spore body, b. Distal focus, large verrucae.

Fig. 16. Cristatisporites verrucosus (Dyb. and Jach.) comb. nov.

Fig. 17. Monilospora sp. A; trilete very faint but extends almost to outer margin of spore.

Fig. 18. Savitrisporites bluntus sp. nov., holotype.

Fig. 19. Thymospora concentrica sp. nov., holotype.

Palaeontology, Vol. 9

PLATE 106

HABIB, Pennsylvanian miospores

D. HABIB; SPORE AND POLLEN ASSEMBLAGES 643

species of Asperispora Stapl. and Janson. It is provisionally placed in Densosporites,
however, because of the distinct verrucae on the distal surface and lack of distinctive
sutural ridges.

Genus cristatisporites Pot. and Kr. 1954
Cristatisporites verrucosus (Dyb. and Jach.) comb. nov.

Plate 106, fig. 16

1957 Densosporites verrucosus Dybova and Jachowicz, p. 166, pi. 50, figs. 1-4.

Remarks. The presence of radially aligned verrucae which are fused at their bases
(Potonie and Kremp ’s ‘ kammen ’) warrants the inclusion of this species in Cristatisporites.

Genus monilospora Hacq. and Barss emend. Staplin 1960
Monilospora sp. A
Plate 106, fig. 17

Description. Trilete spore; roundly triangular in polar view, with convex inter-radial
margins and roundly angular radial corners. Capsula, where it projects beyond central
body, from 13 to 19 p wide; it is characterized by a strongly crenulated outer margin,
giving the specimen a rather serrated overall outline; it is very thin (less than 1 p) where
it encloses the proximal and distal hemispheres of the central body. Central body
roundly triangular, with gently concave inter-radial margins; smooth, and approximately
as thick (approx. 2 p) as the inner margin of the free portion of the capsula. Trilete mark
indistinct, but easily seen with careful focusing; extends almost to outer margin of
capsula. Specimen 66 x 56 p, central body 39 X 38 p.

Remarks. Single specimen from Station 1.

Genus savitrisporites Bhardwaj 1956

Remarks. Bhardwaj ’s diagnosis and schematic illustration (1956, pp. 127-8, text-fig.
3a, b) of Savitrisporites distinguish it as a patellate spore genus. It differs from other
patellate genera described here in its slightly thickened angles (Bhardwaj, p. 127).

Savitrisporites bhmtus sp. nov.

Plate 106, fig. 18.

Diagnosis. Trilete spores; isodiametric triangular in polar view, with slightly thickened,
truncated or rectangular radial corners and straight to gently concave or convex inter-
radial margins. Patella 3-6 p wide at equator, characterized by a gently crenulated
equatorial margin which extends on to the distal surface to form a reticulum in polar
view with wide (approx. 3 /x) but ill-defined muri and smaller lacunae. Proximal surface
smooth. Trilete mark distinct, extends almost to inner margin of patella; usually open,
with thin, raised lips. Size range (fifteen specimens) 30-38 p.

Holotype. Plate 106, fig. 18; 37x 35 fi; slide LKC-12 (7-8) 1, 24-2 118-9.

Remarks. This species is distinguished by its blunted isodiametric shape and small size.
It was encountered most frequently in zones containing Thymospora pseudothiesseni and
Punctatisporites obliquus assemblages.

644

PALAEONTOLOGY, VOLUME 9

Turma monoletes Ibr. 1933
Subturma azonomonoletes Luber 1935
Infraturma sculptatomonoleti Dyb. and Jach. 1957
Genus tuberculatosporites Imgrund 1952
Tuberculatosporites spinoplicatus sp. nov.

Plate 107, figs. 1-3

Diagnosis. Monolete spores; roundly elliptical to bean-shaped in outline. Exine 1-2 /x
thick, somewhat loosely covered with flimsy, usually curved, spinose projections which
are best seen at the outline; from 10 to 20 of the processes present at the outline; each
process characterized by from 2 to 5 small, radiating folds at its base; processes 2-10 fx
long and vary in side view from almost straight-sided with a length of more than twice
its width, to relatively short with tapering sides. Monolete mark distinct, almost always
open, but may be folded along its entire length; extends almost the entire length of the
spore body; lips thin, slightly raised. Size range (twelve specimens) 106-126 ju.

Holotype. Plate 107, fig. 1; 119x81 slide LKC-1 (6-7), wax mount 31.

Remarks. The presence of radiating small folds at the base of spinose protrusions makes
this species unique. Specimens corresponding to this species were observed also in
samples from the Cherokee coals of southeastern Kansas.

Genus thymospora Wilson and Venkatachala 1964
Thymospora concentrica sp. nov.

Plate 106, fig. 19

Diagnosis. Monolete spores; elliptical outline. Exine relatively thick, over 2 p; orna-
mented throughout by thick, low-lying (approx. 3 p high) verrucae which commonly
fuse into sinuous, elongate vermiculae; in some specimens, the ornamentation is near-
concentric. Monolete mark over two-thirds length of spore body; it is usually obscured
by ornamentation, but can be observed with careful focusing. Size range (ten specimens)
60-90 p.

Holotype. Plate 106, fig. 19; 61 x43 slide LKC-8 (24-25) 1, 29-9 117-1.

Remarks. This species is distinguished by its coarse ornamentation and large size.

incertae sedis
Spore sp. A

Plate 107, fig. 4

Deseription. Trilete spore; oval in outline. Crassitudinous thickening at periphery, up to
7 p wide. Exine otherwise approximately 1-2 p thick, covered with very thin, curving,
radially branching cristae, which are low-lying (approx. 1 p high). Very short, irregular
trilete mark is present, though indistinct; it is accentuated by a darkened contact area.
A relatively large lenticular fold trends the length of the specimen. Size 58 x48 p.

Remarks. The presence of a crassitudinous thickening relates this specimen to Crassispora.

D. HABIB: SPORE AND POLLEN ASSEMBLAGES

645

Spore sp. B
Plate 107, fig. 5 a, b

Description. Trilete spore; roundly elliptical in polar view. Exine approximately 1 fj.
thick, covered with sparsely distributed clusters of small hemispherical grana, each
cluster of 15 to 25 grana. Triangular-shaped opening, with flaps folded back, indicating
a torn trilete mark. Size 76x66 p.

Remarks. Specimen is characterized by its clustered grana, and is otherwise very similar
to species of Granisporites.

Spore sp. C
Plate 107, fig. 1 a, b

Description. Trilete spore; triangular in polar view, with near straight-sided inter-
radial margins and sharply rounded radial corners. Exine approximately 2-5 p thick,
and characterized by an irregularly serrated equatorial outline; surface irregular but
apparently without a distinct ornamentation pattern. On the distal surface, three deep
but thin grooves approximately 23 p long form an incomplete triangle. Width of
specimen external to grooves about 12 p. Trilete mark distinct, slightly wavy, extends
almost to the radial margins; lips developed but not prominent. Size 48x41 p.

Remarks. Specimen appears to be unique in its possession of oriented grooves on the
distal surface.

Spore sp. D

Plate 107, fig. 6

Description. Trilete spore; roundly triangular, with convex inter-radial margins and
broadly angular radial corners. Central body roundly triangular, generally smooth and
very thin, probably, less than 1 p \ eight to ten large (to 5 p wide) papillate-type thicken-
ings aggregated in proximal polar area. Covering entire central body, except for a torn
gap on proximal side, is a capsulate (?) structure; it is covered over its entire surface
with short, tapering spines, not over 3 p high; spines broad-based. Trilete mark extends,
on the outer structure, almost to the outer margin of the specimen. Size 51x51 n,
central body 35x35 p.

Remarks. The presence of a capsulate-like thickening relates this specimen closely to
Monilospora.

Anteturma pollenites R. Pot. 1931
Turma saccites Erdtman 1947
Subturma monosaccites (Chitaley) Pot. and Kr. 1954
Infraturma triletesacciti Leschik 1955
Subinfraturma intrornati Butt, and Will. 1958
Genus guthoerlisporites Bhardwaj 1954
Giithoerlisporites erectus sp. nov.

Plate 107, fig. 8

Diagnosis. Monosaccate, trilete pollen grains; overall outline oval to elliptical in
proximo-distal and oblique orientations. Central body circular to subcircular, probably

646

PALAEONTOLOGY, VOLUME 9

less than 1 thick, with a smooth surface; numerous slender folds concentrically
disposed about periphery; outline vague. Saccus covers most of proximal surface of
central body, thereby leaving only a somewhat indistinct outline of the latter structure;
externally levigate and internally moderately reticulate; a darkening of the peripheral
region occurs on most specimens. Trilete mark confined to area of central body, thin
but straight and distinct; it extends from about one-half the distance to the margin to
almost touching it; folds commonly continue the trace of the sutures, to the outline of
the central body; lips very thin. Size range (thirty specimens) 60-89 /x; central body
38-43 ii.

Holotype. Plate 107, fig. 8; 81 X 69 /x; slide LKC-1 1 (24-25) 2, 21-2 115-2.

Remarks. This species is distinguished from Guthoerlisporites magriificus, the type
species, by its smaller size, and tendency of folds to continue the sutures of the trilete
mark to the wall of the central body.

Guthoerlisporites delicatus (Kos.) comb. nov.

Plate 107, fig. 9

1950 Wilsonia delicata Kosanke 1950, pp. 54-55, pi. 14, fig. 4.

1959 Wilsonites delicatus (Kos.); Kosanke, p. 700.

Remarks. Kosanke (1950, p. 54) distinguished Wilsonites partly by its vaguely defined
central body being either completely covered or largely covered by the saccus. His
illustration of the holotype of this species, as well as all the specimens of it encountered
in this study, shows a rather distinct central body. Acknowledging that Wilsonites and
Guthoerlisporites are closely related genera, this species is placed in Guthoerlisporites
on the basis of its distinct central body.

Subinfraturma extrornati Butt, and Will. 1958
Genus endosporites Wils. and Coe 1940

Endosporites globiformis (Ibr.) S. W. and B.

Plate 107, fig. 10

1932 Sporonites globiformis Ibrahim, p. 447, pi. 14, fig. 5.

1944 Endosporites globiformis (Ibr.) Schopf, Wilson, and Bentall, p. 45.

1940 Endosporites ornatus Wilson and Coe, p. 184, fig. 2.

All figures X 500.

EXPLANATION OF PLATE 107

Figs. 1-3. Tuberculatosporites spinoplicatus sp. nov. 1, holotype. 2, specimen with somewhat angular
outline, due to ornamentation; well-developed radiating folds at bases of spines. 3, three distinct
folds radiate from base of a spine.

Figs. 4-7. Unnamed spores. 4, sp. A; contact area present, although ill-defined. 5, sp. B; a. Distal focus;
clusters of grana. b. Proximal focus; triangular gap. 6, sp. C. 7, sp. D; a. Proximal focus; slightly
undulating trilete mark; b. Distal focus, triangularly oriented grooves prominent.

Fig. 8. Guthoerlisporites erectus sp. nov. holotype.

Fig. 9. Guthoerlisporites delicatus (Kos.) comb. nov.

Fig. 10. Endosporites globiformis (Ibr.) S. W. and B.

Figs. 11-12. Endosporites grandicorpus sp. nov. 1 1, holotype. 12, distinct apical papillae.

Palaeontology, Vol. 9

PLATE 107

HABIB, Pennsylvanian miospores

k

£ yn_

D. HABIB: SPORE AND POLLEN ASSEMBLAGES

647

Remarks. Endosporites oniatus is here considered a junior synonym of E. globiformis,
primarily on the basis of Chaloner’s (1958) description of the microspores in the
lycopod cone species Polysporia mirabilis. Although specimens of both species of
Sporae dispersoe could morphologically be distinguished, they invariably occurred in
the same samples, and gradational specimens were observed.

Endosporites grandieorpus sp. nov.

Plate 107, figs. 11, 12

Diagnosis. Monosaccate trilete spores; overall outline roundly triangular in polar view,
with distinctly angular radial corners. Central body circular to sub-circular; it occupies
most of the area of the specimens, extending seven-tenths the radius or more; less
than 1 p. thick; very commonly a number of small, lenticular folds criss-cross its distal
hemisphere. Saccus sharply and densely punctate, showing best at the periphery; a
distinct limbiate thickening occurs at the periphery, approximately 3-4 p. wide, attached
to central body on proximal side. Trilete mark distinct, extends to margin of saccus.
Apical papillae located inter-radially on proximal side, a phenomenon common in
lycopod microspores. Size range (seven specimens) 65-81 p; central body 60-73 p..

Holotype. Plate 107, fig. 11 ; 80x72 p; slide LKC-7 (9-10), wax mount 36.

Remarks. This species is distinguished by its relatively large central body, oblate
triangular outline, and strongly differentiated limbus. It occurs most frequently in
zones containing the Thymospora pseudothiesseni assemblage.

Infraturma vesiculomonoraditi (Pant) Bhardwaj 1956
Genus paleospora gen. nov.

Type species. Paleospora fragila sp. nov.

Diagnosis. Elliptical to subcircular monosaccate grain with a longitudinal slit. Central
body elliptical to circular and occupies an area about one-half or more of the enclosing
saccus; in some cases it carries long, thin, concentric folds. Saccus thin, less distinctive
than central body and very commonly folded or torn. A longitudinal fissure extends
from over one-half to the complete length of the saccus; it forms a line along which
entire specimens are commonly folded. An equatorial enclosing flange is the most
prominent structure, and appears to hold the shape of the flimsy specimens ; it is relatively
wider than the saccus. The single described species ranges upward in size from 150
microns.

Remarks. This genus is distinguished from Potonieisporites by its distinct flange, and
from Speneerisporites by its longitudinal fissure.

Paleospora fragila sp. nov.

Plate 108, figs. 1, 2

Diagnosis. Monosaccate spores or pollen with a straight longitudinal fissure confined
to the area of the saccus; elliptical in outline. Central body more roundly elliptical,
1 p or less thick; it is enclosed by a closely appressing saccus with which it conforms in

648

PALAEONTOLOGY, VOLUME 9

outline. Both saccus and central body appear very flimsy and are almost always folded
in several directions; they both appear to be essentially unornamented. Flange in an
equatorial zonate structure of about the same thickness (approx. 1 jj.) throughout its
width; it is the most prominent and most highly ornamented structure; the ornamenta-
tion consists of very low-lying (approx. 1 ix or less high) irregularly radiating corruga-
tions or cristae, which become more distinctive toward the outer margin. Longitudinal
fissure open in all observed specimens, ranging from a thin slit to a wide gap which
participates in folds which, when they occur, run the entire length of the body. Size
range (twenty-five specimens) 150-195 [x; saccus 105-140 fx; flange width 15-25 /x.

Holotype. Plate 108, fig. 1; 161 X 93 ix- slide LKC-2 (21-22) 2, 36-5 115-2.

Remarks. This species does not appear to be restricted to any particular assemblage
zone.

Genus potonieisporites Bhardwaj 1954
Potonieisporites elegans (Wils. and Kos.) emend.

Plate 108, fig. 3

1944 Florinites elegans Wilson and Kosanke, p. 330, fig. 3.

1964 Potonieisporites elegans (Wils. and Kos.); Wilson and Venkatachala, pp. 67-68.

Emended diagnosis. Monosaccate pollen grains; roundly elliptical in overall outline.
Central body distinct, elliptical to circular in outline, characterized by prominent
crescentic or lenticular folds which commonly transect each other, though on opposite
sides of the central body, at angles approaching ninety degrees; the transverse folds,
always two, appear confined to the distal hemisphere at or near the juncture of central
body and saccus; the longitudinal folds, also two, are present on the proximal (free)
hemisphere of the central body and trend along the maximum dimension of the body;
ornamentation minutely punctate to granulose. Saccus externally levigate and internally
moderately infra-reticulate, attached to central body only on its distal side. Tetrad mark
prominent on proximal (free) side of saccus; it is essentially straight longitudinal, but
occasionally T-shaped with a much reduced third suture, or even symmetrically (Y-shape)
trilete.

EXPLANATION OF PLATE 108

All figures X 500.

Figs. 1-2. Paleospora fragUa gen. et sp. nov. 1, holotype.

Fig. 3. Potonieisporites elegans Wils. and Kos. emend. Monolete mark continued across saccus by
folds, optically superimposed on distally attached central body.

Figs. 4-5. Florinites occiiltus sp. nov. 4, holotype; irregular gap at contact of central body and distal
surface of saccus, and radiating folds. 5a, Distal focus, showing contact of central body and saccus;
saccus torn around thick central body. 5b, Proximal focus, showing rudimentary trilete mark on
saccus.

Figs. 6-8. Vesicaspora wilsonii Schemel. 6, Polar view, showing vague central body. 7, Oblique view;
continuation of saccus across central body illustrated by folds. 8, Equatorial view, showing orienta-
tion of saccus, proximally protruding central body, and tenuitas-like membrane crossing saccus.
distally.

Fig. 9. Vesicaspora saarensis (Bharad.) comb. nov.

Fig. 10. Complexisporites chalonerii sp. nov. Ftolotype.

Figs. 11-12. Pityosporites kittanningensis sp. nov. 11, holotype.

Palaeontology, Vol. 9

PLATE 108

HABIB, Pennsylvanian miospores

D. HABIB: SPORE AND POLLEN ASSEMBLAGES

649

Dimensions. (Twenty-five specimens) 135-215 jn; central body 85-1 10 /x.

Remarks. This species is formally emended because of the additional information
provided by numerous specimens present in the Lower Kittanning seam. These speci-
mens agree in every respect to the species recombined by Wilson and Venkatachala
(1964) and have demonstrated that the suture is confined to the saccus on the proximal
(free of any contact with the central body) side. In a specimen illustrated in Plate 108,
fig. 3, the suture and attendant minor folds pass across the central body and its trans-
verse and longitudinal pairs of folds without disturbing them. This seems possible only
when the suture is superimposed on the central body optically from another surface.
Because no mention of the actual position of the tetrad suture is given in the generic
diagnosis by Bhardwaj (1954, pp. 19-20), this species is retained in Potonieisporites.

This species is almost invariably restricted to the bottom-most samples in the Lower
Kittanning seam.

Infraturma aletesacciti Leschik 1955
Genus florinites S. W. and B. 1944

Florinites occultus sp. nov.

Plate 108, figs. A,5a,b

Diagnosis. Monosaccate pollen grains; roundly elliptical to subcircular outline. Central
body circular to oval, very thick and dark, probably over 2-5 jx thick; an irregularly
spinose or conate ornamentation is distinct at its outline. Saccus externally levigate and
internally coarsely reticulate, reticulation pattern radiating out from area of central
body; attached distally to central body, proximal side free; junction with saccus on
distal side attended by an irregular gap, exposing central body; in some specimens, it is
concentrically torn around the central body, presumably because of the latter structure’s
thickness. A tetrad scar was observed in only one specimen (Plate 108, fig. 5 a, b), where
a short, irregular trilete mark is present on the proximal (free) side of the saccus.
Size range (eight specimens) 60-106 p; central body 35-80 /x.

Holotype. Plate 108, fig. 5 a, b; 91 x76 p; slide LKC-12 (51-52), wax mount 21.

Remarks. This species is distinguished by its thick, dark central body. It occurs most
frequently in zones containing the Lyeospora-Guthoerlisporites erectus assemblage, most
usually in the bottom-most samples.

Genus vesicaspora Schemel emend. Wils. and Venkat. 1963
Vesicaspora wilsouii Schemel emend. Wils. and Venkat. 1963

Plate 108, figs. 6-8

Remarks. Specimens of this species present in the Lower Kittanning coal demonstrate
well the morphological reinterpretation by Wilson and Venkatachala (1963).

Vesicaspora saarensis (Bhard.) comb. nov.

Plate 108, fig. 9

1957 Alisporiles saarensis Bhardwaj, p. 117, pi. 31, figs. 14-15.

Remarks. The species described by Bhardwaj (1957) as Alisporites saarensis differs from
previously described species of Vesicaspora by its slightly darker central body.

■650

PALAEONTOLOGY, VOLUME 9

Subturma disaccites Cookson 1947
Genus complexisporites Jizba 1962
Complexisporites chalonerii sp. nov.

Plate 108, fig. 10

Diagnosis. Striate, capped, bisaccate pollen; roundly elliptical to oval in outline, the
width of either saccus never exceeding that of the central body. Central body smooth,
not accentuated by an equatorial thickening; subcircular to slightly transversely
elongated, outline obscured by attachment folds of sacci. Sacci externally levigate and
internally infrareticulate, distally inclined, attached distally to central body near the
polar area and proximally in a subequatorial position; distinct folds define their distal
attachment to central body. Relatively narrow (approx. 6-8 /u.) germinal furrow lies
parallel to sacci bases. On the proximal side of the central body, a vaguely defined, but
always present, cap is defined by a circular groove; within the cap, from 26 to 35 p in
diameter, is a straight longitudinal fissure, running the length of the cap; it is attended
on either side by from one to three less distinct striae; in some specimens the striae
could be distinguished only with careful focusing. Size range (six specimens) 63-86 p.

Holotype. Plate 108, fig. 10; 66x56 p\ slide LKC-7 (18-19) 1, 22-5 114-3.

Remarks. This species is placed in Complexisporites Jizba on the basis of its striated
proximal cap. It differs from C. polymorphus, the type species, mainly in the structure
of the cap. This species also has sacci which are more distally inclined. It occurs with
C. polymorphus in the Lower Kittanning coal, in the lower samples of the Lycospora-
Guthoerlisporites erectus assemblage.

Genus pityosporites Seward emend. Manum 1960

Remarks. Alete, bisaccate grains with distal orientation of sacci are assigned to
Pityosporites (Sew.) Manum in this study. These grains are similar to Klausipollenites
Jansonius, especially in the crescentic to semicircular outline of the sacci, but can be
distinguished from this genus by their distal attachment of sacci closer to the polar area
of the central body, and narrower germinal furrow.

Pityosporites kittanningensis sp. nov.

Plate 108, figs. 11,12

Diagnosis. Alete, bisaccate pollen; central body almost circular, but appears transversely
elongated due to the mode of saccus attachment; sacci crescentic, free margin circum-
scribing more than half the outline of a circle. Central body indistinct, probably not
more than 1 p thick, characterized by subequatorial, convexly crescentic thickenings at
proximal bases of sacci; surface smooth to minutely punctate, punctae being observed
only with high magnification oil objective. Sacci distally inclined at angles from 20 to 30
degrees, attached proximally to central body in a near equatorial position, and distally
in polar area; in smooth continuity with the central body so that in equatorial view they
help purvey the ‘boomerang’ effect typical of the genus; externally levigate and in-
ternally infrareticulate; no evidence of a thickening at their outer margin. Germinal
furrow relatively thin and only vaguely defined, always present, elongated parallel to

D. HABIB: SPORE AND POLLEN ASSEMBLAGES

651

distal attachment of sacci bases. Size range (ten specimens) 60-78 /x; central body
36-58 jU-.

Holotype. Plate 108, fig. 11; 70x47 slide LKC-12 (51-52) 1, 26-8 116-0.

Remarks. This species is distinguished from closely allied species by its size and less
distinct central body. Almost all specimens were recovered from the lower portions of
the Lower Kittanning seam.

Genus alisporites Daugherty emend. Pot. and Kr. 1954
AUsporites symmetricus sp. nov.

Plate 109, figs. 1-3

Diagnosis. Alete, bisaccate pollen; elliptical to elongate-elliptical in polar and equatorial
views. Central body indistinct; outline discerned in some specimens only with careful
focusing; approximately 1 ft thick, smooth to minutely punctate; greatest diameter
along the greatest dimension of the specimens. Sacci attached proximally in subequa-
torial position and distally nearer the polar area; they appear to be of about the same
thickness as the central body and are moderately infrareticulate without any distinct
radial orientation. Connecting the two sacci distally is a very thin (probably less than 1 fi)
germinal furrow which when observed in equatorial view is always split open, 12-30 ft
wide, oriented parallel to the sacci bases; unornamented and much lighter in colour
than either central body or sacci. The sacci are not pendant, but are isodiametrically
opposed. Size range (eight specimens) 73-91 ft.

Holotype. Plate 109, fig. 1 ; 81 X 58 ft; slide LKC-12 (63-64), wax mount 39.

Remarks. This species is smaller than most species of Alisporites. It was found only in
the lowermost samples of the seam.

Bisaccate sp. A
Plate 109, fig. 4

Description. Fissured, bisaccate pollen grain ; elliptical in outline, with the central body
at least as wide as the sacci which, in turn, are widest at their bases. Central body 71 ft
in diameter, almost circular, although not sharply defined; without appreciably distinct
wall, minutely granulose. Sacci externally levigate and internally moderately reticulate,
attached distally very close to polar area of central body, and proximally in a subequa-
torial position; distal attachment marked by distinct primary folds which are convexly
lenticular and extend the entire distance of the bases. Relatively narrow germinal
furrow, about 7 p, wide, parallel to the primary folds; a narrow but very distinct fold
runs almost its entire longest dimension. On the proximal surface of the central body is
a very distinct longitudinal fissure 23 fi long, bordered by thick raised lips. Size 91 X 71 ft.

Remarks. This specimen is distinguished by its large central body and bordered longi-
tudinal fissure.

Bisaccate sp. B
Plate 109, fig. 5 a, b

Description. Two alete, bisaccate pollen grains; elliptical or dumb-bell shaped in outline,

x X

C 4397

652

PALAEONTOLOGY, VOLUME 9

the central body circular with sacci somewhat hemispherical. Central body very vaguely
defined, probably very thin and smooth. Sacci attached distally by thickened areas near
polar area of central body; attached proximally, although very vaguely, in a sub-
equatorial position; externally levigate and internally incompletely reticulate, the
ornamentation appearing more vermiculate or verrucate ; the effect of the ornamentation
gives the specimens a roughened outline; a darkened peripheral zone occurs at the
outline. Relatively wide but thin furrow between sacci bases on distal side. Specimens
64x 50 (Plate 109, fig. 5 a, b) and 61 x43 /.t.

Remarks. The two specimens are distinguished by their coarse ornamentation and
dumb-bell shape.

Bisaccate sp. C
Plate 109, figs. 6,1 a, b

Description. Two striate, bisaccate pollen grains; elliptical in polar view, with hemi-
spherical sacci wider than the central body. Sacci attached to indistinct central body in
polar area on distal side, and subequatorially on proximal side. On proximal side of
central body, well-defined cap distinguished by circular groove; cap striate, six to
seven distinct striae, a few split longitudinally for part of their length. Central body
roundly elliptical, smooth; sacci externally levigate and internally microreticulate.
Relatively narrow germinal furrow parallels the sacci bases. Specimens 89x62 p
(fig. 6) and 16x60 p (fig. 7).

Remarks. The proximally located striate cap relates these specimens to Complexisporites;
however, the podocarpitidaceous sacci-central body relationship relate them closely to
Striatopodocarpites. Only these two specimens were found, in the lowermost samples
containing the Lycospora-Guthoerlisporites erectus assemblage at Stations 7 and 8.

Bisaccate sp. D
Plate 109, figs, i a, b and 9

Description. Two striate, bisaccate pollen grains; sacci circumscribe an outline greater
than one-half that of a circle, and with a width greater than that of the central body.
Central body characterized by thick (to 8 p wide at outline) sinuous wall ; on its proximal
side, from two to four relatively wide (approx. 3 p) but somewhat less sinuous thicken-
ings are separated by striae; usually one or two thickenings are more conspicuous than
the others. Sacci externally levigate and internally finely reticulate; long, slender,
straight folds radiate out from area of central body, presumably due to thickness of
central body; sacci attached distally and proximally to central body in subequatorial
positions, being probably only very slightly distally inclined. On the distal surface, a

, ^ EXPLANATION OF PLATE 109

All figures X 500.

Figs. 1-3. Alisporites symmetricus sp. nov. 1, holotype. s

Figs. 4-9. Bisaccate spp. 4, Sp. A. 5, Sp. B; a, distal focus; b, proximal focus. 6-7, Sp. C. 8-9, Sp. D.
Fig. 10. Schopfipolleiiites sp. A.

Palaeontology, Vol. 9

PLATE 109

HABIB, Pennsylvanian miospores

D. HABIB: SPORE AND POLLEN ASSEMBLAGES

653

wide unornamented germinal furrow. Specimens, 71 X45 yu. (fig. 8 a, b) and 80x60 /x
(fig- 9).

Remarks. The striate central body and larger sacci relate these specimens closely to
Striatopodocarpites.

Turma plicates Naumova emend. R. Pot. 1960
Subturma praecolpates Pot. and Kr. 1954
Genus schopfipollenites Pot. and Kr. 1954

SchopfipoUeniles sp. A
Plate 109, fig. 10

Description. Monolete, prepollen grain; elongate-elliptical outline. Specimen bordered
by thick flange-like structure approximately 8-10 yu, wide, thickest nearest the polar area,
and thinning rapidly towards the outer margin. Two centrally located thick primary
folds extend longitudinally, each 25 long; located on the distal surface, they taper in
opposite directions, joining with each other at their terminations. Monolete mark
slightly bent, extends over two-thirds length. Exine densely but very minutely punctate,
probably over 2 jj, thick. Specimen 179x 109 ix.

Remarks. This specimen is distinguished by its smaller size, slender elliptical outline,
and thick distal folds.

DESCRIPTION OF SPORE AND POLLEN ASSEMBLAGES

The diagnostic characteristics of the five spore and pollen assemblages occurring in
the Lower Kittanning coal seam are given in the succeeding numbered paragraphs.
Following these are more detailed descriptions and discussions of the assemblages.

1. Densosporites oblatus assemblage. Total of all densospores in samples containing
this assemblage ranges from 15 % to 47 %. The densospore species include Densosporites
oblatus, D. reynoldsburgensis, D. spackmanii, and Cristatisporites verrucosus.

2. Punctatisporites obliquus assemblage. Range of Punctatisporites obJiquus-
Laevigatosporites g/obosus in incremental samples from 19 % to 90 %. Combined
densospore species never exceed 13 % in any sample.

3. Thymospora pseudothiessenii assemblage. Range of Thymospora pseudothiesseni
in incremental samples from 3-5 % to 24 %. Florinites pellucidus always occurs in its
greatest amounts in the assemblage zones.

4. Lycospora-Guthoerlisporites erectus assemblage. Lycospora ranges in incremental
samples from 30 % to 74 %. Guthoerlisporites erectus occurs in every assemblage zone.
Punctatosporites minutus, Granisporites medius, and Crassispora kosankei occur here in
greater amounts. There is the greatest diversity of saccate species in the assemblage
zones.

5. Lycospora assemblage. Lycospora ranges from 40 % to 65 % in incremental
samples. Guthoerlisporites erectus is totally lacking.

Bar histogram profiles of the more common species, with the delineated assemblage
zones, are presented for most of the investigated localities in text-figs. 3-7.

STATION NO. 2 STATION NO.

654

PALAEONTOLOGY, VOLUME 9

□

Lycospora-Guthoerlisporifes e rectus
Assemblage

0 50%

10%

100%

TEXT-FIG. 3. Distribution of spores and pollen in the Lower Kittanning coal at Stations 1 and 2.

The average percentages of species in each assemblage are given in text-fig. 8. Each
of the zones is defined solely on the basis of the spore and pollen assemblage occurring
in it. Because the increase and decrease of species in the profiles are for the most part
gradual, the boundaries of assemblage zones were drawn arbitrarily at the top of the
uppermost sample containing the particular assemblage. Because of the sampling
interval, it was decided to obtain more than one sample for any particular zone, except
where the diagnostic species occurred in numbers well above the required minimum.

D. HABIB: SPORE AND POLLEN ASSEMBLAGES

655

Densosporites oblatus assemblage

This assemblage is characterized by the relatively high percentage of densospores
{Densosporites, Cristatisporites) which range in their respective assemblage zones from

TEXT-FIG. 4. Distribution of spores and pollen in the Lower Kittanning coal at Stations 5 and 6.

20-5 % in Zone III at Station 13 to 41-5 % in Zone IV at Station 10. With one
exception, the lowest occurrence in any sample in a zone is 17 %. Pimctatisporites
obliquus-Laevigatosporites globosus is the most abundant spore species in the assem-
blage, although the densospores are collectively most abundant at approximately
half the localities. Densosporites oblatus sp. nov. accounts for more than half the
number of densospores present. Other species included in this group include
Densosporites reynoldsburgensis, D. spacknianii sp. nov., and Cristatisporites verrucosus
comb. nov. Other species which more frequently occurred in this assemblage include

STATION NO. 8 (TYPE LOCALITY) STATION NO. 7

656

PALAEONTOLOGY, VOLUME 9

n

Lycospora-Guthoerlispo riles e rectus
Assembloge

0 50%

10%

100%

TEXT-FIG. 5. Distribution of spores and pollen in the Lower Kittanning coal at Stations 7 and 8.

STATION NO. 10 STATION NO. 9

Iycospora-Guthoerli$porit9s e rectus
Assembloge

0 50%

10% 100%

TEXT-FIG. 6. Distribution of spores and pollen in the Lower Kittanning coal at Stations 9 and 10.

Lycospora
Assemblage

Anapiculatisporites spinosus, Verrucosisporites sifati, Ahrensisporites vagus sp. nov.,
AcanthotriJetes echiuatoides, Punctatisporites globu/osus sp. nov., and/*, ellipticus sp. nov.

The analytical data on the Densosporites ohlatus assemblage show a relatively small
number of species counted and a small number observed per sample. The species
ranged from as little as 12 to as many as 22, whereas a range of from 20 to 28 was

□

Lycospora-Guthoerlispo rites e rectus
Assemblage

0 50%

100%

TEXT-FIG. 7. Distribution of spores and pollen in the Lower Kittanning coal at Station 12.

observed for the entire population. As many as 38 species were observed in the upper-
most zone at Station 4, but this proved to be an exception. The high incidence of
mineral matter and occurrence of species not found in other zones containing the
Densospohtes oblatus assemblage suggest their having been contributed from outside
the immediate environment.

D. HABIB: SPORE AND POLLEN ASSEMBLAGES

659

60%

50%

40%

307*

Densosporites oblatus Assemblage

Punclatisporites obliquus Assemblage

60% —
50%,—

Thymospora pseudotbiesseni Assemblage

10.

11.

Lycospora Assemblage 12.

TEXT-FIG. 8. Average percentages of species in the

histograms refer ti

Lycospora-Gulhoerlisporites erectus Assemblage

Densosporites oblotus, D. reynoldsburgensis, Crisfotisporites verrucosus
Gronisporites medius

Loevigotosporltes desmoinensis, L. medius, L. vulgaris
Florinites pel lucid us
Thymosporo pseudotbiesseni

Punctotisporites obliquus-Laevigotosporites globosus
Lycospora trigonoreticulata, L. pseudoannuloto
Punctalospo rites minutus

Tuberculofosporites modicus, Laevigatosporites minimus, L. perminutus
Triquitrites exiguus, T. bransonii, T. cheilus
Vesicospora wilsonii

Calamospora multiplicota, C. pseudotriongulara

spore and pollen assemblages. Numbers under
> listed species.

The rather small number of species comprising the Densosporites oblatus assemblage
agrees with the data recorded by Smith ( 1 962, pp. 448-9), who reported a low incidence of
species in those portions of the seam rich in Densosporites. As a result, he concluded
(p. 461) that the vegetation was composed of very few species. It seems reasonable to

660

PALAEONTOLOGY, VOLUME 9

hypothesize an environment represented by relatively few plant species, either forming
dense stands or ‘overproducing’ microspores to a great extent.

Zones containing the Densosporiles oblatus assemblage occur at eight of the fifteen
localities. They always occur in the upper portions of the seam when present, most
frequently occupying the position in contact with the overlying Lower Kittanning shale.
Where the assemblage occupies the uppermost zone, the overlying facies is marine or
restricted-marine.

Punctatisporites obliquus assemblage

This assemblage is similar to the preceding one, but differs markedly in the reduced
occurrence to near-absence of densospores. P. obliquus-L. globosiis is usually most
abundant. At a few localities the lack of densospores is compensated by the large num-
bers of Lycospora and/or small monolete species. P. obliquus-L. globosus ranges from
a minimum of 24-2 % in Zone II at Station 7 to 77-3 % in Zone III at Station 2. Ten
of the thirteen zones containing this assemblage, however, range from 34-5 % to 59 %.
Densospores reach above 5 % in only two of the zones. At Station 7, they reach to 13 %,
a maximum occurrence which is attended by large numbers of Lycospora. Species which
are more frequently encountered in this assemblage include Laevigatosporites perminutus,
L. minimus, Tuberculatosporites modicus, Punctatisporites obesus, P. sphaerorigidus sp.
nov., Endosporites zonalis, Acanthotriletes echinatoides, Ahrensisporites vagus sp. nov.,
and Verrucosisporites sinensis.

The number of species per 200 spores counted ranges from 16 to 28, while the total
number of species observed per sample ranges from 24 to 40. The total number of
species per sample always increased in the lower portions of the assemblage zones.

The Punctatisporites obliquus assemblage occurs at 12 of the 15 localities. It most
usually underlies zones containing the Densosporites oblatus assemblage or occurs as
the uppermost zone of the seam. In either case, it underlies the Lower Kittanning shale
where it contains a marine or restricted-marine facies.

Thymospora pseudothiesseni assemblage

This assemblage is defined by the presence of Thymospora pseudothiesseni, which
ranges from 3-5 % to 19 % in its assemblage zones. Although the percentage is not high,
this species is virtually absent in samples containing other assemblages, reaching in
rare instances, only to as high as 2 %. Concomitant with the increase in T. pseudothies-
seni is the relatively sudden jump in Florinites pellucidus, which ranges to 19-5 %.
At only two localities. Stations 2 and 4, does this species not rise appreciably. Although
it was found in almost every sample of the seam, only seldom did it increase above 2 %
in other assemblages. It is within zones containing the T. pseudothiesseni assemblage
that P. obliquus-L. globosus becomes replaced in the lower portions in being the most
abundant representative. On the average this assemblage is less dominated numerically
by a single species than are the other assemblages. Many of the species frequently en-
countered in lesser proportions in other assemblages consequently are more frequent
here. Vesicaspora wilsonii is found in most zones of the seam, but occurs more frequently
here, as do species of Calamospora, Laevigatosporites, and Triquitrites. Other species
which are regular members of the Thymospora pseudothiesseni assemblage are T.
obscura, Schopfites dimorphus, S. colchesterensis, Cirratriradites annulatus, Triquitrites

D. HABIB: SPORE AND POLLEN ASSEMBLAGES

661

exiguus, Mooreisporites inusitatus, Laevigatosporites vulgaris, L. perminutus, L. nunimus,
and TubercuJatosporites modicus. The latter three species occur in lesser proportions
than in the previously described assemblages. Densospores are present in a few samples.

The number of counted species increases from 19 to 27, while the total number
increases from 22 to 51. Most samples contain between 30 and 35 species.

The relatively larger numbers of Florinites pellucidus, combined with an increase in
number of species per sample and lesser domination by a single species, suggests an
environment such as an open herbaceous marsh, which would contain relatively a
greater number of spores from surrounding plant communities. Chaloner (1958 b)
proposed an ‘upland’ source for Florinites-producing plants, primarily on the results
arrived at by Neves (1958). If a rather constant supply of Florinites pollen is introduced
into the coal swamp from marginal sources, as is evidenced by its low but constant
occurrence in almost all samples of the seam, a swamp environment (open marsh) with-
out significant in situ spore production by the very nature of its inability to mask (i.e.
compare with Lycospora-Guthoerlisporites erectus assemblage) would permit the increase
in expression of this genus. The high number of species may have been brought in
mostly when the environment was covered by the water table, as in floods.

Zones containing this assemblage are present at 12 of the 15 localities. It most
frequently occurs in zones in the middle portion of the seam. The assemblage is best
developed, i.e. highest percentage of T. pseudothiesseni and F. pellucidus, usually where
the seam underlies ancient ‘marine’ sediments.

Lycospora-Guthoerlisporites erectus assemblage

This assemblage is characterized by the high percentage of Lycospora and low but
consistent presence of Guthoerlisporites. Lycospora is represented by L. trigonoreticulata
and L. pseudoannulata which together range over 45 % in twelve of the fourteen zones
the assemblage occurs in. The genus is accompanied in almost every sample of the
assemblage by Guthoerlisporites erectus sp. nov. Although the species does not occur in
appreciable quantity, its near omnipresence serves to distinguish the assemblage. It was
only very rarely found in other assemblage zones, and only in the lowermost portions
which are directly above the Lycospora-Guthoerlisporites erectus. It was not found in
any sample from Station 13, where the Lycospora-Guthoerlisporites erectus assemblage
was not observed. It is in samples of this assemblage that Punctatosporites minutus
becomes most abundant; in some samples, it replaces Lycospora in predominance.
It reaches to 56 % of the population in a sample from Zone I at Station 8.

A higher number of species were found to occur regularly in this assemblage, in-
cluding Granisporites medius, Crassispora kosankei, Endosporites globiformis, species of
Calamospora, Florinites similis, Spackmanites ellipticus gen. et sp. nov., S. facierugosus
comb, nov., Mooreisporites inusitatus, Vestispora levigata, V.fenestrata, and V. profunda.
Other species which are not as widely distributed but still occur most frequently here are
Aculeispores aculeus, Cadiospora magna, Complexisporites polymorphus, C. chalonerii
sp. nov., Alisporites symmetricus sp. nov., Florinites diver siformis, F. occultus sp. nov.,
Potonieisporites elegans emend. Pityosporites kittanningensis sp. nov., Guthoerlisporites
delicatus comb, nov., Wilsonites kosankei, and Apiculatisporis globulus. Of all these
species, G. medius is the most abundant, ranging to 19-5 % in Zone I at Station 6.

This assemblage is characterized by a low number of counted species (16 to 24), but

662

PALAEONTOLOGY, VOLUME 9

in contradistinction to the Densosporites oblatus assemblage, a rather high number of
total species per sample (29 to 38). The low count is obviously due to the great prepon-
derance of lycosporids. Most other species were well below 3 %. That the lycosporid
species were produced by arborescent and presumably anemophilous lycopods may
have masked the production of other smaller trees or herbaceous plants in the com-
munity. The over-all low percentage of gymnospermous pollen, represented, however,
by a number of distinctive genera, suggests either long-distance transportation from out-
side the area of the coal swamp environments, or even in situ deposition during the
very early stages of peat accumulation. Most of these species are confined to the lower-
most samples of the seam.

This assemblage is present at 14 of the 15 studied localities.

Lycospora assemblage

This assemblage is very similar to the last, differing primarily in its total lack of
Guthoerlisporites erectus. Lycospora trigonoreticulata and L. pseudoannulata range from
46 to 59-2 %, with Punctatosporites minutus never exceeding 14 % in any zone. The
saccate gymnospermous species are notably missing, and Endosporites globiformis was
observed in the assemblage only at Station 9. Granisporites medius occurs in larger
numbers than in some other assemblages, but reaches only 3 %. Punctatisporites
obliquus-Laevigatosporites globosus was recorded at approximately 12 %. The
densospores are better represented than in many of the preceding assemblages, extending
to 7 % at Station 9.

The Lycospora assemblage contains a low number of counted species (14 to 22), and
a total number of 25 to 35 per sample. The low number of species counted is similar
to that in the Lycospora-Giithoerlisporites erectus assemblage, which is due here also to
the high occurrence of Lycospora species. The additional low number of species per
sample, however, is in contrast.

The Lycospora assemblage occurs at 3 of the 15 localities and always occurs in the
uppermost zone. It was found only where the seam underlies the non-marine facies of
the overlying shale unit.

DISCUSSION

The most typical sequence in which the spore and pollen assemblages of the Lower
Kittanning seam occur from the bottom to top is: (1) Lycospora-Guthoerlisporites
erectus assemblage; (2) Thymospora pseudothiesseni assemblage; (3) Punctatisporites
obliquus assemblage; (4) Densosporites oblatus assemblage. This sequence may be
terminated at the top by the Punctatisporites obliquus assemblage. This sequence (1-4
or 1-3) is typical of those localities where the marine and restricted-marine facies
lie directly over the seam. It is interrupted, and in one case considerably upset (e.g.
Station 9) at the fresh-water localities. There, the Lycospora assemblage occurs in the
uppermost zone. The vertical variation in abundance of species is less pronounced, and
consequently fewer assemblages are represented. Even those assemblages more typical
of other localities which are present (e.g. Thymospora pseudothiesseni Sind Punctatisporites
obliquus assemblages) here appear to be less well-developed, presumably due in large
part to the preponderance and masking effect of Lycospora.

D. HABIB: SPORE AND POLLEN ASSEMBLAGES

663

CONCLUSIONS

The spatial arrangement of spore and pollen assemblage zones in the Lower Kit-
tanning seam suggests a genetic relationship with the environmental conditions in which
the overlying sediments were deposited (text-fig. 1). Their distribution can be compared
in part with the Recent pollen deposits of southwestern Florida, only on the basis,
however, of their relationship to marine sediments lying directly on buried portions, and
distribution of assemblages according to salinity, water table level, and drainage factors
(see Spackman, Dolsen, and Riegel 1966; Flabib, Riegel, and Spackman 1966).
Although both geo-botanical models cannot be considered completely analogous, they
offer similar evidence for migration of peat environments and their contained spore and
pollen assemblages in response to a transgressing sea.

The salinity factor, resulting from proximity to the sea, seems plausible in attempting
to explain the distribution of assemblage zones in the seam. The arborescent lepidoden-
drids, from which Lycospora was disseminated, were presumably less-brackish paralic
swamp dwellers which migrated towards the margins of the peat accumulating basin in
step with the encroaching more-saline marine waters. Other factors which are also best
manifested in selective zones of a coastal swamp are of equal importance, however.
The position of the water table, relatively higher towards the centre of a basin, may
have had a marked effect on the distribution of the plant communities from which
the Densosporites oblalus and Punctatisporites obliquus assemblages were formed. The
margins of the basin would be better drained and hence better able to support the
retreating lepidodendrids.

The large number of species and more or less sudden jump in frequency of Florinites
pelluddus in the Thymospora pseudothiesseni assemblage, suggest periods of wide-
spread high water conditions in an open marsh environment. A modern day analogue of
this hypothesis can be found in the surface peat deposits of southern Florida, where a
relatively large number of pine pollen is present in an environment represented by
many marsh plants (Spackman et al. 1966, fig. 16). The pollen deposits of this frequently
flooded marsh environment are characterized by a large number of plant species.

As the effect of marine transgression became more intensive, the floral environments
became more competitive, with an overall decrease in number of species. Here, the
number of densospore-producing plants increased, as the water table rose and the area
became more brackish. The apparently anomalous presence of the Densosporites
oblalus assemblage in Zone 11 at Station 9, and increase in number of densospores at
Station 7, may reflect the presence of a tongue of the marine embayment, or more
brackish estuary extending northward. In the modern swamps of the Everglades region
of southern Florida, the salinity tolerant red mangrove (Rhizophora mangle L.) extends
up the banks of the lower reaches of the Shark River into areas containing less-brackish
waters (Spackman and Dolsen 1962).

That the influence of the subsiding basin was little felt in the initial stages of peat
accumulation is attested by the extensive distribution of the Lycospora-Guthoerlisporites
erectus and to a lesser degree Thymospora pseudothiesseni assemblages in the lower zones
of the seam. Factors introduced with the advent of the marine embayment are recorded
as having begun to affect the coal-swamp complex during the time of the Thymospora
pseudothiesseni assemblage, although it would have become intensive in the next higher

664

PALAEONTOLOGY, VOLUME 9

zones. It is quite possible, of course, that the basin was actively subsiding during the
initial stages also, but without presenting evidence into the spore-pollen record.

Immediately prior to, or at the very beginning of, peat formation in the swamp
environment, the area was clothed with a gymnospermous (i.e. Alisporites, Pityosporites,
Complexisporites, Wilsonites, Guthoerlisporites, and Potom'eisporites) and endosporitid-
lycopod forest. The gymnospermous genera may instead represent the allochthonous
elements of extra-swamp (‘highland’) plant communities; restricted occurrence to the
lowermost portions of the seam appears to obviate this alternative, however.

In Great Britain, Smith (1963, 1964) found no apparent relationship between coal
seams containing Densosporites-rich assemblages and the facies of the overlying shales.
He suggested instead, that the sequence of assemblages culminating in one containing
abundant Densosporiles represent a succession of hydrological units in a stable environ-
ment, and that the Deusospohtes-x\c]\ assemblage may have formed partially under
subaerial conditions. This is difficult to envisage for the Lower Kittanning seam, where
in places marine beds rest directly on the coal.

BIOSTRATIGRAPHICAL IMPLICATIONS

The detailed description of spore and pollen assemblages in the coal seams of western
Pennsylvania may well lead to a more meaningful biostratigraphical zonation of the
Pennsylvanian rocks of the eastern United States, on the basis of paleoecological
interpretations. Similarly conducted palynological investigations of several other coal
horizons (e.g. Brookville and Lower Clarion seams by Frederiksen 1961 ; Upper Freeport
and Redstone seams, unpublished work by the author) have demonstrated the occur-
rence of spore and pollen assemblages and/or assemblage sequences unique to each
seam. As a result, two broad categories of assemblages could be established, according
to the position of the seams in marine and non-marine stratigraphical sections.

This type of palynological investigation also brings to light the significance of the
less abundant but nevertheless persistent spore-pollen species, which are of importance
for the determination of more accurate stratigraphical ranges.

Acknowledgements. The author is deeply indebted to Dr. William Spackman, Professor of Paleobotany
at the Pennsylvania State University, for his supervision of this study. He provided all facilities neces-
sary for its completion. Gratitude is also extended to Dr. William G. Chaloner of University College,.
London, for his helpful criticisms, and to Dr. Eugene G. Williams of the Pennsylvania State University,
for pointing out many of the sampled localities.

REFERENCES

ALPERN, B. 1958. Description de quelques microspores du Permo-Carbonifere frangais. Rev. Micro-
paleont. 1, 75-86, 2 pi.

BHARDWAJ, D. c. 1954. Einige neue Sporengattungen des Saarkarbons. Neiies Jahrb. Geol. Paldont.
(11), 512-25, 12 fig.

■ 1956. The spore genera from the Upper Carboniferous of the Saar and their value in Stratigra-

phical Studies. Palaeobotanist, 4, 119-49, 2 pi.

■ 1957. The palynological investigations of the Saar coals, Pt. I, Morphology of Sporae dispersae.

Palaeontograpliica, 101 B, 73-125, 10 pi.

D. HABIB: SPORE AND POLLEN ASSEMBLAGES 665

BUTTERWORTH, M. A. and WILLIAMS, R. w. 1954. Descriptions of nine species of small spores from the
British coal measures. Ann. Mag. Nat. Hist. 7, 753-64, 3 pi.

1958. The small spore floras of coals in the Limestone Coal Group and Upper Limestone

Group of the Lower Carboniferous of Scotland. Trans. R. Soc. Edinb. 63, 353-92, 4 pi.

BUTTS, c. 1906. Kittanning and Rural Valley quadrangles, Pennsylvania. U.S. Geol. Bull. 279.

CHALONER, w. G. 1953. Anewspccics of Lepidostrobus containing unusual spores. Geol. Mag. 90,97-1 10.

1958a. Polysporia mirabilis Newberry, a fossil lycopod cone. J. Paleont. 32, 199-209.

19586. The Carboniferous upland flora. Geol. Mag. 95, 261-2.

1961. Paleo-ecological data from Carboniferous spores. Rec. Adv. Bot. 2, 980-3. (Univ. Toronto

Press. )

CRiDLAND, A. A. and MORRIS, J. E. 1963. Taeiiiopten's, Walchia, and DichophyUinn in the Pennsylvanian
System of Kansas. Univ. Kansas Sci. Bull. 154, 71-85.

DEGENS, E. T., WILLIAMS, E. G., and KEITH, M. s. 1957. Environmental studies of Carboniferous sediments
Pt. I: Geochemical criteria for differentiating marine from fresh-water shales. Bull. Ainer. Assoc.
Petr. Geol. 41, 2427-55.

, 1958, Environmental studies of Carboniferous sediments Pt. II: Application of

geochemical criteria. Ibid. 42, 981-97.

DYBOVA, s. and jachowicz, a. 1957. Microspores of the Upper Silesian coal measures. Warsaw, Inst.
Geol. Prace, 23, 328 pp., 91 pi.

FREDERiKSEN, N. o. 1961. Sporomorphac of the Brookville seam near Brookville, Pennsylvania.
Unpubl. MS. thesis, Pennsylvania State University.

HABIB, D. 1964. Distribution of miospore assemblages in the Lower Kittanning coal (abstr.). Program
Geol. Soc. America Conv., Miami, 1964, p. 81.

RiEGEL, w., and SPACKMAN, w., 1966. Relationship of spore and pollen assemblages in the

Lower Kittanning coal to overlying faunal facies. JL Paleont, 40, 756-9.

HACQUEBARD, p. A. and BARSS, M. s. 1957. A Carbonifcrous spore assemblage, in coal from the South
Nahanni River area. Northwest Territories. Geol. Surv. Canada, Bull. 40, 63 pp., 6 pi.

HORNBAKER, A. L. and HABIB, D. 1962. Palynological correlation of Upper Cherokee coals of Kansas,
Missouri, and Iowa (abstr.). Geol. Soc. America, Spec. Paper 68, 76.

IBRAHIM, A. c. 1932. Sporenformen des Aegirhorizonts des Ruhr-Reviers. Wiirzburg, Triltscli, 46 pp.,
8 pi.

JizBA, K. M. M. 1962. Late Paleozoic bisaccate pollen from the United States midcontinent area. J.
Paleont. 36, 871-87, 4 pi.

KOSANKE, R. M. 1950. Pennsylvanian spores of Illinois and their use in correlation. Illinois State GeoL
Surv. Bull. 74, 128 pp., 16 pi.

1959. Wilsonites, new name for Wilsonia kosanke, Kosanke, 1950. J. Paleont. 33 , 700.

KREMP, G. o. w. 1952. Sporen-Vergesellschaftungen und Mikrofaunen-Horizonte im Ruhrkarbon. Ill
Congres. Advanc. Etud. Stratigr. Carb., Heerlen, 1951 , C. r., 1, 347-57, 1 pi.

LOOSE, F., 1934. Sporenformen aus dem Eloez Bismarck des Ruhrgebietes. Arbeit. Inst. Paldobot. u.
Petrog. Brennsteine, 4, 128-64, 1 pi.

NAUMOVA, s. N. 1939. Spores and pollen of the coals of the U.S.S.R. Intern. Geol. Congr., Moscow
(1937) pp. 353-64.

1953, Spore-pollen complexes of Upper Devonian of the Russian Platform and their significance

for stratigraphy. Trans. Inst. Geol. Sci. Acad. Sci. U.S.S.R., Rel. 143, Geol. Ser. (60), 202 pp., 22 pi.

NEVES, R. 1958. Upper Carboniferous plant spore assemblages from Gastrioceras subcrenatum horizon.
North Staffordshire. Geol. Mag. 95, 19 pp.

pococK, s. A. J. 1961 . Microspores of the genus Murospora Somers from the Mesozoic strata of western
Canada and Australia. J. Paleont. 35, 1231-34.

SCHEMEL, M. p. 1951. Small spores of the Mystic Coal of Iowa. Amer. Midi Nat., 46, 743-50.

1957. Small spore assemblages of mid-Pennsylvanian coals of West Virginia and adjacent areas.

Unpubl. Pli.D. thesis. West Virginia University.

SCHOPF, J. M., WILSON, L. R., and BENTALL, R. 1944. An annotated synopsis of Paleozoic fossil spores and
the definition of generic groups. Illinois State Geol. Surv., Rept. Invest. 91, 73 pp., 3 pi.

SMITH, A. H. V. 1957. The sequence of microspore assemblages associated with the occurrence of cras-
sidurite in coal seams. Geol. Mag. 94, 345-63.

666

PALAEONTOLOGY, VOLUME 9

SMITH, A. H. V. 1962. Paleoecology of Carboniferous peats based on miospores and the petrography of
bituminous coals. Proc. York. geol. Soc. 33, 423-74, 5 pi.

1963. Palaeoecology of Carboniferous peats: Problems of Palaeoclimatology. John Wiley and

Sons, New York, pp. 57-75.

1964. Zur Petrologie und Palynologie der Kohlenfloze des Karbons und ihrer Begleitschichten.

Forstscbr. Geol. Rheinld. u. Westf. 12, 285-302.

SPACKMAN, w., and DOLSEN, c. p. 1962. The characteristics of modern organic sediments and their use
in the identification, description, and interpretation of carbonaceous rocks and rock sequences.
The First Nat. Coastal and Shallow Water Res. Conf, October, 1951. (D. S. Gorsline, ed.)

and RiEGEL, w. 1966. Phytogenic organic sediments and sedimentary environments in the

Everglades-Mangrove complex. Part I: Evidence of a transgressing sea and its effects on environ-
ments of the Shark River area of southwestern Florida. Palaeontographica, 117 B, 135-52.

SCHOLL, D. w. and taft, w. h. 1964. Field guidebook to environments of coal formation in

Southern Florida. Geol. Soc. America, preconvention field trip, November 1964, 67 pp., 57 fig.

STAPLiN, F. L. 1960. Upper Mississippian plant spores from the Golata formation, Alberta, Canada.
Palaeontographica, 107 B, 1-40, 8 pi.

WEBER, J. N., WILLIAMS, E. G., and KEITH, M. L. 1964. Paleocnvironmental significance of carbon isotopic
composition of siderite nodules in some shales of Pennsylvanian age. J. sedim. Petrol. 34, 814-18.

WILLIAMS, E. G. 1959. Aspects of the paleogeography of the coal measures of western Pennsylvania.
Min. Ind. 28, 3-5.

and KEITH, M. L. 1963. Relationship between sulfur in coals and the occurrence of marine roof

beds. Econ. Geol. 58, 720-9.

WILSON, L. R. and coe, e. a. 1940. Descriptions of some unassigned plant microfossils from the Des
Moines Series of Iowa, Amer. Midi. Nat. 23, 182-6.

and kosanke, r. m. 1944. Seven new species of unassigned plant microfossils from the Des

Moines Series of Iowa. Proc. Iowa Acad. Sci. 51, 329-33.

and VENKATACHALA, B. s. 1963. A morphological study and emendation of Vesicaspora Schemel

1951. Okla. Geol. Notes, 23, 142-9.

1964. Potonieisporites elegans (Wilson and Kosanke 1944) Wilson and Venkatachala comb.

nov. Ibid. 24, 67-68, 1 pi.

D. HABIB

Department of Geology and Geography,
Queens College, Flushing,

New York 11367,

U.S.A.

Manuscript received 26 June 1965

A NON-MARINE OSTRACOD FAUNA FROM THE
COAL MEASURES OF DURHAM AND
NORTHUMBERLAND

by JOHN E. POLLARD

Abstract. The occurrence and ostracod fauna of the Hopkins Band, Lower Anthraconaia modiolaris Zone, of
the Lower Coal Measures of Northumberland and Durham are described. In this Band the ostracod Geisina
arcuata is associated with five species of Carbonita, newly recorded, and variants of the non-marine lamelli-
branchs Antliracosia regiilaris, Carbonicola oslancis, Naiadites productus, and Anthraconaia modiolaris. A vertical
sequence of six faunal phases is recognized in this Band and is compared with that previously recorded from
an equivalent stratigraphical horizon in the Yorkshire Coal Measures.

Vertical changes in the number of ostracod species and number of moults present in the Hopkins Band are
recorded and explained in terms of the changing conditions in the environment of deposition. Certain ecological
requirements of the ostracods are suggested.

In the systematic description of the fauna a lectotype for Geisina arcuata and a new species Carbonita
claripunctata are proposed. The species Carbonita lunnilis, C. evelinae, C. pimgens, and C. inflata are briefly
described and their morphology and stratigraphical distribution discussed.

Lectotypes for Carbonita agues, C. evelinae, C. lunnilis, C. inflata, C. pungens, and C. rankiniana are designated
by F. W. Anderson in a postscript.

Despite the profusion of non-marine ostracods at many horizons within the British
Coal Measures, these fossils have received little attention since the latter half of the
nineteenth century. The earliest descriptions of Coal Measures ostracods were by
Bean (1836) from Northumberland and Durham, and Murchison (1839) from North
Staffordshire, but the majority of known species have been described and named by
Jones (1870), and Jones and Kirkby (1867, 1879, 1886n, 18866, and 1890). The Carboni-
ferous ostracod faunas of Scotland were revised by Latham (1932) in the light of the
then current American work (i.e. Ulrich and Bassler 1906, 1908; Harlton 1927, 1928;
Knight 1928; and Roundy 1926), but since then little has been published on British Coal
Measure species, except for occasional lists and records given in the publications of the
Geological Survey.

In the Coal Measures there is a fairly sharp distinction between marine and non-
marine associations of ostracods. Ostracods are relatively rare in marine and Lingula
bands in which only HoUineUa bassleri, Paraparchites, and Cypridina occur with any
frequency. Ramsbottom (1952) recorded a unique fauna from the Cefn Coed Marine
Band of South Wales which included the genera Arnphissiles, Coniigella, Cypridina,
Kirkbya, Knightina and Roundyella, as well as HoUineUa. All these genera are common in
the American Pennsylvanian. Although non-marine assemblages of ostracods are of
much commoner occurrence than marine ones, and occur in greater profusion, the
number of genera represented is smaller. Geisina and Carbonita are the only common
genera, and it is with these that this paper is concerned.

The ostracods that have been recorded by various authors from horizons in the
Northumberland and Durham Coal Measures are shown in text-fig. 1. Apart from
one specimen of HoUineUa bassleri, recorded by Armstrong and Price (1953) from the

(Palaeontology, Vol. 9, Part 4, 1966, pp. 667-97.]

C 4397

Yy

668

PALAEONTOLOGY, VOLUME 9

vcndL-modioIaris or Harvey Marine Band, all the others are species of the non-marine genera
Geisina and Carbonita. Most of these records are of a few specimens of one or two
species found at isolated localities and horizons, but that of the Hopkins Band in the
Lower modiolaris Zone is of a rich fauna which occurs at the same horizon throughout

much of the coalfield. The purpose of this paper is to describe the ostracod fauna of this
Band in detail, and to relate its major vertical changes to the changing conditions in
its depositional environment.

UPPER t

COAL I

MEASURES

To»n End Fon

Q.

I

-E - E~ Otf«rton Eocilherio Bond

MIDDLE

COAL

MEASURES

<

a.

I

V)

CO

q:

%

o

M M P jrho p«

Marin* Band

-L-L- Hfltan Morin* Band

KIrkby’* Marin* Bond
— l?YHOPE FIVE OUAPTEP

RYHOPE LITTLE

Hi^hMain Marin* Bond
HIGH MAIN

META L

FIVE OUAPTEP
MAIN COAL

Blockholl Eu««(htria Bond
MAUDLIN

CO

Q

o

LOWER

COAL

MEASURES

LOW MAIN
BRASS THILL
HUTTON

RULER

Harvey Morin* Band
HA&VEY-^ BEAUMONT

I LLEY

BUSTY

FUhburn Eucfthcrlo Bond
Three Quarler Bond

VICTOR I A

MARSHALL GPEEN
TOW LAW GANISTEP

Bond

OSTRACODS

1. Hollinella basslzri

2. Geisina arcuata

3. Geisina subarcuata

4. Carbonita hunnilis

5. Carbonita ctaripunctata

6. Carbonita evelinae

7. Carbonita inflata

8. Carbonita pungens

9. Carbonita secans

10. Carbonita scatpellus

- ' - common
rare

dubious record

TEXT-FIG. 1. Ostracod species recorded from the Coal Measures of Northumberland and Durham.

THE OCCURRENCE AND FAUNA OF THE HOPKINS BAND

Definition. The Hopkins Band, or Hopkins Shell Bed, was first described by Hopkins
( 1 927) as ‘ The Ostracod Band ’ above the Beaumont or Harvey Seam of the Northumber-
land and Durham Coalfield, and was named after him by Carruthers (1930, p. 69).
A complete definition was given by Hopkins (1928, p. 6):

This Band is unique in the coalfield as wherever it is found it shows a constant threefold division.
(3) An upper division composed of Carbonicola and Naiadites.

J. E. POLLARD: NON-MARINE OSTRACOD FAUNA

669

(2) A middle division composed of Spirorbis and a few Carbonicolae, and

( 1 ) A basal division composed entirely of ostracods.

It is not the presence of the ostracod ' Beyrichia' arciiata (Bean) that makes it distinctive but the
assemblage and order of deposition of the ostracods, annelids and mussels.

A complete account of subsequent reference to this Band was given by Hopkins ( 1960)
in the Lexique Stratigraphique International.

Other ostracod-mussel associations at a corresponding stratigraphical horizon in the
modiolaris Zone were later described from many other British Coalfields; for instance,
from Scotland (Leitch, Absalom, and Henderson 1937), from Lancashire and Yorkshire
(Wright 1931; Eagar 1961), from North Wales (Wood 1937), and from North Stafford-
shire (Melville 1946). The importance of the Hopkins Band is that it was the first band
of its type to be described, is the best known, and is probably the type horizon for the
common Coal Measure ostracod Geisina arcuata (Bean).

Occurrence and stratigraphy. Text-fig. 2 shows the geographical distribution of the
different sediment types associated with the Hopkins Band in the roof strata of the
Harvey-Beaumont Seam, as mapped from borehole information and underground
evidence. The Band can be traced northwards from Ferryhill in County Durham to
Blyth in Northumberland, and eastwards in County Durham from Burnopfield to
Sunderland, covering an area of approximately 450 square miles. Throughout this area
the stratigraphy of the Band is consistently as shown in text-fig. 3, although the thickness
of the fossiliferous strata may vary from 6 in. to 3 ft. 6 in. Magraw, Clarke, and Smith
(1963, p. 162) recorded an association of ostracods and mussels in shales above the
Harvey Seam in Offshore Bore No. 1, four and a half miles east of Blackball Colliery
but the stratigraphy is not typical of the Hopkins Band.

The normal stratigraphic sequence of the Hopkins Band shows that the Harvey-
Beaumont seam is overlain either by a bed of ankeritic mudflake conglomerate between
a quarter and six inches in thickness, or else by a black carbonaceous shale. A black
shale with between 5% and 8% of organic carbon, rich in macroscopic plant remains,
ostracods, and Naiadites succeeds the conglomerate or its lateral equivalent. At the
top of this shale is a densely packed coquina of ostracods, Naiadites, and Spirorbis,
which I have called the Geisina Phase. This is equivalent to the Hopkins (1928) division 1
of the Band. Above the Geisina Phase, the lithology changes abruptly (text-fig. 3) to a
laminated grey shaly-mudstone with less than 1% organic carbon. At this level the
population density of the ostracod fauna declines and the non-marine lamellibranch
Carbonicola appears. This is Hopkins division 2. Clay ironstone nodules and bands
appear in the upper part of the grey mudstone, and the lamellibranchs Anthracosia and
Naiadites replace Carbonicola, as in Hopkins division 3. A fine siltstone succeeds the
mudstone and the fauna then disappears.

Faunal composition. The fossils recorded from the Hopkins Band prior to the present
study have been listed by Hopkins (1960). The present work reveals a more extensive
fauna than was previously known, and a more detailed faunal sequence, as shown in
text-fig. 3. Five species of the ostracod genus Carbonita, namely C. humilis (Jones and
Kirkby), C. eve/inae (Jones), C.pungens (Jones and Kirkby), C. inflata (Jones and Kirkby),
and C. claripunctata sp. nov., are recorded from this band for the first time in association
with Geisina arcuata (Bean).

670

PALAEONTOLOGY, VOLUME 9

Sandstone

Siltstone with
plants

Shales and mudstones
with ostracods ~
Hopkins Band

Black shales with
scattered mussels

Cannel coal or shale
— with ostracods

Seam outcrop

5S0 National Grid

100 km. square NZ

TEXT-FIG. 2. Distribution of the sediment types in the immediate roof-strata of the Beaumont-Harvey
Seam of the Northumberland and Durham Coalfield. (The large black spots and adjacent numbers
refer to the localities mentioned in the text and in Table 5.)

Most of the non-marine lamellibranchs recorded by Hopkins have been found and the
overall mussel fauna is of the osloncis-regidaris type (Calver 1956), which is typical of
horizons below the m.\d-rnodioIaris Marine Band. Naiadites is represented by variants of
the N. triangidaris-produclus group and carinate forms of N. carinatus-flexuosus. The
specimens of Carbonicola are mainly variants of C. oslancis with a low height-length
ratio, but smaller, more orbicular forms close to C. vemista are fairly common. The

J. E. POLLARD: NON-MARINE OSTRACOD LAUNA

671

TYPE SECTION : BEARPARK locality 6

Phases

6

KEY TO FAUNAL PHASE DIAGRAM

Relative Density

Description Representation

RARE

SCATTERED

COMMON

ABUNDANT

I

I

Orientation

MUSSELS

Anthrocosio & CarbonicolQ
Separate valves
Joined valves
O Life position
Naiadites

— » Parallel to bedding
Fragmented

OSTRACODS

Geisina

Complete carapaces
c ,

■f' r Single valves
Corbonito

C> Complete carapaces

Fossils represented

Corbonito pungens

CarbonitQ evelinae
Carbonita cloripunctata

O

Corbonito humilis

Geisina arcuata

Spirorbis

Naiodltes

Curvirimula

Anthraconaia

Carbon icolo
Anthracosia

<^1:::^ Fish

Lepidodendron

bark

Catamites

and

Cordoites

TEXT-FIG. 3. Faunal phases of the Hopkins Band at the type-locality, Bearpark Colliery, County

Durham.

672

PALAEONTOLOGY, VOLUME 9

upper part of the Band is dominated by Anthracosia regularis with rare variants towards
A. cf. aqiiilina and A. cf. ovum. Rare specimens of Anthraconaia also occur and these
are mainly A. modiolaris or forms close to A. robertsoni and A. williamsoni.

Fish remains are confined to the lowest black shale, where scales and spines of
palaeoniscids, crossopterygians and acanthodians, including Megalichthys, Elonichthys,
Strepsodus, Rhadinichthys, and Acrolepis, have been found. The tubicolous annelid
Spirorbis pusiUus Martin occurs throughout the Band, being attached to both mussels
and plant remains at different levels. The plants are represented by lycopods and
Calamites in the black shale at the base, and allochthonous fragments of Calamites,
Cordaites, and rare pteridosperms in the siltstone at the top of the Band.

Hopkins

Divisions

(1927)

(3)

(2)

(1)

TABLE 1. Faunal phases of the Hopkins Band

Faunal phases

6 Scattered Naiadites and plant frag-
ments with attached Spirorbis

5 Anthracosia common, Naiadites
without attached Spirorbis,
Carbonita pnngens and C. hiimilis
rare

4 Carbonicola common, Anthracosia
and Anthraconaia rare, Naiadites
and Spirorbis common, Carbonita
pnngens and C. hnniilis common,
Geisina rare.

3 Geisina, Spirorbis and fragmented
Naiadites abundant

2 Geisina common, Naiadites and
Spirorbis scattered, Carbonita
pnngens common

1 Carbonita pnngens common, fish
scales and plants

Lithology

Siltstone and siliceous mud-
stone

Siliceous mudstone and pale
grey mudstone with ironstone

Dark grey shale and mud-
stone with mussels.

‘’Geisina Phase’; Black shale
laminae

Black carbonaceous shale

Black shale and mudstone
conglomerate

Faunal phases. By detailed analysis of the relative density of the various fossils on
bedding planes throughout the vertical thickness of the Hopkins Band at twelve localities,
it has been possible to recognize a vertical sequence of six faunal phases. Text-fig. 3
shows the faunal distribution and faunal phases at the type-locality of the Band. Each
of these phases, numbered 1 to 6, is characterized by changes in the dominance of the
fossils present, as represented in Table 1.

The ostracod Geisina arcuata has so far only been found in phases 2, 3, and 4 and only
occurs in real abundance in the Geisina Phase, i.e. phase 3. In these phases G. arcuata is
associated with Naiadites, Spirorbis, Carbonita and plant fragments, but rarely occurs
with Carbonicola, and has never been found with Anthracosia (text-fig. 3). The species
Carbonita pungens and C. humilis have a longer vertical range than Geisina arcuata,
occurring from phase 1 to phase 5, and are thus associated at different levels with all
elements of the fauna.

A similar sequence of faunal phases was described by Eagar (1961) from the shales

J. E. POLLARD: NON-MARINE OSTRACOD EAUNA 673

above the Flockton Thin and Flockton Thick Coals near Wakefield in Yorkshire
(Table 2).

Although Geisina is listed in three of Eagar’s phases, he did not record the presence
of associated species of Carbonita. Dr. Eagar has kindly permitted me to examine his
material and as in the Hopkins Band, Carbonita humilis and C. pimgens occur throughout
the sequence. Tn the Flockton succession, Geisina arcuata is again restricted to the more
carbonaceous shale, but in phase 2 occurs in association with abundant specimens of
Anthraconaia, a situation not paralleled in the Hopkins Band. The association of

TABLE 2. Faunal phases of Flockton Shales, Lower A. modiolaris Zone, L. Coal
Measures, Cold Bath Opencast Site, Sheffield, Yorks, {from Eagar 1961)

Fauna

5 Naiadites only, relatively sparse
4 Anthracosia abundant. Relatively
high H/L ratio. Naiadites,
Anthraconaia occasional
3 Carbonicola abundant. Relatively
low H/L ratio. Naiadites,
Anthraconaia occasional. Geis-
ina (Jonesina) very abundant

Lithology

Pale grey mudstone

Grey, typically ferruginous mud-
stone with little carbonaceous
matter.

Dark fine grained richly carbon-
aceous shale marked by
prolific quantities of ostracods,
mainly of Geisina group.

2 Anthraconaia abundant with less Macroscopic pyrite may occur,
common Naiadites. Geisina notably towards the base
(Jonesina) very abundant
1 Naiadites and Geisina (Jonesina)
very abundant
0 Fish scales abundant

abundant specimens of Carbonicola and rare Anthraconaia, with G. arcuata in Eagar’s
phase 3, has a parallel in the lower part of phase 4 of the Hopkins Band in some
localities, where Geisina persists a little above the Geisina Phase. As in the Hopkins
Band Geisina is not known to occur with Anthracosia. Therefore, in both the Hopkins
Band and the Flockton shales Geisina and Carbonita show distinct, but consistent,
differences of range, population density, and faunal association.

DISTRIBUTION AND P AL AEOECOLOG Y OF THE OSTRACOD FAUNA

Vertical distribution. In order to study the vertical distribution of the ostracod fauna of
the Hopkins Band, ostracods were extracted from two series of shale samples obtained
throughout the vertical thickness of the Band at Bearpark Colliery, Co. Durham. The
heights of the samples above the Harvey Seam and numbers of the various ostracods
extracted are shown in Table 3.

There is a distinct vertical change in the species of ostracods present in the Hopkins
Band, and this is independent evidence for the validity of the faunal phases outlined in
the previous section of this paper. Geisina arcuata is restricted vertically, as previously
mentioned, and only occurs abundantly in the Geisina Phase, where it is the dominant
species. Carbonita humilis occurs throughout the vertical sequence, with fairly constant
population density. The small ‘ darwinuloid ’ form C. pungens is most abundant in the

674

PALAEONTOLOGY, VOLUME 9

three inches of conglomerate and shale immediately above the coal, but it persists
vertically, as shown in text-fig. 3. At higher levels, however, it cannot be extracted by the
hydrogen-peroxide sludge technique that was used, because the fragile disarticulated
valves are easily destroyed. A most interesting antipathetic relationship between the
species Carbonita evelinae and C. claripunctata is suggested by this study (see Table 3).
Carbonita evelinae is restricted to the grey shaly mudstone above the Geisina Phase
(3J-4i in.), while C. elaripiinctata is absent from this mudstone, but is present in the
Geisina Phase and in the black shale below.

TABLE 3. Vertical distribution of ostracods in the Hopkins Band, Lower A. modiolaris
Zone, L. Coal Measures, at Bearpark Colliery, County Durham

Numbers indicate the number of specimens extracted

Height above

Species of ostracods

Harvey Seam

Geisina

Carbonita

Carbonita

Carbonita

Carbonita

{i/i inches)

arcuata

humilis

evelinae

pungens

claripunctata

12-14

2

1

9F10

8i-9-|

4

7-J-8i

31

25

61-72

12

3

6-6|

35

6

43-52

9

36

14

T-3 Al_
•^4 ^2

101

38

1

4

{Geisina Phase)

2-3

6

1-2

29

12

2

0-1

6

25

2

Total number of
ostracods extracted

110

199

49

38

8

In the grey mudstone between 1\ and 8J in. above the Harvey Seam, C. evelinae
shows an increase in abundance in both series of samples, and rivals C. humilis as being
the dominant member of the ostracod population (Table 3). A petrographic study of the
enclosing mudstone shows a reduction in the amount of the coarser detrital constituents
and the first appearance of framboidal pyrite, suggesting a decrease in the rate of
sedimentation. Such a change in the rate of sedimentation could account for the
relative abundance of ostracods at this level.

By measuring the height and length of the ostracod carapaces extracted from the
series of samples at Bearpark, the number of moult stages or instars of the various
species present at different levels was obtained. The lowest two inches of the black shale
contain at least two moults of Carbonita pungens and C. humilis respectively, and three
or more of C. claripunctata. In the Geisina Phase, all the ostracods extracted were
adults. The mudstone above the Geisina Phase contains persistently two moults of
C. humilis, while C. evelinae is present generally as the adult and rarely as the preceding
instar as well.

The absence of early moults from the Geisina Phase was confirmed from all localities
studied, not only Bearpark (e.g. the size-frequency distribution of G. arcuata at Epple-
ton, text-fig. 5), and could be due to removal by current sorting, or preservation, or

J. E. POLLARD: NON-MARINE OSTRACOD LAUNA

675

extraction failure. Petrographic examination of the Geisina Phase sediments shows that
at some levels complete and fragmented ostracod carapaces are restricted to lenticular
pockets or laminae within the sediments. This observation, when considered with their
low clay mineral and high silt content, suggests that these sediments may have under-
gone mild winnowing by currents at the time of deposition. A thin band of carbonaceous
shale with ostracods, very similar to the Geisina Phase, occurs above the Eighteen
Inch Seam in the A. modiolaris Zone of the Cumberland Coalfield (Taylor and Calver
1961, p. 9). At least the three largest moults of both Geisina arcuata and Carbonita
hwnilis are randomly distributed throughout the two inches of shale of this band.
This shale is poorer in detrital silt, and richer in clay mineral material, than the Geisina
Phase sediments, and probably accumulated more slowly under less turbulent conditions
of sedimentation. This comparison further supports the suggestion that selective
removal, or fragmentation, of the early instars by mild current action, rather than
preservation failure, accounts for the absence of the young ostracod instars from the
Geisina Phase of the Hopkins Band.

Lateral variation in the Geisina Phase. Samples of the Geisina Phase were examined in
detail from eleven localities in the coalfield, besides Bearpark, to determine if any lateral
changes in the ostracod fauna occurred (Table 4). No major lateral changes were found,
apart from the rare occurrence of Carbonita inflata at two localities near Sunderland.

TABLE 4. Ostracods extracted from the Geisina Band of the Hopkins Band at various

localities in County Durham

Numbers in parentheses refer to localities in text-fig. 2, numbers without parentheses indicate the

number of specimens extracted

Ostracods extracted

Geisina

Carbonita hwnilis

C.

C.

C.

Localities

arcuata

Male

Female

claripunctata evelinae

inflata

Toted

Follonsby (4)

99

3

2

2

106

Pelton (5)

177

36

62

2

6

283

Bearpark (6)

57

15

23

6

2

103

Whitworth (7)

24

5

13

1

6

49

Ferryhill (8)

97

6

7

110

Washington (12)

65

4

7

3

1

80

Flylton (13)

24

5

18

1

3

1

58

Silksworth (14)

122

77

79

3

4

285

Ryhope (15)

120

18

23

4

1

166

Durham Main (18)

34

1

9

1

1

46

Eppleton (19)

212

34

111

1 1

1

5

374

Bowburn (20)

32

6

12

1

51

Total

Approximate %

1063

210

366

35

25

12

1711

of fauna 60 35 3 98

When all the 1711 ostracods extracted from the twelve localities are considered to-
gether the approximate proportions of species present in the Geisina Phase can be
deduced. Approximately 60% of the ostracods are referable to Geisina arcuata, about

676

PALAEONTOLOGY, VOLUME 9

35% to Carbonita humilis, about 3% to C. evelinae and C. claripunctata together, and
about 2% to C. pungens. Carbonita inflata is very rare and only of local occurrence.

As shown in Table 4, sexual dimorphism in Carbonita humilis was recognized at most
localities, although the ratios of the dimorphs extracted varied considerably. Dimorphism
was not recognized in any other species of Carbonita. The greatest population densities
of ostracods in the Geisina Phase were found in the centre of the coalfield, at the localities
Pelton, Follonsby, Eppleton, Silksworth, and Ryhope, where the Phase is between
1 and \ \ in. in thickness. In the localities towards the south and south-western part of
the coalfield, at Durham Main, Bowburn, Ferryhill, and Whitworth Opencast Site, the
shales of the Geisina Phase thin to between J and 4 in. with an associated reduction in
the population density of the ostracod fauna.

Palaeoecology of Geisina and Carbonita. The vertical changes in the ostracod fauna
of the Hopkins Band are believed to be related to changing conditions in the environ-
ment of deposition of the sediments. In these sediments Geisina areuata is associated
with Naiadites, Spirorbis, Carbonita, and plant fragments, and is restricted to the
carbonaceous shale with between 2% and 8% organic carbon. These facts suggest
that in its ecology this ostracod was associated with plant matter, and a pseudoplanktonic
fauna. Eagar (1961) and Heath (1961) reached similar conclusions regarding the
ecology of this species. The genus Carbonita, however, does not show any particular
vertical restriction or geochemical correlation and occurs with both pseudoplanktonic
and benthonic faunas. These facts and the similarity of shell shape may indicate that
in its ecology this genus is similar to the living Cypridopsis (Scott and Summerson 1943;
Kesling 1951).

Several workers have suggested distinct palaeosalinity differences between these two
genera. Jones, Kirby, and Young (1899), Scott and Summerson (1943), and Scott
(1944), all suggested that Carbonita flourished mainly in freshwater lakes or lagoons.
In the Ruhr Coal Measures, however, Geisina is considered to occur only in brackish
or marine shales (Kremp and Grebe 1955; Ernst, Krejci-Graf, and Werner 1958;
Ernst, Michelau, and Tasch 1960; Knauf 1963; and Boger 1964). The gradual pre-
dominance and final survival of Carbonita over Geisina in ostracod bands from the
Lower to the Upper Coal Measures is thought by Calver (1966) to be due to the gradual
reduction of marine influence on the Coal Measure environments.

The solution of the apparent enigma of these two genera occurring together is
probably that salinity was not the sole ecological factor determining their distribution.
Such related factors as types of bottom substrate, available food supply, and depth,
may have been equally important.

In terms of detailed palaeoecology, Geisina areuata appears to have been a browser
on floating or deposited plant matter and to have had a wide salinity tolerance, perhaps
being euryhaline. Some species of Carbonita, namely C. humilis and C. pungens, may
have been ubiquitous bottom crawlers or near bottom swimmers (cf. Cypridopsis,
Kesling 1951), while others show distinct substrate preferences, C. elaripunctata to a
carbon rich substrate and C. evelinae to a carbon poor substrate. Carbonita appears to
have been oligohaline, tolerant of fresh to slightly brackish water, as it has not so far
been found associated with what are normally regarded as marine faunas in the Coal
Measures.

J. E. POLLARD: NON-MARINE OSTRACOD FAUNA

677

Palaeoecology and vertical distribution of the ostracods. This section is an attempt to
reconstruct the changing conditions of the depositional environment from a considera-
tion of changes in the sediments and associated faunas.

After the initial inundation of the swamp that later formed the Harvey-Beaumont
Seam, quiet lagoonal conditions were presumably established over the area of the coal-
field now occupied by the Hopkins Band (text-fig. 2). The sediment coming into this
lagoon contained large and small plant fragments, together with silt and fine clay,
which, after deposition, were compacted to form a carbonaceous shale with between
5% and 8% of organic carbon and up to 10% free silica. Such a lagoonal environment
would have had a richly carbonaceous floor thickly populated by swarms of Carbonita
pungens, scattered C. hwnilis and C. claripunctata, and small groups of Geisina arciiata
clustered around decaying plant fragments.

Petrographic and geochemical examination of the sediments of the Geisina Phase
shows these to be poor in detrital silt and fine clay, but are largely composed of organic
carbon, pyrite, and calcareous shell material. This widespread Phase seems to represent
a coquina of Geisina, Carbonita, Naiadites, Spirorbis, and plant matter, which accumulated
under conditions of reduced inorganic detrital sedimentation, but mild turbulence.

The abrupt change in lithology at the top of the Geisina Phase suggests a change in the
source and type of sediment being supplied to the environment. The faintly laminated
shaly-mudstone that succeeds the Geisina Phase consists largely of clay minerals, less
than 1% organic carbon, and less than 5% free silica. On this muddy substrate, a
benthonic fauna of the lamellibranchs Carbonicola and Anthracosia and the ostracod
Carbonita flourished. Geisina arcuata died out with the failure of the detrital plant
matter, but a new species of Carbonita, C. evelinae, appeared and soon rivalled C.
huniilis as the dominant ostracod.

At a higher level in the mudstone there is a gradual increase in the detrital silt content,
and the mudstone passes into a siltstone. This increase in silt in the water was probably
the major factor responsible for the elimination of the Hopkins Band fauna. The
benthonic lamellibranchs Carbonicola and Anthracosia first became scattered or grouped
(Broadhurst 1964) and then disappeared, to be survived only by Naiadites, which
probably had a pseudoplanktonic mode of life. Carbonita disappeared with the benthonic
lamellibranchs.

The environment of deposition of the sediments and fossils of the Hopkins Band,
described above, represents the basal lagoonal phase of a typical Coal Measure cyclo-
them. These fossiliferous sediments are nowhere more than 6 ft. thick and are succeeded
by 40-50 ft. of deltaic sandstone that forms the main ‘regression phase’ (Payton and
Thomas 1959) of the cyclothem. The ecological changes suggested above are those
normal in such a lagoonal environment undergoing gradual sedimentary infilling with
the approach of the deltaic phase of the cyclothem.

Acknowledgements. I would like to thank Dr. W. Hopkins for help and encouragement during the
course of this work, Dr. F. W. Anderson for advice on ostracod taxonomy, Dr. F. M. Broadhurst
and Dr. J. W. Stanley for offering many helpful suggestions, and Professor K. C. Dunham and his
technical staff for the use of the facilities of the Department of Geology, University of Durham.
I am grateful to the officers of the National Coal Board, Northumberland and Durham Divisions,
for giving access to underground and opencast workings and borehole records; the Assistant Director
and Mr. M. A. Calver of the Geological Survey Office, Leeds, for giving access to material in their
collections; and to the Assistant Keeper and Dr. R. H. Bate of the British Museum (Natural History),

678

PALAEONTOLOGY, VOLUME 9

Dr. W. I. Rolfe of the Hunterian Museum, Glasgow, and Dr. R. M. C. Eagar of the Manchester
Museum, for the loan of ostracod material. The work was carried out during the tenure of a D.S.I.R.
Research Studentship, which is gratefully acknowledged.

SYSTEMATIC DESCRIPTIONS

As well as the new ostracod species proposed in this study, the other species recorded
are briefly redescribed, since this is the first study involving large numbers of these
ostracods from the British Coal Measures. The lectotype of Geisina arcuata is in the
collections of the British Museum (Natural History) and is described here. The
species of Carbonita are compared with lectotype material in the British Museum as
designated by F. W. Anderson in a postscript to this paper. All numbers of specimens
quoted are British Museum (Natural History) registration numbers.

Subclass OSTRACODA Latreille 1806
Order palaeocopida Henningsmoen 1953
Suborder kloedenellocopina Scott 1961
Superfamily kloedenellaceae Ulrich and Bassler 1908
Family geisinidae Sohn 1961
Genus geisina Johnson 1936
Geisina arcuata (Bean) 1836
Text-fig. 4 a-f

1836 Cypris arcuata Bean, p. 377, fig. 55.

18866 Beyrichia arcuata (Bean); Jones and Kirkby, p. 438, pi. 12, figs. 12-14.
non 1889 Beyrichia arcuata (Bean); Jones, p. 381, pi. 17, fig. 7a-c.

1908 Jonesina arcuata (Bean); Ulrich and Bassler, p. 324, pi. 44, figs. 17-19.
non 1927 Jonesina arcuata (Bean); Harlton, p. 205, pi. 32, fig. 6a-c.

1928 Jonesina arcuata (Bean); Knight, pp. 243-6, pi. 31, fig. 6u-c; pi. 33, fig. 6.

1932 Sansabella arcuata (Bean); Latham, p. 366, fig. 12.

1943 Jonesina arcuata (Bean); Scott and Summerson, p. 672, pi. 1, figs. 12-15, 19, 21.

1946 Geisina arcuata (Bean); Cooper, p. 1 10, name only.

1949 Linmoprimitia arcuata (Bean); Kummerow, p. 49, fig. 1.

1952 Jonesina arcuata (Bean); Marple, p. 936, pi. 135, figs. 4-8.

1953 Linmoprimitia arcuata (Bean); Kummerow, p. 15, pi. 1, fig. 7.

1955 Jonesina arcuata arcuata (Bean); Kremp and Grebe, pp. 159-62, pi. 1, figs. 8-10.

1957 Jonesina arcuata (Bean); Vangerow, p. 468, pi. 20, figs. 27-30.

Type specimens. Lectotype, In 43596, specimen 12; complete carapace, adult female, designated here.
Paratype, In 43596, specimens 1-7, 10, 18-23 (Ex Bean Collection, purchased Kirkby 1888).

Dimensions. Lectotype: length, 1-28 mm.; height, 0-81 mm.; width, 0-62 mm.

Type locality. Newcastle area (see discussion).

Type horizon. Coal Measures, probably Hopkins Band, Lower A. rnodiolaris Zone, Lower Coal
Measures (see discussion).

Description. Carapace sub-ovate to sub-rectangular in lateral view. Dorsal margin
straight, hinge line slightly depressed in posterior half; ventral margin convex. Ends
rounded, the anterior more than the posterior. Anterior cardinal angle obtuse, posterior

J. E. POLLARD: NON-MARINE OSTRACOD FAUNA

679

cardinal angle 90° or less. Greatest height median; greatest width median in the male,
posterior of middle in the female, giving a strongly sub-ovate dorsal outline.

Anterior sulcus SI weak, wide, and shallow, or absent. Posterior to SI there is an
indistinct node that fades towards the dorsum. Median sulcus S2 prominent, slightly
anterior of middle, extending from dorsal margin to central area, straight sided, deepest
and widest in the dorso-central area. Posterior to S2, the shell swells dorsally above the
hinge line in a raised area.

Right valve the larger, overlapping the left at the ventral and end margins. On the
exterior of the right valve are two swellings, one on the central anterior border and the
other at the posterior cardinal angle. Anterior swelling is a low pointed bulb which
imparts pointed appearance to the anterior end of the right valve. Posterior swelling is a
rectangular bulb which curls posteriorly and gives the distinctly right-angled appearance
to the postero-dorsal corner. Ventral edge of the right valve is thickened and projects
below the left valve on overlap. Left valve smaller, lacks the swellings of the right, but is
distinctly rimmed.

Hinge line straight, about two-thirds length of dorsal margin, with flexure of right
valve over the left at both cardinal angles. Surface ornament a reticulate pattern of
polygonal pits over the lateral surface of both valves.

Internal features. The internal features could not be seen on the articulated carapace of
the lectotype, so they are described from both complete and fragmentary valves of this
species from the Hopkins Band of Durham and the Flockton Shales of Yorkshire.

Hinge structure is of true Geisina type (Johnson 1936, p. 22). There is a narrow
shallow groove in the dorsal contact margin of the right valve into which the dorsal
edge of the left valve fits. In the anterior half of the hinge of the right valve the groove is
wide, and its lower edge is reflexed ventrally into a sloping flange, so that the anterior
half of the hinge of the left valve is reflexed upwards to overlap the flange of the right
valve, as described by Johnson (text-fig. 4c).

A wide thickened rim runs internally around the contact margin of both valves.
This rim may be the remains of the duplicature, stated by Sohn (1961, p. 182) to be
present in this genus. In the thickened rim of the right valve there is a ridge and groove to
accommodate the overlapped edge of the left valve.

The median sulcus S2 swells and thickens slightly internally to form a ridge that
fades in the central area. Rarely, faint markings suggesting two groups of muscle scars
can be seen on this ridge.

Other features from a study of specimens from the Geisina Phase. Dimorphism can be
clearly recognized in any population of this ostracod. It is of typical kloedenellid type
(Scott and Wainwright 1961), being shown by the inflation of the posterior part of the
carapace of the female, giving distinctly different dorsal outlines to the dimorphs (text-
fig. 4d,f). Johnson (1936) describes differences in lateral shape as well as posterior
inflation between the dimorphs of the type species Geisina gregaria, but no such differences
are discernible in G. arcuata.

In a sample of 180 well-preserved specimens of this species from Eppleton, near
Sunderland, 80% fall within a range of size dimensions: length, 1T0-L30 mm.;
height, 0-60-0-80 mm. The mean sizes are: length, 1T7 mm.; height, 0-72 mm., slightly
smaller than the lectotype. When this range of size dimensions is plotted graphically

680

PALAEONTOLOGY, VOLUME 9

(text-fig. 5), it confirms that only one instar is present. The dimensions given by other
workers for synonymous or allied species of G. arcuata are plotted in text-fig. 5 and
suggest that in many cases they are earlier moults than those represented in the Geisina
Phase.

Discussion. Morphologically this species is fairly close to the American type species
G. gregaria, and although lacking its postero-dorsal spine, has a rectangular bulb in a

TEXT-FIG. 4. Geisina arcuata (Bean). a~d, Lectotype, In 43596, specimen 12. a, exterior view of left valve;
b, exterior view of right valve; c, ventral view of entire carapace; d, dorsal view of entire carapace
showing posterior inflation of the female; e, interior of the right valve showing grooved hinge and
thickened marginal rim, specimen from Flockton Shale of Yorkshire;/, dorsal view of complete male
carapace from Hopkins Band at Eppleton Colliery, County Durham.

similar position. G. arcuata is distinguished from G. subarcuata Jones, the only other
species of this genus recorded from the British Coal Measures, by the smaller size of the
adult, the deeper median sulcus S2, and the presence of the exterior bulbs on the right
valve. In many characters this species closely approaches various of the American
species of Geisina listed by Echolls and Creath (1959), previously described by other
workers.

Because of the confused taxonomic history of this species, as revealed by the synonymy,
it is necessary to discuss the choice of type material and the generic relationships. The
original description by Bean (1836) was brief and in general terms, and the original
figure very poor. The type locality and horizon were likewise vague, being ‘the coal

J. E. POLLARD: NON-MARINE OSTRACOD LAUNA

681

formation near Newcastle’. The specimen proposed here as the lectotype, however,
from the British Museum collection, bears the label 'Cypris arcuata mihi’ in Bean’s
handwriting. This proposed lectotype comes from a piece of shale lithologically and
faunally identical with the Geisina Phase of the Hopkins Band in the Newcastle area.
The historical record {History of Northumberland, 1930, 8, p. 25) shows that the Beau-
mont Seam was being worked about 1836 along the north bank of the Tyne west of
Newcastle, and so it is here suggested that a fragment of the roof shale of this seam,
rich in G. arcuata, was sent by an interested amateur, ‘Mr. Alder’ (Bean 1836, p. 377),
to the well-known Yorkshire naturalist, William Bean. Therefore, it is believed that the

Geisino arcuata
Mean of Eppleton specimens
and inferred instars

TEXT-FIG. 5. Height and length dimensions of Geisina arcuata (Bean) and related species.

specimens of this species described in this paper come from the type horizon and
approximately the type locality of G. arcuata (Bean).

Jones and Kirkby (18866) gave the first detailed description of this species as
Beyrichia arcuata (Bean), and recorded its occurrence from the Coal Measures of the
Midlands, Manchester, Lanark, and County Durham (Ryhope and Claxheugh).
Material from these localities is preserved in the Kirkby Collection in the British
Museum (Natural History), and apart from the Lanarkshire specimens, from Carluke,
they appear to be G. subarcuata, as their localities in the Middle Coal Measures would
suggest. The species was placed in the genus Jonesiua by Ulrich and Bassler (1908) when
they erected this genus on the type species Beyrichia fastigiata Jones and Kirkby without
description of, or reference to, a type specimen. However, when Johnson (1936) erected
the new genus Geisina for a Nebraskan Pennsylvanian form, he borrowed the only
known material of ' Beyrichia' fastigiata identified by Jones and Kirkby, in the British
Museum collections, and distinguished between Geisina and Jonesina. Although
Johnson (1936) described only one species, Geisina gregaria (Ulrich and Bassler),
Jonesina arcuata was referred here by Cooper (1946), but only as a passing mention
without examination of any material.

682

PALAEONTOLOGY, VOLUME 9

The specimens Jonesina ' Beyrichia' fastigiata in the British Museum (numbered 1 1774
and In 32494-6) have been re-examined and compared with G. arcuata. The genus
Jonesina is considered to be valid and cannot be dismissed as suggested by Sohn (1961,
p. 413), and differs from Geisina in the valves being strongly and deeply bisulcate, the
median sulcus S2 being deep and terminating in a pit, the hinge not being depressed in
the posterior half, and the hinge structure only showing overlap at the cardinal angles.
These characters correspond to those given as distinctive of Jonesina in an apparently
little-known revision of the generic diagnosis by Cooper (1941, pp. 55, 56).

Post-war German work on Ruhr Coal Measure ostracods has shown an ignorance of
both American literature and the existence of type material of G. arcuata in British
collections. Kummerow (1949, 1953) proposed and described a new genus, Linmo-
primitia, for G. arcuata, but Kremp and Grebe (1955, p. 161) dismissed the genus as
invalid because of the poor state of preservation of the original material. On the advice
of Dr. B. Kellett, however, Kremp and Grebe (loc. cit.) considered both Limnoprimitia
and Geisina to be synonymous with Jonesina and so have retained the species as
Jonesina arcuata. Their further subdivision of the species into /. arcuata arcuata and J.
arcuata cingulata seems to be due to their confusion of the latter variety with the species
Geisina subarcuata (Jones).

Stratigraphical distribution. Published records of G. arcuata in Britain range from the
Lower Carboniferous to the Upper Coal Measures, but most probably this species is
really restricted to the Lower Coal Measures. The Scottish Lower Carboniferous forms
may well belong to Sansabella, Jonesina, or other Mississippian genera, while the higher
Coal Measure forms are generally close to G. subarcuata. Authors such as Wright
(1931), and Edwards and Stubblefield (1947) are probably correct in referring the range
of the species to the C. communis and Lower A. modiolaris Zones of the Coal Measures,
that is, to horizons below the Clay Cross Marine Band.

A similar stratigraphical range for this species exists in the European coalfields.
Pruvost (1930) stated that it is an excellent zonal indicator of Westphalian A, while
Kremp and Grebe (1955) recorded it as occurring in Namurian C and Westphalian A,
but not above the Katharina (= Clay Cross) Marine Band.

Echolls and Creath (1959) recorded that the range of G. arcuata in the United States
is Lower and Middle Pennsylvanian, from the Morrowan to the Desmoinesian stages.
This is broadly equivalent to the European Westphalian. However, Scott and Sum-
merson (1943) described "Jonesina' arcuata occurring in abundance with species of
Carbonita and Anthraconaia modiolaris in the Hance Eormation of Tennessee. They
correlated this horizon with the Upper Westphalian A stage of the European Coal
Measures.

The general picture therefore seems to suggest that despite its wide distribution from
central and eastern United States, and across Europe as far as the Russian coalfields,
Geisina arcuata (Bean) has a limited stratigraphical range, from Namurian C to the top
of Westphalian A. It is at its greatest abundance in Upper Westphalian A, which is
equivalent to the Lower A. modiolaris Zone of the British Coal Measures in which the
ostracod fauna from the Hopkins Band of County Durham occurs.

E. POLLARD: NON-MARINE OSTRACOD FAUNA

683

Order podocopida Muller 1894
Suborder podocopina Sars 1866
Superfamily cypridacea Baird 1845
Family cyprididae Baird 1845
(Placed here by Cooper 1946, family uncertain Swain 1961)

Genus carbonita [carbonia] Jones 1870

(Preoccupied by Carbonia Robineau-Desvoidy 1863 = Carbonita Strand (1926; 1928, pp. 40, 41)
Type species. Carbonita [Carbonia] agues Jones 1870.

Generic diagnosis. Carapace sub-ovate, ovate-oblong, or elongate. Greatest height
usually in the posterior third; greatest thickness median to posterior. Dorsal margin
slightly to broadly convex; ventral margin straight to slightly convex. Ends unequal,
broadly rounded to acutely pointed. Hinge in the middle third of the dorsal margin,
straight and simple with the left valve usually raised above the right along the hinge line,
closely adpressed or with a narrow groove. Right valve the larger, overlapping the left
on the ventral margin and narrowly overlapping or over-reaching left on the end margins.
Surface smooth to coarsely reticulate. Ornament may consist of longitudinal striae or
concentric reticulation, but may vary with preservation. Muscle scar circular, sunk into
the shell, anterior or antero-ventral of the mid-point. Major muscle scar encloses a
variable pattern of secondary scars. Shell usually thickened internally postero-ventrally
of the muscle scar, in a low ridge that may leave a pronounced furrow on internal moulds.

Carbonita humilis (Jones and Kirkby) 1879

Text-fig. 6

1879
1884
1911
1930
Non 1955
1955
1955

Carbonia fabiilina var. hurnilis Jones and Kirkby, p. 31, pi. 2, figs. 1 1-14.

Carbonia fabulina var. hinnilis Jones and Kirkby, p. 388, pi. 2, fig. 9.

Carbonia fabulina (Jones and Kirkby) in part; Pruvost, pp. 68-70, pi. 2, figs. 1-8.
CythereUa foveolata Wright, p. 49, pi. 1, fig. 2a-b.

Carbonita humilis (Jones and Kirkby); Kremp and Grebe, p. 151, pi. 16, fig. 1.
Whipplella rhenana (Kummerow); Kremp and Grebe, pp. 155-7, pi. 16, figs. 5, 6.
Whipplella cenisa Kremp and Grebe, pp. 152-5, pi. 16, figs. 3, 4.

Lectotype. A figure; Jones and Kirkby 1879, pi. 2, fig. 14. Chosen by Anderson (see Postscript); type
material not found.

Specimens described. Female, lO 2980, Hopkins Band, Bearpark Colliery, Co. Durham. Male, lO 2981,
Hopkins Band, Ryhope Colliery, Co. Durham.

Dimensions. In mm.

Length Height Width

Female lO 2980 0-74 0-48 0-43

Male ID 2981 0-81 0-48 0-41

Description. Female. Carapace tumid, sub-ovate, dorsal margin gently convex, ventral
margin straight to slightly convex. Ends rounded, posterior more obtuse. Greatest
height just posterior of the middle, while the greatest width is well posterior of middle.
Ridge of thickening externally at antero-ventral corner of the left valve. Hinge short,
straight, in central third of dorsal margin, left valve raised above the right along the
hinge with a narrow groove between the valves. Valves unequal, right overlaps the left

z z

C 4397

684

PALAEONTOLOGY, VOLUME 9

a

0.1 r

TEXT-FIG. 6. Carbonita huimlis (Jones and Kirkby). a-d, specimen lO 298 1 , complete male carapace, e-h,
specimen lO 2980, complete female carapace, a, exterior left valve; b, exterior right valve; c, ventral
view; d, dorsal view; e, exterior left valve;/, exterior right valve; g, ventral view; h, dorsal view.

on ventral and both end margins. Ventral overlap wide, end overlap narrow but
distinct. Shell surface distinctly pitted with small polygonal pits. Internally valves are
bordered with a narrow rim and they possess a typical muscle scar and slightly thickened
ridge postero-ventral of the muscle scar.

Male. As female except lateral shape elongate-ovate, maximum height and thickness
median, not posterior of middle. Height to length ratios: male about 60%, female 70%.

J. E. POLLARD: NON-MARINE OSTRACOD FAUNA

685

Discussion. In the original description of Carbonia fabuUna var. humilis Jones and
Kirkby (1879, p. 31) emphasized that this variety differed from C. fabulina (sensu
stricto) in the flatter dorsal margin and the rounded ends being more nearly alike.
These characteristics fit the male of the species as described above, and the original
figures (pi. 2, figs. 11, 12) are of the male. The lectotype, however (pi. 2, fig. 14), is
relatively higher than the others, sub-ovate with a blunter posterior, and although
an internal mould it is probably a female. It thus seems that although only the male
was originally described by Jones and Kirkby (1879), both forms were recognized and
figured.

In the later description of C. fabulina humilis from the Nova Scotia Coal Measures,
Jones and Kirkby (1884) stated that this form is further distinguished from C. fabulina
(sensu stricto) by the thicker shell, the distinct pitting, and stronger amount of overlap.
These differences appear to be persistent, so the ‘var. humilis' of Jones and Kirkby has
been elevated to specific rank by recent workers (Kremp and Grebe 1955; Copeland
1957; Anderson (in press). This present study of the population of C. humilis in the
Hopkins Band confirms that it is a species in its own right.

Although Jones and Kirkby originally recorded C. fabulina var. humilis from both the
Lower and Upper Carboniferous, this distinctly pitted form with strong marginal
overlap seems typical of the Coal Measures, while C. fabulina (sensu stricto) may be
restricted to the Lower Carboniferous. This stratigraphical distinction has been pro-
posed by Dr. F. W. Anderson (pers. comm.), and so the lectotype figure he has chosen
is from the Coal Measures at Pirnie Colliery, Fife.

Cytherella foveolata Wright, described and figured by Wright (1930) is undoubtedly
C. humilis in an inverted orientation (cf. text-fig. 6 with Wright 1930, pi. 1, fig. 2n, b).
This conclusion is confirmed by Dr. Anderson’s examination of Wright’s specimens,
and material from this horizon in the collections of the Geology Department of
Manchester University, examined by me.

The form described and figured by Kremp and Grebe (1955, p. 151, pi. 16, fig. 1) is
not considered synonymous with C. humilis as described here. However, the forms they
describe and figure as Whipplella cenisa and W. rhenana are probably the male and
female dimorphs of C. humilis respectively. The lateral outlines, overlap, hinge, and
dimension ratios are all similar to those of the respective dimorphs described here. The
American genus Whipplella Holland 1934 was erected for forms like C. fabulina (sensu
lato) but with a stronger marginal overlap. The overlap can be so variable in this form
that Whipplella and Carbonita are here considered to be synonymous, in agreement
with Cooper (1946, p. 67).

Stratigraphical distribution. The species C. humilis is common throughout the Coal
Measures, although its acme is probably in the A. modiolaris Zone at about the horizon
of the Hopkins Band. It is the dominant species of Carbonita in the Bottom Three
Quarter Seam ostracod band in the C. communis Zone of Durham (text-fig. 1), and is
abundant in the Upper A. similis-A. pulchra Zone faunas from both Ryhope Colliery,
Durham, and the ‘Foveolata Zone’ of Wright (1931) from the Manchester area. There
are frequent references to the occurrence of this form in the A. phillipsi Zone faunas of
the Ardwick Series of Manchester (Jones and Kirkby 1890) and the Newcastle Beds of
North Staffordshire (Melville 1946).

686

PALAEONTOLOGY, VOLUME 9

Carbonita evelinae (Jones) 1870

Text-fig. 7

1870 Carbonia evelinae Jones, p. 218, pi. 9, fig. 4.

1870 Carbonia agues var. rugnlosa Jones, p. 218, pi. 9, fig. 8.

1870 Carbonia agues var. siibrugidosa Jones, p. 218, pi. 9, fig. 7.

1866 Carbonia wardiana Jones and Kirkby, p. 265, pi. 9, fig. 10a, b.

1943 Hilboldtina evelinae (Jones); Scott and Summerson, p. 870.

1957 Hilboldtina evelinae (Jones); Copeland, p. 28, pi. 1, figs. 5-7.

Lectotype. In 56370, specimen 9; chosen by Anderson (see Postscript).

Specimen described. lO 2985, Hopkins Band, above Geisina Phase, Bearpark Colliery, Co. Durham.
Dimensions. Length, 0-94 mm.; height, 0-42 mm.; width, 0-30 mm.; H/L ratio, 44-7 %.

Description. Carapace elongate, sub-ovate. Dorsal margin slightly arched; ventral
margin straight to slightly convex. Anterior end obtuse, lower than posterior; posterior

TEXT-FIG. 7. Carbonita evelinae (Jones), a-d, specimen lO 2985, complete carapace, a, exterior left
valve; b, exterior right valve; c, ventral view; d, dorsal view.

end more acutely pointed, high and with a steep postero-dorsal slope. Greatest height
median; greatest width posterior of middle giving an elongate-ovate dorsal outline.
Hinge straight, less than half the length, left valve raised along hinge line but no groove
between the valves.

Valves almost equal, right valve narrowly overreaches left on the free ends and over-
laps it along ventral margin. Shell surface ornamented by fine longitudinal striae or
ridges that converge towards the ends. These striae may be weakly developed or absent.
Internally the shell has a circular muscle scar and ridge of thickening posterior to it.
The muscle scar may contain a number of polygonal secondary scars.

Discussion. Dimorphism has not been detected in specimens of this species from the
Hopkins Band.

The lectotype of C. evelinae in the British Museum (Natural History) is badly crushed,
with strong longitudinal striae, a muscle spot, and pattern of cracks very close to those
on Jones’s original figure (Jones 1870, pi. 9, figs. Aa, b). Because of the crushed state
of this specimen, I agree with Dr. F. W. Anderson (in press) that the shape distinction
between C. evelinae, C. agues rugulosa, and C. agues subrugulosa drawn by Jones is not

J. E. POLLARD: NON-MARINE OSTRACOD FAUNA

687

valid. All these three forms are probably one species, C. evelinae, similar in shape to
C. agnes but diflFering in the surface ornament. Likewise, the species C. wardiana Jones
and Kirkby, of which type material is in the Kirkby Collection in the British Museum
(I 1695), is considered synonymous with C. evelinae. The Hopkins Band specimens
described here are a little closer in lateral outline to C. wardiana than to C. evelinae
(sensu stricto) but the variable nature of this character does not warrant specific
distinction.

The genus Hilboldtina Scott and Summerson 1943 was erected for elongate forms of
Carbonita with striated surface ornament from the Pennsylvanian of North America,
This genus is here considered synonymous with Carbonita, since surface ornament is not
a character of generic value, and the other characters mentioned in the original diagnosis
of Hilboldtina are present in most species of Carbonita as described here.

Stratigraphical distribution. The lectotype of this species is from the Upper Coal
Measures, phillipsi Zone, of the South Wales Coalfield. The form C. wardiana was first
recorded from a similar horizon in North Staffordshire. The presence of this species in
the Hopkins Band in the modiolaris Zone of Co. Durham is the first published record of
its occurrence outside the Upper Coal Measures, although similar specimens may be
seen in the Geological Survey collections from the Upper A. modiolaris Zone of the
Cumberland Coalfield.

It seems probable that although this species is predominantly an Upper Coal Measure
form, it occurs rarely throughout the Coal Measures. Similar elongate striated specimens
of Carbonita are reported by Scott and Summerson (1943) from the Pennsylvanian of
Tennessee and by Copeland (1957) from the Upper Carboniferous of Nova Scotia.

Carbonita inflata (Jones and Kirkby) 1879

Text-fig. 8

1879 Carbonia fabulina var. inflata Jones and Kirkby, p. 31, pi. 2, figs. 15-99.

1933 Bythocypris tinnidus Upson, p. 24, pi. 2, fig. 1 \ a-c.

1934 Wbipplella cuneifonnis Holland, p. 344, pi. 25, fig. 5a-c.

1935 Carbonita? tnmida Kellett, p. 160, pi. 16, fig. 9a-c.

1946 Carbonita inflata (Jones and Kirkby); Cooper, p. 66, pi. 8, figs. 40-42.

1957 Carbonita inflata (Jones and Kirkby); Copeland, p. 26, pi. 1, figs. 12-14, pi. 2, figs.
18, 19.

Lectotype. I 1745; chosen by Anderson (see Postscript).

Specimen described. lO 2986, Hopkins Band, Hylton Colliery, Sunderland, Co. Durham.

Dimensions. Length, 0-97 mm.; height, 0-68 mm.; width, 0-63 mm.; ratio H/L, 70 %; width/height,
92-5 %.

Description. Carapace tumid, sub-ovate. Dorsal margin convex; ventral margin straight
to slightly concave. Ends rounded, anterior more obtuse than the posterior, which is
distinctly tumid. Greatest height and width posterior and dimensions almost equal
(W/H ratio: > 90%). Dorsal outline subcuneiform or broad sub-ovate.

Hinge short, in middle third of dorsal margin, depressed in a distinct but shallow
groove. Left valve hardly raised above right along the hinge line. Valves unequal, right
valve overlaps left around free margin. Overlap most pronounced on anterior margin

688

PALAEONTOLOGY, VOLUME 9

and anterior end of ventral margin. Ventral overlap narrow and ventral margin is
flattened along area of overlap. Externally the left valve possesses a low thickened ridge
antero-ventrally. Shell coarsely pitted, and is generally thick with rimmed margins
internally. Muscle scar circular antero-ventral of mid-point.

Discussion. This species is rare in the Hopkins Band and was only found at two localities,
Hylton and Eppleton. No details of moults or dimorphism are known. The specimens
described are very similar to the original description and flgures of Jones and Kirkby,
and to the lectotype in the British Museum (Natural History). The coarsely pitted thick

a

c

d

0.1mm

0.1mm

TEXT-FIG. 8. Carboiiita inflala (Jones and Kirkby). a-d, specimen lO 2986, complete carapace, a, exterior
left valve; b, exterior right valve; c, ventral view; d, dorsal view.

shell, tumid carapace, with posterior width nearly equal to height, and short depressed
hinge, are all sufficiently persistent characters to make the var. inflata of Jones and
Kirkby a species in its own right.

Cooper (1946) examined the type material of the various species named by Upson,
Holland, and Kellett, given in the synonymy, and considered them to be conspecific
with C. inflala Jones and Kirkby. My redescription of this species and comparison with
the descriptions and flgures of the various authors, confirms this synonymy. Many of
the characters by which these authors distinguished their species from C. inflata are in
fact present on perfectly preserved specimens of the latter species but were not included
in the original description. The synonymity of this form with Whipplella cuneifonnis
Holland, the type species of Whipplella, is an additional reason for the rejection of this
genus as suggested in the discussion of C. Immilis.

Stratigraphical distribution. The lectotype comes from the Coal Measures at Pirnie
Colliery in Eifeshire, but the precise level within the Coal Measures is not recorded.

J. E. POLLARD; NON-MARINE OSTRACOD LAUNA

689

Forms similar to those described here from the A. modiolaris Zone are included in C.
fabulina (sensu lato) recorded from the Upper similis-pulchra Zone of Durham, but
they have not been found in the Three Quarter Seam Band in the C. communis Zone.
C. inflata is also a common form in the A. phillipsii Zone faunas of North Staffordshire
and the Manchester area. It seems probable that this species occurs sporadically
throughout the Coal Measures.

Cooper (1946) recorded this species from the Pennsylvanian of Illinois, and the
synonymous forms of other authors come from the Pennsylvanian of Nebraska,
Kansas, Pennsylvania, and West Virginia. This species is also a common one in the
Upper Carboniferous faunas of Nova Scotia described by Copeland (1957).

Carbonita pungens (Jones and Kirkby) 1879

1867 Cythere pungens Jones and Kirkby, p. 222.

1879 Carbonia pungens Jones and Kirkby, p. 37, pi. 3, figs. 21-23.

1884 Cythere (Darwinelld) pungens (Jones and Kirkby); Jones, pp. 319, 325.

1911 Carbonia pungens (Jones and Kirkby); Pruvost, pp. 71-72, pi. 12, figs. 13, 14.

1932 Carbonia pungens (Jones and Kirkby); Latham, p. 386.

1946 Darwinula pungens (Jones and Kirkby); Cooper, p. 78, pi. 10, figs. 39, 40.

1957 Carbonita pungens (Jones and Kirkby); Copeland, p. 26.

Lectotype. I 1731, no. 6; chosen by Anderson (see Postscript).

Specimens described. lO 2987n-c, black shale below the Geisina Phase, Bearpark Colliery, Co. Durham.

Dimensions. In mm.

Length

Height

Width

HjL ratio

lO 2891a

0-53

0-26

0-22

49-2%

lO

28976

0-44

019

018

43-2 %

lO

2897c

0-57

0-29

019

50-8 %

Description. Carapace small, sub-cylindrical in shape, ends equal. Dorsal margin
straight or flatly convex, sloping from posterior to anterior; ventral margin straight.
Anterior end low and pointed; posterior end high, rounded, and somewhat tumid.
Greatest height and width posterior, the height being less than half the length (H/L <
50%). Dorsal and ventral outlines sub-cuneiform, thickness about equal to height.

Hinge line straight, low, longer than half the length. Left valve not raised along hinge
line and no groove. Valves almost equal, the right slightly larger and overlapping the
left narrowly along the ventral margin. Maximum overlap in posterior half of ventral
margin. Shell not preserved on Durham specimens; thin and smooth on type specimens.
Muscle spot antero-ventral of mid-point, arrangement of secondary scars uncertain.

Discussion. Although the specimens described here are too badly preserved to be figured,
their general features are close to the original description and figures of Jones and
Kirkby (1879). These specimens are smaller than those of Jones and Kirkby but com-
parable with the dimensions given for this species by Latham (1932).

The generic relationship of this species has been in doubt since its earliest description,
as the synonymy suggests. In 1884 Jones doubtfully referred it to DarwineUa (= Dar-
M'inula, DarwineUa preoccupied). Brady and Robertson, and several later authors,
followed suit. Cooper (1946), describing American specimens he considered synonymous.

690

PALAEONTOLOGY, VOLUME 9

removed this species from Carbonita and placed it in Darwimila on account of its hinge
structure and the character of the overlap of the valves. The left valve is not raised
above the right along the hinge as in most species of Carbonita, but the variation of this
character within the genus suggests that the simple hinge of C. piingens is more a matter
of structural simplicity in this thin shelled form than a fundamental generic distinction.
Cooper also suggested that in C. pungens the maximum overlap of the valves is around
the ends, not along the ventral margin as is the normal situation in Carbonita. However,
the lectotype specimen shows the normal Carbonita overlap, not that described by
Cooper.

C. pungens is certainly darwinuloid in shape but the distinctive Darwimila pattern of
muscle scar has not been recorded. It is considered, therefore, best to retain this species
in the genus Carbonita.

The frequent occurrence of bedding planes in Coal Measures shale crowded with
myriads of C. pungens is comparable to the mode of occurrence of Darwinula proper in
Mesozoic, Tertiary, and Recent non-marine sediments (Jones 1885; Kaufmann 1900;
Harper and Sutton 1935; and Swain 1955). Mandelstam (1956) also recorded a similar
occurrence, with up to forty variants of Darwinula, from the Upper Permian rocks of
the Kuznetsk Coalfield of Russia. In mode of occurrence, therefore, if not in structure,
C. pungens is very like Darwinula.

Stratigraphical distribution. The lectotype of this species comes from one of Jones and
Kirkby’s original localities, the Coal Measures at Methil in Fifeshire, but the same
authors also recorded this form from freshwater limestones in the Carboniferous Lime-
stone Series of Scotland. In the English Coal Measures, Jones and Kirkby (1890)
recorded its abundance in the Upper Coal Measures of the Manchester area. Edwards
and Stubblefield (1947) found it in the \]px>^x similis-pulchra Zone of Nottinghamshire
and Derbyshire. The present study has shown the abundance of C. pungens in both the
A. modiolaris and C. communis Zones of Durham (text-fig. 1). These records, though
sparse, suggest that this species ranges throughout the Coal Measures in Britain and
occurs locally in the Lower Carboniferous of the Central Valley of Scotland.

This species has also been recorded from the Coal Measures of France by Pruvost
(191 1), the Pennsylvanian of Illinois by Cooper (1946), and the Upper Carboniferous of
Nova Scotia by Copeland (1957).

Carbonita claripunctata sp. nov.

Text-fig. 9

1879 Carbonia rankiniaua Jones and Kirkby, p. 34 (in part), pi. 3, figs. 3, 8.

1897 Carbonia rankiniaua (Jones and Kirkby); Dawson, p. 396, fig. 10.

1932 Carbonia rankiniaua (Jones and Kirkby); Latham, p. 385 (in part).

1955 Carbonita agues (Jones); Kremp and Grebe, p. 148, pi. 16, fig. 2.

Derivation of name. Clams (lat.) — distinct; punctatus (lat.) — pitted, punctate.

Holoty'pe. Complete carapace, lO 2983. Paratypes. 4 complete carapaces, lO 2982 and lO 2984.

Occurrence. Holotype, lO 2983, Geisina Phase, Hopkins Band, Lower A. modiolaris Zone, Lower
Coal Measures, Ryhope, Sunderland, Co. Durham. lO 2982, horizon as holotype, localities Washing-

J. E. POLLARD: NON-MARINE OSTRACOD FAUNA 69t

ton and Follonsby, Co. Durham. lO 2984, shales above Flockton Thin Coal, Lower A. modiolaris
Zone, Lower Coal Measures, Warncliffe Woodmoor Colliery, Doncaster, Yorkshire.

Dimensions. In mm.

Holotype, lO 2983
ID 2982n
b

ID 2984n
b

Length

Height

Width

HjL ratio

1-04

0-55

0-43

52-8 %

0-98

0-50

0-40

51-0%

F02

0-54

0-42

53-0%

F07

0-57

0-47

53-2 %

106

0-55

0-46

52-7 %

a

TEXT-FIG. 9. Carbonita claripimctata sp. nov. a-d, paratype, lO 2984o, complete carapace, a, exterior
left valve; b, exterior right valve; c, ventral view; d, dorsal view; e, internal mould showing furrow and
muscle scar cast, from the Geisina Phase of the Hopkins Band, Ryhope Colliery, County Durham.

Description. Carapace elongate, sub-ovate, slightly tumid posteriorly. Dorsal margin
straight or slightly convex, sloping posterior to anterior; ventral margin gently eonvex.
Posterior end high, broadly rounded; anterior end low, bluntly pointed, with a steep
antero-dorsal slope. External anterior margin of left valve bears a low ridge of thicken-
ing inside the overlap of the right valve; a similar thickening is present on the postero-
ventral margin of the left valve. Greatest height and width posterior, dorsal outline
broadly sub-ovate.

Hinge line straight, or arched slightly in lateral view, about half the length of the
dorsal margin and mainly posterior. Left valve raised very slightly above right along the

692

PALAEONTOLOGY, VOLUME 9

hinge, and there is no groove. Right valve larger and overlaps the left narrowly on the
ventral, anterior, and postero-dorsal margins. Shell surface coarsely pitted except over the
muscle spot. These pits may tend to show a concentric pattern of arrangement. Muscle
scar circular, antero-ventral of mid-point, with irregular secondary scars sometimes
suggesting a rosette pattern. Internal ridge of shell thickening posterior to muscle scar
well developed and leaving a characteristic furrow on internal moulds. A second fainter
ridge is sometimes present anterior of the muscle scar.

Discussion. Dimorphism has not been detected in this species in the Hopkins Band,
probably owing to the relatively small number of well-preserved specimens that were
extracted.

In their original description of Carbonia rankiniana, Jones and Kirkby (1879)
included both smooth and punctate forms from the Lower Carboniferous and Coal

Cgrbonito claripunctatg

1.0-1

HEIGHT
m m. 0.5“

-EG

-i — ■ — . — I — . — I — I — I — . — I — r-

0.5 1.0. 1.5

LENGTH mm.

• Hopkins Band specimens

O Approx. Instar means

X Flockton Shale specimens

rankiniana LATHAM 1932

® C.satteriana JONES 8. KIRKBY 1890

© C. salteriana COOPER 1946

+ e.ggnes KREMP & GREBE 1955

TEXT-FIG. 10. Height and length dimensions of Carbonita ckiripimctata sp. nov. and allied or synonymous

species.

Measures of Scotland and the Coal Measures of England. The new species proposed
here is similar in lateral outline to the form they figure as fig. 2 and, like the shell of
fig. 8, is punctate. Their figured specimens came from the Coal Measures at Provenhall
in Lanarkshire, and material from that locality in the John Young Collection at the
Hunterian Museum, Glasgow, undoubtedly belongs to C. claripimctata sp. nov. The new
species was included in Jones and Kirkby ’s original broad definition of C. rankiniana.
It differs, however, from C. rankiniana (sensu stricto) in lateral outline, the greatest
height being nearer the posterior end, and in having a punctate shell. C. rankiniana, as
recently restricted by Anderson (in press), is a smooth shelled Lower Carboniferous
species.

Carbonita claripimctata sp. nov. differs from the Upper Coal Measures species
C. salteriana, and the synonymous C. roederiana in shape, punctate shell, the ventral
thickening of the shell, and in the presence of a well-developed internal ridge posterior
to the muscle scar.

The form described and figured by Kremp and Grebe (1955) as C. agues Jones from
the Coal Measures of the Ruhr is similar to this new species in lateral outline, overlap,
hinge characters, surface ornament, and size dimensions (text-fig. 10). The type species.

J. E. POLLARD: NON-MARINE OSTRACOD FAUNA

693

<T. agues Jones, although similar to C. cJaripimctata in size and surface ornament,
differs in the arch of the dorsum and rounding of the ends. The German specimens
should, therefore, more correctly be referred to C. daripuuctata sp. nov. than to C. agues
Jones.

Stratigraphical distributiou. Although the type material comes from the Lower A.
fuodiolaris Zone of the Lower Coal Measures of Durham and Yorkshire, this species
occurs in the Bottom Three Quarter ostracod band in the C. comunmis Zone of Durham.
Scottish specimens are likewise from the Lower or Middle Coal Measures.

Kremp and Grebe (1955) recorded their synonymous form ‘C. agues' as occurring
sporadically throughout stages Westphalian A to Westphalian C in the Ruhr coalfield.

POSTSCRIPT
by F. W. Anderson

A complete revision of the genus Carbouita has been undertaken and is being prepared
for publication. It was found necessary during the progress of this work to designate
lectotypes for most of the species described by Jones (1870) and by Jones and Kirkby
(1879).

In the foregoing paper Dr. Pollard has had occasion to refer to some of these species,
and lectotypes are therefore here designated.

1. Carbouita agues (Jones) 1870

Lectotype. In 43513, Adams Coll., Brit. Mus. (Nat. Hist.).

Horizon. Black Band Beds {A. tenuis Zone), Upper Coal Measures.

Locality. Cilfach Bargoed Colliery, Glamorgan, South Wales.

A holotype was not designated by Jones. The lectotype is from the type material and is the nearest
to that figured by him (1870, pi. 9, fig. 6a).

2. Carbouita eveliuae (Jones) 1870

Lectotype. In 56370, Adams Coll., Brit. Mus. (Nat. Hist).

Horizon. Black Band Beds (A. tenuis Zone), Upper Coal Measures.

Locality. Cilfach Bargoed Colliery, Glamorgan, South Wales.

Holotype not designated by Jones (1870), but the original of his fig. 4n, pi. 9, is easily recognized in
the type material and is here chosen as the lectotype.

3. Carbouita hitiuilis (Jones and Kirkby) 1879

Lectotype. The specimen figured by Jones and Kirkby, 1879, pi. 2, fig. 14.

Horizon. Earl David’s Parrot Coal (A. modiolaris Zone), Coal Measures.

Locality. Pirnie Colliery, Leven, Fife, Scotland.

The species was originally figured by Jones and Kirkby as Carbonia jabulina var. humilis. Several
specimens still exist in the type material in the Brit. Mus. (Nat. Hist.) Collection but none of these can
be identified with any certainty as that figured by Jones and Kirkby.

694

PALAEONTOLOGY, VOLUME 9

4. Carbonita inflata (Jones and Kirkby) 1879

Lectotype. I 1745, Brit. Mus. (Nat. Hist.) Coll.

Horizon. A. modiolaris Zone, Coal Measures.

Locality. Pirnie Colliery, Leven, Fife, Scotland.

The lectotype is one of the specimens figured by Jones and Kirkby (1879, pi. 2, figs. 15-19) as
Carbonia fabulina var. inflata, probably fig. 15.

5. Carbonita pimgens (Jones and Kirkby) 1879

Lectotype. I 1731, T. R. Jones Coll., Brit. Mus. (Nat. Hist.).

Horizon. Coal Measures, probably Lower A. similis-piilchra Zone.

Locality. Methil, Fife, Scotland.

The lectotype is taken from the type material and is the nearest to that figured by Jones and Kirkby
( 1879) as pi. 3, fig. 21.

6. Carbonita rankiniana (Jones and Kirkby) 1879

Lectotype. 1 1741, Brit. Mus. (Nat. Hist.) Coll.

Horizon. Calciferous Sandstone Series, Lower Carboniferous.

Locality. Coast west of Pittenweem, Fife, Scotland.

The lectotype is one of the specimens figured by Jones and Kirkby (1879, pi. 3, figs. 1, 6, 7) from
Pittenweem. Those on the same plate, from the Coal Measures at Provenhall, Glasgow, i.e. figs. 2-5, 8,.
are thought to be specimens of C. salteriana (Jones).

REFERENCES

ANDERSON, F. w. Revision of the British species of the Carboniferous ostracod genus Carbonita. Bull,
geol. Surv. Gt Br. (In press.)

ARMSTRONG, G. and PRICE, R. H. 1953. The Coal Measures of north-east Durham. Trans. Instn Min.
Engrs. 113, 973-97.

BEAN, w. 1836. Description and figures of Unio distortus Bean, and Cypris concentrica Bean, from the
upper sandstone and shale of Scarborough; and Cypris arcuata Bean from the coal formation of
Newcastle. Mag. Nat. Hist. 9, 376-7, fig. 55.

BOGER, V. H. 1964. Palokologische Untersuchungen an Cyclothemen im Ruhrkarbon. Paldont. Z. 38,
142-57.

BROADHURST, F. M. 1964. Some aspects of the palaeoecology of non-marine faunas and rates of sedi-
mentation in the Lancashire Coal Measures. Am. J. Sci. 262, 858-69.

CALVER, M. A. 1956. Die Stratigraphische Verbreitung der nichtmarinen Muscheln in den penninischen
Kohlenfeldern Englands. Z. dt. geol. Ges. 107, 26-39.

1966. In SMITH, E. G., RHYS, G. H., and EDEN, R. A. Geology of the country around Chesterfield,

Matlock and Mansfield. Mem. Geol. Surv. U.K. (In press.)

CARRUTHERS, R. G. 1930. Name, ‘Hopkins Shell Bed’. Mem. geol. Surv. Summ. Prog, for 1929, p. 69.

COOPER, c. L. 1941. Chester ostracods of Illinois. Rep. Invest. III. St. geol. Surv. 77, 101 pp., 14 pis.

1946. Pennsylvanian ostracods of Illinois. Bull. III. St. geol. Surv. 70, 177 pp., 21 pis.

COPELAND, M. J. 1957. The arthropod fauna of the Upper Carboniferous rocks of the Maritime Pro-
vinces. Mem. geol. Surv. Brcli Can. 286, 24-28, pi. 1, 2.

DAWSON, J. w. 1897. Notes on Carboniferous Entomostraea from Nova Scotia, in the Peter Redpath
Museum determined and described by Prof. T. R. Jones and Mr. Kirkby. Can. Rec. Sci. 7,
316-23.

EAGAR, R. M. c. 1961. A Summary of recent work on the palaeoecology of Carboniferous non-marine
lamellibranchs. C.R. 4'”^ Congr. Strat. et Geol. carbonifere, Heerlen, 1, 173-87.

J. E. POLLARD: NON- MARINE OSTRACOD LAUNA 695

ECHOLLS, D. J. and CREATH, w. B. 1959. A survey of Mississippian, Pennsylvanian and Permian Ostra-
coda recorded in the United States. Micropalaeontology, 5, 389-414.

EDWARDS, w. and STUBBLEFIELD, c. J. 1947. Marine bands and other faunal horizons in relation to the
sedimentary cycles of the Middle Coal Measures of Nottinghamshire and Derbyshire. Q. Jl. geol.
Soc. Land. 103, 209-53.

ERNST, w., KREJCi-GRAF, F., and WERNER, H. 1958. Parallclisierung von Leithorizonten im Ruhrkarbon
mit Hilfe des Bor-Gehaltes. Geochim. cosmochim. Acta, 14, 211-22.

MiCHELAU, p., and tasch, k. h. 1960. Vergleich des Niveaus von Floz-Zollverein I mit dem von

Wyshagen mit Hilfe der Bor-Methode. C.R. 4""^ Congr. Strat. et Geol. carbonifere, Heerlen, 1,
163-8.

HARLTON, B. H. 1927. Somc Pciinsylvanian ostracoda of the Glen and Hoxbar Formations of southern
Oklahoma and the upper part of the Cisco Formation of northern Texas. J. Paleont. 1, 203-12,
pi. 32, 33.

1928. Some Pennsylvanian ostracoda from Oklahoma and Texas. Ibid. 2, 132-44, pi. 21.

HARPER, F. and SUTTON, A. H. 1935. Ostracods from the Morrison Formation of the Black Hills of
South Dakota. Ibid. 9, 623-38, pi. 76.

HEATH, R. J. 1961. The palaeoecology of the pseudorobusta Band, with a consideration of the lithology
visible in hand specimen. Unpublished M.Sc. thesis, Univ. of Manchester.

HENNiNGSMOEN, G. 1953. Classification of Palaeozoic straight hinged ostracods. Norsk geol. Tidsskr.
31, 185-288.

HOLLAND, w. c. 1934. The ostracods of the Nineveh Limestone of Pennsylvania and West Virginia,
Pittsburg. Ann. Carneg. Mas. 22, 343-50, pi. 25.

HOPKINS, w. 1927. Further modifications of the correlation of the coal seams of the Northumberland
and Durham Coalfield. Trans. Instn. Min. Engrs. 74, 221-53.

1928. The distribution and sequence of the mussel bands in the Northumberland and Durham

Coalfield. Proc. Univ. Durham, phil. Soc. 8, 1-14.

1960. ‘Hopkins Shell Bed’. Le.xiqiie Stratigraphique International, I Europe. Ease. 3a VIII, 183.

JOHNSON, w. R. 1936. The ostracods of the Missouri Series of Nebraska. Pap. geol. Surv. Neb. 11,
52 pp. pi. 1-5.

JONES, T. R. 1870. On the bivalved Entomostraca from the Coal Measures of South Wales. Geol. Mag.
(1) 7, 218-19, pi. 9.

1884. Notes on the late Mr. G. Tate’s specimens of Lower Carboniferous Entomostraca from

Berwickshire and Northumberland. Hist. Berwicksh. Nat. Cl. 10, 319, 325.

1885. On the ostracods of the Purbeck formation; with notes on Wealden species. Q. Jl. geol.

Soc. Loud. 41, 311-52, pi. 8, 9.

1889. Notes on Palaeozoic Bivalved Entomostraca No. 27. On some North American (Canadian)

species. Ann. Mag. nat. Hist. (6) 3, 373-87, pi. 16, 17.

and KiRKBY, J. w. 1866. Some Carboniferous species. Ibid. (3) 18, 32-51.

1867. On Entomostraca of the Carboniferous rocks of Scotland. Trans, geol. Soc. Glasg. 2,

213-28.

1879. Some Carboniferous species belonging to the genus Carbonia Jones. Ann. Mag. nat.

Hist. (5) 4, 28-40, pi. 2, 3.

1884. On some Carboniferous ostracoda from Nova Scotia. Geol. Mag. (3) 1, 356-62,

pi. 12.

1886a. On some undescribed species of British Carboniferous ostracoda. Ann. Mag. nat.

Hist. (5) 18, 249-69, pi. 6-9.

18866. On some fringed and other ostracods from the Carboniferous Series. Geol. Mag. (3)

3, 248-53.

1886c. Notes on the distribution of the ostracoda of the Carboniferous formations of the

British Isles. Q. Jl. geol. Soc. Lond. 42, 496-514.

1890. On ostracoda found in shales of the Upper Coal Measures at Slade Lane, near

Manchester. Manch. geol. Min. Soc. Trans. 21, 137-42.

and YOUNG, J. 1899. On Carbonia', its horizons and conditions of occurrence in Scotland,

especially Fife. Trans. Edinb. geol. Soc. 7, 420-42.

KAUFMANN, A. 1900. Cypdden und Darwinulen der Schweiz. Rev. suisse Zool. 8, 209-423.

696 PALAEONTOLOGY, VOLUME 9

KELLETT, B. 1935. Ostracodes of the Upper Pennsylvanian and Lower Permian strata of Kansas.
J. Paleont. 9, 132-66.

KESLiNG, R. V. 1951. Morphology of ostracod moult stages. Illinois biol. Monogr. 21, 324 pp.

KNAUF, w. 1963. Zur Microfauna im Oberkarbon der Bohrung Miinsterland, 1. Fortsclv. Geol. Rheinld
West/. 11, 113-22.

KNIGHT, J. B. 1928. Some Pennsylvanian ostracodes from the Henrietta Formation of Eastern Missouri.
J. Paleont. 2, 229-67.

KREMP, G. and GREBE, H. 1955. Beschreibung und stratigraphischer Wert einiger Ostracodenformen aus
dem Ruhrkarbon. Geol. Jb. 71, 145-69, pi. 16.

KUMMEROW, E. 1949. t)bcr einiger Siisswasser-Ostracoden des Ruhrkohlen-Gebietes. Neues Jb. Miner,
geol. Palaeont. Abh. 9, 45-49.

1953. Uber oberkarbonische und devonische Ostracoden in Deutschland und in der Volksrepublik

Polen. Geologic, 7, 70 pp.

LATHAM, M. H. 1932. Scottish Carbonifcrous Ostracoda. Trans. R. Soc. Edinb. 17, 351-95.

LEiTCH, D., ABSALOM, R. G., and HENDERSON, s. M. K. 1937. The basal fauna of the Zone of Anthraconaia
modiolaris in the Coal Measures of South Ayrshire. Trans, geol. Soc. Glasg. 19, 404-8.

MAGRAW, D., CLARKE, A. M., and SMITH, D. B. 1963. The Stratigraphy and structure of part of the south-
east Durham Coalfield. Proc. Yorks, geol. Soc. 34, 153-208, pi. 19, 20.

MANDELSTAM, M. I. 1956. Ostracods of the coal-bearing formations of the Kuznetsk Coalfield. (In
Russian.) In Atlas to guide forms of fossil fauna and flora of Permian formations of Kuznetsk Coal
Basin, vzegei, Moscow, 59-109, pi. 9-19.

MARPLE, M. F. 1952. Ostracodes from the Pottsville Series in Ohio. J. Paleont. 26, 924-40, pi. 133-5.

MELVILLE, R. V. 1946. The non-marine lamellibranchs of the North Staffordshire Coalfield. Ann. Mag.
nat. Hist. (11) 13, 288-300.

MURCHISON, R. I. 1839. Silurian System. London.

PAYTON, c. E. and thomas, l. a. 1959. The petrology of some Pennsylvanian black shales. J. sedim.
Petrol. 29, 172-7.

PRUVOST, p. 1911. Note sur les Entomostraces bivalves du terrain houiller du Nord de la Erance.
Annls. Soc. geol. N. 40, 60-80, pi. ), 2.

• 1930. La faune continentale du terrain houiller de la Belgique. Mem. Mus. Hist. nat.Belg. 44,

218-56.

RAMSBOTTOM, w. H. c. 1952. The fauna of the Cefn Coed Marine Band in the Coal Measures at
Aberbaiden, near Tondu, Glamorgan. Bull. geol. Surv. Gt Br. 4, 8-32, pi. 2, 3.

ROUNDY, p. V. 1926. Mississippian formations of San Saba County, Texas. Part II: The Microfauna.
Prof. Pap. U.S. geol. Surv. 146, 5-8, pi. 1.

SCOTT, H. w. 1944. Permian and Pennsylvanian fresh-water ostracods. J. Paleont. 18, 141-7, pi. 23, 24.

1961. In MOORE, R. c. (ed.) Treatise on Invertebrate Paleontology, Pt. Q: Ostracoda. Geol. Soc.

Amer. and Univ. Kansas Press.

and SUMMERSON, c. H. 1943. Non-marine ostracoda from the Lower Pennsylvanian in the southern

Appalachians and their bearing on inter-continental correlation. Am. J. Sci. 241, 653-71,
pi. 1, 2.

and WAINWRIGHT, J. 1961. In moore, r. c. (ed.) Treatise on Invertebrate Paleontology. Pt. Q:

Ostracoda. Geol. Soc. Amer. and Univ. Kansas Press.

SOHN, I. G. 1961. In MOORE, R. c. (ed.) Ibid.

STRAND, E. (1926) 1928. Name Carbonita Strand. Miscellanea Nomenclatoria Zoologica et Palaeon-
tologica. Arch. Naturgesch. 92, 40, 41.

SWAIN, F. M. 1955. Ostracoda of San Antonio Bay, Texas. J. Paleont. 29, 561-646.

TAYLOR, B. J. and CALVER, M. A. 1961. The stratigraphy of exploratory boreholes in the west Cumber-
land Coalfield. Bull. geol. Surv. Gt Br. 17, 1-74.

ULRICH, E. o. and bassler, r. s. 1 906. Notes and descriptions of Upper Carboniferous genera and species.
Proc. U.S. natn. Mus. 30, 149-64, pi. 11.

1908. New American Palaeozoic Ostracods; Preliminary revision of the Beyrichidae with

descriptions of new genera. Ibid. 35, 277-346, pi. 37-44.

UPSON, M. E. 1933. The ostracoda of the Big Blue Series in Nebraska. Bull. geol. Surv. Neb. 8, 1-54,
pi. 1-4.

J. E. POLLARD: NON-MARINE OSTRACOD FAUNA 697

VANGEROW, E. F. 1957. Mikropalaeontologische Untersuchungen in den Kohlschreider-Schichten im
Wurmrevier bei Aachen. Geol. Jb. 12>, 457-97, pi. 20.

WOOD, A. 1937. The non-marine lamellibranchs of the North Wales Coalfield. Q. Jl. geol. Soc.
Loud. 93, 1-22.

WRIGHT, w. B. 1930. Additions to the fauna of the Lancashire Coal Measures. Mem. Proc. Manchr.
lit. phil. Soc. 13, 41-51, pi. 1.

1931. In The Geology of Manchester, Mem. geol. Surv. U.K.

JOHN E. POLLARD

Department of Geology,
The University,

Manuscript received 9 July 1965 Manchester 13.

THE PALAEONTOLOGICAL ASSOCIATION

Extracts from the Annual Report of the Council for 1965-6

Membership. On 31 December 1965 there were 1,342 members (673 Ordinary, 114 Student, and
555 Institutional), a net increase of 73 members during the year.

Finance. Because of the continued growth of Ordinary and Institutional membership, income from
subscriptions increased during the year by £232. Favourable interest rates helped to increase income
from this source by £116, but the figure includes £22 relating to 1964. Donations from authors for
specific papers and from oil companies have been very welcome and still form a substantial part of
income. They show an increase of £188 over 1964. It is hoped this support will continue at the same
generous level into 1966 when the Association has to meet further cost increases and new commitments.

The very large drop in income from sales (£542) is mainly attributed to the Association’s agents
selling stock valued at £930 which had been paid for in 1964; the income from sales in 1964 was there-
fore inflated at the expense of 1965.

The cost of publishing Palaeontology rose sharply, due to increasing the print off from 1,600 to
2,000 copies, to increasing the size by 56 pages (8 per cent) and 2 plates, and to an increase of charges
by the printers.

When the surplus balance of £207, transferred to the Publications Reserve, is used to meet the
remaining cost of Volume 8, the year’s working will have almost exactly balanced.

‘Palaeontology’. The four parts of Volume 8 were published during 1965. They contained 44 papers
and consisted of 767 pages and 1 1 1 plates.

Meetings. Four meetings took place during 1965-6. The Association is grateful to the Council of the
Geological Society of Fondon, Professor J. H. Taylor (King’s College, Fondon), and Professor F. H. T.
Rhodes (University College, Swansea) for generously granting facilities for meetings, and to the
Focal Secretaries for their efficient services.

a. The Eighth Annual General Meeting was held in the Rooms of the Geological Society of Fondon,
Burlington House, Fondon, W. 1 on Wednesday, 3 March 1965, at 5.00 p.m. The Annual Report
of the Council for 1964-5 was adopted and the Council for 1965-6 was elected. Professor T. S.
Westoll, F.R.S., delivered the Eighth Annual Address on ‘Problems of the Arthrodiran Fishes’.

b. A Field Demonstration Meeting was held at Castleton, Derbyshire, on Saturday, 8 May 1965.
The theme was ‘Faunal facies variations within the Fower Carboniferous reefs of Derbyshire’.
Dr. F. M. Broadhurst and Dr. I. M. Simpson acted as Feaders and Miss Joan I. Norcott was
Focal Secretary.

c. A Discussion Meeting on ‘Accumulation of pelagic skeletal remains’ was held at King’s College,
Fondon on Wednesday, 27 October 1965 at 2.00 p.m. About eighty persons attended. Dr. J. E.
Prentice was Focal Secretary.

d. A Joint Syinposiinn Meeting of the Geological Society and Palaeontological Association was held
in the Department of Geology, University College, Swansea on Monday/Tuesday, 20/21 December
1965. The symposium meeting was focused on some aspects of the volume ‘The Fossil Record’
(to be published by the Geological Society). Papers were read and discussed at three half-day
sessions and a Symposium Dinner was held in College House. Dr. V. G. Walmsley was Focal
Secretary.

THE PALAEONTOLOGICAL ASSOCIATION

699

Special General Meeting. Immediately before the Discussion Meeting at King’s College, London,
on Wednesday, 27 October 1965, a Special General Meeting was held at 1.45 p.m. to consider a change
in Rule 2 of the Constitution. The following motion was proposed from the Chair and passed {nein.
con.)\

That, effective from 1 January 1966, Rule 2 of the Constitution be amended to read as follows:
‘There shall be Ordinary members. Institutional members, and Student members, and their
annual subscriptions shall be Three Guineas, Seven Pounds, and Two Guineas respectively. Each
subscriber shall be considered a member of the Association, but Institutional members shall not
be eligible to take part in the Government of the Association.’

Council. The following were elected members of the Council of the Association for 1965-6 at the
Annual General Meeting on 3 March 1965 : President: Dr. L. R. Cox, F.R.S. Vice-Presidents: Dr. W. S.
McKerrow, Professor F. H. T. Rhodes. Treasurer: Dr. C. Downie. Secretary: Dr. C. H. Holland.
Editors: Mr. N. F. Hughes, Dr. Gwyn Thomas, Dr. I. Strachan, Dr. M. R. House. Other members:
Dr. C. G. Adams, Professor P. M. Butler, Dr. W. J. Clarke, Dr. G. Y. Craig, Dr. T. D. Ford, Dr. B. M.
Funnell, Dr. J. M. Hancock, Dr. G. A. L. Johnson, Dr. F. A. Middlemiss, Mr. M. Mitchell, Dr.
W. D. I. Rolfe, Dr. A. J. Rowell, Professor Scott Simpson, Dr. L. B. Tarlo, Dr. H. Dighton Thomas.

BALANCE SHEET AND ACCOUNTS FOR THE YEAR ENDING

31 DECEMBER 1965

Balance Sheet

Liabilities £ s. d.

Publications Reserve Account Balance as per Annexed Account
Amounts received in Advance Subscriptions for 1966 .

Provision for Cost of Publication of Palaeontology Vol. 8 as per
Income and Expenditure Account . ..... 7,147 19 8

Z.C55 expenditure incurred to 1 December 1965 . . . 4,681 2 0

Sundry Creditors .........

Note: No amount has been included in these Accounts for Sub-
scriptions unpaid at 31 December 1965

£ s. d.
6,290 2 3
401 0 6

2,466 17 8
11 4 4

£9,169 4 9

Assets

Office equipment .......

Investments at cost:

Equities Investment Fund for Charities — 958 units .
5 per cent Defence Bonds . . . . .

Sundry Debtors and Payments in Advance:

Authors for offprints ......

Palaeontology, Vol. 9 .

Cash at Bank — Deposit Account . . . .

Current Account ....

£ s. d. £ s. d.

20 15 0

999 18 3
4,000 0 0

173 9 10
47 15 0

4,999 18 3

221 4 10
2,807 8 7
1,119 18 1
£9,169 4 9

Report of the Auditors to the Members of the Palaeontological Association. We have examined the above Balance
Sheet and annexed Income and Expenditure Account which in our opinion give respectively a true and fair view
of the state of the Association’s affairs as at 31 December 1965 and of its income and expenditure for the year
ended on that date.

JOSHUA WORTLEY & CO.
Chartered Accountants.

C 4397

3a

700

PALAEONTOLOGY, VOLUME 9

Income and Expenditure Account
FOR THE YEAR ENDED 31 DECEMBER 1965

Expenditure £ s. d.

To Provision for Cost of Publication of Palaeontology,

Vol. 8, Part 1

1,827

13

10

Part 2

1,820

5

10

Part 3 ........ .

1,750

0

0

Part 4 ........ .

1,750

0

0

7,147

19

8

Add extra cost of Vol. 7, Part 4 . .....

9

17

1

7,157

16

9

Less excess Provision for Vol. 7, Part 3 (not now required)

161

4

0

Administrative Expenses:

Insurance ..........

1

15

3

Auditor’s Fee .........

10

10

0

Post and Stationery ........

162

9

1

Assistance to Treasurer .......

100

0

0

Duplicating and dispatching of circulars ....

109

1

4

Cost of distributing publications ......

60

7

1

Meetings, etc. .........

45

1

9

Depreciation of equipment .......

Excess of Income over Expenditure for the Year. Transferred

to

Publications Reserve Account ......

Income

£

s.

d.

Rv Subscriptions for 1965 .....

5,052

3

11

Subscriptions for 1964 .....

103

2

7

Sales of Publications ......

Interest Received:

5 per cent Defence Bonds ....

200

0

0

Bank Deposit Account .....

120

16

9

Equities Investment Fund for Charities

90

4

3

Specific Donations ......

General donations:

Californian Oil Co. .....

157

10

0

British Petroleum Co. .....

250

0

0

Total Oil Co. ......

50

0

0

Texaco ........

175

0

0

Burmah Oil Co. ......

100

0

0

£ ,y. d.

6,996 12 9

489 4 6
2 6 2

207 14 4
£7,695 17 9

£ s. d.

5,155 6 6
1,243 0 10

411 1 0

96 10 0

732 10 0

57 9 5
£7,695 17 9

Miscellaneous Receipts

INDEX

Pages 1 to 182 are contained in Part 1; 183 to 354 in Part 2; 355 to 522 in Part 3; and 523 to 705 in Part 4.

Figures in bold type indicate plate numbers.

A

Acaiithophylliim, 548; (A.) ae/ieae, 550, 85; aequise-
ptatwn, 85; asper, 85; clermontense, 552, 85, 86;
(Grypophyllum) jenkinsi, 562, 87; (Neostringophyl-
liim), 554; impHcatiim, 554, 86; turni, 559, 87.

Acant/iotriletes, 637; dimorphus, 637, 105; flextius, 637,
105; tenuis, Gil, 105.

Acaste, 187; downingiae s. str., 10, 1-3, 187, 28, 29; d.
macrops, 19, 3 (= Acastocephala, 201 ); inflata, 192,
29; subcaudata, 194, 30.

Acastocephala, 199; dudleyensis, 203, 32; macrops,
201, 31, 32.

Acastoides constricta, 22, 3, 197, 30.

Acritarchs: fine structure of, 351.

Acrocythere, 214; hauteriviana, 214, 33; h. laeva,
214, 33.

Actinodictyon, 99.

Actinostroma, 86; clathratiim, 15.

Aenigmastricklandia contorta, 347, 58.

Ahrensisporites vagus, 640, 106.

Alaska: crustacean Caryocaris from, 371.

Albaillella, 321 ; pennata, 323, 54.

Algae: Silurian Girvanella, 48.

Alisporites symmetricus, 65 1 , 108.

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

Ammodiscus cretacea, 495, 77.

Amphicytherura, 211; bartensteini, 211, 33; roemeri,
212, 33.

Amphipora, 109.

Ampyxina sabnoni, 46, 5.

Annelida: Australian Precambrian, 599.

Anostylostroma, 91; arvense, 14; hamiltoiiensis, 15;
laxum, 14; maculosum, 16.

Apatocythere, 21 1 ; ellipsoidea, 21 1 ; simulans, 211.

Arborea, 618; arborea, 619, 102.

Arctic: Silurian bryozoa, 517.

ArenobuUmina cf. sphaerica, 497.

Arthropoda. See Crustacea, Ostracoda, and Trilobita.

Ataxophragmium variabUs, 497.

Atelodictyon, 87.

Aulacera nodulifera, 19.

Australia: Devonian ammonoid, 458; Devonian
corals, 544; Precambrian fossils, 599.

Australophyllum, 584; bi late rale, 589, 95, 96; spougo-
phylloides, 93; thomasae, 586, 94, 95.

Australosutura aff. gardneri, 271, 42.

B

Barker, D. See Kaye, P.

Barr, F. T. The foraminiferal genus Bolivinoides from
the Upper Cretaceous of the British Isles, 220;
Upper Cretaceous foraminifera from the Bally-
deenlea Chalk, County Kerry, Ireland, 492.

Belocriuus cottaldi, 245, 39.

Beltanella gilesi, 604.

Birgeria acuminata, 137, 20.

Blastoidea: Devonian Belocrinus from France, 244.

Bohemia: new Devonian odontopleurid, 330.

Bolivinitella, 501 ; eleyi, 502, 77.

Bolivinoides, 227; australis, 34, 35; decorata, 231,
34-36, 500, 77; draco, 235; hiltermanni, 229, 36, 37;
laevigata, 237, 34, 38; miliaris, 234, 35; peterssoni,
238, 38; praelaevigata, 237, 38; pustulata, 236, 37;
sidestrandensis, 239, 34, 36, 37; strigillala, 228, 34,
37.

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

Brachiopoda: Devonian schizophoriids, 381; from
Nevada, 152; new Scottish productid, 426; Para-
chonetes gen. nov., 365 ; predation and shell damage,
355; Silurian Eocoelia, 523; unusual stricklandiids,
346.

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

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

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

Bryozoa; micrometric formula in fenestrate crypto-
stomes, 413; Silurian from Arctic Canada, 517.

C

Calamospora, 632; elliptica, 632, 104; mulliplicata,
632, 104; pseudotriangulara, 633, 104.

Canada: Devonian gastropod, 142; Ordovician

trilobites, 39; Silurian bryozoa, 517.

Carboniferous: Australosutura from Oklahoma, 270;
cristate megaspores, 488; Namurian radiolaria,
319; new productid, 426; non-marine ostracoda,
667; population studies in limestones, 252; shell
damage in brachiopods, 355; spore assemblages
from Kittanning Coal, 629.

Carbonita, 683; agues, 693; claripunctata, 690; eve-
linae, 686; humilis, 683; inflata, 687; pungens, 689;
rankiniana, 694.

Caryocaris, 371; curvilata, 64, 65; oklaliomensis, 379;
silurica, 379.

Cephalopoda: Australian CIteiloceras, 458; Creta-
ceous phylloceratid, 455; Liassic ammonites, 312;
variation and ontogeny in Oxfordian ammonites,
290.

Chasmops odini, 482, 75.

CIteiloceras (C.) aciitum, 459, 72.

Cheilotrypa ostiolata, 82.

Churkin, M. Morphology and stratigraphic range of
the phyllocarid crustacean Caryocaris from Alaska
and the Great Basin, 371. See also Lenz, A. C.

702

INDEX

Cibicides beaumontiana, 507, 79.

Cicatricosisporites, 288.

Clarkson, E. N. K. Schizochroal eyes and vision of
some Silurian acastid trilobites, 1 ; schizochroal
eyes and vision in some phacopid trilobites, 464.

Classification: of fenestrate cryptostomes, 413.

Clatlirocoilona, 98; eifelensis, 16.

Claihrodictyoti, 89; striatellimi, 15; vesicidoswn, 14.

CUthrocytheridea brevis, 210.

Coelenterata: Devonian corals, 544; dimorphism in
i'n /a/o/iora, 448; microstructure of stromatoporoids,
74; Precambrian from Australia, 599; Silurian
Microplasma, 148.

Complexisporites chalonerii, 650, 108.

Conomediisites lobatus, 608, 99.

Converrucosisporites pseudovalvus, 635, 105.

Corals: Australian Devonian, 544; dimorphism in
Striatopora, 448; Silurian Microplasma, 148.

Cordyloblastus eifelensis, 39.

Craig, G. Y. and Jones, N. S. Marine benthos, sub-
strate, and palaeoecology, 30.

Crassispora kosankei, 640, 106.

Creniceras renggeri, 301, 50-52.

Cretaceous: Bolivinoides from Britain, 220; foramini-
fera from Irish Chalk, 492; ostracoda from Bar-
remian, 208; phylloceratid from Speeton Clay,
455 ; schizaeaceous spores, 274.

Cristatisporites verrucosus, 643, 106.

Crustacea: phyllocarid from North America, 371.
See also Ostracoda.

Cryptolithoides sp. indet., 44, 5.

Cyclomedusa, 606; plana, 607, 98.

Cyclotrypa silurica, 521, 82.

Cyrtina, 177; sp., 178, 27.

Cytliereis, 215; blanda, 215; cf. geometrica, 215, 33.

Cytherelloidea, 208; dalbyensis, 209, 33; ovata, 208.

Cytheropteron, 213; (C.) reightonensis, 213; rugosa,
213; (Eocytheropteron) nova, 213; (Infracytliero-
pteron) exquisita, 213.

Cytherura reticulosa, 212.

D

Dalmanites, 478; caudatus, 480, 75; vulgaris, 478, 75.

Delepinea destinezi, 60.

Dendrostroma, 92.

Densosporites, 641; oblatus, 641, 106; spackmanii,
642, 106.

Devonian: Cheiloceras from New South Wales, 458;
blastoid from France, 244; corals from New South
Wales, 544; gastropod Orecopia in Canada, 142;
Nevada brachiopods, 152; new odontopleurid,
330; Paraclionetes gen. nov. 365; schizophoriids
from Europe, 381.

Dickinsonia, 620; costata, 620, 101 ; elongata, 621, 102;
tenuis, 622, 103.

Dimorphism: in coral Striatopora, 448.

Diplotrypa franklini, 518, 81, 82.

Distortion: method of analysing, 125.

Ditomopyge fatmii, 67, 13.

Dolocytheridea, 209; intermedia, 210.

Donovan, D. T. The Lower Liassic ammonites Neo-
microceras gen. nov. and Paracymbites, 312.

Dubaria? cf. thetis, 170, 24.

E

Echinocoelia denayensis, 176, 26.

Echinodermata. See Blastoidea.

Edenoxylon, 360; aemulum, 360, 61; parviareolatum,

61.

Ediacara flindersi, 602, 99.

Endosporites, 646; globiformis, 646, 107; grandicorpus
647, 107.

England: Barremian ostracoda, 208; Cretaceous
Bolivinoides, 220; Eocene wood, 360; Namurian
radiolaria, 319; non-marine ostracoda from Coal
Measures, 667; Rhaetic labyrinthodonts, 135;
Silurian coral Microplasma, 148; Speeton Clay
phylloceratid, 455.

Eocene: fossil wood from Britain, 360; new turtle
from Somalia, 511.

Eocoelia, 528; curtisi, 537, 83, 84; hemisphaerica, 533,
83, 84; intermedia, 536; 83, 84; sulcata, 538, 83, 84.

Eomarginifera sp., 60.

Eponides cf. concinna, 505, 79.

Eucytherura nettletonensis, 212.

Europe: Devonian schizophoriid brachiopods, 381.

Euryitycythere sp. B, 211.

Eyes: in acastids, 1 ; in phacopids, 464.

F

Eerestromatopora, 111.

Fistuliporai?) mutabilis, 520, 82.

Florinites occultiis, 649, 108.

Foraminifera: Bolivinoides in Upper Cretaceous, 220;
Chalk from Ireland, 492.

Fossils: distortion in, 125.

France: Devonian blastoid, 244.

G

Gasteropoda: Devonian Orecopia in Canada, 142.

Gavelinella, 506; lorneiana, 506, 79; thalmanni, 506, 78.

Geisina arcuata, 678.

Gerronostroma, 101 ; elegans, 16.

Girvanella, 50; effusa, 60, 11; fragila, 54, 7; incompta,
58, 9-11; media, 59, 8, 10, 11; problematica, 57, 9;
p. lumbricalis, 57, 9; prolixa, 54, 6, 8; pusilla, 58, 10;
ramosa, 59, 12; sarmenta, 53, 6, 7.

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

Globigerinelloides aspera, 503, 78.

Glomospira cf. gordialis, 497.

Granisporites medius, 634, 105.

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

Gravisporites densus, 641, 106.

Gutboerlisporites, 645; delicatus, 646, 107; erectus,
645, 107.

Gypidula, 160; cf. recur tens, 161, 24.

Gyroidinoides umbilicata, 505, 79.

INDEX

703

H

Habib, D. Distribution of spore and pollen assemblages
in the Lower Kittanning Coal of western Pennsyl-
vania, 629.

Hadvorhynchia, 164; eurekaensis, 165, 25.

Hammatostroma, 92; albertense, 15.

Hermatostwma, 106; beiithi, 17; schl uteri, 17.

Heterohelix globidosa, 503, 78.

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

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

Hughes, N. F., and Moody-Stuart, Judith. Descrip-
tions of schizaeaceous spores taken from early
Cretaceous macrofossils, 274.

Hypophylloceras cf. perlobatum, 455, 72.

I

Idiostroma, 105; caespitosum, 16; roemeri, 17.

Ireland: Chalk foraminifera, 492; population studies
on Visean limestones, 262.

Isoprusia, 331 ; corimticauda, 341, 56; laportei, 337, 57;
mydiakia, 332, 55-57; sandbergeri, 341, 56; sperata,
341 ; Ursula, 339, 57.

J

Jenkins, T. B. H. The Upper Devonian index am-
monoid Cheiloceras from New South Wales, 458.

Johnson, H. M. Silurian Girvanella from the Welsh
Borderland, 48.

Johnson, J. G. Middle Devonian brachiopods from
the Roberts Mountains, central Nevada, 152;
Parachonetes, a new Lower and Middle Devonian
brachiopod genus, 365.

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

Jurassic: fish and labyrinthodonts in Rhaetic, 135;
Liassic ammonites, 312; ontogeny and variation in
Oxfordian ammonites, 290.

K

Katliwaia, 69; capitorosa, 71, 13.

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

Kimberia quadrata, 611, 97.

Knoxisporites sp. A, 638, 105.

Koneprusia, 342; subterarmata, 343, 57.

Krotovia spinulosa, 60.

L

Labechia, 118; confer ta, 19.

Lageiia acuticosta, 498, 77.

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

Leiorhyncbus sp., 166, 24.

Leiotriletes sp. A, 631, 104.

Lenz, A. C., and Churkin, M. Upper Ordovician
trilobites from Northern Yukon, 39.

Leptathyris circula, 171, 24.

Lophotriletes interruptus, 636, 105.

Lorenzinites, 608; rants, 609, 100.

Lycospora contacta, 640, 106.

Lyrielasma, 567; aggregatum, 570, 89, 90; inicrum,
567, 89; subcaespitosum, 569, 87.

M

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

Mawsonites spriggi, 607, 99.

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

Medusinites asteroides, 605, 97.

Micrhystridiiim, 352; cf. breve, 352, 59; cf. fragile,
352, 59.

Microplasma lovenianiim, 149, 22.

Microstructure: of stromatoporoids, 74.

Mississippian: Australosutura in Oklahoma, 270.

Mollusca. See Cephalopoda, Gasteropoda.

Monilospore sp. A, 643, 106.

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

N

Neocythere (N.) protovauveeni, 215.

Neomicroceras, 312; commune, 313, 53.

Nevada: Middle Devonian brachiopods, 152.

New South Wales: Devonian corals, 544; Devonian
index ammonoid, 458.

Nodosaria limbata, 498.

North America: assemblages of spores from Kittan-
ning Coal, 629; Caryocaris from Alaska, 371;
Devonian brachiopods from Nevada, 152; Mis-
sissippian trilobite from Oklahoma, 270; Ordovician
trilobites from Yukon, 39.

O

Oklahoma: Mississippian trilobite from, 270.

Oliver, W. A., Jr. Description of dimorphism in
Striatopora flexuosa Hall, 448.

Ordovician: range of Caryocaris in America, 371;
trilobites from Yukon, 39.

Orecopia, 142; cotei, 145, 21; mccoyi, 21; uclitensis, 21.

Ormiston, A. R. Occurrence of Australosutura
(Trilobita) in the Mississippian of Oklahoma,
U.S.A., 270.

Orthonotacytliere, 213; blanda, 213; inversa inversa,
213; problematica, 213, 33.

Ostracoda: Cretaceous from Lincolnshire, 208; non-
marine from Coal Measures, 667.

Ovatoscutum concentricum, 612, 97.

P

Pakistan: Permian trilobites from, 64.

Palaeoecology: marine benthos and, 30; population
studies in Visean limestones, 252.

Paleospora fragila, 647, 108.

Palframan, D. F. B. Variation and ontogeny of some
Oxfordian ammonites: Taramelliceras richei (de
Lori of) and Creniceras renggeri (Oppel) from
Woodham, Buckinghamshire, 290.

Parachonetes, 365; macrostriatus, 367, 62, 63; ver-
neuili, 63.

Paracymbites, 314; dennyi, 315, 53.

704

INDEX

Parallelopora, 1\S', goldfiissi, 19; ostiolata, 19.

Parvancorina minchami, 625.

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

Pelletieria valdensis, 278, 44, 45.

Pehoceras sp., 48.

Pennsylvania: spore assemblages from Kittanning
Coal, 629.

Pentamerella wintereri, 161, 23.

Permian: trilobites from Pakistan, 64.

Phacops, 466; batracheus, 475, 74; boecki, 476, 74;
breviceps, 412, 74; feciindus, 471, 73, 74; /. major,
467; latifrons, 7, 73; rana, 467; r. crassitiibercidata,
470, 73; r. milleri, 468, 73.

Pityosporites kittamnngensis, 650, 108.

Planisporites ? sp. A, 636, 105.

Plants. See Algae, Spores, Wood.

Plicatifera plicatilis, 60.

Pocock, Yvonne P. Devonian shizophoriid brachio-
pods from western Europe, 381.

Podoaiemis somalieiisis, 511, 80.

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

Popofskyelhim wuhdatiim, 326, 54.

Potoniesporites elegans, 648, 108.

Pvaebidimina obtusa, 499, 77.

Praecambridium sigillum, 623, 102.

Precambrian: fossils from Australia, 599.

Productina margaritacea, 356.

Promargbufera, 428; trearnensis, 430, 67.

Protocythere, 215; hecliti, 215; inoniata, 215.

Protozoa: Cretaceous Bolivinoides, 220; Cretaceous
formaninifera from Ireland, 492; Namurian radio-
laria, 319.

Pseiidoaclinodictyon, 99; vagaus, 15.

Pseiidobytliocy there, 214; ornata, 214; vellicata, 214.

Pseudochonophyllum, 563 ; pseiidoheliaiilhoides, 564,

87, 88.

Pseudorhizostomites, 609; liowcliini, 103.

Pteridinium cf. simplex, 616, 101.

Ptilofenestella carrickensis, 413.

Pimctatisporites, 633; ellipticiis, 634, 105; globidosus,
634, 104; obliqiiiis, 633, 104; sphaerorigidus, 633,
104.

R

Radiolaria: Namurian from Derbyshire, 319.

Pangea, 614; grandis, 616, 100; longea, 614, 100.

Ranocythereis caistorensis, 215, 33.

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

Reticidatasporites aletoreticidus, 631, 104.

Robergia yiikonensis, 41 , 4, 5.

Rosenella macrocystis, 19.

Rotaspora? perforata, 641, 106.

Ruffordia, 275; goepperti, 276, 43.

Rugoconites, 610; eiiigmaticiis, 611, 100.

Riigoglobigerina sp., 504, 78.

S

Salt Range: Permian trilobites from, 64.

Savage, R. J. G., and Large, N. E. On Birgeria
acuminata and the absence of labyrinthodonts from
the Rhaetic, 135.

Savitrisporites bhmtiis, 643, 106.

Schizaeopsis americana, 283, 46, 47.

Schizophoria, 381; antiqua, 383, 66; providvaria, 386,
66; pygmaea, 391, 66; schnuri blankenheimensis,
391, 66; striatula, 397, 66; strigosa, 402, 66; vid-
varia, 405, 66; sp., 60.

Schopfipollenites sp. A, 653, 109.

Scliopfites grossus, 635, 105.

^ Schiichertella' sp., 164, 23.

Schideridea, 210; cf. bernouilensis, 201, 33 ; rhomboid-
alis, 210.

Scotland: cristate megaspores, 488; new Visean pro-
ductid, 426.

Sdzuy, K. An improved method of analysing distor-
tion in fossils, 125.

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

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

Silurian: bryozoa from Canadian Arctic, 517; coral
from Monmouthshire, 148; Eocoelia spp., 523;
Girvanella, 48; revision of Acaste, 183; trilobite
eyes, 1, 464; unusual stricklandiids, 346.

Somalia: new Eocene turtle from, 511.

Spackmanites, 637; ellipticiis, 638, 105; facierugosus,
638, 105.

Spinatrypa, 167; aff. lata, 168, 25; sp. B, 168, 25.

Spongophylliim, 573; halysitoides halysitoides, 573, 90;
rosiforme, 575, 91.

Spores: Cretaceous schizaeaceous, 274; cristate mega-
spores, 488; Kittanning Coal assemblages, 629.

Spriggina ? ovata, 622, 98.

Stacliyodes, 116; verticellata, 18.

Steam, C. W. The microstructure of stromatoporoids,
74.

Stictostroma, 96; damnoniensis, 14; mammilliferum,

16.

Striatopora flexuosa, 448, 68-71.

Stricklandia lens aff. progressa, 348, 58.

Stromatopora, 110; bucheliensis, 14; carteri, 18; con-
centrlca collicidata, 17; divergens, 17; mononensis,
14; typica, 14, 18.

Stromatoporella, 93; arachnoidea, 16; granidata, 15;
selwyni, 15.

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

Stylodictyon, 1 1 6.

Syringostroma, 113; densiim, 18; subfuscum, 14.

T

Taimyrophyllum, 591 ; expansum, 592, 96.
Taleastroma, 112; cummingsi, 18.

Taramelliceras richei, 291, 48-50.

Tavener-Smith, R. The micrometric formula and the
classification of fenestrate cryptostomes, 413.

INDEX

705

Tertiary: fossil wood, 360; new Eocene turtle, 511.
Thymospom concentrica, 644, 106.

Triassic: fine structure of acritarchs, 351; Rhaetic
labyrinthodonts, 135.

Tribrachidiimi heraldicimi, 625, 97, 99, 101.

Triletes, 488; pannosus, 488, 76; xnbpalaeocristatiis,
490, 76.

Trilobita: Australosutiira in Oklahoma, 270; eyes and
vision in, 1, 464; new Devonian odontopleurid,
330; Ordovician from Yukon, 39; Permian from
Pakistan, 64; revision of Acaste, 183.

Triqidtrites, 639; cheiliis, 639, 106; magnificus, 639,
106; sp. A, 639, 106.

Trochammonoides sp., 497, 77.

Trupetostroma, 102; warreni 16, 17.
Tiibercidatosporites spinoplicatiis, 644, 107.

V

Vagiania ? sp., 169, 26.

Vallomyoma, 158; devonica, 159, 23.

Verrucosisporites compactus, 635, 105.

Vertebrata: Birgeria from Rhaetic, 135; new Eocene
turtle, 511.

Veryhachium reductum, 353, 59.

Vesicaspora, 649; saavensis, 649, 108; wdsotii, 649,

108.

Visean: shell damage in brachiopods, 355; new pro-

ductid, 426; population studies in Irish limestones,
252.

W

Wade, M. See Glaessner, M. F.

Wales: unusual stricklandiids from, 346.

Walker, C. A. Podocnemis somaliensis, a new pleuro-
diran turtle from the Middle Eocene of Somalia,
511.

Wanenella, 173; cf. cohimbiua colwnbina, 175, 27;
kirki praekirki, 173, 27.

Welsh Borders: Silurian GirvaneUa, 48.

White, D. E. The Silurian rugose coral Microplasma
lovenianum Dybowski from Monmouthshire, 148.

Wood: Eocene Anacardiaceae, 360.

X

Xystriphvllwn, 511 \ diiiistaiii, 578, 91, 92; magnum,
584, 93.

Y

Yukon: Ordovician trilobites from, 39.

Z

Ziegler, A. M. The Silurian brachiopod Eocoelia
hemisphaerica (J. de C. Sowerby) and related
species, 523; unusual stricklandiid brachiopods
from the Upper Llandovery Beds near Presteigne,
Radnorshire, 346.

■)*

■ \

■\ 4 V ■■

■m

THE PALAEONTOLOGICAL ASSOCIATION

COUNCIL 1966-7

President

Professor T. S. Westoll, The University, Newcastle upon Tyne
Vice-Presidents

Dr. W. S. McKerrow, University Museum, Oxford
Professor F. H. T. Rhodes, University College, Swansea

Treasurer

Dr. C. Downie, Department of Geology, The University, Mappin Street, Sheffield 1

Secretary

Professor C. H. Holland, Department of Geology, Trinity College, Dublin

Assistant Secretary

Dr. J. M. Hancock, Department of Geology, King’s College, London

Editors

Mr. N. F. Hughes, Sedgwick Museum, Cambridge
Dr. Gwyn Thomas, Department of Geology, Imperial College of Science, London, S.W.7
Dr. I. Strachan, Department of Geology, The University, Birmingham, 15
Professor M. R. House, The University, Kingston upon Hull, Yorkshire

Other members of Council

Professor P. M. Butler, Royal Holloway College, Surrey
Mr. M. A. Calver, Geological Survey Office, Leeds
Dr. G. Y. Craig, Grant Institute of Geology, Edinburgh
Miss Grace Dunlop, Bedford College, London
Dr. T. D. Ford, The University, Leicester
Dr. B. M. Funnell, Sedgwick Musemn, Cambridge
Dr. R. P. S. Jefferies, British Museum (Natural History), London
Dr. G. A. L. Johnson, The University, Durham City
Dr. F. A. Middlemiss, Queen Mary College, London
Dr. W. D. I. Rolfe, The University, Birmingham
Professor Scott Simpson, The University, Exeter
Dr. A. H. Smout, British Petroleum Company, Sunbury-on-Thames
Dr. L. B. H. Tarlo, The University, Reading

Overseas Representatives

Australia: Professor Dorothy Hill, Department of Geology, University of Queensland, Brisbane
Canada: Dr. D. J. McLaren, Geological Survey of Canada, Department of Mines and Technical
Surveys, Ottawa, Ont.

India: Professor M. R. Sahni, 98, The Mall, Lucknow (U.P.), India

New Zealand: Dr. C. A. Fleming, New Zealand Geological Survey, P.O. Box 368, Lower Hutt
West Indies and Central America: Mr. John B. Saunders, Geological Laboratory, Texaco Trinidad,
Inc., Point ^ Pierre, Trinidad, West Indies

Eastern U.S.A.: Professor H. B. Whittington, Department of Geology, Sedgwick Museum,
Cambridge, England

Western U.S.A. : Professor J. Wyatt Durham, Department of Paleontology, University of California,
Berkeley 4, Calif.

PALAEONTOLOGY

VOLUME 9 • PART 4

CONTENTS

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

related species. By a. m. ziegler 523

Spongophyllidae from the Devonian Garra formation. New South Wales.

By D. L. STRUSZ 544

The late Precambrian fossils from Ediacara, South Australia. By m. f.

GLAESSNER and M. WADE 599

Distribution of spore and pollen assemblages in the Lower Kittanning Coal

of Western Pennsylvania. By d. habib 629

A non-marine ostracod fauna from the Coal Measures of Durham and

Northumberland. By j. e. pollard 667

PRINTED IN GREAT BRITAIN AT THE UNIVERSITY PRESS, OXFORD
BY VIVIAN RIDLER, PRINTER TO THE UNIVERSITY

i

p

M 2 ^ 2 W Z CO

TUTiON NoriniiiSNi NviNosHilws S3iavaan libraries Smithsonian institution Noiir

to \ ^ ^ </) “ to

jvaen libraries Smithsonian institution NoiinmsNi nvinoshiiiais saiavaan libr

(O — w ± to \ ±

ITUTION NOlinillSNI NVINOSHilWS SBiaVaaiT LIBRARIES SMITHSONIAN INSTITUTION NOlil
' (O z w 5 .... w z ,y w

^ .3 ^ »' — ^N.'-N — ^

to .V-;

O

z

X
to
o

> 5 5 Z '■•iiCmsjX 5 >

aVaan^LIBRARIEs‘^SMITHSONlAN INSTITUTION NOliniliSNI NVINOSHilWS S3 I a Va 8 n_ LI B R

= to X ^ CO — CO ^ 2

o x^Npc^ ” o pc^ ^ o

ITUTION^ NOIiniliSNI~’NVINOSHiIWS S 3 I a V a 9 IT L I B R A R I E S SMITHSONIAN INSTITUTION NOlif
Z r- z I- Z - -

.V ,n Nftr ^ x-'S^' ^ m m

- <A - to ± to

dvaen libraries smithsonian institution NoiiniiisNi MviNOSHxms S3 1 avaa n ^li br

2 , to ^ ^ Z > to 2 ^ ^ 5 ■'

— >iv ^ ^ S S Xi^ti vo?v — —

2 H z ^ 2 /^teW^S H 2

I

.:W' I t

^ z — ^ Z to z CO

ITUTION NOliniliSNI NVINOSHill^S S3iavaan libraries SMITHSONIAN INSTITUTION NOlif

oj % </> _n in

o r fee, Jlj o

|.' i I-'

z to 2

O •• W _ \iVAS;^ O _ O

_ Z ^ _J Z _j Z -J

avaan libraries smithsonian institution NoiiniixsNi NViNOSHims saiavaan libr

2 r- Z r- z ""

^ . vm 2 OT 2 03

>

m DC ^ 2 rn 2 rn ^ m

to — to ± to ^. _ to

ITUTION NOliniliSNI NVINOSHillNS S3iavaan LIBRARIES SMITHSONIAN INSTITUTION NOlil
^., to Z to Z .<• to z .V- ^

^ V- ^

x / Jt /* "Z. —I . .r / y.A- 2 —1

W X . X O .mV X

- ^

o

z

to H „ _

avaan libraries smithsonian institution NoiinmsNi NviNosHims saiavaan libr

— #/s rn — rn

wai 3; O

5 >■' S >■

W ■•'* 2: w Z CO '•^' 2 00

NviNOSHiiws SBiyvyan libraries Smithsonian institution NoiiniiiSNi nvinoshiiws

^-rrr~^ ^ ^ ^ ^ ^ \ ^

3 ^ ^ LU

O — O

-J Z -J Z _

SMITHSONIAN INSTITUTION NOIXnillSNI NVINOSHillNS SBinvysn LIBRARIES SMITHSONIAN
F” z F- z r

cn

S) >

JJ ^

^ — to^. ? CO ? co

NVINOSHilWS S3 1 yvy an LIBRARIES SMITHSONIAN INSTl TION NOliDiliSNI NVINOSHillNS

W 2. .-.♦F. W 2

^ . V ^ d ^ ^

,.-/ Z —t /W<rtn)r»AX -»■ ' — j ^ A /y>fTOV

) ^ CO (/5 ^

SMITHSONIAN_INSTITUTION NOlifUliSNI NVINOSHil^VS SBiyvyail LIBRARIES SMITHSONIAN

’ Z ^ <F) — CO ~ to

m

o

O XCVVDC^^ “ "^Uisty o

.NVINOSHillNS S3 I H Vy 3 IT LI B RAR I ES^SMITHSONIAN~*INSTITUTION^NOIiniliSNI~'NVINOSHilVVS^

. ^ ^ [I ^ 5 F~ z F- V z

vv» t (§^ .i?) > (1^ is) H fp.fe 311 2 (§1^ is) > t

m N^osv'jy ^ m ^ x^ivas>^}?^ rn

CO • _ CO ? CO

SMITHSONIAN INSTITUTION NOliDiliSNI NVINOSHimS S3iyVHan LIBRARIES SMITHSONIAN I

Z <c CO z CO z \ z

' E ^ i » § s ». i 3 I

2 ' \T' > ' 2 >■ 2 '" S^’ > ^

<0 ■•'" z to — z CO '^' z ^

NVINOSHillAJS S3iyvyan libraries SMITHSONIAN institution NOIiniliSNI NVINOSHXUMS '

. . Z ^ ^ CO 5 \ ^

, ^ ^^A^ I g S 5 “

SMITHSONIAN INSTITUTION NOliDiliSNI NVINOSHillNS S3 I y Vy 8 ll”' LI B RAR 1 ES SMITHSONIAN I

NVINOSHimS S3 I yvnan LIBRARIES, SMITHSONIAN INSTITUTION NOliDiliSNI NVINOSHIIWS